1. Early-Cult-of-Skanda-in-North-India-From-Demon-to-Divine-Son-Richard-D-Mann (thesis)

THE EARLY CULT OF SKANDA IN NORTH INDIA: FROM DEMON TO DIVINE SON

By

RICHARD D. MANN

A Thesis

Submitted to the School of Graduate Studies

in Partial Fulfillment of the Requirements

for the Degree

Doctor of Philosophy

McMaster University

Copywrite by Richard Mann, April 2003

National Library Bibliothèque nationale of Canada du Canada

Acquisitions and Acquisisitons et Bibliographic Services services bibliographiques

395 Wellington Street Ottawa ON K1A 0N4 395, rue Wellington

Canada Ottawa ON K1A 0N4 Canada

Your file Votre référence ISBN: 0-612-86473-1 Our file Notre référence ISBN: 0-612-86473-1

The author has granted a non- L'auteur a accorde une licence non exclusive licence allowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sell reproduire, preter, distribuer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/film, de reproduction sur papier ou sur format électronique.

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts from it Ni la these ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou aturement reproduits sans son permission. autorisation.

Canada

THE EARLY CULT OF SKANDA IN NORTH INDIA

DOCTOR OF PHILOSOPHY McMaster University (Religious Studies) Hamilton, Ontario

TITLE: The Early Cult of Skanda in North India: From Demon to Divine Son.

AUTHOR: Richard D. Mann, M. A. (McMaster University)

SUPERVISOR: Professor Phyllis E. Granoff

NUMBER OF PAGES: 309

ii

Abstract

This thesis studies the development of the Hindu god Skanda-Kärttikeya from the fourth century BCE to the fourth century CE in north India. I argue that during this time period the deity is transformed from a demonic being associated with childhood diseases to a respected divine general and son to Śiva. I begin with a discussion of the earliest written material about the deity found in the two Sanskrit Epics (The Mahabhärata and The Ramayana) and other texts. These texts establish Skanda-Kärttikeya's origins in demonic beings and illustrate his transformation into a martial deity. These texts also demonstrate how Brahminical redactors assimilated this deity into their own traditions. This process of assimilation takes an inauspicious and unorthodox deity and transforms him into an auspicious and orthodox deity. I go on to argue that this transformation did not result in the increased popularity of this deity, but brings about the end of his popular cult in the north of India. Based on ancient coinage, statuary and inscriptions I demonstrate that this deity's popularity was related to his earlier terrible image and a propitiatory cult designed to appease him. Once the dangerous aspect of his image was removed, so was the main source of his popular cult. As opposed to previous scholarship on this deity, I argue that the Brahminization of this deity's cult brings about its end. l also demonstrate, based on this deity's depiction on ancient coinage, statuary and epigraphy, that there were also political forces at work in this process.My research demonstrates that the most important groups in this process were non-Indian. Primarily, I identify the Kusānas as the main political group who transform this deity. This conclusion related to the foreign influence in the development of this deity lie in stark contrast to previous studies of this deity.

iii

Acknowledgements

I was able to do research at a number of institutions in the process of my research and I would like

to thank those organizations that helped to sponsor my work. I would like to thank Ellen Raven at the Kern

Institute of Indology, Leiden University, who first explained the importance of Indian coinage in a study of

this type and the need to examine the coins first hand. I would like to thank Joe Cribb and Elizabeth

Errington at the British Museum for their help while I studied there in the spring of 2001. I am also greatly

indebted to Dr. Jha and all the scholars and staff at the Indian Institute for Research in Numismatic Studies

where I studied in the summer of 2001. I would also like to thank Dr. Savita Sharma at the Bharat Kala

Bhavan, the Director and his staff at the Indian Museum, Dr. Sharma at the National Museum, the Director

and staff of the Archaeological Survey of India, the Director and staff at the Mathura Museum and Gwalior

Museum and the scholars and staff at the American Institute for Indian Studies for allowing me access to

their coin and photographic collections. The library staff at the Deccan College and the Bhandarkar Oriental

Institute in Pune were also of great help to me. I also received assistance from Hans Bakker and his

colleagues at the University of Groningen. I would also like to thank the participants of the Images in Asian

Religions: Texts and Contexts conference held at McMaster University, May 2001, for their advice on the

early stages of my work. I am also grateful to the Social Sciences and Humanities Research Council of

Canada for funding the period of much of this research as well as the Department of Religious Studies at

McMaster University. I also received the following travel scholarships from various organizations at

McMaster for which I am grateful: James F. Harvey and Helen S. Harvey Travel Scholarship and the Mary

Margaret Scammell Travel Scholarship.

I would also like to thank the members of my committee for the assistance they have given to me.

In particular I would like to thank Dr. Koichi Shinohara and Dr. Alan Mendelson for their careful reading of

the manuscript and their advice. Two past members of my committee also deserve recognition, Dr.

Gretchen Ulmholtz and Dr. David Kinsley. Sadly, David passed away in the early stages of the dissertation,

iv

and Gretchen left McMaster towards the end of its completion. I hope they would have been proud of the

final result.

It is hard to express the gratitude I feel towards my supervisor Dr. Phyllis Granoff. The idea for

this topic, many of the central ideas expressed in this text and its timely completion are due to her input and

direction. It has been an honour to be Phyllis's student, and I can only hope that my future work displays the

richness of my training.

I would be remiss not to mention my fellow students Tina Marie Jones and Lavanya Vemsani.

Both Tina and Lavanya have supported this effort, and I am grateful for the comradery they have provided

me with over the years. I would also like to thank Lavanya's family in Hyderabad for their wonderful

hospitality.

My parents and family have stood beside me throughout my studies and I would like to thank them

for a lifetime of financial and emotional support and encouragement. I would like to close by thanking my

wife, Paula, for her support through the process of my graduate life and the production of this dissertation.

In many ways a dissertation becomes a family affair dragging in to its effort, at times unwittingly, those

closest to us. I am grateful to you, Paula, for your help and encouragement over the years. I am doubtful that

the dissertation could have been finished without your love and support along the way.

V

Table of Contents

List of Abbreviations. IX

List of Illustrations. .X

1. Chapter One: Introduction and Preliminary Material. .1

1 1 Introduction ... .... 1 1.2 The Supposed Indus Valley Skanda-Kārttikeya. .6 1.3 Other Theories on the Origins of Skanda-Kärttikeya. .8 1.4 The Early Textual References to Skanda-Kārttikeya

2. Chapter Two: The Birth of Skanda-Karttikeya in the Aranyakaparvan. 15 2.1 The Mahabhärata and the Birth of Skanda-Kārttikeya ... 15 2.2 The Aranyakaparvan version of Skanda-Kārttikeya's birth and deeds. 19 2.3 The Birth of Skanda-Kärttikeya in the Aranyakaparvan: An Inauspicious Beginning. .21 2.3.1 The Inauspicious Skanda-Kārttikeya: Sexual impropriety and violence in his birth. .21 2.3.2 The Inauspicious Skanda-Kārttikeya: His tie to Grahas, Kumāras and Mātrs .. .29 2.3.3 The Inauspicious Skanda-Kārttikeya: Grahas and Skanda in the Suśrutasamhitā .. .... 38 2.3.4 The Inauspicious Skanda-Kārttikeya: magico-religious aspects of the Suśrutasamhitā .. .39 2.3.5 The Inauspicious Skanda-Kārttikeya: Grahas in the Suśrutasamhitā. 41 2.3.6 The Inauspicious Skanda-Karttikeya: Grahas in the Aranyakaparvan. .46 2.3.7 The Inauspicious Skanda-Kärttikeya: The worship of Skanda-Kārttikeya and Grahas in the Āranyakaparvan .... .50 2.3.8 The Inauspicious Skanda-Kārttikeya: The worship of Dhūrta in the Skandayāga .. .50 2.3.9 The Inauspicious Skanda-Kārttikeya: Final points on Grahas in the Āranyakaparvan .. 56 2.3.10 Conclusion to the Inauspicious Characterization of Skanda-Kārttikeya. 57 2.4 The Auspicious Skanda-Kārttikeya: The frame narratives in the Aranyakaparvan .. .58 2.4.1 The Auspicious Skanda-Karttikeya: The Vedic missing Agni frame narrative 59 2.5 The Auspicious Skanda-Karttikeya: The second frame narrative. .67 2.6 The Auspicious Skanda-Karttikeya: The Brahminical ritual taming of the deity. .70 2.7 The Auspicious Skanda-Kārttikeya: The senāpati appears .... .72 2.8 The Auspicious Skanda-Kärttikeya: Siva's fatherhood in the Aranyakaparvan .. .80 2.9 The Socio-political context of the Aranyakaparvan ... .86

3. Chapter Three: The Birth of Skanda-Kārttikeya in the Salyaparvan and the Anusasanaparvan ....... 90

vi

3.1 The Salyaparvan account of Skanda-Kärttikeya's birth and deeds .90 3.2 The Parentage of Skanda-Karttikeya in the Salyaparvan. 90 3.3 The Auspicious Skanda-Kärttikeya in the Salyaparvan: The military ascetic 92 3.4 Skanda-Kārttikeya's birth and deeds in the Anusāsanaparvan .. 100 3.5 The Parentage of Skanda-Kārttikeya in the Anusāsanaparvan .. 100 3.6 Agni's Parentage of Skanda-Kärttikeya and Gold in the Anusasanaparvan 105 3.7 Skanda-Kārttikeya in the Rāmāyana. 116 3.8 Conclusion to Epic Material 117

4. Chapter Four: The Earliest Depictions of Skanda-Kārttikeya on Coins. 120 4.1 Questioning the Earliest Issues Depicting Skanda-Kārttikeya. 120 4.1.1 Questioning the Earliest Issues Depicting Skanda-Karttikeya: The Punch Marked Coins and Ujjain Coins .. 120 4.1.2 Questioning the Earliest Issues Depicting Skanda-Kārttikeya: The Mitras of the Punjab or the Audumbara Mitra Coins .. .124 4.1.3 Questioning the Earliest Issues Depicting Skanda-Kārttikeya: The Ayodhyã Coins .. .127 4.2 Yaudheya Skanda-Kārttikeya (Brahmanyadeva) Coins. 130 4.2.1 The Yaudheya Coins: The First Series. 130 4.2.2 The Yaudheya Coins: Problems of Classification 134 4.2.3 The Yaudheya Six-Headed Coins: Six-Headed Brahmanyadeva with a bar. ..... 135 4.2.4 The Yaudheya Six-Headed Coins: Six-Headed Brahmanyadeva with bars only on the top row of heads ....... 136 4.2.5 The Yaudheya Six-Headed Coins: Six-Headed Brahmanyadeva with a T 137 4.2.6 The Yaudheya Six-Headed Coins: Six-Headed Brahmanyadeva with dots. 138 4.2.7 The Yaudheya Six-Headed Coins: Six-Headed Brahmanyadeva without features .... 138 4.2.8 The Yaudheya Six-Headed Coins: Unusual Six-Headed Brahmanyadeva. 139 4.2.9 The Yaudheya Single-Headed Coins: Single-Headed Brahmanyadeva with Balls and Dots or a Multi-Headed Brahmanyadeva. .140 4.2.10 The Yaudheya Single-Headed Coins: Single-Headed Brahmanyadeva with Dots and Bars. .... 141 4.2.11 The Yaudheya Single-Headed Coins: Single-Headed Brahmanyadeva with a T-device .. .143 4.2.12 The Yaudheya Single-Headed Coins: Single-Headed Brahmanyadeva with Single Bar ... 143 4.2.13 The Yaudheya Single-Headed Coins: Single-Headed Brahmanyadeva with rays. 143 4.2.14 The Yaudheya Single-Headed Coins: Single-Headed Brahmanyadeva with rays and a T ... 144 4.2.15 The Yaudheya Class Three Coin Legends .. 145 4.2.16 The Foreign Influence on Yaudheya Class Three Issues. 152 4.2.17 Chronology and Context. 162 4.2.18 Yaudheya Class Three Coins: The Origins and Meaning of the Six Heads 174 4.2.19 The Importance of the Single-Headed Class Three Type ... 182 4.2.20 The Worship and Popularity of Brahmanyadeva and Sasthi. 183 4.2.21 Conclusion to Yaudheya Class Three Coinage. 186

vii

Chapter Five: Skanda-Kārttikeya on Kusāna and Class Six Yaudheya Coinage and Statues ..... 187 5.1 Kusāna Statues Depicting Skanda-Kārttikeya. 187 5.2 Kusāna Coinage. 199 5.2.2 The Mahasena Coin of Huviska: A Description. ... 200 5.2.3 Huviska's Mahasena Issue: Economic and Archaeological Perspectives. 200 5.2.4 The Skanda-Kumara with Visākha Coin Type of Huviska: A Description ....... .209 5.2.5 The Skanda-Kumāra with Visakha Coin Type of Huviska: Two Forms of the Same Deity or Two Different Deities ?. 211 5.2.6 The Dioscuri and their Influence on the Skanda-Karttikeya with Visakha coin. .... 214 5.2.7 The Skanda-Kärttikeya with Visakha Coins of Huviska: Concluding Remarks .221 5.2.8 The Skanda-Kumara, Visākha and Mahasena Coins of Huviska: A Description. .222 5.2.9 The Rise of Mahāsena. .223 5.2.10 Skanda-Kārttikeya and Mars in Kusāna Era Material .. 225 5.3.1 Yaudheya Coins: The Class Six Issues .. 234 5.3.2 The Class Six Issues: From Six Heads to One and the Impact of Foreign Design .... 237 5.3.3 Yaudheya Class Six Coins: The Context of the Yaudheyas and the 'Popularity' of Skanda-Kärttikeya at the Time of the Class Six Issues. .240 5.3.4 Yaudheya Coins: The Meaning of Skanda-Karttikeya on the Class Six Coins 248

Chapter 6: Conclusion 6.1 Conclusion. 252

Bibliography. 256

Illustrations. 270

Maps. .. 306

viii

Abbreviations

AIIS American Institute for Indian Studies ASI Archaeological Survey of India AV Atharvaveda AVP Atharvaveda Parisista BM British Museum BP Bhagavata Purāna DA Dharmasūtra of Āpastamba DB Dharmasūtra of Baudhāyana DG Dharmasūtra of Gautama DV Dharmasūtra of Vasistha IIRNS Indian Institute for Research in Numismatic Studies JB Jaiminīya Brāhmana KB Kaušītaki Brāhmana KS Kāthaka Samhitā LIMC Lexicon Iconographicum Mythologiae Classicae Mbl Mahābhārata MS Maitrāyanī Samhitā OED Oxford English Dictionary R Rāmāyana RV Rgveda ŚB Śatapatha Brāhmana SŚ Suśruta Samhitā TB Taittirīya Brāhmana TS Taittirīya Brahmana Uttara Uttaratantra of the Suśrutasamhitā YJ Yavana Jātaka

ix

List of Illustrations

Figure 1. Indus Valley seal (Lannoy 1971).

Figure 2. British Museum, Punch Marked Coin (Allen 1936: pl. 41, fig. 10).

Figure 3 Ujjain coins (Bopearachchi and Pieper 1998: pl. 5, figs. 4-5).

Figure 4. Three-headed Ujjain coin (Bopearachchi and Pieper 1998: pl. 5, fig. 21).

Figure 5. Punjab Mitra, Āryamitra (Dasgupta 1974: pl. 2, fig. 26).

Figure 6. Punjab Mitra, Mahīmitra (Dasgupta 1974: pl. 2, fig. 28).

Figure 7. Kubera/Yaksa (Coomaraswamy 1923:47).

Figure 8. Ayodhyā coins, Aryamitra (Bopearachchi and Pieper 1998: pl. 27, fig. 3), Vijayamitra (Bopearachcchi and Pieper 1998: pl. 27, fig. 4), Devamitra (Smith 1906: pl. 19, fig. 18).

Figure 9 Yaudheya class three coin: Six-headed Brahmanyadeva with a bar (IIRNS photograph 610/2-28).

Figure 10. Yaudheya class three coin: Six-headed Brahmanyadeva with a bar (IIRNS photograph 610/2-28).

Figure 11. Yaudheya class three coin: Six-headed Brahmanyadeva with a bar (IIRNS photograph 605/3-35).

Figure 12. Yaudheya class three coin: Six-headed Brahmanyadeva with a bar (IIRNS photograph 606/2-6).

Figure 13. Yaudheya class three coin: Six-headed Brahmanyadeva with a bar, silver issue (Mitchiner 1976 IV:331).

Figure 14. Yaudheya class three coin: Six-headed Brahmanyadeva with bars only on the top row of heads (IIRNS photograph 606/1-24).

Figure 15a. Yaudheya class three coin: Six-headed Brahmanyadeva with a T (Bopearachchi and Pieper 1998: pl. 23, fig. 1).

Figure 15b. Yaudheya class three coin: Six-headed Brahmanyadeva with a T (Bopearachchi and Pieper 1998: pl. 23, fig, 2).

Figure 16 Yaudheya class three coin: Six-headed Brahmanyadeva with a T (IIRNS photograph 612/2-8).

x

Figure 17 Yaudheya class three coin: Six-headed Brahmanyadeva with dots (IIRNS photograph 606/2-8).

Figure 18 Yaudheya class three coin: Six-headed Brahmanyadeva without features (IIRNS A181/3-2).

Figure 19a Yaudheya class three coin: Single-headed Brahmanyadeva with balls and dots (D. Mukherjee 1983).

Figure 19b Yaudheya class three coin: Single-headed Brahmanyadeva with balls and dots (D. Mukherjee 1983).

Figure 19c Yaudheya class three coin: Single-headed Brahmanyadeva with balls and dots (D. Mukherjee 1983).

Figure 20 Yaudheya class three coin: Single-headed Brahmanyadeva with balls and dots (IIRNS photograph 604/1-13).

Figure 21 Yaudheya class three coin: Single-headed Brahmanyadeva with balls and dots (IIRNS photograph 605/3-19).

Figure 22 Yaudheya class three coin: Sasthi with balls or dots around a single head (IIRNS photograph 607/2-12).

Figure 23 Yaudheya class three coin: Brahmanyadeva with dots and bars (IIRNS photograph 611/2-12).

Figure 24 Yaudheya class three coin: Brahmanyadeva with dots and bars (IIRNS photograph 611/2-13).

Figure 25 Yaudheya class three coin: Brahmanyadeva with dots and bars (IIRNS photograph 609/3-13).

Figure 26 Yaudheya class three coin: Single-headed Brahmanyadeva with a T (IIRNS photograph 612/2-15).

Figure 27 Yaudheya class three coin: Single-headed Brahmanyadeva with a single bar (IIRNS 610/1-13).

Figure 28 Yaudheya class three coin: Single-headed Brahmanyadeva with rays (IIRNS 606/2-13).

Figure 29 Yaudheya class three coin: Single-headed Brahmanyadeva with rays (Bopearachchi and Pieper 1998: pl. 23, fig. 3).

Figure 30 Yaudheya class three coin: Single-headed Brahmanyadeva with rays and a T (IIRNS photograph 612/3-22).

Figure 31 Yaudheya class three coin: questionable attribution (IIRNS photograph 609/2-11).

Figure 32 Ai-Khanum cion (Narian 1973).

Figure 33 Silver issues of Demetrius I depicting Herakles (Gardner 1866: pl. 2, fig. 9).

Figure 34 Silver issue of Euthydemus II depicting Herakles (Gardner 1866: pl. 2, fig. 3).

Figure 35 Issue of Spalagadams depicting Zeus (Gardner 1866: pl. 21, fig. 10).

Figure 36 Issue of Spalirises depicting Zeus (Gardner 1866: pl. 22, fig. 1).

Figure 37 Issue of Spalirises depicting Zeus (Gardner 1866: pl. 22, fig. 2).

Figure 38 Silver issue of Demetrius I depicting Herakles (Gardner 1866: pl. 2, fig. 12).

Figure 39 Silver issue of Euthydemus II depicting Herakles (Mitchiner 1975 I:62).

Figure 40a Issue of Demertius depicting Artemis (Gardner 1866: pl. 3, fig. 1).

Figure 40b Issue of Demertius depicting Artemis (Mitchiner 1975 I:60).

Figure 41 Panel from Mathura depicting Sasthī flanked by Skanda and Viśākha (Joshi 1986: fig. 45).

Figure 42 Statue from Gandhära depicting Skanda-Kārttikeya (Zwalf 1996: fig. 102).

Figure 43 Statue from Mathurā depicting Skanda-Kārttikeya (Rosenfield 1993: fig. 49).

Figure 44 Statue from Sonkh (Mathurā) depicting Skanda-Kārttikeya (Härtel 1993:282, fig. 1).

Figure 45 Statue from Mathurā depicting Skanda-Kārttikeya (Asthana 1999:33, fig. 29).

Figure 46 7th century Gandhära diptych depicting Skanda-Kārttikeya with Sasthī (Czuma and Morris 1985:162, fig. 80).

Figure 47 Panel from Mathurā depicting Skanda-Kärttikeya with Mothers (Joshi 1986: fig. 33).

Figure 48 Panel from the Lucknow Museum depicting Skanda-Kärttikeya with Mothers (Joshi 1986: fig. 28).

Figure 49 Statue from Gandhāra depicting Hāriti (Zwalf 1996: fig. 92).

Figure 50 Statues from Gandhāra depicting Hāritī with Pañcika (Rosenfield 1993: fig. 61).

Figure 51 Gold issue of Huviska depicting Mahäsena (Göbl 1984: pl. 22, fig. 7).

Figure 52 Damaged Huviska die (Rosenfield 1993: fig. 66).

Figure 53 Forgery of a Huviska coin (Göbl 1984: pl. 129).

xii

Figure 54 fig. 135). Gold issue of Huviska depicting the king with a bird standard (Rosenfield 1993:

Figure 55a Issue of Azilises depicting a single Dioscuri (Gardner 1886: pl. 20, fig. 8).

Figure 55b Issue of Hyrcodes depicting an unknown reverse figure (Gardner 1886: pl. 24, fig. 10).

Figure 56a Crude terracotta from Khairadih depicting Naigamesa (Jayaswal 1991: pl. 8, fig. 24).

Figure 56b Crude terracotta from Khairadih depicting Naigamesī (Jayaswal 1991: pl. 9, fig. 28).

Figure 57 Statue from Mathurā of Naigamesa (AIIS photograph 779-80).

Figure 58 Statue from Mathurā of Naigamesī (ASI photograph 1167/92).

Figure 59 Gold issue of Huviska depicting Skanda-Kumāra with Visākha (Göbl 1984: pl. 13, fig. 156).

Figure 60 Gold issue of Huviska depicting Skanda-Kumāra with Viśākha (Göbl 1984: pl. 13, fig. 156).

Figure 61 Gold issue of Huviska depicting Skanda-Kumāra with Visākha (Gobl 1984: pl. 13, fig. 156).

Figure 62 a-c Gold issues of Huviska depicting Skanda-Kumāra with Viākha (Göbl 1984: pl. 14, fig. 166).

Figure 63 Gold issue of Huviska depicting Mao and Miiro (Gobl 1984: pl. 22, fig. 295).

Figure 64a Gold issue of Huviska depicting Oesho and Ommo (Gobl 1984: pl. 23, fig. 310).

Figure 64b Gold issue of Huviska depicting Oesho and Nana (Gobl 1984: pl. 168, fig. 167/3).

Figure 65a Roman coin depicting Mars giving a statuette of Victory to a ruler (LIMC 1986 III: fig. 412).

Figure 65b Roman coin depicting Mars and Rea Silvia (LIMC 1986 III: fig. 313).

Figure 66a Copper coin of Diomedes depicting the Dioscuri (Bopearachchi and Pieper 1998: pl. 43, fig. 163).

Figure 66b Copper coin of Diomedes depicting the Dioscuri (Bopearachchi and Pieper 1998: pl. 43, fig. 165).

Figure 67 Issue of Azilises depicting the Dioscuri (Gardner 1886: pl. 20, fig. 6).

Figure 68 Gandhäran panel possibly depicting the Dioscuri (Gnoli 1963: fig. 3).

xiii

Figure 69a Gandhäran panel possibly depicting the Dioscuri (Gnoli 1963: fig. 7).

Figure 69b Gandhäran panel possibly depicting the Dioscuri (Gnoli 1963: fig. 8).

Figure 70 Gold coin of Huviska depicting Mahāsena, Skanda-Kumāra and Visākha (Göbl 1984: pl. 13, fig. 157).

Figure 71 Gold coin of Huviska depicting Mahāsena, Skanda-Kumāra and Visākha (Gobl 1984: pl. 165, fig. 227A/1).

Figure 72 Yaudheya class six coin (Bopearachchi and Pieper 1998: pl. 23, fig. 5).

Figure 73 Yaudheya class six coin with 'dvi' added to legend (IIRNS photograph A/209-21).

Figure 74 Yaudheya class six coin with 'tri' added to legend (IIRNS photograph 556/3-9).

Figure 75 Not available.

Figure 76 Bactrian seal and sealing depicting Skanda-Kärttikeya as Mars (Callieri 1997: pl. 58, fig. U 7.5).

Figure 77 Bactrian seal and sealing depicting Skanda-Kärttikeya as Mars (Callieri 1997: pl. 19, fig. 7.2).

Figure 78 Panel from Gandhāra (Gnoli 1963: fig. 1).

Figure 79 Third section of the panel illustrated in 78 (Taddei 1966: fig. 1).

xiv

1

Chapter One: Introduction and Preliminary Material 1.1 Introduction

By the seventh century CE the cult of Skanda-Karttikeya' is a minor one in the north of India.

Textual and sculptural representations of him from this time all depict him as the son of Siva and the

general of the army of the gods, who was born to destroy Tarāka, an asura (demon). In these accounts he is

largely viewed as an auspicious but a minor figure when compared to the cults of other Saivite figures like

Śiva, Pārvatī and Ganeśa. Skanda-Kärttikeya was not, however, always primarily recognized as the son of

Siva, as a martial figure or as a minor deity in ancient north India. It is the study of this early figure to which

this dissertation is devoted. While there has been a great deal of speculation concerning the origins of

Skanda-Kārttikeya, no scholar has yet to present a well-supported argument for the exact nature of his

origins, and the early development of his cult is something of a mystery to previous scholarship. This

dissertation will present an argument concerning the origins of this deity and the progression of his early

cult in the north of India, and will question and correct many of the commonly held scholarly assumptions

regarding this deity's cult and characterization.

This thesis will argue that the origins of Skanda-Kärttikeya can be found within unorthodox Graha

and Mätr cults.2 Both of these cults deal with inauspicious and dangerous beings who threaten the health

and safety of children. It will be my contention that Skanda-Kärttikeya begins as an inauspicious Graha-like

figure whose worship was directed towards propitiating him. Once propitiated he was regarded as a

protective deity whose inauspicious powers could be directed towards others. It will be my argument that

'Throughout this thesis I will use the hyphenated name Skanda-Karttikeya when making general remarksabout this deity. The actual hyphenated name 'Skanda-Kārttikeya' does not appear as a proper name of the deity from within the tradition, but I have selected it to avoid any confusion which may arise if I were to use an assortment of his numerous epithets. I will, however, use specific names of the deity when the context calls for such a usage.

2Grahas, or 'graspers', are demons who attack young children and pregnant women in Hinduism. Mätrs, or 'mothers', fulfil a similar function in the traditions this dissertation examines.

2

this inauspicious and dangerous aspect of his character, which could be transformed through ritual, was at

the root of his popular cult in north India beginning at approximately the fourth century BCE. It is within

this non-Vedic and non-Brahminic ritual milieu that, I argue, Skanda-Kārttikeya received widespread

worship.

I will then trace the development of his cult that sees a transformation in his characterization from

inauspicious and Graha-like to auspicious, martial and part of the orthodox Hindu fold. I will argue for two

main forces behind this shift. The first is religious and the second is political. The religious force is

Brahminical, which attempts to transform a non-orthodox figure, from the standpoint of Vedic and

emerging Hindu perspectives, into an orthodox figure. The political forces are primarily foreign in origin,

and they influence the shift from dangerous Graha to respectable general. As we shall see, these two

catagories of religious and political forces are not completely separate entities, but part of the same dynamic

which informs the changes in Skanda-Karttikeya's cult during this period. It will be my argument that this

development occurs from the first century BCE to the fourth century CE, and it is primarily to the Kusāna

era that this re-characterization of the deity can be traced. This aspect of my thesis lies in sharp distinction

to previous scholarship on this deity. Previous scholars have been quick to acknowledge that Skanda-

Kärttikeya was 'popular' with foreign groups, but do not recognize the importance of these groups in

shaping the martial character of this deity.

I will also argue, contrary to the arguments presented by previous scholars, that the shift towards

an auspicious, martial deity with respected parentage does not lead to this god's widespread popularity, but

actually undermines his earlier widespread ritual cult based in propitiation, and replaces it with a cult

focused on royal and elite segments of north Indian society. It will be my contention that the cult loses

widespread popularity in the north well before its supposed golden age in the Gupta period. The apparent

popularity of the deity in the Gupta era is a continuation of royal and elite support for the deity, as opposed

to the widespread support he received before the Kusana era. With the end of the Gupta Empire and the end

of royal support for this deity in the north, his cult all but disappears, having lost most of its widespread

following well before this period.

To trace the history of this cult and to demonstrate my hypothesis I will use evidence from a

mixture of sources including texts, statues, coins and inscriptions. The primary text I will use to demonstrate

my points is the Mahabharata. There are three accounts of Skanda-Kärttikeya's birth and deeds in this text.

The longest, and I will argue the most important, is found in the Aranyakaparvan. I will argue that the

Äranyakaparvan should be read as a mythologized summary of the actual shifts in characterization this

deity undergoes over time. The Aranyakaparvan begins with a depiction of Skanda-Kārttikeya as a

dangerous and inauspicious deity who is closely associated with similar classes of child-afflicting demon-

like figures, and who is controlled through worship. He is, however, transformed into an auspicious military

figure and the son of Siva by the end of the text. I also argue that this version of his birth story should be

read as a conversion story which attempts to transform a non-Vedic deity into one with as many links to the

Vedic tradition as possible. It uses a theme of domestication to explain the dramatic changes this deity

undergoes from the beginning of the narrative to its end. Part of the agenda behind the transformation of

Skanda-Kārttikeya in this text, I argue, is to present a deity who fits into an orthodox or Brahminical

concept of an object of worship without threatening the position of other deities like Agni and Śiva.

It will be my contention that the cult of Skanda-Karttikeya does begin as the Āranyakaparvan

describes, with a dangerous Graha-like deity who is propitiated through worship. While some scholars have

recognized that Skanda-Karttikeya's cult begins in what they call "folk" or "tribal" cultures, they have failed

to recognize that his origins can be localized and given a precise context within a Graha milieu, and they

have failed to recognize the widespread popularity of the cult associated with this aspect of the deity. I will

also use texts such as the Atharvaveda Parisista and the Surutasamhita to support these points concerning

Skanda-Kärttikeya's origins. I will then argue that by the end of the Aranyakaparvan most of the original

reasons for worshiping Skanda-Kārttikeya are removed, as an image of an auspicious military deity and son

of Siva is constructed at the expense of his earlier inauspicious Graha-like characterization.

I next discuss the two other versions of Skanda-Kärttikeya's birth narrative found in the

4

Mahabharata and a single version found in the Ramayana. These descriptions of his birth and deeds will

demonstrate that the end result of the Aranyakaparvan version quickly becomes the normative textual

understanding of this deity. His links to Grahas and Mätrs are largely eliminated, and his role as an

auspicious general becomes the focus of these accounts. He is also used in these accounts as a means to

extol the power of his fathers: Agni and Siva.3 In these narratives the importance of worshiping Agni and

Śiva is emphasized as opposed to that of Skanda-Kārttikeya. These Epic narratives do not reflect the

popularity of Skanda-Kärttikeya as many scholars have supposed, but actually indicate that he was a means

to establish the power and superiority of other deities, and that his cult faltered over time due to his change

of character.

While the textual material does point to a mythologized version of the actual progress of this

deity's characterization and the development of his cult, it does not supply any indication of the historical

groups involved in the transformation of his cult or the specific mechanics of this transformation. My

sources for tracing the historical development of his cult come almost exclusively from archaeological

sources. I begin the section on archaeological sources by arguing against scholars who attempt to trace the

first images of this deity to Punch Marked Coins. I also reject arguments that claim Skanda-Kārttikeya, or

references to him, can be found on the coinage of the Mitra rulers of the Punjab or on the coinage of the

Ayodhya rulers. I will argue that the first numismatic representation of the deity is found on what are

commonly referred to as class three Yaudheya coins. The Yaudheyas were an indigenous group who ruled

in areas of the modern states of Haryana, Himachal Pradesh and the Punjab at different points in time. I will

present a detailed catalogue of these coins and argue for a new date for their production of between the last

decades of the first century BCE to the close of the first century CE. From these coins I will establish that

there were two ways of representing Skanda-Kärttikeya: either as six-headed or single-headed. I will argue

that the six-headed version of the deity reflects the dangerous Graha-like image of him. I will then

3 Skanda-Kärttikeya is described as having multiple parentage, and this aspect of his narratives will be discussed in more detail below.

5

demonstrate that the single-headed versions are strongly influenced by Indo-Greek, Indo-Scythian and Indo-

Parthian coinage and present a more softened understanding of the deity, which is the image that endures

over time.

I then discuss Kusāna era statuary of this deity. Here I demonstrate a separation between the

statuary of Gandhara and the far north-west of the subcontinent, and the statuary of Mathurā. This

comparison will reveal that the martial understanding of the god comes primarily from the far north-west of

the subcontinent and is largely influenced by non-Indian forces. The Mathura region does show some of the

influence of the martial Skanda-Kärttikeya, but the majority of the panels and statues depicting this deity

from Mathura reflect a Matr and Graha context, which I argue is the dominant initial indigenous

understanding of the deity.

I will then present a detailed study of the coins of the second century CE Kusāna king Huviska

which depict Skanda-Kärttikeya. Here I argue that the context of these coins is both martial and directed at

elite society. The Graha and Mätr links to this deity are removed from these coins and with it the reason for

his widespread cult. The Kusana influence on this cult transforms it into an elite and martial cult at the

expense of its popular roots. The Kusanas are, I think, the primary political force behind the transformation

of this deity described in the Āranyakaparvan.

The final group of coins I present are the class six Yaudheya coins which date to the third and

fourth centuries CE. These coins demonstrate that the foreign or Kusäna understanding of Skanda-

Kärttikeya as a martial deity was absorbed by indigenous groups who once recognized the deity as a Graha-

like figure. The Graha and Mātr connections to Skanda-Kārttikeya all but disappear in the post-Kusāna age,

and the iconographic representations of him in the north of India all emphasize his martial aspect. This

section will also explore the 'popularity' of Skanda-Kärttikeya among the Yaudheya people. I will

demonstrate through evidence supplied by seals and sealings that the cult of this deity was already a minor

one at this time when compared to the cults of Siva and Visnu. I argue from this evidence that this deity lost

the base of his widespread propitiation cult due to his re-characterization. He was no longer 'popular'

6

except with royal and elite groups.

The conclusion of this thesis demonstrates from the points mentioned above that the so- called

golden age of this god as coming in the Gupta era is not an accurate depiction. Without doubt he was

popular with kings and other elite figures, but his ritual and the cult to propitiate him continue to decline

during this period. Indeed, he cannot be called popular over a wide spectrum of society. With the fall of the

Guptas his cult also falls without the broad support from the non-elite sections of society that once

worshiped him for their children's safety.

1.2 The Supposed Indus Valley Skanda-Kārttikeya

Before discussing the Mahabharata and my other textual sources, I will briefly review some

arguments presented by other scholars who argue for a much earlier origin for the deity. T. G. Aravamuthan

(1948), B. Y. Volchok (1970) an d W. Doniger (1973; 1975) have argued that scenes found on Indus Valley

seals and sealings relate to the cult of Skanda-Kärttikeya. They refer to five related seals/sealings, one of

which depicts seven females in the foreground and a tree in the background with a figure standing in the

middle of its branches. There is also a kneeling figure before the tree and what may be a severed head,

along with a bull-like animal behind the kneeling figure. This sealing is illustrated in figure one. The other

related seals/sealings show six females and some do not show the severed head. Aravamuthan argues that

the females represented are the Krttikas, which leads him to believe that the kneeling male figure is Skanda-

Kārttikeya (1948:52, 54). As we shall see, the Krttikas (the Pleiades) play a significant role in the birth

stories of Skanda-Karttikeya. Aravamuthan also thinks the figure in the tree is the impersonal brahman

(1948:58-59), but others have suggested it represents Brahmã (Rana 1995:3). Aravamuthan also claims that

the bull is a composite bull-goat and represents Agni, and that the severed head is a trophy from

vanquishing demons (Aravamuthan 1948:52). As we shall see, the narrative Aravamuthan tries to read into

these seals/sealings is found in the Aranyakaparvan. While S. S. Rana accepts this theory (1995:2-4), P. K.

Agrawala is correct in rejecting it as "mere conjecture" (1967:xiii). These Indus Valley seals/sealings are

open to interpretation, and there is nothing to back any one interpretation of them. There is also a significant

7

historical gap of at least one thousand to two thousand years between the production of these seals/sealings

and the first written stories of Skanda-Kārttikeya in the Mahābhārata, which makes any interpretation of

these seals/sealings from the known mythology of Skanda-Kärttikeya dubious at best.

W. Doniger and B. Y. Volchok argue that these seals/sealings represent the celebration of the birth

year of Skanda-Kärttikeya with its six seasons, which are represented by his six heads in later accounts. This

birth year is supposed to have occurred during the new moon at the spring equinox when the sun was in the

Pleiades. In other words, they argue that Skanda-Kärttikeya's birth commemorates a rare astrological

occurrence that has Agni, in the form of the sun, enter the constellation, the Krttikas. This rare astrological

event occurred, they claim, during the third millennium BCE (Volchok cited in Doniger 1973:100).4 Similar

criticisms can, however, be leveled against this theory as those listed above. Even though Volchok is able to

supply evidence for his unusual astrological event, we have no way of knowing if the Indus Valley people

kept track of such events, let alone if such events led to the cult of Skanda-Kārttikeya.5

A. Parpola attempts to relate finds of bangles at Indus Valley sites and an Indus Valley script sign

found on them with the cult of Skanda-Kärttikeya. The argument is speculative and relies on the assumption

that certain Indus Valley signs relate to heavenly bodies and that the Indus script relates to Dravidian

languages (1990:265). The author argues that a particular Indus sign relates to the Old Tamil word muruku,

4Volchok's argument concerning Skanda-Kärttikeya comes from a Russian article he published in 1972, "Protoindiiskie paralleli k mifu o Skande" in Proto-Indica (Akademiia Nauk, SSSR, 305-112), which was summarized by Doniger (1973:100; 1975:104).

5In an earlier article Volchok suggests that certain elements of Indus Valley art reflect notions found in Buddhist and Hindu cosmography, but these points are also speculative (1970:29-53). In this context he argues that part of Skanda-Kärttikeya's mythology can be related to Indus Valley seals. Specifically, he refers to a group of seals that appear to depict a buffalo being killed by a human with a spear. He argues that this image relates to the Aranyakaparvan account of Skanda-Kärttikeya killing the demon Mahisa (1970:45-46). He also claims that the figure on the seal relates to a figure on Punch Mark Coins whom he also understands to be Skanda-Kārttikeya. I will reject this argument in section 4.1.1. His arguments for these seals are not well supported in part due to the unsupportable assumptions he brings to this Indus Valley evidence, and the considerable time gaps between the seals and the later evidence he employs to explain them.

8

which the author than relates to the Tamil name Murukan6 and to a Tamil term for 'ring' (1990:270).

Parpola goes on to speculate, based primarily in modern uses of bangles in India, that these ancient bangles

and the deity, whose name supposedly appears on them, are connected to pregnancy in the Indus Valley

civilization (1990:272-277). This argument is, however, built on a number of unsubstanciated hypotheses

related to the Indus script and the relation of this culture to Vedic and Hindu culture. Hence, any reading of

these seals or script remains without historical basis, and I find these arguments that attempt to trace the

origins of Skanda-Kärttikeya to the Indus Valley to be without merit.

1.3 Other Theories on the Origins of Skanda-Kārttikeya

P. K. Agrawala presents a hypothesis that the eventual deity Skanda-Kärttikeya is an amalgam of

Kumāras ('boys' or 'youths'), a composite of various Vedic beings sharing the name Kumāra (1967:1-11).

One of Skanda-Kärttikeya's main epithets is Kumāra, which provides the basis for Agrawala's argument. I

will briefly review Agrawala's idea here. He begins by noting that the word kumāra occurs seventeen times

in the Rgveda where it usually means son, child or youth (1967:1). It is used as an epithet of Agni to

describe his state when a 'young' or freshly kindled fire as in Rgveda 5.2. He states that in Rgveda 10.135

Kumara appears as an independent deity associated with Yama (1967:2). He also claims that Rgveda

2.33.12 states that the poet says he bows to Rudra as Kumära to his father (1967:4). In both of these cases,

however, Kumära may simply be translated as 'boy', and does not refer to a deity, or some proto-form of

Skanda-Kārttikeya.7 Rgveda hymns 5.2 and 10.135 are also attributed to Āgneya Kumāra who Agrawala

takes to be a personification of the Kumära aspect of Agni (1967:3), but this seems more like a clever use of

epithets than the establishment of a separate entity called Kumara. The point, I think, that we should take

from Agrawala's section on Kumara and the Rgveda is that Agni is at times called Kumāra, meaning

6Murukan is a Tamil deity who is eventually assimilated with Skanda-Kārttikeya, but this connection is not attested to until the fourth or fifth centuries CE (Clothey 1978:62) and will not be discussed in this dissertation.

1Geldner takes the word to just mean boy "wie der Knabe gegen den lobenden Vater, so habe ich mich (gegen dich), Rudra, verneigt, wenn du nahst" (RV 2.33.12 translated by Geldner 1951 I:318).

9

'young'; his other observations are questionable.8

Agrawala then goes on to discuss the following passages from the Śatapatha Brāhmana:

Now, those beings are the seasons; and that lord of beings is the year; and that Ushas, the mistress, is the Dawn. And these same creatures, as well as the lord of beings, the year, laid seed into Ushas. There a boy (kumāra) was born in a year: he cried. Prajāpati said to him, 'My boy, why criest thou, when thou art born out of labour and trouble?'He said, 'Nay, but I am not freed from (guarded against) evil; I have no name given me: give me a name!' ... He said to him, 'Thou art Rudra.' And because he gave him that name, Agni became such-like (or, that form), for Rudra is Agni: because he cried (rud) therefore he is Rudra. (6.1.3.8-10 translated by Eggeling 1882-1900 III:158-159)

This Kumära keeps demanding names, however, and is also named Sarva, Paśupati, Ugra, Aśani, Bhava,

Mahādeva and Īśāna (6.1.3.11 -. 17). The text concludes with:

These then are the eight forms of Agni. Kumara (the boy) is the ninth: that is Agni's threefold state. And because there are eight forms of Agni -- the Gäyatrī consisting of eight syllables -- therefore they say, 'Agni is Gäyatra.' That boy entered into the forms one after another; for one never sees him as a mere boy (kumara), but one sees those forms of his, for he assumed those forms one after another. (6.1.3.18-19 translated by Eggeling 1882-1900 III:160-161)

Based on these passages, Agrawala suggests a developing form of Agni called Kumära who will eventually

become Skanda-Kārttikeya (1967:6-7). The eight names given to Agni-Kumāra here are all epithets of

Rudra-Siva. The idea that Rudra is Agni is repeated in the Satapatha Brāhmana (1.7.3.1-8, 5.2.4.13,

9.1.1.1), and the above passage seems to be designed to make the same point. Kumāra is not likely a

separate deity here or a reference to Skanda-Kärttikeya. It is a form of Agni that dwells in these other forms

of Rudra to account for the equation Agni=Rudra.9

I do not think we can claim as does Agrawala that Skanda-Kärttikeya comes out of an Agni-

Kumāra who has "an amazing array of Kumaras being conceived around him", and who all share the

8Agrawala also cites Atharvaveda 4.31.11 as describing a Kumära in the form of a hairy Gandharva who looks like a monkey or a dog and who pursues a woman (1967:6). I do not, however, find AV 4.31.11 as hymn 4.31 appears to end at verse seven.

9Agrawala also notes Sanatkumāra from the Chändogya Upanisad as a Kumāra that makes up the Skanda-Kārttikeya amalgam (1967:11). I will comment on this idea in my section on the Salyaparvan. He also notes a Kumara found in the Paraskara Grhyasutra, which I will also discuss later.

10

common feature of "Kaumārya or celibacy" (1967:11).10 I think Agrawala has accepted too much of the

Epic's attempt to present Skanda-Kärttikeya as another form of Agni, or as arising out of a Vedic context.

Epic writers may look back to Agni as Kumära in an attempt to write Skanda-Kārttikeya into their tradition,

but that does not place his actual roots within an Agni-Kumāra figure. It merely shows us another strategy

employed by Epic writers to produce a Brahminical version of Skanda-Kārttikeya. Agrawala also does not

emphasize the Graha-like aspect of 'Kumara' which I will discuss as part of the likely origins of the deity.

Agrawala's point about celibacy is also questionable. Skanda-Kārttikeya does become associated with

celibacy, but it is a later Puranic tradition; we will find no explicit evidence of it in this account from the

Epic." What this section also makes clear is that Skanda-Karttikeya is not a Vedic deity. As we shall see,

while the Aranyakaparvan tries to link this deity to the Vedas, he is not Vedic himself and does not develop

out of a Vedic context.12

That Skanda-Karttikeya is derived from Agni-Kumara is only part of Agrawala's thesis concerning

the deity's origins. He states:

As a result of our enquiry it appears that the conception of Skanda as we find it in the Mahābhārata and in archaeology was the outcome of a long evolution in which several streams of cults and folk- beliefs combined. One such tradition was that of the Graha-Devatās or soil-born demoniacal deities in which Skanda is associated with the Mätrkās and other gnomes and ogresses emerged as their chief. As already noted, the second stream was Vedic in which the conception of Kumāra .... (1967:xiv-xv)

0Chatterjee also suggests that Skanda-Kārttikeya's "Kumāra"status implies a celibate status: "[a]s a true Kumāra and ascetic Skanda-Kārttikeya is often pictured as shunning the company of women" (1970:103). He is not able, however, to cite many instances of this shunning and only one, Kālidasa's Vikramorvasīyam, appears in be within the approximate time frame of this thesis.

11See the Brahma Purana 81.1-6 for an account of Skanda-Kärttikeya taking a celebate life and Doniger (1973:204) for a discussion of it. Also see Brhaddharma Purana (2.60.107-108) and Siva Purāna (2.4.20.23-37) for similar references. Shulman also lists some Tamil shrines where Skanda-Kārttikeya or Murukan is worshipped as a brahmacārin (1980:146).

12I should note that forms of Skanda-Kärttikeya do appear in the Atharvaveda Parisista and the Grhyasutra tradition, both of which could be considered part of the Vedic tradition. They are both, however, part of the late Vedic tradition and speak more to the rise of Hinduism than to the Vedas themselves.

11

To this list of what makes up the final form of Skanda-Kärttikeya Agrawala also adds Dhūrta from the

Dhūrtakalpa of the Atharvaveda Parisista, Kārttikeya as a "constellar concept", Guha as a "hilly demon",

and Naigamesa as "a deity of child-birth" (1967:xv). This is something of a 'sum of various parts' theory to

which most other Skanda-Kärttikeya scholars subscribe (Sinha 1979:145; Thakur 1981:14; Ghurye

1977:92). This general idea is not without merit, and I do not wish to contradict it. The eventual post-Epic

version of Skanda-Kārttikeya views names like Visākha and Naigamesa as epithets of one deity as opposed

to separate deities or forms of Skanda-Kārttikeya. Certainly, like many Hindu cults, Skanda-Kārttikeya

absorbs similar deities into his cult over time. Where I differ from Agrawala and others is in their

preference for a Skanda-Kärttikeya that comes out of a Vedic context. I think this Vedic roots assumption is

a naive acceptance of the Mahabharata's rhetorical agenda in constructing an orthodox Skanda-Kärttikeya.

While most scholars do acknowledge a Graha element to Skanda-Kārttikeya, none recognize the importance

of this element of the deity's roots, nor do they recognize his inauspicious past as the location of the origins

of his cult. Finally, other scholars have failed to explain Skanda-Kārttikeya's transformation into a military

deity or the role of foreign groups in bringing about this change.

1.4 The Early Textual References to Skanda-Kärttikeya

The earliest use of the name Skanda, likely occurs in the Chändogya Upanisad (7.26.2). It is

unclear, however, if this reference actually refers to Skanda-Kärttikeya, and it will be discussed in more

detail in the next chapter. For now it is enough just to mention that Skanda-Kārttikeya as a full-fledged

character does not appear until the Mahabharata and very late Vedic ancillary texts. He cannot be traced to

the Vedas themselves and cannot be traced beyond the approximate date of the fourth century BCE from the

evidence supplied by religious texts.13 These points will become clear as we continue.

Several scholars have argued that Skanda-Karttikeya is referred to in Kautilya's Arthasastra

(Meyer 1977:75-76; P. K. Agrawala 1967:17; Chatterjee 1970:30-31; Ghurye 1977:118-119; Sinha

13If we take a more conservative approach and attempt to date the appearance of Skanda- Kärttikeya's cult to the first certain statues of him we cannot trace the cult past the first century CE.

12

1979:41-42). At 2.4.17 of Kautilya's text he states that the king: "should cause to be built in the centre of

the city shrines for Aparājita, Apratihata, Jayanta and Vaijayanta as well as temples for Siva, Vaiśravana,

the Aśvins, Śrī and Madirā" (Kangle 1960:80). Kangle suggests that the names Aparājita, Apratihata,

Jayanta and Vaijayanta all refer to spirits of victory (1960:80). Kangle's suggestion seems reasonable, but

Meyer regards at least one of the names to refer to Skanda-Kārttikeya (1977:26). There is little direct

evidence to support this claim, but there are later commentaries on the text which attribute at least one of

the four to Skanda-Kārttikeya (Chatterjee 1970:30-31; Sinha 1977:42). These names do not, however,

appear as epithets of Skanda-Kärttikeya in any other element of the tradition, and it is clear from the above

quotation that Kautilya knew the normative names of other deities and used them. I suspect that if Kautilya

intended to mention Skanda-Kärttikeya he would have done so in an explicit manner, and I find Kangle's

suggestion to provide the most reasonable account of these names.

Another potential reference to Skanda-Kärttikeya is found in 2.4.19 of Kautilya's text. Here he

mentions that Brahma, Indra, Yama and Senapati should preside over the gates of a city. The scholars noted

above seem sure that Senäpati is a direct reference to Skanda-Kärttikeya, but some caution is required.

While Skanda-Kärttikeya becomes a divine senäpati (a general), it is not clear to me that this understanding

of the deity was current during the fourth century BCE, which is the approximate date of Kautilya's text

(Rangarajan 1990:19).14 I will argue in this thesis that the understanding of Skanda-Kārttikeya as a senāpati

is a later development in his cult that only really becomes evident in the first or second century CE. It is not

clear, then, that Kautilya did mean Skanda-Kārttikeya by his reference to Senāpati.

A more certain reference to Skanda-Kārttikeya and a related deity named Visākha15 can be found

in Patañjali's Mahabhasya, which is a second century BCE commentary on Pänini's fifth century BCE

14There is debate over the date of Kautilya and his text. Some scholars date the text as late as 150 CE. For a summary of these views see Rangarajan (1990:18-21). If the text is as late as the Common Era, I have fewer reservations about associating this Senāpati with Skanda-Kārttikeya.

15Visākha will be discussed in more detail in the next chapter.

13

grammar (V. S. Agrawala 1963:478). In his commentary on Pānini's 5.3.99 Patañjali mentions statues of

Śiva, Skanda and Visākha that he claims were used during the Mauryan Empire. Based on these comments

we can date the appearance of statues of Skanda-Kärttikeya and Visākha to Patañjali's time and perhaps

back to the Mauryan empire, which could place the statues in the fourth century BCE.16 We do, however,

have to be cautious with Patañjali, and we can question his knowledge of the Mauryans or his desire to

represent an accurate history through his grammatical commentary. Patañjali, it must be remembered, is not

trying to write an historical text. It is also difficult to speculate on Patañjali's understanding of these gods

from such a short reference other than to say that they were objects of worship from at least the second

century BCE and possibly back to the fourth century BCE. We can, however, only speculate as to the

apparent production of statues of these deities from this period because no statuary of Skanda-Kārttikeya or

Viśākha survives from that period.17

There may also an early reference to Skanda in the Deva Gayatrīs found in the Maitrāyanī

Samhita. This short Gayatri summons Skanda, who is also referred to as Kārttikeya and Kumāra in the text

(MS 2.9.1-2).18 The text has been dated to be not much earlier than the third century BCE (P. K. Agrawala

16P. K. Agrawala has attempted to push back the date of Patañjali's comments to Pänini himself in the fifth century BCE (1967:16). There is, however, no evidence to back such a claim.

17In relation to this point of early statuary of Skanda-Kārttikeya, P. G. Paul and D. Paul (1989) have argued that there may be a reference to a statue of Skanda-Kärttikeya in Quintus Curtius' history of Alexander the Great. Curtius claims that an image of Heracles/Hercules was carried in front of the army of King Porus when he went to battle Alexander. Paul and Paul argue that this image could have been Skanda- Kärttikeya. They note that a large number of Skanda-Kärttikeya statues have been recovered from the Gandhära area and do not think that the discrepancy of attributes (Skanda-Kārttikeya's spear and Hercles' club) is problematic. They also claim that the image was made of wood and that is why we have no surviving examples of this deity's statuary from this period (1989:114-116). I do, however, think that the discrepancy of attributes is a significant factor because the spear is one of the primary means we have of identifying early statuary of Skanda-Kārttikeya. While Skanda-Kārttikeya does appear to have been popular with certain classes in Gandhära, surviving statues of the deity from that region come 300 to 500 years after Alexander and do not provide a sound basis for their argument. Finally, their wooden statue argument only helps to underline the issue that there is no physical evidence to support their claim, and it must be viewed as speculative at best.

18“tat kumārāya vidmahe kārttikeyāya dhīmahi tannah skandah pracodayāt".

14

1967:12). The authenticity of these Deva Gayatris as part of the original Samhita has, however, been

questioned, and conclusions drawn from these passages regarding the dates associated with the early

worship of Skanda-Kärttikeya are not well supported (Chatterjee 1970:2-3).19 Generally speaking, then,

there is little solid evidence for the cult and characterization of Skanda-Kärttikeya until we reach the Epic

and late Vedic ancillary traditions, which will be detailed in the next chapter.

P. K. Agrawala argues that the origins of the name Skanda can be found in the Paippalāda Samhitā (2.24.1-5). In this section of the text Agni is asked to drive away a demon, and the verb used to describe the action is apaskandayatu. Agrawala suggests that this aspect of Agni is easily named after the action as Apaskanda or Skanda (1967:14). He goes onto argue that the root verb skandir means "'to suck out' or 'to make dry'" (1967:14). He feels this is related to Agni's ability to make things dry "hence demons or disease-demons of a fiery nature were to be named Skanda. Thus, the disease-demons or diseases were named Skanda owing to their fiery nature of sosana. The fire god invoked against them was called Skanda" (1967:14). There is, however, no evidence to support Agrawala's claim that the verb apaskandayatu became a proper name or that Skanda-Kärttikeya is a fire deity. Agrawala is eager to find Skanda-Kärttikeya's roots in Agni, but, as we shall see, such an argument accepts uncritically the agenda of the Äranyakaparvan version of Skanda-Kārttikeya's birth.

15

Chapter Two: The Birth of Skanda-Karttikeya in the Aranyakaparvan 2.1 The Mahäbhärata and the Birth of Skanda-Kārttikeya

The Mahabharata contains three separate narratives of Skanda-Karttikeya's birth (3.213-221;

9.43-45; 13.83-86). This is the formative period of Skanda-Kārttikeya's mythology and tells us the most

about his origins and how he was included into the Brahminical pantheon. It is also the most confused

period of his development, with numerous contradictory versions of his birth and actions being retold in the

Epic. The end result of this textual development is clear: Skanda-Kārttikeya is primarily recognized as

Mahāsena, the surasenäpati, or general of the army of the gods. The stages to this eventual characterization

of the deity can be traced in the various accounts of him found in the Mahabharata and an assortment of

other texts from the same period.

Before entering into this material, however, it is important to discuss the nature of the

Mahabharata and how I will approach working with this text. One of the most important aspects of studying

a text of this type is to attempt to understand its context, but the context of the Mahabhārata is elusive.

Dates for the text vary widely as do theories on how it was produced. A resent review of these problems has

been done by Alf Hiltebeitel (2001), and I will largely be following his discussion here. He begins by

discussing the "epic period", a notion put forth by scholars who proposed to find the actual date for core

events in the Epic and to suggest a period of expansion for the text which resulted in its current form. C.V.

Vaidya is his main representative of this school. He dates the Epic period from 3000 BCE to 300 BCE, or

from well within the Vedic age to just past Alexander the Great (Hiltebeitel 2001:10). Vaidya envisions an

original text produced by Vyasa that swelled under his pupil Vaisampayana and reached its current form

with the sage Ugrasravas during the reign of Chandragupta Maurya (Hiltebeitel 2001:12). In other words,

Vaidya assumes that the characters within the text who are depicted as producing and narrating it were real

people. He accepts the text's own mythology, an approach Hiltebeitel calls a "fanciful extraction"

(2001:12).

16

Another approach taken to the text is the postulation of an "encyclopedic period". This theory

begins with E. Hopkins who proposed a five stage development for the text from about 400 BCE to 400+

CE (Hiltebeitel 2001:14-15; Hopkins 1978:369-370, 397-398). Proponents of this theory often begin with

the assumption that the text is based on an ancient 'original' oral narrative that was steadily embellished

over time, a process that created a chaotic text that attempted to be encyclopedic in its function. Often these

theories postulate a historical core, or warriors' tale, that is taken over by the Brahminical caste who then

radically alter the text.

Representative of this group is Mary Carroll Smith. She argues that a Vedic core in the Epic is

found where tristubh poetic meter (four lines with eleven syllables each) is used. She suggests that tristubh

meter was standard for Vedic poets, but by the historical period when the Mahabharata was being redacted

tristubh meter was replaced by slokas (four lines of eight syllables each) as the dominant verse form (Smith

1992:13-18). Smith argues that those verses in tristubh meter which survived the redaction are a direct link

to the oldest parts of the Epic. As a group the tristubh verses tell a story focused on the dynastic struggle

between the Kurus and the Pändavas. Largely absent from the tristubh meter sections are references to

Brahmins, a situation that is remarkably reversed once the text has been redacted by Brahmins and

presented largely in sloka verses (Smith 1992:7, 13, 54).1 Hiltebeitel rejects this archaeological approach to

the text that attempts to discover its original core. He suggests that it was written over a relatively short

period of time (200 BCE to the year zero) by a few generations of Brahmins working as a committee

(2001:18-26).

It is not my hope here to present my own theory regarding the dating and production of the Epic.

1Hiltebeitel is critical of Smith's approach and conclusions. He claims that she ignores other irregular metres used in the text and does not recognize that the sloka is also a Vedic metre. He also feels that Mahabharata "poets show a mastery of archaization" and could have used the ancient metre well after the Vedic period (2001:18). Smith's work is not fool proof, but we must also be aware that Hiltebeitel is also working out his own agenda in his attack on her and the school of thought she represents. Hiltebeitel hopes to promote scholarship that looks to how the narrative functions as opposed to what he calls "excavative scholarship" (2001:2). We have to read his criticisms, I think, in light of his agenda.

17

What I do hope to point out is that all of these theories work to provide us with an eventual version of the

text which is Brahminical. All of these scholars and all of their theories eventually have Brahmins

producing a text that praises their own caste and its role in the religious world of India. The text positions

itself as anti-heterodox and as a defender of Brahminical orthodox religion (Hiltebeitel 2001:17, 19). While

a specific date of production and a precise mechanism of production elude us, or at least me, one context is

certain: this is a Brahminical text that fulfils Brahminical agendas.

Robert Goldman has demonstrated that one of these Brahminical redactors may have been the

Bhrgus (1977:2). He too takes the approach that the original Epic was a warrior tale that was transformed

into a Brahminical text. Goldman argues that we can see a mythic representation of the shift from warrior

tale to the current Brahminized Epic in the story of the Bhrgu Räma who, in a legend where he destroys the

warrior class, personifies this shift: "through Räma's mythical extermination of the warrior class, the

Bhärgavas have proclaimed themselves the masters of the epic" (1977:140). We do not have to accept all of

Goldman's argument, but he does demonstrate that the Bhrgu tales consistently assert the superiority of

Brahmins over kings (1977:105, 139). These Bhrgu redactors had a clear agenda behind the manner in

which they redacted these stories.

What this manipulation of the Epic means for us is that we must closely read these stories in the

text with an eye for differing voices. There is a dominant view presented in the Mahabharata stories about

Skanda-Kärttikeya and, as we shall see, this view does fulfil certain Brahminical agendas. There are also

differing views within the narration, however, which will cause us to question the dominant portrayal of

Skanda-Kärttikeya and to recognize that other understandings of this deity existed, which this text attempts

to cover over.2

The Brahminical and orthodox stance of this text is significant for this study because the earliest

2Examples of scholarship on the Epic which demonstrate a successful questioning of the Epic's redaction can be found in David Shulman's "Devana and Daiva" (1992), and David Gitomer's "King Duryodhana: The Mahabharata Discourse of Sinning and Virtue in Epic and Drama" (1992).

18

version of Skanda-Kärttikeya that I will argue for does not fit such a Brahminical/orthodox model. Part of

the Brahminical agenda of this text, I will argue, is to transform an unorthodox Skanda-Kārttikeya into an

orthodox figure. By the end of the Epic period almost all of Skanda-Kārttikeya's inauspicious roots have

disappeared; they only survive in a few medical texts. Having stated this much, however, I need briefly to

discuss what this text's vision of orthodoxy is. While the term 'orthodox Hinduism' and 'orthodox religion'

are often used by scholars in relation to the Mahabharata, there is not always a careful evaluation of these

terms in relation to the text. These terms and their implications are important to this study and I will briefly

present my understanding of them in what follows.

In relation to Skanda-Karttikeya's myths in the Mahabharata I would consider orthodoxy to be

informed by two main factors: Vedic traditions and an emerging sense of Hindu order or dharma. As we

shall see, Skanda-Kärttikeya is not part of the Vedic tradition and he seems to present challenges to

established order before his conversion to orthodoxy. To make Skanda-Kärttikeya orthodox the redactors of

the Epic must place him within their own narrative religious history, which is primarily a Vedic religious

history. The stories of this deity must resonate with Vedic themes and have clear links to Vedic precedent

before he can be recognized as orthodox. He is to be Brahminized in the sense that he must be depicted as a

continuation of Brahminic narratives based in the Vedic tradition. As we shall see, the Aranyakaparvan and

the Anusasanaparvan versions of the deity's narratives are particularly concerned with linking Skanda-

Kārttikeya to the Vedic tradition.

Orthodoxy has more to it in the Mahabharata than a narrative link through myth to the Vedic past,

however. While the text resonates with allusions to Vedic myth, the Vedic world of ritual and sacrifice is

not a major aspect of the text, especially in the case of Skanda-Kärttikeya. Part of this deity's apparent

power in this text seems to derive initially from his very non-orthodox and dangerous character. While the

text works hard to situate Skanda-Kärttikeya within an orthodox/Vedic tradition, it is his very non-Vedic

past which appears to attract Epic writers to him. He is part of an emerging Hinduism and a new orthodoxy

that must take into account more than a constructed Vedic past. I would argue, in relation to Skanda-

19

Kārttikeya's depiction in the Mahābhārata, that there are two other features of this new Hindu orthodoxy,

both linked to the notion of a Vedic precedent. The first relates to ritual. That Brahmins would have a

concern for ritual is almost a given within the Vedic or Hindu context, but ritual has specific functions in

the stories of Skanda-Kärttikeya. First, rituals conducted by representatives of the Brahminical order help to

establish that Skanda-Karttikeya is part of that order and controlled by that order. Part of his

Brahminization and shift to orthodoxy is to have his rites become part of Brahminically controlled liturgy.

The second role of ritual in these stories is to limit Skanda-Kārttikeya to a certain position or role. The rites

in this text initiate the deity into certain roles and duties, and this is also the second feature of Hindu

orthodoxy I wish to stress: an emphasis on order, hierarchy and specific duties. A particular anxiety that

seems to be a part of these stories concerning Skanda-Karttikeya is where does he fit within established

hierarchies and roles. There seems to be a need in the process of transforming this deity to catagorize him in

such a way that he complements, or at least does not threaten, established figures like Agni and Siva. This

categorization implies duty or a role to play in cosmic order, an exclusive function for Skanda-Kārttikeya

that will augment existing hierarchies. As we shall see, Skanda-Kärttikeya's role or duty within this

orthodoxy will be military and his position in the hierarchy will be subservient. He will be a perpetual son

to a more powerful father and a general to a more powerful leader. The process of transformation from

unorthodox to orthodox that I will illustrate will revolve around these three themes in relation to Skanda-

Kārttikeya: a construction of Vedic mythological links, a Brahminization of his rituals and assigning him a

proper role in an established structure.

2.2 The Aranyakaparvan version of Skanda-Karttikeya's birth and deeds

The Aranyakaparvan version of Skanda-Kärttikeya's birth is the most significant for this study

because it provides us with a summary of the deity's development and inclusion into the Hindu or Epic

pantheon. The story presents an initial picture of Skanda-Kārttikeya as an inauspicious and dangerous deity

who is transformed into an ordered and obedient military deity. As suggested in the previous section, this

transformation is achieved through three strategies. The first is to place him within a Vedic frame of

20

reference in part through narrative allusions and in part through his parentage. The second is to give

Skanda-Kārttikeya a divine role and to domesticate him by giving him a family. Skanda-Kārttikeya is

literally married to his duty in this text and he is given an assortment of parents. This places him within a

family context of responsibility and hierarchy. Once placed within a family his previous dangerous outsider

image is lost as he becomes a dutiful son and responsible general. He is given responsibilities and duties to

perform that maintain order in the heavenly and earthly worlds. The third is to place him within Brahminical

ritual structures. This text is not always explicit about these changes and how they come about, but a careful

reading of this text below will demonstrate the above points.3

3Very few scholars have recognized this development in the characterization of Skanda-Kārttikeya and none has studied it in detail. The scholar that comes the closest to recognizing this aspect of the narrative is P. K. Agrawala. Agrawala has, arguably, the best summary of textual accounts of Skanda- Kärttikeya and a number of other scholars borrow from him. At one stage in his discussion of chapter 218 of the story at hand where Skanda-Kārttikeya is accompanied by Śrī, he comments: "this is clearly indicative of the fact that a red-eyed, sharp-toothed and fierce-looking kindred deity was being transformed into a high-souled dignified god. Skanda, in the form of a lower deity seems to have been elevated in rank by his association with Śrī-Laksmi" (1967:27-28). His general understanding of this story is that it elevates Skanda-Kārttikeya "from humble status" to that of a general (1967:28). While I agree with Agrawala's general premise here, he does not develop this theme any further and the process is far more complex than an association with Sri. The association with Sri is one of several signs that this deity's status has changed, but does not explain how or why the change occurs, which is largely the focus of my study. I will discuss chapter 218 in more detail below. V. M. Bedekar notes that the Puranic version of Skanda-Kärttikeya show him as assimilated into Vedic theogony and as having a uniform character. The Mahabhārata does not present this uniform characterization, and he regards this presentation as inconsistent with Vedic understandings of deities. He thus views the early stories of Skanda-Kärttikeya as representing an un-Vedic "folk-spirit" who is assimilated into the Vedic and Hindu fold (1975:142). He does note that the Epic form of Skanda-Kärttikeya is associated with malevolent spirits and that these associations may point to his past, but, in the end, all he suggests is that Skanda-Kärttikeya began as a tribal/indigenous village god who was associated with non-Vedic tribal mother goddess (1975:168), but is unable to demonstrate or support these points. Again, these general points are not without their value, but, as I will demonstrate, the process of assimilation is much more complex than Bedekar acknowledges and can be demonstrated in a much more specific manner. K. Sinha views the Epic accounts as evidence of Skanda-Kärttikeya's growing popularity (1979:23-25). He notes that Skanda-Kärttikeya may have originally "been the god of the primitive village folk-tribes", and ends up in the Epic as important, but he does not develop these ideas (1979:26). Rana does recognize some opposition to Skanda-Kärttikeya in these Epic versions of his birth and deeds because he was not a "propitious deity" and was associated with Grahas and thieves (1995:29-31), but he also does not develop these ideas. He also feels the Epic accounts raise Skanda-Kārttikeya's status to that of Siva and Visnu (1995:41). Chatterjee (1970), Navaratnam (1973), Ghurye (1977), Thakur (1981), Gupta (1988), and Mani (1990) only summarize the stories from the Mahabharata with little or no commentary. Absent from all of these previous studies is a careful and detailed account and analysis of the Epic stories, which can demonstrate some of their suggestions and question others. What follows will hopefully provide such a

21

I will be separating the story of Skanda-Kärttikeya's birth into two main parts in this chapter

because the text has two different images of him: the fierce and the tamed. I will begin by discussing the

image of Skanda-Kārttikeya as fierce and inauspicious. I demonstrate this aspect of the deity in two ways.

The first is through a description of the inauspicious circumstances of his birth and his initial violent deeds.

The second is through a study of his associations with three inauspicious beings in this text: Mātrs, Kumāras

and Grahas. I then demonstrate Skanda-Kärttikeya's origins as a Graha-like being through a study of

Ayurvedic literature and its understanding of Grahas and this deity. I also place this inauspicious form of

the deity into a ritual context through a discussion of the Atharvaveda Parisista text, the Dhūrtakalpa. My

conclusion will be that the Aranyakaparvan's association of Skanda-Kärttikeya with Grahas and other

child-snatching beings is actually a reflection of his own past. I then return to the Aranyakaparvan birth

story to explain the second image of the deity presented by this version of his birth, which is as a tamed and

auspicious figure. In this section I illustrate the text's use of frame narratives to help construct an auspicious

deity and the eventual depiction of Skanda-Kärttikeya as the auspicious general of the army of the gods.

Finally, I discuss Skanda-Kärttikeya as the son of Siva.

2.3 The Birth of Skanda-Karttikeya in the Aranyakaparvan: An Inauspicious Beginning

The shift in Skanda-Kārttikeya's characterization from inauspicious and unorthodox to auspicious

and orthodox is fundamental to this account of the deity. His inauspicious character is presented in two

ways. First, his birth and early deeds are replete with images of sexual impropriety, violence and danger.

Second, he is associated with a variety of inauspicious beings who all share a common trait of afflicting

foetuses and children with disease. It is with these disease causing beings that I will argue we find the

origins of Skanda-Kärttikeya. I will begin with Skanda-Kārttikeya as inauspicious because he is violent and

then argue for his origins with disease causing demons.

2.3.1 The Inauspicious Skanda-Kärttikeya: Sexual impropriety and violence in his birth

detailed account and analysis of these important narratives.

22

The manner of Skanda-Kärttikeya's birth and the events that lead up to it are filled with

inauspicious imagery. According to the text, Agni, while serving as the sacrificial fire at a sacrifice of the

seven sages, fell in love with the wives of the sages. In the Vedic and Puranic traditions Agni is, at times,

associated with adultery and seduction (Doniger 1973:91-94). Here Agni is well aware that his desire for

these women is improper. At one point he thinks to himself: "It is not right for me to be excited. I long for

the chaste wives of the best among the Brahmins, and who are without desire" (3.213.45).4 For a while he is

able to console himself by watching the wives all day in the form of their household fires. Eventually, he

becomes sick with love, and once he realizes that he cannot obtain these women he retires to the forest

apparently to kill himself (3.213.35-45)

We are then told that Svāha, a daughter of Daksa, was in love with Agni. She decides to assume

the shapes of the seven wives to fool Agni into having sex with her (3.213.50-52). She begins by taking the

form of Siva, the wife of the seer Angiras, and convinces Agni to have intercourse with her. After their

encounter in the forest she thinks to herself: "Those who see this form of me [as Sivā] in the forest, they will

talk about the offence of the Brahmins' wives with Agni, which would be a falsehood. Therefore, I will

become Garudi [a bird] and prevent that [gossip] and my exit from the forests will be easy" (3.214.8-9).5

Sväha then takes on the form of five of the other wives of the seers and repeats her affair with Agni and her

escape from the forest in disguise.

This episode of the story begins, then, with a theme of infidelity, uncontrolled lust and deceit. Agni

is aware of the impropriety of his desire for the wives of the seers, and Svaha is also aware of her own

impropriety and the potential danger her disguises may have for the real wives of the seers. The text

highlights Sväha's immorality by telling us that she cannot take the form of Arundhati, the seventh wife,

4"sa bhūyaścintayāmāsa na nyāyyam ksubhito'smi yat sādhvīh patnīrdvijendrānāmakāmāh kāmayāmyaham". Unless otherwise noted, all of the translations that appear in this dissertation are my own.

5"acintayanmamedam ye rūpam draksyanti kānane te brāhmanīnāmanrtam dosam vaksyanti pāvake [.8] tasmādetadraksyamānā garudī sambhavāmyaham vanānnirgamanam caiva sukham mama bhavisyati [.9]". My thanks to Phyllis Granoff for helping in the translation of these verses.

23

"because of her austerity and her obedience to her husband" (3.214.14).6 Arundhati becomes a model of

marital fidelity within the Hindu tradition (Doniger 1973:101), and a contrast is created in this text between

the pure Arundhati and the impure Svāhā. A further, more subtle, comparison is also made between order

and disorder. The seven seers and their wives can be viewed as representative of a balance between

Brahminic household order and ascetic prowess in the form of austerities, as the above quotation relating to

Arundhati suggests. Agni with his lust for the sages' wives and his withdrawal from the sacrifice threatens

this Vedic sense of order. Sväha is also a threat to the domestic order represented by the sages and their

wives. It is these forces that threaten Brahminical order which produce Skanda-Kārttikeya.7

6"tasyāstapah prabhāvena bhartrśuśrūsanena ca".

7There is also some additional background to this story regarding Agni's love of these wives and Arundhati's role as a model of marital virtue. In this account of Skanda-Karttikeya's birth the six wives whom Svāha imitates become the asterism the Krttikas. The Satapatha Brāhmana tells us that the Krttikās are Agni's asterism (2.1.2.1). The text notes that some object to setting up these fires under the Krttikās because they are regarded as the wives of the seven sages, who are also depicted as an asterism (2.1.2.4). The text explains, however:

They [the Krttikäs] were, however, precluded from intercourse (with their husbands), for the latter, the seven Rsis, rise in the north, and they (the Krttikas) in the east. But he may nevertheless set up (his fire under the Krttikās); for Agni doubtless is their mate, and it is with Agni that they have intercourse: for this reason he may set up (the fire under the Krttikās). (2.1.2.4-5 translated by Eggeling 1882-1900 I:282-283)

The legend of Agni's relationship with these women that we find in the Epic text thus has a background in the Vedic tradition. The mythology of Arundhati may also be rooted in an actual reduction in the number of stars that formed the Krttikās. The Vedas recognize a group of seven stars as making up the Krttikās, but by the Epic the number has been reduced to six. The shift from seven to six stars also occurs in a number of other cultures as does a mythology of a lost Pleiad (Allen 1963:406-412; Shulman 1980:245). Today, only six stars are clearly visible with the naked eye, the seventh star, having a different magnitude, is harder to see. While we cannot be sure as to the circumstances of the mythological shift from seven to six, some have suggested that the seventh star was brighter before the second millennium BCE and its reduced intensity occurred in recorded history and is reflected in the form of these myths (Harper 1989:25). Whatever the case, the Epic only recognizes six Krttikās and part of the Skanda-Kärttikeya story allows for the separation of the seventh star, Arundhati, from the group. Wendy Doniger has also studied elements of this Agni and Svāha narrative. She locates the origins of the myth in Vedic narratives of Agni uniting with the wives of Varuna. All later versions of the story replace the wives of Varuna with the wives of the seven sages. She provides a number of versions of the myth and notes that while Sväha is dropped from most other accounts of the story, the seven wives are always found in some element of these versions. The point Doniger draws from this material relates to the

24

After each encounter with Agni, Svāha flew to Mount Śveta where she found a "mountain ridge

which was very hard to reach", on which she "quickly threw the semen into a golden pot" (3.214.12).8 We

are told that this mountain is a terrifying place which is guarded "by supernatural seven-headed and venom-

eyed snakes, and by Rāksasas, Pisācas, and bands of violent ghosts, and which was also full of Rāksasīs and

many animals and birds" (3.214.11).9 The six units of semen that will form Skanda-Kārttikeya are discarded

in a dangerous, inauspicious place which is the home to various monsters. The mountain lies outside of the

ordered world of Arundhati and the Brahminical seers. Skanda-Kärttikeya will arise out of an area full of

demonic and fearful beings, and by placing him in this context the text suggests that he shares a great deal

in common with the inauspicious circumstances and place of his birth.

The semen is placed into the golden pot on the first day of lunation.10 A baby is produced on the

second day and is described as having six heads, twice as many ears, twelve eyes, arms, and feet, but one

neck and one body (3.214.17).11 On the third day we are told he is already a child (sisu) and on the fourth

taint of impropriety placed on these wives even though they do not actually engage in sexual intercourse with Agni. She also notes that Arundhati remains chaste in the various accounts of the Saivite Pine Forest story, while the other wives fall victim to the seduction of Siva (1973:94-101). Elements of the Pine Forest story and the Birth of Skanda-Kärttikeya story may have been woven together here in the account of the wives of the sages. Arundhati is also a name used for a healing plant or goddess in the Atharvaveda (8.7, 4.12, 19.38; Zysk 1998:72-75), but it is not clear if this is the same Arundhatī as above. Shulman also notes some Tamil folk myths concerning Arundhatī (1980:245-246).

8"sā tatra sahasā gatvā śailaprstam sudurgamam prāksipatkāñcane kunde śukram sā tvaritā sati.

drstīvisaih saptaśīrsairguptam bhogibhiradbhutaih raksobhiśca piśācaiśca raudrairbhūtaganaistathā rāksasībhiśca sampūrnamanekaiśca mrgadvijaih"

0Elements of this account may be borrowed from the Kausitaki Brahmana (Shulman 1980:65). In that text a story is given where Usas appears before the gods causing them to shed their seed. Prajāpati then

(KB 6.1-2). makes a sacrificial pot out of gold and puts the seed in it. From this golden pot full of semen Rudra is born

l“şatśirā dvigunaśrotro dvādasāksibhujakramah ekagrīvastvekakāyah kumārah samapadyat".

25

day, Guha,12 as he is called here, has all his major and minor limbs.13 The child or embryo develops

preternaturally fast.

The first appearance and deeds of this deity are presented in a terrifying manner:

Surrounded by a great red cloud with lightning, he shone like the rising sun in a very great red cloud. A terrifying and large bow, which cut the enemies of the gods to pieces and which was placed there by the Destroyer of Tripura [Siva], was held by him. Then the powerful one, having grasped that most splendid bow, roared. [With that roar] he stupefied the three worlds [with all its] moving and still creatures. (3.214.19-21)14

This roar causes two divine war elephants, Citra and Airavata, to attack Skanda-Kārttikeya, and he is

described as holding these two elephants with one hand, a spear in another and a huge wild cock in another

(3.214.22-24). He continues to emit terrifying roars and to frolic mightily. He pounds the air with one of his

free hands and blows into a conch shell with another.

The text continues by describing the fear he inspires in other beings and the response of the

frightened to him:

That one of supernatural strength and immeasurable being sat on the summit of that mountain and

12Other versions of the story in the Mahabhärata view Guha as another name of Skanda- Kārttikeya. Guha means 'hidden', or 'reared in a secret place' (Monier-Williams 1999:360). The name may bear some relation to Agni. Guha is never used as an epithet of Agni, but the term does occur in relation to him during hidden or missing Agni narratives (P. K. Agrawala 1967:2). The Anusāsanaparvan claims he has the name Guha because he was born hidden (guha) in a bed of reeds. I think, however, that the name Guha may be linked to Kubera and cults related to child afflicting deities. Kubera is the leader of the Guhyakas (P. K. Agrawala 1967:48; Pal 1977:22), and may have begun as a demonic snatcher of children (Pal 1977:22). The term guhya means secret, or to be covered or hidden (Monier-Williams 1999:360) and is clearly related to guha. In the Buddhist tradition, both Kubera and Kumāra were Yaksas who were converted by the Buddha (Pal 1977:22). Yaksas were linked to childhood dangers and the Buddhist tradition seems to understand these figures as part of such a context. Both Skanda-Kärttikeya and Kubera are also depicted in panels with Mätrs, another child snatching group of deities (Joshi 1986:6-7). Skanda- Kärttikeya and Kubera may, then, have been related as child afflicting deities, but in particular the name Guha may be linked to Kubera or one of his assistants. I think these later attempts to derive Guha's name from some 'hidden' element of his stories reflects latter attempts to justify the name after the period when Kubera and Skanda-Guha are no longer linked together as child snatching deities.

13My thanks to Phyllis Granoff for helping me to understand this passage.

14“lohitābhrena mahatā samvrtah saha vidyutā lohitābhre sumahati bhāti sūrya ivoditah [.19] grhītam tu dhanustena vipulam lomaharsanam nyastam yattripuraghnena surārivinikrntanam [.20] tadgrhītvā dhanuhśrestam nanāda balavāmstadā sammohayannivemānsa trīmllokānsacarācarān [.21]".

26

looked at the region with his various heads. He saw the various sorts of beings and roared again. Having heard his cry the creatures, who were crushed from anguish, repeatedly collapsed and went to him for refuge. Those creatures of various colours who had sought the protection of that god, Brahmins call them the very powerful retinue of that god. That long-armed one stood up and comforted those creatures .... (3.214.27-30)15

Who these 'creatures' (janäh) are is not very clear. The term jana can refer to people and creatures in a

general sense. That these creatures or people are many-coloured may be significant. It may suggest that the

devotees of this dangerous deity where from all segments of society (Bedekar 1975:157). We also get a

sense of the relationship between god and devotee in the above quotation. This dangerous god is

worshipped because he is feared, but once worshipped he becomes protective. This dynamic is, I think,

central to the early cult of this deity. Skanda-Kärttikeya does inspire fear in people, and texts like this one

make it clear that there is a great deal to be frightened of, but if approached and worshipped he can be

benign.

This god's destructive acts have only just begun. Next, he shoots arrows at Mount Śveta and

cleaves off Mount Kraunca, which groans loudly with pain as it crashes to the earth. The other mountains

groan with fear and we are told: "That best of the strong, having heard the crying of the greatly afflicted,

was not distressed and that immeasurable being lifted his spear and roared" (3.214.33).16 He shows no

remorse or concern for these mountains which have caused him no harm and proceeds by throwing his huge

spear at Mount Sveta, cutting off its top. Out of fear Sveta and the other mountains leave the Earth, tearing

her up and causing her great pain. The Earth seeks refuge with Skanda-Kärttikeya and appears to regain her

strength. The mountains also pay him homage and return. The chapter ends with "now the world honours

15"sa tasya parvatasyāgre nisanno'dbhutavikramah vyalokayadameyātmā mukhairnānāvidhairdišah sa paśyanvividhānbhāvāmścakāra ninadam punah [.27] tasya tam ninadam śrutvā nyapatanbahudhā janāh mītāścodvignamanasastameva śaranam yayuh [.28] ye tu tam samśritā devam nānāvarnāstadā janāh tānapyāhuh pārisadānbrāhmanāh sumahābalān [.29] sa tūtthāya mahābāhurupasāntvya ca tānjanān dhanurvikrsya vyasrjadvānānśvete mahāgirau [.30]".

16“sa tam nādam bhršārtānām śrutvāpi balinām varah na prāvyathadameyātmā śaktimudyamya cānadat".

27

Skanda on the fifth day of the bright [fortnight]" (3.214.37).17 Skanda-Kārttikeya's violence in these

episodes seems completely unprovoked and unpredictable. He strikes without remorse and appears

deliberately to seek to engender fear in others. Fear is, however, at the root of his worship at this stage of

the narrative as the text tells us: afraid, they sought refuge with Skanda-Kärttikeya (3.214.28).

