Skip to main content

Full text of "Original Sanskrit texts on the origin and progress of the religion and institutions of India; collected, tr. into English, and illustrated by notes. Chiefly for the use of students and others in India"

See other formats

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project 
to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A pubUc domain book is one that was never subject 
to copyright or whose legal copyright term has expired. Whether a book is in the pubUc domain may vary country to country. PubUc domain books 
are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other marginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 
publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with Libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we have taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 

We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain from automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attribution The Google "watermark" you see on each file is essential for informing people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from countiy to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement Liability can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 

at http : //books . google . com/ 


oir XHX 


tiyai, ft ly /lii'yiytiiaiatiur, PapfiaputSv tout' Itrat. {ti| itiv'fivf Imat or t( TJ i\r)0»{f 
xai TMOVTOi' avnr, lti| t) Kay i«i vmAot JniTW ri irpwra run ifi/uiriar oynipcrm tiviu' 
tick V^ rk rovraxn orp/^ofai Tcl oi'^fuira ivjii' (avfioOTov &y iir) » ^ iraAoui ^>^ 
wpht Ti)y ran JSop^opucqt fiq<<>' iw^^w^ PlATO, Oo^vltw, i. 4£1. 

" We might nqr that tho words which we did not undentand were derived from 
the barbariatia. Some of them might in reality be such ; and it might also be the 
case that, owing to the lapse of time, the earliest forms were nndisooverable : for as 
a result of the circulation of words in all parts of the world, it would not he at all 
strange if the ancient language, as compared with the modem, was in no respect 
different flrom tho speech of the barbarians." 










J. MUIR, ESQ., D.C.L. 








/% 6 

JVIY primary object iii tliis volume, as in its predecessor, 
has bccu to priwluce a work which may assist tlie re- 
searches of those Hindus who desire to investigate critically 
the origin and histoiy of their nation, and of their national 
literature, religion, and institutions; and may facihtiite 
the operations of those European teachers whose busi- 
ness it is to communicate to the Hindus the results of 
modem enquiry on the various subjects here examined.' 
The book (as will at once bo apparent to the Orieiitul 
^p^M^holar) is, for the most part, either a compilation, or, 
at least, foimded on the labours of others ; but while my 
principal aim has been to furnish the reader with a sum- 
mary of the results of preceding enquiries, my plan has, at 
the same time, rendered it necessary for me occasionally 
to institute fresh researches in diflerent directions for the 
hirther elucidation of particular points which were touched 
upon in tlie course of my argument. In this way I may 
have succeeded in contributing a small proportion of ori- 

' This peculiarity in the object of the treatise will account to the Euro- 
pean scholar for the introduction of many details which would othcrwiic 
Lave been quite superfluous. 



ginal matter to the discussion of eome of the interesting 
topics whicli have come under review. M 

The obligations under which I lie to the different^ 
authors, whose labours have furnished the chief materials 
of tliis work, have been, in most instances, so fiilly ac- 
knowledged in detail in the following pages, that it is 
not necessary for me to allude to them here more parti- 
cularly. I must, however, refer to the assistance which 
I have derived from the French version of the Rig-veda 
by M. Langlois, which, with his index, has directed my 
attention to various important passages in the later books, 
which I was then enabled to study in tlie original 

Though a small portion only of the present volume 
consists of " Sanskrit texts," which in some parts are 
altogether wanting, and in others but thinly sciittered, 
(apparent ran nantes in gurgite vasto,) I have not con- 
sidered it necessary to abandon the old title, but it has 
been slightly modifiecL 

Althougli some idea of the object and contents of the 
volume may be gained from a penwal of the introductory 
statement of its plan (in pp. 1 — 3), and from the table at the 
close of this Preface, it may conduce to the convenience 
of those readers, who before entering on a perusal of the 
work desire to obtain a more precise conception of the 
course of the discassion, and of the process by which I 
have sought to estabUah my conclusions, if I subjoin here 
a brief concatenated sumraaiy of tlie principal topics in 

The general object of the present Part is to prove 
that tile Iliiidus were not indigenous in Lidia, but have 






ited into tliat country from Central Asia, where 
their ancestors at one time formed one commmiity Avith 
the progenitors of tlie Persians, Greeks, Komans, Geraiaus, 
&c Li order to establish this result I have sought to 
show that Sanskrit, tlie original language of tlie Hindus, 
exhibits imdeniable marks of close affinity to the ancient 
lauguagea of the other races just mentioned; and that 
the earUest religion, and mythology also, of India are con- 
nected with those of Persia by various points of contact 
and resemblance. Having adduced evidence on both these 
heads, and urgued that these facts imply a common origin 
in the nations in question, and their subsequent dispersion 
from one common centre towards the diflferent regions in 
wliich they idtimately settled ; I endeavour to fortify the 
conclusions to wiiich we are thus conducted by demon- 
strating tliat, in the earUest ages of their history, the an- 
cestors of tlie Hindus appear to have occupied only the 
north-western corner of Hijidusthan ; and that, while they 
were connected on the one hand by allinitiea of language 
ond religion with the nations of the west, they were on 
the other hand distinguished, both by language and by 
institutions, from certain otlier tribes with whom they 
came mto colhsion as they advanced across the north of 
IndijL, and afterwards diiTused themselves to the south of 
the j)eniusida : for if we find that the Hindus origuudly 
possessed oidy the Panjab, the presumj)tii m (tlerived fix)m 
other considerations) that they immigrated from the north- 
west, becomes strengthened; and if, again, on their advance 
to the south-east, they encountered tribes with a diilerent 
language and religion, already in occupation of those 



tracts, the protability that they tlid not grow up in Indk, 
alougside of these alien tribes, acquires additional force. 

Ill order to obtain a basis for carrying nut the philologi- 
eid portion of this argument, viz., for coniparing the ori- 
ginal language of the Hindus with those of Uie Persians, 
Greeks, and Latius, it became necessary for me to prove 
that the Sanskrit, which is now a learned language only, 
was at one time spoken by the ancestors of tlie Hindus. This 
I have attempted to do in the First Chapter (pp. 4 — 223), 
by showing in detail that the original Sanskrit idiom has 
undergone a long series of gi*aduul mutations, of which 
we now see the ultimate result in the modem vernacular 
dialecta of the north of Liditu The method which I have 
adopted to exliibit this, has been to begin (Section i., pp. 
4 — 10 J with the existing spoken dialects, Urdu, Hindi, 
Molu'atti, &c., and to show what the elements are of which 
they are composed, viz., (1) pure Sanskrit, (2) modified 
Sanskrit, (3) Desya or aboriginal non-Siinskrit words, and 
(4) words derived from Arabic and Persian, Tlie fourth 
element is the latest which they have acquired, and dates 
only from the M;ihomcdan mvasion ; wldle the second 
and thii'd (in a more or less different form) are common 
to them with the Prakrits, or older vemacidar dialects, 
out of which tliey grew. 

In the succeeding sections (il— vii., pp. 11 — 137) an 
account is given of these eai'her vernaculars, viz. (1) 
the Prakrits, of whicJi specimens are to be found in 
the different Hindu dramas, and which seem to have 
existed as spoken dialects, at ] from Ihe cummence- 
iiieut of the Chrisitiau era until tliey became jnergcd in 




tho modem vernaculars; (2) the Pali, or sacred langiia^u 
of the BuddJiist b<x)ks of Ceylon and Burmah, whicli 
appears to represent one of the provmcitd dialecb 
of nortlicni ludia existmg at the time when Buddlusm 
began to be propagated in the 6th century B.C., and 
exliibits to us die popular speech of that region at a 
■Somewhat earlier stage tlian the dramatic Prakrits ; (3) 
the dialects (nearly contemporaneous with the Pah) which 
l^e employed in the rock and pillar inscriptioas of A&oka ; 
and (4) the singular ditdect or jargon employed in the 
Gatlias or metrical porlion8 of the Buddhist chronidus of 
uortliern Lidia. In tliis portion of die work some com- 
parative tables are introduced, which exhibit (1) the 
relations, {i.e. the i>ointa of resemblance and of difference) 
between the modem vernaculars, Hindi, and Midiratti, 
and the dramatic Prakrits, and show how tlie two 
former have been fonued by a modification of all the 
various elements of the latter, just as tht-y (the older 
Prakrits) in their turn, have sprung up (if we except a 
small uon-Sanskritic residuum) from the gradual decomposi- 
tion of the Sanskrit; (2) the fonns which are common to 
the dramatic Prakiits, and the Pali, as well as those points 
in which they vary, and which demonstrate tliat the Pidi 
diverges considerably less from the Sanskrit than tlie 
Prakrits do, and must consequently be more ancient than 
they ; and (3) the relation in which the rock inscriptions 
stand to the Pah. In Section viii. (pp. 138 — 153) the 
conclusion is drawn that, as the vemacidar speech of 
India, as for back as we are able to trace it, has been 
undergoing a continual ^eri(:•s <if mutations, and as the 


older the form is in which we find it existing, the ncyirer 
it approaches to the Sanskrit in its words and its gram- 
matical inflections, — it must at some period a Uttle fui'ther 
back have entirely merged in Sanskrit and have been 
identical with it. Tlius Sanskrit havinsr been once the 
Bame with the oldest spoken language of India, must at 
that period have been a vernacular tongue. After some 
speculations on the history of the Sanskrit language and its 
mutations, some further arguments, — drawn partly from the 
parallel of Latin (which though once a spoken tongue, 
wa^ ultimately lost in its derivative dialects, Italian, &c.), 
and partly from certain phenomena in Indian liteiuture, 
or notices occurring in Indian authors, — are adduced in 
Section is. (pp. 153 — 168) in support of the position that 
Sanskrit was once a vemacidar language, and that the 
Vedic hymns were composed in the same dialect which 
their authors habitually spoke, I then go on to argue 
further (Section x., pp. 169 — 223) that as Sanskrit was 
once a spoken tongue, it nmst in its earlier stages have 
been exposed to all tlie mutations to which all spoken 
languages are subject. That such has actually been the 
case, is clear from a comparison of the oldest Sanskrit, that 
of the Vedic hynms, with the form which it took in the 
later literal m^e, and which (as it became exempt fix)m 
fui'tlier modifications by ceasing to be popularly spoken) it 
has continued ever after to retain. As, however, the distinc- 
tion which is here drawn between the older and the more 
recent literatiu-e may be disputed by the Hindu student, 
I have considered it necessary to adduce proof of the 
assertion that the Vedic hynms are the oldest of all the 




Intlian writings ; and with this view to ascend by gradual 
stepB from the most recent commentaries on the Veda, 
through the Niinikta, the Brfihmanas, &c., to the hymn- 
collections, pointing out that each of these classes of works 
presupposes one of the others to liave preceded it in 
regular order, and that such methods were employed by 
the commentators for the interpretation of the hymns as 
to prove that much of tlieir language wasah-eady obsolete 
or obscure, and that consequently tlieir priority in time to 
the very oldest of their expositors must have been veiy 
considerable. To complete the survey of the subject, I 
further show, that there is a difference in the ages of the 
several Vedaa (the Eik, Yajuah, Saman, &c.) themselves, 
as well as between the different portions of each, as is dis- 
tinctly evidenced by their contents. The superior antiquity 
of the Vedas to tlic other Indian writings is next proved 
by a statement of the differences discoverable between the 
religious systems of these two classes of works, the natiire- 
worship of the Vedas supplying the original germ, out of 
which the Puranic mytholog}'- was slowly developed with 
innumtn^ble modifications. The greater age of the Vcnlas 
is then shown by comparing a number of their grammatical 
forms with those of the later Sanskrit. Finally, I revert to 
the conclusion before indicated, that the language in wliich 
the Vedic hymns were composed can have been no other 
than the vernacular speech which was employed by tlie 
rishis and their contemporaries, as it is quite inconceivable 
that in tliat early age, when the refmements of grammar 
were imknown, there could have existed any learned lan- 
guage distinct from the ordinary dialect of the people. 



ELaviiig thus shown cause for belicvmg that Sanskrit, the 
original speech of the early Hindus (or Indo-Ariaus) was at 
one time a spoken language, and consequently liable, like 
all other spoken languages, to continual mutations in its j 
earliest ages, and having by this means paved the way for V 
proving that it is descended fix>m one common mother 
with the ancient languages of the other ludo-Eiiropean 
races, to which it exhibits the most striking family resem- 
blance; — I proceed, in the Second Chapter (pp. 224 — 
372) to produce the evidence wluch c^imparative philology 
fiuTiishes of this resemblance, and to argue from the affinity 
of Ismguages a community of origin between the diflcrent 
nations by which they were spoken. I tlien go on to 
biing fonvard the fiu'ther grounds, supphed by comjiarap 
tive mjtliology and by other considerations, for supposing 
that the ancestoi's of the Hindus belonged to the same ^oal 
family as the Persians, Greeks, Romans, &c,, wliich had ii 
original seats in Central Asia, and that, on the disper^ion^ i 
various directions, of the different branches of that ancient 
family, the Indo-Arians immigrated, into Hindusthan from 
the north-west. The follo>\dng are some of the details of this 
process of proof : In Section i. (pp. 226 — 232), a few simp] 
remarks on comparative pliilology are premised, in which 
it is sho^vn how, by a comparison of dieu* roots and structure, 
languages can be distributed into different fiimilies, of which 
the sevei-al members have a more or less close affinity 
each other, wliile they have httle or no resemblance to thi 
members of any otlier family. Tliis is illustrated by 
comparative ttible, in wliich it is shown that while Sanscrit 
has in many of its words a strong similmity to Persia 




has scarcely any to Arabic ; and by some other particu- 
lars. Section ii. (pp. 233 — 277) supplies dcttiilod evidence 
of the affinities of Sanskrit with the Zend, Greek, and 
Latin, consisting, first, of comparative lists of words bo- 
longing to those languages wliich correspond with encli 
other both in sound and sense; and secondly, of illu.s- 
trations of the resemblances between those langimges in 
their modes of inflection, as well as in the formation of 
wordii. 2\a, however, the mutual diflerences wliich tlicse 
languages also exliibit, might be urged as disproving the 
inference of their derivation from a common source, it 
is shown how, in the course of time, dilierent branches 
of the same original tongue have an inevitable tendency 
to diverge more and more from the prijiiitive type, 
both by modifying their old elements, and by assimilating 
new: and it is further pointed out that it is precisely 
those parte of a language which are the most primitive 
and essential in wliich the diflerent Indo-European 
tongues coincide, while those in which tliey diiTer are 
such as would grow up after the nations whicli spoke 
them had been separated, and had become exposed to the 
action of diverse influences physical and moral. But 
as, admittuig the resemblances between these languages 
n Hindu might feel disposed to draw the conclusion that 
Sanskrit is the source of all the other kindred tongues, 
la^itt'ad of being derived together with them from an older 
language, the common parent of them all, — to obviate 
this erroneous inference, it is next shown that the wliole 
grammatical character of Greek and Latin is that of 
independent languages ; that in this respect they diflVr 



entirely from the Indian Prakrits (which have evidently 
resulted from the decomposition of Sanskrit), and that 
they even contain various forms which are older than those 
of the Sanski'it; while the greater part of their vocabu- 
lary is diflerent. The same considerations apply, though 
not so strongly, to Zend. In Section iii. (pp. 277 — 281 
the inference is drawn that affinity in language implii 
affinity in race ; and that, therefore, the ancestors 
the Hindus must at one time have lived in the sami 
country, as a part of one and the same commimity, wit] 
the forefathers of the Persians, Greeks, and Eoinaus. 
such a ciise as Ls here supposed, those branches of thi 
original nation which separated earliest from the othe: 
would in alter times exhibit the fewest points of ro 
semblance in language and institutions to the rest, while 
those which remained longest together would show in all 
respects the closest mutual affinities. In Section iv. (pp. 
281 — 285) it is argued that there is no objection arising 
from physiological considerations, i. e. from colour or 
bodily structure, to classing the Hindus among the Ind(^ 
European races. Section v. (pp. 285 — 298) exhibits the 
grounds which exist for supposing that the ancestors of 
the Indians and Iranians (or Persians) continued to form M 
one commimity after the other kindred tribes had separated ^ 
from them, and departed to distant regions. These groiuiils 
are, first, tlie closer affinity whicli subsists between Zeud, 
the language of the ancient Persians, and Sanskrit (of 
wWch some illustrations are furnished) ; secondly, tlie fact 
that both nations in former times applied to themselves the 
apj^ellation of Arya ; and, thirdly, the nearer and more 




numerous coincidences which are discoverable between 
the early mythologies of the two peoples, of which some 
details are adduced. Fi'om this more intimate affinity 
between the Indians and Persian.?, uidependent as both 
are of each other in their origin and development (see 
also pp. 314 — 317), a strong confirmation is derived to 
the general conclusion (deduced mainly from language) 
of the common origin of all the nations called Indo- 
Eiiropean. In Section vl (pp. 298 — 304) the theory of 
Mr. Curzon, that India was the original country of the 
Indo-European races, from which tliey issued to conquer, 
occupy, and civihse tlie countries Ijing to the north-west, 
is stated, together with some of tlie arguments by which 
he supports it. The remarks of Mr. Elph'uistone, wlio 
leaves it undecided whether the Hindus were autochthonous 
or immigrant, are also quoted. In Section vii. (pp. 304 — 
322) I cite the opinions of Schlegel, Lassen, Benfey, 
Miiller, Weber, Spiegel, Eenan, and Pictet, who concur 
in the conclusion that the cradle of the Indo-European 
race mitst be sought, not m India, but, as Schlegel, Lassen, 
and Pictet argue, in some central tract, from which the 
different branches of tliis great family could most ea^iily 
have difiiised themselves towards the widely-separatetl 
countries which they eventually occupied ; a condition 
which would not be fulfilled by supposing a remote and 
southerly region, such as Hindiisthan, to be the pohit of 
departure. Some of these writers draw the same in- 
ference from the relation in wliich the Indo-Arians stotnl 
to the aboriginal tribes whom they encoiuiteretl in India. 
Ill opposition to Mr. Curzon, who represents the language 




iiclia as the sources from wliich those of all 
the other Idudi-ed races issued, Professor Spiegel maiiitai 
that the Iranian huiguagc and mj-tiiology, though ovming 
common origin, are hi tlieh- devL-loptneut perfectly ind 
pendent of those of tliu Indians. In the same section it 
further urged that as neither the languages nor the m; 
tiiology of the Greeks and Romans are derived from th 
of the Indo-Ariaus, there is no gi'ound for supposing tliat 
the former nations emigrated from India at any period 
whatever.^ Section viii. (pp. 323 — 339) contains 
few passages I have been able to discover in tlic In^ 
authors "vvliieh may be supposed to embody any refereni 
(in no case, it must be confessed, other than a very ol> 
Bcure one) to the tiaus-Himtilayan origin of their ancesto: 
The chief of these are the iiitcrestiug paragraph of the Si 
tapatha-brfdimana, which conttiins tlie legend of the deluge 
in the oldest fonu in wliich it occurs in any Sanski-it work, 
and some textjs relating to the northerly region of Uttara 
Kuru, the Ottorocorras of Ptolemy. In Section ix. (pp. j 
339 — 344) I have quoted, according to the versions of« 
Spiegel and Haug, the first chapter of the Vendidad, wliich 
contains the oldest tradition of the Persians relative to 
Airyana-vaejo, the supposed primeval abode of their fore- 
fathers. Section x. (pp. 344 — 354) discusses the routt- 
liy which the Aryas immigrated into India. Benfcy 
tliiuks they may have cros.sed the passes of the Hima- 
laya from Little Tiiibet, and following partly the 
various branches of the Ganges, liave occupied first 
of all the tract Ijotween the Jumna and the SarasvatL 

' Compare " Additinnn aud Corrections," pp, 4S12,403. 






354 — 372. Sect. XI. The immigration of the Indo-Arians from the north- 
west rendered probable by the tenor of the Yedic hymns. 

373 — 4G5. Chaptbk III. Tax Ariatis is India : theik Adtarce to the 
£ast akd South. 

374 — 384. Sect. I. Distinction drawn between the Aryas and Dasyus in the 

384 — 113. Sect. II. Addidonal Vedic texts bearing on the relations of the 

Aryas and Dasyus. 
414 — 423. Sect. III. The Arians on the Sarasvati, and their ditrusion oast- 
ward and southward from that point. 
423—438. Sect. IV. Advance of the Arians from the Doab acroiis the 

Vindhya mountains : and their conflicts with the abnrij^inul 

tribes of the Dekhan. 
438 — 440. Sect. V. Indian traditions regarding the tribes in the suuth of 

the peninsula. 
440— 4J7. Sect. VI. Languages of the south of India, and their fun<ia- 

mental difference from Sanskrit. 
457—465. Sect. VII. Results deducible from the preceding section:*. 
4ti7— 487. Appkkdix: 
488 — 495. Additions and Cobrectioks. 





hlegel and Lassen, on tlie otlier liand, are of opinion 
tliat they must liave penetrated into India from tlic west 
by the route of Kabul and across the Indus. Roth and 
Weber also regard the Panjab as the earliest seat of the 
ludoAnans in Uiiidusthan, Li Section xi. (pp. 354 — 
372) I have endeavoured to show by quotations from the 
edas, that at tlie period when the hymns were competed, 
e Indians, though not imacquainted with the central 
I'ovincea of northern Itidta, were most familiar with the 
coimtiies liorderiug on, or beyond the Indus, and the 
north-western parts of Hindusthan generally. From tliis 
L fiict, and from the testimony of later writers to their iutcr- 
Bcourse with tribes, apparently Arian in descent and lan- 
■giiage, residing in the Panjab and on the other side of 
■the Indus, I derive a confirmation of the view that the 
■Hindus entered Litliii from the north-west. 
' In the Third Chapter (pp. 373 — 465) I have sought to 
draAV further arguments in support of the same conclusion, 
(1) from the distinction drawn by the autliors of the Vedic 
liymns between their own kinsmen, the Aryas,and the tribes, 
diflcring from them in complexion, customs, and religion, 
■^vhom they designate as Dasyus ; (2) from the accounts 
occurring in tlie Brfdimanas and post- Vedic writings, 
of the gradual advance of the Aryas from the north-west 
of India to the east and south ; and (3) from the well- 
established fact that tlie south-Indian languages are fun- 
damcnttdly different from the Sanskrit, and imply a 
non-Arian origin in the people by whom they were origi- 
nally spoken. Section i (pp. 374 — 384) contains a 
selection of passages from the Rig-veda, in which the 



Arj-as and the Dasyus are distingiiislied from one another, 
and reference is made to the enmity existing between tlie 
two. In most of these pass£>ge.<, it appears, liuman enemies 
and not demons must be intended mider tlie appellation a^ 
Da8yus,aa linfcr both from the tenorof tlie texts themselveaB 
and because in later writings, the Aitareya-brahmana, the 
Institutes of Manu, &c., this word is always applied to 
barbarous tribes. Section ii. (pp. 384 — 413) supplies a 
fiirther collection of Vcdic texts, l^eaiing upon the re- 
lations of the Aryas and Dasyus, and the characteristics 
of the latter as degraded, dark-complexioned, UTeligious, 
neglecters of sacrifice, Sec There are indeed other texts 
in wliich these Dasyus are regarded as demons, and this 
creates a difficulty. An attempt is made at the close of 
the section to explain, (1) from the original position 
of the Aryas, aa an invadhig tribe in a country 
covered by forests, and from tlie savage character of the 
aborigines, as well as (2) from the lengthened period 
dm-ing which tlie hymjis continued to be composed, — ho 
the same appellations and epithets might come to 
applied to different classes of bemgs, human, etherial, am 
demoniacal, indiscriminately. In Section iii. (pp. 414 
423) I quote the well-known passage from Manu's Insti 
tutes, which adverts to the superior sanctity of the coun 
on the banks of the Sarasvati„ (which is iu consequem 
presumed to have been for some time the scat of tlic m^ 
distinguishal Indian sages, and the locality where the 
Hindu institutions were chiefly developed) and dclini 
tlie limits of the several provinces of Braliuianical lui 
fta then recognized. I next adduce a highly inteivs^ 



legend fiora the Satiipatha-brfihiuaua which narrates liow 
the sacred fire (tj'pifjiiig, of course, the sacrificial rites of 
the Brahmans) travelled from the neighbourhood of tlie 
Sarasvati eastward, across the River Sadtinlrfi mto Videha, 
or north-Behar. Section iv. (pp. 423 — 44:0) presents a. 
selection of passages from the great epic jx)em, the Eaina- 
yaiia, descriptive of the Rakiihasas or gigantic denious 
by whom the Brahman settlers in southern Intlia were 
oppressed and their rites obstructed, aud wliose monarch 
Havana was vanquished aud slain by the Indian hero llama, 
with tlie aid of an array of nionkoys. In these poetic and 
hy|)crbolical descriptions, it is supposed we can discern tlie 
intlistinct outlines of a great movement of tlie Arj'as 
from the Doab southward across the Vindliya range, 
aud theu' conflicts with the aboriginal tribes of the Dekhan, 
the enemies of the Bralmians and their institutions. The 
epitliets applied to the Rakshasas in the Eaniayana cor- 
respond in many respects, it is observed, with those 
employed in tlie Rig-veda to characterise the Das)iis, 
Bakshasas, and Yatudlianas. Section v. (pp. 438 — 440) 
ctmtains some Hindu traditions regarding the tribes in the 
^uth of the peninsula, which however, are not considered 
to tlirow any hght on their real origin. Section vi. (pp. 440 
— 457) 8U{)plie3 a variety of details, derived from Mr. A. 
D. Campbell's Telugu Gmnimar (inclutliug the imjM>rtaiit 
note by Mi*. F. W. Elhs), and Dr. Caldwell's Comparative 
Grammar of the Dravidian languages, by which it is 
clearly shown that the Tamil, Tehigu, Malayalira, and 
Canarese tongues (which are spoken by thirty-one millioiifi 
of people), though, at different periods since the occupation 

•o i 

'■:i ^jQEhcm. LKlia by 'jca Bttcsane. they have received a 
large inf^*-xi 'j( SiEi^ris w-r-cisiw art- mevathdeas, oiigi* 
zlAjIj and fnn'iaratcjally -Tuiuc ojchcc frr-m. and indepen- 
(ieac od. rh^tz kruri^g^. ac^i ^ibu T*mfl ccmpoefhicm in 
partitrokr. i* r>e2aj»i"ai by rfi-e rasvy; acdtx? •< pare and 
cl^<!*ical in. pn:>p:-ni<:«i to is 1^v«d>?ca €r<xa Sao^xh w<»d& 
In the vii*. ami oTocfodrog S«cti»XL (pp. 457 — 165), the 
r&hxili5 ot* the preceding ^eecvxts are sonmied up. From 
the fatet (established b»xh In* philokisical oxtdderaticHis, 
and bv the tesdmoav of the ovrth- TrnKan sTammarians) 
that the Cravidian languages are ^Keenttalhr disdnct from 
^^anskiit. it is argaed that the people by whom the former 
claafe of languages were spoken originally (i.t». before the 
Brahmanical invadoa of the Dekhan) must have bdonged 
\fy a race which had no affinity to the Ssmskrit-speaking 
Aryan ; and could not, therefore, as Manu asserts, have 
hKic-n degraded Kshattriyas. I then endeavour to show, 
how the results obtained in this Chapter, viz.. (1) that the 
Aryaii, when living in the Panjab. came into conflict with 
an alien race called Dasyus : (2) that the Aiyas can be 
shown from their own books to hjive at first occupied only 
tho north-west of Lidia and then to have advanced gra- 
dually to the east and south, and last of all to have crossed 
the Vindhya range into the Dekhan ; and (3) that the ori- 
ginal languages of the south of the peninsula are distinct 
fW)m SaiLskrit, — how, I say, these results harmonize with, or 
corrolxjrate the theory that the Hindus, or Indo-Arians are 
not auU)chtlionous, but immigrated into Hindusthan from 
iIm; nrirth-wost. 
Tlif: Appf'iidix (pp. 407 — 487). and the "Additions and 



Correciious" (pp. 488 — 495) contain some further illustra- 
tions of the subjects discussed in the body of the work, and 
in a few cases,* supply some modifictitions of the text 
which closer rescarcli has rendered necessary. 

In the notes towaixls the close of the Volume, and in the 
Appemlices, the Sanskrit passiiges have been printed in the 
Italic character. The system I have followed is nearly 
that of Sir W. Jones. The distinctions between some 
stitmlai' letters have not always been very carefully in- 
dicated ; but the Sjinskrit scholar will have no difliculty 
in determining the words which are intended. 

Nearly all the Sanskrit texts in this Voliune, have been 
taken from printed editions. The quotations from those 
pai'ts of the Rig-veda which have not yet appeared in 
Professqj- Miiller's edition, have been copied from the MS. 
copy in my possession, alluded to in the Preface to the 
First Part. The quotations from Durgacharj^a, in pp. 
175, 17G, and pp. 183, 184, have been derived from a 
^MSt belonging to the East India House. That in p. 215 

?i, I believe, extracted from a MS. in the Lihraiy of 
the Asiatic Society in Calcutta. The two passages from 
Bhaskara Achaiyj-a, pp. 170 and 189, were obttuned ftx)m 
Piu»<lit Bapu Deva of the Benares College. 

I owe it to the kindness of IVifessorGoldstlicker, that I 
am able to adduce the; extracts from the Nyaya mfdit 
vi«tara, in pp. 6G and 190. 

Thu work of M. Vivien do Sauit-Miutin, entitleil : Etfuie 
Hur la Gi'ographie et les Populations Primitivett du Nord- 
ouMt de tinde dnpres les fli/mues Vidiqiies (which 
discusses many of the subjects han'lled iji the present 


volume), has only now come into my hands, as the last 
sheet, containing part of the Appendix and the ** Addi- 
tions and Corrections," is passing through the press. 

The results at which this author has arrived in his 
valuable and ingenious dissertation, in r^ard to the origin 
of the Aryas, their immigration into India, and the directirat 
of their movements within that country, correspond pre- 
cisely with those which I myself had reached. EQs views 
on some points of detail on which I had adopted a different 
opinion, tell even more strongly than my own in favour of 
the general conclusions in which we both coincide.* 

An Index to this, as well as the preceding Part is now 
under preparation, and will be published separately. 

* I allude to his conclusion that the Saraja referred to in the Veda vu 
a river in the Fanjab (in support of which he refers to Buroeura "Ebag. 
Fur. ii. 4J)5) ; and that the country of the Kikatas must, most probablj, have 
been in Ko&ila or Audh, and not in Magadha, or South Behar. 

I am happy to learn from M. de Saint Martin's work that he intends 
to prosecute further his researches into the ethnography of India. 



i. — ^xx. Fbefacb. 

1 — 3. Flak or the Fbeseitt Fabt. 

4—223. Chapteb L Tub Lakguaobs or Nobtobbr India: theib IIis- 


4 — 10. Sect. L The north-Indian dialects, ancient and modern. 

1 1—43. Sect. II. The Frakrit dialects employed in the dramas. 

43 — 62. Sect. III. On the origin and -vernacular use of the scenic di.i- 

53 — 65. Sect. IV. Views of the Indian grammarians on the relation of 

the Frakrits to Sanskrit, and on the other elements in their 

65—107. Sect. Y. The Fali ; and its relations to Sanskrit and Frakrit 
107^123. Sect VI. The dialects of the rock and pillar inscriptions of 

124—137. Sect YII. The dialect of the Buddhist Gathas, and ite relation 

to the Fali : Summary of the results of this and the preceding 

138 — 153. Sect YIII. On the original use of Sanskrit as a vernacular 

tongue ; on the manner in which the Frakrits arose out of it, 

and on the period of their formation : views of Professors 

Weber, Lassen, and Benfcy. 
153—168. Sect IX. Reasons for supposing that the Sanskrit was originallj 

a spoken language. 



169—223. Sect. X. Various agea of Sanskrit literature, and the ^fierent 
forms in which they exhibit the Sanskrit language :* the later 
Vedic commentators : earlier expounders : the Nirukta : the 
Brahmanas : the Yedic hymns : imperfect comprehension of 
them in later times from changes in the language : the hymns 
composed in the vernacular idiom of their age. 

224 — 372. Chaptsb II. AFinnriBS of thb Irdiahs with thb Fuuxahb, 
Gbbbks and Bomass, ahd deritation of aix thesi 


226—232. Sect. I. Introductory remarks on comparative phildogy : affi« 
nities of Sanskrit and Persian with each other. 

233—277. Sect 11. Detuled illustrations of the affinities of Saoakrit with 
the Zend, Greek and Latin languages. 

277 — 281. Sect. in. That affinity in language implies affinity in race: 
modes in which a greater or less diversity of language and 
institutions would arise in different branches of the same 
stock : Central Asia the birth-place of the Aryaa. 

281 — 285. Sect. IV. That there is no objection arising from physiological 
considerations, to classing the Indians among the Indo-Euro- 
pean races. 

285 — 298. Sect. V. Beasons for supposing the Indians and Fernans in par- 
ticular to have a common origin. 

298 — 304. Sect. VI. Was India the primitive country of the Aryyas or 
Indo-European race ? 

304 — 322. Sect. VII. Central Asia the cradle of the Arians : opinions of 
Schlegel, Lassen, Benfey, Miiller, Spiegel, Renan, and 

323 — 339. Sect. VIII. On the national tradiUons of the Indians regarding 
their own original country. 

339 — 344. Sect. IX. Ancient Fersian tradition of the earliest abodes of the 
Arian race. 

344—354. Sect. X. What waa the route by which the Aryas penetrated 
into India ? 




JJ.ITHERTO I have merely sought to bring together the ac- 
counts given in the Sanskrit authorities, especially the Itihasas 
and Puranas, relative to the origin of the caste system prevailing 
long the people of India ; and to show that these uccoimt&, 
nken in their obvious sense, are inconsistent with each other ; 
id that in consequence of this discrepancy, the theory, cora- 
jonly received by the Hindus, of the original distinctness of the 
jur castes, in virtue of their derivation from different portions 
the Creator's body, is not established as the doctrine of Hin- 
luism, even by a literal interpretation of these popular writings. 
It will now be my endeavour to show by a series of proofs of 
different description, derived from comparative philology, and 
)in on examination of the earliest Hindu writings, the Vedas, 
jt merely that the people of India who belong to the principal 
re and mixed awtes are of the same race witli the neighbour- 
»g nations (which, as we have seen. Part I. pp. 175. ff. is the 
trine of Manu) ; but that they were not originaJly divided 
Ito caefteii, or indigenous in India, but in reality form a branch 
f the great Indo-European lamily, of which the Persians, Greeks, 
^mtuim, and Germanic tribes were, or are, also members ; and 
it while other branches of this great family, (which seems to 
%re had it^ primeval abode in some distant country to the 
>rth-west of ludiii), separated themselves from the main stock 
, migrated to the westward, the progenitors of the Iliudua 



, .tiiMreUed towards HindusthaD, where tlieir origuud reli^ou& 
-./'tem was gradtudly modified, and the system of castes^ and ol 

- iiMtttatioiis and tenets of Brnhmanism were eiowly dereloped. 
The ptoeet of reawnung by which I hope to establish tbeie 
«oocltMiof» is the foUowing. First, I propose to show bj an 
examination of the languages and literature of India that tl>e 
SafMkrit is not, (as the Hindus appear to conceive), an immutable 
form of iqpeech of divine origin, but is very different now frotn 
what it WM when their ancestors first came into India. This will 
be mode apparent by a comparison of the diction of the Vedic 
hymns with the language of the Itihasas and Puranas ; and that 
this difference is the result of gradual development will be proved 
by a reference to the natural laws of speech, and to the analogow 
prooeiw which the tongues of other nations have undergone; 
by Aft indication of the earlier stage through which the Saniikiit 
paned, viz., that shown in the Vedic hymns, before it attained 
ita more modem form; by arguments drawn from the compo- 
iiitioD of such books as the Nighantu, and Nirukta, explaiu- 
tory of obsolete words and phrases in the hymns, and from tlie 
existence of such liturgical commentaries as the Brahmanas, »iui 
such Hpoculative treatises as the Upanishads, which presupposed 
aheody antiquated, or, at least antecedent, the hymns whi 
they qiiotv, aud the nensc of which they explain and de^eli 
The difference in age between the various ludiaa S^astras will 
further briefly adverted to', and established by pointing out the 
great (liscrepancy between the religious ideas, forms of woi 
and state of mannera which they severally represent ; the Vi 
hytiuis Iteing shown liy all these various lines of proof to be 
earlie«t of all the Indian books, and the others to follow 
thcin by a natural course of growth and expiuision. While the 
mutability aud the actual mutations of the Sanskrit languagt 
are demonstrated by this historical outline of Sanskrit Uteral 
the process of proof will be completed by some introdui 



' The detailed treatment of this portion of the subject will bo deferred to 
uUtcr part of this work. 

pla:^ op the present part. 



sectioDs, showing how the spoken Sanskrit became gradually 
broken down and corrupted into the Pali, and Prakrita, of bye- 
gone centuries, till it ultimately took the form of the modem 
vernacular dialects of Northern India. 

Having thus shown the mutations which the Sanskrit has 
undergone since its introduction into India, I propose, secondly, 
to prove by a comparison of that venerable language with the 
Zend, Persian, Greek, Latin, and other western tongues, that 
these forms of speech are all closely related to each other, both 
in respect of roots, and forms of inflection ; and thh in such a 
manner as to show them to be sister-ilialects, derived, by gradual 
alteration, from some more ancient, and now extinct, parent- 
language. From these facts, and others derived from Zend and 
Greek mythology and literature, I shall proceed to argue the 
common origin of the diflferent nations, — generally called the 
Ariou, Indu-Germanic, or Indo-European nations, — by which 
the alwve-mentioued languages have been spoken, and the equal 
footing in respect of civilifatiou, on which, in their earliest stages, 
they stood ; as well as to eyince the strong probability that the 
progenitors of the Hindus immigrated from the north or north- 
west into India. 

I shall then endeavour to fortify these conclusions by ex- 
hibiting the collision of the Indo-Arians, after their arrival 
in India, with certain barbarous tribes, speaking a different 
language, and belonging to a different race, who occupied that 
country before their immigration, and by sketching a history of 
their advance to the south and east* These subjects will be illus- 
trated from the data to be found in the Vedio hymns, the 
most ancient monuments of Indian antiquity, as well as in the 
other Sastraa of later date. 

When the preceding points shall have been all sufficiently 
discussed, the several tc»pic8 adverted to at the close of the intro- 
duction to the first Part of this Work, (pp. 3 — 4.) will still remain 
for consideration. These I shall hope to take up in one or 
more Bucceeding volumes. 



[cair. I- 



Skct. I. — The North- Indian Dialects, Ancient atid Modem. 

A sunvET of the languages of Northern India reveals to us the 
following facts. We find, first, a polished and complicated lan- 
guage, the Sanskrit, popularly regarded as sacred, and in realitj 
of very high antitjuity; wliicli is now, however, understood only 
by a few learned men, and spoken iu their schools as the vehicle 
of discussions on grammar and philosophy, while it is totally 
unintelligible to the mass of the people. We find, secondly, 
a variety of provincial dialects which are employed both by tbc 
learned and the imleamed, v\z. Bengali, Hindi, Mahratti, 
Oiizaratl, &c., all bearing a close resemblance to each other, 
and all composed, in a great measure, of the same roota 

Tlje words of which these vernacular dialects are formed m»y 
be divided into four classes. First, such as are pure Sanskrit, 
as for example parameiwara, devatd, exvarga, Btri, purusha, 
jarut; secondly, words which though modified from their 
original form, are easily recognisable aa Sanskrit, such as lo/j 
from luliU, istri from atn, munJi from muklux, hfuii from bhrdin, 
bhatija from bkrdtrija, bcihin from bhdgini, biydh from mvaha, 
bkuin from bhumi, and innumerable others in Hindi ; thinlly, 
words which have no resemblance to any kno^vn Sanskrit 
vocables, and which we must therefore suppose to Lave ao 
origin independent of that language, such as in Hindi, bap, 
father, beta, son, per, a tree, chauki, a chair, chuk, a blunder, 
kkirki, a window, jkagrd, a dispute, btxkhejrd, the same, dii. 


our,cfiatdi, amat,and amultitudeof other instances. Foiirthly, 
ords derived from Arabic, Persian, or some other foreign lan- 
1, as ddvii, a man, 'aural, a woman, hdkim^ a ruler, kdkliny 
lysician, dxiruat, right, roz, day, dariyd, a river, Toahanit 
igbt, &c, &c. &c. 

Let us now see what is the history of these vernacular dia- 
ects. It is clear, for many reasons, that they cannot have 
sxisted for ever in their present form. Witen, therefore, and 
uno have they been created ? Wbtit do history and the hooka 
if Indian grammarians tell us on the suljject ? 

If we begiu with the Arabic and Persian words which the 
•forth-Indian dialects, such ua Bengali and Hindi contain, we 
khall find it to be tmiversally admitted that words of this kind 
lave only been introduced into tliose languages since the time 
'hen the Miisulmans began to invade India. Now it is well 
nuwn that Mahmud of Ghazni mode his first inroad into Ilin- 
liistban about 850 years ago. Before that time, and in fact til) 
oug afterwards, when the Mahomedans had penetrated from 
;he north-west far into India, and taken possession of that 
country, there could have been scarcely any intermixture of 
Irabic or Persian words in the In d ian dialects.' 

We learn, indeed, from the works of the ancient a«lronoiner Variiha 
klihira, that u few oatrononiical and astrological terms of Greek or Arabic 
Icrivation had been borrowed from the Arabiao astronomers, and introduced 
Oto Saoslu-it books. I allude to such words oa hura, drikana, lipta, anaphd, 
lanapkd, apnklima, rihpha, which arc of Greek origin, and muhririna, 
miJtdviltL, taidi, latli, iic, which are derived from the Arabic. (Colebrooke'a 
Uisc. EMajs II. 525. ff., and Weber's Indische Literaturgeschichie, p. 227. 
ind Indische Studien, If,, pp. 234, and 263.) The following verse of 
iTornha Mihira proves cleorl/ how touch the Indian astronomers were 
debted to the Greeks : — 

l-H^l f% ^IWTT^ Wf^^ UTWfHT %7nTJ "^t%^ 

f sf^ ^M f^ j^nf^^ Htst: II 

For the Yavanas are Mlechhaa ; jet oniony (hem this science is thoroughly 
hiltirated ; and even they are revered like l^iiis : how mucli more a Brahman 

B 3 


_cnAP, t. 

Id the preface to the popular Urdu book, the Bagb o Bahar, 
we have the following account by the author, Mir Ammaa 
of Dehli, (who statefl that his forefathers had served all th» 
kings of Hindusthan from Hiuuayun downwards), of the 
origin of the Urdu language; which I copy in the Bomaa 
character: — 

" Uaqiqni Urdu ki zabwa ki huzurgon ke munh se yun 
auni hai kik Dilli efiahr Hlnd-uon ke iiazdik ctiaujugi hai. 
Unhen ke raja 'parjd qadiTti se t-ahte tfie, aur aprvi bhdkkd 
bolte the. Ilazdr haras se Musulvidnan kd 'amal hud- SulUl 
Mahmud Ghaznavi dyd. Phlr Gkori aur Lodi hddehdh hui 
la dmad o raft ke h'als kuchh zahdrvon ne Hindu Mtisulmnt 
ki dmezifk pdi. AkJivr Amvr Taimur ne . . . Ilindiittd* 
ko liyd. Unke dne aur rahne se laslJio/r kd. bdsdr sha/tr mm 
ddkh'tl hiid. Is wdele skahr kd bdzdr Urdu kaJddyd. . . . 
Jab Akhar hddehdh takhi par bait he, tab cluiron tuixi/ ke 
•mulkon «c eab qaum qadrddni aur faizrasdni u» khdnddn Id- 
mni ki awikar huzur men dkar jama a hue. Lekin luir ek lei 
(joydi aur boli judi judi thi. Tkatthe h(/iie se dpas wen hn den 
eaudd sulf suwdl jaxcdb karte ek zabdii Urdu ki rnuqamtr 

hui. , KUlun sabdn Urdu ki manjtt 

manjte aUi wari/t kth ki^u ehuhr ki holi xis se takkar nafdn 

" I have heard from the lips of my ancestors the following 
account of the Urdu language: — The City of Delhi iu the 
opinion of the Hindus has existed during the four Yugas. It 
Avaa inhabited of old by their kings with their subjects, who 
Bpoke their own hhdkhd (dialect). A thousand years ago the 
rule of the Musulmans began. Sultiin MahmCid of Ghazni came. 
Then the Ghori and I>odi djoja-stics held sway. In consequence of 
this intercourse, a certain nuxture of the languages of the Hindus 

tkillcil in astrology ! " (Colebrookc's Essays, II. 410.) This trifling cxc(?ptiKn. 
liowever, does not invalidate the assi'rtion m.<idc in the text, that it mi 
onljr uftcr tkc seltlcment of the Musulmans in India ibat ArnLic ml 
rtirwion words came to b« used in the dialects of India. 

SECT. 1. 



and Musuliimns took place. At length Amir Tsumur . , . . 
conquered Hindustan. In consetjuence of his arrival aiid re- 
sidence, the bazar of the army was introduced into the city, 
and the bazar of the city came in consequence to be called 

Urdu. When king Akbar ascended the throne, 

all races, learning the liberality of that unequalled family and 
its patronage of merit, gathered round his court from all the 
surrounding countries ; but the language of all these people was 
different. From their being collected, however, trafficking 
together, and talking with each other, a c/iiup (Urdu) language 
became eetiiblished. ... At length, the Urdu language, 
being gradually polished, attained such a degree of refinement., 
that the speech of no city can vie with it*" 

But it is only in the Urdu dialect, which is need by the 
Alahoinedans and by those Hindus in the north-western provinces 
of India who have leamt the Persian Iang\i!^e, that Persian and 
Arabic words are extensively employed. The words derived from 
those sources which exist in the Bengali, Hindi, Mahratti, 
Gozdra^, and other North-Indian dialects in the form in which 
they are generally spoken by the Hindu8f are considerably fewer 
in number. By far the larger portion of words in those tongues 
jure (a« has been alreatly said) either (1) pure Sanskrit, or (2) 
bormpt Sanskrit, or (3) words which can be traced neither to 
Sanskrit nor to Persian or Arabic, and which I shall therefore 
8t>'le indigenous. 

Several interesting questions arise here; as First, how far 
back can wo trace the existing vernacular dialects, Bengali, 
Hindi, Mahratti, Gttzarati, &c, in the form in which they are 
now spoken? Secondly, what has been the process of their 
formation ? and, thirdly, from what soxu-ce have they derived 
those words, as well as forma of inflection, which do not come 
from the Sanskrit ? 

Tlie question regarding the antiquity of the existing verna- 
culars is one which I am not prepared to answer with smy precision. 
ProfeEaor Laaseu (lu»titutioues Lingua; Pracritica?, p. 00) thinks 

o 4 



they have existed for at least 850 years. But it is sufficient for my 
purpose to show that they are not of any very great antiqviity, bat 
have been derived by a gradual process of change from other 
provincial dialects which preceded them; and which in their 
turn have sprung from the Sanskrit. 

There is no difficulty in conceiving that the Indiau verna- 
cular dialectfl should have imdergone great modifications in a 
long course of ages. The mere fact above adverted to, which 
every one recognizes, of their having at a particular assignable 
date admitted into their voc-abulary a large influx of Pereian 
and Arabic words is sufficient to render it probable that they 
may have formerly experienced other mutations of vario' 

The circumstance, too, that the people who inhabit the 
different provinces of northern India make use of different 
provincial dialects, Bengali, Hindi, Mahratti, &c, wliich are al) 
evidently derived from the same common source, is a proof of 
the tendency to cliange whicli is inherent in all spoken lan- 
guage. For as the inhabitants of all these provinces projiess the 
same creed, receive the same religious books, and are divid< 
into the same or similar castas, and for these and other 
appear nmuifestly to be descended frum one common stock ; 
is clear that their common ancestors must, at one time, liavi 
employed one and the same langiuige : and that that langii 
has in process of time undergone various provincial modificatioi 
out of which the several modern vernaculars have been 
dually formed. 

We shall also see, a little further on, that the differences between 
the North-Indian dialects (the old Mabaraslitri, f^auraseni, &c 
which preceded tlie modern verimciiliU' tongues, were few and on. 
important; whereaa the modem vernacular tongues, Bengal 
Hindi, Mahratti, and Guzarati, differ very widely from each oth 
in their forms of inflection and conjugation, lliis greater diver 
gence between the modern than we find to have existed* between 
the earlier dialects, evinces clearly the tendency to continual alter' 



ation, wliich I have remarked as a characteristic of language 
in generah 

1 shall first of all state hriefly the facta hy which it is proved 
that the modem vernaculars are not, comparatively speaking, 
of any great anticjuity, but have arisen out of earlier provincial 
dialects: — and then proceed to establish theee facts more in 

First. In extant Buddhist histories, such as the Lalita Vietara 

iposed in Sanskrit, numerous verses, styled Gathas, ai'e inter- 
'ipersed, the language of which differs from pure Sanskrit, by 
the forms of inflection being altered or mutilated. This 
mutilated Sanskrit, or something akin to it, appears to have 
been at one time the spoken language of India ; or, at least, 
this Gatlia dialect exhibits some specimens of that ancient 
spoken language, and exemplifies the process by which the 
Sanskrit, itself at one time a spoken language, became gradually 

Second. It has been discovered that many inscriptions are 
extant, engraven on rocks in different parts of India, bearing 
date between two and three hundred years anterior to the Chris- 
tian era, in which a language differing both from Sanskrit, and 
the modern vernaculars is used. 

Third, Tliere are extant in otJier countries such as Ceylon 
and Burmah, very ancient Buddhist books written in a language 
called Pali or Magadhi, which also is different from the modem 
vernaculars, aa well as from Sanskrit, while it closely resembles 
the language of the rock inscriptions just alluded to. 

Fourth. In ancient Indian dramas such as the Mrichhakati, 
S^akuntala, &c., while kings and Brahmans are made to spesik 
Sanskrit, various forms of speech called Prakrit and Apabhran^ii 
are employed for tlie inferior castes and for women, which in 
like manner, differ both from Sanskrit and from the existing 
vernacular tongues. 

The four foregoing classes of language have a more or less 
close affinity to each other ; and from the use made of the last 



[CBAP. I. 

three in particular, viz., that used in the rock inscriptions, tliat 
found in the Pali Buddhistical writings, and those employed in 
tlie dramas, it is impossible to doubt that either they, or forms 
of Rpeech closely connected with them, were formerly current, 
during a long course of centuries, as the acttial vernaculars of 
the periods when they were employed for literary, political, and 
religious purposes. 

But while we thus discover that Pali and different forms of 
Prakrit, such as have been described, were employed in former 
times, we can find no traces of the mtnlem vernacular dialects, 
J{iudi, Bengali, or Mahratti, «!cc,, in their present shape, in the 
ancient records of that same period; and we must therefore 
of necessity conclude, that these modern vernaculars did not at 
that time exist, but have been subsequently developed out of 
the above-mentioned Prakrit languages and other pre-existent 
fiinns of speech ; in other words that the former vernaculars (or 
Prakrits) have been grailually altered imtil they have assumed 
the form of the modem Hindi, Bengali, Mahratti, &c 

I shall now proceed to supply a more detailetl account of those 
forms of vernacular speech already alluded to, which appear to 
have preceded tlie existing varieties, and which are now obsolete^ 
In carrj'ing out this design, it will be advisable to begin with 
those dialects which seem to be the most recently formed 
and employed, of the four Indian classes of speech which 
have been Ijefore alludod to, viz., first, that found in the 
Buddhist Gdthas; secondly, that used in the rock inscriptions; 
thirdly, the Pali ; and fourthly, the dramatic Prakrits. Tlie last, 
named class appearing to be the most recent^ I shall first subject 
it to examination, and then proceed to the others. 



8ccT. H. — The Prakrit Dialectt employed in the Dramat. 

With the view of aecertainiug the relation in which the Prakrit 
languages stand to the modern vernaculara of northern India, I 
have gone cursorily over FeveraJ of the dramas in which they 
are employed, such as Mrichhakati, attributed to King SSudraka, 
and the Vikramorvaai attributed to KaHdasa, (both of which, 
though their precise age be disputed, appear to have been re- 
spectively composed, at the latest, about sixteen and fourteen 
hundred years ago",) together with several others. I have also 
referred to the examples given in the Prakrit Grammar of 
Vararuchi, which is considered by Lassen to have been com- 
posed about 1800 years ago*, or rather in its commentary. An 
examination of the Prakrits which are found in these several 
works shows that the languages of India were then in a state of 
tnuLsition, and formed an intermediate link between the Sanskrit, 
and the modem vernacular dialects. The Prakrit forms of 
inBection and declension approach more to the Sanskrit than 
to the modem vernaculars; but yet exhibit a great breaking 
down and modification of the former. I will give some in- 
stances of this which will make rny meaning clearer than any 

' Profcssnr Wilson, reasoning from a variety of consiJerationa, conaiderfl 
the filfic'lilinknti to Imvc Wen probably composed in the interval between 
100 B.C. ami llie end of the second ccnturj, a.d, (iDtroduction to the play, 
pp. 5 — 9.) The same writer thlnka that the Vikrainorviui, which is regarded 
as the work of Kfilidilsa, is more recent than the Mrichhakati, but does not 
nssigD anj probable date. (Introd. tu drama, p. 185, 186.) Loascn holds 
that the Mrichiiakntl was composed towards the end of the first centurr, 
A.D , while the Vikramorvu) and the ^akuntala (which Lst iaolso assigned to 
Kttlidltsa) were compo«ed in the second half of the second century, k.v. 
(Ind. Alt. ii. p. 1160.) Weber, on the other band, io bis latest notice of 
the subject in the Introduction to his Millavikii and Agnitnitra, pp. xxxiii. xl. 
places the age of Kiilidasa, the author of the Vikramorvosi and ^kuntalii, at 
the close of the third century, a.d. The Mrichhakati is held by the same 
author to bo not earlier than the second century, A.d. (Ind. Stud. ii. 14S.) 

* lud. Altctth., vol. ii. p. I ICO. 



J ^--^ ' - T- ■ -r- 

^^>- - i 





[chap. I- 1 

f general statements I do not think it necessary to distinguish 1 

^^ here the different 

kinds uf Fnikrit, which ^yill be specified 1 

^H further on. 


^H Rwikilt. 



EoEllib. 1 

^^M Bbavimi 



lam. 1 

^^m Bhavoai 



Thou art. ^J 

^^H BImvali 



Ue is. ^H 

^^m DbBTonti 



They are. ^^ 

^M TTttiabtha 



Rise. J 

^^M Prupnomi 



I obtain. ^^H 

^^M ^rinoini 



I hear. " 

^^M ^riiju 

Sunn, or SuQohi 


Hear (imper.). 

^^M Kathnja 




^^H Dadumi 



I give. 

^H Dadati 



He giyes. 

^^M Dntuim 


Diya, Din 


^^1 Nrityati 



He dancea. 

^^H Knksbami 



I keep. ^M 

^^m Dliuva 



Wash. ^ 

^^H Bruutaii 



We speak. 

^^H Patiim! 



I fall. 

^^H Nialika^ja 




^^M Gliritam 




^M Mukba 




^^1 Kuryyam 




^^M Kurma 




^^M Karna 




^^H Twnui 



Tbou or you. 

^^M Tubliyam 



To thee. ^H 

^^H YushtniLkam 



or you. ^H 


Atlhi, or A('bfljhi 

Aclichhc (Beng 

) He ^H 



Ackcblien(di«o.) They are. 1 

^H It is manifest 

that in these 

instances we 

see the inter- 1 

mediate forms 

which the words 

took in Prakrit before they | 

aB8umcd tlie s 

lapes in which wc 

' now find th 

em in Hindi or 1 

Bengali, e.g., karma and kdryya became in Prakrit respectively | 

kavima, and kajja 

, and finally in Hindi kdni and kdj. The 1 

Sanfikrit form rakshdmi (I keep) re-appears 

in the Prakrit 1 

rakkhdmi with 

the compound consonant ksh changed into Ickh, m 


but with cumi the final afiSx of the first person singular 
unchanged. In the modern vernacular the former change 
remains, but the word has undergone a farther modification, the 
peculiar affix of the first person singular aim. having disappeared 
in the Hindi rakktd, which does not differ from the second 
and third persons. A fuller exemplification of the points in 
which the Prakrits coincide with, and diverge from, the Sanskrit, 
on the one hand, and approximate to the modem vernaculars, on 
the other, will be found in the tabular statement subjoined. 

