Skip to main content

Full text of "The Cambridge history of India"

See other formats






E ^1 



' 1 


Digitized by the Internet Archive 

in 2007 with funding from 

IVIicrosoft Corporation 





C. F. CLAY, Manager 









t IAS 








179*^ 3S', 

lo. q . S)3. 








THE present volume deals with the history of ancient India from 
the earliest times to about the middle of the first century a.d. ; 
and it attempts to represent the stage of progress which research 
has now reached in its task of recovering from the past the out- 
lines of a history which, only a few years ago, was commonly 
supposed to be irretrievably lost. Well within the memory of 
contributors to this volume it was the fashion to say that there 
was no history of India before the Muhammadan conquests in the 
eleventh century A.D., and the general opinion seemed to be summed 
up in the dictum of the cynic who roundly asserted that all sup- 
posed dates for earlier events were like skittles — set up simply to 
be bowled down again. But this gibe, not quite justifiable even 
when it was uttered, could not be repeated at the present day. 
It has lost its point : it is no longer even approximately true. 

Regarded as a record of the character and achievements of 
great leaders of men, this history indeed is, and must always remain, 
sadly deficient. Of all the conquerors and administrators who 
appear in this volume there are two only — Alexander the Great 
and, in a less degree, A9oka — whose personality is at all intimately 
known to us ; in the case of others the bare memory of some of 
their deeds has been preserved ; the rest have become mere names 
to which research has given a time and a place. But the fragments 
of fact which have been rescued from the past are now sufficiently 
numerous and well established to enable us to construct a chrono- 
logical and geographical framework for the political history of 
many of the kingdoms and empires of ancient India ; and into this 
framework may be fitted the history of social institutions, which 
is reflected with unusual clearness in the ancient literatures. 

The manner in which modern scholarship has succeeded in 
throwing light on the dark ages of India, and in revealing order 
where all seemed to be chaos, is briefly indicated in the latter 
section of Chapter ii which deals with the sources of history. The 
story of rediscovery is a long record of struggles with problems 
which were once thought to be insoluble, and of the ultimate 
triumphs of patience and ingenuity. It begins in 1793, when Sir 
William Jones supplied ' the sheet-anchor of Indian chronology ' 


vi Prejace 

by his identification of the Sandrocottus of Alexander's historians 
with the Chandragupta of Sanskrit literature ; and its great epoch 
is ushered in by the decipherment of the long-forgotten alphabets 
of the ancient Indian inscriptions by James Prinsep in 1834. The 
first comprehensive summary of historical gains appeared in 1858, 
when Christian Lassen published the first volume of his Indische 
Alterthv/mskunde] and in recent years other summaries have been 
made by Dr Vincent Smith {Early History of India, Ist edn. 
1904, 2nd edn. 1908, 3rd edn. 1914), by Dr L. D. Barnett {Indian 
Antiquities, 1913), and, on a smaller scale, by the editor of this 
volume {Ancient India, 1914). 

The Cambridge History of India marks a new departure. The 
literature of the subject has become so vast, and is still gi'owing 
with such rapidity, that the best hope of securing a real advance 
in the study now lies in a division of labour among scholars who 
have explored at first hand the main sources of information. This 
volume therefore follows the plan adopted in the modern and 
medieval histories published by the Cambridge University Press. 
It is the outcome of the combination of a number of investigators 
with an editor whose function it has been to coordinate, so far as 
seemed possible or advisable, results obtained independently. That 
this plan is not without its disadvantages is obvious. AU coopera- 
tive enterprises of the kind involve necessarily some reiteration 
and also some discrepancy; and the questions which an editor 
must decide are how far repeated discussions of the same topic 
contribute to a fuller knowledge or are merely redundant, and how 
far different opinions admit of reconciliation or should be allowed 
to remain as representing the actual state of a study which abounds 
with difficulties and obscurities. In all important cases of the 
occurrence of such supplementary or contradictory views in this 
volume cross-references are given to the passages in which they 
are expressed. 

The general scheme of the work may be explained in a few 
words. The first two chapters are introductory. In Chapter i 
Sir Halford Mackinder describes the India of the present day when 
railways and telegraph lines have taken the place of the ancient 
routes, and gives an account of those geographical features which 
have determined the course of history in past times. The chapter 
is founded on Eight Lectures on India prepared for the Visual 
Instruction Committee of the Colonial Office and published in 1910. 
A similar acknowledgment of indebtedness is due to the Govern- 
ment of India for the use which the editor has made of its official 

Preface vii 

publications, especially the Census Reports of 1901 and 1911 and 
the Imperial Gazetteer of India, in Chapter ii on peoples and 
languages, and the sources of history. In Chapter iii Dr P. Giles 
reviews the evidence which Comparative Philology, aided by the 
ancient inscriptions of Western Asia, supplies for a knowledge of 
the early culture of the Aryans or * Wires,' their original habitat, 
and the date of the migrations which eventually led some of their 
tribes into India. The next five chapters are devoted to accounts 
of political, social, and economic conditions as represented in the 
earliest scriptures of the Brahmans, Jains, and Buddhists — Chapters 
IV and V by Prof. A. B. Keith on the Vedas and Brahmanas, 
Chapters vi and vii by Dr J. Charpentier and Dr T. W. Rhys Davids 
respectively on the earliest history of the Jains and Buddhists, and 
Chapter viii by Mrs Rhys Davids on economic conditions in early 
Buddhist literature. The five chapters which follow extend this 
investigation to the Brahman sources for the history of the post- 
Vedic period — Chapters ix-xii by Prof. E. Washburn Hopkins 
on the Sutras, epics, and law-books, and Chapter xiii by the editor 
on the Puranas. Up to this point the evidence has been gleaned 
almost entirely from Indian sources and confined almost entirely 
to India. In the next four chapters India is viewed in relation to 
other countries — Chapter xiv by Prof. A. V. Williams Jackson on 
the ancient Persian dominions in India, Chapters xv and xvi by 
Mr E. R. Bevan on the invasion of Alexander the Great and the 
early Greek and Latin accounts of India, and Chapter xvii by 
Dr G. Macdonald on the Hellenic kingdoms of Syria, Bactria, and 
Parthia. In Chapters xviii-xx the first great historical empire, 
that of the Mauryas, is described by Dr F. W. Thomas; and in 
Chapters xxi-xxiii the editor deals with the powers which arose 
on the ruins of the Maurya empire — the Indian native states, the 
Greek successors of Alexander the Great, and the Scythian and 
Parthian invaders. In Chapter xxiv Dr L. D. Bamett sums up 
what is known of the early history of Southern India; and in 
Chapter xxv he gives an analysis of the history of Ceylon which 
possesses a continuity in striking contrast to the fragmentary 
records of the kingdoms of the sub-continent. In the final Chapter 
XXVI Sir John Marshall describes the ancient monuments, and 
traces the various phases of Indian art from its beginnings to the 
first century a.d. 

The editor desires to thank all the contributors for the courtesy 
with which they have received and carried out his suggestions. 
He is doubly indebted to Sir John MarshaU, Dr L. D. Barnett, 

viii Preface 

Dr George Macdonald, Dr F. W. Thomas, and Mr E. R Bevan for 
much valuable advice and for their kind help in reading the proofs 
of chapters other than their own. He gratefully acknowledges also 
the assistance which he has received from Colonel Haig and from 
Sir Theodore Morison, the editors in charge of the Muhammadan 
and British sections of the Cambridge History of India (vols, 
iii-iv and v-vi respectively). 

The preparation of vol. ii, which will deal with the period from 
the downfall of the ^aka and Pahlava empire in the middle of the 
first century a.d. to the Muhammadan conquests, is attended with 
unusual difficulties, caused partly by the vast extent and partly by 
the fragmentary character of the historical records ; but it is at 
least to be hoped that its appearance may not be delayed by 
disasters such as that which has impeded the publication of vol. i. 

The printing of this volume began in 1913, and more than half 
the chapters were in type in 1914, when war made further progress 
impossible until the end of 1918. Since then the work has been 
completed, all the earlier chapters have been revised, and no effort 
has been spared to make the book representative of the state of 
early Indian historical studies at the end of 1920. 

The system of chronology, which has been adopted for the 
periods of Qaka and Kushana rule, needs some explanation. The 
chronological difficulties connected with the Vikrama era of 58 E.G. 
and the ^aka era of 78 A.D. are well known ; and it is universally 
admitted that the names which these eras bear were given to them 
at a later date, and afford no clue to their origin. The view 
maintained in this work is that the eras in question mark the 
establishment of the ^aka and Kushana suzerainties. The idea of 
suzerainty, that is to say, supreme lordship over all the kings of a 
large region — 'the whole earth,' as the poets call it — is deeply 
rooted in Indian conceptions of government ; and the foundation 
of an era is recognised as one of the attributes of this exalted 
position. Now there is abundant evidence that the ^^-ka empire 
attained its height in the reign of Azes I and the Kushana empire 
in the reign of Kanishka. It is natural to suppose therefore that 
such imperial eras must have been established in these reigns, and 
that their starting point in both cases was the accession of the 

The story of the foundation and extension of later eras in Indian 
history — the Gupta era and the era of Harsha, for example — can 
be clearly traced. All such undoubted illustrations of the process 
are seen to imply the existence of certain political conditions — the 

Preface ix 

relations of suzerain to feudatories, in fact. It is not necessarily, 
or indeed usually, the founder of a dynasty who is also the founder 
of an era ; but it is that member of the royal house who succeeded 
in asserting 'universal' sway and in reducing his neighbours to 
the status of feudatories. The use of the era can be shown, in 
well-ascertained cases, to have spread from the suzerain to the 
feudatories. Is there any reason to suppose that extension in the 
contrary direction — from feudatory to suzerain — ^has ever taken 
place or could possibly take place ? 

It has been suggested that the Vikrama era originated with the 
Malavas, whose name it sometimes bears in inscriptions. They 
were a people, apparently of no great political importance, who 
can be traced in the Punjab and Rajputana centuries before they 
settled in Malwa, the tract of Central India which now bears their 
name ; and they were almost certainly, like the other peoples 
of these regions, included in the ^^-ka empire at one period of 
their history. Is it conceivable that they could have initiated the 
Vikrama era, and that a great suzerain like Gondopharnes, who 
almost beyond doubt dates his Takht-i-Bahi inscription in this 
era, stood indebted to them for its use? The Vikrama era had 
undoubtedly become the traditional reckoning of the Malavas in 
the fifth century A.D. ; but the most obvious explanation of the 
fact is that they had inherited it from their former overlords. 

In the same way, the later name of the era of 78 a.d. may be 
due to its use for centuries by the ^aka satraps of Western India ; 
but they can scarcely have founded this era. Their very title 
' satrap ' shows that they were originally feudatories ; and they 
were most probably feudatories of the Kushanas. If so, they would 
use the era of their suzerains as a matter of course. 

Thus all a priori considerations favour the views which are 
adopted in this work in regard to the origin of these eras ; and, as 
is pointed out on pages 581-2, the Taxila inscription of the year 
136, which first suggested to Sir John Marshall the possibility of 
an ' era of Azes,' may also furnish positive evidence of their cor- 
rectness. It has been necessary to deal with these chronological 
problems somewhat at length because of their importance. If the 
theories here maintained are accepted, there will be an end to the 
worst of the perplexities which have for so long obscured the his- 
tory of N.W. India during the centuries immediately before and 
after the Christian era, and the dates in all the known inscriptions 
of the period will be determined, with the single exception of that 
which occurs in the Taxila copper-plate of Maues, and which, as is 


suggested, may be in some era which the (^akas brought with them 
from eastern Iran into India. 

The munificence of Sir Dorabji Tata has enabled the Syndics of 
the University Press to illustrate this volume more lavishly than 
would have been possible without such generous help. Mr G. F. 
Hill and Mr J. Allan of the British Museum have most kindly pro- 
vided casts of the coins figured in Plates i-viii; and Sir John 
Marshall has enhanced the value of his chapter on the monuments 
by supplying photographs, which were in many cases specially taken 
for the illustrations in Plates ix-xxxiv. 

The index has been made by Mr E. J. Thomas of Emmanuel 
College and the University Library. Modern place-names are, with 
very few exceptions, given as they appear in the index- volume of 
the Imperial Gazetteer of India. For the spelling of ancient names 
the system adopted by Prof. Macdonell in his History of Sanskrit 
Literature has been followed. This system has the double advant- 
age of being strictly accurate and, at the same time, of offering as 
few difficulties as possible for readers who are not orientalists. The 
vowels should be pronounced as in Italian, with the exception of a 
which has the indefinite sound so common in English, e.g. in the 
word organ. The vowels e and o are always long in Sanskrit, and 
are therefore only marked as such in the non-Sanskritic names of 
Southern India, in which it is necessary to distinguish them from 
the corresponding short vowels. 

E. J. R. 

St John's College, Cambridge, 
18 August 1921. 




By Sir Halford J. Mackinder, M.A., M.P., Reader in 

Geography in the University of London, formerly 

Student of Christ Church, Oxford 


The four sub-continents of Asia 1 

Ceylon; Colombo, the strategic centre of British sea-power in the 

Indian Ocean 1 

The Malabar and Coromandel coasts; the Western and Eastern Ghats 2 

The Carnatic ; Travancore ; Cochin 3 

The Gap of Coimbatore or Palghat 3 

The plateau between the Ghats ; Mysore 4 

Climate of the southern extremity of India 4 

Madras ; some causes of the comparative isolation of southern India . 5 

Burma, the connecting link between the Far East and the Middle East 5 

The geography of Burma 6 

The geography of Bengal 8 

Calcutta 9 

Countries of the Himalayan fringe 10, 25 

Valley of the Brahmaputra 11 

The plain of the Ganges and Jumna 12 

Central India 15 

The situation of Bombay 16 

The Maratha country ; Hyderabad ; the Deccan plateau .... 18 

The Central Provinces ; Baroda 19 

Kathiawar and Cntch 20 

The Himalayan barrier 20 

Bajputana; historical importance of the great Indian desert and the 

Delhi gateway 20 

The north-west frontier 26 

The plain of the Indus 27,31 

Routes leading into N.W. India 28 

Kashmir 32 

Gilgit ; Chitral ; the Karakoram ; the Hindu Rush 33 

Lateral communication between the Khyber and Bolan routes . . 33 

The Hindu Kush and the Indus as boundaries between India and Iran 34 

Summary of the principal physical features of the sub-continent . . 34 

xii Contents 


By E. J. Rapson, M. A., Professor of Sanskrit in the University 
of Cambridge, and Fellow of St John's College 



Varieties of race, speech, and culture 37 

Western and eastern invaders 38 

Natural and ethnographical divisions 40 

The seven chief physical types 40 

The four families of speech 48 

The caste-system 53 


Prehistoric archaeology 56 

Ancient literatures 56 

Foreign writers 58 

Inscribed monuments and coins 60 

The ancient alphabets 62 

Progress of research 62 



By P. Giles, Litt.D., Master of Emmanuel College, and Reader in 
Comparative Philology in the University of Cambridge 

The Indo-European languages 65, 71 

The Wiros and their original habitat 66 

Their migrations 70 

Evidence of the inscriptions of Boghaz-koi 72 

Iranians and Indo- Aryans 73 

Aryan names in the inscriptions of Mesopotamia 76 


By A. Berriedale Keith, D.C.L., D.Litt., Regius Professor 

of Sanskrit and Comparative Philology in the University of 

Edinburgh, formerly Scholar of Balliol College, Oxford 

The hymns of the Rigveda 77, 108 

Geography 79 

Fauna; peoples 81 

Social organisation 88 

Origins of the caste system . 92 

Political organisation 94 

Warlike and peaceful avocations 98 

Dress, food, and amusements 101 

Religion 103 

The beginnings of philosophy 107 

Chronology of Vedic literature 110 


Contents xiii 



By Professor A. Berriedale Keith 


Vedic literature after the period of the Rigveda 114 

Extension of Aryan civilisation to the Middle Country .... 116 

Peoples of the Middle Country . 118 

The more eastern peoples 122 

Changes in social conditions 125, 135 

Grovernment and the administration of justice 130 

Industry ; social life ; the arts and sciences 135 

Religion and philosophy 141 

Language 146 

Criteria of date 147 



By Jarl Charpentier, Ph.D., University of Upsala 

Jainism in its relation to Brahmanism and Buddhism .... 150 

The tirthakaras or ' prophets ' ; Par^sva 153 

Mahavira 155 

Jains and Buddhists 160 

Mahavira's rivals, G-osala and Jamali 162-3 

The Jain church after the death of Mahavira 164 

The great schism : ^vetambaras and Digambaras . . . . . 165 

Settlements in Western India 166 

Organisation of the religious and lay communities 168 

Blanks in Jain ecclesiastical history 169 



By T. W. Rhys Davids, LL.D., Ph.D., D.Sc, formerly Professor of 
Pali and Buddhist Literature at University College, London, and 
Professor of Comparative Religion in the University of Manchester 

Pre-Buddhistic 171 

India in the Buddha's time ; the clans 174 

The kingdoms : 

Kosala 178 

Magadha 182 

Avanti 185 

The Vamsas 187 

The first great gap 188 

Chandagutta 190 

Age of the authorities used 192 

Growth of Buddhist literature from the time of Buddha down to Asoka 197 

xiv Contents 



By Mrs C. A. F. Rhys Davids, M.A., D.Litt., FeUow of 
University College, London 


Rural economy 198 

Cities ; villages ; the land 200 

Agriculture 203 

Labour and industry 205 

Social conditions 208 

Trade and commerce 210 

Trade centres and routes 212 

Methods of exchange and prices 216 

Securities and interest 218 

General conclusions 219 



By R Washburn Hopkins, PLD., LL.D., Professor of Sanskrit 
and Comparative Philology in Yale University 

Language of the later Brahman literature 220 

Social conditions as reflected in the Brahman and Buddhist books . 221 

Outlines of chronology 222 

Causes of the wider political outlook 225 



By Professor E. Washburn Hopkins 

The Sutra literature 227 

Grihya Sutras 228 

Oblations to spirits and gods 229 

Rites to avert disaster and disease 231 

Substitution of images of meal for sacrificial animals . . . ' . 232 

Marriage ceremonies 233 

Caste and family 234 

Contents and arrangement of the Grihya Sutras 236 

The life represented is rural, not urban 237 

Minor superstitions 238 

irya and ^udra ; Dharma Sutras 240 

The beginnings of civil and criminal law ; inheritance .... 243 

Duties of the king 244 

Taxes ; status of women ; ordeals 247 

Legal rates of interest 248 

Religion and philosophy 248 

Relative ages of the Sutras 249 

Contents xv 



By Professor E. Washburn Hopkins 


The two chief varieties of epic poetry 251 

Sources of the Mahabharata 252 

Narrative and didactic interpolations 255 

The characters partly historical and partly mythical .... 257 

Date of the poem in its present form 258 

Features common to the Mahabharata and the Eamayania . . . 259 
Social conditions in the Sanskrit epics and in Buddhist works of the 

same period 259 

The story of the Mahabharata 262 

The Mahabharata and the Eamayafjia contrasted 264 

Earlier and later moral ideals in the Mahabharata 265 

Knights, priests, commoners, and slaves 266 

The king and his ministers 271 

Religious and philosophical views of the epics 272 

Peoples traditionally engaged in the great war 274 

The genealogies 275 


By Professor E. Washburn Hopkins 

The chief codes— Manu, Vislmu, Tajnavalkya, and Narada . . . 277-9 

Growth of the distinction between religious precepts and law . . 280 
Growth of the distinction between civil and criminal law, and the first 

enumeration of legal titles 281 

Ordeals 282 

Dharna 284 

Punishments 284 

Development of civil law 286 

Interest; wages; property 287 

The king as ruler in peace and war 288 

The king as judge 290 

Hereditary traditional law and custom 291 

Infant marriage ; the levirate ; the status of women .... 292 

The law-books and the Arthaeastra 293-4 



By Professor E. J. Rapson 

The classical definition of a Pura^a 296 

Kshatriya literature 297,302 

Scriptures of the later Hinduism 298 

Critical study of the Puranas 299 

The Puranas and Upapuranas 300 

Their chronological and geographical conceptions 303 

Genealogies partly legendary and partly historical 304 

xvi Contents 


Common traditional elements in Vedic literature and the Pura^ias . 306 

Traditional period of the great war between Knrus and Pa^^^is . . 306 

The Purus 307 

Thelkshvakus 308 

Kings or suzerains of Magadha : 

Brihadrathas 309 

Pradyotas, originally kings of Avanti 310 

^i9unagas 311 

Nandas 313 

Contemporary dynasties in Northern and Central India . . . 315 
Later kings of Magadha and suzerains of N. India: 

, The later Nandas, Mauryas, ^ungas, Kanvas, and Andhras . . 317 



By A. V. Williams Jackson, Ph.D., LL.D., Professor of 
Indo-Iranian Languages in Columbia University 

Prehistoric connexions between Persia and India 319 

Common Indo-Iranian domains 321 

Evidence of the Veda and the Avesta 322 

Avestan, Old Persian, Greek, and modem designations of Persian 

provinces south of the Hindu Kush 326 

Early commerce between India and Babylon ...... 329 

The eastern campaigns of Cyrus 329 

Cambyses 333 

Darius : ... 334 

Xerxes 340 

Decadence of the Achaemenian empire 340 

The conquest of Persia by Alexander 341 

Extent of Persian influence in India 341 

Note to Chapter XIV 

Ancient Persian Coins in India 
By Dr George Macdonald 

The rarity of Persian gold coins in India explained by the low ratio of 

gold to silver 342 

The attribution of punch-marked Persian silver coins to India doubtful 343 



By E. R. Bevan, M. A., Hon. Fellow of New College, Oxford 

The E[abul valley and the Punjab in the fourth century B.C. . . . 345 
Alexander's conquest of the Persian empire; settlements in Seistan, 

Kandahar, and the upper Kabul valley ; invasion of Bactria . . 347-8 
The i-aja of Taksha^ila and the Paurava king (Poms) .... 349 

Contents xvii 


Invasion of India from the upper Kabul valley 350^ 

Hill tribes beyond the N.W. frontier 352 

Occupation of the lower Kabul valley 354 

Siege and capture of Aornus 356 

The crossing of the Indus 357 

Reception at Taksha^ila . . . . . . . . . . 358 

The Paurava king 359,368,383 

The battle of the Hydaspes 360 

Foundation of Nicaea and Bucephala 368 

Flight of the second Paurava king, and occupation of his kingdom . 370 

Capture of Sangala 371 

Saubhuti (Sophytes) 371 

The Hyphasis, the eastern limit of Alexander's conquests . . . 372 

Return to the Hydaspes ; expedition to the Indus delta .... 373 

Defeat of the Malavas 375 

Musicanus 377 

Return of Craterus through Kandahar and Seistan 379 

Pattala 379 

Return of Alexander through G-edrosia 380 

Return of Nearchus by sea 381_ 

Alexander's Indian satrapies 383 

Consequences of the invasion , 384 

Note to Chapter XV 

Ancient Greek Coins in India 
By Dr George Macdonald 

Athenian and Macedonian types 386 

Sophytes 388 

Coins attributed to Alexander 388 

Double darics 390 



By E. R. Bevan, M.A. 

The early sources of information 391 

Scylax of Caryanda; Hecataens of Miletus; Herodotus; Ctesias of 

Cnidus 393 

Nearchus; Onesicritus; Clitarchus 398-9 

Megasthenes; Da'imachus; Patrocles 399-400 

Geography and physical phenomena 400 

The mineral, vegetable, and animal world 403 

Ethnology and mythology 407 

Social divisions according to Megasthenes 409 

His description of Pataliputra 411 

Manners and customs ; laws 412 

Marriage ; suttee ; disposal of the dead ; slavery 414 

The king; royal festivals; government ofl&cials 416 

Industries 418 

Brahmans; ascetics; philosophers ,,,,,... 419 

C. H. I. I. 6 

xviii Contents 


Deities 422 

Fabulous creatures 422 

Pearls ; Southern India and Ceylon 423 

Later sources of information— Apollodorus of Artemita; Eratosthenes; 

Alexander Polyhistor ; Strabo; Pliny; Arrian; Aelian . . . 425 

Greek merchantmen 425 



By George Macdonald, C.B., LL.D., F.B.A., First Assistant 

Secretary Scottish Education Department, formerly Scholar 

of Balliol College, Oxford 

Greek and native rulers of the Kabul valley and N.W. India after 

Alexander 427 

Chandragupta 429 

The Indian expedition of Seleucus 430 

Relations of Syria with the Maurya empire 432 

Foundation of the kingdoms of Bactria and Parthia .... 434 

Diodotus 436 

Arsaces 438 

Euthydemus 440,442 

Invasion of India by Antiochus III 441 

Sophagasenus 442 

Demetrius 444 

Eucratides 446, 454 

Euthydemus II ; Demetrius II; Agathocles; Pantaleon; Antimachus 447 

Heliocles and Laodice 453 

Plato 456 

Parthian invasion of Bactria 457 

Scythian invasion of Bactria 458 

Heliocles 459 

Key to Plates I-IV 462 



By F. W. Thomas, M.A, Ph.D., Librarian of the India Office, 
fonnerly Fellow of Trinity College, Cambridge 

Characteristics of the Maurya period and authorities for its history . 467 

N. W. India before and after Alexander 468 

Agrammes, Xandrames, or Dhana-Nanda 469 

Nanda and Chandragupta 470 

Date of the overthrow of Nanda 471 

Plot of the Mudrardkshasa 471 

Chandragupta and Seleucus 472 

Megasthenes 472 

Rule of Chandragupta and extent of his dominions 473 





By Dr F. W. Thomas 

Internal conditions: page 

The land 474 

Towns and fortifications 475 

The imperial capital 477 

The people 477 

Trade 478 

State of society 479 

Literature 482 

Lang^uage 483 

Religion 483 

Law 485 

Government and administration: 

Civil administration founded on village autonomy .... 486 

The army 489 

Foreign policy 490 

Tribal oligarchies 491 

Monarchies 491 



By Dr F. W. Thomas 

Bindusara . . 495 

Events and principal enactments in Anoka's reign 495 

Religious and other foundations 497, 501 

Buddhist Council of Pataliputra and religious missions .... 498 

Duration of Anoka's reign and his family history 499 

Chronology 502 

Acjoka's principles and personal action 504 

His admonitions 507 

His ordinances and institutions 508 

The personality of A^oka as revealed in his edicts 509 

His successors 511 

Probable division of the empire after the reign of Samprati . . . 512 



By Professor E. J. Rapson 

The peoples of India in the inscriptions of Acoka 514 

Internal strife and foreign invasions the result of the downfall of im- 

I)erial rule 516 

Routes connecting Pataliputra with the north-western and western 

frontiers 516 

XX Contents 


Kingdoms on the central route 517 

The ^ungas 517 

Feudatories of the QungBS 523 

Kosala and Magadha 527 

Independent states 528 

Rise of the Andhras 529 

Conquest of Ujjain 531 

Conquest of Vidi^a 533 

The Kalingas 534 

Key to Plate V 538 



By Professor E. J. Rapson 

The Yavana invasions 540 

Extension of Yavana power along the routes from Kabul . . , 542 

Stages on the route leading to Pataliputra 543 

Yavana invasion of the Midland Country 544 

Baetrian and Indian coins 545 

The house of Euthydemus : 

Demetrius; Pantaleon and Agathocles; Antimachus. . . . 546 

Apollodotus; Menander 547 

The kingdom of ^akala 549 

Dominions of this house after the conquests of Eucratidcs and 

Heliocles 551 

The families of Menander and Apollodotus 552 

The (^aka conquests completed in the reign of Azes I . . . 553 
The house of Eucratides : 

Eucratides 554 

The kingdom of Kapi^a 555 

Heliocles 556 

His successors at Pushkalavati 557 

Antialcidas 558 

Archebius his successor at Taksha^ila 559 

^aka conquest of Pushkalavati and Taksha^ila in the reign of Maues 559 

Successors of Eucratides in the upper Kabul valley .... 560 

Hermaeus the last Yavana king 560 

Pahlava conquest of the upper Kabul valley 561 



By Professor E. J. Rapson 

The ^aka invaders came from Seistan and Kandahar into the country 

of the lower Indus 563 

^akas in the inscriptions of Darius 564 

Migration of the Yueh-ohi 565 

Bactria overwhelmed and Parthia invaded by ^akas .... 566 

Being checked by Parthia the ^akas invaded India .... 567 

The title 'Great King of Kings' 567 

Contents xxi 


Pahlava and ^aka suzerains in eastern Iran and India .... 569 

The date of Maues, the conqueror of Pushkalavati and Taksha^ila . 570 
Extension of his conquests by Azes I, who may be the founder of the 

Vikrama era 571 

Azilises ; Azes II ; Vonones 572 

The family of Vonones 673 

^aka satraps 574 

The strategoi 677 

Grondopharnes 577 

Pacores 680 

The transition from Pahlava to Kushana rule in Taksha9ila . . . 580 

V'ima Kadphises 581 

The date of Kanishka 583 

Summary of numismatic evidence for the history of the Yavana, ^aka, 

and Pahlava invaders of India 586 



By L. D. Barnett, M.A., LittD., Keeper of the Department of 

Oriental Printed Books and MSS. in the British Museum, formerly 

Scholar of Trinity College, Cambridge 

The Dravidian peoples 593 

The Tamil kingdoms 594 

The Andhras or Telugus 598 

The Kalingas 601 

Mahai'ashtra, etc 602 



By Dr L. D. Barnett 

Sources of history 604 

The Vatldas 604 

Other races in Ceylon 605 

Earliest colonisation 605 

History from Vijaya to the advent of Buddhism 607 

From Devanampiya Tissa to Kutakauna Tissa 608 



By Sir J. H. Marshall, K.C.I.E., LittD., Director General 
of Archaeology in India, formerly Scholar of King's College, 


Prehistoric antiquities 612 

The age of iron 615 

The mounds at Lauriya Nandangarh; the walls of the old city of 

Rajagriha 616 

The earliest buildings 617 

Monuments of the Maurya epoch 618 


xxii Contents 


Persian influence 621 

Contrast between Indian and foreiicn workmanship also seen in the 

minor arts 622 

Development of art in the (^unga period 623 

Bharhut 624 

Besnagar 625 

More advanced style in the railing at Buddh Gaya 625 

Sanchi 627 

Interpretation of the reliefs 629 

Varieties of style 630 

The pre-Kushana sculptures at Mathura 632 

Decadence of art 633 

Chaityas 633 

Bhaja; Kondane; Pitalkhora; Ajanta; Bedsa; Nasik; Eiirli . . 635 

Vihdras 637 

The caves of Orissa 638 

Sculptures in the caves 640 

Paintings of the Early Indian school 642 

Terracottas 643 

Foreign influence in Indian art 644 

Coins 645 

Architecture 646 

Minor arts 646 

The Gandhara School 648 

Greek and Indian ideals 648 

List op Abbreviations 651 

Bibliographies 653 

Chronology 697 

Index 705 

Plates I— XXXIV at end 


1. Physical Map of the Indian Empire before p. 1 

2. Physical Map of N.W^. India and the Adjacent Countries 

between pp. 26 and 27 

3. The Kabul Valley and the Punjab between pp. 844 and 345 

4. Baluchistan and the country of the Lower Indus between pp. 374 and 375 

5. Bharatavarsha between pp. 514 and 515 

6. The Indian Empire and Ceylon in pocket at end of volume 




Ill, IV. 



























Persian, Athenian, and Macedonian Coins in India. 

Coins of the Seleucid Kings and their Successors. 

Coins of Bactria. 

Coins of Native Indian States. 

Coins of the Greek, Scythian, and Parthian Invaders. 

Palaeolithic and Neolithic Implements. 

Prehistoric Copper Objects. 

Gold leaf from Lauriya Nandangarh; Seals from Harappa; 
A9oka Pillar at Lauriya Nandangarh ; Facade of the Lomas 
Kishi Cave. 

Yaksha (?) Statue from Parkham ; Capital of A^oka Column at 

Yaksha Figures from Patna; Indigenous punch-marked Coins; 
Coin of Sophy tes (Saubhuti). 

Terracotta Figures from Sarnath and Basarh (Vai^ali) ; Jewel- 
lery from Taxila ; Crystal Keliquary from Piprahwa Stupa. 

Gateway and Railings of the Bharhut Stupa. 

Reliefs on the Railing of the Bharhut Stupa. 

Garudia PUlar at Besnagar. 

Railing and Reliefs at Buddh Gaya. 

The Main Stupa at Sanchi (General view from N.E.). 

Stupa II at Sanchi (General view and Pillars). 

Front Face of East Gateway of the Main Stupa at Sanchi. 

Architraves of Gateways of the Main Stupa at Sanchi. 

Jain Votive Tablet and portion of Torana Arch from Mathura. 

The Chaitya Cave at Karli. 

Caves at Kondane, Bhaja, Nadsur, and Nasik. 

Manchapuri Cave and Ananta Gumpha. 

Bani Gumpha and Alakapuri Cave. 

Terracotta Plaque from Bhita ; Copper Lota from Kulu. 

Coins of Bactria and North-western India. 

Head of Dionysus from Taxila ; Stupa Base of Scytho-Parthian 
epoch at Taxila. 

Bronze Statuette from Taxila; Gold Casket from Stupa at 

Intaglio Gems from the North-west of India. 

Life-size Statue of Buddha from Gandhara. 



General Bibliography 653 

Chapter I 654 

Chapter II 655 

Chapter III 658 

Chapter IV 659 

Chapter V 660 

Chapter VI 662 

Chapter VII 663 

Chapter VIII 665 

Chapters IX— XII 666 

Chapter XIII 669 

Chapter XIV 670 

Note to Chapter XIV . . . . . .673 

Chapters XV and XVI 674 

Note to Chapter XV 676 

Chapter XVII 677 

Chapters XVIII and XIX . . . . .679 

Chapter XX 681 

Chapter XXI 683 

Chapter XXII 685 

Chapter XXIII 686 

Chapter XXIV 688 

Chapter XXV 690 

Chapter XXVI 692 


p. 129, 1. 2 from foot. For Himalayas read Himalayas. 

p. 184, 1. 11. For Saman'Ha- read Sdmafina-. 

p. 194, 1. 3. For Moggliputta read Moggaliputta. 

p. 265, note '\ 1. 3. For Krishna read Krishna. 

pp. 320, 321, note *, 11. 7 and 10. For Oldenburg reoo^ Oldenberg. 

p. 501, 1. 15. For Konagamana read Konagamana. 

p. 581, L 11. For protrait read portrait. 

p. 598, 1. 10. For Aravalar read Aruvalar. 

p. 636, L 5 from foot. For Andra read Andhra. 

p. 684, 1. 1. For Fourcher read Foucher. 

MAP 1 



The Cambridge History of India ^ Vol. I 




Scale of Miles 

Above 12,000 feet 
8000 10 12,000 „ 
4000 - 8000 „ 
2000 - 4000 ., 
1000 - 2000 „ 

Below 1000 ., 


Map I 

George Philip & Son.L^" 



The great continent of Asia falls naturally into four parts or 
sub-continents. The east drains to the Pacific, and is mainly 
Buddhist. The north and west centre lie open in an arctic direc- 
tion, and during the past century were united under Russian rule. 
The south-west, or Lower Asia, is the land of passage from Asia 
into Africa, and from the Indian ocean to the Atlantic. It is 
the homeland of Islam. In the middle south is the Indian sub- 

The inhabitants of the United States describe their vast land 
as a sub-continent. As regards everything but mere area the 
expression is more appropriate to India. A single race and a 
single religion are overwhelmingly dominant in the United States, 
but in India a long history lives to-day in the most striking 
contrasts, presenting all manner of problems which it will take 
generations to solve. 

In the past there have been great empires in India, but it is a 
new thing that the entire region from the Hindu Kush to Ceylon, 
and from Seistan to the Irrawaddy should be united in a single 
political system. The one clear unity which India has possessed 
throughout history has been geographical. In no other part of 
the world, unless perhaps in South America, are the physical 
features on a grander scale. Yet no where else are they more 
simply combined into a single natural region. 

The object of this chapter is to give a geographical description 
of India, as the foundation upon which to build the historical 
chapters which follow. We will make an imaginary journey through 
the country, noting the salient features of each part, and will then 
consider it as a whole, in order to set the facts in perspective. 

The most convenient point at which to begin is Colombo, the 
strategical centre of British sea-power in the Indian ocean. Four 
streams of traffic, India-bound, converge upon Colombo from Aden 
and the Mediterranean, from the Cape, from Australia, and from 

C.H.I. I. 1 

2 7 he Sub- Continent of India [ch. 

Singapore and the Far East. From Cape Comorin, in the imme- 
diate neighbourhood of Colombo, the Indian coasts diverge to 
Bombay and Karachi on the one hand, and to Madras, Calcutta, 
and Rangoon on the other. 

Colombo is not, however, in a technical sense Indian. It is the 
chief city of the luxuriant and beautiful island of Ceylon, which is 
about as large as Ireland. Neither to-day nor in the past has 
Ceylon been a mere appendage of India. The Buddhist religion 
of half its population, and the Dutch basis of its legal code are 
the embodiment of chapters in its history ; it is for good historical 
reasons that the Governor of Ceylon writes his despatches home to 
the Secretary for the Colonies and not to the Secretary for India. 

The passage by steamer across the Gulf of Manaar from 
Colombo to Tuticorin on the mainland occupies a night. Midway 
on the voyage the mountains of Ceylon lie a hundred miles to the 
east, and Cape Comorin a hundred miles to the west. The gulf 
narrows northward to Palk Strait, which is almost closed by a 
chain of islands and shoals, so that the course of ships from Aden 
into the Bay of Bengal is outside Ceylon. 

Cape Comorin, the southernmost point of India, lies eight 
degrees north of the equator, a distance nearly equivalent to the 
length of Great Britain. From Comorin the Malabar and Coro- 
mandel coasts extend for a thousand miles, the one north-westward, 
and the other northward and then north-eastward. The surf of 
the Arabian sea beats on the Malabar coast, that of the Bay of 
Bengal on the Coromandel coast. Both the Arabian sea and the 
Bay of Bengal open broadly southward to the Indian ocean, for 
the Indian peninsula narrows between them to a point. 

The interior of the Indian peninsula is for the most part a low 
plateau, known as the Deccan, whose western edge is a steep brink 
overlooking the Malabar coast. From the top of this brink, called 
the Western Ghats, the surface of the plateau falls gently eastward 
to a lower brink, which bears the name of Eastern Ghats. Between 
the Eastern Ghats and the Coromandel coast there is a belt of 
lowland, the Carnatic. Thus India presents a lofty front to the 
ship approaching from the west, but a featureless plain along the 
Bay of Bengal, where the trees of the coastline appear to rise out 
of a water horizon when seen from a short distance seaward. 

As the steamer approaches Tuticorin the land becomes visible 
some miles to the west as a low dark line along the horizon. 
Gradually the detail of the coast separates into a rich vegetation 
of trees and a white city, whose most prominent object is a cotton 

i] Carnatic : Ntlgiris : Coimbatore 3 

factory. India is a land of cotton. Its people have grown cotton, 
woven cotton, and worn cotton from time immemorial. The name 
calico is derived from Calicut, a town on the Malabar coast which 
was a centre of trade when Europeans first came over the ocean. 

On leaving Tuticorin we travel northward over the Carnatic 
plain. It is a barren looking country and dry, though at certain 
seasons there are plentiful rains, and crops enough are produced 
to maintain a dense population. Far down on the western horizon 
are the mountains of the Malabar coast, for in this extremity of 
India the Western and Eastern Ghats have come together and 
there is no plateau between them. The mountains rise from the 
western sea and from the eastern plain into a ridge along the west 
coast, with summits about as high as the summits of Ceylon, that 
is to say some eight thousand feet. The westward slopes of these 
mountains, usually known as the Cardamon hills, belong to the 
little native states of Travancore and Cochin. 

A group of hills, isolated on the plain, marks the position of 
Madura, a hundred miles from Tuticorin. Madura is one of three 
southern cities with superb Hindu temples. The other two are 
Trichinopoly and Tanjore, standing not far from one another, a 
second hundred miles on the road from Tuticorin to Madras. 

A hundred and fifty miles west of Trichinopoly is Ootacamund, 
high on the Nllgiri hills. ' Ooty,' as it is familiarly called, stands 
some seven thousand feet above the sea in the midst of a country 
of rolling downs, rising at highest to nearly nine thousand feet. 
This lofty district forms the southern point of the Deccan plateau, 
where the Eastern and Western Ghats draw together. 

South of the Nilgiris is one of the most important features in 
the geography of Southern India. The western mountains are 
here breached by the broad Gap of Coimbatore or Palghat, giving 
lowland access from the Carnatic plain to the Malabar coast. The 
Cardamon hills face the Nilgiris across this passage, which is about 
twenty miles broad from north to south, and only a thousand feet 
above the sea. 

The significance of the Gap of Coimbatore becomes evident 
when we consider the distribution of population in Southern India. 
For two hundred miles south of Madras, as far as Trichinopoly and 
Tanjore, the Carnatic plain is densely peopled. There are more 
than 400 inhabitants to the square mile. A second district of 
equal density of population extends from Coimbatore through the 
Gap to the Malabar coast between the ancient ports of Cochin and 
Calicut. There are many natural harbours along the Malabar 


4 The Sub- Continent of India [CH. 

coast all the way from Bombay southward, but the precipitous and 
forested Western Ghats impede communication with the interior. 
Only from Calicut and Cochin is there an easy road to the Caraatic 
markets, and this is the more important because the Coromandel 
coast is beaten with a great surf and has no natural harbours. 

To-day there is a railway from Madras through the Gap of 
Coimbatore to Cochin and Calicut, and from this railway a rack 
and pinion line has been constructed up into the Nilgiri heights 
to give access to the hill station of Ootacamund. There are 
magnificent landscapes at the edge of the Nilgiris where the 
mountains descend abruptly to the plains. On the slopes are great 
forests in which large game abound, such as sambar and tiger. On 
the heights the vegetation is naturally different from the lowland. 
The cultivation of the Nilgiris is chiefly of tea and cinchona. 

Northward of the Nilgiris, on the plateau between the Ghats, 
is the large native state of Mysore. The Cauvery river rises in 
the Western Ghats, almost within sight of the western sea, and 
flows eastward across Mysore. As it descends the Eastern Ghats 
it makes great falls. Then it traverses the Carnatic lowland past 
Trichinopoly and Tanjore to the Bay of Bengal. The falls have 
been harnessed and made to supply power, which is carried elec- 
trically for nearly a hundred miles to the Kolar goldfield. 

Around the sources of the Cauvery, high in the Western Ghats, 
is the little territory of Coorg, no larger than the county of Essex 
in England. The best of the Indian coffiee plantations are in 
Coorg, which is directly under the British Raj, although adminis- 
tered apart from Madras. Mysore is separated from both coasts 
by the British Province or Presidency of Madras, which extends 
through the Gap of Coimbatore. 

All the southern extremity of India, except the greater heights, 
is warm at all times of the year, though the heat is never so great 
as in the hot season of northern India. There is no cool season in 
the south comparable with that of the north. In most parts of 
India there are five cool months, October, November, December, 
January, and February. March, April, and May are the hot season. 
The remaining four months constitute the rainy season, when the 
temperature is moderated by the presence of cloud. In the south, 
almost girt by the sea, some rain falls at all seasons, but along 
the Malabar coast the west winds of the summer bring great rains. 
These winds strike the Western Ghats and the Nilgiri hills, and 
drench them with moisture, so that they are thickly forested. At 
this season great waterfalls leap down the westward ravines and 

i] Madras: Burma 5 

feed torrents which rush in short valleys to the ocean. One of the 
grandest falls in the world is at Gersoppa in the north-west corner 
of Mysore. 

The city of Madras lies low on the coast four hundred miles 
north of Tuticorin, but the chief military station of southern India 
is Bangalore on the plateau within Mysore. A hundred years ago, 
when Sultan Tipu of Mysore had been defeated by the British, 
Colonel Wellesley, afterwards the great Duke of Wellington, was 
appointed to command ' the troops above the Ghats.' The expres- 
sion is a picture of the contrast between the lowland Presidency 
and the upland Feudal State. 

Madras city, like the other seaports of modern India, has grown 
from the smallest beginnings within the European period. It has 
now a population of more than half a million. Until within recent 
years, however, Madras had no harbour. Communication was 
maintained with ships in the open roadstead by means of surf 
boats. Two piers have now been built out into the sea at right 
angles to the shore. At their extremities they bend inward towards 
one another so as to include a quadrangular space. None the less 
there are times when the mighty waves sweep in through the open 
mouth, rendering the harbour unsafe, so that the shipping must 
stand out to sea. Almost every summer half a dozen cyclones 
strike the east coast of India from the Bay of Bengal. When the 
Madras harbour was half completed the works were overwhelmed 
by a storm, and the undertaking had to be recommenced. If 
we consider the surf of the Coromandel coast, and the barrier 
presented by the Western Ghats behind the Malabar coast, we 
have some measure of the comparative isolation of southern India. 

From the far south we cross the Bay of Bengal to the far east 
of India. Burma is the newest province of the Indian Empire, if' 
we except sub- divisions of older units. In race, language, religion, 
and social customs it is nearer to China than it is to India. In 
these respects it may be considered rather the first land of the 
Far East than the last of India, the Middle East. Geographically, 
however, Burma is in relation with the Indian world across the 
Bay of Bengal, for it has a great navigable river which drains into 
the Indian ocean, and not into the Pacific as do the rivers of the 
neighbouring countries, Siam and Annam. Commercially it is 
coming every day into closer relation with the remainder of the 
Indian Empire, for it is a fruitful land of sparse population, which 
may perhaps be developed in the future by the surplus labour of 
the Indian plains. 

6" Thi Sub- Continent of India [CH. 

The approach from the sea is unimpressive, for the shore is 
formed by the delta of the Irrawaddy river. The easternmost of 
tlie channels by which that great stream enters the sea is the 
Rangoon river. The city of Rangoon stands some thirty miles up 
this channel. The golden spire of its great pagoda rises from 
among the trees on the first low hill at the edge of the deltaic 

Fifty yeare ago Rangoon was a village. To-day it has a quarter 
of a million people. Like the other coast towns of India and 
Ceylon, it owes its greatness to the Europeans who have come 
over the ocean. In all the earlier ages India looked inward, not 

Rangoon is placed where the river makes a bend eastward. 
The city lies along the north bank for some miles, to the point 
where the Pegu tributary enters. Black smoke hangs over the 
Pegu river, for there are many rice mills with tall chimneys along 
its banks. Rangoon harbour is always busy with shipping. Along 
its quays are great timber yards and oil mills, for the products of 
Burma are first and foremost rice, and then timber, especially 
great logs of teak, harder than oak, and then petroleum. The 
work of the port and mills is largely in the hands of Indians and 
Chinese. The Burmese are chiefly occupied with work in the fields. 

The geogi-aphy of Burma is of a simple design. It consists of 
four parallel ranges of mountain striking southward, and three 
long intervening valleys. The easternmost range separates Burma 
and the drainage to the Indian ocean from Siam and the drainage 
to the Pacific ocean. This great divide is continued through the 
Malay peninsula almost to Singapore, only one degree north of 
the equator. The westernmost range divides Burma from India 
proper, and then follows the west coast of Burma to Cape Negrais. 
This range is continued over the bed of the ocean, and reappears 
in the long chain of the Andaman and Nicobar Islands. In its 
entirety it has a graceful waving lie upon the map, curving first to 
the west, then to the east, and then again to the west. The two 
intervening ranges separate the Salween, Sittang, and Irrawaddy 

The valley of the Salween is less deeply trenched between its 
bounding ranges than are the other two, and therefore has a 
steeply descending course broken by rapids, and is of small value 
for navigation. At its mouth is the port of Moulmein. The valley 
of the Sittang, which is a relatively short river, prolongs the upper 
valley of the Irrawaddy, for the latter stream makes a westward 

i] Mandalay : Bhamo »7 

bend at Mandalay, and passes by a transverse gap through one 
of the parallel ridges. Beyond this gap it bends southward again, 
accepting the direction of its tributary, the Chindwin. The railway 
from Rangoon to Mandalay runs through the Sittang valley and 
does not follow the Irrawaddy. 

The delta of the Irrawaddy bears the name of Pegu or Lower 
Burma. The region round Mandalay is Upper Burma. The coast- 
land beyond the westernmost of the mountain ranges is known as 
Arakan. The coastland south of the mouth of the Salween, beset 
with an archipelago of beautiful islands, is known as Tenasserim. 

The train from Rangoon to Mandalay crosses the broad levels 
of the delta, passing through endless rice or ' paddy ' fields. Only 
the ears of the grain are lopped olF ; the straw is burnt as it stands. 
The Burmans are mostly yeomen, each owning his cattle and doing 
his own work in the fields. Beyond the delta the railway follows 
the Sittang river, with hill ranges low on the eastern and western 
horizons. At Mandalay it comes through to the Irrawaddy again. 

There is a hill in the northern suburbs of Mandalay, several 
hundred feet high, from which you may look over the city. Even 
when seen from this height the houses are so buried in foliage that 
the place appears like a wood of green trees. It has a population 
of about two hundred thousand, so that it is now smaller than 
upstart Rangoon. Mandalay is the last of three capitals a few 
miles apart, which at different times in the past century were the 
seat of the Burmese kings. Amarapura, a few miles to the south, 
was the capital until 1822. Ava, a few miles to the west, was the 
capital from 1822 to 1837. 

The navigation of the Irrawaddy extends for nine hundred miles 
from the sea to Bhamo, near the border of the Chinese Empire. 
As the steamer goes northward from Mandalay the banks are at 
first flat, with here and there a group of white pagodas. Great 
rafts of bamboo and teak logs float down the river. At Katha the 
flat country is left, for the river there comes from the east through 
grand defiles, with wooded fronts descending to the water's edge. 
Bhamo lies low along the river bank beyond the narrows. It is 
only twenty miles from the Chinese frontier. Many of its houses 
are raised high upon piles, because of the river floods. Until 
recently the Kachin hillmen often raided the caravans passing 
from Bhamo into China. 

To realise the antiquity and the splendour of early Burmese 
civilisation we must descend the Irrawaddy below Mandalay to 
Pagan. There for some ten miles beside the river, and for three 

8 The Suh- Continent of India [ch. 

miles back from its bank, are the ruins of a great capital, which 
flourished about the time of the Norman Conquest of p]ngland. 
From the centre of the ruined city there are pagodas and temples 
in every direction. 

Pagan is situated in what is known as the dry belt of Burma, 
the typical vegetation of which is a tall growth of cactus. In 
Burma the winds of summer and autumn blow from the south- 
west, as they do in southern India. They bring moisture from the 
sea, which falls in heavy rain on the west side of the mountains 
and over the delta. At Rangoon there is an annual rainfall of 
more than one hundred inches, or more than three times the rain- 
fall of London. At Pagan, however, lying deep in the Irrawaddy 
valley under the lee of the continuous Arakan range, the rainfall 
is small, as little as twenty inches in the year, and the climate is 
hot and evaporation rapid. 

Elsewhere in Burma are either rich crops, or the most luxuriant 
forests of tall leafy trees, full of game and haunted by poisonous 
snakes. Wild peacocks come from the woods to feed on the rice 
when it is ripe, and tigers are not unknown in the villages. Only 
a few years ago a tiger was shot on one of the ledges of the great 
pagoda in Rangoon. Notwithstanding the age of its civilisation 
Burma is still subject to a masterful nature. Moreover civilisation 
is confined to the immediate valleys and delta of the Irrawaddy 
and Salween. On the forested hills are wild tribes, akin to the 
Burmese in speech and physique — the Shans in the east, the 
Kachins in the north, and the Chins in the west. Burma contains 
but twelve million people — Burmese, Chinese, Hindus, and the hill 

From Burma the passage to Bengal is by steamer, for the 
Burmese and Indian railway systems have not yet been connected. 
The heart of Bengal is one of the largest deltas in the world, a 
gi'eat plain of moist silt brought down by the rivers Ganges and 
Brahmaputra from the Himalaya mountains. But hill country is 
included along the borders of the province. 

To the north the map shows the high tableland of Tibet, edged 
by the Himalaya range, whose southern slopes descend steeply, 
but with many foothills, to the level low-lying plains of the great 
rivers. Eastward of Bengal there is a mountainous belt, rising to 
heights of more than six thousand feet and densely forested, which 
separates the Irrawaddy valley of Burma from the plains of India. 
These mountains throw out a spur westward, which rises a little 
near its end into the Garo hills. The deeply trenched, relatively 

i] Bengal 9 

narrow valley of the Brahmaputra, known as Assam, lies between 
the Garo hills and the Himalayas. The southward drainage from 
the Garo hills forms a deltaic plain, extending nearly to the port 
of Chittagong. This plain, traversed by the Meghna river which 
gathers water from the Garo and Khasi range, is continuous with 
the delta of Bengal proper. 

To the west of Bengal is another hill spur, bearing the name 
of Rajmahal, which is the north-eastern point of the plateau of 
central and southern India. A broad lowland gateway is left 
between the Garo and Rajmahal hills, and through this opening 
the Brahmaputra and Ganges rivers turn southward and converge 
gradually until they join with the Meghna to form a vEist estuary. 
The country west of this estuary is the Bengal delta, traversed 
by many minor channels, which branch from the right bank of 
the Ganges before the confluence with the Meghna. East of 
the estuary is that other deltaic land whose silt is derived from 
the south front of the Garo hills. It is said that the highest rain- 
fall in the world occurs in those hills, when the monsoon sweeps 
northward from the Bay of Bengal, and blows against their face. 
The rainfall on a single day in the rainy season is sometimes as 
great as the whole annual rainfall of London. Little wonder that 
there is abundance of silt for the formation of the fertile plains 
below ! 

The approach to the coast of Bengal, as may be concluded from 
this geographical description, presents little of interest. At the 
entrance to the Hooghly river, the westernmost of the deltaic 
channels, are broad grey mud banks, with here and there a palm 
tree. From time to time, as the ship passes some more solid 
ground, there are villages of thatched huts, surrounded by tall 
green banana plantations. 

Calcutta, the chief port and largest town of modern India, is 
placed no less than eighty miles up the Hooghly on its eastern 
bank. The large industrial town of Howrah stands opposite on 
the western bank. Not a hill is in sight round all the horizon, 
Only the great dome of the post office rises white in the sunshine. 
Calcutta is connected with the jute mills and engineering works of 
Howrah by a single bridge. Below this bridge is the port, always 
thronged with shipping. 

Calcutta has grown round Fort William as a nucleus. The 
present Fort, with its outworks, occupies a space of nearly a 
thousand acres on the east bank of the Hooghly below the Howrah 
bridge. To the north, east, and south, forming a glacis for the 

lO The Sub- Continent of India [ch. 

fort, is a wide green plain, the Maidan, and beyond this is the 
city. The European quarter lies to the east of the Maidan. The 
government offices, and beyond them the great native city, lie to 
the north. Calcutta with more than a million inhabitants exceeds 
Glasgow in size, and is the second city of the British Empire. 

Three hundred miles away to the north, approached from 
Calcutta by the East Bengal railway, is Darjeeling, the hill station 
of Calcutta, as Ootacamund is of Madras. The railway traverses 
the dead level of the plain, with its thickly set villages and tropical 
vegetation. There are some seven hundred and fifty thousand 
villages in India, and they contain about ninety per cent, of the 
total population. The Province of Bengal has a population equal 
to that of Great Britain and Ireland, but concentrated on an area 
less than that of Great Britain without Ireland. Yet it contains 
only one great city, as greatness of cities is measured in the British 

Mid- way from Calcutta to Darjeeling the Ganges is crossed. 
The passage occupies about twenty minutes from one low-lying 
bank to the other. Then the journey is resumed through the rice 
fields, with their clumps of graceful bamboo, until at last the hills 
become visible across the northern horizon. The train runs into a 
belt of jungle at the foot of the first ascent. Passengers change 
to a mountan railway, which carries them up the steep front, with 
many a turn and twist. On the lower slopes is tall forest of teak 
and other great trees, hung thickly with creepers. Presently the 
timber becomes smaller, and tea plantations are passed with trim 
rows of green bushes. Far below, at the foot of the steep forest, 
spreads to the southern horizon the vast cultivated plain. Finally 
trees of the fir tribe take the place of leafy trees, and the train 
attains to the sharp ridge top on which is placed Darjeeling, a 
settlement of detached villas in compounds, hanging on the slopes. 

Darjeeling is about seven thousand feet above sea-level, on an 
east and west ridge, with the plains to the south and the gorge of 
the Rangit river to the north. In the early morning, in fortunate 
weather, the visitor may gaze northward upon one of the most 
glorious scenes in the world. Over the deep valley at his feet, 
still dark in the shade, and over successive ridge tops beyond, rises 
the mighty snow range of the Himalayas, fifty miles away, with the 
peak of Kinchinjunga, more than five miles high, dominating the 
landscape. Behind Kinchinjunga, a little to the west, and visible 
from Tiger hill, near Darjeeling, though not from Darjeeling itself, 
is Mount Everest, the highest mountain in the world, more than 

i] Darjeeling: Sikkim: Assam li 

five and a half miles high. Across the vast chasm and bare granite 
summits in the foreground, the glittering wall of white mountains 
seems to hang in the sky as though belonging to another world. 
The broad distance, and the sudden leap to supreme height, give 
to the scene a mysterious and almost visionary grandeur. It is, 
however, only occasionally that the culminating peaks can be seen, 
for they are often veiled in cloud. 

The people of Sikkim, the native state in the hills beyond 
Darjeeling, are highlanders of Mongolian stock and not Indian. 
They are of Buddhist religion like the Burmans, and not Hindu or 
Muhammadan like the inhabitants of the plains. They are smaU 
sturdy folk, with oblique cut eyes and a Chinese expression, and 
they have the easy going humourous character of the Burmans, 
though not the delicacy and civilisation of those inhabitants of the 
sunny lowland. 

It is an interesting fact that these hill people should belong to 
the race which spreads over the vast Chinese Empire. That race 
here advances to the last hill brinks which overlook the Indian 
lowland. The political map of this part of India illustrates a 
parallel fact. While the plains are administered directly by British 
officials, the mountain slopes descending to them are ruled by 
native princes, whose territories form a strip along the northern 
boundary of India. North of Assam and Bengal we have in 
succession, from east to west in the belt of hill country, the lands 
of Bhutan, Sikkim, and Nepal. From Nepal are recruited the 
Gurkha regiments of the Indian army, the Gurkhas being a race of 
the same small and sturdy hill men as the people of Sikkim. In 
other words, they are of a Mongoloid stock, though of Hindu religion. 

The Rangit river drains from the hills of Darjeeling, and from 
the snow mountains beyond, into a tributary of the Ganges. 
Several hundred such torrents burst in long succession through 
deep portals in the Himalayan foot hills and feed the great rivers 
of the plain. These torrents are perennial, for they originate in 
the melting of the glaciers, and the Himalayan glaciers cover a 
vast area, being fed by the monsoon snows. Nearly all the 
agricultural wealth of northern India owes its origin to the summer 
or oceanic monsoon, which beats against the Himalayan mountain 
edge. That edge, gracefully curving upon the map, extends through 
fifteen hundred miles. The streams which descend from it in long 
series gather into the rivers Brahmaputra, Ganges, and Indus. 

The valley of the Brahmaputra forms the province of Assam. 
Notwithstanding its vast natural resources, Assam is a country 

12 The Suh-Continent of India [CH. 

which, at most periods of its history, has remained outside the 
Indian civilisation. Even to-day it has but a sparse population 
and a relatively small commercial development, for it lies on the 
through road no whither. High and difficult mountains close in 
the eastern end of its great valley. 

The geography of Assam, though very simple, is on a very 
grand scale. The Tsan-po river rises high on the plateau of Tibet 
northward of Lucknow. For more than seven hundred miles it 
flows eastward over the plateau in rear of the Himalayan peaks. 
Then it turns sharply southward, and descends from a great height 
steeply through a deep gorge, until it emerges from the mountains 
at a level not a thousand feet above the sea. At this point, turning 
westward, it forms the Brahmaputra, 'the son of Brahma, the 
Creator.' The Brahmaputra flows for four hundred and fifty miles 
westward through the valley of Assam, deeply trenched between 
the snowy wall of the Himalayas on the one hand and the forested 
mountains of the Burmese border and the Khasi and Garo hills on 
the other hand. The river rolls down the valley in a vast sheet of 
water, depositing banks of silt at the smallest obstruction. Islands 
form and reform, and broad channels break away from the main 
river in time of flood, and there is no attempt to control them. 
The swamps on either hand are flooded in the rainy season, till the 
lower valley is one broad shining sea, from which the hills slope 
up on either side. The traffic on the river is maintained chiefly 
by exports of tea and timber, and imports of rice for the labourers 
on the tea estates. Some day, when great sums of money are 
available for capital expenditure, the Brahmaputra will be con- 
trolled, and Assam will become the seat of teeming production and 
a dense population. The Indian Empire contains three hundred 
and fifteen million people, but it also contains some of the chief 
virgin resources of the world. 

Where the Brahmaputra bends southward round the foot of 
the Garo hills the valley of Assam opens to the plain of Bengal. 
Across that plain westward, where the Ganges makes a similar 
southward bend round the Rajmahal hills, Bengal merges with the 
great plain of Hindustan, which extends westward and north-west- 
ward along the foot of the Himalayas for some seven hundred 
miles to the point where the Jumna, westernmost of the Gangetic 
tributaries, leaves its mountain valley. Hindustan begins with a 
breadth of about a hundred miles between the Rajmahal hills and 
the northern mountains, spreads gradually to a breadth of two 
hundred miles from the foot hills of the Himalayas to the first rise 

i] Hindustan 1 3 

of the Central Indian hills, and then narrows again to a hundred 
miles where it merges with the Punjab plain between the Ridge of 
Delhi and the Himalayas. The great river Jumna-Ganges streams 
southward from the mountains across the head of the plain to 
Delhi, and then gi-adually bends south-eastward and eastward 
along that edge of the plain which is remote from the mountains, 
as though it were pinned against the foot of the Central hills by 
the impact of the successive great tributaries from the north. 
Three of these tributaries are the Upper Ganges itself, whose 
confluence is at Allahabad, and the Gogra and the Gandak which 
enter above Patna. The Jumna-Ganges receives from the south 
the Chambal and Son, long rivers but comparatively poor in water. 

Access to the plains of Hindustan was formerly by the naviga- 
tion of the Ganges and its tributaries. Then the Grand Trunk 
Road was made from Calcutta to Delhi. More recently the East 
Indian Railway has been built from Bengal to the Punjab. Both 
the road and the railway avoid the great bend round the hills by 
crossing the upland to the west of Rajmahal. The road descends 
to the Ganges at Patna, but the railway at Benares, where it 
crosses by the lowest bridge over the Ganges. 

Two great provinces divide the plain of Hindustan between 
them. In the east is Bihar, with its capital at Patna ; in the west 
are the United Provinces of Agra and, Oudh with their capital at 
Allahabad. For administrative purposes Bihar is now joined with 
Orissa, the deltaic plain of the MahanadI river on the coast of 
Bengal. A broad belt of sparsely populated hills separates 
Bihar from Orissa, whereas each of these fertile lowlands opens 
freely to Bengal, the one along the Ganges, and the other along 
the coast. 

When we go from Bengal into Bihar, or from Bihar into the 
United Provinces it is as though we crossed from one to another 
of the great continental states of Europe. The population of 
Bengal is larger than that of France. The population of Bihar 
and Orissa is equivalent to that of Italy. The population of 
the United Provinces is nearly equal to that of Germany since 
the War. 

Five considerable cities focus the great population of the 
United Provinces, Allahabad, Cawnpore, Lucknow, Agra, and 
Benares. Allahabad is built in the angle of confluence between 
the Jumna and the Ganges. A hundred miles above Allahabad, 
on the right or south bank of the Ganges, is the city of Cawnpore, 
and on the opposite or north bank extends the old kingdom of 

14 The Sub- Continent of India [ch. 

Oudh, with Lucknow for its capital, situated some forty miles 
north-east of Cawnpore. Agra, which gives its name to all that 
part of the United Provinces which did not formerly belong to 
Oudh, is situated on the right or south bank of the Jumna, a 
hundred and fifty miles west of Lucknow. All these distances lie 
over the dead level of the plain, dusty and like a desert in the dry 
season, but green and fertile after the rains. Scattered over the 
plain are innumerable villages in which dwell nineteen out of 
twenty of the inhabitants of the United Provinces. 

Eighty miles below Allahabad, on the north bank of the 
Ganges is Benares, the most sacred city of the Hindus. Benares 
extends for four miles along the bank of the river, which here 
descends to the water with a steep brink. Down this brink are 
built flights of steps known as Ghats, at the foot of which pilgrims 
bathe, and dead bodies are burnt. The south bank opposite lies 
low and is not sacred. The word Ghat is identical with the name 
applied geographically to the west and east brinks of the Deccan 

Cawnpore is the chief inland manufacturing city of India, con- 
trasted in all its ways with Benares. But none of these cities are 
really great, when compared with the population of the United 
Provinces. Lucknow is the largest, and has only a quarter of a 
million inhabitants. Notwithstanding the great changes now in 
progress, India still presents in most parts essentially the same 
aspect as in long past centuries. 

If there be one part of India which we may think of as the 
shrine of shrines in a land where religion rules all life, it is to be 
found in the triangle of cities — Benares and Patna on the Ganges, 
and Gaya some fifty miles south of Patna. Benares has been a 
focus of Hinduism from very early times. Patna was the capital 
of the chief Gangetic kingdom more than two thousand years ago 
when the Greek ambassador Megasthenes, first of the westerns, 
travelled thus far into the east. Gaya was the spot where Buddha, 
seeking to reform Hinduism some five hundred years before Christ, 
obtained ' enlightenment,' and then migrated to teach at Benares, 
or rather at Sarnath, now in ruins, three or four miles north of the 
present Benares. The peoples of all the vast Indian and Chinese 
world, from Karachi to Pekin and Tokyo, look to this little group 
of cities as the centre of holiness, whether they be followers of 
Brahma or of Buddha. 

The language of the United Provinces and of considerable 
districts to east, south, and west of them, is Hindi, the tongue of 

i] United Provinces : Central Indian Agency 15 

modern India most directly connected with ancient Sanskrit. 
Hindi is now spoken by a hundred million people in all the north 
centre of India. It is the language not only of Bihar and the 
United Provinces, but also of Delhi and of a wide district in 
Central India drained by the Chambal and Son rivers. Other 
tongues of similar origin are spoken in the regions around — 
Bengali to the east, MarathI and Gujarat! to the south-west 
beyond the Ganges basin, and Punjabi to the north-west. Away 
to the south, beyond the limit of the Sanskrit tongues, in the 
Province of Madras and neighbouring areas, are languages wholly 
alien from Sanskrit. They differ from Hindi, Bengali, Marathi, 
Gujarati, and Punjabi much as the Turkish and Hungarian lan- 
guages differ from the group of allied Indo-European tongues 
spoken in Western Europe. These southern Indian tongues are 
known as Dravidian. The most important of them are Telegu, 
spoken by twenty millions, and Tamil spoken by fifteen millions. 
The Dravidian south, however, and the Aryan north and centre 
agree generally in holding some form of Hinduism or Islam. 

Within the central hills there is a wide district drained north- 
eastward into the Jumna-Ganges chiefly by the rivers Chambal 
and Son. This district, much less fruitful than the plain of Hindu- 
stan, because less abundantly watered, and composed of rocky 
ground instead of alluvium, is ruled by native chiefs. The British 
suzerainty is exercised under the Viceroy by the Central Indian 
Agency. Of the chiefs of Central India the most important are 
Sindhia and Holkar, two Marathas ruling Hindi populations. 
Sindhia's capital, Gwalior, lies a little south of Agra. It is domi- 
nated by an isolated rock fort, flat topped and steep sided, more 
than three hundred feet in height. Indore, Holkar's capital, lies 
in the land of Malwa, on high ground about the sources of the 
Chambal river, a considerable distance south of Gwalior. In the 
neighbourhood is Mhow, one of the chief cantonments of the 
Indian army, placed on the high ground for climatic reasons, like 
Bangalore in southern India. 

The long upward slope to the Chambal headstreams ends on 
the summit of the Vindhya range, a high brink facing southward. 
From east to west along the foot of the Vindhya face runs the 
sacred river Narbada in a deeply trenched valley. Thus the 
Narbada has a course at right angles to the northward floAving 
Chambal streams on the heights above. The Son river occupies 
almost the same line of valley as the Narbada, but flows north- 
eastward into the Ganges. On the south side of the Narbada 

1 6 The Sub- Contineitt of India [ch. 

valley is the Satpura range, parallel with the Vindhya brink, and 
beyond this is the Tapti river, shorter than the Narbada, but flow- 
ing westward with a course generally parallel to that of the sacred 
river. The Narbada and Tapti form broad alluvial flats before 
they enter the side of the shallow Gulf of Carabay. South of the 
Tapti begins the Deccan Plateau. 

Thus a line of hills and valleys crosses India obliquely from 
Rajmahal to the Gulf of Cambay, and divides the rivers of the 
Indian Upland into three systems. North of the Vindhya brink, 
over an area as large as Germany, the drainage descends north- 
eastward to the Jumna-Ganges. Between the Vindhya range and 
the edge of the Deccan Plateau are the two exceptional rivers, 
Narbada and Tapti, flowing westward in deeply trenched valleys. 
From the Western Ghats, and from the hills which cross India south 
of the Tapti and Son to Rajmahal, three great rivers flow south- 
ward and eastward to the Bay of Bengal — the Mahanadi, Godavari, 
and Kistna. The area drained by these three streams of the plateau 
is a third of India. 

The first * factory ' of the English East India Company was at 
Surat on the lower Tapti, but Bombay, two hundred miles farther 
south, long ago supplanted Surat as the chief centre of European 
influence in Western India. The more northern town had an easy 
road of access to the interior by the Tapti valley, but the silt at 
the river mouth made it difficult of approach from the sea. Bombay 
oflfered the security of an island, and has a magnificent harbour 
between the island and the mainland, far from the mouth of any 
considerable stream. 

Two new facts have of recent years altered all the relations of 
India with the outer world, and have vitally changed the con- 
ditions of internal government as compared with those prevailing 
even as late as the Mutiny. The first of these facts was the open- 
ing of the Suez Canal, and the second was the construction, and 
as regards main lines the virtual completion, of the Indian railway 
system. Formerly shipping came round the Cape of Good Hope, 
and it was as easy to steer a course for Calcutta as for Bombay. 
To-day only bulky cargo is taken from Suez and Aden round the 
southern point of India through the Bay of Bengal to Calcutta. 
The fast mail boats run to Bombay, and thence the railways 
diverge south-eastward, north-eastward, and northward to all the 
frontiers of the Empire. Only the Burmese railways remain for 
the present a detached system. But in regard to tonnage of traffic 
Calcutta is still the first port of India, for the country which lies in 

i] Bombay 17 

rear of it — Bengal, Bihar, and the United Provinces — contains 
more than a hundred million people. 

From Bombay inland runs the Great Indian Peninsula Railway. 
The line branches a short distance from the coast, striking on the 
one hand south-eastward in the direction of Madras, and on the 
other hand north-eastward in the direction of Allahabad on the 
East Indian Railway. Each week, a few hours after the arrival of 
the mail steamer at Bombay, three express trains leave the 
Victoria Station of that city. One of them is bound south-east- 
ward for Madras. The second runs north-eastward to Allahabad, 
and then on to Howrah for Calcutta. The third also runs north- 
eastward, but diverges northward from the Calcutta route to Agra 
and Delhi. When the Government of India is at Simla the last 
mentioned train continues beyond Delhi to the foot of the moun- 
tains. The time taken to Madras is twenty-six hours, to Calcutta 
thirty-six hours, and to Delhi twenty-seven hours. Recently a more 
direct line has been made from Bombay to Calcutta which does 
not pass through Allahabad, but through Nagpur. It traverses a 
hilly country, much forested and relatively thinly peopled, in the 
upper basins of the Godavari and Mahanadi rivers. 

The two lines of the Great India Peninsula system approach 
one another from Allahabad and from Madras at an angle. They 
are carried separately down the steep mountain edge of the 
Deccan Plateau by two passes, the Thalghat and the Borghat, 
which have put the skill of engineers to the test. The junction is 
in the narrow coastal plain at the foot of the mountains. Thence 
the rails pass by a bridge over a sea strait into Salsette Island, 
and by a second bridge over a second strait into Bombay Island. 

The island of Bombay is about twelve miles long from north to 
south. At its southern end it projects into the southward Colaba 
Point and the south-westward Malabar Point, between which, 
facing the open sea, is Back Bay. The harbour, set with hilly 
islets, lies between Bombay and the mainland, the entry being 
from the south round Colaba Point. Bombay is now a very fine 
city, but like the other great seaports of India, it is new — as time 
goes in the immemorial East. Calcutta was already great when 
Bombay was but a small place, for a riverway extends through 
densely peopled plains for a thousand miles inland from Calcutta, 
whereas the horizon of Bombay is barred beyond the harbour by 
the mountain face of the Western Ghats. The real greatness of 
Bombay came only with the opening of the Suez Canal, and of the 
railway lines up the Borghat and the Thalghat. 

c. H. I. I. 2 

1 8 The Sub-Continent of India [CH. 

The train works up the Ghats from Bombay through thick 
forests, and if it be the rainy season past rushing waterfalls, until 
it surmounts the brink top and comes out on to the plain of the 
Deccan table-land in the relative drought of the upper climate. 
The Western Deccan in rear of Bombay constitutes the Maratha 
country. The Marathas are the southernmost of the peoples of 
Indo-European speech in India. Their homeland on the plateau, 
round the city of Poena, now forms the main portion of the 
Province of Bombay. The landscape of the plateau lies widely 
open, studded here and there with table-topped mountains, not 
unlike the kopjes of South Africa. These steep-sided isolated 
mountain blocks have often served as strongholds in warfare. 

South-eastward of Poena, but still on the plateau country, is 
Hyderabad, the largest native state in India. It is ruled under 
British suzerainty by the Nizam. The majority of the Nizam's 
subjects speak Telegu and are of Hindu faith, but the Nizam is a 
Muhammadan. Near his capital, Hyderabad, is Golconda Fort, 
rising above the open plateau with flat top and cliff" sides. The 
name of Golconda has become proverbial for immensity of wealth. 
Formerly it was the Indian centre of diamond cutting and polishing. 

The wide Deccan Plateau is in most parts of no great fertility. 
Over large areas it is fitted rather for the pasture of horses and 
cattle than for the plough. Agriculture is best in the river valleys. 
But there is one large district lying on the plateau top east of 
Bombay, and on the hill tops north and south of the Narbada 
valley which is of a most singular fertility. The usually granitic 
and schistose rocks of the plateau have here been overlaid by 
great sheets of basaltic lava. Detached portions of these lava 
beds form the table tops of most of the kopje-like hills. The lava 
disintegrates into a tenacious black soil, which does not fall into 
dust during the dry season, but cracks into great blocks which 
remain moist. As the dry season advances these blocks shrink, 
and the cracks grow broader, so that finally it is dangerous for a 
horse to gallop over the plain, lest his hoof should be caught in 
one of these fissures. 

This remarkable earth is known as the Black Cotton Soil. The 
cotton seeds are sown after the rains, and as the young plant 
grows a clod of earth forms round its roots which is separated 
from the next similar clod by cracks. Wheat is grown on this soil 
in the same manner, being sown after the rainy season and reaped 
in the beginning of the hot season, so that from beginning to end 
the crop is produced without exposure to rain, being drawn up by 

i] Central Provinces : Baroda 1 9 

the brilliant sunshine, and fed at the root by the moisture pre- 
served in the heavy soil. 

Thus in the part of India which lies immediately east, north- 
east, and north of Bombay the lowlands and the uplands are alike 
fertile — the lowlands round Ahmadabad and Baroda, and in the 
valleys of the Narbada and Tapti rivers, because of their alluvial 
soil, and the uplands round Poona and Indore because they are 
clothed with the volcanic cotton soil. 

The east coast of India, where it trends north-eastward from 
the mouths of the Godavari river to those of the Mahanadi, is 
backed by great hill and forest districts, tenanted by big game 
and by uncivilised tribes of men. The Eastern Ghats are here 
higher than elsewhere, and they approach near to the coast, so 
that their foot plain affords only a relatively narrow selvage of 
populated country. Through this coastal plain the railway is 
carried from Calcutta to Madras. 

The reason for the primitive character of this part of the 
country, and of many of the districts which extend northward 
through the hills almost to the valley of the Son, is to be found in 
the conditions of soil and climate. There have been no volcanic 
outpourings on the gneissic and granitic rocks hereabouts, and the 
summer cyclones fi'om the Bay of Bengal strike most frequently 
upon this coast and travel inland in a north-westerly direction. 
Some of the Gond tribes of the forests, who may perhaps be 
described as the aborigines of India, still speak tongues which 
appear to be older than Dravidian. In the more fertile parts of 
the upper Mahanadi and Godavari basins are comprised the Central 
Provinces of the direct British Raj, whose capital is at Nagpur. 
The Central Provinces have an area comparable with that of 
Italy, though their population is but one-third the Italian popu- 
lation. They must not be confused with the Central Indian Agency. 

We return to the west coast. The Bombay and Baroda railway 
runs out of Bombay northward and does not ascend the Ghats, but 
follows the coastal plain across the lower Tapti and Narbada 
rivers to Baroda, and thence on, across the alluvial flats of the 
Mahi and neighbouring small rivers, to Ahmadabad. The Gaikwar 
of Baroda governs a small but very rich and populous lowland. 
His people speak GujaratI, though the Gaikwar is a Maratha, like 
Sindhia and Holkar. His territories are so mixed with those of 
the Bombay Presidency that the map of the plains round Ahmada- 
bad and Baroda city is like that part of Scotland which is labelled 
Ross and Cromarty. Ahmadabad was once the most important 


20 The Sub-Conti7tent of India [CH. 

Muhammadan city of Western India, and contains many fine 
architectural monuments, surpassed only by those of the great 
Mughal capitals, Delhi and Agra. 

Westward of the alluvial plains of Gujarat, and beyond the Gulf 
of Cambay, is the peninsula of Kathiawar, a low plateau, lower 
considerably than the Deccan, but clothed in part with similar 
sheets of fertile volcanic soil. Baroda has territory in Kathiawar, 
as has also the Presidency of Bombay, but in addition there are a 
multitude of petty chieftainships. North of Kathiawar is another 
smaller hill district, constituting the island of Cutch. The Rann 
of Cutch, a marshy area communicating with the sea, separates 
the island from the mainland. Apart from Travancore and Cochin 
in the far south, Kathiawar and Cutch are the only part of India 
where Feudal States come down to the coast. There are a few 
diminutive coastal settlements belonging to the French and Portu- 
guese governments, but these are too insignificant to break the 
general rule that the shores of India are directly controlled by the 
British Raj. The largest of the foreign European settlements is at 
Goa on the west coast south of Bombay. Goa has a fine harbour 
but the Ghats block the roads inland. 

We have now completed the itinerary of the inner parts of 
India. What remains to be described is the north-western land of 
passage where India merges with Iran and Turan — Persia and 
Turkestan. The Himalayan barrier, and the desert plateau of 
Tibet in rear of it, so shield the Indian world from the north and 
north-east that the medieval Buddhist pilgi'ims from China to 
Gaya were in the habit of travelling westward by the desert routes 
north of Tibet as far as the river Oxus, and then southward over 
the Hindu Kush. Thus they came into India from the north-west, 
having circumvented Tibet rather than cross it. Great mountain 
ranges articulate with the Himalayas at their eastern end, and 
extend into the roots of the peninsula of Further India. Thus the 
direct way from China into India by the east is obstructed. To- 
day as we have seen the railway systems of Burma and India are 
still separate. 

The centre of north-western India is occupied by a group of 
large Native States, known collectively as Rajputana. Through 
Rajputana, diagonally from the south-west north-eastward, there 
runs the range of the Aravalli hills for a distance of fully three 
hundred miles. The north-eastern extremity of the Aravallis is 
the Ridge of Delhi on the Jumna river. At their southern end, 
but separated from the main range by a hollow, is the isolated 

i] The Great Indian Desert 2 1 

Mount Abu, the highest point in Rajputana, standing up con- 
spicuously from the surrounding plains to a height of some five 
thousand feet. 

East of the Aravallis, in the basin of the Chambal tributary of 
the Jumna-Ganges, is the more fertile part of Rajputana, with the 
cities of Jaipur, Ajmer, Udaipur, and the old fortress of Chitor. 
Beyond the Chambal river itself, but within its basin, are Gwalior 
and Indore, the seats of the princes Sindhia and Holkar. But 
Gwalior and Indore belong to the Central Indian Agency and not 
to Rajputana. 

West of the Aravalli hills is the great Indian desert, prolonged 
seaward by the salt and partly tidal marsh of the Rann of Cutch. 
In oases of this desert are some of the smaller Rajput capitals, 
notably Bikaner. Beyond the desert flows the great Indus river 
through a land which is dry, except for the irrigated strips beside 
the river banks and in the delta of Sind below Hyderabad. South 
of Mount Abu streams descend from the end of the Aravalli hills 
to the Gulf of Cambay through the fertile lowland of Ahmadabad, 
sunk like a land strait between the plateau of Kathiawar to the 
west and the ends of the Vindhya and Satpura ranges to the 
east. The Aravallis are the last of the Central Indian hills 
towards the north-west. Outside the Aravallis the Indus valley 
spreads in wide low-lying alluvial plains, like those of the Ganges, 
but dry. 

It would be difficult to exaggerate the importance to India of 
the existence of the great desert of Rajputana. Tlie ocean to the 
south-east and south-west of the peninsula was at most times an 
ample protection against overseas invasion, until the Europeans 
rounded the Cape of Good Hope. The vast length of the Himalaya, 
backed by the desert plateau of Tibet, was an equal defence on 
the north side. Only to the north-west does India lie relatively 
open to the incursions of the warlike peoples of Western and 
Central Asia. It is precisely in that direction that the Indian 
desert presents a waterless void extending north-eastward from 
the Rann of Cutch, for some 400 miles, with a breadth of 150 miles. 
In rear of the desert a minor bulwark is constituted by the Aravalli 

Only between the north-eastern extremity of the desert and 
the foot of the Himalayas below Simla is there an easy gateway 
into India. No river traverses this gateway, which is on the 
divide between the systems of the Indus and the Jumna-Ganges. 
Delhi stands on the west bank of the Jumna at the northern 

22 The Sub-Contine7tt of India [ch. 

extremity of the Aravallis, just where the invading forces from the 
north-west came through to the navigable waters. 

Aided by such powerful natural conditions the Rajputs — ^the 
word means 'sons of princes' — were during many centuries the 
defenders of India against invasion by the direct road to Delhi. 
Unable at last to stem the tide of Musalman conquest, they have 
maintained themselves on the southern flank of the advance, and 
to-day some of their princely families claim to trace their lineage 
back in unbroken descent from ancestors before the Christian 
era. The descendants of conquerors who had won their kingdoms 
with the sword, they remain even now proud aristocratic clans 
holding a predominant position in the midst of a population far 
more numerous than themselves. 

Narrow gauge lines branch through Rajputana in the direction 
of Delhi, past the foot of Mount Abu, which rises like an island of 
granite from amid the sandy desert. The top of Abu is a small 
rugged plateau, measuring fourteen miles by four, in the midst of 
which is the Gem Lake, a most beautiful sheet of water, set with 
rocky islands and overhung with great masses of rock. The house 
of the Resident of Rajputana is on its shore, for Mount Abu is the 
centre from which Rajputana is controlled, so far as is necessary, 
by the advice of the Viceroy. The summit of Abu also bears some 
famous iniins of Jain temples. 

Some of the most beautiful cities of India are in Rajputana. 
Udaipur stands beside a lake, with its palaces and ghats reflected 
in the clear waters. Ajmer, now under direct British rule, is set 
in a hollow among low hills, and is surrounded by a wall. Here 
also there is a lake, and upon its banks are marble pavilions. 
Jaipur is a walled city, surrounded by rocky hills crowned with 
forts. The streets are broad, and cross one another at right 

The Rajputana Agency is as large as the whole British Isles, 
but it contains only about ten million people, since a great part of 
it is desert. The Central Indian Agency is about as large as 
England and Scotland without Wales. It has a population only 
a little smaller than that of Rajputana. W€ may measure the 
significance of the more important chiefs in these two Agencies by 
the fact that Sindhia rules a country little less, either in area or 
population, than the kingdom of Scotland. 

From Rajputana we come to Delhi, which may truly be called 
the historical focus of all India ; for, as we have seen, it commands 
the gateway which leads from the Punjab plain to Hindustan, the 

i] The Delhi Gateway 23 

plain of the Jumna and the Ganges. Here the fate of invasions 
from India from the north-west has been decided. Some have 
either never reached this gateway or have failed to force their 
way through it. The conquests of Darius in the latter part of 
the sixth century B.C., and of Alexander the Great in the years 
327-5 B.O., were not carried beyond the Punjab plain. Such direct 
influence as they exercised in modifying the character of Indian 
civilisation must therefore have been confined to this region. On 
the other hand, the invasions which have succeeded in passing the 
gateway and in effecting a permanent settlement in Hindustan 
have determined the history of the whole sub-continent. These 
belong to two groups, the Aryan and the Musalman, distinguished 
by religion, language, and type of civilisation, and separated from 
each other by an interval of probably some two thousand years. 

For the chronology of the Aryan conquests, which may well 
have extended over many generations or even centuries, we possess 
no certain dates. All the knowledge which we can hope to gain 
of the history of this remote period must be gleaned from the study 
of the ancient scriptures of these Aryan invaders. 

The course of JNIusalman invasion, which entailed consequences 
of perhaps equal importance, may be traced with greater precision. 
If we reckon from the Arab conquest of Sind in 712 A.D. to the 
establishment of the Sultanate of Delhi in 1193 A.D., we shall see 
that nearly five centuries elapsed before Musalman conquest spread 
from the confines through the Delhi gateway into the very heart 
of India. During this long period it was held in check by the 
Rajput princes ; and their ultimate failure to impede its progress 
was due to internal discord which has always been the bane of 
feudal confederations. 

So Delhi, founded by the Rajputs in the neighbourhood of 
Indraprastha (the modern Indarpat), the capital of the Kurus 
in the heroic ages celebrated in India's great epic poem, the 
Mahabharata, passed into the hands of the invading Musalmans 
and with it passed the predominant power in India. 

What Benares, and Patna, and Gaya were and are to the 
Brahman and Buddhist civilisations native to India, what Calcutta, 
and Madras, and Bombay, and Karachi are to the English from 
over the seas, that were Delhi and Agra to the Musalmans entering 
India from the north-west. 

More than three centuries and a quarter later another Musal- 
man invasion, more effective than the former, came into India 
by way of Delhi. The Mughals or Mongols of Central Asia had 

24 I'f^^ Sub- Continent of India [ch. 

been converted to Islam, and in the time of our King Henry VIII 
they refounded the Musalman power at Delhi. For a hundred 
and fifty years, from the time of our Queen Elizabeth to that of 
our Queen Anne, a series of Mughal emperors, from Humayun to 
Aurangzeb, ruled in splendid state at Delhi over the greater part 
of India. Agra, a hundred miles lower down the Jumna, became 
a secondary or alternative capital, and in these two cities we have 
to-day the supreme examples of Muhammadan architectural art. 

More than sixty-two millions of the Indian population hold the 
faith of Islam. They are scattered all over the land, usually in a 
minority, but frequently powerful, for Islam has given ruling chiefs 
to many districts which are predominantly Hindu. Only in two 
parts of India are the Musalmans in a majority, namely in the east 
of Bengal about Dacca, and in the Indus basin to the north-west. 
We may think of the Indus basin — lying beyond the desert, low 
beneath the uplands of Afghanistan — as being an ante-chamber to 
India proper. In this ante-chamber, for more than nine hundred 
years the Musalmans have been a majority. 

When the decay of the Mughal Empire began in the time of our 
Queen Anne, the chief local representatives of the imperial rule, 
such as the Nizam of Hyderabad, and the Nawabs of Bengal and 
Oudh, assumed an independent position. It was with these new 
dynasties that the East India Company came into conflict in the 
days of General Clive. Thus we may regard the British Empire 
in India as having been built up from the fragments into which 
the Mughal Empire broke. In one region, however, the Western 
Deccan, the Hindus reasserted themselves, and there was a rival 
bid for empire. From the neighbourhood of Poena the Marathas 
conquered eastward to the borders of Bengal, and northward to 
the walls of Delhi. It was the work of Lord Lake and General 
Wellesley to defeat the Marathas. 

North-westward of Delhi, in the gateway between the desert 
and the mountains, the ground is sown over with battlefields — 
ancient battlefields near the Jumna, where the incoming Musalmans 
overthrew the Indian resistance, and modern battlefields near the 
Sutlej, where advancing British power inflicted defeat upon the 
Sikhs. It is by no accident that Simla, the residence of the British 
Viceroy during half the year, is placed on the Himalayan heights 
above this natural seat of empire and of struggle for empire. 

In the Mutiny of 1857 the Sikhs of the Punjab remained loyal 
to the British rule, although they had been conquered in terrible 
battles on the Sutlej less than ten years before. So it happened 

i] Delhi: Hardwar : Nepal 25 

that some of the British forces in the Punjab were free to march 
to recapture Delhi, which had been taken by the mutineers. Thus 
the Indian Mutiny was overcome from two bases ; on the one hand 
at Lucknow and Cawnpore by an army from Calcutta and the sea ; 
and on the other hand at Delhi by an army advancing from the 
Punjab over the track beaten by many conquerors in previous 

The river Jumna runs past Delhi with a southward course, and 
is there crossed by a great bridge, over which the East Indian 
Railway runs from Delhi through the United Provinces and Bengal 
to Howrah, opposite Calcutta. West of Delhi is the last spur of 
the Aravalli hills, the famous Ridge of Delhi, striking north-east- 
ward to the very bank of the river. The city lies in the angle 
between the Ridge and the Jumna. To the north, in the point of 
the angle, is the European quarter ; in the centre is Shahjahanabad, 
the modern native Delhi ; southward of the modem city is Firozabad, 
or ancient Delhi. Between Shahjahanabad and the river is the 

The plain southward of Firozabad continues to widen between 
the river and the hills, and is strewn over with still more ancient 
ruins. To the west of these, at the foot of the hills, and in part 
upon them, is the site chosen for the new imperial capital of 
British India. Finally, eleven miles south of Delhi are the buildings 
of the Kutb Minar, where are some of the few remains of the early 
Hindu period. 

A hundred miles north-north-east of Delhi is Hardwar on the 
Ganges, at the point where the river leaves the last foothills of 
the Himalaya and enters the plain. Hardwar is the rival of 
Benares as a centre of Hindu pilgrimage for the purpose of ablu- 
tion in the sacred waters. At the annual fair are gathered hundreds 
of thousands of worshippers. The great day at Hardwar is near 
the end of March when the Hindu year begins. Then, according 
to tradition, the Ganges river first appeared from its source in the 
mountains. The water at Hardwar is purer than at Benares in 
the plain. It flows swiftly and is as clear as crystal. 

From near Darjeeling until near Hardwar the foothills of the 
Himalaya for five hundred miles belong to the Gurkha kingdom of 
Nepal, whose capital is Katmandu. Notwithstanding its close con- 
nexion with the Indian army, Nepal is counted as an independent 
state, over which British suzerainty does not formally extend. 
From Hardwar, however, for seven hundred miles north-westward 
to where the Indus breaks from the mountains, the foothills 

26 The Sub-Continent of India [ch. 

belong to the Empire, and upon them stand, high above the plain, 
a series of hill stations. The first of these stations is Mussoorie, 
not far northward of Hardwar. Mussoorie is about a mile above 
sea level. Close by, but lower down, is Dehra Dun, the head- 
quarters of the Gurkha Rifles. Hereabouts the Tarai, an elephant- 
haunted jungle belt, follows the foothills, separating them from 
the cultivated plains. A hundred miles farther along the mountain 
brink is Simla, the summer capital of India, high on a spur above 
the divide between the Indus and the Ganges. The snow often 
rests on the ground in the winter at Simla. 

Immediately to the north of Simla the Sutlej, tributary to the 
Indus, trenches a way out of the mountains, and where it issues on 
to the plain is the oflF-take of a great system of irrigation canals. 
The lowland north-westward of Delhi has a sparse rainfall, for the 
monsoon has lost much of its moisture thus far north-westward 
from the Bay of Bengal. As a result of the construction of the 
irrigation canals colonies have been established between the Sutlej 
and the Jumna, and wheat is grown on thousands of square miles 
that were formerly waste. India has a great population, but with 
modern methods of water supply, and more advanced methods of 
cultivation, there is still ample room for settlement within its 

Two Sikh Feudal States, Patiala and Nabha, are included within 
the area now irrigated from the Sutlej, but Amritsar, the holy city 
of the Sikhs, lying beyond the Sutlej, about two hundred and fifty 
miles from Delhi, is under the immediate British Raj. Fifty miles 
west of Amritsar is Lahore, the old Musalman capital of the 
Punjab. We conquered the Punjab from the Sikhs, but for many 
centuries it had been ruled by the Musalmans. In the break-up of 
the Mughal Empire during the eighteenth century, invaders came 
from Persia and from Afghanistan, who carried devastation even 
as far as Delhi. In their wake, with relative ease, the Sikhs, 
contemporaries of the Marathas of Poona, established a dominion 
in the helpless Punjab. They extended their rule also into the 
mountains of Kashmir, north of Lahore. 

In all the British Empire there is but one land frontier on 
which warlike preparation must ever be ready. It is the north- 
west frontier of India. True that there is another boundary 
even longer, drawn across the American continent, but there 
fortunately only customs-houses are necessary, and an occasional 
police guard. The north-west frontier of India, on the other 
hand, lies through a region whose inhabitants have been recruited 

MAP 2 







The Cambridge History of India, Vol. I 


Map 1 


George Philip & Son,l5.'^ 

i] The North-West Frontier 27 

throughout the ages by invading warlike races. Except for the 
Gurkha mountaineers of Nepal, the best soldiers of the Indian 
army are drawn from this region, from the Rajputs, the Sikhs, the 
Punjabi Musalmans, the Dogra mountaineers north of the Punjab, 
and the Pathan mountaineers west of the Punjab. The provinces 
along this frontier, and the Afghan land immediately beyond it, 
are the one region in all India from which, under some ambitious 
lead, the attempt might be made to establish a fresh imperial rule 
by the overthrow of the British Raj. Such is the teaching of 
history, and such the obvious fate of the less warlike peoples of 
India, should the power of Britain be broken either by warfare on 
the spot, or by the defeat of our navy. Beyond the north-west 
frontier, moreover, in the remoter distance, are the continental 
powers of Europe. 

The Indian army and the Indian strategical railways are there- 
fore organised with special reference to the belt of territory which 
extends north-east and south-west beyond the Indian desert, and 
is traversed from end to end by the Indus river. This frontier 
belt divides naturally into two parts. Inland we have the Punjab, 
where five rivers — the Indus, Jhelum, Chenab, Ravi, and Sutlej — 
emerging from their mountain valleys, gradually close together 
through the plain to form the single stream of the Lower Indus ; 
seaward we have Sind, where the Indus divides into distributaries 
forming a delta. 

Sind is a part of the Bombay Presidency, for it is connected 
with Bombay by sea from the port of Karachi. Of late a railway 
has been constructed from Ahmadabad, in the main territory of 
Bombay, across the southern end of the desert to Hyderabad, at 
the head of the Indus delta. The Punjab is a separate province, 
with its own lieutenant-governor at Lahore, and a population as 
large as that of Spain. 

To understand the significance of the north-west frontier of 
India we must look far beyond the immediate boundaries of the 
Empire. Persia, Afghanistan, and Baluchistan form a single plateau, 
not so lofty as Tibet, but still one of the great natural features of 
Asia. This plateau in its entirety is most conveniently known as 
Iran. On all sides the Iranian plateau descends abruptly to low- 
lands or to the sea, save in the north-west, where it rises to the 
greater heights of Armenia, and in the north-east, where it rises to 
the lofty Pamirs. Southward and south-westward of Iran lie the 
Arabian sea and the Persian gulf, and the long lowland which is 
traversed by the rivers Euphrates and Tigris. Northward, to the 

28 The Sub-Conti?ient of India [CH. 

east of the Caspian sea, is the broad lowland of Turkestan or 
Turan, traversed by the rivers Oxus and Jaxartes, draining into 
the sea of Aral. Eastward is the plain of the Indus. The defence 
of India from invasion depends in the first place on the main- 
tenance of British sea-power in the Persian gulf and the Indian 
ocean, and in the second place on our refusal to allow the establish- 
ment of alien bases of power on the Iranian plateau, especially on 
those parts of it which lie towards the south and east. 

In the north-east corner of Iran, west of the Punjab, a great 
triangular bundle of mountain ridges splays out westward and 
southward from the north-east. These ridges and the intervening 
valleys constitute Afghanistan. Flowing from the Afghan valleys 
we have on the one hand the Kabul river, which descends eastward 
to the Indus, and on the other hand the greater river Helmand, 
which flows south-westward into the depressed basin of Seistan in 
the very heart of Iran. There the Helmand divides into many 
channels, forming as it were an inland delta, from which the waters 
are evaporated by the hot air, for there is no opening to the sea. 
The valley of the Kabul river on the one hand, and the oasis of 
Seistan on the other, might in the hands of an enemy become 
bases wherein to prepare for the invasion of India. Therefore, 
without annexing this intricate and difficult upland, we have 
declared it to be the policy of Britain to exclude from Afghanistan 
and from Seistan all foreign powers. 

There are two lines, and only two, along which warlike in- 
vasions of N.W. India have been conducted in historical times. 
On the one hand the mountains become very narrow just north of 
the head of the Kabul river. There a single though lofty ridge, the 
Hindu Kush, is all that separates the basin of the Oxus from that 
of the Indus. Low ground, raised only a few hundred feet above 
the sea, is very near on the two sides of the Hindu Kush. There 
are several ways into Indisl over this great but single range and 
down the Kabul valley. The most famous is known as the Khyber 
route, from the name of the last defile through which the track 
descends into the Indian plain. 

The other route of invasion lies five hundred miles away to the 
west and south-west There the Afghan mountains come suddenly 
to an end, and an easy way lies round their fringe for four hundred 
miles over the open plateau, from Herat to Kandahar. This way 
passes not far from Seistan. South-eastward of Kandahar it 
descends through a mountainous district into the lowland of the 
Indus. This is now called the Bolan route, from the last gorge 

i] Routes leading into N.W, India 29 

towards India ; but in ancient times the road went farther south 
over the Mula Pass. It debouches upon the plain opposite to the 
great Indian desert. Therefore the Khyber route has been the 
more frequently trodden, for it leads directly, between the desert 
and the mountains, upon the Delhi gateway of inner India. 

Another line of communication connecting India with Persia 
passes through the Makran, or the barren region lying along the 
coast of Baluchistan, This route was much frequented by Arab 
traders in the Middle Ages ; and by it at an earlier epoch Alexander 
the Great led back one detachment of his forces with disastrous 
results. But apart from this return march, and the Indian 
expeditions of Semiramis and of Cyrus which it was designed to 
emulate and which may or may not be historical, this route seems 
not to have been followed by any of the great invasions of India 
in historical times. 

The practical significance of all this geography becomes evident 
not only when we study the history of Ancient India but also 
when we consider the modern organisation of the Indian defensive 
forces. They are grouped into a northern and a southern army. 
The northern army is distributed from Calcutta past Allahabad 
and Delhi to Peshawar, the garrison city on the frontier. All the 
troops stationed along this line may be regarded as supporting the 
brigades on the Khyber front. The southern army is similarly 
posted with reference to Quetta on the Bolan route. It is dis- 
tributed through the Bombay and Madras Presidencies, A'hence 
Quetta can be reinforced by sea through the port of Karachi. 

The conditions of the defence of India have been vitally changed 
by the construction of the North- Western Railway from Karachi 
through the Indus basin, with branches towards the Bolan and the 
Khyber. To-day that defence could be conducted over the sea 
directly from Britain through Karachi, so that the desert of 
Rajputana would lie between the defending armies and the main 
community of India within. 

Karachi stands at the western limit of the Indus delta, in a 
position therefore comparable to that of Alexandria beside the 
Nile delta. The railway keeps to the west of the river for more 
than three hundred miles as far as Sukkur, where is the Lansdowne 
bridge, eight hundred and forty feet long, between Sukkur and 
Rohri on the east bank. This is the very heart of the rainless 
region of India. During twelve years there were only six showers 
at Rohri. A scheme is under consideration for damming the Indus 
near this point, in order that the irrigation canals below may be 

30 The Sub- Continent of India [ch. 

fed, not only in time of flood as at present, but in the season of 
low water as well. 

From Sukkur a branch railway traverses the desert north- 
westward to the foot of the hills below the Bolan pass. This part 
of the desert occupies a re-entering angle of lowland, with the 
mountains of Afghanistan to the north and those of Baluchistan 
to the west. On the map, the Afghan ranges have the effect of 
being festooned from the Bolan eastward and northward. The 
railway ascends to Quetta either by the Mushkaf valley — the 
actual line of the Bolan torrent having been abandoned — or by a 
longer loop line, the Harnai, which runs to the Pishin valley, north 
of Quetta. The latter is the usual way. By the Mushkaf route 
the line is carried over a boulder-strewn plain about half a mile 
broad in the bottom of a gorge, with steeply rising heights on 
either side. Here and there the strip of lower ground is trenched 
and split by deep canyons. At first the rails follow the Mushkaf 
river, and the gradients are not very severe, but once Hirok, at 
the source of the Bolan river, is passed, a gradient of one in 
twenty-five begins, and two powerful engines are required to 
drag the train up. The steep bounding ridges now close in on 
either side, with cliffs rising almost perpendicularly to several 
hundred feet. Occasional blockhouses high up amid the crags 
defend the pass. 

The gradients of the Harnai route are not quite so steep as 
those 6f the Mushkaf. Should either way be blocked or carried 
away by landslips or floods, the other would be available. The 
Harnai line passes through the Chappar rift, a precipitous gorge 
in a great mass of limestone. The old Bolan gorge way of the 
caravans was dangerous because of the sudden spates which at 
times filled all the bottom between the cliffs. 

Quetta lies about a mile above sea-level in a small plain, sur- 
rounded by great mountains rising to heights of two miles and 
more. Irrigation works have been constructed, so that Quetta is 
now an oasis amid desert mountains. It has a population of some 
thirty thousand. The Agent General for British Baluchistan 
resides there. The town is very strongly fortified, for it commands 
the railways leading from the Khojak pass down into India. 
Quetta and Peshawar are the twin keys of the frontier. 

From Quetta there is a railway north-westward for another 
hundred and twenty miles to Chaman on the Afghan frontier, 
where is the last British outpost. This line pierces the Khojak 
ridge by a tunnel and then emerges on the open upland plain of 


i] Plain of the Indus : Peshawar 3 1 

Iran. The rails are kept ready at Chaman for the continuation of 
the track to Kandahar, seventy miles further. 

We return to Rohri on the Indus. The North- Western Railway 
now runs to the east of the river and soon enters the Punjab. Not 
very long ago all this land was a desert. To-day, as the result of 
a great investment of British capital, irrigation works have changed 
the whole aspect of the country. The plain of the Indus has become 
one of the chief wheat fields of the British Empire, for wheat is the 
principal crop in the Punjab, in parts of Sind, and — outside the 
basin of the Indus itself — in the districts of the United Provinces 
which lie about Agra. The wheat production of India on an 
average of years is five times as great as that of the United 
Kingdom, and about half as great as that of the United States. 
In the three years 1910-12 the export of wheat from India to the 
United Kingdom exceeded that from the United States to the 
United Kingdom. 

The brown waste of the plains of the Punjab becomes, after 
the winter rains, a waving sea of green wheat, extending over 
thousands of square miles. Far beyond the area within which the 
rainfall alone suffices, the lower Punjab and the central strip of 
Sind have been converted into a second Egypt. Though the 
navigation of the Indus is naturally inferior to that of the Ganges, 
yet communication has been maintained by boat from the Punjab 
to the sea from Greek times downward. The Indus flotilla of 
steamboats has however suffered fatally from the competition of 
the North-Western Railway, and the wheat exported from Karachi 
is now almost wholly rail-borne. 

At Multan, a considerable mercantile city near the Chenab, the 
railway forks to Lahore and Peshawar. From Lahore the triangle 
is completed by a line to Peshawar along the foot of the mountains, 
past the great military station of Rawalpindi. The lines from 
Lahore and Multan unite on the east bank of the Indus, fifty miles 
east of Peshawar, just below the point where the Kabul tributary 
enters. They cross the Indus by the bridge of Attock. Above 
Rawalpindi is the hill station of Murree. The long tongues of land 
between the five rivers of the Punjab are known as Doabs, a word 
which in Persian has the significance of Mesopotamia in Greek. 
Punjab signifies the land of five rivers. 

Peshawar is the capital of the North-West Frontier Province 
created in 1901, a strip of hilly country beyond the Indus. Unlike 
its sister Quetta, it lies in the Indian lowland at the foot of 
the Khyber pass. It has about a hundred thousand inhabitants, 

32 The Sub- Continent of India [CH. 

chiefly Musalman. In the Bazar are to be seen representatives of 
many Asiatic races, for Peshawar is the market of exchange where 
the great road from Samarkand and Bukhara, over the Hindu 
Kush and through Kabul, by the Khyber meets the road from 
Delhi and Lahore. Here you may buy skeins of Chinese silk, 
brought by the same roundabout ways that were trodden by the 
Chinese pilgrims in the Middle Ages. 

Jamrud, at the entrance to the Khyber, lies some nine miles 
west of Peshawar. In the Sarai at Jamrud all caravans going into 
India or returning to Central Asia halt for the night. The great 
Bactrian camels, two-humped and shaggy, present an unwonted 
contrast with the smaller Indian camels. The fort of Ali Masjid, 
nearly three thousand feet above the sea, crowns the steep ascent 
to the crest of the pass. At Landi Kotal begins the descent into 
Afghanistan. Thus the Khyber is a saddle in the heights, not the 
gorge of a torrent as is the Bolan. The Kabul river flows through 
an open valley until it nears the British frontier. Then it swerves 
through a precipitous chasm by a northward loop. The road is 
therefore carried over the intervening mountain spur. 

The Khyber is protected by its own hill tribes, enlisted in the 
Khyber Rifles. We have brought these Pathan mountaineers into 
the service of law and order by enrolling them in military forces, 
just as the Scottish highlanders were enrolled in the British army 
in the eighteenth century. The Pathans are born fighter«i. They 
love fighting for its own sake, and many a curious tale is told of 
the vendettas intermittently continued when the Khyber riflemen 
of Peshawar return from time to time on furlough to their homes 
in the hills. 

The Indus river rises, like the Brahmaputra, high on the 
plateau of Tibet to the north of Benares, and flows north-westward 
through the elevated valley of Leh until it reaches the 36th parallel 
of latitude. There it turns south-westward and cleaves its way 
through the Himalayas by the grandest gorge in the world. You 
may stand on the right bank of the Indus and look across the river 
to where the summit of Nanga Parbat descends by a single slope 
of four miles — measured vertically — to the river bank, every yard 
of the drop being visible. 

Within the great northward angle thus made by the Indus is a 
second smaller valley amid the mountains, which is also drained 
through a gorge to the Punjab. This is the famous valley of 
Kashmir, whose central plain, sheltered in every direction by lofty 
snow-clad mountains, is a sunny paradise of fertility. Srinagar is 

i] Karakoram and Hindu Kush 3 3 

the capital of Kashmir, whose Maharaja rules also over Ladakh 
(capital Leh) formerly a province of Tibet. 

The northernmost outposts of the Empire are in the valleys of 
Gilgit and Chitral, which diverge south-eastward and south-west- 
ward to the Indus and Kabul rivers. Enframing Gilgit and Chitral 
is a great angle of the loftiest mountain ridge, which may be 
likened, as it appears upon the map, to a pointed roof sheltering 
all India to the south. The south-eastward limb of the angle is 
the Karakoram range, and the south-westward is the Hindu Kush 
range. The north-western extremity of the Himalaya fits into the 
angle of the Karakoram and the Hindu Kush, from which it is 
separated by the valleys of Leh, Gilgit, and Chitral. 

The Karakoram is backed by the heights of the Tibetan plateau, 
here it is true at their narrowest, but none the less almost in- 
accessible, except for one or two passes at heights of 18,000 feet, 
which are traversed in the summer time by a few Yak caravans. 
In the Karakoram is mount Godwin Austen, second only to Everest 
among the mountains of the world. There also are the largest 
glaciers outside the Arctic and Antarctic regions. 

The Hindu Kush, notwithstahding its elevation, is in marked 
contrast to the Karakoram. It is a single broad ridge, backed by 
no plateau, and is notched by some relatively low passes. The 
ridge itself may be crossed in a few days or even hours at heights 
of twelve and thirteen thousand feet. The difficulties of access 
from the valley head of Kabul to the lowland of Bactria on the 
Oxus lie rather in the approaches to the passes than in the passes 
themselves. But human patience has in all ages succeeded in 
surmounting these difficulties ; and the Hindu Kush, although the 
natural boundary of India north-westward, has been no effective 
barrier either in a military or a commercial sense. 

There is lateral communication between the Khyber and Bolan 
routes outside the Indian frontier and yet within the Hindu Kush. 
The route follows a chain of valleys between Kabul and Kandahar 
through Ghazni. Along it from Kandahar to Kabul the army of 
Alexander the Great marched to his Bactrian and Indian cam- 
paigns : and it again became famous in the last generation because 
of the march of General Roberts from Kabul to the relief of 
Kandahar during the Afghan war of 1882. From this Kabul- 
Kandahar road several passes penetrate the mountainous belt of 
the Indian frontier, presenting alternative exits from the two 
trunk routes. But amid the maze of mountains north of the 
Kabul-Kandahar line, there are no practicable alternatives to 

c. H. 1. 1. 3 

34 The Suh-Continent of Iiidia [cH. 

the two ways — over the Hindu Kush and over the plateau from 

The long barrier of the Hindu Kush seems as if it were designed 
by nature to be the protecting boundary of India on the north- 
west. It is the 'scientific frontier' which in the last century 
British policy sought in vain to secure. At the present time it 
lies mostly within the 'buffer state' of Afghanistan which was 
created as the best alternative. But there have been periods in 
history when it has formed the actual, as well as the ideal, limit of 
the Indian empire. In the last quarter of the fourth century B.O., 
within a few years of the departure from India of Alexander the 
Great, it separated the dominions of the Maurya emperor of 
India, Chandragupta, from those of Seleucus Nicator, Alexander's 
successor in the eastern portion of his vast empire. In about the 
middle of the third century B.C. the Seleucid province of Bactria, 
which lay immediately to the north of the Hindu Kush, became 
an independent kingdom, from which, when the Maurya empire 
declined and the barrier was no longer adequately protected, 
a second series of Greek invasions poured into India about 200 B.O. 