The next frightened group to respond to him are the gods. They tell Indra to destroy this new force

lest "that one of great strength subdue the three worlds, us, and you, O Sakra" (3.215.14).18 Indra is,

however, clearly afraid and is described as vyathita, 'trembling' or 'distressed' (3.215.15). He does not

want to fight Skanda-Kärttikeya and tells the gods: "This tottering child is very powerful. He, having

attacked him in battle, could destroy even the creator of the worlds. But, let all of the Mothers of the world

attack Skanda. Let those that have the strength and desire kill him ... " (3.215.15-16).19 These Mothers

(mätrs) accept this task, but when they see Skanda-Kärttikeya they think he is invincible and seek his refuge

(3.215.17).20

Indra must now face Skanda-Kärttikeya and this battle is narrated in chapter 216. The chapter

opens with an image of Skanda-Kārttikeya surrounded by various terrifying celestials (3.216.1), and Indra

with his army marching on him. Skanda-Kärttikeya's might is not to be matched, however. The gods and

Indra enter battle with a roar and Skanda-Kärttikeya answers back with a roar of his own that "made the

army of the gods senseless" (3.216.8).21 Then, Skanda-Kärttikeya belches flames from his mouth, burning

17"athāyamabhajallokah skandam śuklasya paňcamīm".

18"trailokyam samnigrhyāsmāmstvām ca śakra mahābalah".

9"sa tānuvāca vyathito bālo'yam sumahābalah srastāramapi lokānām yudhi vikramya nāśayet [.15] sarvāstvadyābhigacchantu skandam lokasya mātarah kāmavīryā ghnantu cainam tathetyuktvā ca tā yayuh [.16]".

20I will discuss this section with the Mätrs in more detail below.

21The whole verse reads: "He [Skanda-Kärttikeya] wandered here and there among the army of the gods which was senseless and which was as the splendid ocean shaken up by his great noise". "tasya śabdena mahatā samuddhūtodadhiprabham babhrāma tatra tatraiva devasainyamacetanam".

28

the gods and causing them to seek refuge with him. Indra does not give in, though, and throws his

thunderbolt at Skanda-Kärttikeya. The impact, however, does not harm him and actually creates another

being, Viśākha, who appears out of Skanda-Kārttikeya's side. Visākha is described as "a being without

rival, a youth who had golden armour, held a spear, and wore divine earrings" (3.216.13)22 and as "equal in

radiance to the doomsday fire" (3.216.14).23 Faced with this second foe, Indra gives up and seeks refuge,

which Skanda-Karttikeya affords him and his army.24

The implications of Indra's defeat are several for this story. Indra, as the king of the gods, sits at

the top of the Vedic hierarchy of deities. He attained this position by being a great warrior in Vedic texts.

22" .. puruso'parah yuvā kāñcanasamnāhah śaktidhrgdivyakundalah ... ".

23" . kālānalasamadyutim ... ".

24Visākha is a deity of some importance to the Skanda-Kärttikeya cult. He is also mentioned in the Epic at 1.60.23 where he appears to be born from Skanda-Kārttikeya's back. The Salyaparvan version of Skanda-Kärttikeya's birth and deeds, which will also be discussed below, regards Visākha as a form of Skanda-Kärttikeya (9.43.33-38). He is described in the Skandayāga as a brother of Skanda-Kārttikeya (AVP 20.2.9). Visākha is also treated in the Suśrutasamhitā, where he is referred to twice: "That one who is known as Skandäpasmāra, who is the cherished friend of Skanda and who is known as Visakha, let the ugly faced one be friendly to the child" (Uttara. 29.9), and "The one who is known as Skandāpasmāra, he who is equal in radiance to fire was [created] by Agni, and he who is the friend of Skanda is called 'Visākha' by name" (Uttara. 37.7) (skandāpasmārasamjño yah skandasya dayitah sakhā visakhasamjñśca śisoh šivo'stu vikrtānanah), and (skandāpasmārasamjño yah so'gninā-gnisamadyutih sa ca skandasakhā nāma višākha iti cocyate). Meulenbeld understands from this section that Visakha-Skandapasmāra was created by Agni to be a companion of Skanda-Kärttikeya (1999 A1:312). The Suśruta makes a clear separation between Skanda- Kārttikeya and Višākha. Visākha is always referred to as Skanda-Kārttikeya's friend in this text, but he has a separate creation and is certainly not a form or aspect of Skanda-Kārttikeya as elements of the Mahābhārata suggest. His appearance on Kusana coinage will also be discussed below. As will become clearer below, the likely origin of Visākha is as a Graha under the leadership of Skanda-Kärttikeya. Both of these deities are worshipped by people seeking relief from sickness. Viśākha was another feared deity worshipped with Skandagraha to protect children. Visakha is given a similar role to play in chapter 217 of the Aranyakaparvan, which will be discussed below. In the present episode he is closely associated with Skanda-Kärttikeya and described as a wealthy warrior-type of deity. He is introduced during a battle scene and is the product of Skanda-Kärttikeya as a warrior and to some degree of Indra's energy. The implication here is that Visäkha is a warrior deity like the developing depiction of Skanda-Kārttikeya, which is a significant shift from his status as a Graha. There appears to be general confusion regarding Visakha, but all of these texts are clear that he has a very strong relationship with Skanda-Kärttikeya. I would suggest that the Epic's attempt to present him as a form of Skanda-Kārttikeya does not likely represent his original status, but an attempt at assimilation. It is only in medical texts that Visākha has a clearly defined role that is more than just an appendage to Skanda-Kārttikeya. It is likely as an independent Graha associated with the Graha named Skanda where Visākha finds his origins.

29

His defeat at the hands of Skanda-Kärttikeya is an indication that a new warrior god is replacing the old

warrior god.25 If Indra can be viewed as a champion of Vedic order centred on a sacrificial cult, Skanda-

Kärttikeya can be viewed as representing a force that disrupts that order and draws his power from

inauspicious or non-orthodox sources. As we shall see, Skanda-Kārttikeya's origins are indeed inauspicious

and non-Brahminic. He upsets the divine hierarchy, the earth and its creatures. He is a force of disorder and

chaos thrown into a Vedic and Brahminic world, which prides itself on being able to order and control the

universe. From such a Brahminical perspective he is an inauspicious and outside force and chapters 214 to

216 seem to emphasize this aspect of his character.

2.3.2 The Inauspicious Skanda-Karttikeya: His tie to Grahas, Kumāras and Mātrs

Another aspect of Skanda-Kärttikeya's inauspicious image in this version of his birth is his

association with three ghoulish types of beings: Grahas, Kumaras and Mätrs. Much of the importance of

these beings in their relation to Skanda-Kärttikeya is that they are edited out of the majority of Epic and

Purānic legends of this deity, as are the above inauspicious deeds of the deity. As potential clues to Skanda-

Kärttikeya's origins, however, an understanding of them is central to this thesis. Accounts of these beings

are found in chapters 21726 and 219 of the ranyakaparvan, and we have already seen elements of their

characterization in other chapters. I will begin with a discussion of Skanda-Kärttikeya's relationship with

Mätrs and then his relationship with Kumäras and Grahas. Based on these relationships, I will argue that

Skanda-Kärttikeya's origins are located with these inauspicious groups.

I have already mentioned that Indra sends out a group of Mātrs to destroy Skanda-Kārttikeya, but

25Sinha comments that he views Skanda-Kärttikeya's victory over Indra as a sign of his increasing status (1979:22). Rana regards the episode of the battle as an indication of initial opposition to Skanda- Kārttikeya's inclusion into the pantheon (1995:29).

26On the whole I have to admit that I find chapter 217 a jumbled mess that contradicts other sections of this story and even seems to contradict itself. It may have been garbled in transmission. What the section does remind us of is that this account was not written and edited by one hand, but by several over a period of time. It will become clear that the writers or editors of chapter 219 did have a good understanding of the topics that chapter 217 attempts to relate and that they were not produced by the same hand(s).

30

they end up seeking his refuge. These Mätrs become Skanda-Kārttikeya's mothers and begin to protect him.

They are not gentle creatures, but are described as horrible beings. Two of them are described in the

following manner: "The one woman among all the Mothers who arose from Wrath, she, who holds an iron

spike, and who is a midwife, protected Skanda as a mother would protect her son. The cruel daughter of the

Blood Ocean, who feeds on blood, having embraced Mahasena, protected him like a son" (3.215.21-22).27

They are frightening, inauspicious figures and their association with Skanda-Kārttikeya seems appropriate

given his own terrible image. Part of Skanda-Karttikeya's characterization as inauspicious or as auspicious

in this text is through association. Here, obviously, his inauspiciousness is magnified through his association

with, and similarity to, these Matrs.

A great deal of power is associated with these inauspicious and horrific figures. The gods and

Indra seem to be aware that they are no match for Skanda-Kärttikeya, and they call upon other inauspicious

beings to defeat him. The text seems here to suggest that a great deal more power lies in these inauspicious

deities than in the Vedic gods. Their horrific and inauspicious nature seems to be the source of their power,

but there is also a duality in the representation of these beings in this text. While Skanda-Kärttikeya is huge,

destructive and horrible, he is also a child who is playful and appears, at times, to be innocent. Chapter 215

ends with one such image as Agni joins his son and tries to amuse him: "Agni having become Naigamesa,

the goat-faced one, who has many children, delighted the child who was on the mountain as with toys"

(3.215.23).28 As this quotation suggests, this powerful and destructive deity is also depicted as a child who

enjoys the playful pursuits of other children, once placed in a domestic setting.

A similar duality is seen, I think, in these Mothers. While they are horrific, they are also maternal

27“sarvāsām yā tu mātrnām nārīkrodhasamudbhavā dhātrīsā putravatskandam śūlahastābhyaraksata [.21] lohitasyodadheh kanyā krūrā lohitabhojanā parisvajya mahāsenam putravatparyaraksata [.22]". Wendy Doniger suggests that the Mother referred to here is Kāli (1975:114), but I am not convinced of this point.

28“agnirbhūtvā naigameyśchāgavaktro bahuprajah ramayāmāsa śailastham bālam krīdanakairiva". My thanks to Phyllis Granoff for helping in the translation of this passage.

31

and can be protective as is the case of the daughter of the Blood Sea in the quotation above (3.215.22). This

passage makes it clear that this 'cruel daughter' is a horrible figure, but she also seems to be capable of

acting like a mother, at least towards Skanda-Kärttikeya. I would suggest that this dual nature seen in

Skanda-Kärttikeya and these Mtrs is suggestive of the profound change that comes over these deities when

placed into a domestic setting. This text views a domestic relationship, particularly a parent-child

relationship, as a key means to explain the shift towards an auspicious characterization of Skanda-

Kärttikeya and, to some degree, of these Mätrs by the end of the text.

That these monstrous divinities are transformed by these roles tells us a great deal about this text's

assumptions relating to the power of family roles. Even though the Mothers and Skanda-Kārttikeya come

from inauspicious backgrounds and display cruel characteristics, the power of familial responsibility and

duty is depicted as radically changing their approach to each other. This text evokes very strong cultural

assumptions about how parents and children interact in its attempt to shift Skanda-Kärttikeya and these

Mothers from an inauspicious to an auspicious milieu. The gentle and protective side of these deities is also

a part of Skanda-Kärttikeya's cult as described in these opening sections. When worshipped or sought for

refuge he switches from malicious to child-like and from murderous to protective, and, because he and his

'Mothers' are so powerful, they constitute a considerable protective force. These deities are not one

dimensional and are capable of rewarding those they favour.29

Another group of Mätrs is presented in chapter 217. In that chapter a group of "powerful maidens"

(kanyāś mahābalah) were also born when Indra struck Skanda-Kārttikeya with his thunderbolt (3.217.2).

29The idea of dangerous mothers is also taken up by Shulman (1980:223-267). He uses domestication through marriage as a model to explain the taming of dangerous goddesses (1980:141, 198, 212). While the gender roles are reversed in the Skanda-Kärttikeya narrative (the male is dangerous and the female, Devasenā, is safe), the theme of domestication Shulman explores is also demonstrated by this story. He also regards goddesses, especially Tamil goddesses, to express an ambivalence around mothers. These mothers are both violent and nurturing, a point he demonstrates in part through the fluids commonly associated with them: blood and milk (1980:229). He regards a plurality of mothers, as in the case of Skanda-Kärttikeya, as an example of this ambivalence or multivalence of the mother figure (1980:229). He also argues that the Skanda-Kärttikeya story illustrates his three main roles for Tamil goddesses: threatening, mother virgin and incestuous (1980:243).

32

They are also mentioned in the following: "The blessed one,30 having become goat-faced, protects in battle

30What is also not clear from the above 3.217.3-4 is just who the "blessed lord" (bhagavan) is. The whole section reads:

Very powerful maidens were also born from the thunderbolt strike and the Kumāras settled on Visākha as their father. That blessed one, having become goat-faced, protects in battle surrounded by the troops of maidens and all his sons and children, and Bhadrasakha [provides] prosperity to the Mothers as they look on. Thus, people in the world say Skanda is the father of boys (kumāras). People in all places who desire sons perpetually worship Rudra, Agni, the very powerful Uma and Svāhā. (3.217.2-5) (vajraprahārātkanyāśca jajñire'sya mahābalāh kumārāśca viśākham tam pitrtve samakalpayan [.2] sa bhūtvā bhagavānsamkhye raksamśchāgamukhastadā vrtah kanyāganaih sarvairātmanīaiśca putrakaih [.3] mātrnām preksatīnām ca bhadraśākhaśca kauśalah tatah kumārapitaram skandamāhurjanā bhuvi [.4] rudramagnimumām svāhām pradeśesu mahābalām yajanti putrakāmāśca putrinaśca sadā janāh [.5]).

While Visakha could be the 'blessed lord' and the Kumaras have claimed him as their father which could account for the reference to sons, it sounds as though Skanda-Kärttikeya is being spoken of as goat-faced. The text will go on to state that one of Skanda-Kärttikeya's six faces is a goat's face and another head is called Bhadrasakha: "Know, O king, that the sixth face from among his [Skanda's] six heads is like a goat's face, which is always worshipped by the band of Mothers. Now, the best of his six heads is called Bhadrasākha with which he created the divine Sakti" (3.217.12-13). (sastham chāgamayam vaktram skandasyaiveti viddhi tat sat sirobhyantaram rājannityam mātrganārcitam [.12] sannām tu pravaram tasya śīrsāņāmiha śabdyate śaktim yenāsrjaddivyām bhadraśākha iti sma ha [.13]). Given this passage, we might well assume that the 'blessed lord' in verse three is Skanda-Kärttikeya, but the text seems confused. The text is, I think, unsure of just who Skanda, Visākha, Naigamesa (a goat-headed Graha) and Bhadrasākha are and how they are related. I think most of these beings are Grahas whose characters and relationships are made clearer in medical texts. The Aranyakaparvan's treatment of them suggests that it is not completely familiar with them as Grahas or is trying to change their roles. The text is also not clear on just who should be worshipped for the safety and production of male children. While Skanda-Kärttikeya is called "the father of boys", we are told that people worship Rudra, Agni, Umā and Svāha to have a son. The text seems to trip over itself in an attempt to account for all the deities involved in the birth story and in amalgamating several deities into Skanda-Kärttikeya through a confused description of his six heads. There are several narratives of how Skanda-Kärttikeya came to have six heads. In this version of his birth story the six heads seem to be accounted for by the references to the number six in the birth of the deity. We are told that Sväha has intercourse with Agni six times and that she deposits semen in the mountain after each encounter. It seems that these six bunches of semen create one six-headed being, though the text does not go into detail on this point. Other versions of the birth story that deal with the appearance of the six heads differ and there is no agreement on the origin of these six heads within the textual tradition. This text seems confused by the heads in its description of them (3.217.12-13). It seems from this that each of the heads represents a different form of the deity, but such a description is only found in this text and the sculptural and numismatic traditions do not represent the heads as being different. We may also question what the other heads look like and how they are related to Skanda-Karttikeya? Are they all manifestations of the deity, or is one of these heads Skanda-Kärttikeya himself? This text's account of these six heads and their origin creates more questions than it answers. I think that the idea of Skanda- Kärttikeya as a six-headed deity did not originate with these Epic writers and is something they have trouble accommodating or explaining, but was clearly an aspect of the deity's cult that was popular enough that it could not be ignored.

33

surrounded by the troops of maidens ... " (3.217.3). A similar obscure reference to maidens is made in the

Skandayäga: "and the great one who is ever surrounded by a thousand maidens, that Dhūrta, having lion-

equipment, I cause to be brought in" (AVP 20.2.7).31 The Aranyakaparvan appears to equate these maidens

with Mätrs:

Those ones whom the oblation eater [Agni], 'tapas' by name produced, are those maidens. They approached Skanda and he asked: 'What will I do?' The Mothers said: 'We should be the first Mothers of the world and from your kindness we should be worshipped -- you should do us this favour. (3.217.6-7)32

Skanda-Kärttikeya does answer these Mätrs and allows them to become the world's Mätrs, though it is not

clear just what this means, and he separates them into asiva (inauspicious) and siva (auspicious).33 The text

states that these Matrs make Skanda-Karttikeya their son and are named Kākī, Halimā, Rudra, Brhalī, Āryā,

Palala and Mitra. It is not clear if these Matrs are the same Matrs mentioned in 215 and 216 who have

already claimed Skanda-Karttikeya as their son, or a new group of Matrs. The text does call them "saptaitāh

sisumatarah" (3.217.10), or 'the seven mothers of new born sons', which suggests that this text may be

trying to explain the origins of the saptamātrkas or saptamātaras with this particular group.34

Mätrs also appear in chapter 219 of the text. By this stage of the text Skanda-Kārttikeya has largely

been transformed into an auspicious deity, but his connection to these inauspicious beings remains. In the

31"yaśca kanyāsahastrena sadā parivrto mahān tamaham simhasamnāham dhūrtamāvāhayāmyaham". The Skandayāga will be discussed in more detail below.

32"yāstāstvajanayatkanyāstapo nāma hutāsanah kim karomīti tāh skandam samprāptāh samabhāsata [.6] mātara ūcuh bhavema sarvalokasya vayam mātara uttamāh prasādāttava pūjyāśca priyametatkurusva nah [.7]".

33P. K. Agrawala suggests that the mothers of the tapas fire are a blend of several mothers. He feels they are low or folk goddesses and they have a desire for higher status (1967:31).

34Seven female divinities are associated with Skanda-Kärttikeya in early statuary. By the Gupta era Skanda-Kārttikeya is replaced in sculpture by a form of Siva and Ganesa, and the Mätrs appear as female personifications of male divinities. They are given names at this stage like Brahmī and Kaumarī, which appear as female forms of the male deities they represent. The names Rudra and Mitra from the above list appear to suggest the beginning of such a process of assimilation of female and male deities with these Epic 'Mothers'. K. Harper notes that this is "the first septpartite group of chthonic goddesses to be named in Indian literature" (1989:56). She does not, however, explicitly view them as the Saptamātrkās.

34

text he is approached by a band of Mätrs who tell him: "We are celebrated by sages as the Mothers of all

the world. We want to be Mothers to you, honour us" (3.219.14)!35 Skanda-Kärttikeya acknowledges that

they are his mothers and then they ask him: "These ones who were made the first Mothers of this world, that

position should be ours and should not be theirs. Let them be not worshipped and let us be worshipped in

the world, O best of the gods. Because of you they have taken away our offspring, give (them back) to us"

(3.219.16-17).36 What is not clear is if these Mätrs are the same or different from the Matrs encountered in

chapter 217, who asked to be the mothers of the world. This group from chapter 219 appear to be

complaining about the Matrs from 217, but we cannot be sure about this.37 Skanda-Kārttikeya refuses to

give back their progeny, but he will give them other offspring. The Mothers answer: "We want to eat the

offspring of those Mothers with you, give them to us, and the lords of those Mothers too" (3.219.19).38

Skanda-Kärttikeya grants this, but requires: " ... you must happily save their offspring who worship

properly" (3.219.20).39 These passages make it clear that these Mothers are viewed as dangerous to children

and that propitiation based on fear drives this cult. This text seems to understand that there are several

different groups of these terrible Mätrs, but many of them are presented as preying on children and being

controlled by Skanda-Kärttikeya through familial bonds.

35"vayam sarvasya lokasya mātarah kavibhih stutāh icchāmo mātarastubhyam bhavitum pūjayasva nah".

36“yāstu tā mātarah pūrvam lokasyāsya prakalpitāh asmākam tadbhavetsthānam tāsām caiva na tadbhavet [.16] bhavema pūjyā lokasya na tāh pūjyāh surarsabha prajāsmākam hrtāstābhistvatkrte tāh prayaccha nah [.17]".

37P. K. Agrawala regards this hostility between Mothers as a form of class warfare between "low" "indigenous goddesses" and "high" goddesses (1967:32). He feels that in this story the "lower" goddess are given recognition by higher religious society and assimilated in the Skanda-Kärttikeya pantheon, which changes them from child-destroying to child-protecting (1967:32). There may be something to Agrawala's comments here, though I am weary of his high-low distinction and his assumption that all of these Mothers become child-protecting, which is not the case in this story.

38"icchāma tāsām mātrnām prajā bhoktum prayaccha nah tvayā saha prthagbhūtā ye ca tāsāmatheśvarāh".

39"pariraksata bhadram vah prajāh sādhu namaskrtāh".

35

It is important to note that there was likely an active cult of Mätrs in ancient India and Skanda-

Kärttikeya was connected to this lived cult. From the Mathura region during the Kusana era we have

numerous panels which depict a group of Mätrs with Skanda-Kärttikeya. These panels are discussed in more

detail in my section on Kusana coins and statuary. For now we may note that these panels do not directly

relate Skanda-Karttikeya to his military role. What they demonstrate in light of this text is that, at least in

Kusana Mathura, the cult of Matrs with Skanda-Karttikeya was focussed on a propitiatory cult. Why the

Epic text must account for the associations between Mätrs and Skanda-Kārttikeya is now clear; it was an

important element of his early cult, which, I would argue, was centred on the worship of non-orthodox and

inauspicious figures. While the text is unable to ignore these links to these inauspicious Matrs, it does

attempt to domesticate them, and ultimately the text attempts to separate Skanda-Kārttikeya's character

from that of these Matrs. By chapter 219 he no longer shares the inauspicious characteristics given to these

Mätrs and is presented as the auspicious intermediary between Mātrs and humans.

Another inauspicious group related to childhood illness are the Kumāras. Chapter 217 begins by

telling us that a group called the Kumärakas, or lesser Kumäras, were also born when Indra's thunderbolt hit

Skanda-Karttikeya's side. These Kumarakas seem to be the same as Kumaras, which is the term used for

them throughout the rest of the text. These beings are described as "the terrible attendants of Skanda who

have a wonderful appearance .... They are the cruel ones who take children who are born and [those] still in

the womb" (3.217.1).40 The description of these lesser Kumaras makes them sound like Grahas. Just why

these terrible beings are depicted as coming from Skanda-Kärttikeya is not stated by the text, but a clue may

be found in one of the most common epithets of the deity, 'Kumära'. As we have seen (section 1.3), other

scholars argue for a connection between Skanda-Kärttikeya as Kumära and several other Kumāras from the

Vedic tradition. Certainly, this is the understanding of Kumära the Epic endorses in its accounts of Skanda-

40The complete verse reads: "mārkandeya uvāca skandasya pārsadānghorañśrnusvādbhutadarśanān vajraprahārātskandasya jajñustatra kumārakāh ye haranti śiśūñjātāngarbhasthāmścaiva dārunāh".

36

Kärttikeya, but I would like to suggest that Skanda-Kārttikeya's epithet of Kumāra may have begun due to

his similarity to these unorthodox and inauspicious Kumāras.

The Päraskara Grhyasūtra helps to confirm such an understanding of the name Kumāra. This text

describes a household rite for an expectant mother.41 One of the rituals is for a "Kumāra attack" (kumāra

upadravet) (1.16.24). This Kumara may well be Skanda-Karttikeya. It is clear from the context of the

passage that this Kumara attacks the foetus, and it is also clear that the passage understands this Kumāra to

be a dog (śunaka).42 The demon dog-Kumara's mother is Sarama, his father Sīsara and his brothers Śyāma

and Sabala (1.16.24).43 All of these beings are inauspicious divine dogs. This Kumära-dog who attacks

foetuses certainly is an inauspicious Graha-like figure and, as we shall see, sounds much like the

Skandagraha from the Susrutasamhitā. Sarama is also mentioned in the Aranyakaparvan (3.219.33) as a

Graha, who is the divine mother of dogs and who takes the foetuses of people,44 and is among the group of

41 Grhyasūtras describe domestic rites. They are usually simple, performed by the householder and inexpensive. They are based on smrti, or the remembered tradition, which is secondary in terms of authority to śruti, or the heard tradition. These domestic rites are closely associated with everyday life and are often related to expiation, atonement, averting evil, gaining success, curing disease and life cycle rituals (Gonda 1977:468-69, 565).

42I would like to thank Phyllis Granoff for pointing out this passage to me. P. K. Agrawala also notes this passage, but fails to directly state that Kumära is represented as a dog (1967:15). Chatterjee also notes the passage, but also fails to state that Kumāra is a dog in it (1970:6).

43Sīsara is usually regarded as Saramā's husband, and Syāma and Sabala are guard dogs of Yama. These two dogs may be referred to in Rgveda 10.14.10-12, where they are described as sons of Saramā and as "Die zwei breitnasigen Lebensräuber, die ..... , gehen als Yama's Boten unter Menschen um" (10.14.12a-b translated by Geldner 1951 III:144).

44 Sarama is found in the Rgveda where she retrieves the gods' cows who have been stolen by the Panis, a type of asura (10.108). W. Doniger argues that Sarama is also found in Vedic literature as a long- tongued dog who defiles the sacrifice (1985:100). In the Jaiminiya Brahmana a similar story is recorded (2.440-42; Doniger 1985:97-99), but here she is described as "eating the outer membrane" to free the cows (Doniger 1985:99). This is a biological birth metaphor, she eats the membrane of the cave so that the cows can be 'born' (Doniger 1985:99). Doniger points out that this deed is both good, in that it frees the cows, and tainted in that it involves the eating of the impure amniotic sac (1985:99-100). The idea of the impure Saramā continues into the Mahābhārata (1.3.1-10; Doniger 1985:100). She may also be related to the Rāksasī Dīrghajihvī (Long-Tongue) (Doniger 1985:101-102). This demon is not a dog, but licks and defiles Soma (JB 1.161-63; Doniger 1985:101-102). She fulfills a similar role in the Aitareya Brāhamana (Keith 1981:151). Dīrghajihvī also appears in the Kāśyapasamhitā in an account which involves Skanda-

37

Grahas controlled by Skanda-Kärttikeya in the Epic. These links between Saramā, Kumāra and Skanda-

Karttikeya do not appear to be arbitrary, and they suggest that the Kumāra of the Pāraskara Grhyasūtra is

related to the Skanda-Karttikeya of the Aranyakaparvan. As we have seen, the name Kumāra has some

auspicious associations, but it also clearly seems to function as a name for inauspicious beings who attack

foetuses and children. It may well be that Skanda-Karttikeya's original label as Kumara lay in an

understanding that he was one of these inauspicious 'Kumaras'. Certainly, such a previous existence would

explain 'Kumara' as one of his main epithets, and such an understanding helps to explain why the Epic

views it as logical to have these Kumārakas spring from him.

These Kumäras take Visākha as their father. This is an unusual step because beings of this type are

almost always placed under Skanda-Kärttikeya's direction. This section of the narrative may point to an

earlier tradition where Visäkha was a more prominent figure. It may also be the case, however, that the text

is deliberately separating Skanda-Kārttikeya from these inauspicious beings. The text's agenda is to

construct an auspicious Skanda-Kärttikeya, and part of this agenda is achieved by separating Skanda-

Kärttikeya from these beings. While Visäkha and the Kumäras can be viewed as coming from Skanda-

Kärttikeya, they are still separate beings. It is almost as if the thunderbolt struck off the inauspicious

elements of Skanda-Kärttikeya leaving us with a more appropriate object of Brahminical worship. The text

succeeds in acknowledging the existence of Kumärakas and a relationship between Skanda-Kārttikeya and

them, but in the process attempts to prevent Skanda-Karttikeya from being an inauspicious 'Kumāra'

himself.

A final group of inauspicious beings that are linked to Skanda-Kärttikeya in this text are Grahas.

Before discussing the characterization of Grahas in this text, however, I will first discuss Grahas as they are

found in the Susrutasamhita from the Ayurvedic tradition. This text contains sections which discuss Skanda-

Kärttikeya as a deity connected to illnesses affecting foetuses and children. The dating of this text, however,

2001:217). Kärttikeya. In that story she is defeating the gods and Skanda-Kärttikeya is sent out to kill her (Wujastyk

38

presents some problems. I will begin with the Susrutasamhita and a discussion of just where these texts fit

chronologically into this dissertation.

2.3.3 The Inauspicious Skanda-Kārttikeya: Grahas and Skanda in the Surutasamhitā

K. Zysk suggests that the Ayurvedic tradition develops between the second century BCE and the

second century CE (1991:13), and parts of the Susrutasamhita may originate from this time period. The text

itself, however, is a collection of material likely relating to several different time periods and the product of

several different hands (Meulenbeld 1999 IA:348). Even the legends of the transmission of the text state

that the text was revised by human hands well after its original inception (Meulenbeld 1999 IA:347). There

is some general agreement that the Uttaratantra section of the text, with which we are primarily concerned,

is a late addition to the text, though just how late is not clear (Meulenbeld 1999 IA:344, 348). Meulenbeld

supplies a brief survey of the dates for the Susrutasamhita as gleaned from scholarship on the text; these

dates range between 3000 BCE to the twelfth century CE (1999 IA:342-344). The most popular dates for

final compilation seem to fall between the first and fourth century CE (Meulenbeld 1999 IA:342-343).

Wujastyk also suggests the following concerning the date of the text:

... in Suśruta's text we have a work the kernel of which probably started some centuries BC in the form of a text mainly on surgery, but which was then heavily revised and added to in the centuries before AD 500. This is the form in which we have received the work in the oldest surviving manuscripts today. (2001:106)

I have little doubt that the sections we will examine here belong to this later revision. The mythology of

Skanda-Kärttikeya in the Uttaratantra section of the Suśrutasamhitā shows a knowledge of Mahābhārata

versions of the deeds of the deity and certainly does not come before that Epic text. The Suśrutasamhitā

may have early material in it, but separating that material from late material is, as Meulenbeld suggests,

hazardous and often misleading (1999 IA:348). The text must, then, be viewed as providing a post-epic

understanding of Skanda-Kärttikeya and likely a Gupta era view of the deity, but with some valuable

reflection of an earlier age.

Chronology aside, what is, I think, genuinely of value in this text is its context. We should expect

39

to see a different version of Skanda-Kārttikeya in this Ayurvedic text from that found in the Epic text

because they speak to and from different concerns and perspectives. To some degree the Suśrutasamhitā

does not disappoint this expectation, but both texts share a larger context of Brahminization and a concern

for how to fit this inauspicious deity into a religious system increasingly emphasizing purity, even in the

deities it worships. We do, then, see a different vision of Skanda-Kärttikeya in this text, but we can still

trace the effect of Brahminization and Epic versions of his story on him in this text.

For our immediate purposes, the Uttaratantra section of the text demonstrates two aspects of

Skanda-Kärttikeya. The first is it depicts an understanding of magico-religious disease causation and

healing that comes, in part, from a Vedic heritage (Zysk 1998:7). Diseases afflicting pregnant women and

young children are caused by demonic possession and cured through ritual. The second is that the text

discusses a Graha named Skandagraha and his relationship to a deity who is also called Skanda. The text

explicitly separates a Graha Skanda from the deity Skanda-Kärttikeya. The Graha is dangerous and

malicious, the deity is described as not capable of doing evil. The deity is auspicious, and the Graha remains

inauspicious. What I will suggest, however, is that this text's acknowledgment of a Skandagraha is an

acknowledgment of the deity's actual roots, and its insistence that the two are separate an example of the

powerful effect the shift towards orthodoxy has on this deity's narratives. I will begin by discussing the

magico-religious aspect of this text.

2.3.4 The Inauspicious Skanda-Kārttikeya: magico-religious aspects of the Suśrutasamhitā

The rituals and treatments described in this section of the Suśrutasamhita are primarily magico-

religious. In the case of an attack by Skandagraha the text advises sprinkling the child with a decoction,

smearing the child with medicated oils, feeding the child medicated ghee and the fumigation of the child's

body. The child also wears a garland or amulet made from twigs from somavalli, indravalli and śamī trees as

well as the thorns of a bilva plant and the roots of a mrgādani plant (Uttara. 28.7). The physician is also

instructed to worship Skandagraha for three successive nights either in the child's house or at a crossroads.

Offerings of red flowers, red flags, red perfumes and various foods and grains are made to Skandagraha at

40

this ritual. A cock is to be given and offerings of butter made into a fire (Uttara. 28.8-9). The ritual involves

the production of an amulet or garland made of various plants and bali, or offerings, play a central role, all

common elements of the Atharvaveda tradition. The use of red offerings and the cock are in keeping with

the devices used to signify Skanda-Kārttikeya.45

45 Skanda-Kärttikeya is associated with two birds: a cock and a peacock. In sculpture the cock appears as a symbol of a warrior, but the current context seems to also suggest a magical and ritual use for the bird. In the ancient world cocks were often used as part of a poor person's sacrifice in ritual acts (Swartz 2002), and a white cock is the most common animal sacrifice found in Greek Magical Papyri texts (Smith 1995:23-24). While Greek papyri may seem well removed from the context of Skanda-Kārttikeya, there is significant evidence to demonstrate the influence of Hellenistic thought on the development of Skanda- Kārttikeya during the Kusana period. The Greek magical and sacrificial use of birds and cocks should not be automatically dismissed. As I will demonstrate in my chapter on Kusāna coins and statues, the first use of a bird with Skanda-Kärttikeya in statuary comes from the Hellenized region of Gandhāra and was likely influenced by Iranian and other non-Indian traditions. In the Vedic sacrificial tradition (the Srauta tradition), however, references to cocks are rare. Only the Aśvamedha sacrifice requires the use of a cock (krkavāku), and in accounts of that rite the animal is placed in a list of wild animals to be released as opposed to immolated (Bhide 1967:1). Bhide notes that the cock does appear to have been domesticated by the Vedic period in India and was used in domestic rites like those of the Atharvaveda and particularly those involving "sorcery" (1967:1). In relation to these points he cites Atharvaveda 20.1.36; 5.31.2. Atharvaveda 5.31 concerns a counter spell for witchcraft. It reads: "What (witchcraft) they have made for thee in a cock, or what in a kurira-wearing goat, in a ewe what witchcraft they have made - I take that back again" (5.31.2, translated by Whitney: 1962:279). Bhide also notes that the cock may have been excluded from the high sacrificial tradition because it lacked certain limbs and other biological features that are extracted from the sacrificed animal as part of the rites (1967:2). Bhide also discusses the Vedic uses of the word kukkuta, which is the word used in the Epic for cock. The first reference to kukkuta is found in the Samhitas of the Sukla Yajurveda. He notes, however, that the term appears as a qualification of a ritual implement and does not actually refer to a cock in the rites (1967:4). A kukkuta also appears in the Satapatha Brahmana (1.1.4.18), where it is used as a description of a sweet tongued person. Finally, he notes that the term appears in the Yogasikhopanisad (5.21) and the Trisikhabrāmanopanisad (2.42). In both of these late Upanisads, the bird is not mentioned in a ritual context (1967:4). In Dharmasūtra texts cocks are always listed as banned food for Brahmins or as polluting (DA 1.17.32, 1.21.15; DV 14.48, 23.30; DG 17.29, 23.5; DB 1.12.3). In these texts cocks often appear in a list of other domesticated animals, and it is not clear whether the cock itself is the problem or its general class as domesticated. Bhide makes a similar point concerning Manu's Dharmasūtra. He notes that Manu distinguishes between wild and domestic cocks and prohibits the consumption of domestic cocks (1967:2). Bhide argues that wild cocks are acceptable because they were included in the Aśvamedha sacrifice, and what is acceptable for use in a Vedic sacrifice is also acceptable for Brahminical consumption (1967:3). He suggests that the cock may have been used by non-Brahminical villagers as it is today by villagers as an offering to local non-Vedic deities (1967:3). He also notes that a cock is used in Tantric rituals associated with Skanda-Kärttikeya (1967:5). We might guess from this Tantric reference that the injunction against ritual use of the bird was ignored over time. There does appear, however, to have been some inauspicious association with this bird, and the ritual use of a cock may point to non-Vedic and inauspicious origins for this deity and his rites. I will discuss the colour red in relation to Skanda-Kärttikeya in section 3.4.

41

While the practitioner of these rituals is called a bhisaj, healer or physician (28.10), the rites have

little themselves to do with a proto-scientific form of healing. The rituals are designed to propitiate the

Graha, and they rely on a magico-religious act to bring about a magico-religious result. The healer in this

rite engages in a form of spiritual warfare, where the attacking spirit actually receives worship. As we shall

see, the magico-religious aspect of this text and in particular its treatment of Grahas relate to the rituals to

propitiate Skanda-Kärttikeya in the Aranyakaparvan. The rites for Grahas are all for this worldly ends and

revolve around propitiating dangerous beings. The sense that Skandagraha is dangerous and requires

adoration to become benign best fits the characterization of Skanda-Karttikeya we saw in chapters 214-216

of the Aranyakaparvan. They are, I think, intimately related.

2.3.5 The Inauspicious Skanda-Kārttikeya: Grahas in the Suśrutasamhitā

I will now turn to the Suśrutasamhita's treatment of Grahas. The text mentions nine Grahas:

Skanda, Skandāpasmāra, Šakunī, Revatī, Pūtanā, Andhapūtanā, Šītapūtanā, Mukhamandikā and Naigamesa

(Uttara. 27.4-5).46 We are told "[the Grahas] can injure children for worship. They injure children who are

hit, scolded, anxious or upset or for whom auspicious rituals are abandoned, or who are not kept clean, or

whose mother or nurse exercises improper conduct relating to ordained acts" (Uttara. 27.6).47 Thus, a wife

or nurse who has moral failings, or a poorly cared for child can bring about possession by a Graha, but the

prime motive on the part of the Grahas is that they want to be worshiped.

The Grahas are described as having superhuman powers (aiśvarya) and the ability to take on any

form (viśvarūpa). They can also enter a child's body without detection (Uttara. 27.7), and each Graha

causes specific symptoms in a child (Uttara. 27.8-16). They are the focus of everyday concerns of health

and welfare; as such they are connected to the Atharvavedic and Grhyasūtra traditions, which also deal with

everyday concerns and earthly demons.

46The text states that Naigamesa is also called pitrgraha (Uttara. 27.5).

47"dhātrīmātroh prākpradistāpacārācchaucabhrastānmangalācārahīnān trastān hrstāmstarjitān tāditān vā pūjāhetorhimsyurete kumārān".

42

Skandagraha and the deity Skanda-Karttikeya are discussed throughout the section on childhood

diseases caused by Grahas. The text does not state an explicit separation between the two initially.48 Where

a significant change does occur is in Chapter thirty-seven after the various treatments for the nine Grahas

have been described. This chapter claims to discuss the origins of the Grahas and begins in the following

manner:

The nine, who begin with Skanda, those ones who are graspers of children, who are glorious and who are divided into divine female and male forms, are discussed. They were created by Sūlin [Siva], Agni, Uma and the Krttikās for the protection of Guha, who was abiding in the thicket of

48Chapter twenty-seven of the text discusses general treatments for a Graha attack. Several treatments are prescribed as well as a ritual and a mantra. The mantra praises Agni, the Krttikās and Skanda, who is also called Guha in this section. It is primarily the deity Skanda-Kärttikeya to whom the mantra is directed: "Obeisance to god Skanda, obeisance to the chief of the Grahas. I respectfully bow my head to you. May you accept my offering. May my child quickly become free from sickness and normal" (Uttara. 27.21) (namah skandāya devāya grahādhipataye namah śirasā tvābhivande aham pratigrhnīsva me balim nīrujo nirvikāraśca śiśurme jāyatām drutam). This mantra presents Skanda-Kārttikeya or god-Skanda as the lord of the Grahas, but does not directly state that this lord of the Grahas is a Graha himself. Chapter twenty-eight deals specifically with the treatment for an attack by Skandagraha. A number of rites are prescribed, and the only specific mention of who is worshiped in the description of the rituals is Skandagraha: "Thus said the glorious Dhanvantari: [The treatment] of children who are plagued by Skandagraha is declared .... A good cock is of benefit when a Skandagraha is the afflictor" (yathovāca bhagavān dhanvantarih skandagrahopasrstānām praśasyate .... sukukkutah skandagrahe hitāya) (Uttara. 28.2-3, 28.8). The mantra occurring at the end of the chapter, however, makes no direct mention of Skandagraha:

May the god Skanda be satisfied; he who is unchanging and the treasure chest of handsomeness, of splendor, of energy and of ascetic heat. May the glorious Guha, who is the destroyer of the enemies of the army of the gods, the eternal one, the general of the army of the gods, the god who is the general of the army of the Grahas, protect you. May he bestow protection on you; he who is the son of Agni, the great god of gods, the Krttikas, Uma and Ganga. May the god, who is the Krauñca splitter, who is handsome, charming, red, adorned with red sandal paste, who is glorious and who is clothed in red garlands, protect you. (Uttara. 28.11-14). (tapasām tejasām caiva yaśasām vapusām tathā nidhānam yo'vyayo devah sa te skandah prasīdatu [.11] grahasenāpatirdevo devasenāpatirvibhuh devasenāripuharah pātu tvām bhagavān guhah [.12] devadevasya mahatah pāvakasya ca yah sutah gangomākrttikānām ca sa te śarma prayacchatu [.13] raktamālyāmbarah śrīmān raktacandanabhūsitah raktadivyavapurdevah pātu tvām krauñcasūdanah [.14]).

In this quotation 'Skanda' is always referred to as deva, and his ties to the Skanda-Kärttikeya of the Mahābhārata are made clear through references to his role as the general of the army of the gods, his parentage and the mention of his destruction of Mount Krauñca. There is nothing at all here to separate Skandagraha from the deity Skanda-Kārttikeya.

43

reeds, and was protected by their own energy. (Uttara. 37.3-4)49

Guha is Skanda-Kärttikeya, and we have to wonder how the nine Grahas were created to protect him, if he

is one of them? The text continues with "Skanda was created by the glorious destroyer of Tripura (Siva).

That Graha, has another name, Kumära" (Uttara. 37.8).50 The text is subtly distinguishing between a Graha

who has the names Skanda and Kumara, and a deity who has the names Guha, Skanda and Kumara. The

Graha discussed in verse eight only has Siva as his parent, while the Guha of verse four had four parents.

The Skanda-Kumara of verse eight is clearly labeled a Graha, while the Guha of verse four is not. Verse

eight also follows from an explanation of why and by whom the Grahas Naigamesa and Skandāpasmāra

were created: "Naigamesa, who is a ram-faced Graha, who holds a child, was created by Pārvatī as a

beloved friend of the god Guha. That one who is called Skandäpasmära, he [was created] by Agni, and he is

equal in radiance to Agni. He is also called Visakha and is a friend of Skanda" (Uttara. 37.6-7).51 The

sequence of verses leading up to verse eight describe Grahas and their relationship to either a god named

Guha or Skanda. If verse eight follows the same pattern, it is describing a Graha named Skanda, who should

not be the same as the deity Guha-Skanda-Kumära. This ambiguity is resolved in the following verses:

The one who exhibits the divine play of acting like a child is produced from Rudra and Agni. That glorious one cannot himself be occupied in improper conduct. Regarding this matter, physicians, who have little knowledge, who are unlearned in things and who only know about the body declare, 'Kumära is considered as the same as Skanda.' (Uttara. 37.9-10)52

These verses seem to state that the Deva born from Rudra and Agni is incapable of wrong acts, of being a

rogue. He is childish and innocent. This is not a description of Skandagraha, who makes children ill. Verse

49“nava skandādayah proktā bālānām ya ime grahāh śrīmanto divyavapuso nārīpurusavigrahāh [.3] ete guhasya raksārtham krttikomāgnišūlibhih srstāh śaravanasthasya raksitasyaātmatejasā [.4]".

50"skandah srsto bhagatā devena tripurārinā bibharti cāparām samjñām kumāra iti sa grahah".

51“naigamesastu pārvatyā srsto mesānano grahah kumāradhārī devasya guhasyātmasamah sakhā [6] skandāpasmārasamjño yah so'gnināgnisamadyutih sa ca skandasakhā nāma višākha iti cocyate [7]".

52"bālalīlādharo yo'yam devo rudrāgnisambhavah mithyācāresu bhagavān svayam naisa pravartate [9] kumārah skandasāmānyādatra kecidapanditāh grhnātītyalpavijñānā bruvate dehacintakāh [10]".

44

ten goes on to state that it is only due to ignorance that people equate the Deva with the Graha. It is due to

the similarity of their names that, according to this text, causes confusion, a point the commentator,

Dalhana, tries to clarify by stating: "As for the statement beginning with Kumāra, Kumāra is Kārttikeya.

Equal to Skanda means equal to Skandagraha by name" (1963:667).53 In other words, Guha-Skanda-

Kumara is the god, who is not the same as Skanda-Kumara-graha, the Graha who afflicts children. The

differences between the two figures that I would note are titles (one is repeatedly called Deva, while the

other is called Graha), names (Guha is only used for the Deva), parentage (the Graha is produced by Siva,

the Deva by multiple parents), and auspicious or inauspicious actions (the Deva is incapable of doing

improper deeds, while the Graha does improper deeds).

Once the text has established that the Deva and Graha are separate, it continues by explaining how

the Grahas came to afflict children. Siva explains to Skanda-Kärttikeya that gods and humans engage in

mutual assistance to survive. The gods change the seasons at the proper time and, in return, humans worship

the gods. These Grahas will attack the children of parents who do not worship properly or fail on various

accounts to maintain purity or propriety. These parents will, in return, worship the Grahas, and that will be

their livelihood (Uttara. 37.13-20). The propitiatory nature of the Graha cult is maintained here, but their

actions are moralized in that they become forces of divine retribution. While they are inauspicious and

dangerous, they still participate in a Vedic sense of order.

The chapter ends with the following:

Thus children who are known to be afflicted by a Graha are very difficult to cure. In the case of [an attack by] Skandagraha it is known that permanent weakness and even death [will occur]. Of all of those [Grahas], Skandagraha is known as the most dark and horrible, but [an attack from] another Graha, who can take any form, is also not curable. (Uttara. 37.21-22)54

These final comments reflect the text's attempt to make a significant difference between the Deva Skanda-

53“kumāra ityādi kumārah kārttikeyah skandasāmanyāt skandagrahnāmatulyatvādbālam ... ".

54“grahopasrstā bālāstu duścikitsyatamā matāh vaikalyam maranam cāpi dhruvam skandagrahe matam [21] skandagraho'tyugratamah sarvesveva yatah smrtah anyo vā sarvarūpastu na sādhyo graha ucyate [22]".

45

Kärttikeya and the Graha Skanda. The Graha is clearly a dangerous being bent on malicious acts. His

character lies in stark contrast to the description of the deity who cannot do harm. The Suśrutasamhitā does,

then, follow the Epic's agenda of promoting an auspicious version of Skanda-Kärttikeya. What this

Ayurveda text cannot do is ignore the existence of a Skandagraha. While the Mahabhārata never

acknowledges the existence of a Graha named Skanda, it seems clear from the Susrutasamhita that such a

being was understood to have existed and for some people was regarded as the same as the deity Skanda-

Karttikeya. I would argue that such was likely the case, and the separation between the inauspicious and

demonic Skanda from the divine Skanda is part of the process of creating an orthodox deity with which both

texts are engaged.55

The Suśrutasamhita is important to this study because to my knowledge all of the Epic and

Puranic accounts of the deity, with the exception of the Aranyakaparvan, ignore the existence of Grahas

associated with Skanda-Kärttikeya. This Ayurvedic text and others like it help us to understand that the

appearance of these beings in the Aranyakaparvan was not an Epic invention that was quickly lost, but an

acknowledgement of one important aspect of the Skanda-Kärttikeya cult and, perhaps, its roots. We leave

the Ayurvedic tradition with a general understanding that there is a great deal of Skanda-Kārttikeya

mythology that the Epic accounts of him ignore or marginalize. Grahas and Skandagraha likely survive in

the Uttaratantra of the Susrutasamhita because the genre of this text as Äyurvedic and religio-magical

demands that it discuss Grahas. That the remainder of the Epic and Puranic tradition attempts to uproot

Skanda-Kärttikeya from the inauspicious and dangerous aspect of these Grahas suggests that the

Brahminization of this deity made radical changes to his cult.

55I should also note that a very similar conclusion is drawn by Michel Strickmann in his account of this deity from Chinese Buddhist texts. He argues "Other mentions of his name [Skanda-Kārttikeya's] in Buddhist scriptures make it clear that he originally belonged to a particularly fearsome group of fifteen demon-thugs, a gang that attacked infants and young children" (2002:219). He is eventually transformed into "the guardian par excellence of Chinese Buddhist monasteries" (2002:219). Hence, the progression of his cult in India is mirrored in part by the progression of his cult in China within a Buddhist context. That Skanda-Kärttikeya first appears in China as a Graha-like being can be accounted for by the strong Indian influence on early Buddhism in China.

46

The Grahas depicted in these texts are also important because they had an active cult in Kusāna

Mathurā. There are several panels from Mathura which depict a six-headed female divinity flanked by two

male divinities. I will discuss these panels in more detail in my section on Kusana coins and statuary. For

now we may note that the two males have been identified as Skanda-Karttikeya and Visākha and the female

as Sasthī. Sasthi is a Graha; her role in the Skanda-Kärttikeya cult will be discussed in more detail in my

section on Yaudheya class three coinage. As we shall see, there is archaeological evidence to suggest that

the Kusāna era Mathurā cult of Skanda-Kärttikeya and Visākha was focused on him as a Graha, who was

also closely associated with Mätrs. In short, the theory that Skanda-Kārttikeya began as a terrible Graha

named Skandagraha is substantiated when we take into account the evidence from art and the

Suśrutasamhitā.

2.3.6 The Inauspicious Skanda-Karttikeya: Grahas in the Aranyakaparvan

Chapter 219 of the Aranyakaparvan also discusses Grahas. By this stage of the text Skanda-

Kārttikeya has been transformed into an auspicious deity, but his association with these creatures continues.

Grahas and Mätrs are closely related in this text. We have already seen that Sarama is called both a Graha

and Mätr (3.219.33). Mätrs are also depicted as producing Graha/Kumära-like beings. The text states that a

group of Mätrs had:

a son, who was possessed of power, who was very terrible, who had terrible red eyes and who was a gift of Skanda. These are called the group of eight. They sprang from the Matrs and Skanda. Along with the goat-faced one, they are called Navaka ('Consisting of Nine'). (3.217.10-11)56

The text then continues with the section on Skanda-Kärttikeya's sixth-head as a goat's head already cited

above. These eight heroes are, presumably, the seven Matrs and their child. The ninth, "chagavaktra" or

'goat-face,' appears to be like Naigamesa, and the total of nine suggests that this section is an attempt to

explain the origins of the navagrahas, the Nine Graspers already mentioned in my section on the

56“etāsām vīryasampannah šiśurnāmātidārunah skandaprasādajah putro lohitākso bhayamkarah [.10] esa vīrāstakah proktah skandamātrganodbhavah chāgavaktrena sahito navakah parikīrtyate [.11]". My thanks to Phyllis Granoff for helping with the Sanskrit in this section.

47

Suśrutasamhitā.57

A Graha is also produced by Skanda-Kärttikeya to help the Mätrs gain worship in chapter 219.

Skanda-Kārttikeya tells these Mātrs:

Injure the young children of people until they are sixteen years old with your different forms, and I will give to you a violent being from myself. You, who are much worshipped, will dwell along with it in the best of happiness. Mārkandeya said: Then a being which glittered with gold and was mighty flew out from the body of Skanda. He was to consume the offspring of mortals. Then, it fell to the earth senseless and hungry. He had a fierce form and was allowed by Skanda to become a Graha. Brahmins call that Graha Skandāpasmāra ('Skanda's Falling Sickness'). (3.219.22-25)58

In the Ayurvedic tradition Skandāpasmra is considered to be a Graha. The Suśrutasamhitã does not view

Skandāpasmāra as a form or product of Skanda-Kārttikeya, but actually regards Visākha and

Skandäpasmära to be the same being (Uttara 29.9; 37.7). Whatever the case, Grahas and Mätrs are closely

connected in this text as inauspicious beings who are dangerous to children and connected to Skanda-

Kārttikeya.

In this chapter Vinata, who is described as a "very fierce Bird Graha" (3.219.26) (mahāraudrā

śakunigrahah), approaches Skanda-Kärttikeya and asks to become his mother and dwell with him forever.59

57The connection between the Mätrs and Grahas is an important one because, I would argue, that many of the female Grahas that appear in medical texts may have begun as Matrs. The two groups share similar functions and attributes, and it may be the case that some of these Mätrs were considered as Grahas once the Saptamätrkas became established as the group of Mätrs at which point these other Mätr groups begin to appear under different names. The navagrahas are probably the nine planets. Here we see the conflation of two distinct groups based on the similarity of their names.

58“skanda uvāca yāvatsodaśa varsāni bhavanti tarunāh prajāh prabādhata manusyānām tāvadrūpaih prthagvidhaih [.22] aham ca vah pradāsyāmi raudramātmānamavyayam paramam tena sahitā sukham vatsyātha pūjitāh [.23] mārkandeya uvāca tatah śarīrātskandasya purusah kañcanaprabhah bhoktum prajāh sa martyānām nispapāta mahābalah [.24] apatatsa tadā bhūmau visamjño'tha ksudhānvitah skandena so'bhyanujñāto raudrarūpo'bhavadgrahah skandāpasmāramityāhurgraham tam dvijasattamāh [.25]". Skanda's falling sickness may mean epilepsy.

59There may be a few Vinatas in the Mahabhārata. One Vinata is a daughter of Prajāpati and wife of the sage Kaśyapa. Kaśyapa grants her a boon to have two equal and powerful sons. She lays two eggs (it may be the case that Vinata is some sort of bird, but this point is never clearly related in the text), but becomes impatient for them to hatch. She breaks one open to find a half-formed child who curses her. The second child is Garuda (1.14.5-25). A Vinata is also named among the thirteen daughters of Daksa

48

She may be an example of a Graha/Mätr and may be the equivalent of the Graha Sakuni, who appears in the

Suśrutasamhitā (Uttara 27.4; 27.10; 30.1-11) and the Kāśyapasamhitā (Tewari 1996:367) as a bird-shaped

snatcher of children. The Epic may have used the name Vinata because she is more familiar to the Epic

audience as the mother of Garuda. Chapter 219 continues by listing other Grahas most of which are also

found in medical texts.60

(1.59.13). She is also cited as the mother of several mythological birds (1.59.40).

60 All of these Grahas are described as horrific creatures. Pūtanā is described as a Rāksasī and as having "a terrible shape, a Night-Stalker with a dreadful form" (3.219.27) (kastā dārunarūpena ghorarūpā niśācari). Chapter 32 of the Uttaratantra of the Surutasamhitā is devoted to the treatment of an attack from this Graha. The mantra to her in the Susrutasamhita describes her in the following manner: "may the slovenly shag-haired goddess, Putana, who is dressed in dirty clothes, and who loves to haunt lonely dwellings, preserve the child. May the fierce-looking, frightful goddess who is as black as a dark rain-cloud, who loves to haunt lonely and dilapidated human dwellings and whose body gives off filthy odours protect the child from all evils" (32.10-11 translated by Bhishagratna 1963 III:153-154). Pūtanā is also a Graha appearing in the Kasyapasamhitā, a seventh century Ayurvedic text (Tewari 1996:154, 173). The most famous accounts of her are likely from the Harivamsa and Bhagavata Purana, where she is sent by the evil Kamsa to kill the infant Krsna. She is described as "a devourer of children" and as wandering "through cities, villages, and pastures, killing infants" (BP quoted in Doniger 1975:214-215). Krsna, as the boy-god, however, is able to destroy this demon. The relationship between these two boy-gods, Krsna and Skanda- Kärttikeya, is not without interest, but will not be taken up in this thesis. A Graha named Śītapūtanā is also in chapter 219. She is described as a Pisācī and as having a "dreadful shape" (dārunākārā) who "steals foetuses of women" (3.219.27) (garbhānsā mānusīnām tu harate). The treatment for her attack is dealt with in chapter 34 of the Uttaratantra of the Suśrutasamhitā. There she is described as delighting in wine and blood and as residing by the side of rivers or tanks (34.9). Another Graha in chapter 219 is Revati or Raivata: "They call Revati Aditi. Raivata is the Graha of her. This great and terrible Graha attacks young children" (3.219.28) (aditim revatīm prāhurgrahastasyāstu raivatah so'pi bālāñsisūnghoro bādhate vai mahāgrahah). Chapter 31 of the Uttaratantra of the Suśrutasamhitā deals with this Graha. There she is described as "tall, drooping and terrible-looking, and who is the mother of many sons" (Bhishagratna 1963 iii:153). In the Kaśyapasamhitā Revatī is a major goddess with a large section, the Revatikalpa, devoted to her mythology. In this text she is described as a Graha. The recitation of her twenty names twice daily is said to protect a child from attack from other Grahas (Tewari 1996:169). In the Kāsyapasamhitā Skanda-Kārttikeya tells her that he regards her and her five brothers, who are not named, and her sister, Sasthi, as deserving the same worship as himself. He establishes the sasthi day, the sixth lunar day, as the day of their worship (Tewari 1996:171). Chapter 219 also mentions Mukhamandikā as "exceedingly dangerous to approach", and who "rejoices in the flesh of children" (3.219.29) (atyartham śiśumāmsena samprahrstā durāsadā). The Suśrutasamhitā also deals with this Graha (Uttara 35), where she is described as beautiful and dressed in ornaments (35.8-9). Why she is described in such a manner is not clear. The Kāsyapasamhitā also discusses her (Tewari 1996:176). There are also a number of Grahas mentioned in this text who do not appear in medical texts. A child swallowing bird is mentioned which is perched on Surabhi (3.219.32). The divine mother of dogs, Saramā takes the foetuses of people, and the mother of trees who lives in the karañja tree is also worshiped for the birth of a son (3.219.33-34). Kadrü is described as eating the foetus of a woman and causing her to

49

This chapter of the Aranyakaparvan does display a knowledge of Grahas large parts of which fit

well with the Ayurvedic tradition. The writers of this Epic version have, however, remained silent about

certain parts of the understanding of Grahas related to Skanda-Kärttikeya and, in particular, the existence of

Skandagraha. Skanda-Kärttikeya does appear to sanction the actions of Grahas, and these beings do come

from him, but he does not act himself as a Graha in this text and is never called one. Given this text's

apparent knowledge of the Ayurvedic tradition concerning these Grahas, it is odd that the text does not

recognize Skandagraha. We might well think that the omission of Skandagraha is far from accidental, but a

deliberate attempt to present an auspicious deity who rules over Grahas, but is not one himself.

The closest the Aranyakaparvan comes to acknowledging Skandagraha is in calling the collection

of Grahas it lists 'the Skanda Graspers' (skandagrahah) (3.219.42). The plural, skandagrahah, is not a term

that is used in Ayurvedic texts. In those texts there is only the singular, skandagraha, who is a particular

figure. The Mahabharata carefully works around the existence of a Skandagraha. As we have seen in the

Suśrutasamhitā, the problem with Skandagraha is that he is inauspicious and malign, and the

Äranyakaparvan is attempting to present a divine senapati who is brahmanya, Brahminic, and auspicious.

The tradition of Grahas and Skanda-Kärttikeya's association with them must have been a strong one or

these inauspicious beings and Skanda-Kärttikeya's link to them would have been edited out of this text and

the Ayurvedic tradition. What the text does do is to allow for the existence of Grahas and even plays with a

term like skandagraha, but establishes Skanda-Kärttikeya as different from these beings. While the Epic

itself never acknowledges Skanda-Kārttikeya as a Graha, I have argued that we need to pay attention to this

text's agenda in recasting this deity as auspicious and attempt to recognize the aspects of this deity's

give birth to snakes (3.219.36). There are also Gandharva and Apsarās Grahas who snatch away foetuses and children (3.219.36-37). A being called Lohitāyanī, who is described as a nurse of Skanda-Kārttikeya and the daughter of the blood sea is also mentioned (3.219.39). Finally, this section of the chapter ends with a mention of a goddess called Arya: "Just as Rudra is in men, so Ārya is in women. Arya, who is the mother of Kumära, is worshiped separately to obtain wishes" (3.219.40) (purusesu yathā rudrastathāryā pramadāsvapi āryā mātā kumārasya prthakkāmārthamijyate). Āryā appears as a name of Pārvatī (Monier- William 1999:152).

50

character to which the text seems to be reacting. The close association between Skanda-Kārttikeya and

Mātrs, Kumāras and Grahas all point towards a deity who began with close affinities with these beings. If

we take into consideration evidence from the textual and sculptural traditions we cannot escape the

conclusion that Skanda-Kärttikeya's origins are located among the Grahas.

2.3.7 The Inauspicious Skanda-Karttikeya: The worship of Skanda-Karttikeya and Grahas in the Aranyakaparvan

The worship of these Grahas through Skanda-Karttikeya is presented by the Aranyakaparvan as

the key to controlling them:

Offerings of oblations and the worship of Skanda in particular [are to be performed]. They [the Grahas] need to be pacified. They should be bathed, given incense and unguent, and offerings should be made. All of them, who are thus honoured and properly offered obeisance and worship, give prosperity and a vigourous long-life to people, O King of kings. (3.219.43-44)61

As with the inauspicious Skanda-Kärttikeya presented in the early chapters of this account, the key to

controlling these fearful beings is through worship and especially the worship of Skanda-Kārttikeya

(skandasyejya). We cannot be sure of the exact nature of the "skandasyejya", but it may be related to the

ritual described in the Skandayäga chapter in the Atharvaveda Parisista.62 I will briefly turn to this text to

examine this ritual and its links to the current discussion.

2.3.8 The Inauspicious Skanda-Kärttikeya: The worship of Dhūrta in the Skandayāga

The Skandayäga, or, as it is also known, the Dhurtakalpa, is book twenty of the Atharvaveda

Parisista and likely represents an early account of rituals performed for Skanda-Kārttikeya. The text is

short, containing nine kandas and under fifty verses. The exact date of the text is not clear, but I do not

think that it represents the earliest account of Skanda-Kärttikeya. The text shows a clear knowledge of

Skanda-Kärttikeya's multiple parentage and does not represent an account of the deity that predates the

61“tesām praśamanam kāryam snānam dhūpamathāñjanam balikarmopahāraśca skandasyejyā visesatah [.43] evamete'rcitāh sarve prayacchanti śubham nrnām āyurvīryam ca rājendra samyakpūj ānamaskrtāh [.44]"

62The Atharvaveda Parisista is part of the late Vedic ancillary textual tradition.

51

Mahābhārata's Āranyakaparvan version of the deity's birth. An approximate date of around the beginning

of the Common Era for the text has been suggested and will do for our purposes here (Modak 1993:473).

The text places Skanda-Kärttikeya within an Atharvavedic and ritual context, and it helps us to understand

why people worshiped this deity.

The text begins by stating that the ritual should be done every four months, after a fast, and it

should be performed in a north-east direction in a clean place that has no salt (AVP 20.1.3). The ritual

begins, then, by respecting the ideals of ritual purity represented in the Vedic tradition. A hut is made in

which are placed a garland of leaves of all tress, bells, banners, wreaths, mirrors and an amulet (20.1.3).63

Then, the text states: "the one who is carried by white horses should be caused to be brought (invited,

invoked) there" (tatra yam vahanti hayāh śvetā ityāvāhayet) (20.1.3).64 As Dhūrta is invoked or brought in

63 .krtvā madhye mandapasya sarvavānaspatyām mālām krtvā ghantāpatākāsrajah pratisaram ca mālāprsthe krtva madhye darpanāmscopakalpayitrā tatra ... ".

64The next seven verses all mention an attribute of the deity and end in "I cause [Dhūrta] to be brought (invoked)" (āvāhayāmyham) (20.2.1). P. K. Agrawala assumes from these verses that a statue of the god is placed in the circle during these sections (1967:107), and he suggests: "In the ritual the god was probably represented by an image perhaps in clay, which in the end was discarded and with due ceremony was thrown in water" (1967:20). Modak also suggests that an image of Dhürta or Skanda-Kārttikeya was used during the ritual (1993:296). There is, however, nothing within the Skandayäga which states that a statue is to be used (Granoff forthcoming:17), and, given the precise list of elements that are to be placed within the circle, one might expect an explicit statement concerning the use of a statue, or what it was made of, if it were to be used in the ritual. The Atharvaveda Parisista is not without references to the use of images. Several sections on sorcery suggest the making of an image (akrti) to defeat one's enemy (35.1.6; 36.5.1-3). These images are not, however, of deities, and the context of sorcery is not invoked in this rite. We are also usually given specific instructions as to the materials used to make the image in this text if an image is required. The opening verses of Atharvaveda Parisista 20 might be associated with bringing an image into the circle, but they may also be designed to invoke the deity into the ritual area. The verb a-vah used in the verses can mean both 'carry' and 'invoke', and the exact intent of the passages is not clear to me. Phyllis Granoff notes that image worship does appear to have influenced the actions and mantras prescribed in the Skandayäga, but there is nothing explicit within the text to demonstrate the use of an image. She concludes that the details provided by the text suggest that no image was used in this rite (forthcoming:17) This ambiguity concerning the rite may be a deliberate device of the Atharvaveda Parisista collectors and writers. A statueless invocation style of ritual is more in keeping with the Vedic heritage to which this text clings. An explicit mention of an image of a deity might be viewed as un-Vedic and may not have been a step Brahminical redactors were willing to take. If, however, we accept the date of the text to around the beginning of the Common Era, there may have been statues available for such a rite. Some of the early statues of Skanda-Kärttikeya are small and transportable, such as the metal statue found at Sonkh

52

several verses of praise are recited. Once the deity is in the circle, offerings accompanied by more verses of

praise are given. Water for a foot bath, perfumes, flowers, incense, a lamp, leaves and various types of food

are offered. A fire is kindled and consecrated after which more praise is given to Dhürta and requests for

specific goods and worldly prosperity are made. While reciting mantras and praises and holding blossoms,

incense and a garland, the priest turns to the right three times and ties the amulet. The protective properties

of the amulet are eulogized, and we are told that Dhurta leaves or is carried out of the circle. The ritual ends

somewhat obscurely with the ritualist waiting for night fall, entering his home, and while looking at his wife

he says: "O wealthy one give me a gift" (20.6.10) (dhanavati dhanam me dehiti). The end of this rite

suggests a concern for reproduction.

The rite seems to have two goals: the acquisition of wealth and the production of a protective

amulet. The request for wealth appears as a general list of worldly goods:

I will praise the beautiful wish granter [the son] of the Krttikās, Agni and Siva. He who is praised and who has all forms let him grant all my desires. [May] he [grant] me wealth, herds with grain and money, a knowledge of words, male and female slaves, a house, a jewel, and divine ointment. Those who honour the Brahminic and beautiful Dhurta with worship, they all should have wealth and beautiful children. (20.5.1-3)65

Such requests for wealth are not out of keeping with the Atharvavedic tradition or the Vedic tradition in

general. These types of requests are not what we might expect from a ritual associated with the Skanda-

Kārttikeya as he appears in the Mahābhārata and Susrutasamhitā, but the attainment of wealth is, in the

Atharvavedic tradition, simply another aspect of gaining a deity's favour and protection.

The rite closes with the tying of the amulet, and the production of this amulet may be the real point

of the rite:

which is only 9.3 cm high (Härtel 1993:281). The ambiguous wording of this text would allow for such a transportable image to be used in the rite without explicitly referring to the image or without rejecting Vedic ideals of worship.

65" šivāgnikrttikānām tu stosyāmi varadam subham sa me stuto viśvarūpah sarvānarthānprayacchatu [.1] dhanadhānyakulānbhogānsa me vacanavedanam dāsīdāsam tathā sthānam maniratnam surāñjanam [.2] ye bhaktyā bhajante dhūrtam brahmanyam ca yaśasvinam sarve te dhanavantah syuh prajāvanto yasasvinah [.3]".

53

... with [the words beginning] 'a thread, which is cut by the Adityas,' he should tie the amulet .... "I bind this amulet which is a destroyer of all foes and a killer of the inauspicious, gives long life, confers fame and wealth. There is no fear from humans and Gandharvas, from Pisācas and Raksasas [with it], and even from an evil action that has been done. He is released from the acts of others as well as his own. He has no fear from malevolent spells, from witches and from inauspiciousness due to the acts of women. There is no fear of that so long as he keeps the thread. As long as the mountains and cows and waters remain, so long will he who keeps the thread have no fear". (20.6.8, 20.7.2-6)66

This amulet and its properties place this ritual firmly within the Atharvavedic magico-religious tradition

(Zysk 1998:14, 97; AV 2.4, 19.34, 6.85, 10.3, 8.5, 10.6). Just what Dhūrta has to do with the production of

the amulet is not clear. We may assume, however, that it is his presence in the ritual that empowers the

amulet.

The amulet and its properties do fit the Atharvavedic tradition and the general comments about

propitiating the Grahas made in chapter 219 of the Aranyakaparvan. In chapter 219 we are told that Grahas

should be propitiated with gifts and the rite of Skanda. We have seen that the Skandayāga involves a great

deal of gift giving. The rites in the Aranyakaparvan are described as a puja, which is basically what the

Skandayāga is. Once propitiated, the Āranyakaparvan states, these Grahas "give prosperity and a vigourous

long-life to people" (3.219.44). The hoped for results of the ritual described in the Skandayāga are not far

removed from those described in the Aranyakaparvan, though the Skandayaga is much more closely tied to

the Atharvavedic tradition than the Āranyakaparvan.

The opening lines of the text refers to itself as the Dhūrtakapla (20.1.1). The word Dhūrta means a

rogue or a cheater (Monier-Williams 1999:518), and it is the primary name for Skanda-Kārttikeya in this

text. In other texts, however, it is an unusual name for Skanda-Kärttikeya. The text is devoted to this figure

of Dhūrta and as such may give us some insight into how one form of Skanda-Karttikeya was perceived by

66" ādityakartitam sūtramiti pratisaramābandhīyāt [.8] dhanyam yaśasyamāyusyamaśubhasya ca ghātanam bandhāmi pratisaramimam sarvaśatrunibarhanam [.2] raksobhyaśca pišācebhyo gandharvebhyastathaiva ca manusyebhyo bhayam nāsti yacca syādduskrtam krtam [.3] svakrtātparakrtācca duskrtātpratimucyate sarvasmātpātakānmukto bhavedvīrastathaiva ca [.4] abhicārācca krtyātah strīkrtādaśubham ca yat tāvattasya bhayam nāsti yāvatsūtram sa dhārayet [.5] yāvadāpaśca gāvaśca yavatsthāsyanti parvatāh tāvattasya bhayam nāsti yah sūtram dhārayisyatīti [.6]".

54

the Atharvavedic tradition. That Dhürta is understood to be Skanda-Kärttikeya is made clear by the text in a

number of places. Dhürta is described as having a number of characteristics many of which are associated

with Skanda-Kärttikeya. He is carried by peacocks (20.2.3), and cocks with red eyes are described as his

(20.6.3). He is described as carrying his spear (sakti) with bells and banners (20.2.5). He is described as

constantly surrounded by mothers (matrs) (20.2.6-7). Verse 20.2.8 asks him to come with a group and an

army. The same verse describes him as having six mouths and lips, and ten eyes. He also has a golden

colour and he shines. Other verses refer to Siva, Agni, Indra and the Krttikās as his parents (20.5.1; 20.6.4).

All of these features are related to Skanda-Kärttikeya from the Epic and Ayurvedic traditions. Several

names are also given to him besides Dhūrta. He is also called: Kārttikeya (20.2.9), Skanda (20.4.2, 20.6.2,

20.6.3), Brahmanya (20.5.3),67 and Kumāra (20.6.2).68 There are also some important adjectives applied to

67This is an important name for this study because the Yaudheyas use this name on a series of their coins depicting Skanda-Kärttikeya or Brahmanya. The term means religious or pious as well as devoted to knowledge or relating to Brahma (Monier-Williams 1999:741). Within the verse itself, it does not, however, stand alone as a name of Dhūrta, but it is an adjective of Dhūrta: "ye bhaktyā bhajante dhūrtam brahmanyam ca yaśasvinah" (20.5.3).

68The following are also listed as names of Skanda-Kārttikeya in this text: Vināyakasena (20.3.1), Mahīpati (20.4.2), Pinākasena (20.4.2, 20.7.11), Bhrātrstrīkāma (20.4.2), Svacchanda (20.4.2), Varaghanta (20.4.2), Nirmila (20.4.2), Lohitagātra (20.4.2), Śālakatańkata (20.4.2) and Svāmin (20.6.5). Vināyakasena is not without interest. Vināyaka means "'remover (of obstacles)'" (Monier-Williams 1999:972). As a name it is usually associated with Ganeśa (Monier-Williams 1999:972). The term can also mean a certain class of demons or particular formulas made over weapons (Monier-Williams 1999:972). Vināyakasena, then, can carry the meaning of leader of the demons (vinayaka), or leader or chief obstacle remover. P. K. Agrawala feels the name means "having an army of the Vināyakas" (1967:20). Mahīpati means earth-ruler or king (Monier-Williams 1999:803). It is not completely clear to me that Mahipati is intended to refer to Skanda- Kärttikeya in this verse. Pināka means a bow, but especially the bow of Rudra-Siva which in the Vedic tradition caused illness (Monier-Williams 1999:627). Pinakasena means "'armed with a bow"" and appears to be a name of Skanda-Kärttikeya which is only used in this text (Monier-Williams 1999:627). The name has clear associations with Siva and disease within the Atharvavedic context. P. K. Agrawala does not find the term to be clear. He notes the connection to Siva, but suggests that the term meant "an army of bow- wielding soldiers or sharp-shooters" (1967:19). I am not convinced by this definition because it assumes the dominance of Skanda-Kārttikeya as Mahāsena, which is not attested to by this text. Loosely, Bhrātrstrīkāma means the one who desires or lusts after the wife of his brother. It does not occur as a name of Skanda- Kärttikeya in any other text, and its exact meaning here is not clear. As we have seen, the text does state that Visākha is Dhūrta's brother, but no mention of a wife of either of them is directly made. Modak feels this name reflects a libertine nature for Dhürta, and implies he is a rogue and cheat in several aspects of life (1993:296). P. K. Agrawala also feels the name reflects on Dhūrta's character (1967:20). Svacchanda means self-willed (Agrawala 1967:109), uncontrolled or independent (Monier-Williams 1999:1275). The

55

Dhūrta like Sadāsya (six-mouthed) (20.6.2) and Sadānana (six-mouthed or six-headed) (20.2.8), which both

point to his six heads. These names make it clear that the text understands Dhūrta to be Skanda-Kārttikeya.

Whether or not Dhürta was always regarded as Skanda-Kārttikeya is an open question.

The ritual ends with an obscure reference to a creature called Pramoda: "The wicked Gandharva

Pramoda by name runs about. Remove sins, remove, remove, release" (20.7.7-9)!69

This Pramoda is also mentioned in the Salyaparvan version of Skanda-Kärttikeya's birth story. In that text

Pramoda is part of a long list of names that are regarded as part of Skanda-Krttikeya's divine army

(9.44.60). In the quotation above, however, Pramoda is clearly a wicked or bad (pradosa) being, and this

text acknowledges that Dhurta is associated with dangerous beings.

This text does seem to be aware of the Epic version of this deity, but does not always agree with

the account given in the Epic. This text does not emphasize the parentage of Siva and Umā as later texts

will, and, while it acknowledges the role of the deity as the devasenapati, that aspect of the god is a side

note in this text. This text primarily describes him as a deity who is concerned with the protection and well

being of humans. As such, he is a very earthly god who is not obviously connected with divine armies and

battles with asuras. The demons mentioned by this text, raksasas, pisācas and gandharvas, are primarily

demons who pray on humans and their possessions. Asuras are the demons of the gods, and it is these

beings that the later sections of the Aranyakaparvan has Skanda-Kärttikeya fight. The nature of the cult

described in the Skandayāga as a propitiatory cult designed to make Skanda-Kārttikeya protective is all

term seems to suggest a wild character, who, while the son of various deities, remains an independent character. P. K. Agrawala takes "varaghantāya" as it appears in 20.4.2 to mean "to him of choice bells" (1967:109). The name may also be an allusion to a name of Siva. Lohitagätra means red-limbed (Monier- Williams 1999:909). Skanda-Kārttikeya is often described as red in colour. Sālakatankata is the name of a Rāksasa in the Mahābhārata (Monier-Williams 1999:1067). The use of the name here suggests that Skanda- Kārttikeya can be viewed as a Rāksasa or associated with Rāksasas. Svāmin, or Lord, is a name for Skanda- Kärttikeya used by the Yaudheyas and, as we have seen, it is also used in the Mahābharata.

69“pramodo nāma gandharvah pradoso paridhāvati muñca śailamayātpāpānmuñcamuñca pramuñca ca". The intent of "sailamayat" ('from things made of stone') is obscure to me; I have left it out of my translation.

56

reminiscent of the Skanda-Kärttikeya who appeared in the early section of the Aranyakaparvan account and

in Ãyurvedic texts. The Skandayaga account helps us to understand that the depiction of Skanda-Kārttikeya

and his cult from the first half of the Aranyakaparvan account are not Epic inventions, but based in the

actual cultic life of this deity.70

2.3.9 The Inauspicious Skanda-Kārttikeya: Final points on Grahas in the Āranyakaparvan

Returning to chapter 219 of the Aranyakaparvan, the text gives a final list of Grahas, which affect

those over the age of sixteen. The conditions enumerated all deal with various aspects of madness, and all of

the Grahas are named after one of the main supernatural groups of beings: devagraha (3.219.46), pitrgraha

(3.219.47), siddhagraha (3.219.48), rāksasagraha (3.219.49), gāndharvagraha (3.219.50), yaksagraha

(3.219.51) and piśäcagraha (3.219.52). The chapter ends by telling us that Grahas are of three kinds:

playful, gluttonous and lustful, and they afflict people until age seventy, after which fever becomes the

equal of a Graha for people. We are then given the following moral observation by way of conclusion to the

chapter: "The Grahas always avoid the faithful who are pious, always alert, undefiled and whose senses are

not confused. Thus, the explanation of the Grahas of humans has been announced to you. Grahas do not

70Based on this text, P. K. Agrawala concludes that "He [Dhūrta] is revealed in this text as the god of cunning and roguery" (1967:18), and "A careful consideration of the text and the ceremony described leads to the conclusion that Skanda is here a god of rogues" (1967:20). Modak comes to a similar conclusion that Dhurta is a god of roguery and a patron of thieves (1993:296). I have little doubt that, from a Brahminical perspective, this figure comes from an undesirable section of society. The very name of Dhürta, some of his epithets, and the beings associated with him all point to a dangerous and potentially inauspicious being. This does not mean, however, that his worshipers are thieves and rogues or that the rite is designed for successful roguery. As I have discussed above, the point of this ritual is to gain worldly wealth and protection from magical forces. The text does not suggest that these goods and protection are associated with theft or dishonesty. The point of the rite seems to be that Dhurta will supply these things in return for worship. I think a more likely relationship between deity and worshiper than that suggested by Agrawala and Modak can be found in the model of Rudra. Rudra is described as dangerous and worshiped to avoid his anger and to gain his support. There is no suggestion that we also need to view Rudra's worshipers as dangerous or like the deity. Dhūrta may be a dangerous and inauspicious character, but with that danger comes a power that his devotees can use. His devotees need not be thieves or inauspicious themselves; the point is that Dhūrta's power comes from such sources. We need not assume with Agrawala and Modak that Dhūrta's followers are like the god.

57

take hold of men who are worshipers of god Maheśvara [Śiva]" (3.219.57-58).71 These final sentences

present a significant shift in the manner in which these Grahas have been presented, and how we are told

they can be controlled. The morality of the victim was never a question in the other sections of this text.

Grahas attacked the children of those who did not propitiate them and protected those that did. This

understanding had nothing to do with the morality of the protected or the victim. In the early sections of this

story Skanda-Karttikeya attacked randomly without a concern for the morality of his victims. This aspect of

morality is a new one, but it is also found in Ayurvedic texts and plays a significant role in the beings

associated with Skanda-Kārttikeya. Skanda-Kārttikeya and his troop become representatives of divine

justice punishing those who are immoral or have offended the gods. Their inauspicious characterization is

mitigated because they are presented as representatives of divine order. As we shall see, Skanda-

Kārttikeya's recharacterization as auspicious also plays into this new theme. He must be viewed as moral

and Brahminical if he is to punish those who are not; he requires a clear position within the divine

hierarchy. He cannot be the figure in chapters 214-216 who seemed to strike fear into all creatures and

made random attacks on mountains. People who are good and Brahminical now seem to be automatically

protected by this deity, which implies a much more disciplined and ordered Skanda-Kärttikeya than that

found in the opening chapters of his birth story.72

2.3.10 Conclusion to the Inauspicious Characterization of Skanda-Kārttikeya

As we shall see, the introduction of the story works hard to claim that Skanda-Kārttikeya is

Brahminic and linked to Vedic sources and is born to become Mahäsena, a military deity. We will also see

that the text continually mitigates against Skanda-Kärttikeya's inauspicious and unorthodox origins. The

text, in effect, works to minimalize the inauspicious characterization of this deity that it presents in these

""aprakīrnendriyam dāntam śucim nityamatandritam āstikam śraddadhānam ca varjayanti sadā grahāh [.57] ityesa te grahoddeśo mānusānām prakīrtitah na sprsanti grahā bhaktānnarāndevam maheśvaram (I am reading the variant 'deve mahesvare') [.58]".

72I will comment on the final lines of the above quotation relating to the worship of Siva in my section on Siva as the father of Skanda-Kārttikeya.

58

chapters. It is clearly hostile to the deity as inauspicious and dangerous, but has also included this element

of the deity in its account of him. Indeed, it is not inappropriate here to question why this version of his

birth includes this inauspicious material at all. An answer, and, I think, the only meaningful answer we can

arrive at, is that this inauspicious and dangerous understanding of the god was popularly held by some

group or groups in South Asia at this time. The evidence from panels, medical texts and the Skandayāga all

point to a propitiatory based cult associated with protection and inauspicious beings -- much like the figure

of Skanda-Karttikeya found in the early chapters of the Aranyakaparvan. What I think the above

demonstrates is that Skanda-Kärttikeya's roots are found within the related Graha/Kumāra/Mātr cults, and

that he was likely one of these beings before his transformation into Mahasena. Hence, if we are to speak of

Skanda-Kärttikeya's popularity at the early stage of his cult we must, I think, view it from within a context

of propitiation of inauspicious beings for the protection of children and not to a military deity which

emerges later.

We leave this initial inauspicious description of the deity with a few details that deserve

emphasizing before we move on. His cult, as depicted thus far, is based on a mixture of fear and protection.

He is terrifying in appearance, sounds and actions, and his initial actions seem deliberately designed to

inspire fear. If approached, however, and sought out for refuge he becomes a protective and even playful

deity to those who seek him out. He is only inauspicious and dangerous to those who oppose him and his

followers. He and his cult are also terrestrially focused in these opening chapters. He is born on earth, his

fearful deeds are all performed on earth, and his worship is performed by all manner of terrestrial creatures

and for worldly success. Finally, as I have stressed, this version of the god is not something the Brahminical

writers and redactors of this text wish to support. Their goal is to transform this deity, and how this is

accomplished will be discussed below.

2.4 The Auspicious Skanda-Karttikeya: The frame narratives in the Aranyakaparvan

The second image of Skanda-Karttikeya this text presents is as an auspicious and tamed deity.

When we left the story of his birth, Skanda-Kärttikeya had already been born, destroyed Mount Krauňca

59

and defeated Indra in battle. I will resume the narrative here by going back a bit before the birth to the

frame story, because the frame narrative sets the tone for Skanda-Kärttikeya as a deity who fits Vedic

paradigms. One of the ways this text blunts the impact of its initial presentation of the inauspicious Skanda-

Kärttikeya is through its narrative structure. There are two frame narratives before the account of the birth,

which form certain conclusions relating to the character of Skanda-Kärttikeya. The text does two things in

its preamble to the birth story. The first strategy is to place us within a Vedic context of divine births. The

story is trying to deal with a potential objection to Skanda-Kärttikeya and his cult: they are not Vedic.

Brahminical Hinduism tends to legitimate itself by claiming Vedic roots for its beliefs and practices.

Skanda-Karttikeya is not in the Vedas, but a birth story potentially works around the problem of his obvious

absence from these texts. The lack of a Vedic past for this deity seems to have been a paramount concern

for the writers of this text, and much of the introduction is designed to link Skanda-Karttikeya with Vedic

themes and precedents to transform him into an orthodox figure. The text uses the example of Agni and his

births to legitimate Skanda-Kärttikeya's divine terrestrial birth. Agni becomes a Vedic precedent for the

account of Skanda-Kārttikeya.