The books to which reference has been made in this statement 
are the following : — Mr. Cowell's Prakrita Praka^ of Vararuchi ; 
Lassen's Institutiones LingusB Pracritdcse; Delius's Radices 
PracriticsB ; the Mrichhakati, Stenzler's edition ; the Sakimtala, 
Boehtlingk's edition; the Prabodha Chandrodaya, Brockhaus's 
edition ; Malavika Agnimitra, Tullberg's edition ; and the Vikra- 
morvaa, Calcutta edition. 

I t.<g ^ 



ho (^ <^ iZ> $ 



F b? &9 V 


. ■ 00 flO 
•1 "^ CM fi 

CS w 



.J — rf: ji— g ^ ■- j3 — ' ja 





f (^ ^ FF-l&ilK^<pfFl 


jf ^b ¥« b P ^ 








• • 









f— '*— 1 










;- r- 






fr« Oi 




W . 



c-' _ 




• • 



— C< 


ir t» 




w *^ 







P « 


• » 

CO . 







^ ;i2 


:r J3 




. ja 

^ • 





1 *" 

f^ ■"" 


•^ -^ 

— -ts 


^ ► 






. t> 

. o 


' g 





































)» fYy ihy 


>o CO 

CO O) 


OS ■» 

. -a ^ 

— ^ 13 ^ 
us ii 

to ' 





S ^ ^ 
















' 00 






-< o 

-,■ 2 ^ 

w "g ji a S3 ■> 55 2 S 



M •» •« «f4 ««« 










■ r V be 

-o — -^ 

e3 B< s n 



.5 n 

el O 1^ 

^ 3 I 

j3 iQ Qri 




'i t ^ ^ 


^ ^ 





li'if '^ %lri 

B 3 o.. 

P 5 IB 

=> «*- i» ^ 





2 «r 

.s -I 

t a £: 

o an 

5 2 






E E » 






I el &f 



* a; 
. > 

u 3 


CO to 




to n U3 
« 03 w 


a o 


> M 
— M 



s ;9 

:i § si- 





/C /^ 






rfT ir 



:if <^ (fcr 


eo a" 





.::■ 04 





is "5 

^ If If 


I } 









2 .«> M - 

• N -* — « 

:= — ja J u: -i 

> . -a ,d ^ . 

« J£ 'C. "u. X.J" 






[our. I 





1 i 1 






.d So bO 










^ «5 

I -^ 

•3 ^ 

$ a 



_5 H^ 

to rea 

3 8 ^ -B 

5 a s *_. 











/1C CQ 

^IP ^ 






<o O..S ^ S 

u S » S J .2 .3 .2 .2 .2 »; 



^ -^ t* f 

I— I .-< -N »^ -< 

• J3 ,J3 J3 •«-• 'm 

i li I 
2 |-2 J 












to see. 
t: be 






















1 1' 

Lcnxr. r. 









2 S 



g? I 




«o O <n . • t^ ^ 

.2 « 

• • •on _ _ 

■ cs >-• a c o '° 

> k J> SP t^ QS 


















•r c 

^ > cT 

c r= £ 

« H S 

;! -S J 

" S I 

a d S 

g - s 

to ^ - 

^ » 

be o 


r. : 


5 T3 

.t; 8 

^- s S 

to- g 

o « £ 

• -5 

•^ S 

i £ I :l 


r. O 

* a 

a .~ 

i ^ 

bo U 

.2 I 

•5 i 

«3 ri 





-1 -.J ^ 





w so 

2 « 

I— I -S 

•2 I 



1^ *& 

^ t 


.S ~ 







•i: I -3 -^ 

5 -^ 2 , 

t| « to = 5 

" - £ M 

- o 2 

•= I -c .5 

a <:> "a .' 

» 1^ S 



But, while the great majority of Prakrit words can, by the 
application of proper methods, be traced back to a Sanskrit 
Boiurce, there are some others which refuse to yield to the action 
of even the most powerful tests wliich criticism can employ, and 
successfidly assert their claim to a diflferent origin. 

Another fact then which is made clear by the examination of the 
dramatic poems and the Prakrit grammarians is, that tlie Prakrit 
dialects contain a certain number of words which are not Sanskrit, 
but which we also find in the modern vernaculars, such as the 
roots duh, to sink, tharlmr (in Hindi tharthar), to tremble, 
dhakkf to cover or shut, and the nouns (/or, leg, bappa, father, 
&c.^ Tlie greater portion of the wordfi of this class which I 
have discovered, will be found in the subjoined table. 


*" See die Rer, II. Ballnntines paper "On the relation of the Mahmttl to the 
Sanskrit," in the Journal of the American Oriental Society, vol, iii. p. 309 — 
nS. Some of the words considered by Mr. BolluTitinc to be Mohratti 
however, Persian or Arabic, such as rnckh, baghal, nianzila ; others, as 
le, to eat, are Sanskrit. I add the following remarks from Dr. J: 
>ViUon's "Notes on the Constituent Elements," &c. of the Maratbi languoge, 
(prefijied to Mulcsworth's Maratlii Dictionary, 2nd edition), p. xxii. 

[The Morathi language] " has two distinct lineal elements, the Scythian 
(or Turanian), and the Sanskrit" .... 

" The Scythian element .... is obviously the more ancient of the two, 
M far as its present locality is concerned. It is still a good deal in use, 
especially among the lower orders of the people, and in the business of 
common life. It claims almost all the words beginning with the cerebral 
letters, which, as initials, were probably not originally in use in the San- 
skrit ; almost all the words beginning with the letter /ft ,- and a great 
majority of the words formed from imitative particles, both simple and re- 
duplioatc<l, which are often very expressive, and ore not now of an arbitrary 
character, whatever they might have been before they got established in the 

■fli* loquendi of the people, by whom they were originally formed." 

"The Sanskrit element is that which predominates in the MarSthi, as the in- 
•p«ction of the Dictionary at once shows." .... " Colebrooke expresses it 
t» his opinion that ' nine tenths of the Hindi dialect may be traced back to 
ihe Sanskrit ;' and perhsps a similar observation may be justly made as to 
Hie proportion of Sanskrit words in the Alarathi, when both primitive and 
modified forms are taken into the account." 

D a 






<^ S <g 

^ ^& (6 i IS? ^^ 





o 3] 





































































CcHiT. b 









1 I 

2 2 

JS ^ 1 g -S I J 
S S 2 S S 3 3 

a ^ 

rill lf» 1 1 &■ i^ 1, 1^ 


^^^%,% g^ttt 





13 la 

> P 






04 ;o 











g ^ 









o . 

•a •* 


8 »- 








It 18 true that, in the dramas, these non-Sanskrit words are not 
so numerous as might at Brstsight appear; that many vocables, 
very unlike the Sanskrit, which seem, on a hasty inspection, to 
lie of a diflfereut origin, are discovered, on a more careful 
examination, to be derived from Sanskrit by successive steps 
proceeding according to certain recognized rules of mutation ; 
and that the indubitably non-Sanskrit words which remain do not 
bear bo large a proportion to those which are of Sanskrit origin, 
as is the case in the modern vernaculars.** This paucity of 
indigenous words in the dramas is, perhaps, to be accounted for 
by the fact that tbey are polished compositions containing many 
poetical passages, and were written by Pandits, men familLor 
with Sanskrit, who would be likely when they could, to avoid 
vulgar words and phrases, and to employ vocables of Sanskrit 
derivation, wherever it was found possible : just as we see the 
pedantic Pandits of our own time are in the habit of doing-'* 
And there can be no doubt that in the provincial dialects, 
as spoken by the lower classes and by unlearned persons in 
general at the time when the dramas were composed, many 
more non-Sanskrit words would be current than we meet with 
in the dramas. In the same way we find in modem times several 
modifications of language in use among different sections of the 

** Lassen, remarkg, p. 286 : " The roots of the Pmkrita roust be looked 
for in Sanskrit ; and the ftw words which appear to be of extraneous origin 
can, for the most part, be traced to Sanskrit, if the investigation is pur- 
sued on right principles. At the same time I would not entireljr deny 
that some vocables may have passed from the indigenous languages of India 
into the Sanskrit, as well as the Prakrit ; but such words are certainly not 
numerous." Lassen may not underrate here the number of purely indi- 
genous words in the Prakrits, properly so called, as they are exhibited 
in the dramas, but his remarks are not certainly correct if applied to the 
modern vernaculars, in which words not derived frcim Sanskrit-, and which 
roust have come down to them from the vernacular Prakrits, arc very nu- 

*• Compare the case of English like that of Dr. Samuel Juhnson, full of 
Latin and Greek derivatives, with other composii ions in which An;;lo-Saxon 



community in the same provinces of Hindusthan. The Hindo 
Poudite, for instance, use a dialect which ia full of Sanskrit 
words; the villagers use fewer Sanskrit and more indigenous 
words; the lower Mahomedaus use a language approaching to 
that of the Hindu villagers, but with more Persian and Arabic 
words ; while educattKl Maliomedans introduce into their dis- 
course a large number of Arabic and Persian phrases. But the 
ejustence of even a small proportion of such non-Sanskrit words 
in the dramas, when taken in conjunction with the corrupted 
form, — akin to that of the modem vernaculars, — in which we 
find Sanskrit words employed there, is quite sufficient to show tb«t 
the Prakrits, such as we see them in the dramas, were, with some 
modifications, the spoken dialects of their day; and were ooo- 
sequenfly tlie precurnors of the modern vernacular tongues. A« 
we find in these kittor a conaiilerable proportion of words whioli 
Cannot be traced back to Sanskrit, we are driven to conclude 
that these words must have existed in the older vernacular 
dialects, and have been transmitted from them to the later. ITie 
only alternative is that we suppose these non-Sanskrit words 
to have been invented in modern times, a supposition wliich 
is destitute of all probability.*" 

The question now recurs, Whence came these words which arc 
not of Sanskrit origin in the Prakrit dialects ? To answer thisi 
question I must anticipate an assertion which I hope further 
on to prove more in detail, viz., that there are in India very 
manifest traces of a variety of races of men differing widely in 
their origin. 

*" Even if I were to make the admission, (which, however, it is unposalle 
to do), thiit the Pali and the scenic dialects were never octuallj tpoliea 
vernaculars, this would not neutrnlise my nrguiuent. For the latter must 
havchacn used on the stage, and must thc'reforc, havoheen understood. Thi'jr 
could not, however, have been iiitelli^^ible, if they had not appronehed closely 
to some r^irni uf spoken language. And the existence of the Pali as well ai 
of the Prakrits shows both the general tendency of men to break down and 
modify their languages, and the actual process by which they proceeded in 
northern India. 

■wrr. u.] THE 



It appears that the ancestors of the higher classes of northern 
Hindus, who originally spoke Sanskrit and called themselves 
Aryas, must have had their origin in countries to the north or west 
of India, and immigrated into Hindusthan at an early period. 
When they arrived there, they found the country already occu- 
pied by a race of men called in the Veda, Dasyus, who spoke a 
different language from themselves, and with whom they became 
engaged in continual warfare. These Dasyus appear to have 
been partly driven away by the Aryas to the east and south and 
north, where they took refuge in the forests and uiouutains, and 
partly to have been subdued and to have become incorporated 
in the Arya communities as their slaves or dependants. Though 
these earlier inhabitants of India also, had, in all probability, 
immigrated into that country at some period anterior to the 
invasion of the Aryas, I shall for the sake of ready distinction 
style them the aborigines. These aboriginal tribes may not have 
been all of one race, and may have arrived in India at different 
times, but their history is very obscure and can only be con- 
jectured. So much is clear, that their languages are not all 
alike. In the south of India we find still existing a set of 
spoken languages called Tamil, Telugu, Canarese, Malayalim, 
&C., which differ very widely from the vernacular tongues of 
Northern India, viz., the Mahratti, Hindi, Bengali, &c Though 
the southern languages have now a certain intermirtiure of 
Sanskrit words, yet, it is clear that this intermixture is only of 
comparatively recent date, as they differ entirely both in structure 
and in the great bulk of the words of which they are composeil 
from the Sanskrit, and all its derivative languages. The dialects 
of northern and central India, on the other hand, viz., tlie Mah- 
ratti, Hindi, Bengali, &c., are, as we have already seen, mainly 
derived frum Sanskrit, thoUgb they contain a considerable propor- 
tion of words which ore evidently of a different origin. These 
words of non-Sanskrit origin, which we first discover, to a certain 
extent, in the ancient Prakrits, and wliich descended from them to 
the northern vernaculars, must Lave been derived from the 




language spoken by the aborigines, wLo Lad occupied the noi 
of India before the Sauiikrit-speaking race of the Aryas arrive 
After these northern aborigines had been reduced to depen 
by the Aryas, and both clasaea, Arya and non-Atya, had coaleaoed 
in one community (of which the former composed the upper, and 
the latter the lower ranks), the languages of both classes (which 
bad previously been different ) would begin to become assimilated 
and amalgamated; the Sanskrit-speakmg Aryas would 
adopt many words belonging to the speech of the aborigin< 
while the aboriginal race would begin to borrow many woi 
from the Sanskrit, the language of their masters. This prooea 
however, would naturally lead to a great corruption and altCT»> 
tion of the Sanskrit. Many of the compound sounds in San»- 
krit words, such as strl, rakia, kshatriya, seem to have been 
found such as the lower orders of people could not prtmoimce, 
and these compounds became accordingly broken up or am- 
plified, or in some way mo4lified. Thus ati'i became i«triyrakUi 
became rakat, and k«hatriya became khatriya or chhatriya. 
In this manner both languages would become gradually chang<ed, 
according to processes which are seen in operation in all couutriei 
Caprice, alteration of physical circumstances, differences of educa- 
tion, and those varieties in the organs of speech which are 
peculiar to different races, — are all found to produce progre»- i 
sive modifications in the languages of mankind. Various foxma ■ 
of Prakrit would spring up by degrees in different provincen, in ~ 
which Sanskrit and aboriginal words and forms would be com- 
bined, though the more cultivated element, the Sanskrit, hu \ 
remained predominant. At the same time the Sanskrit language 
gradually ceased to be spoken in its pure form, and becoming 
the language of books, and of the learned class exclusively, was 
more and more polished and settled by grammarians ; and lieing 
exempted from the ordinary causes of alteration, continued thence- 
forward unchanged : just as was the case with the Latin languagH. 
It seems at the same time to be very probable that many words 
of indigenous origin as well as words which, though of Sanskrit 




ori^n, had been modified in the Prakrits, were incorporated in 
the Sanskrit ; and that in this way the modern vocabiilai-y of that 
Liugiiage includes many words and roota which were unknown 
to it at an earlier period. ** 

Sect. III. — On the origin and cemacuiar lue of the Scenic DialecU. 

It has been doubted, however, whether the dramatic dialects 
were ever spoken langiiages. Thin view is thrown out by 
Prof. H, H. Wilson in the introduction to his " Select Specimens 
of the Theatre of the Hindus," pp. Ixv., Ixvi. 

" There is one question of some interest attaching to our 
oonstnxction of the Prakrit, which merits a fuller inquiry than 
has been yet given to it, and on which tins is not the place to 
dilate. Does it represent a dialect that was ever spoken, or is 
it an artificial moditicattoQ of the Sanskrit language, devised to 
adapt the latter to peculiar branches of literature ? The latter 
seems to be the most likely ; for there would be no diflSculty in 
the present day in writing it, although it is no longer spoken, 
and highly fmisheil specimens are to be found in plays which 
are modern productions. The Vidagdha Mtvdhava, for instance, 
consists more than half of high Prakrit, and it was written less 
than three centiu-ies ago. On the other hand, many of the mo- 
difications are to be found in the spoken dialects of Hindusthan, 
and the rules of Prakrit grammar account for changes which, 
without such aid, it is difficult to comprehend. The simplifica- 
tion of the granimatictd construction by the disuse of the dual 
number, and the reduced number of verbal conjugations, looks 
also like the spontaneous substitution of practical to theoretic 

^ Dr. Stevenson says, in tbc Journal of the Bombay Brunch Rojai As. 
Society, for January, 1859: "The Brnhmnns scattered through all the 
different provinces of Himlusthan no doabt adopted many of the words 
of the languages of the tribes among whom they resided, and introduced 
them into the sacred tongue." Professor Beofey has dmwn attention to the 
introduction into Sanskrit of words which had become modified in tho 
Prakrits. See Lassen, Ind. Alt. ii. 1 149, note 2 ; and Bcnfey, article " Indicn," 
On Ersch & Grubcr's Encycl), p. 248. 





CBAT. t. ' 

perfection in actual speech, and may tempt us to think the 
Prakrit was once a spokea tongue. The subject is interesting, 
not only in a philological, but in a historical view; for tha 
sacred dialects of the Bauddhas and the Jainas are nothing else 
than Prakrit, and the period and circumstances of its trangfer to 
Ceylon and to Nepal are connected with the rise and progress of 
that religion which is professed by the principal nations to the 
north and east of Hindusthan." 

If the Prakrits be merely artificial modifications of Sanskrit 
for dramatic purposes, my reasoning in regard to the relation of 
the modern verneit^tilars to the Prakrits would fall to the ground. 
Though this view appears to me to be sufficiently refuted by the 
proofs of the derivation of the modern vernaculars from the 
older Prakrits supplied liy the lists of words which I have given 
above, I think it expedient to fortify my conclusions by the an- 
nexed extracts from Professor Lassen's Institutiones Pracritaccv 
pp, 39, ff.j which will, at the same time, illustrate the process by 
which the Prakrits were derived from Sanskrit. 

" If the question regarding the origin of these dialects meedj 
refer to the sowce whence they are derived, it admits of a very 
easy answer : for, as has been already stated, all the scenic 
dialects are drawn entirely from the Sanskrit.*^ If, however, the 
question means by iviiat process these dialects have been drawn 
from the Sanskrit, it will be more diflScult to answer. The 
difficulty does not consist in these languages containing any 
forms or words of which the Sanskrit archetypes are undiscover- 
able ; for, on the contrary, both forms and words are deduced 
from that ancient source by imdergoing certain mutations which 
all languages follow as they become altered and corrupted in the 
course of time ; as, for example, has been the case with all the 
Germanic and Romanic dialects which have sprung from the 
Gothic and the Latin. 

" The difficulty, however, consists in this, that these dramatic 

^ See, however, what has bees said on tlu5 subject above, in pp. 39, tt. 



dialects, sprung from the Sanskrit, and beaaing the names of 
different provinces, are different from the provincial languages 
which have the same name and origin ; e. g. the principal 
Prakrit (which appears to have been called Maharashtri) difters 
firom the modem Mahratti, and the Sauraseni from the 
Brajbhakha. Hence a doubt has been suggested whether the 
dramatic dialects were formerly the spoken tongues of the 
people of the several provinces, who at present use a form of 
speech which though cognate, is yet different; or whether these 
dramatic dialects are nothing more than artificial adaptations, 
either of Sauskrit, or of the provincial tongues to dramatic pur- 
poses. The latter opinion has appeared to Wilson the most 
probable, for this reason that the modem dialects of the Mahratta 
country, of Mathiu-a, and Behar, are diffejent from those which 
were employed on the stage under the same names. He assigns 
another reason, viz. that these dramatic dialects can be com- 
posed even now. But is not the case precisely the same with 
the Sanskrit or the Latin ? both of which can in our day be 
written by men who are skilled in them, though they have long 
ceased to be used in daily life, or to be spoken, except by a few 
scholars. Wilson's first reason is equally inconclusive : for to use 
what I may call an arguniCTnin.'ni ad fiotninern, the learned Profes- 
sor would scarcely succeed in making himself understood, if he 
were to address his coimtrymen in the Anglo-Saxon tongue. His 
argument would, indeed, be sound, if it could be proved that in the 
age when the dramatic dialects were ,/?>'»< brought upon the stage, 
the Maharashtri or any other form of contemporaneous speech 
wns different from the dialect introduced into tlie dramas under 
the same name. For it must be recollected that succeeding dra- 
matic poets, following the example of their predecessors, did not 
change the dramatic varieties of speech, but retained them in their 
original forms; whilst, on the other hand, the popular dialects con- 
tinned to umlergo great alterations, as is the fate of all languages 
which are subjected to the wear and tear of constant use. These 
scenic dialects can be taught even now by grammatical tuition, 



nigTOHY Am) relatioxs op 


juHt 08 the Sanskrit can, though neither the one nor the other catfj 
be learned hy the Indians from a nurse. All chsoige in the scenu! 
dialect8 was guarded against (just as in the case of the Sanskrit)] 
£rum the period when their forms and laws had been fixed brf 
grammarians ; and, consequently, the argument drawn from the 
divcrHJty of the dramatic and modem provincial dialects is of nn 
force, unless it can be shown that the provincial dialects also have 
remained unchanged from the commencement to the present 
day. This, however, can neither be shown, nor was it poesiUei I 
On the other hand, tlie existing condition of the provincial dia- ' 
lects cannot be explained unless we suppose them to have Itad 
another form, more ancient than the present, and more con- 
formable to the Sanskrit. 

" Since, then, it cannot be proved that the provincial dialects 
were originally different from tlie scenic, I shall add some argn- 
meuts by which it will be made probable that the latter (the scenic) 
were actually current in the provinces from which they derived 
their names. And first, I shall make use of the names them- 
selves OS an argument : for the names Maharashtri, Sauraseni, 
would be absurd if they were not referred to provinciul dialects; 
seeing that the names cannot be deduced from any orders of 
men so called, nor from any peculiarities in those dialects. 
The same may be said of the Magatlhi, for though I am aware 
that the word Magadha denotes an order of bards, still the 
Mogadhi dialect is employed on the stage by other classes of men, 
and the l)ards themselves derive their appellation from the pro 
vince which gave it« name to the dialect. 

** In the next- place, I argue that the nature of dramatic poetiy 
renders it scarcely credible that dramas composed in a lauguage 
different from that of common life should have been exhibited 
ou the stage. Tliis, however, is a different matter from the 
supposition that the dramatic dialects have subsequently ceased 
to be spoken, and have become olsolete, while yet they main- 
tained their place on the stage. The same thing holds good 



of the employment of Sanskrit itself in dramas written in a 
comparatively modem period. 

** If these conaiderations be duly weighed, it appears to follow 
that the use of diflereut dialecta on the stage was the result of a 
peculiar condition of Indian life, at the time when the laws of 
dramatic art were first fixed by the Indian poets. 

" To these arguments it must be eulded that there is so close an 
affinity between the primary dramatic dialect and the Pali, as to 
leave scarcely any doubt of their being originally identical. So 
much is undoubted that the sacred language of tlie Jains is not 
different from the primary Prakrit. This language would cer- 
tainly not have been adopted by the adherents of a sect which is 
^^rongly opposed to the Brahmans and their opinions, if the 
dramatic dialect had had no other foundation than tlie fertile 
and subtle genius of the Brahmans. Tlie Jains could, however, 
liave no difficulty in appropriating it to their own uses, if it was 
the language of daily life. How it happened that the Maha- 
raahtri dialect in partictilar came to be selected both by the dra- 
matic poets, and by the Jains, is a point to be explained from 
the history of the Indian stage, and of the Buddhist religion, out 
of which the sect of the Jains sprang. To attempt this here 
would be out of place. 

" The primary argument, however, is to be drawn from the 
structure of the languages themselves. This structure is the 
Bame, as r^ards principles and general rules, in all the provincial 
languages of Sanskrit origin, while it is different (though very 
fiimilar), if the individual forms and the elements of these be 
refjarded I shall therefore do sufficient justice to tlie plan I 
have in view, if I examine .more minutely some of these lan- 
guages, and show what their graiuniatical character is. In doing 
eOf however, I am prevented by the limits of my book from ex- 
hibiting their entire grammar, nor would it better serve the 
end I have in view if I were to do so. I propose, therefore, to 
inquire into the scheme of declensions peculiar to these lau- 




gmgw, which follows the same analogy as the lawrs of coDJi^atM& 
I pta over tlie permutixtions of sounds, which are too varioa 
to be treated here ; nor, if I did treat them, would it condoee 
to my object, which is ao to describe the structure of Uw 
IVfirinciiU dialeotii aa to exhibit the differences between them 
and till' dnunutic languages. For the changes in their elementi 
iiiidtTj^iiiu' Ity the Sanskrit words which have been received into 
llir iiKHlrtrn i1ialt«cti<, follow two very different laws, which, if not 
rim-fully (linliiiffUJHheil, might be used to demonsftrate contrary con- 
rliiMionH, Oiji' Hort t>f mutation prevails in those words which htA 
Ik'cii rciHtivdd into tlic provincial dialects which were ancientlj 
ff»min(l, or nithiir oorriiptctl, from the Sanskrit; such 
UrnjIiJiakliA pothi, n\x>ok, which in Prakrit is pothau, and 
Haiinltiit jmhttil.o. mill numerous others; which would lead us t< 
cnncluilr iliiii 111'' -^^uiio changes in the elements of words liav«! 
taki'ii |iliin- ill the iHodern vernaculars as in the dramatic dialect<;;j 
oad tlittt tilt' fiiruiB of words in the former are derived from, aui 
fiiid tiii'ir cxjilatmlion in, the latter. This I by no means deny. 
Hut tliprc in iiiuitluM' kind of words to be found in the modem dia- 
lootd, which i'i)ini< nearer to the original Sanskrit words than do 
forms used in the dramatic Prakrita. The following are 
oxamplcwi from the Bnybhakha, Panjabi, Mahratti, and Bengali 


igili: ■ 

Hriybliuklirt. Paiyabi. Mahratti. Bengali. 

Prakrit X^, ^M1 U^\H ^TrTr. 4^4^ ^\^, ^Wft 

'* To these might be added numerous other instances. And 
such words alone were regarded, it would not be absozd 
conclude that the modern dialects retain a greater number «l 
Sjinskrit words in their genuine form than the Prakrits do. Bui 


^ Put, Uff, ton is, however, also used in this duUect, u is the pbiMe Uf 
pui, /alher and ton. 


this would be an unsound conclusion: for the modem rema- 
cukrs, especially when spoken by men who are learned in 
Sanskrit, and as they are seen in Looks written by such 
persons (from which the manuals, giummars, and lexicons 
of such dialects which we use, have been derived), are con- 
tinually rcciuring to their sacred and ancient soiurce (the 
Sanskrit), not only when they want words expressive of recon- 
dite ideas, and required for elegance of diction, but also when 
the vernacular form of the word is more corrupt than learned 
men would wish to introfluce into their WTltings, Hence it 
happens tliat twofold forms of the same fsanskrit words are 
found in the same provincial language, one more Sanskrit, the 
other Prakrit; for the parent Sanskrit has never ceased to 
exercise an influence ou the vernacular dialects of India, just aa 
the Latin does on the Romanic tongues; while on the other 
band the Sanskrit has exercised no influence on the forms of the 
dramatic dialects from the period when the dramatic poets, and 
the grammarians following their guidance, had assigned to these 
dialects certain fixed forms. It has hence resulted that these 
dramatic dialects have undergone no change whatever, and are 
just the same in dramas composed within the lajst three centiu-ies 
as in the far more ancient Mrichhakati. For the language of the 
stage is continuzdly borrowing Sanskrit words, but alters and 
inflects them according to rules pecidiar to itself; the ver- 
nacnlar dialects, on the other hand, continue similarly to borrow 
words from the Sanskrit, but leave them unaltered*^, while those 
words which they bad long ago adopted had been altered ac- 
cording to natural laws common to them with the Prakrits. In 
this way the occurrence of pure Sanskrit words in the verna- 
culars, such as e. g., iik«hna, tira^krila, in the Bengali, is to be 
explained." — Pp. 39 — 45. 

•* It ii *lio In be obserrcil, that mmy of tic Sanskrit worda wliich liave 
been borrowed and modilied in tlie I'aii and Prakrit are, in t!ie modern dia- 
lects, replaced, as far as tbe couimoii people are concerned, bj words ol' 
■boriginal origin ; rucb as beta instead ot {uttm fur son ; wbile words lilce the 
littlvr are used chiefly by Bralimans, and other bl^h-cnstc persons. 



Professor Lassea then proceeds to examine the forms 
decleusioa employed in some of the modem vernaculars. Ht 
then goes on to remark aa follows : — 

[In tJie modem vernaculars] " we find the structure of 
Sanskrit and Prakrit declension quite destroyed, the same inBeo 
tioHH applied to the singular and the plural, and a newdif 
introduced in certain declensions hetween the direct and 
ohhque cases. This proves that the provincial decleosions are i 
a later date than tliose of the dialects used in the dramas, whidi 
are derived from the Sanskrit by certain fixed rules, and involve j 
only a few innovations. In the provincial inflections the 
remain, indeed, some traces, partly distinct, partly somewl 
obscured, of Sanskrit and Prakrit declension ; but in other poiot 
there are great iimovations which reveal to us a total disgolatioa' 
of the old grammatical structure, and its reconstruction by i 
of new instruments. 

" As this state of things is perceptible in the whole graromar< 
the provincial diulects which owe their origin U* the tSanskrlt, 
conclude that they are of later origin than the scenic dialect)!. 
But between the Sanskrit langu^e and its existing daugfat«n 
[the modern vernacularHJ, there is so great a diversity of gram* 
matical structure as to make it certain that the pristine language 
cannot have sunk by one fall, so to speak, into that condition in 
which we find the provincial dialects. It follows of necessity | 
that there must have been an intermediate conditioa between ' 
the pristine and the modern speech. This intermediiite candi- 
tion was no doubt very various, and approached at first moral 
nearly to the Sanskrit, and subsequently to the prorindal 
tongues. I 

•'If we except the Pali, the earliest form of the Sanskrit jifter it 
began to degenerate and to alter its character is that which we find 
in the dramas ; from which dramatic dialect, therefore, we are to 
suppose that the first mutation of the Sanskrit, which eventually 
gave rise to the modern vernaculars, was not very different. I 
contend that, though not identical, this earlie^st coniiption of 



Sanskrit was very similar to that which we find in the drarnas. If 
this opinion be correct, there is nothing to prevent out believing 
that the acenic dialects were formerly the current speech of the 
tiifferent provinces. The names which these scenic dialects 
have received from the grammarians, and the conditions of 
dramatic poetry, lead us to the same conclusion. 

*' Here, however, I conceive I must stop, for I could not adduce 
detailed argumenta to prove this opinion without examining the 
whole field both of the scenic and the provincial dialects. 1 
think, Jiowever, that I ought distinctly to add that I should not 
be disposed to dissent from any one who should assert that the 
scenic dialects were not exactly the pure forms of speech which 
were contemporaneously current in the different provinces, but 
were a little modified so as better to harmonise with the cha- 
racter of the persons who were to employ tliem. The principal 
argument for this conclusion is that two forms are sometimes 
found to occur in the dramatic dialects, one having a closer 
resemblance ti5 the provincial language, and another which is 
softer and, bo to speak, more feminine. 

** To bring this disquisition to a close : there are two families of 
degenerate Sanskrit extant; the first more ancient, and not 
much corrupted, to which class the Pali and the scenic dialects 
belong; the second of more recent origin, and dispersed at the 
present day over the [northern] provinces of India, which is 
more diverse from the parent language. The members of the 
former family are daughters of the Sanskrit; those of the latter 
are its granddaughters, though it is in some degree doubtfiU 
whether they are daughters of the first family or grandchiughtera 
descended from sisters. As regards the age of these two Classen, 
it is proved by the history of the Buddhist religion and of the 
Indian stage that the former arose prior to the commencement 
of the Christian era; while it can be made out with consider- 
able probability that the latter (i. e. the modem provincial 
▼eruaculars) were formed before the year 1000 of the Christian 
era."— Pp. 57- GO. 

B 2 


[csAr. L 

I eubjoiB some further remarks on the distinciion between 
the older Prakrits, aiid the modem remaculars, from the Indische , 
Alterthumskunde of the same author. Vol. ii. pp. 1 149, 1 150. 

" We must draw a distinct line of demarcation between tLej 
Indian languages of the middle age, (imder which denominatioaj 
we may fittingly cUfS the Pali, the languages of the di 
and those employed in the oldest inscriptions), and the 
Indian, or existing vernacular dialect& The former had not, i 
to speak, croesed the Rubicon, nor entirely renounced oWdie&o 
to the laws of their mother-language. They conform, it is 
but little to the ancient phonetic laws, and are regulated for i 
must part by such as are of a later dat« ; but their giammatic 
forms, though corrupted and stimted, are inherited immediately 
from their parent. The modern dialects of India, on the otUe 
hand, have almost entirely ceased to obey the phonetic rules < 
the Sanskrit. They conform in part to the phonetic laws of tb4 
Prakrit dialects, but in addition to these the modem diale 
have peculiar phonetic laws of their own, and their words, wfaenj 
not borrowed immediately from the Sanskrit to enlarge thcu 
vocabulary, often manifest more extreme contractions, and 
greater deviations from the ori^nal words, than do the 
responding words in the Prakrit, The grammatical forms 
the modern dialects are with rare exceptions, newly constrnct«ij 
for the case-terminations are chiefly indicated by post-positions^^ 
the old personal terminations have for the most part entirdj 
disappeared, and the tenses are marked in quite a differnot 
manner than in the Prakrit dialects, the past tenses beLnop com* 
moiily shown by participles, with the three personal pronoujjs i 
the instrumental case. Even the lowest of the dramatic 
Prakrits, the ApaLhransa, has not transgressed this line 
demarcation and btands much nearer to the Sanskrit than 
modern vernaculars do." 



Sbct. IV. — View* of the Indian Grammarians on the relation of the Prah-itt 
to i>anikrit, and on the other elements in their composition. 

Vararuchi", the oldest extant grammarian who treats of the 
Prakrit forms of speech, and his commentator Bhamaha (in 
his ALoQorama), distinctly recognise their derivation, mediate 
or immediate, from Sanskrits The former describes in hia 
"Prakrita Praka-sa" four dialects of this description, viz.: Ist, 
Maharashtri, or Prakrit generally so called ; 2ud, Paisachi ; Srdly, 
Magadhi ; and 4thly, sSauraseni. After having in the first nine 
chapt<?rs laid down the rules for the formation of the Prakrit, 
properly so called, from Ssvnskrit, he proceeds to the others; 
and at the commencement of Chapter X. he lays it down that 
"the root of the Paisachi is the Sauraseui." xnTT^I ^HRfTTI 
'JUIi^^'TllI On which the commentator Bhaniaha remarks that 
Pai^chi is the language of the Pi&ichas.*" Tlie Magacihi also 
18 deolaretl by Varanichi in Chapter XI. " to be derived from the 
same i^urniteni." ^TTTtftl ITRf^I ift'^'ftll" Thef^auraseui 
dialect itself ia Bpoken of at the commencement of Chapter XIT. 
as derived immediately from the Sanskrit. 3^ <^«D I 1T15f?T! 
Mf^dH II" At the end of the chapter on the t>auraseni, it is 
stAted that "in other points'' (which have not been specifically 
touched upon) " it is like the Maharashtri dialect" If^ f|^|<^T- 
iffY^fT II*' From this and from some other quotations which will 

** See on hi» o^e, Lnsscn, Instil. Pracr. 4. S ; Addenda, p. G2 ; and Indlsobe 
Alterthi]in.<kunde, ii. p. 1160, where be is declared to Luvc flourisLcd about 
iho middle of the lirsl cenlury, x.d. 

T^nC^'ftll Cowell, p. 8G, imd Lassen Inst. PfAcr. 7. -WS. 
" Cowcll, p. 89, and Lnisen, pp. 8. 391. 
** Covcll, p. 93, and Liusen, pp. 8 ; and 49. of Appendix. 
** Cowell, p. 96, and Lassen, pp. 8 ; and SO. of Appendix. 

C 3 


tcBAT. ll 

bo found below, it is clear that the ancient Maharasbtri, and 
dialect called by way of eminence "the Prakrit," are the same." 
In another work called the" ShadbhashaChandrika,'* by Lakshmi-j 
dliara, it in distinctly stateit that the "Prakrita dialect had its origis 
in Maliarashtra." IIT^rf if^TJT^V^^T II** As thei^auraseBi 
is fiaid to be derived frojn the Sanskrit, the same must be 
of the Mahrirafihtri, or principal Prakrit, to which the S^aurasci 
in Mjoat points cuuforms. And, in fact, at the close of Vi 
rijchi's ninth section on the former dialect we have it th 
stated in the following Sutra, the 18th : " The rest is [to 
learned] from the Sanskrit;" •^^: ij^frrTT^lP" On which ti 
common tJit<ir rcrmtrks, •'The rest nioaua all that has not 1 
uhfowJy refer re<l to. The remaining rules for affixes, compouu< 
tadilkilas, genders, lett«rB, &c., must be leai'ned from Uii 
HanNkrit." ^^I^sy: ^ff. I HriJ«<H+{T*lrir^dfrf;«J-^<»<»l- 

f^Wi: ^m ^ iji \ri\ 4 ^4 * \^ <H :\\ The derivation of Pi 
from SanHkrit is here distinctly implied, and, in fact, the 
tiling resiiIiH from the whole series of rules for forming P; 
wordH, which are nothing but explanations of the manner i 
which the Sanskrit forms are modified in Prakrit. The 
origin is ascrihed to Prakrit by Hemacbandra, who «« 
THRfn: ♦i«d*?,l fT^Vii 7m ^^TRff m ^\mff^\\^ " It hat 
its origin in t^aiiskrit, Prakrit i.s that which springs, or comes 
from Sanskrit," Of the Prakrits handled by Varanichi wi 
tlms see that three derive their names from three proviuces 
India, viz,, Maharashtra, ^lagadha, and t!ie country of 
f^urasenas, the region round MathimT. This, as we have alread 


*" Tbat the MabarTishtiT of that period was not the same u tte moJen 
Maliratti, appeam, (I neeil scarcely Bay), from the character of the fumtf, 
us shown in the dramatic worka in which the Prakrits are emploje«i. 

" LasKn, p. li. 

*' Cowcll, pp. 85. and 176. 

*' Cowcll, [I. xvii. Lausen, p. aC 



seen above, p. 46. is considered by Lassen, and justly so, as a 
strong proof that they were spoken dialects. 

Four kinds of Prakrit only, as we have thus seen, are men- 
tioned by Vararuclii, the oldeflt authority on Prakrit Grammar, 
viz., Mabarfiehtri (or the principal Prakrit), tsaiiraseni, Maga<lhi, 
and Paisachi. Though many other dialectic varieties are referred 
to by later grammarians, it ia not necessary for my purpoee to 
give a detailed account of any of tbeae. 

Vararuchi devotes nine chapters, containing in all 424 aphor- 
isms, to the Maharashtri ; one chapter containing 32 aphorisms 
to the peculLaritiea of the J^auraseni ; another chapter contain- 
ing 17 aphorisms to the Magadhi; and a third chapter con- 
taining 14 aphorisms to the Paisachi. At the end of the 
separate chapter on the i^auraseni, it is said that it agrees with 
the Mabfirashtri in all other points, except those which have 
been specially noted as peculiar to itself; and the same thing 
may be presumed in regard to the other two dialects. 

It ia clear from this mode of treatment alone, that the 
points in which these four dialects, and especially the Maha- 
rashtri and the i^aurascui, agree with each other, must be much 
more mmierous than those in which they differ ; and this con- 
clusion is confirmed by a comparison of the specimens of the 
several dialects wliicli are extant in the dramas. Accordingly, 
Professor Lassen remarks, ( Instit, Prac p. 377) that '• the 
principal dialect, and the ^aiirascni, coincide in most respects." 
The technical distinction made between these two dialects by 
the grammarians Ih, tliat the one (the Sauraseui) is the language 
used in prose, while the Maharashtri is appropriated to verse 
(Laasen, p. 384.) The same author remarks of the Magadhi, 
that it does not depart much fmrther from the Sanskrit than the 
principal Prakrit does (p. 387); and that the Indian gram- 
marians are wrong in deriving the Magtulhi from the Sauraseni, 
Ito the former ia as directly descended from the Sanskrit as the 
latter ; and that the two derivatives coincide with each other in 
most respects (p. 437.) The Paimichi, (a dialect employetl by 

B 4 




, ia Hke manner, to hare 
Jntbraprooeas peculur 

beai dan«d<{*reetfy 
to itself (p. 447.) 

In regard to tboe Pxakm «fi«Wt» gii i illy, Tamra reuftib 
(p.386)MfiaDo«s: - tkat tike SmAiitk lu^wges of Hindw 
4hi Fniper «ae iiwmilj kas difawit froDi each oUier than 
they iMnr are^ is to be ia fc ti e d firom tlie &ct that, at tint 
ear&er period tbcy bad not «iTitwl ao iar from their comiooa 

The ibUoving paoage, qwAed by T.aaBen, Instit. Vnent., 
p. 17^ from a work called Piakritadiplka, by Cbaodidevi, 
•eems also to show that Prakrit was a batgoage in current we, 
as well as employed in the dramas : — V^d^fM wV^RTfTTTnT 

irreri fwrfcfHll "This Prakrit of the Maharashtra ooaiitit,fl 

BO called from its coaformity to popalar usage, and from its 

being employed by poets in dramas and other poems, is 

most excellent form of speech. Thus Dandi says 'The 

which prevails in Maharashtra is considered the Ijest,*'' In 

same way Rama Tarkavagiea, in his Prakritakalpataru, d 

" the MaharJisljtri dialect to be the rout of the others;' 

H^S^f'TTr ^^I^TT^II" »f»ti that "the tfiauraseni is derived^ 
from it.' f^i^d i\*-nU 'sfTT^ ^ Wm Tf^'^ 

f^mfm :!!"^ "The Magadhi is said to be derived from these 

two." ^TTii ?TT»n2iHf^*yd - - - "^mj ''(T^ru^^rs^j^- 

** Prtkrltftkulpafarii, rjuolcd hy Lasacn, p. 20. 
** Ibid. 2nd ^ukbu, UlSUvnka. 

'WT^ "R^^: 3T?5?ft f%^ll*' These languages, together 
witb the Ardhamagadlii and the Dakshinatya, are called 
bhdshas. The author then refers to the second class, called 
vibhdslias, the dialects called Sakari or Chandalika, .^aluu"!, 
Abhirika, Dravida, and Utkali, which, he says, " thovigh charac- 
terised by rusticity (apabhransata), are yet not to be ranked in 
the class of npabhrauHas if they are employe<l in dramas." 

•TT^fTRfV ^ ^PT?1?r^ ^fTT^t^THim^ fT^ij: I) " On 

the other hand, the forms vf those vibhdshas wliich are not 
used in the dramas are reckoned by the author among the apa- 
hhram^a dialects, imder which name he understands the pro- 
vincial languages, such as the Bengali, Gnzarati, &c. A third 
class of languages is called by this author the Paisachi. 

The Kavyachandrika, a work on poetry, has the following 
remarks on language: — 

cT^ «<r^«i f^^Ffi^ ^w iTTfirf fvm\ ^rnJirg f^r^r^ 

" In H'gard to liuiguuge, let it be \iuderst«iod that there are 
four kinds, viz., Sanskrit, Prakrit, Apabhransa, and mixed. The 
Munis call Sanskrit the language of the gods; and consider that 
there are several kinds of Prakrit, viz. {Tadhhava), that which 
is derived from, and {Tatsiima) that which corresponds with, 
Sanskrit, and [DchI) the provincial."*' On this passage the 
scholiast remarks : 

** Ibiil., 2nd Stavaka, 

" Ibid., 3rd Staraka, (Lassen, p. 21.). 

'* Kiivyiicliandrika, quoted by La'tcn, p. 32. 



"The word iadhhava means derived front Sanskrit^ like the 
word l-hagga, &c., for Ihadga. Tataama meanathe words which 
are fdllie in Sanskrit and Prakrit, like kindl., rahanda, S,c, 
Desi means the Maharasbtrl, &c Apabhranea is the speech 
of the Abhiras, &.c. Tlie mixed dialect is that of the dramu, 
&c."*^ On this I would remark that though the Mahara«htn ii 
generally recognised as the principal Prakrit, a Desi elctneiit 
is here recognised as existing in it, or contemporaneously willi 
it. Must not this be an indigenous, non-Sanskrit, element? 
To the same effect is the fallowing passage from the Kavyi- 
dar^ of Dandi : — 

ff^^^ ^T?ir^ ^^: ^M^ T^rwi^ nm\ ^^r^w^ W 
TTTPJT^n ^TTTT "Rfi^ 111^ f%^: I wttt: ^?»<^Mi 

" Writers of autliority say that there are four kinds of lan- 
guage : Saneki'it, Prakrit, Apabhransa, and mixed. Great Rishis 
denominate Sanskrit the langiwge of the gods. There are seTeial 
orders of Prakrit, viz., (ludbftara) that which is derived from, 
and (tatsoina) that which corresponds with, Sanskrit, and the 
provincial (den'i). TliP language of ^laharashtra is culled the prin- 
cipal Prakrit, and it is an ocean of pearls and gems produced in 
the Setubandha, &.c. [This line is corrupt ; nnd the above sense is 
assigned as a conjeclnie. The allusion appears to be a double 

'■"^ Scholiast on the fame passage, ibiJ. 
'■'*' Marginal gloss quoted hy Lassen. 


one : first to the ancient Prakrit poem called ' Setiibandha *' ;' 
and secondly to the reef of Setubandha, a line of rocks between 
India and Ceylon, in the vicinity of the Ceylonese pearl 
fislieries.] [In dramatic poetry] the J5auraseni, the Nati, the 
Gaudi, and such like dialects, follow the law of the Prakrit 
according to their several provincial iiaages. The speech of the 
Abbiras, and other such tribes, when occurring in poems, is 
called Apahhraiiiiu In books on grammar, whatever differs 
from Sanskrit is called Apabhransa."*" 

In bis note to the introduction to Campbell's Telugu Grrammar, 
p. IJ., Mr. F. W. Ellis remarks as follows on the Shaxlbhnsha 
Cbandrika of Lnkshmidhara, above referred to (p. 34 :) •• The 
work here noticed is confined to these dialects [the Mahurilshtrt, 
^aurasent, Magadhi, Paisacbi, Cbulika-paisachi, and Apa- 
bhransa], as they now exist in the Natakas [dramas], and treats 
therefore only of Tat^amam and Tadbhavam t«rma of Sanskrit 
origin ; it is expressly stated, however, that each possessed it^s 
proper Desyam, or native, terms; and it is probable, as many of 
these dialects prevailed in countries far distant from each other, 
that each was connected with Desyam words of various deriva- 
tions, in conjunction with which they produced spoken languages, 
differing considerably from each other. This in fact is declared 
to be the case with respect to Paisachi in the following,' passnge," 
[which I give in the Devanjigari character] : f^lIT^fT^rf'r^rFf 

xnrN^tf^^ f^^: I rM'JHM<aiii^ ^^^n*: i qni^^gr^!!^- 

IITI 1^ ^irn^t^- ^^^T'JRT^T H^ ['". e. Two 
kinds of PaiiJiiehi are recognised, which depend on the different 
Pimicha countries. These are declaimed by the anciente to be the 
following, Pandya, Kekaya, Vahlika, Sabya, Nepala, Kuntala, 
Sudbesha [?], libota, fianilhara, Haiva [?], and Kaiidjana [?]. 

*' Sec role, p. x. and note 3. p. 26 in CowcH's Prilkrita-praka!!*. 

•» From the KnvvuJar^ of Dnncji, hb quoted bj Liistcn, pp. 'ii — 33. 



[CRAF. t. 

These are the Paisacba countries; and the native of each 
country has his o«Ti partjcnlar qualities.] *'The two Pai^ii 
dialects are said to prevail in all the countries here mentioned, 
commencing with Pandyam at the southern extremity of 
India, and extending to Canoj (Canojana) in the north, .... 
and it is added, These are (he Pa!«achi coitntries, and 
the Deiiyarn tenns of each have their cnvn partiaUar quality* 
Tlie concluding phrase is more vague in the original than 
Mr. Ellis has rendered it; but as lanffuaoe is the subject which 
the niithfir is treating, it is to be prcsiiimed that he here alludes 
to the peculiar character of the different provinces in respect i 
their varieties of speech. 

It is irrelevant to our present purpose to inquire particularly 
wlittliLT the various distinctions adopted by Vararuchi and liissuo 
PCKHiirK, of tilts iiiediiile or immediate dei'tvation of the Prakrits 
from Sanskrit, and their classificationa of Prakrit, into that which 
is properly so-called, and Apabhransa, and Paisachi, are merely 
arbitrary and fitctitiouj?, or are founded on any rational principles 
It is ennuyli that I find the following facts, which are important 
to the conclusions I am seeking to eetablish, admitted by the 
native authorities I have just cited ; viz., first, that the Prakrits 
are derived from Sanskrit as their somce ; secondly, that they 
arc comptiKod of athrecfold element: Talsainam, pure Sanskrit; 
Tadbhavam, derived from Sanskrit; and Dm local. As this 
third element, Desi, is di.stinguished both from pure Siuiskrit 
and from words derived from Sanskiit but altered, it must 
fiillow, thirdly, that it deuote-s words which were regarded as 
having an origin different from Sanskrit. Such, at least, i* 
indubitably the sense in which the word Dft-fi is used by Telugu 
writers.^* This confirms, by the authority of the Indian gram- 

•* See Ciiniplieirs Tuliigu Gramnuar (3(1 edit., MofJra«, 1849.) p. 37, 

where it ia saiil : — "The words of the Teloogoo Inngiisge are 

classed by Sanskrit graDiniarians under four distinct head?, let. Di*hyxtmooy 
or, as it is more einphntically termed, Utsu Deshi/nmon, the /)i/re lan''ua<»eof 
the land ; 2nil, Tulsumumoo, Sanskrit words assum'mg Teloogoo tcruiiui' 


marians, what I have already asserted above, (p. 41.) to be 
established on other grounds, viz., that languages exist iu India 
which have an origin independent of Stinskrit. 

To give a complete idea of the artificial manner in which the 
Indian critics classify the different Prakrit dialects, and of the 
different classes of people to whom they conceive the dramatic 
writers ought to assign them, I quote a long passage from the 
Sahitya Darpana: — 

WTfirT f% <0^'il*i.l ai*Kl1't ii^rT^^ ail*lO 
^imtTH TT^iTS^ftft ^gn5jT^ 5**^1 r< HI "^rn^ Tin^ft 

Jjm ^ ^JTT^ ^^^kT lit^l ^^^w y*iTi^ «[TfT- 

tions ; 3rd. Tudhhuvuiiioo, Tt-Ioogoo corrupiions of Sanskrit worda, furmcd 
bjr tJi« substitution, the iTi!>iun, ur addition of letters; 4lh. Grumx/umoo, 
firovincial ternu, or words peculiar to the vulvar. To these wc may also add 
Vnyu Dhkywnoo, or words froin other countries, sometimes pivcn ts a sub- 
division of the first class, and comprl&in^, according to the defiottion of 
kncient writers, words adopted from the dialects current in the CuJcarese, 
Mahrnlia, Guzerat, nnd Dravidn proTiiiccs onlj, bnt now also including 
■evcral of Persian, Ilindoostnnce, and English origia." 


ni^3rroHY ksv relations 

tcuAr. % 

" Let men of respectable rank and cultivated minds s; 
Sanskrit ; and let women of the same description use Sauraser^ 
except in the metrical parts where tliey sliould talk Mahdrdalitri, 
Persons living in kings' palaces should employ Md^adhi, and 
servant"^, kings' sons, and magistrates Ardharndyadhi, The 
eastern dialect (which the scholiast says is Gaudi, or Bengali) 
should be spoken by buffoons; and the Avantihy crafty periions. 
Let Dukshinatyd (the language of Vidarbha, according to the 
scholiast) be employed by soldiers and citizens ; and Sdkih'i by 
$akara8, J^akas, and others. The Baldika dialect is the Otte 
proper for celestial (?) personages, Drdridl for Dravidas^ Ac, 
Abhin for Abhiras, CfidnddU for Pukkasas, &c, the Abkifi 
and Sdvari for those who live by cutting wood and gathering 
leaves, and Paimchi, the speech of Pisachas, for charcoal burners. 
Saurcuieni may be used also for female-servautH, and women [?3 of 
the better sort, for children, eunuchs, and low astrologers; ihe 
same, and occasionally Sanskrif, for madmen and sick persons. 
Prakrit shoidd be employed by those who are intoxicated 
by authority or affected by poverty, by mendicants and pri- 
soners, &c. Sanskrit should be assigned to the better sort of 
I'cmalc mendiaiuts, and also, as some say, to queens, ministere* 
daughters, and harlots. A dialect belonging to the country 
from which each character of low origin coraes should be 
asagned to him; and the language employed even by tbe 
superior personages shoidd vary according to their function. 
Sanskrit, varied by other dialect^i, [?] should be assigned with a 



view to politeness to women, female friends, children, liarlots, 
gamblers, and celestial nynipbs."'"'* 

I shall conclude this section by addhig tlie substance of what 
Professor J.*a»$en says about the Prakjit dialects in the earlier 
portion of his work (pp. 22. 25 — 29.) 

" The word prahrlt comes from/)rfd7/V^'(procreatrix), nature, 
and means 'iJenved;^ the several Prakrit dialects being re- 
garded as derivatives of Sanskrit either directly or mediately. 
The original language from which any other springs is called 
its prakriti, or source. Thus Heraachandra says, ' Prakrit has 
its origin in Sanskrit; that which is derived, or comes from the 
latter, is called prulcrltu,''^ The expressions Sanskrit and 

(Prakrit are opposed to each other in another sense, when 
the former word denotes men of ciUtivated minds, and the 
latter those who are uncultivated. The term Prakrit is therefore 
also applied to vulgar and provincial forms of speech. 

" The grammarians concur in considering Maharashtri as in 
the strictest sense of the word Prakrit, the principal form or 
type of Prakrit. The J^auraseni and the Magadhi approach 
most nearly to the Maharaslitri, and both derive their appella- 
tions from the names of provinces. By these three provincial 
de«ignationa, Maharashtri, Sauraseni, and Magadhi, the Indian 
grammarians appear to have understood the local varieties of 
language employed in those three several provinces, as well as 
the dramatic dialects severally so called. V^araruchi specifies 
only one inferior dialect, the Paisachi, and understands by 
it the form of speech employed by the lowest classes of 
men. This is to be distinguislied from the speech of Pi^has 
(goblins), which, when introduced on the stage, are said 
to tise a ^bbcrish totally ungrammatical. The word is to 
be understood as figuratively used to denote the contempt in 
which the lowest clatacs were held. Hemachandra mentions a 

** SShityn Darpina in Ribliothcca Indies, Xo. 63. pp. 172. 173. (See also 
LoMcn, Inst. Pracr. pp. 35, 36 ) 

•* Heniachnn<lro, viii. I., T.flMen, p.2('i; quoted nbove, p. 54. 



variety of this dialect; the Chulika-paiMchi, which denotes « 
form of speech lower than even the former. In fact, 
varieties of Paieiicbi appear to be distinguished by the gram 
marians", Ixith of them spoken by barbarous tribes, of which 
one seems to belong to northern, the other to southern, India, 
Rama-Tarkavagi4a also mentions two sorta of Pai^achi, signi- 
fying by this name a rude mixture of language drawn fr<jm 
diSVreut idioms. 