The river Indus also appears at first sight to form a natural 
boundary between India and Iran; but in this case it would be 
more correct historically to say that the country through which it 
flows has more frequently been the cause of contention between 
India and Iran. The very name India, ' the country of the Indus,' 
was first known to the West as that of a province of the Persian 
empire. In Herodotus, the Greek historian of the wars between 
the Persian empire and Greece in the early part of the fifth 
century B.O., it bears its original meaning. At a later date, Greek 
and Roman writers, as so often happens in geographical nomen- 
clature, transferred the name of the best known province to the 
whole country and set an example which has since been followed 

Thus we conclude a rapid survey of the historical and political 
geography of a vast region. The south and centre of India is 
structurally an island, whose steep brinks, the Western and Eastern 
Ghats, are continued — beyond the coastal selvage and the strip of 
shallow water ofi' shore — by renewed steep descents into the 
abysses of the Arabian sea and the Bay of Bengal, two miles deep. 
This great island has granitic foundations, although it is clothed in 
places with volcanic rocks. Its landward brinks are marked by 
mount Abu, the Aravalli hills, the ridge of Delhi, and the long low 
eastward curve of hills ending at Rajmahal, where the principal 

i] Controlling Geographical Facts 3 5 

coal seams of India rest on the granitic base. The salient angles 
at Delhi and Rajmahal are received, at a distance, by the great 
re-entering angles of the main framework of Asia, constituted by 
the brink of Iran beyond the Indus, the Himalayan brink of Tibet, 
and the mountains of the Burmese border. Between these rocky 
limits — salient on the Indian side and re-entering on the Asiatic 
side — extends a broad alluvial plain, two hundred miles in average 
breadth, and two thousand miles long, from the mouths of the 
Ganges northward to the foot of the mountains, then north- 
westward along that foot to the Punjab, and then south-westward 
to the mouths of the Indus. 

The Indian heights proper are so relatively low, attaining to 
eight or nine thousand feet only in the far south, that the whole 
geography of India seems to be dominated by the Himalayas. We 
recover our sense of the true proportions only when we reflect that 
even the Himalayas are only five or six miles high, and that India 
is two thousand miles long. None the less the Himalayas and 
Tibet are in a very real sense the controlling fact of Indian 
geography. They pierce upward through more than half the 
atmosphere into highland climates, and therefore constitute for 
man a mighty natural boundary. They also guide and limit the 
winds of the lower air, and thus govern the Indian climate. India 
is an agricultural land, whose tillage is everywhere dependent, 
either directly or indirectly, upon the moisture brought from the 
southern ocean by the great wind swirl of the summer and autumn 
monsoon. That swirl strikes the Malabar coast as a south-west 
wind, sweeps over Bengal as a south wind, and drives up the 
Ganges plains as a south-east wind. The whole movement is 
induced by suction to where the air is rising over the hot plains 
of the Middle Indus. There in the summer is one of the hottest 
places, if not the hottest place in the world. The winds which 
come down to it off the Iranian plateau, thus completing the swirl 
stream off a dry land, and bring no moisture. In the winter a dry, 
bright wind, the north-east monsoon, descends from Tibet over all 
India. Only in the Punjab and in the far south are there con- 
siderable winter rains. The Punjab is in Mediterranean latitudes, 
where it rains in the winter. 

By these physical characteristics India is made fruitful, and is 
at the same time more than half isolated from the rest of the 
world. The most primitive of its inhabitants are the Gonds and 
other tribes, who have been driven into the forest recesses of the 
hills eastward of the Deccan plateau and into other regions difficult 


36 The Sub-Continent of India [ch. i 

of access throughout the sub-continent. The Dravidian languages 
have been preserved in the southern promontory. The Aryan and 
later invaders from western and central Asia have come from the 
north-west through the passage of Delhi, and have thence dispersed 
south-eastward down the Ganges to Bengal, and south-westward 
to the fertile Gujarati and Maratha countries. Through the eastern 
mountains, which sever the Indian Empire from China, have pene- 
trated in historical times few great invasions ; and these have not 
been far-reaching in their political results. But if we may judge 
from the physical types and languages of the populations, and 
from their social characteristics, there has been from prehistoric 
times onwards a constant infiltration of Mongolian stock, not only 
abundantly into Burma, and along the Tsan-po valley to the foot- 
hills of the Himalaya, but also in lesser degree into Assam and into 
the eastern parts of Bengal about Dacca. 

From the days of the Greek pilot Hippalus, the monsoons have 
carried some sea traffic to and fro over the Arabian sea from the 
direction of Aden. Sind was raided by Muhammadans overseas. 
But Sind lies outside the desert of Rajputana. The Malabar coast 
long had commercial intercourse with the Nearer East, and thus 
indirectly with Christendom. But the Western Ghats lie behind 
the Malabar coast. In the south of India, on that coast, are two 
curious relics of this traffic, two small ancient communities of Jews 
and of Christians. But these are exceptional. The one gateway 
of India which signified, until modern times, was the north-western 
land-gate. Most of the history which is to be narrated in these 
volumes bears, directly or indirectly, some relation to that great 
geographical fact. 


The Indian Empire is the abode of a vast collection of peoples 
who differ fi-om one another in physical characteristics, in language, 
and in culture more widely than the peoples of Europe. Among 
them the three primary ethnographical divisions of mankind — the 
Caucasian or white type, with its subdivisions of blonde and dark, 
the Mongolian or yellow type, and the Ethiopian or black type — 
are all represented : the first two by various races in the sub- 
continent itself, and the last by the inhabitants of the Andaman 
Isles. Four of the great families of human speech — the Austric, 
the Tibeto-Chinese, the Dravidian, and the Indo-European — are 
directly represented among the living languages of India, of which 
no fewer than two hundred and twenty are recorded in the Census 
Report for 1911 ; while a fifth great family, the Semitic, which has 
been introduced by Muhammadan conquerors in historical times, 
has, through the medium of Arabic and Persian, greatly modified 
some of the Indian vernaculars. The Austric, Tibeto-Chinese, and 
Indo-European families are widely spread elsewhere over the face 
of the earth. The Dravidian has not been traced with absolute 
certainty beyond the limits of the Indian Empire; but there is 
evidence which seems to indicate that it was introduced into India 
in prehistoric times. 

The drama of Indian history, then, is one in which many peoples 
of very diverse origin have played their parts. In all ages the 
fertility and the riches of certain regions, above all the plain of the 
Gange8,have attracted invaders from the outside world ; while over- 
population and the desiccation of the land have given an impulse 
to the movements of peoples from the adjacent regions of Asia. 
Thus both the attracting and the expulsive forces which determine 
migrations have acted in the same direction. It is true indeed that 
the civilisations which have been developed in India have reacted, 
and that Indian religions, Indian literature, and Indian art 
have spread out of India and produced a deep and far-reaching 
influence on the countries of Further Asia ; but the migrations and 

3 8 Peoples and Languages [ch. 

the conquests which provided the human energy with wliich these 
civilisations were created have invariably come into India from the 
outside. And the peninsular character of the sub-continent has 
retained invaders within its borders, with the result that racial 
conditions have tended to become ever more and more complex. 
The outcome of the struggle for existence between so many peoples 
possessing diflPerent traditions and different ideals is to be seen in 
the almost infinite variety of degrees of culture which exists at the 
present day. Some types of civilisation have been progressive; 
others have remained stationary. So that we now find, at one 
extreme of the social scale, communities whose members are con- 
tributing to the advancement of the literature, science, and art of 
the twentieth century, and, at the other extreme, tribes still 
governed by their primitive constitutions, still using the implements 
and weapons, and still retaining the religious ideas and customs of 
their remote ancestors in the Stone Age. 

The Himalayas form an effective barrier against direct invasions 
from the north : the exceedingly toilsome passes in their centre 
are traversed only by a few patient traders or adventurous ex- 
plorers. But at the western and eastern extremities, river valleys 
and more practicable mountain passes afibrd easier means of 
access. Through these gateways swarms of nomads and conquering 
armies, from the direction of Persia on the one hand and from the 
direction of China on the other, have poured into India from time 

By routes passing through Baluchistan on the west and Afghani- 
stan on the north-west, the country of the Indus has been repeatedly 
invaded by peoples belonging to the Caucasian race from Western 
Asia, and by peoples belonging to the Northern or Mongolo- Altaic 
group of the Mongolian race from Central Asia. But these immi- 
grations were not all of the same nature, nor did they all produce 
the same effect on the population of India. In the course of time 
their character became transformed. At the most remote period 
they were slow persistent movements of whole tribes, or collections 
of tribes, with their women and children, their flocks and herds: at 
a later date they were little more than organised expeditions of 
armed men. The former exercised a permanent influence on the 
racial conditions of the country which they invaded : the influence 
of the latter was political or social rather than racial. 

This change in the nature of invasions was the gradual effect of 
natural causes. Over large tracts of Asia the climate has changed 
within the historical period. The rainfall has diminished or ceased ; 

II J Western and Eastern Invaders 39 

and once fruitful lands have been converted into impassable deserts. 
Both Iran and Turkestan, the two reservoirs from which the streams 
of migration flowed into the Indus valley, have been affected by 
this desiccation of the land. Archaeological investigations in 
Seistan and in Chinese Turkestan have brought to light the monu- 
ments of ancient civilisations which had long ago passed into 
oblivion. Especially valuable from the historical point of view are 
the accounts given by Sir Aurel Stein of his wond«"ful discoveries 
in Chinese Turkestan. From the chronological evidence, which he 
has so carefully collected from the documents and monuments dis- 
covered, we are enabled to ascertain the dates, at which the various 
ancient sites were abandoned because of the progressive desiccation 
during a period of about a thousand years (first century b.c. to 
ninth century A.D.). We may thus realise how it has come to pass 
that a region which once formed a means of communication not 
only between China and India, but also between China and Europe, 
has now become an almost insuperable barrier. The same causes 
have tended to separate India from Iran. The last irruption which 
penetrated to Delhi, the heart of India, through the north-western 
gateway was the Persian expedition of Nadir Shah in 1739. 

The routes which lead from the east into the country of the 
Ganges seem not to have been affected to the same extent by 
climatic changes. The invaders from this quarter belonged to the 
Southern group of the Mongolian race, the home of which was 
probably in N.W. China. They came into India partly from Tibet 
down the valley of the Brahmaputra, and partly from China through 
Burma by the Mekong, the Salween, and the Irrawaddy. To other 
obstacles which impeded their progress were added the dense 
growth of the jungle and its wild inhabitants. Tribal migrations 
from these regions can scarcely be said to have ceased altogether 
even now. But they are held in check by the British occupation 
of Upper Burma. The movements to the south-west and south of 
the Kachins, a Tibeto-Burman tribe, from the north of Upper 
Burma have in recent times afforded an illustration of the nature 
of these migrations {Imp. Gaz. xiv, pp. 253-5). 

Thus have foreign races and foreign civilisations been brought 
into India, the history of which is in a large measure the story of 
the struggle between newcomers and the earlier inhabitants. Such 
invasions may be compared to waves breaking on the shore. Their 
force becomes less the farther they proceed, and their direction is 
determined by the obstacles with which they come in contact. 
The most effective of these obstacles, even when human effort is 

40 Peoples and Languages [CH. 

the direct means of resistance, are the geographical barriers which 
nature itself has set up. We shall therefore best understand the 
distribution of races in the sub-continent if we remember its chief 
natural divisions. 

The ranges of the Vindhya system with their almost impene- 
trable forests have in all ages formed the great dividing line 
between Northern and Southern India. In early Brahman litera- 
ture they mark t;he limits beyond which Aryan civilisation had not 
yet penetrated, and at the present day the two great regions 
which they separate continue to offer the most striking contrasts in 
racial character, in language, and in social institutions. But the 
Vindhyas can be passed without difficulty at their western and 
eastern extremities, where lowlands form connecting links with 
the plains of the Indus and the Ganges. The coastal regions are 
therefore transitional. They have been more directly affected by 
movements from the north than the central plateau of the Deccan. 

In Northern India, natural boundaries are marked by the river 
Indus, by the Thar or Great Desert of Rajputana, and by the sub- 
Himalayan fringe which is connected on the east with Assam and 

The seven geographical regions thus indicated form the basis 
for the ethnographical classification of the peoples of India which 
is now generally accepted. The scheme was propounded by the 
late Sir Herbert Risley in the Ceiistis Report for 1901. Its details 
are the result of careful measurements and observations extending 
over many years. It is conveniently summarised in the Imperial 
Gazetteer (new edition, vol. i, pp. 292 ff.) from which the descrip- 
tions in the following account are quoted. The physical types are 
here enumerated in an order beginning from the south, instead of 
from the north-west as in the original scheme : 

1. The Dra vidian type in the larger section of the peninsula 
which lies to the south of the United Provinces and east of about 
longitude 76° E. 'The stature is short or below mean ; the com- 
plexion very dark, approaching black ; hair plentiful, with an 
occasional tendency to curl ; eyes dark ; head long ; nose very 
broad, sometimes depressed at the root, but not so as to make the 
face appear flat.' 

This was assumed by Risley to be * the original type of the 
population of India, now modified to a varying extent by the 
admixture of Aryan, Scythian, and Mongoloid elements.' It must 
be remembered, however, that, when the term ' Dra vidian ' is thus 
used ethnogi'aphically, it is nothing more than a convenient label. 

ii] Dravidians 41 

It must not be assumed that the speakers of the Dravidian lan- 
guages are aborigines. In Southern India, as in the North, the 
same general distinction exists between the more primitive tribes 
of the hills and jungles and the civilised inhabitants of the fertile 
tracts ; and some ethnologists hold that the difference is racial and 
not merely the result of culture. Mr Thurston, for instance, says : 

It is the Pre-Dravidian aborigines, and not the later and more cultured 
Dravidians, who must be regarded as the primitive existing race.... These Pre- 
Dravidians.-.are difiFerentiated from the Dravidian classes by their short stature 
and broad (platyrhine) noses. There is strong ground for the belief that the Pre- 
Dravidians are ethnically related to the Veddas of Ceylon, the Toalas of the 
Celebes, the Batin of Sumatra, and possibly the Australians. {The Madras 
Presidency, pp. 124-5.) 

It would seem probable, then, that the original speakers of the 
Dravidian languages were invaders, and that the ethnographical 
Dravidians are a mixed race. In the more habitable regions the 
two elements have fused, while representatives of the aborigines 
are still to be found in the fastnesses to which they retired before 
the encroachments of the newcomers. If this view be correct, we 
must suppose that these aborigines have, in the course of long ages, 
lost their ancient languages and adopted those of their conquerors. 
The process of linguistic transformation, which may still be 
observed in other parts of India, would seem to have been carried 
out more completely in the South than elsewhere. 

The theory that the Dravidian element is the most ancient 
which we can discover in the population of Northern India, must 
also be modified by what we now know of the Munda languages, 
the Indian representatives of the Austric family of speech, and the 
mixed languages in which their influence has been traced (p. 48). 
Here, according to the evidence now available, it would seem that 
the Austric element is the oldest, and that it has been overlaid in 
different regions by successive waves of Di-avidian and Indo- 
European on the one hand, and by Tibeto-Chinese on the other. 
Most ethnologists hold that there is no difference in physical type 
between the present speakers of Munda and Dravidian languages. 
This statement has been called in question ; but, if it be true, it 
shows that racial conditions have become so complicated that it is 
no longer possible to analyse their constituents. Language alone 
has preserved a record which would otherwise have been lost. 

At the same time, there can be little doubt that Dravidian 
languages were actually flourishing in the western regions of 
Northern India at the period when languages of the Indo-European 

42 Peoples and Languages [CH.. 

type were introduced by the Aryan invasions from the north-west. 
Dravidian characteristics have been traced alike in Vedic and 
Classical Sanskrit, in the Prakrits or early popular dialects, and in 
the modern vernacular derived from them. The linguistic strata 
would thus appear to be arranged in the order — Austric, Dravidian, 

There is good ground, then, for supposing that, before the 
coming of the Indo- Aryans, speakers of the Dravidian languages 
predominated both in Northern and in Southern India; but, as we 
have seen, older elements are discoverable in the populations of 
both regions, and therefore the assumption that the Dravidians are 
aboriginal is no longer tenable. Is there any evidence to show 
whence they came into India ? 

No theory of their origin can be maintained which does not 
account for the existence of BrahuT, the large island of Dravidian 
speech in the mountainous regions of distant Baluchistan which lie 
near the western routes into India. Is Brahui a surviving trace of 
the immigration of Dravidian-speaking peoples into India from the 
west ? Or does it mark the limits of an overflow from India into 
Baluchistan ? Both theories have been held ; but, as all the great 
movements of peoples have been into India and not out of India, 
and as a remote mountainous district may be expected to retain 
the survivals of ancient races while it is not likely to have been 
colonised, the former view would a 'priori seem to be by far the 
more probable. The reasons why it has not been universally 
accepted is that the racial character of the Brahuis is now mainly 
Iranian, and not Dravidian in the Indian sense of the term. But 
the argument from race is not so conclusive as may appear at first 
sight. The area in which the Dravidian Brahui is still spoken 
forms part of the region which is occupied by Turko-Iranian 
peoples ; and the peculiar tribal constitution of the Brahuis is one 
which, unlike the caste-system, does not insist on social exclusive- 
ness, but, on the contrary, definitely invites recruitment from 
outside. This is clear from the account given in the Gazetteer of 
the 'Baloch and Brahui type of tribe': 

The second type of Turko-Iranian tribe is based primarily not upon agnatic 
kinship, but upon common good and ill : in other words, it is cemented together 
only by the obligations arising from the blood-feud. There is no eponymous 
ancestor, and the tribe itself does not profess to be composed of homogeneous 
elements.... The same principles hold good in the case of the Brahui... whose 
numbers have been recruited from among Afghans, Kurds, Jadgals, Baloch, and 
other elements. {Imp. Gaz. i, p. 310.) 

ri] Indo- Aryans 43 

Such circumstances must necessarily change the racial character 
of the tribe by a gradual process which might well in the course of 
ages lead to a complete transformation. There is therefore nothing 
in the existing racial conditions, and equally nothing in the existing 
physical conditions^, to prevent us from believing that the survival 
of a Dravidian language in Baluchistan must indicate that the 
Dravidians came into India through Baluchistan in prehistoric 
times. Whether they are ultimately to be traced to a Central 
Asian or to a Western Asian origin cannot at present be decided 
with absolute certainty ; but the latter hypothesis receives very 
strong support from the undoubted similarity of the Sumerian and 
Dravidian ethnic types ^. 

2. The Indo-Aryan type in Kashmir, the Punjab from the 
Indus to about the longitude of Ambala (76°46' E.), and Rajputana. 
' The stature is mostly tall ; complexion fair ; eyes dark ; hair on 
face plentiful ; head long ; nose narrow and prominent, but not 
specially long.' 

The region now occupied by peoples of this type forms the 
eastern portion of the wide extent of territory inhabited by Aryan 
settlers in the earliest historical times — the period of the Rigveda, 
probably about 1200 B.c. Their oldest literature, which is in a 
language closely connected with ancient Persian, Greek, and Latin, 
supplies no certain indication that they still retained the recollec- 
tion of their former home ; and we may reasonably conclude, 
therefore, that the invasions, which brought them into India, took 
place at a date considerably earlier. 

The Indo-Aryans came from Bactria, over the passes of the 
Hindu Kush into S. Afghanistan, and thence by the valleys of the 
Kabul river, the Kurram, and the Gumal — all of them rivers well 
known to the poets of the Rigveda — into the N.W. Frontier Pro- 
vince and the Punjab. In the age of the Rigveda they formed five 
peoples, each consisting of a number of tribes in which the women 
were of the same race as their husbands. This is proved con- 
clusively by their social and religious status. We may be certain, 
therefore, that the invasions were no mere incursions of armies, 
but gradual progressive movements of whole tribes, such as would 
have been impossible at a later date, when climatic causes had 
transformed the physical conditions of the country (p. 38). On 

1 For the remains of ancient culture in this region, see Imf. Gaz. i, p. 302 ; xiv, 
p. 300. 

^ Hall, The Ancient History of the Near East (4th ed.), pp. 173-4. The converse 
view is, however, held by the author, viz. that the Sumerians came into Western Asia 
ftom India. 

44 Peoples and Languages [ch. 

this point the evidence of literature receives the support of 
ethnology ; for only thus, according to Risley, can be explained 
the uniform distribution of the Indo- Aryan racial type throughout 
the region which it occupies, and the strongly marked contrasts 
which it presents to types prevailing in regions to the east and 
south. Later settlements necessarily consisted almost entirely of 
men. Such modifications of the racial character as would be pro- 
duced by inter- marriage with the women of the country would, 
in the course of time, cease to be recognisable. They would be as 
difficult to trace as the Roman factor in the population of Britain. 

3. The Turko-Iranian type in the N.W. Frontier Province, 
Baluchistan, and those districts of the Punjab and Sind which lie 
west of the Indus. ' Stature above mean ; complexion fair ; eyes 
mostly dark, but occasionally grey ; hair on face plentiful ; head 
broad ; nose moderately narrow, prominent, and very long.' 

The northern section of the region now inhabited by peoples of 
this type, that is to say, the country of the north-western tributaries 
of the Indus, was, in the times of the Rigveda, occupied by Indo- 
Aryans. The predominant racial character of the whole region is 
due to the invasion of Mongolo-Altaic peoples from Turkestan on 
the one hand, and of Persian Aryans or Iranians on the other. The 
Indus is the ethnographical boundary between the Turko-Iranian 
and Indo-Aryan types, just as in history it has often been the 
political boundary between Iran and India. 

4. The Scytho-Dravidian type in Sind east of the Indus, 
Gujarat, and the western section of the peninsula as far as about 
longitude 76° E., that is to say, the Bombay Presidency or Western 
India generally. 'The type is clearly distinguished from the 
Turko-Iranian by a lower stature, a greater length of head, a 
higher nasal index, a shorter nose, and a lower orbito-nasal 

This type, of which the Marathas are the chief representatives, 
occupies a position between the broad-headed Turko-Iranians and 
the long-headed Dravidians. Its designation assumes that the 
foreign broad-headed element was introduced during the period of 
Scythian (Qaka) rule in Western India (c. 120-380 a.d.). But 
there can be little doubt that its origin must be traced to a period 
far more remote. The ^akas were among those military conquerors 
who broke into the Punjab after the downfall of the Maurya 
Empire ; and it can scarcely be supposed that the extension of 
their power to Western India materially affected the race. The 
fact that their Scythian names, as is shown by coins and inscrip- 

ii] Aryo-Dravidians 45 

tions, became Hinduised after a few generations, is conclusive 
proof that they were forced to adapt themselves to their social 
environment. We must therefore seek the disturbing racial influ- 
ence in some earlier tribal immigration of which no other memorial 
now remains. The invaders probably belonged to the broad-headed 
Alpine race which inhabited the plateaus of Western Asia (Anatolia, 
Armenia, and Iran)^; and they would seem to have conie into 
Western India, as the Dravidians also most probably came, through 
Baluchistan before desiccation had made the routes impassable for 

5. The Aryo-Dravidian or Hindustani type in the plain of the 
Ganges from about longitude 76° 30' E. to 87° E. ; that is to say, in 
the eastern fringe of the Punjab, in the United Provinces, and in 
Bihar. * The head-form is long, with a tendency to medium ; the 
complexion varies from lightish brown to black ; the nose ranges 
from medium to broad, being always broader than among the 
Indo- Aryans ; the stature is lower than in the latter group, and 
usually below the average ' (i.e. it ranges from 5' 3" to 5' 5"). 

The Aryo-Dravidian type occupies the ancient Madhyade9a, or 
* the Midland Country,' extending, according to Manu (11, 21) from 
Vina^ana, where the river SarasvatI loses itself in the Great Desert, 
to Allahabad, together with some five degrees of the country farther 
east. It is a mixed type caused apparently by the Indo-Aryan 
colonisation of a region previously held by a population mainly 
Dravidian. The Indo-Aryan type does not, as might have been 
expected from analogous instances, shade by imperceptible degrees 
into the Aryo-Dravidian type ; but a marked change from the 
former to the latter is observable about the longitude of Sirhind. 
It is evident, then, that the waves of tribal migration must have 
been impeded at this point, and that the Indo-Aryan influence 
farther east must be due rather to warlike or peaceful penetration 
than to the wholesale encroachment of multitudes. 

To explain this abrupt transition, the theory of a second Aryan 
invasion, which is supposed to have come into the plain of the 
Ganges from the Pamirs through Gilgit and Chitral, was pro- 
pounded by the late Dr Hoernle and has been generally accepted 
in the official publications of the Government of India. This theory 
is made improbable by the physical difficulties of the route sug- 
gested, and some of the arguments adduced in its favour are 
demonstrably mistaken. There is no such break of continuity 
between the tribes of the Rigveda and the peoples of the later 

1 Haddon, The Wanderings of Peoples, pp. 12, 17. 

46 Peoples and Languages fcH. 

literature as it presupposes*. At the same time it seemed to be 
supported by the existing distribution of the Indo- Aryan languages ; 
but, as will be seen (j). 50), an equally satisfactoiy explanation of 
this distribution may be suggested. 

Apart from this theory, the conclusions of ethnology are entirely 
in accord with the historical indications of the literature. The 
ethnographical limit is also the dividing line between the geography 
of the Rigveda and the geography of the later Vedic literature. In 
the Rigveda Aryan communities have scarcely 'advanced beyond 
the country of the river Sarasvati (Sirhind), which for ever after- 
wards was remembered with especial veneration as Brahmavarta, 
'the Holy Land.' In the Brahmanas the centre of religious 
activity has been transferred to the adjacent country on the south- 
east, i.e. the upper portion of the doab between the Jumna and the 
Ganges, and the Muttra District of the United Provinces. This was 
Brahmarshideca — * the Country of the Holy Sages.' Here it was 
that the hymns of the Rigveda, which were composed in the North- 
West — the country of the ' Seven Rivers ' as it is called {Rv. viii, 
24, 27) — were collected and arranged ; and here it was that the 
religious and social system which we call Brahmanism assumed its 
final form — a form which, in its religious aspect, is a compromise 
between Aryan and more primitive Indian ideas, and, in its social 
aspect, the result of the contact of different races. After Brahman 
culture had thus occupied what has in all ages been the com- 
manding position in India, its trend was still eastwards ; and the 
country of the ' Seven Rivers,' though not altogether forgotten, 
occupies a place of less importance in the later literature. 

Both of the facts above mentioned — the abrupt transition from 
the Indo- Aryan to the Aryo-Dravidian type, and the extension of 
Aryan influence from Brahmavarta to Brahmar8hide9a — are best 
understood if we remember the natural feature which connects the 
plain of the Indus with the plain of the Ganges. This is the strait 
of habitable land which lies between the desert and the mountains. 
Its historical significance has already been noticed'^. It is in this 
strait that the decisive battles, on which the fate of India has 
depended, have been fought ; and here too we may suppose that 
the progress of racial migrations from the north-west in prehistoric 
times must have been checked. Both politically and ethnographi- 
cally it forms a natural boundary. In the age of the Rigveda the 
Aryans had not yet broken through the barrier, though the Jumna 
is mentioned in a hymn (vii, 18, 19) in such a way as to indicate 

1 See Chapters v, p. 119 and xiii, p. 302. 2 Chapter i, pp. 22 f. 

ii] Mongoloids and Mongolo-Dravidians 47 

that a battle had been won on its banks. It was only at some later 
date that the country between the Upper Jumna and Ganges and 
the district of Delhi were occupied. A record of this occupation 
has been preserved in some ancient verses quoted in the (^atapatha 
Brahraana (xiii, 5, 4, 11-14) which refer to the triumphs celebrated 
by Bharata Dauhshanti after his victories on the Jumna and the 
Ganges, and to the extent of his conquests. In their new home 
the Bharatas, who were settled in the country of the Sarasvati in 
the times of the Rigveda (see iii, 23, 4), were merged in the Kurus ; 
and their whole territory, the new together with the old, became 
famous in history under the name Kurukshetra — ' the Field of the 
Kurus.' This was the scene of the great war of the descendants of 
Bharata Dauhshanti, and the centre from which Indo- Aryan culture 
spread, first throughout Hindustan, and eventually throughout the 
whole sub-continent. The epoch of Indo-Aryan tribal migration 
was definitely closed. It was succeeded by the epoch of Indo- 
Aryan colonisation. 

6. The Mongoloid type in Burma, Assam, and the sub-Hima- 
layan tract which includes Bhutan, Nepal, and the fringe of the 
United Provinces, the Punjab, and Kashmir. * The head is broad ; 
complexion dark, with a yellowish tinge ; hair on face scanty ; 
stature short or below average ; nose fine to broad ; face charac- 
teristically flat ; eyelids often oblique.' 

The term Mongoloid denotes the racial type which has been 
produced by the invasion of peoples of the Southern Mongolian 
race from Tibet and China. We have already seen how these 
peoples have from time immemorial been coming down the river 
valleys into Burma and Northern India (p. 39) ; and we shall learn 
more about them, and about the earlier inhabitants with whom 
they intermingled, when we consider the evidence of language 
(p. 49). 

7. The Mongolo-Dravidian or Bengali type in Bengal and 
Orissa. ' The head is broad ; complexion dark ; hair on face 
usually plentiful ; stature medium; nose medium, with a tendency 
to broad.' 

This type is regarded as 'probably a blend of Dravidian and 
Mongoloid elements, with a strain of Indo-Aryan blood in the 
higher groups.' The region in which it prevails lay beyond the 
geographical ken of the earlier literature. It comes into view first 
in the later literature (the epics and Puranas) when it was occupied 
by a number of peoples among whom the Vangas (from whom 
Bengal has inherited its name) and the Kalingas of Orissa were the 

48 Peoples and Languages [ch. 

chief. On the north-west it is separated from the Aryo-Dravidian 
area by what is now also the political dividing-line between Bihar 
and Bengal. In regard to this limit, as marking the extent of 
Indo- Aryan influence at an early date, ethnology and literature are 
fully in agreement. In the Atharvaveda the Magadhas of the Patna 
and Gay a Districts, and the Angas of the Monghyr and Bhagal- 
pur Districts in Southern Bihar, are mentioned in a manner which 
indicates that they were among the most distant of known peoples 
(see Yedic Index, 11, p. 116) ; while a legend in the ^atapatha 
Brahmana (i, 4, 1, 10 flf.) preserves the memory of the spread of 
Brahmanism from the west into Videha, or Tirhut in Northern 
Bihar. The traces of Indo-Aryan descent, which have been 
observed in the higher social grades of Bengal and Orissa, must be 
due to colonisation at a later date. 

On the south-west the Mongolo-Dravidians are separated from 
the Dravidians by the north-eastern apex of the plateau of the 
Deccan, where, in the Santal Parganas and the Chota Nagpur 
Division, hills and forests have preserved a large group of primi- 
tive tribes, some of whom continue to speak dialects of the oldest 
form of language known in India. 

It is here that we find the Munda languages, which, like the 
Mon-Khmer languages of Assam and Burma, are surviving repre- 
sentatives of the Austric family of speech, the most widely diflxised 
on earth. It has been traced * from Easter Island oflF the coast of 
South America in the east to Madagascar in the west, and from 
New Zealand in the south to the Punjab in the north' {Census 
Report, 1911, I, p. 324). 

The Munda languages are scattered far and wide. They are 
found not only in the Santal Parganas and Chota Nagpur, but also 
in the Mahadeo Hills of the Central Provinces, and in the northern 
districts of the Madras Presidency ; and they form the basis of a 
number of mixed languages which make a chain along the Hima- 
layan fringe from the Punjab to Bengal. 

The Mon-Khmer languages are similarly dispersed. They sur- 
vive in the Khasi Hills of Assam, in certain hilly tracts of Upper 
Burma, in the coastal regions of the Gulf of Martaban in Lower 
Burma, in the Nicobar Islands, and in some parts of the Malay 

Thus Austric languages, which still flourish in Annam and 
Cambodia, remain in India and Burma as islands of speech to 
preserve the record of a far distant period when Northern India 
(possibly Southern India also) and F?^rth^r India belonged to the 

ii] Austria L,anguages 49 

same linguistic area. And there is some evidence that they 
shared the same culture in neolithic times ; for the ' chisel- shaped, 
high-shouldered celts' are specially characteristic of these regions \ 
There can be little doubt that the Indian and Burmese tribes who 
speak Austric languages are descended from the neolithic peoples 
who made these celts. We may regard them as representing the 
earliest population concerning which we possess any definite infor- 
mation. Other tribes may have an equal claim to antiquity ; but 
they have abandoned their ancestral speech and adopted that ot 
their more recent and more progressive neighbours. Their title is 
consequently less clear. 

Invasions from the east, some of them historical, have brought 
into the ancient domain of Austric speech languages belonging to 
two branches of the Tibeto-Chinese family — the Tibeto-Burman and 
the Siamese-Chinese. Tibeto-Burman has occupied the western half 
of Burma, where it is represented by Burmese, and the sub-Hima- 
layan fringe of India ; while Siamese-Chinese has prevailed in the 
Shan States of eastern Burma. The influence of each has, at different 
periods, extended to Assam, where at the present day both have given 
place to Assamese, an Aryan language closely related to Bengali. 

In the same way the Austric languages have been submerged 
by successive floods of Dravidian and Indo-European from the west 
and north-west. Dravidian languages, with the exception of 
Brahul, are now confined to the peninsula south of the Vindhyas 
and to Ceylon ; but it is supposed that, at the period of the Aryan 
invasions, they prevailed also in the north. This inference is 
derived from the change which Indo-European underwent after its 
introduction into India, and which can only be explained as the 
result of some older disturbing element. The oldest form of Indo- 
Aryan, the language of the Rigveda, is distinguished from the oldest 
form of Iranian, the language of the Avesta, chiefly by the presence 
of a second series of dental letters, the so-called cerebrals. These 
play an increasingly important part in the development of Indo- 
Aryan in its subsequent phases. They are foreign to Indo-European 
languages generally, and they are characteristic of Dravidian. We 
may conclude, then, that the earlier forms of speech, by which 
Indo-European was modified in the various stages of its progress 
from the north-west, were predominantly Dravidian. 

At the present time Dravidian languages are stable only in the 
countries of the south where they have developed great literatures 
like Tamil, Malayalam, Kanarese, and Telugu. In the northern 

"^ Chapter xxvi, p. 613. 
C. H. I. I. 4 

50 Peoples and Languages [ch. 

borders of the Dravidian sphere of influence, the spoken languages 
which have not been stereotyped by literature are, as each succeed- 
ing Report of the Census of India shows, still continuing to retreat 
before the onward progress of Indo-Aryan. The process, as it may 
be observed at the present day in India as elsewhere, has been 
admirably described by Sir George Grierson, whose observations 
are most valuable as explaining generally the manner in which the 
language of a more progressive civilisation tends to grow at the 
expense of its less efficient rivals. 

Wlien an Aryan tongue comes into contact with an uncivilised aboriginal one, 
it is invariably the latter which goes to the wall. The Aryan does not attempt to 
speak it, and the necessities of intercourse compel the aborigine to use a broken 
' pigeon ' form of the language of a superior civilisation. As generations pass this 
mixed jargon more and more approximates to its model, and in process of time 
the old aboriginal language is forgotten and dies a natural death. At the present 
day, in ethnic borderlands, we see this transformation still going on, and can 
watch it in all stages of its progress. It is only in the south of India, where 
aboriginal languages are associated with a high degree of culture, that they have 
held their own. The reverse process, of an Aryan tongue being superseded by an 
aboriginal one never occurs. {Imp. Gaz. i, pp. 351-2.) 

But the advancing type does not remain unaffected. Each 
stage in its progress must always bear traces of the compromise 
between the new and the old ; and, as each recently converted 
area tends in its turn to carry the change a step farther, the result 
is that the influence of the progressive language is modified in an 
increasing degree. Thus is produced a series of varieties, which 
through the development of their peculiar features become in 
course of time distinct species differing from the original type and 
from each other in accordance with their position in the series. 

We are thus furnished with a satisfactory explanation of the 
distribution of the Indo-Aryan languages. As classified by the 
Linguistic Survey they radiate from a central area occupied by the 
Midland languages, the chief representative of which is Western 
Hindi. In the north of this area lay the country of the Kurus and 
Paiichalas where, according to the ^atapatha Brahmana (iii, 2, 3, 
15) speech, i.e. Brahman speech, had its home {Vedic Index, i, 
p. 165). This is the centre from which the spread of Brahmanism 
and Brahman culture may be traced historically. From it the 
language of the Brahman scriptures extended with the religion and 
became eventually the sacred language of the whole sub-continent ; 
from it the influence of the Aryan type of speech was diffused in 
all directions, receiving a check only in the south where the 
Dravidian languages were firmly established. 

ii] Indo- Aryan Languages 51 

Immediatel)'^ outside the languages of the Midland come those 
of the Inner Band — Punjabi, Rajasthani and Gujarat! on the west, 
Pahari on the north, and Eastern Hindi on the east ; and beyond 
them the languages of the Outer Band — Kashmiri, Lahnda, Sindhi, 
and KacchI on the west, MarathI on the south-west, and Biharl, 
Bengali, Assamese, and Oriya on the east. 

The Indo-Aryan languages have now extended very considerably 
to the south of Aryavarta, * the Region of the Aryans,' as defined 
by Manu (ii, 22), i.e. the country between the Himalayas and the 
Vindhyas from the Bay of Bengal to the Arabian Sea. Orthodox 
Brahmanism, as represented by Manu, directed that all members of 
the 'twice-born' social orders, Brahmans, Kshatriyas, and Vaiyyas, 
should resort to this region, and enjoined that every man of these 
orders should be instructed in his religious and social duties by 
a Brahman belonging to one of the peoples of Brahmarshide9a 
(Kurus, Matsyas, Paiichalas, and ^^urasenas). These, as we have 
seen, inhabited the northern portion of the Midland linguistic area. 
If we follow the course of the Jumna-Ganges we shall pass from the 
languages of the Midland through those of the Inner and Outer 
Bands, and we shall pass from Brahmarshide^a through Kosala 
(Oudh), Videha (N. Bihar) and Vanga (Bengal), which mark suc- 
cessive stages in the spread of Brahmanism to the eastern limit of 
Aryavarta as they are reflected in the literature ^ 

It is not so easy to trace the relations between Brahmarshideca 
and the earlier Aryan settlements in the land of the Seven Rivers. 
It is possible that further invasions of which no record has been 
preserved may have disturbed both political and linguistic condi- 
tions in the North- West, We know nothing certain about the fate 
of this region until the latter half of the sixth century B.C., when 
Gandhara (Peshawar in the N.W. Frontier Province and Rawalpindi 
in the Punjab) together with the province of the Indus — * India ' 
properly so called — were included in the Persian empire of the 

The base from which this Persian power expanded into India 
was Bactria (Balkh), the country of the Oxus, which in the reign 
of Cyrus (558-530 B.C.) had become the eastern stronghold of 
Iran. From Bactria the armies of the Achaemenids, like those of 
Alexander and many subsequent conquerors, and like the invading 
tribes of Indo-Aryans many centuries before, passed over the 
Hindu Kush and through the valley of the Kabul river into the 
country of the Indus. 

1 Vedie Index, ii, pp. 237, 298. 


52 Peoples and Languages [ch. 

Speakers of the two great sections of Aryan languages, Iranians 
and Indo- Aryans, were thus brought into contact ; and as a result 
of some such contact, whether at this period or at some earlier 
date, we find a group of mixed languages still surviving where they 
might be expected, in the transitional zone between the Hindu 
Kush and the Punjab, that is to say, in the Kabul valley, Chitral, 
and Gilgit. These Pi9acha languages, as they are called, were 
once more widely spread : the Greek forms of place-names, for 
instance, seem to show that they prevailed in N.W. India in the 
fourth century B.c. ; but at the present time they are merely an 
enclave in the Iranian and Indo- Aryan domains. 

They possess an extraordinarily archaic character. Words are still in every- 
day use which are almost identical with the forms they assumed in Vedic hymns, 
and which now survive only in a much corrupted state in the plains of India. 

In their essence these languages are neither Iranian nor Indo- Aryan, but are 
something between both. {Imp. Gaz. i, p. 356.) 

The most natural explanation of these mixed languages is that 
they are ancient Aryan (Vedic) dialects which have been overlaid 
with Iranian as the result of later invasion. The districts in which 
they are spoken were certainly colonised by the early Aryan 
settlers, for both the Kabul river (Kubha) and its tributary the 
Swat (Suvastu) are mentioned in the hymns of the Rigveda. 

The contrary view, expressed in the Imperial Gazetteer (i, 
p. 355), viz. that the Pi9acha languages are the result of an Aryan 
invasion of a region originally Iranian, seems to be less probable. 
It presupposes the existence of an early settlement of Aiyans in the 
Pamirs, distinct from 'the Aryans proper, who had entered the 
Punjab by the valley of the Kabul,' and is thus bound up with the 
hypothesis of a second wave of Aryan immigration. 

Beyond the Pi^acha languages on the north, and beyond the 
Outer Indo- Aryan Band on the west, Iranian forms of speech pre- 
vail. The most important of these, so far as they are represented 
within the limitsof the Indian Empire, are the Pashto of Afghanistan, 
the name of which preserves the memory of the Ila/cTue? mentioned 
by Herodotus, and Baloch, the main language of Baluchistan. 

The diversity of speech in the Indian Empire, like the diversity 
of race, is naturally explained as the result of invasions from 
Western and Further Asia. Such invasions belong to a period 
which was only brought to a close by the establishment of the 
British dominion. The power which has succeeded in welding all 
the subordinate ruling powers into one great system of government 
is essentially naval ; and since it controls the sea-ways, it has been 

ii] Social Institutions 53 

forced, in the interests of security, to close the land- ways. This 
has been the object of British policy in regard to the countries 
which lie on the frontiers of the Indian Empire — Afghanistan, 
Baluchistan, and Burma. Political isolation has thus followed as a 
necessary consequence of political unity. But it must be remem- 
bered that this political isolation is a recent and an entirely novel 
feature in the history of India. It is the great landmark which 
separates the present from the past. 

Man has completed the work which nature had begun ; for, as 
we have seen, climatic changes had for ages past been making 
access into India more and more diificult. The era of tribal 
migration had long ago come to an end, and had been succeeded 
by the era of conquest. All through history down to the period of 
British rule we see one foreign power after another breaking 
through the north-western gateway, and the strongest of these 
winning the suzerainty over India. But the result in all cases was 
little more than a change of rulers — the deposition of one dominant 
caste and the substitution of another. The lives of the common 
people, their social conditions and systems of local government, 
were barely affected by such conquests. Indian institutions have 
therefore a long unbroken history which makes their study especially 

The chief distinguishing feature of Indian society at the present 
day is the caste-system, the origin and gi'owth of which may be 
traced from an early period. It now divides the great majority of 
the inhabitants of Northern and Southern India into hundreds of 
self-contained social groups, i.e. castes and sub-castes. A man is 
obliged to marry outside his family, but within the caste, and 
usually within the sub-caste, to which his family belongs. A family 
consists of persons 'reputed to be descended from a common 
ancestor, and between whom marriage is prohibited.' It is the 
exogamous social unit. A collection of such units constitutes a 
sub-caste or caste. 

A caste may, therefore, be defined as an endogamons group or collection of 
such groups bearing a common name and having the same traditional occupa- 
tion, who are so linked together by these and other ties, such as the tradition of a 
common origin and the possession of the same tutelary deity, and the same social 
status, ceremonial observances and family priests, that they regard themselves, 
and are regarded by others, as forming a single homogeneous community. (Census 
Report, 1911, I, p. 367.) 

The institution is essentially Brahmanical, and it has spread 
with the spread of BrahmanisuL It either does not exist, or exists 

54 Peoples and Languages [ch. 

only in an imperfect state of development, in countries where 
Buddhism has triumphed, such as Burma and Ceylon. It would 
indeed appear to rest ultimately on two doctrines which are dis- 
tinctively Brahmanical — the doctrine of the religious unity of the 
family, which is symbolised by the offerings made to deceased 
ancestors, and the doctrine of sva-lcarma, which lays on every man 
the obligation to do his duty in that state of life in which he has 
been born. 

The orthodox Hindu holds that the caste-system is of divine 
appointment and that it has existed for all time. But the sacred 
books themselves, when they are studied historically, supply evi- 
dence both of its origin and of its growth. The poets of the 
Rigveda know nothing of caste in the later and stricter sense of the 
word ; but they recognise that there are divers orders of men — 
the priests (Brahma or Brahmana), the nobles (Rajanya or Ksha- 
triya), the tillers of the soil (Vi9 or Vaigya), and the servile classes 
(^udra). Between the first three and the fourth there is a great 
gulf fixed. The former are conquering Aryans : the latter are 
subject Dasyus. The difference between them is one of colour 
{varna) : the Aryans are collectively known as ' the light colour,' 
and the Dasyus as 'the dark colour.' So far, there was nothing 
peculiar in the social conditions of North- Western India during 
the early Vedic period. The broad distinction between conquerors 
and conquered, and the growth of social orders are indeed univei-sal 
and inevitable. But, while in other countries the barriers which 
man has thus set up for himself have been weakeiied or even 
entirely swept away by the tide of progress, in India they have 
remained firmly fixed. In India human institutions have received 
the sanction of a religion which has been concerned more with the 
preservation of social order than with the advancement of mankind. 

Before the end of the period covered by the hymns of the 
Rigveda a belief in the divine origin of the four orders of men was 
fully established ; but there is nowhere in the Rigveda any indica- 
tion of the castes into which these orders were afterwards sub- 
divided ^ The word 'colour' is still used in its literal sense. 
There are as yet only two varnas, the light and the dark. But in 
the next period, the period of the Yajurveda and the Brahmanas, 
the term denotes *a social order' independently of any actual dis- 
tinction of colour, and we hear for the first time of mixed varnas, 
the offspring of parents belonging to different social orders. 

* For various views on this subject, see Chapters iv, pp. 92-4; v, pp. 125 ff.; viti, 
pp. 208-10; X, pp. 234-6. 

ii] The Caste System 55 

It is to such mixed marriages that the law-books (cf. Manu, x, 
6 ff.) attribute the origin of the castes {jati) strictly so-called. 
To some extent the theory is undoubtedly con-ect Descent is a 
chief factor, but not the only factor, involved in the formation of 
caste, the growth of which may still in the twentieth century be 
traced in the Reports of the decennial Census. Primitive tribes 
who become Hinduised, communities who are drawn together by 
the same sectarian beliefs or by the same occupation, all tend to 
form castes. Tribal connexion, religion, and occupation therefore 
combine with descent to consolidate social groups and, at the same 
time, to keep these social groups apart. 

The caste-system is, as we have seen, a distinctive product 
of Brahmanism, a code which regards the family, and not the con- 
gregation, as the religious unit. And so strong did this social 
system become that it has affected all the other religions. The 
most probable explanation of the very remarkable disappearance 
of Buddhism from the greater part of the sub-continent, where it 
was once so widely extended, is that Buddhism has been gradually 
absorbed into the Brahman caste-system, which has also, though 
in a less degree, influenced the followers of other faiths — Jains, 
Muhammadans, Sikhs, and even native Christians. We must con- 
clude, then, that the caste-system has accompanied the spread of 
Brahmanism from its first stronghold in the country of the Upper 
Jumna and Ganges into other regions of Northern India and finally 
into Southern India ; and we must expect to find its complete 
record only in Brahman literature. Caste must naturally be less 
perfectly reflected in the literature of other faiths. 

Neglect of these fundamental considerations has led to much 
discrepancy among writers on the early social history of India. 
Students of the Brahman books have asserted that the caste- 
system existed substantially in the time of the Yajurveda (say 
1000-800 B.C.) : students of the Buddhist books have emphatically 
declared that no traces of the system in its later sense are to be 
detected in the age of Buddha (c. 563-483 B.C.). Both parties 
have forgotten that they were dealing with different regions of 
Northern India — the former with the country of the Kurus and 
Panchalas, the home of Brahmanism (the Delhi Division of the 
Punjab with the north-western Divisions of the Province of Agra), 
the latter with Kosala and Videha, the home of Buddhism (Oudh 
and N. Bihar). They have forgotten, too, that the records, on 
which they depend for their statements, are utterly distinct in 
character. On the one hand, the Brahman books are permeated 

56 Sources of History [ch. 

with social ideas which formed the very foundation of their religion : 
on the other hand, the Buddhist books regard any connexion 
between social status and religion as accidental rather than essential. 


The caste-system is the outcome of a long process of social 
diflferentiation to which the initial impulse was given by the intro- 
duction of a higher civilisation into regions occupied by peoples in 
a lower stage of culture. The Aryan settlers, as represented by 
the sacrificial hymns of the Rigveda, were both intellectually and 
materially advanced. Their language, their religion, and their 
social institutions were of the Indo-European type like those of the 
ancient Persians of the Avesta and the Greeks of the Homeric 
poems ; and they were skilled in the arts and in the working of 

The prehistoric archaeology of India has not attracted the 
attention which it deserves, and many interesting problems con- 
nected with the earlier cultures and their relation to the culture of 
the Rigveda remain to be solved ; but there is a general agreement 
as to the succession of cultural strata in Northern and Southern 
India. The discoveries of ancient implements seem to prove that 
in the North the Stone Age is separated from the Iron Age by a 
Copper Age ; while in the South no such transitional stage has 
been observed — implements of stone are followed without a break 
by implements of iron. Bronze, it appears, is not found anywhere 
in India before the Iron Age. If these facts may be held to be 
established, we must conclude that the chief metal of the Rigveda, 
ayas (Latin acs), was copper ; and the absence of a Copper Age in 
Southern India would seem to indicate that the earlier inhabitants 
generally were still in the Stone Age at the time when the Aryans 
brought with them the use of copper. Iron was probably not 
known in the age of the Rigveda ; but it undoubtedly occurs in the 
period immediately following when it is known to the Yajurveda 
and Atharvaveda as ^ydma ay as or 'black copper.' Its use was 
introduced by Indo- Aryan colonisation into Southern India where 
the Stone Age of culture still prevailed. 

Described in its simplest terms, the earliest history of India is 
the story of the struggle between two widely diflferent types of 
civilisation, an unequal contest between metal and stone. All the 
records for many centuries belong to the higher type. They are 
exclusively Indo-Aryan. They have been preserved in literary 

ii] The Literatures of India 57 

languages developed from the predominant spoken languages under 
the influence of the different phases of religion which mark 
stages in the advance of Indo- Aryan culture from the North-West. 
The language of the Rigveda, the oldest form of Vedic Sanskrit, 
belongs to the country of the Seven Rivers. The language of the 
Brahmanas and of the later Vedic literature in the country of the 
Upper Jumna and Ganges (Brahmarshide9a) is transitional. It 
shades almost imperceptibly into Classical Sanskrit, which is the 
literary representation of the accepted form of educated speech of 
the time and region. As fixed by the rules of the grammarians it 
became the standard language of Brahman culture in every part 
of India ; and it is still the ordinary medium of communication 
between learned men, as was Latin in the Middle Ages of Europe. 

In the sixth century B.C., after Indo-Aryan influence had pene- 
trated eastwards beyond the limits of 'the Middle Country,' there 
arose in Oudh (Kosala) and Bihar (Videha and Magadha) a number 
of religious reactions against the sacerdotalism and the social 
exclusiveness of Brahmanism. The two most important of these, 
Jainism and Buddhism, survived ; and, as they extended from the 
region of their origin, they everywhere gave an impulse to the for- 
mation of literary languages from the Prakrits or spoken dialects. 
The scriptures of the Jains have been preserved in various forms 
of Magadhi, the dialect of Bihar, ^auraseni, the dialect of Muttra, 
and Maharashtri, the dialect of the Maratha country. The Buddhist 
canon exists in two chief forms — in Pali, the literary form of an 
Indo-Aryan Prakrit, in Ceylon ; and in Sanskrit in Nepal. Pali 
Buddhism has spread to Burma and Siam. The Sanskrit version 
of the canon has, in various translations, prevailed in Tibet, China, 
Japan, Mongolia, Chinese Turkestan, and other countries of the 
Far East. 

In all the large and varied literatures of the Brahmans, Jains, 
and Buddhists there is not to be found a single work which can be 
compared to the Histories in which Herodotus recounts the 
struggle between the Greeks and Persians, or to the Annals in 
which Livy traces the growth and progress of the Roman power. 
But this is not because the peoples of India had no history. We 
know from other sources that the ages were filled with stirring 
events ; but these events found no systematic record. Of the great 
foreign invasions of Darius, Alexander the Great, and Seleucus no 
mention is to be discovered in any Indian work. The struggles 
between native princes, the rise and fall of empires, have indeed 
not passed similarly into utter oblivion. Their memory is to some 

58 Sources of History [ch. 

extent preserved in epic poems, in stories of the sages and heroes 
of old, in genealogies and dynastic lists. Such in all countries are 
the beginnings of history ; and in ancient India its development 
was not carried beyond this rudimentary stage. The explanation 
of this arrested progress must be sought in a state of society which, 
as in medieval Europe, tended to restrict intellectual activity to 
the religious orders. Literatures controlled by Brahmans, or by 
Jain and Buddhist monks, must naturally represent systems of 
faith rather than nationalities. They must deal with thought 
rather than with action, with ideas rather than with events. And 
in fact, as sources for the history of religion and philosophy, and 
for the growth of law and social institutions, and for the develop- 
ment of those sciences which, like grammar, depend on the minute 
and careful observation of facts, they stand among the literatures 
of the ancient world unequalled in their fulness and their con- 
tinuity. But as records of political progress they are deficient. 
By their aid alone it would be impossible to sketch the outline 
of the political history of any of the nations of India before the 
Muhammadan conquest. Fortunately two other sources of infor- 
mation — foreign accounts of India and the monuments of India 
(especially the inscriptions and coins) — supply to some extent this 
deficiency of the literatures, and furnish a chronological framework 
for the history of certain periods. 

The foreign authorities naturally belong to those periods in 
which India was brought most closely into contact with the civili- 
sations of Western Asia and China. The general fact that such 
intercourse by land and sea existed in very early times is undoubted, 
but detailed authentic records of political relations are not found 
before the rise of the Persian Empire in the sixth century B.C., 
when Greek writers and the cuneiform inscriptions of Darius 
enable us to trace the extension of the Persian power from 
Bactria, the country of the Oxus, to N.W. India. From these 
sources it is clear that the Persian dominions included Gandhara 
(the Districts of Peshawar and Rawal Pindi) and the Province 
of 'India' (the Western Punjab together with Sind which still 
retains its ancient name) ; and it is probable that these countries 
remained tributary to the King of Kings until the Persian Empire 
gave place to the Macedonian. 

Then come the Greek and Roman historians of Alexander the 
Great, whose detailed accounts of the Indian campaign (327-325 B.c.) 
throw a flood of light on the political conditions of N.W. India, 
and carry our geographical knowledge eastwards beyond the 

ii] Greek Writers 59 

Jhelum (Hydaspes), the eastern limit of Gandhara, to the Beas 
(Hyphasis). This marks the extent of Alexander's conquests. 
Far from securing the dominant position of Northern India, the 
country of the upper Jumna and Ganges, these conquests failed 
even to reach the country of the SarasvatI, the centre of Indo- 
Aryan civilisation in the age of the Rigveda. Alexander was the 
conqueror of ' India ' only in the sense that for a very few years he 
was master of * the country of the Indus.' The confusion of this 
geographical term with its later meaning has been the cause of 
endless misconception all through the Middle Ages even down to 
the present day. 

The documents of the Persian and Macedonian Empires are 
succeeded by those of the later Hellenic kingdoms of Syria, 
Bactria, and Parthia. All these are invaluable as supplying a very 
remarkable deficiency in the Indian records. They deal with 
a region which is barely noticed, and with events which are com- 
pletely ignored, in the Brahman, Jain, and Buddhist books of the 
period. These two sources of history are thus independent of each 
other. The Greek view is mainly confined to the North-West, 
while the contemporary Indian literatures belong almost ex- 
clusively to the Plain of the Ganges. 

After the death of Alexander other Western writers appear 
who regard India from the point of view of the Maurya Empire 
with its capital at Pataliputra, the modern Patna. The generation 
which saw Alexander had not passed away before the kingdom of 
Magadha (S. Bihar) had brought all the peoples of Northern India 
under its sway, and established a great power which maintained 
relations with Alexander's successors in Western Asia, Egypt, and 
Europe. And now for the first time the two kinds of historical 
evidence, the Indian and the foreign, come into direct relations 
with each other. They refer to the same regions and to the same 
circumstances ; and the light of Greek history is thrown on the 
obscurity of Indian literature. It was the identification of the 
Sandrocottus of Greek writers with the Maurya Emperor Chandra- 
gupta that established the first fixed point in the chronology of 
ancient India. Our object in the first two volumes of this History 
will be to show how far the progress of research starting from this 
fixed point has succeeded hitherto in recovering the forgotten 
history of India from the records of the past. 

Unimpeded intercourse with the countries of the West was 
possible only so long as Northern India remained united under the 
Maurya dynasty, and Western Asia under the Seleucid successors 

6o Sources of History [ch. 

of Alexander. The process of disintegration began in Western 
Asia with the defection of Bactria and Parthia about the middle 
of the third century, and in India probably some thirty years later 
when the downfall of imperial rule was followed by a period of 
anarchy and internal strife. These conditions made possible the 
series of foreign invasions from c. 200 B.c. onwards, which disturbed 
the North- West during many centuries and severed that region 
from the ancient civilisation of the Plain of the Ganges. The 
political isolation of India was completed by the Scythian conquest 
of Bactria, c. 135 B.C., and by the long struggle between Rome and 
Parthia which began in 53 B.C. After the Maurya Empire, inter- 
course tended more and more to be restricted to commerce by land 
and sea ; and for the West, India became more and more the land 
of mystery and fabulous wealth. Down to the last quarter of the 
eighteenth century nearly all that was known of its ancient history 
was derived from the early Greek and Latin writers. 

Of all the factors which contributed to the severance of relations 
with the West, the extinction of Hellenic civilisation in Bactria 
was by far the most important. But while the fate of Bactria 
closed the western outlook, it prepared the way for communication 
with the Far East ; and it is to Chinese authorities that we must 
turn for the most trustworthy information concerning the events 
which determined the history of N.W. India during the follow- 
ing centuries. The Scythian ((^aka) invaders of Bactria were 
succeeded by the Yueh-chi ; and when, in the first century A.D., 
the predominant tribe of the Yueh-chi, the Kushanas, extended 
their dominion in Turkestan and Bactria to N.W. India, the 
Kushana empire formed a connecting link between China and 
India and provided the means of an intercourse which was fruit- 
ful in results. Buddhism was introduced into China and the other 
countries of the Far East ; and, as the explorations of recent 
years have shown, an Indian culture, Indian languages, and the 
Indian alphabets were established in Chinese Turkestan. The 
most illuminating accounts of India from the end of the fourth to 
the end of the seventh century are the records of Chinese Buddhists 
who made the long and toilsome pilgrimage to the scenes of their 
Master's life and labours. 

The remaining source of historical information — the inscribed 
monuments and coins — is the most productive of all. The inscrip- 
tions are public or private records engraved in most cases on stone 
or on copper plates ; and they are found in great numbers through- 
out the sub-continent and in Ceylon. The earliest are the edicts 

ii] Inscriptions and Coins 6i 

of A9oka incised on rocks or pillars situated on the fi'ontiers and 
at important centres of the Maurya empire when at the height of 
its power in the middle of the third century b.c. Others com- 
memorate the deposit of Buddhist relics. Others celebrate the 
victories of princes, the extent of their conquests, the glories of 
the founder of the dynasty and of his successors on the throne. 
Others again place on record the endowments of temples or grants 
of land. In short, there is scarcely any conceivable topic of public 
or private interest which is not represented. The inscriptions 
supply most valuable evidence as to the political, social, and 
economic conditions of the period and the country to which they 
belong. They testify on the one hand to the restless activity of a 
military caste, and on the other to the stability of institutions, 
which were, as a rule, unaffected by military conquest. One con- 
queror follows another, but the administration of each individual 
state remains unchanged either under the same prince or under 
some other member of his family, and the charters of monasteries 
are renewed as a matter of course by each new overlord. 

Coins also have preserved the names and titles of kings who 
have left no other record ; and by their aid it is sometimes possible 
to reconstruct the dynastic lists and to determine the chronology 
and the geographical extent of ruling powers. But it is only when 
coin-legends appear as the result of Greek influence in the North- 
West that this source of history becomes available. The earlier 
indigenous coinage was little more than a system of weights of 
silver or copper stamped with the marks of the monetary authorities. 
The first Indian king whose name occurs on a coin is Sophytes 
(Saubhuti), a contemporary of Alexander the Great. The legend 
of his coins is in Greek. After his date no inscribed coins are 
found for more than a hundred years. During this interval Greek 
rule in KW. India had ceased. It was resumed about the begin- 
ning of the second century by Alexander's Bactrian successors, 
who issued in their Indian dominions a bilingual coinage with 
Greek legends on the obverse and a translation of these in an Indian 
dialect and an Indian alphabet on the reverse. 

The fashion of a bilingual coinage thus instituted was continued 
by the Scythian and Parthian invaders from Iran in the early part 
of the first century B.c. ; and these bilingual coins have supplied 
the clue to the interpretation of the ancient alphabets, and have 
enabled scholars during the last three generations to bring to light 
the long-hidden secrets of the inscriptions and to retrace the out- 
lines of forgotten history. 

62 Sources of History [ch. 

Both of the alphabets, now usually known as Brahmi and Kha- 
roshthi, are of Semitic origin ; that is to say, they are derived 
ultimately from the same source as the European alphabets. They 
were introduced into India at different periods, and probably by 
different routes. Brahmi is found throughout the sub-continent 
and in Ceylon. The home of KharoshthI is in the North-West ; 
and whenever it is found elsewhere it has been imported. 

Brahmi has been traced back to the Phoenician type of writing 
represented by the inscription in which Mesha, king of Moab 
(c. 850, B.C.), records his successful revolt against the kingdom of 
Israel. It was probably brought into India through Mesopotamia, as a 
result of the early commerce by sea between Babylon and the ports of 
Western India. It is the parent of all the modern Indian alphabets. 

KharoshthI is derived from the Aramaic script, which was 
introduced into India in the sixth century b,c., when the North- 
West was under Persian rule, and when Aramaic was used as a 
common means of communication for the purposes of government 
throughout the Persian empire. That originally the Aramaic 
language and alphabet pure and simple were thus imported into 
Gandhara, as Biihler conjectured in 1895 {W.Z.K.M., ix, p. 49), 
has been proved recently by Sir John Marshall's discovery of an 
Aramaic inscription at Taxila^. When the first KharoshthI in- 
scriptions appear in the third century B.C., the alphabet has been 
adapted to express the additional sounds required by an Indian 
language ; but, unlike Brahmi which has been more highly elabo- 
rated, it still bears evident traces of its Semitic origin both in its 
direction from right to left and in its imperfect representation of 
the vowels. In the third century a.d. KharoshthI appears more 
fully developed in Chinese Turkestan where its existence must be 
attributed to the Kushana empire. In this region, as in India, it 
was eventually superseded by Brahmi. 

The decipherment of the inscriptions and coins, and the deter- 
mination of the eras in which many of them are dated, have 
introduced into the obscurity of early Indian history a degree of 
chronological order which could not have been conceived at the 
time when the study of Sanskrit began in Europe. The bare fact 
that India possessed ancient classical literatures like those of Greece 
and Rome can scarcely be said to have been known to the Western 
World before the last quarter of the eighteenth century. At 
various intervals during more than a hundred years previously a 
few isolated students chiefly missionaries, those pioneers of learning. 

» A. Cowley, J.R.A.S., 1915, p. 346. 

ii] The Study of Sanskrit 63 

had indeed published accounts of Sanskrit literature and Sanskrit 
grammar ; but it was only when a practical need made itself felt, 
and the serious attention of the administrators of the East India 
Company's possessions was directed to the importance of studying 
Sanskrit, that the investigation by Europeans of the ancient lan- 
guages and literatures of India began in earnest. To meet the 
requirements of the law-courts the Governor-General, Warren 
Hastings, had ordered a digest to be prepared by pandits from 
the authoritative Sanskrit law-books ; but when the work was 
finished no one could be found able to translate it into English. 
It was therefore necessary to have it translated first into Persian, 
and from the Persian an English version was made and published 
by Halhed in 1776. The object-lesson was not lost. Sanskrit was 
evidently of practical utility ; and the East India Company adopted, 
and never afterwards neglected to pursue, the enlightened policy 
of promoting the study of the ancient languages and literatures in 
which the traditions of its subjects were enshrined. It remained 
for Sir William Jones, Judge of the High Court at Calcutta, to 
place this study on a firm basis by the establishment of the Asiatic 
Society of Bengal in 1784. 

The inauguration of the study of India's past history came at 
a fortunate moment ; for it is precisely to the last quarter of the 
eighteenth century that we may trace the growth of the modern 
scientific spirit of investigation, which may be defined as the 
recognition of the fact that no object and no idea stands alone by 
itself as an isolated phenomenon. All objects and all ideas form 
links in a series ; and therefore it follows that nowhere, whether 
in the realm of nature or in the sphere of human activity, can the 
present be understood without reference to the past. The first 
manifestation of this new spirit of enquiry, which was soon to 
transform all learning, was seen in the study of language. The 
first Western students of the ancient languages of India were 
statesmen and scholars who had been educated in the classical 
literatures of ancient Greece and Rome. They were impressed by 
the fact, which must indeed be apparent to everyone who opens a 
Sanskrit grammar, that Sanskrit, both in its vocabulary and in its 
inflexions, presents a striking similarity to Greek and Latin. This 
observation immediately raised the question : How is this simi- 
larity to be explained ? The true answer was suggested by Sir 
William Jones, whom that sagacious observer, Dr Johnson, recog- 
nised as * one of the most enlightened of the sons of men \' In 1786, 
Sir William Jones wrote ; 

^ G. Birbeck Hill, Johnsonian Miscellanies, n, p. 363. 

64 Sources of History [ch. ii 

The Sanscrit language, whatever be its antiquity, is of a wonderful structure ; 
more perfect than the Greek, more copious than the Latin, and more exquisitely 
refined than either : yet bearing to both of them a stronger affinity, both in the 
roots of verbs, and in the forms of grammar, than could possibly have been pro- 
duced by accident ; so strong indeed, that no philologer could examine them all 
without beheving them to have sprung from some common source, which perhaps 
no longer exists. There is a similar reason, though not quite so forcible, for sup- 
posing that both the Gothick and the Celtick, though blended with a difi"erent 
idiom, had the same origin with the Sanscrit; and the old Persian might be 
added to the same family. 

These observations contain the germs of the science of Com- 
parative Philology. The conception of a family of languages, in 
which all the individual languages and dialects are related as 
descendants from a common ancestor, suggested the application to 
language of the historical and comparative method of investigation. 
The results have been as remarkable as they were unexpected. 
In the first place, the historical method has shown that living 
languages grow and change in accordance with certain definite 
laws, while the comparative study of the lines of development 
which may be traced historically in the different Indo-European 
languages has confirmed Sir William Jones's hypothesis that they 
are all derived ' from some common source,' which, though it no 
longer exists, may be restored hypothetically. In the second place, 
since words preserve the record both of material objects and of 
ideas, a study of vocabularies enables us to gain some knowledge 
of the state of civilisation, the social institutions, and the religious 
beliefs of the speakers of the dififerent languages before the period 
of literary records. Some indication of the light which Compara- 
tive Philology thus throws on the history of the Aryan invaders of 
India is given in the following Chapter. 


Throughout the greater part of Europe and of Asia as far as 
India there exist now, or can be shown to have existed in past 
time, a great number of languages, the forms and sounds of which 
when scientifically examined are seen to have a common origin. 
The languages in question are generally known to scholars under 
the name of the Indo-Germanic, or Indo-European languages. The 
name Indo-European seems to have been invented by Dr Thomas 
Young, the well-known physicist and Egyptologist. The first occur- 
rence known of the word is in an article by him in The Quarterly 
Review for 1813. Examination of the article, however, shows that 
Dr Young meant by Indo-European something quite different from 
its ordinarily accepted signification. For under the term he in- 
cluded not only the languages now known as Indo-European, but 
also Basque, Finnish, and Semitic languages. The name Indo- 
Germanic, which was used by the German philologist Klaproth as 
early as 1823, but the inventor of which is unknown, is an attempt 
to indicate the family by the furthest east and west members of 
the chain extending from India to the Atlantic ocean. The main 
languages of the family had been indicated in a famous address to 
the Asiatic Society of Bengal, delivered by the President Sir William 
Jones in 1786\ He had the insight to observe that the sacred 
language of India (Sanskrit), the language of Persia, the languages 
of Greece and Rome, the languages of the Celts, Germans, and 
Slavs, were all closely connected. To Sir William Jones, as Chief 
Justice of Bengal, law was his profession and the comparison of 
languages only an amusement. But this epoch-making address 
laid the foundations of Comparative Philology on which Bopp in 
his Comparative Grammar built the first superstructure. But 
the study of this family of languages has fi'om the beginning been 
beset with a subtle fallacy. There has been throughout an almost 
constant confusion between the languages and the persons who 

^ See Chapter ii, p. 64 ^ 
C. H. I. I. 5 

66 The Aryans [CH. 

spoke them. It is hardly necessary to point out that in many 
parts of the world the speaker of a particular language at a given 
time was not by lineal descent the representative of its speakers 
at an earlier period. In the island of Britain many persons of 
Welsh blood, many persons of Irish Celtic and Scottish Celtic 
origin speak English. It is many centuries since it was observed 
that Normans and English who had settled in Ireland had learned 
to speak the Irish language and had become more Irish than the 
Irish themselves. It is well known that by descent the Bulgarians 
are of Asiatic origin, and of an entirely different stock from the 
Slavs, a branch of whose language is now their mother tongue. It 
is therefore clear that it is impossible, without historical evidence, 
to be certain that the language spoken by any particular people 
was the language of their ancestors at a remote period. The name 
Indo-Germanic therefore suffers from the ambiguity that it cha- 
racterises not only languages but also peoples. As has been sug- 
gested elsewhere, it would be well to abandon both the term 
Indo-European and the term Indo-Germanic and adopt some en- 
tirely colourless word which would indicate only the speakers of 
such languages. A convenient term for the speakers of the Indo- 
European or Indo-Germanic languages would be the Wiros, this 
being the word for ' men ' in the great majority of the languages 
in question. 

The advantage of such a term is clear, since all we know 
regarding the physical characteristics of the first people who spoke 
languages of this nature is that they were a white race. We cannot 
tell whether these Wiros were long-headed or short-headed, tall 
or of little stature, brunette or fair. It has been customary to 
imagine them as having something of the characteristics which 
Tacitus describes as belonging to the German of the end of the first 
century A.D. But all the evidence adduced in support of this is 
really imaginary. What, therefore, can we say that we know of 
this early people ? From words preserved in their languages, par- 
ticularly in languages far separated, and in circumstances where 
there is little likelihood of borrowing from the one language to 
the other, we may gather something as to the animals and the 
plants they knew, and perhaps a very little as to their industries. 
The close similarity between the various languages spoken by them 
would lead us to infer that they must have lived for long in a 
severely circumscribed area, so that their peculiarities developed 
for many generations in common. Since the study of prehistoric 
man developed, many views have been held as to the geographical 

Ill] The Wiros 67 

position of this early community. Such a confined area must have 
been separated from the outer world either by great waters or by 
mountains. There are however, so far as we know, no rivers in 
the western half of the Old World which at any period have pre- 
sented an impassable barrier to man. In the evidence for the 
early history of the speakers of these Indo-European or Indo- 
Germanic languages there is nothing which would lead us to suppose 
that they lived upon an island. Indeed, it is very doubtful whether 
they possessed a word for the sea at all. For the word mare which 
in Latin means 'the sea,' has its nearest relatives in other languages 
amongst words which mean ' moor ' or ' swamp.' That the climate 
in which they lived belonged to the temperate zone is shown by 
the nature of the trees which a comparison of their languages 
leads us to believe they knew. To their habitat we may assign, 
with considerable certainty, the oak, the beech, the wOlow, and 
some coniferous trees. The birch seems to have been known to 
them and possibly the lime, less certainly the elm. The fruits they 
knew are more uncertain than the forest trees. Many species of 
fruit trees familiar to us have flourished in Europe since late geo- 
logical times ; but at all periods men have been anxious to improve 
the quality of their fruit, and in all probability the commoner cul- 
tivated forms became known in northern and north-western Europe 
only as introduced by the Romans in the period of their conquests 
beginning with the first century b.c. Cherries have grown in the 
West from a very early period, but the name itself supports the 
statement that the cultivated kind was introduced by the great 
LucuUus in the first half of the first century B.c. from Cerasus in 
Asia Minor, an area to which the Western world owes much of its 
fruit and flowering shrubs. The ancient kings of Persia encouraged 
their satraps to introduce new fruit trees and better kinds into the 
districts which they ruled. There still exists a late copy of an 
early inscription in Greek in which the King of Persia gives praise 
to one of his governors for his beneficent action in this respect. 

These Wiros were in all probability not a nomad but a settled 
people. The useful animals best known to them were the ox and cow, 
the sheep, the horse, the dog, the pig, and probably some species of 
deer. The ass, the camel, and the elephant were apparently un- 
known to them in early times ; and the great variety of words for 
the goat would lead us to suppose that this animal also was of later 
introduction. The argument from language, however, is of neces- 
sity inconclusive, because all nations occasionally give animals with 
which they are familiar fanciful names. The Wiros seem also to 


68 The Aryans [ch. 

have been familiar with corn. If so, they must in all probability 
have lived for a considerable part of the year in one situation; 
for the planting of corn implies care continued over many weeks 
or months — care which the more primitive tribes have not been 
able to exercise. Of birds, we may gather from the languages that 
they knew the goose and the duck. The most familiar bird of prey 
was apparently the eagle. The wolf and bear were known, but not 
the lion or the tiger. 

From these data is it possible to locate the primitive habitat 
fi'om which the speakers of these languages derived their origin ? 
It is not likely to be India, as some of the earlier investigators 
assumed, for neither flora nor fauna, as determined by their lan- 
guage, is characteristic of this area, though some forest trees like 
the birch are more magnificent on Kinchinjunga than in any part 
of the Western world. Still less probable is the district of the 
Pamirs, one of the most cheerless regions on the face of the earth. 
Central Asia, which has also been contended for as their home, is 
not probable, even if we admit that its conspicuous lack of water, 
and consequent sterility in many areas, is of later development. 
If indeed these early men knew the beech, they must have lived to 
the west of a line drawn from Konigsberg in Prussia to the Crimea 
and continued thence through Asia Minor. In the Northern plains 
of Europe there is no area which will satisfactorily fulfil the con- 
ditions. As we know it in primitive times it is a land of great 
forests. No country, however, which had not much variety of 
geographical features could have been the habitat of both the 
horse and the cow. The horse is a native of the open plain ; the 
foal is able to run by its mother from the first, and accompanies 
her always in her wanderings. The calf, on the other hand, is at 
first feeble, unable to walk or see its way distinctly, and therefore 
is hidden by its mother in a brake while she goes further afield to 
find suitable pasture. Is there any part of Europe which combines 
pastoral and agricultural country in close connexion, which has in 
combination hot low-lying plains suitable for the growth of grain, 
and rich upland pasture suitable for flocks and herds, and at the 
same time trees and birds of the character already described? 
There is apparently only one such area in Europe, the area which 
is bounded on its eastern side by the Carpathians, on its south by 
the Balkans, on its western side by the Austrian Alps and the 
Bohmer Wald, and on the north by the Erzgebirge and the moun- 
tains which link them up with the Carpathians. This is a fertile 
and well- watered land with gi'eat corn plains in the low-lying levels 

Ill] Home of the Wiros 69 

of Hungary, but also possessing steppe-like areas which make it 
one of the best horse-breeding areas in Europe, while, in the 
uplands which surround it and run across it, as in the case of the 
Bakony Wald, south-west of Buda-Pesth, and still more markedly 
in Bohemia, there is high ground suitable for the pasturing of 
sheep. The forests of the mountains which engirdle it supply ex- 
cellent mast for the maintenance of swine whether wild or tame. 
The beech which dies out further south is found here and all the 
other great forest trees which have been already mentioned. The 
country is large enough to maintain a very considerable population 
which however was likely in primitive times to migrate from it 
only under the stress of dire necessity, because it is so well bounded 
on all sides by lofty mountains with comparatively few passes, that 
exit from it even in more advanced ages has not been easy. If 
this area indeed were the original habitat — and, curiously enough, 
though it fulfils so many of the conditions, it seems not before to 
have been suggested — the spread of the Indo-Germanic languages 
becomes easily intelligible. No doubt the most inviting direction 
from which to issue from this land in search of new homes would 
be along the course of the Danube into Wallachia, from which it 
is not difiicult to pass south towards the Bosporus and the Dar- 

A popular view locates the home of the Wiros in the southern 
steppes of Russia, but that area, though possessing a very fertile 
soil, has not on the whole the characteristics which the words 
common to the various Indo-Germanic languages, and at the same 
time unborrowed from one to another, postulate. It has also been 
commonly assumed that the eastern branches of the family found 
their way into Asia by the north of the Black Sea and either round 
the north of the Caspian or through the one pass which the great 
barrier of the Caucasus provides. Here we are met by a new 
difficulty. The Caspian is an inland sea which is steadily becoming 
more shallow and contracting in area. Even if it had been little 
larger than it is at present, the way into Turkestan between it and 
the Aral Sea leads through the gloomy desert of Ust Urt which, 
supposing it existed at the period when migration took place, must 
have been impassable to primitive men moving with their families 
and their flocks and herds. But there is good evidence to show 
that at a period not very remote the Caspian Sea extended much 
further to the north, and ended in an area of swamps and quick- 
sands, while at an earlier period which, perhaps, however, does not 
transcend that of the migration, it spread far to the east and 

70 The Aryans [ch. 

included within its area tlie Sea of Aral and possibly much of the 
low-lying plains beyond. Turkestan in primitive times would there- 
fore not have been easily accessible by this route. There is in 
fact no evidence that the ancestors of the Persians, Afghans, and 
Hindus passed through Turkestan at all. Nor is passage through 
the Caucasus probable : to people wandering from Europe the 
Caucasus was a remote and inhospitable region, so remote and so 
inhospitable that Aeschylus selected it as the place of torment for 
Prometheus and tells us that it was a pathless wilderness. There 
is indeed no reason to suppose that earlier men followed any other 
route than that which has been taken by successive waves of 
migratory populations in historical times. That path leads across 
either the Bosporus or Dardanelles, across the plateau of Asia 
Minor, or along its fertile slopes on the south side of the Black Sea. 
A European people which would reach Persia on foot must strike 
the upper waters of the Euphrates and Tigris. The fertile country 
with an alluvial soil of tremendous depth, which lies between these 
two rivers, was the centre of one of the earliest and one of the 
most powerful civilisations of ancient times. Migrants would there 
find their progress to the south obstructed and baulked. But by 
passing south of Lake Van and through the mountains which lie 
between it and Lake Urmia, they would find an access to the route 
which travellers still follow between Tabriz and Teheran. From 
there they would advance most likely along the southern end of 
the Caspian towards Mashhad, whence in all ages there has been a 
well-frequented route to Herat. At one time these peoples cer- 
tainly extended far to the east and north, to the country then 
known as Bactria, now Balkh, and carried their conquests into the 
famous region which lies between the two rivers, the Amu Daria, 
or Oxus, and the Syr Daria. 

What evidence have we of such a migration, and, if it took 
place, what was its date ? In all probability the migration of 
peoples from the primitive habitat, which we have located in the 
areas which we now call Hungary, Austria, and Bohemia, did not 
take place at a very remote period. It is indeed probable that all 
the facts of this migration, so far as we know them, can be ex- 
plained without postulating an earlier beginning for the migrations 
than 2500 B.C. It must be remembered, however, that these 
migrations were not into unpeopled areas, that before they reached 
the frontiers of India, or even Mesopotamia, the Wiros must have 
had many hard struggles with populations already existing, who 
regarded their passage as they would that of some gi-eat cloud of 

Ill] Migrations 71 

destroying locusts which devoured their substance and left them 
to perish by starvation, or to survive in the misery of captives to 
cruel conquerors. We must suppose that success could have been 
achieved only by wave after wave following at no 4ong intervals : 
for if their successors delayed too long, the migrants of the first 
advancing wave were likely to be cut off or absorbed. In historical 
times, we know that many tribes thus passed into Asia from Europe, 
among them the Phrygians, the Mysians, and Bithynians. It has 
been plausibly argued that the Armenian stock was the first wave 
of the Phrygian advance, and evidence can be adduced which 
makes it probable that still earlier waves of conquering tribes 
advancing from west to east were represented by the remote 
ancestors of modern Persians and modern Hindus. 

If, as some scholars suppose, modern Albanian is the descendant 
in a very corrupt condition of ancient Thracian, and not of ancient 
Illyrian, the interrelation of the ancient branches of the Indo- 
Germanic family of languages can be outlined. The family is 
divided by a well-marked difierence in the treatment of certain 
k, g, and gh sounds into two parts, one of which keeps the k, g, and 
gh sounds, though submitting them to a variety of changes in later 
times, while the other part changes k and g into some kind of 
sibilant sounds which are represented in the Slavonic and Iranian 
languages by s and z, in Sanskrit by ^ and J. The ^A sound appears 
as z in Zend, the Iranian dialects confusing together g and gh, 
while in Sanskrit it appears as h. The languages which present 
these changes are the easternmost members of the family : Aryan 
(i.e. Indian and Iranian) ; Armenian ; Slavonic ; and Albanian. 
The Albanian it is suggested has been driven westward through 
the Pindus range into its present position within historical times, 
the ancient Illyrians having in this area been swept away in the 
devastation wrought by a sequence of Roman invasions, initiated 
in the second century B.C. by Aemilius Paulus. The languages 
mentioned would thus have started from the eastern side of the 
original habitat, while the tribes which (with an admixture of the 
population already in possession) ultimately became the Greeks, 
moved through Macedonia and Thessaly southwards, and the Latin 
stock, the Celts, and the Germans westwards and northwards. It 
is more than likely that the ancestors of the Slavs found their way 
from the original home by the ' Moravian Gap.' The exact manner, 
or the exact date, at which these movements took place we cannot 
tell, but there is no reason to suppose that any of them antedate 
at earliest the third millennium, b.c. Nor is it likely that they took 

72 The Aryans [cH. 

place all at once. The same causes, though in dififerent degrees, 
were operative then which have produced movements of peoples 
in historical times, one of the most pressing probably being the 
growth of population in a limited area, which drove sections or 
whole tribes to seek sustenance for themselves, their families, and 
cattle in land beyond their original boundaries, without regard to 
whether these lands were already occupied by other peoples or 
not. The movements of the Gauls in historical times were probably 
not at all unlike those of their ancestors and kinsmen in prehistoric 

If, as has been suggested above, the early speakers of the 
primitive Indo-Germanic language occupied a limited area well 
defended by mountains from attack, this would account for the 
general similarity of the languages in detail; if, forced by the 
natural increase of population, they left this habitat in great waves 
of migration, we can see how some languages of the family, as for 
example, the Celtic and the Italic, or the Iranian and the Indian, 
are more closely related to one another than they are to other 
members of the family ; if, further, we assume that such a habitat 
for the prehistoric stock could be found in the lands which we call 
Hungary, Austria, and Bohemia, we can explain a very large 
number of facts hitherto collected for the history of their earlier 
movements and earlier civilisation. 

Of the earliest movements of the tribes speaking Indo-Germanic 
languages which occupied the Iranian plateau and ultimately passed 
into Northern India, history has as yet nothing to say. But recent 
discoveries in Cappadocia seem likely to give us a clue. In the 
German excavations at Boghaz-kbi, the ancient Pteria, have been 
found inscriptions, containing as it appears the names of deities 
which figure in the earliest Indian records, Indra, Yaruna, and the 
great twin brethren the Nasatyas. The inscriptions date from 
about 1400 B.C., and the names appear not in the form which they 
take in the historical records of ancient Persia, but are, so far as 
writing in a syllabary will admit, identical with the forms, ad- 
mittedly more original, which they show in the hymns of the 
Rigveda. It is still too early to dogmatise over the results of 
these discoveries, which it may be hoped are only the firstfruits 
of a rich harvest ; but the most feasible explanation of them seems 
to be that here, far to the west, we have stumbled upon the Aryans 
on the move towards the east. This is not to say that earlier waves 
may not long before 1400 B.C. have penetrated much further to the 
east, or even to India itself. All that can be gathered from these 


III] Inscriptions of Boghaz-kdi 73 

discoveries is that at this period the Mitani, who were apparently 
not of this stock themselves, had adopted the worship of certain 
deities of this stock — deities who at the time of the composition 
of the Vedic Hymns were still the most important, though to them 
had been added Agni, * Fire,' specially an object of priestly wor- 
ship in the Vedic hierarchy. We have here, however, names 
practically in the form in which they survive in Sanskrit, and 
without the changes which characterise the records of the tribes 
of this stock, who remained in Persia. To this as yet unbroken 
unity the name of Aryan is given. It is borrowed from a word 
which appears as Arya, or Arya in Sanskrit, Airy a in Zend, and 
which means ' of good family, noble.' It is the epithet applied by 
the composers of the Vedic hymns to distinguish their own stock 
from that of their enemies the earlier inhabitants of India, whom 
they call Dasas or Dasyus. The term, by reason of its shortness, 
has often been applied to all the languages of this family, in pre- 
ference to * Indo-European ' or * Indo-Germanic,' but is properly 
reserved for the south-eastern group which, when the phonetic 
changes characterising the language of the Avesta and of the old 
Persian inscriptions of the Achaemenid dynasty (520 B.C.-330 B.C.) 
have taken place, falls into the two branches of Iranian and Indo- 
Aryan. The latter term well characterises the Aryans settled in 
India, while Aryo-Indian conveniently designates these Aryans as 
distinct from the unrelated stocks — Dravidian and other — also 
inhabiting the Indian peninsula. 

As these inscriptions of Boghaz-koi show the language still one 
and undivided, we obtain a limit after which the differentiation of 
Iranian and Indo- Aryan must have begun. These Aryan languages 
have some characteristics in common which distinguish them from 
all others : in particular they agree in confusing together the three 
original vowels a, e, and 0, whether long or short, into one sound 
which is written with the symbols for a and a. In modern India 
at least the short sound is pronounced with the obscure vowel 
found in the English 'but,' a fact which produced the English 
spelling of the Hindu words * pundit ' (pandita) and * suttee ' (satt), 
and disguised the liquor compounded of five (pancha) ingredients 
under the apparently English form of ' punch.' They agree also on 
the whole in the case system of the noun, a system to which the 
Slav and Armenian languages offer the closest approximation, and 
in the elaborate mood and voice system of the verb, to which the 
only parallel is to be found in the similar, though not in all respects 
identical, paradigms of Greek. Here the other languages, except 

74 Tf^^ Aryans [ch. 

the Slavonic, fall far short of the elaborate and intricate Aryan 
verb system, whether it be, as is most likely, that the other tribes 
have lost a large part of their share of the common inheritance, or 
whether some of the languages drifted apart, before the complete 
system, seen in the Aryan and Greek verbs, had developed. Other 
changes may with probability be attributed to the influence of the 
peoples whom they conquered and enslaved. A characteristic, 
which distinguishes the languages of this stock in both Persia and 
India is the tendency to confuse r and I, a tendency which is 
characteristic of practically all the languages of the far east. In 
India r is often found in words where the languages of the same 
stock in Europe show l\ l'\9> also, though not so frequently, found 
for r ; in the Old Persian of the Achaemenid inscriptions I is 
found only in two foreign words, and has otherwise been entirely 
replaced by r. 

The dialects of Iran, the language of the earliest Gathas (Songs) 
which are attributed to Zoroaster himself, the later dialect of the 
other surviving parts of the sacred literature of the ancient Persians 
— the Avesta — and the inscriptions beginning with Darius I about 
620 B.C. and best represented in his time but continuing to the last 
Darius in 338B.C., are all closely related to the oldest dialect 
discovered in India, which appears in the hymns of the Rigveda. 
Not only single words and phrases, but even whole stanzas may be 
transliterated from the dialect of India into the dialects of Iran 
without change of vocabulary or construction, though the appear- 
ance of the words is altered by the changes which time and isolation 
have brought about between the dialects east and west of Afghani- 
stan. It is curious to note that the changes are much greater in 
the dialects that remain in Iran than in this oldest recorded dialect 
of the migrants into India. The Iranians have disguised their 
words by changing (as Greek has also done) s followed by a vowel 
at the beginning of words, or between vowels in the middle of 
words, into h : thus the word for 7, the equivalent of the Latin 
scptem, the Greek kirra, is in Sanskrit saptd, but in Iranian hapta. 
There are many other changes both in vowels and in consonants. 
In particular it may be noted that one kind of original g which 
appears in Sanskrit as j has become in the Iranian dialect z ot 8 
(Greek d'yv6<i 'holy,' Sanskrit yajnd- 'sacrifice,' Avesta yasna), 
and a corresponding aspirated sound gh which is in Sanskrit h has 
become identified with g in Iranian as z (Latin hiems, Greek x''^^i 
;)^et/ia, (8yo-);^6/A09, Sanskrit hima- in ' Himalaya,' Avesta zyam-). This 
loss of aspiration has affected also the other aspirates hh, dh, which 


III] lra7iia?is and htdians 75 

survive in Sanskrit, while Iranian tends in certain combinations 
to change original consonant-stops into spirants, making the old 
name of the deity Mitra into Mithra, and from compounds with a 
second element -parna the numerous proper names which we know 
in Greek transliterations as Artaphernes, Tissaphernes, and the like. 

It has sometimes been made an argument for deriving the 
origin of these tribes from India rather than the West, that the 
sounds and especially the consonants of the language spoken have 
survived in greater purity in India than in Iran or elsewhere. 
The argument however is not sound. Invasions of a similar sort, 
though at a much greater distance from their base, were made by 
the Spaniards in America in the sixteenth century. The civilisation 
of the Spaniards was no doubt higher than that of the early Indo- 
Germanic-speaking peoples who invaded India ; but in both 
Mexico and Peru, if not elsewhere, they met a native population p/jii 
also much more advanced in the arts than the earlier inhabitants ' • \ ' > a^ 
of North-Western India could have been. In all parts of America, lcv\ t> w » ^ 
except Chile, the Spaniards were in so small a minority compared ccf "i 'it $ 
to the natives that they had to be careful to preserve themselves ■/ ,aj £ O'f' 
in isolation, with the result that to-day, except in Chile, where 
greater familiarity with the natives has produced a dialect of J vt4 «^ 1 
Spanish words and native sounds, the local dialects are much more \Ja ' ' ^f 
archaic and much more like the Spanish of the sixteenth century ^4*^ t ^iSjit * 
than is the language spoken now in Spain. If the isolation of the x » \ 

English Colonies in North America had remained as great as it was -^ 

in the seventeenth century, no doubt a much greater distinction 
would now exist between the English dialects of North America 
and the English of the Mother country. Yet in many parts of the 
eastern seaboard of the United States many words survive locally 
which have long been extinct except in local dialects in England, 
and many forms of expression survive which the modern English- 
man now regards as mainly biblical. That an isolation resembling 
that of the Spanish colonies prevailed also in early India is shown 
by the most characteristic feature of Indian civilisation — caste. 
The native word for caste, varna, means colour, and the first 
beginnings of the caste system were laid when the fairer people 
who migrated into India felt the importance of preserving their 
own racial characteristics by standing aloof from the dark skinned 
ddsas, or dasyus, whom they found already established in the 

That the sound changes which have been enumerated are not 
so very old has been shown by the names found at Boghaz-koi. And 

76 The Aryans [CH. in 

this is not the only evidence. To the same period as the Boghaz-koi 
inscriptions belong the famous letters from Tel-el-Amama. In these 
occur references to the people of Mitani in north-west Mesopotamia, 
whose princes bear names like Artatama, Tusratta, and Suttania, 
which seem unmistakably Aryan in form. For five hundred 
years (c. 1746-1180 B.c.) a mountain tribe — the Kassites — from 
the neighbourhood of Media held rule over the whole of Babylonia, 
and amongst these also the names of the princes and deities seem 
Aryan, though the people themselves, like those of Mitani were of 
another stock. Names like Shurias * Sun ' and Marytas seem 
identical with the Sanskrit Surya and Marutas (the mnd-gods), 
while Simalia 'queen of the snow mountains' can hardly be 
separated from the name of the great mountain range Himalaya 
and the Iranian word for snow, zima. To a much later period 
belongs the list of deities worshipped in different temples of 
Assyria, which was found in the library of Assurbanipal (about 
700 B.C.), in which occurs the name Assara-Mazas, immediately 
preceding the seven good angels and the seven bad spirits. The 
combination hardly leaves it doubtful that w^e have here the chief 
deity of Zoroastrianism (Ahura Mazda) with the seven Amesha- 
spentas and the seven bad daivas of that religion. Into the many 
other problems that arise in this connexion it is not necessary here 
to enter ; but it is important to observe that even so late as this 
the first part of the god's name remains more like the Sanskrit 
Asura than the Avestan Ahura. While modern Hinduism is the 
lineal descendant, however much modified in the course of ages, 
of the ancient Aryan worship which we know first in the Rigveda, 
the religion of the Avesta is a reform which, like other religious 
reforms, has been able to get rid of the old gods only by converting 
them into devils, the worship of which was probably none the less 
diligent for their change of title. 

There seems, in any case, to be specific evidence for the sup- 
position that by the fifteenth century b.c. tribes of Aryan stock 
held, or exercised influence over, a wide area extending from 
northern Asia Minor over north-west Babylonia to Media; and 
there seems to be nothing to prevent us assuming that even then, 
or soon after, the Aryans pushed their way still eastwards and 
northwards, mainly confining themselves to the territories south 
of the Oxus, but occasionally occupying lands between that river 
and the Jaxartes. 



The earliest documents which throw light upon the history of 
India are the hymns of the Rigveda. In the text which has come 
down to us this samhitd or 'collection' consists of 1017 hymns 
divided into ten books of unequal size. The motive of those to 
whom the collection is due must apparently have been the desire 
to preserve the body of religious tradition current among the 
priests ; and, early as was the redaction, there are clear signs that 
already part of the material had ceased to be fully understood by 
those who made use of it in their worship. The artificial character 
of the arrangement is clearly indicated by the fact that the first 
and tenth books have precisely the same number of hymns, 191 
each. The collection seems however to have been some time in 
the making. The nucleus is fonned by books ii-vii, each of which 
is attributed to a difierent priestly family. To this were prefixed 
the groups of hymns by other families which form the second part 
(51-191) of book i; and still later were added the first part of 
book I and book viii attributed to the family of Kanva. Book ix 
was then formed by taking out from the collections of hymns 
which made up the first eight books the hymns addressed to Soma 
Pavamana, 'the clearly flowing Soma'; and to these nine books 
was added a tenth, containing, besides h}Tnns of the same hieratic 
stamp as those of the older books, a certain nimiber of a different 
type, cosmogonic and philosophical poems, spells and incantations, 
verses intended for the rites of wedding and burial and other 
miscellaneous matters. The tenth book also displays, both in 
metrical form and linguistic details, signs of more recent origin 
than the bulk of the collection ; and the author of one set of hymns 
(x, 20-26) has emphasised his dependence on earlier tradition by 
prefixing to his own gi'oup the opening words of the first hymn of 
the first book. 

78 The Age of the Rigveda [ch. 

There is abundant proof that, before the collections were finally 
united into the form in which the Rigveda has come down to us, 
minor additions were made ; and, as it is perfectly possible that in 
book X old material was incorporated as well as newer work, 
efforts have been made to penetrate beyond the comparatively 
rough distinction between the first nine and the tenth books, and to 
assign the hymns to five different periods, rei)resenting stages 
in the history of Vedic India, and marked by variations in 
religious belief and social custom ^ But so far these efforts can 
scarcely be regarded as successful. The certain criteria of age 
supplied by the language, the metres, or the subject-matter of the 
Rigveda are not sufficient to justify so elaborate a chronological 
arrangement of its hymns. The results produced by the most 
elaborate and systematic attempts to apply the methods of the 
higher criticism to the Rigveda have hitherto failed to meet with 
general acceptance. 

Tlie mass of the collection is very considerable, approximating 
to the same amount of material as that contained in the Iliad and 
Odyssey, but the light thrown by the hymns on social and political 
conditions in India is disappointingly meagi*e. By far the greater 
part of the Rigveda consists of invocations of the many gods of the 
Vedic pantheon, and scarcely more than forty hymns are found which 
are not directly addressed to these deities or some object to which 
divine character is, for the time at least, attributed. These hymns 
contain much miscellaneous information regarding Vedic life and 
thought ; and other notices may be derived from the main body of 
the collection, though deductions from allusions are always difficult 
and open to suspicion. Some names of tribes, places, and princes, 
as well as of singers, are known to us through their mention in the 
ddnastutis or 'praises of liberality' which are appended to hymns, 
mainly in the first and tenth books, and in which the poet praises 
his patron for his generosity towards him. But the ddnastvtis are 
unquestionably late, and it is significant that some of the most 
striking occur in a small collection of eleven hymns, called the Vala- 
khilyas, which are included in the Samhita of the Rigveda, but which 
tradition recognises as forming no true part of that collection. 

From these materials conclusions can be drawn only with much 
caution. It is easy to frame and support by plausible evidence 
various hypotheses, to which the only effective objection is that 
other hypotheses are equally legitimate, and that the facts are too 

^ Especially by Arnold whose results are summed up in his Vedic Metre (Cambridge, 
1905). For criticism, see J.R.A.S., 1906, pp. 484-90, 716-22; 1912, pp. 726-9. 

I v] Geography 7 9 

imperfect to allow of conclusions being drawn. It is, however, 
certain that the Rigveda offers no assistance in determining the 
mode in which the Vedic Indians entered India. The geographical 
area recognised in the Samhita is large, but it is, so far as we 
learn, occupied by tribes which collectively are called Aryan, and 
which wage war with dark-skinned enemies known as Dasas. If, 
as may be the case, the Aryan invaders of India entered by the 
western passes of the Hindu Kush and proceeded thence through 
the Punjab to the east, still that advance is not reflected in the 
Rigveda, the bulk at least of which seems to have been composed 
rather in the country round the SarasvatI river, south of the 
modern Ambala\ Only thus, it seems, can we explain the fact of 
the prominence in the hymns of the strife of the elements, the 
stress laid on the phenomena of thunder and lightning and the 
bursting forth of the rain from the clouds: the Punjab proper 
has now, and probably had also in antiquity, but little share in 
these tilings ; for there in the rainy season gentle showers alone fall. 
Nor in its vast plain do we find the mountains which form so 
large a part of the poetic imagining of the Vedic Indian. On the 
other hand, it is perhaps to the Punjab with its glorious phenomena 
of dawn, that we must look for the origin of the hymns to Ushas, 
the goddess Dawn, while the concept of the laws of Varuna, the 
highest moral and cosmic ideal attained by the poets, may more 
easily have been achieved amid the regularity of the seasonal 
phenomena of the country of the five rivers. 

Of the names in the Rigveda those of the rivers alone permit 
of easy and certain identification. The Aryan occupation of 
Afghanistan is proved by the mention of the Kubha (Kabul), the 
Suvastu (Swat) with its 'fair dwellings,' the Krumu (Kurram) and 
Gomati (Gumal). But far more important were the settlements 
on the Sindhu (Indus), the river par excellence from which India 
has derived its name. The Indus was the natural outlet to the sea 
for the Aryan tribes, but in the period of the Rigveda there is no 
clear sign that they had yet reached the ocean. No passage even 
renders it probable that sea navigation was known. Fishing is all 
but ignored, a fact natural enough to people used to the rivers of 
the Punjab and East Kabulistan, which are poor in fish. The word 
sa/nvudra, which in later times undoubtedly means 'ocean,' occurs 

' See Hopkins, J.A.O.S., vol. xix, pp. 19-28; Pischel and Geldner, Vedische 
Studien, vol. ii, p, 218 ; vol. in, p. 152 ; Vedic Index, vol. r, p. 468. The older view, 
that the hymns were composed in the Punjab itself, was adopted by Max Miiller, 
Weber, and Muir among others. 

8o The Age of the Rigveda [ch. 

not rarely ; but where the application is terrestrial, there seems no 
strong reason to believe that it means more than the stream of the 
Indus in its lower course, after it has received the waters of the 
Punjab and has become so broad that a boat in the middle cannot 
be discerned from the bank. Even nowadays the natives call the 
river the sea of Sind. 

The five streams which give the Punjab its name and which 
after uniting flow into the Indus are all mentioned in the Rigveda : 
the Vitasta is the modern Jhelum, the Asikni the Chenab, the 
Parushni, later called Iravati, 'the refreshing,' the modern Ravi, the 
Vipac the Beas, and the (^utudri the Sutlej. But of these only the 
Parushni plays a considerable part in the history of the time, for it 
was on this river that the famous battle of the ten kings, the most 
important contest of Vedic times, was fought. Far more important 
was the Sarasvati, which we can with little hesitation identify with 
the modern Sarsuti or Saraswati, a river midway between the 
Sutlej and the Jumna \ It is possible that in the period of the 
Rigveda that river was of greater importance than it was in the 
following period when it was known to bury itself in the sands, and 
that its waters may have flowed to the Indus; but, however that 
may be, it is mentioned in one passage together with the Drishad- 
vati, probably the Chautang, which with it in later times formed 
the boundaries of the sacred land known as Brahmavarta. With 
these two streams is mentioned the Apaya, probably a river near 
Thanesar^. In this region too may be placed the lake (^aryanavant^ 
and the place Pastyavant, near the modern Patiala. 

Further east the Aryans had reached the Jumna, which is thrice 
named, and the Ganges, which is once directly mentioned, once 
alluded to in the territorial title of a prince. 

To the north we find that the Himavant or Himalaya mountains 
were well known to the Rigveda, and one peak, that of Mujavant, 
is referred to as the source of the Soma, the intoxicating drink 
which formed the most important ofiering in the religious practice 

1 Both, St Petersburg Dicticmary, s.v., and Zimmer, Altindisches Leben, pp. 5-10, 
identify the Sarasvati in many passages with the Indus; Hillebrandt, Vedische 
Mythologie, vol. i, pp. 99 sq.; vol. in, pp. 372-8, thinks it is in a few places 
the Arghandab. 

2 The identification of the ancient rivers of Brahmavarta must always remain 
somewhat uncertain. At the present day it is difficult to trace their courses, partly 
because the streams are apt to disappear in the sand, and partly because they have to 
a great extent been absorbed in the canal-systems constructed during the periods of 
Muhammadan and British rule. 

3 Identified however with the Wular Sea in. Kashmir by Hillebrandt* Vedische 
Mythologie, vol. i, pp. 126 sq. 

iv] Fauna and Peoples 8 1 

of the time. The name is lost in modern times, but probably the 
peak was one of those on the south-west of the valley of Kashmir. 
On the south, on the other hand, the Vindhya hills are unknown, 
and no mention is made of the Narbada river, so that it may fairly 
be inferred that the Aryan tribes had not yet begun their advance 
towards the south. 

With the conclusions as to the home of the Aryan tribes 
extracted from geographical names the other available evidence 
well accords. The tiger, a native of the swampy jungles of Bengal, 
is not mentioned in the Rigveda, which gives the place of honour 
among wild beasts to the lion, then doubtless common in the vast 
deserts to the east of the lower Sutlej and the Indus and even now 
to be found in the wooded country to the south of Gujarat. Rice, 
whose natural habitat is the south-east in the regular monsoon 
area and which is well known in the latter Samhitas, is never 
mentioned in the Rigveda. The elephant, whose home is now in the 
lowland jungle at the foot of the Himalaya from the longitude of 
Cawnpore eastwards, appears in the Rigveda as the wild beast 
(mriga) with a hand (hastin), while in the later texts it is com- 
monly known as hastin only, a sign that the novelty of the animal 
had worn away. The mountains from which the Soma was brought 
appear, too, to have been nearer in this period than at a later date 
when the real plant seems to have been more and more difficult to 
obtain, and when substitutes of various kinds were permitted. 

When we pass to the notices of tribes in the Rigveda, we leave 
comparative certainty for confusion and hypothesis. The one 
great historical event which reveals itself in the fragmentary 
allusions of the Samhita is the contest known as the battle of the 
ten kings. The most probable version of that conflict is that it 
was a contest between the Bharatas, settled in the country later 
known as Brahmavarta, and the tribes of the north-west. The 
Bharata king was Sudas, of the Tritsu family, and his domestic 
priest who celebrates, according to the tradition, the victory in 
three hymns (vii, 18 ; 33 ; 83) was Yasishtha^ This sage had super- 
seded in that high office his predecessor Vi9vamitra, under whose 
guidance the Bharatas appear to have fought successfully against 
enemies on the Vipa? and ^utudri; and in revenge, as it seems, 

^ This is the view of Hopkins, J.A.O.S., vol. xv, pp. 259 sq. According to the 
older view the Bharatas were foes of the Tritsus; see Muir, Original Sanskrit Texts, 
vol. 1^ p. 354; Zimmer, Altindisches Leben, p. 127; Bloonafield, J.A.O.S., vol. xvi, 
pp. 41, 42. Ludwig, Rigveda, vol. in, p. 172, identified the Bharatas and the Tritsus ; 
Oldenberg, Z.D.M.G., vol. xlii, p. 207, holds that the Tritsus are the Vasishthas, the 
priests of the Bharatas. But see Geldner^ Vedische Studien, vol. ii, pp. 136 sq. 

C.H.I. I. 6 


82 T'/ie Age of the Rigveda [CH. 

Vi^vamitra had led against the Bharatas ten allied tribes, only to 
meet with destruction in the waters of the Parushni. Of the ten 
tribes five are of little note, the Alinas, perhaps from the north- 
east of Kafiristan, the Pakthas, whose name recalls the Afghan 
Pakhthun, the Bhalanases, possibly connected with the Bolan 
Pass, the ^ivas from near the Indus, and the Vishanins. Better 
known in the Rigveda are the other five, the Anus who dwelt on 
Parushni and whose priests were perhaps the famous family of the 
Bhrigus, the Druhyus who were closely associated with them, the 
Turva^as and Yadus, two allied tribes, and the Purus, dwellers on 
either side of the Sarasvati, and therefore probably close neigh- 
bours of the Bharatas. These tribes are probably the five tribes 
which are referred to on several occasions in the Rigveda and 
which seem to have formed a loose alliance. Sudas's victory at 
the Parushni, in which the Anu and Druhyu kings fell, does not 
appear to have resulted in any attempt at conquest of the territory 
of the allied tribes. He seems at once to have been compelled to 
return to the east of his kingdom to meet the attacks of a king 
Bheda, under whom three tribes, the Ajas, (^igrus, and Yakshus, 
were united, and to have defeated his new assailants with great 
slaughter on the Jumna. It is probable enough that the attack 
on the eastern boundaries of the territory of the Bharatas was not 
unconnected with the onslaught of the five tribes and their still 
more northern and western allies ; but the curious names of the 
Ajas, 'goats,' and the (^Igrus, 'horse-radishes,' may be a sign that 
the tribes which bore thSm were totemistic non- Aryans. 

Not less famous was the father or grandfather of Sudas, 
Divodasa, 'the servant of heaven,' Atithigva, 'the slayer of kine for 
guests \' There are records of his conflicts with the Turva^a, Yadu, 
and Puru tribes ; but his greatest foe was the Dasa, ^ambara, with 
whom he waged constant war. He had to contend also with the 
Panis, the Paravatas, and Brisaya. He seems to have been the 
patron of the priestly family of the Bharadvajas, the authors of 
the sixth book of the Rigveda ; and there is little doubt that his 
kingdom covered much the same area as that of Sudas, since he 
warred, on the one hand, against the tribes of the Punjab, and, on 
the other, against the Paravatas who are located in the period of 
the Brahmanas on the Jumna. The Dasas and the Panis were 
probably aboriginal foes, whom, like every Aryan prince, he had to 

Though defeated in the battle with Sudas, the Piirus were 

» V. inf., pp. 101-2, and Chapter x, pp. 232-3. 

iv] The Taurus 83 

clearly a great and powerfiil people. Tlieir home was round the 
Samsvati, and there is no need to interpret that name as referring 
to the Indus rather than to the eastern SarasvatT. On the Indus 
they would have been removed somewhat Avidely from the Bha- 
ratas, their chief rivals, two of whose princes, Deva^ravas and 
Devavata, are expressly recorded in one hymn to have dwelt on 
the Sarasvati, Apaya, and Drishadvati. The importance of the 
tribe is reflected in the fact that we possess an unusually large 
number of the names of its members. The earliest prince recorded 
seems to have been Durgaha, who was succeeded by Girikshit, 
neither of these being more than names. The son of Girikshit, 
Purukutsa, was the contemporary of Sudas, and one hymn tells in 
obscure phrases of the distress to which his wife was reduced by 
some misfortune, from which she was relieved by the birth of a 
son, Trasadasyu. It is not unlikely that the misfortune was the 
death of Purukutsa in the battle of the ten kings. Tlie new ruler, 
as his name indicates, was a terror to the Dasyus or aborigines, and 
seems not to have distinguished himself in war with Aryan enemies. 
We hear of a descendant Trikshi, and, apparently still later in the 
line, of another descendant Kurugravana, son of Mitratithi and 
father of Upama^ravas, whose death is deplored in a hymn of the 
tenth book. The name is of importance and significance, for it 
suggests that already in the later Rigvedic period the Purus had 
become closely united with their former rivals, the Bharatas, both 
tribes being merged in the Kurus, whose name, famous in the 
later Samhitas and the Brahmanas as the chief bearers of the 
culture of the Vedic period, is not directly mentioned in the 
Rigveda, though it was clearly not unknown. Other princes of 
the Puru line were Tryaruna, and Trivrishan or Tridhatu; and 
later evidence enables us with fair certainty to connect with the 
Purus the princely name Ikshvaku, which occurs but once in a 
doubtful context in the Rigveda. 

Connected with the Kurus were the Krivis, whose name seems 
to be but a variant from the same root, and who appear to have 
been settled near the Indus and the Chenab. Possibly we may see 
the allied tribes of Kurus and Krivis in the two Vaikarna tribes, 
twenty-one of whose clans shared the defeat of the five tribes by 
Sudas. If so, like the Piirus the Bharatas must have in course of 
time become mingled with the Kurus and have merged their 
identity with them. 

Allied or closely connected with the Bharatas was the tribe of 
the Srinjayas, whom we must probably locate in the neighbourhood 


84 The Age of the Rigveda [ch. 

of the Bharatas. One of their princes, Daivavata, won a great 
victory over the Turva^as with their allies, the Vrichivants, of 
whom we know nothing more. Other princes of the line were 
Sahadeva, his son Somaka, and Prastoka, and Vitahavya. They 
were, like the Bharatas under Divodasa, closely connected with 
the Bharadvaja family of j)rie8ts. 

No other Aryan tribe plays a great figure in the Rigveda. 
The Chedis, who in later times dwelt in Bundelkhand to the north 
of the Vindhya, and their king Kayu are mentioned but once in a 
late ddnastuti : the queen of the U9inaras, later a petty tribe to the 
north of the Kuru country, is also once alluded to. The generosity 
of Rinamchaya, king of the Rugamas, an unknown people, has 
preserved his name from extinction. One interpretation adds to 
the enemies of Sudas the tribe of the Matsyas ('fishes'), who in 
later times occupied the lands now known as Alwar, Jaipur, and 
Bharatpur. A raid of the Turva9as and Yadus and a conflict 
on the Sarayu^ with Arna and Chitraratha testify to the activity 
of these clans, which otherwise are best known through their 
opposition to Divodasa and Sudas, and which must probably have 
been settled in the south of the Punjab. The family of the 
Kanvas seems to have been connected as priests with the Yadus. 
Connected with the Turvayas was the Vrichivant Varayikha, who 
was defeated by Abhyavartin Chayamana, who himself was perhaps 
a Srinjaya prince. More shadowy still are Nahus, Tugrya, and 
Vetasu in whom some have seen tribes : Nahus is probably rather 
a general term for neighbour, and the Tugryas and the Vetasus 
are families rather than tribes. 