The second strategy the text employs is to give away the ending of the story in the introduction of

the story. The account of Skanda-Karttikeya's birth is introduced to us in such a way that his eventual role

as a husband and divine general who restores order is established before his appearance in the text. In fact,

the point of his birth is presented in terms of his eventual role as the divine general and husband. The text's

structure invites us to look for the emergence of this benevolent and responsible character and removes

credibility from his initial dangerous appearance because we know he will be transformed. The introduction

works to legitimate the final auspicious characterization of the deity at the expense of his inauspicious side.

Thus, the second aspect of the introduction is part of the transformation theme of the narrative and will be

discussed later in this chapter (section 2.5). I will now turn to Agni as the father of Skanda-Kārttikeya.

2.4.1 The Auspicious Skanda-Kārttikeya: The Vedic missing Agni frame narrative

We are first introduced to the connection between Agni and Skanda-Kārttikeya by a question

60

Yudhisthira asks Mārkandeya:

Tell me why Agni went to the forest, and why formerly the great sage Angiras, when Agni had disappeared and having become Agni [himself], carried the oblation? Though Agni is one, a multitude [of fires] are seen at rites. O holy one, all this I want to know: how was Kumāra born, and how did he become the son of Agni, and how was he born from Rudra by Ganga73 and the Krttikās? This I wish to hear from you, O gladdener of the Bhārgavas .... (3.207.2-5)74

Most of Märkandeya's answer takes the form of a lengthy genealogy. At first glance there is a great deal

that seems odd about Yudhisthira's question and Märkandeya's response. It is not immediately clear how

the elements of the question, Agni's absence, Angiras, Agni's forms and Kumära's birth fit together, and

how Markandeya's response answers the questions. I will argue, however, that the themes of Agni's

multiple forms and absence are Vedic themes, all of which relate to divine birth. The text attempts through

this opening narrative to situate Skanda-Karttikeya's divine birth within established Vedic precedent, and to

place this new god at the end of a long chain of divine births related to Agni.

Agni maintains a very strong link with the terrestrial world as the sacrificial fire, and thus he is the

link between Vedic ritualists and the world of the divine (MacDonell 1963:88). Like most of the Vedic

pantheon, Agni is understood to have been born; what separates him from the rest of the Vedic pantheon is

that he is continually reborn in the terrestrial world (MacDonell 1963:91). Each time a sacred fire is kindled

Agni is, in effect, born, and each time a sacred fire goes out he dies. A number of Vedic sacrifices also used

three fires, and it is not unusual to have references to Agnis at a rite. His physical presence on earth, his

apparent continual re-birth and death and his multiple forms presented Vedic ritualists with certain

problems relating to the exact nature of his divine and immortal status, as well as opening a way for

73 Ganga, or the Ganges, is depicted as a mother of Skanda-Karttikeya in most stories of his birth, but not in this one from the Aranyakaparvan. The mention of it here in the opening of the story suggests that this introduction was not part of the original narrative. It may have been added at a later period when the story of Skanda-Kärttikeya's birth had become more standardized.

74"yudhisthira uvāca kathamagnirvanam yātah katham cāpyangirāh purā naste'gnau havyamavahadagnirbhūtvā mahānrsih [.2] agniryadā tveka eva bahutvam cāsya karmasu drśyate bhagavansarvametadicchāmi veditum [.3] kumāraśca yathotpanno yathā cāgneh suto'bhavat yathā rudrācca sambhūto gangāyām krttikāsu ca [.4] etadicchāmyaham tvattah śrotum bhārgavanandana ... [.5]".

61

speculation regarding the nature of divine births on earth.

The thread that runs through Yudhisthira's question, from a Vedic or Brahminical perspective, is a

concern for divine births on earth. The story of Angiras replacing Agni, which follows Yudhisthira's

question, can be connected to Vedic speculation on the fire god's births. In the Epic story, Agni angrily

retires to the forest to carry out asceticism, leaving the world without the sacrificial fire (3.207.6-20). Such a

narrative echoes a common Vedic theme of the missing or hidden Agni, which in turn was often a vehicle

for Vedic writers to speculate on the nature of the deity's forms and birth. An example is Rgveda hymn

10.51, which presents us with a time when Agni was mortal. To escape his eventual death if kindled as the

sacrificial fire he hides himself from the gods, only to be discovered by them. The Vedic gods lure him back

with a promise of immortality. The hymn states:

(Ein Gott:) "Groß, stark war jene Haut, in die eingehüllt du ins Wasser eingegangen bist. Ein Gott erschaute vielfach alle deine Leiber, o Agni Jatavedas."[1] (Agni:) "Wer hat mich gesehen, welcher ist es unter den Göttern, der meine Leiber vielfach erschaut hat ?... " [2] (Agni:) "Aus Furcht vor dem Hotr-Amt bin ich gegangen, Varuna, damit mich nicht die Götter daran einspannen können .... " [4] (Agni:) "Agni's frühere Brüder haben dieses Geschäft hin und her besorgt, wie ein Wagenroß seinen Weg. Aus Furcht davor bin ich weit fortgegangen, Varuna. Wie der Gaurabüffel vor der Sehne des Jägers bin ich geflüchtet." [6] (Die Götter:) Wir bereiten dir ein Leben, das alterlos ist, Agni, auf daß du eingespannt nicht zu Schaden kommest, o Jātavedas .... " [7] (10.51.1-2,4,6- 7 translated by Geldner 1951 III:212-213)

This quotation makes it clear that Agni is understood as having various bodies and to have hidden them all.

The imagery used to discuss the absence of Agni is biological birth. The opening verse tells us that Agni has

returned to the waters, his place of birth. The "membrane" (ulba) in verse one is the covering of the embryo

of the womb (Doniger 1983:109). Agni has returned to his womb and refuses to be born, or kindled, to

carry out his role in the sacrifice. The imagery of biological birth appears together with mention of the

various forms of Agni and the absence of the deity.75

75I would also note that it is Agni's retirement to the forest in the birth story of Skanda-Kārttikeya that leads to that deity's birth (3.213.35-45). This is another version of the missing Agni theme. The themes of the missing Agni and his birth are also present in Rgveda 5.2. This hymn is enigmatic, but it clearly reflects the themes of the birth of fire through kindling and the disappearance of fire:

62

The theme of the missing Agni and his many forms is not unique to Rgvedic hymns, but is found

throughout the Vedic tradition. A legend in the Satapatha Brähmana attempts to explain why three sticks

are used to enclose a ritual fire when it is being kindled. The legend runs that when the gods asked Agni to

be their Hotr76 he refuses telling them: "Already you have chosen three before, and they have passed away.

Restore them to me: then I shall be equal to this, that I should be your Hotr and that I should carry your

oblation!" (ŚB 1.3.3.13 translated by Eggeling 1882-1900 I:87). As with Yudhisthira's question and the

Die jugendliche Mutter trägt heimlich den verhüllten Knaben und gibt (ihn) nicht dem Vater. Die Leute sehen sein verändertes Antlitz nicht vor sich, das bei dem Rosselenker aufbewahrt ist. [1] Was für einen Knaben trägst du da, o Jugendliche, als Stiefmutter (?)? Die Fürstgemahlin hat ihn erzeugt, denn viele Herbste ist die Leibesfrucht gewachsen. Ich sah den Neugeborenen, als die Mutter (ihn) zur Welt brachte. [2] Ich sah den Goldzahnigen von reinem Aussehen fern von seiner Flur, wie er die Waffen erprobte. Ich gebe ihm das befreiende Lebenselixier. Was können mir die anhaben, die keinen Indra, keine Lieder kenner? [3] Ich sah ihnvon seiner Flur sich entfernen, (ich fand) seine schöne Herde nicht mehr stattlich aussehend. Nicht haben sie ihn festgehalten, denn jetzt ist er geboren. Die Ergrauten wurden wieder jung. [4] .... Denn grollend warst du von mir fortgegangen, so hat mir der Gesetzeshütter unter den Göttern gesagt. Denn der wissende Indra hat dich entdeckt; von ihm zurechtgewiesen bin ich gekommen, Agni. [8] (5.2.1-4, 8 translated by Geldner 1951 II:3-4)

Doniger speculates, with the help of later commentaries on this text, that the story deals with the absence of Agni which, in turn, leads to reflections on how he is produced and fades (1981:101). The young mother and father mentioned in verse one are the kindling sticks used to create Agni. These kindling sticks are often presented as a couple in Vedic texts, who produce their child through a sexual metaphor of rubbing the sticks together, as this text suggests:

He takes the bottom piece of wood, with the text, "Thou art the birth-place of Agni;" for it is thereon that Agni is produced: hence he says, "Thou art the birth-place of Agni" .... Thereon he lays the lower churning-stick (with the top to the north), with, "Thou art Urvashi!" He then touches the (ghee in the) ghee-pan with the upper churning-stick, with, "Thou art Ayu," he puts it down (on the lower arani) with, "Thou art Pururavas." For Urvashi was a nymph, and Purūravas was her husband; and the (child) which sprung from that union was Ayu: in like manner does he now produce the sacrifice from that union. Thereupon he says (to the Hotr), "Recite to Agni, as he is churned". (ŚB 3.4.1.20-22 translated by Eggeling 1882-1900 II:90-91)

Ghee or butter is frequently described as seed or semen in texts discussing the sacrifice (Doniger 1980:24), and this additional information should make the sexual nature of this passage clear. In the case of Rgveda hymn 5.2, Agni is hidden or held back by the lower stick or mother, and verse eight also makes the point that Agni has abandoned this sacrificer or is missing. Again, the missing Agni theme is linked to his earthly births which are conceived of in a sexual or biological manner.

76 A Hotr is a priest who at the sacrifice invokes the gods and recites the Rgveda. Agni is often described as a priest in Brāhmana texts.

63

quotations above, Agni has once again removed his services, and it is necessary that he be reborn. The text

states that three Agnis had already been used by the gods and died out. The gods return these three former

Agnis in the form of these three enclosing sticks. Each of these Agnis is given a separate title: "'To the Lord

of the Earth -- svaha!' 'To the Lord of the World -- svaha!' 'To the Lord of Beings -- svaha!' These, indeed,

are the names of those Agnis" (SB 1.3.3.17 translated by Eggeling 1882-1900 I:89). Each is a form of Agni.

This Satapatha Brahmana story provides another example of the story of the missing Agni that also

mentions his various forms and his kindling or birth.77

In these hymns and passages Vedic ritualists appear to have speculated on many of Yudhistira's

questions concerning the absence of Agni and his ability to be both one and many. The birth of Agni, I

would argue, as an example of a birth of a god on earth, provided a paradigm for the Epic poet into which

he could fit the birth of Skanda-Karttikeya. By utilizing this Vedic paradigm, moreover, the Epic poet could

describe the birth of Skanda-Kärttikeya as something fundamentally in harmony with Vedic notions of a

deity. If we keep in mind that the Vedic speculation on the disappearance of Agni was coupled with a

reflection on his birth and his many forms, we are in a better position to understand the coherence behind

Yudhistira's questions.

In the Epic version of this missing Agni theme Märkandeya narrates that when Agni left for the

forest, he was replaced by the Brahminical sage Angiras, who is described as becoming the god of Fire and

as surpassing Agni in his ability to illuminate the world (3.207.7-8). With the appearance of this new and

better fire, Agni returns and asks Angiras to allow him to remain as a secondary fire. Angiras refuses and

wants Agni to resume his former position and to make him Agni's first-born son (3.207.14-16). Agni agrees

to this, and the gods also seem to accept it (3.207.17-19). Following this story Mārkandeya narrates a

77 Charles Malamoud has also discussed the topic of Agni's bodies or forms. He points out that in the Satapatha Brähmana, Agni's material body is understood as the sum total of the various fires ignited on the sacrificial ground, but each of these bodies is a person in its own right. Some are even depicted as arguing with each other (Malamoud 1996: 220-221). While these fires often appear as independent deities, they are all ultimately Agni.

64

lengthy lineage that is attributed in part to Agni and in part to Angiras (3.208-212). The children are,

however, all really the offspring of Agni because this text makes Angiras Agni's first born son (3.207.7-19).

The point of the story is that this priestly lineage originates with the deity Agni.78 Chapters 208 to 212 then

list the lineage of this priestly family calling some of the offspring fires and some gods and goddesses.

Many bear the names of specific fires employed at Vedic rites. Hence, the narrative follows the Vedic

formula by linking a missing Agni theme with accounts of divine births and forms of Agni.

This introduction establishes a clear narrative link between the Vedic world and Skanda-

Karttikeya. This deity is the last member of the lineage, and as such he is connected to both the Vedic world

through Agni and the Brahminical world through Angiras. My argument is that the text is preoccupied with

Skanda-Kärttikeya's lack of a Vedic past and works hard in this introduction to provide as many links

between Skanda-Kärttikeya and Vedic themes and the Brahminical world as possible. Assuming that the

writers of this text are trying to draw a non-Brahminic deity into the orthodox Brahminic fold, they are

availing themselves here of themes well established in the Vedic corpus. Thus after narrating this story and

lineage, Mārkandeya states:

The manifold geneology of fires has been narrated by me, O faultless one. Now hear about the birth of wise Kärttikeya. I will poclaim [the birth] of the one of unbounded energy, the wonderful son of the wonderful one, who was born by the the seven wives of the sages and who is Brahminic and an increaser of glory. (3.213.1-2)79

Such comments relating to a "Brahminic" son of Agni seem well placed when we consider the introduction

to the birth story.

78Such priestly lineages often involve Agni as the following example from the Satapatha Brähmana, where the officiating priest calls on Agni as the ancestral Hotr priest, demonstrates: "This is the reason why he calls on (Agni as) the ancestral (Hotr). He calls from the remote end (of the sacrificer's ancestral line) downwards .... Having named (him as) the ancestral one, he says, '(thou wert) kindled by the gods, kindled by Manu;"" (1.4.2.3-5 translated by Eggeling 1882-1900 I:115-116). The Satapatha Brāhmana and the Brhadāranyaka Upanisad (6.5.1-4) also end in priestly lineage lists that end in divine origins. These Brahminical lineages are largely about religious legitimation.

79"mārkandeya uvāca agnīnām vividho vamśah kīrtitaste mayānagha śrnu janma tu kauravya kārttikeyasya dhīmatah [.1] adbhutasyādbhutam putram pravaksyāmyamitaujasam jātam saptarsibhāryābhirbrahmanyam kīrtivardhanam [.2]".

65

2.4.2 The Auspicious Skanda-Kärttikeya: The fatherhood of Agni

The fact that Skanda-Kärttikeya is presented as Agni's son also tells us something about Skanda-

Kärttikeya's character in this text. The Vedic world did not hesitate to identify the father with the son: "the

father is the same as the son, and the son is the same as the father" (SB 12.4.3.1, quoted in Gonda 1957:10).

The early parts of this version of the story expend a great deal of effort to establish that Agni is Skanda-

Kärttikeya's father; the text also goes to some lengths to imply that Skanda-Kärttikeya shares in some of

Agni's essential physical qualities. Primarily, the text concentrates on Skanda-Kārttikeya's shine, his visible

fiery energy. A number of scholars have been mislead by the description of Skanda-Kärttikeya as bright and

shining into assuming that he is also a son of, or derived from, the sun (Chatterjee 1970:22; Sinha 1979:33;

Thakur 1981:8; Rana 1995:27, 39).80 Certainly, his shine is often compared to that of the sun:

he shines like the sun rising in a very great red cloud. (3.214.19) ... he is as equal in radiance as the newly risen sun. (3.214.23) That inscrutable being was like the sun rising on a mountain top. (3.214.26) The son of Pāvaka [Agni] was clothed in a pair of dustless red clothes, and his blazing and glorious body shone like the sun covered by two red clouds. (3.218.31) The deities looked at Mahäsena who was annointed as the rising sunhaving destroyed the darkness there. (3.218.39) He shone on the golden mountain blazing accompanied by Sri. That mountain, which has a beautiful forest, shone due to that hero like the way Mandara, which has lovely caves, shines due to the rays of the sun. (3.220.21-22) ... Mahasena appeared blazing from anger like the sun. (3.221.62) ... the son of the Krttikas shone like the sun which has scattered sun beams. (3.221.71)81

The point of this repeated simile need not, however, be read as establishing Skanda-Kārttikeya's

relationship with the sun, but as establishing his relationship with Agni; that he carries some of Agni's

80There may be a solar or astrological connection for Skanda-Kärttikeya, but I will argue in my chapter on Kusana statues and coinage that these connections are supplied through Iranian influence and need not be read into these passages.

81"lohitābhre sumahati bhāti sūrya ivoditah" (3.214.19), "sa bālārkasamadyutih" (3.214.23), "araje vāsasīrakte vasānah pāvakātmajah bhāti dīptavapuh śrīmānraktābhrābhyāmivāmśumān" (3.218.31), "abhişiktam mahāsenamapaśyanta divaukasah vinihatya tamah sūryam yathehābhyuditam tathā" (3.218.39), "sa samvrtah pišācānām ganairdevaganaistathā śuśubhe kāñcane śaile dīpyamānah śriyā vrtah tena vīrena śuśubhe sa śailah śubhakānanah adityenevāmśumatā mandaraścārukandarah" (3.220.21-22), "ājagāma mahāsenah krodhātsūrya iva jvalan" (3.221.62), "śuśubhe krttikāputrah prakīrņāmśurivāmśumān" (3.221.71).

66

essence with him. Skanda-Kärttikeya looks like the sun because Agni also looks like the sun, and, in a

sense, is the sun: "The leading ritualists and philosophers of ancient India were deeply convinced of the

fundamental unity of fire (agni), light and the sun, the source of light and life" (Gonda 1991:216). Agni's

brightness is dwelt upon in the Vedas, where he is also seen as the sun (MacDonell 1963:90, 93). The point

we should take from the likeness of Skanda-Kärttikeya with the sun is that all three (Agni, the sun and

Skanda-Kärttikeya) share in the same essential nature. These references to Skanda-Kārttikeya's brightness

and sun-like splendour are simply reflections of his power and his shared nature with his father, Fire.82

Another link between Agni and Skanda-Kärttikeya is as demon killers and generals. In some Vedic

texts Agni is renown as a killer of rāksasas (demons) (RV 10.162.1; Gonda 1959:91; AV 8.2.28; ŚB

2.4.2.15). Skanda-Kärttikeya will also take on a role as a demon killer, but he will not exactly re-duplicate

his father's role. Raksasas are primarily terrestrial demons who afflict humans, and Agni is primarily a

terrestrial deity who helps humans with such problems. Skanda-Kärttikeya will become a great killer of

asuras and danavas who are primarily non-terrestrial demons who afflict the gods. The implications of

Skanda-Kärttikeya's eventual non-terrestrial focus will be discussed in more detail below. Agni is also

occasionally associated with generalship of an army, under the epithet Agni Anikavat ('sharp-pointed'), and

his worship is advised for the defeat of one's enemies (SB 2.5.3.2, 2.5.4.3). The Satapatha Brahmana also

draws a direct connection between Agni Anikavat and a king's senapati (commander of an army) in its

description of the Rajasuya rite. The text tells us that part of this ritual involves the king going to the

commander of the army with two fires and eight offering cakes for Agni Anikavat: "for Agni is the head

(anīka) of the gods, and the commander is the head of the army: hence for Agni Anikavat" (5.3.1.1

translated by Eggeling 1882-1900 III:58). Agni is also called the senani of the gods in KS 36.8 and MS

1.10.14 (P. K. Agrawala 1967:9). This version of Skanda-Kärttikeya's birth may also be attempting to

82Even the semen of Agni glows in this text. The word used for semen in this section is sukra, which can mean 'bright', 'Agni', and 'semen', among other things. The semen and son of Agni all shine because they all contain his essential, fiery nature.

67

suggest that his eventual role as the senäpati of the gods is inherited from Agni; this would provide another

motivation for making Agni his father (P. K. Agrawala 1967:9, 27). Both Agni and Skanda-Kārttikeya are

also linked because they are also both Kumāras ('young' or 'boys').

A final point about Agni's fatherhood of Skanda-Kärttikeya is that the Fire god is viewed as

auspicious, and his products or children are also auspicious. Making Skanda-Kārttikeya Agni's son almost

demands that the child be auspicious. Agni's parentage of this deity thus also plays into this text's attempt

to present an orthodox and auspicious deity to the reader. The choice of Agni as one of the fathers of

Skanda-Kärttikeya is rooted in concerns relating to legitimating this new deity and his role. Agni provides a

useful link for Skanda-Kärttikeya to orthodox Vedic and Brahminical religion as well as explaining his

terrestrial birth and eventual roles.

2.5 The Auspicious Skanda-Karttikeya: the second frame narrative

The second frame narrative before the actual birth of Skanda-Kārttikeya occurs after Mārkandeya's

description of Agni's and Angiras' lineage. The second frame narrative is a story relating to Indra and a

deity named Devasena. Like the lineage before it, this story influences the manner in which we read and

understand the narrative of Skanda-Kärttikeya's birth. Märkandeya states that in the past the gods and

asuras were fighting each other, and the asuras kept defeating the army of the gods. Indra goes to Mount

Mänasa to think over where he can get a commander for his forces (sainyanäyaka), who will be brave

enough to rescue his army (3.213.2-6). While thinking, he hears the cries of a woman and finds a demon,

Kesin,83 trying to make off with a woman. Indra fights with him and is able to force Kesin to withdraw

83Kesin means "having fine or long hair" (Monier-Williams 1999:310). The name is used in the Rgveda to describe an ascetic (10.136) and is also used to describe Rudra and his attendants (Monier- Williams 1999:310). In the Atharvaveda the term is used in a hymn dedicated to Rudra, but it seems to be applied to a separate character: "We go forward (pūrva) to meet him of dark horses, black, swarthy, killing, fearful, making to fall the chariot of the hairy one (kesin); homage be to him" (XI.2.18 translated by Whitney 1962 II:623). Phyllis Granoff notes AV 8.6.5 where Kesin is described as an asura who attacks foetuses (2002:102). He is also described as being born in a clump of reeds in this passage, which reminds us of the Skanda-Kärttikeya story (Granoff 2002:102). Kesin also appears to be a child-attacking being in some other accounts of him (Granoff 2002:102, 105). The first book of the Mahābhārata recounts a number of lineages accounting for the origins of the gods and raksasas. Kesin is listed as one of Danu's forty sons

68

(3.213.7-14). Indra then asks the woman who she is; she replies that she is a daughter of Prajāpati named

Devasenā (3.213.15-16). This woman's or goddess's name is a Sanskrit compound meaning "army of the

gods". She is the female personification of the divine army. In questioning her Indra discovers that she has

received the following boon from Prajāpati:

I am not strong, but, due to a boon from my father, my husband will be strong armed and powerful. He will be honoured by the gods and demons .... That very heroic and strong one will be considered a conquerer of räksasas, snakes, kimnaras, yaksas, dānavas, gods and of the wicked. He will conquer all beings along with you. He will be my husband, who will be Brahminic and an increaser of fame. (3.213.21, 23-24)84

This husband will be Skanda-Kärttikeya. This introduction tells us exactly how Skanda-Kārttikeya will

appear by the end of the story. He will be the general of the army of the gods, married to Devasena and

Brahminic. There is a sense of divine ordination in this introduction, Skanda-Kārttikeya has a role to play in

the universe, and Devasena has told us what it will be. Any other characterization of the god can be

overlooked after these statements because we now know his divine purpose. Once again, we are being

guided to read this story in a certain direction, and we need to be aware of this process and to question it. As

we have seen, when Skanda-Kärttikeya is first born he is not Brahminic, or immediately the general of the

gods or married to Devasenā.

The story continues with Indra realizing that there is no such husband for Devasena, but then he

witnesses a powerful convergence of astrological and ritual power. He sees the moon entering the sun

during a terrible hour on the New Moon Day. He also sees the gods and asuras fighting and the clouds and

ocean a blood-red colour. He sees Agni carrying the morning oblations into the sun. Finally, he sees the

(1.59.23). Dänu's sons are the Danavas, a class of demons. That this account of Skanda-Kārttikeya's birth begins with the separation of Devasena from a Graha-like being, who may have been related to early understandings of Skanda-Karttikeya, is instructive. The evil and demonic Kesin is to be replaced by the good and martial Skanda-Kärttikeya, and the text is instructing us that it does not regard Skanda-Kārttikeya as like the child snatching Kesin. They are opposites and the text seems to be countering the idea that Skanda-Kärttikeya comes from a child attacking background from the very beginning of this narrative.

84“kanyovāca abalāham mahābālo patistu balavānmama varadānātpiturbhāvī surāsuranamaskrtah [.21] devadānavayaksānām kimnaroragaraksasām jetā sa drsto dustānām mahāvīryo mahābalah [.23] yastu sarvāni bhūtāni tvayā saha vijesyati sa hi me bhavitā bhartā brahmanyah kīrtivardhanah [.24]".

69

twenty-four moon-phase days gathered about the sun and the moon conjoint with the sun. Indra recognizes

that this conjunction of powers is miraculous, and that if the moon or Agni were to have a son, that child

would make a suitable husband for Devasenā (3.213.25-33). He takes Devasenā to Brahmā and asks that

such a husband be created for her. Brahma answers: "O Destroyer of Dänavas, it [will] occur in the manner

thought by you. Thus, a feotus will be made which will be mighty and of great strength. He will be the

commander of the army, along with you, O One of the Hundred Sacrifices, and that heroic one will be the

husband of this goddess" (3.213.35-36).85 Skanda-Kārttikeya's role and importance have already been

predicted twice before his story actually begins. This introduction to the birth story works to establish that

the role of this god is as Mahasena, the general of the army of the gods. It even goes as far as to establish

that his reason for being born is to be the husband to the personification of the divine army. He is also

understood to be Brahminic or brahmanya in this introduction. Mārkandeya describes him as such (3.213.2)

and Devasena also uses the word as an adjective for him in the above quotation (3.213.24).

We leave the various layers of introductory material to the birth of Skanda-Kärttikeya with an

understanding of what it is about his character that makes Brahminical redactors nervous, and how they

hope to steer the reader's understanding of his character in this narrative. The introduction attempts to

present Skanda-Kärttikeya as a deity with strong links to Vedic themes and deities. We are also repeatedly

told that he is Brahminic and the asura killing general of the army of the gods. In going to these lengths in

the introduction of the story, however, the Brahminical redactors signal to us that these characteristics are

points of concern for them. Because they have to argue for this characterization of the deity in the

introduction, we might well imagine that there are other characterizations of him which deviate from that

given here. We have already seen what the redactors are concerned about in the opening chapters to the

birth story itself, but the image of the god in those opening chapters is already softened because this

85“brahmovāca yathaitaccintitam kāryam tvayā dānavasūdana tathā sa bhavitā garbho balavānuruvikramah [.35] sa bhavisyati senānīstvayā saha śatakrato asyā devyāh patiścaiva sa bhavisyati vīryavān [.36]".

70

introduction denies the inauspicious characterization of Skanda-Kärttikeya.

2.6 The Auspicious Skanda-Kärttikeya: The Brahminical ritual taming of the deity

Beyond the story's introduction, the conversion of Skanda-Kārttikeya into a Brahminical deity

starts at the beginning of chapter 215. As we might recall, at this juncture of the birth narrative Skanda-

Kärttikeya has already been born, destroyed Mount Krauñca, and the earth and the other mountains have

taken refuge in Skanda-Kärttikeya. The first suggestion of how the text will go about taming this deity

comes when we are told that the Vedic sages know of these events and take action: "The seers, who were

promoters of the welfare of the world and anxious, having seen the sudden appearance of various very

terrifying events boding calamity, performed expiatory rites" (3.215.1).86 The text continues by describing

the confusion of the people concerning the source of Skanda-Kärttikeya, who is regarded by them as a force

of disaster:

Those people who lived in that forest, which is [called] Caitraratha, said: "This great evil has been brought here by Agni, who united with the six wives of the seven seers". Others, who had seen that goddess [Svāhā] leaving with her [Garudī's] form, said to Garudi: "This evil has been brought here by you". No person knew that the deed had been done by Svāhā. (3.215.2-3)87

The result of these rumours has Garudi telling Skanda-Kärttikeya that she is his mother, the seers divorcing

their wives except for Arundhati, and Svaha protesting that the new deity is her son. The only person who

seems to know the truth is the seer Visvamitra, who has followed Agni since the oblation of the seven seers

and knows the truth of the situation. He is the first to seek refuge with Skanda-Karttikeya and also performs

the samskäras pertaining to childhood for him. He composes a divine hymn of praise for him. He sings of

"the majesty of his six faces and the ritual of the cock and the propitiation of the goddess, who is a spear,

and also the attendants of the god. Viśvämitra also did those rites for the welfare of the world. Due to all

lokānām lokabhāvanāh". 86“mārkandeya uvāca rsayastu mahāghorāndrstvotpātānprthagvidhān akurvañśāntimudvignā

87"nivasanti vane ye tu tasmimścaitrarathe janāh te'bruvannesa no-narthah pāvakenāhrto mahān samgamya sadbhih patnībhih saptarsīnāmiti sma ha [.2] apare garudīmāhustvayānartho-yamāhrtah yairdrstā sā tadā devī tasyā rūpena gacchatī na tu tatsvāhayā karma krtam jānāti vai janah [.3]".

71

this the sage Viśvāmitra became beloved to Kumāra" (3.215.10-11).88 Through this account of Viśvāmitra

and the other seers, we are reminded of the power of the Brahmins and sages to bring order to the world.

They are depicted as the 'prosperers of the world' and capable of dealing with this situation through their

knowledge of ritual. We are reminded that there is an order to the universe which Skanda-Kārttikeya's

coming has disrupted, but this disruption is temporary, and the universe can be ordered again through the

power of these seers/Brahmins. The text is subtly telling us how it will approach controlling Skanda-

Kärttikeya. The means of control, so far in the story, are worship and family roles. Worship is at the centre

of appeasing Skanda-Kärttikeya and one of the means of transforming his character.

There may also be something else in the choice of Viśvämitra as the sage that first engages

Skanda-Kärttikeya and draws him into rituals. There are numerous legends involving Viśvāmitra, but one of

the most important deals with his adoption of Brahminical caste status. The legend has it that Viśvāmitra

begins as a Ksatriya king who fights with the sage Vasistha and loses to him. Once defeated Viśvāmitra

realizes that real power lies in the renunciant life of a Brahmin sage and gives up his kingdom for a life of

penance. Eventually he attains Brahminhood through the power of his tapas. The Mahābhārata makes a

brief mention of this transformation in a narration of famous tirthas or bathing places:

... and the Kausiki River, which has plentiful fruit and roots, bull of the Bharatas [Yudhisthira is being addressed by Dhaumya], where Viśvāmitra the ascetic became a brahmin .... They narrate that in the Pañcala country there is the Lotus Cistern where Viśvāmitra Kausika sacrificed with Śakra. And upon seeing Viśvāmitra's superhuman puissance the blessed Lord Jāmadagnya [Rāma] sang there this chronicle verse: "In Kanyakubja, Kausika drank Soma with Indra and there he withdrew from the baronage, saying 'I am a brahmin!"" (3.85.9, 11-12 translated by van Buitenen 1975 II:399-400)

Viśvāmitra is, then, a reminder of the power of the Brahmin caste and the ability of certain forces to change

a being's character. Viśvāmitra is an ideal Brahminical representative because he embodies such a change

88“sadvaktrasya tu māhātmyam kukkutasya ca sādhanam śaktyā devyāh sādhanam ca tathā pārisadāmapi [.10] viśvāmitraścakāraitatkarma lokahitāya vai tasmādrsih kumārasya viśvāmitro- bhavatpriyah [.11]".

72

and reminds the reader of where true power resides in this text.89

This section with Viśvämitra also emphasizes that not any form of worship will do. Brahminically

supervised worship is presented as the most efficacious. It is unlikely that these rituals performed by

Viśvāmitra are Vedic (see note 45 on the ritual use of a cock), but his performance of these rituals gives

them a Brahminical stamp of approval. Part of Skanda-Kärttikeya's inauspiciousness, from a Brahminical

perspective, lies in his non-orthodox roots, and his general characterization that places him as closer to

raksasas than a Vedic god. This text challenges this inauspiciousness by making him Agni's son and by

connecting him with a specific Brahminical sage who is able to use ritual to calm him. Viśvāmitra becomes

his favourite; we are inadvertently being told of the Brahminization of Skanda-Kärttikeya and his rituals.

Immediately following this account of Visvamitra is the account of the gods pleading with Indra to destroy

Skanda-Kärttikeya, something, as we have seen, they fail to do. The text subtly suggests that the real

religious power in this story lies in the Brahmins and not with the Vedic gods. It is through Brahminical

structures that this new power will be controlled and converted.

2.7 The Auspicious Skanda-Kārttikeya: The senapati appears

There is a very clear demarcation between chapter 218 and what comes before it. We have seen in

earlier chapters a balance between an inauspicious depiction of the deity and forces which attempt to soften

that image. With this chapter the inauspicious Skanda-Kärttikeya is all but gone, and the remainder of the

story describes the auspicious figure of Skanda-Kärttikeya as Mahäsena. If I were to speculate on the

various layers of this story, I would suggest that chapters 206-212, 214-216 represent one unit, which may

be the earliest layer of the text, chapter 217 a second unit and chapters 213, 218-222 a final and likely latest

89The Mahabharata as a whole is anxious to elevate the status of Brahmins particularly in relation to Ksatriyas, the warrior and kingly caste. The text insists that royal power be mediated through Brahminical power. Part of the concern for Brahminical redactors in this account of Skanda-Kārttikeya is that his story is basically a glorification of a martial figure. Skanda-Kärttikeya, by the end of the story, is the saviour of the gods and order because he is a great warrior. Hence, this story, which was likely written and redacted by Brahmins, does not allow for complete Ksatriya domination. Through Viśvamitra and the seers the text reminds us of who has the means to control and worship this deity.

73

unit. This is, I stress, speculation -- proving the relative lateness of parts of the same story or identifying

different hands at work in a final version is an inherently speculative exercise. I make the above suggestion,

however, on the basis of the following. In chapters 206-212, 214-216 Agni is depicted as the father of

Skanda-Kärttikeya, and Vedic precedent seems to be a key concern for this section, all of which suggests an

earlier date for it. It also presents a fairly consistent image of Skanda-Kārttikeya as a dangerous deity who is

associated with a cult of propitiation and this is, I think, the earliest version of his cult. Chapter 217 seems

to disrupt the flow of the narrative and is a jumbled section. It also deals with material that is raised again in

chapter 219. This repetition and the problems with the chapter itself flag it as an addition to the narrative.

Chapters 213, 218-222 not only change the character of Skanda-Kärttikeya, but also redefine who his father

is. Siva is Skanda-Kārttikeya's father in this section, and he is also represented as the force of religious

power in this narrative. While chapters 206-212, 214-216 look to Vedic deities, chapters 218-222 look

ahead to the classical Hindu pantheon.90

Chapter 218 begins the softening of Skanda-Kärttikeya's character with a description of him after

his battle with Indra:

Sri, herself, who had the bodily form of a lotus, honoured Skanda, who was seated and who had golden armour and garland, a golden diadem, shining golden eyes, who was covered in red clothes, sharp-toothed, beautiful, possessed all the auspicious marks, heroic and very dear to the three worlds, who was young, a granter of wishes and wore bright earrings. (3.218.1-3)91

Based on the actions and description of Skanda-Karttikeya from chapters 214-216 one cannot help but

wonder who this charming, well-dressed and wealthy figure is, and what has become of the horrific six-

9Most scholars who attempt to chronologically order the three versions of Skanda-Kārttikeya's birth recorded in the Mahabharata agree that the one found in the Aranyakaparvan is the oldest (P. K. Agrawala 1967:24; Chatterjee 1970:8; Sinha 1979:22). I am inclined to agree that the sections which develop the fatherhood of Agni represent the earliest account of the birth of Skanda-Kārttikeya, but I also think that sections which describe the fatherhood of Siva represent a later tradition, and that all of the Aranyakaparvan cannot be spoken of as clearly before these other accounts of the deity.

91"mārkandeya uvāca upavistam tatah skandam hiranyakavacasrajam hiranyacūdamukutam hiranyāksam mahāprabham [.1] lohitāmbarasamvītam tīksnadamstram manoramam sarvalakşanasampannam trailokyasyāpi supriyam [.2] tatastam varadam šūram yuvānam mrstkundalam abhajatpadmarūpā šrīh svayameva śarīrinī [.3]".

74

headed figure who defeated Indra? Certainly, the text is dramatically changing its depiction of him. He now

wears a golden coat of mail (hiranyakavaca).2 This gold and ornamentation are not mentioned until this

point of the story, but the coat of mail relates to statues of the deity. As I will argue in my section on statues,

many of the early statues depicting Skanda-Kärttikeya show him in armour, but this armour is Scythian or

Kusāna in design. The statues of Skanda-Kärttikeya dressed in armour are all from the far north-west of

India and depict the god as a Scythian warrior. I will have more to say on this topic in my section on

coinage and statuary. What is important to note here is that this image of a noble and charming professional

soldier is different from the image of the deity presented earlier in this text.

Skanda-Kärttikeya is also attended by Srī in the above quotation. Sri is a goddess of wealth and

glory in the Mahabharata, and she is also a bestower of royal power or kingship (Hiltebeitel 1976:149,

153). In this scene she has chosen the next king, but, as we shall see, Skanda-Kārttikeya turns down the job.

There is, however, a more subtle point being made in the opening of this chapter. Sri and the description of

Skanda-Kārttikeya place him into a royal and elite context. There was no suggestion of this in chapters 214-

216, in fact the very opposite impression was delivered in those chapters. This royal theme will continue in

the latter half of this story.

The seers, who are described as "great-spirited Brahmins" (3.218.5) (mahātmāno brāhmaņās),

address Skanda-Kärttikeya and ask him to become Indra. Skanda-Karttikeya is not sure what Indra does and

asks the seers: "Great ascetics, what does the Indra of the worlds do? Tell me, how does the lord of all the

gods always protect the group of the gods" (3.218.8).93 The seers explain the job, but acknowledge at the

end that the position goes to the best warrior (3.218.12). Indra even requests that Skanda-Kārttikeya take

the role of Indra, or king of the gods. Skanda-Karttikeya refuses the role and tells Indra: "I am your servent,

92The association of gold with Skanda-Kärttikeya is a significant element of the version of this birth story as presented in the Anusasanaparvan of the Epic. I will discuss the exact relationship between gold and this deity in my analysis of that section.

93"skanda uvāca kimindrah sarvalokānām karotīha tapodhanāh katham devaganāmścaiva pāti nityam sureśvarah".

75

O Śakra, my desire is not for Indrahood" (3.218.14).94 Indra still protests, but Skanda-Kārttikeya demands

he be assigned a different duty by Sakra. Finally, Indra concedes and tells him: "If these words spoken by

you are from a conviction and true, and if, O Skanda, you desire to do as I instruct [then] listen to me. You

must be anointed to the generalship of the gods, O one of great strength, from your command I will be

Indra, O strong one" (3.218.20-21).95 Skanda-Kärttikeya answers: "Consecrate me to the generalship for the

purpose of protecting cows and Brahmins, and for the success of the gods and the destruction of Dānavas"

(3.218.22).96 Quite remarkably Skanda-Kärttikeya has become a force of order and devoted to the cause of

the gods. The predictions of him made in the frame narratives have suddenly come true. Only a few chapters

ago he appeared to be just like a Dänava tearing apart the ranks of the gods. Now he has duties under Indra

and duties to Brahmins. The assignment of these duties is, I think, an attempt at catagorizing Skanda-

Kärttikeya within the orthodox system. He is being assigned specific roles and duties in relation to existing

hierarchies.

Part of Skanda-Kärttikeya's becoming the general of the army of the gods is his induction

ceremony or abhiseka. Skanda-Kārttikeya's abhiseka is a much more developed theme in the Salyaparvan

version of his birth and deeds, and I will discuss it in more detail there. For now, I will state that abhiseka

rites are closely tied to royal legitimacy and authority. It is a ritual largely reserved for deities and royalty,

and within a royal or courtly context it is designed to mark someone off as different or special. It initiates

someone into an elite circle of leadership and authority. I have raised the topic of a royal or courtly rite here

because this abhiseka seems to participate in both a definition of a deity and a demarcation of a courtly duty

94"aham te kimkarah śakra na mamendratvamīpsitam".

95" śakra uvāca yadi satyamidam vākyam niścayādbhāsitam tvayā yadi vā śāsanam skanda kartumicchasi me śrnu [.20] abhisicyasva devānām senāpatye mahābala ahamindro bhavisyāmi tava vākyānmahābala [.21]".

96In the introduction to the birth story Angiras also turns down the role of replacing Agni when he retires to the forest even though the sage is described as becoming a better fire than the god (3.207.7-19). That narrative seems to foreshadow this one. The Sankrit in the above passage reads: "skanda uvāca dānavānām vināsāya devānāmarthasiddhye gobrāhmanasya trānārtham senāpatye'bhisiñca mām"

76

in taking on the leadership of the army. The most important element of this rite is that it separates Skanda-

Kārttikeya from his ignoble roots and places him within elite divine and social circles. This ritual, in effect,

makes him part of Indra's divine court. Just how this separation from his Graha roots affects his cult will be

discussed in another section. The effect of the abhiseka on his appearance and characterization is, however,

dramatic:

The son of Pavaka [Agni] was clothed in a pair of red dustless clothes. His blazing and glorious body shone like the sun [covered] in two red clouds. The cock, which was given [to him] by Agni and which adorned his banner and which was raised over [his] chariot, shone red like the Doomsday fire. His armour entered his body, which was produced at his birth. It always becomes manifest when the god is fighting. Spear, armour, strength, splendour, beauty, truth, invulnerability, Brahminic faith, lack of confusion, protection of worshippers, and the destruction of enemies, and the protection of all the worlds were all born with Skanda, O King. (3.218.31- 35)97

We have seen some of these images before, but the use of certain adjectives in this post-abhiseka

description creates a different perception of the deity. His clothes are dustless; his shine majestic. His cock

has been transformed into an ornamental standard, part of the pomp of his chariot. Our earlier introduction

to Skanda-Kärttikeya's cock and spear tended to emphasize the horrific appearance and giant size of the

new-born god:

Having heard his roar, which had the sound of a mass of clouds, both the great elephants Citra and Airävata jumped up. Having seen both [elephants] rushing towards [him], he, whose radiance was equal to the newly risen sun, grasped [the elephants] with two hands, with another hand [he held] a spear and with another hand the son of Agni held a cock. Having grasped the huge bodied cock, which was brought near, and was the best of the mighty, the very strong one bellowed and sported [about] terribly. (3.214.22-24)98

In this earlier description the god's red glow, giant cock and spear are all parts of his fearful appearance.

97"araje vāsasīrakte vasānah pāvakātmajah bhāti dīptavapuh śrīmāntraktābhrābhyāmivāmśumān [.31] kukkutaścāgninā dattastasya keturalamkrtah rathe samucchrito bhāti kālāgniriva lohitah [.32] viveśa kavacam cāsya śarīram sahajam tatah yudhyamānasya devasya prādurbhavati tatsadā [.33] śaktirvarma balam tejah kāntatvam satyamaksatih brahmanyatvamsammoho bhaktānām pariraksanam [.34] nikrntanam ca śatrūnām lokānām cabhiraksanam skandena saha jātāni sarvānyeva janādhipa [.35]".

9"tasya tam ninadam śrutvā mahāmeghaughanisvanam utpetaturmahānāgau citraścairāvataśca ha [.22] tāvāpatantau sampreksya sa bālārkasamadyutih dvābhyām grhītvā pānibhyām śaktim cānyena pāninā aparenāgnidāyādastāmracūdam bhujena sah [.23] mahākāyamupaślistam kukkutam balinām varam grhītvā vyanadadbhīmam cikrida ca mahābalah [.24]".

77

After the abhiseka these elements point to his majesty and are signs of his status and authority. The point

seems to be that while the emblems and might of the god have not changed, the manner in which we are to

perceive the god and his attributes has changed, or, at least, this seems to be the hope of the redactors of this

text. Regarding Markandeya's statement that beauty, Brahminic faith, lack of confusion and defence of the

worlds were created with Skanda-Kärttikeya's birth, we have seen from chapters 214-216 (section 2.3.1)

that very little of this statement is accurate, but such is this text's remodelling of the deity in chapter 218. It

may be more truthful to state that these qualities come about through the worship of this deity and his

abhiseka into his new duties. The above quotation from 218 also states that Skanda-Kärttikeya was born

with a coat of armour, which was not suggested in the chapters describing his birth. As I suggested above,

the coat of armour is significant, and this later statement that he is born in it is part of this text's attempt to

limit his role to that of Mahasena -- he is obviously born to be a warrior, or so this text would have us

understand

After the abhiseka Skanda-Kärttikeya receives his army. The army worships and praises him after

which he gives them comfort (3.218.40-41). This last part of the text is significant. It was clear earlier that

people or creatures worshipped Skanda-Kārttikeya because they feared him. He was unpredictable and

chaotic and seemed to be calmed only by worship and surrender. Once he becomes the surasenāpati (the

general of the army of the gods), however, that unpredictability and inauspiciousness are lost. He has

become an auspicious protector of Brahminic values who exists to destroy demons. Indeed, his image as

auspicious, a protector and as a representative of order is in keeping with his role as a general. We might

well question: what do people have to fear from this figure if they are following Brahminic ways, or do not

represent a threat to the gods, and if fear was the basis of his worship, what reason is there to worship him

now? His worship now comes from the army. There may be a hint of fear in their worship, but no direct

indication of his terrible nature is given as it was in earlier chapters.

Once Skanda-Kärttikeya is made the general, Indra remembers Devasenã and introduces her to

Skanda-Kärttikeya with the following: "O best of the Gods, this maiden was declared by the Self-existent

78

One to be your wife, while you were unborn'" (3.218.44).99 The two are married; the story has almost come

full-circle. The marriage to Devasen is a reminder that this was the point of Skanda-Kärttikeya's birth, he

is to marry the divine army and defeat demons. The horrific character that sprang from Mount Śveta can

now be forgotten because the 'true' Mahasena has now appeared. We may also recall that the world of the

married sages was earlier in the story presented as part of what made up orderly Brahminical society.

Skanda-Kärttikeya's marriage is another means for this text to present him as ordered and Brahminical -- he

is now a domesticated being who lives up to the duties of a son, husband and general.

The chapter ends with some interesting comments concerning Devasena: "The wise know that

Devasena is the chief wife of Skanda, her whom the Brahmins call Sasthī, Laksmī, Āśā, Sukhapradā,

Sinīvalī, Kuhū, Sadvrtti and Aparājitā" (3.218.47).100 Six of these eight goddesses are separate goddesses in

other texts.101 We get some hint here of this text's approach to various similar deities. For the most part the

text attempts to assimilate similar deities. Most of the goddesses listed have some association with growth,

prosperity and offspring and are amalgamated in this text. A similar fate awaits Skanda, Visākha, Guha and

Naigamesa. While their exact relationships with each other are not clear in this story, they will eventually all

become forms of Skanda-Kärttikeya without separate identities. We can see in these attempts to amalgamate

these deities a general approach Brahminical writers took to non-Brahminical deities that they assimilated:

9"skandam covāca balabhidiyam kanyā surottama ajāte tvayi nirdistā tava patnī svayambhuvā".

1"evam skandasya mahisīm devasenām vidurbudhāh sasthīm yām brāhmanāh prāhurlaksmīmāšām sukhapradām sinīvālīm kuhūm caiva sadvrttimaparājitām".

1Sasthi is a goddess in medical literature who is usually connected with Skanda-Kārttikeya. She will be discussed later. Laksmī is a well known goddess of prosperity who becomes Visnu's consort. Āsā is 'Hope' the personified wife of Vasu in the Harivamsa. Sinīvali is a goddess of fertility and easy birth in the Rgveda (2.32.7). In the Atharvaveda she is worshipped for offspring and described as the wife of Visnu (AV 7.46). In later Vedic texts she is also the presiding deity of the first day of the new moon (Monier- Williams 1999:1217). Kuhū is also a goddess in the Atharvaveda (7.47) and was likely a goddess of the new moon (Monier-Williams 1999:299). Aparājitā is also a name of Durga and several plants (Monier- Williams 1999:51). P. K. Agrawala focuses on the figure of Sasthi. He notes from J. Gonda that she is also identified with Sri in the Manava Grhya Sutra (2.13) (1967:35), and she is connected with new-born children in the Kauśnakī Grhya Sutra (19.7), (1967:35; Gonda 1977:627).

79

similar gods and goddesses are fused together and then placed into familial relationships with established

deities or presented as forms of more important deities. Small cults and local cults are rolled into a small

number of larger cults on the basis of fabricated family ties.

As I have noted, chapter 219 demonstrates a knowledge of Skanda-Kärttikeya's association with

Grahas, but it also ties up some loose ends created by Skanda-Kärttikeya's birth. The six wives of the seers,

who are described as "six Goddesses" (3.219.1), approach Skanda-Kärttikeya. They explain that they were

divorced through no fault of their own. They want Skanda-Karttikeya to acknowledge that he was not born

from them, but also to acknowledge himself as their son (3.219.1-5). Skanda-Kärttikeya agrees to this, and

it appears that these six goddesses become the Krttikās (3.219.7-11).

Chapter 220 opens with Svāhā, as Skanda-Kärttikeya's 'mother', asking him for a gift. She tells

him that she loves Agni and wants to dwell with him forever. Skanda-Kärttikeya fulfils her wish by telling

her that priests will always say svaha when offering an oblation into the fire, and in that way she will always

dwell with Agni (3.220.1-6).102 The chapter ends with a view of Mount Sveta that is radically different from

that presented in chapter 214:

He, who was accompanied by groups of gods and groups of Pisacas, who was surrounded by Śri and who was glowing, shone on the golden mountain. That mountain, which has a beautiful forest, shone due to that hero like [the way] Mount Mandara, which has beautiful caves, shines due to sun beams. Mount Sveta shone with groups of divine birds, with groups of divine deer, and groups of Kadamba trees, Japā and Aśoka forests, Pārijāta forests, Karavīra forests and Samtānaka forests which abounded in flowers .... There divine Gandharvas danced as did Apsaras, and there the sound of pleased creatures was heard. Thus the whole world along with Indra stood on Mount Śveta. He [Indra] looked at the delighted Skanda and did not tire of looking. (3.220.21-27)103

This is a far cry from the desolate and dangerous place where a disguised Svāha threw the semen of Agni

102The word svaha 'hail' is pronounced at most offerings.

103“sa samvrtah pišācānām ganairdevaganaistathā śuśubhe kāñcane śaile dīpyamānah śrīyā vrtah [.21] tena vīrena śuśubhe sa śailah śubhakānanah ādityenevāmśumatā mandaraścārukandarah [.22] samtānakavanaih phullaih karavīravanairapi pārijātavanaiścaiva japāśokavanaistathā [.23] kadambatarusandaiśca divyairmrgaganairapi divyaih paksiganaiścaiva śuśubhe śvetaparvatah [.24] .... tatradivyāśca gandharvā nrtyantyapsarasastathā hrstānām tatra bhūtānām śrūyate ninado mahān [.26]

[.27]". evam sendram jagatsarvam śvetaparvatasamsthitam prahrstam preksate skandam na ca glāyati daršanāt

80

six times into a pot on a deserted mountain. Such, however, is the magnitude of the transformation from

inauspicious to auspicious carried out by this story from chapter 214 to 220.

2.8 The Auspicious Skanda-Karttikeya: Siva's Fatherhood in the Aranyakaparvan

The fatherhood of Siva is first established in the pivotal chapter 218. Just after the performance of

the abhiseka the gods give Skanda-Kärttikeya various gifts. When it is Siva's turn to give a gift we are told

the following:

The Brahmins declare that Rudra is Agni, hence he is the son of Rudra. Mount Śveta was produced from the poured out semen of Rudra, and (it was) on Mount Sveta that the semen of Agni was cultivated by the Krttikas. All the deities, having seen the honouring [of Skanda] by Rudra, called Guha, who is the best of the excellent, the son of Rudra; this boy was born by Rudra entering into Agni. Therefore, Skanda was born the son of Rudra. Hence, Skanda, who is the best of the gods, was born the son of Rudra by the splendour of Rudra, Agni, Svaha and the six women. (3.218.27- 30)104

As we have seen, the Satapatha Brahmana equates Rudra with Agni, so the claims here are not without a

Vedic basis. Siva does become the recognized father of Skanda-Kärttikeya in most of the post-Epic

accounts of his birth. The introduction of Rudra-Siva as the 'real' father of Skanda-Kärttikeya is sudden in

this text, and also accompanies the sudden appearance of the depiction of Skanda-Kärttikeya as wealthy and

noble. The presentation of Rudra-Siva as father and the induction of Skanda-Kārttikeya as surasenāpati are

related events. The exact effects of these events will become clearer as we continue.

As with the case of Agni, we can expect that Skanda-Kärttikeya will embody some of the

characteristics of his other father, Rudra (P. K. Agrawala 1967:9-10). I will briefly outline Rudra's

character to illustrate how father and son are linked.105 In the Rgveda Rudra is an unusual deity in that he

possesses numerous malevolent characteristics as well as benevolent characteristics. In the earliest Veda

104“rudramagnim dvijāh prāhū rudrasnustatastu sah rudrena sukramutsrstam tacchvetah parvato 'bhavat pāvakasyendriyam śvete krttikābhih krtam nage [.27] pūjyamānam tu rudreņa drstvā sarve divaukasah rudrasūnum tatah prāhurguham gunavatām varam [.28] anupraviśya rudrena vahnim jāto hyayam sisuh tatra jātastatah skando rudrasūnustato-bhavat [.29] rudrasya vahneh svāhāyāh sannam strīnām ca tejasā jātah skandah suraśrestho rudrasūnustato'bhavat [.30]".

105 A more detailed discussion of Rudra may be found in Arbman (1922).

81

Rudra is a cause of healing and a dispenser of medicines (RV 2.33.2,4,7.12; 1.114.5). While Rudra is often

called upon to heal, in other verses from the same hymns he is described as a cause of death (RV

2.33.11,14-15; 1.114.7-8). Rudra is both praised and sought after as a healer, but also feared and avoided as

a bringer of death. The hymns also make it clear that while all people are subject to Rudra's wrath, the

young are frequently singled out as targets of this deity (RV 2.33.14; 1.114.8). The particular concern for

children in seeking protection from Rudra is also a concern relating to Grahas. Due to the malevolent and

unpredictable nature of Rudra he is worshiped out of fear (RV 2.33.4-5, 8; 1.114.3). He seems to show

great kindness to those who are devoted to him and great malevolence to those who are not. The

Atharvaveda and Satapatha Brāhmana relate similar points about Rudra.106

106Not many hymns are solely devoted to Rudra in the Atharvaveda, but those that are concentrate on his ability to cause disease:

The arrow that Rudra hurled at thee, at thy limbs and heart, that do we now thus eject asunder from thee. [1]The hundred tubes that are thine, distributed along thy limbs, of all these of thine do we call out the poisons. [2]Homage to thee, O Rudra, when hurling; homage to [thine arrow] when aimed (pratihita); homage to it when let fly; homage to it when having hit [3]. (6.90.1-3 translated by Whitney 1962 I:347)

Rudra's arrows represent the diseases with which he inflicts people, and, as with the Rgveda examples, Rudra is praised and worshiped from fear. In another hymn he and Soma are associated with curing disease: "O Soma-and-Rudra, eject asunder the disease that has entered our household; drive far to a distance perdition; any committed sin put away from us. [1]O Soma-and-Rudra, do ye put all these remedies in our bodies; untie, loosen from us what committed sin may be bound in our bodies [2]" (7.42.1-2 translated by Whitney 1962 I:415). This passage adds the idea of morality to Rudra's actions. In the Rgveda Rudra had little or no connection with concepts like rta and a concept of sin. If anything, Rudra represented the opposite of the order embodied in Varuna. Here, he and Soma seem to untie the effects of divine or moral retribution. There are some significant differences between the Rgvedic and Atharvavedic perceptions of Rudra. These changes may reflect a historical development in the deity; a softening of his character that makes his punishments less random and in support of divine order. A similar moralizing is applied to Skanda-Kārttikeya in medical texts. Rudra also appears in the Satapatha Brahmana, and his characterization in this text shares much with what I have discussed above. Here Rudra is acknowledged as a dangerous deity who has destructive powers. He is described as causing harm to cattle and to households in general (1.7.3.21, 1.7.4.12), and his attacks seem to affect everyone, even the unborn (2.6.2.2). One group of libations called the Tryambaka are offered to Rudra for the protection of one's children (Arbman 1922:48-63). The text tells us that in giving the offering the ritualist "delivers from Rudra's power both the descendants that are born unto him and those that are unborn; and his offspring is brought forth without disease and blemish" (2.6.2.2 translated by Eggeling 1882-1900 I:438). Rudra is worshiped out of fear in this text. He is violent and associated with disease and harming children and unborn children. In another section of the text we are told that Rudra is

82

A great deal of Rudra's character reemerges in the cult of Skanda-Kārttikeya, particularly as he is

described in the opening chapters of his birth story in the Aranyakaparvan. Rudra's ambivalent nature, his

role as a healer and a cause of illness, and the idea that he is worshiped out of fear all resurface in the

character and cult of Skanda-Kärttikeya and the Grahas. Rudra-Siva is a logical choice for Skanda-

Kārttikeya's father because they are similar figures. As I suggested earlier, the Vedic tradition understood

sons to be manifestations of their fathers' characteristics, and Skanda-Kārttikeya before his abhiseka can be

viewed as a manifestation of Rudra's character. It is clear, I think, that the cults of Skanda-Kārttikeya and

Rudra are linked, and that the choice of Rudra-Siva as Skanda-Karttikeya's father is not coincidental. What

is also clear is that both Rudra-Siva's and Skanda-Kärttikeya's cults are moving in similar directions. Both

of their characters are being softened. Both are amalgamating other related cult figures within themselves.

By chapter 218 of the ranyakaparvan Skanda-Kārttikeya is no longer like the Rudra figure I have

described above.

There is another stage in Rudra-Siva's development in this story. I have quoted above in relation to

Grahas that afflict people over the age of 16 the following: " ... no Grahas touch those who are devoted to

god Maheśvara" (3.219.58).107 This quotation suggests that all of the worship and propitiation towards

Skanda-Kärttikeya and the Grahas can be avoided if one simply worships Siva. Hierarchy and order seem to

left out of the sacrifice and threatens to attack the gods if he is not included in the rite. This threat of attack alone is enough to force the gods to concede the leftovers of the sacrifice to this deity (1.7.3.1-7). Like Skanda-Kärttikeya, Rudra's physical might and ability to destroy are additional points of fear in those he comes into contact, but he can be tamed through offerings. While worshiped to prevent becoming an object of his anger, the Satapatha Brahmana also acknowledges that Rudra is associated with healing and medicine. Rudra's dart or arrow in the text is often understood to be the vehicle through which he delivers disease. The Tryambaka rite mentioned earlier is, in part, designed to remove these diseases and create medicine. The text tells us that the remaining oblation cakes are to be thrown into the air, "thereby they cut out his (Rudra's) darts from their bodies. If they fail to catch them, they touch (those that have fallen to the ground). Thereby they make them medicine ... " (2.6.2.16 translated by Eggeling 1882-1900 I:442). While this text maintains that Rudra is also a god of healing, there is also some suggestion in the above quotations that it may be an illness that Rudra caused that needs to be healed.

107 .... na sprśanti grahā bhaktānnarāndevam maheśvaram".

83

be important to this text, and Skanda-Kärttikeya has accepted a role beneath gods like Indra, and he can also

be controlled by his other superior: his father Siva. In this very brief conclusion we see the end of Skanda-

Kärttikeya's ritual cult. He has been softened, which removes him from the initial point of worshiping him

in the first place, and now we are being told that worship of his more powerful father is more efficacious.

Once acknowledged as the son of Siva and placed in a subservient role to that deity, the point of worshiping

Skanda-Kārttikeya at all disappears.

These points become clear when Skanda-Kārttikeya is told by Brahmā and Prajāpati: "Go to

Mahadeva, the Destroyer of Tripura [and your] father. You, the unconquered one, have been born for the

sake of the welfare of all the worlds by Rudra entering Agni and by Umā taking possession of Svāhā'"

(3.220.8-9).108 He is also told by them: "Your horrible and flesh-eating retinue, they are understood by the

wise to be Ganas, who have various forms " (3.220.12).109 Not only is the text claiming that Skanda-

Kärttikeya is the son of Siva, but that the Grahas are really part of Siva's ghoulish cohort, the Ganas. The

point of worshiping Skanda-Kärttikeya is steadily being removed as his cult is absorbed into Saivism.

The text continues to disassociate Skanda-Kärttikeya from his Graha cult by advising those who

want to be cured from a disease or who want wealth to worship the five Ganas. Those who are concerned

for the well-being of the young should worship Miñjika and Miñjika, "who were born from Rudra"

(3.220.15) (rudrasambhavam). Those who want sons should worship "the man-eating women called

Vrddhikās by name, who are born from trees" (3.220.16).110 In short, one need not worship Dhūrta for

protection and wealth, or Skanda-Kärttikeya for the health of the young because these roles are fulfilled by

other members of Siva's entourage. The only meaningful role left for Skanda-Kārttikeya is as the senāpati,

108“tato brahmā mahāsenam prajāpatirathābravīt abhigaccha mahādevam pitaram tripurārdanam [.8] rudrenāgnim samāviśya svāhāmāviśya comayā hitārtham sarvalokānām jātastvamaparājitah [.9]".

109“ta ete vividhākārā ganā jñeyā manīsibhih tava pārisadā ghorā ya ete pišitāśanāh". o manuçamänst 110"striyo manusamāmsādā vrddhikā nāma nāmatah vrksesu jātāstā ... ".

84

and in this text that role is not one based in a ritual cult except within the army itself.111

Chapter 221 primarily concerns a battle between the army of the gods and the Dānavas. It begins

with a lengthy description of Siva in all his majesty leading the gods back to heaven. It reads as a

glorification of Siva, and the description of him reminds us who controls the gods and where Skanda-

Kārttikeya sits in relation to him:

There that god [Siva], who was accompanied by them went along pleasantly in the front and in the back because his path is not fixed. Mortals worship the divine Rudra with virtuous rites. The one they call Lord Rudra, who carries a bow, he is Siva. They honour Maheśvara with various objects. Thus the husband of Devasena, who was surrounded by the armies of the gods, the brahminic son of the Krttikäs, followed the lord of the gods. (3.221.23-25)112

The glory and worship of Siva are made clear in the above as is Skanda-Kärttikeya's role under him as the

general of the army of the gods. Even as the general of the army of the gods his role is diminished by Siva:

Then Mahadeva said [these] great words to Mahasena: "Always alertly protect the seventh division of the Maruts." Skanda said: "I will. O Rudra, watch the seventh division of the Maruts. Tell me quickly, O god, what else am I to do?" Rudra said: "Son, I am always to be looked to in your duties. You will obtain the supreme good by devotion to me and by looking to me." (3.221.26- 28)113

These quotations give us the sense that the better general is Siva and, again, Skanda-Kārttikeya's role as

subservient to his father in all matters is emphasized.

There is also some significance in the procession to heaven in that it leads Skanda-Kārttikeya away

from the terrestrial realm. I noted in the Skandayäga that the ritual is concerned with human and earthly

111A similar fate awaits Skanda-Kārttikeya's association with Mātrs. Eventually, it is Siva and Ganeśa who are associated with the Mätrs, especially in statuary. Many of Skanda-Kārttikeya's once inauspicious roles are eventually absorbed by Siva and other members of his entourage.

112"ebhih sa sahitastatra yayau devo yathāsukham agratah prstataścaiva na hi tasya gatirdhruvā [.23] rudram satkarmabhirmatyāh pūjayantīha daivatam šivamityeva yam prāhurīśam rudram pinākinam bhāvaistu vividhākāraih pūjayanti maheśvaram [.24] devasenāpatistvevam devasenābhirāvrtah anugacchati devesam brahmanyah krttikāsutah [.25]".

113“athābravīnmahāsenam mahādevo brhadvacah saptamam mārutaskandham raksa nityamatandritah [.26] skanda uvāca saptamam mārutaskandham pālayisyāmyaham prabho yadanyadapi me kāryam deva tadvada māciram [.27] rudra uvāca kāresvaham tvayā putra samdrastavyah sadaiva hi darśanānmama bhaktyā ca śreyah paramavāpsyasi [.28]".

85

matters. The Ayurveda and Grhyasūtra texts also had a household and earthly context to them. One gets the

impression from these texts that Skanda-Kärttikeya is engaged in the world as a deity who requires

propitiation. The same may be said for the image of the deity presented in the first few chapters of this story

in the Aranyakaparvan. He is given an earthly birth, and the first elements of his ritual cult concern his

actions on the earth. The parentage of Agni made sense in such a context because he is also closely

associated with the terrestrial realm. Once made the general of the gods and the son of Maheśvara-Siva,

however, he is drawn away from earthly concerns. The march to heaven is a sign of his new status and

realm of action. This theme is more fully explored in other Epic versions of his birth and deeds and will be

discussed in more detail below. For now, we may note a steady movement away from the earth and

inauspiciousness to heaven and auspicious, but this movement does raise the question of what becomes of

his cult which seemed based on his terrestrial and dangerous nature -- a point with which much of the rest of

this thesis will deal.

The eventual battle between Skanda-Kärttikeya and various asuras is of little interest to us. The

gods do poorly in the battle and are routed by a particularly powerful demon named Mahisa. Skanda-

Kärttikeya is able, however, to kill this demon and defeat the rest of the demon army. Perhaps the most

interesting element of this battle is the appearance and defeat of Mahisa. One of the most prominent

narratives in Hinduism and a prominent subject for artists is the defeat of this demon by Durga, but here the

role is given to Skanda-Kärttikeya. The role is quite literally 'given' to Skanda-Kärttikeya because the text

makes it clear that Siva could have done the job himself. The gods are losing to Mahisa and the demon

attacks Siva's chariot. Then, we are told: "But, the Lord did not lie in wait for Mahisa in battle and he

remembered that Skanda [was to be] the death of this wicked one" (3.221.60).114 At the end of the battle

Skanda-Karttikeya is praised by Siva, part of which reads: "You are unconquerable in battle by your

enemies like the husband of Uma. This, your first deed, O god, will be celebrated. Your renown will be

114“tathābhūte tu bhagavānnāvadhīnmahisam rane sasmāra ca tadā skandam mrtyum tasya durātmanah".

86

imperishable in the three worlds ... " (3.221.75-76).115 The story ends where it began with the appearance of

Devasenā: Skanda-Kārttikeya is Mahāsena, and his other roles are made secondary in this text. The goal of

this story is to make him famed for his martial character and as the son of Siva. The other aspects of his cult

are eroded in this account of him. He leaves the story as a secondary figure when compared with his father,

Śiva.

I have stressed throughout my analysis of this story that it aught to be read as a conversion

narrative. It takes an inauspicious, non-Brahminical deity and transforms him into an auspicious,

Brahminical deity with links to both the Vedic and developing Hindu worlds. I have demonstrated that the

writers and redactors of this version are anxious about his inauspicious character and use themes of

domestication, ritual, Vedic precident and duty to construct an auspicious persona for the deity. The

Brahminical agenda in this account acknowledges the popular roots of this deity, but steadily mitigates

against that initial characterization to produce a purer model of Brahminical worship, who is also not a

threat to the cults of Agni and Siva. The remainder of this thesis, particularly the sections on coinage,

statues and other archaeological data will support this thesis that the conversion depicted in the

Åranyakaparvan is also mirrored by similar developments in the lived cult of the deity.

2.9 The Socio-political context of the Aranyakaparvan

In light of what I have just stated, I would like to comment on the potential socio-political context

of this Aranyakaparvan narrative. I have relied on an assumption that is held by most Epic scholars that the

authors and redactors of the Mahabharata are Brahmins. A second assumption made by most Epic scholars

is that these Brahmins must have received some support from royal or wealthy patrons. The size of the text

suggests that its redactors were not engaged in other professions and likely received financial support from

kings or members of a royal court so that they could devote themselves to this task. In Vedic texts priests

are a major recipient of kingly gifts, and the idea of royal patronage seems well established before the Epic

I>" ... ajeyastvam rane'rīnāmumāpatiriva prabhuh [.75] etatte prathamam deva khyātam karma bhavisyati trisu lokesu kīrtiśca tavāksayyā bhavisyati ... [.76]".

87

period begins.

There are, of course, a number of theories regarding the identity of the sponsoring kings, and just

when the final version of the Mahabharata was produced. Various dynasties have been suggested from the

Mauryas116 and Sungas117 to the Guptas118. Most, however, reject the Mauryas because of their apparent

support of what the Brahminical tradition and the Epic would view as heresies. As I have already noted, the

Mahabhārata positions itself as a defense of orthodox Brahminical religion and takes an anti-heterodox

position. Hence, most scholars think it is unlikely that an emperor like Asoka would have sponsored a text

that was critical of his support of heterodox traditions. Alf Hiltebeitel has also criticized those who support

Pusyamitra or the Sungas as the main sponsors of the text (2001:16-17). He notes that Pusyamitra was a

Brahmin king, and that the Mahabharata is very critical of Brahmin kings, hence: "it should be difficult to

maintain that Brahman kings would patronize epics that disqualified them from ruling" (2001:16-17).

Hiltebeitel suggests that the Epic could have been produced during the Sunga period, but not in the Sunga

court itself. He suggests composition dates of 200 BCE to the year zero (2001:18), and proposes "that the

Mahabharata was written by 'out of sorts' Brahmans who may have had some minor king's or merchant's

patronage, but, probably for personal reasons, show a deep appreciation of, and indeed exalt, Brahmans

who practice the 'way of gleaning': that is, unchavrtti Brahmans ... " (2001:19). Finally, James Fitzgerald

suggests a Gupta era redaction because he feels that an undertaking of this size and the promotion of it

would have taken the support of a major dynasty (1991:154; Hiltebeitel 2001:25-26). Hiltebeitel does not

116N. Sutton argues for the Mauryans as the period of final redaction of the Mahabharata, and that the figure of Yudhisthira could have been modelled after Asoka or Candragupta (1997:334-339). These ideas do not seem likely, however, for points which will be discussed below. Other criticisms can be found in Hiltebeitel (2001:17).

117Hiltebeitel points to E. Hopkins as creating the idea of the Sungas as responsible for the final redaction (2001:16). Hopkins argued that the text may have been redacted during the 2nd century CE, and he notes that these rulers were Brahminical and anti-Buddhist, and would have sponsored the Epics to defend their religion (1977:398-400). Criticisms of this view may be found below.

118T. Oberlies (1998:128) and J. Fitzgerald (cited in Hiltebeitel 2001:25-26) have suggested Gupta dates. Fitzgerald's comments will be discussed below.

88

regard the evidence for a Gupta date as very strong, and does not feel we need the Guptas to explain the

diversity and complexity of the text (2001:26).119 I do not hope to promote one of these suggestions over the

others, though I find those of Fitzgerald and Hiltebeitel to carry the most weight. What I hope to emphasize

here is that royal patronage of the production of the Mahabharata is a widely accepted notion which

emphasizes a backdrop of imperial formations along with a defense of orthodox religion and Brahmins

(Hiltebeitel 2001:17). It is the idea of imperial formations that I will focus on here.

The essential plot of the Mahabhärata is a royal or kingly drama, and it is not surprising that a

sponsoring ruler would take an interest in such a text. While there is no doubt that not all of the

Mahābhārata relates to royal concerns, the text is too large and complex for such a blanket statement to be

true, the Aranyakaparvan version of Skanda-Kärttikeya's birth and deeds does indeed reflect on royal or

courtly themes that are worthy of note. It may reflect actual socio-political concerns regarding this narrative

and the developing cult of Skanda-Kārttikeya.

I will begin with ideas of how the text goes about legitimating Skanda-Kārttikeya as a Brahminical

deity. There are, I think, two forms of legitimation used by this text. The first means of legitimation, as we

have seen, is through parentage and lineage in the narrative of Agni. This type of legitimation through

lineage is described by Romila Thapar as related to pre-state forms of government (2000:8, 11). Ritual as a

form of legitimation is still important in such a society, but genealogical connections, be they real or not, are

the focus of such societies seeking to legitimate their elite members. The Aranyakaparvan lineage itself,

however, is a priestly one, and, as I have suggested, likely relates to Brahminical concerns of legitimation

concerning this deity.

The second form of legitimation employed by this text is primarily through ritual. I have suggested

that the abhiseka of Skanda-Kärttikeya is a ritual of legitimacy connected to royalty. Thapar regards this

119Based solely on the progression of the Skanda-Kärttikeya narratives contained in the Mahabharata, I would suggest dates running from the first century BCE to the Gupta Empire for the period of writing and redaction of the text, though this view is limited by the narrow scope of the material I examine from the text itself.

89

type of legitimation to be reflective of a state (2000:172), and I would like to emphasize this point. In such a

state, she argues, there is a focus on the obedience of officials in running the state, the state controls

succession to high office, justifies social divisions and supports religious systems where of use (2000:12-

13). Lineage connections in such a state are still important, but more often than not fabricated, particularly

in court members who have no obvious link to the ruling clan/family. Rites like the abhiseka certainly use

religion as part of their effort to legitimize a figure, but this form of legitimization is primarily concerned

with socio-political structures and is thus separate from the form of legitimation discussed above. Certainly,

a king with a standing army which has a permanent senapati is reflective of a state with a centralized

political machinery. This centralized machinery is basically the depiction of the socio-political

circumstances in these Epic stories once Skanda-Kärttikeya is made a general and domesticated. Skanda-

Kärttikeya's eventual characterization as an auspicious general who maintains order, respects hierarchy and

is powerful but subservient to Indra and Siva is an appropriate model image of what a king might hope for

in his administrators. The inauspicious and dangerous image of Skanda-Kärttikeya who defeats Indra, the

king, and has no respect for order is not an image with which a real king would find a great deal of comfort.

The transformation of Skanda-Karttikeya from inauspicious to auspicious and from rogue to general maybe

reflective of some Brahminical concerns about him, but it may also reflect the concerns of sponsoring kings.

The character of Skanda-Kärttikeya is developed to be a support for his king and to bring order and

prosperity to the realm for the greater glory of his elders. Clearly, I do not think that the final depiction of

Skanda-Kärttikeya as an idealized general is brought about without motivations or causes. One such

motivation may have been a desire by real kings to have a divine general portrayed who met their own

standards of an ideal, model general. I will acknowledge that much of this is speculation, but as we shall

see, kings and emperors in India do play significant roles in the development of Skanda-Kärttikeya's cult. I

would like to raise here the possibility that Skanda-Kärttikeya's narrative is influenced by such sources.

90

Chapter Three: The Birth of Skanda-Karttikeya in the Salyaparvan and Anusasanaparvan 3.1 The Salyaparvan account of Skanda-Karttikeya's birth and deeds

The Salyaparvan account of the birth of Skanda-Kärttikeya is almost entirely focused on Skanda-

Kārttikeya as the surasenāpati whose duty it is to destroy demons. One of this version's unique elements is

how it deals with the multiple parentage of Skanda-Kärttikeya. The text is inclined to elevate Siva above the

other parents as the most powerful of them, but it does not give him exclusive parentage over the others;

how the text achieves this balance will be discussed below. This version of the birth narrative is also

reluctant to acknowledge Skanda-Karttikeya's inauspicious background and provides an unusual

explanation to account for his power. The text presents him as an ascetic or yogin, and this is the root of his

power. The text does acknowledge that Skanda-Kärttikeya is associated with various ghoulish figures, but

these figures are not directly called Grahas and are all part of his divine army. This story makes no direct

reference to Skanda-Kärttikeya as a Graha or leader of Grahas. The text also suggests a shift in rituals

associated with this deity from those concerned with propitiation and sickness to those concerned with

initiation and legitimacy. The abhiseka, which only receives a brief mention in the Aranyakaparvan, is an

important element of this story and will be discussed in more detail here. The story itself is narrated by

Vaiśampāyana to Janamejaya.

3.2 The Parentage of Skanda-Karttikeya in the Salyaparvan

The parentage of Skanda-Kärttikeya is an issue for this story, as it is in every early account of the

deity. Vaiśampāyana tells Janamejaya that the semen or energy (tejas)1 of Mahevara (Siva) falls (skanna)2

Tejas may be translated in a number of ways. It is a type of energy that makes things glow and is often associated with the divine and glorious humans. It can also mean "semen virile" (Monier-Williams 1999:454), but still carries the sense of a fiery energy.

2Skanna is often used in these accounts of Skanda-Kärttikeya to provide a false etymology for Skanda, which can also mean "spurting, effusing, spilling, shedding" (Monier-Williams 1999:1256). As we shall see, in these versions the claim is made that Skanda is called Skanda because he results from the skanna of Śiva's tejas.

91

into Agni (9.43.6). Agni is not able to handle the energy of this tejas and passes it onto Gang, the goddess

who is also the Ganges River. She is also not able to sustain the energy of Siva's tejas and drops it in the

Himalayas (9.43.7-9). The text begins, then, by acknowledging several parents for Skanda-Kärttikeya, but

by also demonstrating the power of Siva through the potency of his tejas. No other deity is capable of

sustaining his energy; the text seems to make the point that it regards Siva as the most powerful of them.

Once cast onto the mountain, Vaisampayana states, "the son of fire" (jvalanatmaja) begins to grow

(9.43.10). The six Krttikas come across the child, and each claims him as her own, by calling out: "This one

is mine" (mamāyamiti) (9.43.11). To satisfy them Skanda-Kärttikeya suckles from each of them with his six

heads.3 Shortly after this scene Vaisampāyana reports that Skanda-Kārttikeya is also known as "Kārttikeya",

which suggests he is the son of the Krttikas and "Gangeya", which suggests he is the son of Ganga

(9.43.16). The story presents us with Siva, Agni, Gangā and the Krttikās as possible parents for Skanda-

Kärttikeya, and while the text suggests that Siva is the most powerful of them, it does not decide on one of

these as his real parent.

Later in the text Vaisampayana narrates that Skanda-Kārttikeya approaches Siva, Umā, Agni and

Gangā. Each of these deities hopes that he will greet them first. To deal with this problem Skanda-

Kārttikeya creates three additional forms (mūrti) and greets the four 'parents' simultaneously (9.43.33-38).4

Hence, the text avoids singling out one of these parents as Skanda-Kārttikeya's 'real' parent. The Krttikãs

have been dropped from the list, but we have already seen that the text accepts them as mothers of Skanda-

3Some scholars claim the Skanda-Kärttikeya creates his six heads in order to suckle from the six Krttikas at once. This is the sense of one manuscript version of this story from the Salyaparvan which is translated as follows: "Understanding the state of mind of those six mothers, the adorable lord Skanda sucked the breasts of all having assumed six mouths" (Roy:1963 VII:131). The Sanskrit of the critical edition does not, however, suggest that Skanda-Kärttikeya gains his six heads here, but already has them at birth: "tāsām viditvā bhāvam tam mātrnām bhagavānprabhuh prasnutānām payah sadbhirvadanairapibattada" or "Then, the glorious Lord, having understood the affection of those mothers, drank the milk of their milk yielding breasts with [his] six mouths" (9.43.12). The Sanskrit makes it clear that the Krttikas are regarded as his mothers, and that he suckles from them with his six heads, but that the six heads are created at this moment is not stated here.

4The other forms he creates are Visākha, Šākha and Naigamesa (9.43.37).

92

Kārttikeya (9.43.12). Umā has been added, but she is the consort of Siva, and the text seems to assume that

if he is Skanda-Kärttikeya's father, she must also be his mother.5 Chapter 45 of this text ends with

speculation concerning the parentage of this deity: "Some say [that Skanda-Kärttikeya] is the son of

Maheśvara, others claim he is the son of Agni, or of Umā, or of the Krttikās, or of Gangā" (45.86).6 The text

leaves these points open and seems happy to assert multiple origins for the deity.

This version of the story begins where the Aranyakaparvan ends in presenting Siva as the source

of Skanda-Kärttikeya; he is the product of Siva's mighty tejas. The text does not, however, eliminate Agni

as a parent of the child, and it adds Ganga and Uma to the list of parents. In later versions of his birth story

Ganga and Uma remain as his mothers.' The appearance of Skanda-Kārttikeya's various forms does not

allow for Siva's dominance in this text. He is recognized as special to Skanda-Kärttikeya, but as only one of

four or five parents. We have yet to reach a stage in the narration of Skanda-Kärttikeya where the point of

his birth is to demonstrate the power of Siva, but we are clearly moving in that direction with this account of

his birth, and that provided in the latter sections of the Aranyakaparvan.

3.3 The auspicious Skanda-Karttikeya in the Salyaparvan: The military ascetic

The inauspicious attributes of Skanda-Kärttikeya are largely ignored by this version of his birth,

and he is presented as a military deity born to kill the asura named Tāraka. This context for Skanda-

Kärttikeya is introduced before the narrative of the birth actually begins. The larger context of this narrative

5Wendy Doniger regards the parentage of Pärvati as the hardest for these stories to establish because in many of the later versions of the story, and in the Anusasanaparvan, Siva gives a boon to the gods that he will not have a child with her. She states that the parentage of Siva and Parvati leads to more complex rationalizations as they replace the seven wives of the seers (1973:103-105). We have seen some of this rationalization already in the Aranyakaparvan concerning the parentage of Siva, but the Epic itself seems ready to accept Parvati's parentage without explanation. It is only in Puranic accounts that her parentage seems to become an issue.

6There are additional segments of this quotation which will be discussed below. The Sanskrit reads: "kecinmaheśvarasutam kecitputram vibhāvasoh umāyāh krttikānām ca gangāyāśca vadantyuta".

'Why Ganga is added to the list of parents will be discussed in my section on the Anusasanaparvan version of the birth story.

93

is Vaisampāyana explaining various tīrthas, or holy places, to Janamejaya. At the end of chapter 42

Vaisampāyana is describing a tīrtha on the Sarasvatī where Skanda-Kärttikeya, whom he also calls

Kārttikeya, Kumāra, Skanda, Daitya-destroyer and Mahāsena, dwells permanently in a bodily form. He

explains that it is also here that a great battle took place between the gods and demons, and also here that

Skanda-Kārttikeya killed Tāraka. Finally, it is here that Skanda-Kārttikeya became the "senapatyam", or

where he became the general of the gods (9.42.40-41).8 The story of Skanda-Kärttikeya's birth is introduced

through the frame of his role as a demon-killer and Mahasena. The power of the tirtha and its association

with Skanda-Kärttikeya is based on his military exploits, and the account of his birth is also focused on the

birth of a martial deity. Janamejaya's question which follows up on Vaisampāyana's statements echo similar

themes:

Janamejaya said: O first among the twice born that excellence of Sarasvati has been described by you. Now, O Brahman, please explain the anointment of Kumāra. O most excellent of speakers, tell how, where and when and by whom the blessed lord was installed and by what ritual, and, O great one, how Skanda carried out the destruction of the Daityas. My curiosity is great, tell it all to me. (9.43.1-3)9

Janamejaya's references to the anointment and installation of Skanda-Kärttikeya all refer to the abhiseka

ritual that is performed to make him the senäpati of the gods, which I will discuss in more detail below. The

introduction of this story makes it clear that Skanda-Kärttikeya becomes a general through an abhiseka, and

he destroys demons. These are the reasons for his birth according to this version of his story. Devasenā

makes no appearance in this version of the story, and this approach to the narrative is in contrast to the

8“yasyānte'bhūtsumahāndānavānām daiteyānām rāksasānām ca devaih sa samgrāmastārakākhyah sutīvro yatra skandastārakākhyam jaghān [40] mahāseno yatra daityāntakartā senāpatyam labdhavāndevatānām sāksāccātra nyavasatkārttikeyah sadā kumāro yatra sa plaksarājah [41]". "Near which there was the great battle, called the Taraka battle, between the gods and the demons in which Skanda killed Täraka, and where the destroyer of the Daityas was consecrated. Where Kärttikeya, the young one, dwells by the Fig Tree". My thanks to Phyllis Granoff for her help in this translation.

9"janamejaya uvāca sarasvatyāh prabhāvo 'yamuktaste dvijasattama kumārasyābhisekam tu brahmanvyākhyātumarhasi [1] yasminkāle ca deśe ca yathā ca vadatām vara yaiścābhisikto bhagavānvidhinā yena ca prabhuh [2] skando yathā ca daityānāmakarotkadanam mahat tathā me sarvamācaksva param kautūhalam hi me [3]".

94

Åranyakaparvan, where she was the means for introducing these topics. There is no mention of a wife for

Skanda-Kärttikeya in this text and very little of the domestication theme that was present in the

Åranyakaparvan. Much of the domestication theme in that other version was to establish a justification for

the dramatic change in Skanda-Kärttikeya's character. The Salyaparvan account, however, never allows for

an inauspicious Skanda-Kärttikeya and can drop much of the material used to explain his transformation.

This text, we might speculate, is either later than the Aranyakaparvan, and the deity is now regarded only

as the senapati of the gods, or the group(s) this version of the story addresses chose to ignore Skanda-

Kārttikeya's past.