*' I'he terra ajmhhrania is applied by the grammarians to th 
dialects which are the furthest removed from the pure Sanskrit 
original, and have undergone tiie greatest corruption. HeniS' 
Chandra specifies two kinds, of which one has naost affinity with 
the principal Prakrit, and the other with the Sauraseni. The 
older writers assign this dialect to the people who dwell on the 
shores of the western ocean, especially the Abhiras. Rmns 
Tarkavfigisa, departing from the view of the earlier writers, 
ascribes the varieties of the local and provincial dialects to the 
apabhransa, as their source. The same author seems also (whea.| 
he uses (iii. 1 .)tbe words nngddikramdt, "according to the manner 
of those who speak like Ndf/as, or serpents,. &c.^'), to assign a 
mythological name t/o the provincial delects in the same way U 
the older writers talk of certain barbarous tribes aa Pisachaa. 
This designation appears to have proceeded from the writers uu 
rhetoric, who assign Sanskrit to the gods : Prakrit is then left 
for men ; while those whom the Brahmans consider to be 
scarcely deserving of the name of men, Chandalas, Abhinis, 
and such like, are only fit to utter the speech of goblins, or 

"Tlie Prakrit dialects employed in the dramas are rightly 
asserted l>y the grammarians to be of Sanskrit origin ; for both 
the grammatical forms and the words, with very few exceptions, 
&a well as the entire structure of the Prakrits, and the character 
of their syntax, are derived from the Sanskrit. When, however, 

•* See the passage rjnoted in p. 59. 


the more recent gramraariaQSi assert the same of the Canarese 
and other »South-Iiidian dialects, they are in error, as, although 
these languages contain words formed from Sanskrit according 
to certain rules, their grammatical forms and primary words 
cannot hy any possibility have been drawn from that source." 

The later native authority to whom Professor Lassen here 
refers appears to be Eaina Tarkaviigiso, (p. 23.) I will hereafter 
show (when I come to refer more particularly to the South- 
Indian languages) that the Indian grammarians of the south 
claim for the Telugu, and no doubt for the Tamul, Canarese, 
and Malayalim, also, an origin quite independent of the 

Sect. V. — The Pali ; ami its relations to Saiukrit and PrahriL 

The above tabular comparison of the Prakrits with the mo- 
dem vernaculars, will have abundantly shown, that the latter 
are derived from the former, and that both are derived in great 
part from the Sanskrit, the one mediately, the other more 
immediately. Though, however, it be sufficiently clear, both 
firom the authority of the native grammariaua and by a com- 
parison of the Sanskrit and the Prakrits, that the latter are 
derived from the former, yet the Prakrits do not represent the 
derivative form of speech which stands nearest to the Sanskrit ; 
and we are in a position to point out a dialect which ap- 
proaches yet more closely to the latter than the Prakrits do. I 
mean the Pali, or sacred language of the Buddhists; a language 
which is extinct in India, but in which numerous canonical \>f>«V» 
of the Bauddha religion, still extant in Burmah and Ceylon, are 

" Sec Dr. Caldwell's Comp. Grammnr of the Drnvitlian lnng:tispc8, pp. 
30, 31. ; the Introduction lo Cainiibell's Telugu Griuninar, 3ti tA~, Madras, 
IMS, pp. xr. S. ; and the Note, in the same work, hj Mr. Ellis, to Mr. 
Campbell's Introduction, pp. 1 1 — 22. 

•• If anj Brahmanicitl reader BhoulJ think of studying these page*, I hope 
that the connection of the Pull language with the Buddhist religion will not 
deprtTC it of alJ interest in his eyes, much less induce him, with the author 




[coat. 1. 

fate vSM 

Though, however, this language has had the singular fate 
having now disappeared from its native soil, to become a 
language in foreign countries, it is yet nothing more 
one of the ancient vernacular dialects of Northern ludifti 
Magadhi is the appellation which the Buddhists of Ceyloil 
theoiselves give to it. It is, indeed, true, as we are infor 
by Mr. Tiunour, that the '* Buddhists are impressed 
the conviction that their sacred and classical language 
Magadhi or Pali, is of greater antiquity than the Saaskrii; 
and that it had attained also a higher state of refinement tixaaj 
its rival tongue had acquired. In support of this belief ihcyj 
adduce variouB arguments, which in their judgment are qt 
conclasive. They observe that the very word ' Pfdi ' (dgnifie 
original, text, regularity ; and there is scarcely a Buddhist Pi 
scholar in Ceylon, who, in the discussion of this question, will 
not quote, with an air of triumph, their favourite ve 

WRT H'm^l '^f^ >f 1 H < tl * Tliere is a language which is 
the root (of all langtiuges) ; men and Bralimans at the commence- j 
ment of the cj-eation, who had never before heard or uttered i 
human accent, and even the supreme Buddhos, spoke it : it 
Magadhi.'^' This verse" is a quotation from KacUchayauo's 

of tLo Uyay& miila vistora, I. 3. 4, to regard it, though of pure Sanskriftfl 
original, as polluicil, like cow's milk in a dog's skin, (iff^ ^rf WIT 
'n^TT ^5 rll \j rJM I ) ^>y the unholy contact of these heretics. 

'" The idea enlcrlainuU by the ButWhists of the «uperioritj of the Pali U> 
Sanskrit may also be learnt from the following passage of the commentary 
on the Grammar called RupusidJbi, Joscribin;^ the result of the comixMitioti 
of Kachuhayano's Grammiir: TJ^ Hf?l ^T^l <^4:|H l<JI 4j^ril f^ 4^- 

f«rrT?^^^HI*l< i|t^. . .^T^^^^ ^JJlf^i^jpHlg 

"This lieing done, men, overcoming the confusiim and incorrectness 
diction, arising from the mixture of Sanskrit and other dialects of t« 
countries, .... will easily acquire the doctrine of liuddito." Mabivi 
Introd. pp. jtxvi. xxvii. 
^' Preserved in llie grammar called Payogasiddhi. Turaour, p. xxvii. 






Grammar, the oldest referred to in the Pali literature of Ceylon. 
The original is not extant in this island." JMr. Tumour, however, 
18 inclined to "entertain an opinion adverse to the claims of the 
Buddhista on this particular point [tlie priority of Pali to Sans- 
krit]. The general results of the researches hitherto made by 
Europeans, both historical and philosophical, unquestionably con- 
verge," he thinks, " to prove the greater .<ujtit]uity of the Sanskrit. 
Even in this island," he proceeds, " all works ou astronomy, medi- 
cioe, and (such as they are) on chemistry and mathematics, are 
exclusively written in Sanskrit : while the works on Buddhism, 
the histories subsequent to the advent of Gotamo Buddho, and 
certain philological works alone, are composed in the Pali 
language." (Mahawanso, Introd. pp. xxii. xxiiL) There ia no 
• [uestion that Mr. Turnour is right, and that the priests of 
Ceylon, who are no philologists, are wrong. The Pali bears as 
distinct traces of derivation from Sanskrit as any of the other 
nortiiem dialects. Before, however, adducing the proofs of 
this, I must give some account of the manner in which the 
Pali was introduced into Ceylon. 

The appearance of Buddha as a religious reformer in Northern 
Hindusthan seems to have taken place in the earlier part of the 
sixth century before Christ. He is said to have entered on his 
mission in the year 588, and to have died in 543, b. c. (Turnour, 
Introd. to Mahaw., p. xxix.y In strong contrast to the Brahmans, 

" The grounds for preferring the Cingalese date of Buddha's death, 543 
or 544 B c, to that of the Northern liuddhista, are set forth by Lasseu, lod. 
Alt. vol. ii. jip. 51 — 61. See especially pp. 60, 61. The historical value of 
the Buddhiat records i^ according to Mr. Turnour (Introd. p. xxviii.), 
lanired in the following way:— "The age in which we now live is the 
BwUhotpodo of Gotmno [the interval between the manifestation of one 
Buddho and the epoch when his religion becomes extinct] llis religion 
wua de«tined to endure 5,000 years; of which 2,380 have now passed away 

(a. i>, 1837) since his dii»th, ai>d 2,620 are yet to come By 

tbia fortunate fiction, n limitation has been prescribed to the mystification in 
which the Buddliisti(;al creed has involved all the historical data contained 
in its literature anterior to tlic advent of Gotaina. . . . The niyatifica- 

r 3 


he and his followers stroye to disseminate their new doetriaeifl 
in a popular shape among all classes of Bociety, and for thi< ' 
purpose employed, where necessary, the ciurent vernaculari 
dialects of their age and country, though, at the same time 
they may have used both Sanskrit and Magadhi in the oomj 
Bition of their sacred works, (Lassen, Ind. Alt, ii. 492, 3; 
1147, 8: Burnouf, Lotus de la Bonne Loi, p. 862.y» 
Buddhist synods were held at different periods within 3C 
years after Buddha's death, for the collection and arrangeme 
of the sacred works which expounded the doctrines and diHciplii 
of his religion; for the correction of errors and abusea; and 
for the purjKise of propagating the new faith in foreign coua-1 
tries. The revelations of Buddha are stated by his follower 
" to have been orally pronounced in Pali, and orally per- 
petuated for upwards of four centuries, till the close of the 
Buddhistical &ge of inspinition." They consist of the Pit 
kattaya [in Sanskrit Pitakatraya], or the three pitakas, wbicll 
now form the Buddhistical Scriptures, divided into the Mnaj 
Ahhidhiu-ma, and Sutra pitakas. A schism having arisen aH 
Buddha's deatli, the first Buddhist council was held in 
when the authenticity of this Pali collection was establislie 
and commentaries upon it, called Atthakatha,were promulg 
At the second council, in 443, b.c., the authority of 
Pitakattaya was again vindicat^Ml, and the Atlhakatha deliver 
on that occasion completed the history of Buddhism for tfc 
interval giibsequent to the previous council In the year 3C 
B. c, the third council was held in the reign of King Aioka, whd 

tion or tlie Buddhistical data coased a century at least prior to n. c. 5S 
when Prince Siddhattho attained BuddkoLood, >u the character of Gc 

'* Benfey is of a different opinion. lie says (Indien, p. 194), tb 
Buddhist boolcs of Nepal composed in Sanskrit arc, "as we shall bei««to 
show to be probable, merely trunslalions from the Buddhist sources, whkk 
were originally composed in Pali." 



was a zealous promoter of Buddhism [Turnour, p. xxix.]. Various 
missions were consequently undertaken.'* Mahendra, the son of 
King A^oka, was sent on a mission to Ceylon, for the conversion 
of tliat island. 

The following account of his proceedings is given by the 
native authorities, as abstracted by Professor Lassen (Ind. Alt. ii. 
pp. 247 — 253): — Mahendra arrived iu Ceylun Lu the year 
245 B.C., was hospitably received by the king of the island, and 
began by his preaching to convert the inhabitants to the religiou 
of Buddha. The king himself embraced the new doctrine. Re- 
lics of Buddha were transported to the island from Northern 
India, and the Bodbi tree, under which Buddha had attained 
the most perfect knowledge, was trau apian ted thither from 
Behar, and according to the belief of the Buddhists, continues 
to flourish to the present time. Many miracles attended these 
transactions. The conversions to Buddhism continued ; and many 
male and female devotees were consecrated to the Buddhist 
priesthood. Buddhism, thus introduced, has ever since remained 
the creed of Ceylon ; and that island, the head-quarters of 
S>outhern Buddhism, and the seed-plot from which it was pro- 
pagated into BurmaJi and other parts of Transgangetic India, is 
regarded in those countries as a holy land. In Ceylon there 
exists, as has been already mentioned, an extensive Buddhistic 
literature, which fills up an important blank in that of the 
Brahmans. This literature is, as I have stated, in Pali, 
At first, however, the principal sacred records of the 
Buddhists are said to have been handed down by oral tradi- 
tion. Mr. Tiu-nour (p. iLJcix.), gives the following statement on 
this subject from the native authorities : The Pitakattaya, to- 
gether with the Atthakatha, completed to the era of the third 
Council, were orally promulgated in Ceylon by Mahendra, 
the Pitakattaya in Pali, and the Atthakatha in Cingalese, 
with a further Atthakatha of his own. These works were, it 

" See Lass«n, Ind, Alt. ii. pp. 79. 86. 229. ff., «id 234—240. 
r 3 



CHAr. I. 

and snofl 

is said, propounded orallj by his inspired disciples and 
oessors till the close of the period of inspinUion, which 
in Ceylon between 104 and 76 B.C. They were then conK 
mitted to writing, the text (Pitakattaya) in Pali, (in which it 
had before been handed down orally), and ita oommentariei 
in Cingalese. This event is thus celebrated in the Maliawanso, 
chap. 33, p. 207. fMH*Tli<Mlf^^ fTWT ^ ^«ft^^ WI 

f^^m^ll "The wise Bhikkhus of earlier times had handed 
down orally both the text (Pali) of the three pitakas, and their 
atthakatha. But at that period, perceiving the injury which 
would otherwise be sustained by the people, the Bhikkhus 
sembled and caused them to be written down in books for the 
more lasting stability of the faith." About 500 years later, in 
the period between 410 and 432 A.D., Buddhaghosa transferred 
the Cingalese Atthakatha into Pali, as related in the 37tll 
Chapter of the MuLawanso. These Pali versions of the Buddhist 
scriptures and their commentaries are those now extant in 
Ceylon, and they are identically the same with the Siamese and 
Burmese versions. 

Such are the Buddhist traditions regarding the oral 
mission of their sacred books, viz., the ScriptorM themsel' 
in Pall, and the commentaries, &c., in Cingalese, and their 
flxibsequent consignment to writing. It will be seen, however, 
that 80 much of this narrative as records the oral transmission 
of these works, is distinctly rejected by Mr. Tumour, who siys, 
p^ Ivii. " although there can be no doubt as to the belief en- 
tertained by Buddhists here, that these scriptures were per- 
petuated orally for 453 years before they were reduced to writin*', 
being founded on superstitious imposture, originating perhaps 
in the priesthood denying to all but their own order, acceas to 
their scriptures ; yet there is no reaaonable groimd for question- 




ing the authority of the history thus obtained of the origin, 
recognition, and revisions of these Pali scriptures," 

Kegarding the introduction of Pali into Ceylon, different views 
have been taken. In his " lustitutiones Lingufe Pracriticae," 
Professor Lassen remarks as follows, (pp.60, 61): — 

" It is clear that the Pali is the sacred language of the 
Southern Buddhists, {. e. of those who departing, for the most 
part, from the shores of Kalinga, towards the south, carried with 
them, first of all, the doctrines of Buddhism into Ceylon, and 
erentualiy propagated them in India beyond the Ganges," 

And again : — 

" While the Pali is connected with the emigration of Bxul- 
dhism to the south, it was itself, vrithout doubt, produced in 
India. It is by no means clear whether the Buddhist-s, when 
they travelled southwards, made use of the Pali language from 
the first or not; but indeetl, as the commencemeut of the 
emigration to Ceylon can scarcely be placed earlier than from 
628 — 543 before Christ, the application of the Pali dialect as 
a vehicle for communicating the Buddhist doctrines can scarcely 
have taken place earlier than that period. How much more 
recent it may be, I leave to those who may be endeavouring 
to trace the history of this sect, to discover." 

In his later work, however, the " Indian Antiquities," (vol. ii. 
pp.488 — 490), Lassen proposes the following theory on the 
subject, which I translate, with slight abridgements: — 

"The Pali language is calle<l by the Buddhists of Ceylon 
Magadhi, and it ought consequently to have had its birthplace 
in Magadha. This, however, cannot have been the case, as, like 
the majority of the dramatic dialects, it does not possess the 
peculiarities of the Magadhi. Tlie Buddhists are also wrong when 
they declare the Pali to be the root of the Sanskrit, and assert 
that Katyayana restored it to its original perfection by purifying 
it from all intermixture of Sanskrit and the provincial dialecta. 
We shall therefore have to seek for the birthplace of the Pali 
elsewhere than in Magadha. We must necessarily assume 

r 4 



CcBAT. 1,1 

it to have been once a vernacular dialect, as it is otlicnriMi 
impossible to perceive why it should have been selected as the 
language of the sacred writings. There ia, besides, nothing in 
its character whicli is opposed to the supposition that it mt 
once a spoken tougue. If we compare it with the langiuge 
of the Western inscriptions, we find that, generally speakin^J 
they stand both etjually removed from the Sanskrit ; for if the! 
one presents some forms which are older, the other again 
other forms which are more ancient."^ The western inscriptioDttl 
have, in addition to other differences, also the pecidiar phnaeticj 
rule of chauging tvd into ptd, (e.g. dfisayltvd [Sanskrit (/ansa- 1 
yi/i'o] into thtsaylptd), which is unknown to the Pali, as well Ml 
to the dramatic dialects. These discrepancies render it impossibl* 
to identify the Pali with the language of the western inscriptioni 1 
It is besides to be observed, that Buddhism had not its pritt-j 
cipal seat on the western coast, where the dialect in questiaii| 
was vernacular." 

Thus, according to Lassen, the Pali is neither identifial»le with! 
the Magadhi, the language of Eastern Hiiidusthan, nor with 
the dialecta of Western India, as made knoi^Ti by the western j 

" In the abseuce of auy other circumstance to indicate Uiaj 
birthplace of the Pali, (Professor Lassen proceeds,) I 
pose the following conjecture on the subject. I assume 
Katyayana selected the speech of the country in which 
was engaged in propagating Buddhism, ?'. e. of jSIalwa. 
the Prakrits employed in tiie draiuiis, the Saiu-aseni is tlie' 

'* Thus the language of the inscriptions preseryes the # before ( and tk, , 
as in aiti, in sftthe, and in usthdna; and the r in sarvra, where the Pali hu\ 
ttk, tlh, and OP. The inscrijUiona, too, preserve the Sanskrit diitive^ for i 
the genitive is used in Fali, though the grammarians recognise the ei 
of the dative. In Pali the ablative in sma, as well as mha, and the locatiT 
mtmin as well as mhi, are found, though they ore rarelj used in compoMti 
In the inscriptions, on the other hnnd, (he locative has the form mhi, ■ 
the ablative of words in a is a, so that the pronominal declination of Uiis i 
has not yet been transferred to the noun. 


oae most frequently employed, and is the variety UBed in the 
prose passages. Vararuchi derives it immediately from the 
Sanskrit, and from it the other dramatic dialects. He must 
therefore have considered it as the oldest-, thougii he (as well as 
hU sucoeissors), regards the dialect called Mahanlshtri as the 
principal. These two dialects stand the nearest to the Pali, 
though it in decidedly older than they are. I conjecture, there- 
fore, that we may regard it as the oldest form which has been 
preserved of the vernacular language of Western India between 
the Jumna river and the Vindhya range, a tract which includes 
Malwa. The iSauraseni would cousecjuently present a later form 
of tliis language. From Ujjayani a knowledge of Katyayana's work 
was probably diffused over the Dekhan ; and the Cingalese derived 
their acquaintance with the dialect of which it treateil from the 
country of the Dauiilas, i.e. the Taiuiliaus, or the Cholas. In 
that country, Dip;iukaia, surnamed Budtlhapriya, composed Ids 
new arrangement of that work, the oldest Pali grammar now 
extant,^* A« the canonical writings in Ceylon were not translated 

" " The oldest version of the compilation from Kaohchajano's Grammar," 
\j» Ui. Turnour, (Introd. to MjiLiiw. p. xxr.), "is acknowlcdired to be the 
,Spu)ddhi. I quote three {lassages .... The first of tbe»e extract* 
[from the conclusion of the Kuiiaaiddhi] .... proves the work tu be of 
verj considernble antiquity, from its having been composed in the Daksino, 
while Buddhism prevailed there as the religion of the stale." This (luotnliuii 

i* u follows:— f^i^ijy I HI •I'^^K*^*^ <<<'Hv«T rH^MHs^^H 

the aid of Mr. Tumour's version, I trnnslnie as I'ullows : — "The cele- 
brated teacher Anando, who was a rallying point like a standard to 
Tambapanni (CcjtIod), had a disciple called Dipankaro. The latter, who 
bad obtained renown in the land of Dumila, and was the superintendent of 
two religious bouses, calle<l fiiiludichcha, &c., illustrated the religion of 
Buddha. lie was the devotee who bore the appellation of Buddhapijo, and 
composed this perfect Rupaviddhi." 


Lciur. I. 

into this sacred dialect till the beginning of the fiilh century, 
A.D.^', the knowledge of it appears to have been only very 
slowly diffused towards the south. The grammar just referred 
to appears to be more ancient than that translation. A 
more accurate conclusion regarding this portion of the his- 
tory of the languages of India, will perhaps result from n 
complete investigation of the writings of the Southern Bud- 

These remarks of Lassen scarcely afford sufficient grounds for 
denying that the Pali was introduced into Ceylon from Magadha. 
The peculiarities which are enumerated by Vararuchi as the 
characteristics of the Mogadhi, as it existed in his day, such as 
the substitution of i (If) for sh (If), and 8 (^), y (^) for j (^), 
sk (^) for I'sft (^), I for r, are, after all, of no great conse- 
quence, and woxild perhaps be regarded by learned persons 
even in Magadha itself, rather as vulgar provincialisms, than 
essential characteristics of their language. If so, such varieties 
would naturally be discarded by educated men acquainted with 
Sanskrit, when they came to form for themselves a Uteraiy 

The early Buddhist teachers appear to have been in the 
habit of travelling over the whole of the central parts of 
Northern India, and must have been acquainted with the lan- 
guages of ita difibreut provinces. When, therefore, they set them- j 
selves to compose works which were intended for circulation in I 
all these different regions, they would naturally adopt the mo8t 
correct and approved forms of speech which were current any- 
where within those limits. The case ia quite different in regard 
to the dramatic compositions of India, which would preeerve 
the most salient points of every provincial patois, as works of this 


" This statement of Lassen disagrees with the account given hj Mr. 
Tumour, on native authority (quoted above, pp. 69, 70.) that t}ie Pit^kattaTA 
bad been handed down in Pali from the first. See also the Journal of the 
Asiatic Societj of Bengal for 1837, pp. 503. IT. 



lerive a considerable part of their attraction from de- 
picting, or even exaggerating local peculiarities. 

I find it also diflRcult to concur in Lasflen's opinion aa to the 
period at which the Pali, or Magadhi, was introduced into Ceylon. 
Mahendra and his followers, who were no doubt numerous, must 
necessarily have carried with them the language of their native 
country ; and not only so, but must have been the bearers of 
numerous works written in that language. For it is not easy to 
receive literally the account given by the Ceylonese writers 
(which, as we have seen, p. 69, Mr, Tumour also rejects,) of 
the time at which their religious works were first committed to 
writing, or to suppose that the foreign propagatora of BuddhLmi, 
who would at first be ignorant of Cingalese, should, at the period 
of their arrival, have had no records in their own language of 
the new religion which they were introducing, or that these 
records should not have been safely handed down to their suc- 

M. Eugene Bumouf, in the course of a comparison which he is 
instituting between a paragraph extracted from a Pali book, the 
Digha Nikaya, and a parallel passage from a Nepalese Sanskrit 
work, makes the following observation on the language in which 
the former is composed, from which it will be observed, that he 
does not controvert the derivation of the Pali language from the 
dialect of Magadha : — 

" It is quite possible that these two versions may have been 
nearly contemporaneous in India, and have been current there 
from the earliest period of Buddhism, before the events occurred 
which transported them to Ceylon. The Pali version would be 
popular among the inferior castes and the bulk of the people of 
Magadha and Oude, while the Sanskrit version waa used by the 
Brahmans. Still, we should not be justified in supposing that 
we possessed in the Pali text the authentic version of this pas- 
sage in ita true Magadhi form, since a comparison of the 
Indian inscriptions of AAoka, and of the Pali of Ceylon, reveals 
to us certain dififereuces between the forms of these two dialects. 



TCBAT. f. 

Still, while we allow for the d^ree of artificial regularity which 
tlie cultivation of the Pali in Ceylon may have introduced, we 
must hold that the Pali version of this passage approaches very 
closely to the form which it must have had in Magadhi." — (Lotus 
de la Bonne Loi. App. p. 862,). 

Professor Weber, (in the course of a detailed notice of the 
Lotus de la Bonne Loi in his Indische Studien, iii. 176, ff.) re- 
marks as follows on this passage : — " This last explanation 
[that the Pali was elaborated in Ceylon] does not appear to me 
satisfactory, because a language carried by a few persons along 
with them into a foreign country ordinarily retains ita ancient 
character unchanged. It ia further very questionable whether 
the cultivation {u e. the grammatiaxl culture?) of the Pali 
commenced in Ceylon, and probability speaks rather in favour 
of the supposition that the grammar of the language wa« 
fixetl In the coimtry which was its home." Weber proceed* 
to observe, that the Cingalese tradition ascribes the origin 
of their grammar to India; and thinks it may be doubtiiil 
whether Puli was used at all in Ceylon before the arrival 
there of Buddhaghosa in 420 a.d. For though a translation of 
the Sutras is said to have been made into the Cingalese eiity 
years earlier, (which seems to prove that the Pali was understood 
all along), yet it is improbalile, he conceives, that, if it had been 
earnestly studied before Buddliaghosa, the translation of the 
work called Atthakatlia would have been so long deferred. 
At any rate, he thinks the arrival of this teacher appears to have 
given a new impulse to the stmly of Pali, as is attested by the 
compositiou of the Mahavansa in that language, fifty yean 
later. It is clear, however, that Weber maintains the essential 
identity of Pali with the verUiicular dialect of Magadha, in the 
sixth century b. c, as he explains the more aichaic character of 
the language of the Pali books, the Atthakatha and Tripitaka,u 
compared with the language of the Indian inscriptions of Asoka, 
by supposing that (while the popular dialect had undergone 
great alterations in the 300 years which intervened between 


Buddha's death and the date of the inscriptions) the followers of 
Buddha may have made it a rule to retiiin, as far as possible, 
the dialect in which Buddha himself spoke, as the language of 
all the discourscH whicli actually emanated from him, or were 
ascribed to him, as well as of all the narratives of which he 
formed the subject. 

I quote two other authorities on the subject of the early in- 
troduction of Pali into Ceylon. The first is Professor Spiegel, 
who remarks as follows, in the Preface to the Kammavakya (a 
short Buddhist work edited by him, and translated into Latin): 
— "It appears reasonable to believe that the Pali was intrwluced 
by the Buddhists into Ceylon, and carried thence into Trans- 
Jpmgetic India. An extensive intercourse existed between the 
continent of India and Ceylon from the earliest period, and the 
mention of this island in the liamayana is well known. Six 
Brahmanical kings are enumerated in the Mahavansa, who, as 
they lived before the age of A^oka, must no doubt have employed 
another language. That this was the case is proved by the mul- 
titude of words which have been transferred from Sanskrit., not 
from Pali, into the Cingalese langiiage, and which appear to 
have been introduced in consequence of that previous inter- 
course to which reference has been made. Thus we find in 
Cingalese, hirna, not htnna, ear, vaira, not twa, enmity, the 
use of the visarga, which has nearly disappeared from Pali, as 
well as the vowels fi, ri, Ifi, IfV Spiegel proceeds: — "We 
find, from the Cingalese books, that the Buddhists arrived in 
Ceylon, bringing with them the Pali language, in the time of 
Devanampiyatissa, the contemporary of Aioka, who reigned 
from 260 — 219 b.o. It is probable that the Pali was called 
Magadhi in consequence of the mission of Asoka's son IMahendra 
to introduce Buddhism into Ceylon. In fiict, a compaiison of 
the Pali with the language of the inscriptions which have de- 
Bcended to our own time, leaves no doubt that the two forms 
of speech are most closely connected. Both are but comparatively 
little removed from the Sanskrit, since in neither of them is 




elision of letters practised^ nor, witli few exceptions, are aspixiktol 
letters commuted into h, an in tlie Prakrit." 

The other authority I shall ijuote, ia Professor Beufey, wbo 
thus writ«a in hia article on India, (in Ersch and Gruber's Ger- 
man Encyclopoedia, p, 194.) 

" The plaice exterior to India, where Buddhism became first 
established as a state religion (about 240 years before Christ) 
under the special auspices of A^oka, Emperor of India, waa 
Ceylon. It is there/ore to be assuvied that at th/xt period aU 
u'hicli was of importance on Vie subject of Buddfti*niy waa 
brought to Ceylon in the form in xvhidi it then existed. 
Besides, so close a connection existed between Ceylon and the 
head quarters of the Indian empire, viz., the regions lying on 
the Bay of Bengal (Bengal itself and the adjoining province«)y 
that the Ceylonese took at least a passive share in the develop- 
ments of Buddhism. Hence their books appear to me to be 
authorities of the greatest consequence. It is further to be 
observed that these works are composed in Pali, which is tlie 
sacred language of the Buddhists in Ceylon, and in the countries 
converted to Buddhism by the Ceylonese, and which was tbft 
predominating popular dialect in central India." 

I quote another poissage, to a similar effect, from p. 250 of 
the same work ; and although there, at the close, the author 
speaks doubtfully of the derivation of Pali from the province 
of ALigadha, and of the introduction of Buddhism into Ceylon 
from the shores of the Bay of Bengal, he is not to be under- 
stood as throw^ing any uncertainty on the connection of Pali or 
of Buddhism with Northern India in general. 

He characterises the Pali as " the sacred language of the 
Buddhist writings found in Ceylon and Transgaugetic India, 
.... which is shown both by internal and external indi- 
cations to have been the vernacular dialect of central India, 
and which was diffused (dong with the Buddhist religi<^n in the 
coiDitnes above luimed, where it souu nc«|uired the same sacred- 
ness in the eyes of the Buddliists, which Sanskrit possessed, and 





still possesses, for the Brahmans. This language," he continues, 
" (though flistinct proof cannot yet be adduced of the assertion), 
is one of the very oldest of the Indian vernaculars, and was 
already in popular use at the period of the rise of Buddhism. It 
was probably the dialect of a considerable, I mean the western, 
portion of Bengal. It was from this point, from Banga or 
Kaliuga, that according to the Ceylonese accovmt, Buddhism was 
introduced into Ceylon : and yet again this conjecture becomes 
uncertain, owing to the fact that the language of ISIagadha, 
which was spoken only a little to the north of the Bay of 
Bengal, and which (as Asoka's insciiption in Cuttack seerufl to 
have been composed in it) appears also to have extended 
towards the south, varies essentially from the Pali in several 
particulars." Again in p. 246, Benfey speaks of " the Pali, as 
varying in mauy particuhirs from the language of Magadha, 
and approximating to the principal Prakrit or Maharashtri, 

But it matters little in what particular province we suppose 
the Pali to have originated, whether in Magadha, or in some 
countr}' further to the westward : as the fact remains in any case 
indubitable that it represents one of the oldest Prakritic dialects 
of northern India. 

The Buddhist writers assert, as we have already seen, that the 
Pali is not derived from the Sanskrit ; but that on the contrary 
it is the primitive language from which all others are descended. 
The«e BudilhiBt grammarians were no doubt led away by their 
prejudice in favour of the dialect which they or their prede- 
oeasors had adopted as the depositary of their sacred literature ; 
and by a prejudice against the Sanskrit, which was venerated by 
their rivals, the Brahmans. Even Mr. Clough says, (Pali Gram- 
mar, Advertisemeot, p. iii.) without determining the question, "it 
has long been a contested point whether the Pali or Sanskrit be 
the more ancient language of India ;" and contents himself with 
the remark that, " it is cerUiin that Pali was the popular dialect 
of the native country of Buddho, namely, Magadha, before the 




powerful sect, founded by him, was expelled from the continea! 
of India, an eveat prior to the Christian era," 

The real relations of the two languages, the Pali and thi 
Sanskrit, could not, however, escape the notice of any person 
■who had mastered the true principles of pliilology ; and are 
accordingly enunciated with distinctness, and iu a masterly way, 
in the following passage, by MIL Burnuuf and Lassen (Ea sai 
Bur le Pali, pp. 138. fif.) ^^H 

" The Pali is derived from the Sanskrit, according to cerK^^ 
rules, for the most part euphonic, which do not allow the deri- 
vative language to admit certain sounds and combinatiooa of 
ooneonants, common in the parent t<tngue. These modificatioiis 
apply equally to the substantive portions of the words and to 
their terminations and inflections. It hence results that there is 
no grammatical form to be found in Pali of which the origin 
may not be discovered iu Sanskrit ; and that there is no occa^on 
to call iu the influence of auy foreign idiom to explain the 
modificationfs to which t!ie Pali has subjected the Sanskrit. 

" Wlien the Pali, as a derivative from Sanskrit, is oomparad 
with other dialects having the same origin, it is found to approadi 
far more closely than any of those others to that common source. 
It stands, so to speak, on the first step of the ladder of departure 
from Sanskrit, and is the first of the series of dialects which 
break up that rich and fertile language. But it appears that the 
Pali, which contained in itself the germs of alteration already 
greatly developed, was arrested in its progress all at once, and 
fixed in the condition in which we now find it, i. e. in a state of 
almost immediate connection with the language from which it 
proceeded. In fact, the greater part of the words which form 
the basis of the one, are found without modification in th» 
other; those which are modified can all be traced to their 
Sanskrit root; in short, no words of foreign origin are to be 
found in Pali." 

Again : — 

" We shall not enter into new details regarding the manner in 




which the Pali has been derived from the Sanskrit. The laws 
which have guided the formation of that language are the same 
which we find at work in other idioms in different ages and 
countries; these laws are general, because they are necessary. 
Whether we compare the languages which are derived from Latin 
with the Latin itself, or the Later Teutonic dialects with the 
ancient languages of the same stock, or the modem with the 
ancient Greek, or the numerous popular dialects of India with 
the Sanskrit, we shall see the same principles developed, the 
same laws applied. The organic inflections of the parent 
languages are seen to exist in part, but in a state of evident 
alteration. More commonly they \d]l be foimd to have dis- 
appeared, and to have been replaced, the case-terminations 
by particles, and the tenses by auxiliary verbs. The processes 
vary in difiTerent languages, but the principle is the same ; it is 
always analytic, whether the reason of this be that a synthetic 
language happens all at once to become the speech of barbarians 
who do not understand its structure, and therefore suppress its 
inflections, and replace them by other signs ; or whether it be 
that when abandoned to its natural course, and as a necessity of 
ita cultivation, it tends to decompose and to subdivide the 
representative signs of ideas and relations, just as it unceasingly 
decomposes and subdivides the ideas and the relations them- 
Belvea. The Pali appears to have undergone this last sort of 
alteratioD ; it is Sanskrit, not such as it would be spoken by a 
strange population, to whom it would be new ; but pure Sanskrit, 
becoming altered and modified in proportion as it becomes 
popular. In this manner it still preserves its declension, instead 
of replacing it by particles, as the modern dialects of India do. 
One form only, the ablative in to might pass for the commence- 
ment of the analytic declension ; but it is already found in the 
parent language. A great number of Pali forms nught be cited 
to prove that the modifications, which it has made in the Sans- 
krit, are of the same kind as those which the Italian, among 





other tongues, has made in the Latin. Thus the assimilatioa of] 
consonants, which in Italian makes letto from lectus, aud scritto 
for scnptiis, is one of the principles of PalL" 

The Pali, in the precise form in which we find it in the I 
Ceylonese books, cannot have been a vernacular language. Itj 
exhibits a variety of refinements which could not have been 
employed in common speech ; but must have been confined to 
the language of composition, or introduced after the Pali had i 
ceased to be the spoken tongue of the followers of Buddha, and 
had become couseerated to the service of religion and literatiuie: 
just as the grammar of the Sanskrit itself became regulated by I 
more fixed and rigid rules, after it had been removed from 
the deteriorating influences of vernacular use. Such peca«| 
liorities are the use of interpolated letters to obviate the 
inharmonious sounds which would arise from the coUisiou of 
vowels. No less than nine letters, y, v, m, d, n, t, t, 1, and g, 
are employed for this purpose, as is shown in the following 
examples, viz ; — 

+ iinossa becomes 
+ angikam „ 
+ essati „ 

+ attham „ 

-f- ajali „ 

1. y — na 

2. t> — Ji 

3. m — Inhu 

4. (/ — Btta 

5. n — ito 

6. / — tasmn + iha 

7. r — 8abblii+ c?a 

8. I — cha + abhinna 

9. g — puLtba-f-eva 










This peculiarity of attention to euphony is common to d 
Pali with the Sanskrit ; and though the meajis they use art 
for the most part different, yet in neither case could the refine- 
ments employed in writing have been practiced in the language 
of ordinary life. The Pah has other characteristics (borrowed 
from the Sanskrit) which could scarcely have been common in 
the vernacular dialects of Northern India, supposed to have been 

CioDgh's Pali Grammar, p. 11. 


contemporary with it ; such as the use of desiderative, and 
nominal verbs ; like jighachchhati, he wishes to eat ; pabbatdyati, 
he resembles a mountain ; puttiyati, he treats like a son J' 

Fauflboli observes in his introduction to the Dhammapada 
(p. vi.) that the antiquity of that work is proved by the character 
of iie language, which approaches closely to the Sanskrit, even 
in some of its oldest forms, and differs widely from the diction 
of the prose Siitras, and of the commentary of fiuddhoghosa. 
Thus we find in the Dhammapada such forms as these, viz., the 
nominative of the present participle in avi, as ganayam, rodam 
(instead of ganayanto, &c.) ; the third person plural of the 
present middle in dit as sochare, upapnjjare; and the dative 
form of the infinitive, as netave, pafidtave, which is usually 
found only in the Vedas, &c. It is clear from this that the 
Pali appears in various phases of greater or less antiquity. 

Notwithstanding the introduction of various refinements into 
the Pali, after it became the sacred language of the Buddhist 
religion, there can be no doubt, as Burnouf remarks, (Lotus, 
App. 862.) that it substantially represents to us the language 
which was in vernacular use in Behar, and in all the central 
parts of Northern India, at the era when Buddhism was first 
introduced, i. e. in the third, fourth, and fifth centuries b.c. 
Such being the case, we should natiu-ally expect to find that it 
bears a strong resemblance to the Prakrit dialects ; which, as we 
have already eeea (in the preceding section) were spoken in 
the same provinces in the first centuries of the Christian era. 
That such was actually the case, is put lieyoud a doubt by a 
comparifion of these dialects with the Pali. I shall procee<l to 

'* Some desidcr&tive verbs and nouni miut, however, have been in ordi- 
narj lue in the Frakits ; m we find in the modern vemoculara some wordi 
which hare their origin in desidera(ive«. Thus the Ilinili iAuAA, hunger, must 
come from bubhukkfia, a Prakrit corruption of bubhukihu. The Hindi pij/as, 
thirst, too, is probably derived from pipdul, though it may also have been 
compounded otpi + ala, a desire to drink. 

a 2 

84 BISTORT AND BEIATIOira OF [chat.i. 

prove, by some oomparatiTe lista of nouns, pfronouns and yeAm, 
first, that an extensive class of Sansikxit words undengoes pre- 
cisely the same modifications in the Pali as in the Prakrit ; and 
secondly, that in some respects the modification of Sanskrit 
words and forms of inflection had not proceeded bo fiu: in Pali 
as it afterwards did in Prakrit. From this comparison it will 
result that the Pali stands nearer to the Sanskrit, and representB 
a more ancient phase of the vemacidar speech of Korthetm India 
than is exhibited in the Prakrit. 


[cair. L 



i ^ 





t^ ^ 


a s 

^ a fe 




8 t 






e r^EI^ 





1— 1 





































































«l -s 









































































I I 

J :S 




rfsr.rrli^l j 

ffli:|||.ll.|f|| ! 



• • 



• • 



• .J 


:S • 


* u 


J • 

• ^ 


f> . 











B S.J 5 
P O M 








d a 
3 3 

^ -a 



*^ "73 t? 

9 -^ 


=i 6 



.2 a (3 .2 

«• i 









J 4 

f ^llttlll 


' 4, >« 



^' ... 

k/^ ;o ;o 

o5 (J. B^ 2 <^ 2 -i 
CO e^ s '^ 


a) jO 

a d 

Z S S 4 o 6 

S S S C S S 

-S ,3 b <" -^ 


Q Q O O q 









CO <o 

C£> «0 C4 


4^ «« «^ 

S S o 



[cair. t 


%2 ua 

^ a a 







sf ^ S^ 

a' I? IT 

o c 

e * • 

r -0> 
CO « « 


- JS 




o a 


1 tm 


" c. ;2, «s 




. S3 


00 3 

2^. a 



8 S 

i-H eo 

A a 

.a Q 
^ 2 














^ ja 



[CBAP. I. 






3 "S 
^ -6 




■Ji S 

? i ? 

i 1 1 i 

1 1 I I I J 8^ .^ 




00 CO 
«5 ^ 

. r 


. bo 












-= •-- 








3 P C § ^ 















e .J 





u ; 














N '^ 









9 B 
















a 2 






M '&) *8 


2 J Js 






^ ^ <& ¥ ^ 'P 'S 


& C^ <E 
5' 5' »' 

0» eo 









~ ..a '^ 

na "aS -« S 

p< n 

• "O a u5 
H H § w 

























I" 5 2| 








« a o s 
^ SCO 






INDiZ 105 ■ 

The followi 

ng is a comparative scheme 

of the declension 1 


to the Sanskrit one in d, in which it will he seen H 

that the Pali is somewhat nearer 

than the Prakrit to the Sanski'it | 

forms. (Clough, p. 19. ; Cowell, 

p. xxiv.) 



PiMral ^^^1 





1. Buddha 


1. Buddbii. 

Buddba. ^^H 

S. Buddhom. 


2. Buddhe. 

Buddha. ^^M 

S. BuddLena. 


3. Biiddbcbi. 

Buddbebi. ^H 

4. BuddhSja. 

Same as 6tb 

4. Budilbiiiiam 

Same as 6tb ^^H 




^■f. Buddbasnm. 


S. Buddbebi. 

BiiddbShinl^. ^^^U 

^H BudJbti. 


Buddbe bid. 

Biiddbasuiito. ^^^H 

^^^P Buddbnaihu. 



He. Baddliaasa. 


ti. Buddbixnam 

Buddhunam. ^^^H 

7. Buddbasinin. 

7. Buddbcau. 

Buddbc;«u. ^^^1 







The first personal pronoun in 

the two languages is as follows : H 

(Clough, p. 61 

. ; Cowell, p. xxviii.) In most 

cases the Pali is ^^M 

nearest to the Sanskrit : 









1. aham. 


]. majam. 




3. oiam. 


2, andiiikain. 


^^B mumun. 



ambe. ^^^H 

^^BS. mmyi. 


3. amhebhi. 






^^^B>*- mama. 





& S ambakam. 

ambiinnm. ^^^| 

^^H^ may bam. 




H amham. 



H m amain. 



1 5. mayo. 


6, amhebhi. 

amhahinto. ^^^| 



amhasunto. ^^^| 

1 7. oiayL 


7. ambesU. 

amhwu. ^^^| 


fcnAP. t. 

In the Recond personal pronoun, the Prakrit has the fomu 
tujjhe, tuijhehin, tujjliaaam, tujjjhesu, as well as tunihe, tiunhe- 
4un, &c. The first named forms are not given in Clougb's 
rammar, as employed in Pali. 

The Pali verb seems to be far more complete tian the Prakrit 
The fallowing are some of its priucipal tenses, as compared with 
those of the latter: (Clough, p. 100. fif. ; Cowell, p. xxix.) 






or mUtc mood. 


or mldilU mood. 






1. pachorai. 



(wan ting) 


2. pachasi. 



2, pai'hase. 


3. pnchadi. 



3. paebade. 






1. pachiimo. 
pacbimo, &c. 





2. pacbaba. 





3. p.tcbanU. 




The Pali has also, like the Sanskrit, a potential mood, 
three past teuse-s, which iu the partismai-pada or active mood, are 
as follows : — 

II. Koduplicited perfect. 
*<"*■ Pi^. 

papacba. 1. pnpacliimba. 

papacbc. 2. papachitlba. 

papacba. 3. papnchu. 

IV. Third preterite. 
Smf. Plur. 

apachim. ]. apacbimhi. 

apacho. 2. apachittha. 

apachi. 3. apacbum. 


In Prakrit, on the other hand, few traces appear to remain of 
any past tenses at all. Mr. Cowell says, p. xxix., " The only 
tenses of the active voice whicli rem.oin, seem to be the present, 
_the second futiue, and the imperative." In the 23rd, 2-lth, and 

I. Potential. 
Sing. Plur. 

1. pacheyyanu. 1. pacbcyyamn. 

2. pachcyyasi. 2. pachoyyuttlm. 

3. pache. 3. pacheyum. 


III. Imperfect. 
aing. PItr. _ 
1. apacha. 1. apacbamha. 

2 apacho. 2. apachattha. 

3. apacha. 3. apacbii. 




25th aphorisms of the VTIth Chapter, and in the 19th aphorism 
of the Vfllth Chapter of Vararuchi, however, (Cowell, pp. 162, 
1 63) mention is made of a past tense, of which the instances, 
huvia, hoh ia, usl, ' he wae,* haeia, ' he laughed,' kdhia, * he did,' are 
l^trea. Few inatances of the past tense in Prakrit, however, seem 
to occur in the dramas; but it is inconceivable that in the 
Prakrit dialects which were currently sjwken in the long interval 
between the disuse of the Pali and the rise of the modern verna- 
culars (in both of which we find past tenses), there should have 
been no grammatical forms in daily use for expressing past time. 
It is not, however, necessary to pursue this subject further : as 
the details and explanations which I have already furuislied, are 
amply sufficient to show the place which the Pali and the 
Prakrit dialects respectively occupied in the history of North- 
Indian speech. 

[Professor Muller considers the data— derived froai Buddhist sources— 
on which the death of Buddha is placed in 543 b.c, and on which the occur* 
rence of any Buddhist sjnnds before the one in Aioka'a time, is asserted, to 
be fictitious and unsatisfactory. Though he does not try to hring down 
Buddha's death bi.-iow 477 D.c, he regards all the Buddhi«t dates before 
Chandragupta as merely hypothetical. See his " Ancient Sanskrit Litera- 
ture," received while this Section was in the press, pp. 260 — 300.] 

Sbct. TI. — The DialecU of the Roch amd Pillar Iiucriptioni of Atoka. 

Oar knowledge of the vernacular languages of India in the 
centuries immediately preceding the Christian era iB not, how- 
ever, exclusively derived from the Pali books of Ceylon. Certain 
inscriptions, dating from the third century B.C., containing edicts 
of king Priyadarsi or Asoka**, (whose name has been already 

** Professor Wilson thinks it extremely uncertain whether Fiyadasi can be 
identified with Ak)ka, and inclines to the conclusion that the date of the 
riptions is some period subsequent to 205 b.c. (Joum. Koy. As. Soc, 
xiL pp. 243— 2£1; vol. xvi. p. 357). Professor Muller, in his "Ancient 
Sanskrit Literature," p. 320, speaks incidentally, but without any hesitation, 
uf the inscriptions as being those of A^ka, and as dating from the 3rd 




inentioDed above, p. 68.), and written in a ooiTUpt Sansknlj 
appareutly tlie vernacular speech of tliat period, are still 
engraved ou pillars iuul rocks in different parts of India. 

I borrow the following particulars regarding them from tha 
summary given by Lassen (Ind. Alt. ii, 215. ff.).** The inscrip- 
tions are engraved partly upon pillars, partly on rocks. The pilknj 
are at Delhi, Allahabad, Mathiah, and Radhia. The inscri}*- 
tions on these four pillars are partly uniform, wliile those 
Delhi and Allahaliad have additions peculiar to tbemaelre 
The rock iii«criptiou.>< are, Istly, those at Gimar in Giuent,' 
divided into fourteen compartmeuta ; 2ndly, those at Db&aG 
in Orissa, which for the most part agree in purport with the 
at Girnar, though the dialect is different; and 3rdly, those i 
Kapurdi Giri, near Peshawar, which coincide in purport, thou 
they often differ in expression, and in their greater or lessdiBfW* 
ness, from the Gimar inscriptions. Besides these, Asoka appears to 
have caused other similar edicts to be promulgated in tlte mB^| 
way. AccordiQgly another inscription ha-s been discovered *l^ 
Bhabra, not far from Jeypur, which contains a fragment of an 
address to the Buddhist s3'Dod in Magadha. 

These inscriptions were mostly discovered altout twenty 
ago, and the great merit of having first (in 1837 and 1839^ 
deoyphered and translates! by far the larger portion of them 
belongs to the late Mr. James Priusep. His translatioxM 
subsequently revised by Prof. H. H. Wilson, in an article in 
Journal of the Koyal Asiatic Society for 1849 (vol. xii. jwrt 
pp. 153 — 251): and a portion of them were a third tiin« 
&\amined by M. Bumouf in the Appendix to his translatioo «i 
the Lotus de la Bonne Loi, pp. 652 — 781.*« Prof. Wilson btf 

See also the other aulhoritiea cited in the text, a Gttlt 

of an 1 


St ifl 

century b, c. 
further on. 

** See alao IVinsep's Indian Antiqaities, bj E. Thomas, i. 23S, ii. 11. 

'^ In an obituary notice (probably contributed by Profeaaor Wilson) 
M. Burnouf, in the Annual Report of the Royal Asiatic Society for ltii$ 
p. xiii. (published in Part I. rol. xt. of ibe Soi-icly'g Journal), tlie foU< 
rcourks are made on this dissertation : — *" Bringing to the inquiry a 



concluded hia notice of the sulyect in a fiirtlier pjiper on the 
Bhahra inscription, in the Jouru. Royal As. Soc, vol. xvi,, part 
ii. pp. 357 — 367. The importance of these inscriptions, aa 
throwing light on the langiiages of India in the third century 
B.C., is also expressly recognised by Prof. La.ssen (Ind. Alter- 
thumsk., vol. ii.), in passages which will bo (juoted below ; by 
Weber in his review of the Lotus de la Bonne Loi (Ind. Stud. 
iii. 166 — 173.), in the Preface to his MalavikJi and Agnimi- 
tn, p. xxxii., and in his Indische Literaturgeschichte, p. 170; 
and by Benfey, in his Article Indien, in Erach and Gniber'e En- 
cyclopaedia, pp. 194 and 250. 

To give the reader an idea of the nature of these edicts as well 
as of the l.inguage in which they are composed, I shall quote 
the eleventh, which is short and tolerably clear, according to the 
Girnar version, together with the translation of M. Burnouf. 
(Lotus, App. X. p. 736., Wilson, p. 212.) : 

^T^ mfk^ ^*«<i*i %Rnmf|- wt Mw^fkmvi m 

VmM4*-^ ^1 ri7\ Tt *?^ 41*JHrl*f^ T^slMld^fft 

nrpf ^rrv tt^ mtii^i ^rttw wr^ tht ^trj f^T 

4i^l<iyV Trf?T WU\ ^ '^I^ ^3T >f^ ^ ^•«<H-IH 

ledge of Pali and of Duddhiam, the superiority of which hia predecessors 
would be the rir«t to acknowledge, and having the odvantflgeof ilieir prcviotia 
Bpeculalk)n», the taUicoI' vihich M. Burnouf, with liis ncvcr-failiiig candour, 
recognitea, we mar look upon hin researches os conclusive, and feel satisfied 
that they have eliminate<l from these remains of ant iquitj' all the information 
they are capable of affording." Pruf. Weber also in his review of the Lotus 
de la Bonne Loi, (in the Ind. Stud.) apeoks in highly laudatory terms of ihe 
aame diisertntion. 




cair. L 

** Piyadasi, king belonged by the gods, speaks thtiB : There is 
no gift equal to the gift of the law, or to the praise of the law, 
or to the distribution of the law, or to union in the law. This 
gift la thus exhibited : Good will to slaves and hired servante, 
and obedience to one's father and mother are good things: 
liberality to friends, acquaintances, and relations, Brahmans and 
Samanaa, is a good thing : respect for the life of creatiu-es is a 
good thing : this is what ought to be said by a father, by a son, 
by a brother, by a friend, by an acquaintance, by a relation, and 
even by simple neighbours : this is good ; this is to be done. 
He who acts thus is liououred in this world ; and for the world 
to come an infinite merit results from the gift of the law." 

From the age to which these inscriptions appear to belong, we 
might expect that their language, as it is not piu'e Sansknt, 
would coincide in a great degree with the Pali, which, as we have 
already seen, represents what we may suppose to have been the 
spoken language of northern India about the same peiiod. And 
Buch proves on comparison to be to a considerable degree the 
case. In proof of this point I shiill first proceed to <juote the 
general observations made by Professors Wilson, Lassen, and 
others, on the subject of the languages in which the inscriptions 
are composed ; and then supply a comparative table, by which 
an opinion may be formed of the degree in which they coincide 
with and diverge from, the PaLL*^ 

The following are the remarks made by Professor Wilson 
(Journal Royal Asiatic Society, vol. xiL pp. 236. fF.) on the lan- 
guage of the edicts : 

" The language itself is a kind of Pah', offering, for the greatest 
portion of the words, forma analogous to those which are modelled 

*'' I might have been Id a posiuon to treat this subject in a more satufic- 
lory manner llian I can now hope to do from my own cursory uivestigatioB^ 
Lad I buen able to coniiult the Fati Gramiuar, with Appendices oo the dialeoti 
of Dbauli and Girnar, formerly advertised for publicutioD, but never pub* 
lishcd, by Professor Spiegel. (See the cover of hia Aaecdota Palica, pub- 
lished at Leipzig, in 1642.) 


by the rules of the Pali gi-amnmr still in use. There ore, how- 
,,aver, tuany dififerences, some of which arise from a closer adher- 
«Dce to Sanskrit, others from possible local peculiarities, indi- 
cating a yet unsettled state of the language. It is observed by 
Mr. Prinsep, when speaking of the Lat inscriptions, • The lan- 
guage differs from everj'" existing written idiom, and is as it were 
intermediate between the Sanskrit and the Pali.' The nouns and 
particles in general follow the Pali gtructxu"e ; the verbs are more 
frequently nearer to the Sanskrit forms; but in neither, any 
more than in grammatical Pali, Is there any great dissimilai-ity 
from Sanskrit. It is curious that the Kapur di Giri inscription 
departs less from the Sanskrit than the others, retaining some 
compound consonants, as j>r in priya, instead of plya ; and having 
the representatives of the three sibilants of the Devanagari al- 
phabet, while the others, as in Pali, have but one sibilant^*; on 
the other hand, the Eapur di Giri loser iptiou omit« the vowels 
to a much greater extent, and rarely distinguisbefl between the 
long and short vowels, peculiarities perhaps not unconnected 
with tlie Semitic character of its alphabet. 

*' The exact determination of the differences and agreements 
of the inscriptions with Pali on the one hand, and Sanskrit on 
the other, would require a laborious analysis of the whole, and 
would be scarcely worth the pains, as the differences from either 
woidd, no doubt, prove to be comparatively few and unimportaait, 
and we may be content to consider the language as Pali, not yet 
perfected in its grammatical structure, and deviating in no im- 
portant respect from Sanskrit. 

** Pali is the language of the writings of the Buddhists of Ava, 


Weber also reauu-ks (Ind. Stud., iW^ 180) : "The greater parity of 
pronunciation majntttined in the popular dialect of the north-west in com- 
parifon with the east, is shown bj the inscription of Kapur di Giri, in which, 
Uflording to Wilson's remark (The Rock Inscriptions of Kapur di Giri, &c.), 
not onlj the three sibilants of the Sanskrit, but also a number of cotnpouDd 
consonants, containing an r (such as priya, latra, prati, yatra, putrs, savntra, 
krsroa, iu^ilahii, sramana, bramana, bhratu), and some others, such as tt, 
$tr, hate been preserved." 





8iam and Ceylon ; therefore it is concluded it was the langoage 
of the BuddhiBts of Upper India, when the inscriptions were 
engraved, and conseqnenllj they are of Buddhist origin. This, 
howerer, admits of question ; for although the Buddhist aatho- 
ritiea assert that Sakya Sinha and his successors taught in Pali, 
artd that a Pali grammar was compiled in his day ; jet> on the 
other hand, they af&rm, that the doctrines of Buddha fttm 
long taught omily only, and were not committed to writing till 
four centuries aft«T his death, or until b. c. 153, a date, no doabt, 
fflibsequent to that of the ingcriptions."** .... 

It ifl by no means established, therefore, that Pali was the 

cred language of the Buddhists at the period of the inscriptions, 
and its utte conHtitutea no conclusive proof of their BuddhiA 
origin.*' It seems more likely that it was adopted as being the 
spoken language of that part of India where Piyadasi resided^ 
and was selected for his edicts that they might be intelligible to 
the peoi>le." 