More important by far, it may be believed, than the intertribal 
warfare of the peoples who called themselves Aryan were their 
contests with the aborigines, the Dasas or Dasyus as they are 
repeatedly called. The same terms are applied indifferently to 
the human enemies of the Aryans and to the fiends, and no 
criterion exists by which references to real foes can be distin- 
guished in every case from allusions to demoniacal powers. The 
root meaning of both words is most probably merely 'foe' ; but in 
the Rigveda it has been specialised to refer, at least as a rule, to 
such human foes as were of the aboriginal race. Individual Dasas 
were Ilibi^a, Dhuni and Chumuri, Pipru, Varchin, and ^ambara, 
though the last at least has been transformed by the imagination 
of the singers into demoniac proportions. The only peoples named 

^ The identification of this river is uncertain ; see Vedie Index, vol. 11, p. 434. 

iv] The Dasas or Dasyus 85 

which can plausibly be deemed to have been Dasas are the (^imyns, 
who are mentioned among the foes of Sudas in the battle of the 
ten kings, and who are elsewhere classed with Dasyus, the Kikatas 
with their leader Pramaganda, and perhaps the Ajas, Yakshus, and 
^igrus. The main distinction between the Aryan and the Dasa 
was clearly that of colour, and the distinction between the Aryan 
varna, 'colour,' and the black colour is unquestionably one of the 
main sources of the Indian caste system. The overthrow of the 
black skin is one of the most important exploits of the Vedic 
Indian. Second only to the colour distinction was the hatred of 
men who did not recognise the Aryan gods: the Dasas are con- 
stantly reproached for their disbelief, their failure to sacrifice, and 
their impiety. Nor is there much doubt that they are the phallus 
worshippers who twice are referred to with disapproval in the 
Rigveda, for phallus worship was probably of prehistoric age in 
India and by the time of the Mahabharata it had won its way into 
the orthodox Hindu cult. We learn, disappointingly enough, 
little of the characteristics of the Dasas, but two epithets applied 
in one passage to the Dasyus are of importance. The first is 
mridhravdehah which has been interpreted to refer to the nature 
of the aborigines' speech ; but which, as it elsewhere is applied to 
Aryan foes like the Piirus, probably means no more than *of hostile 
speech.' The other epithet, andsah, is more important : it doubt- 
less means 'noseless,' and is a clear indication that the aborigines 
to which it is applied were of the Dravidian type as we know it at 
the present day. With this accords the fact that the Brahul speech 
still remains as an isolated remnant in Baluchistan of the Dravidian 
family of tongues \ But though the main notices of the Rigveda 
are those of conflict against the Dasas and the crossing of rivers to 
win new lands from them, it is clear that the Aryans made no 
attempt at wholesale extermination of the people. Many of the 
aborigines doubtless took refuge before the Aryan attacks in the 
mountains to the north or to the south of the lands occupied by the 
invaders, while others were enslaved. This was so normal in the 
case of women that, in the literature of the next period, the term 
DasI regularly denotes a female slave; but male slaves are often 
alluded to in the Rigveda, sometimes in large numbers, and wealth 
was already in part made up of ownership of slaves. The meta- 

1 In the Imp. Gaz., vol. i, p. 382, it is suggested that the Brahuis who are there 
ethnographically classed as Turko-Iranian show the original type of Dravidian, and 
that the modern Dravidian type is physically due to influeuce by the Munda speaking 
peoples. The Bigvedic evidence does not favour this view. See Chapter ii, pp. 42-3. 

86 The Age of the Rigveda [ch. 

phorical use is seen in the name of one of the greatest of Vedic 
kings, Divodasa, 'the servant of heaven.' In the Purushasukta, or 
'Hymn of Purusha,' which belongs to the latest stratum of the 
Rigveda, and which in mystic terms describes the creation of the 
four castes from a primeval giant, occurs for the first time the term 
^udra, which includes the slaves as a fourth class in the Aryan 
state. Probably enough this word, which has no obvious explana- 
tion, was originally the name of some prominent Dasa tribe 
conquered by the Aryans. 

Of the stage of civilisation attained by the aborigines we learn 
little or nothing. They had, it is certain, large herds of cattle, and 
they could when attacked take refuge in fortifications called in 
the Rigveda by the name pur, which later denotes 'town,' but 
which may well have then meant no more than an earthwork 
strengthened by a pallisade or possibly occasionally by stone. 
Stockades of this kind are often made by primitive peoples, and 
are so easily constructed that we can understand the repeated 
references in the Rigveda to the large numbers of such fortifica- 
tions which were captured and destroyed by the Aryan hosts. 
Some Dasas, it seems, were able to establish friendly relations with 
the Aryans, for a singer celebrates the generosity of Balbutha, 
apparently a Dasa ; nor is it impossible, as we have seen, that the 
five tribes of the Punjab were not above accepting the cooperation 
of aboriginal tribes in their great attack on Sudas. We must 
therefore recognise that in the age of the Rigveda there was 
going on a steady process of amalgamation of the invaders and 
the aborigines, whether through the influence of intermarriage 
with slaves or through friendly and peaceful relations with power- 
ful Dasa tribes. 

Like the Dasas and Dasyus in their appearance both as terres- 
trial and as celestial foes are the Panis. Tlie word seems beyond 
doubt to be connected with the root seen in the Greek pernemi, 
and the sense in which it was used by the poets must have been 
something like 'niggard.' The demons are niggards because they 
withhold from the Aryan the water of the clouds : the aborigines 
are niggards because they refuse the gods their due, perhaps also 
because they do not surrender their wealth to the Aryan without 
a struggle. The term may also be applied to any foe as an 
opprobrious epithet, and there is no passage in the Samhita which 
will not yield an adequate meaning with one or other of these 
uses. But it has been deemed by one high authority^ to reveal to 

1 Hillebrandt, Vedische Mythologie, yol. i, pp. 94 sq. 

I v] India and Iran 8 7 

us a closer connexion of India and Iran than has yet suggested 
itself: in the Dasas Hillebrandt sees the Dahae, in the Panis the 
Parnians, and he locates the struggles of Divodasa against them in 
Arachosia. Support for this view he finds in the record of Divo- 
dasa's conflicts with Brisaya and the Paravatas, with whose names 
he compares that of the Satrap Barsentes and the people Paruetae 
of Gedrosia or Aria. Similarly he suggests that the Srifijaya 
people, who were connected like Divodasa with the Bharadaja 
family, should be located in Iran, and he finds in the Sarasvati, 
which formed the scene of Divodasa's exploits, not the Indian 
stream but the Iranian Harahvaiti. Thus the sixth book of the 
Rigveda would carry us far west from the scenes of the third and 
seventh which must definitely be located in India. But the 
hypothesis rests on too weak a foundation to be accepted as even 

Other references to connexions with Iran have been seen in 
two names found in the Rigveda. Abhyavartin Chayamana, whose 
victory over Vara(?ikha has already been recorded, bears the 
epithet Parthava, and the temptation to see in him a Parthian is 
naturally strong. But the Rigveda knows a Prithi and later texts 
a Prithu, an ancient and probably mythical king, and thus we 
have in the Vedic speech itself an explanation of Parthava which 
does not carry us to Iran. Still less convincing is the attempt to 
find in the word Par^u in three passages of the Rigveda a reference 
to Persians : Parcu occui*s indeed with Tirindira as a man's name, 
but the two are princes of the Yadus, and not a single personality, 
'Tiridates the Persian \' Whatever the causes which severed Iran 
and India, in the earliest period, at least as recorded in the Rig- 
veda, the relations of the two peoples seem not to have been those 
of direct contact. 

As little do the Rigvedic Indians appear to have been in 
contact Avith the Semitic peoples of Babylon. The term Bekanata 
which occurs along with Pani in one passage has been thought to 
be a reference to some Babylonian Avord: though the Indian 
Bikauer is much more plausible as its origin. Bribu, mentioned 
once as a most generous giver and apparently also as a Pani, has 
been connected by Weber ^ with Babylon, but without ground: 
more specious is the attempt to see a Babylonian origin for the 
word mand found in one passage only of the Rigveda where it is 

^ Iranian relations are accepted by Ladwig, Rigveda, vol. ni, pp. 196 sq, ; Weber, 
Episches im vedischen Ritual, pp. 36 sq. See also Chapter xiv, pp. 321 sq. 
2 Op. cit. pp. 28 sq. ; Indische Studien, vol. xvii, p. 198, 

88 The Age of the Rigveda [ch. 

accompanied by the epithet 'golden.' The Greek mina, presumably 
borrowed from the Phoenicians, is a plausible parallel; but the 
passage can be explained without recourse to this theory \ A 
Semitic origin has been claimed for the word para^u, *axe,' but 
this too is far from certain. There is nothing in the Rigvedic 
mythology or religion which demands derivation from a non-Aryan 
source, though it has been urged that the small group of the 
Adityas, whose physical characteristics are very faint and whose 
abstract nature is marked, is derived from a Semitic civilisation^. 
In the succeeding period the Nakshatras or lunar mansions may 
more probably be ascribed to a Semitic source ; but in the 
Rigveda the Nakshatras are practically unknown, appearing as 
such only in, the latest portions. It is therefore impossible to 
assume that the great Semitic civilisations had any real contact 
M'ith India in the Rigvedic age. 

Scanty as is our information regarding the Vedic tribes, yet we 
can see clearly that the social and political organisation rested 
upon the patriarchal family, if we may use that term to denote 
that relationship was counted through the father. The Aryan 
marriage of this period was usually monogamic, though polygamy 
was not unknown probably mainly among the princely class ; and 
in the household the husband was master, the wife mistress but 
dependent on and obedient to the master. The standard of female 
morality appears to have been fairly high, that of men as usual 
was less exigent. Polyandry is not shown by a single passage to 
have existed, and is not to be expected in a society so strongly 
dominated by the male as was the Vedic. Of limitations on 
marriage we learn practically nothing from the Rigveda, except 
that the wedlock of brother and sister and of father and daughter 
was not permitted. Child marriage, so usual in later times, was 
evidently unknown ; and much freedom of choice seems to have 
existed. Women lived under the protection of their fathers during 
the life of the latter, and then they fell if still unmarried under 
the care of their brothers. Both dowries and bride-prices are 
recorded : the ill-favoured son-in-law might have to purchase his 
bride by large gifts, while other maidens could obtain husbands 
only through the generosity of their brothers in dowering them. 
A girl without a protector ran grave risk of being reduced to 

1 For the borrowing see Zimmer, Altindisches Leben, pp. 50, 51 ; Weber, Indiscke 
Studien, vol. xvii, p. 202. Bohtlingk, Dictionary, s.v., recognises only ' desire ' or 
' wish ' as the sense. 

2 So Oldenberg, Religion des Veda, p. 193; Z.D.M.G., vol. l, pp. 43 sq. ; but see 
Bloomfield, Religion of the Veda, pp. 133 sq. 

iv] Family Relations 89 

immorality to maintain herself, and even in cases where no such 
excuse existed we learn of cases of moral laxity. But the high 
value placed on marriage is shown in the long and striking hymn 
(x, 85) which accompanied the ceremonial, the essence of which 
was the mutual taking of each other in wedlock by the bride and 
bridegroom, and the conveyance of the bride from the house of 
her father to that of her husband ^ In this hymn the wedlock of 
Soma, here identified with the moon, and Surya, the daughter of 
the sun, is made the prototype and exemplar of marriage in 
general. Moreover, the Vedic marriage was indissoluble by human 
action, nor in the early period does it seem to have been con- 
templated that remarriage should take place in the case of a 
widow ^. To this there was the exception, which appears clearly in 
the burial ritual of the Rigveda, that the brother-in-law of the 
dead man should marry the widow, probably only in cases where 
the dead had left no son and it was therefore imperative that steps 
should be taken to secure him offspring ; for the Rigveda recognises 
to the full the keen desire of the Vedic Indian for a child to per- 
form his funeral rites. 

The relation of child and parent was clearly as a rule one of 
close affection; for a father is regarded as the type of all that 
is good and kind. There are traces, however, that parental rights 
were large and vague: if the chastisement of a gambler by his 
father may be deemed to be legitimate exercise of parental 
control, this cannot be said of the cruel act of his father in 
blinding Rijracva at which the Rigveda hints. The father pro- 
bably controlled in some measure at least both son and daughter 
as regards marriage ; and the right of the father to adopt is clearly 
recognised by the Rigveda, though a hymn ascribed to the family 
of Vasishtha disapproves of the practice. The son after marriage 
must often have lived in the house and under the control of his 
father, of whom his wife was expected to stand in awe. But, on 
the other hand, as the father advanced in years it cannot have 
been possible for him to maintain a control which he was physi- 
cally incapable of exercising ; and so we find the bride enjoined to 
be mistress over her step-parents, doubtless in the case when her 
husband, grown to manhood, had taken over the management of 
the household from his father's failing hands. 

1 For the marriage ritual, see Weber and Haas, Indische Studien, vol. v, pp. 177-412 ; 
Winternitz, Das altindische Hochzeitsrituell (1892). 

^ See Delbriick, Die indogermanischen Verwandtschaftsnamen, pp. 563-5. Possibly 
remarriage was permitted in the case of a woman whose husband disappeared; see 
Fischel, Vedische Studien, vol. i, p. 27. 

90 The Age of the Rigveda [ch. 

The head of the family appears also to have been the owner of 
the property of the family ; but on this point we are reduced in 
the main to conjecture. It is certain that the Rigveda recognises 
to the full individual ownership of movable things, cattle, horses, 
gold, ornaments, weapons, slaves, and so forth. It seems also 
certain that land was already owned by individuals or families: 
the term hshetray 'field,' is unmistakably employed in this sense, 
and in one hymn a maiden, Apala, places her father's cultivated 
field (urvard) on the same level with his hair as a personal pos- 
session. Reference is also made to the measuring of fields, and to 
hkilya, which appear to have been strips of land between the 
cultivated plots, probably used by the owners of the plots in 
common. The Rigveda has no conclusive evidence that the sons 
were supposed to have any share whatever in the land of the 
family, and the presumption is that it was vested in the father 
alone, as long as he was head of the family and exercised his full 
powers as head. We are left also to conjecture as to whether the 
various plots were held in perpetuity by the head of the family 
and his descendants, or whether there were periodic redistribu- 
tions, and as to the conditions on which, if there were several sons, 
they could obtain the new allotments necessary to support them- 
selves and their families. But there can hardly have been much 
difiiculty in obtaining fresh land; for it is clear that population 
was scanty and spread over wide areas, and wealth doubtless 
consisted in the main in flocks and herds. 

There is no hint in the Rigveda of the size to which a family 
might grow and yet keep together. It is clear that there might 
be three generations under the same roof, and a family might 
thus be of considerable dimensions. But life can hardly have been 
long — so much stress is laid on longevity as a great boon that it 
must have been rare — and, even if we decline to accept the view 
that exposure of aged parents was normal, there must have been a 
tendency for the family to break up as soon as the parent died, 
especially if, as is probable, there was no such land hunger as to 
compel the sons to stay together. The sons would, however, 
naturally enough stay in the vicinity of one another for mutual 
support and assistance. The little knot of houses of the several 
branches of the family would together form the nucleus of the 
second stage in Rigvedic society, the grama, 'village,' though 
some have derived its name originally from the sense ' horde ^' 

1 See Zimmer, AUindisches Leben, pp. 159, 160 ; Feist, Kultur, Ausbreitung, und 
Herkunft der Indogermanen, p. 143. 

iv] Social Groups 91 

as describing the armed force of the tribe which in war fought 
in the natural divisions of family and family. Next in order 
above the grama in the orthodox theory was the vi^ or 'canton/ 
while a group of cantons made up the jana, 'people.' This 
scheme can be supported by apparent analogies not only from 
Greece, Italy, Germany, and Russia, but also from the Iranian state 
with the graduated hierarchy of family or households, vis, %antu, 
and dahyu\ But for Vedic India the fourfold gradation cannot 
successfully be maintained. It is not merely that the various 
terms are used with distressing vagueness — so that for example 
the Bharatas can be called at one time a jana and at another a 
grama — but that the evidence for the relationship of subordina- 
tion between the grama and the vi^ is totally wanting. More- 
over the Iranian evidence tells against the theory that the vi^ is 
removed by the grama from the family in the narrower sense: 
the more legitimate interpretation is to see in the Iranian division 
a step further than that of the Rigveda and to set the jana as 
parallel to the zantu, acknowledging that in the time of the 
Rigveda the political organisation of the people had not extended 
to the creation of aggi'egates of janas, unless such an aggregate is 
presented to us in the twenty-one janas of the two Vaikarnas who 
are mentioned in one passage of the Samhita. The vi^ will thus 
take its place beside the Iranian vis as a clan as opposed to family 
in the narrower sense, and be a real parallel to the Latin getw, 
and the Greek genos. It is possible that the grama is originally 
the gens in its military aspect, but even that is not certain, for the 
word may originally have referred to locality. Nor can we say 
with any certainty for the period of the Rigveda whether the 
grama contained the whole of a vi^, or part of a vi^, or parts 
of several vi^as. But amid much that is conjectural it is clear 
that the vi^ was not a normal unit for purposes of government, 
for the term vi^ati, 'lord of a v^p,' has not in any passage the 
technical sense of 'lord of a canton.' On the other hand, the 
grama as a unit is recognised by the use of the term gramanl, 
'leader of a village,' an officer who appears in the Rigveda, and 
who was probably itivested with both military and civil functions, 
though we have no details of his duties or powers. 

While the sense 'clan' is comparatively rare, the word vi^ not 
unfrequently in the plural denotes 'subjects': so we hear of the 
vi^ of Trinaskanda, a king elsewhere unknown, and of the vi^as 
of the Tritsus, the royal family of which Sudas was a member. In 

1 Zimmer, I. c. ; Geiger, Ostiranische Kultur, p. 427. 

92 The Age of the Rigveda [ch. 

the former case the sense 'clans' is obviously inappropriate, while in 
the latter the rendering 'clans' which was long adopted has resulted 
in the confusion of the relations of the Bharatas and the Tritsus, 
the Tritsus being regarded as a people opposed to the Bharatas, 
instead of taking their place as the rulers of the Bharatas. The 
subjects as a whole made up the jana, a term which in Vedic use 
denotes either the individual man or the collective manhood of the 
tribe as a political unit. Above that unit no political organisation 
can be shown to have existed. The confederacy of the five tribes 
by whom Sudas was attacked was evidently more than a mere 
passing episode, but clearly it did not involve any system of 
political subordination, from which a great kingdom could emerge. 
There was however beyond that a feeling of kinship among all the 
tribes who called themselves Aryan, stimulated no doubt into 
distinct expression by their presence in the midst of the dark 
aboriginal population. 

The question now presents itself as to the extent to which in 
the period of the Rigveda the caste system had been developed. 
The existence of the caste system in any form in the age of the 
Rigveda has been denied by high authority^, though it has been 
asserted of late with increasing insistence^. In one sense, indeed, 
its presence in the Rigveda cannot be disputed. In the Purusha- 
sukta the four castes of the later texts, Brahmana ('priest'), 
Rajanya ('prince' or more broadly 'warrior'), Vai^ya ('commoner'), 
and (^udra are mentioned. But this hymn is admittedly late and 
can prove nothing for the state of affairs prevailing when the bulk 
of the Rigveda was composed. On the other hand, as we have 
seen, the distinction between the Aryan colour (varna) and that of 
the aborigines is essential and forms a basis of caste. The question 
is thus narrowed down to the consideration of the arguments for 
and against the view that in the Aryans themselves caste divisions 
were appearing. On the one hand, it is argued that in the period 
of Vasishtha and Vi9vamitra, when the great poetry of the Rigveda 
was being produced, neither the priestly class nor the warrior 
class was hereditary. The warriors of the community were the 
agricultural and industrial classes, and the priesthood was not yet 
hereditary. It has been held that the Brahman priest was not 
necessarily the member of an hereditary class at all, that the term 

1 Muir, Original Sanskrit Texts, vol. i^, pp. 239 sq. ; Zimmer, Altindisches Leben, 
pp. 185-203 ; Weber, Indische Studien, vol. x, pp. 1 sq. 

2 See Geldner, Vedische Studien, vol. ii, p. 146; Oldenberg, Z.D.M.G., vol. li, 
pp. 267 sq. 

iv] Origins of the Caste System 93 

could be applied as well to any person who was distinguished by 
genius or virtue, or who for some reason was deemed specially 
receptive of the divine inspiration. The growth of the caste 
system is traced on this hypothesis to the complication of life 
ensuing on the further jienetration of the Aryans fi'om the Punjab 
towards the east. The petty tribes found it necessary, in order to 
defeat the solid forces of the aborigines, to mass themselves into 
centralised kingdoms. The petty tribal princes thus lost their full 
royal rank, but found employment and profit instead in becoming 
a standing armed force, ready to resist sudden incursion or to 
crush the attempts at rebellion of the defeated aborigines. On the 
other hand, the industrial and agricultural population, relying on 
the protection of the warrior class, abandoned the use of arms. 
Together with the growth in the size of the kingdom and the 
increasing complexity of civilisation, the simple ritual of an earlier 
period, when the king hunself could sacrifice for his people, grew 
to an extent which rendered this impracticable, while at the same 
time an ever increasing importance came to be attached to the 
feiithful and exact performance of the rites and the preservation of 
the traditional formulae. The result of this process was, it is 
suggested, the growth of a priesthood, of a warrior class, and of a 
third class, the Vai9ya, sharply distinguished from one another and 
strictly hereditary. But the comparatively late date of this de- 
velopment is shown by the fact that in later times the inhabitants of 
the North- West, the home of the Rigveda, were regarded as semi- 
barbarians by those of the Middle Country, in which the Brahmanical 
civilisation had developed itself, on the ground that they did not 
foUow the strict caste system. 

While there is much of truth in this view, it must be admitted 
that it exaggerates the freedom of the Rigveda from caste. As we 
have seen, the probabilities are that the main, though not the 
earliest, part of the Samhita had its origin not in the Punjab proper 
but in the sacred country of later Brahmanism, the land known in 
the Samhitas of the succeeding period as Brahma varta. Moreover, 
there is no actual proof in the Rigveda that the priesthood was 
not then a closed hereditary class. The term Brahmana, ' son of a 
Brahma,' seems, on the contrary, to show that the priesthood was 
normally hereditary, and there is no instance which can be quoted 
of any person who is said to be other than a priest appearing to 
exercise priestly functions. We are told that there is a case of a 
king exercising the functions of domestic priest and sacrificing 
himself for his people, but the alleged case, that of Devapi, rests 

94 The Age of the Rigveda [ch. 

only on an assertion of a commentator on the hymn (x, 98), in 
which Devapi appears, that he was originally a king. Even, how- 
ever, if this were the case, it must be remembered that even after the 
complete establishment of the caste system, it was still the privilege 
of kings to exercise some priestly functions, such as that of the study 
of the nature of the absolute, a practice ascribed to them in the 
Upanishads. The arguments regarding the warrior class rest on a 
misunderstanding. Even in the latest Vedic period we have no 
ground to suppose that there was a special class which reserved its 
energies for war alone, and that the industrial population and the 
agriculturists allowed the fate of their tribe to be decided by con- 
test between warrior bands, but the Rigveda certainly knows of a 
ruling class, the Kshatriya, and the Vedic kingship was normally 
hereditary, so that we may well believe that even then there 
existed, though perhaps only in embryo, a class of nobles, who are 
aptly named in the term of the PurusJiasiikta, Raj any as, as being 
* men of kingly family.' There are traces, moreover, of the division 
of the tribe into the holy power (hrahman), the kingly power 
(kshatra), and the commonalty (vi0, and, while it is true that the 
caste system is only in process of development in the Rigveda, it 
seems impossible to deny that much of the groundwork upon which 
the later elaborate structure was based was already in existence. 

So far, our sources of knowledge, if imperfect, have given us 
material sufficient to sketch the main outlines of Vedic society. 
Unhappily, when we turn to consider more closely the details of 
the political organisation proper, the evidence becomes painfully 
scanty and inadequate. The tribes of the Rigveda were certainly 
under kingly rule: there is no passage in the Rigveda which 
suggests any other form of government, while the king under the 
style * Raj an' is a frequent figure. This is only what might be 
expected in a community which was not merely patriarchal — a fact 
whence the king* drew his occasional style of vi^ati, * Head of the 
vi^^ — but also engaged in constant warfare against both Aryan 
and aboriginal foes. Moreover, the kingship was normally here- 
ditary: even in the scanty notices of the Rigveda we can trace 
lines of succession such as that of Vadhrya^va, Divodasa, Pijavana, 
and Sudas, or Durgaha, Girikshit, Purukutsa, and Trasadasyu, or 
Mitratithi, Kunicravana, and Upamagravas. In some cases it has 
been argued that election by the cantons was possible^; but this 

* See Zimmer, Altindisches Leben, pp. 162 sq. ; Weber, Indische Studien, vol. xvn, 
p. 189 ; Bloomfield, Sacred Books of the East, vol. xlii, p. 330. That sometimes election 
actually took place is quite probable ; but the passages quoted do not show this ; see 
Qeldner, Vedische Studien, vol. ii, p. 303. 

iv] Political Orga7iisation 95 

interpretation rests only on the improbable view that m(^h 
denotes not 'subjects' but 'cantons'; and the idea has no support 
in later literature. The activity of the sovereign on which most 
stress is laid is his duty of protecting his subjects ; and even the 
Rigveda, despite its sacerdotal character, allows us to catch some 
glimpses of the warlike deeds of such men as Divodasa, Sudas, and 
Trasadasyu. Of the king's functions in peace the Rigveda is silent, 
beyond showing that he was expected to maintain a large body of 
priests to perform the sacrifices for him and his people. From his 
subjects he was marked out by his glittering apparel, his palace, 
and his retinue, which doubtless included the princes of the royal 
house as well as mere retainers. To maintain his state he had the 
tribute paid by conquered tribes and the gifts of his people, which, 
once proffered freely, had doubtless become fixed payments, which 
the king could exact, if denied. Doubtless, too, when lands were 
conquered from the aborigines or from other Aryan tribes, large 
booty in land and slaves and cattle would be meted out to the 
king ; but the Rigveda contains no hint that he was considered as 
owner of the land of the people. Nor in that Samhita is there any 
trace that the king has developed from the priest: if that was 
the case in India the distinction lies far beyond the period of the 

Of the entourage of the king and his servants we learn almost 
nothing. The sencini, ' leader of the army,' who appears in a few 
hymns, may have been a general appointed by the king to lead an 
expedition of too little importance to require his own intervention. 
The grdmanl probably led in war a minor portion of the host and 
was identical with the vrdjapati mentioned elsewhere. Far more 
important, in the estimation at least of the composers of the hymns, 
was the purohita or domestic priest, whose position represented 
the height of a priest's ambition. Nor, after allowing for priestly 
partiality and exaggeration, can we deny the importance of the 
Purohita amongst a people who followed the guiding in religious 
matters of an hereditary priesthood. The Vedic Purohita was the 
forerunner of the Brahman statesmen who from time to time in 
India have shown conspicuous ability in the management of affairs ; 
and there is no reason to doubt that a Vi^vamitra or a Vasishtha 
was a most important element of the government of the early 
Vedic realm. It is clear, too, from the hymns which are attributed 
to the families of these sages, that the Purohita accompanied the 
king to battle, and seconded his efforts for victory by his prayei-s 
and spells. In return for his faithful service the rewards of the 

96 The Age of the Rigveda [ch. 

Purohita were doubtless large : the dmiastutis of the Rigveda tell 
of the generous gifts of patrons to the poets, and we may safely 
assume that the largest donations were those of kings to the 
Purohitas. It is significant of the social arrangements of the time 
that the gifts enumerated are all gifts of personal property ; land 
was evidently not then a normal form of gift, though we may 
conjecture that, even at this early period, the king might confer on 
a priest or other servant the right to receive some portion of the 
gifts in kind which were clearly no inconsiderable part of the royal 

The power of the king cannot have been in normal circum- 
stances arbitrary or probably very great. There stood beside him 
as the mode of expression of the will of the people the assembly, 
wliich is denoted by the terms smniti and sahhd in the Samhita. 
It has been proposed by Ludwig^ to see in these two terms the 
designations of two different forms of assembly : the one would be 
the assembly of the whole people, while the other would be an 
analogue of the Homeric council of elders, a select body to which 
the great men of the tribe, the Maghavans, alone would go to take 
counsel with the king. Zimmer^, on the other hand, sees in the 
samiti the popular assembly of the tribe, in the sahhd the assembly 
of the village. But neither view appears to be acceptable. There 
is no distinction in the texts which would justify us in contrasting 
sahhd and samiti in either of the ways suggested : rather it seems 
the samiti is the assembly of the people for the business of the 
tribe, the sabhd particularly the place of assembly, which served 
besides as a centre of social gatherings. The king's presence in 
the samiti is clearly referred to; and there seems no reason to 
doubt that on great occasions the whole of the men of the tribe 
gathered there to deliberate, or at least to decide, on the courses 
laid before them by the great men of the tribe. But we are 
reduced to analogy with the Homeric assembly for any conception 
of the action of the assembly ; for, perhaps owing to the nature of 
the sources, nothing is known of its part in Vedic life. If indeed 
the king was ever elected by the cantons, the election took place 
in the samiti ; but the theory that the king was ever elected has, 
as has been already said, nothing to support it. 

In accordance with the apparently undeveloped condition of 
political organisation, we learn little of the administration of 
justice. That the king exercised criminal and civil jurisdiction, 
assisted by assessors, is a conclusion which must rest for its 

^ Rigveda, vol. iii, p. 253. ^ Altindisches Leben, pp. 172-4. 

iv] Administration of yustice 97 

plausibility on analogy and on the later practice in India ; for no 
passage in the Rigveda definitely alludes to the sovereign as acting 
in either capacity. It is therefore at least probable that his 
functions as judge were still confined within narrow limits. One 
word in the Rigveda shows that the system of wergeld was in full 
force, a man being given the epithet i^ataddya, which denotes 
that the price of his blood was a hundred cows. In one hymn the 
Pani, whose niggardliness made him the chief object of dislike to 
the greedy Vedic poets, is declared to be a man only in so far as 
he has a wergeld, here called vairadeya, ' that which is to be paid 
in respect of enmity.' The crime, however, of which most is re- 
corded in the Rigveda is that of theft, including burglary, house- 
breaking, and highway robbery, crimes which clearly must have 
been of frequent occurrence. The punishment of the thief seems 
to have rested with the person wronged : there are clear allusions 
to binding the thief in stocks, presumably with a view to induce 
his relatives to pay back to the aggrieved man the loss he has 
sustained. In one passage of the Rigveda there is a probable 
reference to the employment of trained men to recover stolen 
cattle, just as the Khojis of the Punjab down to modem times were 
expert at this difficult employment. Of death as a punishment for 
theft, as in later times and in other primitive societies, curiously 
enough nothing appears in the Rigveda. 

There is hardly any mention of other forms of crime in the 
Rigveda. It appears clear that marriage of brother and sister was 
regarded as incest, and apparently marriage of father and daughter 
was placed in the same category of wrongful actions, as it certainly 
was in the later Samhitas, where the union of Prajapati, one aspect 
of the supreme god, with his daughter is at once punished by the 
other gods. Prostitution was certainly not unknown, but in other 
respects morality seems to have been fairly high: there is no 
sufficient ground for attributing to the peoples whose actions are 
reflected in the Rigveda either the exposure of the aged or the 
putting away of female children. 

Our knowledge of civil law is as scanty as that of criminal law. 
As we have seen, land seems not to have been an article of 
commerce. Movable property could change hands by gift or by 
sale, the latter taking the form of barter. The Rigveda records 
that in the opinion of one poet not ten cows was adequate price for 
an image of Indra to be used doubtless as a fetish. The haggling 
of the market is once clearly referred to. The standard of value 
seems to have been the cow, and no coin appears to have been 

C.H.I. I. 7 

9 8 'The Age of the Rigveda [ch. 

known, though the origin of currency may be seen in the frequent 
references to nishkas as gifts: the nishka most probably was an 
ornament in the shape of a necklace of gold or silver : at a later 
date the name was transferred to a gold coin. Property doubtless 
passed by inheritance and could be acquired originally by a man's 
own efforts in creation or discovery, while the dowry and the price 
of the bride played a considerable part in early Vedic economy, as 
is seen by the stress laid upon both in the Samhita. Of forms of 
contract the only one of which we know anything was the loan, 
rina. The Vedic Indian was an inveterate gambler, and for that 
among other causes he seems always to have been ready to incur 
debt. The rate of interest is unknown, a reference to payments of 
an eighth or a sixteenth may be referred either to interest or 
instalments of principal. At any rate, the debtor might as a result 
be reduced to slavery, as we learn from an interesting hymn (x, 34) 
where an unsuccessful dicer recites the fatal fascination for him of 
the dice and his consequent ruin and enslavement with its results 
for his family. Of civil procedure we know only so much as may 
be inferred from a single word, madhyama^i, which may denote 
one who intervenes between two parties as an arbitrator, though it 
has also been referred to the king as surrounded by his retainers 
in his camp. 

In war the Vedic host was led by the king ; and doubtless at 
this time all the men of the tribe took part in it, encouraged by 
the priests, who with prayer and incantation sought to secure 
victory for those whom they supported The king and the nobles, 
the Kshatriyas, fought from chariots of simple construction, the 
warrior standing on the left hand of the charioteer on whose skill 
he so largely depended. The common people fought on foot, doubt- 
less with little attempt at ordered fighting, if we may judge from 
analogy and from the confused battles described in the later epics. 
The chief weapon in honour was the bow which was drawn to the 
ear and not as in Greece to the breast ; but lances, spears, swords, 
axes, and slingstones seem to have been employed. The warrior, 
when completely equipped, wore coat of mail and helmet, and 
a hand or arm guard to save his arm from the friction of the bow- 
string. The arrow had a reed shaft, and the tip was either of horn 
or of metal: poisoned arrows were sometimes used. Though 
horse riding was probably not unknown for other purposes, no 
mention is made of this use of the horse in war. Naturally enough 
the banks of rivers seem to have been frequently the spots chosen 
for the conflict, as in the case of the famous battle of the ten kings. 

iv] Warlike and Peaceful Avocations 99 

All the evidence points to the absence of city life among the 
tribes. The village probably consisted of a certain number of 
houses built near each other for purposes of mutual defence, 
perhaps surrounded by a hedge or other protection against wild 
beasts or enemies. The 'pur, which is often referred to and which 
in later days denotes a ' town,' was, as we have seen, probably no 
more than a mere earthwork fortification which may in some cases 
at least have been part of the village. In certain passages these 
puras are called autumnal, and by far the most probable explana- 
tion of this epithet is that it refers to the flooding of the plains 
by the rising of the rivers in the autumn, when the cultivators 
and herdsmen had to take refuge within the earthworks which 
at other times served as defences against human foes. Of the 
construction of the Vedic house we learn little, but the bamboo 
seems to have been largely used for the beams which borrowed 
their name from it. In the midst of each house burned the 
domestic fire, which served the Indian both for practical and 
sacrificial uses. 

Like the aborigines, the Vedic Indians were primarily pastoral : 
the stress laid by the poets on the possession of cows is almost 
pathetic. The name of the sacrificial fee, dakshind, is explained 
as referring originally to a cow placed ' on^,the right hand ' of the 
singer for his reward. Tlie singers delight to compare their songs 
to Indra with the lowing of cows to their calves. At night and in 
the heat of the day the cows seem to have been kept in the fold, 
while for the rest of the day they were allowed to wander at wiU, 
being thrice milked^. Bulls and oxen on the other hand regularly 
served for ploughing and drawing carts, a purpose for which horses 
were not much used. Second to cattle came hoi-ses, which the 
Indian required both for bearing his chariot into the battle and for 
the horse-race, one of his favouri'je sports. Other domesticated 
animals were sheep, goats, assea, and dogs, the last being used for 
hunting, for guarding and tracking cattle, and for keeping watch at 
night. On the other hand, the cat had not been domesticated. 

Agriculture was already an important part of the Vedic economy. 
The practice of ploughing was certainly Indo-Iranian as the same 
root (Jcrish) occurs in the same sense in the two tongues. But it 
is clear that even in the Rigveda the use of the plough was in- 
creasing in fi'equency. We learn of the use of bulls to draw the 
plough, of the sowing of seed in the furrows thus made, of the 
cutting of the corn with the sickle, the laying of it in bundles 

^ See Geldner, Vedische Studien, vol. ii, pp. 282 sq. 


lOO The Age of the Rigveda [cH. 

on the threshing floor, and the threshing and final sifting by 
winnowing. Moreover, the use of irrigation seems to be recognised 
in the mention of channels into which water is led. On the other 
hand, the nature of the grain grown is uncertain : it is called yava, 
which in the later Samhitas is barley, but it is quite uncertain 
whether this definite sense can be assigned to the word in the 
Rigvedic period. 

Hunting seems still to have played a considerable part in the 
life of the day. The hunter used both bow and arrow and snares 
and traps. There are clear references to the capture of lions in 
snares, the taking of antelopes in pits, and the hunting of the boar 
with dogs. Birds were captured in nets stretched out on pegs. 
Possibly the use of tame elephants to capture other elephants was 
known, but this is very uncertain, for there is no clear proof that 
the elephant had yet been tamed at this early date. Buffaloes 
seem to have been shot by arrows, and occasionally a lion might be 
surrounded by hunters and shot to death. 

There is some evidence that already in this period specialisation 
in industry had begun. The worker in wood has clearly the place 
of honour, needed as he was to produce the chariots for war and 
the race, and the carts for agricultural purposes. He was car- 
penter, joiner, wheelwr ght in one ; and the fashioning of chariots 
is a frequent source of metaphor, the poet comparing his own 
skill to that of the wheelwright. Next in importance was the 
worker in metal who smelted ore in the furnace, using the wing 
of a bird in the place of a bellows to fan the flame. Kettles and 
other domestic utensils were made of metal. It is, however, still 
uncertain what that metal which is called ayas was. Copper, 
bronze, and iron alike may have been meant, and we cannot be 
certain that the term has tl e same sense throughout. Of other 
workers the tanner's art is all^^ded to not rarely ; and to women 
are ascribed sewing, the plaiting of mats from grass or reeds, and, 
much more frequently, the weaving of cloth. It is of importance 
to note that there is no sign that those who carried on these 
functions were in any way regarded as inferior members of the 
community, as was the case in later times. This fact is probably 
to be explained by the growing number of the servile population 
which must have steadily increased with the conquest of the 
tribes, though we cannot conjecture the motives which ascribed 
to inferiors tasks which in the Rigvedic time were apparently 
honourable and distinguished. Presumably even at this time the 
slave population must have been utilised in assisting their masters 

iv] Industries and Domestic Life loi 

in their various tasks, agricultural, industrial, and pastoral ; but 
the Rigveda unquestionably presents us with a society which is 
not dependent on such labour, and in which the ordinary tasks 
of life are carried out by the free men of the tribe \ This is 
one of the facts which show the comparative simplicity of the age 
of the Rigveda as compared with the next period of Indian history. 

Fishing is not directly mentioned ; and the Vedic Indian 
seems to have been very little of a navigator. The use of boats, 
probably dug-outs, for crossing rivers, was known, but the sim- 
plicity of their constioiction is adequately shown by the fact that 
the paddle alone was used for their propulsion. There is no 
mention of rudder or anchor, mast or sails, a fact which inci- 
dentally negatives the theory that the Vedic Indians took any part 
in ocean shipping. 

Of the domestic life of the time we have a few details. The 
di'ess usually worn consisted either of three or of two gamients. 
These were generally woven from the wool of sheep, though skins 
were also employed. Luxury manifested itself in the wearing of 
variegated garments or clothing adorned with gold. Ornaments in 
the shape of necklets, earrings, anklets, and bracelets were worn by 
both sexes and were usually made of gold. The hair was carefully 
combed and oiled. Women wore it plaited, while in some cases men 
wore it in coils : it was a characteristic of the Vasishthas to have 
it coiled on the right. Shaving was not unknown, but beards were 
nonnally worn, and on festive occasions men bore garlands. 

As was natui*al with a pastoral people, milk formed a consider-? 
able part of the ordinary food, being taken in its natural state or 
mixed with grain. Ghee or clarified butter was also much used. 
Grain was either parched or ground into flour, and mixed with 
milk or butter, and made into cakes. As throughout the history 
of India, vegetables and fruits formed a considerable portion of 
the dietary. But the Vedic Indians were a nation of meat-eaters, 
nor need we believe that they merely ate meat on occasions of 
sacrifice. Rather, as in the Homeric age, the slaughter of oxen 
was always in some degree a sacrificial act, and one specially 
appropriate for the entertainment of guests, as the second name 
of the heroic Divodasa Atithigva, ' the slayer of oxen for guests,' 
and as the practice of slaying oxen at the wedding festival 
abundantly show. The ox, the sheep, and the goat were the 

^ The view of Indian civilisation presented by Baden Powell {Indian Village 
Community (1896) and Village Communities in India (1899), etc.) which assumes that 
the Aryans were princely conquerors of agricultural aborigines and not themselves 
cultivators cannot be reconciled with the Kigveda. 

I02 The Age of the Rigveda [CH. 

normal food eaten by men and offered to their gods: hoi'se-flesh 
was probably eaten only at the horse-sacrifice, and not so much as 
ordinary food as with a view to gain the strength and swiftness 
of the steed. There is no inconsistency between this eating of 
flesh and the growing sanctity of the cow, which bears already 
in the Rigveda the epithet aghnyd, 'not to be killed.' If this 
interpretation of the temi is correct, it is merely a proof of the 
high value attached to that useful animal, the source of the milk 
which meant so much both for secular and sacred use to the Vedic 
Indian. The flesh eaten was either cooked in pots of metal or 
earthenware or roasted on spits. 

In addition to milk, the Indians had at least two intoxicating 
drinks. The first was the Soma, which however, by the time of 
the Rigveda, appears almost exclusively as a sacrificial drink. It 
stands, however, to reason that the extraordinary preeminence 
which it acquired for religious purposes can hardly have been 
attained except through its original popular character ; and it 
is difficult to resist the impression that the Soma was at first 
a popular drink in the home whence the Vedic Indians entered 
India, and that in India itself they found no plant which pre- 
cisely coincided with that whence the Soma had first been 
produced, and so were compelled to resort to substitutes or to 
use the original plant after it had been brought fi-om a great 
distance and had thus lost its original flavour. The popular 
drink was evidently the sura, which seems to have been distilled 
from gi'ain. It was clearly extremely intoxicating, and the 
priests regarded it with disapproval : in one hymn mention 
is made of men made arrogant by the surd reviling the 
gods, while another couples it with anger and dicing as the 
cause of sin. 

Of the amusements of the Indian first place must clearly be given 
to the chariot race, a natural form of sport among a horse-loving and 
chivalrous people. The second belongs to dicing, which fonns the 
occasion of a lament, already referred to {v. sup. p. 98). Unhappily, 
the details of the play are nowhere described, and the scattered 
allusions cannot be reduced to a whole without much conjecture ; 
but, in one form at least, the aim of the gambler was to throw 
a number which should be a multiple of four\ Dancing was also 
practised, and the dancing of maidens is several times mentioned ; 
it seems that men also on occasion danced in the open air, as a 

^ See Liiders, Das Wilrfelspiel im alien Indien; Caland, Z.D.M.O., vol. lxii, 
pp. 123 sq.; Keith, J.R.A.S., 1908, pp. 823 sq. 

iv] Pastimes: Religion 103 

metaphor alludes to the dust of the dancing feet of men. Music too 
had advanced beyond the primitive stage ; and already the three 
types of instrument, percussion, string, and wind, were represented 
by the drum, used, among other purposes, to terrify the foe in battle, 
the lute, and the flute, the last-named instrument being said to 
be heard in the abode of Yama, where the holy dead dwell. The 
hymns themselves prove that singing was highly esteemed. 

The comparative simplicity of the life of the Vedic Indian 
stands in striking contrast to the elaboration of the religious side 
of life by the priests. The Rigveda does not present us with any 
naive outpouring of the primitive religious consciousness, but 
with a state of belief which must have been the product of 
much priestly efibrt, and the outcome of wholesale syncretism. 
Nothing else can explain the comparative magnitude of the Vedic 
pantheon, which considerably exceeds that of the Homeric poems. 
In the main, the religion revealed to us is in essence simple. The 
objects of the devotion of the priests were the great phenomena 
of nature, conceived as alive, and usually represented in anthro- 
pomorphic shape, though not rarely theriomorphism is referred 
to. The chief gods include Dyaus, the sky, who is usually coupled 
with Prithivi, the earth, and whose anthropomorphism is faint, 
being in the main confined to the conception of him as father. 
Varuna, the sky-god par excellence, has superseded Dyaus as a 
ix)pular figure, and has acquired moreover a moral elevation, 
which places him far above the other gods. Varuna is the sub- 
ject of the most exalted hymns of the Rigveda ; but it seems 
clear that in this period his claim to divine preeminence was 
being successfully challenged by the much less ethical Indra, the 
god of the thunder-storm which causes the rain to pour, when 
the rainy season long hoped for comes to relieve the parched 
earth. Varuna bears the epithet Asura, which serves to show 
his parallelism with Ahura Mazda, the highest of Iranian gods ; 
nor can there be any reason to doubt that in the Indo-Iranian 
period he acquired his moral elevation and preeminence. But 
in India it seems that his star paled before that of Indra, whose 
importance gi'ew with the advance of the Aryan tribes to the 
regions where the rain was confined in the main to the rainy 
months and the terrors of the storm supplanted in the popular 
imagination the majestic splendour of the sky. With Varuna 
seems to have been bound up in the first instance the concep- 
tion of rita as first cosmic and then moral order, and with his 
lessening glory these conceptions fade from Indian thought. The 

I04 The Age of the Rigveda [ch. 

importance of the sun is shown by the fact that no less than five 
higli gods seem to be solar — Surya and Savitri, who represent the 
quickening power of the luminary, Mitra, whose fame in Iran is 
but palely reflected in India, where he is conjoined with Varuna 
and eclipsed by Varuna's glories, Piishan, the representative of 
the power of the sun in its effect on the growth of herds and 
vegetation, and Vishnu, the personification of the swift moving 
sun and a god destined to become one of the two great gods 
of India, ^iva, his great rival in later days, appears in the name 
of Rudra, seemingly in essence at this time a storm-god, with a 
dark side to his character presaging his terrible aspect in later 
days. Other gods are the Acvins, apparently the morning and 
evening stars, who are clearly parallel to the Dioscuri, the Maruts, 
storm-gods and attendants on Rudra, Vayu and Vata, the wind- 
gods, Parjanya, the god of rain, the Waters, and the Rivers. Ushas 
the Dawn, deserves separate mention, since she has evoked some 
of the most beautiful of Vedic poetry ; but her figure seems to 
belong to the earliest period of Vedic hymnology, when the 
Indians were still in the Punjab ; and after the Rigveda she 
vanishes swiftly from the living gods of the pantheon. 

Next to Indra in importance rank Agni, 'the fire,' and the 
Soma. To the priest indeed there can be little doubt that these 
gods were of even greater importance than Indi'a, but the latter 
was seemingly more of a national god, and more nearly alive in 
the hearts of the people. Agni has three forms, the sun in the 
heaven, the lightning, and the terrestrial fire ; and liis descent 
from his highest form is variously pictured. He seems in his 
growth to have vanquished older gods, like Trita and Apam 
Napat, * the child of the waters,' who were forms of the lightning, 
and Matari9van, a form of celestial fire. The Soma must have 
owed its original divine rank to its wonderful intoxicating power ; 
but priestly speculation by the end of the Rigvedic period had 
succeeded in identifying the Soma and the moon, a tour de force 
which can indeed be rendered less unnatural by recognising the 
potent effect of the moon in the popular imagination on vege- 
tation, but which is none the less remarkable in the success in 
which it finally imposed itself on the religious conscience. The 
Soma hymns are among the most mystical of the Rigveda ; and 
one of the legends, that of the bringing of the Soma fi-om heaven 
by the eagle, appears to be a reflection of the fall of rain to earth 
as a result of the lightning which rends the cloud just when the 
rain begins to fall. 

iv] Deities 105 

The creation of what may be called abstract deities is not far 
advanced in the Rigveda, such deities as (^^raddha, 'faith/ and 
Manyu, 'wrath,' being confined to a few hymns of the tenth 
book. On the other hand, the specialisation of epithets in some 
cases results in the production of what is practically a new figure : 
thus Prajapati, an epithet of such gods as Savitri and Soma, as 
'lord of creatures' approaches the position of a creator. The 
Adityas and their mother Aditi, who may be derived from them, 
present scarcely any physical features and, as we have seen, have 
therefore by Oldenberg been assigned to a Semitic source; but 
this hypothesis has not yet been rendered probable in a mythology 
which else seems so little touched by external influence. Personifica- 
tions like Ratri, ' the night,' are mainly poetic rathei' than religious. 

A characteristic of the Vedic theology is the tendency to group 
gods in pairs, especially Mitra and Varuna, a practice due in all 
probability to the natural union of heaven and earth as a pair. 
Of larger groups there are the Maruts, the Adityas, and the 
Vasus. The last are associated vaguely with Indra or Agni, and 
have practically no individual character. Finally, priestly specu- 
lation has created the class of the Vi?ve devas, ' the All-gods,' who 
first include all the gods, and, in the second place, are regarded as 
a special group invoked with others, like the Adityas and the Vasus. 

Little part is played by minor deities in the Vedic theolog}% 
The predominance of the male element is marked : the goddesses 
are pale reflections of their husbands by whose names, with a 
feminine affix added, they are called: the only one who has 
a real character is Ushas, and more faintly Prithivl, 'the earth,' 
and of rivers the sacred Sarasvatl. The Ribhus are aerial elfs, 
the Apsarasas water nymphs, and the Gandharvas, their play- 
mates, are aerial sprites. The simpler and more primitive side 
of nature worship is seen in the invocation of the plants, of the 
mountains, and of the trees of the forest; but real as these 
beliefs may have been to the common people, they are not the 
true subjects of the priests' devotion. When speculation turned 
to deal with these matters, it found an utterance such as is seen 
in a striking hymn to the goddess of the forest, which exhibits 
much more poetical than religious feeling. 

While the great gods might be conceived at times in animal 
form, for example Indra or Dyaus as a bull, or the sun as a 
swift horse, actual direct worship of animals is hardly found in 
the Rigveda. The drought demon which xjrevents the rain from 
falling is conceived as a snake whom Indra crushes, and we hear 

io6 The Age of the Rigveaa [ch. 

of the snake of the abyss; but, in striking contrast with later 
India, no direct worship of the snake attributable to its deadli- 
ness occurs. Of totemism, in the sense of the belief in an animal 
ancestor and the treatment of that animal as sacred and divine, 
the Rigveda shows not a trace. On the other hand, fetishism is 
seen in the allusion already quoted to the use of an image of 
Indra against one's enemies. Analogous to this is the sentiment 
which deifies the pressing-stones which expressed the Soma, the 
drum and the weapons of the warrior and the sacrificial post. 
The chief opponents of the gods are the Asuras, a vague group 
who bear a name which is the epithet of Varuna and must 
originally have had a good meaning, but which may have been 
degraded by being associated with the conception of divine cunning 
applied for evil ends. On a lower plane are the Rakshasas, demons 
conceived as in animal as well as human shape, who seek to destroy 
the sacrifice and the sacrificers alike, but whose precise nature 
cannot be definitely ascertained. 

To the gods the Indian stood in an attitude of dependence, 
but of hope. The gods are willing to grant boons if they are 
worshipped; and the overwhelming mass of the evidence shows 
that the ordinary Vedic sacrifice was an ofiering made to win 
the divine favour, though thank-ofierings may well have been 
known \ Inextricably bound up with this conception of the divine 
relation is that other which regards the gods as subject to control 
by the worshipper if he but know the correct means, a motive 
clearly seen in the selection of the horse as a sacrifice whereby 
the swift steed, the sun, may regain strength and favour his 
worshippers. The higher and more mystic view of the sacrifice 
as a sacrament is not found except in the quite rudimentary 
form of the common meal of the priests on the sacrificial victim : 
there is no proof that in thus consuming the victim the priests 
deemed themselves to be consuming their god, though doubtless 
they regarded the meal as bringing them into special relation 
with the god who shared it with them and so in some measure 
acquired the same nature as themselves. But if the view of 
sacrifice was less mystic, in some aspects at least, than in the 
case of the Mediterranean peoples, Vedic civilisation at this stage 
was spared the horror of human sacrifice, which can be found in 
the Samhita only by implausible conjecture^. 

1 See Caland and Henry, L'Agnistoma, pp. 469-90; Keith, J.R.A.S., 1907, 
pp. 929-49. 

2 See Hillebrandt, Z.D.M.G., vol. xl, p. 708, who finds it alluded to in x, 18, 8. 
But see Keith, J.R.A.S., 1907, p. 946. 

iv] Sacrifices: Philosophy 107 

Tlie sacrifices offered included offerings of milk, grain, and ghee, 
as well as offerings of flesh and of the Soma, It is impossible to 
adapt the later sacrificial theory, as it appears in the next period, 
to the Rigvedic texts, and it is clear that at this time the sacrifice 
was less elaborate than it became; but there is abundant proof 
that already the Soma sacrifice in particular had been elaborated, 
and that the labour had been divided among several priests, the 
chief being the Hotri who recited the hymns and in earlier times 
composed them, the Adhvaryu who performed the manual actions 
to the accompaniment of muttered prayers and deprecations of 
evil, the Udgatri who sung the Saman chants, and several as- 
sistants, the number seven being found quite frequently in the 
Rigveda. Naturally these elaborate sacrifices could not be under- 
taken by any save the rich men of the tribe and especially the 
king; and we must therefore picture to ourselves the priests as 
maintained by the rich men, the Maghavans, 'bountiful ones,' of 
the Rigveda, their number and rewards rising with the social 
scale of their patron, until the height of the priest's ambition 
was attained, the position of Purohita to the king. Beside all 
this elaborate ritual there was of com'se the daily worship of 
the ordinary Aryan, which he no doubt in this period, as later, 
conducted himself; but the Rigveda is an aristocratic collection 
and contains little of popular religion beyond a few incantations 
in the tenth book, which can-y us into the homely region of spells 
against rivals and to repel diseases and noxious animals. But 
these are not really parts of the main body of the Samhita. 

The late tenth book also gives us the beginnings of the 
philosophy of India. The multiplicity of gods is questioned and 
the unity of the universe is asserted, while attempts are made 
to represent the process of creation as the evolution of being 
from not being, first in the shape of the waters and then in the 
sliape of heat. Other hjTnns more simply consider the process as 
that of a creation by Vicvakarman, 'the all-maker,' or Hiranya- 
garbha, 'the golden genu,' apparently an aspect of the sun. In 
yet another case the sacrificial theory is applied, and in the 
PurushasuMa, the earliest authority for caste divisions, the world 
is fashioned from the sacrifice of a primeval giant whose name 
Purusha, ' man,' reappears in later philosophy as the technical term 
for spirit. These speculations are of interest, not for their intrinsic 
merit, but for the persistence with which the same conceptions 
dominate the religious and philosophical systems of India, 

There is little in the Rigveda that bears on the life after 

io8 The Age of the Rigveda [CH. 

death. The dead were either cremated or buried, and, if cre- 
mated, the ashes were regularly buried. This suggests that burial 
was the older method which was altered under the pressure of 
migration and perhaps the Indian climate. The Rigveda is inno- 
cent of widow burning, though it clearly has the conception which 
gave rise to that practice, the view that life in the next world is 
a reflex of this life, and though in the next period we have 
clear references to the fact that the burning of widows was not 
unknown. The direct authority for the custom, which later days 
sought to find in the Rigveda, owes its existence to a daring 
forgery of quite modern date\ The exact fate of the dead is 
somewhat obscure: they are conceived, at one time, as dwelling 
in peace and converse with the gods of the world of Yama, the 
first of the dead and king of the dead. In other passages, the 
gods and the fathers are deemed to dwell in different places ; while 
a third conception declares that the soul departs to the watei-s 
or the plants. Beyond this last idea there is nothing in the 
Rigvedic literature to suggest that the idea of metempsychosis 
had presented itself to the Indian mind: the fate of the evil 
after death is obscure: possibly unbelievers were consigned to 
an underground darkness; but so scanty is the evidence that 
Roth held that the Vedic poet believed in their annihilation. But 
this vagueness is characteristic of the comparative indifference of 
the Rigveda to morals: the gods are indeed extolled as true, 
though perhaps rather as a means of securing that they shall 
keep faith with their votary than as an assertion of ascertained 
truth. Except in the case of Varuna, the omniscient, whose 
spies watch men and who knows the every thought of man, the 
characteristics of the gods are might and strength rather than 
moral goodness, or even wisdom. 

In its metrical form the Rigveda shows traces of the distinc- 
tion between the recitative of the Hotri and the song of the 
Udgatri^ : thus besides hymns in simple metres, rhythmical series 
of eight syllables, three or four times repeated, or eleven or twelve 
syllables four times repeated, are found strophic effects made up 
of various combinations of series of eight and twelve syllables, 
these being intended for Saman singing. The verse technique 
has risen beyond the state of the mere counting of syllables 

1 See Wilson, J.R.A.S., vol. xvi, pp. 201 sq.; Fitzedward Hall, J.R.A.S., n. s. 
vol. in, pp. 183-92 who traces it to Baghanandana (1500 a.d.). 

* See Oldenberg, Z.D.M.G., vol. xxxvni, pp. 439 sq. ; Prolegomena, pp. 1 sq. ; 
Arnold, Vedic Metre, Cambridge, 1905. 

iv] The Vedic Hymns 109 

which it shared as regards the use of eight and eleven syllable 
lines with the Iranian versification ; but the process of fixing the 
quantity of each syllable, which appears fiilly completed in the 
metres of classical Sanskrit verse, is only in a rudimentary state, 
the last four or five syllables tending to assume in the case of 
the eight and twelve syllable lines an iambic, in the case of the 
eleven syllable lines a trochaic cadence. The poetry of the col- 
lection is of very uneven merit : Varuna and Ushas evoke hymns 
which now and then are nearly perfect in poetic conception and 
expression ; but much of the work is mechanical and stilted, being 
overladen with the technicalities of the ritual : this condemnation 
applies most heavily to the ninth book, which, consisting as it does 
of hymns addressed to the Soma in the process of its purification 
for use, is arid and prosaic to a degree. In style, practically 
all the hymns are simple enough, and their obscurity, which is 
considerable, is due to our ignorance of the Vedic age, which 
renders unintelligible references and allusions clear enough to 
the authors. But there is unquestionably much mysticism in 
the later hymns and still more of that confusion of thought and 
tendency to take refuge in enigmas, which is a marked feature 
of all Indian speculation. 

The language is of the highest interest, as it reveals to us an 
Indo-European speech with a singular clarity of structure and 
wealth of inflection, even if we admit that the first discoverei-s 
of its importance from the point of view of comparative philology 
exaggerated in some degree these characteristics. Historically it 
rendered comparative phOology the first great impetus, and it 
must for all time be one of the most important subjects of study. 
But it is clearly, as preserved in the hymns, a good deal more 
than a spoken tongue. It is a hieratic language which doubtless 
diverged considerably in its wealth of variant forms from the 
speech of the ordinary man of the tribe \ Moreover it shows 
clear signs of influence by metrical necessities which induce here 
and there a disregard of the rules normally strictly observed of 
concord of noun and attribute. It must be remembered that it 
was in a peculiar position: in the first place, it was the product 
of an hereditary priesthood, working on a traditional basis: the 
very first hymn of the Samhita alludes to the songs of old and 

^ Cf. Grierson in J»t_p. Gaz., vol. i, pp. 357 sq.; J.R.A.S., 1904, pp. 436 sq.; 
Wackemagel, Altindische Gravimatik, vol. i, pp. xviii sq. ; Petersen, J.A.O.S., 
vol. xxzii, pp. 414-28; Michelson, J.A.O.S., vol. xzxiii, pp. 145-9; Keith, Aitareya 
Aranyaka, pp. 180, 196. 

no The Age of the Rigveda [ch. 

new poets: in the second place, the language of all classes was 
being affected by the influence of contact with the aboriginal 
tongues. The existence of slaves, male and especially female, 
must have tended constantly to affect the Aryan speech, and the 
effect must have been very considerable, if, as seems true, the 
whole series of lingual letters of the Vedic speech was the result 
of aboriginal influence. Many of the vast number of words with 
no known Aryan cognates must be assigned to the same influence. 
Thus in the period of the Rigveda there was growing up an ever 
increasing divergence between the speech of the learned and that 
of the people. As a result the language of literature remains 
the language of the priesthood and the nobility: it is modified 
gradually, and finally, at an early date, fixed for good as regards 
form and construction by the action of the grammarians : on the 
other hand, the speech of the commoner, in consequence of the 
constant contact with the aborigines and the growing admixture 
of blood, develops into Pali and the Prakrits and finally into 
the modern vernaculars of India. What we do not know is how 
far at any given moment in the Vedic period the gulf of sepa- 
ration had extended. Nor do we know whether at this epoch there 
were distinct dialects of the Vedic speech: efforts to find traces 
of dialects in the Rigveda have so far ended in no secure results 

It is natural, at the conclusion of this survey of the more 
important aspects of the Vedic civilisation, to consider what 
date can be assigned to the main portion of the Rigveda or to 
the civilisation which it records. One fact of interest has been 
adduced from the records of treaties between the Hittites and 
the Kings of Mitani of about 1400 B.C. In them occur names 
which a certain amount of faith may induce us to accept as 
denoting Indra, the two A9vins under the name Nasatya, one of 
their epithets — of unknown meaning — in the Rigveda, Mitra, and 
Varuna. It is right to add that these identifications must not be 
regarded as certain, though they may be correct. It has been 
argued by Jacobi^ that these names must be derived from a 
tribe practising the religion revealed to us in the Rigveda, that 

1 The theory of Hoernle, Grierson, and Kisley {Imperial Gaz., vol. i, pp. 303 sq.), 
which sees in the Rigvedic language the speech of the Middle Country (Madhyadega) 
only is not supported by the Eigveda. Only the N.W. region of the Middle Country, 
■which lay between the rivers SarasvatI and Drishadvati (Brahmavarta) was intimately 
known to the poets of the Eigveda. They show more acquaintance with the Punjab 
and with the Kabul Valley than with the Middle Country generally, that is to say the 
region lying between the Sarasvati and Prayaga, the modem Allahabad. 

■^ J.E.A.S., 1909, pp. 721-6. For these names see also Chapters in and xiv. 

iv] Evidences of Date 1 1 1 

the presence of this tribe at this date is due to a movement on 
their part from India, and that we have a definite date assigned 
at which the culture of the Rigveda existed. Unhappily the 
argument cannot be regarded as conclusive. It is considered by 
E. Meyer ^ and by Oldenberg^ that the gods are proto-Iranian 
gods, aflfording a proof of what has always seemed on other grounds 
most probable, that the Indian and Iranian period was preceded 
by one in which the Indo-Iranians stiU undivided enjoyed a 
common civilisation. This is supported by the fact that the 
Avesta, which is doubtless a good deal later than the date in 
question, still recognises a great god to whom Varuna's epithet 
Asura is applied, that it knows a Verethrajan who bears the 
chief epithet of Indra as Vritrahan, * slayer of Vritra,' that It 
has a demon, Naonhaithya, who may well be a pale reflex of the 
Nasatyas, and that the Avestan Mithra is the Vedic Mitra. It 
is also possible that the gods represent a period before the sepa- 
ration of Indians and Iranians, though this would be less likely 
if it is true that the names of the Mitani princes include true 
Iranian names^ But, in any case, it is to be feared that we 
attain no result of value for Vedic chronologj^ 

Another and, at first sight, more promising attempt has been 
made to fix a date from internal evidence. It has been argued 
by Jacobi* on the strength of two hymns in the Rigveda that 
the year then began with the summer solstice, and that at that 
solstice the sun was in conjunction with the lunar mansion 
Phalguni. Now the later astronomy shows that the lunar man- 
sions were, in the sixth century A.D., arranged so as to begin for 
purposes of reckoning with that called A9vini, because at the 
vernal equinox at that date the sun was in conjunction with 
the star ^ Piscium. Given this datum, the precession of the 
equinoxes allows us to calculate that the beginning of the year 
with the summer solstice in Phalguni took place about 4000 b.C. 
This argument must be considered further in connexion with 
the dating of the next period of Indian history; but, for the 
dating of the Rigveda, it is certain that no help can be obtained 
from it It rests upon two wholly improbable assumptions, first, 
that the hymns really assert that the year began at the summer 
solstice, and, second, that the sun was then brought into any 

1 Sitzungsberichte der k. preussischen Akademie der Wissenschaften, 1908, pp. 14 sq. 

2 J.R.A.S., 1909, pp. 1095-1100. Cf. Keith, ibid. 1100-6. 
' Sayce, ibid. p. 1107, denies this. 

* Festgrtiss am Eothj, pp. 68 sq. = Indian Antiquai-y, vol. xxiii, pp. 154 sq. 

112 The Age of the Rigveda [ch. 

connexion at all with the Nakshatras, for which there is no 
evidence whatever. The Nakshatras are, as their name indi- 
cates and as all the evidence of the later Samhitas shows, lunar 
mansions pure and simple. 

In the absence of any trustworthy external evidence, we are 
forced to rely on what is after all the best criterion, the develop- 
ment of the civilisation and literature of the period. Max Miiller^ 
on the basis of this evidence divided the Vedic period into four, 
that of the Sutra literature, 600-200 RC, the Brahmanas, 800- 
600 B.C., the Mantra period, including the later portions of the 
Rigveda, 1000-800 B.C., and the Chhandas, covering the older and 
more primitive Vedic hymns, 1200-1000 B.C. The exact demarca- 
tion did not claim, save as regards the latest period, any special 
exactitude, and was indeed somewhat arbitrary. But the fact 
remains that definitely later than the Rigveda we find the other 
Samhitas, of which an account is given below, and the prose 
Brahmana texts, which contain comments on and explanations of 
the Samhitas, whose existence they presuppose. It is impossible 
to deny that this mass of work must have taken time to produce, 
especially when we realise that what has survived is probably 
a small fraction as compared with what has been lost. Now in the 
Brahmanas we find only the most rudimentary elements of the 
characteristic features of all Indian literature after Buddhism, 
the belief in metempsychosis, pessimism, and the search for de- 
liverance. The distance between the Brahmana texts with their 
insistence on the ritual, and their matter-of-fact and indeed sordid 
view of the rewards of action in this world, and the later doctrine 
of the uselessness of all mundane efibrt, is bridged by the 
Aranyakas and the Upanishads which recognise transmigration, 
if not pessimism, which definitely strive to examine the real 
meaning of being, and are no longer content mth the explana- 
tion of sacrifices and idle legends. It is unreasonable to deny 
that these texts must antedate the rise of Buddhism, which, in 
part at least, is a legitimate development of the doctrines of the 
Upanishads. Now the death of Buddha falls in all probability 
somewhere within the second decade of the fifth century before 
Christ^: the older Upanishads can therefore be dated as on the 
whole not later than 550 B.C. From that basis we must reckon 
backwards, taking such periods as seem reasonable; and, in the 
absence of any means of estimating these periods, we cannot 

1 Cf. Rigveda Samhitd, vol, iv^, pp. vii sq. 

2 Fleet, J.R.A.S., 1912, p. 240, thinks 483 B.C. is the most probable date. 

iv] Development of Civilisation 113 

have more than a conjectural chronology. But it is not likely 
that the Brahmana period began later than 800 B.C., and 
the oldest hymns of the Rigveda, such as those to Ushas, may 
have been composed as early as 1200 RC. To carry the date 
further back is impossible on the evidence at present available, 
and a lower date would be necessary if we are to accept the 
view that the Avesta is really a product of the sixth century B.C., 
as has been argued on grounds of some though not decisive 
weight; for the coincidence in language between the Avesta 
and the Rigveda is so striking as to indicate that the two 
languages cannot have been long separated before they arrived 
at their present condition. 

The argument from literature and religion is supported also 
by the argument from civilisation. The second period, that of 
the Samhitas, shows the development of the primitive Vedic 
community into something more nearly akin to the Hinduism 
which, as we learn from the Greek records, existed at the time 
of the invasion of Alexander and the immediately succeeding 
years. But we are still a long way from the full development 
of the system as shown to us in the Artha9astra, that remarkable 
record of Indian polity which is described in Chapter xix. The 
language also of the Vedic literature is definitely anterior, 
though not necessarily much anterior, to the classical speech 
as prescribed in the epoch-making work of Panini: even the 
Siitras, which are undoubtedly later than the Brahmanas, show 
a freedom which is hardly conceivable after the period of the 
fiill influence of Panini^; and Panini is dated with much plausi- 
bility not later than 300 RC.^ 

1 Biihler, Sacred Books of the East, vol. ii^, p. xlv, relies on this argument to assign 
Apastamba's Sutras to a date not later than the third cent. B.C., and suggests that they 
may be 150 or 200 years earlier. 

^ See Keith, Aitareya Aranyaka, pp. 21-5. 

C.H.I. I. 



Definitely later than that depicted in the Rigveda is the 
civilisation presented by the later Samhitas, the Brahmanas, the 
Aranyakas, and the Upanishads. It is on the whole probable that 
the total time embraced in this period is not longer, perhaps it is 
even shorter, than that covered by the earlier and later strata 
of the Rigveda; and there are hymns in the tenth book of the 
Rigveda which are really contemporaneous with the later Samhitas, 
just as those Samhitas have here and there preserved work of 
a much earlier epoch. But the distinction between the main body 
of the Rigveda and the rest of the Vedic literature is clear and 
undeniable. Nor is it open to much doubt that the redaction of the 
Samhita of the Rigveda into Avhat, in substance as opposed to 
verbal form, was its present shape took place before the other 
Samhitas were compiled. Of these Samhitas the Samaveda, the 
collection of chants for the Saman singers, is so dependent on the 
Rigveda for its contents, that it is negligible for purposes of 
history. On the other hand, the Samhitas of the Yajurveda, the 
collection of the formulae and prayers of the Adhvaryu priest, | 

to whose lot fell the actual performance of the sacrificial acts, are 
of the highest historical importance. They represent two main 
schools, the Black and the White, the name of the latter being 
due, according to tradition, to the fact that, whereas the texts of 
the Black Yajurveda contain verse or prose foiinulae and the 
prose explanations and comments combined into one whole, the 
text of the latter distinguishes between the verse and prose 
formulae which it collects in the Samhita, and the prose ex- 
planations which it includes in a Brahmana. Of the Black Yajur- 
veda three complete texts exist, those of the Taittiriya, the 
Kathaka, and the MaitrayanI schools, while considerable fragments 
of a Kai)islithala Samhita closely allied to the Kathaka also exist. 

CH. v] Xhe earliest Prose 115 

In the case of the Taittiriya there is a Brahmana which is a 
supplementary work, dealing with matter not taken up in the 
Samhita. The White school has the Vajasaneyi Samhita and the 
^atapatha Brahmana, the latter being one of the most important 
works in the whole Vedic literature. Finally, there is the Samhita 
of the Atharvaveda, which is technically reckoned as appertaining 
to the Brahman, the priest who in the later state of the ritual 
superintends the whole of the sacrifice, and which is a curious 
, repository of most mingled matter, for the most part spells of 
every kind, but containing also theosophical hymns of considerable 

The conjunction of the prose explanation with the formulae does 
not prove the later composition of both the prose and the formulae, 
and there is no ground for attributing the two strata to the same 
date. On the other hand, the prose of the Yajurveda Samhitas is 
amongst the earliest Vedic prose. Possibly somewhat earlier may 
be that of the Pafichavim9a Brahmana, which is the Brahmana of 
the Samaveda, and which, despite the extraordinary technicality 
of its details, is yet not without importance for the history of the 
civilisation of the period. The Brahmanas of the Rigveda are 
probably slightly later in date, the older being unquestionably the 
earlier part (books i-v) of the Aitareya, and the younger the 
Kaushitaki or (^ankhayana\ When the Atharvaveda, which long 
was not recognised as fully entitled to claim rank as a Veda 
proper, came within the circle of the Vedas, it was considered 
desirable to provide it with a Brahmana, the Gopatha, but this 
strange work is in part a cento from other texts, including the 
^atapatha Brahmana, and appears to be later than the Kau9ika 
and Vaitana Sutras attached to the Atharvaveda : its value then 
for this period is negligible. 

Special portions from the Brahmanas are given the title of 
Aranyaka, 'forest books,' apparently because their contents were 
so secret that they had to be studied in the depths of the forests, 
away from possibility of overhearing by others than students. 
The extant texts which bear this name are the Aitareya, the 
Kaushitaki, and the Taittiriya, which are appendages to the 
Brahmanas bearing those names. All three are somewhat hetero- 
geneous in composition, the Aitareya being the most definitely 
theosophical, while the Taittiriya is the least. Still more important 
are the Upanishads, so called because they were imparted to pupils 
in secret session, the term denoting the sitting of the pupil before 

^ See Keith, Aitareya Aranyaka, pp. 172, 173. 


Ii6 Later Samhitas^ Brahmanas^ etc, [ch. 

the teacher. Each of the three Aranyakas contains an Upanishad 
of corresponding name. More valuable however are the two great 
Upanishads, the Brihadaranyaka, which is attached to the (^atapatha 
Brahmana, forming part of its fourteenth and last book in one re- 
cension and the seventeenth book in the other, and the Chhandogya 
Upanishad attached to the Samaveda ; these two are in all 
probability the oldest of the Upanishads. To the Samaveda also 
belongs the Jaiminlya Brahmana, one book of which, the Jaiminiya 
Upanishad Brahmana, is really an Aranyaka, and, like other . 
Aranyakas, contains in itself an Upanishad, the brief but interesting 
Kena Upanishad. The number of treatises styled Upanishad is 
very large ; but, with the possible exception of the Kathaka, 
which expands a legend found in the Taittiriya Brahmana dealing 
with the nature of the soul, none of them other than those 
enumerated can claim to be older than Buddhism ; and the facts 
which they contain cannot therefore prudently be used in sketching 
the life of the period under review. Similarly, the Sutras, which 
are text-books either giving in the form of very brief rules 
directions for the performance of the sacrifice in its various forms 
(the Qrauta Sutras dealing with the great rites at which a number 
of priests were employed, the Grihya Sutras with the domestic 
sacrifices and other duties performed by the householder), or 
enunciating customary law and practice (the Dharma Sutras), 
cannot safely be relied upon as presenting a picture of this period. 
They are however of much indirect value ; for they throw light 
upon practices which are alluded to in the Brahmanas in terms 
capable of more than one interpretation ; and here and there they 
preserve verses, far older than the works themselves, which contain 
historic facts of value. 

We have seen that, in the period of the Rigveda, the centre of 
the civilisation was tending to be localised in the land between the 
Sarasvati and the Drishadvati, but that, though this was the home 
of the Bharatas, other tribes including the famous five tribes 
dwelt in the Punjab, which had in all probability been the earlier 
home of the Indians. In the Brahmana period, as the period 
under review may conveniently be called, the localisation of 
civilisation in the more eastern country is definitely achieved, and 
the centre of the life of the day is Kurukshetra, bounded by 
Khandava on the south, Turghna on the north, Parinah on the 
west In contrast with the frequent mention of the eastern lands 
the Punjab recedes in importance ; and its later name, Panchanada, 
* land of the five streams,' is not fomid until the epic period. The 

v] Extension of Aryan Civilisation 117 

tribes of the west receive disapproval both in the ^atapatha and 
the Aitareya Brahmanas. In the Aitareya Brahmana a geo- 
graphical passage ascribes to the Middle Country, the later Ma- 
dhyade9a, the Kuriis and Paiichalas with the Va9as and the U9inara8, 
to the south the Satvants, and to the north beyond the Himalaya 
the Uttara-Kurus and the Uttara-Madras. On the other hand, 
while the west recedes in importance the regions east of the Kuru- 
Panchala country come into prominence, especially Kosala, corre- 
sponding roughly to the modern Oudh, Videha, the modern Tirhut 
or N. Bihar, and Magadha, the modern S. Bihar. Still further east 
was the country of the Angas, the modern E. Bihar. In the south 
we hear of outcast tribes in the Aitareya Brahmana, probably tribes 
who were not fully Brahmanised: their names are given as the 
Andhras, who appear as a great kingdom in the centuries im- 
mediately before and after the Christian era, Pundras, Miitibas, 
Pulindas, and ^abaras, the last named being now a tribe living on 
the Madras frontier near Orissa and showing, in its language, 
traces of its Munda origin. In the south also was Naishadha. 

It does not seem likely that Aryan civilisation had yet over- 
stepped the Vindhya, which is not mentioned by name in the 
Vedic texts, though the Kaushitaki Upanishad refers to the 
northern and southern mountains, the latter of which must be the 
Vindhya. At the same time geographical knowledge of the north 
is wider : the Atharvaveda knows not only of the Mujavants and the 
Gandharis, but also of the Mahavrishas, and the name of a place in 
the Mahavrisha country, Raikvaparna, is preserved in the Chhan- 
dogya Upanishad. Yaska in the Nirukta, a text of about 500 RC. 
explaining with illustrations certain selected Vedic words, tells us 
that the speech of the Kambojas differed in certain respects from 
the ordinary Indian speech, referring doubtless to the tribes living 
north-west of the Indus who bore that name. Vidarbha, the modern 
Berar, is mentioned, but only in the late Jaiminiya Upanishad 
Brahmana, though a Bhima of Vidarbha occurs in a late passage 
of the Aitareya. 

In addition to a wider geogi-aphical outlook, the Brahmana 
period is marked by the knowledge of towns and definite localities. 
There are fairly clear references to Asandivant, the Kuru capital, 
Kampila, the capital of Panchala in Madhyade9a, to Kau9ambi, 
and to Ka9i, the capital of the Ka9is on the river Varanavati, 
whence in later times Benares derives its name. So we hear in 
this period for the first time of the Vina9ana, the place of the 
disappearance of the Sarasvati in the desert, and Plaksha Prasravana 

1 1 8 Later Samhitas^ Brahmanas^ etc, [ch. 

the place forty-four days' journey distant, where the river reappears 
and which, in the version of the Jaiminiya Upanishad Brahmana, 
is but a span from the centre of the universe. These are clear 
signs both of more developed city life and of more settled habits. 

Corresponding with the change in geographical conditions is 
a stiU gi'eater change in the grouping of the tribes. The Bharatas, 
who are the heroes of the third and the seventh books of the 
Rigveda, no longer occupy the main position, and we find in their 
place, in the land which we know they once held, the Kurus, and 
close to the Kurus the allied Paiichalas. As we have seen already, 
there is little doubt that the Kurus were new comers with whom 
the Bharatas amalgamated, and the Kurus thus reinforced included 
in their numbers the Purus. The mention of the Uttara-Kurus as 
resident beyond the Himalaya is sufficiently accounted for if we 
suppose that a branch of this tribe had settled in Kashmir, just as 
another branch seems to have settled on the Indus and the Chenab. 
The Panchalas, too, seem to have been a composite tribe, as the 
name which is clearly derived ivQvu'pancha, ' five,' shows. According 
to the ^atapatha Brahmana the older name for the Paiichalas was 
Krivi ; and we may at least believe that the Krivis who with the 
Kurus appear to have constituted the two Vaikarna tribes of the 
Rigveda were a part of the Panchala nation^. The same Brahmana 
suggests, if it does not prove, that the Turva9as were another 
element of the people ; and the disappearance from history at this 
period of the Anus and Druhyus may indicate that they also were 
merged in the new confederation. With the Kurus and Panchalas 
must be ranked the Va9as and Ui^inaras, two minor tribes who 
occupied the Middle Country, and the Srinjayas, whose close con- 
nexion with the Kurus is proved beyond doubt by the fact that at 
one time they had a Purohita in common, showing that, for the 
time at least, they must have been acting under the leadership of 
one king. 

In the texts the Kuru-Panchalas pass as the models of good 
form: the sacrifices are perfectly performed in their country: 
speech is best spoken there and, as it seems, among the northern 
Kurus ; and the Kaushitaki Brahmana tells of people going to the 
north for the sake of its pure speech. The Kuru-Panchala kings 
are the example for other kings : they perform the Rajasuya, the 
sacrifice of the royal consecration : they march forth in the dewy 
season for their raids and return in the hot season. Their Brahmans 
are famous in the literature of the Upanishads for their knowledge ; 

1 See also Chapter iv, p. 83. 

v] Kurus and Pahchalas 119 

and the Samhitas and Brahmanas which are preserved seem, with- 
out exception, to have taken definite form among the Kuru- 
Panchalas, even when, as in the case of the ^^tapatha Brahmana, 
they recognise the existence of the activities of the kings and 
priests of Kosala-Videha. It is significant of the state of afiairs 
that in the Samhitas and allied texts of the Yajm-vedas where the 
ceremony of the Rajasiiya is described, the king is presented to 
the people with the declaration, * This is your king, Kurus,' with 
variants of ' O Panchalas ' and ' Kuru-Paiichalas.' 

In the Sanskrit epic the Kurus and the Panchalas are conceived 
as being at enmity; and it is natural to enquire whether this 
tradition goes back to the Vedic period \ The reply, however, 
must be in the negative, for the evidence adduced in favour of the 
theory is of the weakest possible character. In the Kathaka 
Samhita there is an obscure ritual dispute between a certain priest, 
Vaka, son of Dalbha, who is believed to have been a Paiichala, 
and Dhritarashtra Vaicitravirya, who is assumed to have been 
a Kuru king. But apart from the fact that a mere dispute on 
a point of ritual between a Panchala priest and a Kuru king could 
not prove any hostility between the two peoples, there is no ground 
for supposing that this Dhritarashtra was any one else than the 
king of the Ka9is who bears the same name and who was defeated 
by the Bharata prince, Satrajita ^atanika, and in the very same 
passage of the Kathaka allusion is made to the union of the Kuru- 
Paiichalas. A second argument of some human interest is derived 
from the clever suggestion of Weber that in the revolting ceremony 
of the horse-sacrifice, one of the great kingly sacrifices by which 
the Indian king proclaimed his claim to imperial sway, the queen 
of the Kurus is compelled to lie beside the victim, since otherwise 
Subhadrika, the wife of the king of KampTla, the capital of 
Panchala, would take her place. If this were the case there would 
be convincing proof of an ancient rivalry which might weU end 
in the bitter conflicts of the epic ; but, unhappily, the interpretation 
is almost certainly incorrect. With the absence of evidence of 
opposition between the Kurus, assumed to have been specially 
Brahmanical, and the Panchalas, disappears any support for the 
theory 2, based on the phenomena of the later distribution of 
dialects in India, that the Kurus were a fresh stream of immigrants 
into India who came via Chitral and Gilgit and forced themselves 

1 For this view see Weber, Indische Studien, vol. i, pp. 184, 205, 206; vol. ni, 
p. 470; Grierson, J.R.A.S., 1908, pp. 602-7, 837-44, 1143. Arguments against are 
given by Keith, J.R.A.S., 1908, pp. 831-6, 1188-42. 

- See Chapters ii, pp. 45-6, 50, and iv, p. 110, note 1. 

I20 Later Samhitas^ Brahmanas^ etc, [ch. 

as a wedge between the Aryan tribes already dwelling in the land. 
The theory proceeds to assume that, coming with tew or no women, 
they intermingled with the Dra vidian population with great com- 
pleteness and produced the Aryo-Dra vidian physical type. If these 
things were so, the fact was not at any rate known by the age 
which produced the Samhitas and the Brahmanas. 

Though the Bharatas disappear in this period as a tribe, the 
fame of the Bharata kings had not been lost : in a passage in the 
(^atapatha Brahmana which describes the famous men who sacrificed 
with the horse-sacrifice, we hear of the Bharata Dauhshanti, whom 
the nymph (^akuntala bore at Nadapit, and who defeated the king 
of the Satvants and won victories on the Ganges and Jumna, 
showing that the Bharatas, as in the Rigveda, were performing 
their great deeds on the eastern as well as on the western side 
of the kingdom. Another king, Satrajita ^atanika, as we have 
seen, defeated the king of the Ka9is. We hear too of a descendant 
of Divodasa, Pratardana, whose name is of value as tending to 
show that the Tritsus were the family of the royal house of the 
Bharatas : according to the Kaushitaki Upanishad he met his death 
in battle. It is possible that with him perished the direct Tritsu 
line: at any rate, the first king who bears the Kuru name, 
Kuru9ravana, is a descendant of Trasadasyu, the greatest of the 
Puru kings. But of Kuru9ravana and of his father Mitratithi, 
and his son Upama9ravas we know practically nothing; and the 
first great Kuru king is one mentioned in the Atharvaveda, 
Parikshit, in whose reign the hymn tells us the kingdom of the 
Kurus flourished exceedingly. His grandson and great-grandson 
according to tradition were the Pratisutvana and Pratipa whose 
names are mentioned in the Atharvaveda. A later descendant of 
his was the famous Janamejaya, whose horse-sacrifice is celebrated 
in the ^atapatha Brahmana, and who had in his entourage the 
priests Indrota Daivapi ^aunaka and Tura Kavasheya. His brothers 
Ugrasena, Bhimasena, and Qrutasena by the same sacrifice purified 
themselves of the crime of Brahman-slaying. But the history of 
the Kurus was not apparently, at the end of the period, un- 
chequered : there is an obscure reference to their being saved by 
a mare, perhaps a reference to the prowess of their charioteers or 
cavalry in battle ; but the same text, the Chhandogya Upanishad, 
alludes to a hailstorm or perhaps a shower of locusts^ afflicting 
them, and a prediction is preserved in an old Sutra telling that 
they would be driven from Kurukshetra. It is in accord with 

1 See Jacob, J.R.A.S., 1911, p. 510. 


v] Peoples of the Period 12 1 

these hints that the Brihadaranyaka Upanishad sets as a question 
for discussion the problem what has become of the descendants of 
Parikshit: the dynasty must have passed away in some great 
disaster. From the ^atapatha Brahmana we gather that the 
capital of Janamejaya was Asandivant, 'the city of the throne,' 
and that at Mashnara a Kuru king won a victory, and Tura 
Kavasheya, a priest of the Bharatas, sacrificed at KarotL 

Of the Paiichalas apart from the Kurus we hear comparatively 
little : they had however kings like Kraivya and ^ona Satrasaha, 
father of Koka, who performed the horse-sacrifice and thus claimed 
imperial power, Durmukha, who was taught the royal consecration 
by Brihaduktha and conquered the whole earth, and the more 
real Pravahana Jaivali who appears as philosopher king in the 
Upanishads, and who at least must have been willing to take part in 
the disputes of the Brahmans at his court Panchala towns were 
Kampila, Kau9ambT, and Parivakra or Paricakra, the scene of 
Kraivya's exploits. 

The Uttara-Kurus seem already in the time of the Aitareya 
Brahmana to have won a somewhat mythical reputation, for when 
Atyarati Janamtapi, who was not a king, proposed to conquer 
them as well as the rest of the world, he was dissuaded by his 
priest Vasishtha Satyahavya, and for his rashness was defeated by 
Amitratapana (^ushmina, the king of the ^ibis, a tribe no doubt 
identical with the ^ivas of the Rigveda and belonging to the 
north-west. The Uttara-Madras must have lived near them in 
Kashmir ; and the Madras of Avhom we hear in the Brihadaranyaka 
Upanishad were, in the Buddhist epoch, settled between the Chenab 
and the Ravi. In the Middle Country with the Kuru-Paiichalas 
were the Va9as and U9inaras who seem to have been of no 
importance. With them in the Kaushitaki Upanishad are coupled 
the Matsyas, and we hear of one great Matsya king, Dhvasan 
Dvaitavana, who performed the horse-sacrifice and who probably 
ruled in or about Jaipur or Alwar, where lake Dvaitavana must be 
placed. On the Jumna we hear at the end of the period of the 
Salvas, under king Yaugandhari, probably in close touch with the 
Kuru-Panchala people. 

The Srinjayas also stood in this period in close relationship 
to the Kurus, and like the Kurus the Srinjayas seem to have 
sufiered disaster at some period. The Vaitahavyas, the Atharvaveda 
relates, offended the priestly family of the Bhrigus and came to 
ruin : this tradition is confirmed by the notices of disasters in the 
Kathaka and Taittiriya Samhitas. Of their history we have one 

122 Later Samhitas^ Brahmanas^ etc, [ch. 

definite glimpse : they rose against their king, Dushtaritu Paum- 
sayana, despite the ten generations of his royal descent, and 
expelled him with his Sthapati, 'minister,' Chakra Revottaras 
Patava ; but the latter afterwards succeeded in restoring his master 
to power, despite the opposition of Balhika Pratipiya, whose 
patronymic reminds us of the Pratipa who was a descendant of 
the Kuru king Parikshit, showing that the Kuru princes were 
probably anxious enough to use domestic strife as a means of 
securing a hold over a neighbouring kingdom. Perhaps in the 
long run the ruin of the Vaitahavyas took the shape of absorption in 
the Kuru realm. On the other hand, the defeats of the Satvants on 
the south by the Kurus were doubtless nothing more than mere 

Further east of the Kuru-Panchala realm lay the territories of 
Kosala and Videha, which were, however, not allied in any so close 
a manner as the Kurus and the Paiichalas. Para, son of Atnara, 
their greatest king who celebrated the horse-sacrifice, is however 
spoken of as a king of Videha as well as a king of Kosala, showing 
that the kingdoms were sometimes united under one sovereign. 
A well-known legend in the ^^tapatha Brahmana recognises that 
Videha received Vedic civilisation later than Kosala, for it tells 
how Mathava the Videgha, whose name shows the older form of 
the word Videha, passed from the Sarasvati, the seat of Vedic 
culture, to the land of Videha, crossing the Sadanira; this perennial 
stream, as its name denotes, formed the boundary of Kosala on 
the east and, with some plausibility, has been identified with 
the modern Gandak, which rising in Nepal joins the Ganges 
near Patna. Kagi and Videha are also connected in the Kaushitaki 
Upanishad ; and a late text preserves the record that Jala 
Jatukarnya was the Purohita of the Kosalas, Videhas, and Ka9is 
at one time, proving a temporary league. Of other kings we hear 
of the Kosalan Hiranyanabha, of the Videhan Nam! Sapya, and 
beyond all of Janaka of Videha, whose fame leads him to play the 
part of the father of Sita, the heroine of the Ramayana, the 
second of India's gi*eat epics. Janaka appears himself as a king 
ever anxious to seek for the wisdom of the Brahmans ; and among 
his contemporaries are mentioned the great Yajnavalkya, and 
^vetaketu. His contemporary was Ajata9atru of Ka9l, whom one 
account indeed refers to as of Ka9i or Videha, and it is a natural 
suggestion that in this name we have a chronological fact of value. 
It is suggested that in this A]ata9atru we have the Ajatasattu of 
the Buddhist texts, who was a contemporary of the Buddha and who 

v] The Eastern Peoples 123 

therefore reigiied in the sixth century rg.^ But the suggestion is 
not a happy one. In the Buddhist text Ajatasattu never appears 
as king of any other place than Magadha, and the name is merely 
an epithet, *he who has no foe,' which could be applied to any 
king, though it may well be that the Ajatasattu of Magadha gladly 
borrowed an epithet which a king of Kaci had made famous. 
Other kings of Ka^I were Dhritarashtra, whose defeat by a Bharata 
has been mentioned above, and Bhadrasena, a descendant of 

It is very noticeable that the relations of Ka9i and the Bharatas 
seem to have been those of war; and there is evidence of some 
aversion existing between the Kosala-Videhas and the Ka9is on 
the one hand and the Kuru-Panchalas on the other. It is clear 
enough that the Brahmanical tradition came to the Kosala-Videhas 
from the Kuru-Panchala country; but the question remains 
whether the Aryan tribes, who occupied Oudh and Tirhut, were 
a branch of the Kuru-Panchalas or men who were originally settled 
in the Kuru-Paiichala country or on its borders and were pushed 
eastwards by the pressure of the Kuru-Paiichalas. The evidence 
is not sufficient to pronounce any opinion on either view, and, as 
we have seen, still less to show that the Kurus were distinct from 
the Panchalas as a different branch of the Aryan invaders of 

Much more definitely still beyond the pale were the people of 
Magadha, which serves with Anga in the Atharvaveda as a symbol 
of a distant land. The man of Magadha is dedicated, in the 
account of the symbolic human sacrifice given in the Yajurveda, 
to 'loud noise,' suggesting that the Magadha country must have 
been the seat of minstrelsy, an idea supported by the fact that in 
later literature a man of Magadha is the designation of a minstrel. 
If, as has been suggested, the Kikatas of the Rigveda were really 
located in Magadha, the dislike of the country goes back to the 
Rigveda itself. The cause must probably have been the imperfect 
Brahmanisation of the land and the predominance of aboriginal 
blood, which later in history rendered Magadha the headquarters 
of Buddhism. It is significant that the Buddhist texts show a 
subordination of the Brahman to the Kshatriya class which has no 
parallel in the orthodox literature. It is clear however that 
Brahmans sometimes lived there, but that their doing so was 
a ground for surprise. 

^ See Hoernle, Osteology, p. 106. For arguments against, see Keith, Z.D.M.G., 
vol. Lxii, pp. 138-9. 

124 Later Samhitas^ Brahmanas^ etc, [ch. 

The man of Magadha is brought into close connexion with the 
Vratya in a mystical hymn in the Atharvaveda which celebrates the 
Vratya as a type of the supreme power in the universe. A more 
connected account of the Vratyas is found in the Paiichavim^a 
Brahmana of the Samaveda and the Sutras of that Veda^. It is 
clear that, as their name suggests, they were persons regarded as 
outcasts; and ceremonies are described intended to secure them 
admission into the Brahmanical fold. The description of the 
Vratyas well suits nomad tribes : they are declared not to practise 
agriculture, to go about in rough wagons, to wear turbans, to carry 
goads and a peculiar kind of bow, while their garments are of 
a special kind. Their sense of justice was not that of the Brahmans, 
and their speech, though it seems Aryan, was apparently Prakritic 
in form, as is suggested by the significant remark that they called 
what was easy of utterance hard to speak ; for the Prakrits differ 
from Sanskrit essentially in their efforts to avoid harsh consonantal 
combinations. Where they were located is not certain, for their 
habits would agi*ee well enough Avith nomads in the west ; but the 
little information which we have seems fairly enough to lead to the 
conclusion that some at least of- the Vratyas were considered to 
be dwellers in Magadha. 

There is little to be said of other tribes. The Vidarbhas are 
known through one of their kings who received certain knowledge 
from the mythical sages Parvata and Narada, and through a special 
kind of dog found in their country. The list of kings who per- 
formed the horse-sacrifice includes the Qvikna king, Rishabha 
Yajnatura. Mention has been made above of the Paravatas, who 
were found on the Jumna; and the Kekayas with their prince 
A9vapati, and the Balhikas were located in the far north. The 
temptation to transform the name of the latter into a sign of 
Iranian influence must be withstood, as it rests on no sure basis 
and we have seen Balhika as part of the name of a Kuru prince. 
An early Siitra refers to ^aphala, the kingdom of Ritupania. The 
Andhras, and other tribes mentioned by the Aitareya Brahmana as 
outcasts, were probably still Dra vidian in blood and speech, though 
Munda speaking tribes may have been mingled with them as the 
name ^^bara suggests. The Angas, too, may have been com- 
paratively little affected by the influence of the Aryan culture. 
It has been conjectured that in Magadha the wave of Aryan 
civilisation met with another wave of invasion from the east ; but, 

^ Charpentier, V.O.J. , vol. xxv, pp. 355 sq., sees in the Vratyas the precursors of 
the Qivaites of to-day. But see Keith, J.R.A.S., 1913, pp. 155 sq. 

v] . Social Changes 125 

tempting as the suggestion is, it cannot be supported by anything 
in the Vedic literature ^ 

As was to be expected, society was far from unchanged in this 
period of active Aryan expansion. As we have seen, there is good 
reason to believe that in the period of the Rigveda the priesthood 
and the nobility were hereditary. This view receives support from 
the fact that similar class distinctions are to be found in other 
Indo-European communities, such as the patrician gentes in Rome, 
the Eupatridae of Athens, the nobles of early Germany, the eorls 
of the Anglo-Saxons, and the still closer parallel of the Iranian 
classes of Athravas and Rathaesthas, ' priests ' and * warriors.' It 
may even be that these distinctions are earlier than the severance 
of the Indo-Iranians, if not as old as the union of the Aryan 
peoples. But in this period there comes into existence a new 
factor, the introduction of divisions among the ordinary freemen, 
the Vai9ya8, and the development of a large and complicated 
system of caste which converts the simple distinction of Vai^ya 
and (^udra into an ever-increasing number of endogamous hereditary 
groups practising one occupation or at least restricted to a small 
number of occupations. This result was certainly far from being 
reached in the period of the Brahmanas, but the tendency of social 
or racial distinctions to harden into castes is already apparent. In 
this development there must have been two main influences : the 
force of occupation is later revealed clearly enough in the Pali 
texts, and another interesting case is supplied by the Brahmanas 
themselves. In the Taittirlya Brahmana the Rathakaras, ' chariot 
makers,' appear as a special class along with the Vai9yas ; and in this 
special position we can see how the chariot makers, the type of skilled 
workers in the Rigveda, have, through their devotion to a mechanical 
art, lost status as compared with the ordinary freeman. The influence 
of the aborigines must also have been very strong, as intermarriage 
proceeded. To be born of a female 9udra was a disgrace with 
which Kavasha and Vatsa were taunted by their priestly con- 
temporaries: contact with the aborigines seems to have raised 
questions of purity of blood very like those which at present 
agitate the Southern States of the United States or the white 
people in South Africa. In the Rigveda, restrictions on inter- 
marriage seem to have been of the simplest kind, confined to rules 
such as those prohibiting marriage of brother and sister or father' 
and daughter. In the Sutras the rules are still not quite rigid; 

1 See Pargiter, J.R.A.S., 1908, p. 852. Oldenberg, Buddha'^, p. 10, thinks that the 
Anga, Magadlia, Ka^ji, Kosala, and Videha tribes were earlier Aryan immigrants. 

126 Later Samhi^s^ Brahma?jas^ etc. [cH. 

but they insist that there shall be no marriage with agnates or 
cognates, and they require that a man must either marry in his 
own caste, or if he marries out of his caste, it must be into a lower 
caste. But while some authorities so lay down this rule as to allow 
the Brahman to marry into the next two lower castes, the Kshatriya 
and the Vaigya, and the Kshatriya to marry into the Vai9ya caste, 
others also permit marriage with (^udras, and therefore allow 
a Vai9ya to marry into that caste. 

As might be expected, the Brahmana period presents us with 
a stage intermediate between the rules of the Sutras and the laxity 
of the Rigveda. The rule as to marriage within the circle of the 
cognates and agnates seems, by the time of the ^atapatha Brahmana, 
to have extended only to the prohibition of marriage with relations 
of the third or, according to others, of the fourth degree. Similarly 
in the Brahmanas, while we have no reason to doubt that priest- 
hood and nobility were hereditary, these castes seem to have been 
free to intermarry with the lower castes including the ^udra, as 
the cases of Vatsa and Kavasha cited above indicate. The marriage 
of a Brahman with the daughter of a king is attested by the 
case of Sukanya, the daughter of ^aryata, who married the seer, 

The question how far change of caste was possible raises 
difficult problems. The evidence of any change is scanty in the 
extreme. The most that can be said is that it does not seem to 
have been impossible. Thus in the Rigveda, as we have seen, 
Vi9vamitra is a priest, the Purohita of the king Sudas, but in the 
Paiichavim^a and the Aitareya Brahmanas he is treated as of royal 
descent, of the family of the Jalmus. The Panchavim9a Brahmana 
also speaks of certain persons as royal seers ; and the later tradition, 
preserved in the Anukramanl or ' index ' to the composers of the 
Rigveda, ascribes hymns to such royal seers, in some cases at least 
without any real foundation. Yaska, in one instance, represents 
a prince, Devapi, as sacrificing for his brother ^amtanu, the king ; 
but here we can see from the passage of the Rigveda on which his 
narrative is based that he has no warrant for this theory. In the 
Aitareya Brahmana a king, Vi9vantara, sacrifices without his 
priests, the Qyaparnas; but the case has no cogency, for the 
mention of other priests in the context suggests the natural 
inference that he used one or other of these groups. Some kings 
are mentioned in the Panchavim9a Brahmana and elsewhere as 
having been great sacrificers; but this may mean no more than 
that they were the patrons of the sacrifice, the normal part of the 

v] The Four Great Classes 127 

king. We come nearer to contact with fact in the concurrent 
stories of the Upanishads which show kings like Janaka of Videha, 
A9vapati king of the Kekayas in the Punjab, Ajata9atru of Ka9i, 
and Pravahana Jaivali of Panchala disputing with and instructing 
Brahmans in the lore of the brahman, the unity which is the 
reality of the world. Very possibly this attribution is mainly due 
to considerations of the advantage of conciliating the kings who 
were the patrons of the new philosophy ; but, in any case, there is 
no reason to deny that kings could and did take interest in in- 
tellectual movements, and we cannot from such facts infer that 
there was any possibility of interchange of caste : we cannot say 
that, if a king became a seer, as the Jaiminiya Upanishad Brahmana 
asserts in one case, it really meant that he was regarded as ceasing 
to belong to the kingly caste, any more than we can say that, 
if a priest became king, as was not unknowTi later at least, he 
thereby suffered any loss of his priestly position. One case of 
interest remains, that of Satyakama Jabala who was accepted as 
a pupil by a distinguished priest because he showed promise, 
although all he could tell of his ancestry was that he was the son 
of a slave girl; but, evidently, his father might have been a 
Brahman, and the case is only of value as negativing the idea of 
any unnatural rigidity of institutions in the Vedic age. The 
history of later India shows how rigid distinctions might be in 
theory but how ingeniously they might in practice be evaded in 
the individual case. What is more significant, perhaps, is that 
there is no instance recorded in the Vedic texts of a Vai9ya rising 
to the rank of priest or a prince : the two upper hereditary classes 
might to some degree permit closer relations, but they seem to 
have regarded the commoner as definitely beneath them. 

The relations of the four great classes of castes are summed 
up from the point of view of the Brahman in a passage of the 
Aitareya Brahmana^. In that passage the Kshatriya is taken as 
the norm, and the other castes are defined according to the 
relations which they bear to him. 

The Brahman is *a receiver of gifts, a drinker of the Soma, 
a seeker of food, and liable to removal at will.' We can distinguish 
in this period two classes of Brahmans, the priests who, as Purohitas 
of the king or belonging to his entourage, took part in the vast 
sacrifices, some of them lasting for at least a year, which they 
offered for their mastei-s, and the priests of the village who lived 
a humble and more restricted existence, except when they might 

' Vedic Index, vol. ii, p. 256. 


128 Later Samhitas^ Brahmanas^ etc. [ch. 

be called on to serve at the sacrifice instituted by some rich noble 
or merchant. In both cases the priest was, in the long run, at the 
mercy of the political power of the king. To the spiritual claims 
of the Brahmans, so proudly asserted at the ceremony of the royal 
consecration, when the king is announced to the people as their 
king but it is added that the Soma is the king of the Brahmans, 
must be opposed the practical power of the king. 

The Vai9ya is described as ' tributary to another, to be lived on 
by another, and to be oppressed at will.' From the point of view 
of the Kshatriya this indicates the fact that the exactions of the 
king from the commoners of the tribe were limited only by 
practical considerations of expediency: the commoner had no 
legal right to his landholding or to his private property if the king 
decided to take them from him ; and, if he was allowed to retain 
them, he paid for them in tribute and in the duty of supporting 
others. This refers, no doubt, to the king's privilege of assigning 
to his nobles the right to receive food from the common people, 
and thus of making provision for the maintenance of the nobility, 
who assisted him in the protection of the country, and in the 
administration and the conduct of justice. By this means the 
nobles came more and more to occupy the position of landholders 
under the king, while the Vai9yas approximated to the position of 
tenants. Moreover, the nobles may well have received from the 
king, as a result of successful onslaughts on the aborigines, grants 
of conquered lands and slaves, which they would hold in full 
proprietorship, subject to the political authority of the king. 
Among the Vai9yas, again, distinctions were gi'owing up: that 
originally the agriculture was carried on by Aryan tillers is certain ; 
but, in the period of the Brahmanas, the position was changing 
gradually; and, for the peasant working on his own fields, was 
being substituted the landowner cultivating his estate by means of 
slaves, or the merchant carrying on his trade by the same instru- 
mentality, though we cannot with any certainty say how far this 
process was proceeding. The industrial workers, like the chariot 
makers, the smiths, the tanners, the carpenters, were sinking in 
estimation and forming distinct castes of their own. 

On the other hand, the (^iidra was approximating more and 
more to the position to which the humbler freeman was being 
reduced. In the passage referred to, he is still described as * the 
servant of another, to be expelled at will and to be slain at will ' ; 
but in the Siitras we find that, while the Vai9ya has a wergeld of 
100 cows, the ^udra has a wergeld of 10 cows; and, even if we 

v] The Qudras 129 

assume that this is merely for the benefit of his master — which is 
very doubtful — still unquestionably the growing complication of 
the social scheme was abolishing the relation of simple slavery. 
Slaves proper there were, as we see in the Buddhist texts; but, 
where whole tribes were reduced to subjection, the tendency must 
have been to assign villages and their inhabitants to the king and 
to the nobles, sometimes, perhaps, also, though in a less degree, to 
the commoners who at this period must still have formed the bulk 
of the army. While some of the aboriginal inhabitants would 
thus become slaves pure and simple, the rest would rather stand in 
the relationsliip of serfs ; and, as we have seen, there is reason to 
suppose that in many cases the true Vai9yas also were approxi- 
mating to the position of tenants of the nobles. There is an 
interesting parallel in the early history of England, where the 
ordinary freeman gradually fell into feudal dependence on his 
superiors, while the slave as gradually acquired the position of 
a serf, and became more and more assimilated to the position to 
which the freeman had sunk. 

This ambiguous position of the ^udra is amply recognised in 
the Vedic texts: on the one hand, he is emphatically regarded 
as being impure and not fit to take part in the sacrifice: after 
consecration, in some cases, the mere speaking to a 9^dra is 
absolutely forbidden. He was not allowed even to milk the cow 
for the milk needed for the offering to Agni. In the Vajasaneyi 
Samhita illicit connexions between Aryan and Qudra are severely 
reprobated; but, in other places, sin against Arya and ^iidra is 
referred to, prayers are uttered for the glory of Arya and Qudra, 
and we learn of rich Qudras. The Sutras, while they emphasise 
many points not attested by the Brahmana texts, such as the danger 
of sitting near (Qudras, their exclusion from the study of the Veda, 
and the prohibition of eating food touched by them, yet recognise 
that they may be merchants or indeed exercise any trade. 

It seems probable enough that among the Qudras themselves 
there were rules of endogamy; for we may generally assume, in 
the absence of anything to the contrary in the texts, that the 
Vedic Indians and the aborigines alike married within the tribe. 
The Qudras seem often to have been subjugated by whole tribes, 
such as the Baindas, the Parnakas, the Paulkasas, and perhaps the 
Chandalas, who may originally have been members of small and 
degraded tribes living mainly by fisliing or hunting: such tribes 
have survived in the Central Provinces and near the Himalayas 
until the present day, and they must have been much more 

C.H.I. I. • 9 

130 Later Samhitas^ Brahmanas^ etc, [ch, 

numerous in the first millennium B.C. Thus from below as well as 
from above, from the practices of the conquered aborigines as well 
as from the class prejudices of the Aryans, may have come the 
impulse to the development of caste. 

From the political point of view the chief characteristic of the 
new order was the growth in the power of the king. We must 
not assume that, even in this period, there were great kingdoms. 
It is true that the horse-sacrifice as reported in the (^atapatha 
Brahmana and in the royal consecration of the Aitareya Brahmana, 
both of which passages are late, presuppose that the kings who 
performed it set up claims to imperial dignity, and that they had 
won the proud title of 'conquerors of the whole earth,' which is 
applied to them. But real conquest seems not to have been 
meant; and, though the evidence above given proves that there 
was considerable amalgamation of tribes and the formation of 
larger kingdoms than those in the period of the Rigveda, yet it is 
significant that even the Kuru-Panchalas, and still less the Kosala- 
Videhas, never amalgamated into single kingdoms. We may, how- 
ever, safely hold that the king now ruled in many cases a much 
larger realm than the princes of the Rigveda. The hereditary 
character of the monarchy is clearly apparent: in one case, that 
of the Srinjayas, we hear expressly of a monarchy which had lasted 
ten generations. The term Rajaputra, 'son of a king,' is now 
found together with the older Rajanya, which probably covers the 
nobles as well as the king and his family. The importance of the 
kingly rank is emphasised by the elaborate rite of the royal 
consecration, the Rajasiiya. The king is clad in the ceremonial 
garments of his rank, is formally anointed by the priest, steps on 
a tiger skin to attain the power of the tiger, takes part in a mimic 
cattle raid, assumes the bow and arrow, and steps as a conqueror 
to each of the four quarters, an action paralleled in the coronation 
of the Hungarian king. A game of dice is played in which he 
is made the victor. A list of kings who were thus consecrated is 
given in the Aitareya Brahmana: in all but details it coincides 
with the list given in the Qatapatha Brahmana of those who 
performed the horse-sacrifice. 

At the royal consecration the entourage of the king played an 
important part. The list of Ratnins, 'jewels,' given by the Taittiriya 
texts, consists of the Brahman, i.e. the Purohita, the Rajanya, 
the Mahishi, the first wife of the four allowed to the king by 
custom, the Vavata, 'favourite wife,' the Parivrikti, 'discarded 
wife,' the Suta, ' charioteer,' the Senani, ' commander of the army,' 

v] Officials 131 

the Gramani, 'village headman,' the Kshattri, 'chamberlain,' the 
Samgrahitri, ' charioteer ' or ' treasurer,' the Bhagadugha, ' collector 
of taxes ' or ' divider of food,' and the Akshavapa, ' superintendent 
of dicing' or 'thrower of dice.' The ^atapatha Brahmana has 
also the 'huntsman' and the 'courier,' while the Maitrayani 
Samhita adds the Takshan, 'carpenter,' and Rathakara, 'chariot- 
maker.' In an older list of eight Viras, 'heroes,' given in the 
Panchavim9a Brahmana are found the brother, son, Purohita, 
Mahishi, Siita, Gramani, Kshattri, and Samgrahitri. We are faced, 
in the interpretation of the names of several of these officers, with 
the doubt whether we are to recognise in them merely courtiers or 
public functionaries. The Suta is according to native tradition 
the ' charioteer ' ; but it seems much more probable that he was at 
once a herald and a minstrel, and to this conclusion the inviolability, 
which in one passage is attributed to him, clearly points. The 
Gramani has already been met with as a military official in the 
period of the Rigveda. Probably at this epoch a Gramani was, 
both for civil and military purposes, at the head of each village, 
owing, it may be conjectured, his position to the king, while the 
Gramani par excellevice presided over the city or village where the 
royal court was situated. It is also far from unlikely, despite the 
silence of the texts, that the civil functions of the Gramani were 
the more important; for the post is emphatically declared in 
several places to represent the summit of the ambition of the 
Vai9ya. If later analogy is to help us, we may conjecture that 
the Gramani formed the channel through which the royal control 
was exercised and the royal dues received. It may well be then 
that the household officers, besides their more primitive functions, 
carried out the important duties of receiving and disbursing the 
revenues which the king thus obtained; and on them must have 
fallen the duty of seeing that the supplies, which the Vaigyas were 
required to provide for the maintenance of the king's household, 
were duly forthcoming. The condition of these officers is indeed 
probably to be compared with that of the household of the early 
English and Norman kings. 