In the Aranyakaparvan it was Skanda-Kärttikeya's initial outsider and dangerous image that made

him powerful. As the Salyaparvan never acknowledges this dangerous side of the deity, it must find a way

to account for his power and maintain his auspicious characterization. The means this version employs to

account for his power and keep him auspicious is to present Skanda-Kärttikeya as full of ascetic power.

Vaiśampāyana repeatedly calls Skanda-Kārttikeya "mahātman," 'great souled,' or 'great being'

(9.43.5, 9.43.11, 9.44.19, 9.44.37, 9.44.108), and also calls him "mahāyoga," 'great yogin' (9.43.16,

9.43.33), "yogīnāmīśvara," 'lord of the yogis' (9.45.87), "samanvitah tapasā," 'full with ascetic energy'

(9.43.17). It is also his "yoga", we are told, that allows him to appear to his parents in four forms at once

(9.43.36). There are other adjectives which describe his great strength, energy and valor, but this version's

stress on Skanda-Kärttikeya as a yogi endowed with great ascetic power is unusual. I suggest it is an attempt

to explain Skanda-Kärttikeya's power in an auspicious manner. His great strength and martial ability are

linked to his superior being and yogic strength. There may also be a suggestion of his relationship to Siva in

this description of Skanda-Kärttikeya. Siva is more typically described as the great ascetic and the lord of

yogins. Skanda-Kärttikeya's yogic powers do not result from lengthy asceticism or practice and, I think, we

must assume that his yogic ability is inherited from his father. Siva's tejas is described as having a great

95

energy and causing Agni, while he held it, to also have great energy and splendor (43.6-7).10 Siva's seed or

tejas seems to transfer his ascetic energy to Skanda-Kärttikeya.11 As I have already demonstrated, there was

also a belief in ancient India that a son embodies the qualities of his father. Skanda-Kārttikeya's ascetic

ability can be accounted for in this manner.

A final link the text may be trying to make by presenting Skanda-Kärttikeya as a yogin is to the

sage Sanatkumära. I have already discussed that chapter 45 ends with speculation concerning Skanda-

Kärttikeya's parentage. One of these verses is as follows: "There are some who claim that [he is] the lord

Sanatkumara, who is the supreme first born of the sacred source of all, who is the son of the paternal

grandfather" (9.45.85).12 This passage echoes one from the Chandogya Upanisad where the sage

Sanatkumära is equated with Skanda: " ... to such a man who has wiped away all stains Lord Sanatkumāra

points out the way to cross beyond darkness. It is he whom people refer to as Skanda" (7.26.2 translated by

Olivelle 1996:166). If the cited passage from the Chandogya Upanisad is not a later addition, it may well

be the earliest textual reference to Skanda. The idea presented in this quotation, however, is completely

undeveloped by the Upanisad. It is also not clear just what this text understood by the name or term Skanda.

Olivelle notes regarding the appearance of the name here: "It is, however, unclear whether the term is used

in this sense [as a reference to Skanda-Kārttikeya] in this early text" (1996:354). Olivelle seems sure here

that the passage belongs to this text, but questions the exact meaning of the name Skanda. While both the

10This process is actually an inversion or development from Vedic understandings of tejas and its relationship with Agni. In the later Vedas and some sections of the Epic tejas is regarded as a constituent element of Agni's divinity. He was considered to contain this substance, and the visible sign of this was his flaming splendor (Gonda 1957:43, 58-59). Agni also uses his tejas, which can also be translated as semen, to produce offspring like gold. As we shall see in the Anusasanaparvan, Agni's offspring produced by his tejas also display their own fiery energy and glow. Here, however, the fiery power of tejas and its ability to produce offspring have been taken over by Siva. It is now his tejas that burns Agni, who was once the embodiment of this substance. This material is likely indicative of the rise of Siva, part of which is achieved by his taking over some of the roles and powers once held by more prominent Vedic deities.

1This may suggest some relationship between tapas and tejas.

12"kecidenam vyavasyanti pitāmahasutam prabhum sanatkumāram sarvesām brahmayonim tamagrajam".

96

Chandogya Upanisad and the Salyaparvan assert an identification between the two kumäras, the exact basis

of their association is not clear to me, and it is a rare identification in the Epic material.13 In the

Śalyaparvan it seems particularly well placed because of the connections drawn in this text between

Skanda-Kärttikeya and asceticism and yogic powers. There may be an old, but largely undeveloped

connection between Skanda-Kumära and Sanatkumära, though the exact nature of the relationship, other

than the shared name, remains obscure to me.

Another inauspicious link to Skanda-Kärttikeya in the Āranyakaparvan is his associations with

ghoulish creatures. The Salyaparvan does acknowledge various ghoulish creatures who are under Skanda-

Kārttikeya's leadership, but in this version they are all in the divine army. This text provides long lists of

companions who are given to Skanda-Karttikeya by the gods to be part of his army. They are described as

having various animal faces and other terrifying anatomical features, but the text makes it clear that these

companions are not inauspicious, with one group being described as "constantly absorbed in yoga, great

souled and friendly to Brahmins" (9.44.72).14 While these beings are also described as carrying terrible

weapons and being mighty, the text seems to make it clear that these characteristics are associated with their

martial role. The beings themselves are also not referred to as Grahas, nor do they behave like Grahas. The

text also provides a long list of Mätrs who are also part of his troop, and they are treated in a similar

fashion. Most of them are described as having a terrible appearance, but they are also, as a group, called

yaśasvin, 'beautiful or famous' and kalyani'beautiful or virtuous or auspicious' (9.45.2). Their associations

with inauspicious activities are not mentioned by the text, and they appear to be part of Skanda-Kārttikeya's

army. Skanda-Kärttikeya is still associated with ghoulish creatures in this text, but their characters have

been softened, and their role is linked to the divine army. They are instruments of divine order and not

13Phyllis Granoff has pointed out to me that Skanda-Kärttikeya and Sanatkumāra are described as brothers and as sons of Agni in the Vāyupurāna (Uttara. 5.23).

14“yogayuktā mahātmānah satatam brāhmanapriyāh". Similar comments are made about these beings in 9.44.106.

97

obviously connected with harming humans.

Skanda-Kärttikeya's birth is also presented in a different fashion by the Salyaparvan. In the

Āranyakaparvan his birth caused great disruption and fear, in the Salyaparvan his birth brings beauty and

happiness to the world and the gods. He is described as being so full of luster that he causes the mountains

to be transformed into gold (9.43.14). The whole earth is described as becoming beautiful, and the

appearance of Skanda-Kärttikeya is the reason mountains produce gold (9.43.15).15 He is described as

growing up to be as handsome as the moon itself and possessing great beauty. He is praised by Gandharvas

and sages, and danced to by celestial girls (9.43.17-19). Then he is attended by a large list of gods and other

divinities (9.43.20-32). These gods ask Brahma what role should be given to this new deity. He tells them to

make him their senāpati (9.43.48).

Skanda-Kärttikeya's appointment to the generalship is a formal affair in the Salyaparvan.

Immediately after Brahma's decision to make him the senapati the gods arrange an abhiseka rite to install

him ritually to this post. The gods, led by Brahmā, take Skanda-Kārttikeya to Sailendra, the Himālayas, for

the induction ceremony and select a tirtha on the Sarasvati, which is described as flowing from the

Himālayas and called Samantapañcaka, as the spot to perform the rite (9.43.49-52). The description of the

ritual makes it clear that it follows Brahminical norms, and it is a costly affair. All the necessary equipment

for the abhiseka is collected according to the scriptures. The Brahmin priest Brhaspati performs the

necessary rites (9.44.1). Himavat, the Himālaya, gives jewels to Skanda-Kärttikeya, and he is seated in a

divine gem filled seat (9.44.2). The gods ensure that all the materials relating to the rite are present, and that

the rite is conducted in accordance with ritual rules and with the due recitation of mantras (9.44.3). All

manner of gods, sages and supernatural beings attend the rite (9.44.4-16). The rite itself involves pouring

water over Skanda-Kärttikeya, but even this is described as relating to wealth and auspiciousness. The water

15The connection between Skanda-Kärttikeya and gold is not elaborated on in this version of his birth. The Anusasanaparvan does elaborate on the relationship between the two. I will discuss that relationship in more detail in my section on that text.

98

pots themselves are gold and also full of divine materials, and the water used is the sacred water of the

Sarasvatī (9.44.18)

As I have discussed earlier, the abhiseka rite is a form of legitimation that relates to the state and a

king's court. This version of Skanda-Kärttikeya's birth and deeds focuses on this ritual as the event that

demarcates him as a general. This version does not associate any other ritual with this deity. The abhiseka is

also not a ritual like the Skandayaga or the skandasyejya, which were designed to propitiate the deity and

were performed by any class of human. The abhiseka is a one time event and is designed to initiate Skanda-

Kārttikeya into a specifically martial and courtly class. The ritual emphasizes his role as a general and

removes him from the ritual milieu of Grahas and Dhurta which did not share in this royal aspect. I have

already noted that part of the understanding we need to bring to the production of this text is as material that

is sponsored by a royal court. The image of Skanda-Kärttikeya presented in the Salyaparvan, who never

attacks Indra, and who performs his military duties without threat to the rest of the administration of the

state, would very likely meet with the approval of a sponsoring king. We need to be aware that a number of

the Brahminical means of legitimation for this deity have been dropped from the account of him; this story

has become much more focused on the military and courtly status of Skanda-Kārttikeya.

The story ends with a description of Skanda-Kärttikeya defeating various demons in battle. There

are a number of famous demons he kills in this version including Mahisa (9.45.64) from the

Äranyakaparvan and Tāraka (9.45.65), who will be his main foe in other accounts of him.16 As we might

expect, he is very successful against these demons. One account that is of interest in these battles is his

destruction of Mount Krauñca. In the ranyakaparvan Skanda-Kārttikeya destroys Mount Krauñca in a

moment of mindless violence. It is one of his early acts that makes him worthy of worship because he is

feared and dangerous. In this text, however, Vaisampāyana narrates that a demon named Bana hid in the

16In this version Skanda-Kärttikeya kills the following demons: Tāraka (9.45.64), Mahisa (9.45.65), Tripāda (9.45.65), Hrdodara (9.45.66) and Bāna (9.45.71-81). The killing of Bāna will be discussed below.

99

mountain, and that Skanda-Kārttikeya pierced the mountain to destroy that demon. Various animals and

supernatural creatures are described as running from the mountain as do hundreds of demons who are killed

by Skanda-Kārttikeya and his army (9.45.71-81). The point of the destruction or damaging of Mount

Krauñca is radically different between these two versions. In the Salyaparvan it is designed to demonstrate

Skanda-Kärttikeya's military ability and his desire to restore order by eradicating the enemies of the gods.

The destruction of Krauñca is reduced in this text to an example of 'friendly fire' and not an example of the

fear and disruption this deity causes.

The Salyaparvan presents us with a Skanda-Kärttikeya who bears little resemblance to the

dangerous and inauspicious figure described in the Aranyakaparvan. He is an auspicious figure from the

beginning of the Salyaparvan account, and his only role in this text is as Mahasena, the senapati to the

gods. As I have demonstrated in this section, the parentage of Skanda-Kärttikeya remains in the hands of

several parents, but Siva's role among those parents and his connection to this new deity have become

increasingly important. Skanda-Kärttikeya's power is accounted for in an auspicious manner, and the text

ignores his role with Grahas. The deity associated with healing and protective cults does not directly appear

in this text. Finally, the text presents us with a ritual which is focused on political legitimacy and is well

removed from the ritual life of his propitiation cult. The Salyaparvan helps us to understand that the Epic

tradition begins to settle into a role for Skanda-Kärttikeya without the apologetic found in the

Āranyakaparvan. The view that Skanda-Kārttikeya is auspicious, a general and part of 'royal' society is the

dominant version of his characterization. I would argue that the Salyaparvan represents a stage later than

the Aranyakaparvan in which the inauspicious nature of the god has little importance, and possibly one

where his cult has moved towards state or royal sponsorship.

100

3.4 Skanda-Kārttikeya's birth and deeds in the Anusasanaparvan

The third narration of the birth and deeds of Skanda-Kärttikeya found in the Mahābhārata is in the

Anusāsanaparvan (13.83-86). Like most of the early accounts of the deity, the Anusasanaparvan presents

its own understanding of this deity. The most significant difference is that this version does not establish

Siva as the primary father of Skanda-Kärttikeya, but uses the birth narrative to praise Agni as much as, or

more than, Siva. In what follows I will discuss the parentage of Skanda-Karttikeya as described in this story.

I will argue that while the story begins with the marriage of Siva and Uma, and there is significant praise for

Siva in the account, the birth of Skanda-Kärttikeya is largely presented as a vehicle for the praise of Agni

and urging of gifts of gold to the Brahminical caste. I will demonstrate that this account is heavily

influenced by Brahminical redactors who have used legends drawn largely from the Brahmana tradition to

place Skanda-Kärttikeya within a Brahminical frame of reference. This story also presents Skanda-

Kärttikeya as auspicious from birth and gives no sense of the danger associated with him from that seen in

the Aranyakaparvan. Finally, this version reduces his roles to that of divine general and destroyer of Tāraka

and does not mention his associations with Grahas or propitiation cults. This version presents the reader

with a very other-worldly deity who is born to deal with the problems of gods and demons and not worldly

problems of childhood illness and disease. The significance of these shifts and the effects of the

Brahminical redactive hand in the production of this narrative will be explored in more detail below.

3.5 The Parentage of Skanda-Kārttikeya in the Anuāsanaparvan

The story has several layers of narration. The immediate narrator of the text is Bhīsma, who

narrates the account to Yudhisthira, but he presents it as originally told by Vasistha to Parasurama in the

past. There is no textual evidence for this earlier narration, and this narrative framing of the story is likely

an attempt to make it appear as an ancient story. Bhisma's discussion comes about due to a question from

Yudhisthira concerning the origin of gold, and why gold is viewed as the best daksina.1 Bhisma's answer

1Daksinā is the gift given to a priest in return for the performance of various rituals. I will have more to say on the significance of this opening narrative below.

101

eventually leads him to a discussion of the marriage of Siva and Umā and the birth of Skanda-Kārttikeya.

Bhīsma begins after the wedding of Siva and Uma and their desire for a sexual union (13.83.40-41). The

rest of the gods find the prospect of this union very disturbing. They approach Siva and explain that the

product of a union of Siva's tapas with Uma's tapas, and of Siva's tejas with Umā's tejas (13.83.43) would

be more than the three worlds could sustain: "'O god, the offspring of you both would be powerful, O lord.

This offspring will not spare anything in the three worlds" (13.83.44).2 They ask Siva for a boon that he

will hold back his semen (tejas) and restrain from having offspring (13.83.45). Siva agrees to the boon and

becomes known as Urdhvaretas, 'he who has drawn up his semen' (13.83.46-47). This text begins with the

marriage of Siva and Uma and establishes the awesome power of the couple through the gods' apprehension

of the potential product of their union. Later the Anusasanaparvan version will, however, begin to shift its

focus away from the power of this divine couple.

Siva's boon makes Umā very angry, and she curses the gods to go without offspring themselves

(13.83.48-50). Agni, however, was not part of the group who were cursed and, even though Siva is trying to

hold in his semen, some of it spills and falls to the earth and into Agni where it begins to grow (13.83.51-

53). In the mean time, the gods are being defeated by the asura Taraka and go to Brahma to request that he

ordain his death (13.83.54-57; 13.84.1-2). Brahma sees that the Vedas and duties (dharmāh) are threatened

by this demon and ordains his death (13.84.3-4). He already knows that some of Siva's semen has landed in

Agni and predicts that Agni will pass the seed onto Ganga. Eventually a child, who he describes as "like a

second Agni" (13.84.12),3 will be born, and he will kill Tāraka (13.84.11-12). Brahmā stresses to the gods

that this child will destroy the enemies of the gods, and that they have nothing to fear. The story begins,

then, by establishing two groups of parents: Siva and Uma, and Agni and Ganga. The text appears, however,

to lean towards Agni and not Siva as the important parent because the offspring will be a second version of

2"apatyam yuvayordeva balavadbhavitā prabho tannūnam trisu lokesu na kimcicchesayisyati".

3"tattejo'gnirmahadbhūtam dvitīyamiva pāvakam," or "that semen/energy, which was Agni will become like a second Agni".

102

him. The text also establishes Skanda-Kärttikeya's divine role well before he is born. The threat Tāraka

represents is not only a threat to the gods, but also a threat to dharma and the Vedas. From the beginning of

this narrative Skanda-Kärttikeya is depicted as a defender of Brahminical values and order. At no stage in

this version of his birth story is that depiction ever placed in doubt.

The story continues as Brahma describes the virtues of Agni to the gods:

Let Agni be sought out, and let him be employed [in this task]. The death of Tāraka has been narrated by me O faultless ones .... He [Agni] is the indescribable lord of the world. He is all pervading and all creating. He abides in the heart of all creatures, and the powerful one is better than Rudra. Let him, who is the oblation eater and a mass of splendor, be quickly sought out. That god will accomplish the desire which is in your minds. [13.84.14, 17-18]4

The praises for Agni given in this quotation all echo the Satapatha Brähmana,5 but the underlying point of

the quotation is to elevate Agni above Rudra.

Both the Satapatha Brähmana and the Anusasanaparvan share something in that they both elevate

the position of Agni. What may not be so obvious is that the elevation of Agni against other deities is also a

priestly device used to elevate their own position through their role as close to Agni due to their ritual

function. Several verses from the Satapatha Brahmana directly state that Agni is a Brahmin or like a

Brahmin (2.1.4.10, 3.2.2.7, 6.1.1.10). Agni is also described as the ancestral Hotr priest in both the

4"anvisyatām vai jvalanastathā cādya niyujyatām tārakasya vadhopāyah kathito vai mayānaghāh [.14] jagatpatiranirdesyah sarvagah sarvabhāvanah hrcchayah sarvabhūtānām jyestho rudrādapi prabhuh [.17]anvisyatām sa tu ksipram tejorāśirhutāśanah sa vo manogatah kāmam devah sampādayisyati [.18]".

5That text refers to Agni and his forms as the lord of the earth and the lord of beings (1.3.3.16), and it frequently presents Agni as the world itself, the source of all things and within all things: "Agni (the fire), assuredly, represents all the deities, since it is in the fire that they make offering to all deities ... " (1.6.2.8 translated by Eggeling 1882-1900 I:163), and "He then makes offering to Agni, the householder. Agni, indeed, is this world ... " (1.9.2.13 translated by Eggeling 1882-1900 I:259). A similar point is made in 6.1.1.1-11 where Agni is equated with Prajäpati and the Vedas. He is also described as the first thing created and is the basis of the material and Vedic (spiritual) world. While the texts say the Vedas underlie the world/everything, it is in fact Agni who is in everything according to this text. Section 2.2.2.8-20 also states that the gods internalize Fire because he is the immortal element which they place in their innermost being. Humans, the text states, can also internalize Agni and place him in their innermost being. These points likely have their root in the Rgveda: "[Agni is] he who is the immortal in morals, the keeper of order, the Hotr priest" (1.77.1) (yo martyesu amrta rtava hotr), and "Agni is the immortal that is fixed in morals" (7.4.4 quoted in V. S. Agrawala 1961:28). The Anusasanaparvan description of Agni appears to come out of a priestly and sacrificial mileau similar to that of the Satapatha Brāhmana.

103

Satapatha Brāhmana and the Rgveda (RV 1.77.1; ŚB 1.4.2.3). There is a symbiotic relationship between

the priestly caste and the Fire god in the Satapatha Brähmana. The elevation of Agni also results in the

elevation of his earthly representatives. What we need to be aware of as we read this Anusasanaparvan

version of Skanda-Kärttikeya's birth and its elevation of Agni is the reflection of Brahminical concerns and

agenda in this account. Through the Satapatha Brahmana we can also see that this priestly influence is one

that involves the great sacrifices of the 'high' tradition of the Vedas. The lense through which this story is

redacted and written is a Brahminical one from the sacrificial tradition.

The above quotation from the Anusasanaparvan ends with Brahma instructing the gods to find

Agni. From here the narrative shifts to a topic with which we are familiar: the hidden Agni theme. Agni

hides from the gods, and they must search him out (13.84.22-46).6 Eventually, the gods do find Agni and

ask him to produce a child. I have discussed earlier the missing Agni theme, and how it relates to the birth

of Skanda-Karttikeya. The use of this theme in the Anusasanaparvan fulfills a similar function as it did in

the Aranyakaparvan version of the story. This version of the story is still concerned with the birth of this

deity and uses Agni as a medium through which his birth can be explained and set within a Vedic context.

The text proceeds with Agni placing the seed into Gangā (13.84.52-53). Gangā has great trouble

carrying this foetus because of its great energy. She can no longer bear to carry it and casts it off. The

energy of the seed and Ganga's trouble with it are well described in the scene where she casts it off:

Then that best of rivers, while checked by Agni and by the gods, discharged that foetus onto Meru, the best of mountains. Although she was able to bear it, she was overwhelmed by the splendor [tejas] of Rudra and was unable to endure that foetus with such energy [ojas]. In pain she discharged that powerful fetus that was like a blazing fire .... (13.84.64-66)'

6Wendy Doniger regards the hidden Agni theme from this section "as a conscious multiform within the story of Skanda's birth" (1973:285). She regards his hiding in the waters as a multiform of his placing of his seed in Ganga and as related to a theme of distribution (1973:285). I agree with her that the hidden Agni theme does reoccur in a number of these versions of the birth story, but, as I have argued, the point of this theme is to link Skanda-Kärttikeya to Agni and to justify his birth.

1“sā vahninā vāryamānā devaiścāpi saridvarā samutsasarja tam garbham merau girivare tadā [.64] samarthā dhārane cāpi rudratejahpradharsitā nāśakattam tadā garbham samdhārayitumojasā [.65] sā samutsrjya tam duhkhāddīptavaiśvānaraprabham [.66 a-b]". My thanks to Phyllis Granoff for her

104

The term for "fire" here is vaiśvānara, which means 'fire,' but more specifically refers to a form of Agni

called Vaiśvānara (belonging to all men, or omnipresent). The text seems to combine the energies of Rudra

and Agni in this foetus as we are told that Ganga cannot withstand the tejas of Rudra that is like the

powerful blaze of Vaiśvānara. Certainly, Skanda-Kārttikeya as the product of this meeting of energies will

be powerful, but the text also eulogizes the power of Siva and Agni by making their semen so powerful that

even Ganga cannot withstand it. Ganga is often praised for her power and ability to carry things. Even in

this story Agni, in trying to persuade her to hold onto the foetus, reminds her that she is capable of bearing

the whole world (13.84.63). That she cannot stand to carry this foetus speaks of the blazing power of it, but

also of the power of those who created it: Siva and Agni. Skanda-Kärttikeya's power, energy and might are

not his own in this version of the story, but off-shoots of his fathers' powers. This story is subtly not so

much about the glory of Skanda-Kärttikeya as it is about the glory of Agni and Siva.

The foetus appears to survive its release from Ganga. She describes its appearance to Agni in the

following manner:

Gangā said: That foetus just born is golden, it is like you, O Fire, in its energy. It is golden, pure and blazing, and it illuminated the mountain. O best one, that smell of him is cool, like that of lakes together with water lilies and lotuses, and equal to Kadamba blossoms, O best of those who burn. By the energy of that shining foetus, just like the rays of the sun, those objects on the earth or on the mountain which are touched appeared [to be] golden from his contact. (13.84.68-70)8

While the previous section praises both the power of Siva and Agni, this section regards Skanda-Kārttikeya

as a version of Agni. Ganga repeatedly refers to the foetus as golden and as very bright due to its tejas or

energy. It glows so much it is described as being like the sun (sūrya) or a second moon (soma) (13.84.72).

assistance in translating this passage.

8"gangovāca jātarūpah sa garbho vai tejasā tvamivānala suvarņo vimalo dīptah parvatam cāvabhāsayat [.68] padmotpalavimiśrānām hradānāmiva śīttalah gandho'sya sa kadambānām tulyo vai tapatām vara [.69] tejasā tasya garbhasya bhāskarasyeva raśmibhih yaddravyam parisamsrstam prthivyām parvatesu vā tatsarvam kāncanībhūtam samantātpratyadrsyata [.70]". My thanks to Phyllis Granoff for her help in this translation.

105

Once the foetus is produced Agni and Ganga leave (13.84.72-73), and the foetus, which is called

"pāvakodbhavah", or 'born from Agni' and "gangeyah", or 'from Gangā', develops in a divine forest of

reeds (13.84.75).9 There he is found by the Krttikās who raise him as their son. Hence, we are told, he is

called Kārttikeya after the Krttikās, and because he fell (skanna) from Rudra he is called Skanda, and

because his birth took place in a forest of reeds hidden (guha) from view he is called Guha (13.84.76-77).

3.6 Agni's parentage of Skanda-Karttikeya and gold in the Anusasanaparvan

The close of chapter 84 involves an awkward shift from the birth of Skanda-Kārttikeya to the

production of gold from the same event. Just after the section explaining the various names of Skanda-

Kārttikeya Bhīsma states:

Thus, there was formed gold, the son of Agni. Of all forms of gold the most splendid one is from Jāmbūnada10 which is even the decoration of the gods. From that time on it was called 'just born'. That which is gold that is the glorious Agni, who is the supreme one, Prajapati [lord of creatures] Gold is the purifier of purifiers, O best of the twice born. It contains the nature of Agni and Soma, and is known as Jatarupa. It is the best of jewels, the best of ornaments, and the purest of the pure, the most auspicious of the auspicious. (13.84.78-81)1

The transition from explaining Skanda-Kärttikeya's names to an account of the nature and wonder of gold is

abrupt in this text. We might view Skanda-Kärttikeya as gold, which is the sense Wendy Doniger takes from

the above (1973:108). In the above description of the foetus given by Ganga she states that it is golden and

makes the things around it appear to be gold. The only birth mentioned in this text is Skanda-Kārttikeya's,

9A foetus developing in reeds is also linked to Agni. The Satapatha Brähmana describes reeds as a womb for Agni (6.3.1.26), but a number of other plants are also described as Agni's garbha (womb) in this and other texts (Gonda 1957:93-94). Agni himself is also described as a garbha in some texts (Gonda 1957:94). Doniger also regards Skanda-Karttikeya's birth in reeds as part of Agni's mythology, and she notes that a reed is also a hiding place for Agni and Indra in several stories (1973:97-98).

10The word jambūnada means coming from the river Jambu, but it also refers to gold obtained from that river (Monier-Williams 1999:419).

11"evam suvarnamutpannamapatyam jātavedasah tatra jāmbūnadam śrestham devānāmapi bhūsanam [.78] tatah prabhrti cāpyetajjātarūpamudāhrtam yatsuvarnam sa bhagavānagnirīsah prajāpatih [.79] pavitrānām pavitram hi kanakam dvijasattama agnīsomātmakam caiva jātarūpamudāhrtam [.80] ratnānāmuttamam ratnam bhūsanānām tathottamam pavitram ca pavitrānām mangalānām ca mangalam [.81]". My thanks to Phyllis Granoff for her help with this translation.

106

leading to the assumption that Skanda-Kärttikeya is gold or the cause of gold's appearance. The text does

not directly make this connection, however. It is also possible that the connection between gold and Skanda-

Kārttikeya relates to their mutual father, Agni. As we shall see, in Vedic texts Agni is linked to gold through

his colour and brightness, but also gold is described as a product of Agni's semen, his son. The same idea is

also expressed in this section of the Anuśsanaparvan.

That gold is a product of Agni's semen is established in Vedic texts. In the Satapatha Brahmana

we are told that when setting up the fires one must equip Agni with objects which have some of his nature,

specifically, the text advises supplying the fire with splendor, cattle and a mate (2.1.1.1). The mate is water;

the splendor is gold (2.1.1.4-5). To explain the use of these substances to the reader the text states: "Now

Agni at one time cast his eyes on the waters: 'May I pair with them,' he thought. He came together with

them; and his seed became gold. For this reason the latter shines like fire, it being Agni's seed" (2.1.1.5

translated by Eggeling 1882-1900 I:277-278).12 The text continues to explain this is why gold is found in

water (2.1.1.5). The Anusäsanaparvan makes a similar point concerning the parentage of gold: "It is also

held that gold is the offspring of Agni" (13.85.54c-d).13 The above account from the Satapatha Brahmana

helps to explain two rather obscure elements of the Anuasanaparvan account: why Skanda-Kärttikeya and

gold are associated with each other, and why Agni and Ganga are paired up as a couple. Gold and Skanda-

Kärttikeya are related because they are both products of Agni's semen or tejas. The implications of this

relationship will be discussed below. The relationship between Agni and the waters above is also of value

because it suggests an explanation as to why Ganga becomes a mother of Skanda-Kärttikeya. The evidence

from the Satapatha Brahmana suggests that her inclusion in Skanda-Karttikeya's story is based on the

legend cited above. She is the water with which Agni couples to produce gold and, in the Anusasanaparvan,

12The point that Agni's seed is gold is repeated in ŚB 3.2.4.8-9, 3.3.1.3, 3.3.2.2, 4.5.1.14, 5.2.3.6, 5.5.1.8, 7.4.1.15 and 14.1.3.14. A similar myth is recounted in the Taittiriya Brähmana (1.1.3.8), where the waters are called Varuna's wives (Gonda 1991:16). Gonda also notes that gold is identified with Agni's semen or retas in TB 1.2.1.4 and JB 1.56 (1991:16).

13"agnerapatyametadvai suvarnamiti dhāranā".

107

Skanda-Kārttikeya.

As the offspring of Agni, we should expect gold to carry some of its father's characteristics, and

the Vedic tradition does not disappoint our expectations. The Satapatha Brahmana notes that gold is pure

(3.2.4.14), auspicious (3.2.4.14), and there are several references to its shine (3.2.4.14, 7.4.1.15). Jan

Gonda notes that gold is identified with tejas as a "fiery energy or efficacy" in the following texts: TS

5.7.5.2, 5.7.9.4; MS 1.11.8:170.6; KS 14.8:207.17, 21.7:46.11; TB 3.12.5.12, 2.7.9.3 (1991:18). Gold is so

bright that it is often compared with the sun in Vedic texts (Gonda 1991:8, 54). Gold is often used in the

sacrifice itself and is linked to satya (truth or reality) (Gonda 1991:19-20). All of these characteristics are

also shared by Agni. Agni is understood to contain tejas and to shine brightly (Gonda 1957:43). His shine is

also often compared with gold: "O Agni, shining like gold, resembling gold" (RV 6.16.38), and "like

clarified ghee, spotless is (his) body, brilliant (radiant, pure, suci) gold; that of thine is radiant (splendid,

shines) like a golden ornament (rukma)" (RV 4.10.6 translated by Gonda 1991:14). As the above notes,

Agni is also pure, and he is also called Pavaka, the purifier (SB 2.2.1.11). He is the auspicious ritual fire

who, like gold, is also linked to satya (Gonda 1991:19-20). Gold is not just a precious metal, but something

imbued with Agni's qualities; a virtual form of Agni.

Gold as a form of Agni is also a theme in the Anusasanaparvan. We are told that gold can be used

as a substitute for Agni in the sacrifice (13.85.54-56). The Anusäsanaparvan also states that Agni is

regarded as the refuge of the gods; the producer of the divinities (13.85.53-54), and: "from that it is heard

by us that when one who sees the dharma gives gold, he gives all of the deities" (13.85.58).14 Agni is

regularly described as embodying all of the gods or as representing all of the gods (ŚB 1.6.2.8, 3.1.3.1,

5.2.3.6), and this quality seems to be transferred to his offspring, gold.

We can argue that what is true for Agni and gold should also be true for Agni and Skanda-

Kārttikeya. In the Anusāsanaparvan Skanda-Kārttikeya is presented as Agni's son and, like gold, as imbued

14“tasmādyo vai prayacchanti suvarnam dharmadarsinah devatāste prayacchanti samastā iti nah śrutam"

108

with the characteristics of his father. As brothers of sorts, the text also establishes a relationship between

gold and Skanda-Kärttikeya. As we have seen, both Skanda-Kärttikeya and gold are described as pure,

auspicious sons of Agni. They are physically like fire in their brilliance and remarkable shine. Gold is

described as having the essence of Soma and Agni, and Skanda-Kärttikeya is also described as being like

these gods (13.84.72, 13.84.68). The two are related and, we are to understand, of the same stuff, the same

tejas. The character and appearance of Agni are reflected in all of these characteristics of his sons. There is

no ambiguity in the Satapatha Brahmana or in the Anusasanaparvan concerning Agni's characterization.

He is good, holy and auspicious, and those things he produces contain these same virtuous qualities. The

result in the Anuāsanaparvan is a specific version of Skanda-Kārttikeya who is unambiguously good,

moral and auspicious because he is the son of Agni and not because he is worshiped or propitiated. We

must recall that part of Skanda-Karttikeya's ritual cult as displayed in the Aranyakaparvan was designed to

transform an inauspicious deity into an auspicious one. This version of his birth does not allow for such a

ritual transformation; the implications of this will be discussed in more detail below.15

I would also like briefly to comment on the choice of gold as Skanda-Kārttikeya's colour as

opposed to red, which is the colour of his skin in many Puranic and Ayurvedic accounts of him. All three

versions of his birth and deeds from the Mahabhārata make some reference to this deity as gold, golden or

wearing gold. The Skandayäga also describes him as "golden coloured" (suvarnavarno) (20.2.8), but also

states he is "red limbed" (lohitagātrāya) (20.4.2). The Mahābhārata does associate him with red (3.214.19,

15Wendy Doniger argues that the origins of the golden seed is found in the golden seed of Brahmā/Prajāpati called hiranyagarbha, and this develops into the golden egg and then the god of the golden seed (1973:107). The concept of hiranyagarbha (the golden germ) has a long history in South Asian thought beginning with the Rgveda (10.121). Jan Gonda has a detailed study of the term which largely agrees with Doniger's summary above, except he also stresses the identification of Prajāpati with Agni, and that is how Agni becomes associated with the golden seed the origins of the universe (1991:216-246). Doniger also claims that "gold forms a constant tie between Agni Siva in the Skanda story", she regards a number of Saivite stories involving gold to be "multiforms of the Skanda myth" (1973:108). I have some reservations about this 'multiform' idea; I think it is too reductive in nature. I do agree, however, that the golden seed remains an important concept in Puranic accounts of this birth story where it is Siva's seed that is golden (Doniger 1973:108).

109

3.218.2), but never acknowledges it as his skin colour. The Suśrutasamhitā calls Skanda-Kārttikeya red and

mentions that his clothes and ornaments are also red (Uttara. 28.14), and most Puranic accounts refer to his

colour as red. The Epic's choice of gold as Skanda-Kärttikeya's colour may not be random, however,

because gold is an auspicious colour, and red is often inauspicious. Some of the auspiciousness of gold is

seen in accounts of Skanda-Kärttikeya. In the Aranyakaparvan version of his story his golden armor and

golden colour are only mentioned once he is auspicious. The Salyaparvan and Anusasanaparvan versions

begin with the premise that the god is auspicious and establish his golden appearance early. Gold is also a

royal colour. Being hiranyavarna or suvarnavarna (gold-coloured) "is considered to be one of the

characteristics of a nobleman or royal personage" (Gonda 1991:21). As with the abhiseka, being gold

coloured has an elite and royal connotation. Not only does the colour gold demarcate Skanda-Karttikeya as

pure and auspicious, it also marks him as royal or connected to royalty. It is an elite material and colour,

and Skanda-Karttikeya's status as the material or very like it connects him to royal and wealthy circles.

Gold's connection to royalty is also made in the Satapatha Brahmana. In that text gold, as a

priest's fee, is only mentioned in relation to the Räjasūya sacrifice (5.2.3.6), which is an elaborate rite to

inaugurate a new king. The Rajasuya is a complex rite involving several different sacrifices, and the entire

rite could run as long as two years (Eggeling 1882-1900 III:xxvi). Kings could afford gold as a price for

such an extended ritual, but paying gold for everyday rites of protection like those described in the

Skandayaga is unlikely. One of the points I wish to draw from this material concerning gold is its

connection to royality and wealth, and the implication that Skanda-Kārttikeya has been joined with this

elitist group of society. There is something exclusive about gold and the Skanda-Kārttikeya that appears in

this version of his story that was not present in the early stages of the Aranyakaparvan or in the

Skandayaga. The social connections of this deity are being shifted from a god of creatures of all colours

with a concern for protection and the health of their children, to royal and wealthy society and a concern for

protection in the form of an organized army and its leadership, a leadership which is a reflection of

themselves more than everyday society. Who this deity is associated with in the various accounts of him

110

matters. Here he is associated with elite members of society, in part through his connection with gold.

Red, on the other hand, places Skanda-Kärttikeya into rather a suspect context. Red is often

associated with impurity and inauspiciousness. In the Atharvaveda Parisista when one is casting a

malicious spell on someone else the rules of the ritual as regards purity and auspiciousness are reversed.

One selects an inauspicious day and time for the rite. Offerings are made with the left hand while facing

south, and the priest should wear black or red clothes (AVP 31.9.3; Modak 1993:315). Here, wearing red is

connected with a number of ritual taboos that are normally considered as inauspicious ritual actions. Gonda

also notes that a number of Grhyasutra and other non-solemn rites employ red in rites with malevolent

intents (1980:45). Red is also the colour of blood, which is largely regarded as impure (Gonda 1980:45).

We have already seen that blood is related to a number of the inauspicious Matrs from the Aranyakaparvan.

Blood is one of the substances that, if touched, the Dharmasūtra of Apastamba requires a Brahmin to purify

himself (1.16.14).16 Red beings or beings associated with red are often also inauspicious or dangerous as

with the example of the creation of 'Fever' in the Mahabharata. The story goes that when Siva destroyed

Daksa's sacrifice a drop of sweat fell from him and "became a great fire like the doomsday fire; then it

became a man named Fever, short, red-eyed, red-bearded, hair standing on end, very hairy, dark-skinned,

wearing red garments" (Doniger 1973:284). The gods fear that if fever is kept whole the earth would not be

able to withstand Fever, so Siva divides him up into various ailments (Doniger 1973:284).17 The point, I

think, is clear: red is associated with impure substances and dangerous beings. Skanda-Kārttikeya as a red

coloured being fits his Graha-like character, but not his auspicious general character; this may be behind the

16Several other Dharmasutras have rules regarding the ritual cleaning or destruction of household items and sacrificial utensils if they come into contact with blood (DB 1.6.5, 1.8.48, 1.13.11, 1.13.28, 1.13.32, 1.14.3, 1.14.6; DV 3.59).

17Not only is this being's colour like Skanda-Kārttikeya's, but his Graha-like ability to cause illness also relates him to Skanda-Kārttikeya.

111

Epic's choice of gold for his skin colour.18

In the Anusasanaparvan gifts of gold are described as attaining for the giver: heaven, protection

from evil foreshadowed in dreams, destruction of future misdeeds, residence with Brahmā, Vayu, Agni and

Soma, great fame, glory and the fruition of wishes (13.85.59-66). The ability of gold to bring about these

fantastic returns, if given away as a priestly gift, is justified by its close association with Agni. The point of

these passages is to re-enforce the idea that gold is the best gift to give to a priest as payment for his duties.

The section largely reads as a legitimation of an expensive gift to Brahmins by making gold something

divine. Certainly, I think, we can see the redactive hand of the Brahminical caste behind this account.

Chapter 84 ends with a praise for gold and with the embryo becoming the son of the Krttikās.

Chapter 85 is rarely discussed by other Skanda-Kärttikeya scholars because it does not immediately

continue the story of the birth. It does not mention the birth of the deity and appears to be an awkward

interruption in the narrative. I think, however, it functions in a similar manner to chapters 207 to 212 of the

Åranyakaparvan. It is designed to explain divine birth and creation through the Vedic example of Agni.

Chapter 85 is also an apologetic for the multiple parentage of Skanda-Kārttikeya. It presents multiple

parentage as a normative process in divine creation stories.

The text begins when Vasistha tells Parasuräma that in former times Rudra, who had taken the

form of Varuna, held a grand sacrifice which all the gods and sages attended. Even personified forms of the

Vedas and elements of the sacrifice are present at the rite (13.85.2-6). The wives, daughters and mothers of

the gods are also present, and, on seeing them, some of Brahma's (svayambhuva) semen (retas) falls on the

earth. The Vedic deity Püsan then picks up the semen, mixes it with soil and throws it into Agni, the

sacrificial fire. We are also told that part of this semen is also ladled into the fire as part of the offering of

the rite (13.85.7-11).

A number of the elements of this account have a Vedic background. The narrative of Brahmã

18Red is not always inauspicious, however. Gonda also notes that red is a colour of fertility, sexual love and reproduction in Grhya rites (1980:45).

112

dropping some semen which is added to the sacrificial fire is likely rooted in Vedic accounts of Prajāpati

having some of his semen fall. In the Satapatha Brähmana the story runs that Prajäpati has sex with his

daughter Usas. The gods ask Rudra to punish him, and Rudra hits him with an arrow causing some of his

semen to fall to the ground (1.7.4.1-3). Various deities try to deal with the semen so that it can be made a

part of the sacrifice. Pūsan is one such deity; he tries to eat the semen, but it knocks his teeth out (1.7.4.7).

Certainly, part of the Anusasanaparvan account is a modified version of this Satapatha Brahmana account.

The ladling of semen into the fire is also a common metaphor in the Satapatha Brahmana, which evokes the

reproductive theme of the sacrificial cult (1.7.2.11, 6.2.2.27, 6.3.3.18).

The products of this sacrifice are the gunas, the three elements that make up the universe in

Samkhya philosophy. The passages are not in complete agreement with Samkhya thought in that only tamas

and sattva are explicitly mentioned, though the place of rajas may be taken by tejas in this text (13.85.12-

14). Vasistha then states that three ancient sages, Bhrgu, Angiras and Kavi, are produced from the

sacrificial fire. Other famous sages and deities are also produced after these three from the ashes of Agni.

The deities called the Asvins appear from Agni's eyes, from his ears the Prajapatis, from his pores the Rsis

and from his sweat the Chandas, a Vedic meter. The list continues until most of the Vedic sages and minor

divinities are described as originating from parts of Agni's body (13.85.15-24). Due to all of this,

Paraśurāma is told: "For this reason, sages, who are accomplished in sacred knowledge, due to an

examination of evidence from the Vedas, say that Agni is all the divinities" (13.85.20).19

While the text does not draw a direct analogy to the birth of Skanda-Kärttikeya in this section, the

implications of this account for that birth story are clear. Skanda-Kärttikeya's birth as the result of semen

being dropped into Agni is given a Vedic precedent by this account. Skanda-Kārttikeya's birth is not

unusual in this account of his birth, but part of normal creation consistent with Vedic accounts. The use of

these precedents shows the influence of Brahmana and Brahminical thought on this text in an attempt to

19"etasmātkāranādāhuragnim sarvāstu devatāh rsayah śrutasampannā vedaprāmānyadarsanāt".

113

place Skanda-Kärttikeya within a familiar frame-work.

There is, however, in this Epic narrative some dispute over who owns the products of this sacrifice.

Varuna claims that it is his sacrifice, and the products of it belong to him and should be regarded as his

offspring (13.85.25-26). Agni claims that they are his offspring because they came from him (13.85.27),

and Brahmā claims it was his semen so they belong to him (13.85.28-29). Eventually, Bhrgu is given to

Varuna, Angiras to Agni and Kavi to Brahmā (13.85.30-35). The sons of the sages are then listed and

described as forming the first families and tribes. They are destined to populate the world and become great

men who are learned in the Vedas (13.85.37-53). It is clear from this narrative that having multiple claims

to parentage is not unusual where divine creation is concerned. The text is attempting to deal with the

difficult issue of Skanda-Kärttikeya's multiple parents by supplying a myth that demonstrates a similar

scenario of multiple claims to parenthood; a scenario which is legitimated through the context of the

sacrifice.

The text makes the disagreement between these gods over whose children these sacrificial products

are moot when it ends this section with: "Thus it happened at the sacrifice of that great souled one, the best

of gods who had taken the form of Varuna. Agni is Brahma, he is Pasupati, he is Sarva, he is Rudra, and he

is Prajāpati" (13.85.53-54a-b).20 The text, then, establishes that Agni really is Rudra and a variety of other

creator gods. Again, similar relationships are found in the Satapatha Brähmana between Agni and

Prajāpati, and Agni and Rudra (6.1.1.11, 9.1.1.1).21 It is Agni, the text concludes, who is the source of all

this creation; a view which is likely strongly influenced by the Brahmanical and Brahminical traditions. We

have seen the equation of Agni and Rudra-Siva in the Aranyakaparvan, but the point in that text was the

opposite of what we have found here. In that text the equation was to promote the idea that Siva was the real

20“evametatpurā vrttam tasya yajñe mahātmanah devaśresthasya lokādau vārunīm bibhratastanum [.53] agnirbrahmā pasupatih šarvo rudrah prajāpatih [.54a-b]". My thanks to Phyllis Granoff for her assistance in translating this passage.

21 Agni and Rudra are also assimilated in Tāndya Brāhmana (12.4.24) and Taittirīya Brāhmana (1.1.5.8-9, 1.1.6.6, 1.1.8.4) (V. S. Agrawala 1961:29).

114

father of Skanda-Karttikeya, who was working through the person of Agni. In the Anusasanaparvan,

however, the equation is used to establish Agni's supremacy. Here it is really Agni who works through Siva

and others. The use of Agni as Rudra in this text indicates that the cult of Agni was the preferred medium

through which Skanda-Kärttikeya's birth was viewed by some Brahminical writers during the Epic period.

While Siva will ultimately replace Agni as the ultimate father of Skanda-Kärttikeya, and part of that process

was already underway in these texts, Agni is clearly viewed as an important deity by some Epic redactors.

He also provides a useful means of explaining Skanda-Kärttikeya's birth and multiple parentage. Chapter 85

should not be ignored, but is a part of the birth story narrative in that it attempts to place the narrative within

a Vedic or Brahminical context and gives us some insight into the thinking of the writers and redactors of

this text.

Chapter 86 of the Anusasanaparvan returns to the theme of Skanda-Kärttikeya's birth and

destruction of Täraka. The chapter begins with Yudhisthira asking Bhīsma to narrate the killing of Tāraka

(13.86.1-4), and it is clear that this version of Skanda-Kärttikeya's deeds is entirely focused on this single

act. The text picks up the birth story with the gods and Rsis asking the six Krttikās to incubate the foetus

because no other divinity was capable of carrying that foetus which was endowed with such energy

(13.86.5-6). Bhīsma then states that Agni had released his semen (13.86.7), and the six Krttikas nourished it

(13.86.7-8).22 Here the text emphasizes the role of Agni as the father of the child over Siva. The text is clear

in stating that it is Agni's tejas which the Krttikas nourish. The foetus is nurtured by these six even though

the developing foetus and its energy (tejas) makes them unable to find comfort anywhere (13.86.8-9).

Eventually, the child is born to the Krttikās (13.86.11).

The story continues with a description of the child as full of splendor, and he is compared with

22Here the text seems to forget that Agni has already discharged his or Siva's semen into Gangā, and that a foetus has already been produced and left in the reeds. We might regard chapter 86 as a separate narrative of the birth of Skanda-Kärttikeya due to these plot inconsistencies. While I acknowledge the existence of these inconsistent elements, such inconsistencies are not unusual in the Epic or within individual stories in the text. I think the best way to read chapters 83-86 is as one account of the birth, which shows the influence of several layers of redaction and chronological development.

115

Agni and the sun. The affection of the Krttikas towards him is also briefly described (13.86.12-13). Hymns

of praise are sung to him, and the gods present him with various gifts. He is described here as six-headed

and twelve-armed as well as being devoted to Brahmins, or dear to Brahmins (dvijapriyam) (13.86.17-24).

Throughout this section the text continually reminds us that this new god is created to destroy Tāraka: "The

gods along with the sages seeing him lying in a clump of reeds obtained the highest joy and regarded the

demon as dead" (13.86.19).23 The demon is Taraka; this story continually reminds us that killing him is

Skanda-Kārttikeya's only role.

The final element of the story is a brief description of the slaying of Tāraka. Bhīsma states that

while Täraka sought ways of killing Skanda-Kärttikeya, he was never able to do so. Eventually Skanda-

Kärttikeya is honoured (pujayitva)24 by being given the generalship (senapatya) and told of the gods'

trouble with Täraka (13.86.26-27). Dutifully, Skanda-Kärttikeya kills the demon with his spear (śakti) and

is able to re-establish Indra as the ruler of the three worlds. In the closing verses we are told:

That Skanda, who was that general, was resplendent. He was full of valor, a lord, the protector of the gods and pleasing to Sankara [Siva]. He is of golden form, the blessed one, that son of Agni, ever youthful, he obtained the position of lord of the army of the gods. Therefore gold is auspicious, the best, indestructible gem, born together with Kärttikeya the best vital energy [seed] of fire. Thus, O descendant of Kuru, Vasistha narrated to Räma in the past and from it you, O king of men, must give gold as a gift. By giving gold Rama was set free from all of his offences and attained a high place which is rare and difficult to attain by men in the third heaven. (13.86.30- 34)25

Skanda-Kārttikeya is glorified in this closing section, but so is gold and its ability to free people from sin

when given as a priestly gift. As the story closes we are reminded again of the Brahminical hand in the

23“ śayānam saragulmastham drstvā devāh saharşibhih lebhire paramam harsam menire cāsuram hatam".

24There is no suggestion here of the abhiseka rite described in the Salyaparvan.

25“sa senāpatirevātha babhau skandah pratāpavān īšo goptā ca devānām priyakrcchankarasya ca [30] hiranyamūrtirbhagavānesa eva ca pāvakih sadā kumāro devānām senāpatyamvāptavān [31] tasmātsuvarnam mangalyam ratnamaksayyamuttamam sahajam kārttikeyasya vahnestejah param matam [32] evam rāmāya kauravya vasistho'kathayatpurā tasmātsuvarnadānāya prayatasva narādhipa [33] rāmah suvarnam dattvā hi vimuktah sarvakilbisaih trivistape mahatsthānamavāpāsulabham naraih [34]".

116

production of the narrative and the link between the dual products of Agni's tejas: the pure and auspicious

gold and, by extension, the equally pure Skanda-Kārttikeya.

3.6 Skanda-Karttikeya in the Rāmāyana

In this section I will briefly deal with the story of Skanda-Kārttikeya as it appears in the Rāmāyana

(1.35-36). The story of Skanda-Karttikeya receives only a short mention in this text and bears a close

similarity to the Anusasanaparvan version. The narrative begins with the story of Uma's marriage to Siva.

The two have yet to have children, but the gods are anxious about the potential power of their offspring.

They are able to convince Siva to retire to a life of penance with Uma (1.35.1-16). Some of Siva's semen

does, however, spill and falls to the earth, and, we are told:

Then the gods spoke to Agni, the eater of oblations, 'You and Vayu must enter Rudra's abundant semen.' Permeated by Agni, it was transformed into a white mountain on which there was a celestial thicket of white reeds that looked like the sun surrounded by fire. It was there that Kärtikeya came into being, born from fire. (1.35.17-18 translated by Goldman 1984 I:192)

The chapter ends with a very angry Uma cursing the gods to go without offspring themselves.

The next chapter begins, having apparently forgotten that Skanda-Kārttikeya has already been

born, with the gods asking Brahma for help because Siva has left them to do penances. Specifically, the

gods need a leader for their army because, according to this text, that was the role Siva fulfilled (1.36.3).

The solution Brahma offers is to have Agni and Ganga bear a son "who will be a foe-conquering

commander for the army of the gods" (1.36.7 translated by Goldman 1984 I:194). Things proceed as

predicted, but Ganga is unable to bear the splendour of Agni and is unable to hold the embryo. Agni has her

drop the embryo on the Himalayas, a birth which is also described as the birth of gold (1.36.14-18). Not

only is gold produced with this embryo, but a number of other metals as everything the shining embryo

touches is transformed: "Thus, when it touched the earth, it turned into the various elements. The moment

the embryo was set down, the whole mountain forest was pervaded by its splendor and turned to gold"

(1.36.20-21 translated by Goldman 1984 I:194-95). From this splendour Skanda-Kärttikeya is born who is

given to the care of the Krttikas. In short time he defeats demons is battle and is made the general of the

117

army of the gods (1.36.22-30).

This version of the birth story goes into minimal detail on these points, but it is clearly related to

the Anusasanaparvan version. Both begin with the marriage of Uma and Siva, and have the creation of

Skanda-Kärttikeya derive from the spilling of Siva's semen. Both versions also then forget these events and

have the child become the product of Agni and Ganga, a union which also produces gold. The point of both

birth stories is also similar. Both are focused on using the birth of Skanda-Karttikeya to demonstrate the

power of Uma, Siva and Agni. He is also born to fulfil the specific role of becoming a divine general

defeating the enemies of the gods and has only a slight association with the world through his and gold's

birth on the Himalayas. Any connection he may have to Grahas or Mätrs or a non-elite cult is ignored. It is,

however, instructive to note that the Saivite version of the story has yet to dominate the narratives of this

deity as they will in texts like the Kumarasambhava. The narrative has a final step to take, but for our

purposes the narrative of an auspicious divine general whose birth illustrates the power of his parents has

taken root and, with the exception of a few Ayurvedic accounts of the deity, will not change significantly.

Skanda-Kärttikeya has settled into the spot created for him by Brahminical redactors.

3.7 Conclusion to Epic Material

The Epic accounts of Skanda-Kärttikeya can be read as an extended conversion narrative. The

focus of each account is to depict an auspicious deity, who is both a good son and good general. I have

isolated two groups or forces that bring about this change. The first is Brahminical. The Brahminical

tradition attempts through these narratives of Skanda-Kärttikeya's birth to establish him as a supporter of

Brahminical ideals and as subject to rituals. He likely begins as a deity who is outside of Brahminical ritual

practice, a demon of childhood disease. He is, by the end of these stories, and in some cases from the

beginning of them, brahmanya, Brahminic. The second is royal or courtly. Part of the auspicious

characterization of Skanda-Kärttikeya comes about with a shift in his associations with social groups. He

moves from being an everyman type of deity to one associated with elite duties. This shift is, in part, made

possible through the Brahminization of the deity. Once he is absorbed into the orthodox Hindu fold he

118

becomes a suitable object of royal attention. His image as an ideal general and as subservient to his

superiors is one, I have suggested, that may even have been influenced by actual socio-political concerns of

the kings who sponsored the Epic.

This shift in characterization also affects Skanda-Kārttikeya's realm of action. As a dangerous

deity his realm of action is the earth; he is a household deity who reflects household concerns. Texts like the

Āranyakaparvan, Skandayāga, Suśrutasamhitā and Parakagrhyasūtra demonstrate this household nature of

the deity and the terrestrial aspect of his cult. Once made the divine general of the gods, however, his realm

of action becomes heaven and to some degree the elite context of a royal court. His cult becomes one based

in an army and royalty, and his stories involve the destruction of demons who have little impact on the

everyday or the household life of most humans. The strongest evidence of a ritual cult for this deity from

any of these texts depicts a propitiatory cult. Once removed from a terrestrial and inauspicious context, I

would argue, his cult will eventually erode. The evidence of this erosion will become clearer in later

chapters.

The end process, and this is likely the key to the development of his cult, is the creation of a

subservient god, who is always secondary to another deity. While Skanda-Kärttikeya can be viewed as

secondary to Indra and Agni in some of these Epic narratives, Siva is the key figure in this process of

creating a secondary god. As I have already noted, most later accounts of Skanda-Kärttikeya will establish

that Siva is his true father and will use the birth and deeds of his son to elevate the stature of the father. We

have already seen some of this process in the Aranyakaparvan and in other Epic episodes relating to

Skanda-Kärttikeya. Once he is made the general of the army of the gods, the point of worshiping him as a

deity associated with healing disappears, and once made the subservient son of Siva the point of worshiping

Skanda-Kärttikeya at all disappears. Why worship the son when the much more powerful father could be

worshiped instead? I would argue that the transformation of Skanda-Kārttikeya from dangerous to

auspicious, the shift in why he was worshiped and who worshiped him, his separation from his Graha roots,

and the assimilation of his cult into Saivism began a steady process towards obscurity within the Saivism of

119

north India by the middle ages.

It is argued, however, by most Skanda-Kärttikeya scholars that the Epic narratives point to his

increased popularity in South Asia, and his adoption by 'high' religion from some lowly source is part of

this process (P. K. Agrawala 1967:27-28; Chatterjee 1970:14; Sinha 1979:24-25; Gupta 1988:22; Rana

1995:23).26 No scholar has, in fact, viewed the assimilation of his cult into Saivism and his transformation

into an auspicious martial figure as having a negative impact on his cult. All of the scholars cited above

regard these elements as increasing his popularity until his "golden" age in the Gupta era. What I question

here is the dominant scholarly view that the movement to 'high' religion or elite religion from 'low' religion

is proof of Skanda-Kärttikeya's popularity and the success of his cult. I have argued that these

developments in his cult actually remove the basis of his popular ritual cult. After the Epic accounts are

done with him, there is no reason to fear him or to engage in propitiatory worship of him. If one does want

assistance one would be wiser to worship Siva, or Agni. In short, the treatment of this deity in the Epic is

not a sign of his growth and success, but actually an indication that his cult faltered over time as it is

transformed through this text and, as we shall see, other sources. Certainly, Skanda-Kārttikeya must have

been popular with some groups, but what we have to question is what version of Skanda-Kārttikeya was

popular and with whom. These questions will be addressed in my section on coinage and statuary. In those

sections I will also argue that the development of his cult that is outlined in these Epic narratives is also

echoed in the ancient lived tradition of Skanda-Kārttikeya's cult.

26 An extreme example is Chatterjee, who claims that "by the time the two Epics were compiled Skanda became a favourite god of the Indians" (1970:14). He sites as evidence "more verses have been devoted to him in the Great Epic than to any other god (Visnu and Siva excepted)" (1970:14). What Chatterjee fails to acknowledge is that many of the verses devoted to Skanda-Kärttikeya are, in fact, indirect praise for Agni and Siva. It is wise that he lists Visnu and Siva as excepted from his popularity count because most of the Epic makes it clear that Skanda-Kärttikeya is beneath Siva in stature and cannot be compared with that god. Sinha makes similar claims concerning the length of the accounts in the Mahābhārata as reflective of Skanda-Kārttikeya's popularity (1979:24-25). He also argues that the association between Skanda-Kärttikeya and Siva "paved way for the growth and popularity of Kārttikeya" (1979:25).

120

Chapter Four: Coins and Statuary and Skanda-Karttikeya: The Earliest Depictions

What follows builds on the textual analysis I have presented and attempts to ground my argument

in the lived religious tradition of Skanda-Kärttikeya as revealed by coinage, statuary and epigraphy.

Through this evidence I will argue that the most important factor in the early cult of this deity is the

involvement of foreign powers and their desire to present him as a military deity. It is through an analysis

of these non-textual sources that we can witness first hand the transformation of Skanda-Kärttikeya to which

the textual sources allude.

4.1.1 Questioning the Earliest Issues Depicting Skanda-Karttikeya: the Punch Marked Coins and Ujjain Coins

One of the problems when dealing with ancient coinage concerning Skanda-Kärttikeya is

chronology. Some scholars, most notably John Allan (1936:xxxvi), have argued that the first South Asian

issues which depict the deity are silver Punch Mark Coins (PMC),' but there is doubt about the

identification of the figure on these ancient coins. I will begin by reviewing scholarship on this matter, and

then I will provide my own views on the subject.

Allan bases his identification of these PMC on later Ujjain copper coin issues.2 He sees a

correspondence between the PMC image illustrated in figure 2 and an Ujjain coin type illustrated in figure

3. He identifies this Ujjain figure as "Kārttikeya" because he "holds a spear", and "on var. i he is shown

with three heads; the other three are behind and naturally not represented because they cannot be seen, so

that he is six-headed -- which identifies him as Karttikeya" (Allan 1936:cxlii-cxliii). This three-headed

Ujjain figure is illustrated in figure 4. Allan goes on to acknowledge that the three- or, as he would like us

'Dating these PMC is a point of controversy. They are not inscribed and were used for centuries after they were produced leaving dating them open to question. The PMC I will discuss are likely issues of the Magadha kingdom or the later Mauryan empire, which provides a general chronology for them of the fourth century BCE to the second century BCE (Bopearachchi and Pieper 1998:17-19).

2 Allan dates Ujjain coins to the third or second century BCE (1936:cxlv). Bopearachchi and Pieper see Ujjain copper coinage beginning at around 200 BCE (1998:21).

121

to imagine, six-headed figure could be Mahākāla Siva whose cult is known to have been important in

Ujjain, but he argues that the absence of the trident and the presence of only two arms work against such an

identification. He appears to leave the issue undecided by suggesting that the Ujjains may have represented

both Siva and Karttikeya on their coinage (Allan 1936:cxliii). He does, however, continue in the catalogue

to call the figure 'Kärttikeya' and makes no allusion to any doubt concerning this attribution when he

discusses the figure on the PMC (figure 2):

it [the figure on the PMC] is one which is found identically on another series -- the copper coins attributed to Ujjayini3 .... This enables us to identify the figure as Karttikeya, who appears six- headed on some coins of Ujjayini, and may therefore be recognized in the other forms in which he appears there. (Allan 1936:xxxvi)

In short, Allan hinges his identification of the single-headed PMC figure on his identification of the later

three-headed Ujjain figure, which he nonetheless acknowledges may well be a form of Śiva.

Allan's argument that the lack of a trident or multiple arms is evidence countering an identification

of the figure as Siva is not well grounded. The trident battle-axe as an emblem of Siva on coinage may have

originated in the north-west of the Subcontinent around the second to first century BCE on coin types which

show the influence of Indo-Greek issues.4 Ujjain coinage does not appear to be influenced by Indo-Greek

3Ujjayinī is another spelling for Ujjain.

4Srivastava feels the trident-battle axe symbol may have been borrowed from Indo-Greek and Kusāna coinage (1977:150). I think the trident first appears on one of the issues of the Indo-Greek king Demetrius I. The reverse of the coin simply shows a trident (Bopearachchi and Pieper 1998:231). The king's rule has been approximately dated to 200-185 BCE (Narain 1955:4). Antimachus I also has a trident on his coins depicting Poseidon on the reverse. His approximate dates of rule are 190-180 BCE (Narain 1955:5-6). The trident with battle-axe on other Indo-Greek coins is harder to trace; Dasgupta claims it appears on the reverse of a Zoilus coin, but he provides no reference, and I have been unable to locate this coin (1974:62). Both Tarn (1938:319) and Narain (1957:114-115) argue for two Zoiluses, and Dasgupta also does not mention which one of the two he intends. Zoilus I can be dated to the second half of the second century BCE (Narain 1955:23), and Zoilus II may have ruled from 95-80 BCE (Narain 1955:29). I think the trident with battle-axe first appears on indigenous issues of the north-west with the coinage of the Audumbaras from the second to first century BCE. Audumbara coinage was influenced by Indo-Greek coinage (Dasgupta 1977:59), though it is not clear if the trident symbol itself was borrowed from the Greeks. Another possible origin for the trident with battle-axe symbol is Mitra coinage of Pañcala. B. Lahiri dates these coins to the closing years of the second century BCE or the early first century BCE (1974:187) The earliest Pañcala ruler to use the device may have been Agnimitra, who can be dated to the first half of the first century BCE, though this dating is a hypothetical reconstruction (Lahiri 1974:123).

122

issues or north-west coinage in general, but takes its model from Mauryan PMC (Bopearachchi and Pieper

1998:23-24). These Ujjain coins may also predate the first appearance of the trident with battle-axe as an

emblem of Siva on coinage. Thus, the absence of a trident with battle-axe on these Ujjain issues does not

indicate that they are not associated with Siva.

Allan's lack-of-multiple-arms theory is on equally unsure ground. Depictions of deities with

multiple arms do not appear on South Asian coins until Kusana coinage around the beginning of the

Common Era. Thus, not finding multiple arms on these Ujjain issues is not grounds for justifying an

identification of the three-headed figure as Skanda-Kärttikeya.5 There is also room to question the logic

behind seeing six heads on a coin that clearly shows three heads. It is more prudent, I think, to argue from

what is represented on the coin and not from what we can imagine might be represented. Thus, Allan's

identification of the three-headed figure on Ujjain coins as Skanda-Kärttikeya is questionable.

J. N. Banerjea has also taken issue with Allan's identification and has argued that if the Ujjain

coins with three heads are supposed to represent Skanda-Kärttikeya, they ought to show him with six heads.

He notes that some of the Kusana coins depicting Siva show him with three heads, and on these grounds he

thinks the Ujjain image is Siva (1956:117).6 I agree with Banerjea, but for different reasons. Given the

Ujjain region's close association with the Mahakāla cult and given the weak argument for identifying the

figure as Skanda-Kärttikeya, the most likely candidate to be the multiheaded deity on these coins is Siva or

a form of him.7

5In fairness to Allan, he does note that the appearance of only two arms "need not be stressed" (1936:cxliii).

'It should be noted, though, that the Kusana coins to which Banerjea refers come two to three centuries after these Ujjain coins.

7A. M. Shastri has also dealt with this point. He, with Banerjea, argues that the single headed Ujjain figure is Siva because he holds a danda and kamandalu (1968:73-74). He does, however, think that the three-headed figure on Ujjain coins is Skanda-Kārttikeya (1968:76). He argues for a lengthy development in representing this deity's six heads, which begins in "Ujjayinī where three front heads were shown and the remaining three heads were left to be imagined; it ended with the Yaudheya coins where all the six heads are clearly visible, nothing being left for the imagination of the devotees" (1968:77). This

123

This conclusion leaves Allan's remaining attributions for PMC and Ujjain coins in doubt. Some of

these Ujjain coins that depict a single-headed figure do present some resemblance to later images of

Skanda-Kārttikeya on coins and statuary. The figure may hold a spear and has some sort of headdress,

which, as we shall see, are characteristic signs of Skanda-Kärttikeya from certain coin issues and statuary.

The coins are, unfortunately, too worn to identify accurately the type of headdress or to determine if the

spear is really a long staff on some types. The strongest argument against Allan's identification, however, is

the bag or jar the figure holds in his left hand. There are no other iconographic examples of Skanda-

Kärttikeya holding such an object and no hints from textual sources that would suggest such a

representation of Skanda-Kärttikeya is possible. I would argue, this bag or jar is simply not part of Skanda-

Karttikeya's iconographic repertoire, and these images do not represent the deity. Thus, Allan's attribution

of Skanda-Kärttikeya for these PMC and the Ujjain coins does not appear to be correct.8

view is also supported by O. P. Singh (1978:70-71). I find Shastri's reasoning concerning these imagined heads as questionable as the same argument used by Allan above.

8Other Skanda-Kärttikeya researchers have largely followed Allan's lead or taken Banerjea's criticism of the identification into account. Regarding the PMC coins, P. K. Agrawala notes Allan's argument and Banerjea's rejection of it. He does not present a conclusion of his own, but does suggest that the crest-like thing on the PMC figure's head "seems to be the crest of a warrior's helmet. If so, it would be an additional point suggesting this to be a warrior-deity" (1967:39). Agrawala also makes no clear statement on the Ujjain coinage. He notes Allan's argument and Banerjea's opposition and claims with respect to the three-headed figure that coins without any clear Saivite symbols like a trident or bull should not have figures on them identified as Siva. He does not, however, claim that the figure on any of the Ujjain coins is Skanda-Kārttikeya (Agrawala 1967:39-40). In a later paper, however, Agrawala notes regarding the crest- like headdress on the PMC coin that it is "indicative of identity of the figure as warrior or the war-god Skanda", and he seems more certain that Banerjea's criticism of Allan is unjustified (1977:102-104). B. Chattopadhyay sides with Banerjea and regards the Ujjain coins as depicting Siva (1977:51). K. Sinha notes that Allan regards the PMC and Ujjain coins to represent Skanda-Kārttikeya. He notes some criticisms of this view, but does not supply a conclusive statement of his own regarding the identity of the figures (1979:72-73). U. Thakur notes Allan's arguments, but regards the identity of the figure as "doubtful" (1981:44). Thakur does not supply reasons for this doubt. He also shows some doubt concerning Allan's treatment of the Ujjain coins, but is not willing to discredit the idea completely (1981:45-46). V. R. Mani mentions Allan's identification and Banerjea's criticism of it, but does not offer an opinion (1990:68). P. G. Paul and D. Paul present Ujjain figures as representing Skanda-Kärttikeya, but only cite Allan to back their claims (1989:116-117). Bopearachchi and Pieper think the Ujjain coins do depict Skanda-Kārttikeya, but they present no arguments for this assumption (1998:24). A. K. Chatterjee (1970), R. Navaratnam (1973), G. S. Ghurye (1977), S. M. Gupta (1988) and S. S. Rana (1995) make no mention of these coins.

124

4.1.2 Questioning the Earliest Issues Depicting Skanda-Kärttikeya: The Mitras of the Punjab or the Audumbara Mitra Coins

Chronologically, the next group of coins in which Allan sees Skanda-Kärttikeya is the coinage of

the Audumbaras (1936:1xxxvi). The Audumbaras are a group from the Beas River area who, according to

some scholars,9 issued three separate coin types between the second century BCE and the beginning of the

Common Era (Dasgupta 1974:56; Bopearachchi and Pieper 1998:47). The coins in question belong

chronologically to the last group of coin types and are inscribed with the last name Mitra. Allan dates these

Mitra coins to between the first century BCE and the first century CE (1936:1xxxvi).10

The name Audumbara does not appear on the Mitra issues; this has caused some debate over the

correct attribution of these coins to the Audumbaras. The earlier class one and class two coins of the

Audumbaras are all inscribed with the group's name. Allan claims to follow Cunningham who first assigned

these Mitra coins to the Audumbaras based, according to Allan, on their find site (Allan 1936:1xxxiii;

Cunningham 1963:69-70).11 For contemporary scholars, however, the use of find sites alone is not regarded

as adequate evidence for such an attribution.12 Cunningham's assignment of the coins to the Audumbaras is

no longer accepted, and most scholars now assign this group of coins to an independent Mitra dynasty in the

Punjab ruling after the Audumbaras (Ahmad 1977:65-68; Sharma and Kumar 1979:9-10; Bopearachchi and

9I will discuss this in detail below.

ODates for these Mitra kings are uncertain. Dasgupta also dates them to a period after class one and two Audumbara coins (around the first century BCE), but he does not think that they can be placed in the first century CE (1974:58). Ahmad dates them to around the first century BCE (1977:68), while G. B. Sharma and M. Kumar date them to the second century BCE (1979:13). A certain date regarding these coins does not seem possible, but most scholars except Allan seem to agree that the coins cannot be placed in the Common Era.

11A number of noted 'tribal' issue numismatists have gone along with Cunningham's attribution of these Mitra coins including K. K. Dasgupta (1974:57-58) and M. K. Sharan (1982:192-193). I do not, however, find where Cunningham argues for this attribution, but he does include these Mitra kings in his catalogue of Audumbara coinage (1963:69-70).

12O. Guillaume has illustrated the dangers of using provenance in numismatical reasoning for Indo-

(1990:105). Greek coinage (1990:103-107). He regards reasoning based solely on find-spots to be insufficient

125

Pieper 1998:48).13

We have coins of six Punjab Mitra kings: Āryamitra, Agnimitra, Mahīmitra, Bhānumitra,

Mahābhūtimitra and Indramitra (Ahmad 1977:65; Sharma and Kumar 1979:51). Coins of Āryamitra and

Mahīmitra are illustrated in figures 5 and 6. The coin in figure 5 shows a standing male figure who looks to

the proper right of the coin. He holds what appears to be a large spear in his right hand, and his left hand

rests on his hip. His hair is tied into a top knot, and he wears a sash around his waist, part of which flows

down between his legs. His style of dress is Indian. There is a legend in Kharosthi which reads "Raña

Ajamitasa" meaning: of the king Aryamitra (Dasgupta 1974:50). The first coin in figure 6 shows the same

standing male, but with a wavy line to the left of the coin. This coin's Kharosthi legend reads

"Mah(i)mitrasa" meaning: of Mahīmitra (Dasgupta 1974:50). The second coin in figure 6 shows the lower

torso of the figure and is also of Mahimitra. On this coin the figure may stand frontally, and his waist sash

seems to flow down from either side of his hips.

Allan identifies this figure as "Karttikeya, or perhaps simply a warrior holding a spear ... "

(1936:lxxxvi). I can do little better than Allan in terms of presenting a clear identification of the figure on

the coins, but what can be done is to place the figure into an iconographical frame of reference. There are a

number of iconographic features of this warrior figure which relate to statues. The image on the Mitra coins

bears, I think, a strong resemblance to the first century CE statue from the Aligarh district which

Coomaraswamy has identified as "Kuvera, or other Yaksa" (1923:47). An example is illustrated in figure 7.

A certain identification of the statue is not forthcoming, but it does bear some resemblance to the Mitra

figure. They both have the same headdress, Indian style clothing, and both carry a staff or spear.14 They

both fit a sort of yaksa-warrior type of statue that does not obviously identify the figure as Skanda-

13 Ahmad provides a good summary of the various arguments presented by scholars regarding the attribution of these coins (1977:57-65).

14Coomaraswamy feels the statue holds a long staff (1923:47). I think that both the statue and coins are too worn for a clear identification of spear or staff.

126

Kärttikeya, but could represent a prototype image that will develop into Skanda-Kārttikeya.

There may also have been some importance placed on the Mitra coin figure. I make this last

statement because we have moved away from a period of PMC, where images are placed on coins without

obvious reference to coins as political or propaganda tools, and into an age where the ability of coinage to

transmit political and religious messages is recognized.15 Placing a warrior image on coins was likely not a

random choice on the part of the Mitra rulers of the Punjab, but a careful selection designed to present a

political or social message from the issuers to a general public.16

15This shift is likely due to Indo-Greek and Central Asian influence on South Asian coinage. The basic coin type introduced by Alexander the Great has an obverse with the portrait head of a divinity or ruler, and a reverse with the figure of a deity with the name of the ruler inscribed on it (Errington 1992:49) This coin style is used throughout Indo-Greek coinage which influenced the indigenous issues of the north- west and is even continued by the Kusanas. The economic idea behind this type of coinage and one of the reasons for its popularity is that the value of the coin is backed by the king whose name appears on the coin (MacDowall 1987:8-12). This change in approach signals a shift from valuing a coin only on the basis of the intrinsic value of the metal in the coin to one that also takes into account the authority of the issuer in the realm and his reputation for producing a certain quality of coinage. Thus, Alexander's silver issues were accepted as coinage and not as bullion. In fact, some of his silver coinage was valued over bullion by as much as 16% (MacDowall 1987:8-9). Some credence, though, should be given to Holt's criticism of MacDowell and other scholars, who attribute massive economic change in the area singularly to Alexander and his coinage (1999:30-35). His criticisms of MacDowall, however, concern the causes behind the change from a barter economy to a coinage economy in Afghanistan. He claims that this change was actually carried out by the Seleucids and took some time to achieve (Holt 1999:36-37). Holt's arguments are well supported, but it is primarily the coin type and economic idea behind Alexander's coinage which are continued by the Seleucids which concern us here. I think, at least in part, coinage of the Alexandrian type becomes linked to royal reputation and ability. Coins of this type become a means of royal propaganda with kings adding titles to their names and selecting certain gods and goddesses to demonstrate their connections to previous rulers or to imply the support of a certain deity for the king (Errington 1992:49-50). This style of currency differs remarkably from PMC from the Mauryan Empire, which presents no clear means of identifying the ruler or what that ruler might stand for or assume about himself. It is the Indo-Greek pattern of inscribing coins to identify the issuer and of placing meaningful and often divine figures on coinage that north western Indian groups begin to imitate as a means of conveying messages. This practice was also continued by the Kusānas.

16A final point to note is that the figure on the Mitra coins is anthropomorphic. The figure may represent a human warrior, but I will argue later that the use of anthropomorphic figures on coinage of this region is often a sign of borrowing from Indo-Greek and other foreign sources. These Mitras certainly would have had contact with Indo-Greek kingdoms, and the possibility of borrowing should not be ignored here. I will have more to say on this topic in my section on Yaudheya coinage. Other researchers on Skanda-Kärttikeya generally follow Allan regarding these coins. Agrawala summarizes Allan's view and mentions that Aryamitra's name (Agrawala presents the second a in the name as a long vowel, Allan leaves it short; it could be read either way) could be linked to Arya, which is a name given to one of Skanda-

127

4.1.3 Questioning the Earliest Issues Depicting Skanda-Kārttikeya: The Ayodhyā Coins

The Ayodhy coins represent one more group of coins which may or may not be identified as

representing Skanda-Kärttikeya. These coins come from another Mitra dynasty ruling in Ayodhyā who

issued coins between the first and second centuries CE (Allan 1936:xc). While these coins are slightly out

of the chronological order that follows, I will address them here because their connection with the cult of

Skanda-Kārttikeya is questionable. The coins of Aryamitra, Vijayamitra and Devamitra are illustrated in

figure 8. The first coin is of Aryamitra, it depicts a bird, which some have identified as a cock, and a tree,

which some have identified as a palm tree, and a wavy line, or river below. The next coin belongs to

Vijayamitra and is basically the same as Aryamitra's. The final coin belongs to Devamitra and depicts a

bird on a post that V. Smith thinks is a cock along with a palm tree and wavy line below (Smith 1972:151).

It is this coin of Devamitra and Smith's description of the bird as a cock which has led J. N. Banerjea to

relate these coins to a Skanda-Kärttikeya cult (1956:141). The majority of writers on this topic follow

Smith's identification of the bird as well as Banerjea's link of these coins to Skanda-Kārttikeya.17

The jump, however, from a coin which depicts a cock to one that depicts Skanda-Kärttikeya is not

so straight forward. The cock is associated with Skanda-Kärttikeya in statuary and later coinage, but this

Kārttikeya's mothers in the Mahabharata (3.219.40), but this argument does not strike me as very strong (1967:40). Dasgupta does not present a direct discussion of this figure on the coins, but describes him as a "male figure standing to left holding a spear in right hand" in his catalogue (1974:50). Sinha cites Allan and describes the figure on the coins as representing a god who appears as a warrior (1979:72). Thakur follows similar lines and concludes that the figure may be identified as Skanda-Kārttikeya (1981:47). Mani, citing Allan, identifies the figure as Kārttikeya (1990:68). Chatterjee (1970), Navaratnam (1973), Gupta (1988) and Rana (1995) do not mention the coin series.

17P. K. Agrawala notes the cock and post as well as Banerjea's linkage of the cock to Skanda- Kärttikeya. He backs this interpretation of the cock as a symbolic representation of Skanda-Kärttikeya with the Lala Bhagat column which also has a cock capital carved on it (1967:44-45). This column will be discussed below. Chatterjee notes the same coins and column, and suggests it is "permissible to think that both the above mentioned kings [Devamitra and Vijayamitra] were connected with Kārttikeya worship" (1970:34). Sinha notes the same points as Agrawala (1979:71). O. P. Singh also makes mention of these Ayodhya coins and the Lala Bhagat column as evidence of Skanda-Kārttikeya devotion (1978:71). Thakur (1981:46), Gupta (1988:7), Mani (1990:69), and Rana (1995:21) make similar conclusions. Navaratnam (1973) and Ghurye (1977) make no mention of the coins.

128

does not automatically imply that a cock alone represents the deity. While the potential weakness of this

identification of the Ayodhya cock coins as representing Skanda-Kärttikeya is never raised by scholars who

discuss this material, they do, I think, acknowledge its weakness by drawing on the additional evidence

provided by the Lala Bhagat column.18

Lala Bhagat is a village near Kanpur where a carved pillar has been discovered. Near the pillar a

large carved bird was also found. P. K. Agrawala, Banerjea and others assume that the bird is a cock which

was originally mounted on top of the column and, therefore, that the column is dedicated to Skanda-

Kärttikeya and illustrates his worship in the area (Banerjea 1956:106; P. K. Agrawala 1967:45). The

unstated implication is that a lone cock can stand in for the deity. The pillar also has a number of figures

carved onto it. At its top are carved two flying birds, possibly geese. Below them is a chariot which holds

Sürya and two female attendants. The chariot appears to be driving over a demon-like creature. Below the

demon are two females, possibly divine, and below them are a group of gana-like creatures. Still further

down is a peacock and below it an elephant which is pulling on a lotus stalk. At the bottom is a female,

possibly Laksmi, who is turned towards a carving of a bird mounted column which is held by a crouching

dwarf. Banerjea and Agrawala regard this bird as a cock. Certainly, the carving of the bird mounted column

on the column itself suggests that the carved bird found near it originally stood on top of the column. There

is also a fragmentary inscription on one side of the column which reads kumārav(a)ra in Brähmī of the first

to second century CE (P. K. Agrawala 1967:46).

On the basis of this evidence, P. K. Agrawala claims that this column was dedicated to Skanda-

Karttikeya, and that the carvings tell his story as given in the Aranyakaparvan of the Mahabhārata

(1967:45). His arguments for reading the Mahabhärata narrative into this column are weak and strike me as

based on arbitrary assumptions about whom and what the carvings on the pillar represent (P. K. Agrawala

18Indeed, P. K. Agrawala's account of the column follows immediately from his discussion of the Ayodhyā coins (1967:45); most Skanda-Kärttikeya researchers follow Agrawala's example.

129

1967:28, 45-46).19 It is also not clear to me that the bird found near the column, the bird carved on the

column and the bird on the Ayodhya coins can only be identified as a cock. Allan also has his doubts

concerning the bird on the coins. He comments: "and [on the] reverse [of the coin] a bird, usually called a

cock but probably a hamsa" (1936:lxxxix). Indeed, the bird depicted on these coins and on the column is

not obviously a cock and could be a goose (hamsa) or some other bird. I also think it is odd to view this

pillar as special to Skanda-Kärttikeya when it bears no anthropomorphic image of him (by far the most

popular representation of him), but does present anthropomorphic representations of other deities. I think it

is more likely that the pillar could be dedicated to Sürya or Laksmi; we have seen that Sūrya is clearly

represented on the pillar, and one of the female figures may be Laksmi. There are, then, a number of

potential ways to understand the significance of the column.

The strongest evidence Agrawala presents for attributing the column to Skanda-Kärttikeya is the

inscription, kumarav(a)ra, which, as he points out, is an epithet of the deity used in the Mahābhārata

(1967:28, 46). The inscription is, however, fragmentary, and the key phrase can only be read through a re-

construction of the epigraph. What is more, the word kumära, while common in literature as an epithet for

Agni and Skanda-Kärttikeya, is not always used as an epithet for the deity in epigraphy. On inscriptions

kumāra can also mean "prince" or a proper name. Kumära as an epithet of Skanda-Kärttikeya often appears

on inscriptions with other epithets of the deity in order, I suspect, to make it clear that the deity is being

referred to and not a royal prince. Thus, on Huviska's coinage depicting the deity the inscription reads:

skanda-kumāra, and on Yaudheya coins we read: bhāgavata svamino brahmanya devasya kumarāsya.20

19 Agrawala argues that the depiction of Sūrya relates to a solar connection for Skanda-Kärttikeya for which he and Banerjea argue (P. K. Agrawala 1967:45-46; Banerjea 1956:106-107). I have already addressed this issue in my discussion of textual sources for Skanda-Kärttikeya and shown the idea to be incorrect. Agrawala argues that the gana-like creatures are Grahas, and that the elephant represents Indra (1967:46). He also claims that the female figure on the bottom of the pillar is Laksmi, who is "showering her blessings on the pilaster carved on her right, surmounted by a cock evidently [a] symbolical representation of Skanda" (1967:28).

20These coins and inscriptions are studied in more detail below.

130

Without more of the inscription from the Lala Bhagat pillar we cannot say with certainty that the kumāra in

the inscription is really a reference to Skanda-Kārttikeya. Given these problems P. K. Agrawala's comments

that the pillar shows "beyond doubt that this was a column with a cock-capital raised in honour of Skanda"

are not well justified (1967:28). The best that can be said, I think, is that the column may indicate Skanda-

Kärttikeya worship, but we lack conclusive evidence to make a firm statement regarding the attribution of

the pillar to the deity.

Since the Lala Bhagat column is the evidence used to show that a cock alone can represent

Skanda-Kärttikeya, we must also cast doubt on the supposed cock as an emblem of the deity on Ayodhyā

coinage. As with the pillar, the Mitra coins of Ayodhya do not provide incontrovertible evidence that

Skanda-Kärttikeya is represented by a cock on the coins or that he was worshipped by these kings. The

coins only allow for the possibility that Skanda-Kärttikeya may have been represented on them.