" We may, therefore, recognise it as au actually existing form 
of speech in some part of India, and might admit the testimonr 
of its origin given by the Buddhists themselves, by whom it IB 
always identified witli the language of Magadha or Behar, the 
scene of .Sakya Siuha's first teaching ; but that there are eevenl 
differences between it and the Magadhi, as laid down in Prakrit 
grammars, and as it occiu-s in Jain writings. It is, as Menn. 
Burnouf and Ijasscn remark, still nearer to iSanskrit, and maj 
have prevailed more to the north than Behar, or in the upper 
port of the Doal), and in the Punjab, being more analogiius to 
the tSaurascni dialect, the language of Mathura and Delhi, 
although not differiuj,' from the dialect of Behar to such an ex- 
tent as not to be iutelligilile to those to whom Sakya and 

•• Sec, however, the remarks in the precediDg aection, p, 70. 

•• Professor AVilson hrw since, however, from an examinntlon of the Bhabn 
inscription, Arrived at the eunvietion, thnt there is in it "enough sufficieullT 
indisputable to cstuhtJsh the fnct that Prijudarci, whoever he ina^ have been 
WM a follower of Buddha," (Joum. K. A. S., Vol. xv., p. 357.) 



successors addressed themselves. The language of the inscrip- 
tions, then, although necessarily that of their date, and probably 
that in which the first proptagators of Buddhism expounded 
their doctrines, seems to have been rather the epoken language 
of tlie people in Upper India, than a form of speech peculiar to 
a class of religionists, or a sacred language, and its use in the 
edicts of Piyadasi, although not incompatible with their Bud- 
dhist origin, cannot be accepted as a conclusive proof that they 
originated from any peculiar form of religioua belief." 

Some observations of Professor Ijassen regarding these dialects, 
and their relative antiquity as compared with the Pali, liave been 
already quoted in the last section (p. 72.) He remarks in another 
place (Ind. Alt. ii. 221, 222): "These inscriptions are of the 
greatest value for the history of the Indian languages, because 
they exhibit to ua in an authentic shape the mast ancient forms 
assumed by the popular dialects, and furnish us with a secure 
basis for tlie comparative grammar of the great Sanskritic family 
of languages, which became so variously developed." 

" In these inscriptions we possess specimens of three verna^ 
cular dialects, one from tlie liorder country to the north-west, a 
second from western, and a third fi-om eastern Hindusthan. The 
inscriptions on the pillars of Delhi, Allahaba«l, Ac, differ only in 
particular forms from the Dhauli (Cuttiik) inscription, while they 
possess in the main the same character, and may be classed with 
the Alagadhi of the grammarians. As this dialect is used even 
<ak the Delhi column, which is situated beyond the bounds of 
Magadha, A^oka appears to have had a partiality for the verna- 
cular language of his principal province ; and from the pre- 
dominating employment of this particular derivative of the 
Sanskrit, we rnay perhaps explain the fact that, among the 
Gngalese, who receivetl the Buddhist religion from that 
country, their sacred language should have obtained this ap- 

At p. 486, again, lessen says : **It is only the rock inscriptions 
which can l>e admitted nJi authentic evidence of the local 


niarroBY akd relations op 

diiilectK, while the columnar inscriptions evenrwhere exhibit tliel 
flotjie dialect, which ooneeq-uentlj cannot hare been epoken ivj 
erery quarter where such pillars have been disoorered. This^ 
remark is espedally true of the Delhi column. ^Vben we con- 
sider tliat, between CabuL. Quzerat and Magadha (which latter J 
province was the native country of the dialect employed in tlMi| 
pillar inscriptions) a wide region intervenes, inhabited by dififerenM 
branches of the Sanskrit'-speaking race, we are driiren to tlta| 
conclusion that many other dialects must have been current 
there, of which we find no specimens in any of the ioscrip 

The following list of words, from the Delhi and All 
columns, and the Bhal»ra stone, borrowed from M. Bumoofs" 
Lotu« de la Bonne Loi (App. x-, pp. 665, 724, and 741), wiUj 
show the correctness of Lassen's remark, that the dialect of tha 
pillar inscriptions resembles the Magadhi of Dhauli, as exhibited] 
in the comparative list which I shall immediately atlduce. Tht! 
on thiise columns we have dhannnc., tUlnet aadui^ anugaJ^ 
kxUe, piye, kaydne and pdpe, for dkamniOf ddna/m, aatkoM, 
anuf/aho, koto, piyo, kaydnam and papaifn. ; Idjd, vdUchalatiy 
vUuilatain, cldla, Aliya, puUsa and ahhihdJe, for raja, iifm-J 
cliaremi, vlhdratarn, ehinu, Ariya, purisa and ahhiAdtvi 
iUitViasi, dJuvnxmasi and eangJum, for BudJMvJii, 
maniJii and eaUfjliamhi, 

The list of words, which I shall immediately adduce, 
rowed from the article of Prof. H- H. Wilettn, alx)ve alluded ' 
to, in the Xllth VoL of the Journ. Roy. As. Soc, and from 
the Appendix, No. X., to M. Buruouf s Lotus de la Bonne Loi^^ 
when compared with the Pali equivalents which I have add 
will suflSce to show the points in which the languages of 
iuHcriptions agree with the last-named dialect, aa well as the] 
respects in which they differ from one another. I must, how- 1 
ever, frankly state that I do not pretend to have made the 
iucriptions, or the character in which they are written^ thai 
object of particular study; and I, therefore, take it for granlelj 



thot the words have been correctly decyphered by the emioent 
scholars from whom I quote. 

In comparing the dialect of the inscriptions with other kindred 
fomts of language, presumed to be of about equal antiquity with 
them, which have come doMm to us in books, we should recollect that 
the latter may have been retouched from time to time, to render 
them more intelligible to the readers by whom they were studied 
in successive generations, whereas the inscriptions have descended 
to us unaltered, except by the defacing action which ages have 
ejcercised on the rocks on which they are engraved. On this 
srabject I quote the following judicious observations of Mr. 
Tumour, in the Joiunal of the Asiatic Society of Bengal for 
Dec, 1837, p. 1049:— 

*' When we consider that these msoriptions were recorded up- 
wards of two thousand years ago, and that the several columns 
on which they are engraven have been exposed to atmospheric 
influenoea for the whole of that period, apparently wholly 
neglected ; when we consider, also, that almost all the inflections 
of the language in which these inscriptions are composed, occur 
in the ultimate and penultimate syllables, and that these inflec- 
tions are chiefly formed by minute vowel ayrabols, or a small 
anu3u>ara dot ; and when we further find that the Ptlli ortho- 
graphy of that period, as shown by these inscriptions, was very 
imperfectly defined — using single for double, and promiscuously, 
aspirated and unaspirated consonants; and also, without discri- 
mination as to the class each belonged, the four descriptions of 
•n, — the surprise which every reasonable investigator of this sub- 
ject must feel will be occasioned rather by the extent of the 
agreement tlian of the disagreement between our respective 
readings of these ancient records." 

The following is the comparative list 1 proposed to adduce: — 

T i 

M M B M *" 

^ ^ ^ i ^ ^ 

rf a ^ ^ _ _ 
J J J J ^ ^ 
^ ^ ^ g ^ ^ ^ 




CD CO cn 



-« « 2J o* ^ 

OS O OJ — Oi 

llllllls I 




J e i 

^ « ? 

I -I 



Skct. VU.— The Dialect of Ike Bvddkut OdAdt, a»d iU reiatiom to the PaK: 
Summary of the runUs of (Am and &e p teeed u g Section. 

I DOW come to the last of the varietiea of corrupted Sanskrit to 
which I referred in p. 9, viz. the language which we find in the 
Gathas, or metrical portions occurring in such works as the Ia- 
lita Vistara, descriptive of the life and discourses of Gotama 
Buddha. An account of the peculiarities of this dialect, aa it 
is convenient to call it, has been given by Babu Rajendntlal 
Mitra, in the 6tb No. of the Journal of the Asiatic Societv of 
Bengal for 1854. Of the Lalita Vistara, from which the speci- 
mens given by this writer, and those which will be adduced by 
myself are drawn. Professor MiiUer remarks, that though " on 
account of its style and language" it "had been referred bv 
Oriental scholars to a much more modem period of Indian 
literatiu-e," it ** can now safely be ascribed to an ante-Chri.sUjin 
era, for it was translated from Sanskrit into Chinese, as one uf i 
the canonical books of Buddhism, as eaily as the year 76 A-d.** 

I proceed to give the substance of Babu Rajeudralal's di*- 
eertation in his own words, omitting only those portions which 
appear to be of the least importance ; making occasional 
abridgments ; and adding, in some places, to the number of the 
specimens he has given of the Gatha forms. 

" Of the dialects which have proceeded from the Sanskrit, the 
Pali and the Magadbi have hitherto been supposed to bear th« 
closest resemblance to their parent, but the discovery of tlje 
Sanskrit Buddhist Uterature of Nepal has brought to our know* 
ledge a new dialect bearing a still closer affinity to the classic 
language of the East, than either of the former. NepoJese 
chroniclers have named it Gatha (ballad)**, probably, from ib 

•* Buddliigtn and Butldhist Pilgrims, p. 24. 

•• [Tlio untiiiuity of certain compositions, called Gathat, ia proved by Um 
fact that the expression miuiigatha, tlie giithas or yerses of the Muni, or 
Muni*, occurs in the ancient inscription of Pijradasi at Bhabra. Bumouf 
App. X. to Lotus, pp. 724, 725, 729 ; Wilson, Jour. R. A. S., Vol. xri, 
pp. Z69, 363. 367. Biibu Kiijcndralal also refers to the Mohiiwanso, p. ^i, 
where > nicnlioncd, — J. M.] 


having been principally used by the scalds and bards of m^ 
di%val India. For nearly a aimilar reason the Balenese style 
the language of their poets, the Kdivi or poetical, and the lan- 
guage of the Vedas is called Chhandaa (metrical), whence by a 
well-known euphonic law, we have the Zend of the old Persians. 

•' M. Buruouf, the only European scholar who lias noticed the 
existence of this dialect, describes it to be 'a Larlj^^Lrous Sans- 
krit, in which the forms of all ages, Sanskrit, Pali, a:id Pra- 
krit appear to be confounded.''* It diflFers from the Sanskrit 
more in it« neglect of the grammatical rules of tiie latter than 
from any inherent peculiarity of its own. The niceties of the 
Sanskrit forms of declension and conjugation find but a very 
indifferent attention from the Gatha versifier ; he uses or rejects 
the usual case-affixes according to the exigencies of his metre 
with as much veneration for the rules of Panini as the We-st 
Indian Negro baa for those of Lindley Murray ; indeed, tlie best 
illustration that can be given of the relation which exists be- 
tween the Sanskrit, the Gatha and the Pali, would be extracts 
from the literature of the Negroes. 

•* The Gatha exists only in a versified form, and is to be met 
with in that class of Buddhist writings called tiieMahdvaijpulya 
or the • highly developed ' sutras. It occurs generally at the 
end and often in the middle, but never at the commencement of 
a chapter, and contains a poetical abstract of the subject de- 
scribed in the prose portion of the works. The latter is written 
in pure Sanskrit, and comprises a highly amplified version of 
the subject-matter, and often adverts to circumstances unnoticed 
in the former. 

" The Gatha is written in a variety of metres, from the facile 
octosyllabic amishtup, to the most comjiiic&ted Sdrdulavikndita. 
Its peculiarities are tliose of a language in a state of transition ; 
it professes to be Sanskrit, and yet does not conform to its 
rules. In it we find the old forms of the Sanskrit grammar gra- 
dually losing their expressive power, and prepositions and peri- 
*' rnistoirc du Buddhiame, p. 104. 





CBAT. t 

ptiraRtlc tnwfjrww aapplyh^ their plaeeSf and time^talknred 
Tcrb* and eoayngitioos jnxtapoaed to raiffar dmngs and nnetwth 
prmriiictaliaBa. At ooe pboe, ortiiognftby is Bacrificed for the 
aake of prosody and a word of a aitq^ riMct ajrllablo b inflated 
into one of three syliables, arfaile at another the latter yiddfl 
to the former and a moloseiu supplies the place of a pynhie 
or a tribrach. A spirit of economy pervades the whole, and 
iiyllaljles and words are retrenched and modified with an uo- 
•paring; hand. In the Lolita Vistara instances of these peca- 
Uaritiea occur in great profufiion, and tbey may be generally 
referred to (A) exigencies of metre, (B) provineialiaii^ and (C) 
erroni of iiyntox and prosody. 

<* A Of the changes which may be attributed to the exi> 
gencieA of metre, prolongation, contraction and elision of voweb, 
elixion of conHonants, and the segregation of compound ooo- 
Bonanto and lung vowels into their simple elements, appear to 
be the most frequent. We shall quote a few instances: 

ImU "Of the prolongation of vowels the following may be 
t/ikcn riH exanijjifH.'" 

irr ^ for ?I ■^; ^t ^ for ^ ^; 3IirT?fY lor M I «l I d t ; 0<**H 

for ^^pRHf. 

'du(\. "Of contractions of vowels, instances occur almost in 
rvcry sliikfL They are generally effected by the use of short 
for long v<hvi'Ih, and the substitution of i and u for e, ai, o, and 
au: for exam pie, 

irrfi? for tfTTi ; Vrf^ f»r VnTrfrf ; ^*1<<< for ^4{<4<i:t 

Wr^ f'T ^r«rr; ^rn? tWr l^fTJJT; Ijal^rti for ^3JT^?rf; ^^ 

for ipiT : ?TO ^^t ?T^ ; W^ f«r W^. 

3rd. '* £li>«ion8 of vowels and consonants are also very fre- 
qui'iit ; they are effected principally with a view to economy and 
euphony, Kinal ses are invariably elided. Take for instance: 


** Qdoted (Voin the «dit<OQ of tbc Lolita Vislara in the Bibliotheei 



»n^ for »wf%; VHH.\: for ^r^^:; ir^rH^ff^ for ^- 

for fir^^rC; Iffw^f'^f for TrfwWT'lf'fT ; '^TT for W^re; ; 

1?^ for T^rTT. 

4th. " Of the division of long vowels and compound con- 
sonants into their short and simple elementi?, the following are 
instances of constant occurrence : 
ilfrl^i for ?7^«n": or <lom*(^; f^f^fW for (^*5«i: ; f^I- 

^111 f«>r l«!jl«?) ; t1% for "^ ; rjfi.*} for Jft^ ; ^lf*>l I m* 

for ^*M* ; fwf^ir for ^ ; f^fr for ^ ; fjrf^ for t^ ; 

fi|f<«JI for f^nn- ; f^rft^ for f^ ; $f^ for ^T ; ^|^>TTT^ 

for prri?; Tj^iirfsr for q^lfil ; <H^{le^l for <H^44l: 

^f^ for ^Tl^. 

" This tendency to segregation of aspirated consonnntfi, forms 
a principal characteristic of raediteval and modern Indian pho- 
nology. The Pali and the Prakrit owe their origin entirely to 
this cause. The Hindi and the Marhatti indulge in it to a large 
extent, and the Bengali is not exempt from iu influence. 

" B. The provincialisms of the Gatha include n<^lect of 
gender, number, and case, abbreviations and omissions of de- 
clensions, corruption of pronouns, and new forms of conjuga- 

(!.) "Of the neglect of gender, number, and case, the follow- 
ing may be taken as examples : 
t^lj^f^n?^ for f^l5^f«T^WT^ (singular for plural); ^- 

%^ fiT ^^Tj"s||fi(T (lingular for phu-al); ^(4Jp|«|| for ^?r9- 

mfi^ (instrumental for ablative) ; ^^tfv^^T for WTtV^^TTr^^ 

(objective for ablative) ; ^3?^ ^^fT for '*^f ^^ (plural for 

dual); 3»r^<*MI4 f^r ^fg^^rm^ (locative for inatni- 



[ca*r. 1. 1 

mental) ; f^t^T^ for f^i^V^ ( newter for feminine) ; ^fTT'PT 
for ♦HUJTf'I (fiingiilar feminine for plural neuter); si^'^i: 
for "T^^lfW (masculine for neuter) ; "TO for JJl}^; ; ^Tff ^T^ 

for T^rrf'T <**5i(w- f 

(2.) " Under the head of abbreviations and omissions of dc- 
cleufliou, the most remarkable peculiarity appears to be the use 
of ^ in the room of all flettional affixes," a^ ^r1^ for ^TSHl ; 

^ry for ^ff^^ I ^ is also merely bvit for the inhereTit T- 

as ui two of the following cases ; VjUljfj^^ for ^PT^^T'T j 

fiM^I«rl for t%5^iT5fr ; 'iR'^lRrl for MR^lfr^r:. The next 

are instances iu wliiolt tiie case terminations are omitted;! 
I^V(%^ fur ^%^; ; f%i:qr4 ft>r fV^^WW: ; and snch j 
instances are of coutimial occurrence. 

(3.) " The following are the corruptions of pronouns that nw] 
frequently met v\ith in the Ijilita Vistara. They ap|)areDtljr 
lead the way to the formation of pronoims in the modern ver-j 
ITTH f^>r fm and ^rf t ; rP^ for T^Jf, (sic) t^, and rfW ; ^ 

for Tj;^: : ^ for m ; qrf^ for ^^ and ^3f. 

(4.) " The new forms of conjugation observable iu the Gathil 
ore attributahle exclusively to corrupt pronunciation ; they fol 
low no fixed rule, and are the result of that natural teiidencj 1 
abbreviation" which in the English originates "wont' 
"will not,'' and "shant" from "shall not." The following ue' 
a few examples; 

^f^fiT and ^^ for ^J^T^R ; Htf% f<Jr Vf^f^ ; ntf^ for Hsrf^; 
Htf^ for ^f^fs^ ; lilT^f% for f^ ; '%n^f^ for "^TCt^TI^ 
^SRwt or T^ f"r '^IT^; "^fi^T for ^itTE ; <^ for ^^^ 

" [I do not think that ibu cuusos here alleged afford a sufficient, 
a suitable explanation of ail (he changes under consideration. — J. M 



ajuhHS and IJUT fur 37TIT ; TT^jft for ^RfS ; VffrU for ^^f^- 

^rrf^-^-^-fTf-n: ^'^-^ ^:-"«j; vlfr^r^ for Mti^ ^^; 
iti4\ for f^T^: ; ir^T^ for 3?T!!rf^ ; i5f%<^, ^r^mr, 

l^fuifli], ftnJ ^Rir ft>r ^^ ; 'JL^RlttlfH for ^^V^fH '■> Ip^i ^or 
WC^T^l ; "^^njHT fur ^f^^^ ; HMf*<^ for J^m<4l^l^: ; 
«rfV^ for ffm ; ^f^ for ^^. 

" It may be reraarkud that the corruptions above quoteil are, 
in raanj instances, the precursors of forms adopted in other afB- 
liated dialects. In Sanskrit the third person singular of the 
verb to be w bhuvati, which in the Gutha cliaagea to bfiMl by 
the conversion of the i' into o and the eliniou of the a before 
and after it^ {bhrnitl in the plural and hhosi in the Becond per- 
son singular), and thence we have Ju>t{, hosi, and Jundi in the 
Magadhi. Sunitvd for indvd is the first step to the formation 
of afunid in Bengali, while ifunoki passes into s^mo with nothing 
but the elision of an inflexion. 

" C. In the collocation of words and phrases the Gatha strictly 
follows the rules of Sanskrit syntax, but in the formation of 
compound terms it admits of many licences highly offensive to 
the canons of Pauiui and Vopadeva. They seem, however, to 
be the consequence of haste and inattention, and are not re« 
ferrible to any dialectic peculiarity." 

There are, however, some other forms discoverable in the 
Gatha dialect, which have been either passed over, or but cursorily 
noticed by Babu Rajeudralal, and which yet present some points 
of remarkable interest. Thus the plural instrumental in ebhihf 
vhioh is BO general iu tlie Vedaa, is in use in the Gathas also, as 
io the instances </a/r/i/cM//t, sattvebhih, g^unebhifj , ainhdannebhi}}.^ 
ddrakehhih, chetakebliili, employed instead of the form, snkyaih^ 
satTtaiJIf, &c, which is alone current in rawlern Sanskrit. It is 
from this older form in ehhilj, that the Pali form of the same 
case in ebhi, or ehi, is derived, as in the word bud^lhebhi, or builr 




dhehi. (Clougb, Pali Grnra. p. 19.) Agdn, we find in the Gatluk* 
the case-termiiiatioas of the declension in a sulwtitutt^l, in 
case of words ending in consonants, for those proper to the latte 
form of declension. Thus, for jagatah, and joffati (the 
and loc. of Jagat), we have Jaga»ya and jage ; for vdmnd (instr. 
of ndman), we have namena ; for makdimdnam we have mo-a 
hdtviavi; for anantayaSaaain we have anantayaAam ; for" 
kamvanah (gen, of harwan) we have kaiinneya ; and for 
duhitaratti, acciosative of the word duJtitfi (ending in ri), 
have duhitdvif the accusative of feminine nouns ending in 
This change is one to which the Pali inclines (as in the for 
BrahmasacL, f\s one of the genitives of Brahman), and to whic 
a still more decided tendency is observahle in the PrakriL (S 
Cowell'H Prftkrit Graui. IntrrKl. p. xxiii. xxiv.) On the othei 
hand, we find also in the Gathas instances of the quite differeut 
change of e into i in the locative, as loki, gehi, udari, for th^ 
proper form loke, gelie, udare. The particle api {afso) is ooo 
tracted to pi, as in Prakrit; thus we have ahnmpi for ahai 
api. Again, we have the peculiar forms jihmiy jlhmaf fa 
yatJid; yathariva for yatkaiva (precisely as in Pali, Cloughli 
Gram. p. 11); »Ui for smrUi ; pcUfie for patheehxi, and uhtihaii 
for yashiidhdrahhu 

Many of the changes in the Gatha verbs are in part the 
which we find in Pali. Thits, for the correct Sanskrit fo 
chodayanti, tarpayishyanti, nlvarttayati and dhdfayatUiy 
have chodenH, tarpeshyati, nivarttetl, and dItareiUi, which, ii 
Pali, would be dioilenti, tappessati, nivatteti, and dhareuii 
Again, for avalavxbate we have olavibate, which would take th« 
same form in Pali. The mmlifications avachi for aii 
munchi for amunchat, gcichchhl for agachchhat, dkydyi kt 
adhydyat, correspond in some measure to such Pali forms 
akdsi for aJcdrahU, ahdsi for ahdrshit, addai for addt, nfi^i h 
abhut, atthdsi for asthdt, id)adhl for abailkJt, &c : and tnapin' 
8un for »iidpaydmaauh or asisnapan, is nearly the same as (lui| 
Pali form apackhisu, the third person plural of tlie third prrtf-I 



rite. The Gatlia fonns dariishyasi for dnikshyaei, and aunish- 
yaii for ^roshyaii, are closely similar to the Pali form vedissdmi 
for vetsydmi, bhuiijUsdmi for bhohkydini, and d^siasdnd for 
delshydmu The Gatha past indeclinable particles also, Buch 
as bhai'Uwd, raniitwd, hanitwd, labhitwdf etuvUwdy manitwd, 
vijlhitwd, iunitwd, for bfiutwd, mantwd, liatwd^ labclfiwd, atu- 
Iwd, mahud, vi + h!Hvd, ^ruttvd, are formed on the same prin- 
ciple as the similar Pali ones, 'pavisitwd, jdniiwd, bhunjitwd, 
(forpra + veahtwd, jridtwd, and bhuktu'd). Of the forma kaTityOt 
and ka/riydna (for kriiwd) the latter coincides in its termina- 
tion with such Pali forms as aidwdi^a and disvjdua (for stiitwd 
and drialtpvd). Again, we have the form htmpayantd (for 
kampayan), which coincides with the Pali and Prakrit. For 
tyaktu'd I find the word chhorayitwd, which does not seem to 
be much used in Sanskrit, though Wilson, in bis Dictionary, 
gives chhorana in the sense of " leaving." I quote the following 
additional anomalous forms, viz. pithitd for pUhtd, vienapi for 
tfyaamlpdyan, snapit for sndpayitwd, ksKlphisu for kshlpanti, 
bfiaviya for bhavet, pratishthihitwd for prati-ahthdya, dd^nii 
for ddayami, dlyatu for diyatdtiL, daHki for dadataJt, kurumi 
for karond, jana/nii for jdndmi, bhdai for bhdshate, vinetUi for 
vrnwhyati, jaiieshl or jaaaishi for janayishyati, adrUuh for 
a-ilrdktihuh, paayeta for drw/ate, adhye^lUa for adhyetum, 
ckintayd for chintayUwd, aniari for smfitam, atikrayneUiifii for 
atiki-cmdium. (In all these cases, I should observe, the Sanskrit 
equivalents are given according to the notes in the published 
edition of the Lalita Vistara). 

I proceed with the quotations from Babu Kajendralal's Essay. 

*♦ Of the origin of the Gatlia, nothing appears to be known 
for certain. M. Buruouf is iucllnecl to attribute it to ignorance ; 
he says : — 

' This fact (the difference of language of the different parts 
uf the Vaipulya Sutras) indicates in the clearest manner that 
there was another digest (of the Buddhist literattue prejiared, 
beaides tboae of the three convocations), and it agrees with the 

K t 



cn*r. I. 

development of the poetical pieces in which these impiiritii 
occur, in showing that those pieces do not proceed from the 
hand to which the simple Sutras owe their origin. There 
nothing in the boolcs characterised by this difference of Language, 
which throws the smallest light on its origin. Are we to look 
this as the \ise of a popular style which may have develo] 
itself subsequent to the preaching of Sakya, and which wotili 
thns he intermediate between the regular Sanskrit and the Pali,] 
■ — a dialect entirely derived, and manifestly posterior to tin 
Sanskrit ? or should we rather regard it as the crude oompositioi 
of writera to whom the Sanskrit was no longer familiar, and wh 
endeavoured to write in the learned language, which they i 
understood, with the freedom, which is imparted by the babitoal 
use of a popular but imperfectly determined dialect ? It wi 
be for history to decide which of these two solutions is correct 
to my mind the second appwurs to be the more probable one, hul 
direct evidence being wanting, we are reduced to the iuductioi 
furnished by the very few facta as yet known. Now, these fi 
;are not edl to be found in the Nepalese collection; it is im 
pensably necessjuy in order to understand the question in all its'' 
bearings to consult for an instant the Singalese collection and 
the traditions of the Buddhists of the South. \Miat we thence 
learn is, that the sacred texts are there written in Pali ; that ia 
to aay in a dialect derived immediately from the learned idiom 
of the Brahmans, and which differs very little from the dialect 
which is found on the most ancient Buddhist monuments in 
India. Is it in this dialect that the poetical portions of ibe 
great Siitras are composed ? By no meiUis ; the etyle of thi 
portions is an indescribable melange, in which incon-ect SMukrit 
bristles with forma of which some are entirely Pali, and otb<ff| 
popular in the most general sense of the term. There is »o 
geograpliical name to bestow upon a language of this kind ; b«rt 
it is at the same time intelligible how such a jargon may bsf* 
been produced in places where the Sanskrit was not stuified 
Bystemntically, and in the raiilst i)f populations which had uevtr 



8|)i)ken it, or had known only tbe dialects derived more or less 
remotely from the primitive source. I iacliue then to the belief 
that this part of the great Sutras niuet have been writtea oiit of 
ludia, or, to express myself more precisely, in ooimtries situated 
on the western side of the Indus, or in Caahmir, for example; 
countries where the learned language of Brahmanism and Bud- 
dhism would be cultivated with less success than in Central 
India. It appears to me almost impossible that the jargon of 
these poems could have been produced in an epoch when Bud- 
dhism flouriahed in Himlustlian. Then, in fact, tht priests had 
uo otliej" choice but between these two idioms; either the Sans- 
krit, /. e, the language which prevails in the compositions col- 
lected in Nepal, or tiie Pali, that is, the dialect which is foimd 
on the ancient Buddhist inscriptions of India, and which has 
been adopted by the Buddhists of Ceylon."^* 

*' This opinion," (continues Babu Rajendralal) " we venture 
to think, is founded on a mistaken estimate of Sanskrit style. 
Tlie poetry of the Gatha haa much artistic elegance which 
at once indicates that it is not the composition of men who 
were ignorant of the first principles of grammar. Its authors 
display a great deal of learning, and discuss the subtlest qaes- 
tions of logic and metaphysics with much tact and ability, 
and it is difficult to conceive that men who were perfectly 
familiar with the most intricate forms of Sanskrit logic, who 
bAve expressed the most abstruse metaphysical ideas in pre- 

•• rBwtoire du Buddhisme Indien, p. 105. [1 hare introduced a very few 
rerbsl aherations into this tnuMlation froin Btirnouf. Lnssen, Ind. Alt iJ., 
p. 9, concun in these views of Bumouf : — " The M^hitynna Sutras (of the 
BoddhUu) are coui[)oscd in a prose made up of a luixiure of irregular 
Sanskrit, of Pali, and of forms borrowed from the vulgar dialects ; and the 
narrative is repeated in verse. Such a nie<llejr of fitrms could only, a« it 
appears, have arisen in a country where the learned language was no longer 
maintained in its purity ; and, consequently, the writing* in question were 
probably eompooed in tbe countriea bordering on the Indus, and most likely 
io Kashmir, which ploys an important part in the later history of Buddbism." 
(See al»o pp. 491, 492, and p. 1153 of the same volume.)— J. M.] 

■ S 





dse and often in beautiful language, wiio composed with 
and elegance in Aiya^ Totaka, and other difficult measures,^ 
were unacquainted with the rudiments of the lauguage in vhicb 
they wrote, and even unable to conjugate the verb to he, m iH 
its forms. This diflBculty \s greatly enhanced, when we bear ini 
mind that the prose portion of the Vaipulya Sutraa ia written tal 
perfectly pure Sanskrit, and has no trace whatever of the prt»-J 
vincialisms and popidar forms so abundant in the poetry, 
these Sutras be the productions of men beyond the Indus im- 
perfectly acquaintetl with the Sanskrit, how happens one portionJ 
of them to be so perfect in every respect, while the other is my 
impure ? WHiat coiild have been the object of writing the sHue 
subject t>vice over in the same work, once in pure prose and then 
in incorrect poetry ? 

" It might be supposed — what is most likely the case — that 
the prose and the poetry are the productions of two different 
ages ; but the question would then ai-ise, how came they to hej 
associated together ? ^^1lat could have induced the authors of 
the prose portions to insert in their works, the incorrect produc-j 
tions of Trans-Indus origin ? Nothing but a sense of the tratfakl 
fulness and autlieuticity of thoae narratives, could have led to] 
their adoption. But how is it likely to be supposed that 
most authentic account of !§akya within three hundred 
after his death, was to be had only in countries hundreds of 
miles away from the place of his birth, and the field of bil| 
preachings ? The great Sutras are supposed to have been coi 
piled about the time of the third convocation (309 b. c), wl 
it is not at all likely that the sages of central India would have] 
gone to Cashmere in search of data, which could be 
gathered at their own tlireshold. 

"The more reasonable conjecture appears to be that the 
Gatha is the production of bards, who were contemporanee of 
immediate successors of Sakya, who recounted to the de 
congregations of the prophet of Magadha, the sayiugs 
doings of their great teacher, in popular and easy flowing ■ 



which in course of time came to be regarded as the most au- 
thentic source of all information connected with the foimder of 
Buddhism. The high estimation in which the ballads and im- 
provisations of bards are held in India and particularly iu the 
Buddhist writings, favours this supposition; and the circum- 
stance that the poetical portions are generally introduced in cor- 
roboration of the narrative of the prose, with the words : Taire- 
lUwiUchyate, ' ITiereof this may be said/ affords a strong 
presumptive evidence." 

[n a review of " Burnouf's Lotus de la Bonne Loi," Pro- 
feaaor Weber (in the Indische Studien, iiL pp. 139, 140), remarks 
as follows on the views expressed by Biu-nouf in the preceding 
paaaage in regard to the language of the Gathas : — 

•* The Ifwt reason (viz. that Sanskrit was cultivated with less 
liuccess iu Kashmir than in central India), is an incorrect one ; 
lince, on the contrary, it is precisely in the north-west of India 
that, tlie proper seat of Indian grammatical learning appears to 
^ve existed. As regards the fact itself, Burnouf may be right, 
■ad the jargon of those poetical portions may have actually been 
at one time the local »lialect of Kashmir, which would preserve 
a far more exact resemblance to the ancient form of speech, than 
did the Ptili and Prakrit dialects which were developed in India 
proper under the iuHueuce of the aborigines, who spoke differ- 
ently. But as Burnouf lu-ges elsewhere, that the more recent 
a Buddhistic work is, the purer and more correct is it« language, 
it i^pears to me more nattu-al to assume that these poetical por- 
tions are fragments of older traditions-, because, if they were 
more recent than the rest of the text, there is no good ground 
on which to account for their deviating irom them in point of 
language ; or if there were a difference, one would expect that 
the poetical parts would be more, correct than the prose. This is 
in fact the view taken in the Journal of the Asiatic Society of 
Bengal for 1851, p. 283, where the Laltta Vistara is said to have 
been * compiled in Sanskrit about the end of the sixth century 

K. 4 



frum ballads in an obsolete pcUois of that language, compoeedl 
evidently by bards at a much earlier period.' " 

In regard to the point on which Babu RajemlralaJ is at ubw i 
with the views of M. Burnouf, I will not venture to express anyfl 
opinion. The peculiarities of the Gatha dialect are so anomakMu 
that it 18 very difBcult to explain them. In any case, it is cl«Hr 
that, if not a spoken language, it was at least a written langna^ 
in a remote age : and it, therefore, exemplifies to us some portion 
of the process by which the Sanskrit was broken down and cor- 
rupted into the derivative dialects, which sprang out of it. 

I subjoin the concluding passage of Babu Rajendralal's dis- 
sertation, in which he states his opinion in regard to the periods I 
at which the successive modifications of Sanskrit were qx>kea in 

" The language of the Gratha is believed, by M. Bumouf, to] 
be intermediate between the Pali and the pure Sanskrit. Now, 
as the Pali was the vernaciilax language of India from Cuttackl 
to Kapurdfigiri within three hnmlred years after the deatii cf* 
Sfikya, it would not be unreasonable to suppose that the GathA 
which preceded it was the dialect of the million at the time of 
{fljikya's advent. If our conjecture in this respect be right it^ 
wuuld follow that the Sanskrit passed into the Gatha six bondred ™ 
years before the Christian era; that three hundred years sahoe- 
quently it changed into the Pali ; and that thence, in two hundred 
years more, preceded the Prakrit and its sister dialects the Swi- 
raseui, the Dravidi** and the Psinchali, which in their turn formed 
the present vernacular dialects of India." 

I liave thus' (as I originally proposed in Section I.) 
review the various pliases through which the vernacular : 
of Northern India has gone since it began to deviate from 
forms of its parent Sanskrit. Commencing with the provii 

•* [If by ihe'Driiviili i« mcuul the Telupu, or any ol its cognate Uag 
it 18 II miitukc to claai it witli thu northern Frakritd. — J. M,] 



dialects of oui- own day, the Hindi, Mahrotti, Bengali, &c., 

which diverge the most widely from the original tougue, I have 
attenipted to ascend, successively, from the more recent tu the 
more ancient mediaeval vernaculars, and to trace backwarda their 
gnulual approach in form and Btructure to the language of the 
Vedas. In Section I. the modern vernaculars are considered. 
In Sections II. — IV. the dialects entitled Prakrit are described; 
and it has there been proved tliat thej' were the spoken tongue* 
of India before the modem vernaculars came into existence, and 
shown, by a variety of illustratious, that the former approach 
much more closely to the Sanskrit than the latter. In Section V. 
an account has been given of the Pali ; and it lias been demon- 
strated at length that it, in its turn, is more ancient in ita gram- 
matioil forms than the Prakrits are, and departs less widely than 
they do from the Sjuiskrit. In Section VI. I have supplii d some 
description of the language employed in the rock inscriptions of 
Piyarlasi ; by which it is put beyond a doubt that different dialects 
resembling the Pali were in vernacular use in Northern India in 
U»e third century b,c. And, finally, in Section VII., I have de- 
', %gribed a form of corrupt Sanskrit occurring in the Gathas or 
ive poems in which the actions of Buddlia were recounted 
at a period apparently preceding the Clirwtian era. 

It is not necessary that I should be alile to point out the 
exact relative antiquity of the Pali, of the language of the inscrip- 
tions, and of the language of the Gathas. We have seen (p. 72) 
that the Pali has some grammatical forms which are older than 
those of the inscriptions ; and vice versa. It is sufficient to say 
that all these three different dialects exhibit a form of Indian 
speech which is of greater antiquity than the Prakrits of the 
dramatic poems ; and that they illustrate to us some of the 
euliest stages of the process by which the original spoken lan- 
of India, i. e. the early Sanskrit, was disintegrated and 



[cR4r. I. 

SlCf . VriL— 4^ flht mrifitai mte tf Stmiiril at a rvmaaiiar hmgve ; on tkt 
■NWMT is aAMfl air /Vainh mrtm mt iff it, mud on the pcrtod of their 

FVoin tiM ab»v« iwrwir of the ipoken diAlecte of India, oom- 
wtiHwg wMk tlM mtoitm vemacnlan, aod going back to th« 
Pmkrili KaA Um FitM, we <&aei>rer tltat the older these dialect* 
arei, Um OMI* dowly <!• tkqr nMBtble the Sanskrit, in the sab- 
•Immw of tfa* irordB tlMnsahw^ as well as in the forma by which 
itvty are dedincd and cot^jvgated. Judging by the great dtf> 
farencM i whkli ws iad but w ocu the mtxiern Indian languages 
aiul the oldest ftii— i of the vemaeular dialects, and by the 
gHMhiiU changea Ihro^gk whidk the latter have at length passed 
ittttt tlM> former, «e «a& ham no difficulty at all in cuncludiog 
ihal tho yttj oMcai knovn forms of the Prakrits also had, in 
earlier ages, nadmgvam aimtlar mutadooa, and had at one time 
been diffHranfc in aooM wspaots fiwu the languages vrKich have 
been handed down to vs; and that the further back these 
> dialtvtM Wfut, the fewer and smaller were their deviations from 
tho Sanskrit^ till tlK^y at length magged altogether in that parent 
language, and were, in butt, identioal with it. And as there is 
no doubt that those Ptakrit dial«w<B, in the oldest forms in which 
we can trace them, were spoken hmgnage^ 00 we are further 
entitled to conclude that the :5ansknt itself ¥ras at one time, u & 
at the period before the Prakrits broke off from it, a vemaciH 
larly spoken language. 

Before, however, proceeding to the particular proof of this, I 
shall first of all present some general speculations of ProfessoM 
Weber, Lassen, and Ben fey, on the anterior elements out of 
which the Prakrits (under which term I include all the old ver- 
nacular languages derived from Sanskrit) were developed, and 
the process by which their formation was effected. 

The following is Professor Weber's accoimt of the wigr in 
which be conceives the Prakrits to have arisen"*: — 



"* Indiadic Studicn, ii. p. 87, note 


" I take this opportunity of declaring myself distinctly against 
a commonly received error. It has been concluded from the 
existence (in inscriptions) of Prakrit dialects in the centuries im- 
mediately preceding oar era, that the Sanskrit language had died 
ti\it before these dialects were formed ; wher^s we must, on the 
contrary, regard the development of both the Sanskrit and the 
Prakrit dialecte from one common source, viz. the Indo-Arian 
speech, as entirely contemporaneous. For a fuller state- 
ment of this view I refer to my ♦ Vnjasaneyi Sanhitse speci- 
1,' ii. 204 — 6 ; and, in proof of what I have urged there, I 
'adduce here the fact that the priucijml laws of Prakrit speech, 
viz. assimilation, hiatus, and a fonduc8s for cerebrals and as- 
pirates art! prominent in the Vedas, of which the following are 
examples: kuta = krita, R. V. i. 46. 4.; kdta = karta, (above, 
p. 30) ; geha = griha, (above, p. 40) ; guggulu = gungulu, 
Katyay., 5, 4, 17; vivittyai=vivishtyai, Taitt. Arany. x, 58; 
krikalasa, Vrih. Ar. Ma. i. 3. 22. =krikadtUu, Rik. i. 29. 7; 
piirodai5fi=purolttea (comp. dasru = lacrjTna); pa<lbhih = pad- 
bhi^j kflhiiUaka = kshudraka; li]ialla,k);ha=bhadraksha, Chhan- 
dogya, 6. 1. (gloss); vikiri()a=vikiridra (above, p. 31); gabhasti 
= grabhflfiti, or garbhasti ; nighantu = nigranthu ; ghas = gras ; 
bhanj = bhranj = bhuj = bhnij ; bhas = bras. .... Compa- 
rative phiJology exhibits siuiiUu- phonetic prakritizings with- 
in the circle of the Indo-germauic languages as compared 
the one with the other." The same writer says in his Vajaa. 
lb. specimen ii. 203. ff. : "" "I incline to the opinion of 
who deny that the Sanskrit Bhasha, properly so called, 
ever the common spoken language of the whole Ariaa 
people, and assign it to the learned alone. Just as our modem 
high German, arising out of the ancient dialects of the Germans, 
reduced what was common to all to universal rules and laws, 
and by the power of analogy obliterated all recollection of va- 
rieties; and just as, on the other hand these dialects, while they 


'^' Uepriiited in Inditcke Sludieo, ii. pp. 110, 111. 



gradually degenerated, often preserved at the 8ame time fuller and 
more ancient forms; so also the Vetlic dialects, Ijecame partly com- 
bined iu one stream, in wLich their individual existence wu 
lo.«t, and 8o formed the regular Sanskrit Bhasba, and partly 
floweil on individually in their own original (Prakrita) irre- fl 
giilar force, and continued to be the idioms of diflFerent pro 
vinces, in the corruption of which they participated. The 
Sanskrit language and the Prakrit dialects bad, tberefore, a 
common and a Himultjuieous origin : the latter did not spring 
out of the former, but rather, being connected by a natural bond 
with the ancient language, have often a more antique fashion 
than the Sanskrit, which, being shaped and circumscrilxsd by the 
ndes of grammanans, has siicrificed the truth of analogy for the 
sake of regidarity. The Prakrit tongues are nothing else tiian 
ancient Vedtc dialects in a state of degeneracy; while the 
Sanskrit (or Epic) bhdshd is the sum of the Vedic dialects con- 
Btructed by the labour and zeal of grammmians, and polished by 
the skill of learned men. In this way we obtain an explanation 
of two facts: 1st, That the very same exceptions which are con- 
ceiled by grammarians to the Vedic language {cJiharulasi) &re 
often found in the Prakrit dialects, being in fact nothing but 
original forms; and 2nd, That in the Vedic writings, forms anil 
words occur which are more irregular than any Sanakrit wonl 
could ever be ; for as yet no fixed rules of euphony, orthography, 
or formation existed, — rules which were eventually deduced in 
part frttm those very irregularities. All the irregular forma 
which prevail in the Prakrit tongues are to be found throughout 
the Vedas. In the latter, the faculty which creates language 
is seen exuberant in its early power, while in the former 
(the Prakrits ) it is seen in the degeneracy of fuU blown licence, 
luxuriating wantonness, and at last of senile weakness. Assimi- 
Istion, the hiatus, and a fondness for cerebrals and aspirates, 
play an important part in the Vedas, not so much in those por- 
tions which are [>eculiar t-o the Yujur-veda (which, as forming a 
transition from the Vedic to the Epic period, or nether itself 



initiating the Epic period, has also a style of language of a more 
modem cast, and adapted to grammatical rules), as in the older 
forms and words of the Rig-veila, many of which were difficult 
to luidt-rstand in the age of the Aitareya and i>atapatha Brah- 
manas (paroxavrittai/ah: corap. Koth. p. li. Nighantavah.) There 
occur moreover in the Epic poems many words which, however 
corrupted, have been received into the Sanskrit sometimes with 
no change, sometimes with very little, from thu Prakrit lan- 
guages in use among the greater part of the people. Of this the 
word govhuhi is a clear example, which, according to the in- 
genious conjectiure of Gildemeister is nothing but gohinda 
derived froni gopendra.^ 

This theory of Professor Weber, even if it were correct, wo\ild 
not ID validate the conclusion which I hope ultimately to estab- 
lish, viz., that the language out of which the Prakrits grew had 
itself been subject to mutation prior to their evolution out of 
it. It would only prove that no one such language as Sanskrit 
existed during the Vedic era, but was then represented by a 
number of what (to distinguish them from the Prakrits) I may 
call Sanskritic dialects, which, by the continued action of a 
modifying analytic process all along at work in them, were, on 
the one hand, gradually broken down into the dialects which 
received the name of Prakrit, while, on tlie other hand, by a 
reverse process of synthetic and more formal construction, 
another language of a diflFerent character, and previously non- 
existent, became developed out of them, under the appellation 
of Sanskrit. 

Weber's theory, however, taken in its full extent, appears to 
me to be disproved by the fact that, in its forms, the Vedic Sans- 
krit is (excepting a few archaisms) nearly identical with the Epic, 
while it is very different from even the oldest type of Prakrit, It 
is no answer to this that some old Vedic forms, such as the instru- 

E^ntal in «bhis, reappear in the Prakrits ; for no one asserts 
That, at the earliest period when the Prakrits began to be formed, 
the Sanskrit did not still retain many of its Vedic forms. 




LcwAr. ». 

I will now adduce two quotations of considerable length, from 
Lassen's Indian Antiquities, vol. iL pp. 1147 — 1149, and 1151 
— 1153, on the history of the languages of Northern India, in 
the course of which he replies to the preceding observationB of 

" The inacription of the Sinha Prince Ruilradaman, which 
dates from the year 85 B.C., is written in Sanskrit prose, of an 
artiiicial character, with long compound words. From tliis fiict 
we may infer that Sanskrit was no longer spoken by the comtnon 
people, hut only by the Brahmans and other persons in the 
higher classes." 

•*It has been already shown (p. 486) that in AiSoka's time the 
common people spoke dialects derived from the sacred language, 
and that, at that time, there were at least three such diulect« ; of 
which one prevailed in Eastern India, the second in Guzerat, 
anil the third in Eastern Cahul. The existence of a fourth, of 
which the seat was perhaps in Upper Eajasthan, is attested bf 
the inscription of Megliavahaua. It is highly probal>]e that the 
popular Indian dialects existed at a still earlier period than this, 
[the age of A4oka was 263 — 226 B.C., and that of Meghavahana 
110a.d.'°*] ; for the jvccounts of Buddha's sayings and doings appear 
to have existed in a double form, ?*. e. botli in the vernacular 
tongues and in Sanskrit. I do not venture to assume that the 
vernacular tongues originated much earlier. It is inde«d true 
that we find in the Vedic hjnnna some individual traces of thoee 
corruptions which in Prakrit have become the ride and character- 
istic feature of the language. But we must assume a long 
period to have intervened between these isolated appearanoM 
and their full blown development, as exhibited in particnlu 
local dialects. I do not, therefore, believe in a contemporane<.«i!> 
development, side by side, of the Sanskrit and the Prakrit 
tongues out of the one common source of the IndivArian 
language ; hut I assume that it was not till long after the immi- 

"" Las8«n, App. pp. x. ixiii. 



wUl ft 

■ pec) 

gration of the ludo-Arions that the Prakrits were formwi in the 
Beveral provinces of India, I further regard it as improbable 
that the Prakrits arose out of one particular diiUect of the 
ii^skrit; for no dialects of the Sanskrit have yet been pointed 
out. An accoxint is to be found, it is true, in an ancient record, 
according to which the Sanskrit had been preserved in greater 
purity in the northern countries than elsewhere, and Kashmir 
and Badari, at the sources of the Ganges, are specified by the 
comraentatx>r as such regions. This, however, ia not sufficient 
to prove that in the different provinces of India there were then 
fundamental differences in the sacred language. 

"No conclusion in regard to the existence of dialectic varietiefl 
in the Sanskrit can be drawn from the fact that the Prakrit 
dialects: have all preserved the form of the instrximental phmil 
m hi (derived from bfiie), in words ending in a, while the modern 
skrit has lost this form; for the ancient form in ebhiaia not 
peculiar to any particular Vedic writinga The preservation of 
this form only proves that the Prakrit dialects began to be 
formed at an early period, when the termination in question was 
in frequent use. The early adoption, too, into Sanskrit of words 
which had become modified according to the laws of the Prakrit 
dialects, testifies, not so ranch to the early creation of popular 
dialfccts widely different from each other, as to the mere begin- 
nings of such. We have to regaid the causes of the vaiieties in 
the Indian dialects as twofold. The first is that general one, 
which has operated also in other languages, an<l which is indeed 
the principal, viz., those peculiarities connected with the abodes 
and the character of the tribes into which a people becomes 
divided. The reason why they have so operated, as they actnally 
have, is in individual instances often difficult, nay, impossible to 
assign. In this way, five principal modern languages, the 
Provencal, French, Portuguese, Spanish and Italian, have arisen 
out of the Latin. Of these languages, the second, the fourth, 
nnd the tifth are rich in dialects. The second cause is (as has 
been already noticed) a special one, — I mean the influence 



exerdsed on the Prakrit dialects by the languages of the abori*! 
ginal tribes adopted into the Indian political system, who discarde 
their own form of speech and adopted the Indo-Arian langnagel 
of the province in which they dwelt. These aboriginal tribui 
contributed, in some instances, to introduce peculiar varietieifl 
into the Prakrit dialects, 'VNTien these aborigines were |iar- 
ticuhu-ly savage and uncultivated, it could scarcely fail to bappro 
that they occaaioned very great corruptions of sound and fonn [ 
in the Iniio-Arian languages." 

The second jwissage is a« follows : — 

" It is in the period with which we are now occupied {ktA 
that between Vikraniailitya and the later Gupta kings) that the ' 
appellations Sanskrit for the classical language, and PnVrrH 
for the forms of speech springing from it, must have arisen; 
because it was now that the distinction between the clasidcBl 
language (which was no longer employed as a spoken tongoej 
except by the Brahmans and highest classes) and tiie popular! 
dialects became decidedly marked. It has been maintained 
Sanskrit was never the common populiur dialect of tbe Ariaal 
Indians, but owed its origin only to the learned ; and that the 
Vedic dialects coalesced, on the one hand, in a single language, 
and so created the regular Sanskrit, in which they were lost; 
and on the other hand, in virtue of their inherent chanicttf, 
became corrupted and irregular, and in this corrupt furm con- 
tinued to exist as the vernacular dialects of particular provmo& 
I must dissent from this view on the following grounds: First, 
it has not yet been proved (as I have already, p. 1 148, obeerved) 
that there were any different dialects in the Vedie language, 
prove that there were, it must be shown that in ooutemi; 
neous Vedic writings there are found variations of such 
ewential character as to justify us in assuming a varie^ of 
dialects : varieties observable in writings belonging to differcatl 
ages only show that there was a progreiEdTe departure from i 
earlier condition of the language. Secondly, it is nc 
that we be agreed as to what we mean liy tan^ua^ 


thereby we mean the style of expitssion, then it may be 
asserted of many languages which have attained a higher degree 
of perfection by being employed in literatiire or in public as- 
semblies, that they were not popular languages. The Athenians 
and Romans certainly did not, in their ordinary life, express 
themselves in the same style in which their orators spoke ; and 
we Gernums permit om^elves to make vise of many turns of ex- 
pression which we deny ourselves in books. So too we may 
suppose that tl»e Indians of the earliest age did not ordinarily 
speak the same language which their poets employed. If on 
the otlier hand, by language we mean f/rmnm^Uical foitiis, I 
cannot see wljy the Indians should not in the earliest ages have 
made use of the same as the contemporary poets employed. This 
certainly was the case also in the succeeding periods. Tu this it 
must be added that Panini, the oldest of the three grammarians 
who are 8tyle<l saints, uses the word bhushd (speech) to designate 
the ordinary language in contradistinction to the Vedic, and em- 
ploys as synonymous with bhmlidydm the word hikf, i. e. i>t the 
w&rkL The language which the Sanskrit-speaking Indians then 
spoke cannot, therefore, have been difiFerent from this bhdshd, 
or current form of speech. Its fate in contrast to that of its 
daughters has been a peculiar one. WTiilst among the Greeks 
the Attic dialect became the general language of prose composi- 
tion, . . . and the other dialects became less and less prominent, 
• . . and whilst in Germany the new high-German, from its 
in literatm-e and education, has more and more superseded 
popular liiiilectfi, the sacred speech of the Brahmans, on the 
Itx^ary, continued to lose ground, not so much in local exten- 
sion, as in its employment by the different classes of the population 
in the same couutrios. It may be assumed that in the time of 
Asoka the greater part of tlae people in the countries inhabited 
by Arian Indians spoke the local dialects, and that only the 
Brahmans and the principal persons spuke Sanskrit. On this 
circumstance the distribution of the dialects in the dramaa rests. 
As the kings who were inclined to the Huddhist religion permitted 




only the popular dialects to be used in their inscriptions 
coins, it becomes probable that they did the same thing in 
decrees, and for other purposes." 