An officer, not included in the list of the Ratnins but often 
mentioned in the texts of the period, was the Sthapati; and we 
learn that it was the Sthapati of Dushtaritu who restored him to 
the kingdom of the Srinjayas after he had been expelled thence 
by his subjects. He may have been a governor of part of the 
kingdom ; but the more likely interpretation of the term is ' chief 
judge,' an official who doubtless combined executive as well as 


132 Later Samhitas^ Brahmanas^ etc. [cH. 

judicial functions. Later however in the Sutras we hear of a 
Nishada-Sthapati which may mean a 'governor of Nishadas,' 
apparently the ruler of some outlying aboriginal tribes, who had 
been reduced to subjection and placed under the royal control. 

Of the actual functions of the king we hear little detail. He 
still led in war — the Kuru-Pafichala princes sallied forth to raid in 
the dewy season and returned in the hot weather as a matter of 
course — ^but the Senani appears as leader in charge under him. 
From the Sutras and from a stray reference in the (^atapatha 
Brahmana, he seems to have taken a very active part in the ad- 
ministration of the criminal law. There can be no doubt that he 
controlled the land of the tribe. It is not, however, necessary to 
ascribe to this period the conception of the royal ownership of all 
the land, though it appears in the Greek sources from the time of 
Megasthenes downwards, and is evidenced later by the law-books 
of the time. He had, it is true, the right to expel a Brahman or 
a Vai9ya at will, though we do not know expressly that he could 
do this in the case of a Kshatriya. But these considerations point 
to political superiority rather than to ownership proper; and we 
may assume that, when he gave grants of land to his retainers, he 
gi'anted not ownership but privileges such as the right to receive 
dues and maintenance from the cultivators. There is a clear 
distinction between this action and the conferring of ownei*ship; 
and it may be doubted if the actual gift of land was approved in 
this epoch: the only case of which we hear is one reported in 
the ^atapatha and the Aitareya Brahmanas, in which the king 
Vi9vakarman Bhauvana gave land to the priests who sacrificed for 
him, but the Earth itself rebuked his action. It is more probable 
that, at this time, the allotment of land was determined by the 
king or the noble to whom he had granted rights of superiority 
according to customary law, and that gifts not in accordance with 
this law were disapproved. It is hardly necessary to point out the 
close similarity between such a state of affairs and that existing at 
the present day in parts of West Africa, where kings have intro- 
duced for purposes of personal gain the practice of dealing as 
absolute owners with lands, which, according to the strict system 
of tribal law, they had no power to allocate save in accordance 
with the custom of the tribe. Nor is it inconsistent with this view 
that the king had an arbitrary power of removing a subject from 
his land. That power flowed from his sovereignty, and though 
disapproved was acquiesced in, we may presume, just as in West 
Africa ; while the dealing of kings with the land by way of absolute 

v] yudicial Procedure .133 

ownership was regarded as a complete breach of the tribal law, 
the actual removal from his land of any individual was recognised 
as a royal prerogative, even if the power were misused. 

In curious contrast with the comparative wealth of information 
regarding the king, is the silence of our texts on the assembly of 
the people. The samiti or the sabJid is not rarely mentioned in 
these texts ; and we cannot assume that the assembly had lost its 
power, though it may have diminished in importance. Even this, 
however, we cannot absolutely assert; for we hear so often of 
expelled kings that we must believe that the people were far from 
obedient to a yoke which rested on them too heavily. But there 
must have been in the extension of the realm a tendency to 
diminish the possibility of frequent meetings of the samiti, and 
accordingly some diminution in its control over the state. At any 
rate, there are indications, if no conclusive proof, that there was 
growing up within the members of the sahhd a distinction between 
those who attended only at the great meetings and the sabhdsads, 
or 'assessors,' who attended regularly; and it may be that for 
judicial purposes the activity of the sahhd was entrusted to a 
smaller number, the Homeric gerontes, unless indeed we are to 
trace judicial functions to an origin in voluntary arbitration \ 

On judicial matters we learn but little more than in the pre- 
ceding period. Serious crimes like killing an embryo, the murder 
of a Brahman, and the murder of a man occur in lists of sins 
together w ith minor defects, such as the possession of bad nails. 
Other more serious crimes mentioned are stealing gold and drinking 
the surd, while treachery to the king is recognised as a capital 
offence. There are traces of a growing sense of justice in the 
discussions which are recorded in the case of the accidental death 
of a boy through the carelessness of the king and the Purohita, 
who were driving in a chariot. But the procedure in cases of 
crime is still quite uncertain : the king may have presided and the 
tribe or the assessors may have judged ; but for this result we can 
rely only on the fact that the king is said to wield the rod of 
justice, and that in the case of the accidental death of the boy the 
matter is stated to have been referred to the Ikshvakus who decided 
that an expiation was due. In the case of theft in the Chhandogya 
Upanishad we find the axe ordeal applied, apparently under the 

1 Bonner {Classical Philology, vol. vi, pp. 12-36) finds in Homer no criminal law, 
except in the form of the punishment by the whole people of an offender whose wrong- 
doings involved the whole people in danger of reprisals; the function of the king or 
Gerontes he traces in civil cases to voluntary arbitration. It is of interest that Homer 
(p. 32) knows nothing of witnesses ; the Vedic texts likewise seem to ignore them. 

134 Inciter Samhitas^ Brahmanas^ etc, [ch. 

direction of the king; but this is the solitary case of an ordeal 
known in Vedic literature as a part of criminal procedure. In 
the Siitras we hear of the king with his own hand striking a 
confessed thief On the other hand, beside the public organisation 
of criminal justice, there was still the system of private vengeance 
tempered by the wergeld. The Sutras fix the wergeld of the 
Kshatriya at 1000 cows, of the Vaigya at 100, and of the Qiidra at 
10, with a bull over and above for the king, according to the text 
of Baudhayana. This seems to indicate a stage when the royal 
power had extended sufficiently to secure that the wergeld should 
be accepted, and that the insult to the royal peace required the 
appeasement of the king and his reward for his intervention by 
the gift of a bull. The lower position of women is shown by one 
text which assigns in her case only the same wergeld as for a 
^iidra. Unhappily, the texts are so vague that we cannot be 
certain whether the payment in the case of a ^udra was always 
required or whether he might be slain with impunity by his 
master, as the term 'to be slain at pleasure' applied to him in the 
Aitareya Brahmana suggests. 

We have also very little information regarding civil law. The 
use of an ordeal in this connexion is attested only by the case of 
Vatsa who proved his purity of descent, which was assailed, by 
walking unharmed through fire. Presumably, civil cases might be 
decided by the king with assessors ; but this view rests only on the 
analogy of other peoples and on the later practice in India itself. 
We know for certain that a Brahman had preference in his law 
cases ; but whether because it was a moral duty of the witnesses 
to bear testimony in his favour, or for the judges to give judgment 
for him, cannot be decided from the passage of the Taittiriya 
Samhita which records the preference. As regards the substance 
of the law we learn the outlines of the law of succession: a 
father might in his lifetime divide his property among his sons, 
in which case he seems to have had a free hand as to their shares : 
if he grew old and helpless, they themselves might divide it, while 
in the division among the sons on his death the older son received 
the larger share. Women were excluded from the inheritance. 
Similarly, a woman had no property of her own : if her husband 
died, she passed to his family with the inheritance like the Attic 
epiUeros. Her earnings, if any, were the property of husband or 
father. The ^udra seems in law to have been also without capacity 
of owning property in his own right. As in the period of the 
Rigveda, there is no evidence of joint family ownership of any 


v] Position of JVomen : Agriculture 135 

property, even in the case of land, though, as we have seen, land 
at this epoch was not considered a suitable form of gift There is 
a clear reference on the other hand to the allotment of land by 
the Kshatriya, presumably in accordance with the customary law. 
There is no trace of the development of the law of contract: 
much work was doubtless done by slaves or by hereditary craftsmen 
who received customary remuneration from the villagers, not pay- 
ment for each piece of work. 

On the whole, there seems to have been some decline in this 
period in the position of women : as has been seen, in one of the 
Siitra texts her wergeld is assimilated to that of a 9udra and her 
lack of proprietary power must have tended to decrease her 
prestige. The polygamy of the kings is now fully established; 
and, presumably, the practice of the sovereigns was followed by 
the richer of their subjects. In a number of passages in the 
Brahmanas it has been sought to find proof that female morality 
was not highly estimated; but this cannot be established; and it 
is a mistake to suppose that the exposure of female children was 
practised. On the other hand, the preference for sons becomes 
more and more pronounced: *a daughter is a source of misery, 
a son a light in the highest heaven.' Generally speaking, the 
increased complexity of society seems to have been accompanied 
by an increase of crime and moral laxity, as appears from the 
curious litany in the Yajurvedas where Rudra is hailed as the 
protector of every kind of thief and ruffian. 

In agi'iculture and pastoral pursuits progi*ess was doubtless 
made. The plough was large and heavy : we hear of as many as 
twenty-four oxen being harnessed to one : it had a sharp point and 
a smoothed handle. In addition to irrigation, which was known in 
the Rigveda, the use of manure is referred to several times. In 
place of the indeterminate yava of the Rigveda many kinds of 
grain are mentioned, and yava is restricted, in all probability, to 
the sense 'barley.' Among those names are wheat, beans, com, 
sesamum from which oil was extracted, Panicum miliaceum, 
frv/mentoAieum, and italicum, Wrightia antidysenterica, Dolichos 
unifiorus, Ervum hirsutwm, Coix harhata, and various others. 
Rice, both domesticated and wild, was much used. The seasons of 
the different grains are briefly summed up in the Taittiriya 
Samhita : barley, sown no doubt, as at present, in winter, ripened 
in summer : rice, sown in the rains, ripened in autumn : beans and 
sesamum, planted in the time of the summer rains, ripened in the 
winter and the cold season. There were two seasons of harvest 

136 L.ater Samhitas^ Brahmanas^ etc, [CH. 

according to the same authority; and another text tells us that 
the winter crops were ready in March. The farmer had, as now, 
constant troubles to contend with: moles destroyed the seed, 
birds and other creatures injured the young shoots; and both 
drought and excessive rain were to be feared: the Atharvaveda 
provides us with a considerable number of spells to avoid blight 
and secure a good harvest. Cucumbers are alluded to, perhaps as 
cultivated ; but there is no certain reference to tree culture, though 
frequent mention is made of the great Indian trees like the 
A9vattha, the Ficus religiosa, and the Nyagrodha, the Ficus 
indica, and the different forms of the jujube are specially named. 
Even more striking is the great development of industrial life and 
the sub-division of occupations. The list of victims at the symbolical 
human sacrii&ce of the later texts of the Yajurveda provides us 
with a large variety of such occupations; and, after making all 
allowances, it is impossible to doubt that the lists represent a good 
deal of fact. We hear of hunters, of several classes of fishermen, 
of attendants on cattle, of fire-rangers, of ploughers, of charioteers, 
of several classes of attendants, of makers of jewels, basket-makers, 
washermen, rope-makers, dyers, chariot-makers, barbers, weavers, 
slaughterers, workers in gold, cooks, sellers of dried fish, makers of 
bows, gatherers of wood, doorkeepers, smelters, footmen, messengers, 
carvers and seasoners of food, potters, smiths and so forth. Pro- 
fessional acrobats are recorded, and players on drums and flutes. 
Beside the boatman appears the oarsman, and the poleman; but 
there is still no hint of sea-borne commerce or of more than river 
navigation, though we need not suppose that the sea was unknown, 
at least by hearsay, to the end of the period. There is a trace of 
police ofiicials in the Ugras who occur in one passage of the 
Brihadaranyaka Upanishad; and a Gramyavadin or village judge 
appears to have held a court for petty cases in the village. Among 
the priests themselves, we find the sub-division of Chhandogas, the 
singers of Samans, while the Charakas were wandering students, 
a special branch of whom are said to have founded the schools of 
the Black Yajurveda. Moreover, in accordance with the tendency 
to sub-divide and formulate, the life of the priest is now more 
rigidly regulated: he must pass as a preliminary through the 
apprenticeship of being a Brahmacharin. In this stage he is taught 
by a master, for whom in return he does all the necessary work of 
the day and for whom he begs or otherwise provides food. Two 
important features of later village life in India appear in the forms 
of the astrologer and the barber. Of women's work we learn of 

v] The Metals : Social Life 137 

the dyer, the embroiderer, the worker in thorns, and the basket- 
maker. The merchant is often mentioned, and the usurer has 
a special name : it is of interest that the term ^reshthin several 
times occurs, denoting at least a wealthy merchant, and possibly 
already the word has its later technical sense of the head of a 
merchant gild. 

The advance of civilisation is seen also in the more extended 
knowledge of the metals : as compared with the gold and the ayas, 
of doubtful meaning, of the Rigveda, this period knows tin, lead, 
and silver of which ornamented bowls are made, while ay as is 
diflPerentiated as red ayas, presumably copper, and dark or black 
ay as, which must be iron. Another sign ofthe ne w era isjhe defini te 
references to the keeping of tamedephants, the guarding of elephants 
being one of the occupations occurring in the Yajurveda texts. But 
there is no hint that the elephant was yet used for war as it was 
already in the time of Ctesias. The use of horses for riding had 
certainly become more common ; but no clear reference is made to 
the employment of cavalry in war, though that was usual by the 
time of Alexander's invasion. 

Little change can be traced in the social life of the time. The 
use of houses of wood continued; and, as a result, we have not 
a single relic remaining of the architecture of the period. Nor 
have we any coins: it is not probable, indeed, that a regular 
coinage had begun, though the path to this development was 
already opened by the use of the Tcrishnala, the berry of the Ahrus 
preeatorius, as a unit of weight. We hear in the Brahmanas of 
the ^atamdna, a piece of gold in weight equivalent to a hundred 
krishnalas, and such pieces of gold were clearly more or less 
equivalent to currency and must have been used freely by the 
merchants, of whose activities we hear so little in the sacred texts. 
The nishka, originally a gold ornament, was also at this time a unit 
of value ; and the cow as a unit was probably in course of super- 
session. The style of clothing seems to have continued unchanged, 
though we hear more of the details : among other things we are 
told of woollen garments, robes dyed with saffron, and silk raiment. 
The food of the Indian remained unaltered: the eating of meat 
is, indeed, here and there censured, as for instance in a hymn of 
the Atharvaveda where meat eating is classed with the drinking of 
the sura as a sinful act, and meat might be avoided like other 
things by one who was keeping a vow. But it was still the custom 
to slay a great ox or goat for the entertainment of a guest, and 
the great sage Yajnavalkya ate meat of milch cows and oxen, 

138 Later Samhitas^ Brahmanas^ etc, [ch. 

provided that the flesh was amscda, a word of doubtful import, 
rendered either 'firm' or 'tender' by various authorities. The 
doctrine of ahimsd, which forbids the doing of injury to any 
animal, was indeed only in embryo in this period, and was not 
fiilly developed until the growth of the belief in transmigration 
came to strengthen the philosophic tenets of the Brahmanas as to 
the unity of all existence. The amusements of the day were, as 
in the period of the Rigveda, the chariot race, dicing, of which we 
have several elaborate but not very clear accounts, and dancing. 
The term (^ailusha appears in the list of victims at the human 
sacrifice, and the sense ' actor ' has been seen in it Taken in con- 
junction with the dozen or so of hymns which show a dialogue form 
it has been supposed to indicate that the Rigveda knew of a ritual 
drama, the direct precursor of the drama of later India. But the 
evidence adduced is insufiicient to bear the strain of the hypothesis K 
In one respect there seems to have been a distinct retrogression 
since the age of the Rigveda. In that Samhita there is frequent 
mention of the physician's skill, and wonderful deeds are ascribed 
to the A9vin8 as healers of diseases. As early as the Yajurveda 
Samhitas, however, the physician appears to be held in less esteem : 
the A9vins were said to have made themselves inferior to the other 
gods by their practice of medicine, by which they made themselves 
too familiar with all sorts of people. The Atharvaveda contains 
much which gives a sad picture of the medical practice of the 
day: against the numerous diseases which it mentions it had 
nothing better to oppose than the use of herbs and water ac- 
companied by strange spells, based on sympathetic magic. The 
number of diseases recorded by difiering names is large : the most 
frequent was fever, no doubt the malaria which still haunts India ; 
and others mentioned are consumption, haemorrhoids, abscesses, 
scrofula, dysentery, boils, swellings, tumours on the neck, con- 
vulsions, ulcers, scab, rheumatism, tearing pains, headache, leprosy, 
jaundice, cramp, senility, and others less easy to identify. Various 
eye diseases were known; and the use of a sand bag to stop 
bleeding is recorded. The dissection of the animal victims at the 
sacrifices gave the opportunity to acquire knowledge of the bones 
of the body^, but on the whole the facts recorded, especially in the 
Atharvaveda and the ^atapatha Brahmana, give us no very elevated 
opinion of the accuracy of the Vedic physician in this regard. 

^ See von Schroeder, Mysterium und Mimus im Rigveda, Leipzig, 1908; Hertel, 
V.O.J. , vol. xvin, pp. 69 sq., 137 sq., xxni, 273 sq., xxiv, 117 Bq. ; Winternitz, V.O.J. , 
vol. xxni, pp. 102 sq.; Keith, J.R.A.S., 1911, pp. 979-1009. 

* See Hoemle, Osteology, Oxford, 1907. 

v] Astronomy 139 

On the other hand, a distinct advance was unquestionably made 
in regard to astronomical knowledge. The Rigveda knows only, so 
far as we can see, the year of 360 days divided into twelve months 
of thirty days each, which is six days longer than the synodic lunar 
year, and nearly five and a quarter days too short for the solar 
year. To bring the year into something like order, intercalation 
seems to have been attempted quite early : we hear in a riddle hymn 
of the Rigveda (i, 164) of the intercalary month, the thirteenth. In 
the Samhitas the system is slightly more developed ; and possibly 
some efibrts were being made to arrange intercalation in a cycle 
of five years in such a manner that the years and the seasons 
would be made to coincide ; but it is fairly clear that a satisfactory 
method had not yet been obtained. The Samhitas, however, give 
us the names of the twelve months arranged very artificially in six 
seasons, and they introduce to us the important doctrine of the 
Nakshatras, or ' lunar mansions,' groups of stars selected as roughly 
indicating the parts of the sky in which the moon appeared in the 
course of a periodic month of 27-28 days. In the Rigveda the 
term Nakshatra seems usually to mean no more than ' star ' ; and 
it is only in the admittedly late marriage hymn (x, 85) that the 
names of two of the Nakshatras proper are found though in 
altered forms. The number of the Nakshatras is variously given as 
twenty-seven in the Taittiriya Samhita and the Kathaka lists and 
usually later, and as twenty-eight in the lists of the Maitrayani 
Samhita and the Atharvaveda. As the periodic month has between 
27 and 28 days, the variation may be primitive: of the allied 
systems the Chinese Sieou and the Arabic Manazil have twenty- 
eight: the missing star Abhijit in the smaller enumeration may 
have fallen out for a variety of causes; and it seems easier to 
assume this than to regard it as a later addition. The use of the 
Nakshatras offered a simple and efiective means of fixing dates by 
the conjunction of the new or fiiU moon with a particular Na- 
kshatra, and in the Brahmana period a further step was taken: 
on some arbitrary basis which we cannot now determine, twelve of 
the Nakshatra names in adjectival form were chosen to represent 
the months. It might have been expected that the months repre- 
sented by these names would be lunar, but they are, as a matter 
of fact, the twelve months of the traditional year of 360 days. 
The whole series of the new names is not found until the Sutra 
period ; but the vitality of the new system is adequately proved by 
the fact that the old series of twelve given in the Samhitas correspond- 
ing to the six seasons is practically ignored in the later literature. 

I40 Later Samhitas^ Brahmanas^ etc, [ch. 

The origin of the Nakshatras has formed the subject of most 
lively controversy: it is clear that the Vedic Indians knew very 
little about astronomy, for it is extremely doubtful whether the 
planets were known at all in the Brahmana period. But it is not 
impossible that, even at this epoch, the Nakshatras could have 
been discovered, for the achievement is a rude one. The question 
is, however, complicated by the existence of the Arabian Manazil 
and the Chinese Sieou. The Manazil are better chosen as lunar 
mansions than the Indian Nakshatras: borrowing on the part of 
India from Arabia cannot be proved in view of the late date of 
the Arabian evidence, while the superiority of the Arabian system 
seems to make it improbable that it should have been derived from 
India. The Chinese evidence is early enough to allow of borrowing ; 
and the dependence of India on China has been maintained by 
Biot and de Saussure ; but the difficulties in the way of this view 
are really insuperable. It remains therefore as the most plausible 
view that the Nakshatras are derived from Babylon, though direct 
proof of the existence of the Nakshatras there has yet to be 

Compared with the case of the Nakshatras there is little other 
evidence of the contact of India with other civilisations in this 
period. In the ^atapatha Brahmana for the first time there 
appears the legend of the flood and the saving of Manu by a great 
fish; and it is most unlikely that we are to see here any re- 
miniscence of the former Aryan home and the crossing of the 
Hindu Kush\ It is therefore possible that the legend may be of 
Semitic origin; but, if so, as usual the Indians have completely 
appropriated the motive, so that the borrowing cannot be proved. 
It has been suggested^ that the knowledge of iron was derived 
from Babylon ; but this is merely a conjecture which has at present 
no support in evidence. A sea-borne commerce with Babylon 
cannot be proved for this epoch either by the evidence of Vedic 
literature or by the references in the Book of Kings to apes and 
peacocks by names which are believed to have had an Indian 
origin. The history of the alphabet has been used by Biihler^ to 

^ This is held by Weber, Indische Studien, vol. i, pp. 163 sq. ; see Muir, Original 
Sanskrit Texts, vol. n2, p. 323. 

^ See Vincent Smith, Indian Antiquat-y, vol. xxxiv, p. 229; Imperial Gazetteer, 
vol. II, p. 98. 

8 Indische Palaeographie, pp. 17 sq. Biihler relied on references to sea trade in the 
Siitras (Baudhayana, i, 2, 4 ; ii, 2, 2 ; Gautama, x, 33) and in the Jatakas and believed 
these to be authorities for the sixth century B.C.; see Indian Studies, no. in, pp. 15 sq. 
But neither Sutras nor Jatakas can be relied on for information regarding so early a 

v] Writing: Religion 141 

show that it was borrowed by traders fi-om a South Semitic source 
via Mesopotamia about 800 B.C.; but we cannot lay any stress 
upon this date. It seems, indeed, most probable that writing was 
introduced by traders and that it was only gradually adopted into 
its proper foiTa for the expression of the Sanskrit language. At 
what date this took place is not really susceptible of proof : there 
is no certain reference to writing in the literature of a date earlier 
than the fourth century RC; and the real development of writing 
belongs in all likelihood to the fifth century RC. It was the end 
of the sixth century that saw the invasion of Darius and the 
annexation of the territory round the Indus; and, prior to that 
event, there is no strong evidence of a really active contact between 
India and the outer world. It is, indeed, probable enough that 
even before the time of Darius, Cyrus had relations with the tribes 
on the right bank of the Indus, and Arrian asserts that the 
Assakenoi and the Astakenoi were subject to the Assyrian kings ^ ; 
but everything points to the fact that, in the period of the 
Brahmanas, relations with the Gandharas and other tribes in the 
remote north-west were very slight. It is also significant that 
there is no really certain case of an inscription of any sort in India 
before the third century b.c. ^ 

The development in religion and philosophy in the period is 
remarkable. The ritual has grown to very large proportions ; and 
with the ritual the number of the priests required at a sacrifice 
had increased until sixteen or seventeen are enumerated as taking 
part in the more important offerings. The mere offerings of 
vegetable food and milk are comparatively unimportant; but the 
animal sacrifice is increasingly elaborated, and the Soma sacrifice 
has developed largely. In addition to the simplest form of the 
Soma sacrifice occupying one day, there are innumerable other 
forms culminating in the Sattras which might last any time from 
twelve days to a year or years. It is significant that, at the bottom 
of this priestly elaboration, is much really popular religion. Thus 
the Rajasuya, or royal consecration, is fundamentally a popular 
rite for the anointing of the king : the Vajapeya betrays a popular 
origin in the prominence in it of a chariot race, once probably the 
main element; the Gavamayana, a Sattra lasting a year, is dis- 
tinguished by the ritual of the Mahavrata day in which long since 
was recognised a primitive performance celebrating the winter 

' See Dufif, Chronnloyy of India, p. 5; Arrian, Indica, i, 3 (trans. M'Crindle, 
p. 179). 

'^ Vincent Smith, Early History of India^, p. 16. 

142 Later Samhitas^ Brahmaijas^ etc, [ch. 

solstice. The horse-sacrifice is at bottom the elaboration of a 
simple rite of sympathetic magic; but it has been so elaborated 
as to combine everything which could make an appeal to the 
warrior Indian king and induce him to distribute abundant largesse 
on the celebrators. But beside these and other popular festivals, 
which the priests have worked over, stands one of the highest 
interest to the priest, which seems to reflect a new conception of 
theology. It is the building of the altar for the sacred fire; in 
one sense no doubt this was an ancient and simple rite, accompanied 
as so often by the slaying of a man in order to secure the abiding 
character of the structure: the Brahmana texts avoid requiring 
any such actual slaughter, though they record it as a deed of the 
past; but they elaborate the building out of all reason and utility. 
The only explanation of this action must be that offered by 
Eggeling^, that, in the building up of the fire altar, the Brahmans 
sought to symbolise the constitution of the unity of the universe. 
As we have seen, in the Purusha hymn of the Rigveda occurs the 
conception of the creation of the universe from the Purusha, and 
in the theology of the Brahmanas the Purusha is identified with 
Prajapati, 'lord of creatures,' and the sacrifice is conceived as 
constantly recurring in order to maintain the existence of the 
universe. To render this possible is the end of the fire altar, the 
building of which is the reconstruction of the universe in the 
shape of Prajapati. Prajapati, again, is identified with Agni, the 
fire of the altar, and both Prajapati and Agni are the divine 
counterparts of the human sacrificer. But Prajapati is himself 
Time, and Time is in the long run death, so that the sacrificer 
himself becomes death, and by that act rises superior to death, 
and is for ever removed from the world of illusion and trouble 
to the world of everlasting bliss. In this the true nature of 
Prajapati and of the sacrificer is revealed as intelligence, and the 
^atapatha Brahmana urges the seeker for truth to meditate upon 
the self, made up of intelligence and endowed with a body of spirit, 
a form of light and an ethereal nature. 

The same doctrine appears in another form in the Upanishads 
which are engaged with the discussion of the underlying reality. 
They agree in this that all reality in the ultimate issue must be 
reduced to one, called variously hrahman, *the holy power,' or 
atman, 'the self.' Moreover, the Upanishads agree in regarding 
the absolute to be unknowable, and though they ascribe to it 
intelligence they deprive that term of meaning by emptying it of 

1 See Sacred Books of the East, voL XLin, pp. xiv-xxiv. 

v] The Doctrine of Transmigration 143 

all thought. If the real is the absolute alone, the existence of the 
appearance of this world must be explained ; but naturally enough 
the Upanishads do not successfully attempt this task ; and it was not 
until the time of ^ankaracharya in the beginning of the ninth century 
A.D. that it was found possible to reconcile the doctrines of the 
different texts by the view that all existence is merely illusion. This 
is perhaps a logical development of the doctrine of the Upanishads ; 
but the Upanishads were groping after truth and did not attempt to 
deduce all the consequences of their guesses at the nature of reality. 
There was one consequence which followed so clearly from the 
new conception of existence that it is enunciated, though not very 
decidedly, in the Brihadaranyaka Upanishad, namely that there 
was no consciousness after death in the case of him who realised 
the true nature of the self as intelligence without thought. But 
this conception plays a very small part in the texts compared with 
the new theory of transmigration. There is no real sign of this 
doctrine in the Brahmanas proper, but there is a certain amount 
of preparation for its appearance in the gradual development of 
the doctrine that not even after death is the horror of death 
ended : a man may die repeated deaths in the next world. If this 
conception be transferred to the present world, then the doctrine 
of transmigration is produced, and in the Upanishads this doctrine 
is clearly and expressly enunciated. The Chhandogya and Bri- 
hadaranyaka agree in the main outlines of the new belief: the 
forest ascetic who has realised the nature of brahman after death 
goes by the way of the gods to be absorbed in brahman and never 
again to be bom: the man who has done good deeds but has not 
attained the saving knowledge goes to the world of the moon to 
reside there until the fruit of his deeds is exhausted, when he is 
bom again first as a plant and then as man or at once as a man : 
the wicked on the contrary are born as outcasts, dogs, or swine, 
according to the Chhandogya, as birds, beasts, and reptiles ac- 
cording to the Brihadaranyaka. There is a variant version on the 
Kaushitaki which makes all first go to the moon ; but the essential 
point is the acceptance as a matter of certainty of the new doctrine 
of transmigration. The Brihadaranyaka also has an important 
addition to the doctrine in the form of the gospel of harman 
* action,' which determines on a man's death the nature of his next 
birth. In the Buddhist view the idea recurs in the simple form 
that the self, which is recognised as persisting through trans- 
migration by the Brahman, is discarded as needless and the Tmrman 
alone is asserted to possess reality. 

144 Later Samhitas^ Brahmanas^ etc, [ch. 

The origin of this doctrine may have been helped by the widely 
prevalent view among tribes of animists that the souls on death or 
even in life can pass into other forms, animal or vegetable. We 
have seen that in the Rigveda in one hymn the soul is regarded as 
going to the waters or the plants ; and we have no reason to doubt 
that such ideas were prevalent among the aboriginal tribes with 
which the Aryans mixed. But these vague ideas are totally 
inadequate to account for the belief in transmigration, and the 
theory must, it would seem, have been a discovery of the schools 
of seekers after the nature of truth, who arrived at it on the 
one side fi-om the popular beliefs of the peoples among whom they 
lived, and on the other from the conception of the Brahmanas 
that death could be repeated in the other world. The doctrine 
led directly to j^essimism, but the Upanishads are not themselves 
pessimistic; and we obtain thus a valuable evidence of their 
priority to the rise of Buddhism, which is saturated with the 
doctrine of the misery of the universe. The extraordinary success 
of the doctrine shows that it was in harmony with the spirit of the 
Indian people, and suggests what is otherwise probable, that by the 
end of the period of the Brahmanas the influence of the Aryan 
strain was waning, and that the true Indian character of the 
intellectual classes was definitely formed. 

As we have already seen, the tradition makes kings take part in 
the discussions which marked the formation of the doctrine of the 
absolute, and even hints that the doctrine was in some way a special 
tenet of the ruling class ; but it is doubtfiil if we can accord full 
credit to this tradition, or believe that the hrahman doctrine was 
the reaction of the noble class against the excessive devotion of 
the priests to the rituaP. Policy adequately explains the part 
assigned to them by the Brahmans, whose aim it was to make their 
patrons appreciate that their researches were such as to deserve 
support. Parallel with the development of philosophy there was 
proceeding the movement which leads to the religions of modem 
India, the exaltation of Rudra and in a minor degree of Vishnu to 
the position of a great god. Prajapati is indeed the main subject 
of the theosophical speculation of the Brahmana texts, a purpose 
to which his name as * lord of creatures ' especially lent itself ; but 
Prajapati had no claims to be a god of the people, and the position 

1 The tradition is accepted by Garbe, Beitrage zur indischen Kulturgeschichte, 
pp. 1 sq. ; Denssen, Philosophy of the Upanishads, pp. 17 sq. ; Ehys Davids, Buddhist 
India, pp. 256-7. See also Chapter xi, pp. 264-5. Its validity is doubted by Bloom- 
field, Religion of the Veda, pp. 218 sq.; Oldenberg, Buddha^, p. 73; Keith, J.R.A.S., 
1908, pp. 868-72. 

v] Rudra and Vishnu 145 

of Rudra as a popular deity is sufficiently shown by the litanies to 
him in the Samhitas of the Yajurveda, and by the whole outlook of 
such texts as the Aitareya, Kaushitaki, and ^atapatha Brahmanas. 
When Prajapati committed incest with his daughter, the Aitareya 
tells us that the gods were wroth, and from their most dread forms 
produced the god Bhutapati, 'lord of creatures,' who represents 
one aspect of Rudra's activities. He pierced Prajapati and thereby 
acquired his dominion over all cattle. In another passage the 
wording of a Rigvedic verse is altered to avoid the mention of 
Rudra's dread name : in yet another he appears at the sacrifice in 
black raiment and appropriates to himself the sacrificial victim. 
We need not suppose that in this presentation the Brahmanas were 
creating a new figure : rather they were adapting to their system, 
as far as they could, a great god of the people. But the Rudra of 
this period can hardly be regarded as a mere development of the 
Rudra of the Rigveda : it seems most probable that with the Vedic 
Rudra is amalgamated an aboriginal god of vegetation, closely 
connected with pastoral life. 

Vishnu cannot be said to have won any such assured place as 
Rudra, who is already hailed as the ' great god ' par excellence, and 
already bears the name of ^iva, 'propitious,' which is to be his 
final appellation. But the constant identification of Vishnu and 
the sacrifice is, in view of the extraordinary importance attached 
to the sacrifice by the Brahmans, a sure sign that he counted for 
much in Vedic life, and that he shared with Rudra the veneration 
of the people, who may in different localities have been the followers 
of one or the other god respectively. For the rest, while we now 
obtain many details of the lower side of the religion in the spells 
of the Atharvaveda, the pantheon of the Rigveda remains unaltered 
save in such minor aspects as the new prominence of the Apsarasas, 
the mechanical opposition of the gods and the Asuras, and the rise of 
snake worship, which seems to have been due to the imitation of 
the aboriginal tribes. On the other hand, the attitude of the 
priests to the gods as revealed in the sacrifice has lost whatever it 
had of spontaneity and simple piety. It is no doubt possible to 
exaggerate these qualities even in the earlier hymns of the Rigveda ; 
but their absence in the later Samhitas is unquestionable. The 
theory of sacrifice is bluntly do ut des; and even in that theory r ^J^jUi 
the sacrificers had so little trust that the whole sacrificial apparatus 
is dominated by sympathetic magic. So convinced is the priest of 
his powers in this regard that the texts explain that he can ruin as 
he pleases, by errors in the sacrifice deliberately committed, the 

C.H.I. J. 10 

146 Later Samhitas^ Brahmanas^ etc. [ch. 

patron for whom he is acting, and in whose interest he is presumed 
to be at work. It is a sordid picture ; and, as we have seen, the 
higher spirits turned away from a hocus pocus, which they must 
have despised as heartily as any Buddhist, to the interpretation of 
the reality underlying phenomena. Yet it is characteristic of the 
Indian genius that, though it evolved views which must have 
rendered all the sacrificial technique logically of no avail, it made 
no efibrt to break with the sacrifice which was allowed to stand as 
a preliminary towards the attainment of that enlightenment which 
the priests professed to impart. 

The language of the Samhitas in their verse portions is similar 
to that of the Rigveda, especially in the tenth book and in the 
later additions to the other books. The language of the prose 
represents the speech of the Brahman schools of the day : it difiers 
from that of the verse by the removal of abnormalities, and by 
much greater precision shown, for example, in the exact use of the 
tenses, the 'narrative perfect' being at first carefully eschewed, 
and by the disappearance, except in a narrow sphere, of the use of 
the unaugmented past tenses of the verb with modal meaning. 
There seems in one passage of the (^atapatha Brahmana to be 
a curious admission that other tribes had not preserved the purity 
of the Vedic speech : the Asuras are credited in that text with the 
utterance of the words he 'lavo, which may be interpreted he rayah, 
* Ho ! ye foes ! ,' and, if so, can be explained as Prakrit forms. 
Similarly, as we have already seen, the Vratyas are described as 
regarding the Vedic speech as difficult to pronounce, no doubt 
because of its conjunct consonants which the Prakrits avoid. In 
both cases the reference is probably to tribes of the Magadha 
country, and the Magadhi Prakrit is marked by both the points 
alluded to\ There are also signs of this corruption of the language 
through the contact with the aborigines in the fact that in the 
spells of the Atharvaveda are found several forms which can only 
be accounted for as Pi-akritisms. Beyond these generalities we 
cannot afiect to estimate how far the process of the transformation 
of the language in the popular speech had gone: the earliest 
foreign evidence, that from the Greek records, shows that many 
names were reported by Megasthenes and others in Prakrit form ; 
and, in the middle of the third century B.C., the inscriptions of 
A9oka are all written in Prakrit dialects varying considerably in 
detail from one another. It is therefore reasonable to suppose 

* Grierson, Z.D.M.G., vol. lxvi, p. 66, thinks that Pai9achl, a dialect of north-west 
India, is meant ; but see Vedic Index, vol. 11, p. 517. 

v] Criteria of Date 147 

that, beside the language of the Brahman schools, there existed 
more popular forms of speech; but everything points to the fact 
that tlie deeds of princes were still sung in a language of the same 
form as the priestly speech. In metre a significant change can be 
seen : the later hymns exhibit, when written in the eight syllable 
metre, a distinct tendency to be composed of stanzas in which the 
four lines are no longer independent in structure, but the first and 
third and the second and fourth respectively are assimilated. Tlie 
latter pair is made to end Avith a definite iambic cadence, while the 
first and third on the contrary are made to end with an iambus 
followed by a trochee, thus producing an efiect of contrast and 
setting a gulf between the old and the new form of versification. 
This new form is far from being exclusively employed even in the 
latest versification of the period, but in the epic it is firmly 
established, and the variants reduced to narrow limits^. 

Interesting as are the Samhitas and the Brahmanas from the 
point of view of the history of civilisation and religion, as literature 
they are hardly ever of substantial value. Much of the speculation 
of the Brahmanas is utterly puerile and seems to be the product of 
a decadent intellect. On the other hand, the real interest of the 
Upanishads is undeniable : these primitive philosophical fragments 
exhibit a genuine spirit of enquiry, and here and there do not fail 
to rise to real dignity and impressiveness. 

For the date of the epoch of the Brahmanas we are again 
thrown back on those considerations of literary and social develop- 
ment which we have found to be the sole trustworthy criteria for the 
dating of the epoch of the Rigveda. The lower limit is given by 
the fact that Buddhism accepts from the Upanishads the doctrines 
of transmigration and pessimism, the latter of which had been 
developed as a doctrine of obvious validity from the facts of trans- 
migration. Other indications, such as the want of any trace of the 
knowledge of writing, show that we cannot legitimately carry the 
Upanishads of the older type later than 550 or perhaps more probably 
600 RC. The fixing of the language which is posterior to the Brah- 
manas may be dated at latest at 300 B.C. ; and the earlier Sutras 
probably go back to at least 400 B.C. and very possibly earlier. 
These are important considerations and their cumulative eflect is 
harmonious and practically decisive of an early date for the 

' See Oldenberg, Z.B.M.G., vol. xxxvn, pp. 67 sq. ; Sacred Books of the East, 
vol. XXX, pp. XXXV sq. ; G.G.N., 1909, pp. 219 sq.; Hopkins, Great Epic of India, 
pp. 194 sq.; Jacobi, Indische Studien, vol. xvn, pp. 442 sq.; Keith, J.R.A.S., 1906, 
pp. 1-10; 1912, pp. 757 sq. 


148 Later Samhitas^ Brahmanas^ etc, [ch. 

civilisation which has been described. On considerations of probable 
development, the beginning of the Brahmana period may fairly be 
put back to 800 B.C. 

As with the Rigveda, attempts have been made to show that 
these dates are much too low and that astronomical data enable 
us to carry the Brahmanas much further back. The lists of the 
Nakshatras all begin with Krittikas, and we know that in the 
sixth century a.d. the constellation which then headed the Nak- 
shatras was chosen because the vernal equinox took place when 
the sun was in conjunction with that Nakshatra. From the pre- 
cession of the equinoxes, we are enabled to arrive at the conclusion 
that the position of Krittikas at the vernal equinox must have 
taken place in the third millennium B.c. This has been supported 
by a passage in the ^atapatha Brahmana where it is said that 
Krittikas did not move from the eastern quarter at that time. 
But we have no evidence whatever to connect the sun and the 
Nakshatras at this period, and the notice regarding the position of 
Krittikas cannot be taken seriously in a work which shows so little 
power of scientific observation of facts as the ^atapatha. More- 
over if, as it is probable, the Nakshatra system was borrowed ready 
made, we cannot even conjecture for what reason Krittikas was 
placed first. More promising is a definite notice contained in the 
Kaushitaki Brahmana and repeated in the Jyotisha, a late Vedic 
work on astronomy, if indeed it can be dignified with this title, 
that the winter solstice took place at the new moon in Maghas. 
From this datum results varying from 1391-1181 B.C. were early 
deduced by different investigators ; but these conclusions can claim 
no scientific value, as they rest on assumptions as to the exact 
meaning of the passage which cannot be justified. The possible 
margin of error in the calculations is at least five hundred years ; and 
we are therefore reduced to the view that this evidence only indicates 
that the observation which is recorded was made some centuries 
B.C. The same conclusion can be drawn from the fact that in 
quite a number of places the month Phalguna is called the beginning 
of the year. In the view of Jacobi, this shows that the year 
began with the winter solstice at full moon in PhalgunI, and thus 
would correspond with his view that in the Rigveda the sun at the 
summer solstice was in Uttara-Phalguni. But, in this case also, 
the result is unacceptable; for it is nowhere stated that the 
beginning of the year was dated from the winter solstice. The 
most probable explanation is that the full moon in Phalguni was 
deemed to be the beginning of the year, because it marked, at 

v] Astronomical evidence 149 

the time when it was so termed, the beginning of spring. Since 
the new moon in Magha was at the winter solstice, the fiill 
moon in PhalgunI would fall about a month and a half later in the 
first week of February, which is compatible with Feb. 7, the Yeris 
initiv/m in the Roman calendar, and which is a X)erfectly possible 
date for about 800 b,c., especially when it is remembered that the 
division of the year into three periods of four months was always 
a rough one, and the beginning of spring had to be placed early so 
as to allow of the rains, which are definitely marked out by the fall 
of the first rain, to fill the period from about June 7 to October 7. 
With this explanation the theory, that the mention of the full 
moon in PhalgunI as the beginning of the year records an observa- 
tion of the fourth millennium B.C., disappears, and still more the 
theory that the mention of the month Caitra as the beginning of 
the year carries us back to the sixth millennium. Nor can any 
more trust be put in the argument that the mention in the late 
marriage ritual of the Dhruva, a fixed star shown to the bride and 
bridegroom as a symbol of constancy, points to an observation made 
at a period when there was a real fixed pole star, i.e. in the third 
millennium B.C. We do not even know whether this part of the 
rite goes back to the period of the Brahmanas ; and, even if it did, 
for so little scientific a purpose there was no need of anything save 
a fairly bright star not too distant from the pole. Ingenious there- 
fore as all these arguments are, they must be dismissed as affording 
no real certainty of correctness. The most that can be said is 
that they tend to support the period 800-600 B.C. as a reasonable 
date for the period of the civilisation of the Brahmanas \ 

1 The main supporters of the astronomical arguments are Jaoobi, Z.D.M.G., 
vol. XLix, pp. 218 sq. ; l, pp. 69 sq. ; J.R.A.S., 1909, pp. 721-6; 1910, pp. 460-4; 
Tilak, Orion, Bombay, 1893; The Arctic Home in the Vedas, Bombay, 1903. On the 
other side, see Oldenberg, Z.D.M.G., vol. xlviii, pp. 629 sq.; xlix, pp. 470 sq. ; 
L, pp. 450 sq.; J.R.A.S., 1909, pp. 1095 sq.; Thibaut, Indian Antiquary, vol. xxiv, 
pp. 85 sq. ; Whitney, J.A.O.S., vol. xvi, pp. Ixxxii sq. ; Keith, J.R.A.S., 1909, 
pp. 1100 sq. ; 1910, pp. 464-6. On the origin of the Nakshatras, see de Saussure, 
T'oungPao, 1909, pp. 121 sq., 255 sq.; Oldenberg, G.G.N., 1909, pp. 544 sq. 



The later half of the sixth century b.c. seems to have been 
unusually fertile in giving rise to new religious movements in India. 
An old text amongst the sacred lore of the Buddhists^ mentions 
sixty-three different philosophical schools — probably all of them 
non-Brahman — existing at the time of Buddha, and there are 
passages in Jain literature exhibiting a far larger number of such 
heretical doctrines. Although these statements may have been 
influenced by the tendency to exaggerate which is visible in most 
Hindu works, and although many of these sects may have been 
distinguished only by very subtle difierences in matters of doctrine 
and practice, we are still bound to belieye that there was an 
extraordinary impulse shown in the rise and development of new 
theological and philosophical ideas at that time. It is beyond our 
power of investigation to determine whether some of these schools 
may not have owed their origin to a time far more remote than 
that of Buddha. In the few cases where we are in some degree 
able to form an opinion on such points — and the history of the 
Jain doctrine gives us some hints in this direction — it seems most 
probable that this may have been the case. It is certainly diiRcult 
to believe that all these sects should have originated at the same 
time. We may therefore suggest that revolts against the Brah- 
man doctrines date from a much more remote age than the time 
^N*^" of Gautama Buddha, the founder of one of the most important 

religions of the world, and Vardhamana Mahavira, the founder or 
rather reformer of the Jain church. Not only these two religious 
< ,v«."*^ teachers but also a number of others, of whom we know little or 

nothing more than the name, preached in a spirit of most con- 
scientious and determined contradiction against the sanctity of the 
Vedic lore, the sacrificial prescriptions of the ritualists, and the 
claims of spiritual superiority asserted by the Brahmans ; but it is 
a strange characteristic of these sects, so far as we know them, that 

1 Cp. S.B.E., vol. X : 2, p. 93. 


CH. vi] Brahmans and yains 151 

they adopted in their ascetic practices and in their whole mode of 
life the rules which had been already fixed by their Brahman 

In the later law books the life of a Hindu is theoretically 
divided into four successive stages, viz. those of hrahmachdrin or 
student of the sacred lore, grihastha or householder, vdnaprastha. 
or anchorite, and parivrdjaka or wandering mendicant. Now 
there are no express statements in Vedic, or pre-Buddhist, texts, 
concerning the existence of this theory in older times ; but from 
certain passages in the principal Upanishads we may infer that 
at least the germs of this institution existed at a comparatively 
early period, as in them we find the knower of the dtman or 
' Supreme Soul,' that is to say, the parivrdjaka or Brahman ascetic 
contrasted with students, sacrificers and anchorites ^ However, 
the order of the difierent stages — with, the exception of that of a 
hrahmachdrin, which is always the first — seems not at that time 
to have been a fixed one, and it may be doubted if this theory was 
ever on a gi'eat scale adopted in real life in India. But this 
question is for us of no importance, as we have here only to take 
notice of the fourth stage, that of the Brahman ascetic, whose life 
was, no doubt, the standard for the rules of discipline laid down by 
Mahavira for his followers. 

The Artha^dstra or * Manual of Politics ' which may possibly be 
the real work of Chanakya or Kautilya, and therefore written 
about 300 B.c.^, describes in the following words the life of a 
parivrdjaka : ' (the duties) of an ascetic (consist in) subduing his 
senses, withdrawal from worldly things and from communication 
with people, begging for alms, living in the forests, but not in the 
same place, cleanliness external and internal, abstinence from 
injuiy to living beings, and in sincerity, purity, freedom from envy, 
in kindness and in patience^.' These general rules could — perhaps 
with one slight alteration — as well be found in any Jain work, and 
in fact we do find them in many passages of the Jain canon, 
although perhaps not exactly in the same words. But the 
similarity between the life of a Brahman and a Jain ascetic goes pvr*»n* 
much further, and often extends to the most trifling rules of 
discipline as has been shown by Professor Jacobi from a com- 
parison of the rules laid down for Jain monks and for Brahman 
mendicants*. Evidently there is not the slightest reason for 

1 Cp. Macdonell and Keith, Vedic Index of Names and Subjects, vol. i, pp. 68 sq. 

2 See Chapter xrx. 3 Kautilya, Arthaqdttra, p. 8. 
* Cp. S.B.E., vol. xxn, pp. xxii sq. 

152 The History of the yains [ch. 

regarding either the Jains or the Buddhists as innovators in these 
matters ; and the following pages will show that it was in doctrine 
rather than in life, in the attempt to abolish the authority of the 
Brahman scriptures and the rites of sacrifice rather than in any 
efibrt to change the social institutions and conditions of his time, 
, that Mahavira differed more widely from his Brahman predecessors. 
And when both he and his great rival, Buddha, state that a man is 
not merely born a Brahman, but becomes a Brahman through his 
meritorious actions, they seem not even here to be real innovators ; 
for we are immediately reminded of the legend of Satyakama 
Jabala and other similar instances^, that seem to prove that birth 
was not always regarded as the true keynote of sanctity even in 
orthodox circles. Jainism, as well as Buddhism, is certainly to be 
viewed only in close connexion with the Brahman institutions 
existing at the time of its rise ; and from this standpoint we may 
now enter upon a closer investigation of the subject of this chapter, 
the origin and first development of the Jain church. 

For a considerable time European scholars were unable to form 
a clear opinion on the rise and growth of Jainism owing to the 
absence of original texts which were then scarcely available in 
Europe. Thus the older generations of Sanskrit scholars may be 
said to have shared principally two difierent opinions on these 
matters. Colebrooke, Prinsep, Stevenson, E. Thomas, and others 
thought Jainism to be older than Buddhism — an opinion to which 
we may now willingly subscribe — mainly from the reason, that a 
disciple of Mahavira called Indrabhuti Gautama was held to be the 
same person as Gautama the Buddha. On the other hand, other 
distinguished Orientalists such as H. H. Wilson, Lassen, and even 
Weber, were of the opinion that Jainism was only one of the many 
difierent sects into which Buddhism was divided at an earlier or 
later date after the death of Buddha. Such a view might easily be 
held on the basis of certain somewhat striking resemblances which 
are found in the Buddhist and Jain records of which at that time 
only a comparatively small number had found their way to Europe. 
This latter hypothesis has now been thoroughly refuted by the 
works of two eminent German scholars, Biihler and Jacobi, who 
have laid down a sure foundation for our knowledge of Jainism by 
a thorough investigation of its old canonical texts and a coinparison 
of these with the scriptures of the Buddhists and Brahmans. 
Starting therefore from the standard works on Jainism published 
by Professor Jacobi, and making use of the materials, which have 

^ Cp. Vtdic Index, vol. 11, pp. 84 sq. 

vi] Parqva 153 

been collected and examined by other scholars, we are now able 
to obtain a fairly clear view of the early history of Jainism. 

Mahavira is usually regarded as the real founder of the Jain 
religion ; and, as we have very scanty information about the only 
one of his alleged predecessors, who may possibly have had a real 
existence, we are, in our investigation, almost forced to adopt this 
point of view. But the Jains themselves claim for their religion 
a far more venerable antiquity : they tell us that before Mahavira 
there lived not less than 23 tlrthakaras or ' prophets,' who appearing 
at certain intervals preached the only true religion for the salva- 
tion of the world. The first of these prophets was king Rishabha, 
who after laying down his royal power and transferring the realm 
to his son Bharata, the first universal monarch (chakravartin), 
became a holy man and a tirthakara. As the opinions of the 
Jains about time and the ages of the world ^ are absurdly exag- 
gerated, it is almost impossible to express in numbers the time at 
which he is thought to have lived ; it may be enough to say that 
his lifetime is supposed to have lasted for several billions of years 
and his height to have been about two miles. From such state- 
ments and fi'om the flowery descriptions of the blissful state of 
the world in its first ages, it is evident that the Jains, as indeed, all 
Hindus, attributed to the first race of men a longer life, a greater 
strength, and more happiness than fall to the share of their ofispring 
in the present age. As we know, the Greeks and Romans held 
similar opinions. But, of course, the world grew worse and worse 
and the life of man shorter and shorter, so that the 23rd tirthakara^ 
Parqva, the immediate predecessor of Mahavira, is said to have 
lived only for a hundred years, and to have died only 250 years 
before his more celebrated successor. 

This Pargva is assumed, on the authority of Professor Jacob! 
and others, to have been an historical personage and the real 
founder of Jain religion. As he is said to have died 250 years 
before the death of Mahavira, he may probably have lived in the 
eighth century Rg. Professor Jacobi seems to regard this date as 
not improbable, since some centuries must have elapsed between 
his time and the appearance of the last Jain prophet^. But, as we 
have not a single certain date in Indian history before the time of 
Buddha, it is evidently impossible to prove this. Almost as scanty 
is our knowledge of the life and teaching of Parqva, in spite of the 

^ Upon this subject oonsalt Jacobi in Hastings' Encyclopcedia of Religion and 
Ethict, vol. I, p. 202. 

2 Cp. S.B.E., vol. XLV, p. 122, n. 3. 

154 ^'^^ History of the jfains [ch. 

large body of literature which has clustered around his name. In 
the well-known Kalpasutra of the Jains, which is stated to have 
been written by the pontiff Bhadrabahu (perhaps somewhat before 
300 RC), we have in the chapter called ' The life of the Jinas ' a 
short account of the life of Parcva ; but, as it is written in a purely 
formal style and bears too much resemblance to other records of 
the same sort, its value as an historical document is somewhat 
doubtful. However, it states that ParQva, like all tirthakaras, was 
a Kshatriya, a member of the second caste, that of the warriors or 
nobility according to Brahman law, and son of king A^vasena of 
Benares and his wife Vama. No such person as A9vasena is known 
from Brahman records to have existed : the only individual of that 
name mentioned in the epic literature was a king of the snakes 
indga), and he cannot in any way be connected with the father of 
the Jain prophet. Par9va, who is always titled purisddamya, 
which may mean either 'the people's favourite'^ or 'the man of 
high birth,' lived for thirty years in great splendour and happiness 
as a householder, and then, leaving all his wealth, became an ascetic. 
After 84 days of intense meditation he reached the perfect know- 
ledge of a prophet, and from that time he lived for about 70 years 
in the state of most exalted perfection and saintship, and reached his 
final liberation, nirvana, on the top of mount Sammeta surrounded 
by his followers. 

In regard to the teaching of Parcva we are better informed : it 
was probably essentially the same as that of MahavTra and his 
followers. But we have no exact knowledge, except on two prin- 
cipal points, as to how far this creed was due to Parcva, or what 
innovations may have been introduced by his successor. We are 
told that Parcva enjoined on his followers four great vows, viz. not 
to injure life, to be truthful, not to steal, and to possess no 
property^, while Mahavira added a fifth requisition, viz. that of 
chastity. Further we know that Parcva allowed his disciples to wear 
an upper and an under garment. Mahavira, on his part, followed 
the more rigid rule which obliged the ascetic; to be completely 
naked. These seem to have been, in fact, the most important 
differences in doctrine between the founder and the reformer of 
Jainism; for an old canonical text^ tells us about a meeting 
between Gautama, the pupil of Mahavira, and Ke9in, a follower of 

1 Cp. S.B.E., vol. XXII, p. 271. 

^ Gp. S.B.E., vol. XLT, p. 121, and Dr Hoemle in Hastings' Encyclopadia, vol. i, 
p. 264. 

3 Cp. S.B.E., vol. XLV, pp. 119 sq. 

vi] Traditional T>ate of MahavJr a 155 

Par9va, in which they tried successfully to solve those questions 
on which a difference of opinion existed among the religious ; and 
in that account the four vows and the wearing or not wearing of 
clothes form the main points of discussion. From this text we 
may venture to draw the conclusion that followers of Par9va, who 
did not, perhaps, fully recognise Mahavira as their spiritual head, 
existed during the lifetime of the latter, and that a sort of com- 
promise was effected between the two sections of the church. 
Indeed it seems to remain a somewhat unsettled question if 
followers of Par^va and of Mahavira are not to be found even at 
the present day as the (^vetambaras, or * monks in white clothes,' 
and the Digambaras, 'sky-clad or naked ascetics.' However, this 
hypothesis is denied by most authorities ; and as a matter of fact 
the old records place the division of the church into these two 
main sects at a time much later than Mahavira, as we shall see 

Nothing is known about the followers of Par9va until the time 
of the appearance of the last prophet of the Jains, Mahavira. As 
he is not only the most famous propagator of the Jain religion, but 
also after Buddha the best known of the non-Brahman teachers of 
ancient India, we shall have to dwell a little longer upon the records 
of his life, and in the first place we must examine such chronological 
data as exist for the determination of his period. 

The Jains themselves have preserved chronological records 
concerning Mahavira and the succeeding pontiffs of the Jain church, 
which may have been begun at a comparatively early date. But 
it seems quite clear that, at the time when these lists were put into 
their present forai, the real date of Mahavira had already either 
been forgotten or was at least doubtful. The traditional date of 
Mahavira's death on which the Jains base their chronological 
calculations corresponds to the year 470 before the foundation of 
the Vikrama era in 58 B.C., i.e. 528 RC.^ This reckoning is based 
mainly on a list of kings and dynasties, who are supposed to have 
reigned between 528 and 58 B.C. ; but the list is absolutely valueless, 
as it confuses rulers of Ujjain, Magadha, and other kingdoms ; and 
some of these may perhaps have been contemporary, and not 
successive as they are represented. Moreover, if we adopt the 
year 528 RC, it would exclude every possibility of Mahavira having 
preached his doctrine at the same time as Buddha, as the Buddhist 

^ Or 527 B.C. according to those authorities who regard 57 b.o. as the starting point 
of the Vikrama era. Dates are here given on the assumption that the Yikrama era 
began in 58 b.o. 

156 The History of the yains [cH. 

texts assert ; for there is now a general agreement among scholars 
that Buddha died within a few years of 480 b.c.^ ; and therefore 
some fifty years would have elapsed between the decease of the two 
prophets. But we are told that Buddha was 80 years old at his 
death, and that he did not begin preaching before his 36th year, 
that is to say, at a time when Mahavira, according to the tradi- 
tional date, was already dead. Finally, both Mahavira and Buddha 
were contemporaries with a king of Magadha, whom the Jains call 
Kiinika, and the Buddhists Ajata9atru; and he began his reign 
only eight years before Buddha's death. Therefore, if Mahavira 
died in 528 B.C., he could not have lived in the reign of Kiinika. 
So we must, no doubt, wholly reject this date and instead of it 
adopt another which was long ago suggested by Professor Jacobi^ on 
the authority of the great Jain author Hemachandra (d. 1172 A.D.), 
viz. 468 (467) B.C. The dynastic list of the Jains mentioned above 
tells us that Chandragupta, the Sandrokottos of the Greeks, began 
his reign 255 years before the Vikrama era, or in 313 KG., a date 
that cannot be far wrong^ And Hemachandra states that at this 
time 155 years had elapsed since the death of Mahavira, which 
would thus have occurred in 468 B.C. This date agrees very well 
with other calculations and is only contradicted by a passage in the 
Buddhist Digha Nikaya^ which tells us that Nigantha Nataputta 
— the name by which the Buddhists denote Mahavira — died before 
Buddha. This assertion is, however, in contradiction with other 
contemporaneous statements, and forms no real obstacle to the 
assumption of the date 468 B.C. We may therefore adopt this 
year as our basis for calculating the various dates in Mahavlra's 

To give a sketch of Mahavira's life is a somewhat difficult task 
as the oldest existing biography, included in the chapter of the 
Kalpasiitra to which we have referred, is fanciful and exaggerated, 
bearing in these respects a certain resemblance to the tales in the 
Lalita-vistara and Nidana-katha concerning the early life of Buddha. 
If this biography is really the work of Bhadrabahu, it may be 
expected to contain notices of great value, even although its state- 

1 In 483 B.C. according to the system of chronology adopted in this work ; or in 
478 (477) B.C. as appears more probable to the present writer. For a fuU discussion of 
the dates of Mahavira and Buddha, on the assumption that the Vikrama era began in 
57 B.C., see Charpentier, Ind. Ant., 1914, pp. 118 ff., 125 ff., 167 S. 

2 Kalpasiitra, pp. 8 ff. 

3 V. inf., p. 164. 

* D.N., m, pp. 117, 209. Also Majjhima Nikaya, ii, pp. 273 ff. Cp. Chalmers, 
J.B.A.S., 1895, pp. 665 f. 

vi] Mahavtra 157 

ments cannot always be accepted as strictly accurate. There are, 
moreover, in several old canonical works passages which give 
information on various events in Mahavlra's life ; and the Buddhist 
scriptures also give us some valuable hints. 

The capital of Videha, Vesali or Vai9ali\ was without doubt 
one of the most flourishing towns of India about 500 years before 
the beginning of our era. The government, which was republican, 
or perhaps rather oligarchical, was entrusted to the princely family 
of the Licchavis, who are often mentioned in Buddhist and Jain 
writings, and who were certainly mightier at that time than at a 
later date, when an author ^ remarks that they * lived by assuming 
the title of king {rdjan).' Just outside Vai9ali lay the suburb 
Kundagrama — probably surviving in the modem village of Basu- 
kund — and here lived a wealthy nobleman, Siddhartha, head of a 
certain warrior-clan called the Jiiatrikas. This Siddhartha was 
married to the princess Tri^ala, sister of Chetaka, the most eminent 
amongst the Licchavi princes, and ruler of Vai9alT. To them were 
bom, according to the tradition, one daughter and two sons, the 
younger of whom was called Vardhamana, the future Mahavira. 
Through the Licchavis Siddhartha became the relative of a very 
powerful monarch ; for king Bimbisara or ^renika of Magadha, the 
patron of Buddha and the mightiest ruler of Eastem India, had 
married CheUana, daughter of Chetaka; and she was mother of 
Ajata9atru or Ktinika, who murdered his father eight years before 
the death of Buddha, and ascended the blood-stained throne of 

This is what we leam fi'om the Kalpasutra concerning Maha- 
vlra's pedigree ; and there is no reason to doubt this information. 
But the birth of great men — and especially religious teachers — 
has often afterwards been made a theme for the most fanciful and 
supematural legends.* And so the Kalpasutra tells us that Maha- 
vira, when he descended from the heavenly palace of Pushpottara 
where he had led his previous existence, was at first conceived in 
the womb of Devananda, wife of the Bi-ahman Rishabhadatta. This 
couple, too, lived in the suburb of Kundagrama. However, it had 
never happened in the innumerable cycles of previous world- 
periods that a prophet had been bom in a Brahman family ; and 
consequently the god ^akra (Indra) had the embryo removed from 
the womb of Devananda to that of Tri9ala. We must observe, 

^ The site and sarronndingB of YaiQali are indicated by Dr Vincent A. Smith, 
J.R.A.S., 1902, pp. 267 ff. 

2 The Arthafdstra of Kautilya, p. 376. 

158 The History of the yains [ch. 

however, that this tale is only believed by the ^vetambaras, and 
constitutes one of the four main points rejected by the Digambaras, 
who seem here to hold the more sensible opinion. 

Just like the mother of Buddha, the princess Tri9ala had 
auspicious dreams in the very night of conception ; and the inter- 
preters foretold that the child would become either a universal 
monarch or a prophet possessing all-comprising knowledge. So 
the boy, whose birth was celebrated alike by gods and men, was 
received by his parents with the most lofty expectations, and was 
educated to the highest perfection in all branches of knowledge 
and art In due time he was married to a lady, named Ya^oda, 
and had by her a daughter, who became the wife of Jamali, a 
future disciple of his father-in-law, and the propagator of the first 
schism in the Jain church. However, Mahavira's mind was not 
turned towards secular things ; and in his thirtieth year, after the 
decease of his parents, he left his home with the permission of his 
elder brother, Nandivardhana, and set out for the life of a homeless 

The first book of the Jain canon, the Acharanga-sutra, has 
preserved a sort of religious ballad^ giving an account of the years 
during which Mahavira led a life of the hardest asceticism, thus 
preparing himself for the attainment of the highest spiritual 
knowledge, that of a prophet. During the first thirteen months he 
never changed his robe, but let ' all sorts of living beings ' — as the 
text euphemistically says — crawl about on his body ; but after this 
time he laid aside every kind of garment and went about as a naked 
ascetic. By uninterrupted meditation, unbroken chastity, and the 
most scrupulous observation of the rules concerning eating and 
drinking, he fully subdued his senses; nor did he ever in the 
slightest degree hurt or cause offence to any living being. Roam- 
ing about in countries inhabited by savage tribes, rarely having a 
shelter in which to rest for the night, he had to endure the most 
painful and injurious treatment from the barbarous inhabitants. 
However, he never lost his patience, and never indulged in feelings 
of hatred or revenge against his persecutors. His wanderings seem 
to have covered a wide area, and on occasions he visited Rajagriha, 
the capital of Magadha, and other towns, where the utmost honour 
was shown him by pious householders. 

It was during one of these visits to Nalanda, a suburb of 
Rajagriha famous in the sacred history of the Buddhists, that he 
met with Gosala Mamkhaliputta, a mendicant friar, who attached 

Translated in S.B.E., vol. xxn, pp. 79 S. 

vi] The Nirvana of Mahavtra 159 

himself to Mahavira for some years. The consequences of this 
meeting were certainly disastrous for both the teacher and the 
disciple. For six years they lived together practising the most 
austere asceticism; but after that time, on account of a dispute 
which arose out of a mere trifle, Gosala separated himself from 
Mahavira, and set up a religious system of his own, soon afterwards 
proclaiming that he had attained to the highest stage of saintship, 
that of a tirthakara. This claim was put forth two years before 
Mahavira himself had reached his perfect enlightenment. The 
doctrines and views of Gosala are known to us only from notices 
scattered throughout the Jain and Buddhist writings, and his 
followers, the Ajlvika sect, have left no written documents; but 
from the intolerant and bitter sayings of the Jains concerning 
Gosala, whom they stigmatise as merely a treacherous impostor, 
we may well conclude that the cause of dissension between him 
and his foiiner teacher was deep-rooted, and that this quarrel 
must have been a severe blow to the rising influence of Mahavira 
and the establishment of the new religious community. Gosala 
took up his head-quarters in a potter's shop belonging to a woman 
named Halahala at (yJravasti, and seems to have gained considerable 
reputation in that town. We shall hear something about him at a 
later stage ; but for the present we must return to Mahavira 

Twelve years spent in self-penance and meditation were not 
fruitless; for in the thirteenth year Mahavira at last reached 
supreme knowledge and final deliverance from the bonds of pleasure 
and pain. The ipsissima verba of an old text will perhaps best 
show us how the Jains themselves have described this the most 
important moment of the prophet's life: 'during the thirteenth 
year, in the second month of summer, in the fourth fortnight, the 
light (fortnight) of Vai9akha, on its tenth day, called Suvrata, while 
the moon was in conjunction with the asterism Uttara-Phalguni, 
when the shadow had turned towards the east, and the first wake 
was over, outside of the town Jrimbhikagrama, on the northern 
bank of the river ^ijupalika, in the field of the householder Samaga, 
in a north-eastern direction from an old temple, not far from a Sal 
tree, in a squatting position with joined heels exposing himself to 
the heat of the smi, with the knees high and the head low, in deep 
meditation, in the midst of abstract meditation, he reached nir- 
vana, the complete and fall, the unobstructed, unimpeded, infinite 
and supreme, best knowledge and intuition, called kevala (total). 
When the venerable one had become an Arhat and Jina, he was a 

i6o The History of the yains [ch. 

Tcevalin, omniscient and comprehending all objects, he knew all 
conditions of the world, of gods, men, and demons ; whence they 
come, where they go, whether they are born as men or animals, or 
become gods or hell-beings ; their food, drink, doings, desires, and 
the thoughts of their minds ; he saw and knew all conditions in the 
whole world of all living beings ^' 

At this time Vardhamana, henceforth styled Mahdvira (the 
great hero) or Jina (the conqueror), was 42 years old ; and from 
this age he entered upon a new stage of life, that of a religious 
teacher and the head of a sect called the nirgranthaa * free from 
fetters,' a designation nowadays obsolete, and superseded by the 
the term Jainas ' followers of the Jina.' His parents had, according 
to a tradition which seems trustworthy, been followers of Par9va, 
the previous tlrthakara: as has already been pointed out, the 
doctrine of Mahavira was scarcely anything else than a modified 
or renovated form of Par9va's creed. As he was a nirgrantha 
monk, and a scion of the Jnatri clan, his opponents, the Buddhists, 
call him Niggantha Ndt{h)aputta (in Sanskrit Nirgrantho Jnd- 
triputrah). We owe to Professor Jacobi the suggestion, which is 
undoubtedly correct, that the teacher, who is thus styled in the 
sacred books of the Buddhists, is identical with Mahavira, and 
that consequently he was a contemporary of Buddha. 

We possess little knowledge of the thirty years, during which 
Mahavira wandered about preaching his doctrine and making 
converts. He apparently visited all the great towns of N. and S. 
Bihar, principally dwelling in the kingdoms of Magadha and 
Anga. The Kalpasiitra tells us that he spent his rainy seasons, 
during which the rules for monks prohibited the wandering life, at 
various places, e.g. at Champa, the capital of Anga, at Mithila in 
the kingdom of Videha, and at ^ravasti, but chiefly at his native 
town Vai9ali, and at Rajagriha, the old capital of Magadha. He 
frequently met with Bimbisara and his son, Ajata9atru or Kunika, 
the kings of Magadha, and their near relations ; and according to 
the texts he was always treated by them and other important 
persons with the utmost respect, and made many converts amongst 
the members of the highest society. But we must observe that the 
Buddhists in an equal degree claim these kings as followers of their 
prophet; and we may conclude that uniform courtesy towards 
teachers of different sects was as common a characteristic of Indian 
kings in those days as at a later period. The Jains do not tell us 

1 From the Achdrdnga-aUtra, vol. i, pp. 15, 25-26 (translated in S.B.E., Tol. xzu, 
pp. 201 sq.). 

vr] yains and Buddhists i6i 

anything about the Buddhists ; but the latter frequently mention 
discussions and controversies between Buddha and disciples of 
Mahavira. In these accounts Buddha, of course, always has the 
last word, and is said to have inflicted considerable loss on the 
Jain community through the converts which he made amongst 
its followers. Even king Ajata9atru, according to the Pali texts, 
failed to obtain a satisfactory explanation concerning matters of 
religion from Mahavira, and consequently turned to Buddha with 
a far better result; but there seems to be little doubt that the 
Jains have more claim to include the parricide king amongst their 
converts than the Buddhists. Another prominent lay-follower of 
Mahavira was the householder of Rajagriha, Upali, who in his 
enthusiasm embarked on the attempt to convince Buddha of his 
wrong views. We learn, however, that the great teacher easily 
upset his arguments, and gained in his opponent a stalwart ad- 
herent to his creed. Subsequently, Upali is said to have treated 
his former teacher with an arrogance, which so shocked Mahavira 
that 'hot blood gushed fi-om his mouth \' 

But although the relations between the Jains and Buddhists 
were by no means friendly, we must probably not attach too much 
importance to the controversies between them or to the number of 
converts said to have been gained by one sect at the expense of 
the other. Between two contemporary religious communities 
working side by side in the same region and often coming into 
contact there must have occurred skirmishes; but the whole 
doctrine and mode of life adopted by the Buddhists was too widely 
difierent from that of the Jains to give occasion for more than 
somewhat temporary relations. We cannot here enter upon any 
full investigation of the doctrine of Mahavira. It must suffice here 
to point out that it represents, probably, in its fundamental tenets 
one of the oldest modes of thought known to us, the idea that all 
nature, even that which seems to be most inanimate, possesses 
life and the capability of reariimation ; and this doctrine the 
Jains have, with inflexible conservatism, kept until modern times. 
This has nothing in common with the philosophy of Buddha. 
There is, in reality, no resemblance between the two systems 
except in regard to such matters as are the commonplaces of all 
Hindu philosophy. Even for those superficial believers who looked 
more to the exterior appearance and mode of life than to the 
doctrine and faith, the two sects presented an aspect so completely 
different that one could not easily be confused with the other. 

^ Cf. the Updli-sutta in Majjhima Nikdya, vol. i, pp, 371 ff. 
C.H.I. I, ' 11 

1 62 The History of the Jains [CH. 

Buddlm had at first sought freedom from kariman, or the bondage 
of 'works,' and from transmigration in exaggerated self-torture: 
but he soon found that this was not the way to peace ; and conse- 
quently he did not enforce upon his followers the practice of too 
hard self-penance but advised them to follow a middle way, that is 
to say, a simple life but one free from self-torture. Mahavira also 
had practised asceticism but with a different result; for he had 
found in its severest forms the road to deliverance, and did not 
hesitate to recommend nakedness, self-torture, and death by starva- 
tion as the surest means of reaching final annihilation ; and the 
Jains proud of their own austerities often stigmatise the Buddhists 
as given to greed and luxury. Buddha always warned his disciples 
against hurting or causing pain to any living being ; but Mahavira 
fell into exaggerations even here, and he seems in reality often to 
care much more for the security of animals and plants than for 
that of human beings. Such instances of a deep-rooted divergence 
in views could easily be multiplied; but what has been already 
pointed out is sufficient to prove that the Jains and Buddhists 
were in fact too far asunder to be able to inflict any very serious 
damage on each other. But this does not mean, however, that 
rivalry and hatred did not exist between them: such feelings 
certainly did exist, and we need not doubt that these rivals did 
their best to annoy each other according to their abilities and 

A far more dangerous rival of Mahavira was Gosala. Not only 
was his doctrine, although differing on many points, mainly taken 
from the tenets of Mahavira^; but his whole mode of life also, in 
its insistence on nakedness and on the utter deprivation of all 
comforts, bore a close resemblance to that of the Jains. Between 
two sects so nearly related the transition must have been easy; 
and pious people may not always have been quite sure whether 
they were honouring the adherents of one sect or of the other. 
The Jain scriptures admit that Gosala had a great many followers in 
(^ravasti ; and, if we may trust their hints as to his laxity in moral 
matters, it is possible that his doctrine may for some people have 
possessed other attractions than those of asceticism and holiness. 
Although Mahavira is said not to have had any personal meeting 
with Gosala until shortly before the death of the latter, it seems 
clear that they carried on a bitter war against each other through 
their followers. Finally, in the sixteenth year of his career as a 
prophet, Mahavira visited ^ravastl, the head-quarters of his mortal 

1 Cp. Hastings' Encyclopadia, vol. i, p. 261, for farther details. 

vi] Schisms in the jfain Church 163 

enemy. The account given by the Jains tells us that, at this 
meeting, Mahavira inflicted a final blow on his adversary, and that 
Gosala died a week afterwards, having passed his last days in a 
state of drunkenness and mental imbecility, but showing some signs 
of repentance at the last. But the story is rather confiised, and it 
seems doubtful to what extent we may trust it. However, it may 
be regarded as beyond dispute that Mahavira was considerably 
relieved by the death of his opponent; and, according to the 
Bhagavati-siitra, he took a rather strange revenge on the dead man 
by describing to his disciples all the wicked deeds he would have 
to perform, and all the pains he would have to suff*er in future 
existences, thus to a certain degree anticipating Dante's treatment 
of his adversaries. The death of Gosala occurred shortly after 
Ajata9atru had gained accession to the throne of Magadha by the 
murder of his father. 

Even within the Jain church there occurred certain schismatical 
difficulties at this time. In the fourteenth year of Mahavira's 
office as prophet, his nephew and son-in-law, Jamali, headed an 
opposition against him, and similarly, two years afterwards, a holy 
man in the community, named Tisagutta, made an attack on a 
certain point in Mahavira's doctrine. But both of these schisms 
merely concerned trifles, and seem to have caused no great 
trouble, as they were speedily stopped by the authority of the 
prophet himself. Jamali, however, persisted in his heretical 
opinions until his death. 

Mahavira survived his hated rival Gosala for sixteen years, and 
probably witnessed the rapid progress of his faith during the reign 
of Ajata9atru, who seems to have been a supporter of the Jains, if 
we may infer that gratitude is the motive which leads them to 
make excuses for the horrible murder of his father, Bimbisara. 
However, we are not informed of any special events happening 
during the last period of his life, which may have been as 
monotonous as that of most religious mendicants. He died, after 
having reached an age of 72 years, in the house of king Hastipala's 
scribe in the little town of Pawa near Rajagriha, a place still 
visited by thousands of Jain pilgrims. This event may have 
occurred at the end of the rainy season in the year 468 B.C. Thus, 
he had survived both of his principal adversaries; for Buddha's 
decease most probably took place at least ten, if not fifteen, years 

1 For reasons why the Buddhist account, according to which Mahavira died before 
Buddha, is not accepted here, see Charpentier, Ind. Ant., 1914, p. 177. 


164 The History of the yains [ch. 

Out of the eleven ganadharas ' heads of the school,' or apostles, 
of Mahavira only one survived him, viz. Sudharman, who became 
the first pontiflF of the new church after his master. Absolutely 
nothing is known concerning the fate of the community for more 
than 160 years after the death of its founder beyond the very 
scanty conclusions which may be drawn from the legendary tales 
related by later Jain writers, above all by the great Hemachandra. 
According to these authorities, Ajata9atni was succeeded by his 
son Udayin, a prince, who may have reigned for a considerable 
time, and who was a firm upholder of the Jain religion. But the 
irony of fate was visible even here ; for the very favour which he 
had bestowed upon the Jains proved to be the cause of his ruin : a 
prince whose father he had dethroned plotted against his life ; and, 
aware of the welcome accorded to the Jains by Udayin, he entered 
his palace in the disguise of a Jain monk, and murdered him in the 
night. This happened 60 years after Mahavira's decease. The 
dynasty of the nine Nandas, somewhat ill-famed in other records 
which call its founder the son of a courtezan and a barber^, then 
came to the throne of Magadha. However, the Jains do not share 
the bad opinion of these kings which was held by the Buddhists. 
This fact seems to suggest that the Nanda kings were not un- 
favourably inclined towards the Jain religion; and this inference 
gains some support from another source, for the badly mutilated 
inscription of Kharavela, king of Kalinga and a faithful Jain, 
mentions, apparently, in one passage ' king Nanda ' in unmistakable 
connexion with 'an idol of the first Jinal' But the reign of the 
Nandas is one of the darkest even of the many hopelessly dark 
epochs in the history of ancient India. 

The last of the Nandas was dethroned by Chandragupta, the 
founder of the Maurya dynasty, with the aid of the great statesman, 
Chanakya, within a few years of the departure of Alexander the 
Great from India. The Jains put the date of Chandragupta's 
accession in 313 (312) B.C., that is to say, eight years later than 
the Buddhists. This date coincides probably with a year wliich 
marks an epoch in the history of the Jain church. Sudharman, the 
first pontifij had died twenty years after his master, leaving the 
mitre to Jambu, who held his high ofiice for 44 years, dying at a 
time nearly coincident with the accession of the Nandas. After him 
passed three generations of pontiffs ; and, in the time of the last 
Nanda, the Jain church was governed by two high-priests, Sambhii- 
tavijaya and Bhadrabahu, the author of the biogi-aphy of Mahavii-a 
» See however Chapter xni, pp. 313-4. = Cf. Ind. Ant., 1914, p. 173. 

vi] Qvetambaras and Digambaras 165 

quoted above. These two were the last who knew perfectly the 
fourteen purvas or divisions of the most ancient Jain scriptures ; 
and Sambhutavijaya is said to have died in the same year in which 
Chaiidragupta took possession of the throne. At the same time a 
dreadful famine lasting for twelve years devastated the region of 
Bengal ; and Bhadrabahu, seeing that this evil time would provoke 
numerous offences against the ecclesiastical rules, thought it prudent 
to escape. Gathering his followers together, therefore, he emigrated, 
and took up his abode in the country of Karnata in Southern India. 
The whole community, however, did not follow him. Many Jains 
remained in Magadha and other places under the spiritual leader- 
ship of Sthulabhadra, a disciple of Sambhutavijaya. 

At the end of the famine the emigrants returned, but at this 
time Bhadi'abahu seems to have laid down his leadership of the 
church, and to have retired to Nepal in order to pass the remainder 
of his life in penance, leaving the succession to Sthulabhadra. 
There is no reason to believe the account given by the Digambaras, 
according to which he was murdered by his own disciples. But, in 
any case, this time seems to have been one of misfortune for the 
Jain church; and there can be no doubt it was then, i.e. about 
300 B.C., that the great schism originated, which has ever since 
divided the community in two great sects, the (^vetambaras and 
the Digambaras. The returning monks, who had during the famine 
strictly observed the rules in all their severity, were discon- 
tented with the conduct of the brethren who had remained in 
Magadha, and stigmatised them as heretics of wrong faith and lax 
discipline. Moreover, during this time of dissolution, the old 
canon had fallen into oblivion; and consequently the monks who 
had remained in Magadha convoked a great council at Pataliputra, 
the modem Patna, in order to collect and revise the scriptures. 
However, this proved to be an undeilaking of extraordinary 
difficulty, since the purvas or older parts were known perfectly 
only to Bhadrabahu, who had at this time already settled in Nepal ; 
and Sthiilabhadra, who went there in person, although he learnt 
from his predecessor all the fourteen purvas, was forbidden to 
teach more than the first ten of them to others. The canon 
established by the Council was therefore a fragmentary one ; and 
in it, to some extent, new scriptures took the place of the old. In 
some degree it may be represented by the present canon of the 
(^vetambaras, since that too is preserved in a somewhat disorderly 
condition. The returning monks, the spiritual ancestors of the 
Digambaras, seem to have taken no part in the council, and to 

1 66 The History of the jfains [ch. 

have proclaimed that the real canon had been hopelessly lost ; and 
even to the present day they have continued to hold the same 
opinion. They regard the whole canon of the (^vetambaras, the 
Siddhanta as it is called, as merely a late and unauthoritative 
collection of works, brought together by Jinachandra in Valabhi at 
a far later date. 

But probably the difficulties which beset the Jain church at 
this period were not only internal. As is well known, the Jains 
nowadays are settled principally in Western India, Gujarat, etc. 
That they have been there for a very long time is certain, since 
their non-canonical writings, as well as epigraphical documents, bear 
witness at an early date to their influence in these parts of India. 
As the historical records of the sect have very little to tell us of 
the reign of Chandragupta and his son Bindusara, and perhaps even 
stiU less of the great A9oka, it seems probable that they had 
already in the third century B.O. begun to lose their foothold in 
Eastern India. The manual of politics by Chanakya describes a 
purely Brahman society ; and it may perhaps not be too hazardous 
to infer jfrom this fact that the first rise of the Maurya dynasty may 
have marked an attempt to restore the Brahman power and so 
check the rising influence of the heterodox communities. If so, 
this policy was certainly abandoned by A9oka whose zeal for 
Buddhism may have been one of the main causes for the downfall 
of his great empire immediately after his death. It is true that 
A9oka in one of his edicts mentions his protection of the nirgranthas 
as well as of the Buddhists and other pious men ; but any attempt 
to prove a greater interest on his part in the welfare of the Jains 
must fail, unsupported as it is by the scriptures of the Jains them- 
selves. It is true too that Kliaravela, king of Kalinga, who, although 
his exact date may be doubted, certainly lived a considerable 
time after A9oka, displayed a great zeal for the Jain religion ; but 
it seems quite clear that, at the time of A9oka's death, the Jains 
had practically lost their connexion with Eastern India; since 
they apparently know nothing of his grandson Da9aratha, who 
succeeded him in Magadha, and, of the following princes, only the 
usurper Pushyamitra, a patron of Brahmanism,is mentioned by them. 
On the other hand, they tell us that Samprati, another grandson of 
A9oka who reigned probably in Ujjain, was a strong supporter of 
their religion, and his capital seems to have played at this time an 
important rOle in the history of Jainism. 

As we have seen, in about 300 B.C. the division of the Jain 
church into the two great sects of the ^vetambaras and Digambaras 

vi] Western Settlements 167 

had probably already begtin. The final separation between the 
two communities is, no doubt, reported not to have taken place 
before 79 or 82 A.D, ; but the list of teachers and schools in the 
Kalpasutra and the numerous inscriptions from Mathura, which 
date mostly fi*om the time of the later Kushana kings, Le. after 
78 A.D., afford sufficient proof that the ^vetambara community was 
not only established but had become subdivided into smaller sects 
at an earlier period. This is especially clear from the frequent 
mention of nuns in the Mathura inscriptions; for it is only the 
^vetambaras who give women admission into the order. Every- 
thing tends to show that the Jains were probably already at this 
time (300 RO.) gradually losing their position in the kingdom of 
Magadha, and that they had begun their migration towards the 
Western part of India, where they settled, and where they have 
retained their settlements to the present day. Attention has 
already been called to the fact that the later Jain authors mention 
Ujjain as a place where their religion had already gained a strong 
foothold in the age of A9oka and his immediate successors. Another 
locality in which the Jains seem to have been firmly established, 
from the middle of the second century B.C. onwards, was Mathura 
in the old kingdom of the ^urasenas, known at an earlier date, 
e.g. by Megasthenes (300 rc), as the centre of Ki-ishna-worship. 
The numerous inscriptions, excavated in this city by General 
Cunningham and Dr Fiihrer, and deciphered by Professor Biihler, 
tell us about a wide-spread and firmly established Jain community, 
strongly supported by pious lay devotees, and very zealous in the 
consecration and worship of images and shrines dedicated to 
Mahavira and his predecessors. An inscription, probably dated 
from 157 A.D. (= 79 9aka), mentions the Vodva tope as ' built by 
the gods,' which, as Biihler rightly remarks, proves that it in the 
second century A.D. must have been of considerable age as every- 
thing concerning its origin had been already forgotten. 

Except the long lists of teachers, often more or less apocryphal, 
which have been preserved by the modern subdivisions of the Jain 
comjnunity, there exist practically no historical records concerning 
the Jain church in the centuries immediately preceding our era. 
Only one legend, the KalaTcdchdrya-hathdnaka, ' the story of the 
teacher Kalaka,' tells us about some events which are supposed to 
have taken place in Ujjain and other parts of Western India during 
the first part of the first century RC, or immediately before the 
foundation of the Vikrama era in 58 B.C. This legend is perhaps 
not totally devoid of all historical interest. For it records how 

1 68 The History of the yains [ch. 

the Jain saint Kalaka, having been insulted by king Gardabhilla of 
Ujjain, who, according to various traditions, was the father of the 
famous Vikramaditya, went in his desire for revenge to the land of 
the ^akas, whose king was styled ' King of Kings ' {sdhdnusdhi). 
This title, in its Greek and Indian forms, was certainly borne by 
the ^aka kings of the Punjab, Maues and his successors, who belong 
to this period ; and, as it actually appears in the form shaonano 
shao on the coins of their successors, the Kushana monarchs, we 
are perhaps justified in concluding that the legend is to some 
extent historical in character. However this may be, the story 
goes on to tell us that Kalaka persuaded a number of ^aka satraps 
to invade Ujjain and overthrow the dynasty of Gardabhilla ; but 
that, some years afterwards, his son, the glorious Vikramaditya, 
repelled the invaders and re-established the throne of his ancestors. 
What the historical foundation of this legend may be, is wholly 
uncertain — perhaps it contains faint recollections of the Scythian 
dominion in Western India during the first century B.C. In any 
case, it seems undoubtedly to give further proof of the connexion 
of the Jains with Ujjain, a fact indicated also by their use of the 
Vikrama era, which was established in the country of Malwa, of 
which Ujjain was the capital. 

Thus, the history of the Jains during these centuries is 
enveloped in almost total darkness; nor have we any further 
information as to the internal conditions of the community. 
Almost the only light thrown upon these comes from the Mathui-a 
inscriptions, which incidentally mention a number of various 
branches, schools, and families of the Jain community. From this 
source, too, we learn the names of teachers who under difierent 
titles acted as spiritual leaders of these subdivisions, and of monks 
and nuns who practised their austere life under their leadership. 
Much the same religious conditions as are shown by the inscrip- 
tions have been preserved in the Jain church till the present day, 
although the names and external forms of the sects and the 
monastic schools may have changed in the course of twenty 
centuries. Moreover, the inscriptions mention the names of a 
vast number of these pious lay people, both male and female, who, 
in all ages, by providing the monks and nuns with their scanty 
livelihood, have proved one of the firmest means of support for the 
Jain church, and whose zeal for their religion is attested by the 
numerous gifts of objects for worship recorded in the inscriptions. 
Dr Hoernle^ is no doubt right in maintaining that this good or- 

1 Proceed, of the As. Soc. of Bengal, 1898, p. 53. 

vi] Conservatism of the yains i6g 

ganisation of the Jain lay community must have been a factor of 
the greatest importance to the church during the whole of its 
existence, and may have been one of the main reasons why the 
Jain religion continued to keep its position in India, whilst its far 
more important rival, Buddhism, was entirely swept away by 
the Brahman reaction. The inflexible conservatism of the small 
Jain community in holding fast to its original institutions and 
doctrine has probably been the chief cause of its survival during 
periods of severe affliction; for, as Professor Jacobi has pointed 
out long ago^, there can be little doubt, that the most important 
doctrines of the Jain religion have remained practically unaltered 
since the first great separation in the time of Bhadrabahu about 
300 B.C. And, although a number of the less vital rules concerning 
the life and practices of monks and laymen, which we find recorded 
in the holy scriptures, may have fallen into oblivion or disuse, there 
is no reason to doubt that the religious life of the Jain community 
is now substantially the same as it was two thousand years ago. 
It must be confessed fi'om this that an absolute refusal to admit 
changes has been the strongest safeguard of the Jains. To what 
extent the well-known quotation * sint ut sunt ant non sint ' may 
be applicable to the Jains of our days, may be questioned ; but the 
singularly primitive idea that even lifeless matter is animated by 
a soul, and the austerest perhaps of all known codes of disciplinary 
rules seem scarcely congruent with modern innovations. 

In the preceding pages an attempt has been made to give a 
brief sketch of the history of the Jain church from its foundation 
or reformation by Mahavira about 500 B.C. down to the beginning 
of our era. While we possess materials which enable us to con- 
struct a fairly clear biography of the prophet, and while we have 
at least some information concerning the events which preceded and 
were contemporary with the beginning of the great separation 
between ^vetambaras and Digambaras about 300 B.C., the follow- 
ing period is almost totally devoid of any historical record. And 
this is not the only blank in Jain ecclesiastical history. Scarcely 
more is known concerning the fate of the Jain church during the 
early centuries of our era down to the time of the great council of 
Valabhi, in the fifth or at the beginning of the sixth century A.D., 
when the canon was written down in its present form. The Jain 
church has never had a very great number of adherents; it has 
never attempted — at least not on any grand scale — to preach its 
doctrines through missionaries outside India. Never rising to an 

^ Z.D.M.G., vol. xxxvni, pp. 17 sq. 


The History of the yains [CH. vi 

overpowering height but at the same time never sharing the fate of 
its rival, Buddhism, that of complete extinction in its native land, 
it has led a quiet existence through the centuries and has kept its 
place amongst the religious systems of India till the present day, 
thanks to its excellent organisation and to its scrupulous care for 
the preservation of ancient customs, institutions, and doctrine. 



1. Pre-Buddhistic 

The early history of the Buddhists should properly begin far 
enough back before the birth of the Buddha to throw light on the 
causes that were at work in producing the rise and progress of the 
Buddhist reformer. Unfortunately, even after all that has been 
written on the subject of early Buddhist chronology, we are 
still uncertain as to the exact date of the Buddha's birth. The 
date 483 b.c. which is adopted in this History must still be 
regarded as provisional. The causes of this uncertainty which 
were explained by the present writer in 1877 still remain the 
same : 

If the date for Asoka is placed too early in the Ceylon chronicles, can we still 
trust the 218 years which they allege to have elapsed from the commencement of 
the Buddhist era dowa to the time of Asoka ? If so we have only to add that 
number to the correct date of Asoka, and thus fix the Buddhist era [the date of 
the Buddha's death] at 483 B.C. or shortly after. Of the answer to this question, 
there can I think, be no doubt. We can not\ 

This statement was followed by an analysis of the details of the 
lists of kings and teachers, the length of whose reigns or lives, 
added together, amount to this period of 218 years. The analysis 
shows how little the list can be relied on. The fact is that all 
such calculations are of very doubtful validity when they have 
to be made backwards for any lengthened period. Sinologists, 
Assyriologists, Egyptologists have not been able to agree on results 
sought by this method ; and, though Archbishop Usher's attempt 
to discover in this way, from the Hebrew records, the correct date 
of the creation was long accepted, it is now mere matter for 
derision. As is well known, even the Christian chronologists, 
though the interval they had to cover was very short, were wrong 
in their calculation of our Christian era. The Ceylon chroniclers 
may have been as much more wrong as the interval they had to 

^ In the Buddhist chapters names and titles appear in their Pali form. 

^ Ancient Coins and Measures of Ceylon, p. 44 of the separate edition (London, 1877). 

172 The Early History of the Buddhists [CH. 

account for was longer. We must admit that they tried their 
best, and were not so utterly at sea as the Irish church-dignitary. 
But we do not even know who made the calculation. We first 
hear of it in the fourth century A.D., and are only entitled to con- 
clude that at that date the belief in the 218 years was accepted by 
most of those Buddhists who continued in possession of the ancient 
traditions \ 

There have been endeavours, on the basis of other traditions, 
to arrive at a more exact date for the birth of the Buddha^. It is 
sufficient to state that each of these is open to still more serious 
objection. We must be satisfied to accept, as a working hypothesis 
only, and not as an ascertained fact, the general belief among 
modern European scholars that the period for the Buddha's 
activity may be approximately assigned to the sixth century B.C. 

In previous chapters of this volume will be found the story, 
drawn from the Brahman literature, of the gradual establishment 
in Northern India of the Aryan supremacy. For the period just 
before the rise of Buddhism (say the seventh century RC.) this 
literature tells us very little about political movements. The 
Buddhist books also are devoted to ideas rather than to historical 
events, and pass over, as of no value to their main objects, the 
dates and doings and dynastic vicissitudes of the kinglets before 
their own time. The fact that they do so is historically important ; 
and we should do wrong in ignoring, in a history of India, the 
history of the ideas held by the Indian peoples. But the fact 
remains. It is only quite incidentally that we can gather, from 
stories, anecdotes, or legends in these books, any information that 
can be called political. Of that referring to the pre-Buddhist 
period the most important is perhaps the list of the Sixteen Great 
Powers, or the Sixteen Great Nations, found in several places in 
the early books ^. It is a mere mnemonic list and runs as follows : 

































1 For the recent literature from the point of view of those who accept the 218 years 
as correct see Geiger, MahdvaTjisa (English translation), pp. xxii-xxxvi. 

■■* See, for instance, the various results detailed by Winternitz, Geschichte der 
indischen Litteratur, ii, i, 2, note 1. 

3 Anguttara i, 213; iv, 252, 256, 260. Referred to in Mahdvattu n, 2, line 15. 
Cf . the note in Vinaya Texts, ii, 146. 

VI i] Ancient States and Capitals 173 

When a mnemonic phrase or verse of this kind is found in identi- 
cal terms in different parts of the various anthologies of which the 
Buddhist canon consists, the most probable explanation is that it had 
been current in the community before the books M^ere put together 
as we now have them, and that it is therefore older than those 
collections in which it is found ^. As this particular list is found 
in two of the oldest books in the canon it would follow that it is, 
comparatively speaking, very old. It may even be pre-Buddhist — 
a list handed down among the bards, and adopted from them by 
the early Buddhists, For it does not fitly describe the conditions 
which, as we know quite well, prevailed during the Buddha's life- 
time. Then the Kosala mountaineers had already conquered 
Benares (Kasi), the Aiigas were absorbed into the kingdom of 
Magadhas, and the Assakas probably belonged to Avanti. In our 
list all these three are still regarded as independent and important 
nations ; and that the list is more or less correct for a period 
before the rise of Buddhism is confirmed by an ancient rune pre- 
served in the Digha^, and reproduced (in a very corrupt form, it is 
true) in one of the oldest Sanskrit-Buddhist texts ^. It runs : 

Dantapura of the Kalingas, and Potana of the Assakas, 
Mahissati for the Avantis, Roruka in the Sovira land, 
Mithila for the Videhas, and Champa among the Aiigas, 
And Benares for the Kasis — all these did Maha-Govinda plan. 

We have here seven territories evidently, from the context, 
regarded as the principal ones, before the rise of Buddhism, in the 
centre of what was then known as Jambudipa (India). Though 
quite independent of the list just discussed these mnemonic verses 
tell a similar story. Here also appear the Assakas, Angas, and 
Kasis. Only the Kalingas are added ; and the name of their 
capital, Dantapura, 'the Tooth city,' shows incidentally that the 
sacred tooth, afterwards taken from Dantapura to Ceylon was 
believed, when this list was drawn up, to have been already an 
object of reverence before the time of the Buddha. This tradition 
of a pre-Buddhist Dantapura, frequently referred to in the 
Jatakas, is thus shown to be really of much greater age. And it is 
clear that at the time when the four Nikayas were put into their 
present form* it was believed that, before the Buddha's life- time, 

1 Cf. Rh.D., Buddhist India, p. 188. 

2 II, 235, translated in Dialogues of the Buddlm, n, 270. 
* Mahdvastu m, 208, 209. 

^ For the Nikayas and their probable date, v. inf., pp. 195-6. 

174 T^f^^ Early History of the Buddhists [ch. 

the distribution of power in Northern India, had been different from 
what it afterwards became. 

In an appendix to the Digha verse the names of the seven kings of 
the seven nations are given, and it is curious that they are called the 
seven Bharatas. Their names are Sattabhu, Brahmadatta, Vessabhu, 
Bharata, Renu, and two Dhataratthas ; but the record does not 
tell us which of the seven nations each belongs to. In an interest- 
ing story at Jataka in, 470^, the hero is Bharata, king of the 
Sovlras, reigning at Roruva. This is most probably meant for the 
same man as the Bharata of the Digha passage ; and we may 
therefore apportion him to the Soviras. The mention of Renu in 
a list of ancient kings of Benares given in the Dip. in, 38-40 
probably refers to the Renu of our passage since the same rare 
name is given in both places as the name of the father of Renu. 
On the other hand the King Renu of Jataka iv, 144 is evidently 
not meant to be the same as this one. Three of the other four 
names also recur (not Sattabhu) ; but no inference can be drawn 
that the same people are meant. 

There are lists of pre-Buddhist Rajas (whatever that term may 
signify) in the chronicles and commentaries. But they can only be 
evidence of beliefs held at a late date ; they have not yet been 
tabulated or sifted ; and it would not be safe to hazard a prophecy 
that, even when they shall have been, there will be found anything 
of much value. 

2. India in the Buddha's time ; the Clans 

There is no chapter or even paragraph in the early Pali books 
describing the political conditions of North India during the life- 
time of the Buddha. But there are a considerable number of 
incidental references, all the more valuable perhaps because they 
are purely incidental, that, if collected and arranged, give us a 
picture, no doubt imperfect, but still fairly correct as fer as it 
goes, of the general conditions, as they appeared to the composers 
of the paragraphs in which the incidental references occur. They 
were collected in the present writer's Buddhist India ; and to 
that work the reader is referred for a fuller account. Considera- 
tions of space render it possible to state here only the more 
important of the conclusions which these references compel us to 

1 The references are to the Pjlli text of the Jataka. In the English translation the 
volumes correspond, and the pages of the original are indicated in square brackets. 

VI i] The Clans inn 

Of these the most far-reaching, and in some respe^^ the mote- 
surprising, is the fact that we find not only one or tw^e justified 
monarchies, and several kingdoms of lesser importance->rocedure 
German duchies or the kingdoms in England at the time parlia- 
Heptarchy — but also a number of republics ; some with com'iied. 
some with a more or less modified independence ; and oneow) 
two of very considerable power. This reminds us of the political 
situation at about the same period in Greece. We shall find a 
similar analogy, due to similar causes, in other matters also. If 
not pressed too far the analogy will be as useful as it is certainly 

The following is a list of the republics actually referred to by 
name in the oldest Pali records. Some mentioned by Megasthenes 
are added to it. 

1. The Sakiyas, capital Kapilavatthu 

2. The Bulis, capital Allakappa 

3. The Kalamas, capital Kesaputta 

4. The Bhaggas, capital on Suipsumara Hill 

5. The Koliyas, capital Ramagama 

6. The Mallas, capital Pava 

7. The Mallas, capital Kusinara 

8. The Moriyas, capital Pipphalivana 

9. The Videhas, capital Mithila 
10. The Licchavis, capital Vesali 

11-15. Tribes, as yet unidentified, mentioned 
by Megasthenes 

Nos. 1-10 occupied in the sixth century RC. the whole country 
east of Kosala between the mountains and the Ganges. Those 
mentioned, as is reported in other authors, by Megasthenes seem 
to have dwelt in his time on the sea-coast of the extreme west of 
India north of the gulf of Cutch^. It is naturally in relation to 
the Sakiyas that we have the greatest amount of detail. Their 
territory included the lower slopes of the Himalayas, and the 
glorious view of the long range of snowy peaks is visible, weather 
permitting, from every part of the land. We do not know its 
boundaries or how far it extended up into the hills or down into 
the plains. But the territory must have been considerable. We 
hear of a number of towns besides the capital — Chatuma, Sama- 
gama, Khomadussa, SilavatT, Medalumpa, Nagaraka, Ulumpa, 
Devadaha, and Sakkara. And according to an ancient tradition 
preserved in the Commentary on the Digha^ there were 80,000 

1 M'Crindle, Ancient India as described by Megasthenes, p. 144, cf. p. 156. 
^ See ilh.D., Dialogues of the Buddha, i, 147. 

174 The 'Early History of the Buddhists [ch. 

the distributir ^^® clan. This number (it is noteworthy that the 

what it after'^^'^^^ 84,000 was not chosen) would, allowing for 

Inanap^"^ dependents, mean a population of at least half a 

the sever ^^ would be absurd to take this tradition as a correct, or 

seven T^® ^^ official enumeration. We do not even know who first 

Bhar^® the calculation. But it would be equally absurd deliberately 

^ oo ignore it. It is at least interesting to find that even as late 

as Buddhaghosa the traditional estimate of the number of the 

Sakiyans was still, in spite of the temptation to magnify the extent 

of the 'kingdom' which the Buddha renounced, so limited and so 

reasonable as this. 

The administrative business of the clan, and also the more 
important judicial acts, were carried out in public assembly, at 
which young and old were alike present \ The meetings were held 
in a mote-hall — a mere roof supported by pillars, without walls. 
It is called santhagdra, a technical tenn never used of the 
council chamber of kings ^. 

We have no account of the manner in which the proceedings were 
conducted in the Sakiya mote-hall. But in the Mahd-Govinda Sut- 
tanta there is an account of a palaver in Sakka's heaven, evidently 
modelled more or less on the proceedings in a clan meeting. All are 
seated in a specified order. After the president has laid the proposed 
business before the assembly others speak upon it, and Recorders take 
charge of the unanimous decision arrived at^ The actions of gods 
are drawn in imitation of those of men. We may be sure that the 
composers and repeaters of this story, themselves for the most 
part belonging to the free clans (and, if not, to neighbouring clans 
familiar with tribal meetings) would make use of their knowledge 
of what was constantly done at the mote-hall assemblies. This is 
confirmed by the proceedings adopted in the rules observed at 
formal meetings of the Chapters of the Buddhist Order. Quite a 
number of cases are given in the Canon Law* ; and in no single 
case, apparently, is there question of deciding the point at issue 
by voting on a motion moved. Either the decision is regarded as 
unanimous ; or, if difference of opinion is manifest, then the 
matter is referred for arbitration to a committee of referees ^ It 
is even quite possible that certain of the technical terms found in 
the Rules of the Order {natti for 'motion,' uhhdhikd for 'reference 

1 T>. I, 91. ^ See the passages quoted at J.P.T.S., 1909, 65. 

3 Translated in Dialogues, vol. ii, pp. 259-264. 

* Translated in Rhys Davids' and Oldenberg's Vinaya Texts. See especially vol. 
ui, pp. 44 ff. ^ Vinaya Texts, m, pp. 49 £E. 

VI i] Local Administration 177 

to arbitration,' etc.), are taken from those in use at the mote- 
halls of the free clans. But however that may be, we are justified 
by this evidence in concluding that the method of procedure 
generally adopted in the mote-halls was not, as in modem parlia- 
ments, by voting on a motion, but rather as just above explained. 

A single chief (how or for what period chosen we do not know) 
was elected as office holder, presiding over the Senate, and, if no 
senate were in session, over the state. He bore the title of Raja 
which in this connexion does not mean king, but rather something 
like the Roman conmd, or the Greek archon. We hear at one 
time that Bhaddiya, a young cousin of the Buddha was 'raja'S at 
another that the Buddha's father Suddhodana (elsewhere spoken 
of as a simple clansman, Suddhodana the Sakiyan), held that 

We hear of mote-halls at some of the other towns besides the 
capital, Kapilavatthu. And no doubt all the more important 
places had them. The local affairs of each village were carried 
on in open assembly of the householders held in the groves which, 
then as now, formed so distinctive a feature in the long and level 
alluvial plain. 

The clan subsisted on the produce of their rice fields and their 
cattle. The villages were of grouped, not scattered, huts on the 
margin of the rice field. The cattle wandered in harvest time, 
under the charge of a village herdsman, through the adjoining 
forest (of which the village groves were a remnant), and over 
which the Sakiyan peasantry had common rights. Men of certain 
special crafts, most probably not Sakiyans by birth — carpenters, 
smiths, and potters for instance — ^had villages of their own ; and so 
also had the Brahmans whose services were often in request for 
all kinds of magic. The villages were separated one from another 
by forest jungle, the remains of the Great Wood (the Mahdvana\ 
portions of which are so frequently mentioned as still surviving 
throughout the clanships. The jungle was infested from time to 
time by robbers, sometimes runaway slaves. But we hear of no 
crime (and there was probably not very much) in the villages 
themselves — each of them a tiny self-governed republic. 

Tradition tells that the neighbouring clan, the Koliyas, were 
closely related by descent with the Sakiyas^; but we are not told 
much about the former. Five of their townships besides the 

1 Vinaya ii, 181. * Dlgha n, 52. 

* Sumangala i, 258 ff. 
C. H. I. I. 12 

178 The Early History of the Buddhists [CH. 

capital are referred to by name : — Halidda-vasana^ Sajjanela^ 
Sapuga^ Uttara*, and Kakkara-patta^ Every Koliyan was a 
Vyagghapajja by surname, just as every Sakiyan was a Gotama ; 
and in tradition the name of their capital Ramagama, so called 
after the Rama who founded it, is once given as either Kola- 
nagara or Vyagghapajja*'. The central authorities of the clan 
were served by a body of peons or police, distinguished, as by a 
kind of uniform, by a special form of head-dress^. These men had 
a bad reputation for extortion and violence. In the other clans 
we are told only of ordinary servants. The tradition that the 
Koliyans and Sakiyans built a dam over the river RohinI which 
separated their territories, and that they afterwards quarrelled 
over the distribution of the store of water ^, may very well be 
founded on fact. 

Of the form of government in the Vajjian confederacy, com- 
prising the Licchavis, the Videhas, and other clans, we have two 
traditions, Jain and Buddhist ^ They are not very clear, and do 
not refer to the same matters, the Jain being on military affairs, 
while the Buddhist refers to judicial procedure. 

The Kingdoms. I. Kosala 

Kosala was the most important of the kingdoms in North India 
during the lifetime of the Buddha. Its exact boundaries are not 
knoAvn. But it must have bordered on the Ganges in its sweep 
downwards in a south-easterly direction from the Himalayas to 
the plains at the modern Allahabad. Its northern frontier must 
have been in the hills, in what is now Nepal ; its southern 
boundary was the Ganges ; and its eastern boundary was the 
eastern limit of the Sakiya territory. For the Sakiyas, as one of 
our oldest documents leads us to infer, claimed to be Kosalans^". 
The total extent of Kosala was therefore but little less than that 
of France to-day. At the same time it is not probable that the 
administration was very much centralised. The instance of the 
very thorough Home Rule enjoyed, as we have seen, by the 
Sakiyas should make us alive to the greater probability that 
autonomous local bodies, with larger power than the village 
communities, which were of course left undisturbed, were still in 
existence throughout this wide territory. 

1 ilf. 1,387; S.v, 115. 2 ^. 11, 62. » ^. 11, 194. 

*■ S. IV, 340. ^ A.rv, 281. « Sum. i, 262. 

' S. IV, 341. 8 jat. V, 412-416. 

" Jacobi, Jaina Sutras, 1, xii. ^^ Sutta Nipdta, verse 422. 

vii] Kosala 179 

One or two of the technical terms in use to describe such 
powers have survived. Raja-hliogga for example is the expression 
for a form of tenure peculiar to India. The holder of such a 
tenure, the rdja-hhoggo, was empowered to exact all dues accruing 
to the government within the boundaries of the district or estate 
granted to him. But he had not to render to government any 
account of the dues thus received by him. They were his per- 
quisite. He could hold his own courts, and occupied in many 
ways the position of a baron, or lord of the manor. But there was 
a striking difference. He could draw no rent. The peasantry 
had to pay him the tithe of the rice grown ; and though the 
amount was not always strictly a tithe, and by royal decree could 
be varied in different localities, the grantee could not vary it. 
So with the import, or ferry, or octroi duties. The rate of pay- 
ment, and the places at which the levy could be made, were fixed 
by the government. We have not enough cases of this tenure to 
be able to interpret with certainty the meaning of all the details, 
and limits of space prevent a discussion of them here. But the 
general principle is quite clear ^ It shoAvs how easy would be the 
grant to local notabilities of local government to this extent, -and 
how narrow was the line of distinction between the collection of 
dues by civil servants or farmers of the taxes and their collection 
by a grantee in this way. This custom, thus traced back to so 
early a period in the history of India, seems never to have fallen 
into abeyance. It certainly, in the period under discussion, was of 
manifest advantage. But it must be admitted that it is, to English 
ideas, very strange — so strange that our civilians made the mis- 
take, in Bengal, of regarding all such persons legally empowered to 
collect the land-tax as landlords, and of endowing them accord- 
ingly with the much greater privileges and powers of the English 
landlord. In the Buddhist period there is no evidence of the 
existence, in North India, of landlords in our sense of that term. 

It was the rise of this great power, Kosala, in the very centre 
of Northern India, which was the paramount factor in the politics 
of the time before the Buddhist reform. We do not know the 
details of this rise. But there are purely incidental references 
imbedded in the ethical teachings in the Buddhist books which 
afford us at least hints as to the final manner of it, and as to the 
date of it. For instance we have the story of Dighavu in the 

1 D. 1, 114, 127, 130: of. n, 50 and Bivy., 620, Vin. m, 221, with the Old Com- 
mentary at 222, M. m, 133, Jut. vi, 344, Sum. Vil. i, 245, 246. 


i8o The Early History of the Buddhists [ch. 