4.2.1 Yaudheya Skanda-Kārttikeya (Brahanyadeva) Coins: The First Series

The Yaudheyas are commonly referred to as a tribal republic who occupied various regions of the

modern Indian states of Haryana, Punjab, Uttar Pradesh and Rajasthan between the second century BCE to

the fourth century CE. In this study I will not refer to the Yaudheyas as a tribe or as a republic. I think both

of these terms are misleading and suggest a great deal more about the Yaudheyas than the limited historical

evidence for the group supports.21 I will use their name and the terms state or group when referring to the

2'The term 'tribal' has been critiqued by a number of scholars who contributed to the book Seminar Papers on the Tribal Coins of Ancient India (c. 200 B.C. to 400 A.D.), I would direct readers to this text for more details. The idea of the Yaudheyas as a republic has not been widely critiqued, and I will briefly discuss this term in relation to the group here. An understanding of the Yaudheyas as practicing a republican form of government seems to be rooted in some of the earliest references to them in a grammatical works of Pānini. In his Astādhyāyī, Pānini refers to the Yaudheyas and a number of other groups as an ayudhajīvī sangha (V.3.114-117) (a community who live by weapons or are warriors). The term sangha has been taken to mean a form of government by many scholars (V. S. Agrawala 1963:428). The evidence that backs this assumption comes from Astādhyāyī III.3.86, where Pānini seems to use sangha and gana synonymously. The term gana, as a political term, translates: "any assemblage or association of men formed for the attainment of the same aims" (Monier-Williams 1999:343). Pānini calls the Yaudheyas a sangha, and, as we shall see, the Yaudheyas call themselves a gana on their class six coinage. These comments of Panini combined with the class six Yaudheya coin legend have lead many scholars to refer to the Yaudheyas as practicing a republican form of government. All of the following

131

Yaudheyas. The study of this ancient state is complicated by scholarly accounts of them, which were not

always objective. Scholarship on the Yaudheyas began in earnest in the early twentieth century by scholars

who used the group as a tool in their struggle for Indian independence from the British. The resulting

discussions of the Yaudheyas were often clouded in nationalistic rhetoric, and their position as a tool in a

nationalistic enterprise has led to some overstatements concerning their historical importance and their form

scholars make reference to the Yaudheyas or north-western 'tribes' as republican: V. S. Agrawala (1963:432), P. K. Agrawala (1967:42), Thakur (1981:91), Rana (1995:161), O. P. Singh (1978:73), S. M. Gupta (1988:8), Navaratnam (1973:101), Chatterjee (1970:37), Ahmad and Kumar (1993:53), R. P. Roy (1984:76), U. S. Rao (1962:139), A. S. Altekar (1949:50) and J. Prakash (1965:135). What is not clear, however, is just what many of these writers understand a republican government to be, and if republican is the best translation of such terms as sangha and gana. In its most general sense the term republic means: "A state in which the supreme power rests in the people and their elected representatives or officers, as opposed to one governed by a king or similar ruler" (O.E.D. 1985 II:491); such a general understanding of the term may well apply to the Yaudheyas. A number of scholars may also intend by their use of the term republic in connection with this group a style of government akin to the Roman republic, but such a view is never explicitly expressed. The term has also, however, become connected with democracy, and this may not apply to the case of the Yaudheyas. Some nationalistic historians go too far in presenting this Yaudheya 'republic' as the equivalent of a modern democracy (Jayaswal 1933:148; Sharan 1972:79). We lack the evidence to form such a conclusion. Only a few scholars have attempted to present meaningful qualifications for their use of the term or to rigorously assess the arguments of Jayaswal and Altekar, who first began to use the term in connection with the Yaudheyas. In this vein, J. P. Singh comments at length on the nature of the Yaudheya republic. He calls their form of government a "primitive republican" one (1977:6). He regards all members of the group to be part of an assembly, but felt the state was administered like a monarchical system, and membership was probably hereditary. His evidence comes from both coins and seals, and it is his use of seals that causes him to reassess the term republic in relation to this group (1977:6-9). While I agree with Singh's appraisal that the term republic needs to be qualified, I do not feel his other points are well demonstrated by the evidence he supplies. B. P. Mazumdar rejects a number of assumptions regarding this supposed republic. He notes that there is no evidence for Jayaswal's and Altekar's statements that there were 5000 members in the Yaudheya republic or that we know what sort of republic this group may have had (1969:307-310). What Mazumdar is able to demonstrate is that those who first proposed the use of republican in connection to this tribe did so with minimal solid evidence. I think the evidence from coins, texts and seals does not clearly demonstrate that the Yaudheyas were a republic in the manner that Jayaswal and others understand the term. I also do not think that we should translate the words gana and sangha with republican or democratic. The best, I think, that can be said is that there is minimal indication that the Yaudheyas had a monarchical form of government, and that they appeared to have ruled with some sort of an assembly, but the membership and nature of that assembly and its powers remain a mystery to us. Perhaps the wisest course of action is the one Sinha takes, which is to make no reference to the word republican and to simply refer to the Yaudheyas as a gana (1979:147).

132

of governance.22 I will not detail all of these problems in this section, but one of the results of this

nationalistic emphasis has been the failure to recognize foreign influence on this group and the manner in

which they represent Skanda-Kärttikeya. Part of my argument in this section will be that foreign groups

played an important role in the development of Skanda-Kärttikeya's cult even in the case of this indigenous

group, and their depictions of the deity.

In his catalogue of coins in the British Museum, Allan places the Yaudheya issues into six main

classes, with some of these classes having numerous varieties (1936:cxlvii-cli). Other scholars, most notably

K. K. Dasgupta (1974:201-211), K. D. Bajpai (1973:90-94) and N. Ahmad with M. Kumar (1993:49), have

attempted to revise Allan's classification with limited success.23 Allan's classification is the one most

commonly used by scholars, and it is the classification system I will use with some qualifications.24 Allan's

22The history of Yaudheya scholarship and its problems are not my immediate concern here. I will not be exploring this topic in detail. What I will say is that the two scholars who used the Yaudheyas extensively to make what might be considered nationalistic points are Jayaswal (1933) and Altekar (1962, 1967).

23 One of the problems, I think, with tribal numismatics is one of terminology. All of these scholars use terms like class and variety in their classifications, but there does not seem to be general agreement on what these terms mean. What makes a class of a particular coin series, and what makes a variety of that class is not always clearly spelled out. Until tribal numismatists come to some agreement on these terms, presenting a new classification or criticizing old classifications has little meaning.

24These qualifications are largely supplied by the scholars listed above who have tried to revise Allan's initial attempt. Each of these classification systems has its value, but each also presents some problems. Most scholars now reject Allan's attribution of his class one coinage to the Yaudheyas. These coins are small potin coins which do not have the name of the state on them. Most are uninscribed, but one variety has the legend "Mahārājasa" in first century BCE Brāhmī (Allan 1936:cxlvii). Allan states: "The slightly scyphate fabric, metal, and provenance connect them with coins of Class 2 bearing the name of the tribe" (1936:cxlvii). The provenance of the coins is Behat, where they were found "with coins bearing the name of the Yaudheyas" (1936:cxlvii). The first scholar to criticize these class one issues was K. D. Bajpai. He feels they are not Yaudheya issues and bear more in common with the die struck coins of Kausāmbi (1973:90). K. K. Dasgupta does not note Bajpai's suggestion and accepts Allan's attribution of the coins to the Yaudheyas (1974:200-201). D. Handa criticizes Allan's attribution. He points out that the scyphate fabric and metal type of these coins are not unique to Yaudheya coinage, and the coins in question were found in a hoard of Yaudheya and Kuninda coins, which leaves no grounds for assuming that they belong to the Yaudheyas (1991:69). I have already pointed out that find site alone is not a valid means of identification for coinage. Handa also notes that "mere association, however, cannot be regarded as a firm basis of attribution" (1991:69). He feels that these coins in Allan's class one are likely local coins of Behat (Handa 1991:69). N. Ahmad agrees with Bajpai's criticism of Allan's attribution, but he does not think that

133

classification is somewhat misleading because it is not chronologically based. The classes which depict

Skanda-Kärttikeya are classes three and six, and these are the last two coin types issued by the Yaudheyas.

There is a wide chronological gap between the class three coins and class six coins of at least two-hundred

years and as much as three-hundred years.25 Most Skanda-Kärttikeya scholars do not seem to view this gap

as significant, if they acknowledge it at all. Most only note some of the obvious iconographic differences

between the two coin groups and regard the circumstances and understanding of the deity in the two coin

types as the same. I will present a significantly different argument. I regard the iconographic changes this

deity undergoes on coinage as full of implications concerning how he was characterized and understood.

The period between Yaudheya class three and class six coin production is the period of Kusāna domination

over north India, and their reign sees some dramatic shifts in the understanding of this deity, which are

carried over into Yaudheya class six coinage. The context of the Yaudheyas and their understanding and use

of Skanda-Kärttikeya are different on their class three and class six coinage. This specific section on

Yaudheya class three coinage will illustrate and reflect on the differences between various types of class

three coinage.

they are from Kausambī. He argues that they resemble Uddehika coinage and feels they were likely minted at Behat (1988:67-68). Ahmad and Kumar also do not place Allan's class one with their classification of Yaudheya coins (1993:49). Bopearachchi and Pieper agree with Bajpai's attribution of these coins to Kausambī and note "The two more rectangular pieces in the author's [Pieper's] collection were acquired together with a group of Ujjain and Kausambi coins, thus supporting the assumption that they belong to the lower Ganges or Narbada valley rather than to the Punjab region" (1998:51). Whatever the case, these coins do not appear to belong to the Yaudheyas. These 'class one' coins do not present figures that represent Skanda-Kärttikeya and are not of great consequence for this study. The coins of Allan's class four have been identified as issues of local rulers of Garhwal in Uttar Pradesh and should also not be considered as issues of the Yaudheyas (Ahmad 1985:110-111; Handa 1991:72-73). These coins do depict images of Skanda-Kärttikeya and will be treated in this chapter. Coins of Allan's class two and five, which date to the second and first century BCE, are the earliest Yaudheya issues, but they do not present images of the deity and will not be discussed in detail (Allan 1936:cxlix, cl). The first Yaudheya coins that are, I think, definitely from the group and do depict Skanda-Kārttikeya are Allan's class three. Allan's class three, however, is also not without controversy, but I will discuss this in the body of this chapter. Finally, Allan's class six coins are generally attributed to the Yaudheyas, and they also depict Skanda-Kārttikeya. They will be discussed in the next chapter.

25The argument behind this chronology will be discussed later.

.34

Another impediment to understanding these coins relates to problems of chronology. Previous

scholars were not able to state with certainty whether the first group to issue coins depicting Skanda-

Kärttikeya was the Yaudheyas or the Kusanas. I will argue that the Yaudheya class three coinage represents

the first certain numismatic depictions of the deity. There has also been some debate concerning the date

attributed to the production of Yaudheya class three coinage. I will argue for a date beginning at around the

last decades of the first century BCE and ending at the close of the first century CE for the period of

Yaudheya class three coin production.

A final problem is the lack of a comprehensive catalogue of the Yaudheya class three issues. I have

done what I can to correct this problem in this section, though I acknowledge that my catalogue may not be

complete. I will begin this section with the catalogue of Yaudheya class three coinage which depicts

Skanda-Kärttikeya. I will then demonstrate that certain iconographic types for Skanda-Kārttikeya found on

Yaudheya class three coinage are borrowed from foreign Hellenized coinage. From these points I will argue

for a new chronology for these class three issues. I will conclude with a section discussing the potential

meaning of the iconographic styles used by the Yaudheyas, and the implications created by their

iconographic borrowing from foreign coinage.

4.2.2 The Yaudheya Coins: Problems of Classification

There has yet to be a work on Yaudheya coinage like Gobl's work on the Kusanas, Münzprägung

Des Kušānreiches, which collects the known types of the group's coinage and examines the use of dies and

other numismatic features in detail. The task of studying Yaudheya numismatics would be greatly simplified

and placed on much firmer footing if such a study existed. Such a task is not, however, within the scope of

this dissertation, and I will explain the limitations this absence has for my present study as they are

encountered.

The absence of a complete Yaudheya catalogue does force me to spend time grouping and

describing these Yaudheya issues. A number of the varieties I discuss below have yet to be adequately

illustrated and described; such a process must be performed to facilitate an informed discussion of the

135

Yaudheya Skanda-Kärttikeya issues. To do this I will present a catalogue of Yaudheya class three coins

here which depict Brahmanya or Brahmanyadeva (the names the Yaudheyas give to their representation of

Skanda-Kārttikeya on their class three coinage).26 My catalogue of Yaudheya class three coins consists of

two main groups. The first group represents Brahmanyadeva with a single head, and the second group

represents him with six heads. Both of these groups of class three coinage were likely issued at the same

time, and the six-headed variety is more common than the single-headed.

I have been able to examine the Yaudheya coin photographic collections at the British Museum,

the Indian Institute for Research in Numismatic Studies the Indian Museum, as well as some smaller

collections held at various museums and centres in India. I have not examined every coin I describe here in

person. Indeed, more varieties of Yaudheya coinage depicting this deity may well exist that I have not seen.

There are numerous problems involved in gaining access to coin collections in India and around the world.

There are some collections that I was simply denied permission to examine or photograph. I am also sure

that private collections exist that I do not know about. In addition, the class three coins were not well made,

and they are usually damaged and badly worn, making any attempt to account for the images on them

tentative. The legends on them are partially or completely worn away as are many of the figures on the

coins. The legends on copper class three coins have been reconstructed to read: "Bhagavata-svamino

Brahmanya-devasya (or sa) Kumarāsya (or sa)" or "Of Kumāra the divine lord Brahmanyadeva" (Allan

1936:cl). There is, however, considerable debate concerning these legends, which I will discuss later. With

these points in mind, I now present the varieties of Skanda-Kärttikeya images from the first group that are

known to me.

4.2.3 The Yaudheya Six-Headed Coins: Six-Headed Brahmanyadeva with a bar

This variety is well represented by a large number of coins. One of these coins is illustrated in

figure 9. The obverse of this coin shows a six-headed figure with the heads arranged in two stacked rows of

3.213.24). 26Brahmanya is also an epithet or adjective of Skanda-Kārttikeya in the Mahābhārata (3.213.1,

136

three. Above each head is a small horizontal bar, though the heads on the right are slightly off the flan.

Traces of a Brahmī inscription are visible which reads 'bhaga ... '. The six heads and the partial legend

demonstrate that the figure is Brahmanyadeva. On the worn reverse a deer faces the right towards a

building. There is also an uncertain object over the deer's back. Similar examples are illustrated in figures

10-12.27

Also part of this group is the coin depicted in figure 13, which is a unique single silver Yaudheya

issue. Here Brahmanyadeva is depicted with six heads. Each of these heads appears with two small dots

below, which may represent earrings. His left hand rests on his hips, and his right is raised. A large spear

stands to the proper right. An inscription runs around the edge, which Allan reads as "Bhāgavata-svamino-

Brahmanya[-]-Yaudheya" (1936:cxlix). He argues, however, that the legend should be read as "Yaudheya-

bhägavata-svamino Brahmanya (sa or sya) 'Of Brahmanya (a name of Kärttikeya), the divine lord of the

Yaudheyas' (1936:cxlix). There has, however, been no shortage of discussion regarding this legend; I

discuss these legends in section 4.2.15. The reverse of the coin shows a female with a large central head and

five rays or heads arranged around it. Her left hand rests on her hip, and her right is raised. To her left is a

tree in railing device, and to her right is a three-layer arched mound topped by a nandipada device. Beneath

her feet is a vase device and beneath that a wavy line.

4.2.4 The Yaudheya Six-Headed Coins: Six-Headed Brahmanyadeva with bars only on the top row of heads

Related to the variety of coinage described in 4.2.3 are these which depict a six-headed

27There are a number of coins which belong to this type. The obverse of the coin in figure 10 depicts a six-headed Brahmanyadeva with the heads stacked in two rows of three; each head has a bar over it. Traces of a spear can be seen to the right, and the Brahmi letter 'bha' can be made out in the top left. The body of the figure is unclear. The reverse depicts a deer looking to the proper right, but much of the reverse is worn away. The obverse of figure 11 shows a very worn six-headed Brahmanyadeva whose heads are more oval and elongated than that in figure 10. Above the heads the Brāhmī inscription ' ... manabra ... ' can be made out. On the reverse traces of a deer facing the left can be made out, but the rest is worn away. The obverse of figure 12 shows a similar six-headed figure with traces of the brahmī letters 'bra ... ' visible. The reverse shows a deer looking to the left with a pot over its back. Related coins are IIRNS 610/3-17, 611/1- 29, 611/2-25, 609/2-28, 609/2-34, 606/2-6, 605/3-35, 606/1-33, 606/1-11, 612/1-2, 612/2-6, 612/2-20, 612/2-32, 613/1-11, 613/1-15, 613/1-26, BM BAAA 8 and BM 78.

137

Brahmanyadeva with bars only over the top row of heads. The coin illustrated in figure 14 is worn, but a

six-headed Brahmanyadeva is visible on the obverse with his heads stacked in two rows of three. Only the

top row of heads has a bar above them. A spear is visible to the right. The reverse is worn, but a deer or

animal faces the proper left, and there is a pot over the deer's back.28

4.2.5 The Yaudheya Six-Headed Coins: Six-Headed Brahmanyadeva with a T

This next series is primarily distinguished by the placement of a T-like object over some of the

deity's heads. The best example of this type is BM 57 for which I have no illustration. The obverse of the

coin depicts a six-headed Brahmanyadeva with the heads stacked in two rows of three. Each head, with the

exception of the lower middle, has a T-like object coming out of it. The top middle T points straight up,

while the other four point upwards on an angle. The left side of the coin is unclear, but the right side shows

the deity's right hand holding a spear. The reverse of the coin depicts a female figure who has a large

central head with six balls arranged around it. The bottom two balls may represent earrings. What appears

to be her left arm is raised and her right is down, but her body appears in a stylized form made of a

hollowed out shell with three balls inside.29 The two upper balls represent breasts; the single lower ball

likely represents a womb or stomach. There is a four-tier tree in railing to the left. To the right is a ten-layer

mound, which is topped by an unclear object. Below is a wavy line. Figures 15 and 16 illustrate related

coins.30

28Similar coins are IIRNS photograph numbers 606/3-19, 606/1-25 and 607/3-15.

29This iconographic style may be influenced by Kuninda coinage. The female figure on Kuninda coinage is usually identified as Laksmī (Dasgupta 1974:99), but this identification, in my opinion, is not without doubt. What ever the case, this female is usually represented in a similar fashion as this Yaudheya figure.

30Figure 15 depicts on the obverse of the first coin a six-headed Brahmanyadeva with the heads stacked in two rows of three. Each head, except the bottom middle head, has a T above it. His left hand rests on his hip, and his right hand holds a spear. A Brähmī inscription reading "rasyabhagavata" is visible to the right (Bopearachchi and Pieper 1998:147). The reverse is slightly off the flan, and the number of heads on the female figure is not clear. She has a hollowed out body with dots representing her breasts and womb. An arched hill device is visible to the right and a tree in railing to the left. Below her feet is a wavy line. The second coin in figure 15 represents a similar type. Figure 16 shows only two visible heads on the

138

4.2.6 The Yaudheya Six-headed Coins: Six-Headed Brahmanyadeva with dots

I have no illustration for the best example of the next coin (BM 63), but it is described below. The

coin depicts a six-headed Brahmanyadeva with his heads stacked in two rows of three. Above the bottom

centre head is a bar, and over the other heads is a small ball or dot. He holds a spear with his right hand, and

his left hand is on his hip. The reverse depicts a deer looking to the right, and what may be a tree in railing

to the right, but the coin is worn, and this identification is uncertain. A related coin is illustrated in figure

17. On the obverse of this coin only Brahmanyadeva's six heads are visible. The bottom centre head is

larger than the others, and over the other heads sits a ball. The reverse is worn, but a single-headed figure is

visible with a raised right hand and left hand on hip.

4.2.7 The Yaudheya Six-Headed Coins: Six-Headed Brahmanyadeva without features

This group of six-headed Brahmanyadeva issues depict the deity with no additional features over

his heads. I have no illustration for the best example of this type (BM 48), but it is described below. The

obverse of the coin depicts a six-headed Brahmanyadeva with heads stacked in two rows of three. Below

the bottom centre head is another small ball, which may represent an earring. The figure appears to have

sashes coming from his waist, and a spear can be seen to the right, but it does not appear to be held. The

reverse depicts a female figure with a central head surrounded by five balls of equal size. Her left hand rests

on her hip, and her right hand is raised. To the left is a four-tier tree in railing, and to the right a three-layer

hill topped by a nandipada symbol. A similar coin is illustrated in figure 18. Here, on the obverse, a six-

headed (two rows of three) Brahmanyadeva figure is depicted without any additional dots or bars around or

over the heads. The bottom centre head is connected to the body by a neck. His left hand is on his hip, and

his right hand is raised, but not connected to the spear. The spear sits to the right, and its shaft is made up of

dots. The reverse of the coin is partly off the flan, but a deer facing the left is visible, below which is a pot.

obverse due to wear, but each clearly has a T over it. There may be traces of a spear to the right of the coin. The reverse is also worn, but traces of a deer can be seen looking to the right and perhaps traces of a building to the far right. Coins depicting similar images are BM 56, BM 59, BM BAAA 7, BM Deyell 1983-6-2, BM ITY 70, BM ITY 72, and BM plaster casts 12.72, 10.14 and 14.89.

139

Behind the deer to the right is a tree in railing device.31

4.2.8 The Yaudheya Six-Headed Coins: Unusual Six-Headed Brahmanyadeva

I have no illustration for the best example of the next coin (BM ITY 71). The number of heads on

the obverse figure is not clear. The figure is stick-like and of poor craftsmanship. The figure's left hand is

down to his hip, and his right hand holds the shaft of what may be a spear, but the top of the object is worn

away. There are traces of a Brahmi inscription, but only 'va' can be made out. The reverse shows an

irregular looking deer and a wavy line below. Above is an odd two-tiered symbol. The irregularity of this

coin and the lack of a clear spear or inscription make for a questionable identification of the figure as

Brahmanyadeva, though the coin is included in the British Museum's Yaudheya collection.

The second coin (IIRNS photograph 605/3-22) also has no illustration. The obverse of this coin

depicts a six-headed (two rows of three) figure with a worn, but likely hollow body. This hollow body type

is unusual on the obverse of Yaudheya coins. There are traces of inscription visible, but not enough to read.

There is an unclear object to the left, which may be a jar. The coin is too worn to determine if a spear is

held by the figure, but the figure's six heads suggest he is Brahmanyadeva. The reverse also depicts a six-

headed (two rows of three) figure, which has a hollow body. Only the top half of the body can be seen, but

it is similar to those found on female figures on the reverse of some of the coins examined above. The right

arm is raised, and the right hand has a claw-like appearance. Above the right hand is a pot or lamp shaped

object.32

31Similar coins are BM 49, BM 62 and IIRNS photograph number 605/3-24.

32L. C. Gupta illustrates a number of six-headed Brahmanyadeva coins from a find at Tehri- Garhwal. On a number of coins the author sees Sasthi (the female on the reverse of many of these coins) on the obverse of these coins (1985:14-15). It seems clear to me, however, from the illustrations the author supplies that the obverse figure is Brahmanyadeva. With these six-headed coins some mention should also be made of Allan's class four coinage (1936:cl), or coinage bearing the legend "bhānuvasya" (Dasgupta 1974:209). Allan classes these coins as Yaudheya, based on the reports of Cummingham that they were found with Yaudheya class three coinage in Garhwal (1936:cl). That they do not belong to the Yaudheya catalogue and represent issues of local rulers has been argued by N. Ahmad (1985:106-112). Allan's catalogue and the British Museum collection do not have any of these coins which represent Skanda- Kārttikeya, but S. C. Kala reports in 1956 that a hoard of Yaudheya class three coins and Bhanuva coins

140

4.2.9 The Yaudheya Single-Headed Coins: Single-Headed Brahmanyadeva with Balls and Dots or a Multi-headed Brahmanyadeva

With this first group of coins there is some doubt as to whether a single head is represented with

dots around it, or if a central head is depicted with smaller heads arranged around it. I will first describe the

coins and then discuss this doubt in a later section. I have no illustration for the next coin (BM 76). The

obverse of the coin shows a single-headed Brahmanyadeva with five small balls arranged over his head in a

semi-circle. A hook-like object comes out of either side of his head, which points down, and just below each

of these hooks is a small dot. The deity's left hand rests on his hip, and his right hand holds what may be a

spear, but the top of it is worn away. The reverse shows a deer, above which is a double S symbol and a

svastika. The difficulty in identifying what the obverse figure holds makes a certain attribution of this figure

difficult. The head, however, is similar to other Yaudheya depictions of Brahmanyadeva, and I think the

coin depicts this deity. The similarity of this figure to others on Yaudheya coinage of this type will become

clearer as I progress through this section. The hooks by Brahmanyadeva's head likely represent ears and the

dots below them earrings.33 I will comment in detail later concerning the five dots above the figure's head.

and others was found in the Garhwal District. Some of the Bhanuva coins had a six-headed figure on the reverse (1956:46-48). Based on this report both Dasgupta (1974:208-209) and Ahmad (1985:106-107) refer to a six-headed Skanda-Kärttikeya on the reverse of these coins. I have not, however, been able to view these coins or photographs of them, which makes an informed discussion of them difficult. There are no reports of an inscription which would identify the six-headed figure, but an identification of Skanda- Kärttikeya or Sasthī seems reasonable. As they were also found with Yaudheya class three coinage, and in an area of other class three coin finds, we might also assume that they are related to Yaudheya class three coinage coming either just before it or just after it. These Bhanuva coins do suggest that the six-headed form of the god, or goddess, had some local popularity in this region either before or after the class three issues.

33O. P. Singh also sees these dots as earrings on the obverse and reverse figures on Yaudheya issues (1977:121-122). He feels the appearance of these earrings may be related to the samskāra called karnavedha (1977:121), which has a literal meaning of "ear-boring" (Monier-Williams 1999:257). A brief explanation of this samskära can be found in Pandey (1998:102-105). Whether or not the earrings found on Brahmanyadeva and his female counterpart are simply for decoration or reflect the practice of a life cycle ritual, as Singh suggests, is not clear to me.

141

Related coins are illustrated in figures 19c, 20 and 21.34

These obverse depictions are closely related to some of the reverse depictions of the female figure

from these class three Yaudheya coins. This reverse figure is likely Sasthi.35 One such reverse image is

illustrated in figure 22. Here Sasthi has three or four balls over her head and oblongs running down either

side of her head. The implications of Sasthi's appearance on these coins and the iconography of

Brahmanyadeva on these particular single-headed types will be discussed in more detail later.

4.2.10 Yaudheya Single-Headed Coins: The Single-Headed Brahmanyadeva with dots and bars

This series of coins employ dots and bars around the head of Brahmanyadeva. The obverse of the

coin depicted in figure 23 is mostly off the flan, but the head of the figure is clear. There is a single head

with two slightly curved oblong figures running vertically along either side of his head. Above the head are

three small balls arranged horizontally. There are also traces of two Brähmi letters visible, which appear to

read 'bhäga' possibly for bhägavata. On the reverse is a deer looking to the proper right, and there is an

unclear object above the deer's back. Much of the reverse is worn and off the flan. This coin is off the flan.

34The obverse of the coin illustrated in figure 19c shows a single-headed Brahmanyadeva with approximately ten small balls arranged around his head. His left hand rests on his hip, and his right hand holds a spear. The reverse is off the flan, but a deer looking to the right is visible. Over the deer's head is a double S device, and to the far right is a building. The obverse of figure 20 shows a single-headed figure with four to five dots arranged in a semi-circle over his head. The figure holds a spear in his right hand, but the left is worn away. On the reverse is a deer whose head is off the flan, but who likely looked to the left. Beneath the deer is a pot and behind are traces of a tree in railing device. The obverse of figure 21 is slightly off the flan, and the far left of the coin is worn and cut off in the photograph, but a single-headed figure with balls around his head is visible. To the right of his head three balls are visible, and to the left two balls can be seen, but the left side is worn. Traces of a spear to the proper right are visible as is the Brāhmī letter "va" in the top right and the letter "ga" in the lower right. The reverse of the coin is almost completely worn away, but a head and body may be visible to the right. A related, but unillustrated coin is BM BAAA 9. The obverse presents a single-headed Brahmanyadeva with five dots arranged in a semi- circle above his head. He also has hooks along either side of his head, but the ends of these hooks face into his head. Each hook also has a dot attached to its end. The deity's left hand rests on his hip, and his right hand holds a spear. There is an unclear object to the lower left of the coin. This object may represent a vase, but a certain identification is difficult to make. As in BM 76, I think the hooks on either side of his head represent ears and the final dots earrings. The reverse depicts a deer facing to the proper right of the coin towards a building. There is a pot-like object above the deer and a svastika to the upper left.

35The figure of Sasthi will be discussed in more detail later.

142

It is hard to see if the obverse figure holds a spear. The partial inscription, however, suggests that it is a

Yaudheya Brahmanyadeva coin.36

36 Also representative of this group is the coin illustrated in figure 24. The obverse shows a single- headed figure with three dots above the head and a horizontal bar above the dots. On either side of his head are dots on the upper part of his head and what appear to be oblongs down the lower half, but the lower left is worn away. His right arm and the top of a spear are also visible. There is some Brahmi inscription visible, which appears to read 'bhagava', likely the original read bhagavata, part of the usual Yaudheya inscription. Much of the obverse is off the flan and worn away. On the reverse is a deer looking to the right at a building. The reverse is also off the flan. The unillustrated BM ITY 74 coin depicts on the obverse a single- headed Brahmanyadeva, whose posture is similar to later Yaudheya coin depictions of Skanda-Kārttikeya. Above his head, however, sits what may be a stylized turban or top knot with three balls attached to it. One ball sits on top while the other two hang off the side. There are also two balls attached to either side of the figure's head about halfway down the head. Just below the head are two more balls, one on either side. His left hand rests on his hip, and his right hand holds a spear. There is a sash around his waist, and he may wear boots. There is also an object, possibly a vase, in the lower left, but the coin is not clear. On the reverse of the coin is a deer with clear facial features, which looks to a building to the proper right. Above the deer's head is a double S sign, and to the left is a pot with plant as well as a svastika sign. The obverse of figure 25 features a single-headed figure with four oblong shapes (two on each side) along the side of his head. Just below his head are two balls. The top of his head is worn. His left hand rests on his hip, and his right hand holds a crooked spear. On the reverse is an animal whose face is worn off, but appears to face the proper left. The remainder of the reverse is worn away. Related coins are IIRNS photograph number 609/2- 5, 256-3 and 612/2-7. Similar coins have been discussed by A. M. Shastri and D. Mukherjee. Figure 19 (a and b) illustrates two of the coins discussed by Mukherjee. The head of the figure on the first coin is slightly off the flan. Mukherjee describes the head as follows: "The head of the deity is shown with a peculiarly fashioned ornament (or stylized representation of a six-headed Kärtikeya) consisting of five globules at the top and the other two are hanging on either side of the face attached with bar-like objects" (1983:21-22). The figure's left hand rests on his hip, and his right hand holds a spear, though the top of the spear is slightly off the flan. There appears to be a sash around the figure's waist and traces of a Brahmi inscription and beaded boarder around the edge. Mukherjee reads the partial inscription as "brah ... ku", and she feels the original likely read svāmino brahmanya devasya kumārasya (1983:22). Her reading and understanding of the coin appear to be correct, though I will discuss the five balls above the head in more detail later. The reverse of the coin is also off the flan, but a deer looking to the proper right is visible as is a building in the far right. Mukherjee also sees a double S symbol and a vase over the deer, but these are not visible in the photograph she supplies. The second coin I have illustrated from Mukherjee depicts a similar figure. Here four balls over the deity's head are clearly visible as is a small bar over them. Two oblongs run down either side of his head and end in small balls. The reverse depicts a deer facing a building with a double S symbol over its head. The author notes that these coins come from a private collection (1983:21). Shastri discusses Kuninda and Yaudheya coins found in a hoard from Mandi district in Himachal Pradesh. All of the Yaudheya coins found in the hoard are from Allan's class three, and all of the Kuninda coins are post-Amoghabhuti (Shastri 1984:24). Early pre-Kusāna issues of Kuninda coinage have the word or name Amoghabhuti in their legend. Kuninda issues without this legend are usually later and considered post-Kusāna (Bopearachchi and Pieper 1998:49). One of the coins he discusses depicts a single-headed Brahmanyadeva with four to five balls arranged above his head. Over these balls is a single horizontal line. Two oblong shapes are on either side of the deity's head, and there are two balls just below these shapes. His left hand rests on his hip, and his right hand is raised to a spear. There are traces of a Brahmi inscription

143

4.2.11 The Yaudheya Single-Headed Coins: The Single-Headed Brahmanyadeva with a T-device

The next example, which is illustrated in figure 26, depicts a single-headed figure with a T-like

object over its head. Other than the head, only the right shoulder and arm are visible. On the reverse is a

deer looking to the right with a pot over its back. There are traces of a tree in railing device to the proper

left. This coin's worn state makes for a questionable identification of the figure on the obverse as

Brahmanyadeva. While the T-like object over the head is only found on this example from the single-

headed variety, it is common on the six-headed variety of Brahmanyadeva coinage. The existence of the six-

headed type supports the idea that the obverse figure on the coin in figure 26 represents Brahmanyadeva.

4.2.12 The Yaudheya Single-Headed Coins: The Single-Headed Brahmanyadeva with Single Bar

The next type is represented by the coin illustrated in figure 27. On the obverse of this worn coin

only the head and body of a crudely wrought figure and a spear are visible. There appears to be a small

horizontal bar above the figure's head. On the reverse is a deer facing to the right. There is a pot with a

plant coming out of it above the deer. The figure with a spear allows me to suggest that this figure is

Brahmanyadeva. This type of depiction is also common on six- headed Brahmanyadeva coinage, which also

suggests that this single-headed variety is the same deity.

4.2.13 The Yaudheya Single-Headed Coins: Single-Headed Brahmanyadeva with rays

This group have been placed together for their similar use of oblongs or rays coming from a single

head. The first example is illustrated in figure 28. The obverse is worn, but a single-headed figure is visible

with five rays arranged around his head. What may be traces of a spear are visible to the right. The reverse

to the proper left, which Shastri reads as the syllables "bha" and "ya" (1984:32). He feels that the balls above the figure's head represent heads, but I will comment on this in detail later (1984:32). The reverse depicts a deer looking to the proper right towards traces of a building. Shastri also sees traces of a peaked hill on the reverse, but this is not clear from the illustration he provides (1984:32). Another scholar, L. C. Gupta, has documented a single headed Brahmanyadeva coin which has dots around his head, but describes the figure as the Buddha. The reverse of the coin appears to show a single headed female figure with seven rays coming from the head and two dots beneath her head, but the author describes this figure also as a Buddha (1986:36-37). I think the author is badly mistaken. The figures represented are, I think, Brahmanyadeva and a female figure, who is likely Sasthī.

44

of this coin is not available. The coin illustrated in figure 29 depicts on the obverse a single-headed

Brahmanyadeva with seven rays around his head, and these rays are shaped like inverted L-s. His left hand

rests on his hip, and his right hand holds a spear. A pot is visible to his left as are traces of a Brāhmī

inscription. The reverse shows a single-headed female figure with nine rays around her head. Her left hand

rests on her hip, and her right hand is raised. Her breasts are represented by two balls. To her right is an

arched mound topped with a nandipada device. To her left is a tree in railing device. This reverse image

helps to demonstrate that the rayed-head device is also employed on the female appearing on the reverse of

some Yaudheya issues.37

4.2.14 The Yaudheya Single-Headed Coins: Single-Headed Brahmanyadeva Coins with rays and a T

There is only one certain example of the next type, which blends the T-shaped top with ray-like

lines. This coin is illustrated in figure 30. On the obverse a figure's single head is topped by a T, and there

are four (two on each side) lines along the side of the head which point upwards. The right hand is raised,

but the proper right of the coin is worn, and a spear is not visible. The left hand rests on his hip. There are

also traces of a sash on his hips. The reverse is worn, but an animal facing the right is visible, as are traces

of a Yaudheya style tree in railing to the left. The worn condition of the coin does not allow for a firm

identification of the deity, but the similar posture of the obverse figure and the similar reverse pattern

37 An unillustrated example is BM 58. The obverse of this coin shows a single-headed Brahmanyadeva with five oblongs arranged in a semi-circle around his head. The oblong just above his head is blunted and looks more rounded. His left hand rests on his hip, and his right hand holds a spear. On the reverse is a female figure with an uncertain number of heads. Her left hand is on her hip, and her right hand is raised. To the left of the coin is an unclear mound, and to the right is a four-tiered tree in railing. A related unillustrated coin is BM 60. The obverse is similar to BM 58 except the top oblong is not blunted. His left hand is on his hip, and his right hand is raised. There is a spear to the far right, but his right hand does not seem to hold it. On the reverse of the coin is a female figure, but, due to wear, she has an uncertain number of heads. Her left hand is on her hip, and the right hand is raised. To the proper left is a mound, and to the right is a four-tier tree in railing. Below the figure is a worn wavy line. BM 61 also belongs to this group. It is a worn coin, but a single head with five oblong rays (three to the left and two to the right) and a ball over the top of the head are visible on the obverse. The figure's left hand rests on his hip, and his right hand is raised to a spear on the far right. On the reverse is a female figure, who has a single head with five oblong shapes arranged in a semi-circle over the top of her head and two small balls below her head (one on either side). Her left hand is on her hip, and her right hand is raised. To the right is a tree in railing. A similar coin type is BM plaster cast number 13.50.

145

suggest that the figure is Brahmanyadeva. The use of rays combined with a T-like symbol around a figure's

head appear to be Yaudheya innovations.38

4.2.15 The Yaudheya Class Three Coin Legends

A final aspect of the appearance of these class three coins that must be discussed is their legends.

The legends on these coins have sparked a great deal of controversy, which I will cover in this section. No

single copper class three coin has come down to us with a complete legend. All of the copper coins I have

seen are well worn and usually only a fraction of the legend can be seen on any one coin. We do, however,

38I have placed these coins in this note because all of the coins that follow represent questionable varieties, which may or may not be Yaudheya, and which may or may not represent Brahmanyadeva. IIRNS photograph number 611/1-19 is unillustrated. The obverse depicts a single-headed figure who may wear a headdress. There are five dots around the head. Two of these dots are to the left of the head and three are to the right. The right side of the coin is worn away, but the left arm is visible and hangs down towards his hip. There is also some beaded border to the left. On the reverse of the coin in the upper right is a mound which is slightly off the flan. The lower right is worn away. To the left is an uncertain object and a wavy line near the bottom of the coin along with traces of a beaded boarder. Due to the degree of wear on this coin and the unique figure on the obverse, I am not sure if this figure does represent Brahmanyadeva, but it is classed as a Yaudheya coin in the IIRNS. IIRNS photograph number 610/2-10 is also unillustrated. The worn obverse shows a single-headed figure with five dots or balls above it. The balls form a sort of pyramid over the figure's head. Just below the head are two more small balls, one on either side. The remainder of the coin is completely worn away; a certain identification of the figure is not possible. The head does share the above five dots and the attempt to represent earrings with dots below the head that we have seen on other Brahmanyadeva coins, which suggests that this head also belongs to a Brahmanyadeva figure. IIRNS photograph number 609/3-13 is also unillustrated. The coin is worn, but its obverse depicts a crudely formed single-headed figure. His left hand is on his hip, and his right hand holds a spear. The reverse depicts an animal that may be a deer, but its head is worn away. It appears to face the proper left of the coin. The crude artwork on this piece makes it suspect, and it may not be a Yaudheya coin. Figure 31 depicts on the obverse a crude single-headed stick man-like figure with a shaft to the right, which may represent a spear. On the reverse is a crudely carved animal facing the proper left. An exact identification of this figure is not possible. IIRNS photograph number 612/1-18 is unillustrated. The obverse depicts a single-headed figure whose head is either cut in half on a diagonal or who looks to the right and wears a hat of some sort. Out of the top of the figure's head or hat are two lines which arch down to below his head and have small ball finals. His left hand rests on his hip, and his right hand is raised to a spear. There are traces of a sash on his hips. The reverse is very worn, but a pot is visible in the centre. It is not clear if this figure is meant to represent Brahmanyadeva or if it is really a Yaudheya coin. This unusual coin may have an equally unusual prototype. Figure 32 is an Agathocles coin recovered from the Ai-Khanum excavation. A. K. Narain argues that the figures on this coin represent Vasudeva and Balarāma (1973:76). He also suggests that the figure with the cakra could also be the Buddha (1973:76). Both figures have an unusual headdress, part of which includes two lines arching down with ball finals as in the possible Yaudheya coin. While this Agathocles coin is Indo-Greek, it has no other Indo-Greek prototype and may represent an indigenous understanding of the headdress, though both figures are dressed as Greeks. It is not clear if this unique Yaudheya coin is modelled from this equally unique Agathocles coin.

146

have enough of these coins that the legend on the copper coins has been reconstructed by Allan to read

"Bhägavata-svamino Brahmanya-devasya (or sa) Kumarāsya (or sa)" or "Of Kumāra the divine lord

Brahmanyadeva" (1936:cl). On variety h of Allan's class three the word "darma" appears on the reverse of

the coin, which Allan thinks is "presumably for dharma" (1936:cl). None of the copper issues has the name

'Yaudheya' on them, and their exact identity would be a mystery were it not for the single silver issue,

which, as noted in the body of this work, Allan reads as: "Yaudheya-bhagavata-svamino Brahmanya (sa or

sya)" or "Of Brahmanya (a name of Karttikeya), the divine lord of the Yaudheyas" (1936:cxlix). This

legend allows us to connect the name Brahmanyadeva or simply Brahmanya as well as the unique depiction

of the six-headed deity on the obverse with the Yaudheyas.39 Not all scholars, however, concur with Allan's

reading of these legends. Some scholars have questioned the presumed relationship between the single

silver issue and the coppers attributed to the Yaudheyas.40

39 Allan was not the first scholar ro attempt to read these legends. Among the earlier scholars the two most important are Cunningham and Smith. Cunningham read the silver legend as "Bhāgavato Swāmina Brähmana Yaudheya" and the copper legend as "Bhāgavata Swāmina Brāhmana Devasya" (1963:78). He notes that his reading of the copper coins' legend was done from many coins (1963:78). He does not provide specific dates for these coins, but regards what were to become Allan's class six coins as from about 300 CE (1963:76), and what became Allan's class three to "a slightly later date" (1963:77). Cunningham does, however, note that it is the silver issue that allows him to attribute the copper coins to the Yaudheyas (1963:78-79). Smith describes the Indian Museum collection of copper Yaudheya coins and reconstructs the legend to read: "Bhāgavatah svamino Brahmanya devasya" or, as he translates it, "('coin) of Svāmī (a title) Brahmanya-deva worshipper of Vishnu'" (1972:182). He also notes the appearance of the words "drama" and "Kumāra", but he is unable to explain them. He does suggest that Kumāra may represent "a chief distinct from Brahmanya-deva" (1972:182). Clearly, however, Smith makes an error in assuming that 'Brahmanyadeva' refers to a chief "who calls himself Svāmi Brahmanya Yaudheya" (1972:165). It is generally agreed that Allan's readings of these legends represents an improvement over these two earlier efforts.

40The relationship of the single silver issue to the copper issues has been debated by four scholars: N. Ahmad, P. L. Gupta, D. Handa and A. M. Shastri. The debate begins with Ahmad who argued that the copper coins of Allan's class three were issued by a 'tribe' called Kumara, and only the silver coin of class three actually belongs to the Yaudheyas (1977:186). His argument is largely based on the absence of the name Yaudheya on these copper issues. Regarding the typological similarity between the silver and copper coinage, Ahmad argues that these coins are also similar to Kuninda coinage, but this does not mean that both sets of coinage were issued by the Kunindas. He states a similar argument for Indo-Greek coin types which are similar to other local Indian issues, but clearly issued by unrelated groups (1977:187). He does not, then, regard similar typology as a strong argument. He also notes a second potential counter argument in that the word 'Kumära' appears in the genitive singular on these coins, while most state and 'tribal'

147

names appear in the genitive plural on coinage. He counters this argument by suggesting that Kada and Upagoda coinage has the state's name in the genitive singular, though he does not explain any chronological or geographical links between any of these groups (1977:186-187). Ahmad's arguments were first countered by Shastri. Shastri argues that for 'Kumara' to be a state's name it would have to appear in the plural as do other 'tribal' or state names on coinage that appear at this time. Kumära appears in the singular and is, then, an adjective of Brahmanyadeva (1987:40). Shastri also notes that on Huviska's Skanda-Kumara issues, Kumara appears as an adjective of Skanda, giving precedent for Kumara as an adjective of Brahmanyadeva on the copper issues in question (1987:41). Finally, on the obverse of one of the single-headed Brahmanyadeva coins found in the Mandi hoard, he reports seeing the Brahmi word 'ya'. He argues that the 'ya' may be a short form for Yaudheya. He feels

(1984:35). this'ya' counters Ahmad's premise that the name Yaudheya does not appear on these copper issues

Defending Ahmad's argument and countering Shastri was P. L. Gupta, who takes Ahmad's argument a step further. He argues that Allan's reading of 'Yaudheya' on the silver coin is conjecture. Gupta does not find the word on the coin and feels that this coin may have been issued by the Kunindas or by a group near them (1987:28-29). He also argues that Kumāra does not stand for Skanda-Kārttikeya or Brahmanyadeva on these coins. He feels that there is not a single instance of Kumara being used for Skanda-Kārttikeya in pre-Common Era India, and he places these coins in the first century BCE (1987:29- 30). In an earlier paper, however, Gupta claimed that these coins were from the second century CE (1977:63), and he does not explain his shift in proposed dates. This earlier date allows Gupta to claim that the Yaudheyas are issuing their class two coins at this time. This argument then allows him to argue that it is unlikely that the Yaudheyas issued two completely different coin types at the same time, hence the 'kumära' coins are not theirs (1987:29). In response to Shastri's grammatical concern for Kumära being the name of a state, Gupta acknowledges the problem but asks: "can not there be any exception" (1987:30)? Gupta does not, however, present any evidence to allow for such an exception. Next, Gupta argues against Shastri's idea that the 'ya' he reads on a coin stands for Yaudheya. He argues for reading the phase on some of these coins with the reverse darma legend as "darmattaya" (1987:35-37). Gupta agrees with Ahmad's Kumāra theory concerning the copper class three coins, and argues in kind that this term 'darmattaya' occurring on a few of these class three coppers is also the name of a 'tribe' (1987:30, 38). That some other letters were attached to darma or drama on some copper issues was first noticed by L. C. Gupta. This Gupta, however, misread the legend as reading: "Bhagvato Brahmanya Devasya Dve Mätteya", or "of the lord Brahmanya who has 2 mothers" (1985:14). He defends this notion of two mothers by stating that one mother was Parvati and the other the six Krtikās (1985:14). As my textual analysis has demonstrated, however, the number of Skanda-Kärttikeya's mothers can be much higher than two, and most scholars reject L. C. Gupta's reading of the legend (P. L. Gupta 1987:34). In a related article L. C. Gupta agrees with Ahmad's argument concerning the copper coins and returns to his theory that some of these coins read "Dve-Matteya" (1986:36). In a similar vain, he feels that these 'dve-matteya' coins are Kuninda issues and not Yaudheya (1986:36), but L. C. Gupta's understanding of these coins is poor and his reading of the legends unsubstanciated. Handa argues against P. L. Gupta and Ahmad. He begins by stating, with reference to Gupta's doubt concerning the existence of the word 'Yaudheya' on the silver coin, that: "The illustrations in Cunningham's Coins of Ancient India, P1. VI.9 and Mitchiner's Types 598 and 940 are so clear that they do not leave even an iota of doubt regarding the existence of the name of the tribe on the said coin between III o'clock and VI o'clock" (1991:71). Concerning Ahamd's argument for the Kumara tribe and Gupta's support of it, he notes that literary references to the Kumaras indicate that they lived in Daksinapatha far from the area where these coins were found. He also argues that there is "copious literary evidence regarding the association of the Yaudheyas with Kārttikeya depicted on these coins" (1991:72). Here,

148

J. N. Banerjea was one of the first scholars to question Allan's reading. He read the legend on the

silver coin as: "Bhagavata (h or to) svämino Brahmanya (sa or sya) Yaudheya (sa or sya)", which would

mean "'of the worshipful lord Brahmanya-deva Yaudheya'" (1951:161). He reconstructs the copper legend

as: "Bhagavata (h or to) svāmino Brahmanya-devasya (or sa) Kumārasya (or sa)", which translates as, "'of

(really, 'coin of') the worshipful lord Brahmanya- deva Kumara'" (1951:161). On the basis of this reading

Banerjea regards Skanda-Kārttikeya as the Yaudheya par excellence. He claims the Yaudheya state was

dedicated to Skanda-Kärttikeya, and that the deity was regarded by the group as their spiritual and temporal

lord (1951:162; 1956:142; 1960:43). He regards this dedication to have occured sometime after the

Yaudheya defeat by Rudradäman (1951:162). To support these arguments he mentions a sealing from Bhita,

which he suggests demonstrates that a Maharaja of the third or fourth century CE devoted his kingdom to

Mahāsena. He also uses similar legends from Kuninda coinage to back his claim (1956:142).

Banerjea's dating of these coins to Rudradäman will be rejected based on the chronology that I

will propose later. The Bhita sealing creates a greater problem, however. This Bhita sealing is mentioned by

a number of scholars who follow Banerjea and hope to defend his point, and it is worth a brief examination

to evaluate its significance. Banerjea's comments are largely based on Sir John Marshall's translation and

discussion of this sealing. Marshall reads the sealing's legend as: "Srī Vindhyavedhamahārājasya

Maheśvara-Mahāsenātisrstarāyasya Vrsadhvajasya Gautamīputrasya' (Banerjea 1956:142). He translates

however, Handa seems to overstate the amount of textual evidence for the Yaudheyas. He is also critical of Gupta's argument for a 'Darmattaya' tribe. He regards this as "a very ridiculous suggestion", and observes that there are no references to this supposed tribe in literary or epigraphic sources (1991:72). He reads the phrase as "darma-traya", possibly meaning tetradrachm (1991:72; Gupta 1987:37). Thus, Handa regards Allan's attribution of these copper issues to the Yaudheya's as correct. There are other articles put forth by these scholars debating these points, but the main points have been summarized above. More recently W. Pieper has reviewed this debate and concludes: "Typologically, however, at least the Karttikeya/Lakshmi specimens are so closely connected with the silver type that in this case this can reasonably be taken as the decisive argument. And as in spite of design varieties, all other class 3 coppers likewise show the same inscription, they should as well be regarded as Yaudheya emissions" (Bopearachchi and Pieper 1998:52). I think that Shastri's, Handa's and Pieper's arguments carry the most conviction, and the copper coins and silver coin must be considered as issues of the Yaudheya state.

149

it as "Of the illustrious Mahārāja Gautamīputra Vrsadhvaja, the penetrator of the Vindhyas, who made over

his kingdom to the great Lord Kärttikeya'" (Banerjea 1956:142). Marshall goes on to argue that this sealing

demonstrates a custom by which rulers gave over their kingdoms to their personal deity and ruled as agents

of that deity. Marshall cites a similar case for rulers of Travancore, and Banerjea interprets the Yaudheya

and Kuninda coins in a similar fashion (Banerjea 1956:142).

Marshall's translation of the legend, however, requires an immediate correction because

'Kärttikeya' does not appear in the legend. The sealing refers to Mahäsena, which may or may not be

understood as Kärttikeya in this context. There are also a number of other possible readings of the legend.

Mirashi takes apart the compound as reading: "Maheśvarena Mahāsenena atisrishtam rājyai yasmai

tasya", and he interprets the legend as meaning the king thought he had obtained his kingdom through the

grace of Mahasena (Thaplyal 1972:47). Thaplyal reads the compound as: "Maheśvarasya Mahāsenayā

atisrishtam rajyam yena tasya", which would convey the idea that the king created his kingdom with the

help of his great army (1972:47). I cannot offer my own reading of this legend, but each of these three

attempts strikes me as a valid reading of the sealing. The sealing may not, then, refer to a king dedicating

his kingdom to Mahasena. The sealing was also found in Bhita and is at least 200 years after the Yaudheya

class three coins. A direct link between this sealing and the Yaudheyas is difficult to establish as is the

reading of the sealing. Its use in an argument concerning these Yaudheya coin inscriptions is dubious and

does not demonstrate Banerjea's point well.

Banerjea also discusses the legend on Kuninda coins to back his point concerning the Yaudheya

legends. This legend reads: "Bhagavata Chatreśvara Mahātmanah", which translates to: "'(coin) of the

almighty Chatreśvara" (Dasgupta 1974:98).41 The legend is similar to the Yaudheya class three legend, and

Banerjea draws the same conclusion from it as he does with the Yaudheya issues (1956:142). The class

three Yaudheya coins and the Kuninda coins in question share a great deal in common. The Kuninda coins

41Chatreśvara is understood to be an epithet of Siva (Banerjea 1956:118; Dasgupta 1974:98).

150

show a male divinity (presumably Chatreśvara/Siva) on the obverse with the above inscription, which is a

similar coin design as the Yaudheyas. Earlier Kuninda coinage has a female divinity on the obverse, who

closely resembles the single-headed female on the reverse of Yaudheya class three coins. Both coin types

also depict a deer and very similar reverse devices. The two were also geographically close and may well

have been trading partners.

We must, however, be cautious in interpreting these legends. The legend does imply a special

place for Brahmanyadeva within Yaudheya society, but we lack a clear archaeological or textual context for

this group that would help us in determining the exact nature of this deity's role. We must also consider that

the Yaudheyas and Kunindas borrowed their style of coin legend from the Indo-Greeks and Indo-Scythians,

who usually placed the name of their ruler in the genitive on the coin. The Yaudheyas and Kunindas do not

appear to have had a monarchy at this stage of history and may have placed Brahmanyadeva's and

Chatreśvara's names in the genitive in an attempt to copy a recognized currency. Such a placement of the

name could imply an understanding of Skanda-Kärttikeya as a king, but I think it is safer to argue that

Skanda-Kärttikeya was revered by the Yaudheyas at this time.

Another scholar, R. C. Agrawala, challenges Allan's understanding of the term drama in these

legends. Agrawala reads drama as standing for dramma and argues that it is a denominational term current

in north India until the mediaeval period. He traces dramma from the Greek drachma and Persian dirham

and supplies an inscription from Taxila where he claims dramma is used in reference to coinage (1955:64).

He thus reads the coin legend as "Brahmanyadeva drama" or "coin dedicated to Brahmanyadeva"

(1955:64). He does note, however, a personal correspondence with P. L. Gupta in which Gupta doubts any

reference to dramma on Yaudheya coins and points out that Greek drachmas refer to silver currency and

not copper (Agrawala 1955:65).42

42These comments by Banerjea and Agrawala have had a significant impact on those who followed them with most scholars either reacting against or agreeing with them. J. Prakash does not regard Banerjea's reading of the legend as convincing. He states that there is no evidence for such a reading (1965:134). He argues that the silver and copper coins are designed to eulogize the deity. He translates the copper issue as

151

reading: "(In the name or in honour) of the worshipful (or divine) lord, Brahmanyadeva Kumāra, (of the) Yaudheyas" (1965:134). It should be noted, though, that the word 'Yaudheya' does not appear on any of the class three copper issues, and there is no justification for including it in this translation. He also regards the use of 'svamino' on these coins as borrowed from the Western Ksatrapas (1965:134). Finally, he discusses the meaning of the word "dramma" occurring on some of these coins and feels it means "coin" (1965:135). Sircar notes Allan's readings of the coins and supplies his own translation, but the translation does not differ remarkably from Banerjea's (1968:213). Sharan reviews a number of opinions on these legends, including Allan's and Banerjea's, but presents no new discussion of them (1972:93-95). Dasgupta notes Allan's reading of both coin types and several other possibilities for the intended reading. One alternate reading for the silver issue he suggests is: "Yaudheyānām Bhagavatah svāmino Brahmanyasya" or "'(coin) of the Yaudheyas (and) of the divine lord Brahmanya' or '(coin) of bhagavan Brahmanyadeva, lord (svami) of the Yaudheyas'" (1974:216). A second potential reading he offers is "Bhagavatah svāmino Brahmanyasya Yaudheyānām“ or "(coin) of the lord (svāmî) Brahmanya god (bhagavān) of the Yaudheyas" (1974:216). He does, however, conclude: "All these different readings, however, do not make any substantial change in the meaning" (1974:216). He agrees with Allan's reconstructed reading of the copper legend and translates it as: "'of the divine lord Brahmanya-deva (alias) Kumära"" (1974:216). Following Banerjea, he does think that the coins were issued under the name of the deity, and that the Yaudheya territory was administered under the name of the god (1974:216-217). A number of scholars who contributed to the book, Seminar Papers on the Tribal Coins of Ancient Inda, comment on these topics. A. M. Shastri notes regarding these coins that: "We may thus conclude that the Yaudheyas thought of god Brahmnayadeva-Kumāra [sic.] as the real ruler of their state and of themselves as the carriers of the administrative work on his behalf" (1977:93). Another argument Shastri presents concerns the appearance of darma on some of the copper issues, which he claims stands for the denominational term drachma (1984:36, 1987:42). N. Ahmad thinks the word drama is not actually a word at all, but part of the surrounding devices. He does, however, state that if they are letters than they read as "darma" (1977:184-185). P. L. Gupta takes Allan's reading of the coin legends and feels that these legends and the Kuninda legends "suggest some kind of theocratic government for their issuing tribes" (1977:63). J. P. Singh regards the Yaudheya silver as likely inspired by Western Ksatrapa coinage, though he does not explain the exact nature of the relationship he sees between these two sets of coinage (1977:6). He does, however, note the Yaudheya use of the word svami and its association with the Western Ksatrapas, who used it with sub-kings. Singh feels that this Ksatrapa use of the term works against the idea that Brahmanyadeva is used in the sense of a temporal ruler on Yaudheya coinage because he does not think that the Yaudheyas would present him as a sub-king (1977:12). Finally, he regards the Yaudheya coins as commemorative: "issued in honour of the deities to whose grace they attributed their independence" (1977:6). We do not, however, have a sufficient grasp of Yaudheya history to support Singh's notion regarding Yaudheya independence. Skanda-Kārttikeya scholars typically echo Banerjea's conclusion in their accounts of these coins. P. K. Agrawala notes both Allan's and Banerjea's reading of the legend and Banerjea's conclusions. Based on the legend and Banerjea's comments, he states: "Evidently, the Yaudheyas devoted special service to their god. They had dedicated their state to the god and the republic was ruled on his behalf" (1967:42). Chatterjee takes Allan's reading (1970:35), but shares Banerjea's conclusion stating: "These coins clearly show that this tribe dedicated their kingdom to the War-god Skanda-Kärttikeya" (1970:36). Navaratnam feels the Yaudheyas also dedicated their state to the deity based on the coin legend (1973:100). O. P. Singh notes that the coins are issued under the name of the deity, and that the Yaudheya state must have been theocratic (1978:71-72). With Banerjea, Sinha suggests that the deity was regarded as both the temporal and spiritual ruler of the Yaudheyas (1979:77). Sinha also argues that administrating a state in the name of a deity was not unusual during this period, and he mentions the Bhita sealing and the Kunindas in this context

152

Finally, I will comment on the name given to Skanda-Kärttikeya on these coins: Brahmanya,

Brahmanyadeva and Kumära. The term Brahmanya did occur in relation to Skanda-Kārttikeya in the

Äranyakaparvan, but in that text it is not obviously an epithet of the deity, and I have largely taken it as an

adjective meaning Brahminical in that text. On these coins, however, Brahmanya appears as more of a

proper name. Whether 'Brahminical' works as a translation of it in this context is not clear to me. The use

of Brahmanyadeva implies that the Yaudheyas understood themselves to be worshipping a god named

Brahmanya. The name Kumära certainly, however, links the figure on these coins with the cult of Skanda-

Kärttikeya. I will argue below that aspects of these coins relate the figures on them to Graha cults. The use

of the name Kumara on these coins may, then, relate to the inauspicious connotation of that name I have

discussed earlier. The exact meaning and intent of these class three coin legends are not clear, however, and

I am not able to speculate on the Yaudheya understanding of these names.

4.2.16 The Foreign Influence on Yaudheya Class Three Issues

In this section I will argue for direct stylistic borrowing by the Yaudheyas from foreign or

Hellenized numismatic sources. I will demonstrate that the coins discussed in sections 4.2.9 and 4.2.14 are

examples of direct borrowing from Indo-Greek and Indo-Scythian numismatic issues. I will also suggest

that the coins discussed in section 4.2.10 may also be examples of borrowing from these foreign sources.

Since no scholar has yet suggested a relationship between these Yaudheya issues and foreign coins I will

present my case for this in some detail. I will also argue that the very idea of representing Brahmanyadeva

with one head is influenced by foreign coin designs and, possibly, foreign understandings of the appearance

(1979:77). Thakur cites Allan's reading of the legend, but shares Banerjea's conclusion in stating: "The legend is interesting as it shows that the Yaudheyas had dedicated their State to god Kärttikeya whom they regarded not only as their spiritual but also their temporal ruler" (1981:48). This author also notes the Bhita sealing in relation to this point (1981:49). Mani notes that the legend demonstrates that the Yaudheyas regarded the deity as their spiritual and temporal ruler (1990:70), and Rana notes that the legend shows that the Yaudheyas dedicated their state to the god (1995:22).

153

of deities.

I will begin with the coins from section 4.2.9, which show a single-headed Brahmanyadeva with

five or more smaller balls arranged around his head. There are, I think, Indo-Greek and Indo-Scythian

prototypes for this style of Yaudheya coinage. One of the earliest Indo-Greek examples of a deity with balls

or dots arranged around its head comes from silver issues of Demetrius I. This Indo-Greek king's rule is

dated to c. 200-185 BCE (Narain 1955:4).43 An example of his silver tetradrachm coinage is illustrated in

figure 33. The reverse of the coin is described by Gardner in the following manner: "Young Herakles, in 1.

hand, club and lion's skin; with r. hand, crowns himself with ivy-wreath" (1886:6). The 'ivy-wreath' is the

five balls arranged in a semi-circle around Herakles's head, which are similar in size to those depicted on

the Yaudheya coins from section 4.2.9.

That these balls around Herakles's head on the Demetrius coin represent a wreath as Gardner

suggests becomes clearer when we examine a silver tetradrachm issue of Euthydemus II illustrated in figure

34.44 On the reverse of this coin Herakles is depicted holding a wreath made of six balls in his right hand,

and he is also crowned with a wreath made of six balls. At least in the case of Herakles, these balls over the

head of the deity represent a wreath.

This iconographic convention remains on Indo-Greek coinage until the arrival of the Scythians,

who, as we shall see, adopt the convention. The wreath convention is seen on the coins of the Indo-Greek

kings Eucratides, who places it on Herakles; Heliocles, who places it on a standing Zeus; Archebius, who

places it on a standing Zeus; Lysias, who places it on Herakles; Antialcidas, who places it on an enthroned

Zeus; Apollodotus I, who places it on Apollo; Zoilos I, who places it on Herakles; and Hermaeus, who

43Mitchiner provides dates for Demetrius of c. 205 to 171 BCE (1976 I:57).

44Narain gives c. 200-190 BCE (1955:5), and Mitchiner gives c. 190-171 BCE (1976 I:62) as dates for this king.

154

places it on an enthroned Zeus.45

Closely related to Indo-Greek coinage is Indo-Scythian coinage. The depictions of divinities on the

reverse of these Scythian coins are not Scythian inventions, but borrowed from Indo-Greek issues. As

Michael Mitchiner notes, the invading Scythians adopted a number of aspects of Greek culture when they

entered the Indo-Greek kingdoms:

When the Scythians settled in Arachosia they adopted much of the Indo-Greek cultural and economic heritage of their new homeland. This is reflected in their coinage .... Most of their coin types were borrowed from the Indo-Greeks and a basically Indo-Greek form of mint organisation appears to have been retained, as reflected in the control marks they applied to their coins. (1976 IV:439)

The pattern of representing deities with dots over their heads is one Indo-Greek feature the Indo-Scythians

use. The coin illustrated in figure 35 is from the Scythian ruler Spalagadams, who ruled under king

Vonones.46 The coin shows a standing Zeus with three to four visible dots above his head.

45Narain gives the approximate dates of c. 171-155 BCE (1955:9), and Mitchiner gives c. 171-135 BCE for Eucratides (1976 I:86). The Eucratides coin has been illustrated by Mitchiner (1975 I:87). Heliocles's approximate dates are c. 155-140 BCE (Narain 1955:12), or c. 135-110 BCE (Mitchiner 1976 II:160). The Heliocles coin has been illustrated by Mitchiner (1975 II:160). Archebius's approximate dates are c. 130-120 BCE (Narain 1955:21), or c. 80-60 BCE (Mitchiner 1976 III:206). The Archebius coin has been illustrated by Mitchiner (1975 III:210). Lysias's approximate dates are c. 120-110 BCE (Narain 1955:24), or c. 145-135 BCE (Mitchiner 1976 II:143). The Lysias coin has been illustrated by Mitchiner (1975 II:143). Antialcidas's approximate dates are c. 115-100 BCE (Narain 1955:25), or c. 145-135 BCE (Mitchiner 1976 II:147). The Antialcidas coin has been illustrated by Mitchiner (1975 II:147). Narain gives the approximate dates for Apollodotus of c. 115-95 BCE (1955:26), and Mitchiner gives 160-150 BCE (1976 II:116). The intent of these balls on Apollo may not be to represent wreaths, but rays for a radiate head. Most of Apollodotus's Apollo coinage depicts the deity with rays. These coins have been illustrated by Mitchiner (1975 II:118-119). Narain only gives a termination date for Zoilus I of c. 125 BCE (1955:23), and Mitchiner gives 150-145 BCE (1976 II:140). The Zoilus I issues have been illustrated by Mitchiner (1975 II:140-141). Hermaeus's approximate dates are c. 75-55 BCE (Narain 1955:36), or 40-0 BCE (Mitchiner 1976 III:227). These issues of Hermaeus have been illustrated by Mitchiner (1975 III:231-232). This list is based on Percy Gardner's catalogue of coins in the British Museum (1886:1-67 and plates I- XVI) and Mitchiner's catalogue volumes I-III (1975).

46Vonones carved out a kingdom in ancient Arachosia and Bannu (southern Afghanistan and north- west Pakistan) (Mitchiner 1976 IV:439) and ruled from c. 100-65 BCE (Mitchiner 1976 IV:443). The illustrated coin is a joint silver issue of Spalagadames and Vonones. Spalagadames is thought to have been Vonones's viceroy (Mitchiner 1976 IV:439).

155

Another Scythian ruler was Spalirises (Spalyrises), whose coins are illustrated in figure 36.47 The

reverse of the first illustrated coin is described by Gardner in the following manner "Zeus, facing, wearing

himation; holds thunderbolt and long sceptre" (1886:100). Above Zeus's head are five dots arranged in a

semi-circle. The arrangement of these dots or small balls is clearly borrowed from Indo-Greek coinage and

similar to those found on Yaudheya coins depicting Brahmanyadeva. Another example is a joint issue of

Azes and Spalirises. Gardner describes the reverse in the following manner: "Zeus, facing, laur., clad in

himation; holds thunderbolt and long sceptre" (1886:102). Once again, Zeus is depicted with five dots

arranged in a semi-circle over his head.48

From this evidence I think the style of the Yaudheya representations of Brahmanyadeva and Sasthī

from section 4.2.9 is not their own invention, but appears to have been borrowed from Indo-Greek and

Indo-Scythian coinage, where the dots and balls were used initially to represent a wreath. I cannot say

without doubt that the Yaudheyas did not use these dots to represent heads, but the tradition of the coinage

from which they borrow the idea does not lead to such a conclusion. It should also be noted that the number

of dots above Brahmanyadeva's head does not always equal five and, as we have seen, the same type of

dots also seem to represent ears and earrings. It seems possible that the arrangement of dots above the head

of this deity represents some sort of ornament or indication of divinity or victory, which is a direct

47 Spalirises may have been the brother of Vonones and issued coins under three different titles which likely show a chronological progression: Spalirises as a king's brother, Spalirises as king and a joint issue with Azes (Gardner 1886:100-102). Spalirises ruled from c. 65 to 40 BCE (Mitchiner 1976 IV:442) and appears to have lost his kingdom to re-conquests by Indo-Greeks (Mitchiner 1976 IV:441). The first coin illustrated is a silver issue of Spalirises as the king's brother.

48While Gardner describes the dots about Zeus's head in the Azes and Spalirises coins as a wreath ("laur."), it is not completely clear what these dots represent, because a number of Spalirises's Zeus issues represent the deity with rays for a radiate head. An example illustrated in figure 37 is a bronze issue of Spalirises. Gardner describes the reverse of this coin in the following manner: "Zeus, radiate, seated 1. on throne; r. hand extended; in l., sceptre" (1886:101). In each of these examples Zeus is depicted with six or more rays arranged around his head. By extension it is possible that the dots around Zeus's head on the silver issues also represent a radiate head or that dots and rays have become a numismatic convention whose original meaning has been lost over time. In short, it seems possible to me that the understanding of these

representing rays. dots as representing a wreath has faded and may have merged at this stage into another means of

156

borrowing from non-indigenous sources.

It may well be the case that these dots on Yaudheya coinage are deliberately ambiguous. This

deliberate ambiguity would allow the coins to be used both in a Yaudheya setting, where the figure of

Brahmanyadeva with his extra heads represented by dots would be recognized, and in an Indo-Greek or

Indo-Scythian setting, where a wreathed deity would be recognized. There are, however, more implications

to this iconographic borrowing which I will discuss below.

A related variety of single-headed class three coins is discussed in section 4.2.10. These depict

Brahmanyadeva with dots and bars around his head. As I have noted above, A. M. Shastri argues that the

five small balls or dots actually represent heads (1984:32, 34), and D. Mukherjee suggests a similar point,

but she also suggests the possibility that the dots represent an ornament (1983:21-22). Shastri's argument is

not without basis. As we have seen, one of the most common means of representing Brahmanyadeva on

Yaudheya class three coinage is to depict him with six clearly visible heads. That these dots could represent

a stylized version of this pattern is not out of the question. As with the coins discussed above, however,

these coins may have been inspired by foreign currency issues. A direct prototype for these coins with dots

and lines is not available, but I suspect they are influenced by Indo-Greek and Indo-Scythian coins, which

also employ dots around a deity's head. The coins illustrated in figure 38 show silver drachms of Demetrius.

They are physically and denominationally smaller than his tetradrachm illustrated in figure 33, but are of the

same design. Here, likely due to the smaller coin size, the balls are more like dots and have a closer

resemblance to the Yaudheya coins discussed here. The coins illustrated in figure 39 show silver drachm

and hemidrachm issues of Euthydemus II, which also tend to show the balls as dots. The borrowing here is

less direct than that illustrated above, but I suggest a similar ambiguity is employed with these coins to

allow for either an identification of an ornamented deity or a multi-headed deity.

Another example of borrowing can be found in the coins discussed in section 4.2.14, where

Brahmanyadeva is represented with rays or oblongs coming from his head. Allan and other scholars have

noted the existence of this rayed type, but have little to comment about it except that Brahmanyadeva's head

157

is radiate (1936:272).49 This style of depicting a radiate head also has its origin in Indo-Greek coinage.

Demertius is likely the first Indo-Greek king to issue such a coin. He produces one type of coinage which

shows a radiate Artemis on the reverse. On some of these coins she is depicted with many small rays and on

others with a smaller number of larger rays. In figure 40 I have illustrated two coin types which come the

closest to the Brahmanyadeva coin type in question. Also using the device are the Indo-Greek kings Plato,

who places the device on Helios; Apollodotus I, who places it on Apollo; Lysias, who places it on Herakles;

Antialcidas, who places it on an enthroned Zeus; Herliocles, who places it on a standing Zeus; Philoxenus,

who places it on a sun god; Theophilus, who places the device on Herakles; Amyntas, who places it on an

enthroned Zeus; Hermaeus, who places it on an enthroned Zeus; and Telephus, who places it on Helios.50

The Indo-Scythians also borrow this numismatic device. Vonones and Spalirises places it on an enthroned

Zeus (Mitchiner 1975 IV:463). A related device is also used by the Kusanas. Depictions of Asaixso, Helios

49Smith describes three Yaudheya coins as depicting a single headed radiate Brahmanyadeva, but these are not illustrated in his catalogue. He describes them as follows: "Single-headed god, radiate, facing; vase in r. field; traces of Brahmana deva legend", and "Single-headed god, radiate, grasping spear. Legend devasya drama Bra[h-ma]na", and "Similar; nya devasya legible" (1972:182). K. K. Dasgupta rejects Allan's attribution of a single radiate head for the coins from the British Museum and thinks that they represent a six-headed variety (1974:204). R. C. Agrawala also feels that these coins represent the deity with a central head and five heads around it, though he and Dasgupta may also be referring to the coins I have described above, some of which depict Brahmanyadeva with approximately five balls around his head (1966:201). P. K. Agrawala (1967:40), Chatterjee (1970:35) and Ahmad (1977:156) regard the oblongs as rays.

50Niether Narain or Mitchiner is sure of dates for Plato. Narain suggests c. 155 BCE as the start of his rule, but does not hazard a end date (1955:12). Mitchiner suggests c. 150 BCE as the king's date (1976 I:101). Coins of Plato are illustrated by Mitchiner (1975 I:101). These coins of Apolodotus I are illustrated by Mitchiner (1975 II:118-119). A single example of the Lysias type is illustrated by Mitchiner (1975 II:142). These coins of Antialcidas are illustrated by Mitchiner (1975 II:149). These coins of Heliocles are illustrated by Mitchiner (1975 II:163-164). The rays in these examples may actually represent a wreath. Narain dates Philoxenus to c. 125-115 BCE (1955:22), and Mitchiner dates him to c. 110-80 BCE (1976 III:202). These coins of Philoxenus are illustrated by Mitchiner (1976 III:202). Narain dates Theophilus to c. 85 BCE (1955:33), and Mitchiner dates him to c. 80-60 BCE (1976 III:215). These coins of Theophilus are illustrated by Mitchiner (1976 III:215). Narain dates Amyntas to c. 85-75 BCE (1955:35); and Mitchiner dates him to c. 60-40 BCE (1976 III:220). These coins of Amyntas are illustrated by Mitchiner (1976 III:219-221). These coins of Hermaeus are illustrated by Mitchiner (1976 III:227-232). Narain dates Telephus to c. 95-80 BCE (1955:33), and Mitchiner dates him to c. 80-75 BCE (1976 III:269). A single example of this coin type, which has Helios standing with an attendant, is illustrated by Mitchiner (1976 III:269).

158

and Miiro present the deities with a nimbate head with rays coming out of the nimbus (Göbl 1984:4, 6-9,

11, 14, 16, 19, 21-25, 27).51

The Indo-Greek coins that depict Helios and the Kusana coins tend to use the rays as a sign of a

solar deity. The other uses of the device seem to be similar to the use of the wreath; the rays indicate

divinity and sovereignty. A number of the Yaudheya types feature five rays around a single head, which has

most often been taken by scholars as a way of representing the deity with six heads. What the above

demonstrates is that the rays are one more device that the Yaudheyas borrow from Indo-Greek and Indo-

Scythian coinage. I think, the Yaudheyas may have used the rays to represent rays or heads or possibly

allowed the device to stand for both.

There may also have been some borrowing from foreign sources in the production of some six-

headed class three issues, but the nature of this borrowing is indirect when compared to the single-headed

varieties described above. The coins discussed in section 4.2.3 provide one such example. These coins

depict a six-headed Brahmanyadeva with a bar over each of his heads. The bars over the deity's head have

been interpreted by some scholars as crests of matted locks of hair. This suggestion seems to have begun

with R. B. P. Dayal (1940:111) and V. S. Agrawala (1943:29), and has been repeated by most scholars who

discuss the significance of the bars.52 None of these scholars, however, provides any convincing reasons as

to why we should regard these bars as crests of matted locks. Certainly, there is nothing in the appearance of

these bars that would prove such an identification. The most likely source of inspiration for Agrawala's and

Dayal's identification are Gupta era statues of Skanda-Kärttikeya, where his hair is long and matted and tied

back into three bunches, indicative of a youth. This feature, however, does not appear on Kusāna statuary

and is likely a Gupta innovation. Thus, we would have to read these Gupta-era matted locks back several

hundred years to see them on these class three coins -- an anachronistic reading which I find problematic.

51These page numbers for Göbl all refer to the page numbers in the catalogue section of his text.

52R. C. Agrawala (1966:201), Dasgupta (1974:219), Sinha (1979:73) and Thakur (1981:50) all regard them as crests, matted locks or both.

159

While the evidence from these coins does not, I think, allow for an exact identification of these bars, I

would like to suggest that they may represent attempts at indicating a radiate head. I base this argument in

what I have thus far demonstrated from the single-headed coins. I have shown that most of the features

around and above the head of single-headed Brahmanyadeva are borrowed from Indo-Greek and Indo-

Scythian coinage, where these features represent attempts at presenting a radiate or wreathed head. On these

six-headed Brahmanyadeva coins the number of heads presented limits the space for attributes placed above

or around each head and requires that each head be rather small when compared to the single-headed issues.

This lack of space leads me to suspect that these bars over the deity's heads are designed to indicate a

radiate or wreathed head, and that rays and dots are not used because of the limited space on the coin and

the necessarily small size of each head. A related group of coins may be those discussed in section 4.2.5.

These coins depict a six-headed Brahmanyadeva figure with T-s above each head. These T-s may also be an

attempt to represent a radiate head.

I hope the above demonstrates two things. The first is that foreign groups, who likely had contact

with the Yaudheyas, made extensive use of these numismatic conventions, and the Yaudheyas borrowed

these conventions from them. That the Yaudheyas knew of Indo-Greek coinage is attested to by the finds at

the excavation of Rohtak.53 A mould of a Menander coin along with sixteen Indo-Greek coins has been

uncovered there (Kumar 1996:102). Authentic Indo-Greek coinage is die-struck, and this mould may have

belonged to forgers. The sixteen Indo-Greek coins found are also of crude craftsmanship, which has led the

excavators to suggest that these are also the work of forgers (Kumar 1996:102). Not only does it appear that

the Yaudheyas had exposure to Indo-Greek coinage, but members of their community may also have had

53 A mound at Rohtak called Khokhrakot was excavated in the 1990s. The site appears to have been an ancient Yaudheya town. Numerous artifacts relating to the group have been recovered from the site (Kumar 1996). A fuller description of the dig and the importance of Khokhrakot will be given below.

160

experience in trying to reproduce Indo-Greek coin types.54 That the Yaudheyas could have used elements of

Indo-Greek coinage on these class three issues seems very likely given these discoveries at Rohtak. The

second point is that an active process of borrowing coin types and devices existed in north-west India.

Indigenous and foreign groups frequently borrowed from one another, but most groups borrowed from the

Indo-Greeks, which is both a sign of the success of Indo-Greek coinage as a financial tool and the artistic

merit of the coinage itself. In short, stylistic borrowing from other groups places the Yaudheyas into a

normative pattern of South Asian numismatic production for this time period.

I have demonstrated above that the arrangement of balls and dots around the head of a single-

headed Brahmanyadeva indicates foreign influence. In fact, the very idea of representing Brahmanyadeva

with a single head may also come from these foreign sources. There are no multiple-headed representations

of deities on Indo-Greek and Indo-Scythian coinage that I know. The deities represented on these foreign

issues tend towards anthropomorphism, and I do not regard their very human appearance as accidental on

this coinage. Part of the back drop to Indo-Greek, Indo-Scythian and even Kusana coinage was a trend in

the ancient world towards deifying royalty and emperors. Such trends were visible in the Hellenized world

with Alexander the Great and were also found in the Roman and Kusana empires. At times this message of

deification was presented implicitly through art and other media, and at other times the message was an

explicit state policy.

Hints at this deification appear on Indo-Greek and Indo-Scythian coinage. The general trend on

these issues is, in my opinion, more implicit than explicit. The common issues featuring an enthroned Zeus

with radiate or wreathed head or a young Herakles crowning himself with a wreath are designed, I suspect,

54The demarcation between a forger and a mint master at Rohtak is not straightforward. As will be discussed below, thousands of Yaudheya coin moulds have been discovered near Rohtak. There is little doubt that the site was used as a Yaudheya mint. A small number of moulds for Indo-Greek, Kusāna, Gupta and Indo-Sassanian coins, which were likely used to forge coins, have also been discovered (Kumar 1996:102-106). It does not strike me as unlikely that the same group that produced official Yaudheya coinage from coin moulds during the period of class two coin production turned to forging coinage when Rohtak was occupied by these groups or traded with these groups.

161

to imply that these deities support the issuing ruler, but also to imply that the ruler is like the sovereign Zeus

or heroic Herakles. The appearance of the ruler's head on one side of the coin and the deity's depiction on

the other side certainly allows for such a possible understanding of the coins.55 Certainly, a single-headed

anthropomorphic deity allows for such a comparison much more easily than, say, a six-headed figure. It

should, however, be noted that the Indo-Greeks and other foreign rulers in India did not develop multi-

headed and multi-limbed deities anywhere near the extent to which such figures appeared in indigenous

sources. Multiple-headed figures are also rare on Kusana coinage.56 This lack of multiple heads is, as we

shall see, also true of the Kusana coinage depicting various forms of Skanda-Karttikeya. This evidence

helps to demonstrate that representing deities on coinage as anthropomorphic single-headed beings has

strong links to foreign ideas of how to depict deities on the coinage of north-west India.57 What is clear from

the above discussion is that the Yaudheyas may have borrowed the numismatic/iconographic concept of a

55These views are largely influenced by Frank Holt. Holt argues that Greek kings after Alexander the Great tried to display their wealth, courage and wisdom in their palaces, processions, gifts, monuments and campaigns (1999:7). In particular, he discuses the role coinage played in such displays: "The royal coinage was carefully designed to enhance this image, giving the public a chance to see the king's portrait, to read his name and titles, to identify the gods or heroes whose powers the king shared, and to enjoy spending part of the king's great personal wealth" (1999:7).

56The obvious exception is Siva or Oešo on Kusāna issues. The earliest depictions of this deity on Kusāna coinage tend to depict him with a single head and two arms (Cribb 1997:13). There is one exceptional copper issue of Wima Kadphises which shows the deity with two heads (Cribb 1997:14), but the norm appears to have been to depict him with a single head at this time. The iconographic trend during Kaniska's reign was to depict Oešo with a single head, but with four arms (Cribb 1997:14-15). Huviska issues coins depicting a multiple headed Oešo, though Joe Cribb considers these multi-headed types to represent exceptional issues. Huviska's normative Oešo types are borrowed from Kaniska's coinage (1997:15-16). The three-headed Oešo is normative by Vasudeva I's reign, but drops out of use with Kusāna rulers after him (Cribb 1997:17-19). Thus, the Kusanas do not begin with multi-headed representations of Oešo, and such a representation is only normative on one emperor's issues. It is also, I think, no coincidence that these three-headed Oešo coins become more common once the Kusānas have been in India for some time and have adopted some Indian culture, as the name Vasudeva for one of their kings implies. While these multi-headed coin types are important, their rarity helps to demonstrate the reluctance of foreign groups to represent deities on coins with multiple heads.

57Other indigenous coinage depicting humans like the PMC and Ujjain coins discussed above may be examples of single-headed human figures that developed without this foreign influence. My concern here, however, is with the specific time period in question and the general area of north-west India which sees direct contact with these foreign powers.

162

single-headed Brahmanyadeva from foreign sources.58

4.2.17 Chronology and Context

The identity of the political and cultural groups from whom the Yaudheyas borrow certain coin

styles has great implications for the chronology of the group's coins and Skanda-Kārttikeya's worship. In

this section I will review previous attempts to date these class three coins and argue for my own date which

begins at approximately the last few decades of the first century BCE and ends in the last few decades of the

first century CE. A great deal of the argument will rely on the political geography of north India for the

period of class three production and the information I have supplied above. I will begin by discussing the

dates for the production of Yaudheya class three coinage in relation to the period of the Kusāna empire.

As I have mentioned above, Allan ascribes six classes of coinage to the Yaudheyas. Scholars

following Allan have rejected his class one and four types as genuine Yaudheya issues. Most scholars also

regard Allan's class two and five to be an incorrect separation and regard them as different varieties of the

same class.59 These class two and five coins are the first known Yaudheya issues and are usually dated on

palaeographic grounds to the late second or first century BCE (Allan 1936:cxlix). Allan dates the class three

issues on palaeographic grounds to the second century CE (1936:cl). Most scholars have followed Allan's

dates, though some exceptions will be discussed below. It was assumed by most scholars that the gap

between the class two and class three coins was filled by the Kusana empire. This chronology falls apart,

however, if we examine the foreign influences that help shape the appearance of Yaudheya coinage.

There is something of a normative pattern of numismatic stylistic borrowing as one political group

58The Yaudheyas do not, however, copy the association of the deity with a leader, because no Yaudheya kings or leaders are depicted on their coinage. While the exact nature of the Yaudheya state is a point that requires more scholarly debate, it does seem clear that the Yaudheyas of the period of class three production were not a monarchical state and were likely a gana, a collective of various groups. The representation of a single-headed Brahmanyadeva, then, was not likely informed by the same concerns that may have motivated Indo-Greek, Indo-Scythian and Kusāna monarchs to present anthropomorphic figures on their issues.