The following remarks relating to the early extension and 
vernacular employment of Sanskrit, to its subsequent disuse as a 
spoken language, and to its idtimate resuscitation in a somev^iat 
modified form, as a refined and sacred dialect, are translated, 
with occasional abridgment, from Prof. Benfey's article on India 
(above referred to), pp. 245, ff. : — " The language which we no 
call Sanskrit was once, aa both the ancient and modem dial< 
which have issued from it distinctly show, the prevalent popi 
speech in the greatest part of India. Alongside of it there ez< 
isted in the remotest times several dialects of one or more 
guages, not related to it, of the aborigines of India; which 
langiiages hod at first a ^vider and in later times a continually 
decreasing, extension. The period when Sanskrit began 
spread itself over India cannot be determined any more than 
era of the immigiation pf the people who spoke it. We 
only determine the following points : First, in regard to exten- 
sion; (1) the Sanskrit once prevailed over a considerable tract 
west of the Indus, as is shown both by many geogr^bic*! 
names in those regions, by the accounts of Chinese travel! 
and by the languages which are now found existing there 
(2) to the north, the Sanskrit or its dialects prevailed 
far as the Himalaya and the Indian Caucasus ; (3) to the east, 
in the time of Asoka, as far as the region of the Brahmaputra, 
though this region was not entirely Sanskritized: (4) to tite 
south, Sanskrit exercised nearly its fiill sway as far as the sootih' 
em frontiers of Maharashtra : this is proved by the fact ihtt 
one of the dialects which aie most decidedly of Sanskrit origia, 
namely, the Prakrit pre-eminently so named, is also called tbe 
language of Maharashtra, and is manifestly the parent of the 
modem Mahratti. (5) Sanskrit penetrated still further to tbe 
south, where it formed the language of educated people: birt 
this occurred at a time when the SanskxitHspeaking race had ndt 




8u£Bcient power entirely to expel the indigeoous language, as 
they were able to do in Northern India with a few very isolated 

"Second, as to the time when Sanskrit was the langfuage 
of the people we can determine as follows : We find in Asoka's 
time two vernacular dialects, one in Guzerat, and the other 
in Mogadha, which, as their entire structure shows, could 
not have existed alongside of, L e. contemporaneously with, the 
Sanskrit, but must have become further developed in those pro- 
vinces after the Sanskrit had previously prevailed there : conse- 
quently the Sanskrit must have died out before A^oka, who lived 
in the third century B.C., and therefore about three centuries after 
the period to which the rise of Buddhism may with great pro- 
bability be assigned. Now it is related to us of tlie first 
Buddhists, tliat they composed their books not in Sanskrit, hut 
in the vernacular dialects. The sacred language of Buddhism is 
the Fall, which, though ysayiug in many particulars from the 
language of Magadha, and approximating to the principal 
Prakrit (the Mabarasbtri), stands yet in a similar relation to 
the Sanskrit as the latter, and the two dialects of A^ka's in- 
scriptions. It becomes, therefore, highly probable that at the 
period when Buddhism arose, i. e, about the sixth century B. c, 
Sanskrit was no longer the speech of the people. The entire 
character of Sanskrit shows that, at the time of its acme, it was 
fixed by means of something resembling a literature, and it is 
only on this supposition that we can comprehend how it appears 
itt regions so far apart as the north and north-west of India and 
the Mahratta country, as a perfectly uniform basis of the dialects 
which sprang out of it. But a speech which becomes fixed in 
socb a manner does not soon die out. If we assume about tJiree 

jturiee for the time of its gradual extinction, the period when 
rit was the ordinary language of the people is thrown bat-k 
al>oiit the ninth century before Christ. Diu-ing this and 
the immediately preceding period there existed, as we have 
tiieeAy conjectured, a political union which embraceti the entire 

I. a 





Indian empire ; aud tis we now know that Sanskrit must hav( 
been the actual speech of the people in the Malirotta count 
also at this period, we may conceive tliis political union to hsv 
extended from the Himalaya to the south of the Mahrstta 
country. After this pc>litical unity had become severed (till the 
period of its restoration under Ohandragupta), the various ele 
ments of Indian life became separately developed in the differ 
provinces ; and this was the case with the Sanskrit, too, whic 
up to that time had been common to alL Out of this varietf 
of local developments which the Sanskrit imderwent, its difieren 
derivative languages arose, the earliest forms of which bore abou 
the same relation to Sanskrit as the Bomanic dialects to Latin. 
" But while the Sanskrit was being thus developed and modifie 
by popular use into new vernacular dialects, the literature whi< 
had been created in Sanskrit while it was yet a living tong 
was still preserved in the schools'"' of the Brahmans, and along 
with it the Saiiiskrit itself as the sacred language of culture and 
science, "WTien aroused to new energy by the attxick made upon 
their system by the Buddhists, the Brahmans came forward with 
certain writings composed in this sacred language, and declare! 
to be of primeval antiquity : one of the earliest of these was the 
Institutes of Mauu ; and then followed the Ramayana. Bat ex- 
ternal grounds, as well as the mention which they tuake of tlie 
Yavanas (Greeks), prove these works to have been composed at 
a much later period than that to which they are alleged to be* 
long. In like manner the treatment of the language in thci^ 
books, and still more in the Sanskrit hterature which folloir^l 
and is connected with them, demonstrates that they cannot pos- 
sibly have proceeded from a popular dialect, but, on tlie «•• 
trary, are the products of a learned, or rather a sacred languBgib 
ffhich, having died out among the mass of the people, luid bees 

Id* w xhough we have no diitinct external evidence that there men if 
Bach schools at this earlj period, we uajr yet appeid to the whole tntellectail 
development of Indbn life, in the form whicli it must have taken even faeftfc 
the rise of Huddliism, as evidence of their existence." 



preserved in the circle of the educated pri&sthood as the me- 
dium of intercourse with the gods, and of communicating the 
sacred sciences, and was cultivated with the liveliest zeal and 
devotion. Out of this circle again Sanskrit passed over to 
those persons who stood in connection with the priests aa mem-> 
liers of the same administrative caste. When the Brahmans 
recovere<l their predominance, Sanskrit became for a time the 
language of the educated classes, of the court, and the admini- 
stration"'* generally : and even the Buddhists could not abstain 
from employing so valuable an instrument of cultivation. We 
have only to recollect the manner in which the Latin, though 
long a dead language, remained in use throughout the middle 
ages, and even in our own time, in order to perceive clearly how 
the Sanskrit also, though it had died out as a vernacular 
tongue between the ninth and sixth centuries b. c, should yet 
have held its groimd in the highest circles, and continueil in use 
there to such an extent that it can even now be employed ajs an 
instrument for the expression of thought on the highest subjects. 
The Sanskrit had, however, here an important advantage over 
the Latin in this respect tliat, wherever Brahmanism prevailed, 
it waa regarded as a sacred language, as all the most sacred 
books of that religion were composed in it» In consequence of 
opinion, it was considered a religious merit to be even 
juainted with it; and a Sanskrit grammar, or other work 
which contributed to a knowledge of this language, was and is 
looked upon as a sacred book. In the same way a knowledge 
of Hebrew was long preserved among the Jews ; and even so lat« 
as, perhaps, sixty years ago, no one among them could lay 
daina to the character of a learned man unless he had learnt 
the * aacred language.' "^ 

^'* [We have another instance of a language not vernarular in India being 
ated as the language of admiaistration, in the Persian, which, though unin- 
t«Uigible to the maaa of the people, wa5 used bj the ftlabomedans, and 
after them, for mauj years (until about twenty ye»n ago), by the English, 
ai the language of the law courts and the revenue offices. — J. M.] 

L 3 




•* At the period when the dramatic literature assaTOe*! its fixed 
form (a period which cannot yet be determined, but which ni! 
be conjectui'ally placed in the sixth or seventh century a^d.), 
knowledge of Sanskrit must have extended, on the one hand, 
all who laid claim to the character of educated men, for oth 
wise the dramatic poets could not have compoeed in Saiisl 
the leading parts in plays designed for representation before li 
entire public ; and on the other hand it must have been 
stantly used as the language of public documents, of religioi 
and of learned men, for otherwise it could scarcely have 1 
put into the mouth of gods, kings, and priests, Wlietb 
Sanskrit was at that time the proper court-language, I cann' 
determine; but I scarcely think it wag, as the ofiScers of 
fitat*, if not Brahmans, do not use it." 

Professor Benfey then proceeds to specify the differences 
tween the ancient form of the Sanskrit when it was still a v 
cular language, and the later form which it took after its regene- 
ration as a sacred and learned form of speech, so far as he coi 
sidered himself in a position to do so at a period (1840) wheH 
had before him but a small portion of the Vedas, which ftimi! 
us with almost the only means we can have of judging what 
earlier language was.""* He remarks: " Tlic late Sanskrit is disti; 
guisbed from the Vedic by the tise of extravagantly long compoiuii 
Even if the specimens of the Vedas, and the Upanishads whicb 
are known to me, had not shown tliat in this respect there ia 
essential difference in the use of the Sanskrit at the two peril 
to which I refer, it miglit have been concluded with certai 
from the character and length of these compounds, that 
monstrosities coidd not have been created at a tirae when thf 
language was in vernacular use. Such compounds might ooah 
sionally have been used with effect ; but a living language wooU 
have energetically rejected such an abuse of these forms as 

** Had these observations been written now, Professor Benfey would ] 
bably have seen no cause to modify his main conclusions, though he woxJdl 
been ui a position to express himself with gn^aler confidence and prcciti>»j 




find in the late Sanskrit writings, which renders all easy compre- 
hension impossible. On the other hand, the effort to employ 
such compoimds was quite suitable to a lcarne<l language, 
and to a learned poetry, which waa far removed from the real 
life of the people. In like manner the laws of Sandhi, as prac- 
tised in its widest extent in later Sanskrit, must have been 
equally foreign to the ancient vernacular Sanskrit. In late 
Sanfikrit all the words of a sentence are combined in one im- 
meose whole by the assimilation, or other connection, of their 
final and initial letters. This rule does not, in general, prevail in 
the Vedaa; and although it is well known that in actual discourse 
the final and initial letters of words exercise a certiun modifying 
influence upon each other, every one who has considered the 
limited extent to which a vernacular dialect, and even a literary 
work composed in such a dialect^ can obey this law, and who, at 
the same time, knows to what extremes the modem Sanskrit 
pushes the application of this rule, will be convinced that the 
excessive employment of Sandhi cannot have sprung out of any 
popular use, but must have resulted from carrying out to an 
absurd extent a grammutical canon which is correct in itself. 

" Further, when the later Sanskrit is accurately examined, it ia 
foimd to be affected in a most, important degree by the influence 
of the popular dialects derived from the more ancient Sanskrit. 
The Indiana, with their genius for grammar, or philology gene- 
rally, were in general well aware of the modifications which 
the ancient language had received from the dialects which liad 
been developed out of it : they had investigated the phonetic 
laws by which these dialects had been derived from their 
parent^ and could, as it were, transport the latter back to the 
former. This facility threw them off their guard ; and it conse- 
quently becomes possible for us to demonstrate that the Sanskrit 
of the whole Indian literature subsequent to Manu's Institutes, 
cannot be in all respecta the ancient language of the people, with 
a degree of distinctness which none of the Sanskrit authors, 
eonvinced as they were that they were writing correct ancient 

L 4 



Sanskrit, could have imagined. I miiflt confine mv^elf here to 
exhibiting the principal elements of this proof. It is divisible 
into two part*; as we must (1.) inaiutain that the nerw Sanskrit 
has lost'^ much which the older Sanskrit had, and which it could 
only lose from the circumstance that it had died out in the inter- 
mediate period, and had now to be revived in a form which might 
be as intelligible as possible. To this head belong a niunber of 
roots and inflected forms which the grammarianB recognise and 
adduce partly as current, and partly as obsolete, but of which th« 
later Sanskrit makes next to no use. The reason of this ia that 
these roots, as well as these inflecte<l forms, were either entirely 
lost in the vernacular dialects which existed at the time «h«a 
the new Siiuskrit was created, or bod become so disfigured 
that their Sanskrit form could nut have been easily di^ 
covered or understood. (2.) The new Sanskrit contains in 
much that the old Sanskrit could not liave had. To this h< 
belong a number of forms of roots which had become modi6cd 
according to the laws of some one vernacular dialect, and wtucb 
have been employed in the new Sanskrit in this modified shape* 
which the grammarians either hesitated to refer to its pro] 
Sanskrit form, or did not comprehend. Every single example' 
of this which might be adduced would, however, require detailed 
development and proof, which would demand too much space 
be here attempted. 

" I will, therefore, content myself with repeating the 

«oa "Xhc Sanskrit has lost a great m&ny verbal roots, and hu frerjuentljl 
modified the original meaning of those ilill in existence." — Aufrecht. Unidi*; 
ftQlraii, pref. p. viii. "In the course of time »om« bninches of liier«n 
di»upp<?ared, a number of words becnme antiquated, and the tradition •• *» 
ihc-ir nu-niiinp was either entirely lost or corrupted. When commentatoa 
nn»se to explain the Unadisulras,"— supposed by Dr. Aufrocht (p. ix.) tt 
bf considcrohly older than raniiii,— "they found the greater part of «k 
words contained in them slill employed in the literature of their age, or T»- 
eorded in older dictinnnries. But an unknown residuum remained, end to 
these, whenever tradition failed them, they were bold enough to assign qoit»J 
srbilrsry significations"— Ibid. pp. xi. xii. 


di»- J 

SECT, n 




results of the investigations which have been here merely in- 
dicated, and in great part yet remain to be carried out. These 
results are : That from the pei%od when the Sanehntspeakmg 
race immigrated into India down to perhaps the 9th century 
B.C., Sanskrit became diffused as tite prevailvng vernacular 
dialed over the vihale of ITindusflmn, as far as the southern 
borders of tfie Mahratta country. It penetrated no furtfier 
South as a vernacular tongue, but only as tlie language of educa- 
twHy and apparently at a latei' period. From, the 9th ceyUury 
B.C. the Sanskrit began to die out : derivative dialects became 
devehrped from it; and in the 6th century B.C., it fiad Iteco-me 
extinct as a vernacular language. On the other hand, it main- 
tained its ground in the schools of the Brahmans. About the 
third century B.C., in consequence of the regeneration of Rrah- 
maniam in Kanouj, it was brought hack into public life as a 
sacred language, and gained a gi'adually increasing importance 
as the organ of all the higher intellectual development. About 
tlie fifth century a.d., it had become diffused in this character 
over the whole of India. >So long aa the empire of the Hindus 
lasted, it continued to increase in estimation; and even long 
after the Mahomedans had settled in India, it was the sole in- 
strument for the expression of the highest intellectual efforts." 

Sbct. liL^^Reatont for tuppoting that the Sanskrit vku ori^mtUji a tpoken 


It appears from the passages cited from the works of Professors 

Laasen and Benfey, that these distinguished scholars assiime that 

the Sanskrit was once a spoken language, regarding this as a 

fact which admits of no question, and requires no argument to 

prove ita truth: and Professor Weber is of opinion that the 

only [ndo-Arian speech which existed at the early period to 

^^^rbicb I refer had not yet been developed into Sanskrit, but was 

^^Htill a vernacular tongue."*^ As, however, what seems so clear to 

^^^ft **" Indische Litcratur(;e»chich(e, p. 1. 



HiarroKT AND rklahons op 



the Enropean scholar may not be so plain to the Indian reader, 
it becomes neoeaaaiy for me to adduce the most distinct evideace 
of the fact which I am able to discover. 

Fint: — Even though we assume, as we must do, that there were, 
from the earliest times, other forms of gpokeu language currtoit 
in India besides the Sanskrit; yet these would be tiie dialects 
of the Dasyus, or non-Arian tribes ; while the upper classes of 
the p«^pulation of the Arian race, the same order of persona who 
in after times spoke the most refined Prakrit, must have be«n io 
the habit of speaking Sanskrit a few ages previously ; for, in fact* 
no other Arian language then existed in India which they could 
have used. If languages with such a complicated structure u the 
Pali and the Prakrits were employed in common convet8aliaD< 
there is no difficulty in supposing that Sanskrit too, which was not 
much more complex, should have been spoken by ordinary persctiu. 
We must not, of co\irse (as Professor Benfoy has well remarked 
above, p. 151), imagine that all the refined rules for thepermttt»- 
tion of letters which were used in later Sanskrit compositiou were ^ 
then employed in daily discourse, though some few of them might f 
have been ; for the use of these rules is by no means essential to 
the intelligible or grammatical employment of the language; and 
at the time to which I refer, they ha»l not been developed oc 
systematized. Many, too, of the more complicated inflections <rf 
Sanskrit verbs woidd be then little used in conversation ; as, in 
fact, they are now comparatively little used in most litersry 

It is true that we cannot point out the exact forms of all the 
Sanskrit words in use at the latest period, when it was so em- 
ployed as a spoken tongue; for the language of oonTeFsadon 
always differs to some extent from the language of fonnsl 
composition or of books, and the vernacular Sanskrit was no doubt 
undergoing a perpetual alteration till it merged into Prakrit. 

Second: — The case which I have supposed here of Sannkrit 
having been once a spoken language, and having at length censed 
to be employed in ordinary discourse, while the provincial dialeota 




which (sprang out of it, and graduallj diverged more and more 
from it and from each other, have taken ita place as the popular 
vehicles of conversation, — is by no means a singular occurrence, 
unprecedented in the history of language ; on the contrary, the 
manner in which the Italian, French, and Spanish languages (to 
which Professor Lassen refers in a passage cited above, p. 143) 
have been formed out of Latin, presents a very close parallel to 
the mode in which the various medieval Indian Prakrit 
bhashas (which in their turn have given birth to the modem 
Bengali, Hindi, Mahratti, &c.) grew out of Sanskrit. During 
the existence of the Roman empire, Latin, as Is quite well known, 
was the spoken langmige of Italy, and other western portions of 
Europe. It is now in all those countries a dead language, and 
is only known to the learned who study the works of the Latin 
philoflophers, historians, and poets ; just as it is only the Pandits 
of India and other scholars who can understand the Sanskrit 
Sastras. But while Latin has itself ceased to be a spjken lan- 
guage for eight hundred or a thousand years, various vernacular 
dialects have (as I have said) sprung out of it, such as Italian 
and the other modem tongues already specified ; the I^tin 
words which compose almost the whole of their vocabulary being 
varioiLsly modified, and the ancient Latin inflections being either 
corrupted, or dropped, and replaced by particles and auxiliary verbs. 
Of these derivative dialects, the Italian, which is spoken in Italy, 
has retained the closest resemblance to its mother language. 
Many of the changes which Latin words have undergone in Italian, 
resemble very closely the mollifications which Sanskrit words 
have undergone in Pali and Prakrit, as has been already re- 
marked in the passage quoted, in p. 81, from Bumouf and 
Lassen's Essay on the PalL 

To exhibit the wonderful similarity (amounting in some cases 
to identity) of the processes by which these two ancient lan- 
guages, the Sanskrit and the Latin, are modified in tlieir 
modern derivatives, I shall place in juxtaposition a few of the 
most remarkable instances of it which occur to me. 




L Phosetic Changes. 

(I.) Words in which the c or i is dropped from a compound letter ct, »et,ar 
kt, while the t is doubled. 

Sinikrit fornu. modUied In PiUi and ] 
Sanikril. Pall ami PrmkrlL 

mulitM becomes mutto. 

Ladn fonu 

. at modiAcd In ItalUn. 



perfectus becomes perfetto. 































.) Words in wliich the p of pi 

ruptus becomes 
































(With many others.) 

uptas becomes 































kbit to 







(3.) Words in which the / oF a compound letter, pi or U, is dropped."" 

pUnctus becomes pianto. viklavos becomes vikkavo. 

planus ,, piano. 

'"* The Latin c is suunded as k in Sanskrit, 

giu, in Italian, is Sdundoil asju ("Si) in Sanskrit. 
"<^ In Prakrit, however, a cuinpound letter, of which / is the final par> 
tion, is genernily dissolved into two syllables, as glSna becomes gildna. 

SECT. tX.] 



(4.} Words in which the b of the compound letter bj is dropped. 

Bubjectus becomes Boggetto. kubjas becomus kbujjo.'" 

objectus „ oggetto."* abjas „ ajjo, 

(5.) Words in which the letters rejected are not the same in the Italian and 
Prakrit, but in which both languages show the same tendency to 

sboorptiu becomes uaorto. 

utpolam becomes uppalam. 






































































A large portion of the simplifications in Pali and Prakrit arise 
from the rejection of r licfore or after another consonant, as in 
the words kanna for kama, savva for earwa, mitta for initra, 
■putta for putra, &c. This elision of r ia not usual in Italian. 

IL I give an instance or two to show the manner in which 
the Latin case-terminations have l)een dropped in Italian. In 
Latin the word annus, a year, is thus declined. 

Sliigulxr. Plural. 

Nom annua. Nom anni. 

Gen anni. Gen annorum. 

Dat and Abl. . . anno. Dat. and Abl. . . annia. 

Accus. .... annum. Accus annoa. 

[•" Var. II. S4. 
••* Pronounced as if written in English, terffetto, ojjOto. 
*" lean only infer, from the rule in Vararucbi, III. 2, that the » is thrown 
out and the m doubted in this and the two ruUowing words, as X have not met 
them anywhere. 





lu Italian, on the contrary, there is only one form in the 
aingiilar, anno; and one in the plural, anni; the caae-termina- 
tions being supplied by prepositions with or without the article, 
as follows: — 


Nom. and Accus. . gli onnL 

Gen degli uinL 

Dat. ..... agit *nni. 

Abl dagli anni. 


Nom, and Accus. . Y anno. 

Gen dcli' anno. 

Dat air anno. 

Abl dftir anno. 

III. In Italian verbs, the Latin forms of the active voice are pre- 
served in a modified shape, as the following example will show: 

IVM«nt l»in(>. Imperfect t«n»e. 

L.itln. UaHan. 

1. vendebam. 1. rendeir*. 

2. vendebaa. 2. ▼<:nd«vL 

3. vendebat. 3. venders. 

4. yendebamus. 4 vendcvamo. ' 
fi. Tcndcbatis. 5. vendevate. 
6. vendebont 6 vendeTano. 

Pluperfect tttue. 

1. Tt;ndidissein. I. TendeoL 

2. veadidis8«8. 2. Teadeni. , 

3. vendidiiwct. 3. rendcMb 

4. vcndidisBerous. 4. TendMsimOk 

5. vendidiiifetis. S. vendeste, 

6. vendidisgenU 6. Tendetsera. 

B\jt(lV.) in the passive voice the Italian language has entirely 
loat the Latin forms of conjugation. Thus instead of the I^itin 
forma ego laudor, ' I am praised ; ' ego laudabar, • I wu 
praised;' ego laudarer, 'I should be praised,' &c, the Italians 
employ in all tenses (as the Latin had already done in a few^ 
the substantive verb with the past participle, and say lo eoao 
lodatOt lo era lodato, lo earei lodato, ' I am/ ' I was,' * I should 
be, praised.' 

These few instances will suffice to show the Indian reader how 
the Latin words and inflections are modified in Italian. 

It is thus manifest from the history of Italy in anoieDt and 


























Perr«t tenic. 




























modern times that the people of that country ouce spoke Latin, 
and now speak Italian, a vernacular flialect derived from 
I^tin, and differing from it in many respects, as the Indian 
Prakrits do from Sanskrit, while Latin equally with Sanskrit is 
a dead language, known only from ancient books, or from its 
use in the public worship of the Roman Catholic Church, or 
from ita occasional employment by modern scholars in their 
writings, or in scholastic discusaiomj, in that and other coimtriea. 
But if it be true that a language like L&dn, with its numerous 
and varie<l inflections, was once the conitnon speech of the 
whole Roman people, there can be no difficulty in supposing 
that while the modern Hindus (excepting a few Pandits) can 
only gpeak Bengali, Hindi, Mahratti, &c, and while their 
ancestors spoke different Prakrit dialects, which are the imvic- 
diaUt parents of the modern vernaculars, the Hindus of a still 

rlier period should have spoken Sanskrit itself, from which there 
ia no doubt that the older forms of Prakrit were immediately 
derived. If even in our own day Pandits can talk Sanskrit, why 
should not the vernacular use of it, though in a far more simple 
and mitiu«l style, have, in former ages, been conuuon among 
Brahmans, and even among other persons in all the different 
classes of society ? 

Third : — The fact that the dramatic authors put Sanskrit into 
the mouth of Brahmans and other persons of the hi^er ranks, 

aids an argimient of considerable force that Sanskrit was ouoe 
'^loken by the whole community, and by the upper clai^es down 
to a much later period : and even the common employment of the 
same language by learned Indians in their schools and disputa- 
tions down to the present day, may go some way to prove its more 
general currency as a vernacular at an earlier date. For if 
Brahmans did not at one time employ it in their ordinary dis- 
C01UW, how did they ever get into the habit of speaking it with 
«o much ease and fluency? But if Sanskrit was at one time 
ordinarily spoken by Brahmans, the use of it would eas\\y be 
propagated irom une generation of learned men to another. 


rc«Ar< t 


eeeh in andeot 

otfcos the language of 
i of the Ar;aB to wtiicfa he alludes 
if not Sanskrit it^f ! 
I depend on the age in 
t*a iBBtitutes to hare 
This pMBige, at aaj nte, leaTcs the impreastoo 
Aat tikere vaa a beoad finliaetiaB between tiie Arian language 

it vas contrasted ; and 
, if aagr^ neagAmi as eTBsting in the foi 
: fiaiiparatitiJj iiwigiaificant. 
Fifth : — In aoaae «f Aa aJfat Indian gnnunarians» such as 
Tadw aad Fipni* we find tite ohaoiate laogoi^ of the Vedai 
HktimgiuAtA finm the oi&inrr Saankxitof tiieday. The fonoer 
in alladed to or deaigBaAed br the tenns OHradhydtfam (in the 
Vada), ekhandoB (aietre)^ or draAa (the npeedx of the mbis)b 
Ac; while the ooiileiDp<»»7 San^rit is referred to as bhdtki 
(the spoken languageV Thos Yaaka, the andent author of the 
Ninikta, in the introductory part of hia work, L 4, speaking of 
particles {nipatdh), says: ^m*H ^HK ^W*4I^ ^l^^rTlfd l 

T^ Tf% HT^T^ir^ ^J^TJIT^^ ^PtiR'^'J^ T^l ^ 

^4***i4idfd irfTIWr^: <.rtjlt< ll "Of them Uiese four 
are particles of comparifion. • Iva' Las this sense both in the 
<M)trunriii language {blidshd) and in the Veda (anvadhydi/am): 
timo A'jmr ivo., Jndra iva, ' like Agni,' 'like Indra.* • Na' has 
io the bhn«hd a negative sense. In the Veda, it has the aeow 
iMith of a negative and also of a comparative particle. ThuB ia 
the tejt na Iwlavm devam amansatay 'they did not regard India 
(tN a gofl/ it hofl a negative sense/ &c. Again, in the neiEt 


"* Mnnu. X. 4ft. Sec the verse quoted and translated in Port I. of 

wurk, p. I7n, iind iioUj, 





section (L 5.), he nays eimilarly: H»lf*<fr( fwf?rf^<^r5f^ 

particle 'nuTiani' is used in the bltdsJid to signify uncertainty; 
in the Veda, too, it has that signification, and is also a mere ex- 
pletive." Again, Yaska says, Nir. 11. 2: ^R^ft wf^"*^ 

%7T^>rt- Hifwr ^^ ^rffirfFji v mU fl«dy i^^ 

m4j^ ^H^i^^^lig II " There are Vedic {nalgama) nouns 

tdamunoh and hhefidsadhnlt) which are derived from rrmt« 
id in tlie bhashd ; and also formations in the bhdslid, such as 
iishnanif ghritam, which come from Vedic roots. Further, t!ie 
root« only are employed in the speech of some ; the derived 
forms [or nouns] in that of others. Savati, as a verb for (folncf, 
is iised in the language of the K&mbojas only : its derivative, 
£ava (a coi^e\ ia in use in the language of the Aryas. The 
verb tldti is employed in the sense of cutting by the people of 
the East: while the noun ddtram (a aickle) only is known to 
those of the North." Here it will be observed that pure 
Sanskint words are referred to as being used in the speech not 
only of the Aryas, but also of the Kambojas, a people living 
to the north-west, who are distinguished from the Arya«. 

In the Sutras of Panini the Vedic dialect is referre<i to as 
follows: 1.2.36, f^W^T ^-^f^ "in the chhaadas (Veda) there 

is an option :" 1 . 4. 20, TTSfWH^ftt'I S<^f\l " In the c/ikandas 
we have the forms aymmaya [instead of ayomaycC], &c." ; and so 
in numerous other apliorisms. The word imaUra is put for Veda 
in the following Sutras, 2.4. 80; 6. 1. 151 ; fi. 1.210; 6. 3. 131 ; 
6.4. 53; 6. 4.141. The word nujania is similarly used in 6. 3. 1 13; 
6. 4. 9; 7. 2. 64 ; 7. 4. 74 : — and the expressions fifthati (*in a 




lialii '), and richi (' in a Vedic verse '), are employed in the bum 
way, 4. 4. 96; 6. 3. 130; and 6. 3. 133- In contradistinction 
to the Vedic dialect, on the other Laud, the current Sanskrit 
designated by Panini aa bhdskd in the following Sutras, 3. 2. 108 
>||T||«|j «j<444(4:i *in the current language the root 
Avi, vns, and »ni, take kvagu ;" 6. 3. 20 : "^ '<«| Xmi«H<^ ^ 
*' and in the case of aUia in the current laiiyuag&T The sami 
use of the word will be found in Sutras, 6. 1. 181 ; 7. 2. 88; 
8. 2. 98.«»» 

Yiiska is supposed by Professor Miiller to hare lived in 
fourth century B.c"* If this view be oorrect, and if we adopt (aa 
there appears good reason for doing) the opinion erpressed 
Professors lAssen and Benfey (see last section, pp. 142. 147.)tiutl 
Sanskrit had ceased to be vernacular in the time of Biuldlm, 
t. «. in the sixth century B.C., that language must have died oat 
some centuries before the age of Yafika; and a still longer 
period before the time of Panini, who, from the mo>re deve- 
loped state of his grammatical tenninology is considered to hare 
been later than Ya^ka. In this case, these authors conld 
scarcely employ the word hfiashd, when referred to Sanskrit, 
the sense of an universally epoken langui^e ; for the 
then actually in general use must' have been a species of comi] 
Sanskrit, though it may not yet have received the appellation trf 
Prakrit. But still the spoken language of that day had not 
departed so far from the Sanskrit but that its close relstioD 
the latter as its parent, or rather as its standard, woidd bd &n 
to every scholar; and thus Sanskrit would still bo called 
blulahd, or language par excellence. We have coasequentl; 
in the contmued use of this word, an argument of oonsid 
force to show tliat the Sanskrit had at one time been a spok' 

"' See Weber*« Indbcbe Literaturgeschichte, p. 167, note 2, wbenfufllMr 
insUuce* of the game kind in other ancicDt authors arc referred to. 

"♦ " Last Results of Sanskrit Researches," in Bunaen's Outlines of ibt 
Phil of Un. Hist., i. p. 1S7. 



n the Mnbabhashya (pp. 22 aud 63 of Dr. Ballautyne's 
e<fition) we find tLe following passage: — 

T^swJn":! ^ ^TOT ^r<?ii^ ^^m TT^ ^VJPi 

" Incorrect words are the most numerousj and [correct] words 
are the fewest; for of each word there are many corruptions 
(apabhvan^dfjL). Thus there are the following niimerous cor- 
niptions of the word gatih (cow) ; viz., gam, goni, gotd^ 0'^P'>- 
talikdy &c^ This reference to incorrect forms, 8uch as those 
of the word gau, which seem to be Prakrit' ^ appears to indi- 
cate that. Sanskrit w^as still spoken by a considerable class of 
persons, or was at least regarded as the standard of all spoken 
language; and tliat all deviations from it were looked upon as 
mere vulgarisms : for there would have been no ground for such 
a mode of comparison between words which were regarded as 
belonging to different languages; nor would the Prakrit syno- 
nyms of gaa have been wrong l)ecau8e of their variety of form. 

Sixth: — In the 164th hymn of the 1st book of the Rig-veda, 
the following verse (the 45th) occurs : ^p^iH^ ^T^ ^frf'TTTT 

1JJ-<lf^d <40*i ^r^T TJBJT <i<r--«lH "There are four 
measured grades of langtiage : with these intelligent Brahmans 
are acquainto<L Three hidden in secret make no sign. The 
fourth grade of speech is uttered by men." I quote part of the 
comment on this verse, wldch is given in the Pariiisbta, or 
Sapplement to the Ninikta, I. 9. 

*d*<lf^ ffrt'T >|Hlf<, M<lf'II ^4'l0 ^TT^rr^fTTg" 

"^ In the Mrichhakaf i, pp. 98, OJ), the word gona occurs in the sense "f 

M s 



jgTjf^ ij'l^MirwjH TioyinH«<^i<i:i ^rarft sti^ul H?f?n 
^ ^ ^T^ ^^ '^fJVT ^r*R^ ^^ ^t^ "#tfw ^nj^ 

^I^^g^: I ?nFT^ ^l^liT ^JH4f ^ ^^f^ m ^ 

" "\l\Tiat are these foiir grades ? The explanation of the rislii*^ 
is, that 'they are the four mjstie words, om, bhfdi, bhuvah, and 
mvar.^ The grannnarians"^ say 'they are the four kinds of J 
words, nouns, verba, prepositions and particles.' The ceremooj 
nialists declare them to be ' (Ist) the mantras ; (2Dd) the kalpaSai 
(liturgical precepts) ; (3rd) the brahmanas ; and (4tb) the cur- 1 
rent language.' The commentators {nairuktaJi) explain thetaj 
as being *(lst)tiie rich; (2nd) the yajush; (3rd) the sam&nl 
texts ; and (4th) the current language.' Others think they de-j 
note the speech '(Ist) of serpents; (2Qd) of singing-birds (?);] 
(3rd) of small reptiles ; and (4th) the current language.' 
philosophical school explains the four grades as having refer 

to 'beasts, "^ wild animals, and souL' On this poi 

we have also the following text in a Brahmana : * Speech, wh« 
created, became divided into four parts, of which three abide i 
these three worlds (earth, the atmosphere, and the sky), and ' 
fourth among the beasts. Terrestrial speech abides in fire and] 
in the Rathautara texts; atmospheric speech abides in the wind,] 

"• Se« tbc MBbabhoBhjB, pp. 28, 29. 
^'* I Bin unable to ducuver the meaning of tSnaveshit, the word left ImnJ 



aad in the Vamadevya prayers ; celestial speech abides in the 
Bun, in the Brihat metre, and in thunder. The [fourth portion 
of speech was] in the beasts. The speech which remained in 
excess [?] was placed in the Brahmanas : hence the Brahnians 
speak two sorts of language ; both that of gods and that of 
men.' " 

ITie Pari^ishta appended to the Nirukta is more modem than 
the time of Yaska, though it is regarded as a part of his work 
by Durga, the commentator, who refers to the Nirukta as con- 
sisting of 14 parts. (See his comment on Nir., i. 20, which is 
quoted below, pp. 175, fif.) But though it^ielf subsequent in date 
to the Nirukta, the preceding passage refers to the opinions of 
various ancient writers, and may, therefore, be held to carry us 
back to a remote period. Three of the ancient schools which 
are quoted, assert the current language {v^ynvaJidTikl vdk) to be 
the fourth kind of speech alluded to in the Vedic text as being 
spoken by men- By this we can only understand Sariskrit. It 
is true that in the Brahmana which the author of the Pari^ishta 
cites, a remark is mtule (connected with what precedes) that the 
Bralimans speak two languages, that of the gods aiui tluit of 
nten, and this might seem to prove that, as in later times, a 
distinction was drawn at the time when the Brahmana Mras com- 
posed, between Sanskrit, the language of the gods, and Prakrit, 
the language of men. But the reference may be to the Vedic 
and the ordinary Sanskrit; or to some unknown piece of 
mysticism. And, in any case, as we are ignorant of the date of 
the Brahmana from which the citation is made, no conclusion 
can be drawn from the passage adverse to the vernacular tise of 
Sanskrit in the Vedic age. 

Seventh : — In the Ramayana several passages occur in which 
the colloquial use of Sanskrit is mentioned. These are the 
following""*: — 

"" For the references to most of the texts here quoted I am indebted to 
Wflber, Zeitachr. dcr Deutschcn Morgenl. GcsellschalV, for ldS4, p. 8<il, 

M 3 



[ciiAr, u 

Hanuman, the luoukey general, is represented as having found 
his way into the polaoe of Ravana, the Rakshasa king, and as 
reflecting how he is to address Sita, who is there confined. 
He eays (Sundara Kanija, xxix. 16, 17): '^^«1I^14lfe/^|fy{ 

^ ^T^ ^tyipH f^^lfdfi-^ 4t*dIH^I ^<4HIH-^ 

^T^ ^TPnft HTf^TT^ r{\ urn fT^^imm rn ^wm4 

** I shall console her, whose senses are OTerwbelmed with this 
griefl But I am both unknown, and aboye all a monkey. If I 
were to speak in polished (aanekritdm) language, like a twioe^ 
bom man, Janaki (Sita) perceiving mj appearance, and [bearing] 
my words, would think that I was Ravana, and would again 
become terrified; and would scream in consequence of her 
fright.** Considering that this would lead to a diacorery, be 
concludes as follows (verses 33 and 34 of the same section): 

Iff^fn II "Anuoimcing [imitating?] by signs the undaaiited 
Bama, I shall address to her such polished (sondbfitam) 
Ungoage aa an [ordinaiy?] man would [Thus] I shall not 
occasion her any alarm, as her mind is fixed on the tboiighbi 
of her husband,'' 

As the reason assigned in these passages for not addreaaing Sita 
in Sanskrit such as a Brahman would use, ist not that she would 
not understand it, but that it would alarm her, and be ^tntmitaHiy 
to the speaker, we must take them as indicating that ft > M« fc »«* j 
if not spoken by women of the upper classes at the time what 
tbe Bamayana was written (whenever that may have beeu'")^ 
at least understood by them'", and was commonly spoken by 

"■ LaMen, ImL Aiu, voL L, pp. 4M, fL, doea not detenaise iu date. 
'" In Ute MfichhAk&(i, howerer, written protwUjr not much late 






of the priestly class, and other educated persons. By the 
Sanskrit proper to an [ordinary] man, alluded to in the second 
passage, may perhaps be understood not a language in which 
words difierent from Sanskrit were used, but the employment 
of formal and elaborate diction. 

On the other hand, an expression occurs in the Aranya Kan<Ja, 
Jtvi. 14, from which it would ratlier seem as if the use of Sanskrit 
was a characteristic of Brahmans ; and no doubt they were the 
persons who chiefly spoke it : %||<^i|«1^ "STT^^ 4^Mf^W<H: 

" Aasuming the form of a Brahman, and speaking Sanskrit, the 
nithless Raksbasa Hvala invited Brahmans to a funeral cere- 

In the Sundora Kanda, the discourse of Prabasta, one of the Ra- 
kshasaa, is characterised as ^'^rT ^fJ^TFfT^JJrq?^ '«4'{'T»«<I^H 
" polished (sanshiritam), supported by reasons, and judicious in 
its purport"; and in the Yuddha Kfinda, the god Brahma is said 
to have adilressed to Rama a discourse which was 4Jf|^r| «i«i^ 
^T^'^TT^H^ VTT^ftJfl^ll " polisheil, sweet, gentle, profitable, 

d consonant with virtue." But in neither of these two pas- 

;e8 does there appear to be any reference to the special 
leaning of the word Sanskjrita. 

In the subjoined lines (Sundara Kanda, xviii. 19), the word 

woman's pronunciation of Sanskrit is spoken of as something laughable 
(p. 44, Stenzler's ed.):— »R? ^T^ ^^ff^ "^TW ^^ ^TT^rf^ 

4^^^|^Q II which is thus translated by Professor WOson (Theatre of 

the Hindas, i. 60): — "Now, to me, there are two things at which I cannot 
diooae but laugh, a woman reading Sanskrit, and a man singing a song ; 
the woman snuffles like a young cow, when the rope is 6rst passed through 
her nottrils." 

M 4 



aanskdra ia employed, if not in a technical signification, as 
equivalent to eanskrita, at all events in a manner which euablea 
us (as Weber observes) to perceive how that technical sense of 
the word arose:— ^:;%^ ^%i ^^TT '?^'?T\ ^T^TTTTini: I 

WlfT'^j ^^iy*iwi ■^TfSrer ll " Hanuman, Son of the Wind, 
recognised Sita with difficulty, standing, as she was, unadorned, 
radiant only vrith her ovm brilliancy: just as a word is not 
readily understood, when its sense is changed by the want of its 
cort-ed grammatical furm." 

Eighth : — From the researches of M3I. Kuhn*" and Benf^'" 
it appears that many words, which in modem Sanskrit are 
only of one, two, or three, &c., syllables, have, in the Veda, 
to be read as of two, three, or four, &c., syllables, ue. as of one 
syllable longer, in ordei' to make up the ftill length of the linea 
required by the metre employed by the Vedic poets. Thus ttoaim 
has to be read aa ttiami r^t/usktau as viushtau; turyam m 
turtyam; maHydya as viaHidya; varenyam as varrniain; 
amdtyam as amdtiam; su-adhvaram as suadhvaram ; and 
evasiibhih as auastihkih. Now as this mode of lengthening 
words is common in Prakrit, it would appear that the Prakrit 
pronunciation agrees in this respect with that of the old San&krit 
in contradistinction to the more recent. But aa the Prakrit pro- 
nunciation must have been borrowed from a previously existing 
popular pronunciation, which was at the same time that em- 
ployed by the Vedic poet«, we find here another reason for con- 
cluding that the old spoken language of India, and the Sanskrit 
of the Vedas were at one time identical."* 


'** Zeitschrift fiir die Kunrtc des MorgenIiiiid«8, iii. 80. 

'** Sama-vcda, Introduction, p. liii., ff, 

"' I quote some remarks of Benfey, Saraa-V. Introd. p. liiL : — " The 
eily for frequently clionging the liquids g and r Into the correspondent voweli 
I and II, imd been remarked by tbe Indian writers on prosody, who teach that, 
wherever the metre requires it, ig and un should be read ioatead oF^ aad o, lo 



X. — Varioun stages of Sanskrit literature, and the differeni forms in 
which they exhibit the Sanskrit langitnge : the litter Vtdie eommentaturs : 
earlier expounders: the Nimhta: the Brdhmanas : the Vedie hymns: 
imperfect comprehension of them in later times from changes in the 
lan^piage : the hymns composed in the vertiactdar idiom of their age.^^ 

As I have shown in the preceding section that Sanskrit was 
once a spoken language, it must, in that its earlier stage, have 
been exposed to the mutations to whicli all spoken languages 
are suhject. from their very natiu-e. .Sanskrit must, in the course 
of ages, have become very different from what it originally was.'" 

many words the former mode of writing appears to Iiavc prevailed ; as !■ 
rendered probable by the differences of reading between the Sama-veda and 
the Rig-Tcda, the former, for instance, reading tugriya, tulhuvah, sndruoam, 
where the latter reads tugrya, subhvah, svdrvam ; and the latter, on the con- 
trary, rea<ling samudriya, where the former reaih samudrya. . . . But 
the necesiiity of nmking the change in order to obtain a rcailing conformable 
to the metre, U of such ordinary occurrence thut we are soon led to conclude 
that, at the timo when the Vedas were coniposeil, the liquids (y nnd o), which 
ap|)ear in (he Sanhiiiia as we now have them, bad not jet, for the most part, 
begun to be pronounced, but that, in their stead, the corretponding vowels, 
I and u were employed." On the other hand, y and v must sometimes be 
read instead of iy and ue (p. Ivi.) The fifteen verses of the Furusha Sukta, 
(cited in the First Part of ihia Work, pp. 6, ff), which are composed in the 
AnusL^up metre, will be generally found to have the proper nuuiberof feet, 
if not in other respects to scan correctly, — if tlie preceding remarks be 
attended to. Thus in the first verse, line second, the words vritivd and 
atyatishthat must be read apart, and luH united by sandbi. Bhavyam (in the 
first line of the second verse) must be lengthened to bhaciyam; ryahrdmat 
(second line, fourth verse), to viakromat; sddhgu (second line, seventh verse), 
to sadhiyd; djyam (first line, eighth verse, tliongh not in second line, sixth 
rene) to djiam ; grdmyd^cha (second line, eighth verse) to grnmidicha ; 
vgadadXtJk and eyakalpaynn (first line, eleventh verse) to viadadhuh and 
tiakalpayan ; and rajanyah (first line, twelfth verse) to rajaniah. 

'** In revising (his section (composed originally in 1S58,) for the press, I 
hare had the assistance of Professor Mullcr's work on the Veda, which 
has enabled mc to make a few additions, and to modify some of my previooa 
'Ifew that the text of Futanjitli (Mababluishya, p. 104) may be cited against 



[cBAr. I. 

Aad in f^ci we find frvm the records of Indian literature, that the 
SftD^rit, as it is bnM^bt before ns in the different S^astras, ba« 
gone through differeat ph tia a. The most modem is that iu which 
«« find it in the Porinas, Itihiau, and Smiitis. The Itihaaaa and 
Puranas are undoubtedlj not to be ranked with the oldest Sanskrit 
-writings for they all impiv that there were many older record* 
of Hindu antiqnitj existing wha» they were compiled, and often 
quote vaiioos ancient xtxaea.*** The Mahabhaiata frequently 

eternal; mkI io tke caae of ebenal words we most Lave immutable iml 
iaiaovtabl* iettert, frae frooi dSainntion. or increase, or alteration." Bat 
tlie wmxk wbi^ Bhiskara Atharfj* Applied to sstronomj are OjnAlly *\^i- 

cable to frmBiniar:— ITW »Tf«lM<<*^W '^W^^TTW'm: 
l^limiU N "In this astroaoBkal deportmeat Kriptare is onlj aotbori- 
liiiTe when it n supported bj demoDstration." This it true, obo, of all otbcc 
BMtiers which, like Gnunmar, come within the sphere of scLeuce. 

'** Hiat thej are not all of one age is shown bj an enlightened lodiaa 
Fiandit, lawara Chandra Vidyasagar, in the Bengali preface to his l(ijnp&|ha, 
or Sanskrit scloctions, as follows : — 

<jnrT*i (^^RrTTif^ ^f^ ^rf^rfif -wKw.^ fri ^r?T«i ■Jmr^r? ?^ 

f^sj JT^ J«^ (^«r^t? ?r«f ^"^TS f?fw<f5 ^fsiyi if^tfs T*"?! V.^^ II ■ 

^ r f t ? ?fip^ CTW IT? "1 II 

" All the Furanas arc commonlj said to be the competition of VtdA* 
Tjiba. But the style of the difierent Furonaa is so Yorious that Ihej camwt 
be conceived to be the work of one person. After reading a portion of tb« 
Visbna-puriina, another of the Bbigavnta, and a third of the Brahmavnifiita^ 
puriiua, it \i difficult to believe them all to have proceeded from one fit- 
So, too, there is such a discrepancy between the stjle of lh> 



introduces old legendH with the following formula: ^^|UJ^'|-^- 
<«rft*in^ri151H ^4 fd«f*(^ 11 " Here they adduce this ancient 
narrative." (See Part I. p. 33.) In all these different classes 
of works, which, in their present form, are comparatively recent 
partfl of Indian literature, the Sanskrit language is substantially 
the same. At the time when even the oldest of these works 
were reduced into their present form, we must stippose that 
the Sanskrit had nearly ceased to be a spoken tongue, and 
had become gradually stereotyped as a polished and learned 
iangtiago, by the precepts of those grammarians who preceded 
Panini, as well as of that scholar himself and his successors. 
As the language which had thus been polished, improved, 
and fixed by precise grammatical nilee, ceased to be popularly 
spoken, it was preserved from any future changes. In this way 
the iSanskrit language has remained unaltered for two thousand 
years, till it has acquired the appearance of immutability ; while 
its antiquity, and the perfection of form which it eventually 
acquired, and has so long ret^uneil, have caused it to be regarded 
as of divine origin ; just as every science which has descended 
from a remote age, or even from a period comparatively recent, 
is regardetl by the people of India as supernatural.'* Prior to 
this era, however, and as long as it had continued to be commonly 
spoken by the upper clajsses, the Sanskrit had been liable to con- 
sftant fluctuations in the forms of its inflections. Accordingly, in 
the works which are more ancient than the Smritis and Itihtifias, 
we find various differences of grammatical form, and a style 
altogether more antique. This is to some extent the case in the 
Brahmanas and Upanishads, where we encounter a simplicity of 

MafaibbiraU and thnt of the Yishnu-purnna, and the other works mentioned 
above, that it cannot be imagined to be the composition of the same person 
by whom ihcj were written." 

'*» The philosophers Rumanuja and Miiilhwacharyja are called incarnations 
oT f^sha and Vaya (Wilson's tlindn Sects^ pp. 24 & 87.) and dankars 
Ackaryya is celebrated in tJic Vfihud Dharma-pttrana as an incarnation 
orVHshnu. (Colebrookc's Essuyis i. 109, lO-t.) 



Kpitax and a taotologj of style, together with many particleit, 
and some modes of oonstmction, which are foreign to the later 
works.'* Ilie firahmanas, however, are only to be regarded as a 
middle stage between the Vedic hymns (mantras) and the more 
modem Sanskrit. It ii to the hymns of the Rig-reda, most of 
which are separated by an interval of several centuries, even 
from the Brahmanas, that we must reeort if we would disooro' 
how wide are the differences between the Sanskrit in its oldest 
known form and its most modem shape. In these hynms 
we find various forms of inflection and conjugation wliicli 
are not to be traced in more modern ^Titings, and numerous 
words which either disappear altogether in later authors, or 
are used by them in a different sense. These hymns are, in 
fact, by far the oldest parts of Indian literature. That this 
is the case, is proved by the whole nature and contents of 
tlie other portions of that literature which is connected with 
those hymns. The hymns are the essential part of the Veda; 
all the other writings which bear the name of Veda are de- 
pendent on the hymns, and subservient to their explanation or 
liturgical use. This may be made clearer by beginning with 
the most recent parts of the literature connected with theYedsB^ 
and going gradually back to the oldest parts. 

First Two of the most recent commentators on the Vedas are 
Sayana Acharya, who lived in the fourteenth century a.©.,"" and 
wrote a detailed commentary called Vedarthaprakaaa, on tbd 
whole of the Kig-veda ; and Mahidhara, who compiled a ooni> 

"^ Thua e.g. any one wlio 13 rumitiar with modern Sanskrit will recagniM 
iu the puaagc cited from the Kaushilakr-bruliiuiina in the First Part of tbii 
work (p. 1 14-) a dissimiliu'ity of style. The separation of the particle abAi &«■ 
the Tcrb aMupa/, in the phrase ^tIvT ^T^TWT«T 'VH<4ii^ll "Hebecaiae 
superior to the SuudHsas," ia a remnant of the Vedic tisage. In moden 
Sanskrit this preposition would not be thus severed from the verb. 

'*' Professor Wilson's Rig-veda Snnhitii, Vol. I. IntrodacU p. xlfii. 
Miiller, Sanskrit Res. p. 137. Rnth, Introd. to Ntrukta, p. liii. refen 
hidhart (if nut Sajfana also) to the 16th ccnturj. 




mentary entitled Vedadipa on the Vajasaneyi Sanhita of the 

Second. In such works as these we find reference made to 
earlier writers on the Vedas, such as Sauiiaka, the author of the 
Vrilioddevata, Yaska, the author of the Ninikta, and many 
others, Avith quotations from their works. 

Professor Miiller "* divides the \'edic literatiu-e, properly so 
called, into four periods, which, in the inverse order of their 
antiquity, are the Sulra period, Uie BraJimana period, the 
Mantra period, and the ChJiandas period. The Chhandas period, 
during which the oldest hyrons preserved in theRig-vedacoHection 
were written, he supposes to have lasted from 1200 to 1000 b.c. 
Then followed the Mantra period, from 1000 to 800 b.c., in the 
coiirse of which the more recent of the A'edic hymns were com- 
posed, and the whole were gathered together into one Sankitd 
(or collection). Next in order was the Brahmana period, from 800 
to 600 B.C., during which the chief theological and liturgical 
ir&cta bearing this title were composed and collected. And 
lastly, we have the Sutra period, extending from 600 to 200 b.c. 
in which the ceremonial precepts of the earlier tradition were 
reduced, (by men who, however, were no longer, like their prede- 
rs, regarded a.s inspired,) into a more tangible, precise, and 
imatic form than they had previously possessed. The works 
of this period were not all composed in the concise form of 
Sutras, but some were in verse and others in prose. 

Among the latter is the work of Yaska, who (as we have seen, 
p. 162,) is supposed by Professor Miiller ("Sautikrit Researches," 
p. 137,) to have lived in the fourth century b.c. Yaska found 
an earlier work entitled Nighantii, made up of classified lists of 
Vedic, and partly obsolete words, existing in his day ; to which 
be alludes in the following passage, at the very com- 
mencement of his work (i. 1.): dVimiej: WTP^TrT: ^ 

••• See hi« " History of Ancient Sanskrit Literature," which has juat ap- 
peared ; pp. 70. 244. 249. 3)3. 445. 497. 572. 



record (sajudmiidya) has been LaudeU down, wbicli is to beex« 
pounded. Thb is called the .Vi^/iajjfa'.'a?.*' '** 

And again Hn L 20.^ : Ml^l»*f1W^TW ^^*l\ ^>JWS^- 

" The rlshis had an intuitive insight into duty. Th^, by I 
tuition, banded down the mantras to subsequent preceptors, vbo 
worr (U>Htitute of this intuitive perception. These later teacben,! 
dwlining iu the power of conimimicating instruction, have, forj 
facility uf ixnuprohi'iision, arranged this book (the Nighantu), the 
Vodft, and the Vculangas.'* This passage proves at once the priority 
of (he Ni^'biiutu to the Nirukta, and also the still greater antiquity 
of tbu byiuuB which form the subject' of explanation in both.'** 

'" On tbis the commentator Doi^icbaryja annotates 05 followa ; ^ ^ i'^- 

is, ihiit ' This saereJ record wliicb Lad been set lurth by risbis u u spociBWi 
of the mode of cxplftining the sense of the mantras, lias been formed into oM 
collection in five chapters.'" 

'^ Professor Roth in his Introduction to the Kirukta, p. xiii. remaHb 
thus on this passage: — " Here YiisVa ascribes the compilation of the umH 
collection of words and names which forms the basis of his ex])l&nation, ia 
an undefined way to an ancient trndition, not indeed dating from tlie earliest 
period, when faith and doctrine flourished without artificial aids, but friM 
the generations next to that era, which strove by amingeroent and writlBg 
to preserve the treasures which they had inherited. He further puts tki 
Naighantuka in one cla^s with the Tedas and Vednngas. By the e&mpotUti* 
of the Vedas, which Yaska here places in the second period of Indian histoiy, 
he cannot mean the production of tlic hymns transmitted by the p*bia, wUoh 
were always esteemed in India as the essential part of the Vedas, and wtn 
regarded in the same light by Tiska in the passage before us. All, therefore, 
that could he done by later generations was to arrange these hymns, and 
commit th em to writing. We find here a recollection of a comporatirrl; 

•icr. z.] 



The following are the remarka of Diirgucharyya, the com- 
mentator, on thiti passage (Nir. i. 20.) : 

7^i^Twrs«JMi«idtnh«i ^M<ij«< wwr^i «*-ni^:i ^ 
f^ ^>«f Tf^ri ^s^T*9 '^TT^^T firwt^rsiTRnr?rr "^wf 

Ute rciluction into writing of tbc mental productions of enrly ages, on 
erenl which haa not yet attracted sufficient uotiue in its bearing u|K)n the 
hiftory of Indian literature ." 




[cDAr. I. 

HfHVT Wiy^l M.«»airj VT ^IW*4^I ^^^VT Hl*<^< l 

^^ ^¥^^<ri ^if^T^*ii ^r^rrgtft ^r^rsrr t^^ 
\7T^ f^^fiif^ *s*t^ T^^ fwm ^<T*ff i*^ wfr ^IT^ 

" They to whose minds duty was clearly present, i. «. by whom 
through eminent devotion it was intuitively seen, were the 
persons described by the term «aA«/ia<-^7*i7<x?/ian7uf»ia*. 
Agaiu, who were they? The rishia; who aie called 6o because 
they flmp (ri^hanti) ; because from a particiiliir ceremony accom- 
panied by a mantra of such and such import, in a certain waj, ■ 
such and auch a reward results. And the author will afterwards 
declare that the word rishi comes from 'seeing,' {ilar^andt) 
Henoe the faculty of seeing that such a reward springs from 
such a ceremony, is epoken of in a metaphorical way in the 
words, 'who have an Infuiltve vicfio of duty:'' for duty 
cannot be seen, being something entirely invisible. But what of 
these riahis ? ' They handed down the mantras by oral tuition to 
subsequent men, who ha<i not the same intuitive perception of 
dnty,' ?'. e. those rishis who had an intuitive perception of duty 



handed down the mantras to subsequent men, i, e. to secondary 
rishis {Hrutdrsh is) of a later age, and destitute of power; rishis, 
whose rishihood arose from what tliey hatj heard from others, 
and not vnthout hearing, as was the case with tlmse earlier rishis 
who bad an intuitive perception of duty. What did these earliest 
rishis do? They handed down the mantras by tuition (viz., by 
the function of instructing their pupils) according to their teit 
and meaning'"; and the pupils received them through tuition. 
Then 'these later men, declining in their capacity to instruct, 
arranged this book and the Veda, and the Vetlangaa, in portions, 
for facility of comprehension.' Declining, he means, in their 
knowledge of the best mode of rendering theu: instructions in- 
telligible, grieved when their pupils di<l not understand, actuated 
l)y compassion tow.ard8 them, and having regard to the shortness 
of their lives, and to their power of comprehension, which was re- 
duced in correspondence with the times, they compiled this book 
Qthe Nighantu], beginning with ' gati,^ and ending with • (leva- 
patnijuSy in parts, for facility of comprehension. He next tells 
ns wliat works he means ; the Vedas, and the other Vedangas. 
But how did they compile these works? He tells us; listen: By 
tbe agency of Vyasa they arrange<l the Veda, (which being up to 
tliAt period one, was difficult to study, from its extreme magni- 
tude), in a number of different SdJckfls, for the purpose of eovsier 
comprehension. The Rig-veda was arranged in 21 inldMS, the 
VajiLsh in 101, the Sama in 1000, the Atharvana in 9 : and 
nrailarlj the Vedangas ; grammar in 8 books, the Nirukta in 14, 
ud BO on, in order that they might Ix' apprehended in a divided 
, Aate ; ». «. that powerless and shortUved men might easily lie able 


to understand these several t^rcMrl^, when divided and of limited 

t. He now explains the word 'bilmia,^ B'Uma = 

I, means the division of the Veda^B, and division Htands for 
arrangement. Or it means Uuuanavi, elucidation; i.e. the 

'*• So MiUler (Anc. Sansk. Li(. p. 522) rcnJprs gruntftuto arthatrtifha, 
ilcnjring tu the woril graiUha the sense of wrilUii hiHik. 




caAr. bl 

sense of the Vedaa becomes clear from a knowledge of thol 
vetldiiyas, or supplements to the Veda. Tluis bilma is from tb«| 
root bhUt, or the root bhds. In this way this Nirukta ^ngtra^j 
and the other Veddngas have descended from the ri&hls. ThiiaJ 
the scripture has been elucidated." 