Vinaya^ There Brahmadatta, king of Kasi, invades Kosala, when 
Dighiti was king at Savatthi, and conquered and annexed the 
whole country ; but finally restored it to Dighlti's son, with whom 
he had become on very friendly terms. Other traditions inform 
us on the other hand of several invasions of the Kasi country by 
the then kings of Kosala, Vanka, Dabbasena, and Kamsa^. And 
when that most excellent story, the Rdjovada Jataka^ — as good in 
humour as it is in ethics — was first put together to represent two 
kings in conflict, the quite natural idea was to fix upon kings of 
Kosala and Kasi, and the author does so accordingly. 

No references have so far been found in the books as to any 
contests between Kosala and any other tribe or nationality. It 
would seem therefore that the gradual absorption into Kosala oV 
the clans and tribes in the northern part of Kosala as we know it 
in the Buddha's time took place without any such battle, campaign, 
or siege as was sufficiently striking to impress the popular imagin- 
ation; but that when Kosala came into contact with Kasi there 
ensued a struggle, with varying result and lasting through several 
reigns, which ended in the complete subjugation of the Kasi 
country by Kamsa, king of Kosala. 

As to the approximate period of these events, we see that they 
were supposed to have taken place not only, before the time of 
Pasenadi, who was born about the same time as the Buddha and 
lived about as long, but also before the time of his father the 
Great Kosalan. We have four kings of Kosala mentioned as 
taking part in these wars, and cannot be sure that there were not 
others who had quieter reigns. It would be enough and more 
than enough to allow, in round numbers, a century for all these 
kings. And the period cannot be much longer than that. For 
the name Brahmadatta could not have been older than towards 
the close of the Brahmana literature ; and a century and a half 
before the birth of the Buddha would about bring us to that. 

The king of Kosala in the Buddha's time was Pasenadi. He 
was of the same age as the Teacher*; and though never actually 
converted, was very favourable to the new movement, adopted its 
more elementary teachings, and was fond of calling upon the 
Buddha either to consult him or simply for conversation. A whole 
book of the Samyutta^ is devoted to such talks, and others are 
recorded elsewhere. They are mostly on religion or ethics, but 

1 Vinaya Texts, ii, 293-305. 

2 Jat. I, 262; n, 403; ni, 13, 168, 211 ; v, 112. » jataka n, 1. 
* M. II, 124. c The Kosala Sarnyutta, S. i, 68-102, 

vii] Ajatasattu 1 8 1 

some political and personal matters are occasionally mentioned 

For instance five 'rajas' are introduced discussing a point in 
psychology with Pasenadi. Whatever the title may exactly imply 
it is probable that we have the leaders of five clans or communities 
that, formerly independent, had, at that time, been absorbed into 
Kosala. Again we hear of a double campaign. In the first 
Ajatasattu, king of Magadha, attacks Pasenadi in the Kasi country 
and compels him to take refuge in SavatthiK In the second, 
Pasenadi comes down again into the plains, defeats Ajatasattu, 
and captures him alive. Then he restores to him the possession 
of his camp and anny, and lets him go fi-ee^. The commentaries 
inform us that he also gave him, on this occasion his daughter 
Vajira, to wife^ They also give the reasons for the dispute 
between the two kings ; but this will be better dealt with under 
the next heading. Another conversation arises when the king 
comes to tell the teacher of the death of his (the king's) grand- 
mother for whom he expressed his deep devotion and esteem. 
She had died at a great age, specified as 120 years, no doubt a 
round number*. At another talk Sumana, the king's sister, is 
present, and becomes converted. Desiring to enter the Order she 
refrains from doing so in order to take care of this same old lady, 
and attains Arahantship while still a lay- woman ^ The last and 
longest talk between the two friends took place at Medalumpa in 
the Sakiya country. The king, in much trouble with his family 
and ministers, expressed his admiration, and possibly also some 
envy, at the manner in which the teacher preserved peace in his 
Order. He then took his last leave with a striking declaration of 
his devotion^. But even as they were talking the crisis had come. 
The tradition records that the minister in whose charge the in- 
signia had been left when the king went on alone, had in his 
absence, proclaimed the king's son, Vidudabha, as king. Pasenadi 
found himself deserted by all his people. He hurried away to 
Rajagaha to get help from Ajatasattu, and, worn out by worry and 
fatigue, he died outside the gates of the city^ Ajatasattu gave 
him a state funeral, but naturally enough left Vidudabha un- 

The first use the latter made of his new position was to invade 
the Sakiya territory, and slaughter as many of the clan — men, 

^ S. I, 79. 2 5<, i^ 82-85. 3 j^t. n, 404; iv, 348. 

♦ S. I, 97; cf. Jat. iv, 146. « iS. i, 69; Thlg. 16. 

« M. n, 118-124. 7 Jatdka iv, 152. 

1 82 The Early History of the Buddhists [ch. 

women, and children — as he could catch. Many however escaped ^ 
and it is, perhaps, to this remnant that we owe the Piprahwa Tope 
discovered by Mr Pepp^. Elsewhere^ it has been shown that the 
reasons given for this invasion were probably not the real ones. 
But why should the Buddhists have taken pains so elaborately to 
explain away the fact, unless the fact itself had been indisputable? 
This is the last we know of Kosala. We hear nothing more of 
Vidiidabha, or of his successors if he had any. When the curtain 
rises again Kosala has been absorbed into Magadha. 

II. Magadha 

This was a narrow strip of country of some considerable length 
from north to south, and about twelve to fifteen per cent, in area 
of the size of Kosala. Just as Kosala corresponded very nearly 
to the present province of Oudh, but was somewhat larger, so 
Magadha corresponded in the time of the Buddha to the modem 
district of Patna, but with the addition of the northern half of the 
modern district of Gaya. The inhabitants of this region still 
call it Maga, a name doubtless derived from Magadha^. The 
boundaries were probably the Ganges to the north, the Son to the 
west, a dense forest reaching to the plateau of Chota Nagpur to 
the south, and Anga to the east. The river Champa had been the 
boundary between Magadha and Anga^ ; but in the Buddha's time 
Anga was subject to Magadha — it is the king, not of Anga, but of 
Magadha, who makes a land-grant in Anga (that is a grant of the 
government tithe) ^, and an Anga village is one of the eighty 
thousand parishes over which the king of Magadha holds rule and 
sovereignty^. All the clansmen in each of these two countries are 
called by Buddhaghosa, princes^ (exactly as he elsewhere calls the 
Sakiyas and Licchavis). The same writer says that the two 
kingdoms amounted together to 'three hundred leagues^.' It is 
reasonable to suppose, as he was born and bred in Magadha, 
that he was not so very far wrong. But this is said in reference 
to the time of Bimbisara. Later on he estimates the area of 
the whole of the United Kingdom of Magadha, in the time of 
Ajatasattu, at five hundred leagues. We may conclude from this 
that, according to the tradition handed down to Buddhaghosa, the 

^ DTip. A. I, 359; Malidvarnga vin, 18, and the Tikd on it. 

2 Buddhist India, pp. 11, 12, 3 Grierson in E.R.E. vi, 181. 

* Jdt. IV, 454— above, pp. 172-3. " Digha i, 111. « Vinaya i, 179. 

■? Rdja-kumdrd, Sum. i, 279, 294. See Early Buddhism, 27. 

8 Yojanas, Sum. i, 148. 

vii] Magadha 183 

size of the kingdom had nearly doubled in the interval. This 
would be about correct if the allusion were to Ajatasattu's con- 
quests north of the Ganges \ As Buddhaghosa however seems to 
use the larger figures of a date, not after, but at the beginning of 
those conquests, other wars of which we have no record, to the 
east or south, may be meant. 

The king of Magadha in the Buddha's time, was Bimbisara. 
Of his principal queens one was the Kosala Devi, daughter of 
Maha-Kosala, and sister therefore of Pasenadi^ ; another was 
Chellana, daughter of a chieftain of the Licchavis^ ; and a tliird 
was Khema, daughter of the king of Madda in the Punjab*. If .^ 
traditions of these relationships be correct they_^r% eloquent 
witnesses to the high estimate held in ot^li^ xjplptries of the 
then political importance of Magadha. :, Miera f ."^ ; 

Bimbisara had a son known as VecJehiT^tt* Ajaj«i>s«|,ttu;4in the 
canonical Pali texts, and as Kunika bj^the «Jains. ,>.??^ l^ter 
Buddhist tradition makes him a so^^Jf^i||Jl©<lCo6ala{:]p€^v)[ ; the Jaki 
tradition, confirmed by the standiiigrig|^ithet pf Yflfk^i-P^^itto, son 
of the princess of Videha, in the older, Buddhist books, makes him 
a son of Chellana. Buddhaghosa has preserved what is no doubt 
the traditional way of explaining away the evidence contained in 
the epithet^. But the matter cannot be further discussed here. 

One of the very oldest fragments preserved in the canon is a 
ballad on the first meeting of Bimbisara and Gotama. In the 
ballad the latter is called 'the Buddha.' But the meeting took 
place about seven years before he became the Buddha in our 
modern sense ; and this unwonted use of a now familiar title 
would have been impossible in any later document^ Gotama has 
only just started on his search for truth. The king, with curious 
density, ofiers to make him a captain, and give him wealth. It 
will be noticed that the king still resides in the palace of the old 
capital at the Giribbaja, 'the Hill Fort.' Some years afterwards 
when Gotama returns as a teacher, the king was lodged in the 
new palace that gave its name to the new capital, Rajagaha, ' the 
King's House.' The ruins of both these places are still extant ; 
and the stone walls of the Giribbaja are probably the oldest 
identified remains in India, Dhammapala says that the place was 
originally built or planned by Maha-Govinda, the famous architect, 

^ F. inf. , p. 185. 2 jf7^, xi, 403. * Jacobi, Jaina Sutras, i, xii-xv. 

•* Thig. A. on 139-143, and Apaddna quoted ibid. 131. 

5 Sum. I, 139. Cf. Dialogues, ii, 78. 

6 Sutta Nipdta, verse 408. See Dialogues, n, 2. The ballad is translated in Bh.D., 
Early Buddhism, 31-34. 

184 The Early History of the Buddhists [ch. 

to whom it was the proper thing to ascribe the laying out of 
ancient cities \ 

On Gotama's second visit to Rajagaha Bimbisara presented him 
with the Bamboo Grove, where huts could be built for the 
accommodation of the Order ^ — just as he endowed also the 
opposite teaching^ We hear very little about him in the books- 
He is not even mentioned in three out of the four Nikayas, and the 
few references in the fourth are of the most meagre kind. But the 
Vinaya gives a short account of an attempt made by Ajatasattu to 
kill his father with a sword*, and in the closing words of the 
mnorphala there is an allusion to the actual murder which he 
r,«ommitted^ The commentary on that Suttanta gives 
'•io^ aiik^n^MJf how it happened^ The details may or may not 
[1)6 true ; but the main fact^that Bimbisara was put to death by 
\i^ -^brt'^-^viri^bftiv »«-!>vsibe' a6ctet)ted as historical. The Ceylon 
[?^l'^i61r.2 - : I -^/fevfeiit''' eight years before the Buddha's 

Ib'ffithj'at i.h^'^i! • ^^fei*l!fcfeara, who had come to the throne 
when he w^^^feftfefciij- i itid' i^ftl^ifri^d fifty-two years ^. 

On the death oJT MWfe^ti'tt, his wife, the Kosala Devi, is said by 
tradition to have died of griefs. The government revenues of an 
estate in Kasi had been settled upon her by Maha-Kosala as pin- 
money on her marriage. At her death the payment of course 
ceased. Ajatasattu then invaded Kasi. It seems incredible that 
this could have been the real motive of the war, unless the kings 
of that place and time were less expert in inventing pretexts for a 
war which they wanted than modern kings in Europe. The war 
itself is however mentioned in the Canon ^, and with some detail. 
In the first campaign Ajatasattu out-manoeuvred his aged uncle, and 
drove him back upon Savatthl. In the next, however, Pasenadi 
lured his nephew into an ambush, and he was compelled to 
surrender with all his force. But Pasenadi soon set him at liberty, 
gave him back his army, and, according to the commentary, gave 
him also one of his daughters in marriage. 

In the opening paragraph of the Mahd-parinibbdna Suttanta^" 
we hear of Ajatasattu's intention to attack the Vajjian con- 
federacy, and, as the first step in the attack, of his building a 

1 Vimana-vatthu Commentary, p. 82, and above p, 173. 

2 Vinaya i, 39. » Digha i. 111, 127. 

* Vinaya n, 190. i^ Digha i, 86. « Sum. i, 133-136; Peta-v. A. 105. 

7 Dip. ra, 56-60 ; Mhv. 11, 29, 30. ^ jat. n, 403. 

9 Samyutta i, 84-86. Cf. Dhp. A. ra, 259; Jdt. iv, 342. 
" Rh.D., Dialogues of the Buddha, n, 78. 

VI i] Avanti 185 

fortress at Pataliputta, the modern Patna, on the south bank of 
the Ganges, the then boundary between his territory and theirs. 
The minister in charge of this work was a Brahman, known to us 
only by his official title, 'the Rain-maker' {Vassakara). He fled 
suddenly to the Vajjian capital Vesall, giving out that he had 
barely escaped with his life from Ajatasattu. The Vajjians gave 
him refuge and hospitality. He then dwelt among them, carefully 
disseminating lies and slanders until he judged the unity of the 
confederation to be finally broken. Three years after his kindly 
reception he gave the hint to his master, who swooped down on 
Vesali, and destroyed it, and treated his relatives very much as 
Vidudabha had treated his. We can only hope this ghastly story 
of dishonour, treachery, and slaughter is a fairy-tale. The ques- 
tion can only be discussed with profit when we have the whole of 
the conunentary before us. 

The son of Ajatasattu is mentioned in the Canon \ His name 
was Udayi-bhadda, and it follows from the statements of the 
Ceylon Chronicles that he succeeded his father on the throne^. 
This is confirmed in the commentaries ^ The name also occurs in 
medieval Jain and Hindu lists, independent no doubt, both of 
them, of the Buddhist books ^ 

HI. Avanti 

The king of Avanti in the Buddha's time was Pajjota the 
Fierce, who reigned at the capital Ujjeni. There is a legend 
about him which shows that he and his neighbour king Udena of 
Kosambi were believed to have been contemporaries, connected by 
marriage, and engaged in war^ The boundary is not given, but a 
commentary mentions incidentally that the two capitals were in 
round numbers fifty yojanas^ about four hundred miles, apart. We 
have seen that when the Nikayas were composed Avanti was con- 
sidered to have been one of the important kingdoms of India 
before the Buddha's time^. Shortly after the Buddha's death 
Ajatasattu is said to have been fortifying his capital, Rajagaha, in 
anticipation of an attack by Pajjota of Avanti ^ The king of the 
Surasenas, at Madhura, in the Buddha's time, was called Avanti- 
putto ; and was therefore almost certainly the son of a princess of 
Avanti ^ The Lalita-vistara gives the personal name of the king 

^ Digffta i, 50 = Dialogues, i, 68. ^ Dip. v, 97; Mhv. it, 1. 

» Smp. 321; Sum. i, 153-4. * V. inf., pp. 189-90. 

» Buddhist India, 4-7. * Above, p. 172. Cf. Jdt. iv, 390. 

' M. Ill, 7. 8 M. n, 83. 

1 86 The Rarly History of the Buddhists [ch. 

of Madhura in the year of the Buddha's birth as Subahu^ and this 
may be the same person. 

Avanti became from the first an important centre of the new 
doctrine we now call Buddhism (in India it was not so called till 
centuries later). Several of the most earnest and zealous adherents 
of the DhamTYia Avere either born or resided there. Abhaya 
Kumara is mentioned^ and Isidasi^ and Isidatta* and Dhammapala^ 
and Sona Kutikanna®, and especially Maha-Kaccanal The last of 
these is stated to have been called by the Buddha the most pre- 
eminent of those of his disciples able to expound at length, both 
as to form and meaning, that which had been said in short. The 
last but one, Sona, was in a similar way declared to be the most 
eminent of the disciples distinguished for beauty of expression ^ 
In what language were they supposed to have exercised these 
literary gifts ? It was certainly not the religious language then 
current in the priestly schools of Brahmanism. This archaic form 
of speech which has been preserved in the Brahmanas and 
Upanishads was called by the grammarians chhandasa, 'the 
language of chhandas or Vedic poetry,' to distinguish it from 
the laukiha or * secular ' language ; and the Buddha had 
expressly forbidden his 'word' to be put into chhandas. Each 
disciple was to speak the word in his own dialect^. It would 
be a mistake, however, to be misled by the ambiguities of the 
word dialect, and to suppose it to mean here the language 
as spoken by any peasantry. The higher ethics and philosophy 
of 'the Word' could not be discussed in any such dialect 
Now for two or three generations before the birth of the 
Buddha, the so-called Wanderers ^^ were in the habit of passing 
from Avanti to Savatthi, from Takkasila to Champa, discussing 
in the vernacular, wherever they went or stayed, precisely such 
questions. They had invented or adapted abstract words and 
philosophical or ethical terms useful for their purpose, and equally 
current in all the dialects ; while during the same period there had 
been developed in the rising kingdoms, and especially in Kosala 
(in the very centre of the regions covered by the Wanderers, and 
by far the largest and most important of them all) the higher 

1 Ed. Eajendralal Mitra, p. 24. 2 ^hag. A. 39. 

8 Thlg. A. 261-4. " S. iv, 288; Thag. 120. 

6 Thag. 204. « Vinaya Texts, n, 32; Thag. 369; Ud. v, 6. 

7 Samyutta iii, 9; iv, 117; Anguttara i, 23; v, 46; Majjhima ni, 194, 223. 

8 CL Dhp. A. IV, 101. 

» Vinaya n, 139. Cf. the note in Vinaya Texts, ni, 150. 
10 See Buddhist India, 141-146. 

vii] The Vamsas 187 

terms necessary for legal and administrative purposes. Just as 
the Christians adopted for their propaganda, not classical Greek 
but the Greek of the Koine, the varying dialect understood 
through all the coasts and islands of the Eastern Mediterranean, 
which they found ready to their hands ; so the Buddha and his 
followers adopted this common form of vernacular speech, varying 
no doubt slightly from district to district, which they found ready 
to their hands. The particular fonn of this common speech, the 
then Hindustani, in which the Pali Canon was composed, was 
almost certainly, as the present writer ventured to suggest 
nearly forty years ago on historical grounds ^ and as Professor 
Franke contends on philological grounds^, the form that was 
current in Avanti^. If that be so, it could be said that Buddhism, 
bom in Nepal, received the garb in which we now know it in 
Avanti, in the far West of India. It is true that no such curt 
summary of a great movement can be sufficient. But this would 
be nearer to the facts than that other summary, so often put 
forward as convenient, that Buddhism arose in Magadha and that 
its original tongue was Magadhi^ 

IV. The Vamsas 
The King of the Vamsas in the Buddha's time is called in the 
Canon Udena^ His father's name was Parantapa, and his son's 
name Bodhi Kumara^ But Udena survived the Buddha ^, and we 
are not informed whether Bodhi did or did not succeed him on 
the throne. Tradition has preserved a long story of the adventures 
of Udena and his three wives. We have it in two recensions — a 
Pali one, the Udena-vatthu^ ; and a Sanskrit one, the Makandika- 
avadana^. It is quite a good story, but how far each episode may 
be founded on fact is another question. The capital was KosambI, 
the site of which has been much discussed^". It seems to have 
been on the south bank of the Jumna, at a point about 400 
miles by road from UjjenI, and about 230 miles up stream from 

1 Eh.D. in Trans. Phil. Soc. 1875. ^ r. otto Franke, Pali und Sanskrit, 1902, 

3 Cf. Windisch, Algiers Cong, of Orientalists, 1906; and Bh.D., Buddhist India, 

* For this view see the references given by Wintemitz, Gesch. d. ind. Lit. ii, i, 
p. 10, note 3. ® Uddna vn, 10; Samyutta rv, 110-113, 

8 Vinaya n, 127; iv, 198; Majjhima n, 97; Jdtaka in, 157. 

' Peta-vatthu Commentary 140. 

8 In Norman's Dhammapada Commentary i, 161-280. 

8 Divydvaddna 515-544. (Ed. Cowell and Neil, ) 

w For different views see T, Watters, On Yuan Chwang, i, 366-9 and Chapter xxi, 
p. 524, infra. 

1 88 The Early History of the Buddhists [ch. 

Benares ^ One route from UjjenI to KosambI lay through Vedisa, 
and other places whose names are given but of which nothing else 
is at present known ^, There were already in the time of the 
Buddha &ur establishments or settlements of the Order in or near 
KosambI, each of them a group of huts under trees. One of them 
was in the drama or pleasaunce of Ghosita, two more in similar 
parks, and one in Pavariya's Mango Grovel The Buddha was 
often there, at one or other of these settlements; and discourses 
he held on those occasions have been handed down in the Canon. 
King Udena was at first indifferent or even unfriendly. On one 
occasion, in a fit of drunken jealousy he tortured a leading member 
of the Order, Pindola Bharadvaja, by having a basket full of 
brown ants tied to his body *. But long afterwards, in consequence 
of a conversation he had with this same man Pindola, he professed 
himself a disciple. We have no evidence that he progressed very 
far along the path; but his fame has lasted in a curious way in 
Buddhist legends. For instance there is an early list of the seven 
Con-natals {sahajcdd), persons born on the same day as the 
Buddha^ The details of the lists differ; and already in the Lalita- 
vistara it has grown into several tens of thousands, still arranged 
however in seven groups ^ Many centuries afterwards we find the 
name of Udena appearing in similar lists recurring in Tibetan and 
Chinese books ^ 

The First Great Gap 

The passages referred to above tell us a good deal of the 
political condition of India during the Buddha's life, and enable 
us to draw certain conclusions as to previous conditions for some 
time before the birth of the Buddha. There are also one or two 
passages in the Canon which must refer to dates after the Buddha's 
death. Perhaps the most remarkable is the verse in the Parayana 
(a poem now included in the Sutta Nipata) which, referring to a 
time when the Buddha was alive, calls Vesali a Magadha city^ 
Now we know from the Mahd-parinihhdiia Suttanta that (at the 
time when that very composite work was put together in its 
present shape) Vesali and the whole Vajjian confederacy was 
considered to have remained independent of Magadha up to the 

1 Buddhist India, p, 36. ^ Sutta Nipata, 1011. 

' Vin. IV, 16; Sum. 319. ■» Jdtaka iv, 375. 

6 See Eh.D., Buddhist Birth Stories, note on p. 68. 

8 Lalita-vistara ed. Rajendralal Mitra, p, 109. 

' Eockhill, Life, 16, 17; T. Waiters, On Yuan Chwang, i, 368. 

8 Sutta Nipata, 1013. 

vii] Kings of Magadha 189 

end of the Buddha's life\ If therefore the reading in our text of 
the Parayana be correct, the expression 'Magadha city' must be 
taken in the sense of 'now a Magadha city,' and as alluding to the 
conquest of Vesall as described above, p. 185. But it is apparently 
the only passage in the Canon which takes cognisance of that 
event. Again in the Anguttara we have a sutta^ in which a king 
Munda, dwelling at Pataliputta, is so overwhelmed with grief at 
the death of his wife Bhadda that he refuses to have the cremation 
carried out according to custom. But after a simple talk with a 
thera named Narada he recovers his self-possession. We learn 
from the chronicles that King Munda was the grandson of 
Ajatasattu and began to reign about the year 40 a.b.^ It is a 
fair inference from this episode that Pataliputta had already at 
that time become the capital of Magadha. Narada is said to have 
lived in the Kukkutarama, no doubt consisting of a few huts or 
cottages scattered under the trees in the pleasaunce so called. It 
was a well-known resting-place for the Buddhist Wanderers, and 
Asoka is said to have built a monastery on the site of it*. 

The long poem of old Parapariya, a laudator temporis acti, on 
the decay of religion since the death of the Master', adds nothing 
to political history. So also the edifying ghost-story recorded in the 
Peta-vatthu (ii, 10) can only, at most, give us the name of a sort 
of public- works officer at Kosambi shortly after the Buddha's death. 

These few details are all that we can glean from the Thera vada 
Canon concerning the history of India for more than a hundred 
and sixty years. And the chroniclers and commentators do not 
add very much more. They have preserved indeed a dynastic list 
of the kings of Magadha with regnal years of most of the kings. 
The list is as follows : 

Ajatasattu reigned 
Mupda / 




























4L7 -oy 


^oi - 29 y 


^9 7- ^4- 


iH - > 


3 ^i - 34 

His 10 sons 
Nine Nandas 

1 Dialogues, n, 78-80. ^ A. m, 57-63. 

• Mdhdvamia rv, 2, 3 ; Divyavaddna 369. 

* -Sf. V, 171; J. V, 342; M. i, 350; Divy. 368, 434; T. Watters, On Yuan Chwarig, 
n, 98, 99. 

» Thera-gdthd, 920-948. 


190 The Early History of the Buddhists [ch. 

There are other lists extant, not so complete, and not always 
with the regnal years given, in Jain, Hindu, or Buddhist Sanskrit 
works. They have been carefully compared and discussed by 
W. Greiger, in a very reasonable and scholarly way^. He comes to 
the conclusion that, on the whole, the above list is better 
supported than the others. This may well be the case ; but at the 
same time it must be confessed that the numbers seem much too 
regular, with their multiples of six and eight, to be very probably 
in accordance with fact. And we are told nothing at all of any of 
the other kingdoms in India, or even of the acts of the kings thus 
named, or of the extent of the growing kingdom of Magadha 
during any of their reigns. The list gives us only the bare bones 
of the skeleton of the history of one district. 


When the curtain rises again we have before us a picture 
blurred and indistinct in detail, but in its main features made 
more or less intelligible by what has been set out above. 

India, as shown in the authorities there quoted, appeared as a 
number of kingdoms and republics with a constant tendency 
towards amalgamation. This process had proceeded further in 
Kosala than elsewhere ; that great kingdom being by far the most 
important state in Northern India, and very nearly if not quite as 
large as modern France. It occupied the very centre of the terri- 
tories mentioned in those authorities; it had its capital near the 
borders of what is now Nepal ; and it included all the previous 
states or duchies between the Himalayas on the north and the 
Ganges on the west and south. The original nucleus of this 
great kingdom was the territory now the seat of the Gurkhas, and 
these Kosalans were almost certainly, in the main at least, of 
Aryan race. For the heads of houses among them (the gahapatis) 
are called rdjdno, the same as the clansmen (the Tcula-puttd) in 
the free republics. Of the surrounding kingdoms Magadha, though 
much smaller, was the most progressive. It had just absorbed 
Anga, and at the last moment we saw it attacking, and with 
success, the powerful Vajjian confederation. The rise of this new 
star in the extreme South-East was the most interesting factor in 
the older picture. 

The new picture as shown in the Ceylon chronicles and in the 
classical authors (especially those based on the statements in 

1 Mahdvaipsa (English translation), Intr. pp, zl-xlvi. 

vii] Chandagutta 191 

the Indika of Megasthenes) show us Magadha triumphant. All 
the kingdoms, duchies, and clans have lost their independence. 
Even the great Kosalan dominion has been absorbed. And for the 
first time in the history there is one paramount authority from 
Bengal to Afghanistan, and from the Himalayas down to the 
Vindhya range. 

We shall probably never know how these great changes, and 
especially the fall of Kosala, were brought about. And we have 
no information as to the degree in which the various local authori- 
ties retained any shadow of power. Were the taxes fixed by the 
central power and collected by its own officers? Or were the local 
rates maintained and collected by a local authority? If the latter, 
were the actual sums received paid over to the central office at 
Pataliputta, or was a yearly tribute fixed by the paramount power? 
On these and similar questions we are still quite in the dark. 
But our two sets of authorities, which are quite independent of 
one another, agree in the little they do tell us. 

Unfortunately each set is open to very serious objections. 
The Chronicles are quite good as chronicles go, and we have them 
not only complete but well edited and translated. But of course 
we cannot expect from documents written fifteen hundred years 
or more ago, any of that historical criticism that we are only just 
beginning to use in the West. They are written throughout for 
edification, and in the Mahavamsa sometimes also for amusement ; 
they are in verse, and are not infrequently nearer to poetry than 
history ; and though based on a continuous tradition, that tradition 
is now lost. On the other hand, the work of Megasthenes, written 
during the life-time of Chandagutta, is itself lost. What we have 
are fragments preserved more or less accurately, and with the best 
intentions, by later Latin and Greek authors. Where what is 
evidently intended as a quotation from the same passage in 
Megasthenes is found in more than one of these later authors the 
presentations of it do not, in several cases, agree. This throws 
doubt on the correctness of those quotations which, being found 
in one author only, cannot be so tested. A number of the quota- 
tions contain statements that, as they stand, are glaringly absurd 
— stories of gold-digging ants, men with ears large enough to sleep 
in, men without mouths, and so on. Strabo therefore calls Megas- 
thenes mendacious. But surely such stories (and other things) 
only show that Megasthenes was just as ignorant of the modern 
rules of historical evidence as the Chroniclers were, and for the 
same reason. Strabo's idea of criticism is no better than that of 

192 The Early History of the Buddhists [ch. 

those who ignore the Chroniclers on the ground that they are 
mendacious. As will be seen in Chapter xvi which deals more 
fully with the Greek and Latin writers on Ancient India, it is more 
probable that in these fairy-tales of his Megasthenes, like Herodotus 
before him, had either accepted in good faith stories which were 
current in the India of his day, or had merely misunderstood some 
Indian expression. 

Age of the Authorities used 
It remains now to give some account of the literature from 
which our knowledge of early Buddhism is chiefly derived, and so 
form some estimate of its value as a source of history. This 
literature which deals mainly with ethics and religion, grew up 
gradually among those followers of the Buddha who dwelt in the 
republics and kingdoms specified above. There are now 27 books, 
and only three of them deal with the rules of the Order. But these 27 
are mostly anthologies of earlier shorter passages. The Patimokkha 
for instance — one of the earliest documents — has 227 suttas, and 
they are of the average length of about three lines ; and the Silas, 
a string of moral injunctions, are, if taken separately, quite short. 
But neither of these tracts, each of them already a compilation, 
now exists as a separate book. They are found only as imbedded 
in longer works of later date. It took about a century for the 
more important works, the Vinaya and the four Nikayas^, to be 
nearly finished about as we have them. (See p. 195.) 

The next century and a half saw the completion of the supple- 
mentary works — the supplements to the Vinaya and the four 
Nikayas; the thirteen books of the supplementary fifth Nikaya 
(much of it based on older material) ; and the seven Abhidhamma 
books, mainly a new classification of the psychological ethics of 
the four Nikayas. 

So far the books had been divided into Dhamma and Vinaya ; 
that is to say, religion and the regulations of the Order. Now, 
after the close of the canon, a new division begins to appear, that 
into three Pitakas (or Baskets) of Vinaya, Sutta, and Abhidhamma. 
We do not yet know exactly when or why this new division arose 
and superseded the older one^. As late as the fifth cent. a.d. we 

1 The titles of the five Nikayas are as follows: 1. Digha=the long Suttas; 
2. Majjhima=the Suttas of medium length; 3. Samyutta = Suttas forming connected 
groups; 4. Anguttara = Suttas arranged according to a progressive enumeration (from 
one to eleven) of the subjects with which they deal; 5. Khuddaka= smaller works and 

' Perhaps the oldest reference to the three Pitakas is in Eanishka's Inscr., Ep. 
Ind. vm, 176. 

vii] Antiquity of Sources 193 

find Buddhaghosha still putting the Vinaya and the Abhidhamma 
into the supplementary fifth Nikaya^, though he and other com- 
mentators also use the newer phrased 

The authorities on which our account of early Buddhist history 
is based are therefore the four Nikayas, with occasional use of 
other works mainly of such as are included in the fifth or supple- 
mentary Nikaya. Concerning the period to which the Nikayas 
belong we have some evidence, partly internal and partly external. 
To take the latter first : 

Asoka in the Bhabra Edict recommends his co-religionists the 
special study of seven selected passages. Two of the titles given are 
ambiguous. Four of the others are from the four Nikayas, and the 
remaining one from the Sutta Nipata now included in the fifth 
Nikaya. As was pointed out a quarter of a century ago^ it is a critical 
mistake to take these titles as the names of books extant in Asoka's 
time. They are the names of edifying passages selected fi-om an 
existing literature. It is as if an old inscription had been found 
asking Christians to learn and ponder over the Beatitudes, the 
Prodigal Son, the exhortation to the Corinthians on Charity, and so 
on. There are no such titles in the New Testament. Before short 
passages could be spoken of by name in this familiar manner a 
certain period of time must have elapsed ; and we should be 
justified in assuming that the literature in which the passages 
were found was therefore older than the inscription ^ 

Further, in certain inscriptions in the Asoka characters of a 
somewhat later date there are recorded names of donors to 
Buddhist monuments. The names being similar, distinguishing 
epithets are used — X. who knows Suttantas, X. who knows the 
Pitaka (or perhaps the Pitakas, Petdkl), X. who knows the five 
Nikayas. These technical terms as names for books are, with one 
exception, found only in that collection we now call the Pali 
Pitakas. The exception is the word Pitaka. That is not found 
in the four Nikayas in that sense ; and even in the fifth Nikaya it is 
only approximating to that sense and has not yet reached it. One 
would naturally think, if these Nikayas had been put together after 
these inscriptions, that they would have used the term in the sense 
it then had, and has ever since continued to have ; more especially 
as that sense — the whole collection of the books — is so very con- 
venient, and expresses an idea for which they have no other word. 

1 Attha-sdlini, 26. 

2 Ibid. 27; Sum. Vil. i. 15. So also Mil. 21, 90; Thig. A. 199; Dhp. A. ra. 385. 
8 Rh.D., Questions of King Milinda, i, xxxvii ff. * See J.P.T.S., 1896. 

C.H.I. I. 13 

194 ^'^^ Early History of the Buddhists [ch. 

Thirdly, the commentators both in India and Ceylon say that 
the Katha-vatthu, the latest book in the three Pitakas, as we now 
have them, was composed by Moggliputta Tissa at Asoka's court 
at Pataliputta in N. India at the time of the Council held there in 
the eighteenth year of Asoka's reign. At the time when they 
made this entry, the commentators held the Pitakas to be the 
word of the Buddha, and believed also that the Dhamma had been 
already recited at the Council held at Rajagaha after the death of 
the Buddha. It seems quite impossible, therefore, that they could 
have invented this information about Tissa. They found it in the 
records on which their works were based ; and felt compelled to 
hand it on. Being evidence, as it were, against themselves, it is 
especially worthy of credit. And it is in accord with all that we 
otherwise know. Anyone at all acquainted with the history of the 
gradual change in Buddhist doctrine, and able to read the Katha- 
vatthu, will find that it is just what we should expect for a book 
composed in Asoka's time. It has now been edited and translated 
for the Pali Text Society ; and not a single phrase or even word 
has been found in it referable to a later date. It quotes largely 
from all five ISTikayas^ 

The above is all the external evidence as yet discovered, and 
the third point, though external as regards the Nikayas, is internal 
as regards the Pitakas. The internal evidence for the age of the 
Nikayas is very small, but it is very curious. 

Firstly, the four Nikayas quote one another. Thus Anguttara 
V, 46 quotes Samyutta i, 126 ; but in giving the name of the 
work quoted it does not say Samyutta, but Kumdri-panha — the 
title of the particular Sutta quoted. The Samyutta quotes two 
Suttantas in the Digha by name — the Sakka-panha and the 
Brahma-jdla\ It follows that, at the time when the four Nikayas 
were put together in their present form, Suttas and Suttantas 
known by their present titles were already current, and handed 
down by memory, in the community. 

More than that there are, in each of the four Nikayas, a very 
large number of stock passages on ethics found in identical words 
in one or more of the others. These accepted forms of teaching, 
varying in length from half a page to a page or more, formed part 
of the already existing material out of which the Nikayas were 
composed. Some of the longer Suttantas consist almost entirely 
of strings of such stock passages ^ 

1 See the passages collected in Dialogues of the Buddha, i, pp. xi, xii. 

2 S. Ill, 13 (with a difference of reading), and S. iv, 287. 

3 For instance, the Samglti, D. in, 207. 

VI i] The Nikayas 195 

There are also entire episodes containing names of persons and 
places and accounts of events — episodes which recur in identical 
terms in two or more of the Nikayas. About two-thirds of the 
Mahd-parinibhdna Suttanta consists of such recurring episodes or 
stock passages^. This will help to show the manner in which the 
books were built up. 

Several conversations recorded in the Nikayas relate to events 
which occurred two or three years after the Buddha's death ; and 
one passage (Anguttara iii, 57-62) is based on an event about 40 
years after it. 

The four Nikayas occupy sixteen volumes of Pali text. They 
contain a very large number of references to places. No place on 
the East of India south of Kalinga, and no place on the West of 
India, south of the Godavari, is mentioned. The Asoka Edicts, 
dealing in a few pages with similar matter, show a much wider 
knowledge of South India, and even of Ceylon. We must allow 
some generations for this increase of knowledge^. 

At the end of each of the four Nikayas there are added 
portions which are later, both in language and in psychological 
theory, than the bulk of each Nikaya. 

All the facts thus emphasised would be explained if these 
collections had been put together out of older material at a period 
about half way between the death of the Buddha and the accession 
of Asoka. Everything has had to be stated here with the utmost 
brevity. But it is important to add that this is the only working 
hypothesis that has been put forward. It is true that the old 
battle cries, such as 'Ceylon books' or 'Southern Buddhism' are 
still sometimes heard. But what do they mean? The obvious 
interpretation is that the Pali Pitakas were composed in Ceylon — 
that is, that when the Ceylon hhiTckhus began to write in Pall 
(which was about Buddhaghosa's time) they wrote the works on 
which Buddhaghosa had already commented. This involves so 
many palpable absurdities that it cannot be the meaning intended. 
Until those who use such terms tell us what they mean by them, we 
must decline to accept as a working hypothesis the vague insinua- 
tion of question-begging epithets. We do not demand too much. 
A working hypothesis need not propose to settle all questions. 
But it must take into consideration the evidence set out above ; 
and it must give a rational explanation of such facts as that this 

* See the table of references, and detailed discussion, in the introduction to Dxa- 
logves, n, 71-77. 

2 This point is discussed more fully in Buddhist India, pp. 28-34. 


196 The Karly History of the Buddhists [ch. 

literature does not mention Asoka, or S. India, or Ceylon; and 
that, though there is a clear progress in its psychology and its 
Buddhology, it gives no connected life of the Buddha, such as we 
find in Sanskrit poems and Pali commentaries. 

On the last point the evidence, being very short, may be given 
here. There are a large number of references to the places at 
which the Buddha was stopping, when somie conversation or other 
on an ethical or philosophical question took place. These have 
not yet been collected and analysed. Then there are a small 
number of short references, in a sentence or two or a page or two, 
to some incident in his life. And lastly we have two episodes, of 
a considerable number of pages, describing the two important 
crises in his career, the beginning and the close of his mission. 
Out of approximately 6000 pages of text in the four Nikayas less 
than two hundred in all are devoted to the Buddha's life. 

Of the long episodes the first is in the Majjhima\ and describes 
the events of the period fi-om the time when he had first become a 
Wanderer down to his attainment of Nihhoma (or Arahantship) 
under the Bodhi Tree^ The events are not the names and dates 
of kings and battles, but events in religious experience, the 
gradually increased grasp of ethical and philosophical concepts, 
the victory won over oneself. The Vinaya, very naturally, con- 
tinues this episode down to the time of the founding of the Order, 
the sending forth of the sixty and the accession of the most 
famous of the Arahants^ This episode covers about seven years, 
the Vinaya addition to it being responsible for one. The other 
long episode, about twice as long as the first, describes in detail 
the events of the last month of the Buddha's life. It is contained 
in the Digha, and forms a whole Suttanta, the Mahd-parinibbcma 
Suttanta, referred to above as a composite document. 

We have no space to consider the shorter references ; but the 
following table specifies the more important, arranged chrono- 
logically : 

1. Youth ; three residences. Digha 11, 21 ; Ang. i, 145. 

2. The going forth. Digha i, 115; 11, 151; Ang. i, 146; Majjhima i, 163; 

S. N. 405-424. 

3. His teachers. Majjhima i, 163 ; Saijiyutta iv, 83 ; Digha in, 126. 

4. His trial of asceticism. Majjhima i, 17-24, 114, 167, 240-248. 

5. Nibbana. Majjhima i, 23, 116-118, 167, 173, 248-250 ; Vinaya i, 1-4. 

» Vol. I, pp. 163-175 and 240-249. Bepeated at vol. 11, p. 93, and again p. 212. 
2 The word Nibbana occurs, p. 167. 
9 Vinaya i, 1-44. 

vii] Growth of Literature 197 

6. Explanation of the Path. Saipyutta in, 66 ; iv, 34 ; v, 421 ; Majjhima I, 135, 

300 ; Vinaya i, 8-14. 

7. Sending out of the Sixty. Samyutta i, 105 ; Vinaya i. 21. 

8. The last month. Digha ii, 72-168. 

The relative age, within the Canon, of each of these passages, 
has to be considered as a question distinct from that of the books 
into which they are now incorporated. Towards the solution of 
these questions some little progress has been made, and the tenta- 
tive conclusions so far reached are shown in the following table. 

Growth of Buddhist Literature from the time of the 
Buddha down to Asoka. 

1. The simple statements of doctrine now found in identical 
words recurring in two or more of the present books — the stock 
passages or Suttas. 

2. Episodes (not of doctrine only) similarly recurring. 

3. Books quoted in the present books but no longer existing 
separately — the Silas, the Parayana, the Octades, the Patimokkha, 

4. Certain poems, ballads, or prose passages found similarly 
recurring in the present anthologies, or otherwise showing signs 
of greater age. 

5. The four Nikayas, the Sutta Vibhanga and the Khandakas. 
Approximate dates 100 A.B. 

6. Sutta Nipata, Thera- and Theii-gatha, the Udanas, the 
Khuddaka Patha. 

7. The Jatakas (verses only), and the Dhammapadas. 

8. The Niddesa, the Iti-vuttakas, and the Patisambhida. 

9. The Peta- and Vimana-vatthu, the Apadanas, and the 

10. The Abhidhamma books, the latest of wKich is the Katha- 
vatthu'and the oldest, perhaps, the Dhamma-sangani. 



The following analysis is constructed from a number of incidental 
allusions to economic conditions in the great Pali thesaurus of the 
Jataka, and, to a more limited extent in the Vinaya, and also in 
the other books of the Sutta Pitaka, of which the Jataka is a part. 
Dr Fick's admirable monograph Die sociale Gliederung in Indien 
is similarly based. That work deals chiefly with social conditions. 
The present chapter, on the other hand, is mainly economic in 
scope, and only in a minor degree sociological. It is true that the 
evidence is drawn very largely from stories. But it is fairly clear 
that the folk in those tales have given them a parochial setting 
and local colour. And this is frequently borne out by the coin- 
cident testimony of other books not dealing with folk-lore. 

The rural economy of India at the coming of Buddhism was 
based chiefly on a system of village communities of landowners, or 
what in Europe is known as peasant proprietorship. The Jataka 
bears very clear testimony to this. There is no such clear testimony 
in it to isolated large estates, or to great feudatories, or to absolute 
lords of the soil holding such estates. In the monarchies, the 
king, though autocratic and actively governing, had a right to a 
tithe on raw produce, collected as a yearly tax ; and only to this 
extent could considered the ultimate owner of the soil. All 
abandoned, all forest land the king might dispose of ^ ; and under 
this right was included the reversion to the crown of all property 
left intestate or 'ownerless '^ a custom which may or may not be a 
survival of an older feudalism. The sovereign was moreover en- 
titled to 'milk money,' a perquisite paid by the nation when an 
heir was born to him^, and he could declare a general indemnity 
for prisoners at any festal occasion ^ Besides these privileges he 

1 D. I, 87. . 

* S. I, 89 {Kindred Sayings, i, 115) ; Jdt. in, 3026; cf. iv, 485; vi, 348. 

s lb. IV, 323. * lb. IV, 176; v, 285; vi, 327. 

CH. VI 1 1] Tithes and Taxes 199 

could impose forced labour or rdjahdriya on the people, but 
this may have been limited to the confines of his own estates. 
Thus the peasant proprietors enclose a deer-reserve for their king, 
that they might not be summoned to leave their tillage to beat up 
game for him^. A much more oppressive extent of corvee is pre- 
dicted only of a state of civic decay ^. The tithe on produce was 
levied in kind, measured out either by the village syndic or head- 
man (gdma-hhqjaka), or by an official (a mahdmatta) at the barn 
doors ^, or by survey of the crops*. Some of the rice and other 
grain may presumably have been told off for the special granaries 
kept filled for urgency, in war or famine ^ but Buddhist books make 
no clear reference to such an institution. The amount levied seems 
to have varied from \ to ^^, according to the decision of the ruling 
power ^ or other circumstances. And the contributions raised at 
one or more gdmas (villages), rural or suburban, could be made 
over by a monarch (or by his chief queen ^) to anyone he wished to 
endow, e.g., to a daughter on her marriage", a minister^", a Brahman", 
a merchant, etc.'^ Again, the king could remit the tithe to any 
person" or group'*. 

We have no direct evidence of such a tithe or other tax being 
levied on the commonwealth by any of the republics or oligarchies 
mentioned in the Buddhist canon, such as the Sakiyas, Koliyas, 
Licchavis, Mallas, etc.'^ But that they did so raise the state 
revenue, in the case at least of the Sakiyas, seems to be attested 
by Asoka's inscription on the LumbinI or Rummindei pillar^®. 
The tithe thus remitted on the occasion of Asoka's visit to the 
birthplace of the Buddha, must have been imposed by the Sakiyas 
at a date prior to the Mauryan hegemony. The Sakiyas and 
other republics are recorded as meeting for political business at 
their own mote-halls ''', and must inevitably have had a financial 
policy to discuss and carry out. That their enactments could be 
somewhat drastically paternal appears in the case of the Malla 
clansmen of Kusinara, who imposed a fine of 500 (pieces) on any- 
one who 'went not forth to welcome the Blessed One' when he 

* At Benares, Jdt. i, 149 ; the Anjana Wood at Saketa, ib. m, 270. 

8 Ih. I, 339. A certain familiarity with oppressive taxation is suggested by ih. v, 
99 ff ; cf. I, 339 ; ii, 240. =« Ib. ii, 378. 

* Ib. IV, 169. s ind. Ant., 1896, pp. 261 f. 

^ Cf. Gautama x, 24; Manu. vii, 130; Biihler, Trans. Vienna Acad. Jan. 1897; 
V. A. Smith, J'.E.J.S., 1897, 618 f. ''Jdt. in, 9. ^ Ib. v, U. 

» Ib. II, 237, 403. i» 76. i, 354; vi, 261. 

" D. I, 87 ; Jdt. Ill, 229. ^^ Ib. vi, 344. is /j,. iv, 169. 

1* Ib. I, 200. 15 See Buddhist India, 22. 

18 J.R.A.S. 1898, 546 f. " D. i, 91 ; cf. Dialogues of the Buddha, i, 113, n. 2. 

200 Economic Conditions [ch. 

drew near, on his tour, to their town\ These Mallas were also 
possessed of a mote-hall {santhdgdra) for parliamentary discussions'^ 
— ^a class of buildings illustrated by the bas-relief of a celestial 
House of Lords on the Bharhut Stupal 

Land might, at least in the kingdom of Magadha, be given 
away, and in that of Kosala, be sold. In the former case, a Brahman 
landowner oflfers a thousand karlsas of his estate as a gift*; in 
the latter, a merchant (by a little sharp practice) entangles an 
unwilling noble in the sale of a park". And in the law-books 
we read that land might be let against a certain share of the 
produced The holdings too in the arable land, called the khetta, 
of each village would be subject to redistribution and redivision 
among a family, as one generation succeeded another. It is not 
clear whether any member of a village community could give or 
sell any of the khetta to an outsider. It is just possible that the 
old tradition, expressed in the Brahmanas when a piece of land 
was given as a sacrificial fee — 'And the Earth said : Let no mortal 
give me away !"' — ^may have survived in the villages as a communal, 
anti-alienising feeling concerning any disintegration of the basis of 
their social and economic unity. We should anyway expect, from 
what is revealed in the early Buddhist books, to find such a senti- 
ment upheld, less by the infrequent rural autocrat and his little 
kingdom of country-seat, tenant-farmers, and serfs, than by the 
preponderating groups of cultivators, each forming a gcima. 

When, in the Jataka legend, a king of Videha abandons the 
world as anchorite, he is described as renouncing both his capital, 
the city {nagara) of Mithila, seven yojanas (in circumference), and 
his realm of sixteen thousand gdmas^. It may sound incredible 
that a country owning such a wealth of 'villages' should contain 
but one town, and that so vast in extent, as to suggest inclusion 
not only of parks but of suburban gaimas^. There was not, how- 
ever, any such hard and fast line between gdma and nigama (small 
town) to warrant the exclusion, in this description, of some gdmas 
which may have amounted to nigamas. A similar vagueness holds 
between our 'town' and 'village.' 

A gdma might apparently mean anything from a group of two 
or three houses^" to an indefinite number. It was the generic, 
inclusive term for an inhabited settlement, not possessing the 

1 Vin. I, 247 (Mah. vi, 36). ^ j). n, 147. 

3 Cunningham, Stwpa of Bharhut, pi. xvi. * Jdt. iv, 281. 

» Vin. u, 158 f. {Gull. V. vi, 4, 9 f.). 

6 Apast. 11, 11, 28 (1); i, 6, 18 (20). ^ ^atap. Br. xin, 7, 15. 

8 Jdt. in, 365. " lb. vi, 330. " Childers, Pali Dictionary s.v. 

V 1 1 1] Cities and Villages 201 

fortifications of a nagara, or the ruler's palace of a rdjadhdm. 
The number of inhabitants in the gdmas of the Jataka tales varied 
from 30 to 1000 families. And family (hda), it must be remembered, 
was a more comprehensive unit than it is with us, including not 
only father and mother, children and grandparents, but also the 
wives and children of the sons. Gdma, it is true, might be used 
to dififerentiate a class of settlement, as in the compound gdma- 
nigama, 'villages and towns'; but it is also used in the wider, 
looser sense of group as opposed to single house. For instance, a 
fire, when starting in a house, may extend to the whole gdma\ 
When a hhikkhu leaves park, forest, or mountain to seek alms, he 
'enters the gdma'^,' whether it be a neighbouring village, or the 
suburbs of great Savatthi^. 

Of such cities there were but few in Northern India. Less than 
twenty are named*. Six of them only are reckoned by the Thera 
Ananda as sufficiently important cities (mahd-nagard) to be the 
scene of a Buddha's final passing away: — Savatthi, Champa, 
Rajagaha, Saketa, KosambI, Benares. Kusinara, where that event 
actually took place, he depreciates as not a 'village,' but a jungle 
'townlet' {nagarakay. The greatness of Pataliputra (Patna) was 
yet to come. In the absence of any systematic account of this 
rural organisation in ancient records, it is better to refrain from 
laying down any homogeneous scheme. 'No doubt different vill- 
ages, in diflferent districts, varied one from another in the customs 
of land-tenure, and in the rights of individual householders as 
against the conimunity".' The jungles and rivers of the vast 
Ganges valley fostered independent development probably at 
least as much as the hill-barriers in the Alps have done in the 
case of Swiss and Italian peasant communities down to this day. 

Around the gdma, which appears to have been classed as of 
the country (Janapada)\ of the border {paccantaY, or as suburban, 
lay its khetta, or pastures, and its woodland or uncleared jungle: — 
primeval forest like the Andhavana of Kosala, the Sitavana of 
Magadha, the Paclnavamsa-daya of the Sakiya Territory, retreats 
traditionally haunted by wild beasts and by gentler woodland 
sprites, and where Mara, the Lucifer of seductive evil influences, 
might appear in one shape or another". Different from these were 

* MilindapanlM, 47. 

2 Vin. passim, e.g. Cull. V. v, 12 ; 29. Cf . Thig. ver. 304 ; Camm. p. 175. 

3 Jdt. I, 106 ; Psalms of the Brethren, p. 34, cf . p. 24 ; v. inf. p. 208. 

* Buddhist India, 84 ff. » D. n, 146. 

6 Buddh. Ind. 44 f. ^ jat. i, 318. 8 n, i, 215 ; cf . v. 46. 

^ See Psalms of the Early Buddhists i, u passim; cf. n, p. 161, n. 1. 

202 Economic Conditions [ch. 

such suburban groves as the Bamboo Grove belonging to Magadha's 
king, the Anjanavana of Saketa, the Jetavana of Savatthi. Through 
those other uncleared woodlands and moorlands, where the folk went 
to gather their firewood and litter \ ran caravan routes, roads that 
were at times difficult because of swampy passages after rain, and 
here and there dangerous, less on account of aggressive beasts than 
because of brigands, not to mention demonic bipeds^ 

Adjoining or merged into these wilder tracts were supple- 
mentary grazing pastures * of herds of cattle* and goats", — herds 
belonging to king® or commoners^. Commoners customarily en- 
trusted their flocks to a communal neatherd, as we find in the 
Pennine Alps to-day {le fromageur). We find him either penning 
his herds at night in sheds ^, or, more often, bringing them back 
every evening and counting them out to the several owners, vary- 
ing the pasturage from day to day^ The official name, gopdlaha, 
and the context suggest that dairywork was not usually expected 
of him so much as sagacity in minding his beasts^". 

The arable ground of the gdma lay without the clustered 
dwellings, since these were apparently enclosed by a wall or 
stockade with gates (gdmadvdray^. Fences ^^, snares", and field 
watchmen^* guarded the Tchetta or gdmakhetta from intrusive 
beasts and birds, while the internal boundaries of each house- 
holder's plot were apparently made by channels dug for co-operative 
irrigation^". These dividing ditches, rectangular and curvilinear, 
were likened, at least in the Magadha hhettas, to a patchwork 
robe, and prescribed by the Buddha as a pattern for the uniform 
of his Order : torn pieces of cast-away cloth sewn together, ' a 
thing which could not be coveted'^". The limits of the whole hhetta 
might be extended by fresh clearing of forest land'''. And whereas 
the majority of holdings were probably small, manageable single- 
handed or with sons and perhaps a hired man'*, estates of 1000 
Tcarlsas (acres?) and more occur in the Jatakas, farmed by Brah- 
mans'". In the Suttas, again, the Brahman Kasibharadvaja is 
employing 500 ploughs and hired men (bhatikd)^ to guide plough 
and oxen^\ 

1 Jdt. I, 317 ; V, 103. « lb. i, 99. ^ /j. j, 338. * lb. ni, 149; iv, 826. 

5 lb. m, 401. « lb. I, 240. ^ lb. i, 194, 388; cf. Rigveda, x, 19. 

8 lb. I, 388; m, 149. » A, i, 205; M. Dhp. Comvi. i, 157. ^^ A. v, 350. 

" Jat. I, 239; n, 76, 135; iii, 9; iv, 370 (nigama). i^ jj,. j^ 215. 

13 lb. I, 143, 154. " lb. II, 110; iv, 277. 

« Dhp.yer. 80=U5 = Therag. 19; Jdt. iv, 167; i, 336; v, 412. 

16 Vin. Texts n, 207-9 {Mah. viii, 12). Cf. Pss. of the Brethren, p. 152. 

17 Jdt. II, 357; IV, 167. i^ jj,, i, 277; m, 162; iv, 167. i» lb. iii, 293; iv, 276. 
20 S.N. I, 4 ; cf. S. I, 171 ; Jdt. m, 293. "i j^. n^ 1Q5. 300. 

V 1 1 1] Agriculture 203 

Rice was the staple article of food^; besides which seven other 
kinds of grain are mentioned^ ; sugar-cane^ and fruits, vegetables 
and flowers were also cultivated. 

Instances of coUectivist initiative reveal a relatively advanced 
sense of citizenship in the gdmas. The peasant proprietors had a 
nominal head in the hhojaka or headman, who, as their represent- 
ative at political headquarters and municipal head, was paid by 
certain dues and fines*. But all the village residents met to confer 
with him and each other on civic and political matters. And 
carrying the upshot of their counsels into effect, they built new 
mote-halls and rest-houses, constructed reservoirs and parks, and 
took turns at a voluntary Gorv6e in keeping their roads in repair ^, 
herein again followed by Alpine peasants of to-day. Women too 
considered it a civic honour to bear their own part in municipal 
building''. A further glimpse into the sturdy spirit in gdma-\\i& is 
caught in the Jataka sentiment, that for peasants to leave their 
tillage and work for impoverished kings was a mark of social 
decay ^ Relevant to this is the low social rank assigned to the 
hired labourer, who is apparently classed beneath the domestic 
slave ^. 

Scarcity owing to drought or to floods is not infrequently 
referred to, extending even over a whole kingdom I This contra- 
dicts the 'affirmation' recorded by Megasthenes^**, that 'famine 
has never visited India,' unless his informant meant a very general 
and protracted famine. The times of scarcity in Buddhist records 
apparently refer only to brief periods over restricted areas. 

Nothing in all the foregoing evidence has gone to show that, in 
the India of early Buddhist literature, the pursuit of agriculture 
was associated with either social prestige or social stigma. The 
stricter Brahman tradition, not only in the law-books, but also in 
the Sutta Nipata, the Majjhima Nikaya, and the Jatakas, expressly 
reserved the two callings of agriculture and tr^de for the Vaigya 
or middle class, and judges them unfit for Brahman or noble. 
Thus the Brahman Esukari of Savatthi considers village and dairy 
farming as not less the property and province of the Vaigya than 
are bow and arrow, endowed maintenance (by alms), and sickle and 

1 lb. I, 340; n, 43, 135, 378; ra, 383; iv, 276. 

'^ M. I, 57; also yava (barley) in Jdt. ii, 110. 3 Ih. i, 339; Vin. (Mah. vi, 35, 6). 

* lb. I, 199. « Jdt. I, 199 f. 6 16, 7 j^. i, 339. 

8 Cf. D. I, 51 ; A. I, 145, 206 ; Mil. 147, 331 ; trs. ii, 210, n, 6. 

9 Vin. I, 211, 213 ft. ; Vin. Texts, m, 220, n. 1 ; Jdt. i, 329 ; ii, 135, 149, 367 ; 
Y, 193 ; VI, 487. 

10 M'Crindle, Ancient India as described by Megasthenes, 32. 

204 Economic Conditions [ch. 

yoke, the property and province of noble, Brahman, and working 
classes respectively ^ And here and there, in the Jataka-book, 
Brahmans who engage in agriculture, trade, and other callings are 
declared to have fallen from their Brahmanhood^. On the other 
hand, in both Jatakas and Suttas, not only are Brahmans fre- 
quently found pursuing tillage, cow herding, goat keeping, trade, 
hunting, wood-work or carpentry, weaving, caravan guarding, 
archery, carriage-driving, and snake-charming^, but also no reflec- 
tion is passed ui)on them for so doing, nay, the Bi-ahman farmer 
is at times a notably pious man and a Bodhisat to boot*. Dr Fick 
is disposed to think that the North- western {JJ diced) Brahmans 
of the Kurus and Panchalas, some of whom came east and settled 
there, inherited a stricter standard'. Nevertheless it is not claimed 
for the pious ones just mentioned, living near Benares and in 
Magadha, that they were Udicca immigrants. Even the law-books 
permit Brahmans to engage in worldly callings if they are in 
straitened circumstances, or if they take no active share in the 

As for the Kshatriya clansmen of the republics mentioned 
above, they were largely cultivators of the soil. For instance, in 
the Kundla Jataka, it was the workmen in the fields of the Sakiyan 
and Koliya 'bhqjakas, amaecas, and upardjas' who began to 
quarrel over the prior turn to irrigate^. In the earliest Indian 
literature agricultural and pastoral concepts play a great part. 
But even if this implied that a special dignity attached to agri- 
culture, it does not follow that any such tradition survived, if it 
survived at all, associated with any section of society. There was 
among Indo-Aryans little of the feudal tie between land and lord 
with lordship over the land-tillers, which made broad acres a basis 
for nobility in the West. However they accomplished their pre- 
historic invasion of the Ganges basin, 'land-grabbing' does not 
seem to have been carried out pari passu with success in general- 
ship. This may have been because the annexation of land to any 
wide extent meant clearing of jungle. Except among Dravidian 
and Kolarian towns along the rivers, the task of the invaders was 
more like that of pioneering settlers in America. And there we 

1 M. n, 180. The Vdsettha-sutta (M. no. 98 ; S.N. in, 9) in spiritualising the term 
brahmana, reveals the same exclusive sentiment as current. 

a Jat. IV, 363 f. 

3 Jdt. n, 165 ; in, 293 ; iv, 167, 276 ; in, 401 ; iv, 15 ; v, 22, 471 ; n, 200 ; vi, 
170; IV, 207, 457; v, 127. 

* lb. m, 162. ^ Sociale Gliederung in Indien, 138 f. 

« E.g. Manu x, 116. ^ Jdt. v, 412. 

VII i] Arts and Crafts 205 

know that land is not an appanage involving special privileges and 
entailing special claims, but a commodity like any other. 

The slave or servant {dasa, ddsi) was an adjunct in all house- 
holds able to command domestic service ; but slaves do not appear 
to have been kept, as a rule, in great numbers ^ either in the house, 
or, as in the West, at mining or 'plantation' work. Their treat- 
ment differed of couise according to the disposition and capacity 
of both master and slave. Thus we find, in the Jataka, the slave, 
petted, permitted to learn writing and handicrafts besides his 
ordinary duties as valet and footman, saying to himself that, at the 
slightest fault he might get 'beaten, imprisoned, branded, and fed 
on slave's fare ' ^. But of actual ill-treatment there is scarce any 
mention. Two instances of beating occur, and in both the victims 
were maids. One lies a-bed repeatedly (to test her pious mistress's 
temper)^; the other fails to bring home wages*. Presumably she 
had been sent to fetch her master's wage, or else had been hired 
out. But we do not meet with runaway slaves. Slavery might 
be incurred through capture', commuted death sentence, debt®, 
voluntary self-degradation'', or judicial punishment* ; on the other 
hand, slaves might be manumitted", or might free themselves by 
payment^". They might not, while still undischarged, be admitted 
into the religious community (SanghaYK 

The hireling, wage-earner, day-labourer was no man's chattel, 
yet his life was probably harder sometimes than that of the slave *l 
He was to a great extent employed on the larger land-holdings^*. 
He was paid either in board and lodging, or in money-wages". 
Manu prescribes regular wages both in money and kind for menials 
in the king's service ^^ 

In the arts and crafts, a considerable proficiency and special- 
isation of industry had been reached. A list of callings given in 
the Milindapaiiha, reveals three separate industries in the manu- 
facture of bows and arrows, apart from any ornamental work on 
the same'^ In the same work, the allusion to a professional 
winnower of grain indicates a similar division of labour to our own 
threshing machinists and steamplough-owners who tour in rural 

1 Vin. I, 72 {Mah. i, 39) ; D. i, 60, 72, 92 f., 104 ; Dialogues of the Buddha i, 19, 101. 

2 Jdt. I, 451 f . 3 M. I, 125. * Jdt. i, 402 f. 

5 lb. IV, 220 ; VI, 135. 6 zb. vi, 521 ; Therig. ver. 444. 

' Vin. I, 72 (3Iah. i, 39, 1) ; Sum. Vil. i, 168. 8 j^t. i, 200. 

» D. I, 72 ; Pss. Sisters, p. 117 ; Pss. Brethren, p. 22 ; Jdt. v, 313. 

10 lb. VI, 547. " Vin. i, 76 (Mah. i, 46 f.). ^^ jat, i, 422 ; iii, 444. 

13 lb. in, 406 ; iv, 43 ; S.N., p. 12. " lb. n, 139 ; m, 326, 444 ; v, 212. 

« Manu vn, 125 f. w MU. 331. 

2o6 Economic Conditions [ch. 

districts \ As certain grain crops were reaped twice a year^, this 
would afford a fairly protracted season of work every few months. 

Some trade-names, on the other hand, are as comprehensive as 
our 'smith.' As with us, this word (Icammdra) might be applied 
to a worker in any metal. VaddhaJd, again, apparently covered all 
kinds of woodcraft including shipbuilding, cartmaking^, and archi- 
tecture*, thapati, tacchaJca (lit. planer), and hhamakdra or turner 
being occupied with special modes of woodwork ^ A settlement 
of vaddhakis is able to make both furniture and seagoing ships'. 
Once more the same worker in stone (pdsdnd-kottaka) builds 
houses with the ruined material of a former gdma, and also 
hollows a cavity in a crystal as a cage for a moused 

Important handicrafts like the three above named and their 
branches, the workers in leather, i.e., the leather-dressers, the 
'painters,' and others to the number of eighteen were organised 
into gilds (sent), according to Jataka records; but it is to be 
regretted that only four of the eighteen crafts thus organised are 
specifically mentioned, ' the woodworkers, the smiths, the leather- 
dressers, the painters and the rest, expert in various crafts'*. At 
the head of each gild was a president (pamukhd) or alderman 
(jetthaka), and these leaders might be important ministers in 
attendance upon and in favour with the king. Occasionally these 
functionaries quarrelled, as at Savatthil And it may have been 
such quarrelling also at Benares that led to the institution of a 
supreme headship over all the gilds, an office doubled with that of 
treasurer {hhanddgdrika) being founded at that city. It is of 
interest to note that this innovation in administrative organisation 
was made at a time when, according to the legend, the monarchy 
is represented as having been elective, not hereditary, and when 
the king who appointed, and the man who was appointed, were 
the sons, respectively, of a merchant and a tailor"! The nature 
and extent of the authority of the pamiddia over the gilds is 
nowhere clearly shown. Nor is it clear to what extent the duties 
of a hhanddgdrika, lit. ' houser of goods,' coincided with our word 
'treasurer.' It was not confined to the custody of moneys, for the 

1 Mil. 201 (perhaps a doubtful rendering ; yet there is a professional ploughman in 
Jataka, n, 165). 

2 Megasthenes ; cf. M'Crindle, op. cit. 54 ; v. inf. Chapter xvi, p. 404. 

' Jdt. rv, 207. We find ydnakdras, rathakaras, takatakaras also so engaged. 
■» Jdt. I, 201 ; IV, 323 ; Mil. 330, 345. « M. i, 56, 396 ; m, 144 ; Dhp. ver. 80. 

« Jdt. IV, 159. ' lb. I, 479. » Jdt. i, 267, 314 ; in, 281 ; iv, 411; vi, 22. 

» lb. n, 12, 52 ; cf . mahdvaddhaki in Jdt. vi, 332. 
10 Jdt. IV, 43. 

VI 1 1] Leaders of Industry 207 

Saiigha had officials so named* ; hence it is possible that it referred 
to a supervision of the goods made or dealt with by a gild or gilds 
and not only to the king's exchequer. 

Nor can we with any certainty fill up the fourteen unnamed 
gilds. A great many arts and crafts are mentioned in the books, 
some of them held in less social esteem than others. Among the 
latter were trades connected with the slaying of animals and work 
on their bodies, e.g., hunters and trappers, fishermen, butchers, and 
tanners. Yet other such despised callings Avere those of snake- 
charming, acting, dancing and music, rush weaving and chariot- 
making, the last two because of the despised, probably aboriginal, 
folk whose hereditary trades they were. Other more honourable 
crafts were ivory- working, weaving, confectionery, jewelry and 
work in precious metals, bow and arrow making, pottery, garland- 
making and head-dressing. Besides these handicrafts, there was 
the world of river and sea-going folk, the trader or merchant, and, 
corresponding in a limited way to the first-named, the caravan- 
escorts and guides or 'land-pilots' {thcda-niyydmaka). But although 
reference is made in connexion with some of these, to a jetthaka, 
or Elder, no farther evidence of civic organisation is forthcoming. 

Other instances of trades having jetthdkas are seamen, or at 
least pilots {niyydmakaY, garland makers^, caravan traders and 
guards*, and robbers or brigands. We read, e.g., of a little robber- 
grama in the hills, near Uttara-Panchala, numbering 500 families^. 

The learner or apprentice (antevdsika, literally ' the boarder ') 
appears frequently in Buddhist books, one of which indicates the 
relative positions of pupil and master woodwright^ But no condi- 
tions of pupillage are anywhere stated. 

The title of setthi (best, chief), which is so often met with and, 
without much justification rendered by 'treasurer,' may possibly 
imply headship over some class of industry or trading. It is clear 
that the famous setthi, Anathapindika of Savatthi, the millionaire 
lay-supporter of the Sangha, had «ome authority over his fellow- 
tradere. Five hundred setthis, e.g., attended him in his presenta- 
tion of the Jetavana to the Buddha^. Unless these were convened 
from different towns, the number in any one town was not limited 
to one or a few. They are usu£illy described as wealthy, and as 
engaged in commerce. Dr Fick is probably right in alluding to 
them as representing the mercantile profession at court*. The 

1 Vin. n, 176 (Cull. V. vi, 21, 2). * j^t, rv, 137. » lb. m, 405. 

* lb. I, 368 ; n, 295, 335. « lb. i, 296 f. ; n, 388 ; iv, 430, 433 (Comm.). 

« lb. I, 251 ; V, 290 f. ; Attha-sdlini, p. 111. 
^ Jdt. I, 93. « Op. cit., p. 167 f. 

2o8 Economic Conditions [ch. 

word certainly implied an oflfice {thana^) held during life. There 
might be a chief (mahd) setthi, and an anusetthi or subordinate 
officer 2; a commentary even refers to the insignia of a setthi-chatta 
(umbrella of state) ^ 

The remarkable localisation of industries revealed in Buddhist 
literature has already been noticed. This is observable especially 
in the case of craft- villages of woodwrights*, ironsmiths^, and 
potters*. These were either suburban to large cities, or rural, and 
constituting as such special markets for the whole countryside, as 
we see in the ironsmiths' gama just cited, to which people came 
from the gamas round about to have razors, axes, ploughshares, 
goads, and needles made. On the Ganges or further afield there 
were trapper gamas, supplying game, skins, ivory, etc.^ 

Within the town we meet with a further localisation of trades 
in certain streets, if not quarters, e.g., the street (vUhi) of the 
ivory workers in Benares^, the dyers' street", the weavers' 'place' 
(thdnay, the Vessas' (Vai^yas, merchants?) street". 

Combined with this widespread coi-porate regulation of indus- 
trial life, there was a very general but by no means cast-iron 
custom for the son to follow the calling of the father. Not only 
individuals but families are frequently referred to in terms of their 
traditional calling. The smith, e.g., is Smithson; Sati the fisher- 
man's son is Sati the fisherman; Chunda thie smith is called 
Chunda Smithson, etc.^^ This, however, is not peculiar to Indian or 
even to Aryan societies, up to a certain stage of development. 
Even of our own it was said but half a century ago that the line of 
demarcation between different employments or grades of work 
had till then been 'almost equivalent to an hereditary distinction 
of caste '^l In modern India no doubt these lines of demarcation 
have intensified in the course of centuries, and have split up the 
industrial world into a, to us, bewildering number of sections, or, 
as the Portuguese called them, castes. 

The Jatakas reveal here and there a vigorous etiquette observed 
by the Brahman ' colour ' in the matter of eating with, or of the 
food of, the despised Chandalas, as well as the social intolerance felt 

1 Jat. 1, 122 ; cf. Vin. Texts, i, 102, note 3. 

a Vin. I, 19 {Mah. i, 9) ; Jat. v, 384. ^ vimana-vatthu {Comm.), 66. 

* Jat. 11, 18, 405; iv, 159, 207. 6 lb. in, 281. « lb. ra, 376, (408). 

' lb. VI, 71, nesddagdma ; cf. iii, 49 ; Therig. (Comm.), 220, migaluddakagdma. 
8 Jat. I, 320; ii, 197. » lb. iv, 81. i» lb. i, 356. " lb. vi, 485. 

^^ M. I, 256; D. u, 127 L {'kammdraputto' and 'kammdro') ; Jdt. i, 98, 194, 312; 
II, 79. Cf. nesddo = luddaputto = luddo, Jdt. m, 330 f. ; v, 356-8. 
^* J. S. Mill, Political Economy, xiv, 2. 

VI 1 1] Social Distinctions 209 

for the latter by the burgess class ^. The Jataka commentary tells 
the story of a slave-girl, daughter of a slave and a Khattiya, whose 
father pretended to eat with her only that she might be passed off 
before the Kosalans, seeking a nobly born consort for their king, 
as a thorough -bred Sakiyan^. 

On the other hand, a great many passages from both Jataka 
and other canonical books might be quoted to show that the four 
'colours' are on the whole to be taken in no stricter sense than we 
speak of ' lords and commons,' ' noblesse, ^glise, tiers-^tat,' * upper, 
middle, lower classes.' That Brahmans claimed credit if born of 
Brahmans on both sides for generations back^, betrays the existence 
of many born from a less pure ' connubium.' In the Kusa Jataka, a 
Brahman takes to wife the childless chief wife of a king without 
'losing caste' thereby ^ Elsewhere in the Jataka-book princes, 
Brahmans, Setthis are shown forming friendships, sending their 
sons to the same teacher, and even eating together and inter- 
marrying, without incurring any social stigma or notoriety as 
innovators or militants ^ The following instances may be 
quoted: — 

A king's son, pure bred, cedes his share of the kingdom to his 
sister, turns trader and travels with his caravan^. A prince, whose 
wife in a fit of displeasure has returned to her father, apprentices 
himself at that father's court, without entailing subsequent social 
disgrace, to the court potter, florist, and cook successively, in order 
to gain access to her I Another noble, fleeing from his brother, 
hires himself to a neighbouring monarch as an archer^. A prince 
resigning his kingdom, dwells with a merchant on the frontier, 
working with his hands ^. A commentarial tradition represents a 
child of the Vaccha Brahmans as the * sand-playmate ' of the little 
Siddhattha, afterwards the Buddha ^°. A wealthy, pious Brahman 
takes to trade to be better able to afford his charitable gifts ^^ 
Brahmans engaged personally in trading without such pretext ^^, 
taking service as archers ^^, as the servant of an archer who had 
been a weaver^*, as low-caste trappers ^^, and as low-caste carriage- 

1 Jdt. u, 83 f . ; ni, 233 ; iv, 200, 376, 888, 390-2. 

2 Jat. IV, 144 ff. » D. I, 93 ; M. ii, 156; Thera-gdthd, vv. 889, 1170. 
* Jdt. V, 280. 

6 Jat. II, 319 f.; in, 9-11, 21, 249-54, 340, 405 f., 475, 517; iv, 38; vi, 348; 
421 f. ; Fick, op. cit., vi-xii ; Dialogues i, 96 ff. 

« Jdt. IV, 84 ; Peta-vatthu Comm. Ill f. '' lb. v, 290-3 ; cf. i, 421 f. 

8 Ih. II, 87. ^ Ih. IV, 169. " Psalms of the Brethren, 17 {Vanavaccha). 

" Jdt. IV, 15 f. 12 j5. V, 22, 471. ^^ j?,. m^ 219 ; v, 127 f. 

" lb. I, 356 f . 15 Ih. II, 200 ; vi, 170 ff . »« lb. iv, 207 f . 

C.H.I. I. 14 


2IO Economic Conditions [ch. 

Again, among the middle classes, we find not a few instances 
revealing anything but caste-bound heredity and groove, to wit, 
parents discussing the best profession for their son: — writing, 
reckoning, or money-changing (rupa^, no reference being made 
to the father's trade ^; a (low-class) ideer- trapper becoming the 
prot€g4 and then the ' inseparable fiiend ' of a rich young Setthi, 
without a hint of social barriers^; a weaver looking on his handi- 
craft as a mere make-shift, and changing it ofi-hand for that of an 
archer^; a pious fanner and his son, with equally little ado, 
turning to the low trade of rush weaving* ; a young man of good 
family but penniless, starting on his career by selling a dead mouse 
for cat's meat at a * farthing,' turning his capital and his hands to 
every variety of job, and finally buying up a ship's cargo, with his 
signet-ring pledged as security, and winning both a profit of 200 
per cent and the hand of the Setthi's daughter ^ 

This freedom of initiative and mobility in trade and labour finds 
further exemplification in the enteiprise of a settlement {gdma) of 
woodworkers'^. Failing to carry out the orders for which prepay- 
ment had been made, they were summoned to fulfil their contract. 
But they, instead of ' abiding in their lot,' as General Walker the 
economist said of their descendants, 'with oriental stoicism and 
fatalism^,' made *a mighty ship' secretly, and emigi-ated with their 
families, slipping down the Ganges by night, and so out to sea, till 
they reached a fertile island. Stories, all of these, not history ; 
nevertheless they serve to illustrate the degree to which labour 
and capital were mobile at the time, at least, when these stories were 
incorporated in the Buddhist canon, and before that. And they 
show that social divisions and economic occupations were very far 
from coinciding. There was plenty of pride of birth, which made 
intermarriage and eating together between certain ranks an act 
more or less disgraceful to those reckoning themselves as socially 
higher. And sons, especially perhaps among artisans, tended to 
follow the paternal industry. This was all. 

The trade of the trader, dealer, or middleman (vdnija) may well 
have been largely hereditary^. Traditional good- will handed on 
here would prove specially effective in commanding confidence, 
and thus be a stronger incentive than the force a tergo of caste- 
rule. There is, however, no instance as yet produced from early 
Buddhist documents pointing to any corporate organisation of the 

1 Vin. I, 77 (Mah. i, 49, 1) ; iv, 128 (Pdc. lxiv, 128). * j^t, m, 49 ff . 3 jj,. a, 87. 
4 lb. IV, 318. 6 lb. I, 120 ff. e Jdt. iv, 159. ^ The Wages Question, p. 177. 
8 Jdt. II, 287 ; III, 198. It is noteworthy that mining and miners never came on in 
the Jataka scenes. 

VII i] Partnerships 2ii 

nature of a gild or Hansa leagued The hundred or so of 
merchants who, in the Chullaka-Setthi Jataka^, come to buy up the 
cargo of a newly arrived ship, are apparently each trying to ' score 
off his own bat,' no less than the pushful youth who forestalled 
them. Nor is there any hint of syndicate or federation or other 
agreement existing between the 500 dealers who are fellow 
passengers on board the ill-fated ships in the Valahassa and 
Pandara Jatakas^ ; or the 700 who were lucky enough to secure 
Supparaka as their pilots beyond the fact that there was concerted 
action in chartering one and the same vessel. Among merchants 
travelling by land, however, the rank of satthavaha or caravan- 
leader seems to imply some sort of federation. This position was 
apparently hereditary, and to be a jetthaka or elder, in this 
capacity, on an expedition, apparently implied that other mer- 
chants (vdmja), with their carts and caravan-followers, were 
accompanying the satthavaha, and looking to him for directions 
as to halts, watering, precautions against brigands, and even as to 
routes, fording, etc.^ Subordination, however, was not always 
ensured*^, and the institution does not warrant the inference of any 
fuller syndicalism among traders. 

Partnerships in commerce, either permanent, or on specified 
occasions only, are frequently mentioned : the former, in the 
Kutavdnija'' and Mahdvdnija^ Jatakas, the latter in the Pdydsi 
Suttanta^and theSerivdnija Jataka^". In the Jarudapdna Jataka^^ 
there is, if not explicit statement, room for assuming concerted 
commercial action on a more extensive scale, both in the birth- 
story and also in its introductory episode. The caravan in question, 
consisting of an indefinite number of traders (in the birth-story, 
under a jetthaka), accumulate and export goods at the same time, 
and apparently share the treasure trove, or the profits therefrom. 
In the episode the firm also wait upon the Buddha with gifts 
before and after their journey. These were traders of Savatthi, of 
the class who are elsewhere described as acting so unanimously 
under Anathapindika, himself a great travelling merchant. The 
Guttila Jataka^^, again, shows concerted action, in work and play, 
on the part of Benares traders. It is conceivable, however, that 

^ The compound vaniggrdma is rendered 'merchants' guild' in Macdonell's Sanskrit 

a Jdt. I, 122. 3 j5. 11^ 128 ; v, 75. * 16. iv, 138 fl. ; cf. also vi, 34. 

5 Fick, p. 178 ; D. u, 342 f. ; cf. Jat. i, 98. « jj, i^ io8, 368 ; n, 295 ; m, 200. 
' Jdt. I, 404 also ii, 181. « lb. iv, 350. 9 D. ii, 342. " Jat. i, 111. 

" lb. n, 294 ff. 
" lb. n, 248 ; cf . i, 121 for concerted action between dealers in freights. 


212 Economic Conditions [ch. 

the travelling in company may have been undertaken as much for 
mutual convenience in the chartering of a conunon ship, or the 
employment of a single band of forest-guards, as for the prevention 
of mutual under-selling or the cornering of any wares \ Merchants 
are represented, at least as often, as travelling with their own 
caravan alone. Thus in the first Jataka" two traders, about to 
convey commodities to some distant city, agree which shall start 
first The one thinks that, if he arrive first, he will get a better, 
because non-competitive price ; the other, also holding that com- 
petition is killing work (lit. 'price-fixing is like robbing men of 
life'), prefers to sell at the price fixed, under circumstances 
favourable to the dealer, by his predecessor, and yields him a start. 

The little aper^s which we obtain from the Jatakas of the 
range and objective of such merchants' voyages are so interesting 
as side-lights on early trafficking as to create regret at their 
scantiness. The overland caravans are sometimes represented as 
going 'east and west'^, and across deserts that took days, or rather 
nights to cross, a 'land-pilot' {thala-niyydmaka) steering during the 
cooler hours of darkness by the stars ^ Drought, famine, wild 
beasts, robbers, and demons are enumerated as the dangers severally 
besetting this or that desert routed Such caravans may have been 
bound from Benares, the chief industrial and commercial centre in 
early Buddhist days, across the deserts of Rajpiitana westward to 
the seaports of Bharukaccha, the modern Broach *, and the sea 
board of Sovira (the Sophir, or Ophir, of the Septuagint ?), and its 
capital Roruva^ or Roruka^ Westward of these ports there was 
traffic with Babylon, or Baveru. 

At a later date, say, at the beginning of the first century A.D. 
the chief objective of Indian sea-going trade is given in the 
Milinda^ as follows : — 

As a shipowner who has become wealthy by constantly levying freight in 
some seaport town, will be able to traverse the high seas, and go to Vanga or 
Takkola, or China, or Sovira, or Surat, or Alexandria, or the Koromandel coast, 
or Further India, or any other place where ships do congregate. 

Tamil poems testify to the flourishing state of Kaviri-pattinam 
(Kamara in Periplus, Khabari of Ptolemy), capital of Chola, on 
the Kaveri river, at about the same period, as a centre of inter 
national trade, especially frequented by Yavana (Yona, Ionian) 

1 On a local 'comer in hay' see Jdt. i, 121. 

2 lb. I, 99; cf. 194, 270, 354, 368, 413; ii, 109, 335; in, 200, 403; iv, 15 f.; 
v, 22, 164. 

» lb. I, 98 f . * lb. I, 107. 6 lb. i, 99. « lb. in, 188 ; iv, 187 ; Dip, ix, 26. 

7 lb. m, 470. 8 D. 11, 235 ; Divy. 544. 

9 Milindapanha 359 ; trans, n, 269 (S.B.E. xxxvi). 

VI 1 1] Trade Routes 213 

merchants^. According to the Jataka it was practicable to attain 
to any of these ports starting fi*om up the Ganges, not only 
from Champa (or Bhagalpur, about 350 miles from the sea) but 
even from Benares. Thus the defaulting woodwrights mentioned 
above- reach an ocean island from the latter city ; Prince Mahaja- 
naka sets out for Suvannabhumi from Champa^, and Mahinda 
travels by water from Patna to Tamalitti, and on to Ceylon*. It is 
true that the word samudda, sea, is occasionally applied to the 
Ganges^, nevertheless, if the foregoing stories be compared with the 
Saiikha Jataka ^ it becomes probable that the open sea is meant 
in both. In this the hero, while shipwrecked, washes out his mouth 
with the salt water of the waves during his self-imposed fast 
Again, in the Sllanisamsa Jataka, a sea-fairy as helmsman brings 
'passengers for India' by ships 'from off the sea to Benares by 
river '^. Other traders are found coasting round India from 
Bharukaccha to Suvannabhumi^, doubtless putting in at a Ceylon 
port ; for Ceylon was another bourae of oversea conunerce, and one 
associated with perils around which Odyssean legends had grown 
up®. The vessels, according to Jataka tales, seem to have been 
constructed on a fairly large scale, for we read of 'hundreds' 
embarking on them, merchants or emigrants. The numbers have 
of course no statistical value ; but the current conceptions of 
shipping capacity are at least interesting. 

The nature of the exports and imports is seldom specified. The 
gold which was exported to Persia as early at least as the time of 
Darius Hystaspes, finds no explicit mention in the Jatakas. Gems 
of various kinds are named as the quest of special sea-farers 
anxious to discover a fortune^". 'Silks, muslins, the finer sorts 
of cloth, cutlery and armour, brocades, embroideries and rugs, 
perfumes and drugs, ivory and ivory- work, jewelry and gold 
(seldom silver) : — these were the main articles in which the 
merchant dealt'". 

As to the inland routes, the Jatakas tell of Anathapindika's 
caravans travelling S.E. from Savatthi to Rajagaha and back 

^ Kanakasabhai, The Tamils 1800 years ago, quoted bySubba Eao {v. Bibliography), 
p. 81 f. On Chola see Mahdvamsa, xxi, 13. 

2 Jdt. IV, 159. 3 J6. Yi, 34 f. ■• Vin. m, 338 [Samantapasadika). 

« Jut. I, 227 ff. ; IV, 167 f . ; vi, 158, but cf. M. i, 493 ; S. ii, 32, where sdgara is 

• Jdt. IV, 15-17. 7 lb. n, 112. « jj, ni, 188. 

* Ih. XI, 127 ff. 'The name Lanka does not occur. Tambaparmi-dipa... probably 
meant for Ceylon.' Buddhist India, 105. lo Jdt. rv, 21, 139-41. 

" Rh. D. Buddhist India, p. 98; Fick, op. cit., 174. 

214 Economic Conditions [ch. 

(about 300 miles)*, and also to the 'borders,' probably towards 
Gandhara^. The route in the former journey was apparently 
planned to secure easy fording of the rivers by following * the foot 
of the mountains to a point north of Vesali, and only then turning 
south to the Ganges '^ 

Another route south-west fi'om Savatthi to Patitthana^, with six 
chief halting places, is given in the Sutta Nipata, verses 1011-13*. 
From east to west, traffic, as we have seen, was largely by river, 
boats going up the Ganges to Sahajati®, and up the Jumna to 
Kosambi^ Further westward the journey would again be over- 
land to Sind, whence came large imports in horses and asses*, and 
to Sovira* and its ports. Northward lay the great trade route 
connecting India with Central and Western Asia, by way of Taxila 
in Gandhara (Pali Tahhasild), near Rawalpindi", and presumably 
also of Sagala in the Punjab. This great road and its southern 
connexions with the leading cities of the Ganges valley" must have 
been, even in early Buddhistic days, relatively immune from 
dangers. Instances abound in the Jatakas of the sons of nobles 
and Brahmans faring*^, unattended and unarmed*^, to Takkasila to be 
educated at this famous seat of Brahmanical and other learning". 

There were no bridges over the rivers of India. The setu or 
causeway of Buddhist metaphor^^ is a raised dyke built over shoal 
water ^^ Only fording-places and ferries for crossing rivers are 
mentioned in Buddhist literature*^, and cart-ferries in Manu*^ 

Food-stuffs for the towns were apparently brought only to the 
gates, while workshop and bazaar occupied, to a large extent at 
least, their own special streets within*®. Thus there was a fish- 

1 Jdt. I, 92, 348. 2 j6. i^ 377 f . 

3 Buddh. Ind. 103. The road followed by the Buddha on his last ministering tour 
is from Rajagaha to Kusinara, crossing the Ganges at Patna, with halts at twelve inter- 
mediate towns (gdmas or nagaras), including Vesali. The remainder of this circuitous 
route to Savatthi lay W.N.W. D. ii, Suttanta xvi, 81 fif. 

* Paithan. See map and p. 30, Buddh. Ind. 

6 Cf . the list in Spence Hardy, Manxml of Buddhism, 334. 

« Vin. Texts, in, 401. ' jj,. p. 382. 

8 Jdt. I, 124, 178, 181 ; ii, 31, 287 ; cf. Hopkins, J.A.O.S. xiii, 257, 372 ; Fick, op. 
cit. 176. 

9 Vimdna-vatthu {Comm.) 336. 

10 J. H. Marshall, Archaeological Discoveries at Taxila (1913) ; Guide to Taxila (1918). 
" Of these the route to Rajagaha lay past Saketa. Vin. Texts n, 176 (Mah. 

vm, 1, 8). 

" Jdt. I, 259 ; n, 85, 282, 411 ; ra, 122 ; v, 457, etc., etc. " jj^ ^^ 277. 

1* Biihler, Indian Studies, No. 3. Fick, op. cit., 62; Vin. Texts n, 174 f. {Mah. 
vni, 1, 6ff.). 

» E.g. Thera-gdthd, ver. 7, 615, 762; M. i, 134; A. i, 220; n, 145; Dh. S.,% 299. 

i« Vin. Texts n, 104 (Mah. vi, 28, 12 f.) = D. ii, 89. " Jdt. m, 228. 

18 vin, 404 ff. {S.B.E. xxv.) " V. sup., p. 208 ; Buddh. Ind. 76. 

VI 1 1] Bazaars 215 

monger's village at a gate of Savatthi\ greengrocery is sold at the 
four gates of Uttara-Paiichala^, and venison at the cross-roads 
(siiighdtaka) outside Benares ^ 

The slaughter-houses (silnd) mentioned in the Vinaya* were 
presumably outside also, and near them the poor man and the 
king's chef bought their meat^, unless by sihghdtaka we under- 
stand street-corners as the places where meat was sold^. The great 
city of Mithila was, according to the Mahd-ummagga Jataka, 
composed in part of four suburbs extending beyond each of its four 
gates, and called not gdmas, but nigamas. These were named 
respectively East, South, West, and North Yavamajjhako, translated 
by Co well and Rouse 'market-town 'I The workshop in the street 
was open to view, so that the hhilikhu coming in to town or village 
for alms, could see ^fletcher and carriage-builder at work, no less 
than he could watch the peasant in the fields Arrows and 
carriages and other articles for sale were displayed in the dpana^^ 
or fixed shop, or, it might be, stored within the antardpana^'^. 
In these, or in the portable stock-in-trade of the hawker", retail 
trading constituted a means of livelihood, independently, it might 
be, of productive industry. The application, judgment, cleverness, 
and ' connexion ' of the successful shopkeeper ^^ are discussed in the 
Nikayas^^, and among trades five are ethically proscribed^* for lay 
believers : — daggers, slaves, flesh, strong drink, poisons. 

Textile fabrics ^^, groceries and oiP^, gi'eengi-oceries^^, grain ^^, 
perfumes and flowers", articles of gold and jewelry^, are among the 
items sold in the bazaars of Jataka stories and Vinaya allusions, 
and for the sale of strong liquors there were the taverns {pdndgdra^ 
dpdnay\ But there is no such clear reference made either to a 
market-place in the town, or to seasonal market-days or fairs. 

1 Psalms of the Brethren, 166 ; cf . Jdt. i, 361 : ' they went for alms to a village just 
outside the gates of Benares, where they had plenty to eat.' 

2 Jdt. XV, 445. 3 J?;, ni, 49; cf. M. i, 58 ; iii, 91. 

* Mah. VI, 10 ; Cull. V. x, 10, 4. « Jat. v, 458 ; vi, 62. 

" But cf. Psalms of the Brethren, 254 : ' out of the four gates to the cross roads.' 

7 Jdt. VI, 330 (trans, p. 157) ; Cunningham, Stupa of Bhdrhut, 53. On these 
bas-reliefs the Jataka is called Yava-majjhakiya. 

8 Psalms of the Brethren, 24. 

9 Jdt. n, 267; iv, 488; vi, 29; Vin. iv, 248 ; cf. Gull. V. x, 10, 4. Cf. Apana as 
the name of a nigama, M. i, 359, 447; S.N., Sela-Sutta (called a Brahman gdma, 
Pss. of the Brethren, 310). 

10 Jdt. I, 55, 350; m, 406. " Ih. i. 111 f., 205; n, 424 ; ni, 21, 282 f. 
^"^ Apanika pdpanika. " j. i^ 115 f. '^ A. m, 208. 

1* Vin. IV, 250 f. ^« lb. iv, 248-9. " Jdt. i, 411, " 75, ^^ 267. 

19 lb. I, 290 f. ; IV, 82 ; vi, 336; Vin. Texts, in, 343. «> Jdt. iv, 223. 

" lb. I, 251 f . ; 268 f . ; vi, 328 ; cf . Dhp. Crnnm. in, 66. 

2i6 Economic Conditions [ch. 

Such an institution as the hath, or barter fair, taking place on the 
borders of adjacent districts, finds, curiously enough, no mention in 
the Jataka-book, though as the late Wm. Irvine wrote, *it is to this 
day universal to my personal knowledge, from Patna to Delhi, and, 
I believe, from Calcutta to Peshawar.' The f§tes often alluded to^ 
do not appear to have included any kind of market^. 

The act of exchange between producer and consumer, or 
between either and a middleman, was both before and during the 
age when the Jataka-book was compiled, a 'free' bargain, a 
transaction unregulated, with one notable exception, by any 
system of statute-fixed prices. Supply was hampered by slow 
transport, by individualistic production, and by primitive machinery. 
But it was left free for the producer and dealer to prevail by 
competition^, and also by adulteration ^ and to bring about an 
equation with a demand which was largely compact of customary 
usage and relatively unafiected by the swifter fluctuations termed 

Instances of price-haggling are not rare^, and we have already 
noticed the dealer's sense of the wear and tear of it^, and a case of 
that more developed competition which we know as 'dealing in 
futures"'. The outlay in this case, for a carriage, a pavilion at the 
Benares docks, men {purisd\ and ushers (pdtihdra), must have 
cut deep into his last profit of 1,000 coins, but he was 20,000 per 
cent, to the good as the result of it ! After this the profit of 900 
and 400 per cent, reaped by other traders® falls a little flat, and 
such economic thrills only revive when we consider the well-known 
story of the fancy price obtained by Prince Jeta for his grove near 
SavatthI from the pious merchant Anathapindika, limited only by 
the number of coins (metal uncertain) required to cover the soil®. 

At the same time custom may very well have settled price to a 
great extent. 'My wife is sometimes as meek as a 100-piece slave- 
girl'^" reveals a customary price. For the royal household, at least, 
prices were fixed without appeal by the court valuer {agghahdraka), 
who stood between the two fires of oflending the king if he valued 
the goods submitted at their full cost, or price as demanded, and of 
driving away tradesmen if he refused bribes and cheapened the 

1 Jdt. I, 423 ; in, 446 ; Dialogues i, 7, n 4. 

2 « Market ' and ' market-place ' are frequently used by translators, but rather 
inferentially than as literal renderings. 

3 Cf . Jdt. Ill, 282 f . * Cf. lb. I, 220. 

6 lb. I, lllf., 195; II, 222, 289, 424 f. « lb. i, 99. 

' lb. I, 121 f . 

« Ib.i, 109 ; cf . IV, 2. » Vin. n, 158 f. (Cull. V. vi, 4, 9) ; Jat. i, 92. 

" Jdt. I, 299. 

VI 1 1] Means of Exchange 217 

wares ^ On the other hand the king might disgust him by too 
niggardly a bonus ^. It may also have been the duty of this official 
to assess the duty of one twentieth on each consignment of native 
merchandise imported into a city, and of one-tenth, plus a sample, 
on each foreign import, as stated in the law-books of Manu^, 
Gautama*, and Baudhayana^. Such octrois are alluded to in one 
Jataka, where the king remits to a subject the duty collected at the 
gates of his capitaP. Finally, it may have been his to assess 
merchants for their specific commutation of the rajakariya, 
namely, one article sold per month to the king at a discount 

The ' sample' mentioned above is suggestive of a surviving pay- 
ment made in kind. That the ancient systems of barter and of 
reckoning values by cows or by rice-measures had for the most 
part been replaced by the use of a metal currency, carrying well 
understood and generally accepted exchange value, is attested 
by the earliest Buddhist literature. Barter emerges in certain 
contingencies*, as e.g. when a wanderer obtains a meal from a 
woodlander for a gold pin', or when among humble folk a dog is 
bought for a hahdpana {hdrshapana) plus a cloak ^''. Barter was 
also permitted in special commodities by the law-books ascribed 
to Gautama" and Vasishtha^'^, and was prescribed in certain cases 
for the Saiigha^^, to whom the use of money was forbidden". 
Moreover, as a standard of value, it is possible that rice was still 
used when the Jataka-book was compiled ^^ 

But for the ordinary mechanism of exchange we find, in that 
and all early Buddhist literature, the worth of every marketable 
commodity, from that of a dead mouse and a day at the festival up 
to all kinds of prices, fees, pensions, fines, loans, stored treasure, and 
income, stated in figures of a certain coin, or its fractions ^^ This 
is either stated, or implied to be, the kahdpana. Of the coins 
called purmms this literature knows nothing. Other current 
instruments of exchange are the ancient nikhha (nishka — a gold 
coin, originally a gold ornament) ^^, the suvanna, also of gold, and. 

1 Jdt. I, 124 f. ; II, 31 ; Pss. of Brethren, 25, 212. 2 lb. iv, 138. 

» VIII, 398-400 ; cf. Jat. iv, 132. * x, 26. 6 i. x. 18, w. 14, 15, 

« Jdt. VI, 347. 7 Gaut. x, 35. 

8 There seems to be nothing in the text of Jataka i, 251 (Vdruni Jdt.) to justify the 
translator's inference that barter was normal ; see J.R.A.S., 1901, p. 876. 
» Jdt. VI, 519. i» lb. II, 247. " vii, 16 f. 

" 11, 37 f. 13 Yin, n, 174 (Gull. V. vi, 19). 

" Vin. Ill, 237 ; 11, 294 ff. [Cull. V. xii, 1 ff.). is Jdt. 1, 124 f. 

" For details of prices see Mrs Bhys Davids, J.R.A.S., 1901, pp. 882 f. 
" Zimmer, Altindisches Leben, 259. 

21 8 Economic Conditions [ch. 

such bronze or copper tokens as the katnsa, the pdda, the 
mdsaha {md8ha\ and the hdhanikd. Cowry shells (sippikdni) are 
once mentioned^, but only as we should speak of doits or mites, 
not as anything still having currency. 

That there was instability as to the relative value of standard 
or token coins in place and time we learn from the Vinaya : * At 
that time [of Bimbisara or Ajatasattu], at Rajagaha, five masakas 
were equal to one pada'^. Again, the nilchha was valued now at 
five^, now at four suvannas^ 

Of substitutes for money, such as instruments of credit, we read 
of signet rings used as deposit or security^, of wife or children 
pledged or sold for debt^, and of I Oil's or debt -sheets {ina- 
panimniy. The bankrupt who, in the Jataka tale, invites his 
creditors to bring their debt-sheets for settlement, only to drown 
himself before their eyes*, appears in a Milinda simile anticipating 
the crisis by making a public statement of his liabilities and assets'^. 
The entanglement and anxieties of debt as well as the corporate 
liability belonging to communistic life in a religious order rendered 
it necessary to debar any candidate from admission to the Sangha 
who was a debtor^". And the sight of a deposited security recalling 
the past circumstances of the pledging is instanced in the Milinda 
as a case of the psychical process of recollection (satiy\ 

No definite rates of interest on money loans appear in the early 
books. But the term which appears in the law-books as 'usury' 
(vrddhi, Pali vaddhi) is found. Meaning literally profit or increase, 
it may very early have acquired the more specialised import. There 
is a tolerant tone concerning the money-lender in a Jataka tale, 
where a patron, in enabling a huntsman to better himself, names 
money-lending (ina-ddna), together with tillage, trade, and har- 
vesting as four honest callings ^^. Gautama is equally tolerant^^ 
But the general tendency of this profession to evade any legal or 
customary rate of interest and become the type of profit-mongering 
finds condemnation in other law-books^*. Hypocritical ascetics are 
accused of practising it^\ No one but the money-lender seems to 
have lent capital wealth for interest as an investment. For in- 
stance, only bonds (pannd) are spoken of in the case of the generous 

1 Jdt. I, 425 f. 2 Yin, m, 45. 3 Childers, Pali Dictionary, s.v. nikkho. 

* Manu, vin, 137. For a more detailed discussion see J.B.A.S., 1901, p. 877 ff. 

5 Jdt. I, 121. 6 lb. VI, 521 ; Therig. 444. 

7 jat. I, 230 ; cf . 227, panne dropetvd. » jj. jy^ 256. 

» Mil. 131 (text) ; cf. 279." '<> Vin. i, 76 {Mah. i, 46) ; cf. D. i, 71 f. " p. 80. 

" Jdt. TV, 422. " X, 6 ; xi, 21. 

" Vas. n, 41, 42 ; Baudh. i, 5, lO^s-S; Manu, m, 153, 180; vin, 152, 153. 

" Jat. IV, 184. 

VI 1 1] General Conclusions 219 

Anathapindika's 'bad debts '\ Capital wealth was hoarded, either 
in the house — in large mansions over the entrance passage {dvdra- 
hotthakay — under the ground^, in brazen jars under the river bank*, 
or deposited with a friend ^ The nature and amount of the 
wealth thus hoarded was registered on gold or copper plates ^ 

Fragmentary as are the collected scraps of evidence on which 
the foregoing outlines of social economy have been constructed, 
more might yet be inferred did space permit. It should, however, be 
fairly clear from what has been said, that if, during, say, the seventh 
to the fourth century b.c. it had been the vogue, in India, to write 
treatises on economic institutions, there might have come down to 
us the record both of conventions and of theories as orderly and as 
relatively acceptable to the peoples as anything of the kind in, say, 
the latter middle ages was to the peoples of Western Europe. But 
it is a curious fact that often where the historian finds little 
material to hand wherewith to rebuild, he judges that there 
never were any buildings. Thus in a leading historical work on 
economics, revised and enlarged in 1890, the whole subject of the 
economic ideas of the 'Orient' is dismissed in a single page as 
being reducible to a few ethical precepts, and as extolling agriculture 
and decrying arts and commerce ; further, that division of labour, 
though politically free, stiffened into a system of hereditary caste, 
arresting economic progress, and that the Chinese alone, and only 
from the seventh century A.D., had any insight into the nature of 
money and its fiduciary substitutes^. But we have been looking 
behind the ethical precepts of the preacher, and the sectarian 
scruples of a class, at the life of the peoples of North India, as it 
survives in the records of their folk-lore, and of the discipline of 
the brethren in orders who lived in close touch with all classes. 
And we have seen agiiculture diligently and amicably carried on 
by practically the whole people as a toilsome but most natural and 
necessary pursuit. We have seen crafts and commerce flourishing, 
highly organised corporately and locally, under conditions of 
individual and corporate competition, the leading men thereof the 
friends and counsellors of kings. We have found 'labour' largely 
hereditary, yet, therewithal, a mobility and initiative anything but 
rigid revealed in the exercise of it. And we have discovered a 
thorough familiarity with money and credit ages before the 'seventh 
century A.D.' 

1 Jdt. I, 227. 2 ji,. i^ 351 . n, 431. 

3 7b. I, 225, 375 f., 424 ; n, 808 ; in, 24, 116. * Ih. i, 227, 323. 

« Jb. VI, 521. 6 Ih. IV, 7, 488 ; vi, 29 ; cf . iv, 237. 

'^ L. Cossa, Introdtwtion to Political Economy. 



The later Brahman literature which, whatever may be the age 
of its representative works in their present form, undoubtedly had 
its roots in a period at least as early as the rise of Jainism and 
Buddhism, may be classified under the four headings — Sutras, Epic 
poems, Law-books, and Puranas. These belong to two distinct 
species of literary composition, the Sutras being broadly dis- 
tinguished from the others both in form and object 

The purpose of the Sutras, so called from the word sutra which 
means *a thread,' is to afford a clue through the mazes of Brah- 
manical learning contained in the Brahmanas. In the form of a 
series of short sentences they codify and systematise the various 
branches of knowledge sacred and secular. They are intended to 
satisfy the needs of a system of oral instruction, so that each step 
in the exposition of a subject may be learnt progressively and a 
convenient analysis of the whole committed to memory by the 
student. The earliest Sutras are in the priestly language and 
represent a phase which is transitional between the language of 
the Brahmanas and Classical Sanskrit as fixed by the grammarians. 

The Epics supply the model both for language and form which 
is followed by the LaAV-books and the Puranas. Their source is to 
be traced to the traditional recitations of bards who were neither 
priests nor scholars. Their language is thus naturally more i)opu- 
lar in character and less regular than Classical Sanskrit \ In many 
respects it does not conform to the laws laid down by the gram- 
marians, and is ignored by them. This became the conventional 
language of epic poetry, which was used also in the Law-books, 
the subject-matter of which was taken to a great extent from the 
Sutras, and in the Puranas, which, as they stand at present, belong 
to a period not earlier than the fourth century A.D. The metres of 
the Law-books and the Puranas are also substantially those of the 
Epic poems. 

1 Wackernagel, Altind. Oram. vol. i, p. xlv. 

CH. ix] Brahman and Buddhist Society 221 

The period of the Sutras, Epics, and Law-books thus overlaps that 
of Buddhist India on the one hand, and reaches well into the period 
of the extant Puranas on the other. The earliest known Purana 
precedes the later law-books probably by centuries, as the Sutras 
precede the earliest works of Buddhism. Nevertheless it is not 
only new matter which is offered by the literature, whether legal 
or epic, but virtually a new phase, a fresh point of view, the life of 
India as it shows itself under the dominion of the Brahmans, who 
have been the real masters of Indian thought for more than three 
millenniums. It is in fact the continuation under new conditions 
of the history depicted above, before Jain and Buddhist had arisen. 

As [we read the works of these important sects we receive the 
impression that the world of India was one in which the ancient 
priestly caste had lost its authority ; that nobles and wealthy 
merchants were more regarded than Brahmans. But it must be 
remembered that, despite the wide reach of Buddhism when in its 
full power, it influenced at first only that part of the country 
where it arose, and that the earlier writings depicting the life and 
teaching of Buddha represent chiefly the circumstances found in a 
very circumscribed area, in fact just the area where Brahmanism 
was weakest. The elements of social life were the same here as 
elsewhere, but they were not arranged in the same way. The 
stronghold of Brahmanism lay to the West, and there the priest 
had had his say and built up his power among clans boasting 
direct descent from Vedic heroes and more inclined to bow to the 
mysterious Vedic word of which the only custodian was the Brah- 
man priest In short, as Brahmanism exaggerates the power of the 
priest, so Buddhism belittles it unduly, not because it sets out to 
do so but because each represents fa special point of view based 
more or less upon geographical position. Owing, however, to a still 
later interpretation of caste, our modern ideas on the subject are 
apt to be peculiarly confused. To understand the social order into 
which we enter as we begin the study of the Sutras, epics, and law- 
books, we must renounce altogether the notion of caste in its strict 
modern sense, as on the other hand we must free ourselves from 
the thought that the whole caste-system is merely a priestly 
hypothesis disproved by the conditions revealed in Buddhistic 

In point of fact, even the Buddhist writings recognise the 
formal castes ; and it is simply impossible that a social structure 
so widely pervading as that of the so-called castes, a structure 
revealed not by didactic works alone but implicitly as well as 

22 2 The Period of Sutras^ Epics y etc. [ch. 

explicitly presented to us in every body of writings whether ortho- 
dox or heterodox, should have been made out of whole cloth. 
What we loosely call by this name to-day are later refinements ; 
and we do not need to turn to Buddhist works to show that in 
ancient times the castes were merely orders socially distinct but 
not very strictly separated or ramified into such sub-divisional 
castes as obtain at the present time. 

Yet before giving the proof of this in detail, it will be well to 
consider briefly the chronology of the works to be reviewed in 
relation to the general character and history of the states in which 
they arose. The legal literature which begins with the Sutras and 
is represented in the epics does not really end at all, as works of 
this nature continue to be written down to modern times, chiefly 
by eminent jurists who comment on older works. But, after elimi- 
nating the modern jurists and confining ourselves to the law-books 
which may be called classic, we still find that the terminus falls 
well into the middle of the first millennium of our era ; and as the 
beginning of this literature in Sutra style reaches back at least as 
far as this before the beginning of our era, the whole period is 
rather more than a thousand years, about the middle of which 
must be set the time to which the epic poems are to be assigned 
as works already known and perhaps nearly completed. 

The cycle thus designated as a millennium is one of very varied 
political fortunes ; and the social, political, and religious material 
of the legal and epical literature must necessarily be explained in 
accordance with the outward changes. What these changes were 
is described in detail in other chapters of this work. For our 
present purpose it is necessary only to recount them in outline. 
At the end of the sixth century B.C., early in the period to which 
the Sutras belong, the Persian Empire held two provinces in N.W. 
India — Gandhara, the present districts of Peshawar and Rawal- 
pindi, and the * Indian ' province, that is to say, the country of the 
Lower Indus : and the northern part of India generally was domi- 
nated by peoples of the Aryan race who had descended from the 
Punjab and spread eastward for centuries, but not so that the 
recently acquired territory was thoroughly assimilated to the cults 
and culture of the invaders, nor so that any one of these invaders 
had established an empire. Long before the end of this same 
period, Buddha, Mahavira, and other reformers had broken with 
the cult derived from the Vedic age, and the great empire of A9oka 
had made a new epoch in political life. This alteration, however, 
had been introduced, though adventitiously, through outer rather 

ix] Outlines of Chronology 2 2^ 

than inner conditions. After the short campaign in the Punjab, 
made by Alexander as the conqueror of the Persian Empire, his 
Indian dominions were, within a few years, absorbed by the grow- 
ing power of Magadha (S. Bihar) then under the sway of a usurper, 
Chandragupta (c. 321-297 B.C.) the low-born son of Mura and the 
founder of the Maurya empire. This empire extended from Pata- 
liputra (Patna) to Herat and was maintained by an army of 
approximately 700,000 men, the first real empire in India. His 
successors, Bindusara and A^oka, enlarged the empire, annexing 
Kalinga on the eastern coast and ruling as far south as Madras. 
This dynasty continued in power till the end of the Sutra period ; 
and under it, during the reign of A9oka (c. 274-236 B.C.) Buddhism 
became the court-religion. A9oka's period is determined by the 
mention in his edicts of certain Hellenic princes who were his 
contemporaries, but after his reign there comes a period of less 
chronological certainty. The different versions of the Puranas are 
not in agreement as to the exact number of his successors; but 
they are unanimous in asserting that the Maurya dynasty lasted for 
137 years ; that is to say, it is supposed to have come to an end 
c. 184 RC. For over a century after its fall the ^^nga dynasty, 
whose founder, Pushyamitra, had slain Brihadratha Maurya and 
usurped his throne, held sway, despite forcible inroads of the 
Yavanas (Greeks) and the Andhras; and we learn that both 
Pushyamitra and the Andlira king, ^atakarni, performed the famous 
'horse-sacrifice,' in accordance with the ancient Vedic rite, thus 
challenging all opponents of their authority. The son of this 
Pushyamitra was Agnimitra, who conquered Vidarbha (Berar), 
then a province of the Andhra Empire of S. India, and the 
grandson, who guarded the horse, was Vasumitra. These names, 
as also the re-establishment of the 'horse-sacrifice,' are highly 
significant in that they show a renascence of the Vedic religion 
and a consequent decline in Buddhism. The same thing is indi- 
cated by the fact that Kharavela, a king of Kalinga, who boasts of 
having invaded the Andhra dominions as well as Northern India, 
was a Jain. Sumitra, the son of Agnimitra, was, according to 
Bana's historical romance, the Harshacharita, miserably slain 
by Mitradeva, who may perhaps have been a Brahman of the 
Kanva family which eventually gained the chief power in the 
state. The account given by the Puranas states that the minister 
Vasudeva slew the tenth and last of the ^unga kings and inaugu- 
rated a new dynasty, called the Kanva dynasty, which lasted for 
about half a century ; but, since the Kanvas are definitely styled 

224 ^^^ Period of Sutras^ Epics ^ etc, [ch. 