59These points have been reviewed in detail in footnote number 24.

163

replaces another in north India. The incoming power usually borrows from the coin designs of the previous

rulers to maintain an already accepted currency for trade. What is clear from Yaudheya coinage is that their

class three issues do not borrow from Kusana coinage, while their class six coinage does. It will be my

argument below that the class three coins are pre-Kusana, and the class six coins are post-Kusana. As will

become clearer later in this section, Yaudheya class six coinage is strongly influenced by Kusāna coinage.

This Kusana influence is not my own discovery, but a point that is generally agreed upon by scholars (Allan

1936:cl; Dasgupta 1974:212). There are also Yaudheya class six over-strikes and re-strikes of original

Kusāna copper currency. Yaudheya coins of this type have been discussed by Handa (1982), Kumar (1991)

and by Kumar and Ahmad (1993). Over-struck coins take an original issue, heat it and hammer it with a

new die (Kumar and Ahmad 1993:50). In this case the original issue was a Kusana copper coin which was

over-struck by a Yaudheya class six die. A re-struck coin has part of its original metal shaved off and new

metal added, forming a new blank coin. The new blank is then re-struck with a new die. At times the added

metal falls off from these re-struck coins or becomes so worn that elements of the original show through

indicating a re-struck coin (Ahmad and Kumar 1993:50-51).60 The point of over-struck and re-struck

coinage may be both political and economic. Politically, taking the coins of previous rulers and striking

your own image over the old is a very public sign of a change of power. Economically, the old coinage

presents a ready-made supply of coinage whose weight and general appearance have already been accepted.

Re-striking existing coinage is simply an efficient means of creating one's own coinage in the immediate

aftermath of a change of power. The influence of Kusana numismatic style on these class six issues and the

60Handa regards coins of this type to be suspicious and concludes that the re-used Kusāna issues "are the creation of imitators and counterfeiters" (1982:47). M. Kumar reports a similar group of coins found at Bishan near Rohtak (1991:80). These coins were also Kusäna coppers which were re-struck by the Yaudheyas, but he does not suggest the work of forgers. Kumar also feels that the Yaudheyas minted Kusāna copper coinage from Sunet to meet economic demand, though his claim here is not well supported (1991:80). In a later article N. Ahmad and M. Kumar make a closer examinatin of the coin types in question. They find two types of re-used coinage by the Yaudheyas, one that is over-struck and another which is re-struck (1993:50). They do not regard these issues as the work of forgers, but as issues by the Yaudheyas to meet demand for coinage (1993:51).

164

class six over-strikes and re-strikes of Kusāna coinage suggest that these class six issues follow the fall of

the Kusana empire. They are usually dated to the third or fourth century CE (Allan 1936:cl), and I agree

with this date.

The class three issues are not influenced by Kusana numismatic issues. There are also no reports of

class three issues which are reused Kusāna issues. As I have demonstrated, the groups that influence the

Yaudheya class three issues are either Indo-Greek or Indo-Scythian. These groups pre-date the period of

Kusāna domination in north India. The pre-Kusäna influence on these coins and the lack of class three over-

strikes or re-strikes of Kusana coinage lead me to think that these class three coins come before the

Kusanas. Hence, my chronology for Yaudheya coinage begins with the class two and class five period

which is followed by the class three period, after which there is a break in Yaudheya coin production while

the Kusanas rule. The class six coinage then appears after this break which lasted for several hundred years.

Establishing a beginning date for Kusana rule in the north of India can, then, provide us with a terminus for

Yaudheya production of their class three issues.

Establishing dates for Kusāna rule and Yaudheya coin production have been points of controversy.

A wide range of dates has been proposed for both, and the seemingly simple question of who first minted

coins depicting Skanda-Kārttikeya, Yaudheya or Kusana, was left without a clear answer.61 I think the

61Previous Skanda-Kärttikeya scholars deal with the problem in one of two ways. G. S. Gurye 1973:21), S. M. Gupta (1988:6-8), V. R. Mani (1990:68-70) and S. S. Rana (1995:21-22) place Huviska's coinage before the Yaudheyas. Gupta does not state that he feels the Kusāna's Skanda-Kārttikeya coins are earlier, but he does present Huviska's coinage first and supplies 106-138 CE as dates for the king, although he does not explain how he came to these dates (1988:6). He follows Allan in dating the Yaudheya coins to the second century CE, and, by placing them after his account of Huviska's coinage, he implies that they are after the Kusāna king (1988:7). Rana also states no explicit chronology, but places his account of Huviska's coinage before his account of Yaudheya sources (1995:21-22). Mani dates Huviska to the "first half of the second century A.D" (1990:69) and the Yaudheya issues to the second and third centuries CE (1990:69-70). He does not, however, critically attempt to assess or justify these dates. None of the three authors, however, is able to justify his implied chronology, because the dating of Huviska is either unsubstantiated or ignores the wide variety of potential dates offered by various scholars for the Kusānas. Another group of scholars appears to assume that the first coins which depict the deity belong to the Yaudheyas. As with the scholars listed above, this chronological assumption is rarely stated or justified, but implicit in their presentation of numismatic evidence for the deity. P. K. Agrawala (1967:40-44), U. Thakur (1981:47-53) and K. Sinha (1979:73-84) all begin their reviews of the numismatic sources

165

argument I have presented above largely answers this question, but the issue of precise dates for these

matters will be dealt with in what follows. I will begin with the dates for the Kusāna empire which will

provide a terminus for the period of class three production.

Dates for the Kusāna king Huviska, the only Kusāna king to employ Skanda-Kārttikeya on his coin

designs, have varied dramatically from one scholar to another in the past. Dates for Huviska depend largely

on attempts to date the earlier Kusāna king Kaniska, who begins a new era for dating inscriptions which

Huviska follows. We did not know, however, when Kaniska's year one began. The following dates for

Kaniska's year one have been proposed: 58 BCE, 78 CE, 103, 128, 132, 144, 232 and 278 CE (Errington

1992:17).62 Scholars were not able to agree on dates for the various Kusana kings, leaving the question of

dating historical eras in relation to Kusana chronology open to speculation. This uncertainty has been

reduced with the discovery of the Rabatak inscription and the subsequent scholarship on it. This scholarship

now allows us to present a reasonable hypothesis for the dates of Huviska's reign as either 126-164 CE or

146-184 CE and Kaniska's year one to either 100 or 120 CE (Sims-Williams and Cribb 1994/95:106; Cribb

1999:188). While there is some disagreement over these dates,63 they do provide a probable time line for

concerning Skanda-Kärttikeya with the Yaudheya issues, then the Huviska issues and finally those of Kumāragupta. They all date the Yaudheya issues to the second century CE and simply do not date the Huviska issues. The implicit message such an order of presentation conveys is that the Yaudheyas are the first to present the god on coins, but this chronological assumption is not justified by these scholars.

62The arguments for a number of these dates can be found in Papers on the Date of Kaniska (Basham 1968), and some are summarized in Dynastic Arts of the Kushans (Rosenfield 1993:253-258). As we shall see, however, the controversy over dating these kings has recently come into clearer focus.

63There has been some criticism of Sims-Willaims's and Cribb's reading of this inscription by B. N. Mukherjee, but he does not appear to have any opposition to the chronology they propose based on the inscription (1997/98:5-10). Other scholars, however, have taken issue with Cribb's proposed dates. R. Gobl presents the strongest case against Cribb's chronology (1999:151-171), and his criticisms are supported by M. Alram (1999:45-46). Neither of these scholars thinks the Rabataka inscription solves the questions concerning Kaniska's year one. A valuable review of these problems concerning Kusāna chronology is provided by D. E. Klimburg-Salter (1999:3-18). She concludes that non-numismatists have had success using hypothetical dates for the Kaniska era of 78 to 144 CE, with most scholars picking between 78 or 100 to 110 CE (1999:14). These dates are not far from those proposed by Cribb. Given the criticisms of Cribb we must, I think, regard his dates for Kaniska's year one, either 100 or 120 CE, as hypothetical. Despite all of these reservations, however, I regard Cribb's arguments to be persuasive, and his Kusāna chronology to

166

these kings and a start towards dating the Yaudheya class three issues.

We know that all of north India was held by Kaniska and Huviska, and that parts of the empire

began to collapse under Vasudeva I. It is likely after Vasudeva I that the Yaudheya class six coins begin to

appear. While we can establish that Kaniska and Huviska ruled north India for most of the second century,

establishing the beginning of Kusana rule in north India is not as clear. As with most dates relating to this

period in South Asian history, there is debate over this chronology. Cribb has recently argued that Vima I

Takto was the first Kusana to reach Mathura. Taking Cribb's dates for Vima I of 78-90 or 78-120 CE

provides us with a potential terminus for class three coinage (1999:188). Other scholars, however, prefer to

see the conquest of India as occurring under the rule of Vima II Kadphises. His dates according to Cribb are

either 90-100 or 110-120 CE (1999:188). Vima II's dates give us a second potential terminus for Yaudheya

class three coinage.

Reaching Mathura is a significant marker, because coming south from the Punjab and Sind area to

Mathura would likely have driven Vima I's or Vima II's army through the Rohtak-Naurangabad area, the

seat of Yaudheya power during the production of their class two coinage (see map number one).64 It is also

likely that the substantial wealth the Kusanas displayed was due to their control of east-west trade flowing

from China to the Mediterranean and from the Ganges Basin, Gandhara and Bactria to the Mediterranean.

Such a control of trade goods flowing out of the Ganges Basin and Mathura required control over points

north of Mathura, which would have included areas of Yaudheya control and coin production, so that goods

could flow out to the silk route. This suggests that, not long after the taking of Mathura, the Kusana king or

general responsible for taking the town also secured the area between Mathura and routes north as well as

looking further east towards the Ganges Basin proper. In short, it is unlikely that the Yaudheyas could have

continued as an independent power within this region long after the Kusana conquest of Mathura. Whether

be the leading hypothesis currently available to us.

64I will discuss Naurangabad and Rohtak later in this section.

167

we take Vima I's or Vima II's dates for the conquest of India, we can position the terminus of class three

coinage to the last few decades of the first century CE, if not the close of the first century itself.

Now that we have established the likely terminus of Yaudheya class three coinage by establishing

the approximate beginning of the Kusāna era of domination over north India, we can try to establish a

beginning date for the coinage. I will attempt to do this by examining find site locations for class two and

class three coins. These find sites will indicate that the Yaudheyas were forced out of the Rohtak-

Naurangabad area, where they produced their class two coinage, into the foothills of the Himalayas, where

they produced their class three coinage. By examining a combination of what forces may have caused this

geographical relocation and the resulting borrowing from foreign sources on class three coinage, I will

suggest an approximate start date for this currency of the last few decades of the first century BCE.

We know from the find site locations of their class two and five coins that the Yaudheyas were

primarily located in the modern state of Haryana near the modern towns of Rohtak and Naurangabad.

Almost one thousand coin moulds of class two coins have been found at Khokrakot near Rohtak, and about

eight thousand five hundred of the same moulds have been found at Naurangabad (Handa 1991:67).65 Class

65 A mound near modern Rohtak called Khokhrakot has been excavated, which seems to contain the ancient remains of the town. Excavators discovered four layers of civilization at Khokhrakot, but only the third layer is of relevance to us. This third layer has been subdivided into three periods: pre-Kusana, Kusāna and post-Kusāna (Kumar 1996:96). The pre-Kusana period is indicated by the use of sun-dried and burnt bricks, a decrease in the use of red ware pottery and an increase in grey and black ware, the use of tiles "and [the] interaction of coins, seals, moulded terracotta human figures, etc" (Kumar 1996:97) Unfortunately, Kumar does not supply a complete list, or pictures of these coins, seals and terracottas, but he does note one seal with a second century BCE inscription and states that one of the terracottas is Kubera, while the others are female figures (1996:97). The artifacts from the pre-Kusāna layer indicate that the Yaudheyas dominated over the area. He suggests that the third layer of the site in general reflects growth and a shift from an agrarian based economy to one based in trade and commerce. He reports finds of guild seals and evidence of a textile industry along with sixteen Indo-Greek coins, which seem to suggest that the town evolved as Kumar suggests (1996:97-102). The Kusana phase of the third layer shows the following features: the introduction of 37x23x7 cm burnt bricks, Kusāna copper coins, copper rings and one gold ring, terracottas "etc" (Kumar 1996:97). The excavation demonstrates that the Kusanas now control this town, though there is evidence of the site being used by forgers of Kusana gold coins, which may suggest a lack of complete Kusāna domination. V. S. Agrawala reports finding a mould for an Ardoxšo gold coin of the type produced by Kaniska and Huviska (1953:68). Kusāna gold coins were die struck, and the existence of this mould is suggestive of counterfeit currency. Kumar feels that Rohtak was a Kusana mint town, though I see no evidence to support such a

168

two coins have also been found in limited numbers to the north-east of these towns in Jagadhri, Behat and

further east in Meerut in the modern state of Uttar Pradesh (see map number one) (Handa 1991:67).66

Several scholars have argued that Rohtak was the capital of the Yaudheyas, but this argument is not well

justified. The argument is largely based on the class two coin legend, which reads in Prakrit: "Yaudheyānām

Bahudhañake" or in Sanskrit: "Yaudheyānām Bahudhānyaka" (Dasgupta 1974:215). The legend may be

translated: "(coin) of the Yaudheyas (living) at Bahudhanyaka" (Dasgupta 1974:215). Bahudhānyaka seems

to refer to a place and has been taken to mean: "a fertile land rich in corn" (Dasgupta 1974:215). On the

basis of the finds of coin moulds at Rohtak and the fertile region there, people have assumed that

Bahudhānyaka refers to Rohtak and represents a capital city for the group (Dasgupta 1974:215). The

reasoning behind this assumption is not strong. Bahudhänyaka does not have to refer to Rohtak. A larger

amount of coin moulds and administrative seals and sealings have been recovered from Naurangabad,

which suggests that that town was a larger administrative centre than Rohtak during the production of class

two coinage.67

claim (1996:103). What the dig at Khokhrakot suggests is that Yaudheyas were replaced as the dominant power in the Rohtak-Naurangabad region by the Kusānas.

66Behat is also the site of finds of Kuninda coinage from the same period. Kuninda coins are also found in Karnal and Srinagar, all part of the same basic region in Uttar Pradesh (Dasgupta 1974:88). This region in Uttar Pradesh may have been alternately ruled by either of these groups over the second to first centuries BCE.

67 A seal found at Naurangabad has also been used to argue for Rohtak as the Yaudheya capital city. The seal has been read in a number of ways. Swami Saraswati first read it as "rapata yaudheya janapa da prakritānāka nagara" (1970:155). He feels the seal begins with a yūpa symbol, and he translates the legend as "the seal of Prakritanaka Nagara of the Yaudheya republic of the masters of wealth" (unpublished:2). He also notes that it was found at Naurangabad (1970:155). Thaplyal suggests that the yūpa Saraswati sees is actually the letter vi, which would change the first line to "Virapati" (1972:21). Shastri, however, corrects the first word to "rahata", which he feels stands for Rauhīta or Rohītaka. From this he concludes that the town of Prakrtānaka was in Rohtak (1976:119). He does not seem to acknowledge, however, that the seal was found in Naurangabad. He translates the legend as "the seal of Prakritānākanagara of the prosperous Yaudheya Janapada", or "the seal of the Ānakanagara, the capital of the prosperous Yaudheya Janapada" (1976:119). I am not convinced, however, that rahata need be read from the seal itself or that it refers to Rohtak. Both of these seals and sealings suggest to me the existence of administrative units outside of Rohtak proper and weaken the claim the Rohtak was a capital for the Yaudheyas. Handa also seems to recognize a difficultly in viewing Rohtak as the capital of the Yaudheyas

169

The class two and five coins themselves do not tell us a great deal concerning Skanda-Kārttikeya.

On the obverse of the class two coins is a bull standing to the left or right of a railing with a curved object.

Some scholars feel the curved object represents a yupa. The legend reads as mentioned above. The reverse

shows an elephant to the right, a nandipada symbol above and an uncertain pennon-like object behind the

elephant (Dasgupta 1974:202). Another variety shows a horse instead of the bull on the reverse (Dasgupta

1974:202). There is only one example of the class five coin. The legend reads: "(Yau)dheyān(ām)", above

which is a bull (Allan 1936:cl). Neither coin type reflects Yaudheya class three issues, indeed there is a

remarkable iconographic shift between the coin types.

With this dramatic shift in coin style from class two and five to class three comes a shift in

geographical location. The find sites that have been recorded for Yaudheya class three coins are Tehri,

Dehradun, Jagadhri and Behat all in Uttar Pradesh and the new state of Uttaranchal in the foothills of the

Himalayas (see map number two) (Handa 1991:68).68 During the production of class two currency the

Yaudheyas seemed well established in Rohtak and Naurangabad, and their movement out of this area was

not likely voluntary. As we have seen, these class three coins are poorly made and of inconsistent style and

of a radically changed design from the class two and five types. Inconsistent and low quality coin

production are signs of a state undergoing economic or political turmoil. Hence, the class three coins

suggest difficult times for the Yaudheyas. Clearly another group forced the Yaudheyas out of their base in

for the entire class two period and suggests: "The Yaudheyas of Rohtak seem to have been eclipsed for some time under the pressure of the Indo-Greeks and probably shifted their capital to Naurangabad from where they continued issuing coins of Class 2" (1988:132). Handa's evidence for this Indo-Greek pressure is not clear to me, but Naurangabad is only about 35 kms to the west of Rohtak and does not strike me as a very great distance in which to seek safety if there was genuine "pressure" placed on the group by some outside force. I think scholars have been strongly influenced by textual references to a place called Rohītaka, which is described as special to Skanda-Kārttikeya (Mahābhārata 2.29.5-6), and this has lead them to insist that this place must also have some particular importance to the Yaudheyas and hence to have been their capital. The evidence, however, seems stronger for Naurangabad as a more important Yaudheya centre than Rohtak, though Rohtak was certainly inhabited by the group.

68It should be noted, however, that Allan notes a lack of accuracy in recording the find sites of Yaudheya currency in general (1936:cli). The above locations only reflect find sites that have been accurately recorded.

170

Haryana at around the beginning of class three production. We also know that the Yaudheyas copied from

Indo-Greek and Indo-Scythian designs, and it is with these pre-Kusana groups that we will likely find the

group or groups that forced the Yaudheyas north.69

The first group to force the Yaudheyas out of Haryana may have been the Indo-Greeks. There is

speculation that the Indo-Greeks invaded as far east as Pataliputra under Demetrius (Tarn 1938:132-133),

and that Menander ruled as far south as Mathura and over the territory of the Yaudheyas and Kunindas

(Tarn 1938:239-240). These claims have been criticized by some, and the evidence for these claims is not

strong (Narain 1968:92-93). It is certain that the Yaudheyas knew of Menander's coinage and of other Indo-

Greek issues because of finds of Indo-Greek forging equipment at Rohtak (Kumar 1996:97), but if the

Greeks ever took Rohtak or fought the Yaudheyas is open to speculation.

What is more certain is that after the Indo-Greeks the Indo-Scythians and Indo-Parthians were in

the Punjab region, and some of their kings ruled from Mathura. The Scythians Maues, Azes and his

descendants may have been in the Punjab area, though the centre of their kingdoms lay to the north-west of

the Punjab. These Scythians likely ruled towards the beginning of the first millennium (first century BCE -

first century CE). The Parthians Hagana and Hagamasa appear to rule in Mathura, and after them the

Scythians Rajuvula and Sodāsa are ruling from Mathurā (Narain 1989:116).70 Rajuvula was likely a

Ksatrapa of Maues who invaded various parts of the eastern Punjab before settling in Mathura, where he

becomes a Maha Ksatrapa. Exact dates for him are unclear, but Rosenfield suggests he ruled from circa 75

69It may not be a coincidence that the Yaudheyas begin placing Skanda-Karttikeya on their coins once they are near the Himalayas. As we have seen, most textual accounts of the deity present him coming out of these mountains. It may be the case that the characterization of Skanda-Karttikeya as a six-headed deity began in these mountain areas. Such a conclusion is also suggested by the few finds of six-headed statues of Skanda-Kärttikeya from the north all of which come from Kashmir and Himachal Pradesh, though these statues date to a period well after these class three coins.

70There is, however, doubt concerning the order of these rulers in Mathura. Lahiri has Rajuvula and Sodāsa before Hagāna and Hagāmasa (1974:162-166). Precise dates for Hagāna and Hagāmasa are not clear to me and just how they relate to Gondophares, largely regarded as the founder of the Indo-Parthian kingdom, is also not clear to me.

171

BCE to the last quarter of the first century BCE (Rosenfield 1993:135). Mitchiner, however, argues that he

ruled in the last few decades of the first century BCE and the first decade of the first century CE (1976

VII:581-584). He issued three coin types. The first is a copy of Indo-Greek issues found in the eastern

Punjab and western Uttar Pradesh (Lahiri 1974:163). The second group consists of imitations of Maues'

Herakles issues and are found from the Punjab to Taxila (Lahiri 1974:163). The third type is local to

Mathura and borrows from indigenous coin patterns (Lahiri 1974:163).71 His coin finds suggest he ruled

over a large territory, and that territory would have included Rohtak and Naurangabad. Whether or not it

was Rajuvula who pushed the Yaudheyas out of Haryana or an earlier invader is not clear, but the area

between Mathura and the western Punjab saw a great deal of disruption during the last period of the first

century BCE and the early period of the first century CE. It is likely during this time that the Yaudheyas

were forced out of the Rohtak-Naurangabad area to the north-east.

We cannot discredit the idea that the scope of disruption may go all the way back to Menander and

the middle of the second century BCE. The beginning of class three coinage may, then, go back to the

middle of the second century BCE, but a more likely hypothesis sees the Yaudheyas leave Rohtak-

Naurangabad under pressure from Scythian and Parthian groups just before the Common Era, which is

when we have more solid evidence for these groups in the area. If we accept Menander and the Indo-Greeks

as the force that pushed the Yaudheyas north, we also leave very little time for the production of class two

coinage and cannot account for the paleographical differences between the second century BCE legends of

class two currency and the first to second century CE legends of class three currency. While paleography is

an inexact science and there is some room for movement in the second century CE dating of the class three

coins, I doubt that there is enough room in the palaeological dating to push the date back by 200 or 300

years. Hence, I favour a beginning date for class three coinage of the decades just before the Common Era.

Most scholars have, however, proposed different dates for these coins. I would briefly like to

71The female deity or Laksmi found on these local Mathura issues closely resembles the female found on Kuninda coinage and the single-headed female on Yaudheya class three coinage.

172

account for their arguments here. As I have mentioned, the Yaudheya class three series is dated on

paleographic grounds to the second century CE by Allan (1936:cl) and to the "later second century" by D.

Handa (1991:70). There is, however, some disagreement over the date of these class three coins. O.

Bopearachchi and W. Pieper offer a significantly different date, placing the Yaudheya class three coinage in

the second or first century BCE (1998:51). M. Mitchiner also argues for a much earlier date for the single

Yaudheya silver and the rest of the class three coinage. He suggests c. 150/100 BCE as dates for class three

coinage (1975 IV:331).

In favour of the early date proposed by some for class three coinage is the existence of a single

class three silver issue. Based on this silver issue Pieper argues:

Of this silver type, only the British Museum specimen is known, and without further argument Allan places it together with all the copper coins of the same class 3 to the 2nd century A.D. More probable to me, however, seems to be a chronological placement to the 2nd/1st century B.C., a time when the other tribal silver coins of the Punjab states have been struck. (Bopearachchi and Pieper 1998:51)

Pieper does not explain, however, where he sees the class two coins fitting into this chronology. They date

to the second to first century BCE and have no resemblance to the silver issue in question. He also does not

address problems relating to differing paleography and find sites for the class two and class three coinage.

As we have seen, the silver coin is clearly related to the copper class three coins and cannot be grouped with

the class two series. Pieper also provides no explanation as to how or why the very different class two and

class three coins could have been produced during the same period.

Mitchiner tries to overcome this problem by suggesting two separate Yaudheya groups operating at

the same time. He suggests that a northern group issued the silver coin and other class three issues, and a

southern group issued the class two type at the same time around 150 to 100 BCE (1975 IV:331). What

works against Mitchiner's hypothesis is the lack of class three coin finds in his proposed southern region. If

there were two sections of this group, one would expect to find their coinage in both sections of their

territory. Class two coins are found in the northern area, but class three coins are not found in the southern

area. It seems unlikely that the class two coinage would be accepted as currency by the proposed northern

173

group, while the proposed southern group rejected the class three currency. Mitchiner is also unable to

explain the iconographic, paleographic and quality differences between the two coin types. It seems much

more reasonable to me to view the two types of coinage as existing at different periods in time.

The single silver issue does raise some chronological problems, but we should also be cautious

about the single silver issue. This single coin is an exceptional issue and should not be made the basis for

determining the minting process and chronology of the Yaudheyas. Silver coinage stops being produced in

South Asia by about 50 CE.72 This single issue may come at about the close of the first century BCE and the

beginning of the first century CE, and this is the general date I give to class three coins. It does not have to

be placed in the second to first century BCE as Pieper and Mitchiner claim, and it may not deserve the

amount of attention from scholars that it receives in relation to other evidence relating to these issues.

The late date for the Yaudheya class three coinage of the second century CE is based on

palaeographical evidence, which is not always completely reliable. These class three coins are worn, and no

single coin has the full inscription on it. The legend has to be reconstructed by comparing a large number of

coins, and the poor quality of these coins presents a barrier in determining the precise date of the inscription

on palaeological grounds. The date of the second century also falls within the period of Kusāna domination.

I have already argued that it is unlikely that the Yaudheyas could have issued coins during this period.

Proposing a date of the second half of the first century BCE to before the last few decades of the first

century CE does not require too great of a revision of Allan's dating on paleographic grounds.73

We are left, I think, with only one workable hypothesis for the dating of these coins. It is clear that

72Personal communication with Joe Cribb.

73 Allan's and Handa's rather specific date for these coins is likely not based on palaeography, which is usually thought of as unable to give precise dates. What likely informs their dating is a common theory that the Yaudheya class three coinage came about after a Yaudheya defeat at the hands of Rudradaman. This theory appears to begin with Allan (1936:cliii) and is echoed by Banerjea (1951:162) and others. Our understanding of this defeat comes from the Junagadh inscription where Rudradāman mentions defeating the Yaudheyas, and the inscription has been dated to c. 150 CE (Dasgupta 1974:197); hence their second century date for these coins. I will demonstrate below that this hypothesis for class three coinage as linked to Rudradāman's victory is not well founded.

174

Indo-Parthian and Indo-Scythian groups likely pushed the Yaudheyas out of Rohtak and Naurangabad and

further north into Uttaranchal Pradesh where, I think, they began to produce their class three issues around

the last few decades of the first century BCE.74 They produce these coins until the coming of the Kusāna

dominance in all of north India around the end of the first century CE.

The importance of all this is that the Yaudheyas are the first group to issue coinage depicting

Skanda-Kärttikeya or one of his forms, and they do so between circa 20 BCE to 80 or 100 CE. These coins

are issued after the group has been displaced and were likely issued for trade with indigenous groups like

the Kunindas as well as Scythian, Parthian and what remained of the Greek powers in the area. I should,

however, emphasize that much of this chronology is speculative and built on the chronological hypotheses

of others that have yet to be proven beyond doubt. Some new evidence about the Indo-Greeks, Indo-

Scythians, Indo-Parthians, Kusanas and so on could easily upset this chronology I have proposed. Evidence

for the existence of the Yaudheyas during the period of their class three coin production is limited to the

coins themselves. Having stated this much, however, I feel the hypothesis provided above presents the most

reasonable argument given the state of our current knowledge.

4.2.18 Yaudheya Class Three Coins: The Origins and Meaning of the Six Heads

No Yaudheya statuary depicting Skanda-Kārttikeya has ever been recovered, and, beyond the

examples from foreign coinage, we have no other examples of what may have inspired this Yaudheya coin

style. Here, I will summarize an argument that the six-headed style of representation of Brahmanyadeva is a

Yaudheya innovation and may represent the indigenous understanding of the deity's appearance.

These Yaudheya class three coins are the only South Asian numismatic issues to represent Skanda-

Kärttikeya or one of his forms with six heads. R. C. Agrawala argues that the numismatic presentation of six

74Based on the find sites of a small number of class two coinage in the region of class three production, it is possible that the Yaudheyas could have moved north and continued to produce the class two coins for a brief period before switching to class three coinage. It may also be the case that elements of the Yaudheya state existed in the northern area during the period of class two production and remained independent when Rohtak and Aurangabad fell.

175

heads as two rows of three heads on a deity is a Yaudheya innovation. He notes that the earliest Mathura

sculptures depicting Skanda-Kārttikeya show him with a single head (1966:200). He argues that the first

available six-headed statues of this deity in the Punjab and Kashmir region are post-Gupta. He also states

that there is no evidence for a depiction of Skanda-Karttikeya with six heads in the pre-Kusāna period

(1967:41). From this evidence he thinks the Yaudheyas introduced this particular six-headed trait, though

they may have been inspired by Mathura sculptures, which did place extra heads on top of other heads

(1966:201-202). My sections on statuary in the next chapter will confirm Agrawala's observations here.

That this convention is likely indigenous to the Yaudheyas is important because, as I have

demonstrated above, many of the single-headed Brahmanyadeva coin features of this period are borrowed

from Indo-Greek and Indo-Scythian sources. This aspect of a six-headed deity appears to place us within an

indigenous religious and numismatic context."5 As this six-headed depiction appears to be less influenced by

foreign sources, it may well represent an earlier understanding of the god's physical appearance than the

75The appearance in Indian art of multiple limbs and heads has been a topic of some scholarly discussion in recent publications. One of the most important of these publications is D. M. Srinivasan's Many Heads, Arms and Eyes: Origin, Meaning and Form of Multiplicity in Indian Art. One of Srinivasan's main arguments is that the depiction of multiple heads and arms in statuary is based on a Vedic theory of numbers. Specifically, she argues that the numbers four, five and eight have special 'symbolic' meaning in the Vedas and related ritual texts, and this 'symbolic' significance is the basis of what she calls the multiplicity convention in emerging Hindu art (1997:162-175). She regards six as outside of this Vedic symbolic number system and states: "Six' is used as a mathematical, not symbolic number" (1997:293). Skanda-Kārttikeya's or Brahmanyadeva's six-headedness does not fit into Srinivasan's Vedic model, and she does not seem to know what to do with him, except to suggest that their is no special meaning behind his number of heads. She regards six as an "actual number" associated with Skanda-Kārttikeya's and Sasthi's worship, and as such it does not function on a "symbolic level", but "functions as an arithmetical number" (1997:335). She also attempts to explain Skanda-Kärttikeya's six heads in relation to legends about him. As such she regards his six heads as a result of being suckled by the six Krttikās (1997:333). In a similar fashion she regards Sasthi's multi-headedness to be a reference to her worship, which is supposed to occur on the sixth day of the lunar month (1997:333). What is not clear to me in Srinivasan's discussion is the difference between a 'symbolic' number and an 'arithmetic' number, except that all symbolic numbers fit into her Vedic number theory and arithmetic numbers do not. Simply because Skanda-Kārttikeya's number of heads relates to his depiction in textual accounts and relates to part of his and Sasthi's ritual cult does not, to me, imply that the number six is void of meaning beyond counting the number of his and her heads. While I must admit that the exact meaning of Skanda-Kärttikeya's six heads is obscure, I still have strong reservations about Srinivasan's division between symbolic and arithmetic numbers. Hans Bakker also expresses a number of reservations concerning this book (1999:339-342).

176

single-headed variety. In my section on Kusāna coinage I will argue that the image of Skanda-Kārttikeya as

a single-headed deity becomes the standard representation of him under Kusāna rule and is a part of his

transformation into Mahasena, a military god. The numismatic tradition moves towards eliminating the six-

headed variety on coinage over time, which suggests a chronological progression from six heads to one.

This progression appears to emerge with these class three coins and is the result of foreign influence. What

this progression may signify will be discussed below.

The exact origin and meaning of Skanda-Kärttikeya's six heads is not clear, but I will attempt here

to shed some light on this problem. I will do this primarily through the reverse depiction of Sasthi on the

Yaudheya class three coins.76 I have left out discussions

76There has been some debate over just who this female figure on the reverse of Yaudheya coinage is. A number of early scholars did not seem to know how to identify the reverse figure. There is no inscription on the reverse which identifies her, leaving an absolute identification difficult to achieve. Cunningham simply calls her a "six-headed figure" (1963:78). Smith only refers to her as a "goddess" and makes no attempt to elaborate on this point (1972:181-182). The first genuine attempt to identify the goddess comes with Allan, who thought the reverse figure was Laksmī (1936:cxlix-cl). Allan seems to have been influenced by the similarity between the Yaudheya reverse female and the Kuninda and Mathurā reverse females, who are often identified as Laksmī (1936:cxlix). Some caution is, however, required, because the Kunindas never portray a six-headed goddess and the Yaudheya six-headed goddess is clearly different from the Kuninda single-headed goddess. The Kunindas also never identify the reverse image on their coins, and calling her Laksmī is an academic guess and not a fact. V. S. Agrawala appears to be the first scholar to suggest that the reverse six-headed deity might be Sasthi (1943:29-32). Agrawala claims that Sasthi is also known as Devasena and Laksmi in the Mahabharata and he regards Sasthi's account as dependent on Skanda-Karttikeya's (1943:29-31). In other words, he uncritically accepts the Mahabhārata's attempt at amalgamating various deities through the Skanda-Kärttikeya narrative and understands Sasthī's six-headedness to rely on Skanda-Kärttikeya's six heads. I am not convinced that we need to accept the Epic's attempt at amalgamation in relation to these coins. It may well be that Sasthi is being blended with other fertility figures like Laksmi on these coins, but the case of goddesses like Devasenā requires a more critical eye. Devasena appears to be an Epic invention used primarily to domesticate Skanda-Kārttikeya and to, quite literally, wed him to his role as the general of the army of the gods. Also, Devasena is only mentioned in the Aranyakaparvan, and she does not seem to have endured as a goddess associated with Skanda-Kärttikeya. The idea that Laksmī or another goddess related to fertility may be represented on these coins is not out of the question, but I will discuss this in more detail below. In a separate publication V. S. Agrawala discusses passages from the Kāśyapasamhitā which describe Sasthī as six-faced as further evidence for his identification of her on these coins (1970:92-93). Agrawala's identification does not seem to have immediately caught on, however, with Banerjea still identifying her as Laksmī (1956:141; 1960:45). D. Pandey suggests that the goddess on these coins is Krttikā (1967:7). He notes that in the Mahābhārata Skanda-Kärttikeya is called the son of the Krttikas, there are six of them and Skanda-Kärttikeya's six heads are a result of their suckling of him in the Epic version of the story (1967:5-7). Pandey's theory was strongly opposed by R. C. Agrawala, who defended V.

177

S. Agrawala's identification of the figure as Sasthi and rejected Pandey on the basis of a lack of textual or sculptural support for his claims (1968:181-182). P. K. Agrawala supports V. S. Agrawala's identification, and, noting the epic passage which assimilates Sasthi with Laksmi and Devasena, claims: "and we would feel rather no difficulty in identifying the single-headed goddess on other Yaudheya coins as Devasena- Laksmī, i.e., Devasenā as Laksmī" (1967:41-42). Navaratnam only refers to the reverse of the silver issue and calls that figure Laksmi (1973:100). Dasgupta reviews the various opinions of the above scholars and concludes "it is more than certain that the one-headed or six-headed female figure is that of Shashthi or Devasenã" (1974:222). He also refers to the Mahabhärata passage cited by V. S. Agrawala which amalgamates a number of goddesses (1974:221). Thakur regards the reverse figure to represent Laksmī (1974:305). Chattopadhyay regards the goddess as Laksmī, but notes that there have been attempts to identify her with Sasthī and Devasena (1977:59, 201). Shastri notes that most scholars identify the single- headed goddess as Laksmī, and V. S. Agrawala has identified the six-headed goddess as Sasthi, whom he also calls Devasena (1977:92). He takes the single-headed goddess as Sasthī /Devasena as well and notes " ... because of her being identified, or rather confounded, with Laksmi there was little difference left between the features of the two goddesses" (1977:93). Ahmad reviews the various arguments and concludes "it is definite that the one-headed or six-headed female figure shown on these coins is that of Shashthī or Devasenā" (1977:159). O. P. Singh argues that the six-headed goddess is Sasthi /Devasena and that the single-headed goddess is Laksmī (1977:134; 1978:24). Sinha reviews the various opinions on the topic and concludes that the reverse figure is Sasthī /Devasena, whom he also links to Laksmī (1979:74-77). Finally, Gupta mentions that some scholars regard the reverse figure to be Laksmī, but he seems to prefer an identification as Devasenā (1988:7). The general trend has been to accept the Epic's amalgamation of these goddesses as related to these Yaudheya coins and to view either all the goddesses on the reverse as an amalgamated Sasthi- Devasenā or to view the single-headed issues as Laksmī, a point of view with which I do not agree for a number of reasons. The first is that we do not view the obverse images with one and six heads as representing different deities and not to use the same approach for the reverse images strikes me as inconsistent and does not seem to recognize the nature of these class three coins. The Yaudheyas did not seem to find it problematic to represent Brahmanyadeva as both six- and single-headed and the iconography of this deity seems to have been malleable during this time. The same is likely true for the reverse image. The second is that V. S. Agrawala is correct in recognizing that Sasthi is described as six-headed in the Kāyapasamhitā. She is a Graha in these accounts. She is repeatedly associated with Skanda-Kärttikeya in textual accounts of him, and she is the only six-headed goddess associated with him. The third is that the 'Laksmi' on the Kuninda coins used to demonstrate that the Yaudheya single-headed female is also Laksmī may not be that goddess. There is no legend on the Kuninda coin to identify the female on it, and she might be better labelled as an unnamed fertility or prosperity goddess. It is also the case that not all of the Yaudheya reverses are models of the Kuninda reverse types. The single Yaudheya silver coin has the most obviously Kuninda-influenced reverse. A number of class three Yaudheya coins seem to want to emphasize the reverse image's reproductive and nursing ability by clearly displaying her breasts and womb. This womb is depicted differently from the Kuninda coinage. Some of the Yaudheya six-headed and single- headed depictions of the reverse image show a female figure whose body is hollow except for breasts and a womb. Some of the art of this period was preoccupied with depicting female reproductive capacity. Both the multi-headed and single-headed goddesses on these Yaudheya coins seem to reflect this preoccupation and hence a link to Sasthi. In conjunction with this issue is the evidence from Yaudheya glyptics. Yaudheya seals and sealings show a preoccupation with Mätrs. Sasthī is absent from Yaudheya seals and sealings, but references to Mätrs are common and may be a reference to their broad based popularity (Handa 1985:133). I think this information suggests that if there is any amalgamation occurring on the reverses of these class three coins it is Sasthi as a Graha with Mätr cults, both of which are concerned with fertility, but specifically

178

of the reverse types on Yaudheya coins, but the cults of Sasthi and Skanda-Kārttikeya are closely related,

and her appearance on these coins with Brahmanyadeva provides some valuable material evidence for their

combined cult. The significance of Sasthi on these Yaudheya coins is twofold for this study. First, I will

argue that her six-headedness need not be viewed as dependent upon Skanda-Kārttikeya as V. S. Agrawala

suggests (1943:31), but that the reverse relationship is a more defensible argument. Second, I will argue that

Sasthi's appearance on these coins helps to place these issues in a context of concern for procreation,

Grahas and Mātrs.

While we have no statues of a six-headed Skanda-Kārttikeya from the pre-Kusāna and Kusāna

eras, we do have panels depicting a six-headed Sasthi from the Kusana period. An example of one of these

panels is illustrated in figure 41. In this panel, and others like it, she is depicted as having a central head

with five other heads arranged over it in a semi-circle, reminiscent of her depiction on some of the

Yaudheya coins. She is flanked by two males, who are single-headed and each of whom holds a spear.

These two males are often identified as Skanda-Kārttikeya and Visākha. (P. K. Agrawala 1971:327; R. C.

Agrawala 1971:83-84). These panels may come slightly after the Yaudheya class three coinage, but they do

suggest that Sasthi was understood to have six heads, and that this feature had a stronger association with

Sasthī than with Skanda-Kārttikeya in artistic representations of them.77

Sasthī also has a closer relationship to the number six and its value in early childhood healing cults

than does Skanda-Kärttikeya. Her name literally means 'sixth,' and she is worshipped on the sixth day of an

infant's life (Tewari 1996:171). In the Kāsyapasamhitā Sasthī is also described as having six faces (Tewari

1996:171). Infant mortality in the ancient and not so ancient world was high, especially during a child's first

few days of life. The cult of Sasthi seems to be tied to this early and dangerous period of a child's life, and

human reproduction. There may be some reverse Yaudheya types which represent Laksmi and devices associated with her, but these appear to be influenced by Kuninda coinage and may reflect a need to produce coinage that other groups outside of Yaudheya society would recognize. The majority of class three coins depicting a female on reverse likely present us with a depiction of Sasthi or a Sasthī-like Mātr.

77These panels will be discussed in more detail in the next chapter.

179

her cult seems to revolve around the number six. The coins also echo her connection to fertility and child

birth by the iconographic highlighting of her breasts and womb on some Yaudheya issues.

While Sasthī is described as six-headed in the Kāsyapasamhita, she is not so described in the

Mahābhārata. Skanda-Kārttikeya is described as six-headed in that text, and the narrative concerning the

two is focused on Skanda-Kärttikeya. Sasthi is only mentioned in the Aranyakaparvan version of Skanda-

Kärttikeya's narrative and is a minor figure in that text. It is largely this account of both deities from the

Epic tradition that leads V. S. Agrawala, I think, to suggest that Sasthi's iconography is dependent on

Skanda-Karttikeya's. We must, however, be cautious of the Epic's account of these deities in relation to

their actual lived cult.

One of the tendencies of the Epic Skanda-Kärttikeya narratives is to control a large number of

female deities through a male or several male deities. As we have observed, a number of ghoulish goddesses

are made Skanda-Karttikeya's mothers and placed under his control in the Epic story. While the Epic text

presents these female deities as secondary to Skanda-Kärttikeya and controlled through his worship, the

archaeological evidence suggests a different scenario. Seals and sealings found at Yaudheya sites,

particularly Sunet, suggest that the cult of Mätrs had broad-based popularity during Yaudheya occupation of

the town (Handa 1985:133). In fact, there are many more seals and sealings containing the name Mātr from

Sunet than there are with the name of Skanda-Kärttikeya.78 A similar pattern is found concerning the cult of

Naigamesa, a form of Skanda-Kärttikeya in the Epic. Archaeological evidence tells us the cult of

Naigamesa is also a cult of Naigamesī (Jayaswal 1991:41). Terracottas of this female version of the goat-

headed deity are usually found in higher numbers than the male version,79 but the female version of the deity

is excluded from the textual tradition. The Naigamesa/Naigamesi cult is another example of the disparity

between Epic accounts of female deities and their association with the Skanda-Karttikeya cult on the one

78These seals and sealings from Sunet will be discussed in more detail in the next chapter.

79 At Khairadih, for instance, ten Naigamesīs and three Naigamesas have been recovered (Jayaswal 1991:37-38).

180

hand and what may have been the actual significance of these deities in lived religious practice on the

other.80 Certainly I am not the first person to suggest that the Brahminical tradition or Hinduism in general

has a misogynistic aspect to it, but the specific mechanics of this case are important to the present study. To

suggest that Sasthi's iconographic features are dependent upon Skanda-Kārttikeya's because the Epic seems

to present Sasthi and Matrs as under his control is to ignore the Epic redactors' at times obvious agenda in

narrating these stories as they do.

I acknowledge that we do not have enough evidence to state conclusively which deity influenced

which. While the Epic evidence can be questioned, it cannot be ignored, and the exact origins of both

deities' six heads remains unclear. I do think, however, that the archaeological evidence presents a

defensible argument that Sasthi's six-headed nature influenced Skanda-Karttikeya's in the early period of

their cults. Certainly, Sasthi's appearance need not be viewed as secondary on these class three issues, nor

as dependent on Skanda-Kärttikeya.

The second important thing Sasthi's image on the reverse of these coins does is to link the obverse

figure of Brahmanyadeva to a Graha or Matr cult. Sasthi, particularly as six-headed, is a Graha associated

with Skanda-Kärttikeya also as a Graha-like deity. In the Kāyapasamhitā she is described as a sister of

Skanda-Kärttikeya and as deserving the same worship as he (Tewari 1996:169-170). While there are some

positive sentiments expressed about her in this text, there is also an underside of negativity. Her worship

day is after the sixth day of delivery and every sixth day of every fortnight; she is also described as killing

on the sixth day of delivery (Tewari 1996:171). While the text does not explicitly state it, it seems clear that

her worship is designed to prevent her from killing new born children. She is a dangerous Graha and Mātr

figure, who is also connected with poisoning breast-milk as a form of Graha attack along with Skanda-

Kārttikeya (Tewari 1996:9), and with diseases afflicting wet-nurses (Tewari 1996:252). Sasthi as a

80Even today, while Skanda-Kärttikeya has largely disappeared as a deity concerned with child birth and disease, Sasthi is still an important goddess associated with the protection of children in Bengal. See T. Stewart (1995) for more details on this modern cult.

181

dangerous six-headed being helps us to understand part of the context of these Yaudheya coins and the six-

headed Brahmanyadeva also found on them. The Yaudheya use of these six-headed figures suggests a

Graha or Mätr context for them.

The appearance of Brahmanyadeva on these coins, especially the six-headed type, is, I think, a

frightening image that best fits the early inauspicious versions of the deity I have outlined in my textual

section. The sense of fear and inauspiciousness, but also power invoked by versions of Skanda-Kärttikeya

like Skandagraha seems to be well displayed by this frightening Yaudheya six-headed version of the deity

carrying a giant spear. The textual accounts suggest that it is only those who do not worship Skanda-

Kārttikeya who need to view him as inauspicious and to fear his temper. We may presume, I think, that in

the case of these class three coins the Yaudheyas, or at least some aspects of their society, worshipped

Brahmanyadeva or Skanda-Kärttikeya and felt that his violent power was on their side. Part of the message

of these coins implies that the Yaudheyas or their leadership have harnessed the power of this deity through

worship of him. I must stress, however, we have no physical evidence of this worship. We can only create a

hypothesis based on the limited evidence of the coins themselves. I will return to this topic of the worship of

Skanda-Kärttikeya or Brahmanyadeva by the Yaudheyas later in this section.

Another context that should also be considered here is political. I have already discussed the

relocation of this group to the foothills of the Himalayas and the economic and military stress this likely

forced relocation may have caused. The period was a turbulent one involving a steady stream of new

invaders. The dramatic shift from the Yaudheya class two coins to the class three coins can be viewed as a

response to these hard times. The desire to take on a protective force in Skanda-Kärttikeya and to direct his

malicious energy on to others may be reflected by these class three coins.

There is also a larger context for the Yaudheyas and that is as a warrior group. Textual references

to the Yaudheyas imply their association with warrior ideals. As noted earlier, Pānini's Astādhyāyīrefers to

the Yaudheyas as an ayudhajivī samgha, a group living as warriors or by arms. There may be a hint of

banditry or living by raids in this term. V. S. Agrawala seems to suggest as much when he discusses the

182

Skandayāga in relation to the Yaudheyas (1967:21). The Mahabharata speaks of them in relation to war,

and Arjuna is said to have defeated them in battle (7.18.16 and 8.4.46).81 Frequent references to war are not

unusual in the Mahabharata, but it is significant that the Epic continually mentions this group in reference

to battle. Finally, there is an inscription of Rudradäman, which mentions the group's defeat by him and

describes them as "rendered proud as they were by having manifested their title of heroes among all

Kshatriyas" (Dasgupta 1974:197).82 Thus, the Yaudheyas' martial character seems well attested to in a wide

number of sources. That this group would present Brahmanyadeva with a giant spear seems in keeping with

their reputation as warriors.

While we can regard the six-headed figure on these class three coins as a reflection of Skanda-

Kärttikeya as a Graha-like deity, this need not exclude entirely his martial aspect. Grahas are often violent

and feared, but they can be controlled through ritual. I suggest that the Yaudheyas may have viewed

Brahmanyadeva in more of a continuum as a powerful protective deity with close associations with Grahas,

but also with close associations with a warrior's physical might.

4.2.19 The Importance of the Single-Headed Class Three Type

Given what I have suggested above concerning the importance of the six-headed figure, the co-

existence of the single-headed figure and its origins in foreign depictions of deities are significant. I would

like to suggest that iconographic representations can affect the understanding of a deity's character. A

depiction of a six-headed deity and a depiction of the same deity as single-headed express or emphasize

different things and present different understandings of the deity. I do not think that these points would have

81Some scholars also include the section from the digvijaya section of the Mahabharata where Nakula conquers Rohītaka in their discussion of this topic (2.29.5-6). The land is described as "beloved of Kārttikeya", but the people of the land are called "Mattamayüraka", and the word Yaudheya does not occur in the section (2.29.5-6). Certainly the western travel of Nakula, the name Rohītaka (much like Rohtak) and the association with Skanda-Kärttikeya are suggestive of the Yaudheyas, but the lack of their name in this section leaves the matter open to question.

82The original reads: "sarva-Kshatrāvishkrita-vīra-śabdajātotsekāvidheyānām Yaudheyānām." (Sircar 1965:178).

183

been lost on the everyday Yaudheya populace who used these coins. The single-headed type is a much more

human and softer depiction than the six-headed figure. While I cannot claim to know exactly what South

Asians living in north India would have understood from these images, the six-headed image likely inspired

more fear than the single-headed image.83 The six-headed Skanda-Kärttikeya and Sasthi also has a closer

association with a Graha milieu, which was a fear based protitiation cult. Given that part of Skanda-

Karttikeya's appeal was his potentially horrific nature, this softening of his appearance is a major

development.

That both the single- and six-headed forms seem to exist at the same time suggests that this deity's

iconographic form was malleable during this period, as may have been his character. The inclusion and, as

we shall see, eventually complete adoption of the single-headed form are also significant. As we shall see,

the exclusive adoption of the single-headed representation of the deity comes about under Kusāna rule and

results in a very different understanding of the deity than that depicted in the six-headed variety of class

three coinage. The final Yaudheya issues, Allan's class six, all show Brahmanyadeva with one head. As we

shall see, Brahmanyadeva shifts from a horrific six-headed figure with a massive spear to an attractive

young single-headed warrior figure by the class six coins. This process of iconographic humanization and

beautification seems to begin here with these class three coins. A similar softening of his character has also

been observed in my section on texts. The end point of this softening trend is best summarized by the

Suśrutasamhitā, which claims that the deity Skanda-Kumära is incapable of doing anything wrong or

inauspicious and is separated from the malicious Graha named Skanda (SS 37.9-10). The eventual role of

Skanda-Kārttikeya denies his inauspicious roots, and, I suggest, the shift from six heads to one head is part

of this shift from inauspicious to auspicious.

4.2.20 The Worship and Popularity of Brahmanyadeva and Sasthi

We do, I think, have to assume that the Yaudheyas of the period of class three coin production

83In the Epics and Puranas multiple heads are most often associated with demons. When they are associated with gods it is often to reinforce a fearful or awe-inspiring aspect of them.

184

worshipped Brahmanyadeva and Sasthi, and that these deities held some special place for them. Yaudheya

class three coinage was produced almost exclusively in copper, a low denomination and, therefore, likely

accessible to a wide spectrum of society. Hence, the images on these coins must, to be an effective means of

communication, be recognizable to a wide spectrum of society.

A potential clue for the worship of Brahmanyadeva by the Yaudheyas is the device which looks

like a building on the reverse of some of the class three issues. A number of scholars have claimed that

these buildings are temples and indicate the worship of Skanda-Kārttikeya. Thakur suggests primarily from

Yaudheya coinage, but also from non-Yaudheya inscriptions that " ... it may be presumed that the area

[Rohtak] probably abounded in shrines dedicated to him and the cult image enshrined in them was used as a

coin-device" (1974:305; 1981:49). Sinha comes to an almost identical conclusion as Thakur regarding the

existence of temples dedicated to Skanda-Kārttikeya (1979:79). A. M. Shastri also argues that the temples

depicted on the coins are evidence of temples for Skanda-Kärttikeya (1977:93). D. Mukherjee regards the

temples found on Yaudheya coinage to be evidence of actual temple structures (1983:21), and that these

coins show the advanced "architectonic" knowledge of the Yaudheyas (1983:22).

Chatterjee claims that "It may also be safely conjectured that there existed a number of temples

dedicated to this god in the kingdom of the Yaudheyas" (1970:38). Curiously, Chatterjee does not make

reference to the temple or building that appears on the reverse of some class three Yaudheya coins, but

takes notice of a number of inscriptions, which either make some reference to a Skanda-Kārttikeya temple

or make reference to individuals who were devoted to the deity (Chatterjee 1970:38-41). While his

collection of inscriptions is of some use, what should be noted here is that none of these inscriptions relates

to the Yaudheyas, their territory or their apparent worship of this deity. Indeed, as Chatterjee acknowledges

(1970:40), many of these inscriptions refer to wealthy foreigners as devotees of the god. He concludes with

the following statement:

The above mentioned inscriptions discovered from three different parts of Northern India go far to prove the solid popularity enjoyed by this god both among the foreigners as well as the local people. We have already seen that he was regarded as a tutelary deity by such an influential and

185

powerful tribe as the Yaudheyas. A number of early images of this god discovered from Northern India also testify to his popularity. (1970:41)

Without doubt Skanda-Kārttikeya as a military deity had some popularity with foreign groups as we shall

see in my section on the Kusanas, but the inscriptions he supplies do not indicate the existence of Yaudheya

temples during the period of class three coin production.

Ghurye also discusses the existence of temples known from inscriptions and regards these as

evidence of Yaudheya popular worship (1977:138). These temples may refer to popular worship, but the

earliest inscription he provides dates to the early fifith century CE and the next dates to the early sixth

century CE (1977:138). These dates are well after the Yaudheyas are absorbed by the Gupta empire and, in

all likelihood, refer to Gupta and not Yaudheya devotional practices. The arguments raised by Chatterjee

and Ghurye remind us that, other than these coins, there is no evidence of Yaudheya temples.84

It is also important to notice that the temple or building is always associated with the deer on the

reverse of Yaudheya class three coins and never directly with Brahmanyadeva or Sasthi.85 Who or what this

deer represents is not clear. J. N. Banerjea suggests that the stag or deer on Kuninda coinage is a

theriomorphic representation of Laksmi (1956:134). Other scholars have disagreed with this assessment

because a female figure, whom many scholars understand to be Laksmi, appears on these Kuninda coins

with the deer, and some scholars will not accept the double representation of the goddess on the same coin

(Ahmad 1977:159). Whatever the case, the building on Yaudheya coinage is associated with this deer and

not directly with Brahmanyadeva or Sasthi. What is more, a very similar building appears on Audumbara

coinage, where it is largely understood to be a Saivite temple because of the trident and battle axe device

placed before it (Dasgupta 1974:62-63). We only have this deer or stag before this building, and we are not

84 Some scholars have suggested that the temples represented on Yaudheya coins were made of wood and thus would have not survived (Singh and Kumar 1978:3-5), but how one determines the building material of these structures based on these coins is not clear to me

85O. P. Singh regards the temples on the reverse of the class three coins to be shrines for the goddess also shown on the reverse (1977:135).

186

sure what the animal is meant to signify. Texts like the Skandayaga and the Susrutasamhitā also do not

require the use of a temple or shrine for worship of the deity. While these buildings or temples on class

three coinage are often used to demonstrate Yaudheya devotion to Skanda-Kārttikeya, we have to be

cautious about their meaning. What we can suggest is that the appearance of this deity on these coins during

this period of hardship for the group and the nature of the inscription on the coins suggest that this deity

held wide popularity among the Yaudheyas.

4.2.21 Conclusion to Yaudheya Class Three Coinage

These class three Yaudheya coins allow us to place the origins of the six-headed

Brahmanyadeva/Skanda-Kārttikeya in the geographical region of modern Uttar Pradesh and Himachal

Pradesh India and with an indigenous group. We can say that the first artistic representation of the deity

with six heads comes from the Yaudheyas at a point no later than the first century CE and likely earlier. I

have tried to locate these six-headed depictions of the deity within a context of worship offered for

protection and propitiation of a Graha-like figure. The existence of single-headed representations of this

deity and the influence of foreign iconography on these single-headed issues also suggests that the cult was

developing with other influences, and that several iconographic understandings of the deity were already

merging into one figure at this period. I have suggested that the shift to a single head is part of a general

softening of the characterization of this god. Far from reacting against amalgamating and foreign forces

within the Skanda-Kärttikeya cult, the Yaudheya leadership and mint masters either accepted or promoted

such an amalgamation through their coinage.

187

Chapter Five: Skanda-Kārttikeya on Kusāna and Class Six Yaudheya Coinage and Statues

The Kusanas were a group who were foreign to South Asia, but who ruled over a large empire

within South Asia from approximately the first century of the Common Era to the end of the fourth century,

though the empire was greatly reduced after the third century CE (Cribb 1999:191-193). The height of the

empire comes under king Kaniska, who rules as far east and south as Pataliputra in India, and as far west as

the ancient kingdoms of Seistan and Khorasan and as far north as Sogdiana (see map number three). He

ruled from either 100 to 126 CE or 120 to 146 CE (Cribb 1995/96:106). While statues existed in India

before the Kusanas, they become much more common in India during their rule, and a number of new

depictions of deities are introduced during this period. The Kusānas are also significant because they

assimilated a number of cultural traditions into their art work and royal regalia. The exact cultural or ethnic

background of the Kusänas is not known, but they did absorb aspects of the cultures with which they had

contact. Hence, they display a complex amalgam of Hellenistic, Parthian, Scythian and Indian cultures in

their art and politics. As we shall see, the wide variety of cultures that influenced the Kusanas will play a

significant factor in their understanding and use of Skanda-Kārttikeya. The Kusāna era is crucial to the

development of Skanda-Kärttikeya's cult for reasons that I will discuss below.

5.1 Kusana Statues Depicting Skanda-Kārttikeya

In my textual section I suggested that there was evidence from statuary to suggest that Skanda-

Kārttikeya originated as a Graha-like figure who is associated with similar Graha-like groups of Mätrs. In

this section I return to this point and argue that the depiction of Skanda-Kārttikeya as a military deity is

based on non-indigenous sources. Generally, there are two styles of depiction of Skanda-Kärttikeya in

statues from Kusana India. The two types can be geographically separated into those from Mathura and

those from the far north-west or Gandhara region of the Subcontinent (see map four).1 Figure 42 is typical

1It should be noted that the far north-west and Gandhara are separate geographical regions from the Yaudheya territory that I have been referring to as in the north-west of India.

188

of the Gandhäran type. Here Skanda-Kārttikeya stands in a frontal posture and is dressed in Scythian-style

armour2 with a dhoti beneath. He holds a spear with his right hand and a bird in his left hand. The bird is a

cock (Zwalf 1996:121; Czuma and Morris 1985:180). He also wears an ornamented turban and a short

sword on his left leg. There is also a halo around his head. This is a Gandhära piece from Käfir-kot (Zwalf

1996:121). It has been dated to the Kusana period (Srinivasan 1997/98:264) at around the late first to early

second century CE (Czuma and Morris 1985:180). This style of Skanda-Kärttikeya statuary from this region

and time-period were studied in detail by D. Srinivasan (1997/98). It is clear from her article that far north-

west Kusana statues of the deity all represent him in a fashion similar to figure 42.3

As we shall see, the two features that differentiate this style of representation from the Mathurã

type are the cock and armour.4 Certainly, I think the armour on these statues indicates the god is being

depicted as a warrior. In relation to this armour Srinivasan makes the following comments: "This figure

"Srinivasan notes schists that depict Scythian warriors in a similar posture and dress as Skanda- Kārttikeya in this statue (1997/98:239). There have also been archaeological finds of plate armour, which she regards as the model for the dress of these statues (1997/98:239-240, 253-254). Goetz also describes these figures as a warrior type and notes similar figures from schists. He regards the armour as typical of the later Kusānas (1948:18). Pal also thinks that these images are modeled after Scythian guards depicted in contemporary reliefs (1977:22). P. K. Agrawala recognizes the warrior's dress, but does not mentioned the foreign aspect of it (1967:84).

3Srinivasan discusses sixteen examples of her "warrior god" type of Skanda-Kärttikeya from the first to seventh centuries CE from the far north-west (1997/98). About eleven of these examples come from the Kusäna era. The number of examples may be slightly smaller or larger depending on what we consider to be enough evidence for a positive identification. There are, for instance, a number of broken statues that are dressed as Skanda-Kärttikeya, but the spear or cock has, presumably, been broken off. There are also Skanda-Kärttikeya-like statues that do not depict the bird. These have been variously identified as a warrior or as Skanda-Kārttikeya by various writers. For discussions of this issue see Goetz (1948), Zwalf (1996:284) and Srinivasan (1997/98:252).

4Srinivasan states that the earliest depiction of Skanda-Kārttikeya with a cock comes from Mathurā and dates to the Kusāna era (1997/98:236). She cites Chatterjee to back her point, but Chatterjee regards the Kusāna date for it as doubtful (1970:115). The Kusāna date comes from V. S. Agrawala, and R. C. Kar also doubts it: "It may be a century or two later, though no definite opinion should be passed regarding its age in view of its very worn condition" (1954:82). The photograph Chaterjee supplies of this statue (figure four in Chatterjee 1970) depicts a very worn piece. I do not think that we can safely include this statue among the Kusāna era statues. Hence, I think we can suggest that the use of the cock in statuary is limited to the far north-west during this period.

189

dressed in the Scythian garb, implies that the god had followers among foreigners" (1997/98:237), and

"Dressed in the manner of his foreign devotees suggests that the god, too, may have a foreign component in

his make-up" (1997/98:238). I think this idea of a foreign component in the cult of this deity is correct, but I

also will argue that this foreign interest in the deity is much more significant than Srinivasan and other

scholars acknowledge. I will discuss this in more detail below.

Regarding the cock Srinivasan suggests: "The cock may have been incorporated into

Skanda/Kärttikeya's iconography because of its maneuverability; cocks jump with agility when fighting and

this habit may be symbolic of the military prowess of the warrior god" (1997/98:246). There is, however, no

evidence for this understanding of the cock in South Asia from this time period. I think the cock points to

foreign influence particularly given the martial appearance of the statues. A bird is often associated with

warrior-gods in Iranian literature, and, as we shall see, this association is also used in the Kusāna

numismatic representation of Mahasena. The Huviska coins depicting Mahäsena show him holding a bird

standard.5 Birds appear on the coins of two other Kusāna military deities: Orlagno and Pharro. Orlagno

appears on gold coins of Kaniska, where he is depicted with a bird in his headdress and a sword with a bird-

like hilt. His name is an older version of the Pahlavi Varahran/Verethraghna, who was the national lord of

Iranians in arms. He has ten forms that symbolize victory, one of which is a bird form. In the later Sasanian

period he was worshipped by the martial classes as a god of victory (Rosenfield 1993:95). Pharro appears

on coins of Kaniska and Huviska, where he is depicted with a small wing ornament on his helmet. He is an

Iranian deity and the personification of khvareno, a concept of the glory and legitimacy of kings (Rosenfield

1993:96). Rosenfield regards the winged headdress as a significant feature of Iranian symbolism that is not

highly developed in the Kusāna era, but becomes common under the Sasanians (1993:95).

P. Granoff notes that this concept of khvareno was the embodiment of the powers of kingship as

well as a tutelary god of the king and the legitimating factor in his rule. In Iranian literary sources this

5These coins will be discussed in detail below.

190

khvareno is usually represented by a bird (1970:163).6 Birds are thus Iranian emblems of kingly power and

were employed by the Kusānas in their representations of Mahāsena and other martial deities. This

information should change our view of Mahäsena's bird standard and the depiction of a bird with Skanda-

Kärttikeya in the statuary of Gandhära. The bird presented with Skanda-Kärttikeya in statuary is likely used

in the manner of Orlagno's and Pharro's headdress as an Iranian influenced Kusana sign of victory and

royal support. All of these statues with the bird/cock from the Kusana era come from an area geographically

close to Parthia where Scythian and Parthian culture spread, and I think it owes its presence on these statues

to the influence of those cultures. As Goetz (1948:19-20) and Pal (1977:21, 26) note, there is a blending of

foreign and indigenous characteristics and iconography in the Kusana art of the far north-west; Skanda-

Kārttikeya may be part of this "assimilative attitude" (Pal 1977:21). The use of the bird with a military

figure is, in the context of Kusana numismatics, an Iranian device. Its use with Mahäsena on coinage and in

statuary is an indication of foreign influence on this deity and his depiction.7

The cock itself also relates these images of Mahäsena to an Iranian martial theme. Verethraghna is

associated with a cock, but is more typically associated with a raven or crow (Ackerman 1964/65 II:793).

This link between a cock and Verethraghna appears to come about, however, through a link between this

deity and solar cults. Such a link is important to the cult of Skanda-Kärttikeya, but will be explored later

(section 5.2.10). A more direct connection between a cock and an Iranian martial deity is found in the case

of Sraoša. S. Sen (1950) was the first scholar to note a connection between Sraoša and Skanda-Kārttikeya,

but his understanding of this connection is problematic and will be discussed in a later section (5.2.10).

6I would like to thank Dr. Granoff for drawing my attention to this material.

'Srinivasan thinks this warrior depiction may be "the apotheosis of the heroic ideal .... Exposure to steady warfare could have inspired a warrior or hero cult as the invaders became rulers " (1997/98:253- 254). What I think Srinivasan misses is that the adoption of Skanda-Kārttikeya as a military figure represents a dramatic remodelling of the deity that is driven by royal agendas. These points will become clearer as we continue.

191

Sraoša is a martial deity in Zoroastrianism,8 and in Pahlavi texts he is described as repelling evil powers at

night with the help of a cock:

The cock is created to oppose the demons and sorcerers, as a collaborator of the dog. As He says in the Religion: among the material creatures, those are the collaborators of Sroš [Sraoša], the dog and the cock .... for that cock they call the bird of righteous Sros. And when it crows, it keeps misfortune away from the creation of Ohrmazd [Ahura Mazdā]. (Gbd.XXIV.48 and Jfr.II.25 translated by Kreyenbroek 1985:118)

Sraoša's role as a martial protector of the righteous and as under Ahura Mazda is well attested to in the

Zoroastrian tradition (Kreyenbroek 1985:164-183). It is also clear that this cock is associated with his

protective and martial role.

Of great significance to us is that Sraoša was assimilated with Skanda-Kārttikeya in the Upapurāna

tradition and possibly on a Kusāna inscription. In the Sāmba Purāna the figures Pingala, Dandanāyaka,

Rajña and Stosa are described as standing on either side of Surya. Hazra notes that in the almost identical

Bhavisya Purāna Rājña and Stosa are called Rājā and Śrosa, and Rājña and Srausa (1958 I:39). In the

Avesta Mithra is flanked by Rašnu and Sraoša, which appears to be where these Purānic passages find their

origins (Hazra 1958 I:39). The Sanskrit Srausa or Srausa is the same figure as the Avesta's Sraoša. Later in

these Puranas the following equivalences are made: Dandanāyaka is Indra, Pingala is Agni, and Rājña and

Stosa /Śrausa are Hara (Siva) and Skanda-Kārttikeya, though which Iranian deity is equated with which

Hindu deity is not made clear with this last couple (Hazra 1958 I:44-45). I will argue below that it is likely

Śrausa who is equated with Skanda-Kärttikeya. A passage from the 12th century text, the Tikāsarvasa of

Sarvānanda, supports this point. The text reads: "yamo'pi daksine pārśve khyāto māthara samjñāya

pūrvadvāre haraguhau rājaśrausau kramena tau" (Sen 1950:27). Sen regards the compound rājaśrausau to

be a single name, however, meaning "obedient messengers of the king" (1950:27). It is much more likely

that this passage resembles the Bhavisya Purana, which gives Rājā as a name for Rājña (Hazra 1958 I:39).

Hence, the section should read as 'Raja/Rājña and Śrausa, who are Hara and Guha'. Hence, I would argue

8For more details on the cult of Sraoša see Kreyenboek (1985).

192

that it is Sraoša/Śrausa who is assimilated with Skanda-Kārttikeya.

There is another potential link between Sraoša/Śrausa and Skanda-Kārttikeya. This connection

relates directly to the Kusānas, who, I think, first used the connection between the two deities. The link is

found on the Rabatak inscription, which lists a series of Iranian deities who were installed at this shrine

belonging to the Kusāna king Kaniska (Cribb and Sims-Williams 1995/96:77-79). The Zoroastrian deities

listed on the Bactrian inscription are Umma, Aurmuzd, Sroshard, Narasa and Mihr (1995/96:79), but there

is an interlinear inscription, which is not complete, but reads: " ... and he is called Mahasena and he is called

Visakha" (1995/96:79). In short, some point after the original inscription, it was deemed necessary to

identify the Iranian deities with these Indian deities. Due to the worn nature of the interlinear inscription,

however, we cannot be sure which deities Mahasena and Visakha were assimilated (Cribb and Sims-

Williams 1995/96:79). I would suggest that a reasonable candidate for Mahasena is Sroshard. Sims-

Williams explains that the Bactrian Sroshard is cognate with the Avestian Sraošo (1995/96:85), the

Sraoša/Śrausa we have examined above. It is, then, possible that the links between Sraoša/Śrausa and

Skanda-Kārttikeya, particularly in his martial form of Mahāsena, come out of a Kusāna context in the far

north-west.9

On the basis of the above, I would argue that there is more than an iconographic borrowing of a

cock between Sroaša and Skanda-Kärttikeya in the far north-west of India; we should see here a deliberate

attempt at assimilation on the part of the Kusanas. We might also well question the identity of the deity

presented in figure 42. While it and others like it have always been identified as Skanda-Kārttikeya by

modern scholars, it may well also represent Sraoša. I would like to suggest, however, particularly given the

religious eclecticism of the Kusanas, that both Mahasena and Sroaša are intended by the statues. It is, I

suspect, another case of deliberate ambiguity designed to allow a single figure to appeal to a wide range of

people within the Kusana empire. It is clear that the two deities were assimilated to each other, and I think it

9The Rabatak inscription was found in the Afghan province of Baghlan (Cribb and Sims-Williams 1995/96:75).

193

is clear that the Kusānas used martial figures like Sraoša, Orlagno/Verethraghna and Pharro, as well as

concepts like khvareno in their understanding and representation of Skanda-Kärttikeya. The result is the

production of a deity who very much favours the Mahasena aspect of the god, and whose iconography is

linked to Parthian martial devices.

I suggested above that these figures may be another example of deliberate ambiguity. I have also

pointed to the idea of an 'multiple personalities' for Gandharan statuary. I would briefly like to account for

these comments, and to justify in a more general manner the link I see between Parthian/Iranian art and

religion with Indian art and religion in the Gandharan context. An example of how these forces interact can

be explored through various scholarly discussions on the panels illustrated in figures 78 and 79, which

originate from the Swat Valley in Gandhära. Debate over these figures began with Gnoli (1963), who first

discussed the panel in figure 78 and the identity of the deity represented in the third section of the panel

which is illustrated in figure 79. He described this deity as an Indian male divinity with six arms, a spear,

sword and vajra. He noted the existence of these weapons and the similarity of his costume with other

figures from the same site and concluded that he is a warrior deity connected to some local tradition

(1963:31-32). He went onto suggest that the figure could also be Siva (1963:36). The overall point of his

discussion, however, is to demonstrate the influence of Hellenistic art on Gandhāran art (1963:36).

R. C. Agrawala (1966a) and M. Taddei (1966) challenged Gnoli' s account of the figure, and his

exclusive attribution of Hellenistic influence on the art of Gandhara. Agrawala does not think that the image

is Siva, but suggests a blending of Visnu, Siva and Indra in this one form (1966a:82). Taddei agrees with

Agrawala's remarks that the image needs more investigation. He argues, however, that, from an

iconographical point of view, the image relates to Syrian (west Parthian) iconography of the 1st century CE

(1966:84). In general he notes a "stylistic resemblance that links as well other Gandharan products and

Western Parthian sculpture" (1966:84). He attempts to make iconographic links between the Palmyran deity

Shadrafa and the Gandharan Skanda-Kärttikeya because both wear a similar style of armour and carry

spears (1966:85). There may be something in this iconographic link, but I think it is no more than an

194

iconographic link in this case. Taddei also speculates that the six armed deity from figure 79 may be

Skanda-Kärttikeya, but this idea is speculative and not well developed (1966:85). What is of great value in

Taddei's paper is his conclusion that Syrian/Parthian art of the Roman age did influence Gandhäran

representations of Indian deities (1966:88).

Some years later, Gnoli revisited his original article and agreed with Taddei's assessment of

Parthian and Zoroastrian influences on the panel and in Gandhäran art in general (1992:30-33). He also

expanded this line of thinking to Buddhist statues from Gandhara and saw, at least on an artistic level, "a

reflection of a Parthian-Buddhist religious syncretism" (1992:33). He accounted for this artistic syncretism

with the following:

I think, rather, that the explanation for this substantial presence of Western Parthian elements in Gandharan art must be sought in the tendency and probably also in the desire of the Gandharan Buddhist artists to present their religion in a guise that was iconographically familiar and comprehensible in a cultural milieu that was pervaded by Iranian influences in which Western elements played a considerable part. (1992:33)

He thinks, however, that there is nothing more to the borrowing than this iconographic borrowing.

While I agree with Taddei's and Gnoli's general conclusion that Parthian/Zoroastrian art

influenced the art of Gandhära, I also think that we have with Skanda-Kārttikeya and Sraoša a case of more

than simply iconographic borrowing. There is a direct assimilation of iconographic attributes and

'personalities' in the case of these two figures in the Gandhära milieu. I would also not completely dismiss

the possibility of Hellenistic influence on this cult. As we shall see, I think that the Kusäna understanding of

Skanda-Kärttikeya involved a blending of Hellenistic, Parthian and Indian concepts and deities. What

Gnoli's and Taddei's comments do illustrate is that the Zoroastrian or Parthian links I have argued for in

relation to the statuary of Skanda-Kärttikeya or Mahäsena in the far north-west place him within normative

iconographic and assimilative patterns in Kusāna Gandhāra.

The Mathura statues of Skanda-Kärttikeya are not as uniform as the Gandhāra statues and can be

sub-divided into three classes: a Mahāsena type, a Graha type and a Mātr type. The Mathurā Mahāsena, or

martial, type is illustrated in figure 43. The figure holds a spear in his left hand and raises his right in an

195

abhayamudrã. He wears a dhoti, ornamented headdress, earrings, armlets and bracelets. There is an

inscription on the base which reads: "in the eleventh year, fourth month of winter, on the last day, was

installed the image of Kārttikeya by the brothers Viśvadeva, Viśvasoma, Viśvabhava and Viśvāvasu, the

sons of Viśvila, the Ksatriyas, in their own home ... " (P. K. Agrawala 1967:47). There is debate over the era

referred to by the date mentioned in the inscription.10 It is, however, a Kusäna era piece. The brothers

clearly identify themselves as warriors, and their caste helps us to understand who was worshipping this

deity in this particular form during the Kusna era. The military background of the worshippers leads me to

call this type of representation a Mahasena statue. Stylistically, the statue is modelled after Bodhisattvas (P.

K. Agrawala 1967:84; Rosenfield 1993:295), which suggests a lack of indigenous martial prototypes for

Mahasena in this area. This type is also rare in Mathura. There are only two other Kusana examples of this

type.11

The Mathura Graha type is illustrated in figure 41. Here two identical male figures stand on either

side of a female. The two males hold large spears in their left hands and have their right hands raised in an

abhayamudrā. They both wear dhotis and appear to have their hair tied into a bun. The female holds a bag

in her left hand and raises her right in an abhayamudra. In an arch over her head are five other smaller

heads. This female has been identified as Sasthī and the two males as Skanda-Kārttikeya and Visākha (P. K.

Agrawala 1971:326-327; R. C. Agrawala 1971:83; Joshi 1986:11; Härtel 1987:155-157). An exact

identification of the two males is not completely certain, but the similarity between these figures and that

10P. K. Agrawala regards the year 11 to refer to the Saka era and supplies the date of 89 CE. for it (1967:47). It may also refer to the Kaniska era, which would place its origins to the early stages of his rule in the second century CE. Others have argued that the date belongs to the second Kaniska era and have assigned to it a date of 189 CE (Czuma and Morris 1985:115).

This second example is illustrated in figure 44. This small bronze (9.3 cm high) was excavated from Sonkh which is near Mathura (Härtel 1993:281). The figure is worn, but appears with a spear and dressed in a dhotī and elaborate headdress. There is no bird shown with the figure, but there is also speculation that there might have been a second figure attached to this one that broke off at some point (Härtel 1993:281). It has been dated to 100 CE by the excavators of Sonkh (Härtel 1993:281). The third example is held by the National Museum in New Delhi. It is illustrated in figure 45 and is similar to figure 43.

196

illustrated in figure 43, together with the appearance of Sasthi does support such an identification.12 The

appearance of these figures with Sasthi suggests a devotional cult associated with Grahas. As I have already

discussed in the context of Yaudheya class three coinage, a six-headed Sasthi in association with Skanda-

Kārttikeya is evocative of a depiction of Grahas. As we have seen, Visākha also fits well into a cult of

Grahas.

Mathura panels from this period also depict Skanda-Kārttikeya with Mātrs. Figure 47 illustrates

one such panel. This panel is worn, but shows a single male to the far right holding a spear in his left hand

and his right held up in an abhayamudra. He appears to be dressed in a dhoti. To his left is a row of seven

females. They are all dressed in the same manner and have no distinguishing features. Each raises her right

arm in an abhayamudra and her left rests by her side. The male has been identified as Skanda-Kārttikeya

and the females as Mātrs or the Saptamātrkās (R. C. Agrawala 1971:79-80; Joshi 1986:7-9; Harper

1989:68).13 Panels of this type also evoke a Graha or Mätr related cult. Mathurā panels which link Skanda-

12This type of panel is also closely related to panels from Mathura which depict a Vrsni triad. See André Couture and Charlotte Schmid for a detailed study of these Vrsni panels (2001). This type of panel was also misidentified by Cunningham, Vogel, A. K, Coomaraswamy and V. S. Agrawala as a representation of a Nagi or a Queen of the Nagas, but a number of scholars have since corrected this mistake (P. K. Agrawala 1971:325; R. C. Agrawala 1971:82; Härtel 1987:153). Joshi lists four of these panels from Mathura during the Kusāna era (1986:126-128). R. C. Agrawala would add to this number with a unique Gandhäran diptych held in the Cleveland Museum (figure 46). Only half of the original diptych remains, and Agrawala argues that the upper panel represents Skanda-Kärttikeya with Sasthī (1993:274). Later he claimed that both the lower and upper panels represent the two divinities (1995:329-330). He dates the panel to the second to third century CE (1993:271), and Czuma and Morris date it to the second century CE (1985:162). Srinivasan, however, doubts this date and suggests the sixth or seventh century as a more realistic date (1997/98:163). I agree with Srinivasan on the date and do not regard the diptych as relevant to the current discussion.

13Joshi describes ten of these figures from the Kusana period in Mathura (1986:8). While the depiction of Skanda-Kärttikeya is mostly as described above, the number and depiction of the Mätrs or Mātrkās (Joshi and Agrawala use this term to describe these figures) varies. Some are depicted as seated, in which case only as many as five females are shown. Some are shown with human faces, but others with animal and bird faces. Figure 48 depicts one unusual panel from the Lucknow Museum. Here four devotees are in the far right. Next to them is a jar with an animal shaped head. Next to that is Skanda-Kārttikeya in his usual Mathura depiction, and next to him is an animal shaped female with a baby cradled in her left arm. Her right arm is raised in an abhayamudrā. She is identified as a Mātr (R. C. Agrawala 1971:81; Joshi 1986: 8; Harper 1989:68). Joshi also regards this panel as a depiction of Skanda-Kārttikeya's birth. According to him the pot represents the golden pot into which Svāha dropped Agni's semen. The animal

197

Kārttikeya to Graha and Mätr cults are also found in much larger numbers in that region than the statues of a

lone Mahāsena-like figure from Mathurā.

As the above demonstrates, all of the Kusäna statues and panels from the Mathura area that depict

Skanda-Kārttikeya are iconographically related. They all show a deity dressed in Indian clothing with a

spear and without the Iranian or foreign attributes of the bird/cock and armour. The Mathura cult for

Skanda-Kärttikeya also seems much more oriented towards his associations with Grahas and Matrs and less

so to his military aspect. The Graha and Mätr cultic context is not found in the statues from the far north-

west of this period, and Gandhäran or Iranian culture only seems to be interested in the deity as a military

figure. The statues can, then, be generally separated into Mahāsena-like statues from Gandhāra and

Skandagraha-like examples from Mathura.

The conclusion I draw from these statues is that the indigenous understanding of Skanda-

Kärttikeya is as a Graha and protective deity. The Mathura statuary and the Yaudheya six-headed class three

coins all point to such an understanding of this god from this period. The appearance of Mahasena, or the

martial Skanda-Kärttikeya, in statuary seems to come from, or at least is dominant in, the far north-west

where Scythian, Parthian and other foreign cultures were prolific. The Scythian armour, the cock and

complete absence of a Graha or Mätr context for these Gandhära statues all suggest a foreign interest in

presenting Skanda-Kārttikeya or Mahāsena as a military figure at the expense of his Graha-like origins. The

promotion of Skanda-Karttikeya as a military deity is, I think, a Kusana idea that was heavily influenced by

non-Indian sources and reflect royal and martial attributes like khvareno. Where Mahāsena-like figures do

appear in Mathura, they appear to be connected to martial families as the inscription on figure 43 suggests.

head on the pot represents a ram's head and is meant to represent Agni who amuses Skanda-Kärttikeya by appearing as a ram. The Matr represents Lohitayani who protects the new born (1986:8, 20). While Harper agrees with this analysis (1989:69), I am not convinced, and some of his identifications seem arbitrary. The panel certainly suggests that Skanda-Kärttikeya is associated with some form of horrific female who is associated with children. The pot may also represent a Mätr figure or some form of fertility motif. Pots and

Shulman 1980:250). pot-shaped females were used in South Asian art to evoke an idea of fertility (Srinivasan 1997:190-192;

198

The progression I have illustrated in my textual section from Skanda-Kārttikeya as a Graha to the martial

Mahāsena seems to find its origins in foreign interest in his cult. Such a conclusion is significant because

while a number of scholars have suggested that foreign groups were attracted to this deity, none has

recognized the strong foreign influence in the actual development of his cult and his iconography.

We might well question why far north-western groups did not take an interest in Skanda-Kārttikeya

as a deity associated with Graha and Mätr cults. I think the answer lies in the popularity of deities in the far

north-west who fulfilled similar Graha/Mätr functions as Skanda-Kärttikeya and his associates, and also a

different understanding of deities associated with child-birth in that region. Figure 49 illustrates a statue of

Hariti from the far north-west of India. There are numerous statues of this goddess from the far north-west,

and her association with the protection of children is well established. In one Buddhist account of her, she is

a dreadful Yaksī with hundreds of children who dwells in Rajagrha. Each day she eats a baby from the city.

The people of Räjagrha ask the Buddha for help, and he hides one of Hariti's babies from her. She becomes

distraught, and the Buddha uses her anguish to illustrate to her the pain she has caused the people of

Räjagrha. She converts to Buddhism and becomes a protector of children (Peri 1917:1-102; Bivar 1970:10-

21). The figure of Häriti is clearly similar to the Grahas and Mätrs I have discussed above.

Häritī, or a figure like her, is often associated with a male in art from Gandhāra as in figure 50.

Typically, this couple is identified as Pañcika and Häriti, but there is some debate over the identity of these

figures.14 Pañcika was also a military deity in Buddhism and may be related to Skanda-Kārttikeya. Some

have even suggested Pañcika is derived from Skanda-Kärttikeya, but the evidence for this point is not

14Goetz regards the female in this type of statue to have been Hariti or Ardokhsho or Nanaia, and Pañcika as Mihira or Pharro or Verethraghna, or some amalgam of these (1948:20). Pal suggests that the identity of the figures changes from one region to the next (1977:18). He also argues for a blending of Kubera, Pañcika, Pharro and Sūrya characteristics in the male and Nana, Durga and other goddesses in the female (1977:21).