The Nighantu, the list of words which forma the subject of the \ 
pfeceding remarks, was prefixed by Yaska to his own work, the 
Ninikta, in which he endeavours to throw light on the obscurities 
of the ^'edii.'^* Wlien this work of Yai^ka was written, and even M 
at a much earlier period, it is evident that the sense of rnnny of ■ 
the Vedic words had been forgotten. This appears from the very 
fact of such works as the Nighantu and Nirukta being composed 
at all. For what occasion was there for compiling vocabu- 
laries <jf Vedic words, if the sense of these words had con- 
tinued all along familiar to the students of the Yedais? The 
necessity for works like his own is argued by Y'aska in tb9 
fo^liiwiiig paA-iape (Nir. i. 15.) 

" Now witliout this work the meaning of the hynina cannot 
be understood ; but he who does not comprehend their meaning, 
cannot thoroughly know their accentuation and 


"• "The NiphiiiiVu." siiys Professor Ro(U, (Introd. «« Niruku, p. liL), 
" cjipccirjllj the second portion of it, WM a coUeciiou of difficult and obtoleta 
worils, Mrkicli fonnvd a basis for instructioa in the mode of ezpoaai 
the Veda, sucli as was usually given in tbe sdinoU of tbe Brahmans. Al 
that period do need was felt of continuous commentaries; and in fact lean* 
iii|t hnd not then become separated into so many branches. A memamtdam 
of the terms denoting the ideas of most frequent occurrence in (be y«dik , 
and of the princiiial passages wliich required elucidation; a sfanplc list of fl 
the j?od« and the objects of worsliip, such as we find in the Nighanfo, ■ 
iuffio«d as a manual for oral instruction. At a later era thia aunil 
b«camc the »ubj.ct of formal and written explanation. To tlda pviad 
yMhrf tl)« Niniku." 

iECT. X.] 



forms. Therefore, this department of science is the complement 
of grammar, and an instrument for gaining its object." '" 

Tlie name thing is also clear from many passages in his work, 
in which heatterapts to explain Vedic words by thtiretymologies,'" 
(a process, often tentative, which would have been tmnecefisary 
if their meanings had been perfectly known), or in which he 
cites the opinions of dififereat classes of interpreters who had 
preceded him, and who had severally propounded diSerent ex- 
planations. This further shows that in Yaaka'a time the signi- 
fication of the hymns had formed the suliject of investigation 

'" This passage is translated bj Roth, Nirukta, Erliiuterungen, p. 1 1. And 
Saj-ana sajs in the Inlrod. to bis Coiumentary on the Rig-veda, vol. i, p. 39. 

tim\^ '^4HJl4< tVT^ "^M*^^ U\^*{ \\ " Hence the Nirukt* 
if aerviceabic for the understanding nf the meaning of the Veda." 

•** See Roth's Ertiiuleningen to Nirukta, p. 219. ff. "■Ve<lic interpretation 
could impose on itself no greater obslniclion than to imagine that the Indian 
commentators were infallible, or that the/ had inherited traditions which 
were of onj value. Even « supcrficiAl examination shows that their plan of 
iaterpretation is the very opposite of traditional, that it is in realitj a gram* 
maticol and etymological one, which only ngrcos with the former method in 
the erroneous system of explaining every verse, erery line, every word by 
itself, wiliiout inquiring if the results so obtained harmonise with those derived 
from other quarters. If the fact that none of the commentators are in pos- 
•euioB of anything more than a very simple set of conceptions regarding, e.g, 
the functioiu of a particular god, or even the entire contents of the hymns, 
wrhioli ifacy are continually intruding into their interpretations, be regardeti 
as a proof of their having inherited a tradition, it will at least be admitte<l 
that this poverty of ideas is not a llifng which we have any reason to covet. 
In this set of conceptions are included those scholu^tic ideas which were in- 
troduced at an early period indeed, but not until the hymns had already 
become the nibject of learned study, and the religious views and social 

ctrcunistances on which they are based had lost all living reality 

What is true of Sayann, or any of the other later eonimentatort, applies 
CMcntiaily to Ya«ka also. He, too, is a learned interpreter, who works with 
tlte materials which his predecessors had collected, but he possesses an in- 
calculable advantage, in point of time, over those compilers of detailed and 
continuous commentaries, and belongs to a quite dlHeronl literary period; 
viz. to that when Sanskrit was si ill undergoing a process of natural growth.' 

K X 




by learned men of different schools for many ages preceding*' 
The fallowing passage will illustrate tim, as well as afford soma 
insight ir»to the subjects and maimer of discussion at the fieriod 
when he lived. In the Nirukta, L 15, 16, he thus (in oontioua- 
tion of the passage last cited) alludes to the heretical opiniona 
•which had been entertained by his predecessor Kautaa, regarding 
the value of the hymns : 

nf^ iHr^irar^m ^rt§^ ^T^?f!f^ ^^:i ^pt^^ 

w^ »fif^i 'v^i^ «rr^^ <^M4iMiv<i f%ifr?T^ 

^^mT w(f^\ ^t^ry' ^^ny-iAt^i '©fw^ ^ N^- 
f?:c?Tr? r^nr^i -^rqrf^ urdftf^i'^T vi^f^\ v^ Tj? 
^^s?rf^ ^ f%ff^:i ^«i«iifir ^^4^iR» ^ ^^t 

«iHi«j ^rg^^frfTTi ^iTPJirfi ^frt^r: ^^frif^i ^rf^- 
f%^rTTr€t ^J^^ f5i^<<Hi|syT »T?nprftf?r wtf***«ifvi 




^TOT ^TOwrs^ irri[^:i ^Rf^r^ TT^tfiTi «wr t^ 
iRrr ^4<«i ^^MiHi: m«fl*if5Rrf^i ^fwt t^t^ ^f^- 

1 will, in my translation, place the answers of Yaska opposite 
to the objections of Kautsa (though thej are separated in the 
text) and tkua economize space, as well as make the discussion 

Kautsa objects. 
1. "If the science of inter- 
pretation is intended to make 
the sense of the mantras clear, 
it is useless, for the mantras 
have no sense." 

Tdaka replies, 
1. " The mautraa have a 
sense, for their words are the 
same (as those in the ordi- 
nary language). A Brnhmana 
says, ' that is the perfect rite 
of sacrifice which is accord- 
ing to the form prescribed in a 
text. A rich or a yajna de- 
clares the ceremony to be per- 
formed.' (An example of the 
identity of the Vedic language 
with the ordinary speech) is 
this, ' krUantau,^ ilk,c., ('sport- 
ing with sons and grandsons'). 

■• See Dr. Roth's translation of tliis passajje in the first of liis Abband- 
InBgen, p. 21, and in his Erliiuterungen to the Nirukta, pp. 11-13. There 
■re. however, some ports of the pusiif^ of which 1 do not clearly under- 
•Uod the bearing. 

N .1 



2. "The propositions [in the 
hymns and textes], ba?e cer- 
tain fixed words, and a certain 
fixed arrangement;" [and so 
are mere arbitrary formula?. ?] 

3. "The znaatras hare the 
ritual forms to which they 
refer fixed and enjoined by the 
Brahmanas [and, therefore, do 
not admit of critical exposi- 
iton] : thus • Spread thyself 
widely out ;' and so he spreads ; 
' Let me pour out,' and bo he 

4. "They prescribe what is 
impracticable : thus, * deliver 
him, plant :* *Axe, do not in- 
jure him,' thus he speaks while 

5. " Their contents are at va- 
riance with each other: thus, 
* There exists but one Rudra, 
and no second;* and again, 
' There are innumerable thou- 
sands of Rudrns over the earth :' 
and, * Indra, thou hast been 
bom without a foe ;' and aguLu, 
' Indra vanquished a hundred 
armies at once.' " 

6. *' A person is enjoined to 
do an act with which he is al- 
ready acquainted : thus, ' Aii- 
rlrMa the li} lun to the fire 

fiat Kautsa's objections miut 
be more closely examined." 

2. " This is th« c&se in ordi 
nary language abo^ «. g. In- 
draffniy pUdputrau. ('India 
and Agni,' • father and son.*)" 

3. " This is a mere repetitioQ 
[by the Brdhma'na], of what 
had been already said [in the 
TTiatifra .■] (and consequently 
the latter must have had an 
ascertainable meaning. ?]" 

4. " According to tlie sacr 
tradition it mufit be underwood 
that no injury ia to be in- 

5. " The same thing oocure in 
ordinary language : thus, ♦ This 
Brahman is without a rival;' 
< The king has no enemies.'* 

6. " In the same way pcopl? 
are saluted by their name 
though they already know 
them; and the vuuUiuparbi, 



which is being kimlled.' [This 
ia said by the adhwuryu priest 
to the hotri. Roth.] '*" " 

7. " Again it is said ; * Aditi 
is everything; Aditi ia the sky; 
Aditi ia the atmosphere.'" 

8. " The signification of the 
manti-aa is iudistinct, as in the 
case of such words as amijid; 
yddriihMn, jdrat/dyi, kd- 

(a dish of curds, ghee and 
honey) ia mentioned to those 
who are well acquainted with 
the custom." 

7. "This will be explained 
further on. The same thing 
is said in common language, 
thus, ' All fluids, (or flavours), 
reside in water.'"'" 

8. " It is not the fault of the 
post, that the bliud man does 
not see it. It is the mau'a 
fault. Just as in respect of 
load usages men are distin- 
guished by superior knowledge; 
BO too, among those learned 
men who are skilled in tradi- 
tion, he who knows most ia 
worthy of approbation." 

Durga, the commentator, does not enter on a detailed erpla- 
natioo of this passage. He merely refers as follows to its 

gen«ral scope : — - 

Mf*4HI MiM^^rR: Mrtj^l^a f%<T»M^ it^TWT^J 

'" See MuUer. Hist, of Anc. Sansk. Lit, p. 472. note 1. 

'*' Compare Rogbu Vania, x. IG. 4. 4 i ^d 4, IHil 4l ^ ^ i^ljj) 

M^ S^ffH "'Ab rain water, wbicb Las but one flavour, [wbea it ha* 
fallen] imbibes other davoiin," &c. 

R 4 



UcHAF. ». 

^IHMH4444 H'^Mil fkr^ flpg^l ^ THRT^TW- 

" The student being supposed to Lave an occasion and a right j 
to enter on the study of this Sastra, and the proposition haviog | 
been laid dovni that without the NLrokta, the sense of the man- 
tras cannot be understood, Kautea adduces many reasons for 
declaring the mantras to have no meaning, and on these he 
groimds an assertion, that the Nirukta ia TiselesB. Yaska in 
reply states the reasons on the other side in support of the man-j 
tras having a meaning, which point he accordingly estaUisbeft. | 
And as this work (the Nirukta), which is being commenced, is J 
useful for the explanation of their meaning, its utility is demon- 
strated. Tlius under the guise of an author who gtimulatesJ 
[inquiry by raising difficulties] an opportxmity is taken of stating] 
the arguments on lx>th sides, with the view of increasing the 
student's intelligence. For how is that student, of immature I 
intelligence, ignorant too of reasons and conclu-sious, when h» 
encoimters difficulties connected with the proper explication of] 
words, and is even hindered by other persons, to explain with-j 
out perplexity the meaning of words and sentences ?" 

It would seem from this that Diu-gacharyya looked npoBJ 
Kautsa as being merely a man of straw, into whose month | 
objections against the significance of the Vedas, were put by 
Yaska, in order that he might himself refute them. It does not,| 
however, appear why Kautsii, whose name appears in the old 
genealogical Hsta of teachers in one of the Brahmanas (Mullw, 
Ancient SansL Lit. pp. 181. 442), should be viewed in the ligbtuf 1 
a fictitious " Devadatta," any more tlian any other of the numer* j 
ous eurlitT writers referred to in the Nirakta. There seem5 tn ] 
bo uo other reason than tliis, that Durga did not, perha{>8, wisb 



liis contemporaries to believe that there had been in early times 
any old grammarian who fipposed the authority of the Vedas. 

In Nirukta ii. 16., Yiiska refers to the opiuions of various 
former schoola regarding the meaning of the word Vritra : — 

fwr:i ^^-^ wttwg f^rsTtHN*4^in ^^wn ^r^mi 
rrmiTT^ ^44*^\ »?^f^i ^!rfw^ ^ w^wh 
«ii^^Ni<nyi Ut^i 'n{\\*m ■^t?Ttf% fimK<<l^*KI 

*' WTio waa Vritra ? * A cloud/ says the Nairuktaa : • an Asura, 
son of Twashtri/ say the Aitihasikas. The fall of rain arises 
from the mingling of the waters and of light. This is figura- 
tively depictwl as a conflict. The hymus and Brahmanas 
describe Vritra as a serpent. By the expansion of his body, he 
blocked up the streams. When it was destroyed, the waters 
flowed forth." 

In Nir. iii. 8, he alludes to the views of older writers, regaid- 
ing the Vedic word jjanchajanali ; — M^lf^l iW '^V^ ^T^^ I 

f^tll'^: ^^^Tf ^^Mi{«lJ<4: H "'Ye five classes of beings, 
frequent my sacrifice.' These five classes of beings are the 
' Oandharvas, Pitris, Devas, Asiiras and Kakshases,' say some : 
They are * The four castes with the Nishadas for a fifth,' saya 

In Nir. viii. 22, Yaaka thus speculates on the feelings which 
had led some of hia predecessors to interpret certain Apri hymns 
as addressed to other deities than Agni : — 



T frirnmii 


** * The praydjas and the anuyo^fcu (introductory and 
cloding sacrificial acts), hare Agoi for their deity,' says a Brah 
mana. Another opinion is that they have chhandaa (metre) for 
their deity. For a Brahmana says, 'metres are the praydjas 
•nd amtj/tifai.* [After referring to three other opinions, simi- 
larly Bitpported, Yaska goes on] : * Another view is that they 
have soul for their deity, for a Brahmana says, * soul is the 
praifdjas and anitydjasJ I maintain the opinion that the hymns' 
have Agni for their deity. The other views arise from mete 
devotion [to particular gods]. But why are these rarions view» 
put forward ? Because it is well known to be a precept that 
the person who is about to offer an oblation should meditate on 
the particular deity for whom it is intended." 

In Nimkta xii. 1, he states the different views which had beoi 
put forward regarding the gods called Aswins : — ^'^^f^^Tf^" 

" ' The Aswins are so called from their horses (aefwaw),' eays 
Aurnavabha. But who are the .Aswins ? ' Heaven and earthy* aay 
some J ' day and night,' say others : while others again say, * The 
B\m and moon.' ' They were virtuous kings,' eay the Aitihad' 



"* Sco Roth's Erlaut, pp. 220-221, for some remarka on tiese old inter-] 
pretcrs of the Veda. '^ Older expounders of the Vedaa in general are," kai 
■ays, "called hj Yaska simply Nairuktas; and when be notices any dideroM 
in the conception of the Yedlc gods, those interpreters who take the cuh«- 
meristic view are called Aitihoslkaa. In addition to the exposition of the 
Veda in the stricter sense, there existed also Liturgical interpretations of 
numerous passages, such as we fuid in the Brubmanaa and other kindred 
treatises, in which it was attempted to bring the letter of the received text 



la Nirukta xii. 19, he states the various erpositions givea of a 
passage regarding Vishnu: — ^^^ f^Tg ff^ f^r^RfJT fw^! I 

strude over all this Bpaco : thrice he plants his foot. This he 
doefl in order to his threefold existence, * on earth, iu the atmo- 
sphere, and in the sky,' says i^akapilni : ' At his risiDg, in the 
zenith, and at his setting,' says Aurnavabha." 

In Nir. xii, 41, we have another reftrence to the Brahnianas: — 

?fm^9|«Hf7r ■'^ KI^UIT^ I " The gods sacrificed to Agni 
(fire), with fire. 'Agni was the victim; him they seized^ with 
him they sacrificed ;' so says a Brahmana." 

We thus see that in various passages of his work Yaska refers 
also to the Brnhmanas ; they must therefore have been older than 
his time. 

The following is a list of the writers whom Yaska quotes, aa 
having preceded him in the interpretation of the Vedas ; — Agra- 
yaua, Audumbarayana, AurnavaViha, Katthakya, Kautaa, Krau- 
shtuki, Gargya, Galava, Charmasiras, Taitiki, Varshyayani, 
i^atabalaksha the Maudgalya, ^akatayana, iSakapuni, i^akalya, 
and Sthaulashtivi.'" 

The sultjoined passage from the 12th Sect, of the first Parisishta 

into harmony with the existing ceremonial. Such liturgical interpretations 
■re collcil by Yaska those of the Yiijiiikas, or 'persons skilled in sacrificial 
ritea.' Akin to theirs appears to have been the mode of interpretation 

adopted bj the Naidanos Under thia head we must probably 

understand that method of explanation, which, differing from the gram- 
aatiflal etjroologicg, referred the origin of the words and conceptions to 
WXMJWH which were in a certain sense historical. The Brahmanas and 
Upanishads abound in such historical or mjthological etjmologies, which are 
to be found in all ages and among all nations ; etymologies which their own 
inventors do not regard as serious, but which, from their connection with 
other ideas, obtain a certain importance in the religious system." 
>" Roth, ErUuler. pp. 221, 222. 


nisroKT A^D 

or supplement to the Nirukta (considered by Professor RoUi, 
Kir. ii. p. 208, to be the work of some author subsequent to^ 
Yaska), refers to the antiquity of the Mantras, and the qiuJifi 
cations necessary for expounding them. 

\J X vj 

"This reflective deduction of the sense of the hymns ia 
effected by tlie help of oral tradition and reasoning. The hymns 
are not to be interpreted aH isolated texts, but according to their, 
context For a person who is not a rishi or a devotee-, has no 
intuitive insiykt into their meaning. We have said before 
among those men -who are versed in tradition, he who is m 
learned deserves especial commendation. ^Tien the rishis wi 
ascending [from the earth], men inquired of the gods, 
Bhall be our rishi?' The gods gave them this science of reaaooi 
to serve as a rishi, and to deduce by reflection the sense of 
hymns. Tlierefore, whatever meaning any learned man dedi 
by reasoning, that possesses authority equal to a rishiV 

Here there is to be remarked a recognition of the necessity of 
reason as a co-factor, in the ascertainment of religious truth, « 
the definition of ceremonial practice. With this may be com- 
pared, the whole tendency of the Sankliya doctrine, whidi b 
virtiiaUy, if not avowedly, founded on reasoning; and Um 
usertion of Bhaskara (see above, p. 170, note), that in the mv 
thematical sciences, scripture, if unsupported by d 
is of no authority. 

The same confidence in the inherent force of the humaD iu- 






tdlect is exhibited by Bhaskam in another place, in tJjese 
memorable words: ^J^ V«l4i$di 3(rra«T T^^'fTT *Tf^*yf<f 

4tlM f%W^ •! ^iIhI) "Wlien, again, after a long period, 
there shall be a great distance [observable in the position of 
the stars], then intelligent men of like character with Brahnia- 
gupta and otlier mathematicians will arise, who, admitting a 
movement in consonrtnce with observation, will compose treatises 
accordingly. Hence the science of astronomy, being mmn- 
tainetl by men of great ability, shall never fail in time, though 
it has no beginning nor end." (See Colebrooke's Misc. Ess. ii. 

In the First Part of this work some passages have been already 
adduced from Yaska, regarding the origin of particular Vedic 
hymns which he explains. One of these texts relates to the 
Kishi Viswamitra, and another to the Kishi Devapi. See Part 
First, pp. 124, and 14.3, 144. 

Third. I now proceed to the Brahmanas, to which we have 
been led back through the ascending series of more recent works, 

the oldest expository writings on the Vedic hymns. They are 

nsequently later than tlie hymns, the most ancient portion of 
Indian literature. But while the other explanatory and pre- 
scriptive books connected with the Vedas, such as the gramma- 
tical and ceremonial Sutras, &c, are not regarded as having any 
independent divine authority, the Brahmanas, on the contrary, are 
conMi<lered an a part of the Veda itself. This will appear from the 
following passages from Sayana's commentary on the Kig-veda : 

Rig-veda, Miiller's edition, vol. i. p. 4- 



'* The definition of the Veda, as consisting of Mantxn 
Brahtnana, is unobjectionable. Hence Apastamba says:, io t 
Yajna paribhiishrk, ' Veda is the name applieii bo ilautra am 
Brabmana.'" Again: ITT^T^IW^^V T^ ^4*iHnQtijjf\- 

*<I^^Jl*^rti^i'j^^ ^ i nf^^lf^ l '** " it being admitted that th. 
are two parts of the Veda, viz. Mantra and Brahinana, as 
Mantra has l)een abeady defined, the definition of the Brabmaoi 
will be, that it is the remaining portion of the Vetlo." 

In regard to the Sutras and Smritis, the author of the Njavv 
mala-vistara says : — 

q Hn<J'l i m^HIHJ*il<l«1'» l ftj | T|«1lf<'<mir4-di : ^-^Hi ' 
Y^^ifT^T ^ Xf-s^J \i^ni[ ^ ^ qrr3^Tf<4J**I^Hd, 

mw^ii . • . ^^1 w^^tm 1^ "i i ^ i fii fi^nj 

" Some persona have asserted that the Kalpa-sutras and other 
•works designated by the names of Baudhayana, Apa«tamb«, 
A^waluyana, Katyfiyana, &c., and the Nigama, NirukU, and 
six Vedangas, together with the Smritis of Mauti and otlien, 
'^* fiig-veda, Muller's e<lition, vol. i. p. 23. 


niter. X.J 



are Buperhuman, because they impart to men a comprehension 
of duty, like the Vedas; and that they are not ta be suspected 
of dissimilarity to the Vedas, from the fact of their appealing to 
the authority of the original text; for the knowledge of duty 
which they impavt is independent, because it is admitted to be 
Belf-e^idencing. But this view is incorrect; for the inference in 
question is set aside by the lapse of time. These works are 
called by the names of men ; as, * the Sutias of Baudhayana,' 
'the Sutras of Apastaniba,' &c.; and these designations cannot 
properly be derived from the fact that these works were studied 
by those wboae luuxkes they bear, as is actually the case in regard 
to the Knthaka juid other parts of the Veda : for it was known to 
^jome of their coutemporariea at the time of the composition of 
Sutras and Smritis, &c., that they were then being com- 
! ; and this knowledge hiis come down by unbroken tra- 
dition. Hence, like the works of Knlidnsa and others, the books 
in question are of human origin. Nevertheless, from being 
founded on the Veda they are authoritative." .... And again : 
" It ia not yet proved that the Kalpa-siitras are part of the Veda ; 
aod it would require great labour to prove it ; and, in fact, it 
is impossible to prove it. For the human origin of this book is 
established by its name, and by its being observed to have had 
an author." '*« 

The Bralimana"?, however, as I have said, notwithstanding 
their antiquity, and the authority which is ascribed to them as a 
oonstitueut part of the Veda, are very far indeerl from being so 
old as the hymns. On the subject of these works Professor Koth 
makes some remarks in his Introduction to the Nirukta, p. xxiv. 
ff. ; which I translate with some abridgments.'" 

"The difference in contents Injtween the Brahm.ina8 and the 

^n ihe difTerence in authority between the ItrShniftnos nml the Siilraa, 
I MiiUcr's " Ancient Sanskrit Literature," pp. 75-107. 
•** I refer for further information to Prof Muller's Bcction on the Brih- 
In hU "Ancient San»kt-it Literature;" particularly to pp. .142, ffn 



[cmjir. V. 

Kalpa-books, if judged according to detached passages, might 
aj>f)ear to Itc very small and indeterminate, tbongh even at tirsfc 
Bight it is undeniable tbat the two classes of writings are easily 
distinguishable as regards their position and estimation in the 
whole lx)dy of religious literature. In fact, the differeuce l^etween 
them is most essential. Though both treat of diviae worship in 
its widest extent, yet in the Brahmana it is the subject of 
description in quite a different sense from what it is in tbo| 
Kalpa-^tra. The object of the latter is to represent the whola 
course of the sacred rites which have a place in any particular 
department of worship, e.g., it defines exactly which of 
priests present at a ceremony has to perform a part at eadi 
point of the sacred rite. This is a very essential matter in InJiau 
sacrifices. .... It is further prescribed in these works what 
hymns and invocations are to be uttered, and bow. As a nU(^ 
however, the strophes are indicated only by the initial wordi, 
and pre-suppose other collections in which they must have Ixxa 
put together according to the order of their employment in 

worship Finally, these works prescribe the time, th» 

place, the forms, of the rites of worship, with all the pn 
and following practices. In short, the Kalpa-boijks are comi>lete 
8yst<?ma of ritual prescription, which have no other oViject tlum 
to designate the entire course of the sacred cerenioninl witb iJl 
that accuracy which is demanded for acta done in the 
of the gods, and to their honour, 

" The aim of a Bruhmaua is something very dififerent. As 
name indicates, its subject is the hrahma, the sacred elrmeni iU' 
the rite, not the rite itself. Something holy, the conception ofj 
theDiviue, lies veiled beneath the ceremony. It has now obtaii 
a sensible form, which muBt, however, remain a mystery 
those t«i whom that conception is unknown. He only who kno« 
the Divinity, its manifestation and its relation to men, can ex- 
plain the significaUon of the sjTubol. Such an explanadi 
the Brahmana aims at giving; it proposes to unfold tlie 
of theologiaU wis«lom, which is hidden under the mode >( 



worship inherited from ancient times. From this cause arises 
the mysterious, concise, often dark, style of the language which 
we find in these books. Tht^y are, indeed, the oldest prose 
which is preserved to us in Indian literature. 

"An example of these symbolical explanations is subjoined, from 
the beginning of the Aitareya-bruiiinaJia : — At the commence- 
ment of certain sacrifices, clarified butttr is ofi'ered to Agni and 
Vishnu in eleven pktters. This is done by preference to these 
two deities, the Briihmana explains, because they embrace the 
whole paatheon, Agni as the lowest of the gods, (the fire of 
the hearth and altar), and Vishnu as the highest (the sun in 
the zenith) ; and thuB sacrifice is offered to all tlie gods in the 
persons of these two. Eleven platters are presented, though 
there are only two gods ; eight of the platters are claimed by 
Agni, because the gayatri, the metre sacred to hiju, has eight 
syllables ; three platters belong to Vishnu, because he traverses 
the heavens in three strides (the three stations of his rising, 
his culmination, and his setting). 

•* Such explanations may as frequently be the mere inventions 
of a religious philosophy (encoimtered by us here in its oldest 
form), which delights in bold parallels, and a pretentious exegesis, 
as actual recollections of the beginnings of the litiu-gy, in which, 
among a people like the Indians, we may reasonably expect to 
find delicate and thoughtful references, books will always 
continue to be to us the most valuable sources for tracing the 
beginnings of thought on divine things ; and, at the same time, 
sources from w^hich we may draw the most varied information 
regai'ding the conceptions on which the entire system of worship, 
aa well as the social and hierarchical order, of India, are founded. 
In proof of this, I will ouly refer to the lights which may bo 
derived from the seventh and eighth books of the Aitareya-brah- 
mana on the position of the castes, and on the regal and sacer- 
dotal dignities. The Brahmanas are the dotpnatical books of 
the Brahmahs ; not a scientifically marBhalled system of tenets, 
but a collection of dogmas, as they result from religious prao 




[CSAT. I. 

00%. ThtBfwtn not written as a oompleie ezpodtioa of the 
psiMfplw ct befief ; bat they are DeceaBarj towards suob an ex- 
{MJIwa^ bccaaae thej wen meant to explain and mtahlhh tht 
vhfola ciMtomary cereBunual of worship. 

*^lti>iaifHBhlBBSfet»yaReiTethttttheBrahinana8 ar« bt&eed 
lyaa • fa»«adli^^ tridrfy ^j amificd, and bigfaly-developed 
iQ^alMa of woniiipk Thtf further the practice of sacred institittioaB 
baa ailvaiHMd* the tan disdnctlj- are thoee who practise them cxmi- 
MMMM wf thair m i iaiiin g. Gndaallj, around the central poition 
«f ^ rfliB7> vUch in ite origin was perfectlv tmspntki 
•ad kiritffiM^ tlten giows up a ousb of mbordinate obeerr- 
aMM^ whk'h m pmyartMO a« they are developed in detail, 
bMUM Bion IfMaa^ eoBBeeted with the fondamcntal thowlrt. 
lb* ftm^ hmiiMiiBg ■am iDdependenti^ loses its sjrmbolkal p^i 
port. Tbtt tmiita ^wrufcip had alreadj reached soch a a(i^| 
lihMi lib* Ml^fHM• liAMliia Mi^ftted in the Brahmanas begin 
b» v«tlt «9a» iL Ab% as m aBAa other religiooB Bjatonsof 
•l»lki|iiil^» th* l A i i - wriiin » lailiiiil Ibal it is not relif^ 
^om ibflH^ tbai gire birth to fonns of 
>kia II Kywai ■■iiiiy> which (itaelf the prodncl 
«f i i% i n — fci liag. aMpved by, and become sabaenrient to, a 
BeaesptiuM of the Divne]^ banaMB^ in ila tarn, the parent of a 
men d efc lu pc d and final j defied ibeoiogy. Sach was tha ia> 
laAunof tbeBcihBagastotheciineBKvatalup. The Brahnvp 
doea not appeal to Ae dicta of the aaoad bjrmns as its own 
and most immediate aoorce, bat ratho* rests apon the c 
wffvTT***"'^^ and npoci the earlier coDoepdoos of that cerenumal 
The Aitareya-biahmana, for imttaaoWj from which I borrov 
details, appeals not only to authoritieB, (to whom, written esBh 
podtions are never ascribed,) such as the Bishi Srauta {rn. 1); 
Saiijata, son of Aralha (ni. 22); RAma, Bon of Mrigu (riL S4>; 
Maitreyo, son of Kushdru (viiL 38), dc, or to preceding acn- 
fices of the same kind ; but further, the whole foirm of itt l^ 
presentation is based upon the tradition of earlier 'iwtmn- Ik 
customary formula for this, which is continually recurring at the 


head of a new pasf^nge, is taddhus, 'it is further said,' or atho 
khulv dkiis, * it is moreover said ; ' aud frequent reference is 
made to difference of opinions ;' so do or say the one set of 
persons, and the others otherwise.' But I have never met with 
a citation of an older writing. 

" Taking all this into consideration, we may conclude that the 
Brahmanas belong to a stage in the religious development of 
India when the Brahmanical faith was full-blown. Those reli- 
gions conceptions and sacred usages, which, even in the hymns 
of the Rig-veda, we can see advancing from a simple and uncon- 
nected form to compact and midtifonu shapes, have now spread 
themselves over the entire life of the people, and, in the hands 
of the priests have become a power predominant over everything 

It thus appears that the Brahmanas, though they have come 
to be regarded as parts of the Veda, are yet in time far posterior 
to the hymns, and in fact could have had no existence or use 
without the latter, on which they are either directly or indirectly 
founded, and to which they allude in every page. Thus in the 
Brahmanas we have such expressions contimuilly recurring as, 
** Tlius did the pshi say;"'** " Hence this has been declared by 
the rishi ;" '^' " Rishis of the Rig-veda have uttered this hymn of 
fif)«en verses."'" And Siiyaiia has the fcllowingpossage in reference 
to the priority of the Rig-veda (Comment on R,-V,, Introd, vol. i. 

p. 2.): rnrr^ *j4^<'i«iif5j "srT^wrf'r^TfHf^S"^ f^raro- 

<I«I<4 <t?<1^ -^^SlJ^OflfTr ^^^^T^T^rf^l "And 
so the Briilimanas couiifoted with all the Vfdjis, in order to 
atrettgthon belief in their a^ertions, refer to the Rig-veda, say- 
ing, ' This is declared by the Rich.' " 

"• flmUp.-br. xiii. 5. 4. 5: d<«l<<i<^ '"efwr S"WR^ II 

•*> Weber's Hist of Ind- Lit., p. 1 18. IWht, & llotL Diet, .rub coce ^Lihi 

u a 



The Satapatha-brahmana, in a passage at p. 1 052. ofWeber s edi- 
tion (coiTespondiug with the Vrihadaranyaka-upanishad, p. 213)^1 
refer8 as fullows to a hytnn of Vamadeva in the fourth miuidal:i| 
of the Rig-veda: fi^df(^ ^IJI^f^f^T*!^: Hpi^"^ ^ 
^•]4^^<4M I " Wherefore the riahi Vamadeva in vision obtained 
this text, ' I was Manu.' " Again, the Kaushitaki-brahmaQa 
refers to Vasishtha, in a passage already quoted in the First Partj 
of this work, p. 114. Now, as VasiBhtha was a Vedic 
the author of numerous hymns, tUs Brfihmana must have 
later llian those hymns. I may refer also to the S^adyayana i 
Tandaka Brahmanas, as quoted in Part Fiiat^ p. 115^ to profB 
the game point. 

To illustrate the manner in which the hymns are qnot 
n the Brahmanas, I will only cite further a portio 
of the passage from the Aitareya-brahmana, relating 
story of bunal.isepa, which is given in ori^nal in 
Appendix to Professor Jliiller's Ancient Sanskrit Uteraturt^j 
pp. 581, 582: -^RI ^ l^ifllN t^T^I ^*<M t <r*<^ ^ HT 

C(4>1^l 'V 'Villi [When he saw the preparations made for 
immolation] " iSunali^epa reflected, * They are about to slay i 
as if I were not a man. I shall resort to the gods.' He 
dingly addressed himself to Prajapati, the first of the godfl» vitk 
this fu-A, (Rig-ve<ia i. 24. 1.) * Of whom now, of whidi of J 
the immortals,' &c. Prajapati said to him, ' Agni is the neanrt 
of the gods resort to him.* He addressed himself to Agni with 
this rich, (Rig-veda L 24. 2.) ' Of Agni, the firet of the us- 
mortals,* &c." lu the same way he is represented aa nddffwiqf 
to various deities iu succession the verses composing the remainAr 



of the 24th, and the whole of the 25th, 26th, and 27th hymns 
of the First Book of the Rig-veda, ending with the last verse 
of the 27th siikta; " Salutation to the great! Salutation to the 
little I " addressed to the Visve-devah. '" 

That the Brahmanas were separated from the hymns by a 
able interval of time, is manifest from the various conside- 
which are urged in the passage just quoted (pp. 191 — 5) 
from Professor Roth; who informs us, for instance, that the 
Brahmanas, besides alluding to texts in the hymns, appeal on the 
subject of the ritual to various preceding unwritten authorities; 
and states his opinion, that the " Brahmanas belong to a stage 
in the religious development of India, when the Brahmanical 
faith was full-blown ;" and that "those religious conceptions and 
aacred usages which, even in the hj-mns of the Rig-vetia, can be 
seen advancing from a simple and unconnected form to compact 
and manifold shapes, have now [in the Brahmanas] extended 
themselves over the entire life of the people," This process was 
no doubt one which required several centuries for its accom- 

And Professor Miiller says, (Anc Sansk. Lit pp. 432. 434.) 
" There is throughout the Brahmanas sxich a complete misunder- 
gtanding of the original iutentiun of the Vedic hymns, that we 
can hai'dly understand how such an estrangement could have 
taken place, unless there had been at some time or other a sud- 
den and violent break in the chain of tradition." And again : 
r" Every page of the Brahmanas contains the clearest proof that 
Ibe spirit of the ancient Vedic poetry, and the purport of the 
original Vedic sacrifices were both beyond the comprehension of 
the authors of the Brahmanas .... we thus perceive the wide 
chasm between the Brahmana period and that period by which 
*"' Wt TTS^ ^^ ^!rI^^: See MiiUer'a Anc. Sansk. Lit. 
pp. 413, ff: Prof. RuUi'ii Article in Weber'g Ind. Stud. i. 4G1: Prof. Wil- 
son's Article in Jour. R. A S. vol. xiii. 100, and translation of the Uig- 

Teda, i. pp. 69-71, 

o 3 




it is preceded-" The Brabmana period we have already ae*n 
173 aViove), is placed by him in the 200 yean following 
second Vedic period, that of the Mantras. 

As time still passed on, and a further derelopment of languid 
and institutions took place, the Vedic hymns became lef» and 
less intelligible; and owing to the growth of formal and acaxpa- 
louB ceremonial prescriptions, the application of the 8aci«d texti 
to public worship became more and more difficult. As a natanl 
consequence, the literature connected with the explanation of the 
Mantras, their pronimciation and their ritual uses, continued to 
augment. Then the different grammatical Pratieakhj-a aphoriiou^ 
the Srauta and Grihya ritual Sutras, the Nighantii and Nirukta 
were composed. These works, as we liave already seen, wbw 
the growth of several successive ages subsequent to the date of 
the oldest Brahmanas.^'* 

"s On Uiis subject Professor Roth remarks (Introd. to Kirukts, p. liL) 
follows : — " In Greece a similar state of things prevailed. There, with 
exception of Ilesiod (who never rose to tlie same degree of consideratiioo)! 
Ilomur was the only source of the highest knowledge, and preeminent]/ 
the book of the schools;— the book which gave the first occasion to gram- 
matical, and almost every other sort of fcience to devclope itself. In India 
the Veda occupies the place of Homer. It was to the Veda that the Bnh* 
manical people looked as the sole repository of intellectual culture. As ft 
sacred book it was the more naturally a subject of research to the kwael 
uian, 88 be was at the same time a priest, and it became the first prohlea to 
be solved by grammar,— a science which was far more commonly studied, ud 
at M) earlier period attained a far higher stage, in India than iu Greece. At 
the same time, the Yedo, both as regards its language and its sabject maMo; 
stood far further removed iron) the Indian of the two centuries immei&lalj 
preceding Buddha (700 and 600, n.c.) — in which the sacerdotal ijttim 
reached its climax — than Homer did from the Greek of the Periclcaa tnL. 
At that perio<l, or even earlier, were formed the collection of Uotueric 
which had become obsolete, — the y\u«aai ; while in India, the nig/kttmiat^a 
word which I conceive to he identical in meaning with yXutaum') had 
compiled to illustrate the Veda. In both cascsi the collections bad the 
origin; but in the short interval from Pericles to the end of the Alexaadrai 
era, the Greeks hud done more for the explanation of Homer than tks 
Indians could accomplish for the comprehension of the Veda, in the ioeg 




Fourth. When at length we ascend above the oldest of the 
Brohmanas and arrive at the still more ancient collections 
{SiinhUds, as they are called in Sanskrit) of the Vedic hjTnns 
themselves, we shall find even here distinct proufs of a dif- 
ference of age not only between the several collections viewed 
as aggregates, but also between diflFerent component parts of the 
same compilntiona. Of the four Vedic Sanliitas, the Rik, Yajiish, 
Saman, and Atharvan, the Rig-veda is by far the most complete 
and important collection. Before, however, proceeding to give 
some account of its contents, I must premise a few words about 
the other Sanhitos. 

(L) Although the Vedas were originally considered to be 
only three in number, and the Atharvan was not denominated 
a Veda, yet many of the hymns or incantations of which it is 
made up appear to be of great antiquity.'" 

That the title of the Atharvan to be reckoned as one of the 
Vedas u not so incontestable as that of the tliree others, will 

Beriei of ages down to the time of SSjana $u\d Mahidhara, in tlie 16th 
century a., d. The task of the Indians was, in truth, hj far the more dlA- 
ficult ; and besides, Indian scholarship lay under an incapacity of un- 
fettered movement. It was necessary for orthodoxy to deny the facts of 
history, and to discover only the circumstances of the present in the monu- 
ments of antiquity ; for the present was both unablo and unwilling to rest on 
Mf Other foundation than the traditions of an earlier age, surrounded as these 
"VCN with a halo of glory, and only half understood. The priesthood supplied 
the recju'ired authentic explanation, wiMou/ which the reader of those ancient 
books would never have found in them that which be k> easily discovered 
wiA that assistance. The spirit of the nation, which had been so injuriously 
treated, became accustomed to the yoke, and henceforward walked onwards 
in the track which bad been marked out for it ; men's feeling for historj 
became irrecoverably lost, and they consoled themselves with the harmless 
enjoyment which was still allowed them, of solving grammatical questions. 
We can therefore, at least, boast, by way of compensation, on behalf of the 
Indians, that they have far outstripped the Greeks in the department of 

"* See, on the subject of this Veda, Miiller's Anc Sanak, Lit pp. 88, 446 
C Weber's Hist, of Ind. Lit. p. 10., and Mr. \\'hitney's papers in the 
Journal of the American Oriental Society, iii. 303 0°., and iv. 224 IT. 

o 4 


A5D numosB or 

■ppov firoB ike UOammf eoaadentaoM. It is not 

in the waA vcne of the Piuu Afc SOcta (B.-V. x. 90,) quoted 

in Pkit Flat of tfaii Work, p. 7, S. 

The anae OHHBMi •■ Mitieealile in thf fnllniriTu; pawipi finiii 
the Chhiadopm PjaniAwr^ : VWHlfi tfl<lil«f^ ^BWrRfl^l 

crnrt diy«MMt TT^ TfTT^ ^ "^^ ^T^ 4^ 

^TTDQII Tiajspad burnt [or meditated upon] the worlds; 
and from them, bo borat, he dreir fiacth their easezioes, FSf8| 
from the earth. Air fitYMn the atmayhere, the San fiom the akr. 
lie bomt these three deities; and from than so bamt be drev 
forth their essences: Bik-t«xts from Fire, Yajuah-testi from 
Air, and Sama-textfi from the Sun. He burnt this triple Ki- 
enoe; and from it so burnt, he drew forth its essence^ the par- 
ticle BhvJk from the Rik-teits, BhuvaA from the Yajosh-tezK 
and Su>ar from the Sama-texts." 

A similar passage occurs in the datapatha-brahmana, xL 5. & 1. 
In the following verse (i. 23.), ALiQu ropeat^ the aoooont given 
in tlie Chhilndogya Upanisbad : ^flJ<4l^<R^^ "^^ TPfl 

♦' From P'ire, Air, and the Sun, he drew forth (milked) for the 
accomplishment of sacrifice the eternal triple Veda, disti 
aa Bik, Yajusb, and Saman." 

It is not, however, to be denied that the Atharvan is meni 
in other passages as a Veda: bi& e^g. in the Satapatha-bvvii- 
mana, xiii. 4. 3. 7 and 8, p. 984. '" Madfausudana Sarasrad, 



'^ Cblmodogya Uponishod in BiblioUs. Ind. p. 388. 
u« See Muller's Aac. Sansk. Lit. p. 88. 


author of the Prasthana-bheda, while he calls it a Veda, notices 
at the same time its difference in cliaracter from the other tliree. 

^l<lR**^HniMl<*w|«1 ^rtlTif^'St^ll TJ^I "The Veda is 
divided into Rik, Yajush, and Silmau, for the purpose of carry- 
ing out the sacrifice under its three different forms. . . . The 
Atharva Veda, on the contrary, is totally different It is not 
used for the sacrifice, but only teaches how to appease, to bless, 
to curse, tfec" (Midler, Sansk. Lit, p, 445.) In regard to this 
Veda, Mr. Whitney remarks : " The Atharva is, like the Kik, a 
historical and not a liturgical collection." It was, lie thinks, 
originally composed of only eighteen books. A sixth of tlie matter 
of which these books consist is not metrical. " Of the remainder, 
or metrical portion, about one sixth is also found among the 
hymns of the Kik, and mostly in the tenth book of the latter; 
the rest is peculiar to the Atharva .... The greater 
portion of them are plainly shown, both by their language and 
internal character, to be of much later date than the general 
contents of the other historic Veda, and even than its tenth 
book, with which they yet stand nearly connected in import and 
origin. The condition of the text also in those paffiages found 
likemse in the Rik, points as distinctly to a more recent period 
as that of their collection. Tliis, however, woidd not neceasarily 
imply that the main body of the Atharva hymns were not already 
in existence when the compilation of the Rik took phuse. Their 
character would be ground enough for their rejection and exclu- 
sion from the canon, until other and less scrupulous hands were 
found to undertake their separate gathering into an independent 
collection. The nineteenth book is . . . made up of matter 
of a like nature which hatl either been left out when they were 
compilefl, or had been since produced." (Joum. of the Amer. 
Orient. Society, iv. 254, 255.) The priority of the Rig-veda to 
the Atharva may also be argued from the fact that the rishis 




ca^r. 1.1 

of the hymns in the Rig-veda are referred to in the Atliarra- 
vcda as men of an earlier period ; in proof of which I may refer h 
to the passages quoted in the First Part of this work, (p. 13I,| 
notes 50 and 51.) It is true that the same thing is noticeable 
to some degree in the Rig-veda itvself, in some later bjiiiDS of 
•which the rialiis of earlier hynms are referred to by name. In the 
Atharva-veda, however, the namea so specified are chiefly tho«6 
of the more recent rishis, while many of the personages referred 
t^ in the Rig-veda hppear to belong to a more primitive age. 
(See Roth's Litt xind Gesch. des Weda, p. 13.) . In the former 
Veda, too, the Indian institutions appear in a somewhat raon 
developed state than in the Rig-veda. There is one point at 
least in which this development is visible, viz. in tlie caste 
system. The following extract from Weber's " History of Indian 
Literature," (p. 10,) will exhibit his opinion of the general dif- 
ference which exists between the Rig-veda and the Atharva-veda: 
'• Tlie origin of the Athar\'a-sauhit^i falls within the period wfaeo 

lirnljmauisra ha<I l>ecome dominjuit Many of the hjnutt 

which it contains are to be found also in the Rik-sanhita, but 
there they are recent int<>rpolation8 originating in the period when i 
its compilation took place, while in the Atbarva collection they ' 
are the just and proper expression of the present. The spirit of 
the two collections is entirely different. In the Rik there | 
breathes a lively natural feeling, a wai'm love for nature; while 
in the Atharva, on the contrary, thei'e predominates an 
anxious apprehension of evil spirits and their magical powers: 
in the Rik wc see the people in the exercise of perfect freedom 
and voluutiu-y activity, while in the AthaiTa we observe them 
bound in the fetters of tlie hierarchy and of superstitiou." 

(ii.) The SamJi-veda is a collection of separate texts to be 
chanted at particular parts of the sacrifice;'" which with th« 
exception of a very few^ are all to be found in different part* of 
the Rig-veda, especially tlie 8th and 9th Mandalas. In th« 

iM See MUller, pp. 472-3. 



Big-veda we find the entire hymns; in the first part of the 

■ fianuk-veda we find only isolated verses of those hymns, dis- 

I located from their natural cunnection, though in the second 

^part the extracts are connected, and of greater length. It is not, 

however, quite clear whether the Sama collection or the Rik 

collection may have been first made, Weber remai'ks (Hist of 

I«d. Lit. pp. 9, 62), that the texts of the Sama-veda frequently 

exhibit more ancient grammatical forms tlian those of the Rig- 

^Teda, and suggests that as the former contains no extracts from 

any of the later hymns of the latter, it may have been compiled 

before these later pieces ha<i \>een composed ; but adds that this 

point has not been yet investigated. IMiitney also leaves the 

question undecided. (Joum. Am. Or. Society, iv. 253, 254.) 

Midler, on the other band, says (Anc. Sanak. Lit., p. 457.) 
" The other two Sanhitas were more likely the production of the 
Drahmaua period. These two Vedas, the Yajur-veda and the 
Sama veda were, in truth, what they are called in the Kau- 
ethitaki-brahmana, the attendants of the Rig-veda."'" Hesupposes 
that the hymn.s found in the three Vedas were not " collected 
three times by three independent collectors. If so, their differ- 
ences would have been greater than they are." Tlieir actual dif- 
ference*? are rather those of {^akhaa or branches, he thinks, than 
of independent Sanhitas or collections. 

(iii.) Both the Sanhitas of the Yajur-veda are collections of 
sacrificial formulas in prose, as well as of verses which are partly 
extracttxl from the Rig-veda. Many parts of the Yajur-veda 
exhibit a more advanced development of religious ideas and 
observances than the Rig-veda. Professor Weber, the editor 
of this Veda, considers (Hist, of Ind. Lit. p. 10), that it " belongs 
to a period when the Brahmanical element had already become 
predominant, though it was still exposed to strenuous opposition ; 
and when at all events, the Bnilimanical hierarchy and the 
system of castes had been completely formed." The same writer 

"' drt4r</««<«.MQd<l "^M vi. 11. 




cnnr. T. 

tells UB (pp. 106, 107), that "the 30th book of the Vajaeaoejri 
Kinhita of the Yajur^veda, in enumeratiug the different classM J 
of men who are to be consecrated at tbe rur\i3ha/-medha, or 
Human Sacrifice, refers to the names of most of the Indian mixed 
castes, so that we may thence conclude that a complete consoli- 
dation of the Brahmanical system had taken place." The Vaja- 
saneyi-sanhita is divided into forty Adhyayas or sections, of which 
Weber thinks (Hist of Ind. Lit. pp. 103, 104), that those from 
the 19th to the 25th may be later than the first eighteen; 
while there is, he remarks, no doubt that the last fiiteen 
Adhyayas are later, and perhaps much later, than the rest of 
the collection. This is proved by this portion of the Sanliits 
being called a Kfiila, or supplement, both in the anuknunani or 
index, which is ascribed to Katyayana, and also in ]Viahidh8is*s 
Commentary on the Veda.'" A further proof of the posteriority 
in date of the last parts of the Yajasaneyi-sauhita is derived 
from the fact that they are not found in the Sanhita of the 
Black Yajur-veda, but only in its Brahmana or in its Aranyaka 
parts, which by their very character are in date subsequent to 
the Sanhita (p. 104.) Weber is further of opinion (pp. 44, 105, 
and 106), that the names (Isana and Mahadeva) assigned to 
the god Eudra in Adhyaya 39, in addition to those by whidi 
he is designated in Adhyaya 16, (where he is regarded as tbe 
divinity of fire, though adtii-eased by many of the epithets which 
were subsequently applied to the god Siva,) indicate a more 
advanced stage in the worship of the deity in question at the 
time when Adhyaya 39 was composed, than at the period when 
Adhyaya 16 was written. Professor Miiller thinks that " there 
is nothing to prove that tlie hymn book of the VajaaaDeyiss 
[the Vajasaueyi-sanhita] existed previous to their firahman&* 
[the Satapatha.] Sansk. Lit^ p. 360. 


'"• The worJa of Mabidlmra at the commencement of the 26lh adhyiji 
■re u follows : ^TT^ f^^f^TH'T^pfl II " The KMl«s are now U> bt 
explained." See also Miiller'a Saiuk. Lib p. i6%. 


(iv.) We come now to the Rig-veda-sanhita, which contains 
the most extensive collection of the most ancient Vedic hymns 
in their complete form. It is divided into ten mandalas, and 
contains in all 1017 hynma (Miiller, p. 497.) "The Vedas," says 
Mr. Whitney, (Journ. Am. Or. Soc, ill. 295), " contain the songs 
in which the first ancestors of the Hindu people, at the very 
dawn of their existence as a separate nation, while they were still 
only on the threshold of the great country which they were 
afterwards to 611 with their civiliaation, praised the gods, extolled 
heroic deeds, and sang of other matters which kindled their 
poetical fervour.'*' . . . The mass, as it lies before us, is almost 
exclusively of a religious character ; this may have had its 
ground partly in the end for which the collections were after- 
wards made, but it is probably in a fiu- higher degree due to the 
character of the people itself, which thus shows itself to have 
Deen at the beginning what it continued to be throughout its 
whole history, an essentially religious one .... Hymns of a 
very dififerent character are not entirely wanting, and this might 
be taken as an indicsition that, had they been more numerous, 
more would have been preserved to us." '*• These hymns are said, 
by later Indian wTiters, to have been " seen " by the ancient riehla 
or bards. Thus the Nirukta says, (ii. 11.): "^f^«^a(^l<1^| T^- 

. ^|€|(*f II *" A rishi is so called from seeing. He saw the hymns :' 
• This is Aupamanyava*s explanation. They became rishis, be- 
» Brahma, the self-existent, approached them when they were 
Bimk in devotion. From this, as is generally understood, they 
^»cqxiire<l their character of rishis," There is, however, no doubt 

•*• See ttUo, for an account of the contents of the hymn*, Profc«»or 
Wilson's Introduction to his translation of the Rig-vetla, p. xxiT. AT. 

•••Dr. Aufrccht remarks (lud. Scudicn. iv. 8.) that "possibly only a 
■all |>ortion of the Vedic poems niny hare been preserved to us in the 




that the rishis were themselves the proper authors of ther^e 
ancient songs, which they addretised to the gods when they were 
solicitous to obtain any blessing ; or composed on other occasions. 
The scope of these hjrmna or mantras is well summed up in th« 
following passage from the Anukramaui (index) to the Rig- 
veda, quoted by Mr. Colebrooke (Misc. Ess, i, p. 26.) ^^5^^ 
W^^ ?«lr|l^»^lfHT>«lVT^II " The rishis, desiring [nk- 
rious] objects, hastened to the gods with metrical prayers," It b 
also said in the Nirukta, vii. 1.: ijr4i[^ -iiftl HmJ ^^- 

V^f^ll "Tlie hymn has for its deity the particular god to 
whom the rUhl, seeklnrj to obtain any particxdar object whiA 
he longa for, addresses his praises." 

For many ages the successive generations of these ancient 
rishis continued to make new contributions to the stock of 
hymns, while they carefully preserved those wliich had been 
handed down to them by their forefathers. "' The fact of this 

101 u ffjg Indian Aryaa were disposed to piety both bj their nstaral 
character, and by the irtBtitutioits of Mauu. They were sustained in theM 
sentiments by the chiefs of ccrtniu fannlics in vrhich their religious tr»ditioM 
had iK'fii more especially preserved. In thojie primitiTe ages the political 
■ystcin was precisely tlie same as that which llomer depicts ; — kingi tfac 
veritable Rlieplierda of their people ; cultivators or herdsmen united ■roaml 
their chiefs, and prcpired, whenever necessity arose, to transform theaiselvH 
into warriors; numerous lloi'ka and a profusion of rural wealth; towns 
which were only large villages. Some of these villages served as retreat* 
to renowned sages, who, while their dependants were tending their fiel<k 
Mid flocks, were themselves engaged in the cultivation of sacred scieaM, is 
the company of their sons, or their pupils, and fulfilled the fuDctiona of a 
Citlchiui or a Tiresius to some IndiAn Ai^nmeuinon or CEdipus in their dskIi- 
bourhood. Invited by the chiefs to perform sacrifice, tliey arrived with 
th«lr aacred retinue ; they ascended the mountain where an cnclasure of 
lattice-work had been constructed ; for temples were then unknown. ThfK, 
beneath the vault of heaven, they recited their hereditary songs, or a ne«tj> 
competed hymn ; tlicy invukcil the grand agents in nature to grant i 
to the labours of the field, increase to the flocks, and a racceasioo of ! 




ire compoflition of the hymns ia evideDt from the ancient 
idex (anukramani) to the Rig-veda, as coutinually quoted in the 
commentary of Sayana, which shows that these corajwsitions are 
ascribed to different generations of the same fiamilies, as their 
*' Beers." For example, some of the hymns of the 3rd mandala 
are asai^ed to Gtlthi, the father of Viswiiraitra, others to Viswa- 
mitra himself, others to Bishabha, his son, others again to Kata, 
his descendant, and others to Utkila, of the race of Kata. Here 
■we have the " seers " of hymns extending over five generations 
or more. The same fact, viz. that a long interval elapsed between 
the composition of the different hymns, is manifest from various 
passages in these compositions themselves. Thus the 2nd verse 
of the first hymn of the 1st ■mandala of the Rig-vetla, is to the 
following effect: ^rflf: ^fif -^f^f^rKW" ^^f^^l ^ 

^Wr 1^ W^f?! II " Let Agui, who is to he worshipped by the 
former rishis, arid by the recent ones, bring hither the gods."'** 

and virtuoTU descendant*. Tlicj implored, tbej tLreatened their gods ; and 
when the sacred rites bad been scrupulousljr perfornied, they retired loaded 
with gifts, corrjing away cows, horses, and cars filled with provisions, gold, 
and preciuns slufls. We see thus by what fortune these hj'inns have been 
preserved, forming as they did, a patrimony to certain families, a species of 
productive capital, which it wua their interest to turn to the very best 
•ccount. Com[io8ed on certain recognised and venerable themes, and some- 
times retouched and renovated by the imagination of a new bard, ihey grew 
old, as they were transmitlcd from age to age, bearing on them, sometiiiicf, 
the date of their composition, which was indicated by the name of the in- 
Bpircd author, or of some generous prince." Langlois, French translation of 
Kig-vcda, Vol. I. pref, pp. x. xi. See also Mr, Whitney's remarks in the 
Journal of the Am Or. Soc iv. 24£>. 