'servants of the (^ungas' and for other reasons, it seems more 
probable that the later Qunga kings had been reduced to sub- 
jection by their Brahman ministers, and that the lists of these 
contemporary rulers nominal and actual were wrongly regarded 
by some late editor of the Puranas as successive. It is further 
related that one of the Andhra kings ^ slew Su9arman, the last of 
the Kanvas, and thus brought Magadha under the sway of the 
sovereigns, whose names and titles, as well as their sacrificial 
inscriptions, show them to have been followers of the ancient 
Vedic religion. But here again it appears that dynastic lists 
have been brought together and arranged in an unreal sequence. 
There can be little doubt that the first of the Andhra kings was 
earlier in date than the first of the ^uiigas, and not 157 years 
later as would appear from the Puranas. It is indeed doubtful if 
the Andhras ever ruled in Magadha : but their sway in Central and 
Southern India lasted until the middle of the third century a.d.^ 

In the meantime, on the decline of the Maurya empire which 
must have set in soon after the death of the Emperor A9oka 
(c. 236 RC), the Punjab passed into the hands of foreign invaders — 
first, Greeks from the kingdom of Bactria to the north, and sub- 
sequently Scythians ((^akas) and Parthians (Pahlavas) from the 
kingdom of Parthia to the west. The kingdoms established by 
these new-comers in the Punjab were overwhelmed by still another 
wave of invasion from the north. The Kushanas, a people from 
the region of China who had driven the ^^-kas out of Bactria, 
began their Indian conquests with the overthrow of the kingdom 
of Kabul about the middle of the first century A.D., and extended 
their power until, in the reign of Kanishka (probably 78 A.D.), the 
patron of that branch of the Buddhist Church which is called the 
Mahayana, the Kushana empire was paramount in N. India^ 

In Western India we can to some extent trace from inscriptions 
and coins the varying fortunes in the conflict between the Andhras 
and the invaders of N. India, and the establishment in Kathiawar 
and Cutch of a dynasty of (^aka satraps, originally no doubt 
feudatories of the Kushanas, which lasted till c. 390 A.D. when it 
was overthrown by the Guptas. 

The period of the Gupta empire which dates from 319 a.d. is a 
most important epoch in the history of Sanskrit literature. It is 

^ The Puranas say the founder of the dynasty, Simuka, but the chronological 
difficulties which this statement involves seem to be unsurmountable. 

2 See Chapters xiii (the Puranas) ; xviii-xx (the Maurya Empire) ; xxi (Indian 
Native States) ; xxiv (the earlier Andhras). 

^ For these foreign invaders of India see Chapters sxii, xxm. 

ix] Wider Political Outlook 225 

the golden age of Classical Sanskrit ; and in it most of the Puranas 
and the works belonging to the later legal literature appear to 
have assumed their present form. 

This brief conspectus of the conditions obtaining in India 
during the time to which we have to assign the Siitras, epics, and 
legal works will show that other influences than those with which 
we have been dealing hitherto are to be expected ; and these are 
indeed found, but not to such an extent as might have been antici- 
pated. These influences are indeed to be traced rather in the 
general enlargement of vision of the writers than in specific details. 
The simple village life with which for the most part the Sutras are 
concerned, the government of a circumscribed district by a local 
raja, are gradually exchanged for the life reflected from large 
towns and imperial power. Though this is more noticeable in the 
epics, it may be detected in the later Siitras and again in the still 
later law-books. During this period the power of Buddhism in- 
creased and then, reaching its culmination, began to wane. The 
world of India by the second century before Christ was already 
becoming indiflerent to the teaching of Buddliism and was being 
reabsorbed into the great pemianent cults of Vishnu and ^iva, 
with which in spirit Buddhism itself began to be amalgamated. 
The Brahman priests reasserted themselves ; animal sacrifices, 
forbidden by Agoka, were no longer under the royal ban; and 
with this open expression of the older cult the whole system of 
Brahmanism revived, fostered alike by the temple priests and their 
ritualism and by the philosophers, who regarded Buddliism as both 
a detestable heresy and a false interpretation of life. 

But there is little apparent influence from outside, despite the 
wider political outlook ; and where such influence might be looked 
for with greatest certainty, namely in the efiect of Greek domina- 
tion, it is practically nil. Only the Yavanas, literaUy 'lonians,' 
a people or peoples of Greek descent who may be traced in Indian 
literature and inscriptions from the third century B.C. to the second 
century A.D., and who were manifestly a factor of no small im- 
portance in the political history of Northern and Western India — 
they are celebrated as great fighters in the Mahabharata and other 
literature — remain to show that the conquest of Alexander and 
the Greek invasion from Bactria had any result. Other indications 
point rather to Persia than to Hellas. Thus the title Satrap, which 
was continued in use by Alexander, still remains under ^akas and 
Kushanas to testify to the long Persian dominion in N.W. India, 
Apart from this, political and social relations do not appear to be 

C.H.I. I. 15 

2 26 The Period of Sutras^ Epics ^ etc, [ch. ix 

affected at all either by Hellenic or by Persian influence. The 
native army remains of the same sort, though greatly enlarged. 
The social theory remains practically the same, save that a place 
among degraded 'outcastes' is given to Yavanas as to other 
barbarians. Architecture and the arts of sculpture, gem-engraving 
and coinage do indeed bear witness, especially in the N.W. region 
of India to the influence of Persia and Greece during this period, 
just as, at a later date, native astronomy was affected, and indeed 
practically superseded by the system of Alexandria. But the period 
with which we are dealing at present does not make it necessary 
to inquire into the relation between India and the outer world in 
respect to science. The idea that Indian epic poetry itself is due 
to Hellenic influence has indeed been suggested ; but as a theory 
this idea depends on so nebulous a parallel of plot that it has 
received no support. 



The general period of the Sutras extends from the sixth or 
seventh century before Christ to about the second century. It is 
evident that the different Vedic schools had Sutras which were 
revised, or replaced by ncAv Sutras, at various periods, and that 
some of these extended into later centuries than others. Thus it 
would be a mistake to limit all the Siitras of all the schools to 
certain centuries. The Siitras are manuals of instruction; and 
those which are of interest historically formed but a part of a large 
volume, which was intended primarily for the guidance of religious 
teachers and treated mainly of the sacrifice and other religious 
matters. Except for students of ceremonial details these sacrificial 
works (^rauta Siitras) are of no interest. What concerns us at 
present is that portion of the whole which goes by the name 
of Grihya and Dharma Sutras, that is, manuals of conduct in 
domestic and social relations. In some cases the rules given in 
these two divisions are identical ; and the two divisions are treated 
in such a way as to condense one division for the sake of not 
repeating directions given in the other. For our purpose they 
may be regarded as forming one body containing rules of life not 
especially connected with the performance of the greater sacrifices. 
They differ mainly as representing the views of different schools on 
minute points or as products of different parts of the country, and 
as earlier or later opinions. All of them claim to be based upon 
Vedic teaching. Thus the Grihya and Dharma Sutras of Apas- 
tamba form but a few chapters of a work called the Kalpa, of 
which twenty-four chapters teach the proper performance of sacri- 
fice and only two treat of the sacred law, while one abridged 
chapter gives the rules for the performance of domestic ceremonies. 
Again this special ' law-book ' is not a law-book having universal 
application, but is a product of a Vedic school belonging to the 
Andhras in the south-east of India; and, thirdly, it combats some 
of the opinions expressed by writers on the same subject. Some- 
what similar conditions prevail in the case of the other Siitras. 


22 8 ^ifi ^^^ Customs in the Sutras [ch. 

They are, in short, local manuals which form complete wholes only 
by virtue of their subject-matter, but which, to their authors, were 
merely sections of a greater work, the chief importance of which 
lay in the handing down of traditional knowledge in regard to 
religious practices. They may be regarded, however, as the first 
steps in the evolution of legal literature ; for the metrical (^astras 
or law-books are only the extension and completion of the rules of 
the Dharma Siitras, with a gradual increase in the part allotted 
to civil and criminal law and a relaxation of the bond connecting 
the Sutras with definite Vedic sects. The Dharma Siitras are more 
universal ; the Grihya Siitras reflect individual schools. But even 
the Grihyas are not (^rauta (divinely revealed), but Smarta (sacred 

The content of the Grihya Siitras, as is implied by the name, is 
narrower than that of the Dharma Sutras. The first contain, how- 
ever, to all students of folk-lore a store-house of material in regard 
to rites and superstitions connected with home life, such as no 
other body of literature in the world presents. In the first place, 
the life of man is traced religiously fi-om boyhood to burial. Every 
important phase of a man's existence is accompanied with its 
appropriate rite ; and, incidentally, what to do and what not to do, 
injunctions, prohibitions, taboos, are taught as general rules of 
conduct. The greater events, birth, marriage, death, are described 
in their religious setting, each with minute detail, so that not only 
are the sacred texts cited which should be repeated on every occa- 
sion, bAt the physical acts to which the texts are ancillary are 
described. For example, such a text must be rei)eated while a 
dead man's bones are being collected. The one who collects them 
must pick them up with such and such fingers and place them in 
just such a jar. The wedding verses are indicated ; the bride must 
make just so many steps and pour out grain with her hands held 
in just such a position, etc. Some of the Vedic schools, instead of 
embracing all the Siitras in one work as a Kalpa Sutra, have 
apparently laid so much stress on these domestic rites that the 
manuals have become independent works, thus fore-shadowing 
what happened later in the case of the ^astras. The complete 
work, embracing all kinds of Siitras, belongs, as was to be expected, 
to the Yajurveda schools, since the priests of this Veda were from 
the beginning particularly concerned with manual exercises, in 
arranging the altar, etc., and the details of sacrifice; while the 
priests of the other Vedas had to do more with the recitation and 
chanting of the sacred texts. Nevertheless, the literature of the 

x] The Grihya Sutras 229 

Rigveda also contains both ^rauta and Grihya Sutras, as does that 
of the Samaveda. Finally, the Atharvaveda possesses not only 
a Vaitana (^rauta Sutra but a Kau9ika Siitra, which is in part a 
Grihya Sutra but contains also directions for carrying out the 
many magic ceremonies connected with the text of that unique 

The preponderance of domestic ceremonies in the Grihya Siitra 
results in Dharma, or social, matter being introduced rather adven- 
titiously, as when the rules concerning the choice of wives are 
given, whereas Grihya, domestic, rules belong as much to the 
Dharma Sutras as to the Grihya Siitras themselves. The difference 
is that the weight in the Dharmas is laid on the wider relation of 
man to the state, so that those sections which deal with the family 
become condensed and subordinate. Specimens of southern Grihya 
Sutras are also not lacking. Thus as the Dharma of Apastamba 
reflects a South-Indian origin, so also the Grihya Sutra of Khadira 
belongs to Southern India; and it is an indication that Siitra 
literature extends far beyond the time of Buddlia that this should 
be the case. Such also may be surmised to be the fact (rather 
than that Vedic schools were domiciled in South India at a much 
earlier period) from the circumstance that the Siitra of Khadira is 
a later and more concise version of the Siitra of Gobhila. There 
are other examples of this endeavour to revise a Siitra on lines of 
economy, each later writer reducing the work of his predecessor as 
much as possible or convenient, conciseness being the test of Siitra 
excellence. Gobhila's work is detailed and lengthy ; Khadira's is 
virtually the same work in condensed form. Everything that could 
be omitted, such as explanatory digressions, smaller details of 
ceremonies, etc., was left out, solely to make the work easier to 
remember. But clearness as well as conciseness was aimed at and 
attained by a fresh arrangement of the older matter. 

An example of the scope and method of a Grihya Sutra may be 
taken from the directions of Khadira regarding the little oblations 
to spirits and gods required from a wedded pair. After describing 
the wedding ceremony, Khadira passes directly to this question of 
offerings and oblations, describing first briefly the fire used for the 
purpose of receiving the oblation, thus : 

The domestic {grihya) fire is that at which he has taken her hand (in 
marriage) or that on which he has put the last piece of wood (as a student before 
marriage) or a (fresh) fire twirled out (of wood), the last being pure but not tending 
to prosperity ; or he may get his domestic fire from a frying-pan or from the house 
of a man who makes many sacrifices, Qudras excepted. The service begins with 
an evening oblation. After (the fire) has been set in a blaze before sunset or 

230 Liife and Customs in the Sutras [cH. 

sunrise, the sacrifice (is performed) after sunset (and) aft^r or before sunrise. He 
should make an oblation of rice-food fit for sacrifice after washing it, if raw, with 
his hand (but) with a brass bowl if it is (not rice but) curds or milk, or with the 
rice-pot With the words ' Hail to Fire ' (he makes oblation) in the middle (of the 
fire, at eve) ; secondly in the north-eastern (part of the fire) ; in the morning, with 
the words, ' Hail to Sun ' (he makes the first oblation). The wiping round and 
other (acts) except sprinkling (of water round the fire) are here left out. Some 
say ' let the wife make the oblations,' for this fire is the house-fire and the wife is 
the house (home). When (the meal) is prepared, evening and morning, she (the 
wife) must say ('It is) ready,' (and he) must say aloud 'Om'^, but softly 'May it 
not fail ; to thee^ be reverence.' Of rice-food fit for sacrifice he should make 
(oblations) to Prajapati ; and to (the form of the Fire-god called) Svish^akrit 
(i.e. good sacrificer) make a hali (offering), depositing it outside or inside (the fire- 
place) in four places : (one) at the water-barrel ; (another) at the middle door ; 
(another) at the couch or privy ; and (finally, one) at the heap of sweepings ; sprinkle 
each (oflfering or the gi'ound with water) both (before and afterwards) and pour 
out what is left with the water toward the south. Of chaflF, water, and scum of 
boiled rice (let him make a 6a^^ off"ering) when a donation has been made. The 
gods to whom the tali off'erings belong are Earth, Wind, Prajapati, the All-gods, 
Water, Herbs, Trees, Space, Love or Wrath, the hosts of Rakshasas, the Fathers, 
and Rudra. He should make the off'ering in silence ; he should make it of any 
food (but) make it only once in case (a meal) is prepared at difi'erent times ; and 
if (prepared) at difi'erent places (then he should make the offering of) what belongs 
to the house-holder (himself). But of all food he should off'er (some) in the fire 
and give the best to a priest ; this he should do himself. He should oflFer the 
offerings himself from rice (-harvest) to barley (-harvest) or from barley (-harvest) 
to rice (-harvest) ; (yea,) he himself should offer them^. 

It will have been observed that the religious ceremony of the 
6ct/*-offering implies a cult midway between that of the Vedic 
sacrifice and the sectarian sacrifice not countenanced by the ortho- 
dox. Tlie hali is a bit of food cast upon the ground at the places 
named, the recipients being supposed to be the Vedic divinities of 
a lower order, ending with Rudra, and the hosts of harmful spirits 
who are thus propitiated. Each divinity has a hcdi in his appro- 
priate place and at the right time. Thus the offering by the couch 
is for Love ; that flung to the north is for Rudra ; that by the door 
is for (personified) Space ; and the offerings to the harmful spirits 

^ Om is the sacred syllable, answering in cases like that above to 'very good* 
(Amen). The evening and morning are mentioned in this order because the evening 
precedes the day ; and only two meals are mentioned because the Hindus eat bat twice 
a day. 

* In the Sutras clarity is often sacrificed to brevity. It is not clear here whether 
the wife or husband speaks or to whom the word 'Thee' refers. Presumably the 
husband addresses the words to the food itself (compare Gobbila's Orihya Sutra, 
1, 3, 18). The text and translation (by Prof. Oldenberg) of Khadira are published 
in 8.B.E. vol. xxix. 

' That is from spring till autumn the householder offers barley, and from rice-time 
till barley-harvest time (autumn till spring) he offers rice. The passage quoted is also 
translated by Prof. Oldenberg, in S.B.E. vol. xxix, p. 385. 

x] Rites to avert Disaster and Disease 231 

are given at night. The sprinkling of the offering means (probably) 
the sprinkling of the ground or place where the offering is cast. 
The Dharma Siitras also take up this question of offerings. The 
citation above by implication recognises only the wife as preparer 
of the meal. But a rich householder may have his meals jjrepared 
by a priest or other member of the ' reborn ' castes or even by a 
^Gdra. Special rules are necessary in the last case. The slave- 
cook, being impure, must have his hair and beard and nails cut 
daily or at least at stated intervals, and it must be the householder 
who places on the fire the food prepared by (^iidras. Then in this 
case it is the cook who says (when the meal is prepared), 'It is 
ready ' and the householder who responds (as Apastamba gives the 
rule, with a slight variation) 'Well-prepared food bestows splen- 
dour ; may it never fail. ' 

The rites involving the goblins of disaster and disease have 
naturally a prominent place in the domestic ritual of the Grihya 
Siitras and afford us glimpses of an otherwise unknown pantheon. 
The wife herself, who has so little to do with texts, must go outside 
her house and offer food to ' the white demon with black teeth, the 
lord of bad women,' and if she bears a child the husband must 
daily, till the wife's confinement ends, offer rice and mustard in 
the fire near the door where the wife is confined, dispersing demons 
whose names are given: 'Qanda, Marka, Upavira, (^aundikeya, 
Ulukhala, Malimlucha, Dronasa, Chyavana,' all indicative of trouble, 
as are those that follow (apparently a supplementary list), ' Alikhat, 
Animisha, Kimvadanta, Upagruti, Haryaksha, Kumbhin, (^atru, 
Patrapani, Nrimani, Hantrimukha, Sarshaparuna, Chyavana, 
avaunt^' But if the child falls ill with epilepsy, the 'dog-disease,' 
the father cures him by covering him with a net and murmuring, 

Kurkura, Su-Kurkiira, Kurkura (it is) who holds fast the children ; scat 
{check chet !), dog, let him go ; reverence to thee, Sisara, barker, bender ; true the 
gods have given thee a boon, and hast thou chosen my boy ? Scat, dog, let him 
go (as before). True, the Bitch of heaven, Sarama, is thy mother, Sisara is thy 
father, and Yama's black and speckled dogs thy brothers ; but scat, dog, let 
him go 2. 

The demon attacking the boy is here called Kumara, the cult is 
obviously demoniac. In general, the Siitras of this class are con- 
cerned not with the greater sacrifices, which are discussed in the 
^rauta Siitra, called the Havis and Soma sacrifices, but with the 
so-called great sacrifices of food cooked {paka) and offered on 

^ Paraskara, Grihya Sutra, i, 16, 23 f. 
2 lb. 24. 

232 J^^f^ ^'^^^ Customs in the Sutras [ch. 

special noon-days and at funeral feasts, four or seven in all, in- 
cluding offerings to serpents as well as to demons and gods^ 

The last of these domestic ' cooked-food ' sacrifices introduces a 
new feature : 

On the full moon day of the month Chaitra he makes (images of) a pair of 
animals out of meal ; (he offers) them and jujube leaves (to the gods) ; to Indra 
and Agni a figure with prominent navel ; and balls to Rudra (^ankhayana, Qrihya 
Sutra, TV, 19). 

These images of meal representing living beings are partly due to 
the new feeling of pity for animals and the desire not to injure 
life, which plays a part in Brahmanism as well as in Buddhism. 
It must be admitted, however, that economy had something to do 
with the substitution of animals of meal for real animals, but 
ostensibly it is a Vishnuite trait. The general rule in this regard 
is that attributed to Manu : * Animals may be killed (so said Maim) 
at the Madhuparka and Soma sacrifice and at the rites for Manes 
and gods.' But it is an old rite of hospitality to kill a cow for a 
guest ^ ; and, as a matter of form, each honoured guest is actually 
ofiered a cow. The host says to the guest, holding the knife ready 
to slay the cow, that he has the cow for him ; but the guest is then 
directed to say : ' Mother of Rudras, daughter of the Vasus, sister 
of the Adityas, navel of immortality (is she). Do not kill the 
guiltless cow; she is (Earth itself), Aditi, the goddess. I speak to 
them that understand.' He adds, * My sin has been killed and that 
of so-and-so ; let her go and eat grass.' But if he really wants to 
have her eaten, he says, ' I kill my sin and the sin of so-and-so ' (in 
killing her), and though in many cases the ofier of the cow is thus 
plainly a formal piece of etiquette, yet the offering to the guest 
was not complete without flesh of some sort ; and it is clear from 
the formulas that any of the worthiest guests might demand the 
cow's death, though as the ' six worthy guests ' are teacher, priest, 
father-in-law, king, friend, and Aryan 'reborn' man, and all of 
these were doubtless well grounded in that veneration for the cow 
which is expressed above by identifying her with Earth (as Aditi), 
there was probably seldom any occasion to harrow the feelings of 
the cow-revering host^ Paraskara mentions only the cow but 
(^aiikhayana {G.S., 11, 15, 1) already substitutes a goat as a possible 
alternative ; he also mentions the gods to which this animal is 

^ From the full moon of the month ^ravana, offerings to snakes have to be made 
daily till it is safe to sleep on the ground again. This is called the Pratyavarohana 
and occurs on the full moon day called Agrahdyani, when one may ' descend again ' 
(from the high couch), 

" Chapter iv, p. 101. ^ Paraskara, Qrihya Sutra, i, 3, 26. 

x] Marriage Ceremonies 233 

sacred, that is, he seeks to make the animal oflfered to the guest a 
sacrifice to a god. Thus he says that if the animal is ofiered to the 
teacher and killed it is ' sacred to the Fire-god ' ; if it is offered to 
a king, it is sacred to Indra, and if to a friend {mitra) it is sacred 
to Mitra. Similar additions may be traced in many particulars, 
sometimes found by comparing one text with another, sometimes 
clearly interpolated. 

The Sutras, while they do not recognise the sects of later days, 
yet point to the difierent conception of deity embodied in the two 
great modern sects worshipping Rudra-(^iva and Vishnu. Thus, as 
above, Rudra and the Rakshasas are also associated in the rule : 
* When one repeats a text sacred to Rudra, to the Rakshasas, to the 
Manes, to the Asuras, or one that contains an imprecation, one shall 
touch water' ((^ankh., G.S., i, 10, 9). On the other hand, when the 
bridegroom leads the bride to take the seven steps, which form 
part of the wedding ceremony, he murmurs a blessing at every 
step: 'One for sap, two for juice, three for prosperity, four for 
comfort, five for cattle, six for the seasons. Friend ! be with seven 
steps (mine) ; be thou devoted to me.' And after each clause he 
says 'may Vishnu lead thee.' Similarly, the fact that Vai9ravana 
(Kubera) and l9ana (Rudra-Qiva) are worshipped 'for the bride- 
groom ' point to the phallic nature of these cognate spirits (Par., 
G.S., I, 8, 2; gankh., G.S., i, 11, 7). 

The Grihya Sutras show that there was no one rite of universal 
acceptation in those ceremonies most intimately connected with 
domestic felicity. Indeed, the author of the A9valayana Grihya 
Sutra (i, 7, 1) says expressly that in the matter of weddings, 'cus- 
toms are diverse,' and he gives only that which is common usage. 
Thus he tells how the bride is to go about the fire, mount the 
stone, pour out grain, ga^e at the pole-star, etc., but does not 
mention other rites which other Grihya Sutras enjoin. Some of 
these, however, are of universal interest ; and a comparison of the 
Hindu ceremonies with those of other Aryan-speaking peoples 
shows that in all probability the Indian ritual has preserved 
elements reaching far back into prehistoric times \ 

^ On this point, cf. Haas and Weber, Indische Studien, vol. v ; L. von Sohroeder, 
Die Hochzeitsgebrduche der Esten und einiger anderer Jinnisch-ugrischer Volkerschaften 
in Vergleichung ndt denen der indogermanischen Volker (1888) ; M. Winternitz, Das 
altindische Vergleichung der Hochzeitsgebrauche bei den ilbrigen 
indogermanischen Volkem (1892) ; also a paper by the last writer on the same subject 
in the Transactions of the National Folk-lore Society (Congress, 1891-2), and one by 
Th. Zachariae, ' Zum altindischen Hochzeitsrituell' {Wiener Zeitschrift filr die Kunde 
des Morgenlandes, vol. xvii, pp. Ij55 f., and 211 f.). 

2 34 -^if^ ^'^^ Customs in the Sutras [ch. 

Thus in the ceremony it is universal usage to walk the seven 
steps together and for the bridegroom to murmur, as he takes the 
bride's hand : * This am I, that art thou, that art thou, this am I ; 
Heaven am I and Earth art thou; the (feminine) Rich (Rigveda 
verse) art thou, the Saman am I. Be thou devoted to me,' and to 
make the bride mount a stone as an emblem of firmness. But 
special rules are that women shall come to the bride's house and 
eat and drink brandy and dance four times ; and that merry girls 
shall escort the bridegroom to the bride's house, and that he must 
do all the foolish (?) things they tell him to do (except when taboo 
is concerned). (Qankh., G.8.^ i, 12, 2.) Some measure of values may 
perhaps be obtained from the statement that the fee to the priest 
who performs the marriage-ceremony is a cow, given by the bride- 
groom, if the groom is of the same caste as the priest, but a village 
if the groom is 'royal,' Rajanya, that is a nobleman of 'kingly' 
order, and a horse if the groom is of the third estate (farmer, trader). 
Obviously the succeeding rule, which is not unique, countenances 
a sort of sale in that it adds : * (The bridegroom must give) to the 
one who has the daughter one hundred (cows) together with a 
chariot' The same rule is found in the Dharma Siitras (Apas- 
tamba, ii, 13, 12) with the explanation that the gift must be 
returned, as a sale is not allowed — which only points back to an 
earlier period when the sale of daughters was allowed. 

The distinction among the orders mentioned in the gifts above 
is only one of innumerable passages in which, as a matter of course 
and without thought of any other social order, the castes are 
named as priest, noble or warrior, and 'people,' the last term 
embracing all those 'reborn,' who are not priests or warriors or 
slaves. The slaves, ^udras and lower orders, are recognised as 
part of the social structure. The name itself suggests that the 
(^iidras were originally a conquered people, as Karian became 
synonymous with slave at Athens. Yet the ^udras were not 
Pariahs but members of the household, who took part in some of 
the domestic rites. 

The test of caste is not marriage alone but defilement by eating 
and touching what is unclean. In this regard the Sutras show 
only the beginning of that formal theory of defilement which 
results in a pure man of the upper castes being defiled by the 
shadow of an impure man, and in the taboo of all contact with the 
impure. According to Gautama {Dharrrm Sutra, xvii, 1 f.), a 
Brahman may eat food given by any of the ' reborn ' who are worthy 
members of their caste, and if in need of food to support life he may 

x] Caste and Family 235 

take food and other things even from a (^udra. Food forbidden is 
that defiled naturally by hairs or insects falling into it and that 
touched by a woman during her courses, by a black bird (crow), or 
by a foot, etc., or given by an outcast, a woman of bad character, 
a person accursed, an hermaphrodite, a police-officer {dandika), 
a carpenter, a miser, a jailer, a physician, a man who hunts without 
using the bow (i.e. a non-Aryan snarer of animals), a man who eats 
refuse or the food of a multitude, of an enemy, etc. The list con- 
tinues with the taboo of food offered disrespectfully and of certain 
animals. Apastamba {DJiarma S., i, 6, 18, 1 f ) allows the acceptance 
of gifts, including a house and land, even from an Ugra (low caste 
or mixed caste), though, like the later law-books, his code states 
that a priest may not eat in the house of anyone of the three 
orders (varnas) below him ; but he may eat the food of any other 
priest, and according to ' some ' he may eat the food of people of 
any caste except ^udras and even their food in times of distress. 
Forbidden by him is the food of an artisan, of people who let 
houses or land, a spy, an unauthorised hermit (Buddhist?), besides 
that of surgeon, usurer, and others. Caste is varna or jdti, 
'colour' and 'kin,' the former embracing the latter, as a social 
order including clans or families. Even in the all-important 
matter of marriage, caste is not so important as family. The only 
test, when one seeks a wife, according to ^^ankhayana, is that of 
the family: 'They ask the girl in marriage, reciting the clan- 
names.' The text of A^valayana expressly mentions as a form of 
marriage that in which the bridegroom kills the relatives and 
rapes the weeping girl, evidently a form once countenanced as 
well as enumerated among possible forms ; at any rate it bars out 
all examination of the bride's social position. Indeed the marriage 
rules permit the marriage of a (^udra woman, though as the last of 
four wives, with a member of the highest caste (e.g. Par., G.S., i, 
4, 11), whose offspring, of course, being 'mixed' or impure, is not 
a member of the xlryan 'reborn,' but nevertheless is recognised 
legally. And what shall we say of those who are not ' reborn ' 
although Aryans? The rule in this case is universal that, if priest, 
warrior, or member of the third estate fail to be ' reborn in the 
Veda,' i.e. if such a one is not duly initiated into his social order 
at the proper time, he loses his prerogatives and becomes an 
outcast: 'No one should initiate such men, nor teach them, nor 
perform sacrifice for them, nor have intercourse with them,' and 
further, 'A person whose ancestors through three generations have 
been thus outcast is excluded from the sacrament of initiation and 

236 Liife and Customs in the Sutras [ch. 

from being taught the Veda,' that is, they become Vratyas or 
entirely outcast persons with whom one may not even have inter- 
course unless they perform special rites ^ 

In general the Grihya Sutras may be said to be the later 
scholastic codification of rules, formulas, and rites long practised, 
concerned chiefly with the orderly progress of an individual ideal 
life, and incidentally with such ceremonies as naturally occur in 
such a life, that is, besides rites from babyhood to marriage, 
fixed moon-rites, etc., those concerned with building, holidays, 
burial, etc. That they are not of Vedic age in their present fonn, 
though in substance reverting in part to Brahmana beginnings, 
may be concluded from their obvious posteriority in respect of 
language and metre (where verses are cited) to the Brahmanas, 
not to speak of earlier Vedic texts, as well as from the fact that 
several Sutras emanate from districts scarcely known even by 
name to the Brahmanas. The general order of an-angement in 
the Grihya Sutras' is one conditioned by the subject-matter, which 
is to reveal the whole duty of man as a householder. Most of 
them begin with the marriage and continue with the birth of a 
child, the ceremonies at conception and at various stages before 
birth, at the birth itself, at the naming of the child, when he sees 
the sun, when he is fed, when his hair is cut, when he becomes 
a student, and when he returns home from his Guru (tutor) and 
becomes a householder. Then the child, now grown to a man, 
marries and the circle begins again. Finally the rite for the burial 
is described. A few texts take up the round of life at another 
point, that where the student-life begins. This is the procedure 
in the case of some of the Black Yajurveda texts (for example, the 
Manava and Kathaka Sutras), but it makes no difference where 
one begins; each Sutra follows out the life to the end, and the 
general uniformity shows that, whatever be the minor discrepancies 
and divergences of opinion (of which the authors are themselves 
well aware), the Grihya Sutras as a whole are based upon one 
model, and that, whether in the northern or southern districts, the 
lives of orthodox Aryans were governed by a remarkable conformity 
of ritual. It is not improbable that, as has been suggested by 
Professor Oldenberg, many of the rites prescribed as general rules 
were nothing more than formulas of secret magic owned at first 
by certain families and afterwards become universal property^. 

* See Paraskara, Grihya Sutra, 11, 5, 40 f., and Weber, Ind. Literaturgeach. p. 78 f., 
Eng. trans., p. 67. 

2 Compare the admirable discussion of the position of the Grihya Sutras by 
Prof. Oldenberg in S.B.E. vol. xxx. 

x] Religious and Magical Beliefs 237 

The specimen given above will suffice to show the artless style 
of these didactic Sutras. They have in fact no style save that 
attained by scrupulous brevity. In the following paragraphs we 
shall seek rather to illustrate certain phases of the Grihya Sutras 
as indicative of religious and magical beliefs and of the social 
environment in which they were produced, or at least for which 
they were intended. 

We may begin with reverting to the cure of epilepsy already 
mentioned. In the course of childhood the boy may be attacked 
by the dog-demon (epilepsy). What is the father to do? The 
names of the canine demons have been mentioned above with a 
parallel passage containing more of the same sort. These are to 
be averted by a sort of honorific propitiation. They are lauded ; 
but their objectionable behaviour in this special case is deprecated. 
The author of our Sutra contents himself with this. But a rival 
author or two (Hiranyake9in, G.S.^ 11, 2, 7, 1 f. ; Apastamba, Grihya 
Sutra, VII, 18, 1) are not content with the method here advocated. 
According to them, the father must make a hole in the roof of the 
royal gaming-hall and pull the boy through it, lay him on his back 
on dice strewn about, and then, while a gong is sounded, recite the 
deprecatory words to the dog-demons and pour curds and salt over 
the boy. Several items of this recipe are of interest, the avoidance 
of the door, the use of salt and curds to frighten demons, the gong 
for the same purpose to be beaten on the south side of the hall. 
These may be said to be universal antidotes ; peculiar is the use 
of the dice, which has no parallel in the similar situations offered 
by the Sutra. Finally the fact that the father makes a hole in the 
roof of the gaming-hall shows that it is made of thatch (easily 
repaired) and leads to the question what sort of architecture is 
normally to be found implied in the Sutras. The gaming-hall is 
the public gambling-place which a king is directed to build for the 
use of his subjects, and curiously enough, with the exception of 
the householder's own dwelling, it is almost the only reference to 
edifices found in the Siitras. On the other hand, all the dicta of 
the Siitras show that such life as is depicted is supposed to be 
country life; the district and the village are the geographical 
entities. Cities are not ignored but are despised \ Thus there 
are no ceremonies for urban life. But there is a rite for ploughing, 
when sacrifice is made to A9ani (the thunder-bolt) and to Sita 

* Apastamba, Dharma Sutra, i, 32, 21, 'let him avoid going into towns,' and 
Baudhayaua, Dharma Sutra, ii, 3, 6, 33, ' It is impossible for one to obtain salvation, 
who lives in a town, covered with dust.' 

238 Liife and Customs in the Sutras [ch, 

(the furrow), as well as to other bucolic deities, Arada, Anagha, 
etc., as to the greater bucolic gods, Parjanya and Indra and Bhaga, 
with similar offerings on the occasion of the ' furrow sacrifice,' the 
* threshing-floor sacrifice,' when one sows, reaps, or takes in the 
harvest, all indicating that the life portrayed is that of the village 
agriculturist, who must even 'offer a sacrifice at mole-heaps to 
Akhuraja, the king of moles ' (Gobhila, Grihya Sutra, iv, 4, 28 £ ; 
ibid. 30 f.). So the constant injunctions to 'go out of the village,' 
to sacrifice at a place where four roads meet, or on a hill, etc., 
imply life in villages even for householders and scholars rather 
than in towns (Gobhila, iii, 5, 32-35). 

Besides the introduction of evil spirits and bucolic divinities 
into the ritual of the domestic service, we find in the Sutras for 
the first time the recognition of images of the gods, which must be 
implied by the regulations concerning the deities l9ana, MidhushI, 
and Jayanta ('lord,' 'bountiful one,' 'conqueror') as well as the 
' lord of the field,' Kshetrapati, who are moved about and given 
water to drink (Ap., G.S., vii, 19, 13; ibid. 20, 1-3 and 13). 

When a boy is initiated he is made to mount a stone with the 
adjuration to be ' firm as a stone ' which elsewhere is confined to 
the bride, and is then given in charge to 'Kashaka (Ka9aka), 
Antaka, Aghora, Disease, Yama, Makha, Va9ini, Earth and 
Vai9vanara, Waters, Herbs, Trees, Heaven and Earth, Welfare, 
Glory, the All-gods, all the Bhuts, and all the gods ' (Hiranyake9in, 
G.S., I, 2, 6, 5). In this list of demons and deities to whom the 
boy is given in charge, Va^inl as the 'ruling goddess' is notice- 
able. She is probably the mother-goddess who despite all Vedic 
influence always was the chief spiritual village-power identified 
with (^iva's wife in various forms. Perhaps too the recognition 
(in a rite to procure increase of cattle) of a god described merely 
as 'He who has a thousand arms and is the protector of cow- 
keepers ' {Gaupatya), may be a veiled allusion to Krishna- Vishnu 
(Gobhila, iv, 5, 18). 

As the Grihya Sutras in distinction from the Dharma are 
concerned with domestic superstitions, these may rightly be 
considered their peculiar contribution to the history of India. 
Of political and social life they contain almost nothing except as 
confined within the bounds of the family. The regular routine of 
the normal life contains a sufficiency of such superstitions, though 
the underlying reason for them is due in some cases more to 
mechanical adjustment to a supposed harmony than to spiritual 
fears. This is the case for example in the regulation that the 

x] Minor Superstitions 239 

initiation of the Brahman, Kshatriya, and Vai9ya shall take place, 
respectively, in spring, summer, and autumn, in the eighth, 
eleventh, and twelfth years after conception, the respective 
seasons being supposed to represent the castes, as the years 
represent the metres regarded as peculiar to these castes. Deeper 
lies the origin of the following : — the rite to drive out of the bride 
the influence deadly to the husband and to convert it into an 
influence deadly to her possible paramour (Hir., G.S., i, 7, 24, 1 f.) ; 
the prayer that the 'weeping women' (demons) and Vike9i may 
not torment her, nor the Pi9achas of the womb, who devour flesh 
and bring death {ihid. 6, 19, 7); the scattering of rice and other 
grains on the heads of the newly wedded pair {ihid. 21, 6); and 
the corresponding rite according to which the husband ties barley 
about the wife's head, here expressly ' to have ofi^spring ' (Ap., G.S., 
VI, 14, 7). Naturally the conjugal relations off"er a fruitful field for 
this sort of thing. Thus we have a rite to make a husband subject 
to his wife as well as to make her co-wives subject to her {ibid. 
Ill, 9, 5 f ) and another very peculiar rite, the object of which is to 
keep the wife faithful, in which she is regarded much as is the slave 
around whom, when suspected of estrangement, urine is poured 
from a horn to keep him magically at home (Hir., G.S., i, 4, 14, 2). 
Another subject claiming the attention of the Sutra-maker is 
the efficacy of amulets. These are tied upon the priests, as a sort 
of final expression of good-will, in the A9vayuja rite. They are 
made of lac and herbs (Gobh., iii, 8, 6). Minor superstitions 
abound. If one yawns, one must say, 'May will and wisdom 
abide in me,' evidently a phase of the popular belief that the 
soul may escape in a yawn or sneeze (Hir., G.S., i, 5, 16, 2). 
Signs of ill-luck which must be averted by a sacred formula are 
found in the presence in the house of a dove, of bees, or an anthill, 
in the budding forth of a post, etc. {ii)id. i, 5, 17, 5). The trans- 
mission of sin is illustrated by the dictum that if one touches 
a sacrificial post the faults committed at the sacrifice are incurred 
{ibid. 16, 16) ; also by the injunction that when one's hair is cut a 
well-disposed person should gather it up and hide it away, as the 
well-disposed person (the mother, for example) thus ' hides the sin 
in the hair,' probably a refinement on the original notion of not 
losing one's soul-strength at the hands of some ill-disposed person 
{ibid. I, 2, 9, 18; cf Agv., i, 17, 10, etc., where the formula is 'for 
long life'). Whether the objection to certain trees as liable to 
cause eye-trouble, etc., is grounded in fact or fancy, causing the 
injunction to transplant them, may be questioned, but the original 

240 Life and Customs in the Sutras [ch. 

cause has been lost in the maze of superstition, which makes the 
A9vattha tree injurious on the east side of the house, the Plaksha on 
the south, the Nyagrodha on the west, and Udumbara on the north. 

Before speaking of the Dharma Sutras in particular it will be 
necessary here to settle the question as to what is meant by the 
Aryan, so often mentioned in all the Sutras. While not lacking in 
moral connotation, so that as a common adjective dry a meant 
noble in heart as well as in race, it is only in the democracy of 
religious philosophy that such a person as an Aryan slave or bar- 
barian was conceivable. Practically Arya was synonymous with 
'reborn' and indicated a person of the three upper castes in 
good standing, antithetic to (^^udra and other low-caste or out-caste 
persons. Yavanas (Greeks) are the most esteemed of foreigners, 
but all Yavanas are regarded as sprung from (^Judra females and 
Kshatriya males. Gautama says that sundry authorities hold this 
view\ Such rules as that given by Gautama (xii, 2) in the case of 
the violation of an Aryan woman by a (^^udra, when compared with 
Apastamba, Dh.S., 11, 26, 20, and 27, 9, prove conclusively that 
Arya is ' noble in race ' as distinguished from the ' black colour ' 
(ibid. I, 27, 11, with the preceding 'non-Aryans'). Mr Ketkar in 
his History of Caste in India (p. 82), is rather rash in stating 
that there was no racial discrepancy felt between Aryan and 
Dravidian. It is true that those who were out-caste were no 
longer called Aryans, but no Qudra was ever regarded as Aryan, 
any more than he could be 'reborn.' Arya indicated racial dis- 
tinction from the times of the Rigveda onwards. 

We have seen that the Grihya Sutras practically recognise 
life only as lived in villages. In the Dharma Sutras, as these are 
later and have to do with wider relations, the town {jpwr, nagara), 
appears as a larger unit, though how much larger it is not easy to 
say ; and when we remember that pur is after all only a stronghold 
or fort, and nagara is anything larger than a village, we must be 
cautious of too ready belief in large cities. Everything indicates 
on the contrary that life was still chiefly that of small places and 
kings were only petty chieftains. There was not supposed to be 
any school or even studying done in town. The Dharma Sutra of 
Gautama, regarded as the oldest of extant Dharma Sutras, says 
expressly that one should not recite the holy texts at any time in 
a town ; and it is assumed, as in the Grihya Sutras, that such life 

^ Dh. (Rostra, iv, 21 (erroneously rendered ' offspring of male ^udras and female 
Kshatriyas' in S.B.E. vol. ii, p. Ivi). This passage referring to Yavanas is unique 
in the Sutras. They are Bactrian and other Asiatic Greeks. See Chap, xxu, pp. 540-1. 

x] The Dharma Sutras 241 

as is described passes normally in villages. Even in the descrip- 
tion of the royal residence {v. inf. p. 247), the hall has a thatched 
roof. The king still stands up in propria persona and hits a thief 
with a cudgel; and, if the king fails to strike, the 'guilt falls on 
the king' (Gaut, DhS., xii, 43). The commentators, apparently 
aware of the incongruity in applying such a rule to the kings 
of their day, attempt to restrict its application as intended for 
specially evil thieves (of gold); but it is in fact a general rule 
even as late as Apastamba {Dh.S., i, 25, 4), who says: *A thief 
shall loosen his hair and appear before the king carrying a cudgel 
on his shoulder. With that (cudgel) he (the king) shall smite him ; 
if he dies his sin is expiated, but, if the king forgives him, the guilt 
falls on him who forgives ; or he (the thief) may throw himself into 
a fire or die by starvation.' Thus the later author seeks to excuse 
the king (but not the thief). 

The Dharma Sutras add to the data of social life material 
evidence which shows that there were recognised customs not 
approved in one part of the country but doubtfully admitted as 
good usage because locally approved in other parts. For, in dis- 
cussing usage, Baudhayana (Dh.S., i, 1, 17 f.) expressly says that 
customs peculiar to the South are to eat in the company of an 
uninitiated person, in the company of one's wife, to eat stale food, 
and to marry the daughter of a maternal uncle or of a paternal 
aunt, while customs peculiar to the North are to deal in wool, 
to drink rum, to sell animals that have teeth in the upper and 
in the lower jaws, to follow the trade of arms, and to go to sea. 
He adds that to follow these practices except where they are con- 
sidered right usage is to sin, but that for each practice the local 
rule is authoritative, though Gautama denies this^ Baudhayana 
also admits the doctrine that a priest who cannot support himself 
by the usual occupations of a Brahman may take up arms and 
follow the profession of a warrior ; though here again his opinion 
is opposed to that of the earlier Gautama, who argues that such 
an occupation on account of its cruelty is not fitted for a priest. 
Whether the Gautama here represented as opposed be the Gautama 
whose Sutra has come down to us may be doubted, but the two 
passages show that caste-integrity was not regarded as essential, 
for no one could be a warrior and retain the mode of life deemed 
proper for a priest. 

^ See Biihler, S.B.E. vol. n, p. xlix. The river Narmada (Narbada) is the boundary 
between North and South. 'Making voyages by sea' causes loss of caste (Baudh., 
Dharma Sutra, n, 1, 2, 2). 

C.H.I. I. 16 

242 Life and Customs in the Sutras [CH. 

Tlie geography of the Sutras illustrates very forcibly the limited 
reach of interest at the same time that knowledge of a wider 
country was thoroughly disseminated. Kaliiiga on the eastern 
coast is even the subject of versification, * He sins in his feet who 
visits the Kalingas/ and one who travels to their country must 
perform a purificatory sacrifice ; as must they who visit the Arattas 
(in the Punjab) or the Pundras and Vangas (in Bengal), while the 
inhabitants of the country lying about Multan, Surat, the Deccan, 
Malwa, western Bengal, and Bihar are all declared by Baudhayana 
to be of mixed origin ; and (by implication) their customs are not 
to be followed. The * country of the Aryans ' embraces in fact only 
the narrow district between the Patiala district in the Punjab 
and Bihar, and between the northern hills (Himalayas) and those of 
Malwa ; some even confine the definition of Aryavarta (country of 
the Aryans) to the district between the Ganges and Jumna ^ 

Constant references to the opinions of earlier authorities, in- 
definitely cited as * some,' show that our extant Sutras are but a 
moiety of the mass lost. Naturally the later authors know by 
name more authorities than do the earlier. Apastamba discusses 
* those whose food may be eaten ' and cites a certain Kanva who 
declares that ' who wishes may give ' ; then a Kautsa, whose opinion 
is that he who is holy {punya\ may give ; then Varshyayani who 
says that ' anybody may- give,' because, if it is a sinner and the sin 
remains with him, the receiver cannot sufifer, but if it does not remain 
with him (the giver), then the giving acts as a purification (Ap., 
Dh.S., I, 19, 3 f.). Again the same author discusses theft. Any- 
one who takes what belongs to another is a 'thief; so teach 
Kautsa, Harita, and Kanva ; but Varshyayani says that there are 
exceptions. ' Seeds ripening in the pod and food for a draught-ox' 
may be taken (without theft), though * to take too much ' is a sin. 
Harlta's opinion is that the owner's permission must first be given 
(Ap., Dh.S., I, 28, 5). 

These texts in any case are more or less erroneous transmitters 
of older law. Thus the Sutra law for manslaughter or murder 
enjoins that one who has killed a warrior shall give for the expia- 
tion of his sin a bull and a thousand cows. To whom? The 
commentator (a priest) says that the passage means give to the 
priests (Ap., Dh.S., i, 24, 1), whereas the corresponding rule in 
Baudhayana (i, 10, 19, 1) says that the fine shall be given to the king ; 

^ Bandh. i, 1, 2, 9f. Baudhayana may be the Kanva referred to (in the next 
paragraph) as an authority. He was probably himself a southerner of the eastern 
coast. Cf. Biihler, S.B.E. vol. xiv, p. xxxvi f. 

x] Inheritance 243 

and in both passages the commentator explains that the ' expiation 
for sin ' may mean ' to remove the enmity of the murdered man's 
relatives,' which latter explanation is historically the earlier and 
probably the true explanation, as it is a parallel to the law per- 
mitting compensation for murder as found among other Aryan 
nations \ 

Since, in distinction from the Grihya Sutras, the Dharma 
Siitras have to do with society rather than with family, it is here 
that we find the beginning of civil and criminal law, although 
legal punishments are still retained in part under the head of 
penance, and the conditions of inheritance, which depend on the 
family, are partly explained under domestic duties, for these 
include (as we have seen) the rite of marriage, apropos of which 
is first defined the family {gotra, gens) into which one may marry. 
The rule is that a man shall not give his daughter to one belonging 
to the same gotra, that is, having the same family name^, or, in 
the case of priests, descended from the same Yedic seer, or to one 
related on the mother's side within six degrees. Then the rules 
for inheritance, assuming the meaning of the Sapinda as one 
within six degrees, make Sapindas the heirs after or in default of 
sons. The Sapindas here are males only. The widow is excluded, 
and the daughter (according to Apastamba) inherits only in default 
of sons, teacher, or pupil, these, however, being recommended to 
employ the inheritance for the spiritual good of the deceased. 
Probably the general rule anticipates not the death of the owner 
but a division of property among the sons during his lifetime. 
The king inherits in default of the others named, and some say 
that among the sons only the eldest inherits. These rules are 
sufficiently vague, but local laws are also provided for in the 
additional rule: *In some countries gold, (or) black cattle, (or) 
black produce of the earth (grain or iron ?) is the share of the 
eldest' (Ap., Dh.S., 11, 14, 7). Then in regard to what the wife 
receives, the Sutra leaves it doubtful whether the rule * the share 
of the wife consists of her ornaments and wealth received from 
her relations, according to some (authorities),' is to be interpreted 
in such a manner that 'according to some' refers only to the 
last clause or to the whole. 

What is obvious is that the whole matter of inheritance was 
as yet not regulated by any general state law. Different countries 

^ Cf. Biihler's explanation, S.B.E. vol. n, p. 78. 

" Generally speaking we may say that exogamy is the rule, but epic literature 
records cases of maniage between near relations (cousins). 


244 ^\f^ ^^</ C us 1 07ns in the Sutras [ch. 

or districts of India have different laws; different authorities 
differ in regard to the interpretation of these laws; and, finally, 
different texts of Vedic authority contradict by inference the rule 
to be got from them. Thus because one Vedic text says * Manu 
divided his wealth among his sons,' it is implied that there should 
be no preference shown to the eldest; but, on the other hand, 
another Vedic text says 'they distinguish the eldest by the 
heritage/ which countenances the preference shown to the eldest. 
Now this last point, despite the desire for conciseness, demands 
consideration at length, so the maker of the Sutra takes it up, 
arguing that a mere statement of fact is not a rule. For example 
(he says), the dictum ' a learned priest and a he-goat are the most 
sensual beings' is a statement, but cannot be taken as a rule. 
Hence, he says, the statement * they distinguish the eldest ' is not 
a rule. But the question remains, why then should the other 
statement, *Manu divided his wealth,' be regarded as a rule? 
The subject of inheritance is treated first by Baudhayana under 
the head of impurity, where he says simply that Sapindas inherit 
in default of nearer relations, and Sakulyas (remoter relations) in 
default of Sapindas ; but afterwards he adds that the eldest son, 
in accordance with the quotations cited by Apastamba, may 
receive the best chattel, or the father may divide equally among 
his sons. Here also the fact that the same subject is treated in 
different sections shows that as yet the matter of civil law was 
not treated systematically but incidentally. 

It is no part of the present discussion to enter into the 
confusing details of the laws of inheritance; only to show in 
what state were these laws at the time of the Sutras. The latest 
Siitra, however, already stands on a level with the formal law- 
books, and, for example in this matter of inheritance, is not 
content with the vague 'sons' of the earlier authors but makes 
a formal classification of the (later legal) 'twelve sons,' six of 
whom are entitled to inherit as 'heirs and kinsmen' while six 
(kinds) are ' kinsmen but not heirs,' among the last being the son 
of a (^udra wife. 

Civil law is in general discussed in the Sutras under the head 
of royal duties ; for it is assumed that the king administers justice 
both civil and criminal. It is his part to pay attention to the 
special laws of districts, castes {jciti), and families, and make the 
four orders (varnas, castes in a general sense) fulfil their duties. 
The summary, in the following order, includes punishing those 
who wander from the path of duty, not injuring trees that bear 

x] Royal Duties 245 

fruit, guarding against falsification of weights and measures, not 
taking for his own use the property of his subjects (except as 
taxes), providing for the widows of his soldiers, exempting from 
taxation a learned priest, a royal servant, those without protectors, 
ascetics, infants, very old men, students, widows who have returned 
to their families, unmarried girls, wives of servants, and pradattds 
(doubtful, perhaps girls promised in marriage) ; but first and fore- 
most, the king is to protect all in his realm (Vasishtha, xix, 1-24). 
This quaint summary of royal duties does not even belong to the 
early Sutra period but derives from a text, which in some regards 
is practically, as it is called, a law-book (^astra). It reflects, as 
do the elaboration of details and additions casually made, the fact 
that even at this comparatively late period the king was still a 
small local raja, not an emperor. 

Although we may agree in general with the judgment of 
Buhler to the effect that the Dharma (^astra of Gautama takes 
temporal precedence over the extant Dharma ^^astras and Dharma 
Sutras \ yet it is historically as important to remember that this 
judgment was tempered by the consideration that interpolations 
occur in the work of Gautama, and that in its present form the 
language 'agrees closer with Panini's rules than that of Apastamba 
and Baudhayana.' The title itself of Gautama's work is (^^astra 
not Sutra, and it is obvious from his chapter on kings that sundry 
works called Dharma ^astras were in vogue, for he says: 'The 
administration of justice (shall be regulated by) the Veda, the 
Dharma ^^^tras, the Aiigas, and the Puranas (and Upavedas)' 
(xi, 19), and though the word 'Upavedas' occurs in but one 
manuscript, and logically Dharma is included under Anga, yet it 
is not necessary to assume an interpolation for these words, 
especially as Gautama mentions Manu among teachers of the 
law, from 'some' of whom he cites, though not by name. The 
Atharva^iras, a late work, is also known to him (xix, 12). It 
may then be questioned whether each and every rule of 
Gautama can be cited as being an integral part of the 'earliest 

The royal duties as described by Gautama are few. After 
stating that all the ' reborn ' (men of the three upper castes) are 
to study, ofler sacrifice, and give alms, and that the priest in 
addition is to teach, perform sacrifice for others, and receive alms, 
or, if he does not do the work himself, to practice agriculture and 

1 S.B.E. vol. u, p. liv. 

246 Life and Customs in the Sutras [ch. 

trade ^, Gautama says that a king's special additional duty is to 
protect all beings, to inflict proper punishment, to support learned 
priests and others unable to work, those free of taxes and 
temporary students, to take measures for ensuring victory, to 
learn how to manage a chariot and use a bow, to fight firmly, 
to divide the spoils of battle equitably, to take a tax of one-tenth, 
one-eighth, or one-sixth (of produce), to force artisans to pay one 
day's work monthly, to proclaim by crier lost property, and, if the 
owner be not found in a year, to keep it, giving one-fourth to the 
finder (but all treasure-trove belongs to the king), and to protect 
the property of infants^. In the following section the author 
says that the king is the master of all except the priests ; that he 
is to be moral and impartial, worshipped by all except Brahmans, 
who shall honour him {ihid. xi, If.); that he must protect the 
castes (orders) and different stages of life {cU^ramas), and, with 
the assistance of his chaplain fulfil all his religious duties, as 
enumerated above. Authoritative in the realm shall be all rules 
of castes (jciti), and families {hula), as well as district-rules not 
opposed to (Vedic) tradition, while for their respective orders 
(varga) ploughmen, traders, herdsmen, money-lenders, and artisans 
may make their own rules {ibid. 21). 

In this r4sum4 of royal duties there is no indication or impli- 
cation of any power greater than that of a small king. But the 
later Sutra of Apastamba indicates the beginning of that system 
of government by proxy which obtains in the ^astra of Manu and 
other Smritis. Nor is Apastamba's account of royal duties other- 
wise without interest, since it shows just such a combination of 
old and new as characterises the Sutra period. To begin with, 
after discussing caste-duties in general, Apastamba describes the 
town where the king is to live : 

I will now explain the duties of a king. He shall build a town {pur), and 
a dwelling (vegma), each with a door facing south. The dweUing (Biihler, ' palace ') 
is within the pur, and to the east of the dwelling shall be a hall called the 
' invitation ' (guest) place. South of the pur shall be an assembly-house (sabha), 

1 This and the permission to teach for money are not in accord with the usual 
rules of the Sutras. The practice of Brahmans becoming 'gentlemen farmers and 
sleeping partners in mercantile or banking firms managed by Vaipyas' is not 
countenanced in other Sutras (see Biihler's note to Gautama, x, 5) and probably 
the permission to teach for money is intended only for priests in distress. 

* An exception in the case of treasure-trove is made in the case of a priest being 
the finder, and ' some ' say that anybody who finds it gets one sixth. In the rules for 
taxes, if the stock is cattle or gold the tax according to ' some' is one fiftieth and if it 
is merchandise one twentieth, while one sixtieth is the tax on roots, fruits, flowers, 
herbs, honey, meat, grass, and firewood (Gaut., x, 25 f.). 

x] Taxes ^ Status of Women ^ etc, 247 

having doors on the south and north sides so that it shall be in plain view within 
and without. There shall be fires in all these places (burning) perpetually, and 
offering to the Fire-(god) shall there be made regularly, just as to the sacred 
house-fire. He shall put up as guests in the hall of invitation learned priests... 
and in the assembly-house he shall establish a gaming-table, sprinkle it with water, 
and throw down on it dice made of Vibhitaka (nuts), sufficient in number, and let 
Aryans play there (if they are) pure men of honest character. Assaults at arms, 
dances, singing, concerts, etc., should not take place except in houses kept by the 
king's servants... Let the king appoint Aryans, men of pure and honest character, 
to guard his people in villages and to^vns, having servants of similar character ; 
and these men must guard a town inagara) from thieves for a league {yojana), in 
every direction ; villages for two miles (a kos or quarter of a league). They must 
pay back what is stolen within that distance and collect taxes (for the king). 

A learned priest and women are not taxed, nor are children 
before puberty, temporary students, or ascetics, or slaves who 
wash feet, or blind, dumb, deaf, and diseased persons. The king 
goes personally into battle and is exhorted not to turn his back 
and not to use poisoned weapons or to attack those who supplicate 
for mercy or are helpless (Apastamba, ii, 5, 10, 11), such as those 
who have ceased to fight or declare themselves cows (by eating 
grass, a sign of submission) (Baudh., i, 10, 18, 11; Gaut., x, 18). 

Taxes and inheritance form the chief subjects of civil law, 
together with the vexed question of the status of women. Women 
may not on their own account ofier either the Vedic (^rauta 
sacrifices or the Grihya sacrifices. A woman is 'not independent' 
(Baudh., II, 2, 3, 44; Gaut., xviii, 1), either in respect of sacrifice 
or of inheritance. Widows, if sonless, are expected to bear sons 
by the levirate marriage (Baudh., ii, 2, 4, 9). Suttee is not acknow- 
ledged. Women are property and come under the general rule : 
* A pledge, a boundary, the property of minors, an open or sealed 
deposit, women, the property of a king or of a learned priest are 
not lost by being enjoyed by others ' (Vas., xvi, 18). 

In proving property, documents, witnesses, and possession are 
admitted as proof of title by the late Sutra of Vasishtha (xvi, 19), 
and if the documents conflict, the statements made by old men 
and by gilds and corporations are to be relied upon (Vas., xvi, 15), 
an interesting passage as it shows what importance was ascribed 
to the gilds {^reni) of the time. 

In criminal law, only Apastamba recognises the application of 
ordeals {Dh.S., ii, 11, 3; cf. 29, 6). The ordeals, here merely 
referred to, consist in the application of fire, water, etc., according 
to the later law-books (v. inf. p. 282 ff".), but are not defined in the 
Siitras. Assaults, adultery, and theft are the chief subjects dis- 
cussed in the Sutras under this head. The fines of the later law 

248 ^\f^ ^^(^ Customs in the Sutras [ch. 

are generally represented here by banishment or corporal injury. 
Most of the regulations are dominated by caste-feeling. A ^udra 
who commits homicide or theft or steals land has his property 
confiscated and suffers capital punishment (Ap., Dh.S., 11, 27, 16); 
but a Brahman priest for such crimes shall be blinded (ibid. 17). 
A Kshatriya (warrior) who abuses a Brahmana (priest) is fined 
one hundred (coins); a Vai9ya (farmer) must pay half as much 
again for the same offence ; but if a Brahman abuses a Kshatriya 
he pays only fifty coins (kdrshdpanas), and only twenty -five if he 
abuses a Vai9ya, while if he abuses a ^udra he pays nothing 
(Gaut., XII, 8f.), etc. The same caste-interest works outside of 
criminal law. 

Thus the legal rate of interest is set at (the equivalent of) 
fifteen per cent, per annum (five mdshas a month for twenty 
kdrshdpanas, Gaut, xii, 29; Baudh., i, 5, 10, 22); but according 
to Vasishtha (11, 48), 'two, three, four, five in the hundred is 
declared in the Smriti to be the monthly interest according to 
caste.' This means that the highest caste pays two, the next 
caste three, and so on (limited by the scholiast to cases of loans 
without security). The same author prohibits Brahmans and 
Kshatriyas from being usurers ; but Baudhayana says that a Vai9ya 
may practise usury (Vas., 11, 40 and Baudh., i, 5, 10, 21). That 
there was, however, a notable laxity in carrying out the supposed 
inflexibility of caste-rules is evident from the fact that the law- 
makers expressly permit the upper castes to take to the occupa- 
tion of the lower when in need of sustenance. Even the Brahman 
priest who neglects to say his prayers may at the king's pleasure 
be forced to perform the work of ^udras (Baudh., 11, 4, 7, 15). 
Thus, with certain restrictions as to what he sells, etc., a priest or 
warrior may support life by trade and agriculture (Vas., 11, 24 f.). 
But a man ' reborn ' who persists in trade cannot be regarded as 
a Brahman, nor can a priest who lives as an actor or as a physician 
(ibid. Ill, 3). In other words, as may be concluded from the very 
laws inveighing against them, at the time of the Sutras there were 
many nominal members of the priestly and royal orders who lived 
as farmers and traders, perhaps even as usurers (a special law 
prohibits this, Vas., 11, 40; c£ Mann, x, 117), not only acting the 
part of gentleman farmers but living as humble ploughmen (Vas., 

As touching the outer world, as one is directed to avoid going 
into towns, so one should avoid visiting foreign places and 'not 
learn a language spoken by barbarians ' (Vas., vi, 41 ; Ap., i, 32, 18). 

x] Relative Ages of the Sutras 249 

In religion, as was to be expected, denying the authority of the 
Vedas, carping at the teaching of the Vedic seers, and wavering in 
regard to any traditional duty is to 'destroy one's soul' (Vas., xii, 
41), and there is no salvation for a man who devotes himself to 
epicurean ways or to captivating men or to philology {'^abda^dstra, 
Vas., X, 20). On the other hand the Upanishad doctrine that a 
priest who is learned and austere and repeats the sacred texts is 
not tainted with sin, though he constantly commit sinful acts, is 
a morally destructive teaching already legalised (Vas., xxvi, 19). 
The highest named god is Brahma or Prajapati, to whom, after 
the manner of the epic, verses of legal character are assigned. 
Philosophically the Sutras are dominated by the Vedanta Atman- 
theory, which appears to be known as a system to Apastamba, 
whose Sutra seems to have been a work which arose among the 
Andhras of the south-eastern coast, and probably is not older 
than the second century B.C. It recognises, alone among Sutras, 
a named Purana (the Bhavishya, 11, 24, 6) and its archaic 
effect linguistically, which in large measure determined Biihler in 
his conjecture that this Sutra might revert to the fifth century, 
may well be due to the fact that the Andhras retained linguistic 
peculiarities long after Panini fixed the northern usage. Apastamba 
knows the Atharvaveda, as does Vasishtha, who appears to have 
been a still later writer. It is true that Biihler arranged a 
chronological series of Siitras of the law in the order Gautama, 
Baudhayana, Vasishtha, and Apastamba ; but in doing so he mini- 
mised the late characteristics of Vasishtha (who alone mentions 
* documents ' as legal proofs) ; and in his remark {S. B. E., xiv, 
p. xvii) concerning the fourth Veda he appears to have over- 
looked the passage at vi, 4, where the four Vedas are mentioned. 
It is also quite probable that the passage which seems to make 
Baudhayana earlier than Vasishtha is interpolated, and Biihler 
himself admits that many other passages have been tampered 
with. Whatever the earlier text may have been, the present 
text, with its free use of (^loJca verse, its recognition of Dharma 
^astras, its citations from Manu, Vishnu, etc., and its possible 
allusion to the Romans {Romaka, xviii, 4), seems to be the latest 
of the legal Sutras, though containing much older material. In 
general, the age to which the Siitras may be assigned cannot well 
be earlier than the seventh or later than the second century b.c. 
They represent both the views of different Vedic schools and 
different localities, from the Andhra country in the S.E. to the 
countries of the KW., where probably the school of Vasishtha is to 

250 Life and Customs in the Sutras [CH. x 

be sought ^ Probably the Grihyas represent the earlier Sutras; 
the Dharmas as a whole come later : perhaps 300 b.c. would 
represent the earliest. 

^ For Biihler's views regarding Apastamba, as dating from the third to the fifth 
century b.c, see S.B.E. vol. xiv, p. xlii. The strongest proof that Apastamba was 
a Southerner lies in ii, 17, 17, where he says that ' Northerners ' pour water into 
a priest's hand at funeral feasts. That he followed Baudhayana is undoubted ; but for 
historical use it must also be remembered that only the first two of the four books of 
Baudhayana are genuine and the latter half may be much later. 


The Sutra literature does not lack connexion with the epics, 
to which we now turn. In the Grihya Sutra of ^ankhayana, for 
example, occur the names of Sumantu, Jaimini, Vai9ampayana, 
and Paila, who are teachers of the great epic Mahabharata ; and 
the list of revered teachers, and no less revered species of literature, 
mentioned in the Siitra of A^valayana includes the Bharata and 
Mahabharata, while the (^ambhavya Sutra also mentions the 
Mahabharata (it omits Bharata, perhaps as included in the greater 
name). Although the words are assumed by modern scholars to be 
interpolated, the reason given, ' because otherwise it would make 
the Siitra too late^,' has never been very cogent, since the end of 
the Sutras and beginning of the epics probably belong to about 
the same time. As an indefinite allusion not to a special epic 
poem but to the kind of poetry are also to be noticed such early 
references as that of A9valayana (iii, 3, 1) to Gathas, hero-lauds, 
tales, and ancient legends. 

Epic poetry is divided by the Hindus themselves into two 
genera, one called 'tales and legends' (Itihasa and Purana) and 
the other called 'art-poem' or simply 'poem' (Kavya, the pro- 
duction of a Kavi or finished poet) ; but the compilation named 
Mahabharata is both Itihasa-Purana, its original designation, 
and then Kavya, though it is not recognised as a Kavya till the 
introductory verses exalt it as such. In its origin it was un- 
doubtedly a popular story of the glorified historical character 
which attaches to tribal lays even to-day. The second epic, the 
Ramayana, has always stood as the type and origin of the refined 
one-author poem, and whatever may have been the date of its 
germ as a story, as an art-product it is later than the Mahabharata. 

Thus the oldest references which may indicate epic poetry 
point rather to the story of the Bharatas than to the story of 
Rama. These references, however, in any event are not nearly 

1 Weber, Ind. Lit., p. 63=Eng. trans., p, 58. 

252 Princes and Peoples of the Epic Poems [ch. 

old enough to warrant the assumption of immense antiquity made 
by the native tradition. The language of both epics is not Vedic 
but a popular fonn of Sanskrit, which was developed by the bards 
and became the recognised language of narrative poetry ; and 
their metre is the final reproduction of Vedic metres in modem 
form. Both language and verse are not widely difierent from 
those of the latest Sutras. We may reasonably conclude, then, 
that the latest Siitras and the epics belong to the same period, 
and that they represent two contemporary styles of literature, 
the former priestly and the latter secular. 

There can be no doubt that, so far as much of their subject- 
matter is concerned, the epics and the Puranas are the literary 
descendants of the stories and legends (Itihasas and Puranas) 
which are mentioned in literature from the time of the Atharva- 
veda onwards ; and the particular legend or historical tale (the two 
are confused) which is embedded in the Mahabharata or 'great 
epic of the Bharatas ' is also not wholly without scholastic 
aflBnities. Just as the Brahmanas held the kernel of the Grihya 
Siitras, so the great epic through its promulgator, as traditionally 
recorded, is connected with the school of the Wliite Yajurveda. 
Para9ara is a name especially common in this Veda, occurring 
often in its genealogical lists ; and the epic acknowledges the 
^atapatha as the greatest of Brahmanas, while the heroes of the 
epic are particularly mentioned in the Brahmana, and indeed in 
such a way that Janamejaya, prominent in the epic, is treated as 
a recent personage by the authors of the latter part of the Brahmana, 
though the epic treats him as a descendant of the chief epic hero. 
The explanation of this is not such a mystery as it seemed to Weber, 
who was unable to reconcile the facts that the same person was 
the descendant of the later family and yet appeared as an 
immediate predecessor or contemporary of the earlier. The ex- 
planation is simply that at the time of the eleventh Kanda of the 
(^atapatha Brahmana, Janamejaya to the priestly author was an 
historical character, while to the epic poet he was legendary, 
and the poet himself was, if not a bard, a domestic chaplain 
probably incompetent to analyse history, but anxious to give his 
tale a noble frame. 

Other early allusions to epic characters only show that the 
epic which we now possess was unknown. Vai9ampayana and 
Vyasa are mentioned as early as the Taittiriya Aranyaka, but not 
as authors or editors of the epic which is now their chief claim to 
recognition. The word mahabharata is used by Panini, but only 


xi] Kurus and Pandus 253 

as an adjective which might be applied to anything great con- 
nected with the Bharatas, a hero or town, as well as a war or a 
poem. But above all, the Mahabharata epic is at bottom the 
story of a feud between Kurus and Pandus, and the Pandus are 
unknown to the early literature, either Brahmanas or Sutras. 
The idea that the original epic was a poem commemorating a war 
between Paiichalas and Kurus, which was ably developed by 
Lassen {Ind. AlterthiimskuTide, i, pp. 692 f.), and adopted with 
modifications by Weber {Ind. Literaturgesch. pp. 126 and 203 
= Eng. trans., pp. 114 and 186), is an ingenious attempt to account 
for what is assumed to have existed. As a matter of fact a 
Mahabharata without Pandus is like an Iliad without Achilles 
and Agamemnon ; we know of no such poem. The Kurus and 
Paiichalas are foes in the epic but only as the Pandus ally them- 
selves with the latter. The Kurus of the epic, however, are 
doubtless the Kurus celebrated in ancient times ; even the family 
records show that the epic reflects the glory of these old aristo- 
crats. Thus the names Amba and Ambika as wives of a Kuru 
in the (j^atapatha Brahmana are preserved in the name Amba 
(Ambika) as mother of the king of Kurus in the epic. The first 
occurrence of the name Pandu which can be dated seems to be in 
a vdrtika or supplementary rule to Panini iv, 1, 44, attributed to 
Katyayana (c. 180 B.C.). The Pandus, whatever may have been 
their antiquity, first come into view with the later Buddhist 
literature, which recognises the Panda vas as a mountain clan, and 
possibly in the myth mentioned by Greek writers in regard to a 
Hindu Heracles and his wife Pandaia, though the latter is indeed 
of little weight The epic Pandus are not a people but a family. 

It is not till the second century b.c. that we find unmistakable 
allusion to what we may probably call our epic poem, in the account 
of the Mahabhashya, which alludes to a poetic treatment of the epic 
story and speaks of epic characters. The second century b.c. is 
also the period to which those portions belong in which the 
foreign invaders of the Punjab — Yavanas, (^akas, and Pahlavas — 
are mentioned {v. sup. p. 224). These foreigners are represented 
as fighting on the side of the Kurus. As for the Paiichalas 
being opponents of the Aryan Kurus, the (^^atapatha Brahmana 
represents them as allies, and in early literature they are frequently 
mentioned as forming one people, the Kuru-PaQchalas. A single 
reference in a formula may, indeed, imply disdain of the Paiichalas 
on the part of the Kurus ^, but it is not certain that any racial 

^ Weber, hid. Lit., p. 126 = Bng. trans., p. 114. V. sup. Chapter v, p. 119. 

2 54 P^i^ces and Peoples of the Epic Poems [ch. 

antagonism existed between the two. We may say with Weber ^ 
that * the epic commemorates a fight between Aryans in Hindustan 
after the time when the original inhabitants had been overthrown 
and Brahmanised/ only on the assumption that Kurus, Pafichalas, 
and Pandus were Aryans ; but this is doubtful, and the force of 
the remark is in any case somewhat impaired by the fact that 
contests between Aryans. are no indication of late date, since such 
contests are commemorated even in the Rigveda, 

It is possible that the Pafichalas represent five Naga clans 
(with ala ' a water-snake ' cf. Eng. eel) connected with the Kurus 
or Krivis (meaning 'serpent' or 'Naga'), and that none of the 
families is of pure Aryan blood, for the Nagas in the epic are 
closely related to the Pandus ; but all such considerations at 
present rest on speculation rather than fact. 

Whether we are to suppose that, anterior to our extant epic, 
there was a body of literature which had epic characteristics, must 
depend also largely on speculation regarding the few well-known 
facts in the case. These are briefly as follows. At certain 
ceremonies, not chiefly heroic, Gathas, 'strophes,' in honour of 
great men are sung with the lute as accompaniment. These 
verses apply to men of the past or present, that is, they are 
laudatory verses of a memorial character^. Further, the Grihya 
Sutras recognise Nara9amsis, a sort of K\ia dvZpSiv, 'hero-lauds,' 
as a literary genre. These may have served as nuclei for the 
stories of heroes preserved in epic form. In the epic itself 
genealogy forms an important sub-division, and such a genealogy 
includes the origin of gods as well as of men. Now the Brahmanas 
also know what they call the Devajana-vidya, 'knowledge of the 
gods' race ' ; and since the epic genealogy of gods is in many ways 
indicative of respectable antiquity, it is possible that it derives 
from such a vidyd or science. The stories told in the Brahmanas, 
like that of Hari^chandra in the Aitareya Brahmana, often have 
epic fulness and likeness, being composed in the later epic verse 
though in ruder metre. In these also we get a form of narrative 
told in verse which might presumably have evolved into epic 
form. A great deal of the inflated epic is didactic, and much of 
this is derived from didactic sources older than the present epic. 
Thus dramatic tale, genealogy, and instruction in pedagogic form 
have all aided in the making of the epic. Even the theology of 
the epic has its prototype in the Brahmanas, where Vishnu is 

^ Op. cit., p. 204= Eng. trans., p. 187. 
'' Compare Indische Studien, vol. i, p. 187. 

xi] Early Heroic Poetry 255 

already the ' best ' or most fortunate god {<^rtsMha\ and ^iva is 
already called Mahadeva. 

In the hymns of the Rigveda we find stories in verse which 
appear to need the complement of explanatory prose, and as the 
epic also has examples of this mingling of verse and prose in the 
telling of a story, it is possible that we may have the right to 
presuppose a sort of epic narrative even in the time of the Rig- 
veda. Yet this presumptive epic of the Rigveda is so entirely a 
matter of theory, and not undisputed theory, that it may be left 
out of consideration when discussing the historical epic, as the 
presumptive drama of the Rigveda may be ignored in discussing 
the origin of Hindu historical drama. 

The element in ancient literature which seems at first most 
likely to have contributed to the rise of epic poetry is that already 
mentioned under the name of NaracamsT or 'hero-lauds,' withal 
not so much on account of the subject-matter as on account of 
the circumstances in which the lauds were sometimes sung. At 
the yearlong celebration preparatory to the horse-sacrifice ten 
days were devoted to a series of lauds of gods and heroes, whereby 
the nobility and great deeds of kings were sung by priest and 
warrior musicians in Gathas of an extemporaneous character, 
while the recitation of legends in verse accompanied various 
events of life\ 

Now there are certain scenes in the great epic which lend 
themselves especially to such an interpretation. One can well 
believe, for example, that the story of Amba, who was carried ofi' 
by Bhishma irom her home and given to Qalya (v, 173 f.), was best 
rendered as a thrilling lay; its intensity is almost equal to that 
of the gambling-scene^ (11, 60 f ). But there are many others 
not suited for anything save recitation, not to speak of the in- 
terminable didactic material loaded upon the epic by the bookful. 
How are we to reconcile this mass with a theory of lyric recitation 
or song ? 

A study of the interpolations in the so-called Southern text 
shows that thousands of verses of narrative and didactic material 
have been added to the epic text, and that the redaction comprises 
a shameless incorporation of material drawn from the Puranas and 
from the Harivam9a, a sort of Purana which was added to the Maha- 
bharata, as well as elaborations of the original text, sometimes by 

^ Cf . especially Weber's article in the Proceedings of the Berlin Academy for 1891 
(Episches im vedischen Ritual) and that of Lviders in Z.D.M.G., vol. Lvm, p. 707 f. 

- Cf. also the half-forgotten tale of Vidula, revivified to-day by Professor Jacob! 
{Ueber ein verlorenes Heldengedicht der Sindhu-Sauvlra) in the Album Kern (1903), p. 53. 

256 Princes and Peoples of the Epic Poems [ch. 

the insertion of a dozen or so verses, sometimes by the addition 
to a chapter of half a dozen new chapters narrating feats of the 
heroes or insisting on the godliness of a demi-god Now there 
is no reason not to suppose that the same process has made 
the Mahabharata what it is from the beginning. It contains at 
present a hundred thousand verses, with some prose admixture, 
but internal evidence shows that this is an accumulation ; and 
the text itself admits that it was originally less than nine thousand 
verses in length. As we have seen above (p. 251) the Grihya 
Sutra of A9valayana mentions both a Bharata and a Mahabharata, 
no doubt a shorter and a longer version of the same poem. The 
theme of the epic as a story, the conflict between Kurus and 
Pandus, is at most not so long, about twenty thousand verses, as 
the whole Ramayana, or twenty-four thousand verses. In short, 
in the great epic of India we have a combination of matter, partly 
epical, partly pedagogic, partly narrative or historical. The 
genealogies and the religious-didactic parts are not necessarily 
later in date, but they are later additions to the original materiaL 
Some of the additions may be as old as the original or even older, 
but this does not entitle us to maintain that the epic was originally 
didactic, nor is this the best explanation of the heterogeneous 
mass which we call the epic, and which in its present form 
resembles such a combination as, barring dialectal difi*erences, 
might be effected by combining a few books of the Iliad with 
Hesiod, exti^acts from Euripides, Theocritus, Aristotle, and a few 
chapters of the New Testament. With this exception, most of 
the didactic material is not for the everyday man, but distinctly 
for the military caste. Even the philosophy is not for the 
philosopher, the priest, but for the king and his nobles. The 
predominative religion, too, is that of the kingly caste ^. Indra is 
their sovereign Lord : and the heaven of Indra, with his celestial 
nymphs, the Apsarasas, is the reward for kingly duty faithfully 
performed on earth. The lower castes, Vai9ya and ^udra, the 
agriculturist, the trader, the slave, are scarcely recognised except 
adventitiously, as it becomes convenient to refer to them. The 
epic is thoroughly aristocratic, a work completed by priests for 
warriors, to recount the deeds of warriors and show them the 
need of priests, who convert to orthodoxy the service of popular 
gods dear to the local aristocracy. The epic has thus become 
what it calls itself, the 'fifth Veda,' and may be regarded either 
as a didactic storehouse (it calls itself a Dharma ^'astra) or as a 

* Cf. Eapsou, Ancient India, p. 72. 

xi] Characters of the Mahabharata 257 

magnified Itihasa-Purana, which even before the epic existed was 
regarded as supplementing the Vedas. Both elements are miited, 
religious-didactic and legendary, in such parts as treat of the 
demons, gods, and seers of old. How ancient may have been 
collections of such material prior to our extant epic is uncertain ; 
but the evidence for earlier collective works does not appear to 
be convincing. That a mass of legends existed and that this mass 
was used by Brahmans and Buddhists alike as they needed them 
may be granted, just as the mass of fables known to the ancient 
world was utilised by the epic writers and by those who composed 
the Buddhist Jatakas, though India had no Aesop. 

Many of the characters of the Mahabharata appear to be real, 
historical figures. Others are mythical, in that they represent a 
personality evolved from a divine name or a local hero-god. Thus 
the name Arjuna is first a title of Indra, whose son the epic 
Arjuna is ; but his cousin Krishna is a local demi-god hero, and 
there is no reason to doubt the historical character of the king 
of Magadha who was a foe of this pair and a ^ivaite, though what 
is said about him in the epic may be merely the exaggeration of 
legend, as sung by the bards who made expeditions with the army 
and sang the exploits they themselves had seenK The stories of 
historical characters, like king Janaka, also reflect history through 
the mists of legend. The complete anthropomorphisation of 
heavenly beings, which some scholars are reluctant to admit as 
a possible phenomenon in the best of cases'^, is found in the 
Hindu epic, especially in the inserted tales of the gods ; but it 
does not appear at all certain that any epic hero represents a 
heavenly being in either of the Hindu epics. Krishna in the 
Mahabharata and Rama in the Ramayana are forms of the sun- 
god only as being identified with Vishnu as All-god ; and in the 
case of the Ramayana tliis is a palpably late procedure, while it 
is doubtful whether Krishna was ever a form of the sun. Both 
Rama and Krishna appear to have been tribal heroes, mythical 
perhaps but not products of divine mythology. But, as no attempt 
has ever been made to separate myth from history in India, it 
is impossible to say whether Krishna, the divine hero of the 

* The Sutas or bards were also charioteers. They made a special sub-caste and 
lived at court, while the KuQilavas learned the songs of the bards and wandered among 
the people at large singing them. This name was resolved into Kuga and Lava who are 
represented as two singers, sons of Rama. They learned the poem of Valmiki and 
recited it among the people, as the later story goes {Ramayana, i, 4). The Magadha 
king Jarasandha was the ruler of the East, as the Pan4us were his rivals in the West. 

2 Chadwick, The Heroic Age, p. 265. 

C.H.I. I. 17 

258 Princes and Peoples of the Epic Poems [ch. 

Mahabharata, ever really existed, though this is probable. Krishna 
served as the charioteer of Arjuna, the chief Pandu and epic 
hero ; and though he promised not to fight in person he did all 
he could to keep up and intensify the enniity between the Pandus 
and their related foes, the Kurus, not avoiding even tricks opposed 
to knightly honour. It is not likely that such shameful acts as 
those recorded of him by his own followers would have been 
invented of a god, but rather that the tricks belonged to him as 
a hero, and that no amount of excuse, of which there is enough 
ofiered, could do away with the crude facts of tradition, which 
represented the man-god Krishna as a clever but unscrupulous 
fighter. A later age exonerated him by ofifering various excuses, 
the higher morality of imperative need^, the tit-for-tat rule (one 
sin to offset another), etc., just as it offered various explanatory 
excuses for the polyandry of the Pandus, who, however, as a 
northern hill-tribe or family, probably were really polyandrous 
and needed no excuse^. 

Although the epic age in India must necessarily be an epoch 
too elastic for historical purposes, since it is not at all certain 
that any one epic statement may not be many years later than 
another, yet the effect of this now trite observation is to exaggerate 
the relation between isolated cases and the epic mass. It is true 
that we have additions to the greater epic which are hundreds of 
years later than the mass, but it is possible from the mass to get 
an impression which will represent conditions on the whole, and 
we are tolerably sure that this whole is bounded by the space of 
from three to four centuries, since external evidence, inscriptions, 
the Greek reference to the Indian Homer^, etc., prove that the 
great epic in nearly its present extent existed before the fourth 
century A.D., and negative evidence in India makes it improbable 
that any epic existed earlier than the fourth century rc. Since 
the length of the work requires the assumption of several centuries 
for its completion as it now exists, the centuries immediately 
preceding our era seem to be those to which it is most reasonable 
on general grounds to assign the composition of the Mahabharata 

1 Thus Krishna is made to say, * If I had not done this (unknightly deed) our side 
would have heen beaten,' and this is accepted as an excuse ; but an excuse was 

2 Polyandry is not denounced in the Sutras ; but this is no proof that the Pandas 
lived before they were composed. The custom is found among the hill-tribes and also 
sporadically on the plains. Strictly speaking, epic polyandry is the marriage of one 
woman to a family of brothers. 

s Chrysostom, a.d. 347-407 ; see Ind. SVud., n, pp. 161 f. 

xi] The Mahahharata and the Ramayana 259 

as a whole. This agi-ees best also with the external data to which 
reference has been made in the preceding chapter. During these 
centuries we find a revival of Brahmanism, a cult of Vishnuism 
by the masses and a return to Brahmanism in a modified form 
indicated by the ^ivaite faith of the kings of the north-western 
part of the country. Now Vishnuism is the cult that permeates the 
great epic, though it contains tales showing an older Brahmanism, 
and the (^ivaite portions are chiefly late in character. Again it is 
not unreasonable to assume a certain connexion between the two 
epics. We cannot think of them as isolated productions of the 
western and eastern parts of the country. That they represent in 
general a western and eastern cycle of epic material is true, but 
there are sundry considerations which make it impossible to 
believe that they arose independently. In the first place, while 
the metre of the Mahabharata represents a less polished verse 
than that of the Ramayana, that metre is so nearly that of the 
Ramayana, especially in its later portions, that the two are 
practically the same. Secondly, there are many tales, genealogies, 
fables, etc., which are identical in the two epics. Thirdly, the 
phraseology of the two epics is so cast in one mould that hundreds 
of verse-tags, phrases, similes, etc., are verbally the same. These 
correspond to the iterata found in Homeric verse, and indicate as 
do the Grecian parallels that there was a certain common epic 
body of phrase and fable. Fourthly, the economic conditions and 
social usages as represented in the two epics are sufficiently alike 
for us to be able to draw on both together for a picture of the 
times showing few discordant elements. In detail, the references 
in the Ramayana betray a later or more advanced stage in some 
particulars, such as architectural elaboration, plans of temples, 
etc., which may be due to a higher civilisation ; but in general 
the life of priest, noble, people of the lower castes, slaves, etc., 
is the same in both epics, and except for the use of caste-names 
does not differ from that exhibited by Buddhistic works of the 
same period. The chief difference here is that the Buddhists speak 
more of householder and gildman as if they were separate orders. 
But the Gehapati or householder is also a common expression for 
the ordinary man of affairs in Sanskrit works, and the gilds as 
shown above in discussing the Siitras (p. 247) have their importance 
admitted by the authors of the Sutras and epics alike. It is 
therefore more a question of terminology than a vital distinction 
when we find that the social order is reckoned as composed of 
priest, warrior, householder, gildman, instead of priest, warrior, 


26o Princes and Peoples of the Epic Poems [ch. 

and 'people's man,' Vai9ya, as the Brahman priests divided the 
* regenerate ' members of the community \ 

The main diflference in the presentation of social data given 
by the Brahman and the Buddhist is the one already referred to. 
The Buddhist does not accept the spiritual authority of the 
Brahman and belittles him as a caste-member ; but he cannot 
rid himself of inherited faith and phrase, and so constantly 
recognises him as member of a caste or order like that of the 
monks. On the other hand, the Buddhist state was a democracy 
in spirit ; the teaching of the church (to use the word) was apt 
to exalt the humble and lower the aristocracy. The emperor 
himself was humbled by himself, and his nobles became subject 
to the religious law of love and kindness, while any common 
person was magnified for piety and could obtain high office in the 
council chamber. This was not only theoretically true ; it affected 
the whole constitution of the State. The merchants and farmers 
and the mass of working people were endowed with a new in- 
fluence, which superseded for a short time the influence of priest 
and noble. It is sometimes said that this was no supersession ; 
that Buddhism arose before the four orders were recognised as 
state constituents, and that in the freer use of householder and 
merchant (such was really the Setthi or gildman) we have the 
expression of a freer life not yet bound in four-caste orders. It is 
probable that at all times the third * caste ' was an elastic term for 
every Aryan not priest or warrior ; but it connoted pure blood 
and hence excluded those 'mixed castes' which were sometimes 
higher, but more often lower, than the house-slave. A great mass 
of these people were the hill-tribes reduced to servitude or to low 
pursuits, such as leather-workers, fowlers, etc., all those useful but 
dirty and disagreeable people whom the Brahman despised and 
the Buddhist afiected to love and honour. But the consideration 
shown to the low orders and the dignity attained by the merchants 
under a king who had no use for war are no proof that these traits 
were antecedent to an acknowledgment of the aristocratic classes. 
In fact, in the same district in which Buddhism arose and where 
the Buddhist emperors reigned, some at least of the Upanishads and 

1 For the nomenclature of the Buddhists, cf. Fick, Die sociale Gliederung in 
Twrdostlichen Indien zu Buddha's Zeit (1897), pp. 19 f. and 162 f. Cf. also Senart, 
Let Caites dans VInde, where the contention is upheld that castes (so-called) are really 
social orders. Pick's expression Zu Buddha's Zeit is used with the freedom which 
characterises almost all Buddhist scholars when writing of Buddhist literature. He 
means no more by it than early Buddhist literature, and under that head are included 
the Jatakas which, in their present form, are centuries later than Buddha's time. 

xi] The Social System 261 

Brahmanas were composed, and these pre-Buddhist works all 
acknowledge as a matter of course the high rank of the two upper 
castes and the vulgarity of the lower, who exist, especially the 
farmers, ' to be eaten ' by the king. The Buddhist attitude then 
is not an archaic attitude or one subsequently followed by the 
evolution of a theory of * four castes,' but is due to a revolutionary 
insistence on virtue and use as tests of nobility. It is clear from 
both epics that the attitude toward the lower castes was not 
dissimilar to that held by every aristocracy toward the useful but 
undesirable proletariat. Both epics are from the beginning court- 
epics, to be recited before nobles and kings and priests at the 
great sacrifice which designated a supreme ruler, as the earlier 
texts indicate ; but, as the epics themselves intimate, to be recited 
first at court and then popularised and recited among the people. 
The description of a recitation of the Mahabharata given in the 
work itself implies, however, that this was not such a popular 
recitation as occurs to-day (for the great epic of India is still 
recited dramatically to village throngs), but one conducted in the 
house of a gentleman of leisure for his private entertainment. 

Before discussing the conditions found in the epics it will be 
necessary to mention adversely two hypotheses in regard to the 
time in which the great epic was composed. Both are exaggera- 
tions, based partly on neglect of pertinent data, of views already 
considered. The first of these is the theory that the Mahabharata 
is a product of our middle ages, that is, that it was a late output 
of the renascence. The discovery of inscriptions showing that 
the epic was essentially the same as it is now centuries before the 
middle ages of course disproves this ill-considered theory, but the 
great work in which it is elaborated will always remain a mine of 
useful information \ On the other hand, the theory that the 
Mahabharata is a work of the fifth or sixth century before Christ 
and the product of one author who composed it as a law-book^, is 
a caricature of a fruitful idea of the late Professor Blihler. As it 
violates every known principle of historical criticism it may be 
passed over without discussion. The epic was composed not by 
one person nor even by one generation, but by several ; it is 
primarily the story of an historic incident told by the glorifier of 
kings, the domestic priest and the bard, who are often one ^ 

^ Adolf Holtzmann, Das Mahabharata und seine Telle (1892-95). 

^ J. Dahlmann, Das Mahabharata als Epos und Rechtsbuch (1895) ; and Genesis 
des Mahabharata (1899). 

' That besides the professional bards the domestic priests were eulogisers of the 
king may be remarked from the epic tale of the king's daughter who reproaches the 

262 Princes and Peoples of the Epic Poems [ch. 

The germ of the Mahabharata is the description of the over- 
throw of the Kurus, a Bharata clan, at the hands of the Pandus. 
A thinly veiled genealogy represents the Pandus as cousins of the 
Kurus. In reality, they were a new family or clan, who built up 
a kingdom and then obtained supreme power by allying them- 
selves with the Panchalas and attacking the Kurus, who are 
represented as living about sixty miles north of the Pandus' 
settlement, which was the present Indarpat (Indraprastha), near 

The 'cousins' called Pandus first excited the jealousy of the 
Kurus when the latter were obliged to come south and oflfer 
tokens of submission to the Pandu king, who had crowned himself 
as emperor and performed the horse-sacrifice establishing this 
title. Resorting to trickery, the Kurus invited the Pandus to 
make them a visit. The somewhat uncouth Pandus, who are 
described as good examples of nouveaux riches, flaunting in the 
eyes of their guests all the evidence of their wealth and making 
the lowly but aristocratic Kurus objects of ridicule^, despite their 
sudden rise to power were not yet adepts in courtly arts, and the 
chief art for a knightly gentleman of that day was gambling. As 
the Pandu king says, no gentleman (warrior) can refuse to fight 
or gamble when challenged. The Kurus were an old house and 
had the skill of the court at their command, however poor they 
might be in worldly goods. The Kuru prince, who had been 
humiliated, concocted a scheme to overthrow the Pandus by 
gambling. The old king, his father, was a noble at heart as well 
as by blood and made what protest he could against this scheme, 
which he knew implied cheating at dice. But he was old and 
blind ; and it was not the custom to pay any regard to what a 
man said after he grew old. When any man's hair grew grey he 
was expected to abdicate his power in favour of his son and retire 
from active life. What regard was paid to him thereafter was 
a matter of courtesy. He usually made over his property to his 
sons ?ind disappeared literally or to all intent, becoming a wood- 
dweller. If such was the fate of the ordinary old man, the fate 

daughter of the domestic priest : ' I am the daughter of a kiqg, who is lauded ; thou 
art only the daughter of the laudator.' The first priests who handled the epics were 
of this sort, domestic priests, royal chaplains, indifferently well read in theology and 
philosophy but conversant with the rites of the Atharvaveda, which as a popular work 
of its day is associated with the earlier form of epic {Chhandogya Upanishad, m, 4). 

^ The Kuru prince complains that mirrors were so set in the floor of the Pandas' 
palace that he was made to think them ponds, etc. Every effort was made to humiliate 
the Kurus. 

xi] The Story of the Mahabharata 263 

of kings was worse, as there was more to gain by their suppression. 
No regard at all was paid to the old king, who was king only in 
name. The Pandus were challenged to a friendly game of dice to 
be played in the Kurus' city. It may be remarked here that the 
old site of the Kurus at the famous Kuru Plain had evidently been 
given up, as the Kurus were pushed back to Hastinapur, where 
they lived at the time of the epic story. The Pandus vaingloriously 
assented to make this return visit and see their kinsmen in the 
north. On arriving they were courteously received, and after 
spending a night with their hosts proceeded to the gambling-hall, 
where in one throw after another the Kuru prince, playing by 
proxy and thus securing the aid of the best gambler at court, 
won all the wealth, family, and kingdom of the Pandu emperor, 
who, however, ventured to play once more for the stake of banish- 
ment. As the emperor had already played the lives of his brothers 
and wife and lost, this last throw was an effort on the part of the 
Kurus to get them out of the way without imprisonment or other 
disgrace which might have occasioned a rising of other allies of 
the emperor. As it was, the Pandu king gave his word that, if he 
lost the last throw, he would go into banishment for twelve years 
with all his family. After the twelve years were over, he and his 
brothers took refuge with the Matsya clan, and from that vantage- 
point collected other allies, marched to the Kurus' land, were met 
at Kuru Plain, defeated the Kurus, and regained the old power. 
It is noteworthy that in all the twelve years of banishment the 
bitterest note in the lamentations of the Pandus is not the loss of 
the kingdom but the insult to their wife. As related above, they 
were a polyandrous race, and the king and his four brothers were 
husbands of Krishna. When the king had gambled away his 
brothers and himself, he offered to gamble their wife and did so, 
though the proceeding raised the legal question^ whether one who 
had already made himself a slave could gamble away anything, 
slaves possessing nothing. The question being over-ruled, however, 
the wife was dragged off and insulted by the brother of the Kuru 
prince. Now whenever the Pandus, who are fulfilling the pledge 
to remain in banishment, begin to bewail and plan revenge, it is 
the former plight of Krishna DraupadI which evokes most anger. 
Not the cheating at dice, though that is not forgotten, but the 
insult to Krishna, who was dragged into the assembly of men and 
made a slave dishonoured, animates the Pandus in their despair 

^ No legal authority is cited in this scene, however, though the question is argued 
by the old men who sit and look on during the gambling. 

264 Princes and Peoples of the Epic Poems [ch. 

and causes Bhima to vow that he will drink the blood of the Kuru 
prince, a threat which he fulfils on the field thereafter. 

There is, under another form, the violation of the rite of 
hospitality and virtual abduction of Krishna, the same nucleus 
of tragedy here which makes the simple Ramayana appear like an 
echo of the Iliad. In the Ramayana, the heroine is carried off by 
a treacherous fiend, whom Rama pursues and slays after a long 
interval. But the Ramayana difiers essentially from the Mahabha- 
rata not only in its style but in its spirit. Its most spirited scenes 
occur before the epic plot begins. After the introduction, in the 
history of Sita, Rama, and Havana, turgidity replaces tragedy, and 
descriptions of scenery and sentimentality take the place of genuine 
passion. The didactic overload is indeed lacking, and the Ramayana 
gains thereby ; but in this epic the note of savage lust and passion 
which is the charm of the Mahabharata, as it reveals genuine 
feeling of real men, is replaced by the childish laments and pious 
reflections of Rama, whose foes are demoniac spirits, while his 
allies and confidants are apes. It is a polished fantasia, the first 
example of the Kavya or * artificial ' poetry, which appeals to the 
Hindu taste much more than does the rough genuineness of the 
Great Epic. The Ramayana is in truth artificial in both senses, 
for one cannot possibly believe the tale ; whereas the Mahabharata 
makes its tale real and one believes it as one believes that the 
Achaeans overthrew Troy, however embellished the account may be. 
The fact is that the Great Epic is the one human document after 
the appeal of religious sincerity in the primitive hymns of the Veda. 

The reason for this lies not alone in the fact that literature 
after the early Vedic age is chiefly liturgical and didactic, for this 
only shifts the explanation. Sanskrit literature is without power 
of literary expression from the hymns of the Rigveda to the 
Upanishads, and again from this time to that which produced 
the dramatic scenes of the epic, because it was in the hands of 
priests whose whole interest lay apart from real life. The same 
spirit which produced the best Vedic hymns, the spirit reflecting 
independence and freedom, appears in the royal literature, if we 
may so call it, which stamps the age of the Upanishads and of the 
great epic in its earlier parts. The Upanishads are in part the 
product of unpriestly, or at least anti-ritualistic, thought, and the 
epic also emanates from the throne and not from the altar. As the 
Upanishads embody the cultured philosophy of king and noble \ 
so the epic scenes of love and war reflect the life of court and 

* For another view, see Chapter v, p. 144. 

xi] Ear Her and Later Moral Ideals 265 

camp. They breathe a different spirit, as they come from a 
different source than does the literature of the Brahman, until 
indeed the all-grasping hand of the priest seized even the epic 
tales, and, stifling all that was natural in them, converted them 
into sermons, to teach the theology of the priest and impart to 
the king the teaching best calculated to further priestly greed ^. 
The sociological data of the epic period show that society had 
advanced from a period when rude manners were justifiable and 
tricks were considered worthy of a warrior to one when a finer 
morality had begun to temper the crude royal and military spirit. 
This is sufficient explanation of that historical anomaly found in 
the Great Epic, the endeavour on the part of the priestly redactors 
to palliate and excuse the sins of their heroes. Arjuna shoots 
his rival, Karna, while the latter is helpless. But an act like this, 
which was doubtless considered clever at first, became repugnant 
to the later chivalry. Then the demi-god hero Krishna is made 
to be the source of the sin on the simple ground that if divine 
Krishna commands, it is right. Arjuna is now made to shoot 
reluctantly, in obedience to the divine command. But this may 
not be cited as a precedent against the later code, because it was 
a special case in which the act was inspired by God from occult 
motives outside the sphere of human judgment. So with many 
other sins committed by the heroes. They reflect an old barbarity 
later excused. It is not necessary to assume with Holtzmann, 
von Schroeder, and others that the epic tale has been ' set upon 
its head,' that is, that the whole poem was originally in honour 
of the Kurus, and was then rewritten to honour the Pandus, 
and that in this last process the * sins of the Pandus ' reveal the 
original attitude of reproach taken by the Kuru poet^. There 
is a difference morally between the Kurus and Pandus. The 
Pandus offend against the later military code. Thus the Kurus 
reproach the Pandus because their chief warrior interfered in 
a combat between two warriors and killed his friend's foe, who 
was being worsted in the fight. The Pandu simply laughs at the 

^ Thus whole sections of the Ana^asana (the thirteenth book of the Mahabharata) 
are devoted to instilling the moral grandeur of those kings who give land-grants, cows, 
gold, and clothes to the priests. At the same time, much that is didactic is imbedded 
in the poem without this aim. Only the tendency is apparent to extend moral teaching 
to instruction calculated to subserve the ends of cupidity. 

* For detailed criticism of this theory, see the present writer's monograph on the 
Position of the Ruling Caste in Ancient India {J.A.O.S., 1888). The explanation of the 
poem as a myth of nature, Krishna representing earth wed to the five seasons, etc., is 
unnecessary though ingenious. It was proposed by Ludwig in the Transactions of the 
Royal Bavarian Academy (vi Folge, 12 Band). 

266 Princes and Peoples of the Epic Poems [CH. 

reproach. 'Why' (says he) 'of course I killed him. I saw my 
friend worsted, and interfered just in time to save him/ intimating, 
as is clearly stated afterwards, that a conflict on a field of battle 
is not a polite duel (' That is no way to fight '). But the Kurus 
are just as wicked as the Pandus, only they are diplomatic. Their 
sins smack of cultivated wickedness. They get an expert gambler 
to ruin their rival. They secretly seek to burn their enemies alive. 
They form a conspiracy and send out ten men under oath to attack 
Arjuna. Tliey slay Arjuiia's son first, in order to weaken Arjuna's 
heart. In a word, they are cunning and sly ; the Pandus are 
brutal and fierce. Two types of civilisation are embalmed in 
the poem. 

The most striking difference between the knights of the epic 
and the priestly power, which in the end controlled them, is that 
the warrior-caste w^as the royal caste and hence represented state- 
power, a political body, whereas the priests were never more 
than a caste of individuals. They represented no church-power. 
There is thus a fundamental lack of priestly organisation ; there 
is nothing parallel to the Church of Rome in its contests with 
European state-power. Individual priests, without financial re- 
sources but dependent on the local raja for support, could do 
nothing save persuade the raja. But superstition aided them; 
and persuasion aided by superstition became a compelling power, 
which, however, was exerted only for two objects, the exaltation 
of the individual priest or of the priestly caste and the inculcation 
of religious and moral precepts, never for the formation of a 
worldly power within, but independent of, the State. There was 
no caste-head. When strife arose between priests, as it constantly 
arose apropos of a fat office to be enjoyed (the epic furnishes 
examples), each individual priest fought for his own hand ; he 
had no bishop over him ; and there was no pope to oppose a king. 
Thus, while the priestly law-book says that * the priest is the norm 
of the world,' the epic says ' the king is the norm.' The law says 
that a priest has the right of way even over a king; the epic 
narrates that a king meets a priest and calls out to him ' get out 
of my way,' and despite the law, as cited, smites the priest with 
his royal whip. Such scenes show that the king is not yet the 
creature of the priest, but that the epic unconsciously reflects a 
freer life than that depicted as ideal by the later priests, who teach 
that the king is a steward divinely appointed to provide for them. 

Somewhat as in Buddhist literature we must therefore reverse 
the importance of the two 'upper castes,' and regard the epic 

xi] Knights^ Priests^ and Commoners 267 

state as consisting in a military power, whose head is the raja; 
then a priestly power, politically unorganised, but divided into 
schools; then the merchant-power, represented by gilds, whose 
powerful heads {mahdjana) are of political importance ; then the 
farmers, unorganised but tenacious of certain religious rights and 
boasting of Aryan blood. The two last classes form one body 
only because they are neither of them noble (royal) or priestly 
or un-Aryan. No other tie unites them. The merchants in 
general belong to the town, the farmers to the country ; the two 
are the historical divisions, brought about by economic conditions, 
of that order called 'the people,' in distinction from noble and 
priest. This was the Aryan state. Below the Aryan constituents 
were the many who were either remnants of wild tribes or slaves, 
descendants of conquered clans of other blood. They are all 
mentioned in the epic, as well as foreigners or barbarians. 
Although town-life is well known, yet the farmers and cattlemen 
were perhaps more generally typical, on account of their numerical 
superiority, of the order to which each belonged. So it is said : 
' Work is for the slave ; agriculture for the people-caste' (Mbh., xii, 
91, 4), or again ' The work of the Vai9ya is to tend cattle ' ; less 
commonly 'The duty of the priest is to beg for sustenance; of 
the warrior, to defend the people; of the people-caste, to make 
money; of the slave, to work (manually)' (ibid, v, 132, 30). It 
will be observed that the cattle-raising 'people' are ignored in 
favour of traders in the last citation, though 'to make money' may 
imply farmers and cattlemen as well as traders. 

The slave possessed nothing; his tax was paid in manual 
labour, for he had no money or other possessions, 'there is no 
suum in the case of a slave ' (ibid, xii, 60, 37). The slave comes 
' from the foot of God ' (as the warrior is born of God's arm) and 
hence is ' born to servitude.' The (^udras are especially the slaves 
of the merchants and farmers; for though they are told to be 
'faithful to priest and warrior' they are said in particular to 
'serve the people-caste ' (ibid, i, 100, 11). They are also marked as 
the 'blacks' in distinction from the priests who are white ^. The 
military character of the epic precludes much attention to the slaves, 

1 It is doubtful whether the finer distinction here made {Mbh., xii, 188, 5), namely 
that the warrior(-caste) is red and the people-caste yellow, indicates a real racial 
distinction ; especially since there is no other indication that these Aryans are racially 
sub-divided ; whereas the distinction between white and black is an early mark of the 
difference parting the Aryan and un-Aryan and goes with the nasal distinction noticed 
in theVedas between ' good-nose ' and ' no-nose ' people. The epic poets still speak of 
their Aryan heroes' ' fair-noses.' See also Chapter iv, p. 85. 

268 Princes and Peoples of the Epic Poems [ch. 

who as a fighting host are naturally not of importance, though they 
may be referred to under the designation ' the black mass,' for the 
great hosts led into the field comprise many of the slaves as camp 
followers and helpers. What is very important is that the lowest 
Aryan caste, the body of farmers, is on the verge of mingling with 
the slave-caste. No priest may become a slave, however distressed 
for sustenance he may become ; but a slave may become a herdsman 
or trader if he cannot support himself by service (this is the epic and 
legal rule), and in fact the farmer population was largely composed 
of slaves. In the ethical parts of the epic, where caste-distinctions 
are theoretically abolished in favour of the rule that * there is no 
distinction of caste' (religiously), the slave is even allowed to 
study and may get a reward for practising religious exercises 
(Mbh., XII, 328, 49; xiii, 132, 14), and a learned slave gives moral 
instruction ; but this does not seem to correspond to real conditions 
where the slave is reckoned next to the beast (ibid, xiii, 118, 24). 
The old spirit of the Brahman period, which declares that * priest, 
warrior, and people constitute the whole world ' is still practically 
in force. 

The people are settled in small villages around a fort, which 
remains as a grama or ' crowd ' (village) or expands into a town, 
nagara. Small settlements are called ghoshas or pallls, some of 
them 'marches' {prdntas, *on the border'). The distinction 
between these and the places called Tcharvatas and pattanas is not 
clear, though the grama seems to be smaller than the kharvata, 
which in turn is smaller than the nagara. Perhaps village, 
town, city would represent the series. The villages were largely 
autonomous though under the 'overlord' of the king, who ad- 
ministered justice and laid taxes. In all smaller affairs of life, 
' authority rests with the village,' according to law (Par., Grihya 
Sutra, I, 8, 13) and the epic seems to uphold even family custom 
as legally sufficient. Thus as one man says (v. inf. p. 291) that he 
demands a price for his daughter, because that is his 'family- 
custom,' so another defends his occupation of killing animals on 
the same ground. It has always been the custom of Indian rulers 
to leave afikirs as much as possible in the hands of the local 
authorities; and the headman of the village or the group of five 
elders were practically independent, provided the village paid its 
revenue as assessed by the adhipati or overlord. 

The king rules not because of might alone but by virtue of his 
morality. " A wicked king may be deposed ; a king who injures 
his people instead of protecting them should be killed ' like a mad 

xi] Cowboys and Herdsmen 269 

dog.' Taxes there must be, because the people must be defended, 
and this costs; but they must be light, and vary according 
to need. The tax in kind is common. The merchant pays in 
kind and the ranchman pays in kind, but the town-people are 
fined in copper money for offences, though bodily punishment 
takes the place of fines in all cases where there is intent to 
deceive. Thus the shipping-duties paid by * merchants coming 
from afar' are probably in kind {Mhh., 11, 5, 114). Frequent 
allusions to merchants * using false weights ' (cf. i, 64, 21 f.) show 
that a careful supervision of the market-place was necessary. 
The merchant-gilds were of such authority that the king was 
not allowed to establish any laws repugnant to the rules of these 
trade-unions. The heads of gilds are mentioned next after the 
priests as objects of a king's anxious concern ^ 

The large part of the population employed as ranchmen in 
tending cattle has scarcely been alluded to as yet. They were 
perhaps the original 'people,' before agriculture was much 
practised and when merchants were few. At the time of the 
epic they seem to have become partly cattle-raisers and partly 
farmers, while the occupation of ranchman proper had fallen into 
the hands of barbarians who could not understand Aryan speech. 
Yet the one example of which the epic takes note shows that 
these were merely the cowboys who guarded royal cattle {Mhh., iv, 
10, 1). The king is here represented as having a royal picnic on 
the occasion of a * cattle-branding,' when the court goes into the 
country and the 'ears of the cattle are marked' for the year. 
It is on this occasion that the Kurus lift the cattle of the Matsyas. 
Though accounts of such border-raiding in the old Vedic style 
are rare and this passage in particular can by no means claim 
special antiquity, yet it doubtless reflects a not uncommon 
state of affairs^. Very little in regard to these lowly members of 
the State, the cowboys and herdsmen, is to be gleaned from the 
epic ; but one passage states what the low labourer of the 'people- 
caste ' is to earn per cmnum : ' he should receive the milk of one 
cow for the care of six cows ; and if he tend a hundred head he 

1 On the gilds, see Mbh., m, 249, 16; xii, 54, 20; Ram., vi. 111, 13; of. Hopkins, 
India Old and New, p. 169. Their power may be guessed from the fact that the 
didactic epic recommends the king to circumvent them by bribery and dissension since 
• the safe-guard of corporations (gilds) is union. ' 

* Compare the incidental cause of Arjuna's breaking his promise not to visit the 
king his brother while the latter was engaged with their common wife. A robber had 
come and driven off a priest's cow, and the good knight went into the palace to get his 
arms to attack the robber, doubtless an armed band. 