199

strong.15 While Pañcika is usually regarded as a military deity, in these Gandhäran statues of him with this

female his martial aspect is not emphasized. He looks like a portly seated version of the Mathurā Skanda-

Kārttikeya. Pañcika's portliness relates him to Kubera (Pal 1977:14-18); he is more a deity of prosperity in

these statues than a military deity. The same can be said of the female figure. She holds a cornucopia and

displays links between a fertility and protective goddess as well as prosperity in general. It is the symbol of

material prosperity that tends to separate these figures from the Kusāna era Mathurā Skanda-Kārttikeya with

Sasthī or Mātr figures. Skanda-Kärttikeya and his associates are not directly related to wealth and prosperity

during the Kusana era. The Gandhäran view of deities associated with children, however, prefers to

associate them with fertility and prosperity in general. This preference for fertility and prosperity deities

may be behind the rejection of Skanda-Kärttikeya as a Graha in the far north-west at this historical stage. As

we have seen, however, in the Aranyakaparvan, and will see in Yaudheya class six coinage, Sasthi is

amalgamated into a fertility and prosperity goddess cult, but this is, I think, a post-Kusāna development

within the Hindu tradition.

5.2 Kusāna Coinage

The Kusāna king Huviska, who rules immediately after Kaniska, can be dated to either 126 to 164

CE or 146 to 184 CE (Cribb 1995/96:106). It is primarily with the coinage of Huviska that we are

concerned. I will argue that during this period the standard numismatic representation of Skanda-Kārttikeya

is as a military general, and that we can see the political forces at work that bring about this change. I will

also demonstrate that the evidence from coinage furthers my argument that Skanda-Kārttikeya's cult

becomes associated with the ruling elite of the time and disassociated from his roots in Graha and Mātr

cults.

Skanda-Kumāra, Visākha and Mahāsena are all represented on three separate gold issues of

15P. K. Agrawala regards the iconography of Pañcika and Skanda-Kārttikeya to be related. He also notes that Skanda-Kärttikeya is known as the fifth (Pañcika), but he supplies no citation to back this point (1967:48)

200

Huviska. These deities do not appear on any copper issues of the king. While Kusāna kings represent a wide

variety of deities from various cultures on their coins, only Huviska makes use of Skanda-Kārttikeya on his

currency. Compared with other gold coin types of Huviska, finds of these coins are rare, which suggests that

they were not minted in great numbers and may well represent a minor interest for Huviska. With these

points in mind, I will now present the Huviska gold coin which depicts Mahāsena.

5.2.2 The Mahasena Coin of Huviska: A Description

A Huviska gold coin which depicts Mahasena is illustrated in figure 51. On this coin we have

Mahäsena, or the general of the army of the gods, standing in the middle of the coin facing front. His left

hand is on his sword, which is on his hip. His right hand holds a standard topped by a bird and fillets.16 He

wears a long cloak, dhoti and boots. The cloak and dhoti are indigenous, but the boots are a sign of a

foreign or Kusana warrior. He is nimbate and wears a turban or top knot with fillets or side locks flowing

out from the sides. The turban or top knot is often used in Kusana art to indicate an Indian noble. He is

adorned with bracelets on his wrists and biceps, and his cloak is held by a large double clasp on his chest.

The Greek inscription identifies him as MAASENO, or the Sanskrit Mahāsena. Göbl identifies five other

versions of the Mahäsena die type, but there are no significant variations between these versions (1984:22).

5.2.3 Huviska's Mahāsena Issue: Economic and Archaeological Perspectives

The name given to this figure certainly suggests that the Kusanas were interested in the martial

aspect of this deity.17 There is, however, much more to this selection than first meets the eye. The choice of

Mahāsena to appear on these coins may tell us a great deal about the role these divine images played in

16 An exact identification of the bird is difficult, with scholars suggesting both a peacock and a cock, both of which are emblems of Skanda-Kärttikeya. Gardner (1886:138), Smith (1972:76), Rosenfield (1993:79), P. K. Agrawala (1967:43), Chattopadhyay (1967:80) and Göbl (1984:41) all simply describe it as a bird. Singh (1978:74), Sinha (1979:81), Thakur (1981:51) and Gupta (1988:6) suggest the bird is a peacock. Navaratnam suggests it is a cock or peacock (1973:100). I think the identification of a peacock is incorrect because what may have been interpreted as the long plumage of a peacock are actually fillets attached to the standard Mahäsena holds. The only concrete statement that can be made, however, is that the figure on the end of the standard is some sort of bird.

17The name Mahasena literally means 'possessed of a great army'

201

Kusāna coinage, and how best to study these coins. I think we can understand the appearance of Mahāsena

on these coins if we take into account the wider artistic, archaeological and economic context of these coins

and images.

The Kusänas only issued copper and gold coins, and the gold coins were worth a great deal of

money (MacDowall 1960:68; Holt 1999:116-117).18 These coins were likely produced for, and circulated

among, people of considerable wealth with enough education to identify the deity by the Bactrian legend

inscribed in Greek letters. Unlike the copper Yaudheya class three issues, these Skanda-Kārttikeya gold

coins issued by Huviska were designed to send a message to members of the aristocracy. While we are

looking for religious figures on the reverses of Kusana currency, contemporary users of the coinage must

have seen political statements in them as well as religious images. The political statement here is addressed

to a very small and elite segment of Kusāna society.19

Such a conclusion is important because it is widely assumed in studies of Skanda-Kārttikeya that

these coins demonstrate that the god was widely popular during this period. Some argue that Huviska must

have been a special devotee of Skanda-Kärttikeya. For some scholars this assumption is simply based on the

existence of Huviska's Skanda-Kārttikeya issues (Chatterjee 1970:33; Navaratnam 1973:99; Ghurye

18MacDowall argues that the Kusana gold coins were worth about the same as Roman gold coins of the time period, and that they "would probably be exchanged principally in large scale commercial transactions by international traders" (1960:68). In relation to Indo-Greek kingdoms Holt states: "Ancient silver and gold often circulated over the heads of most farmers and poor tradesmen, who relied upon a bronze token coinage in order to participate in the Greek's world of monetary rents, tolls, taxes, and fixed price trade" (1999:117). Similar conditions to those Holt describes likely prevailed in Kusana India.

19These comments are primarily inspired by the work of Cribb, Errington and others. They comment regarding the devices used on South Asian coinage of this period: "Another important factor could be the need to present through the coin designs the political power and public image of the ruler by whose authority the coins are issued .... The divine images and divine symbols used within this tradition were, like the royal portraits, intimately related to the intentions of presenting a public statement about the ruler and his authority" (1992:49-50).

202

1977:137; Gupta 1988:6; Rana 1995:21).20 Other scholars have been mislead by an obverse coin pattern

which appears to show Huviska holding a cock standard. This obverse image has lead Singh (1978:76) and

Thakur (1981:53) to argue that the king is trying to depict himself as Mahasena, and that this imitation

shows his devotion to the god. The appearance of this cock standard is, however, the result of a damaged

die and forgery (Rosenfield 1993:64; Göbl 1984:129).21 It is this flawed and forged coin to which Thakur

and Singh refer, lending little credence to their arguments concerning Huviska's supposed devotion to

Mahāsena.22

More accurate are the arguments presented by Chattopadhyay (1967:91) and Sinha (1979:83).

Sinha's argument comes primarily from Chattopadhyay's, and I will only discuss Chattopadhyay's views

here. He argues that Huviska was probably a devotee of Skanda-Kärttikeya because he represents himself as

20J. N. Banerjea notes that even though Skanda-Kärttikeya is not found on the textual lists of what he calls "the five principle gods of the five cults," the 'tribal' and Huviska coins which portray this deity demonstrate that he was "highly venerated by a good many people of ancient India" (1956:140). Chatterjee argues that the manner of representation of Skanda-Kärttikeya on these coins "proves beyond all shadow of doubt" that Huviska "had some special sentiment for the Indian War-god" (1970:33). He goes onto suggest that Huviska must have come into conflict with Nahapana and "invoked the aid of Skanda-Kārttikeya the great Indian War-god" (1970:33). There is no evidence to support this last claim. Navaratnam claims, based on these coins, that "The Kushanas and the Kshatrapas who ruled North-west frontier provinces of India were great devotees of Brahmanya-deva" (1973:99). Ghurye claims based on these coins that the Kusānas "accepted Hinduism as their faith" (1977:137). Gupta claims that these coins demonstrate that "the popularity of Skanda grew substantially" during the Kusana period, and that the Kusana kings viewed Skanda-Kārttikeya "as a patron deity" (1988:6). Rana comments "The representing of Skanda on the coins of a foreign dynasty indicates how important this god was at that time" (1995:21).

21The coin with the damaged die is illustrated in figure 52; the modern forgery of it is illustrated in figure 53. Rosenfield comments regarding this die type: "This series includes the interesting examples of an obverse die which became damaged, its flaw becoming increasingly worse until the legend had to be recut in part. This in turn inspired a clever forgery ... which gave rise to the erroneous notion that Huvishka holds a bird standard in his hand" (1993:64).

22Singh argues that the obverse die type in question is evidence of Huviska's faith in the deity (1978:76). He also claims the coin was issued after he defeated the Yaudheyas "in order to make his stable government in conquered area [sic], paid respect to the national god of the people" (1978:76). He supplies no evidence to back this last point. Thakur suggests that the Skanda-Kärttikeya issues of Huviska illustrate that the Kusanas worshiped the god (1981:51). Regarding these obverse types of Huviska he claims they depict the king's "unflinching faith" in the deity (1981:53). In a separate article Thakur claims that the popularity of Skanda-Kärttikeya is attested in the Gandhära region by these coins of Huviska (1974:308).

203

Mahäsena on the obverse of some of his coins, but he does not appear to be referring to the

damaged/forgery die discussed above. Chattopadhyay provides no illustration, but he seems to refer to an

obverse type illustrated in figure 54. This obverse die is genuine. Huviska appears in a seated position and

holds a staff in his right hand "with apparently bird insignia at the top" (Rosenfield 1993:62). Göbl is,

however, unsure of what sits on top of the staff, calling it simply a "Tierszepter" (1984:36). The reverse

types which go along with this obverse type either depict Nana or Ardoksho. The staffs of Huviska and

Mahäsena are not identical, but there is reason to believe that Huviska holds a bird-topped standard similar

to Mahasena's.

We do, however, have to be cautious when coming to conclusions regarding this obverse type. We

must also consider the reverse depictions associated with this obverse type. If the king really wanted to

draw a clear comparison between himself and the god, it would have made more sense for him to place this

obverse type with a Mahasena reverse. This obverse die type is also rare. It is one of thirty obverse types

used for the king's coinage and certainly not a common one (Göbl 1984:36-37). The vast majority of

Huviska's obverse dies show a bust of the king from the chest up. This unusual sitting die type seems to

refer back to earlier Kusana obverse coin dies of Kujula Kadphises and the still earlier Saka issues of Maues

and Azes II (Rosenfield 1993:62). Cribb advises caution when dealing with exceptional issues of Huviska

and Vasudeva I because "they might represent the uncontrolled whims of individual die engravers or

officials working at Kushan mints. Only the normal types are likely to show the general requirements placed

upon the imagery of the coinage emanating from the king or his court" (1997:20). He does not advise that

we ignore exceptional issues, but we cannot use only rare issues to argue for a clear propaganda campaign

by the king or his court from such issues.

This obverse die type does not, I think, tell us about Huviska's devotional tendencies as Sinha and

Chattopadhyay think it does, nor does it inform us about the general popularity of Skanda-Kärttikeya in the

king's realm. This coin type may well suggest that one local minter thought that the king and the god were

related. This minter also chose to have the king resemble Kujula and two Saka rulers for unclear reasons.

204

The type may also simply tell us that the bird topped standard had become a symbol of military prowess. I

have already argued for the use of birds in such a context; this Huviska coin may be another example of it.

The rarity of this obverse type, however, means that we should not overemphasize or overgeneralize the

potential meaning of this issue type.

Returning to the points of other scholars regarding Skanda-Kärttikeya's popularity during this

period, I do not wish to contend that Skanda-Kärttikeya was not popular during this period, but only that

these coins are not clear evidence of the extent of the deity's popularity. They also give no indication of the

circles (beyond the ruling elite) in which the deity was popular during this time. What I will argue for,

however, is that the elite aspect of the use of these coins suggests that they have more to do with the

intersection of religion and politics than with broad-based popular religion or devotion to the military deity

placed on them.

An examination of archaeological evidence can help us to build an argument concerning the

message that these coins were supposed to deliver. There are two potential numismatic forerunners to this

Mahäsena type. They are both illustrated in figure 55. The first coin is from the Indo-Scythian king Azilises,

and it depicts a single Dioscuri.23 This Dioscuri appears to stand frontally and to hold a spear in his right

hand. The second coin in figure 55 is of king Hyrcodes. The reverse shows a figure standing frontally in a

similar posture as Mahäsena. The identity of this reverse figure is not clear, and Gardner describes him as

follows: "Figure of a Deity (?), facing; holds spear in r. hand; flames on shoulders" (1966:117).24 Just who

23The Dioscuri were twin gods and to have only one appear on coinage is unusual. The Dioscuri will be examined in more detail latter in this chapter. Azilises' dates are not clear, though he was likely ruling in Gandhära and the Western Punjab from the middle of the first century BCE to approximately 30 BCE (Mitchiner 1976 V:481).

24Some numismatic depictions of Kusäna kings also show the king with flames coming from his shoulder. It may well be that this figure on the Hyrcodes' coin is meant to represent the king. Rosenfield notes that the figure may represent Pharro (1993:17).

205

Hyrcodes was, though, and where he ruled are not clear.25 Whether or not these coins were used by

Huviska's minters as prototypes is open to conjecture.

Another potential prototype is, of course, the Yaudheya class three issues. The six-headed

Yaudheya style is never used by the Kusanas. As I have suggested earlier, the Kusānas promoted the

depiction of anthropomorphic deities, which may, in part, lie behind their rejection of the six-headed type.

The single-headed Brahmanyadeva type, however, may have influenced the Kusana coin in question. Both

Brahmanyadeva and Mahasena stand frontally in a similar posture. Beyond the similar posture, however,

the two figures have little else in common, and iconographically the Brahmanyadeva on the Yaudheya class

three coins is not the equivalent of the Mahasena on the Huviska coins. Whether or not the Kusānas

borrowed stylistic traits from the Yaudheyas must be left open to conjecture.

A more likely source of inspiration comes from statuary. Cribb has pointed out that the image of

Mahāsena on these coins appears to be copied from contemporary statues of Skanda-Kärttikeya (1985:67;

1997:32). We have already seen examples of the statues Cribb may have in mind.26 While none of the

statues illustrated in the previous section is an exact match for the image on the coins, they do share some

significant common features with the coin images, especially in that both face front. This "rigid frontality"

on Kusana coinage is rare, and for Cribb such an attitude indicates the use of a sculptural prototype for the

coin image (1985:67).27 Borrowing from contemporary sculpture was an unusual practice in Kusāna coin

25Gardner places him with late Indo-Scythian rulers in Bactria, but makes no direct claims concerning the identity of this ruler (1966:xlviii). Rosenfield places him a the period of the rise of the

(1993:17). Kusäna dynasty, but notes that this is tenuous due to the paucity of information we have on the figure

26In his paper, "A Re-Examination of the Buddha Images on the Coins of King Kaniska: New Light on the Origins of the Buddha Image in Gandharan Art", Cribb convincingly argues that the iconography of the Buddha on Kaniska's coinage was inspired by sculptural prototypes. He mentions as an aside that Huviska's Mahasena coin fits the same frontal conventions as the Kaniska Buddha type of coin, and that sculptural images could have also provided the prototype for the Huviska coin (1985:67).

27I should note, however, that the Azilises, Hyrcodes and Yaudheya coins illustrated earlier also present their figures with this same rigid frontality. We do not, however, know if they were copied from statuary.

206

design and is only found on coins representing Mahäsena, the Buddha and Heracles (Cribb 1997:32). The

image of Mahäsena on these coins may be one of the few Kusāna coin images which directly relates to what

may have been cultic images. This observation may help us to come to a better understanding of the extent

and nature of this god's popularity. I do not, however, subscribe to the argument that claims we can evaluate

the popularity of this god simply based on the numbers of his statues from this time-period. Srinivasan, for

instance, uses this type of numbers logic to make the following claims:

During the time specified, images of the Hindu god Skanda/Kärttikeya in the Northwest outnumber most other major Hindu gods represented in the region .... In sum, an assessment of the images discussed below could lead to the conclusion that Skanda/Kärttikeya is a major Hindu god in the Northwest during the incipient phase of Hinduism there. (1997/98:233-234)

I will not deny that this type of representation of Skanda-Kärttikeya must have been popular with some

aspect of Gandhäran society, but a much more nuanced understanding of this deity's 'popularity' can be

arrived at, which does not suggest that these statues indicate widespread popularity for the cult of Skanda-

Kārttikeya, or Mahāsena, in a martial form.

So far, I have stressed that these coins belong to an elite social context. I have also noted that the

Mahäsena statues from the far north-west, and in particular the cock emblem on them, indicate a martial and

royal interest. I have also demonstrated that textual accounts of this deity stress his martial nature and also

suggest an elite context of him. What I argue for in the following is that Mahäsena's, or the martial Skanda-

Kärttikeya's, 'popularity' is limited to a royal, elite and military clientele and did not have the widespread

support of Skanda-Kārttikeya as a Graha. Mahasena's popularity is a highly circumscribed one, and this

shift into a narrow social group and specific role in that group will have great implications for the long-term

evolution of his cult and characterization.

One of the unusual elements of Skanda-Kärttikeya's statuary from this period is the complete

absence of terracotta figures that could be identified as Mahasena or as martial in nature. My own survey of

existing sculpture indicates that the first evidence of indigenous warrior terracotta figures and Skanda-

Kārttikeya terracottas appears at around the third century CE. These figures show the influence of Gupta

207

sculpture. While terracottas of Skanda-Kärttikeya are produced during the Gupta period, few could be

described as resembling the stone statuary of the deity produced during the Kusana era. As we have seen,

the statues of Mahasena from the first and second century CE are all stone or metal. They show a

sophisticated level of craftsmanship and were, in all likelihood, expensive to produce and to purchase. It is

also not clear that the statues of Mahasena were meant for public use. With the exception of the Mathurā

statue (figure 43), these statues are all small (under one foot in height) and may have been intended for

private or portable shrines. The inscription on figure 43 also makes it clear that the statue was set up in the

house of the brothers and may not have been for public worship.

The known examples of Mahäsena statuary thus appear to lack indigenous and folk precursors and,

like the gold issues of Huviska, belong to elite and wealthy sections of Kusana society, who could afford

stone and metal statues for private use. Without clear signs of a popular or folk cult in terracotta or other

forms we must, I think, conclude that Mahäsena was popular during the Kusana era, but only with the

martial and wealthy elite of that society and, in particular, with foreign sections of that society.28

A counter-argument to my hypothesis is that the Kusana era is a period when a number, if not all,

of the religious iconographic forms appear on the Subcontinent without obvious indigenous precursors.

Thus, the statuary of Skanda-Kärttikeya simply fits this model of sudden iconographic development, and

does not, therefore, tell us about the nature or manner of his cult beyond the fact that the cult follows the

same mysterious developmental pattern as other cults at the time. Indeed, some art historians regard the

sudden appearance of a well-developed statuary in South Asia without precursors as the central mystery of

the development of art in the region (Srinivasan 1997:185).

28This argument is the opposite of that presented by Srinivasan regarding these Mahāsena statues. She suggests that "Presumably both the local and foreign segments of the population were followers of the god's cult" (1997/98:238). We have, however, no evidence for this presummed 'local' worship of this deity in this form. We also have a third century epigraph from the Hazara district which states that a foreign individual named Sāphara established a shrine to Kumāra (Chatterjee 1970:39; Sircar 1953/54:59-62; Srinivasan 1997/98:237). We simply have no evidence from the far north-west to suggest that the martial form of Skanda-Kārttikeya was worshipped by local groups.

208

My reply to such arguments comes from examining the statuary of one of Skanda-Kārttikeya's

'forms', Naigamesa.29 The importance of Naigamesa, for the current argument, is that we find crude

terracottas of him along with well-developed terracottas and stone figures of the deity. Terracotta images of

the goat-headed deity as both male (Naigamesa) and as female (Naigamesi) are found in substantial

numbers in archaeological finds in north-west India and the Ganges basin (Jayaswal 1991:41). Figures 56a-

b illustrate some crude terracottas from Khairadih. Similar figures have been found at Rajghat, Kumrahara

and Vaisali in relatively large numbers (Jayaswal 1991:36). There is reasonable textual and archaeological

evidence to suggest these Naigamesa and Naigamesi forms had a ritual function associated with pregnancy

and childbirth (Jayaswal 1991:42-45). They are also, I hasten to add, cheap to produce, and their find

locations at archaeological dig sites suggest that they were used across all levels of society from the

approximate period of Kusana rule until the fourteenth century CE (Jayaswal 1991:41). Besides these crude

terracottas there are also some sophisticated terracotta and stone images of these deities, two of which are

illustrated in figures 57 and 58. These statues come from the Mathura museum and date to the Kusāna era.

The existence of the crude terracottas as well as these more sophisticated statues suggests that the

Naigamesa/Naigamesī cult was widely popular during the Kusana period. They also suggest that these

deities were worshipped by all segments of society. The contrast between the iconographical traditions of

the two figures supports my hypothesis that Mahasena was worshipped primarily in wealthy or official

circles within the Kusana empire and did not experience the broad-based popularity of

Naigamesa/Naigamesī.

We can also draw the Mathura panels which depict Sasthi and Mätrs with Skanda-Kārttikeya into

this discussion. Some of these images have a clear cultic context, if we judge from the representation of

devotees on them (Joshi 1986:8). Some of these panels are also relatively large and may have been intended

29In the Suśrutasamhitā Naigamesa is a goat-headed Graha under Skanda-Kārttikeya's leadership. In the Salyaparvan of the Mahabharata Naigamesa is presented as a goat-headed form of Skanda- Kārttikeya.

209

for public worship.30 As I have suggested above, however, these panels do not evoke the warrior Mahāsena.

They may indicate a cult with similar motivations as the Naigamesa/Naigamesi cult, but this cult is not

directly related to martial ideals in the manner of the north-west statues and Mathura statues like figure 43.31

These Graha and Mätr statues are likely a reflection of the original cult of Skanda-Kärttikeya and reflect the

basis of his popular support. This characterization of the deity will, however, largely disappear from

iconographic sources as the martial Skanda-Karttikeya comes to dominate his cult. The context of these

statues and this cult differs from what we have seen of the Mahasena cult and statues.

Hence, we need to be careful how we use the word popular in relation to this deity during the

Kusāna era. We also need to be specific as to which form of the deity we are referring to in discussions of

him. It is clear from the above that a least two different depictions and understandings of this deity were

current during this time period. These two understandings of the deity are also divided by the cultural and

social groups to whom they appeal. The Mahasena statuary and coinage is focused on a martial figure

whose popularity seems limited to, or primarily directed towards, a socially elite group of non-Indian rulers.

The Graha and Mätr related Skanda-Kärttikeya statuary, and related Naigamesa/Naigamesī cults, seem to be

a reflection of an earlier more broad-based popularity, which is not obviously focused on a particular social

group and appears to be indigenously based.

5.2.4 The Skanda-Kumāra with Visakha coin type of Huviska: A Description

Another gold issue of Huviska depicts Skanda-Kumāra with Visākha. This coin type is illustrated

30Mathura Museum piece number F. 39 is 79x40 cm, and number F. 38 is 120x40 cm (Joshi 1986:123). Both depict Mätrs with Skanda-Kärttikeya. The panel illustrated in figure 42 is 65 cm high. Panel F. 3 from Mathura is 1.17 m high, and the State Museum of Lucknow piece (number J. 84) is 51x54 cm (Joshi 1986:127). These panels depict Sasthī with the two male attendants.

31 As we have seen, however, parts of the Naigamesa cult are absorbed into the cult of Skanda- Kārttikeya. Naigamesa does become an epithet of Skanda-Kārttikeya, but the Naigamesa/Naigamesī cult also seems to continue on separately of the Skanda-Karttikeya cult. There are no icons which suggest that the cults were amalgamated in the lived tradition, and Naigamesa/Naigamesi seem to have an enduring popularity until a late period.

210

in figure 59.32 Skanda-Kumära stands to the right of this gold coin and is identified by the Bactrian

inscription which reads CKANAOKOMAPO (Skando-Komaro), or the Sanskrit: Skanda-Kumāra. He is

turned towards Visakha, with his right hand holding Visakha's right hand. His left hand is outstretched and

rests on Visakha's hip. He appears to have a short staff which rests in the crook of his right arm, and a

sword is on his left hip.33 He is nimbate and dressed in a dhoti, cloak and what appears to be a skull cap. He

is well ornamented with bracelets on his wrists and biceps and two necklaces .

Visākha stands to the left of the coin. He is identified by the Bactrian inscription which reads

BIZATO (Bizago), or the Sanskrit Visākha. He is turned towards Skanda-Kumära with his right hand

holding Skanda-Kumara's right hand; his left hand holds a spear, and a sword is on his left hip.34 He is

nimbate and wears a dhoti and cloak. He is adorned with bracelets on his wrists and biceps and two

necklaces. He may have a mustache and sideburns. He also appears to have some sort of headdress which

may be a wreath or band with small fillets.

Göbl identifies a second and third type of this issue illustrated in figures 60 and 61 (1984:13).

These coins are basically identical. The most obvious difference between the coins is that the two deities in

figures 60 and 61 do not hold hands. Skanda-Kumära's staff is longer, and it is topped with a ball in figures

60 and 61. Also, both of Skanda-Kumra's hands are on his hips and his headdress may be different. The

Kumära part of the inscription is now below the platform on these coins. In figures 60 and 61 Visākha has

his right hand is on his hip. The symbol between the two deities is also different, as is the monogram

32 As in previous sections of this thesis, when I refer to a specific depiction of Skanda-Kārttikeya I will use the specific name given to him in the source and not the general label Skanda-Kärttikeya.

33Gardner (1886:149), Chattopadhyay (1967:80), P. K. Agrawala (1967:43), Chatterjee (1970:32), Singh (1978:74), Sinha (1979:82) and Thakur (1981:51) all describe, or cite Gardner's description of, Skanda-Kumära as carrying a standard surmounted by a bird, though Sinha acknowledges some doubt on the point. I, however, agree with Banerjea, who argues that what appears to be a bird is actually part of the letter M and part of the headdress of the deity (1956:145). There are two other examples of this coin type illustrated below which certainly do not depict the deity with a bird standard.

34Rosenfield sees him holding a trident; I think it is a spear (1993:99).

211

between their heads. Göbl also illustrates three types of half stater gold issues (1984:14). These coins are

illustrated in figure 62. These die types are basically the same as in the coins illustrated in figures 60 and

61

5.2.5 The Skanda-Kumara with Visakha issue of Huviska: Two Forms of the Same Deity or Two Different Deities?

Whether or not these coins depict two separate deities (as in Ayurvedic texts) or two aspects of the

same deity (as in the Mahabharata) has been a topic of scholarly debate concerning these coins. There is

general scholarly agreement that the coins depict two separate deities;35 I concur with this view.

In addition to textual evidence that supports an independent identity for Skanda-Kārttikeya and

Visākha there is the evidence of the coins themselves. First, I think, the hand shake and hand positions from

the coin in figure 59 imply a friendly relationship and do not suggest that the deities are the same. An allied

relationship is the impression that best fits the images presented on the coins.

Second, we have other examples from Kusäna coinage of two deities presented on the same coin

which allows for a point of comparison. The coin illustrated in figure 63 depicts Mao and Miiro (Mithra),

deities associated with the sun and moon on Kusāna coins. This is one of the most common Kusāna golds.

There is no doubt that the two deities are related, but they are also clearly different deities in the Iranian

35This argument is primarily concerned with comments by D.R. Bhandarkar on another coin of Huviska which depicts Skanda-Kumara with Visakha and Mahäsena. Bhandarkar claims that the three figures were really forms of Skanda-Karttikeya and not separate deities (1921:22-23). Most scholars, however, follow J.N. Banerjea, who rejects this assertion and argues for separate identities for the deities (1956:145). John Rosenfield regards them as "still separate entities not yet integrated into a single figure" (1967:100). Chattopadhyay (1967:179), P. K. Agrawala (1967:43), Chatterjee (1970:33), Singh (1978:75- 76), Sinha (1979:83), Thakur (1981:53), Gupta (1988:6), Mani (1990:69) and Shastri (1997:47-50) follow Banerjea's argument and comment that the deities have yet to be merged into one figure. S. S. Rana does not present a clear discussion of the coins, but suggests the figures are "aspects" of Skanda-Kārttikeya (1995:21). Banerjea, Chatterjee and Thakur also claim that iconographically the figures are so close they may as well represent the same deity. I cannot agree with such comments. While the figures on the coins are iconographically similar, there are easily recognizable differences between them. Navaratnam seems confused in his account of the coins citing Allan who gives no account of the issues (1973:100). Navaratnam likely intends to cite Gardner, but he does not present his own opinion on the nature of the deities presented. Finally, G. S. Ghurye does not present a conclusive discussion of the coin only stating: "[it is] the Kushans, that has left us the earliest actual representations of Skanda, Mahasena, Visakha or Kumara" (1977:137).

212

religion from which they are taken. Other Kusāna examples are of Oēšo -- Ommo and Oēšo -- Nana, which

are illustrated in figure 64. Again, these coins depict two deities who share a relationship, but are not the

same deity.36 Thus, similar coin types from the Kusana catalogue of coins do not suggest that placing two

deities on the same coin implies that the deities are the same.

Another example of two deities shaking hands comes from Roman coinage. These coins are

illustrated in figure 65. The first coin depicts Mars giving a statuette of Victory to a ruler, and the second

depicts Mars and Rea Silvia. On both of these coins the deities are understood to have a relationship, but

are not the same deity. These Roman coins may well have been the prototype used by Kusāna minters for

the Skanda-Kumāra with Visäkha issues. Göbl has convincingly argued that one source of inspiration for

Kusāna numismatics comes from Roman imperial coinage produced in Alexandria (1960:75-96;

36A. C. D. Munshi also sees a connection between these coins, but regards their connection to be political. Munshi argues that Huviska first issues the Nana-Oeso coin in an attempt to "Indianise NANA" (1986:92). The Mahäsena coin is then issued, he claims, to represent Mahäsena as the son of Nana and Oešo in a move to confirm that Nana is Oešo's consort (1986:96). Munshi provides no solid evidence to back this claim. He goes onto claim that the Skanda-Kumara with Visakha coin is issued to complement the Umma-Oešo coin by providing them with offspring (1986:97). The author claims that this was a response to Indian outrage over Nana being presented as Oešo's consort, but again there is no evidence for this. Munshi then claims that Huviska "retaliated" against this Indian religious sentiment by issuing the Mao-Miiro type in an attempt to "deliberately slight ... OMMO, the greatest of all Indian Goddesses" (1986:97-98). The Mao and Miiro coin was issued first under Kaniska's rule, and Munshi's arguments are simply unfounded. Munshi then supposes a conflict between an Indian camp and an Iranian camp, which ends in a reconciliation celebrated by the Mahasena with Skanda-Kumāra and Visakha coin which will be discussed later (1986:98-100). He feels this coin demonstrates:

definite proof that the Kushana commaner-in-chief [sic.] had to accept the terms which are set fourth by Indian commander-in-chief as a condition for release of king Huvishka and restoration of kingdom to him and the most important of these terms are (a) Huvishka must cease forthwith his anti-Indian religious policy especially in context to female deities' representation in his coin issues. (b) Huvishka must Indianise his policies in administering the Kushāna possession within India. (c) In return, the Indianised Kushana king shall receive spontaneous and wholesome support and protection of the Indian Camp. (1986:100-102)

Munshi has no evidence to support any of these assumptions. I regard his comments on these coins as fanciful speculations.

213

1984:174).37 Most numismatists agree that Huviska either imported Roman coin designers or that his Indian

coin designers had access to Roman style manuals or coins.

Another potential numismatic prototype from South Asia may be the Dioscuri (Dioskouroi) issues

of the Indo-Greek king Diomedes and the Saka king Azilises. Diomedes' copper coins are illustrated in

figure 66. The illustrated coins depict the twin deities standing facing frontally. The figure on the proper

right holds a long spear in his right hand, and his left hand rests on his hip. The figure to the proper left

holds a long spear in his left hand and has his right hand on his hip.38 These coins show a strong similarity

with the Skanda-Kumara and Visakha issues discussed above and may have influenced the Skanda-Visākha

coins.39 Greek and Roman mythology concerning these twins certainly places them in a close relationship,

but, even as twins, they are still regarded as separate beings.

Third, there is also textual evidence for paired deities in Vedic religion and Hinduism. J. Gonda

argues that the 'pair' was regarded as a fundamental unit in Vedic and Indo-European culture in general

(1974:17). He notes that separate deities were often joined into one grammatical and ritual unit in Vedic

texts (1974:19). The pairing-up of co-operative gods on these coins may, then, have met with a sympathetic

audience in India. While many of the influences affecting these Kusāna issues may have been foreign to

South Asian culture, there was still a strong tradition within South Asian culture that would have accepted

37Göbl's "Roman Patterns for Kushana Coins" (1960:75-96) presents a general hypothesis that some Kusana issues were strongly influenced by Roman issues of the same time period. He provides some specific examples, but not for the Skanda-Visakha coins in question. His later monumental work, Münzprägung Des Kušānreiches, does provide some specific examples of Roman prototypes for these Skanda-Visakha coins as well as Huviska's Mahasena-Skanda-Visakha coin which I will discuss later (1984:174).

38Other issues of Diomedes present the twins in their more common representation as lance carrying warriors on horseback.

39I am not the first scholar to have noticed the similarity between the two coin groups. B. Chattpadhyay comments regarding these coins: "Iconographically the depiction of two armed Dioskuroi standing facing, side by side with weapons in their hands, as we find on the coins of Diomedes, has striking similarity with Skanda-Kumära and Visakha represented on the reverse of the coins of Huviska" (1977:108- 109). He does not, however, expand on this point. I will be exploring the similarity in more detail later.

214

and may have influenced the depiction of dual divinities on these coins. The conclusion that must be

reached regarding these Skanda-Kumāra with Visäkha coins is that they depict two related, but different

deities.

5.2.6 The Dioscuri and their influence on the Skanda-Kumara with Visakha coin

As I suggested above, I think the Dioscuri issues of Diomedes and Azilises influenced the Skanda-

Kumāra with Visakha issues of Huviska. I will now try to argue for this hypothesis. The earliest Indo-Greek

coins that depict the Dioscuri belong to Eucratides. On his issues the Twins are presented in their more

usual depiction: lance carrying warriors on horseback (Errington 1992:60). Narain argues that this king

began ruling in Bactria at around 171 BCE,40 and that he conquered the Paropamisadae, small parts of

Gandhära and regions south of the Paropamisadae to Kandahar (1957:53, 69).41 The king's early dates,

however, may indicate that his coins probably did not directly influence Kusāna minters.

Diomedes issued coins which depict both mounted and unmounted Dioscuri. Unfortunately very

little is known about this king. Tarn thinks he was a king of Gandhära (1938:315), and Narain places him in

western Gandhära (1957:104, 156). Tarn dates him at around the first century BCE (1938:316), while

Narain suggests 95-85 BCE (1955:34), and Mitchiner suggests c. 110 to 80 BCE (1975 III:203). While

there are no significant finds of his coins, he is one of the last Indo-Greek kings to rule in Gandhāra before

the Sakas, Parthians and Kusānas invade. Thus, some of his currency may have been present during Kusāna

40Mitchiner dates this king to c. 171 to 135 BCE (1975 I:86).

41Tarn and Narain have very different views on Eucratides. Tarn sees great significance in the Dioscuri coins used by the king. He argues that the Dioscuri type imply a Seleucid link for Eucratides, and that he was related to the Seleucid king Antiochus IV. Tarn also thinks that the king saw himself as a soter figure and placed the twins on his coins to make that point. Finally, Tarn suggests that the Dioscuri as twins represent the close relationship between two men, and he feels these issues of Eucratides represent his close relationship with Antiochus (1938:204-206). Narain rejects most of Tarn's theories regarding Eucratides. He rejects the idea that the king is of Seleucid decent, that he is related to Antiochus IV, and that he saw himself as a soter figure (1957:54-57). Tarn's understanding of the king and his reading of his coins are hypothetical and open to criticism.

215

rule. 42

The final South Asian king to issue Dioscuri coins is the Scythian king Azilises. Azilises follows

Azes I, whose era began in 57 BCE (Errington 1992:14), and who ruled over Gandhära, Arachosia and the

western Punjab (Rosenfield 1993:127). The issues of Azilises depict the Dioscuri both as mounted and

standing figures as on the Diomedes coin. As noted above, he also produces a single Dioscuri coin. The

standing Dioscuri coins of Azilises are illustrated in figure 67. These coins depict unusual images of the

Dioscuri. Their headdress, which are usually considered to be one of their defining features, is particularly

altered. Their usual pilei head gear has been replaced with bent cone style caps, which Gardner thinks

represent Phrygian caps (1966:93). In one issue they also make contact with each other with their free hands

as in some of the Huviska Skanda-Kumara with Visakha issues. The single standing Dioscuri coins are,

however, unusual. With these coins we must keep in mind that we might not really be looking at the

Dioscuri, but another set of gods presented in the guise of the Dioscuri. Certainly, the single Dioscuri figure

represents a willingness to manipulate these figures to meet the needs of Azilises. These coins are present in

the Gandhära area not long before the Kusanas enter the area. We have already seen that die makers made

use of coin types of much earlier kings.43 Thus, the presence of these coins allows for their use as

numismatic models for the Huviska Skanda-Kārttikeya type.

The Dioscuri also appear in a temple site near Dil'berdž in on the northern fringe of the Bactrian

oasis. This temple was founded by the Greeks and enlarged by the Kusanas and Sasanians (Boyce 1975

III:172). The temple has been named "Temple of the Dioscuri" by some scholars because the only well

preserved detail of the temple is a painting in the porch of the Dioscuri standing on either side of the

42One of his coins has been found as far south as Sonipat with other Indo-Greek coins (Narain 1957:89). It seems possible that his coin types were known over most of what would become the Kusāna empire. Nonetheless, whether or not the Kusana minters of Huviska would have known of the coin type is only conjecture on my part.

43Specifically, we have seen that Huviska's obverse die type which depicts him sitting with a bird standard is borrowed from issues of Kujula Kadphises, Maues and Azis II.

216

entrance (Boyce 1975 III:173). The argument that this temple was dedicated to the Dioscuri has been

questioned. The Dioscuri often act as guardians of doorways; this may have been their function at this

particular temple (Boyce 1975 III:173). What the temple does suggest, be it dedicated to the Dioscuri or

not, is that the Dioscuri had a presence in Bactria, and that the Kusanas were aware of them and did not

remove them from the temple when they enlarged it.

The Dioscuri may also be found on reliefs from Butkara in the Swat Valley. G. Gnoli has

illustrated a relief from Butkara which depicts four panels, which I have already partially discussed. He

argues that the second panel depicts the Dioscuri. He also claims that three other reliefs from the site also

depict the same twin deities (Gnoli 1963:31). These panels are illustrated in figures 68 and 69. As on one of

the Huviska Skanda-Kumara with Visākha coins (figure 59), the panel in figure 68 depicts the deities hand

in hand. Gnoli goes on to argue that these representations of the Dioscuri "are evidently barbarized", and

that they must come after the Greek period, probably from the Saka-Pahlava period (1963:35). All of these

Swat Valley depictions of the Dioscuri, however, show them without their usual attributes of horses, palm

branches or pilei, though their headdress is not clear in most of the statues. They are youths and carry

spears, which does link them to the cult of the Dioscuri, but the panels in question are damaged. The

condition of the panels and the absence of these other attributes make for an uncertain identification.44 This

uncertainty along with the coins of Azilises suggests that the form of the Dioscuri was malleable in the eyes

of Scythian leaders and may have been used as the iconographic model for other paired deities.

The Indo-Greek and Saka coins and possibly these reliefs strongly suggest that Dioscuri or

Dioscuri-like images were known to the general populace of Gandhara, north-west Pakistan and the western

Punjab, and that minters would have been familiar with the standing Dioscuri type not long before the

44 A. Hermary in the Lexicon Iconographicum Mythologiae Classicae agrees with Gnoli's identification (1986 III:594). MaurizioTaddei is not as sure of the identification saying: "In the fourth panel are two standing youths (Dioscuri?)", and he notes "Both figures (Dioscuri?) are badly damaged, but the characteristic clothes are clearly visible" (1964:107). Later, however, Taddei comments that the relief may

detail (1966:85). be related to the Aglibol-Malakbel couple on Palmyran reliefs, but he does not explore this suggestion in

217

entrance of the Kusanas in the region. These factors heighten the likelihood that these Dioscuri coin types

could have been used as models for the Skanda-Kumāra with Visakha type.45

I would like to argue further that the choice of the Dioscuri as a model for Skanda-Kumāra and

Visākha coins was motivated by certain features these two pairs shared. An examination of the Greek and

Roman cults of these Hellenistic deities indicates that they shared some important elements with Skanda-

Kärttikeya's cult, and lends credence to the supposition that the Kusānas deliberately blended Skanda-

Kumāra and Visākha with the Dioscuri.

In Greek religion the Dioscuri are divine twins called Castor and Polydeuces (Pollux in Latin

sources). They are the sons of Zeus and brothers of Helen. The twins are always described as youthful

warrior heroes (Farnell 1995:187-188). In legendary accounts, the Twins often originate in Sparta, where

their cult was particularly strong. The Dioscuri were connected with the Spartan double kingdom; images of

them were carried by the Spartan army into battle (Poulsen 1991:140). Under Roman rule the cult remained

45In relation to this point, I have already noted that Göbl suggests that some Kusāna coin patterns were influenced by Roman coin types, and that the Kusanas may have employed die cutters and minters from the Roman Imperial mint at Alexandria. He argues that the main trade route from India to the Mediterranean went through Alexandria, and that these Kusana coins based on Roman patterns were designed to facilitate that trade (Göbl 1960:77-78). Göbl's argument is accepted by most numismatic scholars. I think Göbl's argument is supported in three ways: the similarity between certain Roman and Kusāna coin types, the documented contact between the two cultures, and a hypothesised pragmatic need for particular coin types for trade. These points also hold true in the specific case of the Dioscuri and Skanda-Kumära with Visākha coins. The standing Dioscuri type and the Skanda-Kumāra with Visākha type are similar. There is clear contact between the Kusanas and Indo-Greek and Saka cultures, who issue Dioscuri coinage in Bactria, Gandhära, north-west Pakistan and India. From a pragmatic stand point, one of the main trade routes in Ancient India began in the Gangetic valley and went north through Pakistan and Bactria and on westward. The resemblance between the Dioscuri type and Skanda-Kumāra with Viśākha type may have assisted trade between Indian groups and groups in what is now Afghanistan by providing coinage with familiar motifs for both groups. Finally, it seems more likely that the Kusanas would have employed die carvers from the Bactrian and Gandhäran area whose forerunners had produced remarkable and artistic coinage for the Indo-Greek, Indo-Parthian and Saka rulers, than import skilled labourers from Alexandria. Where Roman influence is seen on Kusāna coinage, Kusāna minters may simply have copied from actual Roman coin samples. It seems likely to me that the same artistic lineages or guilds that produced the earlier Dioscuri coins for Greek and Saka rulers also produced these Skanda-Kārttikeya issues for their later Kusäna masters. Whatever the case, the same logic that allows Göbl to link Roman issues with Kusāna issues can also allow us to link Indo-Greek and Saka issues with Kusāna issues, specifically the Dioscuri and Skanda-Kumāra with Visākha coins.

218

strong among elite Spartan society, with several wealthy families inheriting the priesthood for the Dioscuri

for generations (Poulsen 1991:136). The Roman cult for the Twins was also primarily focused on their

youthful military aspect. They were regarded as the tutelary gods of the Roman equestrian order, and in

various legends they appear in battles to support the Roman army (Poulsen 1992:48-52). In my chapter on

textual sources for Skanda-Kärttikeya I have demonstrated that one of the deity's primary attributes is

youth; he is a kumara. Especially in Epic and Puranic sources, Skanda-Kārttikeya and Visākha are also

warrior figures.

The Dioscuri are also regarded as soters or saviours in both Greek and Latin traditions. The twins

are regarded as companions of those in danger especially in battle and at sea (Burkert 1985:212-213). In the

Buddhist Mahavastu Skanda-Kärttikeya is one of the gods people call on when shipwreck threatens (Jones

1956 III:71).46 Skanda-Kārttikeya can also be viewed as a saviour figure in battle. As I have demonstrated

in my section on textual sources, Skanda-Kärttikeya's primary role becomes to save the gods from the asura

Täraka as the general of the army of the gods.

The Twins also have strong links to cosmological or astral phenomenon. By the fourth century

BCE they are commonly depicted with stars over their heads and are identified with the constellation known

as the Twins or Gemini (Farnell 1995:186). As I will demonstrate later, Skanda-Kārttikeya is associated

with Mars in the Yavanajataka of Sphujidhvaja. He is also identified with the pole star in the

Kāśyapasamhitā (Wujastyk 1998:217) and with the astral deity Srausa in the Sāmba Purāna. The

Yavanajātaka and Sāmba Purāna are clearly influenced by Greek and Iranian sources; indeed the

Yavanajātaka is a Sanskrit translation of a Greek text. Further connections between Mars and Skanda-

Kärttikeya in Indo-Greek sources will become apparent further on in this section.47 I will also demonstrate

461 am grateful to Phyllis Granoff for pointing this out to me.

47There may have also been a link between the Dioscuri and Mars as military deities within the Roman world. Both the temple of Mars and the Dioscuri temple were involved in the Roman cavalry parade celebrated each July 15 (Scullard 1981:164-165). During the time of Augustus the procession began at the temple of Mars in Rome and went past the temple of Castor and Pollux, where the cavalry was reviewed by

219

that Skanda-Kärttikeya is closely associated with other astral deities in textual sources. These associations

are not explained within the texts and likely come from external sources. Finally, Visākha can also mean

"born under the constellation Viśākhā" (Monier-Williams 1999:953). Hence, Visākha's name seems to

relate him with stars.

There are also some links between the Dioscuri and healing cults. In an inscription at Akrai in

Sicily they are connected to a child-birth goddess. There is also some evidence of them as healers at

Constantinople (Farnell 1995:225-226). The same is true of the Roman cult, where the Dioscuri temple at

the Roman Forum was also associated with healing, though, this association may come as late as 166 CE

(Poulsen 1992:52). Healing appears, however, to have been a minor aspect of the Twin's cult. As already

demonstrated, Skanda-Karttikeya and Visākha have very strong associations with graha cults.

The Roman cult for the Dioscuri is also important as an example of how one Imperial group of

rulers used the Twins in propaganda campaigns.48 Even in Republican Rome the cult is closely linked to

military figures and politics. Their temple in the Roman Forum was regarded as the symbol of successful

military campaigns in Republican Rome (Poulsen 1992:49), and they were closely linked to the Roman

Equites (Scullard 1981:65). Their temple was also used in Republican Rome for public meetings and votes;

it was also the scene of a number of political disturbances (Scullard 1981:66-68).49 The first Roman

denarius issued in 211 BCE depicts the twins on horseback on the reverse and may have been a symbolic

Roman officials (Scullard 1981:164-165)

48A complete discussion of this topic is provided by Birte Poulsen in her paper, "The Dioscuri and Ruler Ideology" (1991).

49The legend behind the founding of the Castor and Pollux temple in the Roman Forum is linked with the battle at Lake Regillus in the fifth century BCE between the Romans and the Latins. Apparently, the Twins appeared on the battlefield as youths on horses and fought with the Romans. Later that day they reportedly appeared at the site of their future temple in Rome to report the victory (Poulsen 1992:46-47). Another part of this legend has it that when the Twins touched the first person to whom they report the victory in Rome, his beard turned red (Poulsen 1992:46). One of the links between Mars and Skanda- Kärttikeya is that they are both red. The exact meaning of the Dioscuri story concerning this red beard is not clear to me, but if this colour is one of their attributes it provides another link between their cult and the cult of Skanda-Kārttikeya.

220

representation of the expanding power of Rome (Poulsen 1992:49). Many of the inscriptions and

dedications to the gods come from Roman generals (Poulsen 1992:49). B. Poulsen argues that in the

Imperial age the emperors Augustus, Tiberius, Hadrian and Antoninus Pius used the Twins extensively in

statuary, coins and inscriptions to promote themselves and their potential successors. Rulers and heirs

apparent were depicted as parallel to the divine Twins on a number of medallions and coins from the age

(Poulsen 1991:119-126, 133-136; 1992:51-52). The appearance of the Dioscuri on the coins of Augustus,

Hadrian and Antoninus Pius is especially important, because it is these issues that Göbl thinks were current

during the Kusana period and available to minters as prototypes (1960:88). The cult of the Dioscuri in

Rome suggests that there were strong political reasons for placing a military deity on a coin. Such military

figures seem to have been used by the military and ruling elite to solidify their own positions.50 I have

argued for similar political motivations behind the Mahasena issues of Huviska.

It is often assumed by those who study Kusana coinage that beyond their obvious financial purpose

coins have little other roles to play, except to suggest the religious devotion of a ruler. Those who study

Roman and Hellenistic coinage have not limited their conclusions to this narrow context. As the comments

above demonstrate, coinage and inscriptions were viewed as valuable political tools in the ancient world.

Coins provided a means for rulers to send their subjects a message about themselves or their realm. It would

have been odd indeed if this propaganda aspect of coinage was not noticed by the Kusäna kings. The

Kusāna kings borrowed Imperial and royal titles from almost every culture with which they came into

contact, including Rome, and placed these titles on their coins and epigraphs (Rowland 1993:vii).51 They

also copied ideas of what an emperor was from other cultures and assimilated these ideas into their own

50Indeed, we have minimal evidence for widespread or popular worship of these deities in Republican or Imperial Rome outside of military or political circles. This may have led H. H. Scullard to comment: "Thus the temple of Castor and Pollux often formed the centre of stirring events, but how well it was attended by worshipers on each 27 January [a festival day for these gods], we do not know" (1981:68).

51For additional points on Kusäna contact with the Roman world, and their use of Roman ideas see Benjamin Rowland's forward to Rosenfield (1993:vii-xvi).

221

rule, and expressing these ideals through monuments and coinage. As I have demonstrated and as we shall

continue to see, the use of coinage as propaganda is a concept the Kusanas understood and used.

Thus, there may be more than simply borrowing of an iconographic coin type between the Dioscuri

and the Skanda-Kumāra with Visakha coins. The gods, a least in a general way, share some close affinities

with each other. Both groups are young warriors who are associated with stars and healing cults. Kusāna

minters may have created the Skanda-Kumara with Visakha coin with these affinities in mind, and Huviska

may have recognized the political value of placing these Dioscuri-like deities on his coinage.

The implications of such a blending are significant. While the Dioscuri have a role in healing cults,

it is a minor one, and their appearance on Roman, Indo-Greek and Saka coinage is focused on their military

character. An implicit link is also made between these deities and the rulers who placed them on the reverse

of their coins. By analogy, what the Huviska coins suggest is that the indigenous cultic background for

Skanda-Kumara and Visākha as Grahas is being replaced by an emphasis on the gods as warriors who

support a royal figure, and that this process may have been facilitated by a deliberate assimilation of the two

Indian deities to the Dioscuri figures from Indo-Greek and Śaka coins.

5.2.7 The Skanda-Kumāra with Visākha Coins of Huviska: Concluding Remarks

Non-numismatic archaeological evidence is of little help in determining the origin of these images.

I know of no surviving statues or plaques from which these icons were copied. They appear to come

completely from other numismatic sources. Other information we can gather from these coins is scanty.

Both deities appear to be dressed as Indians, though Skanda-Kumara's skull cap is not distinctively Indian.

Both figures are armed, but are not dressed as Kusäna or foreign warriors. Neither bears any resemblance to

the depictions of Kusana kings on coins or to depictions of Scythian warriors found in Kusana art. They do

not, however, closely resemble the images of Skanda-Kārttikeya and Visākha found on panels from

Mathurā. The impression we receive from these coins is of a group of designers or rulers who understood

these deities to be paired in a close manner and understood that they were of South Asian origin. What they

did not seem to be concerned with is the indigenous understanding of these deities as Grahas.

222

I would conclude by suggesting that the representation of Skanda-Kärttikeya is greatly softened

from that found on the Yaudheya six-headed coins. Earlier, I suggested that the Yaudheya six-headed issues

reflected cults of inauspicious gods who were controlled through ritual. These coins of Huviska remove

both the healing and inauspicious aspect of these deities. We have seen a similar removal in the

Mahabharata. It is possible that the single-headed Yaudheya class three coins are evidence of a similar and

earlier shift in the understanding of the character of this deity. I would venture to argue that the softening of

Skanda-Kärttikeya's appearance begins with the Yaudheyas under foreign influence and reaches its fruition

under the Kusānas.

5.2.8 The Skanda-Kumara, Visakha and Mahasena coins of Huviska: A Description

The final Huviska gold coin which depicts Mahāsena also depicts Skanda-Kumāra and Viśākha on

the same coin. It is illustrated in figure 70. Mahäsena stands in the centre of the coin facing the front. He has

a large nimbus, and his turban or top knot from the previous coin appears to have been replaced by a crown.

He has both of his hands on his hips and appears to carry a sword on his left hip. Due to wear on the coin it

is difficult to tell if he wears a dhoti, or has changed into a tunic, which would be more indicative of a

Scythian warrior. A Bactrian inscription identifies him as Mahāsena. Visākha stands to Mahāsena's proper

left. His body and head are turned towards Mahasena. He is nimbate, but his nimbus is smaller than

Mahasena's. His left hand holds a spear, while his right hand is on his hip. The Bactrian inscription

identifies him as Visākha. To Mahāsena's proper right is Skanda-Kumāra. Skanda-Kumāra's body and head

also face towards Mahasena. His nimbus is also smaller than the central god's. Both of his hands are on his

hips, and he has a sword on his left hip. The Bactrian inscription identifies him as Skanda-Kumāra. The

three deities stand on a two layered decorated platform and are enclosed by what may be a structure of some

sort.

A second version of this coin exists and is illustrated in figure 71. This coin presents the same

three figures. Here Mahasena holds a spear in his right hand and may wear a turban. Visakha appears to be

essentially the same. Skanda-Kumära holds a spear or staff in his right hand. The inscriptions are arranged

223

differently around the coin, and the structure looks much more like a building or shrine with pillars topped

with fillets with a roof-like structure overhead.

5.2.9 The Rise of Mahasena

Göbl has demonstrated that this coin may also be inspired by Roman prototypes (1984:174). The

appearance of three deities on one coin and the appearance of a shrine-like structure are unique elements for

the Kusana numismatic catalogue. Much has been made of the appearance of the shrine on these coins, but I

will not discuss this element here. My immediate purpose here is to suggest that these coins do not present

an image of three equals, but elevate Mahasena above the other deities. While concrete conclusions are

always hard to draw from the limited evidence on coins, I do think that the central position of Mahasena on

the coins and the image of the other two deities facing towards him imply his superiority. His crown on the

first coin and his larger nimbus on both coins also suggest to me an attempt to elevate his status above the

other two gods depicted on the coin. A Kusna preference for Mahasena is borne out by examples of statues

of this god from the north-west.

The potential implications of representing Mahäsena on these coins becomes, I think, clearer once

we consider the argument I have been making here that the focus of the Kusāna Skanda-Kārttikeya cult was

elitist and centered on the military aspect of the deity at the expense of his inauspicious forms associated

with Grahas and Mātrs. Mahāsena provided Huviska with a valuable tool for royal propaganda. His

placement on the coins of Huviska lent the support of a divine general to the king. He was elevated above

Skanda-Kumāra and Visākha on the coin depicting all three deities because this purely military figure was

of greater political use for the king. These coins provide us with a specific example of J. Rosenfield's

general hypothesis regarding why certain deities appear on Kusana coinage. He comments: "In general, I

shall develop the hypothesis in this and succeeding chapters that most of these deities were the Kushan

comes augustii -- the divine companions and supporters of the monarchy ... " (1993:70).

The context of the Rabatak inscription is also instructive in this matter. The site appears to be a

royal shrine, which has been combined with a divine sanctuary similar to those found at Surkh Kotal and

224

Mat (Cribb and Sims-Williams 1995/96:109). Regarding the divine images referred to in the inscription

Cribb states "the Rabatak inscription shows that they are represented as the source of Kushan kingship"

(1995/96:110). The combination of a royal family shine and a divine shine suggests such a conclusion.

Mahasena's and Visakha's inclusion on this inscription, even though an after-thought, suggests their

participation in a royal and elite cult. The other material I have presented in combination with this

inscription all points towards a Kusäna interest and use of Skanda-Kārttikeya as part of elite and royal cults

based within Iranian and Hellenized traditions. As I have already suggested, it is this elite and martial

characterization of the deity that significantly removes him from his initial base of popular worship. It is this

Kusāna use of Skanda-Kārttikeya and the concomitant adoption of martial devices for the god that brings

about significant change in the characterization of the deity and the locus of his 'popularity' in the north of

India.

What I think my study of Huviska's Skanda-Kärttikeya coin types demonstrates is that the

transformation we have already seen in textual sources, where the deity is transformed from primarily a

Graha-like deity to a military deity, can be located in a definite political and religious context during the

Kusāna era. What I have attempted to do here is to place these coins into a context that recognizes their

financial value and their potential religious and political symbolism; in so doing I have also considered the

backdrop of archaeological finds relating to the period. Through such an approach I have demonstrated that

within official Kusäna circles, particularly those of the north-west, the cult of Skanda-Kärttikeya was

focused on the image of Mahäsena, and that this focus was likely driven by political concerns. The message

these icons was meant to deliver was largely directed to an elite section of Indian society. The presentation

of this deity likely had less to do with religious motivations than issues pertaining to the symbols of foreign

royal power. The image of Mahāsena seems particularly to make this point. Mahāsena does not appear to

have ever been widely popular. This deity's appeal to Huvishka as a divine support for the king's army

does, however, have important political implications and implications for the future 'popularity' of Skanda-

Kārttikeya.

225

5.2.10 Skanda-Kārttikeya and Mars in Kusāna Era Material

An aspect of Skanda-Kärttikeya's character to which I have only alluded is related to astrology and

astrological deities. Some Skanda-Kārttikeya scholars have argued that the only way to explain his

association with the Krttikas, a naksatra or constellation, is to suggest that his own background is as an

astrological deity. Indeed, most scholars who subscribe to the 'sum of various parts' theory of Skanda-

Kārttikeya's origins usually regard Karttikeya as an originally separate astrological deity who is

amalgamated into Skanda-Kārttikeya's cult (P. K. Agrawala 1967:xv; Thakur 1981:10-11).52 The weakness

I see in this line of thinking is that no text or inscription ever actually claims that Karttikeya is or was an

astrological deity. That the name Kārttikeya is likely derived from the name Krttikā seems likely, but this in

and of itself does not make Kärttikeya an astrological deity. The involvement of the Krttikās in the birth

narrative is actually better explained through their connection with Agni found in the Satapatha Brahmana

than as an attempt to link Skanda-Kärttikeya with some astrological origins.53 In short, I do not find solid

evidence to support the idea that Kärttikeya is an astrological deity. It is just as possible to suggest that the

name came about as a means of justifying the inclusion of the Krttikäs in the birth story of Skanda-

Kärttikeya. While I do not agree with those scholars who see an astrological deity in the name Kārttikeya, I

would like to suggest that the strongest evidence for a link between Skanda-Kārttikeya and astrology is a

Kusāna and Scythian view that he was, or was like, Mars/Aries. To date only one scholar has recognized

this link between Skanda-Kärttikeya and Mars,54 and no scholar has recognized the involvement of foreign

powers in bringing it about, so I will briefly explore the evidence that supports my hypothesis here.

52There are also scholars who trace the astrological origins of the deity to the Indus Valley Civilization. I have already discussed these arguments in the introduction to this dissertation and found them lacking.

53I have already cited the section from the Satapatha Brähmana that links Agni and the Krttikās in my chapter on the ranyakaparvan.

54P. Pal (1969) mentions the possibility that Skanda-Kärttikeya's iconography may have been influenced by that of Mars. Pal's points will be given in more detail below.

226

Astrology has ancient roots in India, and tracking the progress of the stars can be traced to Vedic

India. The Vedic study of naksatras was, however, largely based on keeping ritual time. Vedic astrologers

were primarily concerned with the movements of the moon and stars so that they knew when to begin

certain sacrificial rites. There is only a limited sense within the Vedic tradition that these naksatras were

deities worthy of independent worship. The Vedic people were also not interested in the movements of the

planets (Kaye 1998:33).

There is, however, a significant shift in astrological texts in India between the second century CE

and the seventh century CE. Between approximately 200 BCE and 200 CE astrological texts called

Vedangas are produced which largely follow the astrological traditions of the Vedas (Kaye 1998:7-8); after

this period a remarkable change in the tradition occurs:

There is a very marked differentiation between the works of the type of the Jyotish Vedanga on the one hand, and those of the type of the Surya Siddhanta on the other, and this differentiation is not merely one of time -- it is so fundamental that continuity of development appears to be altogether out of the question. Somewhere between the Vedanga period and the period of the composition of the original Surya Siddhanta a distinct break occurred -- the old methods and rules were discarded and new method were introduced and new phenomena treated. (Kaye 1998:7)

Astrological Siddhanta texts emerge at around the seventh to eighth century CE. The shift in method and

subject mentioned above occurs between 200 and 500 CE (Kaye 1998:7). During this period Greek

astrology and astronomy enter India; also during this time the planets and their movements become a topic

of study. The study of Mars/Aries is not something that the indigenous Indian tradition develops

independently, but is borrowed from Greek sources.

Textual references that link Skanda-Kärttikeya and Mars/Aries are first found in the Yavanajätaka

of Sphujidhvaja, or The Greek (or foreign) Horiscopy of Sphujidhvaja. This text is also our earliest record

of a Sanskrit translation of a Greek text on astronomy (Pingree 1964/65:250). The text itself gives us its

history describing its initial translation from Greek into Sanskrit by Yavaneśvara (Lord of the Greeks) in

149/150 CE, and then its rendering into verse form by Sphujidhvara in 269/270 CE (Pingree 1964/65:250;

YJ 79:60-62). It is Sphujidhvara's text that survives today. Other than their brief mention in this text,

227

however, history does not record either of these figures, but on the basis of their names, they are likely

foreign rulers or sub-kings, perhaps belonging to the Western Ksatrapas (Pingree 1978 I:3-4). While we

must exercise some caution in accepting this text's own version of its history, the first appearance of this

text in 149/150 places it in the reign of Huviska, whose reign, as we have seen, has a dramatic impact on the

cult of Skanda-Kärttikeya. This text may be another example of how foreign powers in India perceived

Skanda-Karttikeya. This text may also have played a role in Kusāna perceptions of the deity.

The Yavanajätaka links Skanda-Kärttikeya with Ares/Mars in two ways. First, it characterizes

them in a similar fashion and even points out this similarity. Second, it describes Skanda-Kārttikeya as the

divine ruler of the planet. I will begin by reviewing their similar characterizations. The description of Mars

fits best with the description of the inasupicious Skanda-Kārttikeya:

Mars is a hot and passionate man with flaming curly hair and a terrible red body. The corners of his eyes are bloodshot, and he shines like blazing fire; he is powerful in his vehemence and terrifying like Kumära .... He is a hero, used to killing, taking and opposing; clothed in red, he commits acts of violence and strength. (YJ 1.133-134 translated by Pingree 1978 II:11)

Certainly, the physical description of the two figures is similar: they are both red, associated with fire,

described as bright or shining, clothed in red, quick to anger, violent, strong and heroic, and the text seems

to be well aware of the similarity and encourages the link between the two figures. The analogy only works,

however, if we accept the image of Skanda-Kärttikeya as depicted in the early chapters of the

Äranyakaparvan and not the later domesticated figure. The link between Mars/Aries and Skanda-Kärttikeya

is between two powerful and feared deities.

The text has a number of other links between the physical appearance and characterization of the

two figures. In describing the first Hora of Aries55 we are told the following:

[he] wears red clothes and is flaming like the sun at Doomsday. He holds a sword and a firebrand in his hands. His hair is tawny and sticks up, and his ear-rings are of gold. He is a fierce man who

55There are 24 Horas in this text. A Hora is associated with the rising of a sign and the decline of the sign, hence each sign has two Horas. One Hora rules over 15' of the zodiac and each has a lord. In the case above the first lord of the first Hora of Aries is Mars; the conjunction of the two produces this creature or Hora described in the quotation (Pingree 1978 II:209).

228

has raised the staff of Death for the sake of protection. This is a man shaped creature whose cry is loud and who has a long, thin face. Standing in the midst of flocks of goats and sheep, and mounted on a goat, he rules his host. (YJ 2.2-3 translated by Pingree 1978 II:11-12)

This Hora is something of the product of both Aries and Mars, but it is clearly influenced by the

characterization of Mars and may be considered a form of the planet/deity. The above description of the

Hora is not unlike that of Visakha and Skanda-Karttikeya from the Aranyakaparvan. Visākha is described

in that text as having the glow of the Doomsday fire (3.216.15), as wearing earrings and as wearing gold

(3.216.13-14). The loud cry, as well as the fierce and protective character of the Hora are also reminicent of

Skanda-Kärttikeya's depiction from that text. The association with sheep and goats also fits with Skanda-

Kärttikeya from the Epic. The Hora is also described as the ruler of a host in the above quotation, and this

also relates to Skanda-Kārttikeya's general description. A similar description is made for the first Drekāna

of Aries:56 " ... a man garbed in red and having a red complexion, a fierce man whose limbs and hands are

wounded and who attacks in anger. He bears golden mail and bright arrows, and his hand is upraised with

an axe" (YJ 3.2 translated Pingree 1978 II:15). Not everything about these astrological figures is linked to

Skanda-Kärttikeya, but there are the repeated themes of a red colour, golden adornments (in this case

golden armour which is linked to Skanda-Kärttikeya Mbh 3.218.1) and violence.

Another common topic for this text is predicting the character of people based on the conjunction

of a planet with a sign of the zodiac. As we might expect, the appearance of Mars in a particular sign

produces a certain type of person: "One should know that a man born under sunapha of Mars is fierce and

strong, often engaging in battle and thievery and devoted to riot and enmity, the destructive but rich leader

of an army or tribe" (YJ 10.11 translated by Pingree 1978 II:33).57 All of these aspects of people born under

the influence of Mars also relate to aspects of Skanda-Kärttikeya's character. Mars is also viewed as a lord

56There are 36 Drekanas in this system, three for each sign. As with the Horās, Drekanas produce a being that is influenced by the lord of the particular Drekana in question. The lord of the first Drekāna of Aries is Mars, which produces the being described above (Pingree 1978 II:209).

57 Similar comments may be found at YJ 15.1-2, 20.6, 33.1, 70.2.

229

of armies (YJ 4.32) and as a general (YJ 1.116). While Mars, particularly under the influence of Aries, is

associated with wealth, gold, and military leadership, it is clearly understood as an inauspicious planet.

Planets can also be the lord of a year. The description of a year lorded over by Mars demonstrates the

inauspicious nature of this planet/deity: "A year of Mars involves kings fierce in battle; a scarcity of grain;

dried up and waterless trees, flowers and shrubs; many snakes and fires; it is ruined by diseases, thieves,

hunger, and misfortunes" (YJ 78.13 translated by Pingree 1978 II:186). Like the early version of Skanda-

Kärttikeya, Mars is something to be feared.

A final point to make before discussing Skanda-Karttikeya as the regent of Mars is the association

between this text and childbirth and disease. This text suggests that the conception and birth of a child and

the safety of the mother were all influenced by the planets. Mars along with the Sun, Venus and the Moon

all produced a strong fetus if they were in their own vargas during intercourse. If these planets were in a

malignant sign during intercourse, the semen would die. The same was true for the survival of the fetus and

mother at birth. If one of the above planets was in a bad sign, the result was viewed as disastrous for both

the birth of the child and survival of the mother (YJ 5-6). The causes of disease were also traced back to the

influence of planets on people, but it appears that all of the planets could cause disease if in the right sign,

and the ability of Mars to cause illness is not unique. Thus, Skanda-Kārttikeya and Mars do share some

aspects of their characters which relate to child birth and disease, but these links appear to be indirect. The

text primarily associates the two through their violent and martial characteristics.58

The final aspect of the Yavanajätaka I wish to mention is that the text regards Skanda-Kārttikeya

58The Sanskrit term for these planets is graha. As we have already seen, this is also the name given to child snatching disease demons. Graha means grasper and is used in connection with planets because they grasp people and determine their fate, and it is used for disease demons because they grasp children and pregnant women. That the two types of grahas might be related is a tempting interpretation, but it is one that Wujastyk rejects outright (1999:259). I am not as convinced as Wujastyk on this point. Several prominent disease demons share their proper names with naksatras and stars. I do not think that this can be reduced to a coincidence. While Wujastyk is correct to point out that there is no direct link between the two groups of grahas, there does appear to be some assumed relationship between the two groups, even if that relationship begins as an unrelated sharing of the name graha. Certainly, the status of Skanda as a Graha and Mars as a graha, and their similar characterizations suggest that the two catagories can be blurred.

230

as the regent or deity of the planet Mars. We learn of this when the text suggests performing sacrifices to the

gods of the planets which it lists as: "Jala (Water), Vahni (Fire), Visnu, Prajāpati, Skanda, Mahendra, and

Devī -- in signs which belong respectively to the Moon, the Sun, Mercury, Saturn, Mars, Jupiter and Venus"

(YJ 77.1 translated by Pingree 1978 II:183). Here Skanda is the god of Mars. This passage does not suggest

that the planet Mars and Skanda-Kärttikeya are the same entity, but it does make a relationship between the

two to be clear. Pingree argues that the names of the Indian gods/goddesses associated with each planet are

based on the characters of the Greek gods whose names are given to these planets (Pingree 1978 II:404). In

other words, 'Skanda' is chosen as the god of Mars because he reminded the Yavanajataka authors of the

Greek martial deity Ares or the Roman Mars.

The characterization of Ares/Mars in the Yavanajätaka is in keeping with Greek views on Ares. In

Greek mythology Ares is presented as the god of war, but he is a brutal and ferocious figure who delights in

destruction. He is also implicated in various illicit affairs and is described as not well liked by other deities.

Ares was not a major figure for the Greeks, who preferred Athena, the armed goddess of wisdom, as their

ideal of military prowess. The Romans, however, did elevate Mars as one of the chief gods within their

pantheon praising his miltary ability and glory. Little of this sense of Mars as an auspcious deity is evoked

in this text.

The Yavanajataka is, then, important for two reasons. The first is that it helps us to understand how

Hellenized Scythian and possibly Kusāna groups understood Skanda-Kārttikeya. He was like Ares, a red,

fiery military figure who was known for his brutality and violence. As such he is closely associated with

astrological deities. The second is that this text helps to bridge the gap between Skandagraha, the leader of

the Grahas, and Skanda-Karttikeya, the general of the army of the gods. As we have seen in textual

accounts, the characterization of these two figures is remarkably different, so much so that by the

Suśrutasamhitā any identification between the two figures is denied. Just what allowed for, or initiated the

dramatic shift in characterization from inauspicious Graha to auspicious Senapati is not clear in the

mythology of the deity. In other words, just how the figure of the Graha was linked to the figure of the

231

Senapati is not clear. The Yavanajataka, however, helps us to see that the ground shared by Skandagraha

and the martial Ares was their brutal character. Hence, the beginning of Skanda-Kārttikeya as a military

figure may be located in the resemblence Hellenized rulers in India saw between the two violent figures.

While the violent and dangerous characterization of Skanda-Kārttikeya is steadily removed from his

depictions in Sanskrit sources, his role as a military figure endures. What the Yavanajātaka helps us to

understand is that the element of continuity between the roles is, I think, their initial inauspicious and

violent nature.

The association between Skanda-Kärttikeya and Ares/Mars does not endure, however, in these

texts. Later Indian astrology texts will not copy the Yavanajataka's understanding of Mars as Skanda-

Kārttikeya-like. Even though Mars remains an inauspicious planet, red and associated with violence,

Skanda-Kārttikeya does not. The Epic texts and later Purānic traditions do not represent Skanda-Kārttikeya

as like Mars. The dominent characterization of Skanda-Kārttikeya becomes the auspicious martial son of

Siva, and Mars does not play a role in this characterization. The link between the two deities in textual

sources is short lived.

The association between Skanda-Kärttikeya and Ares/Mars is not just an anomaly found in the

Yavanajātaka, however. Kusāna era seals and sealings from the far north-west (Gandhāra and Bactria) also

attest to a link between Mars and Skanda-Kärttikeya forged by Hellenized groups in the region. Figure 76

illustrates a seal and its sealing from Kusana era Gandhāra (Callieri 1997:191). The seal depicts a standing

male deity holding a spear in his right hand and a shield in his left. His dress is foreign, and he may wear a

Roman-style tunic and breastplate. Standing on the shield is a cock. Callieri has tentitively identified this

figure as "Kārttikeya" (1997:191). I think Callieri is right; the figure is likely a representation of Skanda-

Kärttikeya, but of more importance is Callieri's suggestion that the figure on the Kusana seal is borrowed

from Roman depictions of Mars (1997:191). The only element that is different between the two seal styles

is the addition of the cock on the Kusana version. Callieri also illustrates a number of other seals which

likely depict Skanda-Kārttikeya. Figure 77 depicts another example which shows Skanda-Kārttikeya in a

232

similar posture as that found in figure 76, though this figure has a more rigid stance. His hair is arranged

into a bun with loose locks hanging down. He is dressed in a tunic, breastplate and high boots. In his right

hand he holds a spear and in his left a cock which faces him. The Kharosti inscription reads:

"Pavratigasa' or "'(seal) of Parvatika" (Callieri 1997:106). Callieri also notes that this type of seal is

typical of north-western Kusāna era seals from the Gandhara region which depict tutelary deities

(1997:106).59 The point Callieri takes from these seals is that the Gandharan seals were the work of Roman

craftsmen living in the north-west of India. I am not certain about this point, but the relationship between

Mars and Skanda-Kärttikeya is certainly well demonstrated by these seals. The Yavanajātaka is not an

anomaly, but likely a representation of common Hellenized perceptions of Skanda-Kārttikeya.

That these seals, which were likely produced for officials in the north-west of the Kusāna empire,

represent specifically Mars and not Ares is also significant. Mars is a more appropriate official image

because the Roman deity was more closely associated with state politics and symbols than Ares. The Roman

Mars is closer in characterization to Skanda-Karttikeya as Mahasena in that they are both auspicious martial

figures who played a role in state/elite driven cults. While Ares and Skandagraha were linked and this may

have provided the initial connection between the cults, Mars and Skanda-Mahäsena were also connected. It

is likely to the figure of the Roman Mars that the Imperial Kusanas drew upon when they encountered

Skanda-Kärttikeya. As with the examples from north-west sculpture I discussed earlier, these seals all depict

Skanda-Kärttikeya as a martial figure, in foreign armour with a cock. They all also come from a blended

Hellenized and Parthian context. Once again the evidence from the art of the north-west demonstrates that

the initial characterization of Skanda-Kärttikeya as martial comes from this area and from foreign groups.

The association between Skanda-Kärttikeya and Mars, or the Sanskrit Mangala, does not die out

completely with the fall of the Kusana empire. Sculptural images of Mangala resemble Skanda-Kārttikeya,

59Callieri illustrates ten seals which may depict Skanda-Kärttikeya from this region and time

figure 8). period. Srinivasan also illustrates a single sealing of this type from the same area and era (1997/98:264,

233

and some are clear copies of later images of Skanda-Kärttikeya. P. Pal has also noted the similarity between

the figures. He thinks the images of Mangala are taken from those of Skanda-Kārttikeya (1969:36, 45-46).

In this context he also suggests the following:

It may be of interest here to recall that Mars in Classical mythology is considered a war-god and his attribute often is a spear. Kärttikeya, the adhidevatā of Mangala, is also a war-god, with the spear as his emblem, and possibly the concept of Mars had somewhat influenced the iconic type of both Mangala and Kārttikeya. (1969:45-46)

Pal does not develop this idea, but I certainly think his idea has merit based on the above material. While

these images of Mangala are late relative to the time period studied here, they do suggest that some

association between Skanda-Kārttikeya and Mars/Mangala endured over time.

On the basis of the above, I would argue that one Kusāna understanding of Skanda-Kärttikeya was

as a Mars-like figure. The image of the deity seems to be borrowed from Roman and Greek depictions of

Ares/Mars from both textual and artistic sources that were current during the Kusana period. As I have

already suggested, the Kusanas may also have viewed their numismatic representations of Skanda-Kumāra

and Visakha as like, or a version of, the Dioscuri, figures from another Hellenistic astrological and martial

cult. All of this evidence suggests a Hellenized reading of Skanda-Kärttikeya and his related forms by the

Kusānas which had a dramatic impact on the cult itself, particularly in the reorientation of the cult towards a

martial, state sponsored deity.60

60In this note I will review other comments by scholars that try to link Skanda-Kārttikeya to another astrological deity -- the sun. A number of scholars have also attempted to explain the cock on Skanda-Kärttikeya statuary as an Indian solar symbol and have presented an argument to suggest that Skanda-Kārttikeya is related to Sūrya (Kar 1954:84; Banerjea 1956:106; P. K. Agrawala 1967:45-46; Rao 1968 II:432; Chatterjee 1970:23). These arguments are largely based on passages from the Åranyakaparvan which compare the shine of Skanda-Kärttikeya to the sun and the Lala Bhagat pillar. I have already questioned the common interpretation of the Lala Bhagat pillar and these passages from the Epic. While the association between a cock and the sun is a logical one, I have found nothing within Indian literature from this period that would support such an idea. There may be some relationship between Skanda-Kärttikeya and the sun, but if there is, it is through Iranian religion. The connection with Iranian religion begins with Verethraghna and his association with a cock. I have already linked Skanda-Kärttikeya with this deity in relation to their shared martial character and shared emblem of a bird. Now I would like to argue that the specific emblem of a cock may also relate to a shared astrological link from an Iranian context. The primary birds associated with Verethraghna are a raven and crow, but a cock is also regarded as a bird associated with Verethraghna, likely due to its role as a herald of

234

5.3.1 Yaudheya Coins: The Class Six Issues

As the reader may recall, the Yaudheya class six issues are post-Kusāna. Thus far we have

examined the Yaudheya class three issues as examples of pre-Kusāna depictions of Skanda-Kārttikeya, the

Kusāna depictions of the deity, and these class six issues provide an understanding of the influence of

Kusāna depictions of the deity on this indigerous group. The re-birth of the Yaudheya state seems to come

just after the decline of the Kusana empire, and their class six coinage provides us with a final step in the

the sun (Ackerman 1964/65 II:793). The cock's association with Verethraghna relates to his status not only as a warrior god, but also as a sun deity. This association between Verethraghna and the sun appears in Sasanian religion where he was regarded as the same as Mithra (Ackerman 1964/65 II:788). The cock is regarded in Iranian religion as a sun bird (Ackerman 1964/65 II:803), and the sun/fire was an instrumental aspect of Iranian religion. The cock is, then, linked to Verethraghna as a solar symbol. As I have already suggested, the cock in relation to Skanda-Kärttikeya does fit in his Graha and early non-Brahminical ritual cult, but with the Kusana era statuary of the far north-west the cock is more likely a martial emblem that appears to be influenced by Iranian or Parthian culture. This link, I would suggest is more than martial, but also a link to sun worship and both figures as solar deities within the north-western context. In short,

Verethraghna. Skanda-Kärttikeya's characterization in the north-west appears to have borrowed much from that of

The Iranian solar connection is further made by the Sambha and Bhavisya Puranas. As I have already discussed, in these texts Skanda-Kärttikeya is identified with an attendant of Sūrya named Śrausa, and this figure is the Iranian Sraoša (Hazra 1958 I:45; P. K. Agrawala 1967:72-73). The shared emblem of a cock also relates Skanda-Kārttikeya and Sraoša with a solar cult. Some scholars use these texts to establish an Indian religious solar connection for Skanda-Kārttikeya (Kar 1954:84; Rao 1968:431-432; R. C. Agrawala 1968:163-165), but, as P. K. Agrawala (1967:73) and Chatterjee (1970:23) point out, these deities and associations are based in Iranian religion. While these are Indian texts, they are heavily influenced by Iranian solar cults, and the deities to which Skanda-Kārttikeya is assimilated are also Iranian. While there are Indian connections to Skanda-Kärttikeya and a solar cult, they do not point to an Indian origin for this aspect of his cult, as a number of the above scholars have argued, but to the Iranian influence in the development of his cult. There is significant evidence to suggest that Skanda-Kārttikeya was amalgamated with Iranian warrior and solar deities, and it was likely under the Kusanas that this occured. One Scholar has taken these points seriously. S. Sen has argued for an identification between Śraosa and Skanda-Kārttikeya. Sen does, however, argue that this correspondence between Śrausa and Skanda-Kärttikeya goes back to an original connection found between the Rgveda and Avesta that has survived. He attempts to link a Kumāra found in Rgveda 10.135 with a "prototype of post-Vedic Kumāra and a counterpart of Iranian Śrausa" (1950:27). His argument here is not strong, and he has fallen into the trap of trying to establish the roots of Skanda-Kärttikeya in Vedic material. I would conclude that whatever solar and astrological connection that exists for Skanda-Kārttikeya it is one based in Iranian and Hellenistic religion through an association between Skanda-Kārttikeya and martial/astral deities like Verethraghna, Śrausa and Mars/Ares. It is also clear that the Kusānas borrowed widely from Parthian and Hellenistic cultures in their depictions and use of Skanda-Kārttikeya. To some degree we must view the Kusāna era Skanda-Kārttikeya depicted in the far north-west as a composite of various martial deities from several cultures.

235

development of this deity before the rise of the Gupta empire. We do not really know what the

circumstances of the Yaudheyas were during the rule of the Kusanas. Archaeological digs at previous

Yaudheya centres during class two coin production like Rohtak indicate that they became Kusāna centres. I

have also argued that the Yaudheyas did not produce coins during the Kusana era. The group did not

disappear, though, and is mentioned in the Junagadh inscription, which is dated to circa 150 CE (Dasgupta

1974:197), but beyond this we have no records of the group until after the Kusanas. After the decline of the

Kusana empire the Yaudheyas produce their last series of coinage called 'class six' by Allan (1936:cl).

These coins are usually dated on the basis of palaeography to the third or fourth century CE, and they

borrow their style from the Kusānas (Dasgupta 1974:209-210, 212-213). Yaudheya class six coinage is

much more uniform than their class three issues and has sparked less debate among scholars.

Archaeological sites suggest that the Yaudheya state reaches its height during this period with finds in the

Punjab, Haryana, Rajasthan and Uttar Pradesh (see map number four). I will begin by describing and

illustrating the varieties of class six coinage. Then, I will discuss their importance to this study.

The obverse of the coin illustrated in figure 72 depicts a single-headed deity holding a spear in his

right hand and his left hand on his hip. There are small oblongs on either side of his head which probably

represent ears and below these, dots, which represent earrings. He wears a headdress that appears as a

horizontal bar over his head, above which is a triangle-shaped object resting on the bar on one of its points.

The headdress may represent a turban or top knot. He also wears a necklace, and there is a sash around his

waist which runs down between his legs. There are circles just above his feet which either represent the

bottom of his clothing, or the top of boots. To his left is a bird which faces him. The bird has been variously

identified as a cock or a peacock. There is a Brähmi legend around the outer edge which reads:

"Yaudheyaganasya jaya" (Dasgupta 1974:209), which I take to mean 'glory to the Yaudheya assembly (or

group)'. With Prakash (1965:136-137) I take jaya to be a general term for benediction or laudation among

236

warriors and not to refer to an actual victory.61 The reverse of the coin shows a single-headed female figure

who stands frontally and looks to the right. Her right arm is raised, and her fingers are arranged to have a

claw-like look to them. Her left hand rests on her hip. She wears a necklace, and the pleats and end of her

garment are visible around her legs. There is a beaded boarder around the outside of the coin.

The remaining three varieties are illustrated in figures 73-75. The obverse on the coins in figures

73-75 are almost identical to that depicted on 72. The main difference is that the obverse of the coin in

figure 73 has "dvi" attached to the legend, and the obverse of the coin in figure 74 has "tri" attached to the

legend.62 The reverse of the coin in figure 73 shows the same female figure with an inverted nandipada

61 Jaya has mislead a number of scholars. Altekar seems to champion an understanding of jaya as a marker of a victory over the Kusānas (1967:30). Altekar hypothesizes that "The credit of giving the first blow to the Kushana empire really belongs to the Yaudheyas", but there is no direct evidence for such a 'first blow' or any 'blow' for that matter between the two groups (1967:28). J. Prakash is critical of the above notion. He argues for a more general sense for jaya as a term of benediction among ksatriyas (1965:136-137). He also notes that the term is used by the Malavas and Arjunayanas on their coinage (1965:136). P. L. Gupta is critical of those who translate jaya as victory on these coins. He uses grammatical rules from Paninī to argue for jaya as a technical term implying an instrument of victory (1977:64-65). He states that the term is used as a marker of authority, and its use demonstrates that issuing coins "had now become the prerogative of the state" (1977:65). Jaya may not, then, mean victory and likely does not suggest a specific historical event.