"' The comment of Yaska on this passage (Nirukta vii. 16) is as follows : 

IJ 7<4|fi{^|c|^f^f?T II " Let Agni, who is to be worshipped, rever^ 
enced, by the former pshis, and by us the more modem ones, bring the 
godB hither." SSyana annotates thus on the passage : ^ij^flj: ^f^' 




CB^rb u' 

There are many other verses alluding to a difference of antiqai 
in the hymns and their authors. Such are the following ( Ii.-Vi 
L48. 14): ^ f%f^ Hl4|J(«<j: ^ ^ajT?^ ^a^TH 
former riahis who invoked thee for saocour," &c, (R.-V. L 6: 

13.) ^srnm 'TTrm ttt ^ruttt^ v^ -^fr^rar^rr^i 

" Nodhas, son of Gotama, has fabricated this new prayer to tht>«^ 
India, who art eternal, and yokest thy coursers," SicJ"' { IL-V.^ 

iii. 32. 13.) jf: ^V^ 4^ qwfH ^ iPij^tH^ 

^r1*ffW: II "Who [Indra] has grown through praises, audent, 
middle, and modem."'" (R.-V. vi. 44. 13.): ^: i ^4lf*f4 ^<1 
^r1*llf* T ^rff^ ^T^ ^WrTT^^^WT^II " He [Indra] wh9^ 
has grown by the ancient and mmlern hynins uf the rishis wlitf ™ 
praised him," (R.-V. vii. 22.9.): ^ ^ ^ ^^^ ^ ^^ ^TBfJ 
T^ ^r^rrf^ ^•1<<*H fWffj: II " Indra, the wise riahis, hot 
ancient and modem, have generated prayers." (R.-V. x, 23. 6.) 

"The Vimadas have generated,'" O Indra, for thee, the lieue- ^ 
ficent, a copious hymn, before unheard." | 

In the Vajasaneyi-sanhita of the Yajur-veda, (xviiL 52.) we taeti 
with the following text : ^;^ ^ M^Nai<\ MdPl'^V ^TTW 


•ITWrf^r^f^ ^(H' I " This Agni, who u to be worshipped, t. e. eele^l 
brated, by the former, i.e. the ancient risliis, Bhfigu, Angiras, and (be r&t, 
and tiy the reuent, i. e. the present [rishia], ourselves also," &c. 

■^^li^^rt 11 Sayana. " Fabricated i. e. made for us this new brakmt, i 
praise in the form of this xukta.^ The same verb takth 'w aI»o applied to 
composition of hjmns in R.-V. i. 109. 1.; ii. 19. 8 ; and vi. 3'2. 1. 

4^1 ■:( I II " Praises ancient, intermediate, and formed by us at prcscat* 
This verse is referred to by Milller, p. 482. 

•" The verb ajijanan, "generated," as applied to hymns, also ooGort i»j 
R.-V. viii. 77. 4., and viii, 84. 5. See also Siima-ved«, ii. 108, 109, 
1059, with Bcnfcy's translation and note, p. 245. 



3|JM: ^HTT'ITr M 4, 1 Ul I ; II "But these undecaying feathered 
pinions, with whicVi, Agni, then slayest the Rakshases, wth 
them let iis ascend to the wurld of the rigliteous, whither the 
earliest-born ancient rishis have gone."'^ The writer of this verse 
was himself a rishi, and it is dear, according to his statement, 
that long before his time other risbis had gone to the regions of 
the blessed." 

And in the Rig-veda we find references made in numerous 
hymns to earlier rishis (who themselves are yet declared by later 
writers to have been authors of liymns included in the same 
Veda) having been delivered l>y the gods in ancient times. 
Thus Atri, the author of several Vedie hymns (37. 38. 39. 
40. 43. 85. and 86. of the 5th Mandala), Kanva, the author 
of hymns 37 — 43 of the 1st Maudala, and Vasishtba, the author 
of the greater part of the 7tb Mandala, are spoken of in several 
other hymns, e.g. iu M^nJala i. 112. 7, 9. 16.; i. 117. 3. 8.; 
i. 118. 7., as if they belonged to an earlier age. 

The descendants of the most celebrated rishis, would, no doubt, 
form complete collections of the hymns which had been com- 
posed by their respective ancestors. After being thus handed 
down, with little alteration, in the families of the original au- 
thors for several centimes, during which many of them were 
continually applied to the purposes of religious worship, these 
hymns, which had been gathering an accumidatetl sanctity 
throughout all this period, were at length collected in one great 

ly of Racred literature, styled the Sanhitdy of the Rig-veda 

-& work which in the Puranas is assigned to Vedavyasa, and 
one of his pupils.'" 

••• On lliia Uie Commentator annotates : Tf^^diC Tf^f|| it|i;||: 

< j<niii : jrrfi 'RT ^ai<m<i "^^^ ^r^ ^t^ ^irg: ii 

** Xltc world mLiiher the i'uHi\B,Jirit-boni, i. e. earlicst'produced : ancieut.i. e. 
in former times, joun^; im]H2rl:>Labk% nnd immortal, have gone." 
fic " I su{ipoiic thai at tlilltirent an<l unknown epoc-La, on the invitation of 





As the process of hymn composition continued thiis t«5 go oo 
for many centiuies, it was likely that the collection, when floalljH 
completed, would contain many comparatively new hymns, writttu^ 
just before the canon was dosed. Even after this latter event 
took place we find that some hymns were composed which mm 
hare had some pretensions to a sacred character, as, though ua 
admitted into the canon of the Big-veda, they are found copie 
as Khilas or later additions, at the end of some of the secticoe 
in the manuscripts of that work ; and some passages from them 
are, as Professor Jliillcr informs me, inserted in the other three 
Vedas, and ai"e enjoined by Asvalayana to be employed on par- 
ticular occasions, in the ceremonial of sacrifice. Whether or nu 
these Khilas are the oldest, extant compositions after tboee a>\ 
eluded in the Vcdic collections, (and their style shows them 
to be all 80,) they must at least, from the position which thef^ 
have gained of the Vedic apocrypha, be regarded as a link, 
connecting the Vedic hymns with the later parts of Indian 

The hymns in the Hik-sanhita which bear the most mode 
character, and which from theii- age stand chrouologicaUfl 
nearest to the Khilas just alluded to, are (according to 
fessor Miiller, p. 484.) those in which reference is made to 
complicated ceremonial, to a great variety of priests with dif-J 
ferent functions and appellations, or in which the liberality of 
royal patrons to the sacerdot-al class is the theme of eelebmtion. 
One composition of which the modem character is acknowledged 
by all critics, is the so-called Punisha Siikta, the 90th hymn of 
the 10th Mandala (quoted in pp. 6 — 11 of the First Part of J 


Boine prince, learned] and pious pcr&ons must Lave been charged to 
the Iiytnng composed for the use of the seTeruI sacerdotal fkmilioa, 
orrangi; tlicni in n certain order consistent with tlic tnaintenanceof tiic ieM» 
When we observe the spirit which has dircclcd these collectors, »c ca« 
comprehend how there should be bo many repetitions both ia the idea* aal 
the words. The ancient bards had borrowed from each other nutajthtn^ti 
which tie compilers of different eras hare scrupulouslj reproduced." Lant" 
loia, French translation of Rig-vcda, Vol. I. pref. p. xiii. 

SBCT. X.] 



work), which ]VIr. Colebrooke'^* characterizes in the following 
terms: — 

"That remarkable hymn is in language, metre, and style, 
very tUflferent from the rest of tlie prayers with which it ia 
associated. It haa a decidedly more modem tone; and must 
have been composed after the Sanskrit language had been re- 
fined, and its grammar and rhythm perfected. The internal 
evidence which it furnishes, serves to demonstrate the important 
fact that the compilation of the Vedas, in their present ar- 
rangement, took place after the Santfkrit tongue had advanced 
from the rustic and irregular dialect in which the multi- 
tude of hymns and prayers of the Vedas was composed, to the 
polished and sonorous language in which the mythological poems, 
sacred and profane, {puranas and ciivyas,) have been written." 
(See also the remarks made on this hymn by Professor Miiller, 
Anc Sansk. Lit p. 571.) The loat named author thinks it is a 
mistake to regard any hymn as modern, merely from the pre- 
sence in it of philosophical ideas. But I must refer to his own 
work, pp. 556, ff., for a statement of his views on this point. 

The sketch which I have now given of Sanskrit literature shows 
that we can trace it back, by a series of almost continuous links, 
up to tlie period of its rise.'*' If the Vedic hymns cannot be con- 
nected immediately with the literature which follows next after 
them, they are at least separated from it by no very distant 
interval ; and they are evidently the natural pro<luct of the same 
le Indian mind which aftemards gave birth to the Brah- 

*** MiM. Em. i. 309, note. 

'** It m*y, perh«ps, be iLnugliC that Uiis sulject hu been treated at a 
tk duproportioned to tlie purpose which I hnvc ini mediate) jr in view, viz. 
I' trace the mutalions of the Sanskrit language. But a full exhibition of the 
ehorncter and autiqiiitj of the Vedic hymns, and (jf the relation in whieb they 
vtand to the other parts of Indian literature, will be found to form a neccd" 
nry baiia for various other discussicus wliiek will appear in tbc sequel of 
tliis work, and I have deemed the present a convenient opportunity fur its 

r 9 




manas, the Upanishads, the Dar^nas, and tlie diiferent epic aoi 
mytlioloipcal poeius. 

In the Rig-veda we possess^ as has been alreadj remarked, 
a collection of hymns which were composed during many sui 
cessive generations, but its most ancient portions constitut 
the earliest of all the extant remains of Indian authorship, and 
not only display to us the Sanskrit language in the oldest ph&fe 
in which we can ever Bee it exhibited, but also afford us some ol 
the most authentic materials which we can ever obtain for oi 
researches into the earliest history, religious and political, of the! 
Indian people, and into their pre-historical relations with the 
other branches of the Indo-Eiuropean family. 

Fifth. If any further proof be wanted of the greater antiqoity 
of the Vedic hymns, as compared with the other books esteenn 
more or less sacred by the Hindus, as for instance, the epic pocmf' 
and the Puranas, it may be found in the great difference betwijen 
the mythological systems which are discoverable in theee two 
clafifles of works res]>ectively. As I hope to return to this sub-, 
ject in a future part of this work, I must content myself with > 
very summary notice of it at present. The following extracto 
from Professor H. II. Wilson's introduction to the first volume of 
Ills traualation of the Rig-veda, pp. xxvi. xxvii. will give 
idea of the difference to which I alhide: — 

" The next question is, who are the gods to whom the praisa 
and prayers [in the Kig-Veda] are addressed ? And here we fiml 
also a striking diflference between the mythology of the ^<J' 
Veda and that of the heroic poems and Puranas, The divinitirt 
worshipped are not unknown to later systems, but they t! 
perform very subordiuate part^, whilst those deities who are U* 
great gods — the Dil majores — of the subsequent period, »» 
either wholly imnamed in the Veda, or are noticed in an infcricc 
and different capacity. The names of s5iva, of MAHiDBVi, d 
DuRGA, of KiLi, uf Rama, of Krishna, never occur, as fv** 
we are yet aware: we have a Rudba, who, in after timet, tf 
identified with J^iva, but who, even in the Puranas, is of wfj 



douUfiil origin and identification, whilst in the Veda he is 
described as the father of the winds, and is evidently a form of 
either Aasi or Intra ; the epithet KAPAnnorN,'''''' whicli t3 applied 
to hioij appears, indeed, to have some relation to a characteristic 
attrib\it« of SrvA, — the wearing of his hair in a peculiar braid ; 
but the term has probably in the Veda a different signification 
— one now forgotten, — although it may have suggested in after- 
time the appearance of }5iva in such a head-dress, as identified 
with AoTXi; for instance, Kaparddin may intimate his head 
being surrounded by radiating fiaroe, or the word may be an in- 
terpolation ; at any rate, no other epithet applicable to f^rv'A 
occurs, and there is not the slighteyt allusion to tlie form in 
which, for the last ten centuries at least, he seems to have been 
almost exclusively worshipped in India — that of the Linga or 
Pkallua: neither is there the slightest hint uf another important 
feature of later Hinduism, the Trimurtti, or Tri-une combination 
of Brahma, Vishsp, and ^iva, as typified by the mystical syl- 
lable Gni, although, according to high authority on the religions 
of antiquity, the Trhnviiti was the first element in the faith of 
the Hiudtis, and the second was the LingamJ' — Creuzer, Re- 
Ugiona de VAntiquiU, book i. chap. i. p. 140. 

Even so late as the time when the Satapatha-brahmana was 
composed, the names afterwards appropriated to Mahadeva were 
applied to Agni, as appears from the following passage, i. 7. 3. 8. 

p. 70:-- ^rfir^ w <ii^Hrfi^ "ii^iiR T^ Tf^ "enrr 

"* [Tliis epithet occurs in tl>c following pijssage, verse ». of Siikta 114 
MaiultJa Ist. : — 

i.e. " We "ffiT tbcso praises to the tniglity Rudrn, with iLe braiJcJ |»>iir, tLc 
destroyer of heroes, in order ihat health may be ciijojwd \>j bipeds and 
qud'lru|x.<ds, ntid tlmt all beings in this village xoaj be (well nourisboil, nnd 
exctnpt from dieease." — J. M.] 


r 3 



[cn&r. I. 

f^TU^ "Srn^TfW^II " Agni is a god. These axe his luunes, 
viz., Sarwa, as the eastern people call him, aud BhaiKif as the 
fiahikas. * The lord of animals,' (pa^iin-flm patis), and the 
' terrible Agni,' (jnidro 'giiis); these are liis other and ill- 
omened names. Agni is his mildest appellation." (See Weber'« 
Indische Studien, i. 189. ii. 19-22. 37. 302 ; the l^tapatha^brah- 
mana,vi. 1.3. 10-17. > ix, 1. 1. 1,2, and Jour. Am. Or. Soc iiL3l9.) 
Again, in p. xxxiv. of his Introduction, Professor Wilson says, ia 
regard to Vishnu : — " There is no separate hymn to Vishnu, but 
he is metjtioned as Trivikrama, or he who took three steps or 
paces, whicli Mr. Colebrooke thought might have formed the 
groundwork of the Paurdnik legend of the dvtirf Avatnr. It 
may have been suggestive of the fiction ; but no allu.sion to Uid 
notion of Avatars occurs in the Vcilii, and there can be little 
doubt that the three steps here referred to are the three perit 
of tbe sun'a course — his rise, culmination, and setting."'^' The 
passage here alluded to by Professor Wilson is as follows : llig- 
vedai. 22. 16-21 :— 

Mi Hill "^^ ^^ f^^^ fk^ ^\hn ^5R{T*^'i ^^ 
Mwrf^ VTK^ii fwt: ^r^frf^ m?TfT ^ttY wf^ 

^f^ ^gT^:i f^^ '•l^iMrjIi fTf^HIHT f^tnq^ 

911'igiH: Ff^wi fwr^ q^^ q^ii 

'* May the gods preserve us from that (place) whence Vi 

'" " II IS expressly so slated bj DurgiicLaryya, in his commenlary en iVo 
Nirukta. — S«o nurnouf, Introduclioa to the 3rd vol. of Uw Bl^n*sti 
Puriifia, p. xiii." 

'" Instond of »aptn dMdmabMis, the Sama-Tcda li. 1024, readt oOt 
"OTer tlie $urf«co." 

8JtCT. Z.] 



strode across (?) the seven regions of the earth. Vishnu tra- 
versed this (world) : thrice he planted his foot, and the world 
was enveloped in his dust. Vishnu, the preserver, the unin- 
jurable, stepped three steps, upholding thereby righteous acts. 
Behold the deeds of Vishnu, from whom the worthy friend of 
Indra has received the sacreii ceremonies. The wise ever con- 
template that supreme station of Vishnn, placed like an eye in 
the sky. The wise, ever vigilant and diligent in praise, amply 
glorify that which is the supreme station of Vishnii."' — (See 
Wilson's translation, pp. 53-54 ; and Benfey's translation of the 
Sama-Veda, pp. 2*23, and 287 ; and his glossary, p. 191, under 
the word sapta : see also, Rig-veda ix. 114. 3.) 

The remarks of Yaska on this passage have been already 
quoted above (p. 187.) The following is the note of the com- 
mentator, Durgacharyya, on Yaska's explanation of the above 
pusage of the Rig-veda : — 

" Vishnu is the Sun. How ? Because he says, ' thrice he 
planted his footJ Where did he do so? • On the earth, in the 
firmament, and in the heaven,' says i^akapuui. Becoming 
estrial fire, he paces or resides a little upon the earth, in the 

ipe of lightning in the firmament, and in the form of the sun 
in heaven. As it is said, ' they triply divided him, placing a 
part on earth [?].' &c. Aurnanabha Acharyya thinks the meaning 
is ' He plants one step on the Samdivhana (point of ascension), 
when rising over the eastern muuutuin, (tuiotLer) at noon on the 






the wieiirfiMi dcr, (a third) on Gaj/a^iras, vht 
Betting beiMatli the tmtera hilL*"* 

Anr one who has the slightest ac<)iuuntance with the In 
Hindu mvtbology will petceire at oooe how widely differec 
tiMse Yedic RpnoeBtatkies are from the Pnnuiic aocoonts 
divm and VnAinn. Sneh diai^eB as these, in the cfinception 
the gods, most hare been the work of ages. Here, tlierefore, 
bare anotber proof of the great antiquity of the Vedic hytotu i 
compered with the dther partioos of Indian literature. 

Sixth. How different the Sanskrit of the Vedic Ago waiT 
many of its forms from those which the later San^jkrit aseiuiae 
and still retains, may be seen from the subjoined specimein 
taken from the Rig-reda : — 

Kig-veda, L 2. 1. with modem Sanskrit interpretation under'<< 

VeJic _c^. ^^ "j " Come, Vayn, these somas 
'- prepared. Drink of them ; hear 




Sftoskrit ^ 

Modern __^. ___«_ n 

Jem ^ntTt s f 

Sanskrit. 34^ ^^*V*J invocation.' 

Here it will be observed that four Vedic words, dar^tcUn^ i 

krituh, pdhi, 4rudhi differ from the modem Saniskrit fo 

Rig-veda, L 3. 7 : 

'^nrre^Wwr f%^ It^th '^r^mi < i sl«il 

*' O Vi.s\ve divas, |irt«erv'ers of men,I 
stowers [of rewards], come to the 111 
tion of him who gives you ([oblatious]." 
Here the Vedic forms onulsah, devasaf^ and dffoia stand 
omdh, and devdhf and dgachcIJiata. 
Rig-veda, vii. 35. 5. : 



Vedic VjT l 4i TT II '' Indraheard Vasisbtha when he uttered 

Mijilern -• *^ ' P'"^''^''» ^^^ opened up a wide space to the 

Sanskrit ^ ^t^f^ll J Tritmis." (See Part L p. 122. note 29.) 

Here we have the Vedic forms cuSrot and akrinot, for the 
moderu aSfinot and akaroU 

This fact of the frequent diversity between the Vedic and 
ordinary Sanskrit, is recognized, in every page of his work, by the 
great grammarian Panini. I will quote one of the Sutras, in 
which he refers to some instances of this, together with theiUuj)- 

trations given in the Varttika: (vii. 2. 64.) W*l *l M r| ""iJ' 
^srqr^^^^ f%7mil T^tttOt "^ f^MlrtJ"5%l ^1 ^"TrTT 

^T^ II " The exceptional forms babhutha, dtatantha, jif/rivi- 
blut and vavartlia are employed in the Veda instead of the ordi- 
nary forms, 6(/Wt»i'//Aa,«/r/i(7/i(T, jaf/rlhhna, and vavantha; an 
in the texts, 'thoinrn#< the first priest,' 'whereliy thoudidst stretch 
oxd the wide firmament,' • we have seized, Indra, thy right 
hand,* • thou dUlM enveitrpe with light.* " 

In Sutra, vi. 4. 102, other instances are alluded to of gram- 
matical forms, which are peculiar to the Veda, viz. the impera- 
livea imdki for iniiu, • hear ;' Icfidhi for iruru, 'do;' vri^hi for 
"vriiiUf ' cover;' purdhi for pfiuihi, * fill.' 

In the Vivaraija of Nagesa Bhatta on the Mahabhashya, the 
following reference is made to certain forms which are employed 

in the Veda only: TTT^ «<< *IH1 *'d'?r|*Tjif»|^ <4I^t ^VJTT- 

" The term 'secular' (^iit^vta^ refers to words diflferent from 
such as kaitieb/tik (for karnaiht), d«vdea^ (for devdh), and r/rlbh- 
ndnii (for (jrihiuim!), which are to l>e found in the Veda alone, 
for we never see them employed by eecular people.*' 

'^rnr^j? ^m^T^ "^'i-t**!^ T^F^l. 

No philologist will suppose, from these differences in form 
■which we discover to exist between the Vedic and the laUr 
Sanskrit, that the one language was different from the other. 
A great portion of the substance, and much of the form of the lan- 
guage, waa the same at both periods: a part of the Vedic roots and 
nouns only have in later times fallen into disuse; and the peculiar 
Vedic varieties of form are merely the ancient modes of infleo- 
tion which were in common currency at the time when the 
hymns were composed, and which gradually became obsoletA la 
tlie course of ages.'" Some of them, however, continued for a 

i>> The following \a Professor Whitney's acconnt of the differences betmca 
the Yvilic and the mo<Icrii Sannkrit : 

" The language of the Vedas is an older dialect, varjing verj conaideraUj, 
both in itd grammatical utid lexicfil character, from the classical Sianskrit 
Its grammatical peculiarities run tlirougli all departments: euphonic rulvf, 
word-formation and composition, declension, conjugation, sjntax . - 
[These peculiarities] are partly such as characterize an older Luigaagc, 
consisting in a greater originality of forms, and the like, and partly sueli m 
charactcrtsc a language whicli is still in the bloom and vigour of ltA>, its 
freedom untrammelled by other rules than those of common usage, UhI 
which has not like the [modern] Sanskrit, passed into oblivion as a nattre 


long time in popular use, as we find in the case of the form 
of the instr. pi. ebhis for ai^, which we meet with unchanged 
in the Gathas of the Lalita Vistara, (see the inntances given 
al)ove, in p. 129), and somewhat modified in the Pali forma 
ebld and ehL 

As the hymns of the Veda were the compositions of the 
ancient Indian rishis or bards, who, as we have seen above, 
(p. 206.) freijueutly speak of having " fashioned," or " generated " 
them, they could not possibly have been composed in any other 
language than that which these rishis and their contemporaries 
were in the habit of using for every-day conversation. 

There are, no doubt, in the hymns some apparent traces of an 
idea that the authors were inspired ; as in the following texts : 
B.-V.i.37. 4= "B^^^tT TT^mir'Singthegod-givenprayer.""* 

R.-V. i. 109. 1 : •TT^jn- ^^ Tmf?ITl% TT^ ^ ^ fw^ ^T^^- 

^ffr^TrRl^^l 2. . . . ^7JJ H\*im TT^Trfr g?>«ITf*^-J? I iH 

■pokcn dialect, become merely a conTentional tnediam of comrounicaiioa 
among the learned, being forced, as it were, into a mould of regularity bj 
long and exhausting grammatical treatment. . . . Tiie dijsimtlnrity existing 
between the two, in respect of the stock of words of which each is made up, 
ia, tu nj the least, not less marked. Not single words alone, but whole 
classes of derirations and rootji, with the faniilies that are formeil from them, 
which the Veda exhibits in frequent and familiar use, are wholly wanting, or 
have left but faint traces, in the classical dialect ; and this to such an extent 
u seems to demand, if the two be actually related to one another directly as 
mother and daughter, a longer interval between them than we should be 
inclined to ossumo, from the character and degree of the grammatical, and 
more especially the phonetic, differences." — Joum. of the Amer. Orient. 
Soc. iii. 296. 297. 

"* It is not, however, certain that the word hrahma, in this verse, means 
"prayer." In the Nighnntu, ii. 7, 10, it is explained as signifying "food," 
or "wealth;" and Saynnn, in bis iuterpretation of the passage, seems to 

adopt the former of these two senses. His words arc : If^ TjH^'sm.'- 
♦j<^»|f^ » jj TT^rraTT '^ffVJTT II "Offer praise in reference to brahma, 
I e. the sacrificial food." On the other hand, the word is used in the setiae 
of "hymn " in R.-V. i. 62. 13 ; and viL 22. 9, cited above p. 206. 


tcntr. I.' 

■^V^i ^•fEUf^ <1^7f II " From no other but you, (O ladra and 
Agui) do I derive int«^I]igence: to you have I fabricnted a hymn 
fi)r sustenance.'^'. . . WTiile presenting the soma, I generate for 
you a new liymn, O Iiidraand Afijni." In R.-V. x. 71, it is said of tht? 
goddess Vach: ^^%^ Tf-^: M<4^*l*i^ '^IT^'l^ rfn{^ 'H'^U-<\ 

Mifyu nf^^^ll '' By .sacrifice they followed the path [?] of 
Vach : they found her residing in the riahisT 

In R.-V. X. 1 25. 5. again, Vach i& made to say : ^SPI^^ <«<i<f4^,{' 

r! "^^m rT^f^ ^ ■q^WT^II " I myself declare this, which 
is desired both by gods and men. Every man whom I love, I 
make him terrible ; [I make] him a priest ; [I make] him a 
X'lshi; [I make] him intelligent."'^* 

In a Viilakliilya (or aj^ocrjijlml hymn) which ia to be found 
inserted betweeu the 48th and 49th hjTuns of the 8th manrlala 
of the Rig-veda, the ftillowing verse fK5cnrs: 

For the complete text of this verse I was first indebted to Pro- 
fessor ^liiller, who .supplied also the following version of it: 
*' Indra and Varuna, I have seen through devotion that whicli, 
after it was heard in the beginning, you gave to the poets — 
•wisdom, understanding of speech ; and I have seen the {susieA) 
places which the sages created in performing the sacrifice."*" 

'" fV^ ySiX^^ fsTW^ ^f^^m^^ ^^^ II " I \tn 
fnliricateil, i.e. mudc, a Lj^mn, i.e. n gong oF praise, composed by medilatkni.* 

'" Vach ihns nppears partlv, though not entirely, in the character ftf a 
Muse. Compare wliai Homer saje of Deniodocus, Odyssejr yiii, 63, 64 : 
IV.i' rt'pi Morff' iftXiint^ filov r' ayaOov Ti icatiov r», 
'0|i''i(^/(iJ»' (itV (ifitpn, fifiiv i' tjfiiav rioiS^r, 

'" If the word ijtttam in this vorse be taken aa a substantiTO, 
Iransliilion would run thus, '' I hare seen tlirouj;h devotion that wLLli 


SECT. X.] 


Though, however, some traces of an idea that the rlshls were 
insijired liy the goda, by Vacb, or Indra and Agni, or Indra and 
Vanina, (but not, in any of the passages which I have here 
quoted, by Brahmti, who in later times was regarded as the 
source of inspiration,) may thus be detected in the Rig-veda, 
there is no doubt, on the other liand, that these ancient bards 
generally speak of the hj'mns as the creation of their own minds : 
and it is impossible to a<lmit that they were anything else. But 
as even an inspired composition, to be generally intelligible, 
niust be delivered in the language current among the people 
to whom it is first promulgated, there is no pretence for sup- 
posing that the Sanskrit of the Vedas was not the vernacular 
liujguage of the age in which they were first recited. 

At that early period there was no language current among the 
Aryas but the Vedic Sanskrit, A learned language, different from 
the spoken tongue, was a thing then unknown; and the refine- 
ments of grammar had no existence. This may be gathered from 
the following pa-ssage of an ancient Brahmana, referred to by 
Sayapa in the iutro<lnction to his commentary on the Rig-veda, 
p. 35 : TT^rr ^^TT^I^IJTWr^W «l*4l«ill<4d I ^fTV^% m.1^- 

^rarTTS^^T^I ^\T[ 1t?,*1 •l^f'^il!] nX WT*^ <MT^r«4r?TI ^T 

dmir<<i wT?irTT ^WJi^ TffTi ^ifimfaB MORrif^TTzrrf^ 

*e formerly gave to lie rUhii, trudum, undcrdtandiii^ o/ ipecch, Ira- 


>RT A2n> KEt 



tTOfT T7ir^:i "It ifi thus related in the Aindva^-Vdyaixi^ 
ffraha-bruJtrnana: •Vach (speech) spoke confusedly, and without 
articiilatioii. Tlie gods said to Indra, Make this Vach to become] 
articalate to us. Indra replied. Let me choose a boon ; let th4 
Boma be given to me and Vaju together. Hence the itonukj 
of Indra and Vaju is taken together. Indra then, Btepping^j 
into the midst, divided speech in sunder. Hence she is spoken' 
articulately.' The sense of this quotation, says Sayana, is 
this: Speech, siich as in the verse ^flrnim i7« ptirohitam, &c.^ 
(the first verse of the Eig-veda), was originally confused i. e. un- 
varied like the roar of the sea, &c., and undistinguished, i, &j 
without articulation to denote crude forms, inflections, wordt^j 
and sentences, &c. Then Indra, being solicited by the gods, and! 
gratified by the permi^ion to t&ke the soma-juice in tJie samff' 
vessel with VajTi, divided speech, which had previously been 
without division, and introduced everywhere the distinction of 
crude forms, inflections, &c. In consequence, this speech, bong 
now distinguished in it« parts by Paniui and other great sag(% 
is pronounced by all men." 

It may be asked, however, If the Vedic Sanskrit was onoe tbe 
spoken language of India, how did it ever cease to be spoken. 
To this I reply as follows : — 

By the time when the collections of the Vedic hyinns were j, 
formed, the Sanskrit, the vernacular speech of the rishis aod I 
their descendants, had undergone a considerable alteration, which 
had gradually resulted, as we have alreatly seen, both from the 
general laws of change to which all language is subject, and 
also from the action of local causes, such as the intercooise 
of the Aryas, or Sauskrit-speakiDg race, >vith the Dasyus, or 
Mlechhas, who spoke a quite different tongue. In this w»f, 
words which had formerly been commonly employed in Sanskrit, 
became obsolete, or acquired new meanings, while other new 
words, borrowed from the dialects of the Mleclihas, were intro- 
duced into currency ; and forms of inflection which wct« onte 




current got gradually into disuse, and made way for other novel 
forms. Til US a twofold alteration was produced in the ancient 
Indian language (the Sanskrit of the Vedas), First, the Prakrit 
or vernacular dialects were formed out of it in the manner which 
has already been described (pp. 80, 143, ff., 148, 154) ; and 
secondly, a learned language, based upon the Sanskrit of the 
Vedas, but variously modified (see pp. 151, 152) and polished, 
was gradually constructed by grammarians, which being re- 
moved from the corrupting influences of popular use, has thence- 
forward continued unchanged (p. 171). 

When the process of change had beeu going on for many 
generations, the Vedic hymns became exceedingly difficult to 
understand. The obstacles to compreheusion, arising from 
these intermediate changes of language, were greatly augmented 
by the ol)scure and elliptical style in which the hymns were 
originally composed, wbicli rendered it hard for the men of 
subsequent ages to understand the brief allusions to ancient 
ideas, practices, and events with which they abound. 

These con.sideratione will aufBciently account for the difficulty 
w^hich wns experienced in the comprehension of the Vedic hymns 
in later ages, without there being the least necessity for our sup- 
posing that they were composed in a language at all dififerent 
from that which was ordinarily current in India, among the 
common people of the Arya race, at the time of their 






From the preceding review it is clear that the Sanskrit lang 
has beeu uudergoiiig a contimial change, from the very eurlu 
times up to which we can follow its course. But if this be the! 
case, it would he contrary to all analogy to suppose that tl: 
huiguage Usui rernaiued unaltered in those yet earlier ages beforel 
the Vedaa were composed. It must, therefore, now become iny| 
object to inquire, whether we can discover any means of fol- 
lowing it hack to its origin. We are not, it must be Confe 
iu a position to do this in any other way than that of reasoning' 
and inference; for, in the absence of any Sanskrit writings an- 
terior to the Vedas, we possess no direct means of tracing fh& I 
histoiT of the Sanskrit language and its mutations any further | 
bock than the date of the cumposition of those hymns. There ' 
is, however, another way iu which we can arrive at some con- 
ception of that history. 

Learned men have remarked, that there is a great resemblaoce 
between the Sanskrit and other languages, some of which, like 
it, are now no longer spoken, but were formerly the current 
and popular speech of ancient nations, and are preserved in 
written records which have descended to us from a remote anti- 
quity. These are 1st, the Zend and other varieties of the aucieut 
Persic; 2nd, the Greek; and .Srd, the I^tin.' The Zend lan- 
guage is preserved in the Zend Avesta, a collection of writings 

' It is not nccessarjr for my purpose to insist much on the affimti ei rfi 
the Sanskrit to anj other lunguoges besides those I have named. 



connected with the ancient religion of Persia. The poems of 
Homer which form the oldest relic of the extensive literature of 
ancient Greece, are supposed to Imve been wiitten about 2700 
years ago. And there are many I^atLn books which are 2000 
years old. From the great eimilaiity which exista between these 
languages and the Sanskrit, of which proofs and instances will be 
presently adduced, learned men have inferred : 1st, That these 
forms of speech have all one common origin, L e., that Sanskrit, 
Zend, Greek, and Latin are all, as it were, sisters*, the daughters 
(some perhaps elder and some younger, but still all daughters,) of 
one mother who died in giving them birth, or, to speak without a 
figure, that they are derivatives from, and the surviving represen- 
tatives of, one older language, which now no longer exists ; and 
2ndly, That the races of men who spoke these several languages 
are also all descended from one stock, and that their ancestors at 
a very early period all lived together in some country (situated out 
of Hindu8t.han), speaking one language ; but ailerwards separated, 
to travel away from their primeval abodes, at different times and 
in different directions j the forefathers of the Hindus southward 
or south-eastward to India; the ancestors of the Persians to 
the south ; and those of the Greeks and Romans to the west.' 
The langtiages of those branches of this great Indo-European 
stock which remained longest together in their earUest home, 
viz., the Persians and the Indians, continued to bear the closest 
resemblance to each other ; while the tongues of those offshoots 
which separated earliest from the parent stock, exhibit in later 
times the least amount of resemblance, the divergencies of dia- 
lect becoming wider and wider in proportion to the length of 
time which had elapsed since the separation. 

* Facies non omnibus una, nee diversa taincn, qualcm decet ease 

* For an account or the Greeks and liomaiu, I refur Ihc Indian student 
to u\y of the ordinary Listorical tnanuols. 

A50 mEmwkfWff OP 


fkiUogg: agktitua of i 

1 locrtabBiA tbeasMaertkns M to the resemblance ^ 
to the Zaid, Greek and Latio; after first pre' M 
■MBg a fev mmif^ itwrt'i oa oompantiTe philology in generaL 
A eompMina cf tbe variooa kngoages which are gpokeo in 
(fiflbfeni ci iwMrti i w of Eoiope wad Aaa, has brought to light the 
Curt that tbej bdoa^ to di ffmat Ikmilies or claasee ; and that 
Hie diffi etc u t m«aben vi ttie none fifimilv, while they exhibit a 
■tore or lesi doae iBWinWinrfi to each other, have either uo re 
iwiialilanrii. armrtxj xcaiote one, to those beloogicg to any of 
ofUier fiuailiea. It will be sufficient for the purpose of ill 
if I refer to the two great, and uiuT0«allj reoognise^l, familiea 
■peech, the Se«nitic and the Indo-Oemmnic. The languages 
which belong to the Semitic branch, are the Arabic, Hebrew, 
Syriac, &c Now all who have studied these languages are well 
aware that tbej closely resemble each other in respect of their 
roots and general diaracter ; while they have scarcely any affinity 
at all in any respect with the langxiages of the Indo-Germi 
stock, in which are included Sanskrit, Zend, the later forms of 
Pernan, Greek, Latiu, and the Teutonic and Scl.ivooic languagHL 
Any person who knows both ^Vrabic and Saa^krit is pertedJy 
aware that they have no resemblance to each other eitltcr in 
verbal root*, or nouns, or in the forma of conjugation and de- 
clension. Now, here we discover the very remarkable 
that two languages, both very perfect and polished in theif 
forms and utructure, and both of which are spoken by learneil 
men, of the Hindu and ^lahomedan religions respectively, 
living together, side by side, in the same cities of loiiiv 
are totally different from each other in almost every respect 
in which one elaborate and complicated language can b© dis- 
tinguished from another language of the same character. Afti 
what is the explanation <jf this, at first sight, so staztlilig » 
nhenonieuon ? It ie, of course, that Arabic is, fas its name fa>- 


their ^ 



SECT. I ] 



plies,) the language of the Arabn, a Semitic tribe; and waa 
introduced into India by the Mahoinedan invaders of that 
country, who, though not Arabians by descent, have yet, as their 
designation imports, been converted to the faith of the Arabian 
conqueror Mahomed, and have learned the language in which 
their sacred volume, the Koran, is written : while Sanskrit, on the 
other hand, is the language of the Brahmans, who are descended 
from a race which has uo affinity with the Semitic, viz. : the 
Arian family of nations. It ia not, therefore, wonderful that the 
Saoakrit and Arabic languages, which, though they meet in 
India, have been introduced into that country from quarters 
so perfectly distinct, should be totally different from each other. 
But the MuRulmaus of India are not only acquainted with the 
Arabic tongue, but with the Persian aTso, which is the living 
dialect of Persia, one of the countries which lie intermediate 
between Arabia and India. The Persian language which the 
Persians now speak, and which the learned Musulmans of India 
write, is a composite form of speech, i. e, one cliiefly woAe up of a 
mixture of Arabic with the aacient Persic, which was originally 
devoid of Arabic words. Now in that portion of the modem 
Persian language which has not been borrowed from Arabic, but 
inherited from the ancient Persic, we find many words which are 
manifestly of the same origin as the Sanskrit nouns or verbs of 
the same siguitication. 

The following list of words may suffice to prove the aasertion 
just made, that the Persian language, has, in its purely Persic 
element, an affinity with Sanskrit, while Arabic has no such 
affinity: — 


Comparative Table of Sanskrit, Persian, and Arabic Words. 

BmlUlu SanUrit. Pcniu. Arabic, 

father ))itri padar ilbu 

motber mitfi mwlnr &inm 

•laughter iluhiirl dukbtar bint 

«on-in*lnw jSmiiiri damnd bafid 

« S 



[cBAr. II. 





young man 












































































^^V I subj 

oin some 



of aflSnitj between 

Persian i 

md Sanskrit worda, omitting 

all reference to Ui« 










is tad an 

to stand 

kri (kar) 


to do 



to cook 

bhfi (bhar) burdan 




to die 
to Bleep 



to ask 



to give 



to know 

■to go, eat, 
. graze 







1 _. 



to run 


' (bandad 




to tear 

M. tm.>lnt. ira.) J 1 



to bear 



to fight 

grabU'( Vtdic) giriftan 

to take 






to be 




* InthJi 

caie the Arabic word regci 

nbles the Sanskrit 

* May we not consider the later forn 

u grah OS an early instance of tb* f^ ' 

procen by 

which in Prakrit k was sub 

stituted for bh f See Vararuchi, it- 2*- 




B«rr. t.] 






Eagllib. Sanikrft. Pertluu 




to buy 

« I'damidan 
(dhamati) J 

to ^^H 



to dig 




to draw 

srij airishtan 

to create ^^H 



to milk 

kup ? guftan 

to speak ^^^| 

r to like 
it© seek 

asmi ha«t^^nn 

am ^^H 



asti hast, ast 

he ^^H 



to obtain 

bhavami buvam 
bhavati buvad 

am ^^^H 
he ^^H 


(sanad *^ 

^n ' to strike 

abbuvam budam 
abhut bud 

I was ^^1 

he was ^^H 







Suikrtt. PrnUn. 



U8btar,shutr, camel 

tara sitaruh 

star ^^^1 




bala bal 

hair, feather V 




godhuma gandum 





yava jau 

barley ^^H 




varsha biirish 

rain ^| 



f brother-in- 

kflthti klsht 

fcultivalioa, ^^H 
L field ^H 




ioka sog 





twam tu 

thou ^^H 
you ^^^1 




manaa man$h 

1 - disposition, ^^^1 
! mind ^| 



city, bridge 




prishtha pusht 





durnama dushnSm 

r dutmanas, 
durmanas 1 , , 

I dushman 

bad name ^^^| 

1 disturbed, ^^^| 

enemy ^^^H 

Jan a 








durvara? • dush- 

diiScult ^^1 




V khwar 





mahattara uiihtar 

greater ^^^| 




kiima kam 

wish, desire ^| 



' thumb, 

. finger 

ay as ah an 
mitr roihr 




ffikha shiikh 


• Could this Sanskrit word hare been originally dkatta f 


' War 

is a Persian sulEx, probably unconnected wiUi the Sanskrit vara ; | 

but tbcrc 

is no doubt of the identity of the Persian particle 

! dmh and the H 

Sanskrit du*. 


«J 3 



AFPirnriK hvji jyi 

I OBAr. tb 















garden, reet 



door J 

Irish na 


tub, Up 


f thirst, 
1 thirsty 
r corner, 
\ arbour 

face, body 




f nail (of 
t finger) 






















guru, ■ 
gariyns . 










































char in on 




in ah 
















I tive M 









shell, stone 






Note. — On the other hand I may epecirj the instance ofqftU (Ar.) $fii 
ujmd ($an!i.)t >n which a word of limilar sound has the aame seSM if 
calamity in Arabic and Sau&krit- 


Now the old language of Persia, from which the 
in the above list, still forming part of the modern Perai 
must all be derived, was a language closely connected wilh 
the Sanskrit, That language, at a certain stage of its progrves, 
•was the Zend, which we find employed in the ZendaTe«?ta, 
sacred volume of the Zoroastrians, or Parsia, a ■work whii 
Btill exists, and has recently begim to be studied by Euroj 

8KCT. 1.] 



In the same way, if we compare Sanakrit with the language 
of the ancient Greeks (who lived to the north-west of Persia, 
on the eastern and western shores of the iEgean Sea,,) and with 
that of the Romans, who inhabited Ituly, we shall find a close 
resemblance, and frequently, an almost perfect, identity in very 
many words, hoth as regards the roots and the inflection. 

The resemblances between languages may be twofold. First, as 
regards the roots of the words. For instance in Sanskrit, we have 
the word ndma, * name,' and we find the same word nam in the 
same sense both in Persian and Hindi. The second resem- 
blance is in the mode of iuflection. Here we do not find 
any resemblance iu regard to the way in which this word 7Mi»ia 
is declined between the Sannkrit and t!ie Persian and Hindi 
languages. The Sanskrit has three numbers, singular, dual, 
and plural, and seven cases in each number, whereas the 
Persian ami Hindi have only two numbers, singular and plural, 
and tlie cases are fonneil in quite a different way from those of 
the Sanskrit, To prove this it will be sufficient to give the 
different cases of the singular number of this word in each 
of the languages. 















nam se 



and supplied 
by prepoei- 




niin ae 









If now we compare the Latin word for "name" with the 
Sanskrit we shall find that the root is the same, and that the 
mode of inflection is very similar : thus, 

SanikrK. Latin. 

NoM. nima (from crude form natnao) nomen 

Ace. noma noinen 

lasT. Doinni nomine 

Dat. Damao nouiini 
4 4 


AFPnnriBS and derivation of 












itin language haa no 
























We see here that, while the same root expressing the word] 
* name,' is common to all these languages, the Persian and UiitdiJ 
have lost the ancient forms of inflection, while the Sanskrit- 
aud Latin have preserved them. There thus exists a duublc' 
resemblance, viz., first of roots, and second of inflections, he- 
tweeu the Latin and the Sanskrit, and the same remark is equally 
true of the Greek and the Zend. 

Now, when we find that a multitude of roots coincide in anj 
two languages, of which the one does not derive them from the ] 
other, we may be sure, (even though the one may have no ' 
complex system of inflections, while the other bos,) that those 
two languages have a common origin, especially if we can ahow.j 
that the one which ia deficient in inflections has grodiially Iwt 
them by a particular process of alteration which can still \ie 
traced. But if any tw<i languages reseuible one another MA in 
root* and inflections, the proof of their affinity is then grwatl/ 

r. n.] 


Sect. IT. — Detailed iQiulratioru of the affinities of Sanskrit with the Zend, 
Greek, and Latin languages. 

I proceed now to furnish, first, some epecimena of words 
which as roots correspond to each other in Sauakrit, Zend, 
Greek, and Latin ; and I shall afterwards exhibit the mutual 
resemblances of these four languages in point of inflection 

The following is a list of words (derived from the publica- 
tions of Bopp, Benfey, Aufrecht, Pictet, and othera, and in some 
instances from my own conjectures) which correspond both in 
sound and sense in Sanskrit, Zend, Greek, and Latin. 

No. IX. 





Graek. Lalln. 
paler pater 
tuSt^r inatcr 
pbratria(a clan) frater 






. « 


naptri 1 
napat J 





j husband's 
I brother 

j amatol 




father- in-law 












father's brother 
son, boy 







• • 

hero, man 






^ ^ 

• • 




H 234 



^^M Santkril. 









oM age 

^^M mansfl 





^^H yuvon 




young man 











^^B Borpa 




serpent, repUl< 

^^m patatrin 




a bird 

^^1 patis 










^^B Divospntifl 




lord of tbe sky 

^^^B Yuruna 




Varuna. beavM 

^^M jwan 



can is 


^^^B ^tvannm 




dog (ace) 

^^H maksliika 




flj _ 

^^H paJu 





^^H a^wa 




horse ^^H 


» * 



bog ^^H 

^^H bniisa 



goose ^^1 

^^M riksba 



bear ^^H 

^^m apas 




^H aluha 




^^B barbarn 



borbariaa ^^H 

^^1 dama 



bouse ^^H 

^^M okas? veia 


, • 

bouse ^^B 

^^1 BTindama 



jBubduer offbe^l 
Iborsea M 


cris? (strife) 


enemy ^^^ 


■ ■ 



^^B dru, druma 

dru, drutaos 


tree, wood^^H 

^^H dbuma 



smoke, <><)i^^| 

^H prasUra 



stone ^^^1 

^^H bhru 



eyebrow ^^^ 

^H dantam (ace.) 



tooth ^^1 

^^H nakliA 

onus, onucbos 



^^H Tiaman' 



name ^^H 

^H fiankba 




shell, cockle 1 

^^H * Nwnan is 


by Benfey, G. W. L. ii. 144, ma ^jnaman^ "that by J 

^^H wbicb any one is known." 

The Latin nomai. would in 

like manner be — !•■ 

^H fpiomeiL, (compare cogntmen). The Sanskrit »aoiaA 

might also be = laa 

^^^ mndman, from 

miM, "that 

wbercby any one is remeaibeied." M 

UCT. U.] 







Tach . 












arkhon P 







queen ^^^| 





knee ^^^| 


• < 






• » 

honey, wine ^^H 







• a 



joke ^^^1 




lacryma ? 



■ « 








yesterday ^^^| 




h extern us 

of yesterday ^^^| 








• . . 


day ^^^1 





rby day, ^H 
l under the sky ^^^| 

naktatn, nakti 

mukta (ace.) 

Doctem (ace 






strength ^^^| 





bed, litter ^^M 






kalpa, (krip, 
to make) 


mi-"! J 


body ^^^1 




brilliancy ^^^| 



locus ° 

world, place ^^^| 




barley ^^^H 


mSioi QAar^) 


dirt, ^^^H 


skia * 

■ * 

shadow ^^^H 








light, bright ^^H 



darkness ^^^| 


• 1 


armour ^^^| 


• . 


thunder ^M 




black ; darkness H 




place, grouivd ^| 

Talmlka (ant- 


murmox ? 








giver (mue.) ^^H 




giver (fern.) ^^H 



gen i tor 

father ^^M 


Si quia p 

iorum manibus locv*. 

Tacitus, Agcic. 46. ^^M 

^H 236 "AiTiifm^uS^DERivASo^o^^^tSI^^ 

^^H Suiikrit. Zend. 




^^B janitri 



mother • 

^^H jata 

« • 



^^1 jnatA 




^^M puros 




^^H araa 


. * 


^^1 prithu 


. • 


^^M gurna 




^H garijaa 




^^F garisb^ha 




^^H Tarisbtlia 




^^H la^bu 


« • 

light, small 

^^M lagliisbtha 




^^H tuabfin 




^^H nobtyan 




^^H msrihisbthas \ 
^H (VfsUc.) J 



r greatest, vene* 
1 able 

^^M bahu 



great, thick 





^^1 mpdu 



■ofl, slow 

^^M tanu 

tanu P 



^^m rudhira 



blood, red 

^^1 gbarma 

, thermos 


beat, hot 

^B «ushka 



di7 I^H 

^^H pQrna 



Ml ^H 

^^B dirgba 


• • 

lo»K ^H 

^^H Bwapnn* 



sleep 1 

^^H nabbas 



sky, cload 1 

^^M abbra 

" ombros 
. aphros ? 


f cloud, rain, 1 
1 foam ^^fl 

^^H an (to breathe) 



wind, iiui^^l 



ventus P 





r vapour, tri^^P 
l soul ^^H 

^^H sainaj 




^^P (oka, takman 

tcknon, tekos 


^V tira'o itnre 

aster, astron astrum 

star ^H 

^^B <^ T1)C original form 

of this word waa 

probably ttdra; 

as may be argncfl 

^^1 n-om the i being presc 

rved in Zend, and the s in Greek 

, as well aa ^||fl 

^^P Persian titarah. Ben 

fey, Griech. Wurz. Lexicon, i. 

661. In th^H 

P edition of his Comp. Gr 

am. par. 49, Bopp 

gives the Sanskrit word as dM^ 

B^^^^k {xwm. pi.) in the V'eilic 












. dom 

■ • 

wood, tree 








f • • 











. ■ 

babitation, citj 

igra (Yedio) 







. . 












iron, copper 









. • 




















• . 





. . 





• . 


















• • 



liquid, dew 












. , 


» • 



• ' 

• • 


. belly 


, , 









kratu (Vedio) 






. • 



strength, strong 








peran, peras 


the other side 





right (side) 







. . 








skull, head 





water, wave 





water, sea 


• • 


• • 








• • 





. , 


, . 


H 238 


Zend. Greek. 



^^^^m nrargatas 




from heaven 

^^H^ nodi 




^^H matram 





^^m aksLi 


5k8s, okkSa 



■ i>7* 




face ^M 

^H bShtt 



arm ^^M 

^1 uthi 




bone ^^H 



. » . 


sword ^^M 

^H kravya 




flesh ^M 

^B ekalara 




one of t»0 1 

^H inadhja 


i mesoa 


middle 1 

^H mrityu 




death 1 

^H mrita 


• > • 


dnwl ^m 

^H martyas 




mortal ^^M 

^H auiritas 




immortal ^^M 

^B Bwadu 




Bweet ^^1 

^H pida 




foot fl| 

^H padiili 



r pedes, pe- 
1 ditis (gen.] 


^H anyos 





^H antaras 





^H -ubba 





^H vabana 



• * 

carriage ^^H 

^H aaroit 



■ t 

marsh ^^^ 

^H arbbaka 


orpbanoa ? 


small, bereaved | 

^H phulla 




flower, leaf 1 

^H smilrlta 

martu8 P 

f who rcmeailM|^J 


^H arj^iana 




earning, *'4^^| 

^H Luuus 



• • 


^1 {>atha 






r ground, oa V^ 
\ ground 1 


^^ bharas 




a load 1 



b agios 


venerable, hoi^j 

^H anna (from 

ad) . 


, , 

eaten, eataU^H 

^H Siayana P 



, ^ 

ocean ^^| 

^P guha 


gupe, kupe 


cavern, I><>I<^H 

^B kakudinnt 



cacumcn ? 

mounuun« |^^| 

^B kuliivara 



cadarer P 

body, corali^H 

P aiman 


stone, aoTil'^^l 

8KCT. U.] 




Zeuil. Greek. 



«aru ? 


, » 









tooUi, grinder 


. . sitos 

, , 





pure wine 






• Foinoa 


soma juice' wine 




insect, flea 




jackoll, fox 


k ar abos 


locust, beetle 



. . 


udra, unlra 

. enudria 






■ Indian cuckoo, 
. cuckoo 


« « ■ 





corvus , 





f Indian caokoo, 
I magpie 


• . • 






upari " 

upa " 




a pa 






















pro*, proti 

• . 






inter, intus 







towards, on 

am phi 

a • 

towards, round 

" Tbe Latin forms luper and tub, (eem to be more genuine and original 
than either tbe Sanskrit or the Greek, as Uiey preserre tbe initial i. That 
these wordd must have bad an initial 4 is rendered probable by the other 
cacea in which the Sanskrit and Latin both retain the f, when the Greek 
changes it into h, as in srip and scrpo, compared with herpa ; and taptan 
and tepttm, aa compared with hep^ ; see, however, Bcnfcj's Gr. W. Lex. 
i. 384. 





^^H SauLxit. 




B«tfMu 1 

^^1 P*'^ 



. V^ J 

^H tann(«) 1 
^H (from trO J 



acroM ^^H 








iU ^ 

^^H sumaoas 




^^H durmanas 




^^B •n 



priratire partic 




bow many? 




so many 





^H twla 


. , 

then ^H 

^H yada 



when ^^m 

^^H anyadi 



at oiber ttaflfl 

^H Utaa 



Ibenee ^^| 

^^B yatas 



wlieoce ^^H 

^^1 ittham 

• . 1 


tLus ^H 

^^H pa^clwt 



afker ^^ 

^^^^^ inokshu 



quickly ^^M 



^^V Swikrit. 




Bnfllth. ^^H 






^^M trayoa 
^^1 lisaras (fvm.) 

lisAro ( 



three ^H 

^^H cbatwaras 


ro tessares 


four ^^1 

^^H pnnchaa 


L pente 



^H shat 





^^H laptan 




seven ^^M 

^^H ashman 




eight ■ 

^^M navan 



no vein 

nine ^^M 

^^B daian 









twenty ^^H 

^^B Aatam 




hundred 1 

^^1 pralliainu 




first 1 

^^M dwitiyas 



secundos s«cood | 

^H triayaa 




third J 

^^1 chaturibaa 1 
^^H turyas J 





fourth ^^M 

^H^ putoliuaaa 




fifth ^M 

MCT. II.] 







Engltih. ^1 





sixth ^1 





seventh ^| 





eighth ^1 





ninth ^| 





tenth ^1 










Uirice ^| 


, , 


• , 

in two waja ^H 


, a 



in three ways ^| 


, • 



in four wajs ^| 


■ • 


• • 

in five ways ^H 







Rngllili, ^1 




• ■ 

I tear, flay ^| 





I give ^1 





I place ^1 





I staud, place ^| 


• • 



I mix ^1 





I spread ^| 


, . 



I bear ^1 





to be H 

(lib) lehmi 








I stretch H 





I stretched ^H 





I beget ^^ 

( jna) jiinami 




I know 


• • 



I wound or beat 


» • 



I have beaten 

(»5t) scve 1 



• • 


I reverence ^H 

f he desires, (S.) ~ 
L it pleases (L.) 

(tup) tupomi 




I hurt, beat 





I eat ^ 



• ■ 


I carry ^| 





he carried ^^ 


. , 

• • 


I go, ascend ^| 





I annoiut ^H 

» See Benfcy's Glossary to S.V. ; and 


aaapanla in R 

•Y. vu. 83. 8. H 




Sanikrit. Zciul. Grerk. 