270 Princes and Peoples of the Epic Poems [ch. 

should, at the end of the year, receive a pair. If he acts for the 
master as overseer of flocks or in agricultural labour, he should 
have one-seventh of the proceeds or increase, but, in the case of 
small cattle, a small part (' one-sixteenth ' ; Mhh., xii, 60, 24 f.). 
The six * distresses ' of a farmer do not include excessive taxation, 
but raiding by a foreign king is included among them\ 

The royal soldiery includes not only the nobles of military 
standing supported by the king but the poor members of the 
same Aryan order who with the un- Aryan ' servants ' (not slaves) 
formed the rank and file of the foot-soldiers. In battle they are 
mentioned merely as hosts of nameless archers, slingers, rock- 
throwers, etc., and outside of battle-scenes they are scarcely 
mentioned at all. It is stated that a rathin's, 'car-man's,' 
wage is 'one thousand,' that is, one thousand (coppers) a 
month, and that the king pensions the widows of fallen 
soldiers^. The chief moral laws for members of the military 
caste were hospitality, the sacredness of the refugee, the law 
' not to forget ' a kindness or a hurt, and the rule already referred 
to, that when challenged to fight or gamble it was inglorious to 
refuse. The captured warrior becomes the slave of his captor for 
a year ; if the captor allows him to go free, the captor becomes 
the captured one's Guru or his 'father.' The sign of submission 
is to eat grass (v. sup. p. 247). When the Yavanas were conquered 
(in Brihannar. Pur., viii, 35) they ' ate grass and leaped into water.' 
The epic gives this grass-eating sign as a military rule. As com- 
pared with a member of the 'people-caste,' whose life is valued 
at a hundred head of cattle, the warrior's life is valued at a 
thousand (paid in case of murder). As for the prominent sins of 
the royal military caste, they are mentioned as hunting, drinking, 
gambling, and sensuality withal in a sort of versus memoralis 
which has come down as an apophthegm of law and epic {Mhh., xii, 
59, 60, etc.). Dancing-girls and prostitutes were a part of the 
royal retinue, and hunting was the chief recreation of kings, deer 
and tigers, killed by a king with his sword, being the favourite 
game. Lions were hunted with dogs, as attested by Aelian and 

1 The six distresses {iti) are not defined in either epic ; bat since they are 
mentioned (Mhh., in, 279, 35) and the Puranas define them, it is probable that they 
already include those classified later as too much rain, drought, grasshoppers, mice, 
birds, and neighbouring kings (invasion). 

* The warrior may have three wives, but probably one suflSced in most instances. 
For the pension, compare Mhh. , ii, 5, 54, and for the wage, ib. 61, 20. The wage exactly 
equals the legal ' fine for manslaughter.' The epic copies the law in permitting destitute 
priests to become soldiers, as they may become farmers, but it is considered a disgrace 
for the king to allow priests to depend on such occupations for a livelihood. 

xi] King and Councillors 271 

mentioned in the epic {Mhh., ii, 40, 7). The Buddhist prohibition 
of meat-eating remains as a rule of propriety, but the tales show 
that eating meat was as common as drinking intoxicants and that 
this was the regular court practice, while the story of the crowds 
surrounding a meat-shop (Mbh., iii, 207, lOf.), where the com- 
placent owner boasts that he sells but does not himself kill, shows 
that vegetarianism was by no means universal. 

Passing to a wider point of view we must pause to record the 
fact that certain allusions in the epic to fire-weapons have been 
adduced to prove that the Hindus used gunpowder in the great 
war. How baseless is this supposition has already been demon- 
strated by the present writer, and he can only repeat that all 
mention of fire-weapons in the Hindu epic refers to arms magically 
blazing such as arrows or wheels. No gun or cannon is mentioned 
and gunpowder is unknown \ 

The epic king is no autocrat; he is upbraided and reproved 
by his brothers and ministers. If born to the throne and yet 
defective he is not permitted to become king ('the gods do not 
approve of a defective king,' Mbh., v, 149, 25); but if elected he 
is the leader at home and in the field. He is consecrated by 
baptism with water poured over him from a sacred horn, and is 
crowned 'lord of the earth' (Mbh., xii, 40 and Bam., ii, 69). 
Although the didactic part of the epic emphasises the importance 
of councillors and ministers, without whose sanction the king 
should undertake no important business, yet actually each king 
is represented as doing what seems good to him without advice, 
as the various warriors of the family make raids and rape young 
women from foreign districts without consultation. Indeed, the 
priest supposed to be special adviser is scarcely mentioned in that 
capacity, only as an agent in spiritual matters. Resolving on war 
the kings and allies decide the matter as they will, in the presence 
of priests, indeed, but the priests are ignored (Mbh., v, 1 and 
I, 102). The sabhd or assembly is here simply a military body 
for consultation. Both priests and people are silent in the face 
of force. The king's city was defended by battlemented towers 
and seven moats. It was laid out in squares and the well-watered 
streets were lighted with lamps (Mbh., iii, 284, 3; xv, 5, 16, etc.). 
Only four squares are mentioned in the Ramayana (ii, 48, 19), but 

^ See, in opposition to Oppert, J.A.O.S., 1888, p. 296 f. Since the publication of 
this article Oppert has had published a correspondence with Mr Oscar Guttmann 
{Mitteilungen zur Geschichte der Medizin und Naturwissenschaften, No. 16, iv Band, 
No. 3, 1905), in which he upholds his contention, adopted without question by S. M. 
Mitra in his Anglo-Indian Studies (1913). 

272 Princes and Peoples of the Epic Poems [ch. 

the Mahabharata recommends six. The king's palace included 
or was near to the court of justice, the official gambling-hall, the 
music-room, the place for contests with wild beasts and for exhibits 
of wrestlers. Outside of the inner city were booths for traders, 
etc., and the less pretentious dwellings, with pleasure parks {Mhh.^ 
IV, 22, etc.). Apparently four gates were the usual number, but 
nine are mentioned and even eleven in other literature, and the 
Ramayana gives eight to Lanka (vi, 93). 

For the common members of the military caste to die in bed 
was a disgrace (vi, 17, 11 and often). The mass of the soldiers 
figlit for their chief and when he falls they are disorganised and 
run away. The knights, however, contending for glory as well as 
for their king, remain fighting though the mass desert them. 
Their motto is, ' Sweet it is to die in battle ; the path to heaven 
lies in fighting' (Mbh., viii, 93, 55 f.). In peace the warrior, 
supported by the king, lived at ease and the nobles spent the 
time carousing and enjoying themselves. In war the warrior 
lived and fought for glory as well as for his chief. In the case 
of Karna, who was an independent king, revenge and desire for 
glory are blended ; but most of the epic kings are in the war as 
allies of one side or the other and have no personal motive in 
fighting exoept to win renown. ' A hero lives as long as his fame 
reaches heaven' (Mbh., Iii, 313, 20); 'Glory is preferable to life' 
(ibid. 31). And again, 'Only he who has glory wins heaven' (says 
Karna, ibid, iii, 300, 31). The exhortation to fight valorously 
is based upon the precept that whether slaying or slain one is 
blessed, 'for he who is slain in battle obtains heaven, and if he 
slays he obtains fame ' (ibid, xi, 2, 14). Every hero boasts of his 
great deeds performed and to be performed, even while depre- 
cating boasting as a folly. The heroes boast of their families as 
well as of their prowess^. 

The religious and philosophical views of the epics represent 
every shade of opinion from Vedic theism to philosophical 
pantheism with later forms of Sun-worship (in both epics) and 
sectarian cults of Durga, (^iva, and Krishna- Vishnu in the Maha- 

1 For examples of these and other traits shown by the epic warriors, see the 
specimens collected in the writer's monograph on The Position of the Ruling Caste. 
Interesting parallels may be drawn between the attitude of Homeric and Indie warriors 
in these respects, parallels which may now be complemented by those between Greek 
and Teutonic ideals, as shown in Chadwick's Heroic Age (pp. 325 f.). Prof. Ghadwick 
compares the Anglo-Saxon dom with the Greek Kkia dvdpQv, and the same may be said 
of the kirti and ya^s of the Hindu, as the personal combat of king with king, which 
is the leading characteristic of Hindu epic fighting, may be compared with the style of 
fighting in Homeric and Teutonic poetry (ibid. p. 339). 


xi] Religion and Philosophy 273 

bharata, and Rama-Vishnuism superimposed upon the cult of 
Rama as a hero demi-god in the Ramayana The religion assumed 
as orthodox in both epics is that which we call Brahmanical. The 
Vedic gods with Brahma at their head are to be worshipped, 
as a matter of course. In addition comes the constantly growing 
tendency to exalt the chieftain demi-god from his position as 
clan-hero god to a higher power, till he is identified with Vishnu, 
the popular god of many clans. The cult of Vishnu in this form 
comes under the hands of philosophers, who we may be sure had 
nothing to do with the original epic; and as god he is then 
interpreted according to the philosophical systems of the Sankhya 
and Vedanta, wliich are united with the aid of the Yoga system. 
Of late years it has become usual for scholars to follow the lead 
of Professor Garbe, who has interpreted the chief philosophical 
tract of the Mahabharata, the famous Bhagavadgita, as a rewritten 
Sankhya document of theistic tendency manipulated to serve the 
ends of Vedanta schoolmen. By excluding all the verses which 
teach the Vedanta doctrine, Garbe is naturally enabled to show 
a document which is not Vedantic ; and it may be admitted that 
such a process makes a clearer and more attractive theological 
tract. But the historical efiect produced is fallacious. Exactly 
the same mixture of Sankhya and Vedanta permeates the teaching 
of the philosophical epic in many other passages ; and unless one 
is willing to apply the same process and excise all objectionable 
matter in favour of a theory of Sankhya priority in the philo- 
sophical disquisitions of Qanti or ' quietism,' one has no right to 
dissect the Bhagavadgita into its supposititious prius and ' later 
additions.' The epic philosopher is never a Sankhyan ; ho is a 
Sankhya- Yogist, and it is this connecting link of the Yoga which 
to his mind makes it possible to unite two radically different 
systems \ It must at least remain quite doubtful whether the 
philosophical parts of the epic, most of which have no radical 
connexion with the poem, were not originally composed in their 
present form, representing an attempt, on the part of later 
redactors, to weave into the epic a system of philosophy incul- 
cating the belief in a theistic pantheism derived from Sankhyan 
principles improved by the Yoga and then combined with the 

' For a review of these systems as given in the epic, see the writer's Great Epic of 
India (1901). That the Gita was originally theistic throughout can be proved only by 
rejecting stanzas which are otherwise unassailable. Only four passages out of the 
twenty selected to prove the case in Garbe's Bhagavadgita (1905) show any sign of 
interpolation, and of the four only one is a really striking case of breaking the 

C.H.I. 1. 18 

2 74 Pi^i'^c^^ and Peoples of the Epic Poems [ch. 

All-soul principle later called Vedanta. Vishnu and ^iva both 
served the purpose of the philosophical interpretation. Both 
were popular gods who became the One God in turn (sectarian 
differences probably representing geographical distinctions), that 
One God who even in the Upanishads is also the All-god. For 
this reason many passages of the epic are on the philosophical- 
religious level of the Qveta^vatara Upanishad. 

Two notable attempts to extract historical material from the 
epic have been made in the last few years. They enlarge the 
vision of the fighting hosts on the plain of the Kurus both 
geographically and historically and demand careful examination. 
The first is the result of a study of the forces named in the 
epic itself as allies. As already mentioned, the fighting of the 
Ramayana consists in combats between fiends and monkeys, and 
unless the monkeys are interpreted as southern Hindus speaking 
an alien tongue, and for this and other reasons regarded as little 
better than apes by the Aryan leaders, there is no profit in 
endeavouring to guess at their real significance. In the Mahabha- 
rata, which deals with real people, it is different. The human 
hosts marshalled as friend or foe by the Pandus and Kurus may 
be set against each other geographically. There is a certain 
amount of fiend-fighting, and Nagas of unknown habitat are 
mentioned as contestants. There are also some allies of unknown 
geographical provenance. But the chief factors in the great hosts 
can be distributed geographically. For making such a classifi- 
cation it will be convenient to use the Indian term Madhyade9a, 
the Middle Country, to denote 'the whole of the Ganges basin 
from the Punjab as far as the confines of Bihar,' and to arrange 
the various peoples who are said to have taken part in the war in 
relation to this region. The Pandu forces included the king of 
Magadha associated with the Ka^is and Kosalas, the king of 
Panchala, the king of the Matsyas with mountaineers, the king 
of Chedi — all representing peoples in Madhyade9a — with some 
adherents from the north and south, but especially all the Yadus 
of the west The Kurus, on the other hand, had as allies the king 
of Pragjyotisha, the Chinas, and the Kiratas in the north-east; 
the Kambojas, Yavanas, (^akas, Madras, Kaikeyas, Sindhus and 
Sauviras in the north-west ; the Bhojas in the west ; the king of 
Dakshinapatha in the south; the Andhras in the south-east; and 
the kings of Mahishmati and Avanti in Madhyade^a. Therefore, 
since the Yadus of Gujarat came from Mathura, the statement 
holds that ' the division of the contending parties may be broadly 

xi] Interpretation of Historical Data 275 

said to be South Madhyade9a and Panchala against the rest of 
India^' That this is an important conclusion must be admitted. 
But if it follows that the war was one between southern Madhya- 
de9a, united with Panchala, and the rest of India, how far may we 
assert that this represents earlier epic conditions before the 
nations of the Indian sub-continent were all brought into the 
frame of the epic ? Obviously it would not be safe to make too 
much of a list based on factors of doubtful age, but it is perhaps 
safe to assert that the central plan, so to speak, is historical, 
namely the opposition of the less civilised Pandus and the old 
Panchalas to the orthodox Kurus. 

In the opinion of Sir George Grierson we may make a further 
induction and assert that the Brahmanism of the Kurus represents 
a later tide of immigration as compared with the anti-Brahmanism 
of the Panchalas as earlier Aryan immigrants into India. In a 
way, the anti-Brahmanical party may be said to represent the 
warrior-spirit as opposed to the priestly, which was defeated in 
the contest but revenged itself by manipulating the epic to its 
own glory ^. It is, however, doubtful whether the Panchalas were 
earlier immigrants or in early days were regarded as in any way 
anti-Brahmanical. The further contention, that this unorthodox 
warrior-spirit produced the work of the Bhagavatas and that the 
Bhagavadglta emanates from an un-Brahmanical source, is based 
upon the supposition that the Bhagavadglta and its underlying 
system of Saiikhya philosophy is an exponent of the free eastern 
anti-Brahmanical or un-Brahmanical life which produced the 
great heresies of that region. Buddhism and Jainism. One wishes 
that the veiled history of Hindu thought might be traced back 
so clearly, but the data at our disposal do not justify us in so 
summary a method of reconstructing the past. There is no cogent 
evidence to show that a difference of religious belief had anything 
to do with the war, or that any racial antagonism lies behind the 
division of parties, certainly not of parties opposed as primarily 
Panchalas and Kurus. 

Whether the genealogical lists of the epic may impart ti*u8t- 
worthy information is a second question of importance. It has 

^ F. E. Pargiter, The Nations of India at the Battle between the Pdndavas and 
Kauravas {J.B.A.S., 1908, p. 334), gives a complete analysis of the forces. The 
author admits that the ethnological value of the general statement made above is 
diminished by the fact that the nations on either side were not of the same stock ; also 
it must be remembered that kings were not always of the same stock as the people they 
ruled and brought to war. 

2 J.R.A.S., 1908, p. 606. 


276 Princes and Peoples of the Epic Poems [ch. xi 

been answered affirmatively by Mr Pargiter in the second of his 
valuable papers on the epic^ though with due conservatism in 
view of the contradictions in the epic itself. The later lists found 
in the Puranas may be combined with epic data to make a fairly 
consistent chronological table, but there remains much to be 
taken for granted. Although the names of kings are given, the 
length of their reigns must be assumed on some common basis. 
On the probability that the average length of a Hindu reign was 
fifteen years and on the assumption that unimportant kings have 
been omitted once in so often from some of the lists, Mr Pargiter, 
taking the more complete list of the Solar dynasty as his guide, 
finds that a period of fourteen hundred years intervened between 
the first king, * son of Manu ' (Ikshvaku) and the great war ; that 
Rama, the hero of the Ramayana, lived in the fifth century 
before the great war of the Mahabharata ; Bharata in the eighth 
century, etc. The great war itself marks the beginning of the 
present age (Kali Yuga), 'about 1100 b.c.'^ 

1 Ancient Indian Genealogies and Chronology (J.R.A.S., 1910, p. 1). 

^ See more particularly the work of the same author, The Purdna Text of the 
Dynasties of the Kali Age (1913). For the evidence of the Puranas as to the date of the 
war between the Eurus and the Pan^us v. inf., pp. 306-7. 



The law-books, Dharma ^astras, and especially trained experts 
in law, Dharma-pathakas, are recognised in the didactic parts of 
the epic ; and codes of law are assigned to various ancient worthies, 
among whom Manu generally, but not always, holds the chief place. 
The difference between the formal law-book, (^astra, and the Sutra, 
also concerned with Dharma, is due mainly to the gradual exclusion 
of irrelevant matter in the law-book. Whereas in the Sutra the term 
Dharma embraces all domestic duties, religious and ethical, with 
slight attention paid to formal law, in the completed ^astra law 
itself is the sole subject discussed. But this difference marks only 
the extremes, the primitive Dharma Siitra and the law-book of the 
fifth century A.D. Between the two comes a number of works 
bearing the title of law-book but still retaining in large measure 
the characteristics of the Sutra. Likewise the formal distinction 
between a prose Siitra and a metrical law-book is bridged by a 
period when legal works were partly prose and partly verse. In 
the end, it was found more convenient to versify the rules as the 
Hindus versified all knowledge, and the metre chosen for this 
purpose was the later ^oJca, which ousted both prose and the older 
trishtubh metre still used in early Sutras and ^astras. The name 
also is not absolutely fixed. The Siitra is sometimes called ^astra. 
Vishnu's law-book, for example, is both Sutra and ^astra, as well 
as Smriti, a general term for traditional teaching. 

As the Dharma Siitras emanated from Vedic schools, so, though 
less surely, it may be said in general that the law-books at first 
represented certain schools of Brahmanical teaching. The law- 
books of Vishnu and of Yajiiavalkya are thus exponents of 
Yajurveda schools; but in the end the popular works of this 
class lose all connexion with any one school and become uni- 
versally authoritative. There are not many of the long list of later 
law-books which really deserve the name. As time went on, a 

2yS Growth of Law and Tuega I Institutions [ch. 

large number of works appeared, claiming as their authors sages 
of old, or divine beings, but they are all without historical value 
and usually are sectarian tracts inculcating special religious 
observances. Besides these pseudo-law-books may be mentioned 
the later legal works, Dharma Nibandhas, of the eleventh century 
and later, and the leanied commentaries, like the Mitakshara, 
which have become as authoritative as the text itself But these 
later law-books do not come into our present purview. They belong 
to the age of the later Puranas and subsequent literature. The 
great law-books which we have to examine revert to the beginning 
of the Puranic age or before it. Whatever is of value in the later 
works is taken from the older, which are still authoritative. 

By far the most important of these is the law-book of Manu or 
the Manava Dharma (^astra, a work closely connected with the 
law-book of Vishnu, which has no less than 160 verses of Manu, 
and with the didactic chapters of the epic, which contain numerous 
verses found in the code. Moreover, the epic recognises Manu 
as a law-giver and refers to the Dharma Qastra of Manu. The 
relationship between the two works is made doubtful for the reason 
that we do not know when the later parts of the epic embodying 
these allusions may have been composed. An analysis of all the 
passages in the epic referring to Manu shows that the law-book 
was probably unknown to the early epic but that it was not 
unknown to the later epic. This indicates at least that the 
fabulous age ascribed to the law-book by the Hindus and by early 
European scholars may be disregarded in favour of a much later 
date. On the other hand, the present tendency is to exaggerate 
the lateness of the law-book and bring it down even to the third 
or fourth century A.D. Professor Jolly thinks that the code and 
the epic belong to about the same time, not later than the second 
or third century \ The code in any case may not have been 
identical with the work known to-day as Manu's law-book, for 
all these metrical works have suffered, as has the epic, from 
unnumbered additions. 

Nevertheless, from the contents of the extant law-book of 
Manu some noteworthy data may be extracted which seem to 
show that the work is earlier than any other Dharma ^astra. 
There is not the slightest allusion to any sectarian cult ; docu- 
ments are not cited in the rules on evidence; widow-burning is 

1 Jalias Jolly in "Recht und Sitte, pp. 16 and 30. Burnell in his translation of 
Mann contended for a still later date ; but this (1883) was before the relation of Mann 
to other law-books was understood. 

xii] Manu^ Vishnu^ Yajhavalkya^ Narada 279 

not recommended ; there is no recommendation of the cult of idols 
(service, etc.), though idols are known as objects of veneration; 
the position of the law-giver in regard to titles of law, evidence, 
ordeals, etc., is more primitive than that of any other author of 
a Dharma ^astra and even than that of Vishnu in the Dharma 
Sutra. The law-book of Vishnu belongs to the third century A.D., 
and that of Yajiiavalkya to the fourth century, and the advance 
on Manu in order, method, and detail of legal matters of these 
law-writers is very great. Hence, as in the case of the epic, it is 
probable that the date now currently assumed is too late, and that 
the Manava Code belongs rather to the time of our era or before 
it than later \ 

The law-book of Vishnu, which because of its Sutra form might 
be thought to be earlier than Manu, is so largely interpolated that 
in its present condition it must rank decidedly as secondary to 
that code. It appears to have been an expansion of a Sutra 
belonging to the Kathaka school of the Yajurveda enlarged in the 
hope of making it a general code favouring the cult of Vishnu. It 
mentions books under the modern name pustaka, recognises the 
burning of widows, knows the names of the days of the week, 
evidently borrowing here from Greek sources, acknowledges the 
Hindu Trinity, recommends the Tirthas or pilgrimages, which are 
decried by Manu, and in the matter of debts and legal procedure is 
later than that code. At the same time it contains much ancient 
material, especially in regard to legal penalties, the rights of kings, 
inheritance, etc. A large part of the work is not legal, but treats 
of sacrifice, impurity, sin and atonement, etc.^ 

The codes of Yajnavalkya and Narada are probably to be 
referred to the fourth and fifth centuries, respectively. The 
former was a learned pundit, probably of Mithila, whose work 
is so closely connected with that of Manu and at the same time 
is so clearly a condensation of this code, that it may be taken as 
certain that the author desired to better an original rather than 
make a new work. Yajnavalkya pays more attention to legal 

^ The contention of Mr Ketkar in his History of Caste in India (1909) that Manu 
Ib at least as late as the fall of the Andhras (third century a.d.), because they are 
mentioned as a low caste is not cogent, because the verse may well have been one 
inherited from a list of degraded tribes (castes) and preserved. The Andhras are 
regarded as barbarians in early Brahman literature. Cf. Aitareya Brahmana, 
VII, 18. 

* The connexion of the Manava code with the earlier Manava Sutras is not so close 
as that of the Vishnu code with the Kathaka Sutra, and it is even doubtful whether, 
as first thought probable, the Manava ^astra reverts to a sectarian Manava school. 

2 8o Growth of Law a7id Legal Institutions [CH. 

matters and improves on his model in his views regarding the 
rights of women, whom he permits to inherit equally with men. 
He elaborates the subjects of trade and ownership, and recognises 
written documents in evidence where Manu relies on ocular 
witnesses. He recommends the use of several new ordeals in 
testing truth, and shows a more conservative social feeling in 
objecting to the union of a Brahman priest and a slave-woman. 

Of Narada, who belongs to the fifth century and seems to have 
been from Nepal, it may be said that he is the first to give us 
a legal code unliampered by the mass of religious and moral 
teaching with which and out of which the earlier works on 
Dharma arose ^, a code which in its fine sub-divisions of the 
titles of law, as well as in its elaborate treatment otherwise of 
slaves, inheritance, witnesses, ordeals, etc., is the first in which law 
itself is the subject-matter. Narada's evident posteriority to Manu 
and Yajnavalkya does not show that it was an independent work, 
rather that it was based on these prior works. In addition to 
these legal lights it is necessary to mention only Brihaspati, who, 
as he extols Manu as the first of law-givers, also proves himself to 
be a sort of commentator rather than an original writer. His 
work is in fact a brief for Manu, and proves that in his day (about 
600 or 700 A.D.) Manu was recognised as the original and greatest 
law-giver. His citations from Manu also show that our text has 
not changed essentially since his day^. 

We have already seen that the four castes are regarded as the 
frame of social life, and that the young student, after, spending 
several years with a priestly preceptor, the length of time 
depending partly on caste and partly on aptitude, marries and 
becomes a householder, with numerous religious duties to perform. 
Twelve years of study is regarded as the minimum, forty-eight 
years as enough even for the most studious priest. Megasthenes 
tells us that the Hindus studied for nearly this length of time, but 
it is clear that only priests practised such zeal. The epic warriors 
are supposed to have finished their education by their sixteenth 
year, and the fact that a few words of a hymn are admitted as 
substitution for this part of the education (consisting in memorising 
verses) shows that for practical purposes a smattering of Veda was 
deemed enough in the case of all except the priest. The early 

1 Dharma means 'law' only as law is an expression of right, duty, etc. It is 
based upon revelation and custom, the first perfunctorily, the second actually. Local 
usage is the basis of law and may overrule laws made without regard to custom. 

2 On Brihaspati, see Jolly, Tagore Lectures, and the introduction to the translation 
of Narada and Brihaspati in S.B.E., vol. xxxiii. 

XI i] Civil and Criminal Law 281 

law-books devote no little space to the early youth and conduct in 
later life of the orthodox Aryan. Manu, for example, gives six of 
his twelve books to rules of life before he comes to discuss royal 
life and legal matters. Noteworthy is the early date at which a 
man retires from practical life. As the youth marries early, in the 
warrior caste as early as sixteen, though Manu recognises twenty- 
four or thirty as the usual (priestly ?) age, it may happen that he 
becomes a grandfather before he is forty, by which time, to be 
sure, the Hindu is often grey. Now it is expressly said that when 
a man becomes grey and a grandfather he is to enter the third 
cL^ama or stage of life and become a hermit, either accompanied 
with his wife or not, as he chooses. Severe asceticism marks this 
period of life (it is described in ftill by Manu, Book vi), and 
probably it was reserved generally for the priestly caste; some 
law-givers omit it It is likely that instead of this stage many 
priests became mendicants. The act of renouncing the world is 
introduced by a sacrifice of worldly goods and other ceremonies 
prescribed by the Siitras and law-books. But the latter, in 
distinction from the former, if indeed they devote much time to 
such matters at all, now turn to that part of Dharma or Right 
which is included under the head of Royalty and Vyavahara. The 
latter term means law in the modern sense, business intercourse 
legally interpreted, legal procedure. There is no formal distinction 
between civil and criminal law till the term vyavahara is divided 
by later writers into ' cases of property ' and * cases of hurt.' The 
first enumeration of legal titles is found in Manu and is as 
follows: (1) Recovery of debts; (2) Deposits and pledges; (3) Sale 
without ownership; (4) Partnership; (5) Resumption of gifts; 
(6) Non-payment of wages ; (7) Breach of contract ; (8) Annulling 
of sale and purchase ; (9) Disputes between the owner and tender 
of cattle; (10) Disputes regarding boundaries; (11) Assault; 
(12) Defiimation; (13) Theft; (14) Robbery (with violence); 
(15) Adultery; (16) Duties of man and wife; (17) Partition 
(inheritance); (18) Gambling (with dice) and betting (on cock- 
fights, etc.). In this category, criminal law is represented by the 
titles eleven to fifteen and eighteen, while the first nine and the 
sixteenth and seventeenth titles belong to civil law. There is also 
no distinction between laws affecting things and persons, and, to 
follow the indictment of Mill in his History of India, 'Non- 
payment of wages stands immediately before breach of contract, 
as a separate title, though it ought to be included under that head.' 
But the eighteen titles are remarkable as the first attempt to 

282 Growth of Law and Legal Institutions [ch. 

separate different cases ; to demand that Manu should have given 
us a perfect or even a perfectly clear list is unreasonable. 

The titles and the arrangement of Manu are followed by later 
writers, though with sub-divisions. Thus Brihaspati (li, 8), after 
giving the eighteen titles says that they 'are divided owing to 
diversity of lawsuits'; and other writers give ten chief crimes 
(killing a woman, mixture of caste, adultery, robbery, causing 
illegitimate birth, abuse, insults, assault, procuring abortion) 
headed by disobedience of the king's commands. It is, too, 
only later writers who assert that a lawsuit cannot be instituted 
mutually between father and son, or man and wife, or master and 
servant (Narada, i, 6). Although the titles begin with civil cases, 
there is no doubt that primitive procedure had to do with criminal 
cases before civil cases were known. Thus the earliest trials are 
for theft and perjury, and it is probable that theft was the first 
crime to be recognised legally. We have seen that even in the 
Siitras the thief is brought before the king and punished by him, 
and theft is the chief crime mentioned in the Vedas (more 
particularly theft of cattle, or robbery). There are a thousand 
forms of theft, according to Brihaspati, who makes theft one of the 
kinds of * violence,' of which there are four — homicide, theft, assault 
on another's wife, and injury (either abuse or assault). Thieves are 
of two sorts, open and concealed, 'and these are sub-divided a 
thousandfold, according to their skill, ability, and mode of cheating ' 
(Brih., XXII, 2). Those who cheat at dice or cheat a corporation 
are to be punished as impostors. The punishment for breaking 
into a house to steal is impalement ; highwaymen are hanged from 
a tree by the neck ; kidnappers are burned in a fire of straw ; one 
who steals a cow has his nose cut off ; for stealing more than ten 
measures of gi*ain the thief is executed ; for less he is fined eleven 
times what he has stolen {ibid. 9 £; Manu, viii, 320). The proof of 
theft is possession of the stolen property, or a track leading to the 
house of the suspected man ; but excessive expenditure, intercourse 
with sinners, and other ' signs ' may make a man suspected ; then 
he may have to clear himself by oath or ordeal. 

Manu recognises only two ordeals. Later authors add several 
more and some admit the application of an ordeal to the plaintiff 
as well as to the defendant. The oath of a witness is virtually an 
ordeal, as the oath invokes divine power, which punishes the guilty. 
The oath is taken according to the caste of the witness. For 
example, a farmer swears by his cattle, etc. Or one may simply 
swear that a thing is so, and if his house bums up within a week it 


xii] Ordeals 283 

is a divine conviction of perjury. Later authors also prescribe that 
in ordeals a writing be placed on the head of the suspected man 
containing the accusation and a prayer, so that the divine power 
may understand the matter. The two earliest ordeals are those of 
fire and water (Manu, viii, 114 f.). As the Siitras do not notice 
ordeals, except for a general recognition of them as 'divine' 
proofs on the part of the late Apastamba, and as the later writers 
Yajnavalkya and Narada describe five ordeals, adding the plough- 
share, scales, and poison, it is reasonable to conclude that Manu 
stands in time, as in description, midway between the two sets of 
authors and is the first to describe ordeals already known and 
practised. This is the judgment of BUhler and of Jolly ^, but the 
implication that the mention of daiva in older literature makes 
probable the existence of all the forms of ordeal mentioned only 
in later literature is not safe. Fire and water were first used, then 
come the elaborate trials with balance, etc., till eventually there 
are nine formal ordeals^. 

The nine ordeals are as follows, arranged in the order chosen 
by Brihaspati (xix, 4): the balance, fire, water, poison, sacred 
libation, grains of rice, hot gold-piece, ploughshare, and the ordeal 
by Dharma and Adharma. AVlien Professor Jolly says that no one 
of these can be judged later than any other on the ground that 
the growth from two to five and then to nine ordeals does not 
necessarily imply that one named later did not exist before the 
two named first, he exaggerates the probabilities. Is it likely, for 
example, that the ordeal by Dharma and Adharma is as old as that 
by fire and water ^? 

The ordeal by ploughshare is especially for those suspected of 
stealing cattle; the piece of heated gold is reserved for cases 
involving a theft * over four hundred ' ; that by poison, for one 
worth a thousand, etc. All such restrictions are late emendations 
and additions. In the fire-test one carries a hot iron ball, and if 
unburned is innocent. In the water-test, one plunges under water 
and to prove innocence must remain under as long as it takes for 

1 S.B.E., vol. XXV, p. cii; and Jolly, Recht und Sitte, p. 145. 

2 Compare Stenzler, in the Z.D.M.G., vol. ix, p. 661 ; E. Sohlagintweit, Die 
Gottesurtheile der Inder (1866) ; and A. Kaegi, Alter und Herkunft de$ gentian. 
Gottesurtheils (1887). 

^ The ordeal by Dharma and Adharma consists in painting pictures of Justice or 
Eight and Injustice or Wrong (abstract divinities) upon two leaves, one picture being 
white, the other black. The two images are then worshipped and invoked with sacred 
verses, and, after the leaves have been sprinkled with perfumes and the five products of 
the sacred cow, they are rolled in balls of earth and set in a jar without the accused 
observing them, who then extracts one and ' if he draws Dharma he is acquitted.' 

284 Growth of LiUw and Legal Institutions [ch. 

a dart, shot at the moment of diving, to be brought back. These 
two are alterations of old material, in which the accused walks 
through fire, as in epic tests, or is thrown into water to see if he 
drown. The balance is an easy ordeal and hence is used in the 
case of priests and women. It consists in seeing whether the 
accused weigh less or more the second time the test is made ; if 
heavier, one is guilty. Probably the weight of sin weighs one 
down. So in the Mahabharata, when a truth-telling man lies, his 
chariot begins to sink. 

Another method of exacting justice, used generally in the case 
of debt, was called * the custom ' (Manu, viii, 49) and consisted in 
what is now known as dharna. The guilty man (debtor) is 
besieged in his own house by his opponent, who fasts on him 
till the guilty one yields or the accuser dies. This method of 
punishing'an injurer is well known in the epics, where fasting to 
death against a person is an approved form of retaliation. The 
one who has committed the offence (or owes the money) usually 
yields in order to prevent the ghost of the dying creditor from 
injuring him. 

The punishment for murder, as already noticed {v. sup. p. 242), 
is at first a compensation paid to the relatives or the king (perhaps 
both) and later paid to the priests. The compensation is reckoned 
at a hundred cows (with a bull). This is in the case of a man ; 
in the case of a woman, the punishment is no more than if a slave 
is killed. Manu treats the compensation as a penance (paid to a 
priest) instead of a 'royal right,' as in the earlier Sutra period. 
The custom of appraising death at so much a head for which 
compensation is exacted existed into modern times and is 
mentioned by Tod in his Annals of Rajasthan\ 

Treason of all kinds is punishable by death, whether it consist 
in attacking the king or falsifying an edict or bribing the ministers 
of the king or helping his foes (Manu, ix, 232, etc.). Instead of 
other penalties, the guilty man, especially a priest, may be outcasted, 
that is, formally thrown out or banned from society, for in losing 
his caste he loses all social rights ; though in certain cases through 
established ceremonies he may be taken back. One who is outcast 
loses all right to primogeniture, inheritance, etc.^ 

1 See also Roth, Z.D.M.G., vol. xli, pp. 672 f. ; and other refereuces in Jolly, Recht 
und Situ, p. 132. 

' Primogeniture is not absolutely the cause of preference among heirs. An 
unworthy son may be passed over even if he be the eldest, in favour of a worthier 
junior. On banishment in lieu of capital punishment, see Manu, vni, 380. 

X 1 1] Punishments 285 

Except for treason, all crimes are judged relatively, that is, 
there is no absolute penalty, but one conditioned by the social 
order of the criminal or the victim of the crime. Thus in cases of 
defamation, if a warrior defame a priest, he is fined one hundred 
panas ; if a man of the people-caste do so, one hundred and fifty ; 
if a slave, he shall be corporally punished ; but if a priest defame 
a warrior, fifty ; if he defame a man of the people, twenty-five ; if 
he defame a slave, twelve, and this last fine is that imposed upon 
equals defaming equals within the Aryan castes. But if a slave 
insult a 'regenerate' (Aryan), his tongue is to be cut out. 
Especially is this the case in relations between the sexes, for 
though the rule of death for adultery is general (the woman is 
devoured by dogs in a public place and the man is burned alive, 
Manu, VIII, 371 f.), yet its antique provisions are really preserved 
only out of respect for tradition, the real law being that the 
offending man shall be fined and the woman have her hair cut off 
and be treated with contempt (Narada, xn, 92), unless the crime 
be one that outrages caste-sentiment. Thus a slave who has 
intercourse with a guarded high-caste woman may be slain; a 
Vai9ya shall lose his property ; a warrior be fined a thousand and 
be shaved with urine (Manu, viii, 384 f.). The old general rule of 
the Sutras to the effect that the woman be eaten by dogs and the 
man killed is preserved under the form, explicit in the later works 
but already implied by Manu, that this be the punishment if 
*a wife who is proud of the greatness of her family' (that is a 
woman of high caste) commit adultery, while Narada restricts 
the ferocious penalty to the impossible case of a priest's wife 
deliberately going to a low-caste man and seducing him. 

The general lex tcdionis is similarly confined to thieves or 
robbers (Manu, viii, 334), though another restriction limits it to 
intercourse between low and high caste (if a man of low caste 
injure a man of high caste the limb corresponding to the one hurt 
shall be cut off, ibid. 279). In one particular, however, the rule of 
increased fines is reversed, for in any case where a common man 
would be fined one penny {kdrshapana) the king is fined a thousand 
(Manu, viii, 336), probably on the principle (Manu, viil, 338) that 
he who knows more should suffer more^. 

^ The slave of the rules cited above is a ^udra-slave. The law defines slaves as of 
seven kinds, war-captives, daily workers for food, slaves bom in the house, men bought, 
given, inherited, and those enslaved for punishment. Slaves of war are known iu the 
epic {v. sup. p. 270) and there is no reason for supposing that a captive warrior may not be 
a slave (the commentator confines the captive to the Qiidra caste). According to practice, 
the warrior-caste slave is in bondage only for a year. The ' slave by punishment ' means 

2 86 Growth of L,aw and Legal Institutions [CH. 

In the province of civil law the later law-books show the 
greatest advance over the earlier. For example, where trade is 
concerned, the Sutras know nothing of legal business partnership, 
apart from the united family and its obligation as a whole to pay 
debts. Manu has the idea of a partnership, but his whole dis- 
cussion of the title concerns only the amount of fees payable to 
priests who together perform a ceremony ; and he merely raises 
the question whether all the religious partners or the one who 
performs a special act shall take the traditional fee for that one 
part. He decides that the four chief priests out of the sixteen 
shall get a moiety, the next four half of that, the next set a third 
share, and the next a quarter (the commentators are not unanimous 
in appraising the amounts), and adds ' by the same principle the 
allotment of shares must be made among men on earth who 
perform work conjointly' (Manu, viii, 211). In other words, 
except for stating that one should be paid in accordance with the 
work one does, Manu has nothing to say regarding ' partnership,' 
the formal fourth title of the list. Yajiiavalkya on the other 
hand includes agriculture and trades in his rule (ii, 265). Narada, 
while retaining the matter concerning priestly partnership, ex- 
presses the axiom above in this way: 'Loss, expense, profit of 
each partner are equal to, more than, or less than those of other 
partners according as his share (invested) is equal, greater, or 
less. Storage, food, charges, (tolls), loss, freightage, expense of 
keeping, must be paid by each partner in accordance with the 
terms of agreement,' etc. (iii, 3 f.). Finally Brihaspati begins his 
title ' Partnership ' thus : * Trade or other occupations should not 
be carried on by prudent men jointly with incompetent or lazy 
persons or with such as are afilicted with illness, ill-fated, or 
destitute. Whatever property one partner may give, authorised 
by many, or whatever contract he may cause to be executed, all 
that is (legally) done by them all. Whatever loss has occurred 
through Fate or the king shall be borne by all in proportion to 
their shares. When artists practise their art jointly, they share 
according to their work. If a number of men in partnership build 
a house or a temple, or dig a pool, or make leather articles, the 
headman among the workmen gets a double share. So too among 
musicians: the singers share and share alike, but he who beats 

a debtor unable to pay. It may be observed that prisons are for malefactors and 
traitors rather than for debtors. Manu speaks of prisons situated by the roadway 
where all who pass may see the punishments suffered by the wretches within, and 
the tortures of hell have the appearance of being copied from models nearer home (Manu, 
vni, 288). 

xii] Interest^ Wages ^ Pr^operty 287 

time gets a half share over.' And (still under the head of 
Partnership), ' when freebooters return from a hostile country 
bringing booty, they share in what they bring after giving a sixth 
to the king, their captain getting four shares, the bravest getting 
three, one particularly clever getting two, and the remaining 
associates sharing alike' (Brihaspati, xiv, 32). 

Regarding the use of money, an old Sutra rule confirmed by 
Manu permits interest at fifteen per cent, annually, but for men 
(debtors) of low caste the interest may be sixty per cent. ; yet this 
is where there is no security. The amount differs in any event 
according to caste, as already explained (p. 248). No stipulation 
beyond five per cent, per mensem is legal. Debts unpaid shall be 
worked out by labour by men of low caste. These rules obtain 
from the Sutra age and vary scarcely at all. Megasthenes 
erroneously reports that the Indians do not take interest {ovre 
Savd^ovai ovt€ Xaacti Bavei^ea-Oac, Fr. 27). Possibly he has in mind 
the provision that no Brahman shall be a usurer. Wages are often 
paid in kind ; one fifth of the crop or of the increase in flocks goes 
to the man who cares for the work. The tender of cattle, in 
contrast to the epic rule (v. sup. p. 269 f,), gets the milk of one cow 
out of ten (Manu, viii, 231). If a man work without food or 
clothing given to him he may take a third of the produce ; other- 
wise a fifth (Brihaspati, xvi, 13). But Narada gives a general rule 
to the effect that the servant of a trader, a herdsman, and an 
agricultural servant shall respectively take a tenth part of the 
profit, whether from the sale of merchandise, the increase of 
flocks, or the grain-crop (Narada, vi, 3). This is also the provision 
of Yajnavalkya (ii, 194) ^ The agricultural servant is a (^udra 
slave or a member of a mixed caste ^. 

The family represented in the law-books as the usual family 
is one where all the brothers live together as heirs of the father, 
who may or may not, as he or they prefer, divide his property 
during his life-time. The eldest son has certain rights of primo- 
geniture, but, as said above (p. 284, note), they may be taken from 
him in case he is unworthy (Manu, ix, 213). The property of a 
childless wife belongs to her husband, unless she is married by 
a rite not countenanced by the law; in that case her property 

^ This is expressly the wage ordained by the king in case there has been no especial 
stipulation between master and man. It represents therefore the normal percentage 
of gain (i\j) as wage for the hired assistant of a petty merchant, herdsman, or farmer. 

^ According to the commentator on Vishnu, lvii, 16, where the practice of renting 
land for half the crop is referred to, the herdsman is usually the son of a warrior by 
a slave-girl. These ' mixed castes ' really did most of the general work of a village. 

2 88 Growth of Law and L,egal Institutions [ch. 

reverts to her parents. Woman's property consists only in 
wedding-gifts, tokens of affection, and gifts from her brothers, 
father, and mother, as also what is given her after marriage by 
her husband. All this goes to her children at her death. 

As the preferred family is the joint-family, so the village is 
possessed as a whole of its holdings in land. Thus the only full 
discussion in Manu regarding boundaries (the tenth title) has to 
do with boundaries between two villages. Yet it is clear from 
other passages that private ownership in land under the king was 
recognised. He who first cultivates wild land, owns it (Manu, ix, 44). 
There is also a Sutra rule : * Animals, land, and females are not lost 
by possession of another ' (Gaut, xii, 39), which appears to imply 
individual ownership in land. The land around a village on all 
sides for one hundred * bows ' (about 600 feet) is common ; and if 
crops are grown there and cattle injure the crops, no damage can 
be exacted (Manu, viii, 237 f.) ; but the fields appear to be private 
property as they are fenced va}. 

The government of the country described in legal literature is 
not different from that of the Sutras, and in most respects agrees 
with the conditions represented in the epics, where government 
without a king is so well known as to be the object of the most 
severe condemnation ; and it is regarded as essential that a king 
of good family should be at the head of the state. Slave-bom 
kings are known in history but tabooed in law. The king is treated 
in the law-books under two heads, as general lord of the land and 
as judge and executioner. 

As lord of the land the king is a Zeus Agamemnon, a human 
divinity incorporating the essence of the deities Indra, Vayu, Yama, 
Varuna, Agni, etc., that is of the gods who protect the world in the 
eight directions. In other words, his chief function as lord is 
to protect, and he protects as 'a great deity in human form' 
(Manu, VII, 8). He has, to aid him, seven or eight councillors of 
hereditary office (* whose ancestors have been servants of kings '), 
with whom he daily consults as to affairs of state and religion. 
His prime minister should be a learned priest ; he should appoint 
ofiicials over all public works, mines, manufactures, storehouses, 
etc. Various royal monopolies are mentioned (salt is one of them). 
His officers must be brave and honest, and he himself must be 
brave and lead his troops personally into battle, where he is to 

^ This is not certain evidence that they were private possessions, but such appears 
to have been the case, as the rules regarding flowing water, ' seed cast in another's 
field,' etc. also presuppose private ownership (Manu, ix, 52 f.). To ' let land ' renders 
one impure (Ap., Bh. S., i, 18, 20). 

xii] Provincial Government^ War 289 

make it his duty to * kill kings,' for those kings go to heaven who 
seek to slay each other in battle and fight strenuously for that 
purpose (Manu, vii, 89). As overlord, the king receives a share of 
the booty won in battle, and it is his duty to distribute such booty 
as has not been taken singly among the soldiers. One military 
officer and a company of soldiers he should place as a guard over 
each village and town, to protect them. There should be a lord 
of one village, a lord of ten, (of twenty), of a hundred, and a lord 
(or lords) of a thousand. It is the duty of the lord of one village, 
grdmika, to report all crimes to the da^pa or lord of ten, and 
the lord of ten shall report likewise to the (lord of twenty, and he 
to the) lord of a hundred, and he to the lord of a thousand. As 
much land as suffices for one family shall be the income of the 
lord of one village and so on to the lord of a thousand, who shall 
enjoy the revenue of a town. All these men (it is said) are 
probably knaves and must be spied upon continually through the 
agency of a general superintendent in every town, who shall 
scrutinise the conduct of all the governing lords, * for the servants 
of kings appointed to protect generally become rascals who steal 
the property of others' {ibid. 123). The sum collected from his 
subjects by a just king (as taxes) is a fiftieth part of the increment 
on cattle and gold, and the eighth, sixth, or twelfth part of the 
crops ; while common artisans pay tax by a day's work monthly. 

These provisions (of Manu) are followed by Vishnu, who how- 
ever omits the intermediate lords of twenty villages and recognises 
only the decimal system throughout \ Instead of a thousand villages, 
Vishnu speaks of the ' whole country,' and probably the two ex- 
pressions were synonymous. Vishnu also specifies eunuchs as 
guards of the king's harem, not mentioned by Manu in connexion 
with the palace. Another point which brings Vishnu into line 
with the Sutra authorities (Baudh., i, 10, 18, 1 ; Vas., i, 42) is found 
in his rule regarding taxes. He gives no such option as Manu, but 
specifies one-sixth as the tax on grain and seeds and one-fiftieth 
on cattle, gold, and clothes (all authorities exempt priests from 

The men of war, according to Manu, are to be selected for 
prominent places (in the van) from Kurukshetra, the Matsyas, 
Panchalas, and those born in ^urasena — all districts in the 
neighbourhood of Delhi, Jaipur, Kanauj, and Muttra — a provision 
sufficiently indicative of the geographical origin of his code. It is 

^ The army divisions are also arranged decimally, in squads of ten and companies 
of one hundred or of other multiples of ten {Vas., xix, 17 f.). 

C.H.I. I. 19 

290 Growth of Law and Legal Institutions [ch . 

interesting to note that both Manu and Vishnu state that when 
a king has conquered a foreign foe he shall make a prince of that 
country (not of his own) the king there, and (Vishnu adds, iii, 49) 
he shall not destroy the royal race of his foe unless that royal race 
be of ignoble birth. He is to honour the gods and the customs of 
the conquered country and grant exemption from taxation (for 
a time) (Manu, vii, 201). 

In his capacity as judge the king tries cases himself or appoints 
a priest in his stead (Vishnu, iii, 73) ; but this latter provision is a 
later trait, though found in the Sutras. The earlier rule is that 
the king himself shall try cases daily and have built for that 
purpose a special hall as part of his palace in the inner city, 
and even, as we saw in the Sutra period {v. sup. p. 241), act as 
executioner. The fact that the king has also the pardoning power 
is implied in the provision that if the thief come before the king 
and the king smite him or let him go he is thereby purified^, a 
provision which also brings up the intricate question of the relation 
between legal punishment and religious penance. For many of 
the legal punishments for gross crimes are set down not as such 
but as religious expiations, and it is said that the king has to see 
to it that these religious obligations are fulfilled. In some cases 
without doubt punishment as a matter of law began as a matter 
of priestly religious law. The business of the king as judge was 
not unremunerative, as every debtor who was tried and convicted 
paid a tenth of the sum involved into the royal treasury (Vishnu, 
VI, 20). According to Manu (viii, 59), if plaintiff" or defendant is 
found guilty of falsification in regard to a contested sum, twice 
the sum itself shall be paid as a fine (to the king). The king's 
chaplain has an important place in the court of justice; he is chief 
of the councillors who as a body may include members of other 
Aryan castes. If a deputy act for the king, later authorities state 
that he should carry a seal-ring of the king as sign of authority 
(Brihaspati, i, 3). The right of appeal is also admitted in later 
law-books, which assume that a case may come up first before 
a family, or corporation, when if the judgment is questioned the 
case may be tried by assemblies (of co-inhabitants or castes) and 
then by judges duly appointed {ibid. 39). Yajnavalkya (11, 305) 
and Narada also (i, 65) say that, when a lawsuit has been 
wrongly decided, the trial must be repeated. According to 

1 Apparently a murderer might expiate his crime by dying for the king in battle 
(Apastamba, i, 24, 21), and even, 'if he fights three times, when not slain, he 
is freed' (Vas., Dh. Q., xx, 28). This antique provision is not preserved in the 
later law. 

XI i] Family Law 291 

Yajnavalkya appeal may be taken from corporations, etc., to 
'the judge appointed by the king' (11, 30). Such a judge is 
one appointed to act for the king in his own city or in the 
provinces, a provision found also in epic literature. All the law- 
books acknowledge the importance of the law of family (kida), 
gild or corporation (^reni), and assembly or greater corporation 
(piiga, gana\ of caste or co-inhabitants in making their own laws, 
which the king must not contrg^vene. 

There is one aspect of legal literature which is very significant 
of the origin of the completed codes. The laws, namely, frequently 
contradict one another either by implication or directly, not only 
the laws in general but those of the same code and even the laws 
placed in juxtaposition. An example of such contradiction is what 
may be found in Manu's code respecting the sale of a daughter. 
In VIII, 204, * Manu declares ' that if one girl has been shown to a 
prospective bridegroom and another is given, he may marry them 
both for the same price. In iii, 51, the same code (presumably 
the same Manu) says ' Let no wise father take even a small price 
for his daughter. , .for small or great this would be a sale'; and 
finally in ix, 97, we read : * If the giver of the price die after the 
price for a girl has been paid, she shall be given to the (bride- 
groom's) brother if she is willing,' and immediately after (ix. 98), 
'Even a slave should not accept a price in exchange for his 
daughter,' with a couple of verses following in the tone of the 
passage above, repudiating the 'sale of a daughter.' Yet in 
VIII, 366, under the head of the fifteenth title of law, it is stated 
that a low-caste man courting a woman of the highest caste 
deserves death (or corporal punishment); but one who courts an 
equal shall ' pay the price ' (and take her) if her father consents. 
It was an old provision that a fee or price (a yoke of oxen) should 
be paid to the father, and though this was softened down to a 
' fee ' or ' tax ' (^Ika), yet the advanced code objects formally to 
this business transaction. At the same time the old provision is 
retained, because it was a part of hereditary traditional law. In 
the epic also, the rule against selling a daughter is recorded ; but 
so strong is the feeling against violating family-law that the man 
who purposes^ to sell his daughter, ' because it is the custom in my 
family,' is upheld in doing so by a saint, who even declares that 
the sale is justified by the ancients and by God (Mbh., i, 113, 9f.). 
Here the girl is bought with gold and elephants and other costly 
things. On the other hand, as a matter of dignity, the father of an 
aristocratic girl, more particularly a princess, has in efiect heavy 


292 Growth of Law and Legal Institutions [ch . 

expenses. Thus when king Virata weds his daughter he bestows 
upon his son-in-law seven thousand horses and two hinidred 
elephants {Mhh., iv, 72, 36). The didactic epic says that a man who 
sells his daughter goes to hell(xiii, 45, 18); there is a general Sutra 
rule against selling any human being (Gaut., vii, 14)\ 

In regard to infant marriages the Sutras generally admit the 
advisability of marrying a girl when she is still too young to wear 
clothes, that is, before she becomes adult, or shows signs of 
maturity. The later law and practice are all at variance on this 
point. One of the epic heroes marries at sixteen a princess still 
playing with her dolls but old enough to become a mother shortly 
afterwards. The epic rule is that a bridegroom of thirty should 
marry a girl of ten, a bridegroom of twenty-one a girl of seven 
(xiii, 44, 14). Arrian (23, 9) reports that Indian girls were married 
at seven. Sita is said to have married Rama at six ! The rule of 
Manu is that a bridegroom of thirty shall marry a girl of twelve, 
one of twenty-four, a girl of eight (ix, 94); he also recommends 
that a girl shall not marry at all unless a suitable bridegroom 
appear; but again he countenances infant-marriages (ix, 88 
and 89). 

The rule in regard to the levirate, or the assignment of widows 
to another man to raise up sons for the deceased husband, is 
another instance of the way in which the codes were assembled 
out of contradictory material. In Manu, ix, 64-68, there is a flat 
contradiction of the preceding provisions on this point. No 
remarriage and no assignment of widows are permitted in a 
passage directly following the injunction that a widow shall be 
so assigned, for the purpose of giving her dead husband a son to 
pay him the funeral feast, etc. 

These laws regarding women are on the whole the most self- 
contradictory in the later codes. As the position of woman is 
more or less indicative of the state of civilisation, it is important 
to notice that the high regard paid to woman is confined to her 
function as a mother of sons. The bride must be a virgin (not 
a widow, Manu, ix, 65) and the remarriage of widows is generally 
not countenanced; but the codes do not sanction the custom of 
suttee till late, and the provisions for widows show that, though 
they probably lived miserably and without honour, they were not 
expected to die with their husbands. The Mahabharata and the 

^ The purchase of a wife is the ' demoniac ' form of marriage formally permitted in 
the case of a Vaigya and slave (Manu, in, 24). These two classes ' are not particular 
about wives' (Baudh., Dh. S., i, 11, 20, 14). 

XI i] " Marriage : Suttee 293 

Ramayana both recognise the custom of suttee, but only the 
former (and probably not in an early part) gives a case of a royal 
widow burning herself Avith her husband. It is perhaps the 
extension of a royal custom, as in the epic, which has made the 
rule general, so that later law and practice recommend suttee for 
all. A parallel would be the Self-choice {svayamvara) or election 
of a husband by a princess, afterwards regarded as an election- 
rule in the case of other maidens. The mother is praised as equal 
to the father in honour, and in default of sons she may inherit 
(Manu, IX, 217), but if she bear only daughters or has no children 
she may be divorced {ibid. 81 )^ In general, a woman receives 
respect only as potential or actual mother of sons. Manu repeats 
with unction the dictum of the Sutras that a woman is never 
independent (ix, 2 f.), and says that she may be slain for unfaithful- 
ness and divorced for barrenness ; he also regards women as too 
'unstable' to be called as witnesses (viii, 77). The view that 
women are chattel is yielding in the ^astra to a more enlightened 
opinion. In the epics also the rigidity of the law is not upheld by 
the tenor of tales showing women in a very different light from that 
exhibited in the didactic parts of the epic. Even at a much later 
age women were students, as they were wise in antiquity, and the 
annals of the law itself testify to the ability of the sex, for in the 
eighteenth century one of the great legal commentaries on the 
Mitakshara was written by a woman, Lakshmidevi. At what time 
the Purdah (' curtain ') rule came to confine women to the house 
is uncertain ; but probably not before foreign invasions had com- 
pelled the Hindu to adopt it. The epics and law-books speak of 
confining a woman as a punishment for ill-conduct (e.g. Manu, 
VIII, 365), but Manu insists that ' no man can really guard women 
by force' (ix, 10). To go veiled is only a court-custom alluded 
to in both epics. 

Deficient as are the legal text-books in arrangement and self- 
contradictory as are their enactments, they form a priceless 
heritage of a past which would otherwise have been largely lost 
to us, for they may be accepted as reflecting real and not artificial 
or invented conditions of life. Very material evidence has been 
furnished in the last few years as regards the trustworthy character 
of the information given by authors of the law-books. As remarked 
above concerning the Sutras {v. sup. p. 221), the idea that Brahman 

* The property of women forms too complicated a subject to be discussed here but 
it may be said in general that Manu represents an advance on the older denial of the 
Sutras that women, and in particular widows, could inherit. Baudhayana and 
Apastamba exclude widows from the husband's inheritance (e.g. Apast., n, 14). 

294 Growth of Law and Legal Institutions [ch . 

tradition is manufactured in order to glorify the Brahmans and 
that in the time of Buddha there were no castes, is rendered 
inadmissible by the fact that all Hindu literature acknowledges 
the main facts as stated in the epics and law-books. The fresh 
evidence on this point is supplied by the text of the Artha9a8tra 
called the Kautiliya, which may date from about 300 b.o. and is in 
accord with the Sutras and ^astras in all the chief points which 
these works have in common. This Artha^astra, which forms the 
subject of Chapter xix in this work, recognises castes and mixed 
castes and agrees with the ^astra of the law-givers in a multitude 
of instances, showing that the scheme of law arranged by the 
Brahmans was neither ideal nor invented but based upon actual 
life\ Here for example is repeated almost verbatim the rule against 
debts between father and son ; the kinds of marriage are the same ; 
the antithesis between Arya and ^udra is maintained; the rule 
that the wage is one-tenth the gain ' without previous agreement ' 
is identical with that of Yajiiavalkya cited above, etc. As the 
Kautiliya is a manual of rules imposed by a practical statesman, 
it is impossible to suppose that the conditions it depicts are 
imaginary, yet the same conditions are found in the Sutras, etc. 
If it was indisputable that this work belonged to the third or fourth 
century B.C., it would be of the utmost importance historically. 
As it is, some of the provisions of the Kautiliya agree with those 
of later rather than earlier law-books, and for the present it is not 
advisable to accept all its rules as belonging to the time assigned 
to the work as a whole ^. 

We see in the law-books the king of a limited realm still more 
or less of a patriarch among his people^; a people divided into 
general orders representing the military, priestly, and agi'icultural 
or mercantile classes, still mingling freely with each other, inter- 
marrying, but with due regard for the respect paid to the higher 
orders, and utterly devoid of the 'caste' rules later adopted in 
respect of food and marriage. The family is usually monogamous 
though it may be polygamous, and there are traces of the family- 
marriage, in which a wife marries a group of brothers. The menial 
work of house-life is carried on by slaves and half-breeds, who also 

1 Cf. the articles of Prof. Jacobi in Sitz. K.P.A., 1911, pp. 732, 954 f, ; 1912, 
pp. 832 f.; also the parallels published by Prof. Jolly in Z.D.M.G., lxyii, pp. 49 f. 

2 A sketch of law and government as presented by the Kautiliya Arthaijastra is 
given by Dr Barnett in his Antiquities of India, pp. 98 f. (1914) ; also by Mr M. N. Law 
in his Studies in Ancient Hindu Polity (1914). 

3 Thus the king has personally to go to market and 'settle the price of goods' 
every five days (Manu, vm, 402), 

XI i] The Law- Books and the Arthaqastra 295 

do most of the village labour and serve as petty craftsmen. More 
skilled workers like chariot-makers are of almost Aryan rank and 
are not excluded from society. The laws are harsh and cruel as 
regards punishment (the worker in gold who defrauds the king, 
for example, is, according to Manu, ix, 292, 'to be chopped to 
pieces with knives '), but a regard for truth and justice is the 
dominant trait of the law, which, if it may be personified, has at 
times a naive air of blandly but perplexedly seeking to steer a 
course between that which it thinks is right and ought to think is 
right, because the one has been reasoned out and the other has 
been handed down as part of * revelation ' or law divined 

^ In his Lectures on the Ancient History of India (Calcutta, 1919), Prof. D. R. 
Bhandarkar argues that the legal parts of the twelfth book of the epic revert to a period 
earlier than Kautilya; and that the 'beginnings of Indian thought in the sphere of 
Arthaqastra ' are to be assigned to the seventh century b.c. The first thesis is based 
on the theory that verse precedes prose in legal diction, which is certainly not demon- 
strable. The second is only another way of saying that the subject-matter of the 
Dharma Siitras is probably older than their present form, and that Kautilya had 
numerous predecessors, which is probable. The chief discrepancy between Manu and 
Kautilya is that the former represents a state conceived as a smaller kingdom; the 
latter's purview is not only more exhaustive but wider, e.g. he discusses the ' Arabian 
steeds ' in the king's stud (known to both epics) and cites as authorities later writers. 
On the whole, as with the Jatakas, it would be well not to accept as undoubtedly of 
' c. 400 B.C.' all the data of the Kautillya ArthaQastra. 



The Puranas, or collections of 'old-world ' legends, contain the 
traditional genealogies of the principal ruling houses of the Middle 
Country. They are closely connected both in form and substance 
with the epics and law-books. All three varieties of literature are 
written in the same kind of verse and in the same kind of Sanskrit ; 
and they have much of their subject-matter in common. Not 
isolated verses merely but long passages recur word for word in 
them all. They are all alike inheritors of the same stock of legen- 
dary and traditional lore ; and, so far as the nature of their con- 
tents is concerned, it is not always possible to draw any hard and 
fast line of distinction between them. Thus from different points 
of view the Mahabharata may be regarded, as indeed it regards 
itself, as an epic, a law-book, or a Purana. 

Any old-world story may in fact be called a Purana; but the 
term is specially applied to certain works which, in conformity 
with the classical definition, deal, or are supposed to deal, with 
the following five topics: (1) Sarga, the evolution of the universe 
from its material cause; (2) Pratisarga, the re-creation of the 
universe from the constituent elements into which it is merged at 
the close of each aeon {halpa) or day in the life of the Creator, 
Brahma ; (3) Vam^a, the genealogies of gods and rishis ; (4) Man- 
vantara, the groups of 'great ages' (mahayuga) included in an 
aeon, in each of which mankind is supposed to be produced anew 
from a first father, Manu; (5) Vamgdnucharita, the history of the 
royal families who rule over the earth during the four 'ages ' {yuga) 
which make up one ' great age.' 

With this ideal scheme none of the existing Puranas is in com- 
plete agreement. All difffer from it in various degrees by defect 
or by excess ; but, in spite of this, they profess generally to con- 
form with the old definition, and are thus made to give a descrip- 
tion of themselves which is no longer in accordance with the facts. 
It is evident, then, at the outset that their original form has been 
modified. Only seven out of the eighteen still retain the fifth 

CH. XI 1 1] Literature of the Kshatriyas 297 

section, which should contain an account of kings who have reigned 
during the historical period. For the purposes of political history 
all the rest are therefore without value. 

Orthodox Hinduism regards these works as of divine origin; 
and their framework is stereotyped in accordance with this view. 
The chief speaker is some ancient seer who has received the tradi- 
tion through Vyasa, who himself received it from the Creator. The 
narrative is introduced by a dialogue between the chief speaker and 
his audience, and is continued in the form of a series of reported 
dialogues between the characters of the stories told. 

Most commonly, though not invariably, the narrator is Loma- 
harshana or his son, Ugracravas. The former is called * the Suta,' 
and the latter 'Sauti' or 'the Suta's son' — titles which clearly 
indicate that the traditional lore, out of which the Puranas have 
been fashioned, was of Kshatriya, not of Brahman, origin; for 
the Sutas, its custodians, were a mixed caste who were entrusted 
with various important functions in royal households. In the 
Brahmanas the Suta is the royal herald and minstrel, and possibly 
also 'master of the horse.' He is one of the king's 'jewels' 
(ratnin) and ranks with the commander-in-chief of the army and 
other high officers of state; and in his character as herald he 
was inviolable. In the law-books he is described as the son of 
a Kshatriya by the daughter of a Brahman. The Puranas say 
that he was born to sing the praises of princes and that he was 
entrusted with the care of the historical and legendary traditions ; 
but they state definitely that he had no concern with the Vedas 
(Vdyu Pur., i, 1, 26-28). In later times he appears as the king's 
charioteer ; but he still retains his exalted rank, and in the dramas 
he speaks Sanskrit — the sign of high birth or education — while the 
inferior characters speak some Prakrit dialect. 

In the interval between the Brahmanas and the dramas the 
Siita had evidently been deprived of some of the most important 
of his ancient functions ; and this change in his fortunes reflects 
a change which had taken place in Indian society and in the 
character of the Puranas. In the heroic age, when the Siita was 
the chronicler of kings, the Kshatriyas, as we gather from the 
Upanishads and from early Jain and Buddhist literature, occupied 
a position of considerable intellectual independence. But this 
position was not maintained. In India, as in medieval Europe, 
the priestly power eventually asserted its supremacy, and all the 
old Kshatriya literature was Brahmanised. The record of the 
lineage of princes tended to disappear from the Puranas, and its 


'The Pur anas [CH. 

place was taken by endless legends about holy places, or hymns in 
praise of the divinities who were worshipped there. The Puranas 
had passed from the Kshatriyas to the Brahmans, from the royal 
bards to the priests who waited on temples and pilgrims' shrines — 
a class mentioned with contempt in the law-books (Manu, iii, 152). 
But, in spite of this transference and the radical changes which 
it involved, some of the old terms and some fragments of the old 
literature still remained to testify to a state of things which had 
passed away. 

Thus the Puranas, like the Mahabharata, have undergone a 
complete transformation. Just as the Mahabharata, originally the 
story of a war, has been made into a Dharma ^astra, the main 
object of which is to inculcate duty, so too the Puranas are no 
longer mere collections of ancient legends. Like the * Lives of the 
Saints' they have been applied to purposes of edification. For 
them the kings of the earth have existed simply to point a moral — 
the vanity of human wishes : 

He who has heard of the races of the Sun and Moon, of Ikslivaku, Jahnu, 
Mandhatri, Sagara, and Raghu, who have all perished ; of Yayati, Nahusha, and 
their posterity, who are no more ; of kings of great might, resistless valour, and 
unbounded wealth, who have been overcome by still more powerful Time, and are 
now only a tale : he will learn wisdom, and forbear to call either children, or wife, 
or house, or lands, or wealth, his own. ( Vishnu Pur., trans. Wilson, iv, p. 240.) 

The chief object of the Puranas is to glorify Qiva or Vishnu, 
the great divinities who, in their manifold forms, share the alle- 
giance of India. They have become sectarian and propagandist, 
exalting their own particular deity at the expense of all others. 
In a word, they have become the scriptures of various forms of 
the later Hinduism, and bear to these the same relation that the 
Vedas and Brahmanas bore to the older Brahmanism. But while 
the scriptures of the ancient sacrificial religion have remained 
unaltered and have been protected from textual corruption by the 
elaborate devices of priestly schools, the Puranas have adapted 
themselves to the changes which have taken place in the social 
and religious life of the people, and their text has been perverted 
by generations of editors and transcribers. 

They are made up of elements old and new. However late 
they may appear in their present form — and some of them are 
said to have been altered in quite recent times — there can be no 
question that their main source is to be traced back to a remote 
antiquity. The ancient lore of the bards from which, like the 
epics, they are derived is known to the Atharvaveda (xv, 6, 11 £) 

xiii] Modern Scholarship 299 

as a class of literature with the general title Itihdsa-Purana 
* story and legend ' ; and both in the Upanishads {Chhdndogya, vii, 1 
and 7) and in early Buddhist books {Sutta Nipdta, iii, 7) this litera- 
ture is called the fifth Veda. It was in fact the Veda of the laity ; 
and as such the epics and Puranas have been universally accepted 
all through the classical period even down to the present day. 

The attitude of modern scholarship towards these documents 
has varied at different times. In the early days of the study of 
Sanskrit in Europe they were accepted as historical. But it was 
soon evident that no satisfactory system of Indian chronology could 
be established by their aid alone; and for a long time scholars 
seem to have agreed to ignore their evidence unless when sup- 
ported from other sources. After having been unduly appraised, the 
Puranas were unduly neglected. In recent years a reaction has 
set in, and there is a growing belief that these works are worthy 
of more serious attention than they have hitherto received. It 
has been shown that the historical information which they convey 
is not so untrustworthy as was formerly supposed. Dr Vincent 
Smith, for example, was able in 1902 {Z.D.M.G., pp. 654, 658 ff.) 
to prove that both the dynastic list of the Andhra kings and the 
duration of the diflferent reigns as stated in the Matsya Purana are 
substantially correct. 

The critical study of the Puranas, which was inaugurated by 
Mr Pargiter's Dynasties of the Kali Age (1913), is still in its 
infancy. When this important branch of literature has been 
examined by the methods which have been applied to the Vedas 
and Brahmanas, there can be little doubt that valuable historical 
results will be obtained. The Puranas are confessedly partly 
legendary and partly historical. The descriptions of superhuman 
beings and of other worlds than this are glorified accounts of the 
unknown founded on the analogy of the known. They find their 
counterpart in that Christian Purana, Milton's Paradise Lost. The 
descriptions of ancient monarchs and of their realms are essentially 
historical. They may be compared to the Sagas and the medieval 
chronicles of Europe. They are the products of an imaginative 
and uncritical age in which men were not careful to distinguish 
fact from legend. It is the task of modern criticism to disentangle 
the two elements. Its first object should be to remove from the 
existing Puranas all later additions, and then from a comparison 
of their oldest portions to determine the relations in which they 
stand to one another, and thus, as far as possible, to restore their 
common tradition to its original form. 

300 The Puranas [ch. 

As yet this necessary preliminary process has not even been 
begun ; and until it is completed the real value of the Puranas as 
historical evidence cannot be estimated. They still continue to 
be dated by scholars according to the latest indications which can 
be discovered in them, and they are too often rejected as incom- 
petent witnesses for the events of any earlier period. The ele- 
mentary fact that the date, whether of a building or of a literary 
production, is not determined by its latest addition is in their case 
generally ignored. 

The eighteen Puranas are associated with an equal number of 
Upapuranas. Traditional lists, in which all of these Puranas and 
Upapuraiias are arranged in a definite order of precedence, have 
been preserved in the works themselves. In these the Brahma 
Purana stands first ; and, as this position and its alternative title 
*Adi' or 'the First' would alike seem to indicate, it is probably 
the oldest. There would appear to be nothing in its earlier 
portions to discountenance this claim; but it has received late 
additions, and on the evidence of these Wilson ascribed it to the 
thirteenth or fourteenth century. This afibrds a signal instance 
of the misconception which may be caused by failure to dis- 
criminate between the ages of diflferent parts of a work. All the 
Puranas without exception have been altered. The Vishnu Pur., 
which stands third in the list, has apparently suffered less than 
the others. 

Comparatively little is known about the Upapuranas. Few 
of them have been published or thoroughly investigated. They 
appear to be, as a rule, still more narrowly sectarian than the 
Puranas, and to be intended to further religious interests which are 
more purely local . They probably have little, if any, historical worth. 

The total number of couplets comprised in the eighteen Puranas 
as given in the lists is 400,000, the length of the different versions 
varying from 10,000 to about 81,000 couplets. These statements 
were no doubt accurate at the time when the computation was 
made; but great changes have since taken place. On the one 
hand, whole sections have been lost. The Vishnu Pur., usually 
regarded as the best conserved of all, has now less than 7000 
couplets : in the lists it appears with 23,000. On the other hand, 
numerous more recent works claim to belong to one or other of 
the Puranas, so that it is now sometimes impossible to define the 
precise limits of the latter. If all the productions which profess 
to form portions of the Skanda Pur., for instance, were included, 
the total given in the lists would be greatly exceeded. 

XI 1 1] Different Versions and Age 301 

As to the history of these eighteen versions of a common tra- 
dition, it seems certain that they were moulded into their present 
form at various centres of religious activity. The case has been 
clearly stated by the late Mr A. M. T. Jackson in the Centenary 
Volume of the Jour, of the BoTiibay Branch of the R. A. S. 
(1905), p. 73: 

A very striking analogy to the mutual relations of the various Puranas is to be 
found in the case of our own Saxon chronicle, which, as is well known, continued 
to be written up in various monasteries down to the reign of Stephen, though 
the additions made after the Roman conquest were independent of each other. 
Similarly the copies of the original verse Purana that were possessed by the 
priests of the great centres of pilgrimage were altered and added to chiefly by 
the insertion of local events after the fall of a central Hindu government had made 
communication between the different groups of Brahmans relatively difficult. In 
this way the Brahma Purana may represent the Orissa version of the original work, 
just as the Padma may give that of Pushkara, the Agiii that of Gaya, the Varaha 
that of Mathura, the Vamana that of Thanesar, the Kurma that of Benares, and 
the Matsya that of the Brahmans on the Narmada. 

At what period the eighteen Puranas assumed their distinctive 
titles is uncertain. It was no doubt long after they had ceased 
to be regarded as repositories of historical information, for they 
are grouped in the traditional lists entirely according to their 
religious character. It has sometimes been supposed that one of 
their number is the immediate source of all the others; but it 
seems more probable that they belong to several groups which 
represent different lines of tradition. Possibly the Puranas which 
are narrated by the Suta may belong to one such group, and those 
which are narrated by Maitreya to another. One at least of the 
present titles may be traced back to an early period; for the 
Bhavishya or Bhavishyat Pur., the ninth in the list, is quoted in 
the Dharma Siitra of Apastamba (11, 9, 24, 6) which cannot be 
later than the second century B.C. and may possibly be still more 
ancient ^ But as a rule early references to this traditional lore 
describe it generally as Purana or Itihasa-Purana, a class of lite- 
rature which, as we have seen, undoubtedly goes back at least to 
the time of the Atharvaveda. 

Some such antiquity is implicitly claimed by the Puranas in 
their prologues. Para^ara, who narrates the Vishnu Pur., is the 
grandson of Vasishtha, the rishi of the seventh mandala of the 
Rigveda; and his narration takes place in the reign of Parikshit^ 

1 Chapter x, pp. 249 f. 

^ The name appears as Parikshit in the earlier, and as Parlkshit in the later, 

302 The Pur anas [cH. 

who is celebrated as a king of the Kurus in the Atharvaveda. 
Nearly all the other Puranas are attributed to the Suta and to 
a period four generations later. Of the prologues to these that 
of the Vayu Pur. may be selected as typical. The rishis are per- 
forming their twelve-year sacrifice in the Naimisha forest on the 
bank of the sacred river Drishadvatl. To them comes the SUta, 
the custodian of the ancient Kshatriya traditions. At their re- 
quest he takes up his parable and retells the legends entrusted to 
his care by Vyasa. The scene is laid in the reign of the Puru 
king Adhisimakrishna, who is supposed to have lived before the 
beginning of the Kali Age, or, as we should say, before the his- 
torical period. But the genealogy assigned to him indicates a 
more definite date; for of his immediate forbears — A9vamedha- 
datta, Qatanika, Janamejaya, Parlkshit — all but the first, his father 
A9vamedhadatta, are no doubt to be identified with kings of the 
same names who appear in the Brahmanas. 

Whatever may be the historical value of these prologues, they 
certainly carry us back to the same period, the period of the 
Atharvaveda and the Brahmanas, to which modern research has 
traced the existence of an Itihasa-Purana literature. To suppose 
that they are altogether concoctions of the Middle Ages is to place 
too great a strain on our credulity. They can scarcely have been 
reconstructed from the fragmentary evidence supplied by Vedas 
and Brahmanas at a period when no one could have dreamed of 
treating Vedas and Brahmanas as historical documents — a task 
reserved for the nineteenth century. We cannot escape from the 
only possible conclusion, that the Puranas have preserved, in how- 
ever perverted and distorted a form, an independent tradition, 
which supplements the priestly tradition of the Vedas and Brah- 
manas, and which goes back to the same period. This tradition, 
as we may gather from the prologues, was handed down from one 
generation of bards to another and was solemnly promulgated on 
the occasion of great sacrifices. 

The Kshatriya literature of the heroic age of India has for 
the most part been lost. Such of it as has survived has owed its 
preservation to its association with religion. The commemoration 
of the lineage of kings found a place in religious ceremonial, as, 
for instance, in. the year-long preparation for the 'horse-sacrifice,' 
by the performance of which a king ratified his claim to suzerainty 
over his neighbours. It is no doubt to such commemorations 
that we owe the dynastic lists which have been preserved in the 

XI 1 1] The Manus 303 

The historical character of these works is disguised by their 
setting. They have been made to conform with Indian ideas as 
to the origin and nature of the universe and its relation to a First 
Cause. The effect of this has been to remove the monarch, who is 
represented as reigning when the recital takes place, and all his 
predecessors from the realm of history into the realm of legend ; 
and it has been found necessary to preserve the illusion through- 
out the subsequent narrative. The Suta is invited by the sacri- 
ficing rishis of the Naimisha forest to describe the Kali Age which 
is still to come. It is evident that he can only do so prophetically. 
He can only reproduce the foreknowledge which has been divinely 
implanted in him by Vyasa. Accordingly he uses the future tense 
in speaking of kings who have actually reigned and of events which 
have actually happened. History has been made to assume the 
disguise of prophecy. 

When this pretence is set aside, and when all legendary or 
imaginary elements are removed, the last two sections of the 
Puranas afford valuable information as to the geography and 
history of ancient India. 

The fourth section, the manvantara, deals with the * periods of 
the different Manus.' These form part of a chronological system 
which is purely hypothetical. Time, like soul and matter, is a 
phase of the Supreme Spirit. As Brahma wakes or sleeps, the 
universe wakes or sleeps also. Each day and each night of Brahma 
is an ' aeon ' (kalpa) and is equivalent to a thousand ' great ages ' 
(mahdyuga), that is to say, 1000 x 4,320,000 mortal years. During 
an ' aeon ' fourteen Manus or ' fathers of mankind ' appear, each 
presiding over a period of seventy-one * great ages ' with a surplus. 
Each 'great age' is further divided into four 'ages' {yuga) of 
progressive deterioration like the golden, silver, brazen, and iron 
ages of Greek and Roman mythology. These are named, from the 
numbers on the dice, Krita, Treta, Dvapara, and Kali, and are 
accordingly supposed to last for periods represented by the pro- 
portion 4:3:2:1. We need not follow this subdivision of time 
down to its ultimate fraction ' the twinkling of an eye ' (nimesha) 
or dwell on the sectarian zeal which leads some of the Puranas to 
assert that an ' aeon ' of Brahma is but ' the twinkling of an eye ' 
in the endurance of Qiva or Vishnu. 

The account of the manvantara of Manu SvayambhUva, the 
first in the series of fourteen, includes a description of the universe 
as it now exists or is supposed to exist. The greater part of this 
description is, like the chronology, imaginary. The world, according 

304 T'he Puranas [ch. 

to this primitive geography, consists of seven concentric continents 
separated by encircling seas. These are the ' seas of treacle and 
seas of butter ' at which Lord Macaulay, with his utter inability 
to understand any form of early culture, scoffed in his celebrated 
minute on Indian education. The innermost of these continents, 
which — and here we come to actuality — is separated from the next 
by salt water, is Jambudvipa; and of JambudvTpa the most im- 
portant region is Bharatavarsha or Bharata, that is to say, the 
sub-continent of India : 

The country that lies north of the ocean, and south of the snowy mountains, is 
Bharata ; for there dwell the descendants of Bharata... 

The seven main chains of mountains in Bharata are Mahendra, Malaya, Sahya, 
^ktimat, Riksha, Vindhya, and Paripatra... 

On the east of Bharata dwell the Kiratas (the barbarians) ; on the west, the 
Yavanas : in the centre reside Brahmans, Kshatriyas, Vai9yas, and ^Qdras. 
{Vishnu Pur., trans. Wilson, 11, pp. 127-9.) 

General descriptions such as this are followed by lists, more 
or less detailed, of the rivers which flow from the Himalayas and 
the seven great ranges, and of the tribes inhabiting the various 
regions. As in all early geography, the district is known by the 
plural of the tribal name. Similar lists are found also in the 
Mahabharata and elsewhere. This extensive geographical litera- 
ture gives a remarkably full account of the whole sub-continent. 

The geographical, like the dynastic, lists have evidently been 
brought up to date from time to time, since foreign invaders of 
very different dates appear in them. These seem to range from 
the Yavanas, 9^kas, and Pahlavas, who came into India in the 
second and first centuries B.C., to the Hunas, who broke up the 
Gupta empire at the end of the fifth century A.D. 

The fifth and last section of the Puranas, the vam^dnucharita, 
gives an account of the kings of the earth, the descendants of 
Manu Vaivasvata, the 'son of the Sun.' The narrative uses all 
three tenses, past, present, and future; for it recounts the kings 
who have been, the kings who are, and the kings who are to be. 
The earliest of these genealogies, like the most ancient chronicles 
of other peoples, are legendary. They trace the descent of the 
rulers of this world from the Sun and Moon, and through them 
from the Creator — a claim inherited and still maintained by the 
Surajbansi and the Chandrabansi families of Rajput princes. Such 
pedigrees have been pieced together from fragments of religious 
lore or from fancied etymologies on to which old-world traditions 
and speculations have been engrafted. Ila, the daughter of Manu, 

XI ii] Traditional Genealogies 305 

from whom the Lunar family is derived, personifies, as her name 
denotes, the sacrificial ofiering made by Manu in the legend of 
the Flood ((^ata. Br., i, 8, 1, 11). Such legendary characters are 
everywhere the result of man's early speculations on the origin of 
the world. The first glimpses of authentic history only appear 
when tribal names are inserted in the genealogies under the dis- 
guise of eponymous ancestors. These, too, are the outcome of 
hypothesis, but of hypothesis founded on facts. All the members 
of a tribe are presumably descended from a common ancestor, and 
related tribes are descended from related ancestors. On these 
supposed individuals the names of the tribes are conferred; and 
they supply a sort of genealogical framework which continues to 
be filled in by tradition until the age of records. Once fashioned 
in this way such genealogies are accepted without question until 
the period when critical scholarship arises and undertakes its first 
duty, which is to discriminate between legend and fact in the story 
of past ages. 

In the Puranas, which were the common scriptures of the 
ruling Aryan peoples of Northern and Western India, the tradi- 
tional genealogies of the royal houses have been collected and 
made to form a consistent whole. Not only are the ancient tribes 
of the Rigveda and the kingdoms immediately descended from 
them represented here, but the realms of Kosala (Ayodhya), 
Videha, VaigalT, and Magadha, which were not Aryanised until 
a later date, have also been brought into the scheme and fur- 
nished with a still longer and more august pedigree. They belong 
to the Solar family and are derived directly from Manu through 
Ikshvaku. A family of princes bearing this name is known from 
Vedic literature ; and it is quite possible that the Solar dynasties 
of Kosala and other kingdoms to the east of the Middle Country 
may have been descended from this family. If so, the Ikshvaku 
of the genealogical tree must be regarded as an eponymous an- 
cestor ; and as his superhuman origin had to be explained, a myth 
founded on a far-fetched etymology of his name was invented. 
Ikshvaku was so called because he was born from the sneeze 
(kshava) of Manu ( Vishnu Pur., trans. Wilson, iii, p. 259). 

Fragments of historical fact may no doubt be found embedded 
even in the earliest lists ; but these fragments have been carried 
down the stream of time and deposited far away from their original 
home. Thus, for instance, Purukutsa and his son Trasadasyu, who 
in the Rigveda are Piirus living on the Sarasvati, appear in the 
Puranas among the Solar kings of Kosala ; Vadhryagva, Divodasa, 
C.H.I. I. 20 

3o6 The Pur anas [ch. 

Pijavana, and Sudas, who form a direct line in the succession of 
Bhai'ata princes ruling in the country between the Saras vati and 
Drishadvati, appear in this order, but with intervening reigns, 
among the kings of N. Pafichala^ It is probable that these ap- 
parently conflicting statements are not really contradictory: the 
chain of evidence which might bring the tradition of the Puranas 
into substantial agreement with the Rigveda has been broken. 

But it is clear that documents of this kind can only be used 
with the greatest caution. To some extent at least they have 
unquestionably been fabricated in accordance with preconceived 
opinions. How these pedigrees have been elaborated, even at a 
comparatively late date, by court poets who sought to magnify the 
ancient lineage of their lord, may sometimes be seen at a glance. 
For example, in the genealogy of the Ikshvakus of Kosala the im- 
mediate predecessors of Prasenajit, the contemporary of Buddha, 
are (^akya, ^uddhodana, Siddhartha, and Rahula. That is to say, 
the eponymous hero of Buddha's clan, Buddha's father, Buddha 
himself, and his son have all been incorporated in the dynastic list 
of the kings of Kosala^. 

It seems impossible to bring the Puranic genealogies into any 
satisfactory relation with the Vedic literature or with one another 
until we approach the period at which they profess to have been 
recited, that is to say, the reign of Parikshit in the case of the 
Vishnu Pur. and the reign of Adhisimakrishna in the case of most 
of the others. Then certain synchronisms seem to afford a more 
secure chronological standpoint. Parikshit is celebrated as a king 
of the Kurus in the last and latest book of the Atharvaveda : ac- 
cording to the epic, as usually interpreted, he was appointed king 
of Hastinapura more than thirty-six years after the great war 
between the Kurus and Pandus. Adhisimakrishna, the great 
great grandson of Parikshit is represented by the Puranas as 
contemporary with Divakara of Kosala and Senajit of Magadha. 
Between the last mentioned and his predecessor Sahadeva, who 
was killed in the great war, six reigns intervene. The length of 
each reign and the total duration of the different dynasties of 
Magadha are given in some versions. Unfortunately the state 
of the text is so corrupt and the numbers are so discrepant that 
this evidence is generally of no value. Leaving out of account an 
impossible reading which attributes a reign of one hundred years 
to Niramitra, the Mss. as they stand give a maximum of 289 and 

1 Pargiter, J.R.A.S., 1910, p. 28. 

^ See Pargiter, Dynasties of the Kali Age, pp. 11, 67. 

XI 1 1] The Great War : Purus 307 

a minimum of 259 years to the six reigns which separate the great 
war from Senajit of Magadha ; and even the lesser of these esti- 
mates would seem to be excessive. We must be content with the 
general conclusion that the tradition of the Puranas, according to 
the dynastic lists of Hastinapui-a and Magadha, places the great 
war early in what we know as the Brahmana period, say about 
1000 B.C. 

That the war between the Kurus and Pandus is historical and 
that it took place in ancient times cannot be doubted, however 
much its story has been overloaded with legend, and however late 
may be the form in which it has been handed down. The legend 
of the war of the Mahabharata in India finds its exact parallel in 
the legend of the Trojan war in Europe. Each became the great 
central point to which the nations of the Middle Ages referred 
their history. To have shared ancestrally in the fame of Kuru- 
kshetra or of Troy was for nations the patent of nobility and ancient 
descent. The remotest peoples of Eastern and Southern India and 
the late invaders of the North- West alike claimed a place in the 
story of the Mahabharata, even as the royal houses of Western 
Europe tmced their origin to Trojan heroes. Until the close of 
the sixteenth century no historian of France or Britain doubted 
that the kings of these countries were descended from the Trojan 
Francus or Brutus, both of whom Avere in reality eponymous heroes 
like Yadu and his brothers in the Puranas. Milton in his History 
of England (1670) repeats the story of Brutus at length and in 
detail ; but a chance phrase — ' they who first devis'd to bring us 
from some noble ancestor ' — shows that historians were beginning 
to recognise the origin of such legends. And so far as the Maha- 
bharata associates most of the nations of India with the great war 
it has been ' devis'd ' in the same manner and for the same pur- 
pose. A nucleus of fact has been encrusted with the legendary 
accretions of ages. 

After the great war detailed dynastic tables continue to be 
given in the case of three royal lines only — the Purus, the Iksh- 
vakus, and the kings of Magadha. Other kingdoms are mentioned 
summarily with a bare statement of the number of contemporary 
reigns. The Puranic history is thus, professedly though not actually 
(pp. 311, 318), confined in its later stages to the regions now 
represented by the United Provinces and S. Bihar. 

In the Purus or Pauravas of the Puranas the Bharatas of the 
Rigveda and the Kurus of the Brahmanas have been merged. In 
the Rigveda both the Purus and the Bharatas live in the land of 


3o8 The Pur anas [ch. 

the Sarasvati (Brahmavarta or Sirhiud). But already the Aryan 
occupation of Kurukshetra, the adjacent country of the upper 
Jumna and Ganges on the south-east, was beginning : for a victory 
on the Jumna gained by Sudas, king of the Tritsus, over a native 
leader called Bheda is referred to in vii, 18, 19. In the Puranas, 
Sudas and his family appear in the list of the kings of N. Paiichala 
to the east of Kurukshetra. That is to say, the later kings of N. 
Panchala (p. 316) claimed descent from the Tritsus of the Rigveda, 
who are regarded by the Puranas as a branch of the Purus. 

But the gi'eat conqueror of Kurukshetra was Bharata Dauh- 
shanti, whose victories on the Jumna and Ganges are commemorated 
in an old verse quoted by the ^atapatha Brahmana (xiii, 5, 4, 11); 
and the extension of Bharata's conquests to Ka9i (Benares) is 
attributed by another ancient verse (xiii, 5, 4, 19) to ^^tanika 
Satrajita. In the Puranic list of Puru kings, Bharata and his 
father, Dushyanta, appear long before, and ^atanika soon after, 
the beginning of the Kali Age. Between the periods of the two 
conquerors, Bharata and ^atanika, came the war of the Maha- 
bharata, which for the Puranas marks the division between the 
third and fourth ages of the world. 

The later list contains the names of twenty-nine Puru kings, who 
lived after the war. They reigned first at Hastinapura, the ancient 
capital of the Kuru princes, which is usually identified with a ruined 
site in the Meerut District 'on the old bed of the Ganges, lat. 29^ 9' N., 
long. 78° 3' E.' (Pargiter, Mdrh. Pur., p. 355) ; but when this city 
was destroyed by an inundation of the Ganges in the reign of 
Nichakshus, the successor of Adhisimakrishna, they removed the 
seat of their rule to Kau9ambi, possibly the present Kosam in the 
Allahabad District. Another of their capitals was Indraprastha in 
the Kuru plain, the ancient city of the Pandu princes : it is the 
modern Indarpat, near Delhi. The Purus therefore, with their 
capitals in the north, east, and west, ruled over a large portion of 
the present province of Agra from the Meerut Division on the north 
to the Benares Division on the south-east. The dynasty came to an 
end with Kshemaka, the fourth king to reign after Udayana, the 
contemporary of the Buddha (p. 310)^ 

From the evidence both of Vedic literature and of the Puranas 
it appears that the Ikshvakus were originally a branch of the Purus. 
They were kings of Kosala, the country which lay to the east of the 

1 For the historical details here summarised, see Vedic Index, i, pp. 153, 155, 165, 
169; II, pp. 12, 96, 110, 186; Pargiter, J.R.A.S., 1910, pp. 26-29; Kali Age, 
pp. 4 ff., 65 fi. 

XI 1 1] Kosala: Magadha 309 

Kurus and Paiichalas and to the west of the Videhas, from whom 
it was separated by the river Sadanira, probably the Great Gandak. 
This territory was practically the modern province of Oudh. The 
chief cities were Ayodhya (Ajodhya on the Gogra in the Fyzabad 
District) with which the Saketa of Buddhist writers was probably 
either identical or closely associated, and ^ravasti (Set Mahet in 
the Gonda District). In story Ayodhya is famous as the city of 
Da9aratha, the father of Rama, the hero of the Ramayana. Both 
of these characters, who may possibly have been historical, are as- 
signed by the Puranas to a dim and distant period long before the 
beginning of the Kali Age. 

Although the extension of Brahmanism from the land of the 
Kurus and Panchalas to Kosala was comparatively late (p. 117), the 
Aryan occupation of the country goes back to an earlier period. 
In the later Vedic literature two kings of Kosala, Hiranyanabha 
and Para Atnara, probably father and son, seem to be mentioned 
as performing the horse- sacrifice in celebration of their victories ; 
and, as the former of these appears in the Puranic list before the 
Kali Age, the conquest of Kosala was evidently attributed to the 
period before the great war. 

In the time of the Buddha, Kosala was the predominant king- 
dom in Northern India, but it was already being eclipsed by the 
growing power of Magadha, Such incidents in its history as can 
be recovered from early Buddhist literature have been narrated in 
Chapter VII (pp. 1785*.). 

The Puranic list of Ikshvaku kings in the Kali Age concludes 
with Sumitra, the fourth successor of Prasenajit, who was contem- 
porary with the Buddha. The royal houses of Piiru and Ikshvaku, 
the sovereigns of Agra and Oudh, thus disappear from the scene at 
about the same time (p. 308). Henceforth the historical interest of 
the Puranas centres in Magadha which had become the suzerain 
power in the Middle Country ^ 

The Magadhas, who inhabited the Patna and Gaya Districts of 
S. Bihar, are unknown by this name to the Rigveda ; but, together 
with their neighbours, the Angas, in the Districts of Monghyr and 
Bhagalpur, they are mentioned in the Atharvaveda as a people 
living on the extreme confines of Aryan civilisation. Their kings 
claimed to be Piirus : they traced their descent from Kuru through 
the great conqueror, Vasu Chaidya^, whose son, Brihadratha, was 
the founder of the dynasty which is known by his name. 

1 Vedic Index, i, pp. 75, 190, 491 ; n, p. 506 ; Pargiter, J.R.A.S., 1910, pp. 27, 
29 ; Kali Age, pp. 9, 66. * Possibly the Ka9a Chaidya of Rigveda, vni, 5, 37. 

3IO The Puranas [ch. 

Magadha is the most famous kingdom in ancient and medieval 
India. Twice in history did it establish great empires — the Maurya 
Empire in the fourth and third centuries b,c., and the Gupta Empire 
in the fourth and fifth centuries A.D. The long line of kings attri- 
buted to Magadha by the Puranas consists of a series of no fewer 
than eight dynastic lists furnished with a statement of the number 
of years in each reign and the duration of each dynasty. If all these 
dynasties could be regarded as successive, and if the length of reigns 
could be determined with certainty, the chronology of Magadha 
would be a simple matter of calculation. But this is not the case. 
Some of the royal families included in the series were undoubtedly 
contemporary, and the text of the Puranas has become so corrupt 
that the numbers as stated by the different mss. are rarely in 

Brihadratha himself and nine of his successors are supposed to 
have reigned before the Kali Age. It is recorded that, when Saha- 
deva, the last of these, was slain in the great war, Somadhi, his 
heir, became king in Girivraja, 'the fortress on the hill,' at the 
foot of which the old capital of Magadha, Rajagriha, grew up. 
The site is marked by the ruined town of Rajgir in the Patna Dis- 
trict. In the reign of Senajit, Somadhi's sixth successor, most 
of the Puranas claim to have been recited. . No other event is 
connected with the twenty-one successors of Sahadeva. 

The next two dynasties, the Pradyotas and (^i9unagas, were 
almost certainly contemporary. The Pradyota dynasty may be 
identified with the Paunika family mentioned in the Harshacharita 
(trans. Co well and Thomas, p. 193). According to the Puranas, the 
founder, Punika (Pulika), slew his master, Ripuiijaya, the last of 
the Brihadrathas, and anointed his own son in his stead. After 
five reigns, the duration of which is given by some versions as 
52 years and by others as 138 years, tlie Pradyota dynasty is sup- 
planted by ^igunaga, who, after placing his son on the throne of 
Kagi (Benares), himself takes possession of Girivraja. 

But this is history distorted. Some editor has evidently placed 
independent lists in a false sequence and supplied appropriate 
links of connexion. This is clear from the evidence of Buddhist 

The Pradyotas were kings of Avanti (W. Malwa) and their 
capital was Ujjain. Pradyota (Pajjota) himself, like Bimbisara 
and Ajatagatru (Ajatasattu), the fifth and sixth in the list of 
^igunagas, and like the Puru Udayana (Udena) of Vatsa (Vamsa) 
and the Ikshvaku Prasenajit (Pasenadi) of Kosala, was contem- 

XI 1 1] Avanti: L,ater Qiqunagas 311 

porary with the Buddha \ The first of the Pradyotas, and the 
fifth and sixth of the 9i9unaga8, who are separated by more than 
150 years at the least according to the Puranas^, were therefore 
ruling at the same period in different countries. 

That the Pradyota of the Puranas and the Pradyota of Ujjain 
were one and the same person does not admit of question. The 
fact is implied in the statement of the Matsya Pur.^, and is clear 
when the Puranas are compared with other Sanskrit literature. 
Udayana, the king of Vatsa, is the central figure in a large cycle 
of Sanskrit stories of love and adventure ; and in these Pradyota, 
the king of Ujjain, the father of the peerless Vasavadatta, plays 
no small part. In some of the stories he appears also as the father 
of Palaka and the grandfather of Avantivardhana*. Now of the 
five members of the dynasty in the Puranas the first two are Pra- 
dyota and Palaka {v.l. Balaka), and the last is probably Avantivar- 
dhana ; for the various readings of the mss., as given by Mr Pargiter 
{Kali Age, p. 19), indicate that this may be the correct form of the 
name which appears in his text as Nandivardhana. 

This intrusion of kings of Avanti in the records of Magadha 
is probably to be explained, as in the similar case of the Andhras 
(p. 318), as the result of a suzerainty successfully asserted by Avanti ; 
and this may have been the outcome of the attack on Ajata^atru 
which Pradyota was reported to have been contemplating shortly 
before the Buddha's deaths If so, the supremacy of Avanti, which 
may have been temporary, was not established until some years 
after the beginning of Ajata9atru's reign, and the Pradyotas of the 
Puranas were contemporary with the later Ql^unagas — Ajata9atru, 
Dar^aka, and Udayin. 

It is only when we come to the reigns of Bimbisara and Ajata- 
9atru in the ^i9unaga dynasty that we find the firm ground of 
history. At this period lived Mahavira and Buddha, the founders, 
or perhaps rather the reformers, of Jainism and Buddhism; and 
now the Puranas are supplemented by two other lines of tradition 
which are presumably independent In the Jain accounts Bim- 
bisara appears as (^renika and Ajata9atru as Kunika : the former 
began the expansion of Magadha by the conquest of the kingdom 
of Anga (Monghyr and Bhagalpur), and the latter is said to have 
come to the throne after the death of Mahavira and a few years 
before the death of Buddha. 

» See Chapter vii, pp. 180, 183, 185, 187. ^ Xgii Age, pp. 18-21, 68-9. 

* Mr Harit Krishna Deb in Udayana Vatsardja (Calcutta, 1919), p. 4. 

* Lac&te, Gunddhya et la Brhatkatha, p. 154. 5 Chapter vn, p. 185. 

312 The Pur anas [ch. 

Unfortunately on one important point the three sources of 
information are not in agreement. The first eight kings in the 
Puranic genealogy may be arranged into two groups, the first 
headed by ^i9unaga and the second by Bimbisara. This arrange- 
ment is reversed in the Buddhist lists, while ^i9unaga's group is 
omitted altogether by the Jains. It is difficult to see how the 
three traditions, each of which has its champions among modern 
scholars, can be reconciled. 

The Brahman and Buddhist books record the length of the 
reigns of Bimbisara and Ajata9atru ; but they are not in agreement 
with one another, and moreover the Brahman accounts are not con- 
sistent. In the present corrupt condition of the text the various 
MSS. of the Puranas attribute a reign of either 28 or 38 years to 
Bimbisara, and one of 25, 27, or 28 years to Ajata9atru {Kali Age, 
p. 21). Until the text has been restored by critical editing the 
authentic tradition of the Brahmans cannot be ascertained. In 
contrast with this discrepancy the Buddhist chronicles of Ceylon, 
the Dipavamsa and the Mahavamsa, oflfer a consistent and more 
detailed account of these reigns and of certain important events 
in the lifetime of Siddhartha, the (^akya prince who became the 
Buddha. Whether this tradition is to be accepted as correct in 
preference to the other may be questioned ; but it affords the best 
working hypothesis which has yet been discovered. The chronology 
as determined by Prof. Geiger in the introduction to his translation 
of the Mahavamsa (pp. xl-xlvi) may be tabulated as follows : 

<^i<pinaga Kings. 

Siddhartha {the Buddi 


Bimbisai-a's birth 

558 B.C. 


563 B.C. 

„ accession . . . 

543 „ 

Leaves his father's house... 

534 „ 

„ death 

491 „ 

Becomes Buddha 

528 „ 

Ajatafatru's accession... 

491 „ 

Meets Bimbisara (for the 

second time) ^ 

528 „ 

„ death 

459 „ 

Attains mr??awa 

483 „ 

After these two reigns we come once more to a period of 
conflicting authorities and chronological uncertainty which lasts 
until, the reign of Chandragupta. The Buddhist genealogy pre- 
served in the Mahavamsa is certainly not above suspicion^; for 
each of the five kings from Ajatasattu to Nagadasaka is said to 
have kiUed his father and predecessor within a period of fifty-six 
years, and we are solemnly told that, after the last of these, Naga- 
dasaka, had occupied the throne for twenty-four years, the citizens 
awoke to the fact that 'this is a dynasty of parricides' and appointed 

1 See Chapter vn, pp. 183 f. « Chapter vn, pp. 189 f. 

XI 1 1] Nandas 313 

the minister Susunaga ((^icunaga) in his stead The Jain tradition 
recognises only Udayin and the nine Nandas as reigning during 
this interval ; and the Puranic list {Kcdi Age, pp. 21-6, 68-9) is 

as follows : 

Dargaka reigned 24, 25, or 35 years 
Udayin „ 33 „ 

Nandivardhana reigned 40, or 42 years 
Mahanandin „ 43 „ 

Mahapadma „ 28, or 88 „ 1 ™ , , ,„^ 

His eight sons „ 12 „ } Total, 100 years. 

Dar9aka appears not to be mentioned by the Buddhist writers, 
unless indeed he is to be identified with Nagadasaka whom they 
place before Udayin (Udayi-bhadda) ; but he is known to Sanskrit 
literature as a king of Magadha and the brother of Padmavati, the 
second queen of Udayana, king of Vatsa\ Udayin, or Udayi- 
bhadda, is known to all the three traditions. To him the Brahmans 
and Jains attribute the foundation of Kusumapura on the south 
bank of the Ganges. The new city, which was either identical with 
the later Pataliputra or in its immediate neighbourhood, was built 
near the fortress which Ajata^atru had established at the village 
of Patali as a protection against the Vajjian (Vriji) confederacy of 
Licchavis, Videhas, and other clans of N. Bihar. The foundation 
of Pataliputra is ascribed by the Buddhists to Kalasoka. 

The ten 9i9unaga kings are expressly called Kshatriyas by the 
Puranas, but the last of these, Mahanandin, became, through his 
marriage with a (^udra woman, the founder of a Qudra dynasty 
which endured for two generations — Mahapadma and his eight 
sons. One of the latter, usually supposed to be named Dhana- 
nanda, was on the throne in 326 B.C., when Alexander the Great 
was obliged by the unwillingness of his army to abandon his 
scheme of attacking the Prasioi, or 'eastern nations,' then united 
under the suzerainty of Magadha. Within a few years of 
Alexander's retirement from India, this suzerainty passed from 
the Nandas to the Mauryas, probably c. 321 B.C. 

The period of the nine Nandas is thus determined. According 
to the Puranas they represent no new family : they are the direct 
descendants of the 9i9unagas, the last and the last but one of 
whom, Mahanandin and Nandivardhana, bear names which indicate 
their connexion. There are, therefore, two groups of these kings, 
which seem to be distinguished in literature as the 'old' and the 
*new' Nandas; and, as Mr Jayaswal has suggested, 'new' and not 

* Svapnavdtavadatta, Act. i (eA Trivandrum Series, pp. 4, 5). 

314 The Puranas [ch. 

'nine' may have been the correct designation of the later group ^ 
The Puranas know no break of political continuity between the 
9i9unagas and the Nandas ; but they recognise that a great social 
and religious gulf has been fixed between the earlier and the later 
Nandas by the flagrant violation of caste law which placed Maha- 
padma,the son of a ^udra woman, on the throne ; and they mark their 
sense of this chasm by interpolating after the reign of Mahanandin 
a summary of the number of reigns in other contemporary dynasties 
before proceeding with their account of the rulers of Magadha. 

As to the origin of the Nandas we have no certain information ; 
but the name is probably tribal, and it may be connected with the 
Nandas who lived near the river Ramganga, between the Ganges 
and the Kosi in the Himalayan region of the United Provinces*. 
The countries of the Himalayan fringe at this period were occupied 
by innumerable clans governed by tribal constitutions which may 
best be described as aristocratic oligarchies. Like the Rajputs, 
they were conquerors ruling in the midst of subject peoples ; and, 
as Dr Vincent Smith has suggested^, many of these clans may 
have been of Tibeto-Chinese origin. It is possible that the 
^i9unagas and Nandas may have been the descendants of moun- 
tain chieftains who had won the kingdom of Magadha by conquest. 

A Nanda king is twice mentioned in the Hathigumpha inscrip- 
tion of king Kharavela of Kalinga (Orissa). The inscription, which 
is a record of events in thirteen (or fourteen) years of the king's 
reign, has been badly preserved. Considerable portions have been 
lost, and both the reading and the interpretation of many passages 
are uncertain. The record in its present state can only be used as 
a basis for history with the utmost caution. It is clear, however, 
that in his fifth year Kharavela executed some public work which 
was associated with the memory of king Nanda ^, and that in his 
twelfth year he gained a victory over the king of Magadha and, 

1 Jour. Bihar and Orissa Research Soc, September 1915, p. 21. 

2 Pargiter, Mark. Pur., pp. 292, 383. » Oxford History of India (1919), p. 49. 
* The different versions of this passage in line 6 of the inscr. depend chiefly, 

though not solely, on the translation of ti-vasa-sata. The following renderings have 
been proposed : 

(1) ' He opened the three-yearly almshouse of Nandaraja ' (Pandit Bhagvanlal 
Indraji, Trans. Inter. Or. Cong., Leiden, 1884, Part 3, p. 135. Sata = sattra or gatra, 
cf. Ep. Ind., X, Appendix, no. 967, p. 100, and no. 985, p. 102) ; 

(2) ' He had an aqueduct, that had not been used for 103 years since king Nanda 
(or since the Nanda kings), conducted into the city' (Prof. Liiders, Ep. Ind., x. 
Appendix, no. 1345, p. 161. Sata=gata, as also in the next translation) ; 

(3) ' He brings into the capital... the canal excavated by king Nanda three centuries 
before ' (Mr J. P. Jayaswal and Mr R. D. Banerji, Jour. Bihar and Orissa Research Soc, 
Dec. 1917, pp. 425 ff.). 

XI 1 1] Growth of Magadha : Other Powers 315 

according to Mr Jayaswal's translation^, recovered certain trophies 
which had been carried away by king Nanda. 

These statements of the inscription, coupled with the somewhat 
enigmatical testimony of an ancient Sanskrit MS. quoted by Mr 
JayaswaP, seem to show that Kalinga had been conquered by one 
of the Nanda kings and lost by another. Kalinga was undoubtedly 
conquered by A9oka, the third of the Maurya emperors, c. 262 RC' 
We must infer, therefore, either that it was not included in the 
dominions of the first two emperors, Chandragupta and Bindusara^ 
or that it had revolted and was reconquered by A9oka. 

Certain stages in the growth of the power of Magadha from 
its ancient stronghold in the fortress of Girivraja may thus be 
traced. The expansion began with the conquest of Anga (Monghyr 
and Bhagalpur in Bengal) by Bimbisara, c. 500 RC. The establish- 
ment of a supremacy over Ka9i (Benares), Kosala (Oudh), and 
Videha (N. Bihar) was probably the work of his son and successor, 
Ajata^atru, in the fii*st half of the fifth century. Kalinga (Orissa) 
was, perhaps, temporarily included in the empire as a result of its 
conquest by a Nanda king. It remained for Chandragupta to 
extend the imperial dominions by the annexation of the north- 
western region which for a few years had owned the sway of 
Alexander the Great and his satraps, and for A9oka to conquer, 
or reconquer, Kaliflga. 

The summary of reigns, which comes in the Puranas between 
the description of the earlier and later Nandas, has reference to 
ten dynasties in Northern and Central India which were con- 
temporary with the kings of Magadha. It is a bare list of. names 
and numbers without any orderly arrangement, and, as usual, the 
numbers given by the different MSS. are not consistent. The 
summary may be rearranged geographically as follows (cf. Kali 
Age, pp. 23-4, 69). 

( United Provinces : Agra) 

1. Kurus : 36 (19, 26, 30, or 50) reigns. 

2. Panchalas : 27 (25) „ 

3. ^urasenas: 23 „ 

4. Kaqis : 24 (36) „ 
{United Provinces : Oudh) 

5. Ikshvakus : 24 „ 

10. Kalingas : 32 (22, 24, 26, or 40) „ 

1 Op. eiU, pp. 447, 464-5. ' Ibid. p. 482. 

3 Chapter xx, p. 495. 

(Central India and Gujarat) 



Haihayas 28 (24) reigns 
Aqmakas : 25 „ 
Vltihotras: 20 „ 


(N. Bihar) 
Mithilas : 28 (18) „ 



The Pur anas [ch. 

1. The Kurus are no doubt the Parus of the detailed list ; but the number of 
reigns differs. 

2. The Pafichalas, a confederation of five tribes, were neighbours of the Kurus. 
The capital of N. Pafichala was Ahicchatra, now a ruined site still bearing the 
same name near the village of Ramnagar in the Bareilly District. The capital of 
S. Pafichala was Kampilya, now represented by ruins at the village of Kampil in 
the Farrukhabad District. 

3. The peoples living to the south of Kurukshetra claimed descent from 
YadiL Of these the ^urasenas occupied the Muttra District and possibly some of 
the territory still farther south. This capital was Muttra (Mathura), the birth- 
place of the hero Krishna. 

To the west of the ^urasenas dwelt the Matsyas. The two peoples are constantly 
associated, and it is possible that at this time they may have been united under 
one king. The Matsyas occupied the state of Alwar and possibly some parts of 
Jaipur and Bhartpur. Their capitals were Upaplavya, the site of which is uncer- 
tain, and Vairata, the city of king Virata, the modem Bairat in Jaipur. 

4. The little kingdom of Ka§i (Benares) was bordered by Vatsa on the west, 
Kosala on the north, and Magadha on the east. Some details of its relations with 
these countries may be recovered from early literature. According to the 
(^atapatha Brahmana (xiii, 5, 4, 19), its king Dhritarashtra was conquered by the 
Bharata prince (^atanika Satrajita (p. 308). Such incidental notices of its later 
history as have been preserved by Buddhist writers have been collected in 
Chapter vii, pp. 180 flF. 