62The terms dri and tr have caused significant debate, but their meaning on these coins remains obscure. Cunningham suggests: "the coins of the Yaudheyas show that they were divided into three tribes" (1963:76). He does not specifically mention the coin legends in relation to his hypothesis, but we can assume that this is his intent. Allan thinks the terms refer to the second and third sections of the group (1936:cli), but like Cunningham he has no evidence to support such an idea. Altekar initially suggests that the terms refer to different units of a Yaudheya federation of republics (1949:52). Later he and Majumdar note that "No satisfactory explanation of these terms can at present be proposed" (1967:31-32). They do, though, go onto suggest that the terms refer to an alliance or confederation between the Yaudheyas, Kunindas and Arjunäyanas and even suggest how their administration was conducted (1967:32). They do not, however, supply any evidence for this hypothesis beyond the dvi and tr legends. U. S. Rao seems to suggest that the coins demonstrate three consecutive Yaudheya republics, and because there are so few of the tr coins that this third republic was short lived (1962:139). Lahiri suggests that the three legends represent "three different administrative units of the Yaudheya republic" (1974:212). He does, however, acknowledge that his suggestion is a conjecture (1974:212). Dasgupta argues against Altekar's confederating units hypothesis and suggests the terms refer to a second and third republic in the history of the Yaudheyas (1974:218). J. P. Singh suggests the terms refer to separate Yaudheya mints (1977:11). Handa argues against such a view. He argues that the three separate coin legends refer to three separate units of the Yaudheyas in three separate regions (1978/79:30-33; 1983:7). He uses the evidence of coin moulds from Sunet which demonstrate that all three legend types were cast simultaneously working against the idea that the terms dvi and tr refer to separate victories, republics, mints or members as others have

237

device to her left and a pot with plant to her right. The reverse of the coin in figure 74 places a double S

device to the left of the figure and a sankh shell to her right. The reverse of the coin in figure 75 also places

a double S to her left. All of these coins are copper.

5.3.2 The Class Six Issues: From Six Heads to One and the Impact of Foreign Design

These coins differ in a number of ways from the class three issues. Their uniformity of type and

quality is suggestive of a more controlled minting process, which in turn is suggestive of a more centralized

and prosperous state administration.63 The six-headed figures on the obverse and reverse are gone and have

been replaced by attractive single-headed figures. I have suggested above that the switch to a single-headed

deity has certain religious implications as well as being suggestive of borrowing from foreign iconographic

sources.

The general appearance of these coins is based on Kusāna coinage, but the obverse image of

argued (1983:7; 1991:73). Ahmad and Kumar argue for dvi and tr to indicate the existence of second and third Yaudheya republics (1993:53). They argue against a number of other arguments including Handa's simultaneous casting hypothesis (1993:53). Their argument, however, is based on the weights of re-struck Yaudheya issues from the Bishan hoard. Their study of these re-struck coins shows that the re-struck dvi legend coins weigh more on average than the legend without dvi or tr. From this they conclude: "So it is certain that Class 1 [plain legend] preceded 2 [dvi]" (1993:52). It is not clear, however, how this evidence leads them to this conclusion, or that a number of other conclusions are not also inferable by the evidence of these weights. They also note that the tr legend coins do not fit their model, but regard the small number of tr coins found in the hoard not to allow for an adequate statistical sample (1993:52). Ahmad's and Kumar's hypothesis is not well demonstrated. Their study is only based on one very specific type of Yaudheya coin production; one group within their sample does not fit their argument; and their conclusions do not take into account all the potential options. I have no conclusive statement concerning these dvi and tr legends. I regard their meaning as obscure, even with the extensive discussion of these terms indicated above. Indeed, what I have supplied in this note is only a survey of the major opinions on these legends and several other arguments and counter arguments exist. If anything, these terms are a reminder that caution is always required when arguing that these coins relate to historical facts, and at times it is best to acknowledge that we do not know enough of the group to understand these legends rather than to build an elaborate, but unsubstantiated hypothesis.

63B. Chakravarty has analyzed nine class six coins and finds them all to be of similar weight and to have a high copper content of 99.5% (1987:149-150). This scholar feels that these results show "that Yaudheyas used a high quality of scientific and technical skill in large scale production of these coins" (1987:151). The technical soundness of the coins also leads me to believe that the government that produced them was in an administrative position to exercise control over its various units and had the economic and administrative ability to ensure the production of adequate high quality currency.

238

Skanda-Kārttikeya64 does not have an exact Kusāna prototype. Similarities between the Huviska Mahāsena

issue and the Yaudheya class six Skanda-Kärttikeya are clear. Both figures stand frontally with similar hand

positions. Both are associated with birds, but Mahāsena does not carry a spear on his coin, and Skanda-

Kärttikeya does not carry a sword on the Yaudheya issue.65 They also both wear a headdress, which may be

a turban or top knot. While the two figures are similar, Mahäsena is not an exact match for the class six

Skanda-Kärttikeya. The reverse image is influenced by Kusana representations of deities. Her general

posture is borrowed from Kusana images of Helios, Mithra and Mao (Lahiri 1974:205; Kumar 1993:50).

These class six coins are clearly post-Kusana and show the influence of Kusāna numismatic style.

As I have suggested above, these numismatic borrowings from other sources may well be rooted in

economic necessity. Kusana coinage was successful, and these Yaudheya issues may have sought initial

acceptance by borrowing from the established currency in their region after the fall of the Kusanas. In doing

so they present a very 'Kusanized' or 'Hellenized' representation of this deity. His single-headed status

would have been embraced by Hellenized groups, and this iconographic shift away from the six-headed

form implies, I think, the growing influence of Scythian or Kusana society on how this deity is perceived

and represented. While the single-headed depiction may have been adopted for economic reasons, the

exclusive use of it by the Yaudheyas acts to assert it as the representation of the deity. The six-headed

horrific creature from Yaudheya class three coins has been transformed into an approachable, human figure

who does not, at least in appearance, inspire fear. As I have discussed earlier, this shift from horrific and

feared to human and beautiful is a significant shift, which has implications for the popularity of the deity.

64The inscriptions on these coins do not identify the deity, and I will refer to him as Skanda- Kärttikeya. He can be identified as Skanda-Kärttikeya because of his spear, bird and on the strength of his resemblance to the single-headed figure in past Yaudheya issues. It is not clear, however, if his name is left off because it is assumed that people will know who he is, or if a considerable degree of amalgamation has taken place between figures like Brahmanyadeva, Mahāsena, Skanda-Kumāra, Viākhā and so on, and the name is absent so that this one figure can be seen as representing them all.

65The representation of Mahasena from figure 71 does show the deity with a spear, though his headdress is not clear.

239

The reverse image also now appears exclusively as single-headed. While a sound argument can be

made to identify the six-headed reverse image on the class three coins as Sasthi, this figure is much more

ambiguous and harder to identify. As I have discussed above, there does appear to have been an

amalgamation of various goddesses of prosperity into one form in the north-west; a similar pattern seems to

occur in post-Kusana India. Among these goddesses who are amalgamated into one form are Sasthi and

Laksmī.66 Part of the end of this steady emphasis on the general theme of prosperity, as opposed to the

specific roles of the individual goddesses themselves, is that many of these minor goddesses become just

like Laksmi and are absorbed into her cult. The ambiguity of this reverse figure on the class six coins may

be a reflection of this trend.6 This shift is significant because it demonstrates the fluidity of the cults at this

time and in this region.

Sasthī's once fearful appearance has been replaced by a beautiful one. An identifiable six-headed

Sasthī figure has been replaced by a generalized fertility figure, who was stereotypical on South Asian

numismatic issues during this time. The reverse figure on the class six coins could be any benevolent

goddess or woman for that matter and is a far cry from the six-headed goddess whose worship is

recommended to parents to prevent her from taking an infant's life. This process towards amalgamation and

generalization seems to have resulted in the loss of what made Sasthi a recognizable goddess with her own

distinctive personality and links to the Graha cult. The absence of a clearly identifiable Sasthi figure may

66While Sasthi may not seem like an obvious choice as a goddess of prosperity, most of these goddesses share a common theme of fertility, with which Sasthi is concerned and hence her relationship to a general theme of prosperity. This amalgamation of goddesses into one prosperity goddess may have a great deal to do with the influence of the north-western understanding of these goddesses already discussed above.

67P. Callieri observes a similar trend in seals and sealings of the north-west and Gandhāra at approximately this time. He points out the popularity of Ardoxso on these seals and argues that a number of female deities, whose main characteristic is fertility, have been merged into the single iconographic type of Ardoxšo (1997:254). The later Kusānas also stop producing coinage with a wide variety of deities on them and start to produce coinage that only depicts Śiva/Oēšo and Ardoxšo. Their use of Ardoxšo is an example of the stereotypical use of such a fertility/prosperity goddess on coinage. The use of such goddesses on coinage becomes a normative numismatic device. Similar trends in India result in the absorption of a number of female deities into figures such as Laksmī and Pārvatī.

240

also signal that the obverse single-headed Skanda-Kārttikeya has undergone a similar transformation from a

Graha to a more benign protective figure.

Another point that should be mentioned is the addition of the bird on the obverse of these class six

issues. The bird has usually been identified as a peacock and in some cases a cock.68 I think that viewing

the bird as a cock is more defensible than as a peacock. As we have seen, all of the Kusana era statuary

from the north-west show Skanda-Karttikeya with a cock. The appearance of the peacock in statuary can be

traced to the Gupta period (Srinivasan 1997/98:240). This class six coinage is close to the beginning of the

Gupta era. Hence, we should not dismiss the possibility that the bird is a peacock on these class six coins,

but most of what we have seen from the Yaudheya class six coinage looks back to the influence of the

Kusanas and not ahead to the Guptas. The Kusana and Yaudheya style shows the deity standing with a

smaller bird. Hence, I suggest that the bird shown on the class six coins is likely a cock and likely modelled

after Kusana iconographic designs. The addition of the bird or cock is another sign of Kusana influence on

these coins and points to a Mahäsena-like figure. It indicates that the changes seen in the character of

Skanda-Kārttikeya in the Kusāna era were not just limited to Huviska's rule, but were part of, or influenced,

a much larger shift in the depiction and characterization of Skanda-Kärttikeya in the north of India. In my

material above, I have separated an indigenous understanding of this deity as a Graha from a foreign

understanding of him as a military figure. What the Yaudheya class six coins suggest is that the

understanding of Skanda-Karttikeya as a military god has now been adopted by indigenous groups as well.

What we must now question is whether or not these numismatic iconographic shifts on the Yaudheya coins

indicate a shift in Skanda-Kārttikeya's popularity base as I have suggested in relation to Kusāna coinage and

statuary.

68Most scholars have followed Allan who identifies the bird as a peacock (1936:cl). Banerjea (1956:143), Sinha (1979:80), Thakur (1981:49), Gupta (1988:7) and Rana (1995:22) all identify the bird as a peacock. Dasgupta states that in most cases the bird is a peacock, but in some cases may be a cock (1974:220-221), and Mani states it may be a cock or peacock (1990:70).

241

5.3.3 Yaudheya Class Six Coins: The Context of the Yaudheyas and the 'Popularity' of Skanda- Kārttikeya at the Time of the Class Six Issues

The majority of historians reviewing this period, or Skanda-Kärttikeya scholars examining these

coins, have come to the conclusion that Skanda-Karttikeya was 'popular' amongst indigenous groups in

India and especially with the Yaudheyas.69 I will argue

69In this note I will review the comments regarding the popularity of this deity made by scholars. As we have already seen in sectio 4.2.15 most scholars think that the Yaudheyas dedicated their lands to Skanda-Kärttikeya on the basis of the class three inscription. V. S. Agrawala feels that Skanda-Kärttikeya "received special patronage" from the Yaudheyas (1943:31). He also mentions three seals that use names and epithets of Sasthi and Skanda-Kārttikeya as personal names. He regards this use of divine names for personal names as evidence of these deities' popularity (1943:31-32). It should be noted, though, that the seals he refers to come from Gupta era Rajghat (V. S. Agrawala 1943:31). They are, then, geographically and temporally removed from the context of these Yaudheya coins. A. S. Altekar states: "The Yaudheyas had an unsurpassed reputation for bravery, and no wonder; for they were the devotees of Kärttikeya, the generalissimo of gods" (1962:119). This scholar also makes a number of claims about the history of the Yaudheyas, but few of these claims are well substantiated (1962:119-220, 1967:28-33). P. K. Agrawala states in relation to Yaudheya coinage: "Evidently, the Yaudheyas devoted special service to their god" (1967:42). Chatterjee begins with claims like Skanda-Kärttikeya was the "guardian deity" of the Yaudheyas (1970:35), and that the coins demonstrate that the Yaudheyas had "dedicated their kingdom to the War-god Skanda-Kārttikeya" (1970:36), and moves on to such unsubstantiated claims as: "In their wars against the foreigners they were certainly inspired by their tutelary deity Skanda-Kārttikeya, the War-god" (1970:38). He does not, however, supply evidence of these wars. Navaratnam follows similar lines in describing the Yaudheyas as "the most ardent devotees of Skanda" (1973:100). Dasgupta states that " ... the presiding god of the Yaudheyas is known to have been Karttikeya ... " (1974:216), and " ... coins of the Yaudheyas prove beyond doubt that this god was venerated by them" (1974:220). Ghurye argues that the Yaudheyas "have been known as great warriors and to have had Kumara-Karttikeya as their patron deity since at least the 2nd century A. D." (1977:134). Elsewhere, however, he states it is "probable that as early as the 5th century B.C. Karttikeya-Kumara had become the tutelary deity of the Yaudheyas" (1977:135), though he does not supply much evidence for this conclusion beyond these much later coins and references from the Mahabhārata. Sinha also argues that the Yaudheya coins and coin legends suggest the popularity of Skanda- Kärttikeya and also allude to historical events (1979:76-80), points which he is largely unable to substantiate. A. M. Shastri argues that the Yaudheya coins illustrate that Skanda-Kārttikeya "was popularly worshiped by people" (1977:92). J. P. Singh thinks the Yaudheya coins were issued "in honour of the deities to whose grace they attributed their independence" (1977:6). Singh does not attempt to justify this claim or to state the historical period of Yaudheya coin production to which he refers. O. P. Singh regards Skanda-Kärttikeya as "the deity of the tribe [the Yaudheyas]" (1977:136), and that the Yaudheya coins "throw interesting light on the popularity of the deity" (1978:71). Roy regards the coins as evidence of worship of Skanda-Kärttikeya by the Yaudheyas and asserts that the legend on the class six coins refers to a victory over an unnamed group of Saka-Ksatrapas (1984:76). As with other scholars who make similar claims, Roy is unable to provide any direct evidence of this military success. Gupta simply suggests these coins demonstrate that Skanda-Kärttikeya was the tutelary deity of the Yaudheyas (1988:7). Mani (1990:70) and Rana (1995:22) seem to come to a similar conclusion as Gupta. Ahmad and Kumar regard the placing of Skanda-Kärttikeya on the obverse of class six coins "as the mark

242

here that this is not supported by archaeological evidence for the Yaudheyas at the time of the class six

coins.

Only a small fraction of Yaudheya seals and inscriptions of this period make indisputable direct

reference to the deity, no Yaudheya statuary depicting the deity has come to light; no Yaudheya temples to

the god have ever been unearthed. Outside of the Yaudheya coinage, Skanda-Kärttikeya is a virtual non-

entity in the archaeological sources for this group. Having stated this, however, I must admit that there is

not a great deal of archaeological work being done on this group, and we must proceed with some caution.

Artifacts that have been recovered in some numbers are seals and sealings, which present us with some idea

of popular religious devotion during this period.70

A major site for the discovery of Yaudheya seals and sealings is Sunet. Sunet is a village located

approximately five kilometers west of Ludhiana in the Punjab (see map number two). The site was

excavated in 1983-84, but not all of the findings of the dig have been well documented. Unfortunately, the

of special veneration to the tutelary deity Kārttikeya" (1993:51). As the above demonstrates, a number of scholars regard the Yaudheya people as special worshipers of Skanda-Kärttikeya because of these class three and class six coins. A great deal is made of the legends on these coins by scholars, the appearance of the deity himself and the apparent appearance of temples on the reverse of the class three coins. I do not disagree with these comments in relation to the Yaudheya class three coinage. Skanda-Kärttikeya or Brahmanyadeva along with Sasthi were likely the guardian or tutilary deities of the group at that time. What I disagree with is that all of these scholars naively assume that their statements apply to both types of coinage and most make no attempt to separate the two coin types or to place their comments into a historical context. At least two-hundred years separates the class three and class six issues. The intervening period had immense significance for the development of religions in South Asia. The context of worship and the character of Skanda-Kārttikeya saw dramatic changes between the period of class three and class six coin production. To generalize a meaning for all Yaudheya coinage simply ignores the importance of the historical context in the development of the cult of Skanda-Kārttikeya.

70Seals and sealings have been defined in the following manner:

A seal is an engraved stamp bearing, singly or collectively, a device, mark or letters pertaining to the owner. A sealing is an impression of such a stamp on paper, parchment or some such substance as clay, wax, etc., and capable to being attached to a document to authenticate it or serve as signature or cognizance .... At times, it also contained a pious formula, ethical or religious, in which case it was used for purposes of moral propaganda or kept in a shrine as a votive-offering or given to a pilgrim as a memento. (Thaplyal 1972:1)

243

mound at the site contained a large quantity of Kusāna bricks,71 which were dug-out and hauled away by the

British to be used as ballast in railroad construction projects (Handa 1987:334). The result, I suspect, is that

many of the artifacts from the site were long gone by the time of the excavation. Cunningham, an English

archaeologist of early India, found about one-thousand Yaudheya class six coins at the site. He also found a

single coin of Hermaeus, 269 coins of earlier Indo-Scythian kings, 132 coins of later Indo-Scythian kings,

one Gupta coin and 126 coins from the Indo-Sassanian period (Handa 1987:333). A number of seals,

sealings, beads and other artifacts have also been recovered. A female terracotta figure has been recovered,

and fragments of other statuary are also mentioned by Cunningham, but not recorded in detail (Handa

1987:333-334). Many of the seals and sealings recovered from the site belong to the Vrsnis, who appear to

have ruled there before the Kusanas."2 The village was founded at the close of the first quarter of the second

millennium BCE and reached its peak during Kusana rule (Handa 1987:333-334). The Yaudheyas appear to

occupy the site during the period of their class six coin production. Approximately 30, 000 coin moulds of

class six Yaudheya coinage have been recovered from Sunet, which demonstrates the presence of Yaudheya

administration in the town at the time of class six production (Handa 1988:132; 1991:73).73 Finds of

71Bricks measuring 32x22x5 cm are typical of the Kusāna era and are often called Kusāna bricks.

72There are numerous mythological references involving a group of people called Vrsnis. Whether or not this historical group is related to the mythical group is not clear to me, but many scholars assume that both the historical and mythical Vrsnis are one and the same (Dasgupta 1974:188-189). Only one coin from the group has been recovered (Dasgupta 1974:189), but numerous seals and sealings from Sunet use the name Vrsni and use similar devices as found on the coin suggesting that the same group issued both and were located in Sunet. Handa thinks that the Vrsnis ruled Sunet after the Kusanas (1987:335), but the chronology that Cribb has proposed for the Kusanas does not allow for a lengthy gap between Kusāna and Yaudheya occupation. I think it more likely that the Vrsnis ruled the area before the Common Era and lost the land to successive waves of invading Indo-Greeks and Indo-Scythians. The group may have briefly re- emerged after the Kusanas, but could not have ruled for a long period before the Yaudheyas control the area.

73 As with most topics relating to Yaudheya currency and history, the nature of these moulds for both class two and six coins has sparked some debate. Surviving examples of class two and six coinage indicate that they were produced by both the die struck method and the casting method. The existence of both techniques of coin production has lead some scholars to argue for the massive production of caste coinage by ancient forgers. In two similar articles A. M. Shastri (1993, 1995) argues that all of the Yaudheya coins produced in moulds represent the work of forgers (1993:102). He regards only the die

244

Yaudheya seals and sealings also suggest the Yaudheyas controlled the area during the third to fourth

century CE (Handa 1984:155). Handa also suggests that Sunet may have been the Yaudheya capital during

this time period (1991:73).

In the town 212 seals and sealings have been found (Handa 1985:99-112). Among the numerous

seals and sealings recovered from this town are those that reflect religious concepts or sealings which may

have been issued by a religious group. D. Handa, who has studied the glyptics, notes a number of seals and

sealings from Sunet which depict the Saivite device of a trident/axe with legends like Maheśvara, which he

argues are clear indications of Saivism (1984:161). He also notes that a number of personal names reflect a

Saivite background, such as seal 106 in his catalogue with the legend "Bhavadeva" and sealing 168 with the

legend "Rudra" (1985:133).74 Regarding personal names he comments:

They [names] certainly have a socio-cultural background. Children are named on the basis of their physical traits and other characteristics or according to the religious beliefs, socio-economic and/or political status, educational background, avocations, likes and dislikes of the parent or guardians .... As elsewhere in India, children in Punjab also seem to have been named after favourite deities, rivers, heroes, sages, saints, sacred objects, etc. (1985:132)

struck coins as representing official Yaudheya issues, and because he finds moulds at Yaudheya sites for Gupta and Kusana coins he sees a long history of counterfeiting in the Yaudheya territory (1993:101-102). Shastri's arguments are compelling, and the small number of Gupta and Kusāna moulds is suggestive of some forgery at some point, but the large number of Yaudheya moulds found, 30, 000 in Sunet alone, represents, I think, different circumstances. A large number of coin moulds (approximately 8 500) of class two Yaudheya coins were also found near Rohtak. The moulds at Rohtak were found within the fortified township walls (Kumar 1996:98-102). M. Kumar, who excavated the Rohtak site, reports that he doubts the ability of forgers to carry out such a large operation within official walls (1996:102), and I agree. The evidence from seals and sealings from Sunet suggests that this town was a Yaudheya centre during the production of class six coins. I think it likely operated in a similar manner as the earlier Rohtak model. It appears that the Yaudheyas produced both die struck and cast coins, and this double production may indicate a degree of haste in coin production or the farming out of minting to various groups within the Yaudheya territory, who employed different minting methods.

74Handa has published two versions of this catalogue. The sealing in question is number 66 in the 1985 catalogue and 62 in the 1984 catalogue. I will list the numbers from the 1985 catalogue for the rest of this section.

245

In other words, the use of divine names as personal names is suggestive of a deity's popularity.75 There are

at least twelve seals and sealings with Saivite names listed by Handa in his catalogue (1985:133). There are

also large numbers of seals and sealings which use devices associated with Saivism such as tridents (41

seals/sealings) and bulls (12 seals/sealings) (Handa 1985:99-112). He also comments on a number of

legends ending in īśvara like "Śrī-Kundeśvarasya" and "Śrī-Yyaumīśvarasya", which he thinks are

suggestive of sealings issued from Saivite shrines (1984:161-162). The example of the seals and sealings

featuring Saivite names and symbols demonstrates that we can use this evidence to indicate the popularity of

a cult in this area. I will now review the seals and sealings which may relate to the cult of Skanda-Kārttikeya

from Sunet to gage this deity's relative popularity.

Only a few seals and sealings from Sunet allude to the existence of a cult of Skanda-Kārttikeya in

this area. The only name of the deity to appear on these seals and sealings is Kumāra. Sealing 66 in Handa's

catalogue is one such possible reference to the deity. This single sealing carries the legend: "Kumāra-

Kumāra" with no device on it (Handa 1985:100). This sealing may be intended as a laudatory praise for

Skanda-Kumära. There are six sealings bearing the name Kumāra, but it is not clear if all of these uses of

the name refer to the deity.76 Another sealing which Handa mentions as possibly connected to a Skanda-

Kārttikeya cult is a single sealing (number 229) depicting a dancing peacock (1985:111). This sealing does

75v. S. Agrawala makes a similar suggestion concerning seals and sealings from Rajghat (1943:31). Thaplyal also comes to a similar conclusion regarding devices and legends on glyptics (1972:10, 136-137).

76Other sealings with the name Kumara are: 47 which reads "Devabhadrasa Kāndelaka Kumārasa" with no device; 127 and 128 which read "Srī-Kumara" with no device; 129 which reads "Śrī-Kumārabodhi" with a crescent device; and 131 which reads "Śrī-Kumārasarma" with a lion device (1985:99-104). It is unlikely that any of these refers to a shine. I have suggested before that Kumära in inscriptions and sealings often refers to a prince or the proper name of an individual, and when the issuer of the inscription, coin or seal wished to specifically indicate Skanda-Kärttikeya they used Kumära with another epithet of the deity to make the meaning clear. Certainly, the few devices placed on these sealings do not link the name Kumāra with a shrine to Skanda-Kärttikeya. These sealings are regarded by Handa as possible indications of Skanda-Kārttikeya devotion (1985:124), and, as noted earlier, the use of a deity's name as a personal name can be read as a sign of devotion.

246

not, however, have any legend on it which could help us understand the intent of the peacock. Given that the

peacock is an emblem of Skanda-Kärttikeya, however, we may assume that this sealings relates to his cult.

Thaplyal mentions two sealings from Sunet which show a female standing by a tree feeding a bird

at her feet. Thaplyal suggests that the tree may represent Siva, the female Parvati and "the bird may be a

cock. Its presence in the composition may be easily explained. The bird is a favourite of Kärttikeya, son of

Parvatī, and what is dear to him is equally dear to her" (1972:175-176). While creative, Thaplyal's

argument is speculative at best. What these sealings suggest is that the cult of Skanda-Kumara does exist

among the Yaudheyas in Sunet, but is only a minor cult when compared with other cults, which I will

discuss below.

I have already discussed seals and sealings from Sunet which suggest Saivite leanings. I noted that

at least twelve seals and sealings carry a name associated with the cult and fifty-three make use of a Saivite

device. There are also numerous seals and sealings from Sunet which suggest Visnavite leanings. Handa

argues for a shrine devoted to the five Vrsni heroes at Sunet, based on a single seal with a Kharosthi legend

which reads: "Jaya-Pachalaya". The seal also has a half elephant and half lion on post and discus as its

motifs (1984:163). There is also a later third century CE legend which reads: "Siddham Jaya Pañcha

Nāyākānām" (Handa 1984:148). Handa interprets the frequent use of the half-elephant and half-lion motif

along with the use of the pestle, mace and discus on seals and sealings as an indication of the worship of

Visnu and related cults (1984:163). There are also thirty-one personal names on sealings that relate to Visnu

or his avataras. There are also at least thirty-four seals and sealings which make use of the name Visnu or

an epithet of the deity like: "Jitain Bhagavata", "jaya-svāmi", and "Jitam Bhagavatā Svāmi Nārāyanena"

accompanied by Vaisnavite devices (1984:163). All of these suggest to Handa the existence of Vaisnavite

shrines in Sunet from its period as a town controlled by the Vrsnis through to Gupta control by the fourth

centuries and on (1984:163-164).

Another important cult in Sunet appears to have involved the Mätrs. As I have noted in my section

on textual sources, the Mätrs and Skanda-Kārttikeya are closely associated. The existence of a Mātr cult in

247

Sunet may point to additional evidence of Skanda-Kärttikeya worship. There are ten seals and sealings

which make use of Mätr in their legend. Two (numbers 73 and 74) have only "mätrh" with a trident device

(Handa 1984:147). Five seals/sealings use some form of mtrsarma with either a Nandi device or a trident

device, and other seals/sealings use a combination of Matr with soma or bhūta (Handa 1985:99-112).77

Handa regards these as indicating a Saivite association (1984:165), and I agree with him based on the use of

Saivite devices used with these legends. These Mätr seals and sealings do not provide evidence of a cult of

Skanda-Kärttikeya, but suggest that the Matr cult has been absorbed into the cult of Siva. I have noted in my

Kusāna sculpture section that the earliest depictions of Mätrs show them with Skanda-Kārttikeya, but by the

Gupta period Skanda-Kārttikeya has been replaced by Śiva and Ganeśa (P. K. Agrawala 1967:48; R. C.

Agrawala 1971:85). With these seals and sealings we may be witnessing the same pattern. We must also

consider the possibility that the cult of Skanda-Kärttikeya has also been absorbed into the cult of Siva, and

that is why we find so few epigraphic references to him at this site; people had replaced worship of Skanda-

Kärttikeya with worship of his more powerful father.

Thus, Sunet, a town which produced class six coins depicting Skanda-Kārttikeya by the thousands,

issued by a group who are supposed to have had a special devotion to the deity in question, has revealed

only a small number of glyptics that could be linked to the cult of Skanda-Kārttikeya. This situation is not

unique to Sunet; there are no known Yaudheya seals, sealings, inscriptions, statues or temples which depict

the deity or make direct references to Brahmanyadeva or Skanda-Kärttikeya. Only the term Kumāra, as

noted, is met with on a small number of sealings from Sunet.78

77These seals and sealings are number 75, which reads: "mātrsarma" with a Nandi device; number 76, which reads: "mātrsarma" with a trident device; number 92, which reads: "śrī mātrśarmanah" with no device; number 163, which reads: "śrīr mātrsarmasya" with no device; number 165, which reads "śrīr mätrśarmasa" with a trident device; number 133, which reads "śrī matrsomasya" with a lion before a tree device; number 134, which reads "srīr mätrsomasya" with a trident and lion device; and number 164, which reads "śrīr mātrbhūtasya" with no device (Handa 1985:99-112).

78Not all of the glyptics from centres like Rohtak have been recorded, and Handa does not attempt to date many of the seals and sealings he discusses. It may be that the 'Kumara' glyptics I have referred to above do not come from the Yaudheyas or that other glyptics from Rohtak might suggest more about

248

I am not the first person to observe the paucity of seals and sealings associated with this deity from

Yaudheya sources. Thaplyal, after making such comments as "The Yaudheyas worshipped Kärttikeya, the

general of the gods, whose name and figure occur on some of their coins" (1972:21), and "he [Skanda-

Kärttikeya] was a favourite deity of illustrious kings like Huvishka and Kumāragupta and of the Yaudheya

tribe" (1972:194-195), comments in a footnote: "We would have expected the representation of the deity on

the seals of the Yaudheyas whose coins portray his figures, but no such examples are met with" (1972:195).

Such an expectation is only created if we assume, as Thaplyal does, that the class six coins of the

Yaudheyas are a reflection of popular and widespread religious sentiment. I have already argued in relation

to Kusana issues that coinage does not always reflect wide spread 'popular' devotion, and I will not repeat

those arguments here. It does, however, seem clear from the evidence of the seals and sealings from Sunet

that the apparent special devotion many scholars attribute to the Yaudheyas on the basis of their coinage is

not well supported in the case of these class six issues. There is evidence from some seals and sealings

which suggests the existence of a cult to Skanda-Kärttikeya, but that cult must be seen as minor when

compared to cults devoted to Siva and Visnu at this time.

5.3.4 Yaudheya Coins: The Meaning of Skanda-Karttikeya on the Class Six Coins

We might well question why Skanda-Kärttikeya is represented on these Yaudheya class six coins at

all if, as I have argued, his cult has become minor by this period. I think the answer to this is three-fold. I

have suggested that Skanda-Kärttikeya or Brahmanyadeva was popular among the Yaudheyas during the

period of their class three coin production. It may have been the case that part of this group's sense of

identity was tied to the cult of Skanda-Kärttikeya. If Banerjea's argument that the class three coin legends

indicate that this deity was regarded as the spiritual and temporal ruler of the Yaudheyas is correct, then

Skanda-Kärttikeya's position as a sign of the group becomes more important. While the popularity of

Skanda-Kärttikeya may have faded among non-elite sections of society over the period between class three

Yaudheya devotional practice than is currently known.

249

and class six coin production, the deity may still have acted as a sign of the group's sense of identity and

remained a useful reminder of the group's past.

There is also a conservative nature to coinage production that tends to promote consistent uses of

figures and devices at the expense of ingenuity. It requires a significant political power to issue coinage with

new deities on it. Weaker powers, like the Yaudheyas, rarely change the deities on their currency unless

they hope to signal a change in leadership. The Kusana empire is an example of this numismatic theory in

practice. At the beginning of the empire only a small number of deities appeared on coinage. At the height

of the empire under Kaniska and Huviska a large number of deities appear on coins, while during the

empire's decline only Siva and Ardoxso are represented on Kusāna coinage. Only the wealthiest and most

powerful kings are able to risk dramatic changes in currency. The Yaudheyas of the class six period only

produced copper coinage and cannot be described as holding the same wealth and power as the Kusanas,

even during the decline of that empire. Hence, the Yaudheyas do not present us with a model of a state in a

position to change the deities represented on its coinage.

There may also be something about the nature of the Yaudheya government which prevented

changing the figures on the coinage. The nature of Yaudheya leadership is not clear to us, but if they had

kings, the group did not allow those kings a similar position as other states at the time. No king ever appears

on Yaudheya coinage; this is unusual in the ancient world. Shifts in who or what is represented on coinage

were often driven by new and strong kings who wished to present a certain image or signal a significant

change from past rulers. No such powerful individual appears to be present within the Yaudheya context.

This absence may have encouraged the continued use of past symbols for the group. I must acknowledge

that much of what I have presented above is speculative. Just as there are not clear sources to tell us why the

Yaudheyas placed Brahmanyadeva on their class three coinage, there are also few clear sources to tell us

why they place him on their class six coinage.

Having stated this much, however, I would suggest that the appearance of the figures on the class

three and six coins is different. While they may both be recognized as Skanda-Kārttikeya and Sasthi, their

250

character, as displayed through iconography, is vastly different. I have suggested earlier that the appearance

of these class six images resembles the Kusāna images of Mahāsena, Skanda-Kumāra and Viśākha. It is,

then, possible that Skanda-Kärttikeya is employed by the class six Yaudheyas and their leadership for the

same reason Huviska employed him: a military figure is politically useful. I have already suggested that

warrior or military ideals were important to this group. Their leadership may also have been drawn from

military circles, and the Kusana transformation of Skanda-Kārttikeya into Mahasena may have suited the

leadership of the Yaudheyas well.79

What we do know is that far from reacting against the iconographic shifts brought in by foreign

groups, the Yaudheyas embraced these changes. We should not, I think, underestimate the overall impact

that Greek, Scythian and Kusana culture had on the Yaudheyas and indigenous groups in north India in

general. The titles of their leaders, their coinage and how they choose to represent Sasthi and Skanda-

Kärttikeya are all heavily influenced by these foreign sources. The Yaudheyas had exposure to foreign

79There are two seals from this time period which may shed some light on the role of the military in the Yaudheya leadership. The first seal is a controversial one because it is grammatically flawed and difficult to read. The seal reads: "Yaudheya-gana-puraskritasya Šankararāja Mahākshatrapa Mahāsenāpaterindramitra-grīhītasya Mahārāja Mahākshatrapa Senapateraprati-hata asanasya Dharmamitra Nandavarmmanah" (Saraswati 1970:154). The legend has been translated as: "(the seal is of) Mahārāja Mahākshatrapa Senāpati Dharmamitranandavarman, who was made the ruler of the Yaudheya gana (and) adopted by Mahārāja Mahākshatrapa Mahāsenāpati Indramitra (and) whose rule was undisputed" (Shastri 1974:117). This seal appears to give us some idea of how the Yaudheyas appointed a new leader. It seems that the new leader was appointed and then, perhaps, adopted by the leader he is to replace. The seal also suggests that the Yaudheya leader was called a great king (maharaja), which does not seem to fit what we understand of this group or their gana status and is not an element of the seal I completely understand. What the seal does seem to suggest is that the Yaudheya leadership was drawn from its generals (mahāsenāpati/senāpati). Another sealing has been found at both Sunet and Naurangabad which seems to relate to how the Yaudheyas administered their state. The sealing reads: "Yaudheyānām Jaya-mantradharānām" (Handa 1988:131). Ahmad and Nadooshan feel that this seal "likely hints at the existence of the permanent war operational organization formed as the council consisting of the Generals of the tribe", and they read the legend as: "of the war-ministers of the Yaudheyas'" (1993:153). This reading of the sealing and the interpretation of the nature of the Yaudheya administrative structure do not strike me as improbable as the Yaudheyas themselves and other ancient sources seem to focus on their existence as warriors. While the meaning of this sealing is also not completely clear, military officials may have played an important role in governing the Yaudheya gana and the transformation of Brahmanyadeva into a Mahasena-like figure may have been of political use for these leaders, or at least a shift that appealed to them.

251

groups for at least over 400 years by the time of their class six coinage, which display the greatest

borrowing from these groups. We can only speculate as to just how much of this foreign culture the

Yaudheyas assimilated, but the evidence from coinage suggests that the Yaudheyas absorbed a great deal.

252

Chapter 6: Conclusion 6.1 Conclusion

The cult of Skanda-Kärttikeya does not stop developing after the Kusäna and Yaudheya eras. The

Gupta era adds a great deal to the iconography of this deity, and the eventual assimilation of his cult with

the south Indian Murukan' cult during, or just after the Gupta era, also has a significant impact. What will

not change in the north of India is the location of his cult in elite and royal circles and the steady erosion of

his 'popularity' at a more everyday and widespread level. As most scholars acknowledge, his cult does

receive great patronage from the Gupta emperors Kumaragupta and Skandagupta, but I would argue that

this promotion of the deity serves clear political and elite ends and does not reflect a widespread popularity

for the deity. It is simply the continued promotion of a military deity for royal support. By the 7th century

CE, with the disappearance of this royal support, the cult of the deity in the north of India all but disappears.

The remarkable fall of his cult has been a mystery to previous scholars of the deity who see nothing but his

increased popularity from the Kusanas through to the Guptas. What my study demonstrates is that Skanda-

Karttikeya was 'popular', but not with all groups within north Indian society. Understanding the social

location of his cult, or popularity, over time and the motivations behind his worship are essential to

understanding this deity's development over time. His cult declines, I think, because he was transformed

into an auspicious martial deity who was subserviant to Siva. His cult lost its popularity with everyday

people once it was removed from its origins among Graha cults and became the object of political use. By

the 7th century CE the cult had lost so much of its widespread support that, without the support of a royal

group, it could only endure as a minor aspect of Saivism in the north of India.

I have demonstrated that the origins of this deity's cult lie among the inauspicious Grahas. It is

within this context that he receives widespread devotion. People worshiped him out of fear to gain his

See note 6 in section 1.2 for more on Murukan. Also see Clothey (1978) and Zvelebil (1981) for more on this deity.

253

protection for their children. I have argued that the opening birth narrative of the Āranyakaparvan

illustrates this Graha persona of the deity, and this view of the deity is also illustrated by the Pāraskara

Grhyasūtra, the Susrutasamhita and in part by the Skandayaga. I have also argued that this Graha persona

of the deity and its tie to related Mätr cults is the indigenous understanding of the deity. I have used panels

and statuary from Mathura to substantiate this point as well as the six-headed Brahmanyadeva figures from

Yaudheya class three coinage. This version of the deity, I have argued, was widely popular in the pre-

Kusāna age and was also not obviously associated with a particular group in society. He was a terrestrially

focused deity who dealt with the everyday evils of disease and welfare in a world apparently full of dangers.

I have gone on to argue that this initial image of the deity is transformed into an auspicious martial

deity who is also depicted as the son of Siva. I have argued for two groups who bring about this change.

One group is Brahminical and is primarily responsible for the depiction of the deity in the Epic texts and

later textual sources. The main concerns of this group seem to have been placing Skanda-Kärttikeya within

their own religious and textual world, which largely resulted in a dramatic shift in his characterization

towards an auspicious deity with links to Vedic figures and cosmic duties. He becomes an auspicious

general of the army of the gods as well as the dutiful son of Siva, all of which removes him from his initial

inauspicious Graha-like cult base. This shift in characterization separates him from the support of his once

popular ritual cult and with other influences leaves him the focus of a much restricted cult. His cult base is

particularly weaken when he becomes the son of Siva, whose own cult quickly surpassed that of Skanda-

Kärttikeya. Once placed in this subserviant role he becomes a support for the great deeds of Siva and is no

longer allowed a cult that threatens the status of his father. This point is particularly made in final chapters

of the Aranyakaparvan account of his birth and deeds where all of his actions are orchestrated by, and

made subservient to, Siva.

I have also suggested that this Brahminization of the deity in the Epics may have been driven in

part by socio-political concerns. I have argued that the sponsoring kings of the Epic would have preferred

the eventual auspicious, subserviant and martial version of the deity over the dangerous and disorderly

254

Graha-like deity. This royal or political aspect is the second influence in bringing about the shift in Skanda-

Kärttikeya's characterization. I have gone on to argue that royal and political groups did take an active

interest in the cult of Skanda-Kärttikeya, and it is largely through these political groups that the shift from

Graha to Mahāsena seems to have taken place. In particular, I have argued that the Kusānas played a vital

role in changing his characterization towards a martial deity and placing the context of his cult within elite

and royal circles.

Throughout this thesis I have argued against most numismatists, historians and scholars of Skanda-

Kärttikeya who regard certain coin issues of Huviska, the Yaudheyas and others as evidence of this deity's

wide spread 'popularity' in ancient India. I have demonstrated that such a conclusion ignores what we know

from the classical world about the nature and use of coinage as political tools. As I have argued, Skanda-

Karttikeya's personality, as displayed by these coins, shifts from an inauspicious Graha-like god to an elitist

military god. We move from a hideous and frightening image on Yaudheya class three coins to a handsome

and approachable figure on class six coins. The catalyst that seems to bring about this change is foreign

influence. It is the ruling and military elite of Kusana and Yaudheya class six society to whom the images of

Mahäsena and the class six Skanda-Kärttikeya speak. Such a narrowing and politicizing of his support base

was ultimately to leave him as a minor figure. The Yaudheya class six archaeological evidence seems to

suggest that such a process was already underway. My chapters on coinage and statuary also demonstrate, I

think, that the devlopment of Skanda-Karttikeya's cult I have outlined in the Mahabharata and other texts is

based in the lived cult of the deity. The material I have presented above strongly suggests that royal and

elite concerns come to dominate the representation of Skanda-Karttikeya. This, in turn, can help us to

understand the textual accounts of this deity.2

2This need not imply that the Kusanas are the political force responsible for the Mahabharata's account of Skanda-Kärttikeya. Indeed, it is extremely unlikely that the Kusänas played any direct role in the production of this text. The Kusanas are excluded from Hindu texts, and while I have demonstrated above the foreign impact on this cult, there is never any acknowledgement of this foreign presence in the Mahābhārata's accounts of the deity. As we have seen, however, post-Kusāna Indian groups assimilate the Kusāna understanding of this deity. Hence, it could have been any Indian royal group or groups that

255

The transformation undertaken by the Kusanas is achieved by assimilating the character and

iconography of Skanda-Kärttikeya with a number of foreign martial and astrological deities like Ares/Mars,

Verethraghna and Sraoša. His nature as Mahāsena was emphasized over his role as Skandagraha through

this process. The existence of the Yaudheya class six coins made the point that this shift in characterization

influenced indigenous groups, and the evidence from Yaudheya seals and sealings indicates that his

popularity on an everyday level seems to vanish with this shift in characterization. The idea of Skandagraha

never quite leaves the northern tradition, and the popularity of Skanda-Kärttikeya never completely

disappears, but his widespread popularity as a Graha is no longer obvious.

This study suggests a number of things about the emergence of early Hinduism and, in particular,

the emergence of Saivism in the north of India. In the case of Skanda-Kārttikeya his pre-classical and

classical characterizations and cult are dramatically different. It is difficult to see any of the origins of the

deity in the depictions of him from the Ramayana, the Anusasanaparvan of the Mahabharata and the

Śalyaparvan of the Mahabhärata, and impossible in a classical text like the Kumārasambhava. These

dramatic changes are not brought about completely by indigenous Brahminical forces, but owe a great deal

to foreign influence and interest in this cult. While this foreign component is largely ignored by the

Brahminical textual tradition, its impact on the characterization of this deity and the trajectory of his cult

cannot be denied. The impact of foreign influence was largely ignored by previous Skanda-Kārttikeya

scholars, and its discovery and analysis here are among the significant elements of this dissertation. This

study implies that it may well be time to re-examine the whole of Saivite development from pre-Kusāna to

post-Kusāna India, as I suspect a similar conclusion may be reached in other elements of this cult.

promoted the textual representation of Skanda-Kärttikeya as a military deity in the post-Kusāna age.

256

Bibliography

Agrawala, P. K. 1967. Skanda-Kārttikeya [A Study in the Origin and Development]. Varanasi: Banaras Hindu University.

1971. "Identification of the so-called Nägi Figures as Goddess Sasthi" in East and West. Vol. 21, pp. 325-329.

1972. "Gods and Goddesses on Tribal Coins" in Seminar Papers on the Tribal Coins of Ancient India (c.200 B. C. To 400 A. D.). Eds. Jai Prakash Singh and Nisar Ahmad. Varanasi: Banaras Hindu University, pp. 102-106.

1973. "Tetrad of Skanda Deities in a Relief at Bhubaneshwar" in East and West. Vol. 29, pp. 157- 161.

Agrawala, R. C. 1955. "Dramma in Ancient Indian Epigraphs and Literature" in Journal of the Numismatic Society of India, Vol. 17, pp. 64-82.

1956/57. "Skanda-Kārttikeya in Sculptures from Rajasthan" in Lalit Kala, Vol. 3-4, pp. 109-111.

1966a. "An Interesting Relief from the Swat Valley" in East and West, Vol. 16, pp. 82-83.

1966. "Six Headed Skanda on Yaudheya Coins" in Journal of the Numismatic Society of India, Vol. 28, pp. 200-202.

1967. "Six Headed Skanda" in Journal of the Numismatic Society of India, Vol. 29, p. 41.

1967/68. "A Rare Bronze of Skanda Kumara from the Panjab Hills" in Vishvesharand Indological Journal, Vol. 5, pp. 206-208.

1968. "Six Headed Goddess on Yaudheya Coins" in Journal of the Numismatic Society of India, Vol. 30, pp. 181-183.

1969. "Gandhara Skanda with Flames" in East and West, Vol. 18, pp. 163-165.

1970. "Skanda from National Museum, New Delhi and U. P. Hills" in East and West, Vol. 18, pp. 319-322.

1971. "Matrka Reliefs in Early Indian Art" in East and West. Vol. 21, pp. 79-89.

1972. "Skanda with Cock and Peacock: A Rare Device in Gandhara Art" in Eastern Approaches: Essays on Asian Art and Archaeology. Ed. T. S. Maxwell. Delhi: Oxford University Press, pp. 130-132.

1973. "Sasthī in Gandharan Art" in East and West, Vol. 43, pp. 271-275.

1974. "Skanda-Sasthī in Gandharan Reliefs -- A Review" in East and West, Vol. 45, pp. 329-332.

257

Agrawala, V. S. 1943. "Goddess Shashthi on the Yaudheya Coins" in Journal of the Numismatic Society of India, Vol. 5, pp. 29-32.

1953. "A Kushana Coin-Mould from Rohtak" in Journal of the Numismatic Society of India, Vol. 15, pp. 68-69.

----- 1961. "Fire in the Ridveda" in East and West, Vol.11, pp. 28-32.

1963. (second edition), India as Known to Panini. Varanasi: Prithvi Prakashan.

1964. Ancient Indian Folk Cults. Varanasi: Prithivi Prakashan.

Ahmad, N. 1977. "The Kumaras -- A Forgotten Ancient Indian Tribe" in Seminar Papers on the Tribal Coins of Ancient India (c. 200 B.C. to 400 A.D.). eds. Jai Prakash Singh and Nisar Ahmad. Varanasi: Banaras Hindu University, pp. 154-192.

1977. "The Mitras of the Punjab", in Journal of the Numismatic Society of India, Vol. 39, pp. 52- 68.

1978. "Two Local Rulers of Garhwal (U.P.)" in Jounal of the Numismatic Society of India, Vol. 47, pp. 106-112.

1979. "The Coins of the Regel Titled Janapadas" in Jounal of the Numismatic Society of India, Vol. 50, pp. 64-70.

and Kumar, M. 1993. "The Reminted Type IV Coins of the Yaudheyas" in Numismatic Studies, Vol. 3, pp. 49-54.

and Nadooshan, F. K. 1993. "A Lead Signet of the Yaudheyas and Reappraisal of their Certain Published Seals" in Jounal of the Numismatic Society of India, Vol. 55, pp. 152-154.

Allan, J. 1936. Catalogue of the Coins of Ancient India. London: Trustees of the British Museum.

Altekar, A. S. 1949. "Some Rare and Unique Coins in the Prince of Wales Museum, Bombay" in Journal of the Numismatic Society of India. Vol. 11, pp. 44-63.

[1949]. 1962. State and Government in Ancient India. Delhi: Motilal Banarsidass.

Alram, M. 1999. "Indo-Parthian and early Kushan chronology: the numismatic evidence" in Coins, Art, and Chronology. Essays on the pre-Islamic History of the Indo-Iranian Borderlands. Eds. M. Alram and D. E. Klimburg-Salter. Wien: Verlag der Österreichischen Akademie der Wissenschaften, pp. 19-48.

Aravamuthan, T. G. 1948. "More Gods of Harappa" in Journal of the Bihar Research Society. Vol. 34, pp. 31-82

Atharvavediya Paippalādasamhitā. 1964. Ed. D. Bhattacharya. Calcutta: Sanskrit College.

258

Atharvavedapariśista. Chaukhambha Prachyavidya Granthamala. No. 1. 1976. Eds. George Melville Bolling, Julius von Negelein and Ram Kumar Rai. Varanasi: Chaukhambha Orientalia.

Bajpai, K. D. 1973. "Yaudheya Coins -- A Critical Study" in Journal of the Numismatic Society of India. Vol. 35, pp. 90-94.

Bakker, H. 1999. "Book Review of Many Heads, Arms and Eyes: Origin, Meaning and Form of Multiplicity in Indian Art, by D. M. Srinivasan" in Artibus Asiae, Vol. 58, pp. 339-342.

Bandyopadhyay, P. 1986. "Did Kalidasa Complete the Kumarasambhava?" in Journal of the American Oriental Society. Vol. 106, pp. 559-564.

Banerjea, J. N. 1951. "The Reading and Interpretation of Some Yaudheya Coin Legends" in Journal of the Numismatic Society of India, Vol. 13, pp. 160-163.

1956. The Development of Hindu Iconography. Calcutta: University of Calcutta.

1957. "Interpretation of a few symbols on some tribal coins of ancient India" in Journal of the Numismatic Society of India, Vol. 22, pp. 43-46.

Bedekar, V. M. 1975. "Kārttikeya (Skanda) in Sanskrit Literature, with Special Reference to the Mahabharata: From A Folk Spirit to the Chief War-God" in Annals of the Bhandarkar Oriental Research Institute, Vol. 56, pp.141-177.

Bhide, V. V. 1967. "Cock in Vedic Literature" in Bhāratīya Vidyā. Vol. 27, pp. 1-6.

Bivar, A. D. H. 1970. "Haritī and the chronology of the Kusanas" in Bulletin of the School of Oriental and African Studies, Vol. 39, pp. 333-40.

Bloomfield, M. (Translator) [1897]. 1964. Hymns of the Atharva-veda together with Extracts from the Ritual Books and the Commentaries. Sacred Books of the East, vol. XLII, ed. F. M. Müller. Delhi: Motilal Banarsidass.

Bodewitz, H. W. 1976. The Daily Evening and Morning Offering (Agnihotra) According to the Brahmanas. Leiden: E. J. Brill.

Bopearachchi, O., Pieper, W. 1998. Ancient Indian Coins. Turnhout: Brepols.

Boyce, M. 1975. A History of Zoroastrianism. Vol. 3. Leiden: Brill.

van Buitenen, J. A. B. (Translator) 1975. The Mahabhārata. Vols. 1-3. Chicago: The University of Chicago Press.

Burkert, W. 1985. Greek Religion. Trans. John Raffan. Cambridge: Harvard Univeristy Press.

Callieri, P. 1988. "Review of Robert Göbl, System und Chronologie der Münzprägung des Kušānreiches", in East and West, Vol. 38, pp. 328-330.

259

1997. Seals and Sealings from the North-West of the Indian Subcontinent and Afghanistan (4th Century BC - 11th Century AD). Local, Indian, Sasanian, Graeco-Persian, Sogdian, Roman. Naples: Istituto Universitario Orientale Istituto Italiano Per L'Africa e L'Oriente.

Chakravarty, B. 1987. "A Study in the Chemical Compositions of Copper-Based Yaudheya Coins", in Journal of the Numismatic Society of India, Vol. 49, pp. 148-152.

Chatterjee, A. K. 1970. The Cult of Skanda-Kārttikeya in Ancient India. Calcutta: Punthi Pustak.

Chattopadhyay, B. 1967. The Age of the Kushanas -- A Numismatic Study. Calcutta: Punthi Pustak.

1977. Coins and Icons, A Study of Myths and Symbols in Indian Numismatic Art. Calcutta: Punthi Pustak.

Chaturvedi, S. N., Pathak, V. S. (Unpublished paper), "The Three-headed Divinity in Literature and Archaeology", manuscript held at the IIRNS box xliiiA/36.

Clothey, F. W. 1978. The Many Faces of Murukan: The History and Meaning of a South Indian God. The Hague: Mouton Publishers.

Couture, A., Schmid, C. 2001. "The Harivamsa, The Goddess Ekänamsa, and the Iconography of the Vrsni Triads" in Journal of the American Oriental Society. Vol. 121, pp. 173-192.

Cribb, J. 1983. "Kanishka's Buddha Coins" in Numismatic Digest, Vol. 7, pp. 44-53.

1984. "The Origin of the Buddha Image -- The Numismatic Evidence" in South Asian Archaeology 1981. Proceedings of the Sixth International Conference of the Association of South Asian Archaeologists in Western Europe, 5-10 July 1981. Ed. Bridget Allchin. Cambridge: Cambridge University Press, pp. 231-244.

1985. "A Re-Examination of the Buddha Images on the Coins of King Kaniska: New Light on the Origins of the Buddha Image in Gandharan Art" in Studies in Buddhist Art of South Asia. Ed. A. K. Narain, pp. 59-87.

and Sims-Williams, N. 1995/96. "A New Bactrian Inscription of Kanishka the Great" in Silk Road Art and Archaeology, Vol. 4, pp. 75-142.

----- 1997. "Shiva images on Kushan and Kushano-Sasanian coins" in Studies in Silk Road Coins and Culture: Papers in honour of Professor Ikuo Hirayama on his 65th birthday. Eds. Katsumi Tanabe, Joe Cribb and Helen Wang. Kamakura: The Institute of Silk Road Studies, pp. 11-66.

1999. "The early Kushan kings: new evidence for chronology. Evidence from the Rabatak Inscription of Kanishka I" in Coins, Art, and Chronology. Essays on the pre-Islamic History of the Indo-Iranian Borderlands. Eds. M. Alram and D. E. Klimburg-Salter. Wien: Verlag der Österreichischen Akademie der Wissenschaften, pp. 177-205.

Cunningham, A. [1899]. 1963. Coins of Ancient India from the Earliest Times Down to the Seventh Century A.D. Varanasi: Indological Book House.

260

Dasgupta, K. K. 1974. A Tribal History of Ancient India. Calcutta: Nababharat Publishers.

Dayal, R. B. P. 1940. "A New Hoard of Yaudheya Coins from Dehra Dun District" in Journal of the Numismatic Society of India, Vol. 2, pp. 110-112.

Doniger O'Flaherty, W. 1973. Asceticism and Eroticism in the Mythology of Siva. London: Oxford University Press.

1980. Women, Androgynes and Other Mythical Beasts. Chicago: The University of Chicago Press.

(Translator) [1981]. 1983. The Rig Veda. An Anthology. Suffolk: Penguin Books.

1985. Tales of Sex and Violence: Folklore, Sacrifice, and Danger in the Jaiminiya Brahmana. Chicago: University of Chicago Press.

Errington, E., Cribb, J., Claringbull, M. Eds. 1992. The Crossroads of Asia: Transformation in Image and Symbol in the Art of Ancient Afghanistan and Pakistan. Cambridge: The Ancient India and Iran Trust.

Farnell, L. R. [1921] 1995. Greek Hero Cults and Ideas of Immortality. Chicago: Ares Publishing.

Fleet, J. F. 1970. Corpus Inscriptionum Indicarum: Inscriptions of the Early Gupta Kings and their Successors. Vol. 3. Varanasi: Indological Book House.

Gardner, P. 1886. Catalogue of Indian Coins in The British Museum. Greek and Scythic Kings of Bactria and India. London: Trustees of the British Museum.

Geldner, K. F. (Translator) 1951. Der Rig-Veda. Aus dem Sanskrit ins Deutsche Übersetzt und mit einem Laufenden Kommentar Versehen. Vols. 1-3. Cambridge, Massachusetts: Harvard University Press.

Ghurye, G. S. 1977. Indian Acculturation: Agastya and Skanda. Bombay: Popular Prakashan.

Gitomer, D. 1992. "King Duryodhana: The Mahabharata Discourse of Sinning and Virtue in Epic and Drama" in Journal of the American Oriental Society. Vol. 112, pp. 222-232.

Gnoli, G. 1963. "The Tyche and the Dioscuri in Ancient Sculptures from the Valley of Swat" in East and West. Vol. 14, p. 29-37.

1992. "The Tyche and the Dioscuri in Gandharan Sculptures: A Retrospect" in Eastern Approaches: Essays on Asian Art and Archaeology. Ed. T. S. Maxwell. Delhi: Oxford University Press, pp. 30-37.

Göbl, R. 1960. "Roman Patterns for Kushana Coins" in Journal of the Numismatic Society of India, p. 76 ff.

1984. System und Chronologie der Münzprägung des Kušānreiches. Wien: Verlag der Österreichischen Akademie der Wissenschaften.

261

1999. "The Rabatak Inscription and the date of Kanishka" in Coins, Art, and Chronology. Essays on the pre-Islamic History of the Indo-Iranian Borderlands. Eds. M. Alram and D. E. Klimburg- Salter. Wien: Verlag der Österreichischen Akademie der Wissenschaften, pp. 151-171.

Goetz, H. 1948. "Iranian or Buddhist Deity? A Gandhära Statuette and its Problems" in Bulletin of the Baroda State Museum and Picture Gallery. Vol. 3, pp. 17-22.

Goldman, R. P. 1977. Gods, Priests, and Warriors: The Bhrgus of the Mahabharata. New York: Columbia University Press.

----- (Translator) [1984] 1990. The Ramayana of Valmiki: An Epic of Ancient India. Vol. 1, Bālakānda. Princeton: Princeton University Press.

Gonda, J. 1957. Some Observations on the Relations between 'Gods' and 'Powers' in the Veda, A Propos of the Phrase Sunuh Sahasah. The Hague: Mouton and Co.

1959. Epithets in the Rgveda. The Hague: Mouton and Co.

1960. The Dual Deities in the Religion of the Veda. Amsterdam: North-Holland Publishing.

1961. Vedic Literature (Samhitas and Brahmanas). A History of Indian Literature. Vol. 1, Fasc. 1. Wiesbaden: Otto Harrassowitz.

1962. The Ritual Sūtras. A History of Indian Literature. Vol. 1, Fasc. 2. Wiesbaden: Otto Harrassowitz.

1963. Vedic Ritual: The Non-Solemn Rites. Leiden-Köln: E. J. Brill

1964. The Functions and Significance of Gold in the Veda. Leiden: E. J. Brill.

Granoff, P. 1970. "Tobatsu Bishamon: Three Japanese Statues in the United States and an Outline of the Rise of This Cult in East Asia" in East and West, Vol. 20, pp. 144-167.

----- 2002. "Paradigms of Protection in early Indian religious texts, or an essay on what to do with your demons" in Essays in Jaina Philosophy and Religion. Ed. Piotr Balcerowicz. Delhi - Warszawa, pp. 97-128.

Forthcoming. "Images and their Ritual Use in Medieval India: Hestitations and Contraditions" in Images in Asian Religion: Texts and Contexts.

Guillaume, O. 1990. Analysis of Reasonings in Archaeology: The Case of Graeco-Bactrian and Indo- Greek Numismatics. Trans. O. Bopearachchi. Delhi: Oxford University Press.

Gupta, L. C. 1985. "Some Copper Coins from Tehri-Garhwal" in Journal of the Numismatic Society of India, Vol. 47, pp. 14-17.

1986. "Some Rare Kuninda Coins" in Journal of the Numismatic Society of India, Vol. 48, pp. 36- 40.

262

Gupta, P. L. 1977. "Geographical Distribution of the Tribal Coins" in Seminar Papers on the Tribal Coins of Ancient India (c. 200 B.C. to 400 A.D.). eds. Jai Prakash Singh and Nisar Ahmad. Varanasi: Banaras Hindu University, pp. 49-69.

1987. "The Brahmanya Coins" in Numismatic Digest, Vol. 11, pp. 28-39.

Gupta, S. M. 1988. Karttikeya the Son of Shiva. Bombay: Somaiya Publications.

Handa, D. 1978/79. "Sectional Yaudheya Coins" in The Journal of the Academy of Indian Numismatics and Sigillography. Vol. 2, pp. 30-33.

1982. "A Note on Re-struck and Re-used Kushana and Yaudheya Coins" in Journal of the Numismatic Society of India, Vol.44, pp. 46-47.

1983. "The Bearing of the Yaudheya Coinmoulds from Sunet on Sectional Yaudheya Coins" in Panjab University Research Bulletin (Arts). Vol. 14, pp. 3-9.

1984. "Seals and Sealings from Punjab -- A Study" in Panjab University Research Bulletin (Arts), Vol. 15, pp. 143-184.

1985. "Seals and Sealings from Punjab -- A Study" in Studies in Indian Coins and Seals. Ed. D. Handa. Delhi: Sundeep Prakashan, pp. 93-138.

--- 1987. "Some Ancient Cities and Towns of Punjab" in Indological Studies (Essays in Memory of Shri S. P. Singhal). Ed. D. Handa. Delhi: Caxton Publications, pp. 303-345.

1988. "The Yaudheya Sealings -- A Review" in Journal of the Numismatic Society of India, Vol.50, pp. 130-133.

1989. "The Yaudheyas and their Coins" in Numismatic Studies, Vol. I, pp. 63-78.

Härtel, H. 1987. "Die Kusana -- Göttin Sasthi" in Hinduismus und Buddhismus: Festschrift für Ulrich Schneider. Ed. Harry Falk. Freiburg: Hedwig Falk, pp. 152-162.

1993. Excavations at Sonkh. 2500 Years of a Town in Mathura District. With contributions by Hans-Jürgen Paech and Rolf Weber. Berlin: Dietrich Reimer Verlag.

Hazra, R. C. 1958. Studies in the Upapuranas. 2 volumes. Calcutta: Sanskrit College.

Hermary, A. 1986. "Ares/Mars" in Lexicon Iconographicum Mythologiae Classicae. Eds. ?, Vol. ?, pp. 505-580.

Hiltebeitel, A. 1976. The Ritual of Battle: Krishna in the Mahabharata. Ithaca: Cornell University Press.

----- 2001. Rethinking the Mahabhrata: A Reader's Guide to the Education of the Dharma King. Chicago:The University of Chicago Press.

Holt, F. L. 1999. Thundering Zeus: The Making of Hellenistic Bactria. Berkeley: University of California Press.

263

Hopkins, E. W. [1901]. 1978. The Great Epic of India. Its Character and Origin. Calcutta: Punthi Pustak.

Jayaswal, V. 1991. Kushana Clay Art of Ganga Plains: A Case Study of Human Forms from Khairadih. Delhi: Agam Kala Prakashan.

Joshi, N. P. 1970. "Kārtikeya's Sakti as Ayudha-Devī" in Viśveśarand Indological Journal. Vol. 8, pp. 214- 215.

1986. Mātrkas Mothers in Kusana Art. New Delhi: Kanak Publications.

1987. "Mātrkā Figures in Kusāna Sculptures at Mathura" in Investigating Indian Art: Proceedings of a Symposium on the Development of Early Buddhist and Hindu Iconography Held at the Museum of Indian Art Berlin in May 1986. Eds. Marianne Yaldiz and Wibke Lobo. Berlin: Staatliche Museen Preussischer Kulturbesitz, pp. 159-172.

Kala, S. C. 1956. "Some Interesting Coins of the Yaudheyas" in Journal of the Numismatic Society of India. Vol. 18, pp. 46-48.

Kangle, R. P. (Translator) 1960. The Arthaśastra: A critical edition with a glossary. Bombay: University of Bombay.

Kar, R. C. 1954. "An Early Image of Karttikeya from Taxila" in Indian Historical Quarterly. Vol. 30, pp. 81-85.

Kāšyapa Samhitā, Vrddhajīvakīyam tantram vā. 1938. Ed. T. Yādavaśarma. Bombay: Jadavji Trikamji Acharya.

Kaye, G. R. [1924]. 1998. Memoirs of the Archaeological Survey of India, No. 18, Hindu Astronomy. New Delhi: The Director General Archaeological Survey of India.

Klimburg-Salter, D. E. 1999. "From an art historical perspective: problems of chronology in the Kusāna period" in Coins, Art, and Chronology. Essays on the pre-Islamic History of the Indo-Iranian Borderlands. Eds. M. Alram and D. E. Klimburg-Salter. Wien: Verlag der Österreichischen Akademie der Wissenschaften, pp. 3-18.

Kreyenbroek, G. 1985. Sraoša in the Zoroastrian Tradition. Orientalia Rheno-Traiectina. Ed. J. Gonda. Leiden: E. J. Brill.

Kumar, M., Singh, S. B. 1978. "A Note on Temple Architecture as Gleaned from Coins" in Kurukshetra University Research Journal (Arts and Humanities), Vol. 12, pp. 1-5.

Sharma, G. B. 1979. "Mitra Coins from Ghuram (Patiala)" in Numismatic Digest, Vol. 3, pp. 4-16.

1991. "Yaudheya Coin Hoard from Bishan (Rohtak)" in Numismatic Studies, Vol. 1, pp. 79-98.

----- 1996. "Ancient Mint at Rohtak" in Numismatic Panorama: Essays in Memory of Late Shri S. M. Shukla. Eds. K. K. Maheshwari and B. Rath. New Delhi: Harnam, pp. 95-114.

264

Lahiri, B. 1974. Indigenous States of Northern India (Circa 200 B. C. To 320 A. D.). Calcutta: University of Calcutta.

MacDonell, A. A. 1963. The Vedic Mythology. Varanasi: Indological Book House.

MacDowell, D. W. 1987. "Impact of Alexander the Great on the Coinage of Afghanistan and North West India" in Numismatic Digest, Vol. 11, pp. 5-12.

Mahabharata. For the First Time Critically Edited. 1927-1959. Eds. V. S. Sukthankar and S. K. Belvalkar. Poona: Bhandarkar Oriental Research Institute

Maitrāyanī Samhitā: die samhitā der Maitrāyanīya-śākhā. 1970-72. Ed. L. von Schroeder. Wiesbaden: F. Steiner.

Mani, V. R. 1990. Sons of Siva: A Study in the Religious Cults of Ganesa and Kärttikeya. Delhi: Sharada Prakashan.

Malamoud, C. 1998. "Bricks and Words: Observations on the Bodies of the Gods in Vedic India" in Cooking the World: Ritual and Thought in Ancient India. Trans. David White. Delhi: Oxford University Press, pp. 207-225.

Mazumdar, B. P. 1969. "Political Theory and Practice in the Malava and Yaudheya Republics" in Journal of Indian History, Vol. 47, pp. 303-311.

Meyer, J. J. [1926]. 1977. Das Altindische Buch Vom Welt - und Staatsleben, Das Arthaçästra Des Kautilya. Graz: Akademische Druck - u. Verlagsanstalt.

Meulenbeld, G. J. 1999-2002. A History of Indian Medical Literature. Groningen Oriental Studies. Volume XV. Ed. H. T. Bakker. Vols. 1-3. Groningen: Egbert Forsten.

Mitchiner, M. 1975/76. Indo-Greek and Indo-Scythian Coinage. Vols.1-5. London: Hawkins Publications.

Monier-Williams, M. [1899]. 1999. A Sanskrit-English Dictionary. Delhi: Motilal Banarsidass.

Mukherjee, B. N. 1997/98. "Vima Taktu, An Alleged Kushana King" in Numismatic Digest, Vol. 21/22, pp. 5-10.

Mukherjee, D. 1983. "A Note on Some New Coins of the Audumbaras and the Yaudheyas" in Journal of the Numismatic Society of India, Vol. 45, pp. 20-22.

1986. "Glimpses into the Temple Architecture on some Early Indigenous Coins of Northern India" in Rupanjali, in Memory of O. C. Gangoly. Eds. K. K. Ganguly and S. S. Biswas. Calcutta: O. C. Gangoly Memorial Society, pp. 153-158.

Munshi, A. C. D. 1986. "Indian Pantheons and Experiences of Huvishka" in Journal of the Numismatic Society of India, Vol. 48, pp. 91-103.

Narain, A. K. 1957. The Indo-Greeks. Oxford: Clarendon Press.

265

[1955]. 1968. The Coin Types of the Indo-Greek Kings 256-54 B.C. Chicago: Argonaut.

1973. "The Two Hindu Divinities on the Coins of Agathocles from Ai-Khanum" in Journal of the Numismatic Society of India, Vol. 35, pp. 73-77.

1974. "Ancient Mathura and the Numismatic Material" in Mathura The Cultural Heritage. Ed. D. M. Srinivasan. New Delhi: American Institute of Indian Studies.

Navaratnam, R. 1973. Karttikeya: The Divine Child. Bombay: Bharatiya Vidya Bhavan.

van Nooten, B. A. And Holland, G. B. (Eds.) 1994. Rig Veda. A Metrically Restored Text with an Introduction and Notes. Cambridge, Massachusetts: The Department of Sanskrit and Indian Studies, Harvard University.

Olivelle, P. (Translator) 1996. Upanisads. New York: Oxford University Press.

(Translator) 1999. Dharmasūtras: The Law Codes of Ancient India. New York: Oxford University Press.

Pal, P., Bhattacharyya, D. C. 1969. The Astral Divinities of Nepal. Varanasi: Prithivi Prakashan.

1977. "Dhanada-Kubera of the Vishnudharmottara Purāna and some images from North-West India" in Lalit Kalā, Vol. 18, pp. 13-26.

Pandey, D. 1967. "Six-headed Krittika on Yaudheya Coins" in Journal of the Numismatic Society of India, Vol. 29, pp. 5-8.

Pandey, R. [1969]. 1998. Hindu Samskāras: Socio-Religious Study of the Hindu Sacraments. Delhi: Motilal Banarsidass.

Paraskara-Grihyasutra, with the commentaries of Karkopadhyaya, Jayaramā, Hariharācharya, Gadadhara and Dikshita. 1895. Kāshi: Medical Hall Press.

Parpola, A. 1990. "Bangles, Sacred Trees and Fertility Interpretations of the Indus script relating to the cult of Skanda-Kumara" in South Asian Archaeology 1987. Vol. 1. Ed. Maurizio Taddei with Pierfrancesco Callieri. Rome: Istituto Italiano Per Il Medio Ed Estremo Oriente.

Paul, P. G., Paul, D. 1989. "Brahmanical Imagery in the Kusana Art of Mathura: Tradition and Innovations", in East and West, Vol. 39, pp. 111-143.

Peri, N. 1917. "Hāritī la mère-de-démons", Bulletin de l'École Française d'Extrême-Orient, Vol. 17, pp. 1- 102.

Pingree, D. 1964/65. "Representation of the Planets in Indian Astrology" in Indo-Iranian Journal, Vol. 8, pp. 249-267.

(Editor and Translator) 1978. The Yavanajataka of Sphujidhvaja. Harvard Oriental Series. Ed. Daniel H. H. Ingalls. Volume 48. Two Volumes. Cambridge Mass .: Harvard University Press.

266

Poulsen, B. 1991. "The Dioscuri and Ruler Ideology" in Symbolae Osloenses. Vol. 66, pp. 119-146.

1992. "Cult, Myth and Politics" in The Temple of Castor and Pollux. Eds. Inge Nielsen and Birte Poulsen. Rome: Edizioni de Luca, pp. 46-53.

Prakash, J. 1965. "Observations on the Tribal Coinages of Ancient India" in Journal of the Numismatic Society of India, Vol. 27, pp. 123-141.

Rana, S. S. 1995. A Study of Skanda Cult. Delhi: Nag Publishers.

Rao, T. A. G. [1914]. 1968. Elements of Hindu Iconography. 2 volumes. New York: Paragon Book Reprint.

Rao, U. S. 1962. "The Yaudheya Coins" in Journal of the Numismatic Society of India, Vol. 24, pp. 138- 139.

Rapin, C. 1990. "Greeks in Afghanistan: Aï Khanum" in Greek Colonists and Native Populations: Proceedings of the First Australian Congress of Classical Archaeology held in honour of Emeritus Professor A. D. Trendall. Ed. Jean-Paul Descæudres. Canberra (Humanities Research Centre): Clarendon Press (Oxford), pp. 329-342.

Roy, B. P. 1984. "Bearing of the Mahabharata on the Yaudheya Coins" in Journal of the Numismatic Society of India, Vol. 46, pp. 75-77.

Rosenfield, J. M. [1967]. 1993. The Dynastic Arts of the Kushans. New Delhi: Munshiram Manoharlal.

Saraswati, Swami, O. 1970. "Yaudheyagana kī kuch mahatnapūrn muharem" in Journal of the Numismatic Society of India, Vol. 32, pp. 154-157.

Scullard, H. H. 1981. Festivals and Ceremonies of the Roman Republic. London: Thames and Hudson.

Sen, S. 1950. "Iranian Śraoša and Indian Skanda" in Indo-Iranica, Vol. 4, pp. 27.

Sharan, M. K. 1972. Tribal Coins -- A Study (The Yaudheyas, the Malavas, the Audumbaras and the Kunindas). New Delhi: Abhinav Publications.

1982. "Coins of the Audumbaras" in Visvesarand Indological Journal, Vol. 20, pp. 188-196.

Sharma, B. N. 1970. "An Image of Skanda-Kārttikeya from the Panjab Hills" in Viśveśarand Indological Journal, Vol. 8, pp. 211-213.

Shastri, A. M. 1968. "Religious Study of the Symbols on the Local Coins of Central India" in Seminar Papers on the Local Coins of Northern India c. 300 B.C. to 300 A.D. Editors A. K. Narain, J. P. Singh and N. Ahmad. Varanasi: Banaras Hindu University.

1974. "Note of a Yaudheya Seal" in Journal of the Numismatic Society of India, Vol. 36, pp. 114- 118.

267

1976. "Notes on Some Coins and Seals" in Journal of the Numismatic Society of India, Vol. 38, pp. 115-120.

1977. "Symbols on Tribal Coins: An Interpretative Study" in Seminar Papers on the Tribal Coins of Ancient India (c. 200 B.C. to 400 A.D.). eds. Jai Prakash Singh and Nisar Ahmad. Varanasi: Banaras Hindu University, pp. 87-101.

1978. "Interesting Kuninda and Yaudheya Coins" in Numismatic Digest, Vol. 8, pp. 24-36.

1979. "Some Observations on the 'Mandi' Hoard of Yaudheya and Kuninda Coins" in Numismatic Digest, Vol. 11, pp. 40-43.

----- 1993. "Contemporary Forgeries of Yaudheya Coins" in Journal of the Numismatic Society of India, Vol. 55, pp. 100-108.

1995. "Contemporary Forgeries of Yaudheya Coins" in Numismatic Studies, Vol. 4, pp. 95-110.

1996. "Skanda-Kumara on Huvishka Coins" in Numismatic Studies, Vol. 5, pp. 47-51.

Shulman, D. 1980. Tamil Temple Myths: Sacrifice and Divine Marriage in the South Indian Saiva Tradition. Princeton: Princeton University Press.

1992. "Devana and Daiva" in Ritual, State and History in South Asia: Essays in Honour of J. C. Heesterman. Eds. A. W. Van Der Hock, D. H. A. Kilff, M. S. Oort. Leiden: E. J. Brill, pp. 350- 365.

Singh, J. P. 1977. "Coinage of the Tribal Republics" in Seminar Papers on the Tribal Coins of Ancient India (c. 200 B.C. to 400 A.D.). eds. Jai Prakash Singh and Nisar Ahmad. Varanasi: Banaras Hindu University, pp. 1-20.

Singh, O. P. 1977. "Some Syncretic Gleanings from the Coins of the Kunindas and the Yaudheyas" in Seminar Papers on the Tribal Coins of Ancient India (c. 200 B.C. to 400 A.D.). eds. Jai Prakash Singh and Nisar Ahmad. Varanasi: Banaras Hindu University, pp. 110-111.

1978. Religion and Iconography on Early Indian Coins. Varanasi: Bharati Prakashan.

1979. "Ornaments as Depicted on Early Indian Coins" in Journal of the Numismatic Society of India, Vol. 43, pp. 121-126.

Sinha, K. 1979. Kārttikeya in Indian Art and Literature. Delhi: Sundeep Prakashan.

Sircar, D. C. 1965. (second edition), Select Inscriptions Bearing on Indian History and Civilization. Vol. 1. Calcutta: University of Calcutta.

Smith, J. Z. 1995. "Trading Places" in Ancient Magic and Ritual Power. Eds. Marvin Meyer and Paul Mirecki. Leiden: Brill, pp. 13-27.

Smith, M. C. 1992. The Warrior Code of India's Sacred Song. New York: Garland.

268

Smith, V. A. [1906]. 1972. Coins of ancient India; catalogue of the coins in the Indian Museum Calcutta. Vol. 1. Delhi: Indological Book House.

Srinivasan, D. M. 1997. Many Heads, Arms and Eyes: Origin, Meaning and Form of Multiplicity in Indian Art. Leiden: Brill.

1997/98. "Skanda/Karttikeya in the Early Art of the Northwest" in Silk Road Art and Archaeology. Vol. 5, pp. 233-264.

Strickmann, M. 2002. Chinese Magical Medicine. Ed. Bernard Faure. Stanford: Stanford University Press.

Susrutasamhita of Susruta with the Nibandhasangraha Commentary of Sri Dalhanacharya and the Nyāyachandrikā Pañjikā of Šrī Gayadāsāchārya on Nidānasthāna. 1938. (Revised third edition) Eds. Vaidya Jādavji Trikamji Āchārya and Nārāyan Rām Āchārya. Bombay: Nirnaya Sāgar.

Sutton, N. 1997. "Asoka and Yudhisthira: A historical setting for the ideological tensions if the Mahabharata?" in Religion, Vol. 27, pp. 331-41.

Taddei, M. 1964. Sculptures from the Sacred Area of Butkara I (Swat, W. Pakistan). Vol. II, Part 3. Rome: Istituto Poligrafico dello Stato.

1966. "An Interesting Relief from the Swat Valley" in East and West. Vol. 16, pp. 84-88.

Tarn, W. W. 1938. The Greeks in Bactria and India. Cambridge: Cambridge University Press.

Tewari, P. V. (Translator) 1996. Kāśyapa-Samhitā or Vrddhajīvakīya Tantra (Text with English Translation and Commentary). With N. Kumar, R. D. Sharma and A. Kumar. Varanasi: Chaukhambha Visvabharati.

Thakur, U. 1974. "Kārttikeya in Literature, Art and Coins" in East and West. Vol. 24, pp. 297-310.

1981. On Kärttikeya. Varanasi: Chaukhambha Orientalia.

Thapar, R. [1984]. 2000. From Lineage to State: Social Formations in the Mid-First Millennium B.C. in the Ganga Valley. New Delhi: Oxford University Press.

Thaplyal, K. K. 1972. Studies in Ancient Indian Seals. A Study of North Indian Seals and Sealings from circa Third Century B.C. to Mid-Seventh Century A.D. Lucknow: Akhila Bharatiya Sanskrit Parishad.

Volchok, B. Y. 1970. "Towards an Interpretation of Proto-Indian Pictures" in Journal of Tamil Studies. Vol. 2, pp. 29-53.

Whitney, W. D. (Translator) 1962. Atharva-veda Samhitā. Vols. I and II. Delhi: Motilal Banarsidass.

Wujastyk, D. 1999. "Miscarriages of justice: demonic vengeance in classical Indian medicine", in Religion, health, and suffering. Eds. John Hinnells and Roy Porter. London: Kegan Paul International, pp. 256-75.

269

[1998]. 2001. The Roots of Ayurveda. Selections from Sanskrit Medical Writings. New Delhi: Penguin Books India.

Zvelebil, K. V. 1991. Tamil Traditions on Subrahmanya-Murugan. Madras: Institute of Asian Studies.

Zwalf, W. 1996. A Catalogue of the Gandhara Sculpture in the British Museum. Vols. I and II. London: British Museum Press.

Zysk, K. G. 1991. Asceticism and Healing in Ancient India. Medicine in the Buddhist Monastery. New York: Oxford University Press.

[1985]. 1998. Medicine in the Veda. Delhi: Motilal Banarsidass.

270

Illustrations Figure 1

Figure 2

271

Figure 3 obverse a obverse b

reverse a reverse b

Figure 4 obverse a obverse b

Figure 5 obverse reverse

272

Figure 6

obverse a obverse b

reverse a reverse b

273

Figure 7

274

Figure 8 obverse obverse obverse (Aryamitra) (Vijayamitra) (Devamitra)

reverse reverse reverse

Figure 9 obverse reverse

Figure 10

obverse reverse

275

Figure 11 obverse reverse

Figure 12 obverse reverse

Figure 13 obverse reverse

276

Figure 14 obverse reverse

Figure 15 a obverse reverse

Figure 15 b obverse reverse

277

Figure 16

obverse reverse

Figure 17 obverse reverse

Figure 18 obverse reverse

278

Figure 19 obverse a obverse b obverse c

reverse a reverse b reverse c

Figure 20 obverse reverse

Figure 21 obverse reverse

279

Figure 22 reverse

Figure 23 obverse reverse

Figure 24 obverse reverse

Figure 25 obverse reverse

280

Figure 26 obverse reverse

Figure 27 obverse reverse

Figure 28 obverse reverse (not available)

Figure 29 obverse reverse

281

Figure 30 obverse reverse (not available)

Figure 31 obverse reverse

Figure 32 obverse reverse

282

Figure 33 obverse reverse

Figure 34 obverse reverse

Figure 35 obverse reverse

Figure 36 obverse reverse

283

Figure 37 obverse reverse

Figure 38 obverse reverse

Figure 39 obverse reverse

284

Figure 40

obverse a obverse b

reverse a reverse b

Figure 41

285

Figure 42 Figure 43

286

Figure 44 Figure 45

287

Figure 46

288

Figure 47

Figure 48

289

Figure 49

290

Figure 50

291

Figure 51 obverse a obverse b

reverse a reverse b

Figure 52

obverse reverse

292

Figure 53 obverse reverse

Figure 54

obverse reverse

Figure 55 obverse a obverse b

reverse a reverse b

293

Figure 56 a 24

Figure 56 b

28

294

Figure 57

295

Figure 58

296

Figure 59 obverse reverse

Figure 60 Figure 62 obverse obverse a obverse b obverse C

reverse reverse a reverse b reverse c

297

Figure 61

obverse reverse

Figure 63 Figure 64 obverse obverse a obverse b (not available)

reverse reverse a reverse b

Figure 65

298

Figure 66 obverse a obverse b

reverse a reverse b

Figure 67 obverse reverse

299

Figure 68

300

Figure 69a

Figure 69b

301

Figure 70 Figure 71 obverse obverse (not available)

reverse reverse

enlargement of reverse enlargement of reverse

302

Figure 72 obverse reverse

Figure 73 obverse reverse

Figure 74 obverse reverse

Figure 75 not available

303

Figure 76

Figure 77

304

Figure 78

305

Figure 79

306

Map One

Sindhu Chab Beas ARiNAC

SuHlej Kiue

Jaqaghri Behat

Rohtak Meeruty .

Naurangabad

Mathura Yamuna Aiver

307

Map Two

Jagadhry Behat

308

Map Three

SOGDIANA

NA .......... Bokhara dMaracanda

Kashgar

Termez TARIM Bactra o Al Khanum YARKAND BASIN BACTRIA Herat HINDUKUSH Khotan .. Begram. Kabul .... BamyanO Dielalabad DAROPAMISADAE PAR GANDHARA. U

SEISTAN MIr Zakah Peshawar Taxlla", M Alexandria PANJAB

of Arachosia A ARACHOSIA CHENAB sagal A RAVI SATLEJ ........

GANGES Mathur

p.a putra

INDIAN usambl Sar iath

OCEAN

Broach

309

Map Four

Sura

Asandh Rangrahal Bijnor Agroha Dhanasia Hisar Meham Bha laut Mees .. Muradaba d Poltu Bhiwasi Naurangabad .- Markhauda Mure

Rewari Baghaula