At different periods KaQi came under the sway of the three successive suzerain 
powers of Northern India — the Purus of Vatsa, the Ikshvakus of Kosala, and the 
kings of Magadha j but it seems to have enjoyed its period of independent power 
in the interval between the decline of Vatsa and the rise of Kosala, when king 
Brahmadatta, possibly about a century and a half before the Buddha's time, con- 
quered Kosala. The fame of Brahmadatta has been kept alive in Buddhist 
literature ; for in his reign the Jatakas, or stories of the Buddha in previous 
births, are conventionally set. 

,:, The account given in the Puraijas of the accession of ^iQunaga to the throne of 
Magadha shows that this king was associated also with Ka§I (p. 310). 

5. The number of Ikshvaku kings given in the summary is 22. This is not 
in accordance with the detailed list which (pp. 308 f ) contains 30. 

6. 7, 8. The Haihayas, AQmakas, and Vltihotras, like the ^urasenas, belonged 
to the great family of the descendants of Yadu who occupied the countries of the 
river Chambal in the north and the river Narbada in the south ; but it is difficult 
to identify with precision the kingdoms indicated by these different names. 
Haihaya is often used almost as a synonym of Yadava to denote the whole group 
of peoples ; and the Vitihotras are a branch of the Haihayas. Both the Vlti- 
hotras and the Agmakas are closely associated in literature with the Avantis of 
W. Malwa, whose capital was Ujjain (Ujjayini) on the Sipra, a tributary of the 
Chambal (Charmaijvatl)^. 

It would be strange if the rulers of a city so famous both politically and com- 
mercially as Ujjain should have found no place in this summary. The most 
plausible explanation of their apparent absence from the list is that they are here 
called Haihayas. 

^ For these peoples, see Pargiter, Marli. Pur., pp. 344-5, 371 ; J.Ii.A.S., 1914, p. 274. 

XI 1 1] Quhgas^ Kanvas^ and Andhras 317 

9. The Mithilas take their name from Mithila, the capital of the Videhas, one 
of the numerous clans, possibly of Tibeto-Chinese origin, who inhabited Tirhut (the 
districts of Champaran, Muzaffarpur, and Darbhanga in N. Bihar). Videgha 
Mathava, to whom the Brahmanisation of this region is attributed by the 
^atapatha Brahmana (». sup. p^l22) is probably its earliest recorded monarch. 
According to the Puranas the Aryan kings of the Videhas were a branch of the 
Puru family. They are derived from Mimi, the son of Ikshvaku and the remote 
ancestor of Siradhvaja Janaka, the father of Sita, the heraine of the Ramayaija. 
Like Rama himself, he is supposed to have lived before the Kali Age. It is 
possible that he may be the King Janaka of Videha who is celebrated in the 
Brahmanas and Upanishads ; and, if so, the story of the Ramaya^a has its origin 
in the later Brahmaria period. In the time of the Buddha, the Videhas together 
with the Licchavis of Vaigall (Basarh in the Hajlpur sub-division of Muzaffarpur) 
and other powerful clans formed a confederation and were known collectively by 
their tribal name as the Vrijis (Vajjis). The reduction of their power marks an 
epoch in the expansion of the kingdom of Magadha^ 

10. In the Puranas the monarchs of the five kingdoms of Aiiga (Monghyr and 
Bhagalpur), Vanga (Birbhum, Murshidabad, Bard wan and Nadia), Pun^ra (Chota 
Nagpur), Suhma (Bankura and Midnapur), and Kalinga (Orissa) are derived from 
eponymous heroes who are sxipposed to be brothers belonging to the family of 
Anu''. With the exception of Ahga, none of these kingdoms is mentioned in early 
literature. The earliest monument which throws light on the history of Kalinga 
is the Hathigumpha inscription of Kharavela (». sup. pp. 314 f.) 

After this summary the royal genealogies are resumed, and 
detailed lists of the later Nandas, the Mauryas, the 9'tiwgas, the 
Kanvas, and the Andhras follow^. The continuous record then 
ceases ; but genealogies more or less fragmentary and summaries 
of ruling powers, both native states and foreign invaders, continue 
to appear until about the end of the fifth century A-D. when the 
Puranas cease to be historical. 

The five dynasties just mentioned are, as usual, regarded as 
successive ; but this can only be true of the Nandas, Mauryas, and 
^ungas. The (^^ungas, Kanvas, and Andhras were contemporary, 
although no doubt they claimed the suzerainty of N. India suc- 
cessively. That the first two of these were ruling at the same time 
may be inferred from the incidental statement that the first Andhra 
king destroyed the last of the Kanvas and * what was left of the 
^ungas' power' {Kali Age, pp. 38, 71). But it is certain that the 
^ungas were flourishing after the reign of the first Andhra king. 
Both powers, (^Junga and Andhra alike, arose on the ruins of the 
Maurya empire — the former in the Midland Country and the latter 
in Southern India. It was probably not until the reign of the 

1 Vedic Index, i, pp. 271-3 ; ii, p. 298 ; Pargiter, J.R.A.S., 1910, pp. 19, 27, 29 ; 
Rhys Davids, Buddhist India, pp. 25-6, 40-1. 

2 Pargiter, Mark. Pur., pp. 324-9, 334. 

* For the history of these dynasties, so far as it comes within the limits of the 
present volume, see Chapters xviii-xxi, and xxiv. 

3i8 The Puranas [ch. xiii 

third Andhra king, Qatakarni, that they came into collision ; and 
then their political association appears to have been transient. 

The Puranas, however, state or imply that ten (^uiiga kings, 
reigning for 112 years, were succeeded by four Kanvas, who reigned 
for 45 years, and that then the first of the Andhras, Simuka, having 
wrested the kingdom from the last of the Kanvas, Su9armau, 
became the founder of a dynasty of thirty kings who ruled over 
Magadha during a period of 460 years. This is manifestly in- 
correct. It is evident that by piecing together three separate 
lists some editor has constructed an entirely false chronology and 
has perverted historj^ The Andhras had probably no connexion 
with Magadha. Their only possible claim to a place in its records 
must have been founded on a conquest which transferred to them 
the suzerainty previously held by Magadha \ 

In order to understand the situation we must consider what the 
consequences of a triumph of this kind must have been. Under 
the Nandas and the Mauryas Magadha had established a suzerainty 
which passed by conquest to the first ^unga king, Pushyamitra, 
and was solemnly proclaimed by his performance of the 'horse- 
sacrifice' {a^vamedha)\ This suzerainty, and with it the proud 
title of chafcravartin, 'universal monarch,' was contested success- 
fully by the Andhra king who, as is known from the Nanaghat 
inscription of his queen, Naganika, celebrated the A9vamedha on 
two occasions^ ; and, as we have seen (p. 302), there is good reason 
for believing that the genealogies preserved in the Puranas have 
their origin in the proclamation of the king's lineage which accom- 
panied the performance of this sacrifice. 

The rank of a chakravartin must, at this period, have conferred 
on his family an hereditary distinction which entitled all his suc- 
cessors to be commemorated in the records of Magadha. Imperial 
and royal dignities of this kind, when once established, are not 
readily abandoned, however shadowy and unreal they may have 
become. It must be remembered that the sovereigns of our own 
country continued to use the title and the arms of France until 
the beginning of the nineteenth century, nearly two centuries and 
a half after the loss of Calais, the last of their French possessions. 
Regarded as historical documents, the British coin-legends of the 
eighteenth century, with their purely hereditary titles, are as mis- 
leading as the Puranas, which, arranging all in one long series, 
ascribe to Magadha both its own kings and the families of the 
suzerains of Northern India. 

» Chapter xxiv, p. 600. * Chapter xxi, p. 520. 

3 Biihler, Arch. Sur. West. Ind., v, pp. 60 ff. 



The connexions between Persia and India date back to the 
gi'ay dawn of the period of Indo-Iranian unity, when the Aryan 
ancestors of the Hindus and Persians still formed an undivided 
branch of the Indo-European stock. Though the separation of 
these two kindred peoples, through their migrating into the respec- 
tive countries they have occupied in historic times, must have 
taken place more than three thousand years ago, nevertheless there 
long remained a certain community of interest, which had a bearing 
upon the early history of the north of India, where Persian influence, 
and even dominion, was strongest. The aim of the present chapter, 
therefore, is to bring out the main points of contact between the 
two nations from the earliest times and to indicate the eflfect of 
the sway exercised by Persia in Northern, or rather North-western, 
India prior to the invasion of Alexander the Great and the fall of 
the Achaemenian Empire of Iran in the latter part of the fourth 
century b.c. 

To begin the sketch with the most remote ages, it may be 
assumed that every student is familiar with the evidence that 
proves the historic relationship between the Hindus and the Per- 
sians through ties of common Aryan blood, close kinship in language 
and tradition, and through near affinities in the matter of religious 
beliefs, ritual observances, manners, and customs. 

An illustration or two may be chosen from the domain of reli- 
gion alone. The Veda and the Avesta, which are the earliest 
literary monuments of India and Persia, contain sufficient evidence 
of the fact of such connexion, even though each of these works 
may date from times long after the period of Indo-Iranian separa- 
tion. A certain relationship, for example, is acknowledged to exist 
between the Vedic divinity Varuna and the Avestan deity Ahura 
Mazda, or Ormazd, the supreme god of Zoroastrianism. Equally 
well known are the points of kinship between the Indian Mitra and 
the Iranian Mithra, and, in less degree, between the victorious 

320 Persian Dominions in N. India [CH. 

Indra Vritrahan of the Rigveda and the all-triumphant Vere- 
thraghna of the Avestan Yashts. Nor need more than mention be 
made of the parallels between Yama and Yima or of the cognate 
use made by the Indians and the Persians of the sacred drink soma 
and haoma in their religious rites. Scores more of likenesses and 
similarities might be adduced to prove the long-established con- 
nexion between India and Iran, but they are generally familiar \ 

Additional evidence, however, has comparatively recently been 
furnished by certain cuneiform tablets which the German professor 
Hugo Winckler discovered, in 1907, at Boghaz-koi in North-eastern 
Asia Minor. These documents give, in their own special language, 
a record of treaties between the kings of Mitani and of the Hittites 
about 1400 B.C. Among the gods called to witness are deities 
common in part to India and Persia, whatever the relation may be. 
The names involved in the tablets are Mi-it-ra, U-ru-w-na, In-da-ra, 
and Na-sa-at-ti-ia, corresponding respectively to Mitra, Varuna, 
Indra, and Nasatya (the latter regularly a dual in the Veda, and 
representing the two A9vins) in the Indian pantheon. They answer 
likewise in due order to the Persian Mithra and to those elements 
common between the Zoroastrian god Ahura Mazda and the Vedic 
Varuna, as explained above ; but on the other hand Avestan Indra 
and Naonhaithya (a singular in Av., Vd. x, 9 ; xix, 43) appear as 
demons in the Zoroastrian scriptures. It is not the place here to 
enter into a discussion of the question as to whether the super- 
natural beings thus mentioned in the Boghaz-koi clay tablets are 
to be interpreted as being * proto-Iranian,' 'Vedic,' 'Aryan,' or 
even * Mitanian ' alone, because the matter is still open to debate 
by scholars. It is sufficient to draw attention to the general 
bearings of such a discovery upon the subject of relationship be- 
tween India and Persia, however direct or indirect the connexion 
may be^ 

^ A convenient summary of these now familiar facts will be found in F. Spiegel, 
Die arische Periode, Leipzig, 1887. Throughout the present chapter the terms • Iran ' 
and ' Iranians ' are to be taken broadly, so as to comprehend Persia and its people in 
the widest significance — whether Medes, Persians, or Bactrians — as forming a special 
division of the Indo-Iranian branch of the great Indo-European, or Indo-Germanic, 
stock. The designation ' Aryan ' should really be restricted (as is done by scholars) 
to the common bond represented historically by the Hindus and the Persians. 

s This valuable find of the tablets by Winckler (who died April 19, 1913) was first 
reported in his Vorlaufige Nachrichten iiber die Ausgrabungen in Boghaz-koi im Sommer 
1907, in Mittheilungen der deutschen Orientgesellschaft, No. 35 (1908). The import- 
ance of the discovery was at once recognised by scholars and has since received wide 
attention j compare, for example, Eduard Meyer, Zt. filr vergleichende Sprachforachung, 
Neue Folge (1908), xui, 1-27 ; idem, Sitzb. d. kgl. preuss. A had. d. Wiss., 1908, pp. 14-19 ; 
also H. G. Jacobi, J.R.A.S., 1909, pp. 721-726 ; H. Oldenburg, ibid. pp. 1095-1100 ; 

xiv] Common Indo- Iranian Domains 321 

The geographical connexion between India and Persia histori- 
cally was a matter of fact that must have been known to both 
countries in antiquity through the contiguity of their territorial 
situation. The realms which correspond to-day to the buffer states 
of Afghanistan and Baluchistan formed always a point of contact 
and were concerned in antiquity with Persia's advances into 
Northern and North-western India as well as, in a far less degree, 
with any move of aggrandisement on the part of Hindustan in the 
direction of Iran^. Evidence from the Veda and the Avesta alike 
attests the general fact. 

Vedic scholars, for example, will agree with Avestan students 
that the partly common Indo-Iranian domains comprised in the 
river-system above the Indus basin, and verging toward the north- 
western border adjacent to Iran, are referred to in the Rigveda in 
certain allusions to the district indicated by the rivers Kubha 
(Kabul), Krumu (Kurram), and Gomati (Gumal). They will equally 
unite in emphasising the fact that there are other incidental allusions 
in the Veda, such as those to Gandhara and Gaudhari, which may 
certainly be interpreted as referring to the districts of Peshawar 
and Rawalpindi S.E. from KabuP. A part of these districts 
has belonged rather to Iran than to India in historic times, but it 
is equally impossible to deny or to minimise the role they have 
played in India's development ever since the remote age when the 
tribal ancestors of the present Hindus occupied them on their way 
into their later established horned For the earliest period, we 

A. B. Keith, ihid. pp. 1100-1106; A. H. Sayce, iUi. pp. 1106-1107; J. Kennedy, 
ibid. pp. 1107-1119; H. G. Jacobi, ihid. 1910, pp. 456-464; A, B. Keith, ibid. 
pp. 464-466 ; H. Oldenburg, ibid. pp. 846-850 ; see also M. Winternitz, Globus 
(1909), xcv, 126 ; Macdonell and Keith, Vedic Index, i, p. viii ; and most recently J. H. 
Moulton, Early Zoroastrianism, pp. 5-7, 45, 139, 235 ; Eduard Meyer, Geschichte des 
Altertums, 3rd ed., vol. i, pt 2, §§ 455, 585, 590. 

1 Arrian, Indica, 9, 12, for example, may be cited in support of this statement ; for he 
avers, on Indian authority, that ' a sense of justice, they say, prevented any Indian king 
from attempting conquest beyond the limits of India.' The assertion certainly seems 
true for the earliest times. 

2 For references to passages in the texts and for bibliographical allusions consult 
Macdonell and Keith, Vedic Index, i, 162 (Kubha), 199 (Krumu), 238 (Gomati), and 
218-219 (Gandhara, Gandhari). In regard to the territory to be located by the Vedic 
river Sarasvati, the present tendency among Sanskrit specialists (most recently, for 
example, Macdonell and Keith, op. cit. n, 434-437) is to confine it to India itself and 
not to follow the suggestions that have been made, on etymological grounds, to connect 
the region thus watered by the Sarasvati directly with the region around the Iranian 
river Harahvaiti of the Avesta, or Hara(h)uvatl of the Old Persian Inscriptions, as a 
designation of the ancient land of Arachosia. 

* The student of history, with an eye to the significance of territorial location, will 
at once recall the part played in after ages by Kabul as a strategic centre, and as the 
doorway into India from the north, in the annals of Hindustan. 

C.H.I. I. 21 

322 Persian Dominions in N, India [ch. 

may well agree with the opinion expressed by Eduard Meyer in 
an encyclopaedia article on Persia : ' The dividing line between 
Iranian and Indian is drawn by the Hindu-Kush and the Soliman 
mountains of the Indus district. The valley of the Kabul {Cophen) 
is already occupied by Indian tribes, especially the Gandarians ; 
and the Satagydae (Pers. Thatagu) there resident were presumably 
also of Indian stock '^. These facts, because of their importance in 
regard to this bridge between India and Iran, will be touched upon 
again below (pp. 338-9). 

Regarding the interpretation of certain other references in the 
Eigveda as containing allusions, direct or implied, to Persia in a 
broader sense, there is a wide divergence of opinion among San- 
skritists, even though the Iranian investigator may feel assured of 
the truth of so explaining such passages. Vedic specialists are at 
variance, for example, as to whether an allusion to the Parthavas 
in Rv. VI, 27, 8, is to be understood as a reference to the ancestors 
of the Parthians, and as to whether the Persians are really referred 
to under the designation Par9avas (e.g. Rv. x, 33, 2), especially as 
the diflBculty is increased by the uncertainty in determining the 
real significance historically of the names Prithu and Parcu from 
which the terms Parthavas and Par9ava8 are derived. The name 
Balhika (Atharvaveda, v, 22, 5, 7, 9) has been interpreted by some 
Indie scholars as containing an allusion to the ancient Iranian tribe 
of the Bactrians, especially because it is mentioned in connexion 
with the Mujavants, a northern people ; but other specialists oppose 
this view and deny an appeal to certain other Vedic words that 
might be cited. Nevertheless, and in spite of the diflferences among 
Sanskrit authorities, there is more than one Iranian investigator 
who feels positive that some at least of the Rigveda references in 
question allude to Persia or to Persian connexions in by-gone days. 
The assumption may reasonably be made that scholarship in the 
future will tend to prove the correctness of the attempts (wide of 
the mark though some of them may have been in the past) to show 
through the Veda the continuity of contact between India and 
Persia during the period under consideration^. 

From the Iranian side, if we may judge by the sources available, 

1 Encyclop. Brit., 11th ed., xxi, 203, art. 'Persia.' 

2 For complete references to the Vedic passages involved in the discussion, including 
full bibliographical citations, see Macdonell and Keith, op. cit. i, 29 (Abhyavartin), 
347-9 (Dasyu), 450 (Ninditacjva), and especially 504-5 (Parpu), 521-2 (Parthava) ; ii, 
63 (1. Balhika). Macdonell and Keith join with those Sanskrit scholars who oppose 
the attempt to find any allusions to Iran in the Veda. The extravagant endeavours of 

xiv] Evidence of Veda and Avesta 323 

the evidence seems to be much stronger in favour of Persian 
influence upon India and modifying control over the northern 
part of the country than it is for a reverse influence of India upon 
Iran. Throughout ancient history, as indicated above (p. 321), 
Persia was the more aggressive power of the two. Yet it is un- 
certain how far the sphere of Iranian knowledge and authority in 
India may have extended prior to the time of the Achaemenian 
Empire, at which era our information takes on a more definite 
form. At no time, however, does the realm of Persian activity 
in this direction appear to have extended much beyond the limit 
of the Indus. 

As already intimated, the Avesta is in general the oldest source 
showing Persia's interest in India, although the greatest uncertainty 
still prevails among specialists in regard to assigning any precise 
date or dates. The present writer shares the opinion of those 
scholars who believe that, however late may be some of its portions, 
the Avesta in the main is pre- Achaemenian in content ; in other 
words, even though it is possible to recognise Achaemenian, Par- 
thian, and, perhaps, Sassanian elements in the collection, the general 
tenor of the work and the material on which it is based represent 
a period antedating the fifth century B.C., or the era when the 
Persian Empire reached its height '^. For that reason (and with 
due emphasis on the broad latitude that is to be allowed 
in the matter of dates) it is appropriate to cite the Avestan 

Brunnhofer, Urgeschichte der Arier, 3 vols, Leipzig, 1893, to identify every remote 
Yedic term that had a possible geographical content as an Iranian allusion are bizarre 
in the extreme, even though there are grains of truth in the author's views when he 
touches more conservatively on the domain bordering between India and Iran. The 
writer of the present chapter sympathises strongly with certain of the pleas made by 
the Vedic scholars Ludwig, Hillebrandt, and Weber to recognise Persian allusions in the 
Bigveda ; the titles of the special articles on the subject by these scholars are duly cited 
by Macdonell and Keith in the pages of their Vedic Index, referred to above. It seems, 
for example, that some Avestan student may yet make more use than has been done of 
the material collected by E. W. Hopkins, Prdgdthikdni, i, in J.A.O.S. 1896, xvii, 

^ For a convenient presentation of the various views regarding the date of Zoroaster 
and the age of the Avestan Gathas, as well as concerning the relative antiquity of other 
portions of the sacred canon, see J. H. Moulton, Early Zoroastrianism, London, 1913. 
Dr Moulton summarises his opinion as follows, on p. viii : * The traditional date [of 
Zoroaster] (660-583 b.c.) is a minimum, but there are strong reasons for placing 
Zarathushtra and his Qathas some generations earlier still. The Yashts may be 
placed in the later Achaemenian age, and the prose Avesta, in particular the ritual of 
the Vendidad, probably after Alexander. ' He elaborates this view further on pp. 8-22, 
78, 87, 103, 198, 204, 240. It is important throughout to bear in mind the fact that 
the material may sometimes be very old even though the/or7« is late, and that different 
chapters as well as sections of the Yashts, Vendidad, and Yasna may vary considerably 
in age. 


324 Persian Dominions in N. India [ch. 

references to India, or the region of the Indian frontier, 
directly after the possible allusions to Persia in the Veda already 

The name for India in the A vesta is Hindu, which, like the Old 
Persian Hi(n)du, is derived from the river Indus, Sanskrit Sindhu, — 
the designation of the stream being transferred to the territory 
adjacent to it and to its tributaries. The first chapter of the 
Avestan Vendldad (whatever may be the age of the chapter) con- 
tains an allusion to a portion of Northern India in a list which it 
gives of sixteen lands or regions, created by Ahura Mazda and 
apparently regarded as under Iranian sway^ The fifteenth of 
these domains, according to Vd. i, 18, was Hapta Hindu, 'Seven 
Rivers,' a region of ' abnormal heat,' probably identical with the 
territory of Sapta Sindhavas, 'Seven Rivers,' in the Veda (see 
especially Rv. viii, 24, 27)^. The district in question, which was 
more comprehensive than the modern Punjab, or 'Five Rivers,' 
must have included the lands watered in the north and north-west 
of Hindustan by the river Indus and its afiluents — answering, 
apparently, to the Vedic Vitasta (now Jhelum), Asikni (Chenab), 
Parushni (later named Iravati, whence its present designation Ravi), 
Vipa9 (Beas), and (^Jutudri (Sutlej), the latter being the easternmost 
stream ^ 

In connexion with this Avestan passage (Vd. i, 18), moreover, 
in its bearing on Persian domains in Northern India, it is worth 
while to call attention to the Pahlavi gloss of the Middle Persian 
rendering of the paragraph in Sassanian times. Whatever may be 
the full import of this difficult gloss, the passage may be literally 
translated as follows : ' The Seven Hindukan ; the expression 
"Seven Hindukan" is due to this fact, that the over-lordship 
{sar-xutdt) is seven ; and therefore I do not say " Seven Rivers," 
for that is manifest from the Avesta [passage] " From the Eastern 

1 One might be inclined (as the writer has been led, especially through a study of 
the Pahlavi commentary and other Sassanian sources) to regard Vd. i, though late in 
form, as containing older material that might antedate in substance the division which 
Darius made of his empire into twenty satrapies ; but Darmesteter warns against the 
attempts that have been made to discover much antique history in the chapter. His 
rather strong statement (Vendidad Translated, 2nd ed., S.B.E. iv, 1) is: ' We have here 
nothing more than a geographical description of Iran, seen from the religious point 
of view.' 

2 See Bartholomae, Altiranisches Worterbuch, col. 1814 ; Macdonell and Keith, 
Vedic Index, ii, 424; Hopkins, J.A.O.S. xvi, 278; xvii, 86-88. 

3 Cf. Spiegel, Die arische Periode, pp. 112-118 ; Macdonell, History of Sanskrit 
Literature, p. 140; see also above, p. 321, n. 2 (on the question of Sarasvati=Harab< 

xiv] T*he Eastern and Western Indus 325 

Indus (or, India) to the Western Indus (India)" \' In partial sup- 
port of the scholiast's interpretation as 'the over-lordship is seven' 
it has been further pointed out that a tradition as to the dominions 
involved may have lingered down to Firdausl's time, inasmuch as 
he mentions in one passage seven princes of India, namely the lords 
of Kabul, Sindh, Hindh, Sandal, Chandal, Kashmir, and Multan ; 
but too much stress need not be laid on the points 

The Avestan fragment above cited from the gloss to Vd. i, 18 — ■ 
* from the Eastern Indus (India) to the Western Indus (India) ' — is 
best interpreted as alluding to the extreme ends of the Iranian 
world ; for Spiegel has clearly shown by sufficient references that, 
at least in Sassanian times and doubtless earlier, there prevailed an 
idea of an India in the west as well as an India in the east^ This 
is borne out by a passage in Yasht x, 104, in which the divine 
power of Mithra, the personification of the sun, light, and truth, 
is extolled as destroying his adversaries in every quarter. The 
passage (Yt. x, 104), which is metrical and therefore relatively old, 
runs thus : * The long arms of Mithra seize upon those who deceive 
Mithra ; even when in Eastern India he catches him, even when 
in Western [India] he smites him down ; even when he is at the 
mouth of the Ranha (river), [and] even when he is in the middle 
of the earth ^' The same statement is repeated in part in Yasiia 
LVii, 29, regarding the power of Sraosha, the guardian genius of 
mankind, as extending over the wide domain from India on the 
east to the extreme west : ' even when in Eastern India he catches 
[his adversary], even when in Western [India] he smites him down.' 

There is still another Avestan allusion which may possibly be 
interpreted as referring in a general way to Indian connexions ; it 

^ For the Pahlavi text of the passage, and especially the variant readings, see 
the edition by D. D. P. Sanjana, The Pahlavi Vendiddd, p. 9, Bombay, 1895 ; and the 
earlier edition by F. Spiegel, Avesta sammt der Huzvdresch- Ubersetzung, vol. i, pt 2, p. 7, 
Leipzig, 1851. 

2 The passage, Firdausi, Shdhndmah, ed. Maoan, p. 1579, was pointed out by W. 
Geiger, Die pehleviversion des ersten Capitels des Vendidad (1887), p. 62, and likewise 
by Spiegel, Die arische Periode, p. 117. 

8 Spiegel, op. cit. p. 118. Compare also the remarks made below, p. 340, n. 3, 
on Esther, i, 1. 

* The Av. word niyne, here translated ' smites down,' is best so taken as a verbal 
form ; so also by Bartholomae, Air. Wh. coll. 492, 1814, followed by P. Wolff, Avesta 
Sbersetzt, pp. 79, 214. J. Darmesteter, Le Zend-Avesta, i, 366, also n. 52 (and of. n, 
469) has ' il abat k la riviere du Couchant.' Others have taken niyne as a loc. sg. ; 
thus F. Justi, Handbuch der Zendsprache (1864j, p. 171, renders 'im westlichen 
Niniveh'; F. Spiegel, Die ar. Per. p. 119, 'im westlichen Nighna' (i.e. the Nile). 
Opposed to the explanation as a proper name is C. de Harlez, Avesta traduit (1881), 
p. 461, who gives ' dans les profondeurs de I'occident,' with a footnote ' dans I'enfonce 
ment nocturne' ; cf. also ibid. p. 377, n. 4. 

326 Persian Dominions in N. hidia [CH. 

is the mention, in Yt. viii, 32, of a mountain called Us-Hindava, 
which stands in the midst of the partly mythical sea Vouru-kasha 
and is the gathering place of fog and clouds. The name Us-Hindava 
means ' Beyond (or, Above) India,' according to one way of trans- 
lating; but another rendering makes it simply 'the mountain from 
which the rivers rise.' Owing to this uncertainty, and to a general 
vagueness in three passages in which the mountain is referred to 
as Usind and Usindam in the Pahlavi, or Middle Persian, texts of 
Sassanian times (Bundahishn, xii, 6 ; xiii, 5 ; Zatsparam, xxii, 3), 
it seems wiser for the present to postpone an attempt to decide 
whether the allusion is to the Hindu Kush or possibly the Himalaya, 
or even some other ranged 

Precisely as was noted above (p. 321), in considering the Vedic 
material as sources for the historian's review of the distant past, 
there are likewise a number of Avestan names of places located 
south of the Hindu Kush in the territory that once at least was 
common in part to the Indians and the Iranians and has had, as 
a natural borderland, an important influence upon India's history 
in later ages. A portion of these domains corresponds to a con- 
siderable section of Afghanistan and possibly to a part of Balu- 
chistan, realms now under direct British influence or included 
politically as a part of the Indian Empire. One of the proofs of 
this community of interest is the fact that the territory of Arachosia 
(Av. HarahvaitI, O.P. Hara(h)uvatl), which corresponds to the 
modern province of Kandahar, was known, at least in later Parthian 
times, as 'White India' (IvSikt) AevKij). This we have on the 
authority of the geographer Isidor of Charax (first cent. A.D.), who, 
when mentioning Arachosia as the last in his list of Parthian pro- 
vinces, adds {Mans. Parth. 19), 'the Parthians call it "White 
India ".' As a supplement to this statement, in its historic aspect, 
may be quoted a pertinent observation made by the French savant 

1 The interpretation as Hindu Rush is given by Geldner, Grundriss d. iran. Philol. ii, 
38 ; the rendering of Bartholomae, Air. Wb. col. 409, is ' jenseits von Indien gelegen ' ; 
Darmesteter, Le Z.-A. ii, 423, n. 70, remarks : 'Le mot us-hindu signifie litt^ralement 
" d'oii se Invent les rivieres." II est douteux que ce soit une montagne r^elle: Us- 
hindu est le repr^sentant de la classe. ' For translations of the Pahlavi passages in 
which Usind, or Usindam, is mentioned, see E. W. West, S.B.E. v, 35, 42 ; xlvii, 160 
(and cf. v, 67, n. 3). It may be noted incidentally that an attempt has been made to 
connect the meteorological phenomena described in the myth of the Tishtar Yasht 
(Yt. viii), in which this allusion occurs, with the breaking of the monsoon. See the 
articles by Mrs E. W. Maunder, The Zoroastrian Star-Champions, in The Observatory, 
Nov. and Dec. 1912, March 1913 ; and the similar view by Mr E. W. Maunder, of the 
Royal Observatory, Greenwich, quoted by Moulton, Early Zoroastrianism, pp. 25, 26, 
n. 2, 436-7. 

xiv] The Persian Provinces 327 

James Darmesteter in touching upon the realms of Kabul and 
Seistan. He regards the language of Vd. i as indicating that 
' Hindu civilization prevailed in those parts, which in fact in the 
two centuries before and after Christ were known as White 
India, and remained more Indian than Iranian till the Musulman 
conquest \' 

All of the realms concerned in the next Avestan references to 
be cited have their historical and political bearing, important for 
the statesman as well as for the historian of India ; and they can be 
identified with the provinces under the imperial sway of Darius I of 
Persia, as mentioned in his cuneiform inscriptions. The dominions 
are equally included in the account of the ancient Persian satrapies 
given by Herodotus and are comprised in the geographical descrip- 
tions of Iran by his successors. For that reason, in the following 
enumeration, the Old Persian, Greek, and modern designations are 
recorded in every case together with the Avestan. 

To confine attention first to the land that is now Afghanistan, it 
may be noted that the Hindu Kush range may possibly be referred 
to in the Avestan allusion to Us-Hindava, mentioned above (p. 326). 
It is likewise possible to conjecture that the ridge of Band-i-Baian, 
somewhat to the west, may perpetuate the old Avestan name 
Bayana in the list of mountain names enumerated in the Nine- 
teenth Yasht (Yt. XIX, 3) ; while the chain familiarly known from 
the classics as Paropanisus or Paropamisus appears to be included 
under the Avestan designation Upairisaena, lit. ' Higher than the 
eagle '^. To the north of these barriers lay Bactria (Av. BakhdhI, 
O.P. Bakhtri, Gk. Ba/cT/oot, Ba/cT/jtav?;, Mod. Balkh), a centre which 
was destined to play an important part in India's history ^ 

Herat, on the west, including the district watered by the Hari 
Rud, was known in the Avesta as Haroiva (O.P. Haraiva, Gk. 'Ape/a). 
Kabul, to the east and nearer the present Indian frontier, appears 
as Vaekereta (answering to the western part of O.P. Ga(n)dara, 

1 Darmesteter, S.B.E. (2nd ed.), iv, 2 ; and cf. Le Z.-A. ii, 13, n. 32, where he bases 
his statement about the character of the civilisation on Mas'udI, Les Prairies d^or, 
ed. and tr. Barbier de Meynard, ii, 79-82, Paris, 1863. 

^ Cf. El. Bab. Paruparesanna, substituted for O.P. Ga{n)ddra in these versions 
of the Bahistan Inscription, 1. 16 (6). It is quite possible to see in Av. iSkata and 
pouruta, Yt. x, 14 (cf. Yt. xix, 3 ; Ys. x, 11), the names of two mountain branches 
of the Hindu Kush and Paropanisus; so, among other scholars, Sarre and Herzfeld, 
Iranische Felsreliefs, Text, p. 31 ; somewhat differently Bartholomae, Air. Wb. coll. 
376, 900. 

* For references to the passages in which the ancient Iranian names of the pro- 
vinces occur, consult Bartholomae, Air. Wb., under each of the separate Avestan or 
Old Persian names involved. 

328 Persian Dominions in N. India [ch. 

Gk. Vav8apiTc<;, or El. Paruparesanna, and possibly in part to O.P. 
Thatagu, Gk. l^arrayvBai,^). The region corresponding to the modem 
province of Kandahar, as already stated, is represented by Av. 
Harahvaiti (O.P. Hara(h)uvatT, Gk. 'Apaxoa-io). In the territory to 
the south-west, the river Helmand and the lagoon districts of Seistan 
around the Hamun Lake (which the natives call Zirrah, i.e. A v. 
Zrayah, * sea ') are respectively known in the Avesta as the Haetu- 
mant and as Zrayah Kasaoya (cf, O.P. Zra(n)ka or Zara(n)ka, Gk. 
Zapdyr^oL, 'Eapdyyoi, or Apayyiavrj); while the river systems that 
empty into this lagoon depression from the north are mentioned in 
Yasht XIX, 67, by names that can be identified exactly with their 
modern designations in almost every case^. It is worth noting that 
the majority of these particular allusions are found in the Nineteenth 
Yasht, which is devoted to the praise of the ' Kingly Glory ' of the 
ancient line of the Kayanians, heroes who are known to fame also 
through Firdausi's epic poem, the Shahnamah, and from whom some 
of the families in the regions named still claim to be descended. 

With regard to Avestan place names that may be localised in 
parts of Baluchistan there is more uncertainty. It is thought by 
some, for example, but denied by others, that Av. Urva (Vd. i, 10) 
may thus be a locality near the Indian border^ It might also be 
possible to suggest that the Avestan name Peshana (Yt. v, 109) may 
still survive in the Baluchi town Pishin, near Quetta, but it would 
be difficult to prove this. 

The quotations above given from Avestan sources serve at least 
to show the interest or share which Persia had traditionally in 
Northern India and the adjoining realms at a period prior to 
Achaemenian times, provided we accept the view, already stated 
(p. 323), that the Avesta represents in the main a spirit and con- 
dition that is pre- Achaemenian, however late certain portions of 
the work may be^ 

^ The position of the Sattagydai is not quite certain ; according to Sarre and 
Herzfeld, Iranische Felsreliefs, Text, pp. 27, 256, they are to be located in Ghazni and 
Ghilzai ; but Dames, Afghanistan, in Encyclop. of Islam, i, 158, places them in the 
Hazara country further to the north-west. Other authorities differ ; e.g. J. Marquart, 
Untersuch. z. Gesch. von Eran, ii, 175. 

2 See M. A. Stein, Afghanistan in Avestic Geography, in The Academy, May 16, 
1885, pp. 348-349 (also in Indian Antiquary, xv, 21-23). Consult also Geiger, 
Grundr. d. iron. Philol. ii, 388, 392-4. On the possibility of locating the tribal name 
Av. Sdma, cf. Gk. QaiJ-avaioi in Afghanistan, compare Sarre and Herzfeld, op. cit. p. 27 ; 
Marquart, Unters. z. Gesch. v. Eran, \i, 144, 176. 

3 For references see Bartholomae, Air. Wb. col. 404. 

* Lack of space prevents including here certain supplementary allusions to India in 
early times as found in the Pahlavi literature of the Sassanian era and in such later 
sources as Firdausi's Shahnamah ; but they will appear in the Festschrift WindiFch. 

xiv] Karly Relations with India 329 

Prior to the seventh century B.C., and for numerous ages after- 
wards, there is further proof of relations between Persia and India 
through the facts of trade in antiquity, especially through the 
early commerce between India and Babylon, which, it is believed, 
was largely via the Persian Gulf\ Persia's share in this develop- 
ment, although hard to determine, must have been significant even 
in days before the Achaemenian Empire. Beginning with the sixth 
century B.C., however, we enter upon the more solid ground of 
recorded political history. From unquestioned sources in the 
classics we know that the Medo-Persian kingdom, which was para- 
mount in Western Asia during that century, was brought into more 
or less direct contact with India through the campaigns carried on 
in the east of Iran by Cyrus the Great at some time between 558 
and 530 B.C., the limits of his reign. The difficulty, however, of 
determining exactly when this campaigning occurred and just how 
the domains between the rivers Indus and Jaxartes came under the 
control or sphere of influence of the Persian Empire is a problem 
accounted among the hardest in Iranian history^. 

In the following paragraphs of discussion, which may be con- 
sidered as a critical digression, statements or inferences from 
Herodotus, Ctesias, and Xenophon, with other evidence, have to be 
compared with those of Strabo and with the seemingly more con- 
servative views of Arrian, in interpreting the question of the possible 
or probable control of the Indian borderland touching upon Iran. 

In the first place, Herodotus says (i, 177) that 'Cyrus in person 
subjugated the upper regions of Asia^, conquering every nation 
without passing one by' ; but this statement is so broadly compre- 
hensive that it is difficult to particularise regarding North-western 
India except through indirect corroborative evidence. In fact, 
most of the allusions by Herodotus to India refer to the times of 
Darius and Xerxes, It is certain, however, that Cyrus, by his own 
personally conducted campaigns in the east, brought the major part 
of Eastern Iran, especially the realms of Bactria, under his sway*. 

^ See J. Kennedy, The Early Commerce of India with Babylon, 700-300 b.c, in 
J.R.A.S. 1898, pp. 241-288; and cf. V. A. Smith, Early History of India, 3rd ed., 
p. 28, n., Oxford, 1914 ; likewise W. H. Schoff, J.A.O.S. xxxui, 352 ; Bhys Davids, 
Buddhist India, p. 104. 

■■' See Pr&sek, Geschichte der Meder und Parser, i, 224 ; and compare How and 
Wells, Commentary on Herodotus, i, 177 (vol. i, 135), Oxford, 1912. 

* I.e. the regions in the east, more distant from Greece and contrasted with those 
subdued by Cyrus in Asia Minor through his general Harpagus. 

* For the Bactrian and ^aka conquests, see Herodotus, i, 153 compared with i, 177 ; 
and cotisult Ctesias, Persica, fragms. 33-34 (ed. Gilmore, pp. 127-129). For certain 
problems raised by the question of the ^akas, see F. W. Thomas, J.R.A.S, 1906, 
pp. 181-216, 460-464. 

330 Persian Dominions in N. India [ch. 

His conquests included the districts of Drangiana, Sattagydia, and 
Gandaritis, verging upon the Indian borderland, though we may 
omit for the moment the question of the extent of Cyrus's suzerainty 
over the Indian frontier itself. 

In the same connexion may be mentioned the fact that Ctesias, 
especially in the tenth book of his lost Persica, if we may judge 
from quotations in later authors regarding the nations involved, 
appears to have given an account of the campaigns by Cyrus in 
this region ^ The stories, moreover, regarding the death of Cyrus 
differ considerably ^ ; but the account recorded by Ctesias (fragm. 37, 
ed. Gilmore), which reflects local Persian tradition, narrates that 
Cyrus died in consequence of a wound inflicted in battle by ' an 
Indian,' in an engagement when * the Indians were fighting on the 
side of the Derbikes and supplied them with elephants.' The Der- 
bikes might therefore be supposed to have been located somewhere 
near the Indian frontier, but the subject is still open to debated 

Xenophon, in his romance of the life of Cyrus, entitled Cyro- 
paedia (i, 1, 4), declares that Cyrus 'brought under his rule 
Bactrians and Indians,' as forming a part of his wide-spread empire. 
In the same work (viii, 6, 20-21) he furthermore says that Cyrus, 
after reducing Babylon, * started on the campaign in which he is 
reported to have brought into subjection all the nations from Syria 
to the Erythraean Sea ' (i.e. the Indian Ocean) ; and for that reason 
he repeats that ' the Erythraean Sea bounded the empire of Cyrus 
on the east*.' This reference, though indefinite, certainly contains 
a direct allusion to control over the regions bordering on the 
Indian Ocean ; but it would be unwarranted to interpret it as 
indicating any sovereignty over the mouth of the Indus, such as 
could be claimed in regard to the Persian sea-route to India in the 
time of Darius and his successors. 

In a general way, however, as possibly supporting the idea of 
some sort of suzerainty over Northern India by Cyrus, we may note 
the fact that Xenophon {Cyrop. VI, 2, l-Il) introduces an account 

^ See the passages in Gilmore's edition of the Persica, pp. 133-135 ; also G. 
BawUnson, Five Great Monarchies, n, 371, n. 22 ; but cf. Marquart, Unters. z. Getch, 
V. Eran, n, 139. 

^ Consult G. Bawlinson, op. cit. iv, 378-380 ; E. Katz, Cyrus des Perserkonigs 
Abstammung, Kriege, und Tod, Klagenfurt, 1895 ; Pra§ek, Gesch. der Meder und 
Perser, i, 236, n. 1. 

^ The notices of classical authors regarding this widely distributed people are 
collected by Tomaschek, art. Derbikes, in Pauly-Wissowa, Real-Encyclopcidie, v, 237- 
238, Stuttgart, 1905. 

* In regard to the term ' Erythraean Sea ' as a designation for the Indian Ocean, 
see W. H. Schoff, J.A.O.S. xxxiii, 349-362. 

xiv] Cyrus 331 

of an embassy sent to Cyrus by an Indian king. This embassy 
conveyed a sum of money for which the Persian king had asked, 
and ultimately served him in a delicate matter of espionage before 
the war against Croesus and the campaigns in Asia Minor. It may 
be acknowledged that the value of this particular allusion is slight, 
and that the Cyropaedia is a source of minor importance in this 
particular regard ; but yet it is worth citing as showing, through 
Xenophon, a common acceptance of the idea that Cyrus was in a 
position to expect to receive direct consideration, if not vassalage, 
from the overlord of Northern India. 

Descending to the Hellenistic age, when the Greeks began to have 
knowledge of India at first hand, we find that two of the principal 
authorities, Nearchus, who was Alexander's admiral, and Megas- 
thenes, the ambassador of Seleucus I at the court of Chandragupta, 
are at variance regarding an attempted conquest of India by Cyrus. 

The account of Nearchus, as preserved by Arrian {Anah. vi, 
24, 2-3), links the names of Cyrus and of Semiramis, the far-famed 
Assyrian Queen, and states that Alexander, when planning his 
march through Gedrosia (Baluchistan), was told by the inhabitants 
* that no one had ever before escaped with an army by this route, 
excepting Semiramis on her flight from India. And she, they said, 
escaped with only twenty of her army, and Cyrus, the son of 
Cambyses, in his turn with only seven. For Cyrus also came into 
these parts with the purpose of invading India, but was prevented 
through losing the greater part of his army, owing to the desolate 
and impracticable character of the routed' 

Megasthenes, on the other hand, as quoted by Strabo {Geogr. xv, 
1, 6, pp. 686-687 Cas.), declares that 'the Indians had never engaged 
in foreign warfare, nor had they ever been invaded and conquered 
by a foreign power, except by Hercules and Dionysus and lately 
by the Macedonians.' After mentioning several famous conquerors 
who did not attack India, he continues : ' Semiramis, however, 
died before [carrying out] her undertaking ; and the Persians, 
although they got mercenary troops from India, namely the 
Hy drakes^, did not make an expedition into that country, but merely 
approached it when Cyrus was marching against the Massagetae.' 

We may also take Megasthenes to be the authority for the 
statement of Arrian {Indica, ix, 10; and cf. v, 4-7) that, according 
to the Indians, no one before Alexander, with the exception of 

1 Strabo, Geogr. xv, 1, 5, p. 686 Cas. (and cf. xv, 2, 6, p. 722 Cas. ), likewise quotes 
Nearchus, but merely to the effect that Cyrus escaped with seven men. 

2 I.e. the Oxydrakai or Kshudrakas in the Panjab; see Chapter xv, p. 375. 

332 Persian Dominions in N. India [CH. 

Dionysus and Hercules, had invaded their country, ' not even Cyrus, 
the son of Cambyses, although he marched against the Scythians 
and showed himself in other respects the most enterprising of 
Asiatic monarchs\' 

It appears, therefore, that both Nearchus and Megasthenes 
deny, the former by implication and the latter expressly, that Cyrus 
ever reached India, although Nearchus regards him as having made 
an unsuccessful campaign in Baluchistan. We must not, however, 
overlook the fact that Strabo and Arrian, our proximate sources, 
consider the river Indus to be the western boundary of India 
proper ; and the foregoing accounts consequently leave open the 
possibility that Cyrus made conquests in the borderland west of 
the Indus itself. Indeed, Arrian elsewhere {Indica, i, 1-3) ex- 
pressly states that the Indians between the river Indus and the 
river Cophen, or Kabul, 'were in ancient times subject to the 
Assyrians, afterwards to the Medes, and finally submitted to the 
Persians and paid to Cyrus, the son of Cambyses, the tribute that 
he imposed on them.' 

In regard to the supposed campaign of Cyrus in Baluchistan, 
we may note that Arrian {Anab. iii, 27, 4-5) mentions the story, 
recorded elsewhere in connexion with Alexander's exploits, that 
Cyrus had received substantial help from the Ariaspian people (a 
tribe dwelling in a region that corresponds to the modern Seistan) 
when he was waging war in these territories against the Scythians^. 
This folk received from him in consequence the honorific title 
Euergetae, 'Benefactors,' a term answering to the Persian desig- 
nation Orosangae mentioned by Herodotus (viii, 85) ^ 

One further point may be cited from a classical source. Pliny, 
Hist. Nat. VI, 23 (25), credits Cyrus with having destroyed a city 
called Capisa in Capisene, a place supposed to be represented 
by Kafshan (Kaoshan, Kushan) in the modern Ghorband valley 
district, somewhat north of Kabul, and in any case it could not 
have been far from the Indian frontier*. 

,1 Of. also Justin, Historiae Philippicae, i, 2, 9, who says that no one invaded India 
except Semiramis and Alexander. 

2 Arrian, Anab. iii, 27, 4-5 ; Strabo, Geogr. xv, 2, 10, p. 724 Cas. ; Diodorus 
Siculus, Bibl. Hist, xvii, 81, 1 ; Quintus Curtius, Hist. Alex, vii, 3, 1-3. For a special 
consideration of this subject, see F. W. Thomas, Sakastana, in J.R.A.S. 1906, pp. 181- 
216, 460-464. 

^ For the interpretation of this word as 'active in spirit,' cf. Thomas, op. cit, p. 196. 

* See Thomas, J.R.A.S. 1906, pp. 191, n. 1, 460-461, and the works there cited, 
especially E. J. Rapson, J.R.A.S. 1905, pp. 783-784; J. Marquart, ErdnSahr, pp. 280- 
281; and cf. idem, Unters. z. Gesch. v. Eran, n, 180, Leipzig, 1905. Capisa is the Kia- 
pi-shi of Hiuen Tsiang and the Ki-pin of other Chinese texts. The name is found in 

xiv] Cyrus: Cambyses 333 

To sum up, we may say that, even if there are just grounds for 
doubting that Cyrus actually invaded Northern India, there can be 
no question that he did campaign in the territories corresponding 
to the present Afghanistan and Baluchistan. It seems likely that 
Alexander's historians may have been inclined to minimise the 
accomplishments of Cyrus the Great, especially in the light of his 
apparent set-back in Gedrosia^, in order to bring into greater 
prominence the achievements of the famous Greek invader. 

The view above stated, to the eifect that Cyrus advanced at 
least as far as the borders of the Indus region, will be better under- 
stood from the ensuing paragraphs, in which the holdings of his 
successors and their control of regions integral to the Indian 
Empire of to-day are shown. The main point of this opinion is 
likewise in agreement with such an authority on the subject as 
Eduard Meyer, who expressly says : ' Cyrus appears to have subju- 
gated the Indian tribes of the Paropanisus (Hindu Kush) and in 
the Kabul valley, especially the Gandarians ; Darius himself 
advanced as far as the Indus ^.' 

Cambyses, whose activities were almost wholly engaged in the 
conquest of Egypt, could hardly have extended the Persian 
dominions in the direction of India, even though he may have 
been occupied at the beginning of his reign in maintaining 
suzerainty over the extensive realm inherited fi-om his father. 
Xenophon, or his continuator {Cyrop. viii, 8, 2), speaks of almost 
immediate uprisings by subject nations after the death of Cyrus, 
and these revolutions may have caused the postponement of the 
Egyptian expedition of Cambyses until the fifth year of his reign, 
526-525 B.C. ; but it would be hazardous to suggest any direct 
connexion of India with these presumable campaigns. Herodotus 
makes two very broad statements ; one (iii, 88, cf. i, 177) to the 
effect that, wlien Darius became king, after the death of Cambyses 
and the assassination of the false Smerdis, 'all the peoples of Asia, 
with the exception of the Arabians [who were already allied as 
friends], were subject to him, inasmuch as they had been subdued 

the first element of the compound O.P. Kapisa-kani, the name of a stronghold mentioned 
in the inscriptions of Darius (Bh. 3, 61). Marquart {Unters. ii, 180), with others, 
inclines to regard the two places as identical, although objections may be raised that 
Kapisa-kani was located in Arachosia (the El. version, 3, 37, 25 expressly adding 'in 
Arachosia '). Still much depends on determining the extent of the confines of Arachosia 
in the time of Darius. 

1 Cf, the passages of Arrian and Strabo cited above, p. 331, and n. 1. 

* Ed. Meyer, Geschichte des Altertums, in, 97, with note. See also Max Kiesshng, 
Zur Geschichte der ersten Eegieruiigsjahre des Darius Hystaspis, in Quellen u. Forsch. 
z. alt. Gesch. u. Geogr. p. 28, hrsg. W. Sieglin, Heft 2, Leipzig, 1900-1901. 

334 Persian Dominions in N, India [ch. 

by Cyrus and afterwards by Cambyses in his turn.' Again he says 
(ill, 67), with reference to the death of the usurper Smerdis, that 
* all the peoples of Asia felt regret, except the Persians themselves.' 
Although it would be a forced interpretation of these passages to 
construe them as including India proper among the subject nations 
of the Persian Empire^, it seems clear, nevertheless, that Darius, 
when he assumed the sovereignty in 522 B.O., had, as an Achae- 
menian, an authentic claim to the realms immediately bordering 
upon India, if not to that land itself. 

For the reign of Darius (522-486 B.c.) we have documentary 
evidence of the highest value in the inscriptions executed by that 
monarch's command and containing his own statements. From these 
inscriptions, especially when they are compared one with another, 
we can trace the general outline of the Persian dominion in 
Northern and North-western India in the time of Darius, and we 
can even infer that he annexed the valley of the Indus early in his 
reign, a conclusion which is confirmed by the testimony of various 
passages in Herodotus. The three records in stone which require 
special consideration in this connexion are the following^ : 

1. The famous Bahistan Rock Inscription (1, 16-17; 2, 7-8; 
3, 54-76), which is presumably to be assigned to a period between 
the years 520 and 518 B.O., with the exception of the fifth column, 
which was added later. 

2. The second of the two Old Persian block tablets sunk in 
the wall of the Platform at Persepolis (Dar. Pers. e. 15-18). It 
was probably carved between 518 and 515 b.c. 

3. The upper of the two inscriptions chiselled around the Tomb 
of Darius in the cliff at Naksh-i-Rustam (NR. a. 23-26), which 
must have been incised some time after 515 B.c.^ 

1 Equally doubtful would be the attempt to connect the name of Cambyses 
(0. P. Ka(m)bu]iya) with the frontier people of Kamboja, though consult the references 
given by A. Hoffmann-Kutschke, Die altpersischen Keilinschriften, p. 21, Stuttgart, 
1909; and idem, Indogermanisches, in Recueil de Travaux igypt. et assyr. 31, 66. 

* A mutilated clay tablet, Dar. Sus. e, exhibits the remains of a list of provinces, 
which seems, however, to have been the same as that which is found in NR. a. 

3 The dates assigned to these three inscriptions by different scholars vary somewhat, 
especially in regard to the record on the Bahistan Eock, although they are included 
approximately within the limits given. In respect to dating the Bahistan edict, much 
depends upon the interpretation of the O.P. phrase hamahydyd(h) tharda(h) ; for if, 
following Weissbach, we take it to mean ' in one and the same year,' all the events 
chronicled must have taken place within about a year after Darius succeeded to the 
throne, whereas otherwise they may be regarded as extending over two or three or even 
more years. See F. H. Weissbach, Zur neubabylon. u. achamenid. Chronologic, in 
Z.D.M.G. Lxn, 640-641 ; idem, Keilinschr. d. Achameniden, pp. Ixix-lxxiii, Leipzig, 
1911 ; idem, Zum bab. KcUender, in Hilprecht Anniversary Volume, pp. 285-290 (with 

xiv] Darius 335 

The Bahistan Inscription itself (1, 13-17) does not include 
India in the list of the twenty-three provinces which 'came to 
Darius,' as the Old Persian text says, or 'obeyed him,' as the 
Babylonian version expresses it\ The inference to be drawn, 
therefore, is that the Indus region did not form a part of the 
empire of Darius at the time when the great rock record was 
made, though it was incorporated shortly afterwards, as is shown 
by the two other inscriptions in question. Both of these latter 
(Dar. Pers. e. 17-18, and NR. a. 25) expressly mention Hi{n)du, 
that is, the Punjab territory, as a part of the realm. The Northern 
Indian domain must therefore have been annexed sometime 
between the promulgation of the Bahistan edict and the comple- 
tion of the two records just cited. The present tendency of 
scholarly opinion is to assign the Indus conquest to about the year 
518 B.C.2 

In addition to the evidence of the inscriptions, the fact that a 
portion of Northern India was incorporated into the Achaemenian 
Empire under Darius is further attested by the witness of Herodotus, 
who, in giving a list of the twenty satrapies or governments that 
Darius established, expressly states that the Indian realm was the 
'twentieth division' (Hdt. iii, 94, cf. in, 89). Some inference 
regarding its wealth and extent may furthermore be gathered from 
the tribute which it paid into the Persian treasury. Herodotus is 
our authority on this point, when he explicitly narrates (in, 94) : 
'The population of the Indians is by far the greatest of all the 
people that we know ; and they paid a tribute proportionately 
larger than all the rest — [the sum of] three hundred and sixty 
talents of gold dust.' This immense tribute was equivalent to 
over a million pounds sterling, and the levy formed about one- 
third of the total amount imposed upon the Asiatic provinces ^ 
All this implies the richness of Persia's acquisition in annexing the 
northern territory of Hindustan*; and it may also be brought into 

Table), Leipzig, 1909 ; refer also to King and Thompson, Inscr. Behistun, pp. xli-xliii ; 
Prasek, Gesch. d. Med. u. Pers. ii, 37-38 ; Sarre and Herzfeld, Iranische Felsreliefs, 
pp. 17-33, 106-107 ; cf. also Justi, Grundr. d. iran. Philol. n, 430. 

1 Cf. Weissbach, Die Keilinschriften der Achdmeniden, p. 11, n. 6a. 

2 See Sarre and Herzfeld, Iranische Felsreliefs, pp. 106-107 (with references) ; Max 
Kiessling, Zur Geschichte...des Darius, pp. 56, 57, 60; Pra§ek, Gesch. d. Meder «. 
Terser, ii, 37, n. 5. 

3 See V. A. Smith, Early History of India, 3rd ed., pp. 37-38, n. 1 ; and cf. also 
F. H. Weissbach, Zu Herodots persisch^r Steuerliste, in Philologus, 71 (N. F. 25), 479- 
490 ; idem, Keilinschr. d. Achdmeniden, pp. Ixxiv-lxxv. 

* V. A. Smith, op. cit. p. 38, is of the opinion of those who hold that, owing to the 
changes in the courses of the rivers since ancient times, ' vast tracts in Sind and the 
Panjab, now desolate, were then rich and prosperous.' 

33^ Persian Dominions in N, India [ch. 

connexion with the curious story of the gold-digging ants in this 
region, which Herodotus tells directly afterwards (iii, 102-105). 

There is likewise another passage in Herodotus (iv, 44) which 
affords further proof, both of the Persian annexation or control of 
the valley of the Indus from its upper course to the sea, including 
therefore the Punjab and Sind, as well as of the possibility at that 
time of navigating by sea from the Indus to Persia. Sometime 
about 517 RC, Darius despatched a naval expedition under Scylax, 
a native of Caryanda in Caria, to explore the Indus. The squadron 
embarked at a place in the Gandhara country, somewhere near the 
upper course of the Indus, the name of the city being Kaspatyros 
(Hdt. IV, 44, cf. Ill, 102) or, more accurately, Kaspapyros (Hecataeus, 
Fragm. 179). The exact location of this place is still a matter of 
discussion, but the town may have been situated near the lower 
end of the Cophen (now Kabul) River before it joins the Indus K 
The fleet, it is recorded, succeeded in making its way to the Indian 
Ocean and ultimately reached Egypt, two and one-half years from 
the time when the voyage began. From the statement of Herodotus 
(iv, 44) it would appear that this achievement was accomplished 
prior to the Indian conquest, for he says that 'after {/xerd) they 
had sailed around, Darius conquered the Indians and made use of 
this sea ' [i.e. the Indian Ocean] ; but it seems much more likely 
that Darius must previously have won by force of arms a firm hold 
over the territory traversed from the headwaters of the Indus to 
the ocean, in order to have been able to carry out such an expedi- 
tion 2. This conclusion appears still more convincing when we 
consider the difiiculties which Alexander encountered in his similar 

1 Sir M. A. Stein suggests Jahangira, an ancient site on the left bank of the Eabal 
Biver, some six miles above the point where it flows into the Indus at Attock (see Stein, 
Memoir on the Anc. Geogr. of Kaimir, pp. 11-13, Calcutta, 1899, reprinted from 
J.A.S. Bengal, vol. lxviii, pt. 1, extra No. 2, 1899). Marquart, Untersuch. z. Gesch. 
V. Eran, u, 178-180, 242, and n. 8, 246, n. 3, favours as the location an ancient town 
known in Sanskrit as Pushkalavati. Compare also Prasek, Gesch. d. Med. u. Perser, 
n, 38; and V. A. Smith, Early Hist. India, 3rd ed., pp. 37-38, n. 1. Sarre and 
Herzfeld, Iran. Felsreliefs, pp. 26, 253, seem inclined to revive the old idea of associ- 
ating the name with Kashmir, cf. H. H. Wilson, Ariana Antiqua, pp, 136-137, 
London, 1841. 

' The early Greek geographer Hecataens, who flourished in the reign of Darius, 
seems to have possessed considerable information regarding the Indus valley, which 
may have come to him from Scylax himself. Cf. Fragments 174-179, in Fragmenta 
Historicorum Graecorum, ed. C. Miiller, i, 12, Paris, 1841, especially Fra^jw. IVo, where 
Hecataeus says that a tribe called the Opiai ' dwell by the Indus Kiver, and there 
there is a royal fort. Thus far the Opiai extend, and beyond there is a desert as far as 
the Indians. ' If ' royal fort ' means a fort of the Great King, as is likely, we have 
evidence here for the presence of a Persian frontier garrison on the Indus. 



XI v] The htdictn Realm of Darius 337 

undertaking of voyaging down the Indus to the sea, two centuries 
later, even after having first subdued most of the tribes of the 
Upper Punjab before starting on the voyaged 

The dominion of Persian authority under Darius, therefore, as 
is clear from the Greek sources in connexion with the Inscriptions, 
comprised the realm from the embouchment of the Indus to its 
uppermost tributaries on the north and west. Regarding the 
Indians towards the south, we have the express statement of 
Herodotus (iii, 101) to the efiect that 'these were never subject to 
King Darius.' Herodotus also evidently considers the sandy 
wastes in portions of the present Sind and Rajputana, to the east 
of the Indus, as the frontier in that direction ; for he says (iii, 98) 
that 'the part of the Indian territory towards the rising sun is 
sand,' and he adds immediately afterwards that 'the eastern part 
of India is a desert on account of the sand.' How far eastward the 
Persian dominion may have extended in the Panjab cannot be 
exactly determined ; but it is significant that Herodotus never 
refers to the Ganges valley^, and not one of our sources makes any 
mention of the famous Indian kingdom of Magadha, which was 
coming into prominence under the Buddhist rulers Bimbisara and 
Ajata^atru during the reign of Darius and simultaneously with the 
Persian conquests ^ On the whole, so far as the extent of the 
Persian control is concerned, no better summary need be given 
than the cautious expression of Vincent Smith, when he says: 
'Although the exact limits of the Indian satrapy [under Darius] 
cannot be determined, we know that it was distinct from Aria 
(Herat), Arachosia (Kandahar), and Gandaria (North-western Pan- 
jab). It must have comprised, therefore, the course of the Indus 
from Kalabagh to the sea, including the whole of Sind, and 
perhaps included a considerable portion of the Panjab east of 
the Indus*.' 

At this point it may not be out of place to refer briefly to the 
information that is afforded by the Inscriptions and by Herodotus 
regarding the sway exercised by Darius over the peoples of the 
Indian borderland. Of the twenty-three tributary provinces the 
names of which appear on the Bahistan Rock (Bh. 1, 14-17) and 
are repeated with some slight variations in the Platform and the 
Tomb Inscriptions (Dar. Pers. e. 10-18; NR. a. 22-30), three pro- 

' See Chapter xv, pp. 374 ff. ; cf. V. A. Smith, Early Hist. India, 3rd ed., pp. 88-104. 

* He says, for instance (iv, 40) that ' from India onward the country to the east is 
desert, and no one can tell what it is like.' 

* On this point see V. A. Smith, Early Hist. India, 3rd ed., p. 37. 

* Op. cit., p. 38. 

C.H.I. I. 22 

338 Persian Dominions ^ N. India [cH. 

vinces, namely Bakhtrl (Bactria), Haraiva (Herat), and Z(a)ra(n)ka 
(Drangiana, or a portion of Seistan) as noted above (pp. 327-8), 
form a part of the present Afghanistan lying more remote from the 
Indian frontier. The five that are directly connected with the 
region of the Indus itself are, as partly indicated earlier in the 
chapter {ihid.), Ga(n)dara (the region of the Kabul valley as far as 
Peshawar)^, Thatagu (either the Ghilzai territory to the south-west 
of Ghazni or the Hazara country further to the west and north- 
west), Hara(h)uvatl (the district about Kandahar in the broadest 
sense), Saka, and Maka^. The term Saka may possibly allude to 
Sakastana (Seistan) and the dwellers around the region of the 
Hamun Lake^ ; but the distinction made in the Tomb Inscription 
of Darius (NR. a. 25-26) between the Saka Haumavarga, answer- 
ing to the Amyrgioi Sakai of Herodotus (vii, 64), and the Saka 
Tigrakhauda, 'wearing pointed caps,' an attribute corresponding 
to the term Orthokorybantioi of Herodotus (iii, 92), may indicate 
a special division of the ^^akas, or Scythians, living between the 
extreme northern sources of the Indus and the headwaters of the 
Oxus*. The district Maka is believed to be identified with Makran, 
once occupied by the Mykans of Herodotus (iii, 93 ; vil, 68) and 
now a part of Baluchistan ^ 

Herodotus (iii, 91-93) mentions in his list of peoples that were 
subject to Darius — corresponding in a general way to the satrapies 
of the empire — four or five more which may be identified as 
having occupied districts in or near the present Afghanistan, in 
some cases adjoining the Indian frontier. The Sattagydai and 
Gandarioi (cf. OP. Thatagu and Ga(n)dara), for example, have the 

1 For Greek references to Gandara consult Pauly-Wissowa-Kroll, Real-Encyclo- 
pddie, VII, 696-701, Stuttgart, 1912. 

2 The slight variations in the lists of the three inscriptions, as regarding these 
provinces, are as follows : (1) Bh. 1, 16-17, Ga(n)dara, Saka, 6atagu§, Hara(h)uvati§, 
Maka ; (2) Dar. Pers. e. 17-18, 0atagu§, Hara(h)uvati§, Hi(n)du§, Ga(n)dara, Saka, 
Maka; (3) NB. a. 24-26, Hara(h)uvatis, Batagus, Ga(n)dara, Hi(D)dug, Saka Hauma- 
varga, Saka Tigraxauda. 

3 For such a view see F. W. Thomas, J.R.A.S. 1906, pp. 181-216, 460-464 ; but 
compare the observations by Sarre and Herzfeld, Iranische Felsreliefs, pp. 252-263. 

* For a general discussion of the Qaka question (with bibliographical references), 
see Sarre and Herzfeld, op. cit. pp. 23-24, 30, 36-40 (with cuts), and 252-253, also 
maps 1 and 2 at the end of the same volume. Consult likewise Marquart, Untersttch. 
z. Gesch. V. Eran, ii, 86, 136, n. 5. It may also be noted that Polyaenus, Strategemata, 
VII, 12, refers to an expedition of Darius against the ^a^kas, apparently north of the 
region of Bactria, and mentions Amorges or Omarges (i.e. Haumavarga?) as one of the 
Qaka kings. 

' So also Eduard Meyer, Persia, in Encyclop. Brit., 11th ed., xxi, 202; and Sarre 
and Herzfeld, op. cit. pp. 28-29 ; refer likewise to J. J. Modi, The Country of Mekran, 
its Past History, in East and West, May, 1904, pp. 1-12, Bombay. 

XI v] Peoples of the Indian Frontier 339 

Dadikai and the Aparytai linked with them in the same enumera- 
tion. Of these latter tribes, the Dadikai may be identified with 
the Dards of the Upper Indus valley, somewhere between the 
Chitral district and Kashmir ; and the Aparytai are to be con- 
nected with the inhabitants of the mountainous regions of the 
Hindu Kush, north of KabuP. The Kaspioi, who, according to 
Herodotus (iii, 93, cf. also vii, Ql, 86) constituted together with the 
Sakai the fifteenth division of the empire (and who are to be dis- 
tinguished from the Kaspioi of the eleventh division (iii, 92), by 
the Caspian Sea), must likewise have been an easterly people, and 
they are perhaps to be located in the wild tract of Kafiristan, to 
the north of the Kabul River I The Thamanaioi, whom Herodotus 
(hi, 93, 117) mentions as forming a part of the fourteenth division 
of the tributary nations, occupied a section of Afghanistan not easy 
to define precisely, but presumably in the western or west-central 
region, as noted above (p. 328, n. 2). The territory of Paktyike in 
the thirteenth division (Hdt. iii, 93 ; cf. iii, 102 ; iv, 44) and its 
people, the Paktyes (Hdt. vii, 67), are to be located within the 
borders of the land now called Afghanistan ; but whether the name 
is to be regarded as a tribal designation of the Afghans in general, 
and as surviving in the term Pakhtu or Pashtu applied to their 
language, is extremely doubtfuF. 

Finally, for the sake of completeness, it may be noted that 
India appears as one of the limits of the Persian Empire under 
Darius in the apocryphal Greek version of the Book of Ezra known 
as I Esdms. The passage (iii, 1-2) runs as follows : ' Now King 
Darius made a great feast unto all his subjects, and unto all 
that were born in his house, and unto all the princes of 
Media and of Persia, and to all the satraps and captains and 
governors that were under him, from India unto Ethiopia, in 
the hundred twenty and seven provinces ^' Inasmuch, however, as 
the apologue of the Three Pages, in which this reference .is 
embodied, seems to be subsequent to the age of Alexander, we 
must regard the passage as merely a general tradition concerning 

1 Cf, Marqoart, Untersuch. z. Gesch. v. Eran, ii, 175; Sarre and Herzfeld, op. cit. 
p. 31. 

* So Marquart, op. cit. ii, 140-142 ; but consult Sarre and Herzfeld, op. cit. p. 253. 
Thomas, J.R.A.S. 1906, p. 191, n. 1, suggests reading Kdiriaoi (cf. Capisa, p. 382, 
above) for Kdtririoi. 

3 Consult Marquart, op. cit. ii, 171-180 ; Sarre and Herzfeld, op. cit. pp. 26-27 ; 
Ed. Meyer, Persia, in Encyclop. Brit., 11th ed., xxi, 203 ; Dames, Afghanistan, in 
Encyclop. of Islam, i, 149-150. 

* Cf. also the paraphrase in Josephus, Jewish Antiquities, xi, 3, 2 (33), and the 
passages from Esther cited below, p. 340, n. 3. 


340 Persian Dominions in N. India [ch. 

the extent of tlie Achaemenian Empire without insisting upon the 
chronological allusion to Darius P. 

For the reign of Xerxes (486-465 B.C.) the continuance of the 
Persian domination in Northern India is proved by the presence 
of an Indian contingent, consisting of both infantry and cavalry, 
among the troops from subject nations drawn upon by that 
monarch to augment the vast army of Asiatics which he marshalled 
to invade Greece. Herodotus (vii, 65) describes the equipment of 
the Indian infantry as follows : * The Indians, clad in garments 
made of cotton, carried bows of cane and arrows of cane, the latter 
tipped with iron ; and thus accoutred the Indians were marshalled 
under the command of Pharnazathres, son of Artabates.' It is 
worth remarking, perhaps, that the commander of these forces, as 
shown by his name, was a Persian. Regarding the Indian cavalry 
Herodotus (vii, 86) says that they were 'armed with the same 
equipment as in the case of the infantry, but they brought riding- 
horses and chariots, the latter being drawn by horses and wild asses ^.' 

It may be observed, moreover, that a number of the tribes who 
inhabited the Indo-Iranian borderland in the time of Darius (see 
above, pp. 327-8, 338) were represented in the host of Xerxes as 
well; namely the Bactrians, Sakai, Ar(e)ioi, Gandarioi, Dadikai, 
Kaspioi, Sarangai, Paktyes, occupying the Afghan region, and the 
Mykoi of Baluchistan (Hdt. Vii, 64-68). On the whole, therefore, 
we may conclude that the eastern domain of the Persian Empire 
was much the same in its extent under Xerxes in 480 RO. as it 
had been in the reign of his great father^ 

The period following the defeat of the Persian arms under 
Xerxes by Greece marks the beginning of the decadence of the 
Achaemenian Empire. For this reason it is easy to understand why 
there was no forward movement on Persia's part in India, even 
though the Iranian sway in that territory endured for a century 
and longer. Among other proofs of this close and continued 
connexion may be mentioned the fact that Ctesias, who was 
resident physician at the Persian court about the beginning of the 
fourth century B.O., could hardly have written his Indica without 
the information he must have received regarding India from 

^ See the note on this passage by S. A. Cook, in Apocrypha and Pseudepigrapha oj 
the Old Testament, ed. Charles, i, 29, Oxford, 1913. 

* As a matter of curiosity it may be noted that Herodotus (vii, 187) says that an 
immense number of Indian dogs followed the army of Xerxes in his Grecian invasion. 

3 Later Jewish tradition has the same formulaic description for the empire of Xerxes 
(Ahasuerus) as for that of Darius (cf. p. 339, above) ; thus in the Book of Esther, i, 1 
(cf. also VIII, 9), Xerxes is styled 'Ahasuerus which reigned from India even unto 
Ethiopia, over an hundred and seven and twenty provinces.' 

XI v] Extent of Persian Influence 341 

envoys sent as tribute-bearers to the Great King or from Persian 
officials who visited India on state business, as well as from his 
intercourse with travellers and traders of the two countries^. If 
the work of Ctesias on India had been preserved in full, and not 
merely in the epitome by Photius and in fragmentary citations 
by other authors, we should be better informed to-day as to Persia's 
control over Indian territory during the period under consideration I 

The fact, however, that this domination prevailed even to the 
end of the Achaemenian sway in 330 b.c. is furthermore proved by 
the call which Darius III, the last of the dynasty, was able to issue 
to Indian troops when making his final stand at Arbela to resist 
the Greek invasion of Persia by Alexander. According to Arrian 
{Anah. Ill, 8, 3-6), some of the Indian forces were grouped with 
their neighbours the Bactrians and with the Sogdians under the 
command of the satrap of Bactria, whereas those who were called 
* mountainous Indians ' followed the satrap of Arachosia. The 
Sakai appeared as independent allies under their leader Mauakes. 
These frontier troops were supplemented by a small force of 
elephants ' belonging to the Indians who lived this side of the Indus.' 

Emphasis may be laid anew on the fact that the sphere of 
Persian influence in these early times can hardly have reached 
beyond the realm of the Indus and its affluents. We may assume, 
accordingly, that when Alexander reached the river Hyphasis, the 
ancient Vipa9 and modern Beas, and M'as then forced by his own 
generals and soldiers to start upon his retreat, he had touched the 
extreme eastern limits of the Persian domain, over which he had 
triumphed throughout^. The interesting articles by Dr D. B. Spooner 
in the Jmir. RA.S. for 1915 (pp. 63-89, 405-455), entitled 
Tfie Zoroastrian Period of Indian History, make the strongest 
possible plea for a far wider extension of Persian influence upon 
India in the early historic period. AVhile scholars are fully agreed 
to allow for the general and far-reaching theory of Persian influence, 
they have not found themselves prepared to accept many of the 
hypotheses put forward in Dr Spooner's two articles, as the 
criticisms which succeeded their publication show*. 

* In this connexion compare M'Crindle, Ancient India as described by Ktesias, 
pp. 3-4, London, 1882, noting certain details, for example, in §§ 3-7. 

* The extant remains of the Indica are to be found in Ctesiae...Fragmenta, ed. 
C. Miiller, pp. 79-105 (in his edition of Herodotus, Paris, 1844). 

3 For the situation, see Chapter xv, pp. 372-3, and refer to the map. 

4 V. A. Smith, J.R.A.S. 1915, pp. 800-802 ; Keith, ibid. 1916, pp. 138-143 ; Thomas, 
ibid. pp. 362-366; 'Nimrod,' The Modern Review, Calcutta, 1916, pp. 372-376, 490-498, 

342 Ancient Persian Coins in India [ch. 

With the downfall of the Achaemenian rule before the onslaught 
of the conqueror from Macedon ends the first chapter in the story 
of the relations between India and Persia. It belongs elsewhere 
to indicate those which existed under the successors of Alexander, 
under the Parthian and Sassanian sovereigns, and down through 
Muhammadan times, until, in the eighteenth century, a Persian 
invader like Nadir Shah could carry oflf the Peacock Throne of the 
Mughals and deck his crown with the Koh-i-Nur. 

Ancient Persian Coins in India 

Whatever were the actual, limits of Persian power in India, it 
is certain that within these limits the money of the Persian kings 
must have been current. At the same time it is not easy to 
support the general statement by definite facts. Properly authen- 
ticated records of finds are virtually unknown. Nor can over- 
much reliance be placed on deductions drawn from the occurrence 
of individual specimens in collections that have been formed in 
North-western India. Before the construction of the Russian 
railways in Central Asia the waifs and strays of commerce, like 
gold and silver coins from Bukhara and Khorasan, naturally 
drifted over the mountain-passes of Afghanistan into the Punjab 
as the nearest profitable market. Once they had arrived there, 
however, the dealers into whose hands they came were free to 
assign to them the provenance that seemed most likely to enhance 
their price, a circumstance that renders it difficult to appraise the 
value of the scanty evidence available. For reasons that will 
presently appear, the two precious metals can best be considered 

The standard gold coin of Ancient Persia was the daric, which 
bore upon the obverse a figure of the Great King hastening 
through his dominions, armed with bow and spear ; and upon 
the reverse an irregular oblong incuse. It weighed about 130 
grains (8"42 grammes), and was in all probability first minted by 
Darius Hystaspes, the monarch who was responsible for adding 
the valley of the Indus to the empire. From its infancy, there- 
fore, the daric would have ready access to the country beyond the 
Hindu Kush. At the same time there was an important economic 
reason which would militate against its extensive circulation in 
these regions. Gold was abundant there, so abundant that for 
many centuries its value relatively to silver was extraordinarily 
low. There are grounds for believing that during the period of 

XI v] Ancient Persian Coins i?i India 343 

the Persian dominion the ratio was no higher than 1 : 8, as com- 
pared with the norm of 1 : 13 "3 maintained by the imperial mint. 
Such darics as made their way thither would thus constitute 
an artificially inflated currency, and would tend to be exported 
again on the earliest possible opportunity. There was no tempta- 
tion to accumulate them, when they could be exchanged elsewhere 
for silver at so very substantial a profit The conclusion here 
suggested is fully borne out by the actual phenomena. Persian 
gold has never been discovered in any quantity in India ; the 
hoards of * darics ' sometimes said to have been found in the 
eighteenth century can be shown to have consisted of Gupta coins. 
Isolated examples have, indeed, been picked up sporadically ; the 
daric reproduced on PI. I, 1, is fi-om the Cunningliam Collection. 
But it is significant that in no single instance do these bear 
countermarks or any other indication that could possibly be inter- 
preted as suggestive of a prolonged Indian sojourn. 

The corresponding silver coinage consisted of sigloi or shekels, 
twenty of which were equivalent to a daric. They had a maximum 
weight of 86*45 grains (56 grammes), and had the same types as 
the gold (PI. I, 2, 3). Sigloi are frequently ofiered for sale by 
Indian dealers, and it is a reasonable inference that they are 
fairly often disinterred from the soil of India itself. That is 
precisely what might be expected from the working of economic 
law. The relative cheapness of gold would act like a lodestone. 
Silver coins from the west would flow into the country freely, and 
would remain in active circulation. At one time confirmation 
seemed to be provided by the surviving sigloi. Many of them — 
including, it should be added, a very large proportion that are not 
directly of Xw^vxa. provenance — are distinguished by the presence of 
peculiar countermarks which were thought to have their closest 
analogy on the square-shaped pieces of silver that constitute the 
oldest native coinage of India ^. The punch-marks on the native 
Indian coins (PI. I, 4, 5) appear to have been affixed partly by 
the local authority of the district in which the money was used, 
but to a much larger extent by the merchants or money-changei"s 
through whose hands it passed. The practice was plainly designed 
to obviate the necessity for repeated weighing. As this advantage 
would be as pronounced in the case of the sigloi as in the case of 
the indigenous issues, it would not have been surprising to find that 
they had been subjected to similar treatment. M. Babelon has, 

1 BapsoD, J.R.A.S. 1895, pp. 865 ff. 

344 Ancient Persia^i Coins in India [ch. xiv 

however, expressed the view that the punch-marked sigloi should, 
as a rule, be associated with Lycia, Pamphylia, Cilicia, and Cyprus. 
And it must be admitted that the results of the most recent 
investigation^ rather tend to bear out his opinion. The resemblance 
to the Indian punch-marks remains noteworthy, but proof of 
absolute identity is lacking. 

1 Hill, J.H.S. 1919, pp. 125 ff. 

MAP 3 




The Cambridge History of India, Vol. I 

Map 3 

Cambridge University Press 



In the fourth century B.C. there is a sudden rift in the mists 
which envelop the ancient history of India. The regions disclosed 
are the Kabul Valley, the foothills through which the Five Rivers 
come down into the plains of the Punjab, the plains themselves, 
and the lower course of the Indus. The country, as we see it, is 
held partly by a number of independent tribes, governed by their 
own headmen and owning the authority of no king. But this 
primitive aristocratic type of community is holding its own with 
difficulty against another type of government, the monarchic. In 
parts of the country principalities have been formed under des- 
potic rajas, and between the different elements a struggle with 
varying vicissitudes is going on. The rajas are fighting to extend 
their authority over the free tribes and the free tribes are fighting 
to repel the rajas. The rajas are also fighting amongst themselves, 
and mutual jealousies lead to politic alliances according to the 
necessities of the moment ; we divine in this little world a conflict 
and shifting of antagonistic groups such as we can follow on a 
larger scale in the history of Europe. It is into this world that the 
Western invader plunges in 326 b.c. 

About ten miles north-west from where Rawalpindi now stands 
stood, in the fourth century B.C., the city of Taksha9ila (Taxila), 
long eminent among the cities of India as a great seat of learn- 
ing. In the year 327 it was the capital of a raja, whose principality 
lay between the Indus and its tributary the Jhelum (the ancient 
Vitasta, the Hydaspes of the Greeks) ^ Like Rawalpindi to-day, 

^ Although the courses of the great rivers of the Punjab have greatly changed in 
historical times and are still changing, their names may be traced with certainty from 
the Age of the Sigveda down to the present day. Those which are chiefly important 
in the history of Alexander's Indian campaign are : 

Ajicient Indian. 

Greek and Latin. 



'I>'56s, Indus. 



Ko)<p-^v, Cophen. 



IlodaTOi, Soastus, 



'T8d<XTrris, Hydaspes. 


Asiknl, later Chandrabhaga. 

' K.Kealvrjs, Acesines. 


Parushni, later Iravatl. 

'TdpaJjTTis, Hydraotes, 


Vipa?, later Vipa9a. 

"T0a(rts, Hyphasis. 



ZapdSpos, Zaradrus, Hesydrns. 


34^ Alexander the Great [CH. 

Taksha^ila guarded the chief gate of India from the north-west : 
it was the first great Indian city at which merchants who had come 
down the Kabul Valley and crossed the Indus about Attock arrived, 
three days' journey beyond the river. Its ruler was the first among 
the kings of the Punjab to hear any tidings which might come 
down from the highlands of Afghanistan of events happening 
behind those tremendous mountain walls. For many generations 
now the Punjab must have had some knowledge of what went on 
in the dominions of the King of Kings. For the Persian Empire 
founded two centuries before by Cyrus had been a huger realm 
than had ever, so far as we know, existed in the world under the 
hand of one man, and the power and glory of the man who ruled 
it, the splendour of Ecbatana and Persepolis, must have been 
carried by fame over the neighbouring lands. 

The rajas of Taksha9ila must therefore have long lent an ear 
to the rumbling of wars and rebellions which came across the 
western mountains. They may indeed have known next to nothing 
of what went on at the further extremities of the Persian Empire ; 
for the same realm which at its utmost extension eastward touched 
the Indus reached at its other end the Aegean and Black Seas; 
and the great monarchic Empires of the east are conglomerations 
too loosely organised for the troubles of one province to be neces- 
sarily felt in the more distant ones. The Indian princes may 
therefore have been ignorant of the fact that the Persian king at 
the other end of his realm had come into contact with a singular 
people settled in a quantity of little republics over the southern 
part of the Balkan Peninsula, along the coasts of Asia Minor, and 
in the intermediate islands, the people whom the Persians called 
collectively Yavanas (lonians). We do not know whether it even 
produced any considerable shock on the banks of the Indus, when 
a century and a half before 334 B.C. the Persian king had led his 
armies to disaster in the land of the Yavanas, although those 
armies included Indian tribesmen torn by Persian officers from the 
frontier hills, whose bones were destined to find their last resting- 
place on the field of Plataea thousands of miles away. Of the long 
struggle which went on for generations after that between the 
Yavana republics, especially the one called Athens, and the western 
satraps of the Great King perhaps no rumour was brought down 
the Kabul valley to Taksha9ila. 

But in 334 B.c. and the following years the struggle between 
Persia and the Yavanas took a turn which must have made talk 
even in the palaces and bazaars of the Punjab. The Indian 

xv] From Kandahar to Kabul 347 

princes learnt that a Yavana king had arisen in the utmost West 
strong enough to drive the Great King from his throne. It may 
be that the western provinces, Asia Minor and Egypt, were torn 
away in 334, 333 and 332 b.c. by the invader without yet bringing 
the Indian princes to realise that so huge a fact in the world as the 
Persian Empire was about to vanish. But there can have been no 
mistaking the magnitude of the catastrophe, when Darius III was 
flying northward for his life, when Alexander had occupied the 
central seats of government and set Persepolis on fire (330 B.C.). 
If this man from the West was going to claim the whole heritage 
of the Achaemenian kings, that would make him the neighbour of 
the princes of India. It must have been a concern to the raja of 
Taksha^ila and his fellow-kings to learn in what direction the 
victorious Yavana host would move next. And in fact the tidings 
came before long that it was moving nearer. When the winter of 
330 fell, it was encamped in Seistan, and with the spring moved to 
the uplands which to-day constitute the southern part of Afghani- 
stan. Here the awe-struck inhabitants, Pashtus probably, ancestors 
of the modern Afghans, saw the European strangers set about a 
work which indicated a resolve to make themselves at home for all 
time in these lands won by their spear. They saw them begin to 
construct a city after the manner of the Yavanas at a point com- 
manding the roads ; and when the rest of the host had gone onward, 
there a body of Europeans remained, established behind the fresh- 
built walls. If we may judge by analogies, some thousands of the 
native people were induced by force or persuasion to settle side by 
side with them in the new city. It was only one of the chain of 
cities which marked the track of conquering Hellenism. Like 
many of the others, this too was given the name of the conqueror. 
In the speech of the Greeks it was known as Alexaudria-among- 
the-Arachosians. To-day we call it Kandahar. 

A mountain barrier still separated the Yavana host at Kandahar 
in the summer of 329 from the Kabul valley, that is to say, from the 
river system of the Indus. And it would seem that, when the passes 
filled with the first winter snows, the Yavanas had not yet crossed it. 
But the army led by Alexander was one which defied ordinary 
obstacles. In winter^, under circumstances that made regular pro- 
visioning impossible, by extraordinary endurance"^ it pushed through 
the hills and descended into the Kabul valley. The princes of the 
Punjab might feel that the outlandish host stood indeed at the door. 

1 "Tir6 nXeidSoj Uaiv, Strabo, xv, C. 725. 
a Diod. xvn, 82; Curt, vii, 3, 12. 


Alexander the Great [ch. 

But Alexander, having reached the Kabul valley in the winter 
of 329-8, did not make an immediate advance upon India. Beyond 
the mountain range which forms the northern side of the valley, 
the Hindu Kush, lay the extreme provinces of the old Persian 
Empire towards the north-east — Bactria (whose name still survives 
in the city of Balkh) and the country now called Bukhara. Not 
only were these [irovinces still unsubdued, but the Persian cause 
was upheld there by a prince of the old blood royal. Alexander 
must beat down that opposition, before he could think of in- 
vading India. He waited therefore for the rest of the winter in 
the Kabul valley, till the spring should unblock the passes of the 
Hindu Kush. And again here the inhabitants saw the Europeans 
make preparations for permanent settlement. At the foot of the 
Hindu Kush, whence three roads to Bactria radiate^, on the site 
probably of the still existing village of Charikar, rose another 
Alexandria, Alexandria-under-the-Caucasus. In support of the 
Yavana colony to be left in this town, other little settlements were 
established at points a day's journey off in what were henceforth to 
be Greek towns ; Cartana, noted for the rectangular precision 
with which its walls were traced out (modern Begram, according 
to Cunningham) and Cadrusi (Koratas ?) are names given us. In 
this case we have an express statement that 7000 of the people of 
the land were to be incorporated as citizens of the new towns with 
those of Alexander's mercenaries who cared to settle in this 
region 2800 miles away from their old home^ Another new 
city, or old city transformed with a new Greek name, Nicaea, 
occupied apparently some site between Alexandria and the Kabul 
river ^ 

As soon as the snow was melted enough to make the Khawak 
Pass practicable, the Yavana army trailed up the Panjshir valley*, 
leaving little bodies of Europeans behind it to hold the Kabul 
valley under a Persian satrap and a Macedonian episkopos. The 
main body of the army once more contended with the hardships of 

1 Cunningham, Ancient Geography, p. 24. 

'^ Diod. XVII, 83 ; Curt, vii, 3, 28, according to the mss. has ' vii millibus Beniorum 
Macedonum.' Hedicke in the Teubner text amends this, perhaps too boldly, as ' vii 
millibus subactarum nationum.' 

3 The discussions of Dr Vincent Smith and of Sir Thomas Holdich as to the site 
of Nicaea — the former puts it at Jalalabad and the latter at Kabul — are invalidated by 
the fact that Nicaea, if we follow Arrian, was not on the river Kabul at all. Alexander 
from Nicaea advances towards the Kabul; d<piK6/iei>os is T^iKaXa.v...irpoiix^p^i- us iirl t6i> 
Kw^^va, IV, 22, 6. Mr M'Crindle curiously omits the words in his translation. Not 
Nicaea, but some place on the way to the river Kabul, was where the army was divided. 

* Holdich, Gates of India, p. 88. 

xv] The Raja of Takshaqila 349 

a passage over the high ridges and disappeared to the northwards. 
During the following twelve months (May 328 to May 327) such 
news of it as reached India showed that the Yavana king still pre- 
vailed against all enemies. As far as the Syr Daria (Jaxartes) the 
peoples of Eastern Iran were broken before him. In the early 
spring of 327 he was again moving to the south. 

The i-aja of Taksha9ila must have realised at this juncture that 
a momentous choice lay before him. It may be that the idea of a 
common Indian nationality, in whose cause he and his brother 
kings might stand together against the stranger, did not even 
occur to him : India was too large and too disunited for the mind 
to embi-ace it as a unity. But he might well tremble for his own 
power, if this new resistless deluge came bursting into the land. 
On the other hand it might perhaps be turned to his account. His 
policy was largely governed by his antagonism to the rival prince 
of the Paurava^ house (Porus), who ruled on the other side of the 
Hydaspes (Jhelum). The Paurava was indeed a neighbour to be 
dreaded. He is described to us as a man of gigantic and powerful 
build, a warrior-chief, such as in an unsettled world extends his 
power by aggressive ambition and proud courage. He had con- 
ceived the idea of building up for himself a great kingdom, and he 
was the man to realise it. He had already made an attempt to 
crush the free tribes to the east, pushing his advance even beyond 
the Hydraotes (Ravi), in alliance with the raja of the Abhisara 
country (corresponding roughly with the Punch and Naoshera 
districts in Kashmir) and with many of the free tribes whom he 
had drawn into vassalage swelling his army, although the resist- 
ance he there encountered from the Kshatriyas^ had made him 
temporarily give back*. His hand had perhaps also reached west- 
ward across the Hydaspes into the country which the raja of 
Takshaqila considered his own*. It might well seem to the raja of 
Takshagila that, threatened on the one side by the Paurava and on 
the other side by the European invader, his safest course lay in 
allying himself with the European, riding on the crest of the wave 
that would sweep his rival to destruction. 

And yet the European host which had emerged out of the 
unknown West to shatter the Persian Empire may have appeared 
too unfamiliar and incalculable a power to make the decision easy. 

' Paurava is a title denoting the chief of the Puras, a tribe known in Vedic times 
{v. 8up. Chapter iv, pp. 82 f.). 

2 In Greek Kathaioi, see Lassen, vol. ii, p. 167. The general designation of the 
warrior caste seems to be applied in this case to a particnlar people. 

* Arrian v, 22. * See Anspach, note 125. 

350 Alexander the Great [ch. 

But, if the raja hesitated, his son Ambhi (Omphis)^ had a clear 
opinion as to what the situation required. He pressed his father 
to place his principality at the Yavana king's disposal. While 
Alexander was still in Bukhara, Ambhi began to negotiate on his 
own account Envoys from Taksha9ila made their way over the 
ridges of the Hindu Kush. They were charged with the message 
that Ambhi was ready to march by Alexander's side against any 
Indians who might refuse to submit. Thus the European, at his 
first arrival at the Gates of India, found India divided against 
itself. It was the hand of an Indian prince, which unbarred the 
door to the invader. 

The summer of 327 b.c. was almost come^ before the hillmen of 
the Hindu Kush saw the Yavana army re-appear on the ridges, 
cross probably by the Kushan Pass^ and stream down to the new 
Alexandria. The satrap who had been left here was found to 
have done badly, and Alexander appointed another in his place, 
Tyriespes, a Persian like his predecessor. The population of the 
city was enlarged by drawing in more of the people of the land 
and settling down there more war-worn European veterans. The 
work of making a city of Greek type had really only been begun, 
and a Macedonian of high rank, Nicanor*, was now appointed to 
see it carried through. 

The army moved on from Alexandria to Nicaea, where Alex- 
ander sacrificed to the Greek goddess Athena. From Nicaea he 
sent on a herald to the raja of Taksha9ila and the native princes 
west of the Indus to meet him in the Kabul Valley. We know of 
one Indian chief, 9a9igupta (Sisikottos), already in the conqueror's 
train. His had been probably some little hill-state on the slopes 
of the Hindu Kush, whence he had gone two years since, to help 
the Iranians in Bactria against Alexander. When their cause was 
lost, he had gone over to the European. Messengers now summoned 
the other chieftains of the lower Kabul Valley to meet their over- 
lord. At Tak8ha9ila too messengers appeared with the summons. 
And the raja, acting on the policy which his son had espoused 
so decisively, rose up to obey. 

Encamped in the Kabul Valley at some place not named the 
raja of Taksha9ila saw the host destined for the invasion of his 
mother-land. It numbered, at the lowest estimate, from twenty- 

^ See Sylvain L^vi in Journal Asiatique, 8"" S6rie xv (1890), p. 234 f. 

2 'E^-^KovTos ijSr) ToO ijpos, Arr. iv, 22, 3. 

3 Strabo xv, C. 697 ; Cunningham, Ancient Geography, p. 25. 

* Dr Vincent Smith {Early History of India, 3rd edition, p, 49) seems to be in error 
in identifying this Nicanor with the son of Parmenio. 

xv] From Kabul to the Indus 351 

five to thirty thousand men^ — a strangely compounded army, which 
can only be called European with qualification. Its strength 
indeed consisted in the Macedonian regiments, stout yeomen and 
peasants carrying the long spear of the heavy-armed foot-soldier, 
and troops of splendidly disciplined cavalry drawn from the aris- 
tocracy of the country, the ' Companions ' of the national King. 
European too were the thousands of soldiers from the Greek cities, 
serving as mercenaries, on foot or mounted, and the contingents of 
semi-barbarous hillmen from the Balkans, Agrianes and Thracians, 
serving as light troops — slingers, javelineers, and bowmen — invalu- 
able for mountain warfare. But mingled with the Europeans 
were men of many nations. Here were troops of horsemen, repre- 
senting the chivalry of Iran, which had followed Alexander from 
Bactria and beyond^, Pashtus and men of the Hindu Kush with 
their highland-bred horses^, Central-Asiatics who could ride and 
shoot at the same time* ; and among the camp-followers one could 
find groups representing the older civilisations of the world, 
Phoenicians inheriting an immemorial tradition of ship-craft and 
trade, bronzed Egyptians able to confront the Indians with an 
antiquity still longer than their own. 

There was nothing to arrest this army between the point they 
had noAv reached and the Indus. The local chieftains had indi- 
cated their submission. All along the north side of the Kabul 
however lay the hills, whose inhabitants in their rock citadels, in 
the valleys of the Kunar, the Panjkora, and the Swat, were un- 
schooled to recognise an overlord, and as prepared to give trouble 
to anyone who tried to incorporate them in an imperial system as 
their Pathan successors of a later day. But it was not Alexander's 
way to leave unsubdued regions beside his road. His army there- 
fore broke up into two divisions. One, commanded by Hephaestion, 
the king's friend, and Perdiccas, the proudest of the Macedonian 
nobles, moved to the Indus by the most direct route. This would 
probably mean a route along the south bank of the Kabul, whether 

^ The numbers in the ancient texts are often untrustworthy. The estimate in the 
text is Delbriick's, Qeschickte der Kriegskunst (1900), vol. i, p. 184. Anspach (note 20), 
combining Arrian, Ind. 19, 5 with Diod. xvii, 95, reckons the army in the Kabul Valley 
at about 85,000. Delbriick denies that so large an army with the necessary camp- 
followers could have got across the Hindu Kush. This is a point for practical strategists. 
Whether Plutarch's number {Alex. 56) is correct or not, he does not say, as 
Dr Vincent Smith, p. 49, inadvertently quotes him, that Alexander entered India 
with 120,000 foot and 15,000 horse, but that Alexander left India with that number. 
Reinforcements had been arriving from the West in the meantime. 

2 Arr. IV, 17, 3. ^ 75, y^ n^ 3, 4 7^. ly^ 24, 1. 

35^ Alexander the Great [ch. 

through the actual Khyber Pass or not^ ; the other, led by the king 
himself, turned up into the hills. The two divisions were to rejoin 
each other upon the Indus; Hephaestion and Perdiccas, arriving 
there first, it was calculated, would have made all preparations for 
the passage of the great river. 

The Europeans who had followed Alexander so far into Asia 
now entered the region in which the armies of the English operate 
to-day. At that season of the year^ the hill-country must have 
been bitterly cold, and probably to some extent under snow. It 
was the same hill-country whose contours and tracks and points of 
vantage are studied now by British commanders ; the tough high- 
lander of the Balkans or of Crete climbed and skirmished with 
bow and javelin in 327 B.C. where the Scottish highlander was to 
climb and skirmish with rifle and bayonet two thousand two 
hundred years later. And yet it is impossible to follow the track 
of Alexander over these hills with any precision. We hear of little 
mountain towns stormed, of others abandoned by their inhabitants. 
But their sites cannot be identified. One must however note that 
at this point Alexander, in an ethnograpliical sense, entered India ; 
for these hills, whose population at the present day is either 
Afghan or Kafir, seem then to have been possessed by Indian 
tribes. The A^vakas, as their name apparently was in their native 
speech, were the first Indian people to receive the brunt of the 
invasion ^ The fighting seems to have been of exceptional ferocity. 
At one place, where Alexander was wounded, the whole popula- 
tion was put to the sword. At another place we hear of a huge 
massacre, and 40,000 men taken captive. At a third place a body 
of Indians from the Punjab had come to help the local chieftain 
for hire. When the town capitulated, it was agreed that these 

^ Dr Vincent Smith says that he did not go by the Khyber and cites Sir Thomas 
Holdich in support of the assertion. Sir Thomas in his more recent book, Gates 
of India (p. 94), says that he ' undoubtedly followed the main route which... is sufficiently 
well indicated in these days as the " Khaibar ".' 

2 MerA Sv<t/jAs EXetdSwc, Aristobulus ap. Strabo xv, C, 691. 

' In the Greek accounts a people called Aspasioi are found in the Choes (either the 
Alishang or Kunar) Valley and a people called Assakenoi in the Swat Valley. Both 
names are supposed to represent the same Indian name A9vaka, connected with agva 
• horse. ' If so, the two Greek names may be due to local varieties of pronunciation, 
and it may be noted that the form Aspasioi would then approximate to Iranian speech, 
in which aspa is the equivalent of agva. Strabo, according to the mss. (xv, C. 691, 
0. 698), calls the Aspasioi Hypasioi ; this is often emended in modern texts to 
Hippasioi, on the supposition that the Greeks knew their hippos to be the etymological 
equivalent of aspa and attempted a translation. This is extremely unlike the Greek 
way in these matters. The confusion is made worse by another people called Astakenoi 
appearing in the PushkalavatI region, whose name is supposed not to be connected 
etymologically with that of their neighbours, the Assakenoi. 

xv] Tribes beyond the N,W, Frontier 353 

mercenaries should transfer their services to Alexander. They 
encamped on a little hill apart. There, as they talked together, it 
seemed to them a horrible thing that they should march with the 
Yavanas against their own people. They determined to slip away, 
when night fell, and make across the hills for home. But when 
night fell, they found the hill beset on all sides with the soldiers 
of Alexander; for some one had betrayed their design. The 
Macedonians suffered none of them to live till morning \ 

The town with which this incident is connected the Greeks call 
Massaga. We know only that it was situated east of the Guraeus 
river and apparently not far from the stream. The resistance 
which the easternmost branch of the hill-people, those called by 
the Greeks Assakenoi, offered to the invader seems to have been 
concentrated at this place. All these tribes, as far as the Indus, 
recognised as overlord a chief whom the Greeks call Assakenos. 
His organisation for defence included an alliance with the king of 
the Abhisara country beyond the Indus, who sent contingents to 
his support^. Assakenos had himself taken command at Massaga, 
and fell there, struck by a missile from one of the European siege- 
machines ^ His mother and daughter were left in the enemies' 
hands* ; but it was not among Alexander's faults to fail in chivalry 
to the women whom war put at his mercy ^ 

The loot in cattle in these regions was enormous, and we are 
told that a herd of the finest animals was actually given by Alex- 
ander into the charge of drovers who were to drive them all the 
way from the Hindu Kush to Macedonia. A town called by the 
Greeks Arigaeon, which apparently commanded the road between 
the Kunar and the Panjkora Valleys, was selected for recolonisa- 
tion — a number of war-worn Europeans and a number of the native 
people were to form the population, as in similar cases before. 

One curious incident relieves the story of blood-shed. Some- 
where among these hills* — probably on the lower spurs of the 

1 Arr. IV, 27; Diod. xvn, 84; Plut. Alex. 59 ; Polyaen. Strateg. iv, 3, 20. 

2 Arr. IV, 27, 7. » Arr. iv, 27, 2. •* Arr. iv, 27, 4. 

* A strange story is given by Justin, xn, 7, 10, that it was the wife of Assakenos who 
fell into Alexander's hands and that he had a son by her, who afterwards became king 
of the Indians (!) (cf. Curtius, vni, 10, 35). It may be that the story was concocted in 
later times in the interests of some petty king of this region, who wished to establish 
a claim to be descended from Alexander. That is a claim which is still common in 
the Indian frontier hills. 

* Holdich in discussing the site of Nysa (Gates of India, p. 122) gives a mis- 
translation of Arrian. Arrian does not say that Alexander ' then entered ' that part 
of the country, but that somewhere in the country which Alexander had already 
traversed there was a place called Nysa. 

c. H. 1. 1. 23 

354 Alexander the Great [ch. 

three-peaked Koh-i-Mor — dwelt a people who told the Yavanas, 
or so the invaders understood them, that they were descendants of 
the western people who had come into those parts with their god 
Dionysus ; for Dionysus, the Greeks believed, had gone conquering 
across Asia, at the head of his revellers, in the old heroic days. 
The Greeks always experienced a keen joy of recognition, when 
they could connect foreign things with the figures of their own 
legends, and they were delighted with the suggestion. The asson- 
ance of names lent itself immediately to confirm the theory as 
usefully as it does to confirm the adventurous speculations of 
modern archaeologists. In the legend the name Nysa was specially 
connected with Dionysus — it was the name of his nurse or of the 
place where he was born or of his holy hill — and the name of this 
little town in the Hindu Kush, as it was pronounced to Alexander, 
had a similar sound. Again the legend said that Dionysus had 
been born from the thigh (meros) of Zeus, and a neighbouring 
summit, the Greeks discovered, was called Meru. What could be 
clearer? And when they saw the sacred plants of the god, the 
vine and ivy, running wild over the mountain, as they knew them 
at home\ no doubt could be left. Modena travellers have come 
upon certain fair Kafir tribes in this region, whose religious pro- 
cessions with music and dancing have a Bacchanalian look, and 
the Nysaeans discovered by Alexander, they suggest, may have 
been the ancestors of these Kafirs ; their processions may have led 
the Greeks to connect them with Dionysus. This is possible, but 
in the Greek books we hear nothing of the Nysaeans going in 
procession. It is the Macedonian soldiers themselves, who wreathe 
their heads with ivy and range the hills in ecstasy, calling on the 
god by his sacred names, as their people had done from old time 
on the woody spurs of the Balkans. Hostilities, at any rate, 
with these interesting kinsmen could not be thought of, and the 
'Nysaeans' were themselves prepared to act in character; three 
hundred of them on their mountain horses joined the army of the 
Yavana king and followed him to battle in the plains of the Punjab. 
Whilst Alexander was fighting in the valleys to the north of 
the Kabul, the other division of the Macedonian army under 
Hephaestion and Perdiccas, accompanied by the raja of Taksha9ila, 
made its way along the Kabul to the Indus. It may have been 
through the Khyber Pass that, one day in the cold weather season 
at the end of 327 or beginning of 326 B.C., the glitter of strange 
spears, long lines of mailed men, were seen emerging into the 

^ Holdicb, Gates of India, p. 133. 

xv] Occupation of the Lower Kabul Valley 355 

plain about Peshawar — the advance guard of the European in- 
vasion of India. A few days' march farther, and they came to the 
Indus. Arrived there, the Europeans set about collecting material 
for the bridge which was to transport their fellows into the interior 
of the land. But their hold on the country west of the Indus was 
not yet secure. The region in which the division of Hephaestion 
and Perdiccas was now encamped formed part of the realm of a 
raja, named by the Greeks Astes^, whose capital was the town 
of Pushkalavati (Charsadda) to the north of the Kabul river. 
The raja at this moment declared himself an enemy of the for- 
eigners. He was not strong enough to hold the open field against 
Hephaestion and Perdiccas, and shut himself desperately in some 
walled town. For a month he held it against the besiegers, and 
then the greater strength of the Europeans beat him down, and 
destroyed him. The principality was given to one who had been 
his enemy and become a hanger-on of the raja of Taksha9ila, a 
certain Sangaya. He was a man upon whose loyalty the Yavanas 
could count. 

In the hills to the north, after a few months' fighting, the tribes 
generally had submitted to Alexander and the strong places were 
in his possession. He constituted the lower Kabul Valley and 
the recently conquered hills a special satrapy, distinct from the 
satrapy of the Paropanisadae, which Tyriespes ruled from Alex- 
andria-under-the-Caucasus. The new satrapy, whose official name 
we do not know, but which can be most conveniently described as 
India-west-of-the-Indus, got for its governor a Nicanor, probably 
the same man who had been left a few months before to superintend 
the building of Alexandria. The king himself came down to Push- 
kalavati at the lower end of the Guraeus (now usually called the 
Swat) valley, which was not in a position, after the defeat of its raja, 
to ofi*er any resistance. He set a Macedonian garrison in the town 
under an officer named Philip. 

But the effective occupation of the lower Kabul valley by the 
Yavanas required still more to be done. The division of Hephaes- 
tion had meantime fortified and garrisoned a place the Greeks call 
Orobatis, and Alexander, accompanied by two Indian chieftains, 
'Cophaeus' and ' Assagetes'^, moved about to take possession of 
various small towns between Pushkalavati and the Indus. But 

^ One guess is that this represents the Sanskrit proper name Hasti; a more 
probable one is that it is short for Ashtakaraja, king of the Ashtakas. 

"^ Anspach suggests, p. 65, note 200, that Cophaeus = raja of the region about 
Pushkala near the Cophen, and Assagetes=the raja of the Assakenoi (successor of the 
raja killed in Massaga). 


356 Alexander the Great [ch. 

one great labour remained. The reduction of a certain mountain 
citadel, which crowned Alexander's work during that winter, 
always seemed to the Greeks the great glory of the campaign. 
The Greek books described the siege and storming at greater 
length than any other episode in this region. The story was 
started that Heracles had attempted to storm that very rock and 
failed. Unfortunately, it has so far been impossible to fit the 
Greek description of Aornus to any rocky height noted in the 
country to-day ^ 

Aornus, we are told, was not far from the modern Amb ; it was 
a great isolated mass of rock, 6670 feet high, flat on the top with 
precipitous sides, which on the south went down straight to the 
river Indus. On the summit were woods and watersprings and 
fields whose cultivation could keep a thousand men employed. 
It seems plain that an object of this kind can hardly have escaped 
modern geographers in search of it. The inference is that some 
particulars in the Greek account are due to imagination. But 
when once we begin to trim it so as to suit the actual topo- 
graphy, it depends on a more or less arbitrary selection which 
particulars we eliminate and which we retain. There is at any 
rate no reason to doubt that the final conquest of this mountain 
region did involve the reduction of some exceptionally strong 
rock-citadel, in which fugitives of the defeated tribes made a last 
stand. The citadel, when taken, was held for Alexander by a 
garrison under the Indian (^a9igupta. The capture of Aornus had 
to be followed by another short expedition further up into the 
hills, in pursuit of the flying defenders of the fortress". They were 
led by the brother of the Assakenian chief killed in Massaga^ and 
had with them a herd of fifteen war-elephants. To the Greeks the 
idea of getting hold of these animalsTso strange and wonderful to 
them, of whose value in battle they had probably formed an even 
exaggerated notion, made their pursuit the more eager. The hills 
were deserted before them, and Alexander pushed on as far as a 
town which the Greek books call Dyrta. It was found empty of in- 
habitants. Alexander learnt that the fugitive prince was dead by 
the evidence of his severed head, brought by some hillmen one day 
into camp. He had fallen a victim to some hostile tribe or to his 
own followers. Two bodies of light troops were detached to scour 
the hills yet further, and Alexander himself turned back with the 

1 See the note in Vincent Smith, pp. 56, 57. 

* Arr. IV, 30, 5 ; Diod. xvii, 86 ; Curt, viii, 12, 1. His name is variously given aa 
Aphrikes, Aphices, Erices in different texts. 

XV] The Crossing of the Indus 357 

rest of his division to the Indus. Some natives of the region were 
caught by the Macedonians on their way. They reported that the 
fugitives from Aornus and the people of the hills had escaped into 
the country of Abhisara, whose raja was watching the progress of 
the Yavanas with a doubtful mind. As for the elephants, they 
had turned them loose in the country bordering the Indus, more 
swampy in those days than it is now\ An elephant hunt accord- 
ingly followed ; Alexander had already, with his quick interest in 
new things and his Macedonian sporting propensities, collected 
about him Indians whose special business was elephant-hunting, 
and by their means the scattered herd was driven in, and attached 
to the Macedonian army^. The point at which Alexander's division 
struck the Indus on its descent from the hiUs was some way above 
the point where Hephaestion and Perdiccas had by this time con- 
structed the bridge. Between the two, the right bank of the river 
was largely overgrown with forest, which, if in one way it impeded 
the advance of Alexander's division, in another way helped the 
transport by furnishing timber for boats. Part of Alexander's 
force floated down the river, and when he arrived at Hephaestion's 
bridge the number of new boats was swelled by those brought 
down from up-stream. The two divisions of the Yavana host now 
re-united for passage into the heart of India. The place at which 
the bridge had been made has been fixed by the most recent 
opinion at Ohind, about 16 miles above Attock. The Greeks 
felt that they were crossing the threshold of a new world. Sacri- 
fices to the Greek gods, games and horse-races in their honour on 
the river bank at Ohind, marked their sense that they were about 
to begin a new enterprise of formidable magnitude. Alexander 
was approaching the bourne of the old Persian Empire, and it was 
evident that he meant to press still onward towards the sunrise. 
The Greek divinei*s announced that the omens were favourable. 
In the early dawn one day in the spring of 326 b.c.^, the host began 
to defile over the bridge, the mingled line of many races streaming 
all day into the Indian world. And the composition of the army 
became now more singularly mixed by the contingents of native 
Indian troops sent by the raja of Taksha9ila, squadrons of Indian 
horse and thirty elephants, endless trains moreover of oxen and 
sheep for sacrifice and food, and silver brought in masses from his 

The raja of Taksha^ila was now none other than Ambhi himself ; 

1 Holdich, Qatet of India, p. 122. ^ Cf. Diod. xvu, 86. 

3 Stiabo XV, C. 691. Cf. Vincent Smith