I fall, fly 

sidami from 1 
shad) J • 



I siuk, sit 

chhinadmi *' 



I cut 




they cut 


« • 

findo . 

I cleave 



1 damnemi 

• • 

f I am satisfied, 
I please 
I subdue 



. , 

I take 




I anoint 




to anoint 



fluo, pluo 

f I swim, m 
1 flow, rain 

manaini (mnu) 



I remember 

juhomi (hu) 



I sacrifice 







, , 

I bite 

karomi (kri) 



rl do, fulfil, 
1 create 




I sit 




I vomit 




J ventris cre}Htui 
l edo 

swede I 
swidyami J 



I sweat 




J I afflict, am o 
I fire 




I sound 




f I cut, looK 
I pay 
I burn 




I am, turn 




lie is, tumi 

meliami (mih) 



I make water 

emi (from i) 



I go 

'' This root may originnlly hare had an initial i. Sec, however, Benfe; 
1. W. L. i. 166. 

8SCT. H.] Tira INDIAN AND COGNATE NATK)N?!^^^^^f^^^^B 

S«mlirlc. Zead. 




(IrL^yAmi (olil 





see ^^H 

from dr'ik) 


▼edmi (vid) 



know, see ^^H 



. . 





cleanse ^^^| 




am ^^^1 

prichliomi pSrSSmi 



Task ^H 


skeptoiniii i 


(later form? 

■ (by aieta* V 


I make clear, see V 

pn^yami) - 

Ihesis) J 





I Tear, tremble ^^^| 


nekr03? (dead) iioceo 

I perish, hurt ^^H 



apargo ? 

I touel), sprinkle ^^^| 



tncrgo ? 





I touch, lay ^^^B 




I touch, twined H 




touched, twiiiL-J ^M 

rijG '• 



I rule ^^H 

locbe. loke 



I look, shine ^^^| 




I speak, call ^^^| 

takshami " 

tikto, teukho 


I fabricatc,bc^ct H 




llhlnkiiisccrtain H 

Tapomi " 



1 weave ^^^H 



« ■ 


bUanj " 



I ^H 

bhuj " 

• • 






enjoyed, fruit ^^^| 


. • 


run ^^^^ 


• • 

cucurri ? 

went, ran ^^^| 

kalpayamt ■> 


(from krip; 
Pali kap. 



I cut, pluok ^^^M 

p5ini) J 


«♦ Nighantu, ii. 21. 


>» Compare the word 

1 toka, laJtman, teknon 

in the list of nouns. ^^^| 

'• Dr. Aufrecbt finds 

in the word urvaeabhi the trace ot an old root rabh, ^^^| 

•* to weave," which ig sli 

11 closer to the Greek form. S<>eB 

jhtlingk and Both a ^^^| 

Dicliimwy, *ub roce urn 



" These root» were, 

perhaps, originally 
» s 

bkranj, and 

Mmj, the r Uing ^^H 




Suubit. Zend. Greek. 




• . • 

I eat 

■I am ezetti 




angry, dca 








I live 


. . 


I worship 



. . 




, , 

I kill 

gup (root) 



I hide 








I hide 



. cemo 

r I scatter, 
I separate 








I have dmnk 

patum (to drink) . 







I wake, rouse 

pinashmi • 



I ponnd 





kampe, (I "l 
tremble) J 



I bend 




I bend, nod 




rl seek, am 
1 greedy, rage 

ban (orig. dhan ?) . 



to kill, die 








I speak 




I sew 




I bind 




I went, ran 




I show, tell 




I showed, tAld 




ye showed 

mime, maml'i 
(from ma)/ 



I measure 

trape " 

trepo " 


rl am ashame 
1 Itum 

aflerwards dropped for the sake of euphony. Benfey, G. W. L. iL 14, i 
Weber, Ind. Stud. ii. 88, note. 
■8 These two roots differ in sense ; and perhaps have no affinity. 



Sanikrit. Zend. 





tropftoii J 

* thrupto 


I hurt, break 





I strive, seek 





I rub, bite 





i I rublwl, bit 
■ I unilcrstanil, 
. lb ink on 




• • 


ninjG (nij) 



I cleanse 




(ad)ip)ciscor I obtain 


band]) bandb 



biad (root) 




I jungo 

I yoke, join 





f X roar (Sans. 
"Ji wander) 






luin|>ami . 



I cut, b»eak 





I follow 





I sliinc, burn 





, I roast 




• • 

I run 





I bite 









CO quo 

I cook 


' pakva 






* . . 





. bazomai 


I venerate 





I flow 




Btepho? . 

r I slop, stomp, 
\ crown 





I begin, sew 


t»i tni 



f I JeliTCr, keep, 


Where the Zend 

word has been omitted in 

the proper cohimu 

of the preceding 1 

ist, I have not found it 

readily accessible. 


It will be gathered 

from the list that in many cases where the 

Greek language fu 

rniahes words 


both in sound and 

senae to certain Sa 

jiskrit words, 

the Latin, 

08 preserved to us, 

has no words of co 

rresponding form : and that, viw versa, the 

Latin has often for 

ms corresponding to the 

'Sanskrit, where the 

Greek has none. 

In all the instances I 

have adduced, the 








affiuitj is, of course, not equally certain. Doubtful cases I hart 
generally indicated by a mark of interrogation. 

It will also be oljserved that certain letters in Sanskrit 
uniformly, or generally, replaced by certain other difTereutl 
letters in Zend^ or in Qreek, or in Latin. Thus the Sanskrit i 
becomes h in Zend and Greek, as in the case of aaptan, 
and hepta. The Sanskrit « again becomes I: in Greek ; dh 
replaced by tk ; bh by ph ; gh by kli ; j hy g ; y hj z or (i 
Latin) j ; and so on. 

I now proceed, secondly, to exhibit the resemblances whic 
exist between Sanskrit, Zend, Greek, and Latin, in respect 
their modes of declension and conjugation, as well as geuenlljl 
ill the fonnatioii of wonls from nominal and verbal roots. 

1 shall first tif all adduce as an instance of this aiuiilarity, ' 
firut and second pei'sonal pronouns. 




■ iTiiiliyara, 

ni&mri, niG 


manm, mu 

{iBfthyara, mnibja 1 

Cmoi, moi mihi 


ti]ai]a,me,mOJ emou, mou mei 
eiuoi, moi me 

\a«ni«! ivi)dic)j 



Ave. iisniun, nos no 

IXST. &9IIirkllULS . . . • • . 

I* AT- J ^ I mnibyu, no beoiin nobii 

\ iVin.or iiiu J 

Auu ibmill - • • • • • 








I or nas J 




of us. 

Loc. Bsmuu 

• • 




in us. 

NoM. tw&m 





AcG. twam 





Ihst. tw&ya, 



• • 

by thee. 

DAT. f^^i'y'™' 

tor te 

tbwo!, t6i, 



to thee. 

Abi. twSt 



, , 

from thee. 

Gbn. t&vft, tii 

I thwdi, te 

J- sou 


of thee. 

Loc. tw&yi 





in thee. 

NoM /y"™*^' 

yuzbem 1 
) yu8 J 




Ace. |y"«'»°^' I 





IxsT. yushmabhis 



, , 

by you. 

^ fyusbma- 

Lbhyam, ras 



to you. 

AbIn yusbmat 




from you. 

Gbb. yiuhmakam 


▼0 bumSn 


of you. 




10 you. 

The following are examples of the similarity as regards the 
declension of nouns between the four languages in question. 

NOUNS UASCULINE, ending in a. 
Vrika, "a wolf." 


























B 4 



[cHAr< n. 






















. luke 














1 vrikayos 






■ vrikabhyam 






N. & 

j- vrikas 















Trikcbhyas "1 
vrikebLyas 1 
















" tongue." 


























A III.. 












bizvaya ? 




jib we 




^KeTu^ TUE 



ending in 


Pitri, " a fatber." 







Latto. ^^^H 



brita (brother) 







patrcni. ^^^| 




paturi, palri 





patdri, patri 



pit LIS 


paterl, patri 









briilhri ? 

patCri, patri 






pater. ^^^| 

















p!)trlbu8. ^^H 





patribus. ^^^H 





patribua. ^^^| 




pateron, pati 

on patrium. ^^^H 





patribus. ^^^H 


eading in fi. ^^H 






LaUD. ^^M 









datorem. ^^^H 





datore. ^^^^^H 












dat 11 8 








datore. ^^^^ 







datorcs. ^^^| 



daibrcus ? 


datores. ^^^| 





datoribus. ^^^| 





datoribus. ^^^H 





datoribua. J^^^| 





datorum. ^^^H 




daluribus. ^^^| 











Uharat, "supporting." 








Urtia. ^^B 

" NOK. 




feren». ^H 

ii Ace. 

bharniitam barCnlcm 


fcrenteni. ^H 

^K Il«8T. 




ferente. ^H 





ferentL ^H 

^P Abl, 




ferente. ^H 

H GliN. 




ferentis. ^^M 

H Loc. 




ferente. ^H 

■ Voc. 



Dana, " 



a gift." 

ferena. ^H 












ilonum. ^^m 

■ Ace. 




donam. ^^H 





dono. ^^H 





donu. ^B 



da tad 


dono. ^H. 





doni. ^^^Kt 










donunu ^^^^H 


Neuter Noun ending with a consonant. 



Nainan, " 

a name." 








U NoH. 




noinen. ^^M 

^1 Ace. 




nomen. ^^M 





Domine. ^^H 





nomini. ^Hl 

^1 Abl. 




nomine. ^^M 

^^ Gbn. 




uotiiinia. ^^M 

^P Loc. 
■ Voc. 




nomimnise. ^H 


























n&milbbyas, ' 


r ononioai 



namabhyas . 

I onomiisi 












The forms of conjugating verbs in Sanskrit and Greek have 
a remarkable resemblance, jmrticularly in those Greek verbs in 
mi, in which reduplication of the consonant of the root takes 
place in the present and imperfect tenses, Greek as well as 
Sanskrit has the augment in e = a in the imperfect and aorist, 
and the reduplication of the consonant in the perfect. The 
most striking instance of resiemhlance is, perhaps, the root dd 
or do 'to give;' which I subjoin together with several other 
examples; adding occasionally the Latin forms and the Zend 
alao, where they are easily accessible. 

THE VERB to yive. 
Present tense. 











- ditiliisi 









, , 

• • 



' datthu 





• . 









, , 










^ a 




' adadua 








252 -'] 

\ppixrnE3 Ain) DERivATioy 

OP [caar.H^ 

^^^^^L SaiMkriU 

Z«nd. Greek. 

L«Un. ■ 

^^^^V adodva 

■ > ■ • 

• • J 

^^V Ddal - adattom 


- - ■ 

^^^^^H adaltam 


• • ^ 



dabamua. ^^B 

^^V Plubal ■ ailatta 


dabatia. ^^H 

^^^^^_^ laJadus 


dubanu ^^H 


Third Preterite. 


^^^^" i-adim 


. . ■ 

Si5G. • tiat 


• ■ 



' ' ■ 



- - H 

Dual • odntam 


- • ■ 

I. ailatam 


n j-adama 


• ■ 

^^B FluhaL' mlata 


- - ■ 

^^H l-adus 


- • ■ 


Reduplicated Preterite. 






















^^^^H Dual 





■ dildAtus 







^^^^^M Flukajl • 








^^M Tlic aulijimctive and 

precative moods of the Sanskrit ain 

^^M answer nearly 

to the 

ptatives of the present aud aorut ia 

^H Greek : thus. 
















^^^^H^. tiMyit 




gtCT. 11.] 













doie, &c. 


There is also a resemblance in the Greek future doso, "I will 
give," and the future particle doson, tti the Sanskrit ddsydmi 
and ddayan; and a perfect identity in the Latin gerund, daiuvfi^ 
with the Sanskrit infinitive ddtuvu The affinity between the 
Sanskrit form ddtri, *'a giver," or "one who will give," (wliicli 
makes ddtdras in the plural,) and the Latin future particle 
daiurus, is also striking. 









to place. 

Present Tense. 


a trek. 



' dadhiat 





« • 

- dljattboa 






H dhattba 


. ilndLatt 



|- adadbam 


■ adadbus 





•t adhattam 






J adhalta 





AFFmrriES anb dsbivakion of 

[CBAP. n. 

Third Preterite. 





■ adhas 



r adhava 



' adhatam 







■ adhata 





VERB to spread. 
Present Tense. 







Smo. • 







• strinuvas 

. • 

, . 




. * 



• . 

■ strimumns 










' oatrinavam 



SiNO. ■ 




. astrinot 





, , 

Dual - 



. . 




«• astrinuraa 



Plur. - 

1 astriiiuta 



I astrinvon 





THE YEBB to creep. 

Present Tense. 















































. • 




• • 














Snbjunctive, optative, and future (Latin). 

f sarpeyam 






























Suukrit. Ontk. 


sarpan herpSa 



sarpantam herp5nta 



sarpate Iieip5nti 




sarpantas herpontes 



sarpadbhjas herpoasi 

THE VBBB to be. 




Ztxtd. Greek. 



ahmi esmi 




ahi eis, essi 



a£ti esti 




. . 









hmahi esmen 




Sta este 



hSnU. eisi 





















• • 

, , 




• • 



• 1 




Plural ■ 











VERB to Stand. 





















' tKitithiiiiias 




Plokai. ■ 










THE VERB to show 

or say. 







, adilcshBm 







■ aJifcshat 




cde!.xamen diximiu 

TLUSAt -| ailikshata 



1- adikabaii 



The follnmng are additional examples of similarity of form 
in the past tenses, combined in most cases with identity of 

Sanikrlt. GTMk. 





he carried 

akshipsi [I threw] . 


I wrote. 

apaptam cpipton 

I fell. 

apatam epe^on 

I fell. 

astbum «?stcn 

I stood. 

The subjoined instances exliibitthe similarity in the formation 
of the reduplicated perfect between the Sanskrit and Greek. 











I anointed 






I was able 



I saw. 


tutdpa ' 
tutopha . 

I injured 



I struck. 



I heated 

[from tapbo.) 



I buried. 




I acid some examples of conformity between the Saaskin 
infintttve aud the Latin supine. 









to sUind. 



to hegtU 



lo anoint. 



to go. 



to vninit. 


sod! turn 

to sound. 



to know. 






to join. 



to creep. 



to pound. 

The furm of the Sanskrit deaiderativea, though not the 
Bignification, is found in Greek and Latin: thos we bare Ji- 
gn69c6, (Greek,) atid no$ro, (Latin,) anfiwering to jijun-aavt >,** I 
desire to know;" and again, mimn^A-6 and [re]7nini«!or, 
answering to rnimndadmi, "I desire to remember.** 

Ajpvin, Greek words like paipcillo, daidalloy paiphatto, 
jflrnplnni, phnpremi, &c, though without the meaning, htm 
the form of Sanskrit intensives, like bobhUf bambhram. 

In regard to the participles, also, there is a remarkable ooin- 
cideuco lietween the Sanskrit and tiie Greek. Some of th<' 
pfu-ticiplfs of the active voice have been alrea«iy giren. TV 
fiillowiug are some other specimens. 



t«*«r^A* totoplHua Map fcc a . | tatupiran tnUiptlshi tntupiru 



Swidtrtt (ucttter mk] mamlue) fawa in man oorra^ond # 

" w""": ^'**^*»**»«*»*^ •»«»»•» oananoo to SM*ifc 
i:^*^' UUu : thus »k. SM«fait mriimm. n«r«m. M^ 



to the Greek niptron, pledron, lehtrun, pheretron, lutron, 
arotran, and the Latin viulctram, epedrum, aratt-um. 

The nomiual form in nas, is common to Greek and Sanskrit: 
thus, the hupnos, (sleep,) of the one answers to the svapnas 
of the other. 

Paasive past participles in ta are common to Sanskrit with 
the other languages: thus, 

S«uikr<l. Zend. 



datta* dato 



Compare also bhaffnas in Sanskrit, with stugiioe, terpiws, in 

Abstract or other Rtihstantives in td, tdt, te^, taa, are also 
found in them all : thus, 














Zfnd. nreck. 



amSrStat neo-tSa 



uparatat homo-tes 



astatat platu-tSs 


Forms in tie occur both in Sanskrit and Greek : thus, 









Instances of adjectives similarly formed • 







H kulinos 






^B pitryas 



^^ jaUtjA* 



m Forms in las and 


1 Santkni. 



■ chap&Iaa 



t tarala* 




« it 




^^1 Forms in roe : 


^^^M .Sanikrit. Ore»k. 


^^H tnadhuras phoheros. 


^^H subhraa psuchros 

gnums. ^^H 

^^^H blmitras Inmpros 


^^^^^ Feininine nouns are also similarly 1 

brmed, as follows: ^^H 

^^^^^H Sontkrlt, 

LrtiD. ^^^1 


matron a. ^^^| 

^^^^^H v&runani lukaina 

patrono. ^^^| 

^^^^^1 rudrfini despoino. 


^^^^ Abstract nouns are also formed in 

Greek, as in Sanskrit, ^^B 

^^m changiug the vowel of the root: thus, from the roots bhid, ' 

^^M krudli, and lubk, are formed the nouna bkeda, krodha, and 

^^H lobha; and so in Greek we have 

tromoe, pkoboe, trokkot. 

^^1 nomoB, hnpofi, from tremo, jihebomai 

, trel-ho, nemo, and hipo. 

^^m We have examples of nouns in Latin and Greek resembling 

^^1 Sanskrit nouns in ya, such as these: 

^^^H Saniltrit. 


^^^^^^ mailhurjam mondacium 


^^^^^H naipanjam principium 


^^V Simple radicals are used in all three languages at the end of 

^^1 compound nouns and adjectives: 

^^H Smiikrlt. Gre*k. 


^^H dharran-viil p(<i1otripa 


^^H netra-mush proftphux 


^^H brahraa-dwish bnuplC-x 


^^1 The use of eu. and rftts in Greek, 

corresponds to that of * 

^H »i^<l '^"« in Sanskrit: thua. 



^^^^^k Bukaru 


^^^^^^H suUbhaa 


^^^^^^H duxtaras 

dustrnpos. , 

^^^^^^ duMalias 

dttspboros. fl 

^H . '^^ following arc iiigtances of the employment of a, on, ♦,« B 

^^^ »n privative, in the three languages: 


^^^^^|L '^•krlt. „ . 


^^^^^H . . Or««k, 

^^B agnoto. 

ignotus. ~ 

^^^^^H an-osios, 



The subjoined adjectivea are similarly formed in Sanskrit and J 

Latin from adverbs of time : 




' byutanu 

Lestcrnus. ^^^H 

^was tanas 





sempiternus. ^^^| 

The Ufle of various sort-H of conipouud words is. common to ^^( 

Sanskrit with Greek and Latin. Thus 

we have, ^^H 

■wukttt. Greek. Latin. 


trirutrain trinuctioD lriuo<:tiuui 

a period uf tbrce nights. ^^H 

vwapnokuroa bupnophorus Boiunifer 

bringing sleep. ^^H 

Bftdabbramos aciplanos. 

always wandering. ^^^| 

arindamas i]ip<>daiuoB 

fnc-, Stced-subdutiig. ^^^H 

devadattas ihetxlotos 

god-given. ^^^H 

uiahani&lia inegdometis mngnanimus 

higb-gouI«d. ^^^H 

bhuridhaoas poluchrusos 

very ricb. ^^^H 

bAhuinurttiB polumorphoa multiformis 

multifomi. ^^^H 

cbatushpad tetrnpous quadrupea 

four-footed. ^^^H 

MTUpas sumo^orpbos conformt« 

of the same form. ^^^| 

Forms in atut, nouns and adjectives : 



Omk. ^^H 


drepanon. ^^^H 




Forms in aJca or ika : 


Suilirlt. GrtNik. 


nayaku polciuikos 

medicus. ^^^H 

dhirmiku rhetorikua 

bellicus. ^^H 

Forms in ant: 





doloeis. ^^^P 


dolo«nta. ^^^| 

Snnskrit nouns ending in as, corresponding to Greek and ^ 

Latin nouns of the 3rd declension : 


Sjuukrit. GfMk. 

Utln. ^^H 

ayaa pscudoa 

feed us. ^^^H 

^^^ ya&u modoa 


^^ft apa» kSdoa 








In Greek and Latin the comparative and i$uperlatiTe 
are formed very much as in Sanskrit. The Greek bam, Iwv^ 
ever, two forms, like Sanskrit; the Latin only one. 

Latin. EofUah. H 

SiDikrit. Z«Dd. 

hbaUra husko 

bbadra-tara husko-tara 

bliadra-tama ^pentotama, 








bed US 









1 difierent oeto- 



In Greek and Latin, as in Sanskrit, verbs are oom| 

with prepositionB. 












con-vent t. 

In Latin, as m Sanskrit, verbs are 'compounded with 
or adjectives. 

ainikrft. LMo, 

parikhikaroti significat. 

krtahnikaroti magnificat. 

In Greek aud I^tin, adjectives agree in gender and nt 

with the uouu, just aa in Sanskrit : thus, 

Sanskrit. Greek. Latin. B^Itife, 

NoM. Slit. iiwiidQs swilpuus bOdua hupnos suavis somnus swiM litefk 

Acu, Sin. gwiidum awi^nilm hcJun bupiion suavcm somnum sweet iltff^ 

NoM. Plu. awnd&vna swapnaa Ledues bupnoi suaves somni sweet liecf*' 

NoM. Sin. n&.vo data neos dutcr norua dutor new giver. 

Acc. StM. niivuiu datar&m neon dolvra novum datorem new ginr. 

We must, therefore, conclude from the illustrations which hiW 
been given above, of the resemblances existing both in roots luKi 
inflections, between the Sanskrit, the Zend, the Greek, nod the 
I.Atin, (viewed in contrast with the almost total want of giiuiU' 
rity between the Sanskrit :md other tongiies, e. 4j. tlie .Ajubic,) 
that there is a close atliaity between the various membtf« of tbf 



former group of languages ; and that in fact, they are all de- 
Hcended from one commou utuck. 

It may, however, be objected that the affinity which I have 
been seeking to establish between the Sanskrit, the Greek, and 
the Latin is disproved by the fact that, (while a portion of the 
words in these laugiiagea are identical with, or akin to each 
her,) the great majority of their words are different. If these 
iges had in reality had a common origin, their vocabularies 
must, it may be urged, have been entirely homogeneous, i. e. mxist 
have consisted of the same identical words, just &r is the case 
vnth the Bengali, the Hindi, and the Mahratti, which are con- 
fessedly kindred dijdect& To this I reply, P'irst, that even a small 
proportion of conimon words, combined with great similarity in 
point of structure and inflection, is sufficient to demonstrate the 
common derivation of any two languages from one original stem, 
provided it can be :jhown (as it assuredly can in the case under con- 
sideration) that neither the words nor the inflections have been bor- 
rowed by the one language from the other. For how could tlie 
common possession by these two supposed languages, of even a 
small stock of words be otherwise accoimted for ? This community 
of words oould not be accidental ; for had there been anything of 
accident in the case, we should, lieyond a doubt, have discovered 
the same casual resemblances between oUier lanyumjes — between 
Saoakrit and Arabic for instance, or between Greek and Arabic 
— as we discover between Sanskrit and Greek ; whereas in point of 
fact we discover scarcely any such resemblances. The difterenc* 
between Sanskrit, Greek, and Latin, in regard to the large majority 
of the words of which their vocabularies are compocied, admits 
of an easy explanation. The speech of different branches of 
every gieat race of men, has (jw I have already in part shown 
in p. 42.) an inevitable tendency, arising from. a great variety of 
cauaes, to diverge more and more from tbe original type. This 
tendency is visible even in India itself, among men of the aame 
branch of the Arian family. The vocabu!»i-y of the Vedas is 
very different from that of the later Sauskrit writings. Many 

i 4 

words which are common in the former have been entirely dis- 
used in later times, while new words imknown in the Vedaa^ 
have been introduced. If the Nighantu be compared with the 
Amara-kosha, (which may be supposed to contain the words in 
most general use in later Sanskrit,) many nouns will be found in 
the former, which are wanting in the latter, as well as in all 
other more recent vocabuhkries. T may instance such words as 
tuvi, ' much ;' napnt, ' ofTspring ;' apas, ' work ;' gmd &nd jnM, 
* earth ;' keias, ' knowledge ;' dJcenipaSf ' wise;' takma, 'oflfepring;* 
&c. ; which occur in the Nighantu, but will be sought for in vain 
in the Amara-kosha. In fact, many of the words in the Ni- 
ghantu owe their insertion in that vocabulary entirely to the fact, 
that they had become obsolete in later times. Again, any one who 
is familiar with tlie different modern vernaculars of Indiii, must 
be aware how much they differ from each other, not only in their 
grammatical forms, but also, ft'efjuently, in the words themselves 
which are employed by preference in each to denote any particular 
objects. Now, as we have already seen (pp. 8. 56.), all these dia- 
lects must at one time have issvied from one and the same parent 
form of speech. But if such a divergence as this has actually taken 
place iu dialects spoken by the different branches of one people, 
living ill the same country, under nearly the same influences of soil 
and climate, and professLug tlie same religion ; must not a much 
wider divergence have of necessity arisen between the languages 
of tribe-s separated for thousands of yeais, and living in regions 
fjir apart from each nthor, under different physical conditions, and 
subject to the modifying action of different social, political, and 
religious institutions ? 

tSucb divergences between the languages of any two or more 
nations which have sprung from one common stock have, aa I 
have alrea^ly intimated, au inevitable tendency to become wider 
and more marked ; so that, two dialects derived from the same 
original form of speech, though they at first differed but little 
from each other, will thas almost necessarily become more and 
more dissimilar from each other, the longer they have been 
parated from the ptu*ent rout. 





Peculiar circumstances, such as constaut intercourse, and the 
poesession of a common religion and a common literature, may, 
indeed, for a period of greater or less duration avert such a 
gradual divergence in language between two separate nations. 
This state of tilings is at present actually exemplified in the case 
of England and America. But these two nations have only 
become separated from each other for a comparatively short 
period ; and it would be difficult to predict how long their 
identity of language may continue. So powerful, however, are 
the causes which operate in this to maintain an absolvite 
community of speech, that (notwithstanding the adoption in 
America of some new words, and a considerable number of 
phrases unknown in England) the two nations will, in all like- 
lihood, continue to etuplny the same dialect for many ages to 
come. Tliis result will, however, more probably arise from the 
English language undergoing a parallel alteration in both coun- 
tries, than from its continuing entix'ely unchanged in either. 

But we must be careful not to underrate the extent of the 
fundamental affinity in roots and words between the .Sanskrit, 
the Greek, the Latin, and the other western languages of the 
same family. Even a cursory examination of such works as 
Professor Benfey's " Greek-Kadical-Lexicon," '* is sufficient to 
show that these coincidences are more numerous than might at 
first sight have been supposed, and that it is only an insufficient 
study of the variations undergone by different words in the 
several languages under review which prevents our perceiving 
that a considerable, though prubably undeterminable, proportion 
of their vocabulary is essentially common to them all. 

But, Secondly, there is a further circumstance by which the 
original affinity between Sanskrit, Greek, and I^atin, and their 
ancient derivation from one parent stock are proved ; which is 
this, that it is precisely those words and forms which are the most 
primitive, the most fundamental, and the most essential parts of 

'* Uric-AiliJscbva Wurzcl Lckicoiu 2 vuls. Berlin, 1639 and 1842. 

260 AI'T^lrflTlffl! Am hmiVAfminSv [ceAr. w. 

each language which they have in comraon. I mean. First, thu^ 
words which expreas the natural relations of father, mother, &c,, 
and kiudreil generally; SecmuUy, the pronouns; Thirdly, the 
words expressing number, and Fourthly, the fomie of inflection. 
Thus, the words which Sanskrit has in comnaon with Latin, Greek, 
and tlie other members of the Indo-Ger manic stock are those which 
would be in iLse in the earliest stages of society, when men were 
simple and uniform in their habits and ideas, when they had few 
wants, few arti, little knowledge, no sciences, no philosophy, and no 
complicated institutions. But after the different tribes of the ludo- 
Germanic stock ha*l departed iu different directions, fiom their 
primevEkl abodes, and had settled in distant cotLQtries, they 
beciime in the C(jurse of time more and more different from each 
other in their religions, in their manners and customs, and in all 
their modes of life. The climates under which they lived were 
different; some settling within the torrid zone, while others 
migrated into temperate or even frigid latitudes. The aapecta of 
nature, too, were very dissimilar in these different r^ons, some 
of tbem being level and fertile, others moimtainous and unpro- 
ductive ; some situated on the shores of the ocean, and others 
at a distance inland. The natural productions of these different 
tracts, too, were various, as well as the animals by whom they were 
tenanted. Some of these countries, for instance, produced rice and 
the sugar cane, and were frequented by the elephant, the camel, the 
lion, and the tiger, while in others, these plants and animals were 
not indigenous. In consequence of all tJiese local influences, 
the temperaments and habits of the men of different nations 
became exceedingly diversified. In level and fertile countries with 
hot climates, men became less active and energetic, owing to the 
relaxing effects of the heat, and the diminished necessity for 
labour; while the frames of those who settled in colder countries 
were both braced by the greater severity of the climate, and by 
the necessity of labour for extracting a subsistence from tiie 
less genial soil. Men settled on the shores of the sea, naturally 
became addicted to uiaritime pmsiiits, from which those living 







inland were debarred. In this way different arts arose, different 
sciences were cultivated, and different social and jiolitical insti- 
tutions were established. In some countries the active energies 
of the people were fostered by the existence of free forms of 
government: in others, the feeling of independence, perhaps 
originally feeble, was altogether crushed by despotism ; while, on 
the other luind, the thoughtful tendencies wliich were native to 
the race, found their full scope in scientific pursuits, or in philoso- 
phical and religious contemplation. With these great and mani- 
fold changes in all the conditions of life, corresponding alterations 
in language, intended to express new objects and new ideas, would 
be necessarily introduced, which became more and more exten- 
sive and decided as centuries rolled on.™ The different stages of 
this process which I have been <lescribiug, are more or less dis- 
tinctly exemplitied in the different languages which have been 
specified as connected by affinity with the Sanskrit. Of these 
languages, the Zend (or language of the Zend Avesta) is that 
which had been seijarated from the Sanskrit for the shortest H{)ace 
of time, and subjected to the action of the smallest modifying 
influences, at the period when it took the form in which the most 
ancient of the extant Zoroastrian writings are composed ; and 
accordingly, it has a far closer resemblance to the Sanskrit 
than either the Greek or the Latin. Tliis will l>e made clear 
by the evidence which I shall hereafter adduce. The Greek and 
Latin languages, on the other hand, bad been separated from 
the Sanskrit for a much longer interval of time, and affected by 
novel influences of far greater potency, when they became embo- 
died in the oldest compositions which have descended to us; and 
they accordingly differ from the Sanskrit, in most respects, 
much more widely than the Zend does. 

^ The divergences, apparent or reAl, between the Arian langiugea, are due 
" to alterations, to loiisos occasioned hj the Inpae of time, and also to the in- 
cessant efforts (go to gpfok) of the language to replace the luat forms, und to 
follow step by step the gradual dcvelopiuenta of the several nalionalitics." — 
Pictet, ''Ori;;incs Indo-£uropceane6, ou lea Aryos Priiaitirs," p. 5. See Ap- 
pendix, Note A, 

I conclude, therefore, from the foregoing considenttions, that 
the differences which exist between the Sanskrit, the Greek, and 
the Latin languages, as we find them in their later stages, afford 
no reason for doubting that they had, at an earlier period, a much 
more intimate connection, and were, in fact, originally identical. 

Another objection may, however, perhaps be raised by some 
person looking at the subject from an Indian point of view. It 
is quite true, he may urge, that an aflinity exists between the 
Sanskrit, the Zend, the Greek, and the Latin ; but this quite tallies 
with what our hastras reward (Mami x. 43. 44 ; \'^ishnu-purana, iv. 
3,, p. 375 of Wilson's translation)"', that the Yavanas ( Greeks), 
Falilavas (Persians), and Kambojas, were originally Kshatiiya 
tribes, who became degraded by their separation from Brahmaus 
and Brahmaaieal institutions; and it is also quite clear from 
the proofs which you liave adduced of aflinity between these 
languages and our sacred tongue, that the former are mere 
Prakrit or Apabhran^ dialects, derived from Sanskrit. Your 
byputhesifl of these languages, as well as the Sanskrit, being 
derived from some earlier form of speech now no longer extant, 
is quite gratiutous ; for, what the heretical Bauddhas falsely say 
of their Apabhransa, which they call Pali, is literally true of Sans- 
krit, the language of the gods, that it is that primeval and eternal 
form of speech -' from which all others are derived. 

To this I reply that, even if Zend, Greek, and Latin could be 
shown, on the ground of their affinity with Sanskrit, to be 
derived from it, it would still be quite impossible for the objector 
to prove on the same ground, that Sanskrit was the parent of all 
the languages which are spoken by all the tribes which have 
inhabited India or tho adjacent countries. Arabic, as has been 
shown, is quite distinct from Sanskrit, and has scarcely any per- 
ceivable affinity with it of any kind. And the same is the case 
with the languages current in the south of India, the Tamil, the 
Telugu, the Cauarese, and the Malayalim (the tongues spoken by 

" See Part First oftliia Work, p. 177. 

** See Miihabbfishya, as tjuoLed ubovc, p. 170, 




the inhabitants of Dravida, Telinga, Karnata, &c.) For Mauu 
himself (as we have already seen p. 160), makes a distinction 
etween the languages employed by the people of India; which 
liows that forms of speech of a non-Arian i. e. non-Sanskrit 
character were spoken by part of the population. So that the 
point which the objector is, perhaps, really seeking to establish, 
viz., that the Arian- Indiana are theorigiiuil progenitors of all the 
surrounding nations, and their language, Sanskrit, the parent of 
all other languages, could never be proved. It cannot be ad- 
mitted, however, a» I have alreatly remarked, th.^t Greek and 
Latin are derivatives from Sanskrit There is no proof of this 
theory, and all probability is against it. The whole grammatical 
character of Greek and Latin is that of independent languages ; 
and any one who will compare their strnctiue and composition 
with that of tlie Indian Prakrits, which every one allows to be 
derived from Sanskrit, will at once perceive the difference of 
the two cases. 

First. — The g^rammatkal fon^is of the Prakrits (as we have 
already seen, p. 81), have evidently resulted from a disintegration 
or simplification uf the older Sanskrit forms. Thus the Sanskrit 
words vmktn, gupia, autra, mdrga, ariha, AresHha, drishti, 
jyushpa,<^al>'8hinn, riiailhifa, saij/a, hishnlm, laffhu, suxUm^ sabhd, 
are in Prakrit softened down into mutta, ffxUtn, »ntta, mufiga, 
attfutfSett.ha, di^.hi,pupphn,dakkhina, vtajjhcifSachcha, tunJiim., 
laJiu, adhti, and sahn. The further back we trace the Prakrit 
forms, the more nearly do they resemble the Sanskrit, till tliey 
are found to l.)e almost identical; while the more modern the gram- 
matical forms are which the Prakrits have taken, the more widely 
do they diverge from their Sanskrit prototypes. The case is 
quite different with the I^atin and Greek. A few instances 
may, no doubt, lie discovered where the modes in which the 
I^tin or Greek forms vai-y from the Sanskrit correspond in some 
degree to those changes of softening or simplification** which the 

" There are very few of the Prakrit forms which »re not simplifications 
of the Sanskrit. Even in such a case rb that of the word itthi, or iithiyn. 


SaDskrit forms have imdergone in Prakrit, Thus the Latin 
word /tw WHS, "the ground," diffi:'rs from the Sanskrit bhtmi 
in the same way that the Prakrit sahd differs from the Sanskrit 
8ab}id; the Greek doldkhoa "long," varies from the Sanskrit 
fUrgha somewhat in the same manner as the Prakrit liri or 
^^{7*1, vary from the Sanskrit »ri and hri ; and the Greek hnpno$ 
" sleep," appears to simplify the Sanskrit svapna, by much the 
same process as that hy which the Prakrit reduces the Sanskrit 
etJulna " place," to thdna. But the few instances which can be 
addiiced, are quite insufficient to prove that even in these cases 
the Greek or the Latin words are borrowed from the Sanskrit.** 
They may with quite equal probability have been derived from 

" woman "(from Bin), the change is a Binipliilcation, as one or more conEionanto 
are thrown oat-, and the vowel i is prefixed to facilitate pronunciation. But 
the great majority of Sanskrit words commencing with a double consonant 
are modified in Prakrit, not hy prefixing a vowel, but cither by rejecting one 
of the members of the compound consonant, or by int«rposii]g a vowel 
between them. Thus tbe Sanskrit a^a becomes in Prakrit tha,ttkala becomes 
Ihala, tkandha becomes hanilha, sprU becomes phtiru, tuhama becomes khama^ 
tnana becomes tthitna, meha becomes stttuha, mlana becomes malana. 

** It may, however, be further objected that my argument is incomplete, as 
all Prakrit or derivative dialects do not modify the original language in iJie 
name manner. Thus French and Spanish, it may be said, do not corrupt 
the Latin in the same way as Italian does, Now, as it has been stated above 
(p. 153) that the Indian Prakrits corrupted Sanskrit very much in the same 
way as the Italian corrupted Latin, so (the objector may urge) Zend, and 
Greek, and Latin, may have modified Sanskrit in a somewhat different way, 
as French and Spanish modified Latin. To this I reply that in the case of 
all these derivatives of Latin, viz. Italian, French, and Spanish, it can be 
shown (1.) that the people who spoke these languages wore either entirely or 
in port descended from (he Romans ; or that, at least, they received their 
language from the Romans who conquered and colonised their respective 
ooontries ; but it cannot be shown either that the Greeks or Romans were 
descended from the Indians, or in any way received their languages from 
Hindustban. (2.) In the case of the French and Spanish languages, as well 
a« in that of the Italian, the exact process and the very steps can be pointed 
out by which they changed the forms of the Latin words; but it cannot be 
shown, in regard to the Greek or Latin, fhat their words are in any wsy cor- 
ruptions of Sanskrit originals. 






an earlier language from which the Sanskrit is also drawn. 
There is no appearance of Greek and Latin words having re- 
sulted from any modification of the Sanskrit; for, while many 
of their forms have a close resemblance to the Sanskrit forms, 
they ore at the same time, for the moat part, equally original 
with those of that language ; and many of them are so different 
from the Sanskrit, and so peculiar, that they could not be de- 
duced from it according to any laws of mutation recognized by 
philologists. The Greek and Latin forms can, therefore, only 
be derived from another and anterior source, from which the 
Sanskrit forms also, as well as they, have flowed. It is, fiu^her, 
the opinion of distinguished comparative philologists, that 
Latin and Greek liave preserved some forms of inflection, 
which are more ancient than those preserved in Sanskrit; 
and represent more exactly the original forms of the sup- 
posed parent language. For instance, the Latin has pre- 
served the nominative of the present participle ending in 
em, such as /evens, (carrying), while Sanskrit has only the 
form in at, hharat for example, which seems to have been origi- 
nally bharcma or bharaid.^ The same is the case with various 
roots, nominal and verbal, in which the Sanskrit appears to have 
lost the original form of the wr>nl, while it has been preserved in 
Greek or Latin, or lx)th. Thus the Sanskrit word Uiray " a star," 
seems to have been originally atdra, a form which has been 
preserved in the Greek aster and astron, and in the Latin 
astntm, as well as in the Zend tttdre, and the Persian sitdnik. 
Again, it seems probable that the Sanskrit root bhanj, "to 
break," may have been originally bhranj, with an r, which 
has been preserved in the Lrfitin fi-ango, and the Greek 
reffnumi or prefjnumi. And the liatin forms of the prepositions 
»ub and supei; (corresponding to the Greek hxipo and htiper,) 
appear to be more ancient than the Sanskrit forms upa and 

" Bojpp, Comp. GnunraM-, ptra. 129. Ad. Rcgnier, Trtit^ de la fomi»- 
Uon des inoUi dans la langue Grecque, note 1, pp. 68, 69. 


jVffixities and derivation of 

CCBAP. tr. 

languages ore 
ly be furttier 

Second : But the fact that the Greek and L« 
in their origin independent of the Sanskrit 
shown by the fullowing considerations:"' 

• (1.) On a careful examination of the roots oontained in 
the Dhatupathas, or lists of radicals in the classical or modem 
Sanskrit, it will lie found that many of these verbal roots are 
compounded, or resolvable into simpler forms. But as those 
roots, notwithstanding their composite character, are treated 
by the Indian grammarians as idtimate radicals, it is clear that 
those gramraariaus have forgotten the simpler forms from which 
the others have been derived. Of this remark the following 
roots are exemplifications, viz. : vya/nj, vyay, x% vyadh^ pyush 
or ri/usk, pntsh, veksli, and itjhh, which, though evidently com- 
pounded of vi + anj, vi + ay, t'/ + i, vi + adJi, pi or vi+ush, 
pra + iish, va (or ava-*-lksh, ut-^hd {jahtlti), are yet treated 
by the Indian grammarians as if they were simple roots. 

• (2.) The Sanskrit has not only undergone alterations such 
as the above, but the modern language has actually lost some 
fuller forms of roots, which are still discoverable in the Vedic 
•hymns. As an instance of this may be mentioned the root 
grabh, (see al>ove, p. 228,) •* to seize," which in the modem 
Sanskrit has become prakrliised into gvah. Other instances 
are the Vedic dhurv, and dbtyrif as compared with the modem 
hvri; and the Vedic ^mdh, as compared with the modem 
^udlu The following Vedic roots are not to be found in 
modem Sanskrit at nil, viz." : «R^, TS » ^^» '^^ ^^> 

" I am tn(k'btcil fur the substanue of the paragraphs marked with an aste* 
risk (*) to the kindness of Profcasor GoMstiicker, who is dissatiafied vith 
the views propounded in the passage immediately preceding, aa he rejecti 
the theory vliieh has hitherto been in favour with philologists that the fuUett 
forms are the oliiest. Compare for tbo roots given in paragraph • (1.) 
Professor Benfey's " Complete Sanskrit Grammar," pp. 73, ff. 

*^ On (he hypothesis that the fuller form is the more ancient, I may alao 
cite the Vedic forms ichnm (as compared with the modern chant) and ichamd 
(ai compared with the modern chand), aa given in Professor Benfey's 
" Complete Grammar," p. 73. 





^^. 'JI^. W^> TSfv »T*^' ^^. ^. ?54^. H^. 

• (3.) But it is not only a fact that tlie modern Sanskrit has 
lost some of the oldest verbal roots ; the siiiae appears to be the 
case with the more ancient Veilic Sanskrit also, from which 
some primitive radicals had already disappeared. This is 
indicated by the circumstance that there erist certiuu Sanskrit 
nouns, which must have been derived from radicals which in 
their verbal form are not discoverable even in the Vedas, Thus 
from the existence of the word virudh, " a shrub," and nya- 
ffrodha (a particular tree), we may infer that there once existed 
a root rudli, *' to grow," which in this sense (for the modern 
Sanskrit has still rmlh in the sense of " to stop,") now survives 
only in its weakened form ru/t."* In like manner it appears 
from the nouns dhanns, "a bow," pra-dharuiy "battle," and 
ni-i-dhana, "death," that the root han, "to kill," must once 
have existed in the stronger form rf/ia7i = Greek, 6av. 

* (4.) Some of the verbal roots which have been lost by both 
the modern and the Vedic Sanskrit, and which cannot be traced 
there even through their preservation in derivatives, may yet be 
recovered from oblivion by the aid of the Greek or Latin. 
Thus tlie Sanskrit /iJi, " to sacrifice," must have originally ex- 
isted in the stronger form dhu, as we may infer from the 
Greek 6v<a; and in the same way the earliest form of the Sans- 
krit f/uh, "to hide," was probably ffudh, as the Greek Ktvdw 
would lead us to suppose. So too from the Greek forma 
vrjdo), " to spin," and \sixto, " to lick," we may argue that the 
original Sanskrit forms of nah and iih must have been nadh 
and ligh.^ Several forms of substantives and other words also 
can be shown, in which the Greek forms are stronger than the 

«'* See Piotet'« "Origines Indo-Europ6cnne«," p. 145. 

** So too the root duh " to milk," must have once been dugh, as is proved 
not only by its passive participle dugdha, but •Jso by the Zend fubatantive 
dttghdhar and tbc Greek Ovyarip, " daughter," a word which moat philolo- 



IcifAr. n. 

Sanskrit. Thus, instead of the Sanskrit kima, " winter," ahi, 
" a serpent," A ?/rt.8, " yesterday," we find in Greek the stronger 
forms \etfMav, ixis or o<^{r, ^Bes or ^x^es. 

From the facts detailed in the preceding paragraphs, which 
prove that compoimd root« have been taken by the Indian 
grammarians for simple ones, and that old forms have been 
modified or lost in the modern, or even in the Vedic, Sanskrit, it 
is clear that that language (especially in its modern form) can- 
not be always regarded as a fixed stAndard, according to which the 
origiudity of the Tiatin and Greek forms could be estimated. 
And the supposition that any of the Greek or Latin words" are 
borrowed from Sanskrit by a prakritizing process is satisfactorily 
disproved by the fact tliat various instances have been adduced 
of the very opposite natm-e, where the Greek and Latin forms, 
instead of being like the Prakrit ones, weaker or simpler than 
the Sanskrit, are stronger or more complex. For, whether or 


gists think, originally signified " milker." Professor Goldstiickcr Is of opinion 
that ail Ihc Sanskrit root» ending in h are weakened or prakritixed from 
stronger forms. TLiw he thinks gah, vrih, *prih, for instance, were once 
gddh, vridh, apndli, (compare sparddh). Dah, on the contrary, he con- 
ceives was once dadh, ns is shown by the substantive antardadhana ; but here 
we have also the noun nidtigha^ which would lead us to infer a form dagh. 

Conversely we sometiraes see the aspirated consonant of the root changed 
into A, as iu the case of the participle hUa {vi-hita, ni-hilA, &c.) from 
the root dhii, " to liolii." This weakening process commenced in S.-inBkTit has 
been perpetuated in Prakrit, where the aspirated consonants of Sanskrit are 
softened into h, as where the root kath " to say" becomes kah. SecYaromchi, 
il. 27. 

*' I except, of course, such words as have evidently passed from Saoakrit 
into Greek at a period comparatively modem ; such as capnarros from har- 
puna, and others of the same kind. But, on the other hand, a good many 
Greek words can be shown to liave been received into the Sanskrit astro- 
nomical literature within the two thousand years, such ns hora, kendra, 
iiptn, dfikana, anapha, naiaphS, apokiima, panaphara, jamitra, methurma, 
and rihpha, derived from Uie Greek upa, Ktvrpav, Xtrra, iuiavoc, avaf^^ 
ttvvnftf, AiratXtpa, lirnvafopa, tiofitrpov, fitoovpaviiiia, and pt^lf. — Colebrookc 
Misc. Ess. il. 526, ffi Weber. Ind. Stud. ii. 254. 





not the existence of these stronger or more complex foians in 
Greek and Latin proves that the Sanskrit once had similar fdrms, 
which have now disappeared, it is at least sufficient to neutralise 
the argument, — drawn from the presence of certain other stronger 
or more complex forms in Sanskrit than we encounter in the 
corresponding words in Greek and Latin, — that those languages 
are derived from Sanskrit : for, hy parity of reason, the pre- 
sence of some forms (which we havt? actually seen to exist) in 
Greek and Latin stronger or more complex than those dis- 
ooveirable in correspontling cases in Sanskrit, would prove that 
theee weaker Sanskrit forms were mere corruptions of the Greek 
and Latin words. 

Third : — The Indian Prakrits have derived by far their 
kirgeet stock of words from the Sanskrit ; the few which they 
contain that are not Sanskrit, having been derived from the 
langTiages of the indigenous tribes who inhabited Northern India 
before the arrival of the Aryas. On the other hand, ordy a certain 
proportion, as we have seen, of the words which compose the 
vocalnilai-y of the Greek and Latin languages are common to 
them with the Sanskrit : the greater part of the words are, 
if not different, at least, difficult to identify as the same. Now, 
had Latin and Greek been derived from the modern, or even 
from the Vedic Sanskrit, the number of words indisputably 
common to all three languages must have been very much 
greater. It is true that more may be said in favour of the 
hypothesis that the Zend has been derived from Sanskrit; but 
there are sxifficient reasons for believing that Zend is a sister 
and not a daughter of Sanskrit ; and consequently, that both 
have a common mother of a more primeval date. 

I therefore conclude, that Greek and Latin, as well as Zend, 
are not derived from Sanskrit, but have, together with it, grown 
out of some older parent language,"* which was superseded by its 

•* " An indubitable result of tUe rescorches which have recently been pur- 
sued into tlie Arian tonnes is that, notwithstanding the various alterations 

T 2 



[ciur. n. 

daughters, and became extitict, because it ceased to be emplo) cd 
a& a spoken tongue, and because (as being the language of a very 
early stage of society) it has not been preserved in any literary 
records. To render this supposition conceivable, I may remark 
that tlie same fate, — extinction — might have befallen the Sans- 
krit itself, and the Latin, when they, in like manner, gave birth 
to the various dialects which have 8U|jer8eded them as living 
and popular forms of speech, had it not been that they flourished 
at periods of much more advanced civilisation tlian the unknown 
primeval language to which I have referred, and have been 
perpetuated by means of the numerous writings, secular and 
sacred, of whieh they are tlie vehicles. 

The primitive language to which I have just alluded is thuH 
characterised by JI. Pictet, in the work above referred to, pp. 
1,2: "\Nniile thus au^nen(iag in numbers and in prosperity, 
that prolific race was labouring to create for itself, as a powerful 
means of development, a language admu-able by its richness, its 
force, its harmony, and the perfection of its forms ; a language 
in which were spontaneously reflected all its impressions, not 
merely its mild affections and its simple admiration, but also its 
nascent aspirations toward a higher world ; a language abound- 
ing in images and in intuitive ideas, bearing within it, in germ, 

which thej bnrc undergone, tbej all bear the dear impress of r>nc common 
tjpe, and are consequently descendcil from one real, living, primeval laa- 
j^uage, which was complete in itself, and wliich was employed hy a whole 
nation as its common organ of coiiimuiiication. I'hia is not a mere hypo- 
thesis deviled to cxfihiin the reluiiona by wh'ch Ihnse languages are con- 
nected wiih each other : it is a conclusion wbi<'h forces itself irresistibly 
on our belief, and which possesses all the validity of the best established 
fttct When vre perceive so large a number of languages, of a character so 
marked, converging in all the rlcCails of their structure towards a common 
centre in which every parlieuhir fuct 6ncls ils cause, it becomes impossible 
to admit that that centre has never hud any otlier than a purely imaginary 
existt'Dce, and that that marvellous agreement arises solely from an instinc- 
tive impulse peculiar to a certain race of men." — A. Pictet, Origines Indo- 
£urop6cnncs, p. 43. 






all the future .nffluence both of the most sublime poetry and of 
the most profound reflection. At first one and homogeneous, 
that language, already perfected to a very high degree, served as 
a common instrument of expres-sion to this primitive people, aa 
long as it continued within the limits of its native country." 

SecT. III. — TTiat I'ffiitity in larigvage implies nffinitij in race : nuhlr* in which 
a grtuter or lesx dirertity iif latigiiuge and iimlituiiong would arise in different 
branches of the lame stock: eeutnd Asia the birth-place of the Arijus. 

The fact* and considerations a^lductd in the preceding section 
have, I think, proved beyond a doubt that the Sanskrit language 
has a common origin with the Zend, the Greek, and the Latin ; 
and that all these tongues have sprung (like branches from one 
stem) out of tlie same parent language, now extinct. This con- 
cliisiou being e;itabli8he<l, it follows as a necessary corollfirj' that 
the Indians, the Persians, the Greeks, and the Komans also, that 
is, the nations who spoke the languages derived from that one 
common source, were themselves also descended from one and 
the same stock ; /. e., that they had for their common ancest^^rs 
the ancient people who spoke the extinct language to which I 
hare referred." It must, therefore, be taken for an established 
fact that the anc&stors of the Intlians at one time lived together 
with the ancestors of the Persians, the Greeks, and the liomaus, 
in one country, aa one nation. It is tnie that we have no histo- 

" " The radical offiuitr of all the Arian languages necessarily le&iTs ua," aa 
we have seen, " to regard them aa having tiprung from one single primitive 
language. . . ■ 

"Now as a language presupposes always a people to speak it, it further 
follows that all the Ariaii nations have isxued from one single vource, 
though they may hare become occasionally blended at a later perio<i with 
some (breign elements. Hence we mny witii certainty infer the existence, 
at a prc-historic period, of an Arian penple, free, originnlly, from all foreign 
intermixture, suCEciently numerous to have supplied those swarms of men 
which issulmI from its bosom, and sufEciently endowed by nntiire tn have 
created fur itself the most beautiful perhaps of all languages. This [teople, 
though unknown to tradition, ia in a certain degree revealed to us by phi- 
lological science."— I'ictet, pp. 5, 1?. 

T 3 



[CHAF. n. 

rical record of this primeval period ; but we are inevitably 1«1 
to assume the existence of au anterior state of things such as I 
have asserted, by the factj that no other supposition will account 
for the philological phenomena which we encounter in later his- 
tory. From the eflfecta, we are entitled to reason back to the 
existence of a cause adequate to their production. We have to 
explain the fact, that there have been found in dififerent regions 
of the world, lying far apart from each other, various nations 
speaking languages which evince au unmistakeable aflEinity to one 
another ; and there is no other mode of explaining this circum- 
stance but by supposing that these nations have radiated in dil- 
ferent directions from one central country which was their com- 
mon birthplace, and where their ancestors ail employed one 
common language. 

If we piu-sue our inquiries further, we shall find that oertain 
data exist by means of which we can discover with some proba- 
bility, what was the order of time in which these several nations 
sejwM'ated themselves from the original stock, and departed towards 
those new countries which they continued to occupy in later time& 

Let us suppose a numerous and powerful nation occupying an 
isolated position in the centre of a vast region as yet thinly 
peopled. Let us next imagine what would be the probable 
comse of affairs in such a society, and then compare this hypo- 
thetical picture with the traces remaloiug to us in history of the 
actual event's. We have first then to contemplate this origiual 
race, gifted with vigorous pow^ers both of mind and body,*" as 
still residing in its primitive abode. When in the natural order of 
events, this active and gifted race began to multiply, the coun- 
tries which they at first occupied either as shepherds or agricul- 
turists, would soon be found too narrow for tJic supply of their 
growing necessities.'* If fresh lauds fit for pasture or for cultiva- 

"* M. E. Renan, however, tbinks that the Ariau race was not originally 
superior in intelligence to the ScmitJc, Ilaiuitic, and other racea, but the 
contrary. Ilistnire des langues S^mitiqucs, p. 487. 

^ ^'iiut a constant and rapid increase of the population could not but 




tion existed unoccupied in the vicinity of their original territory, 
they would insensibly extend their borders as occaaion required. 
If there was not territory near at hand which would yield them 
a subsistence, the raore energetic and adventiu'ous members of 
the community would be driven by the pressure of necessity to 
inquire whether ampler posseesiona might not be found at a dis- 
tance ; and they would depart in larger or smaller detachments 
in quest of new abodes. This process of migration, when once 
commenced, would go on without intermission. The first adven- 
turers would be speedily followed by other successive bands till 
at length new nations were formed at a greater or less distance 
from the original country. 

The eailiest emigrant^!, who thus departed to distant regions, 
passing proliably through countries differing in climate and pro- 
ductions from their primeval abodes, eucountering new and strange 
objects, and inured to new pursuits, would gradually lose many 
of their ancient customs; and in exchange would acqvure new 
habits, and along with them, also, new modes of speech. Those 
portions of the origiual nation, on the contrary, which continiied 
to live together in their ancient country, or had gradually ex- 
tended themselves together over adjacent regions, would preserve 
more nearly their original customs, religion, and language. But 
length a period might arrive when the same causes which had 
icasioned the separation of the earlier emigrants, or some other 
causes of a different nature, would lead to a disruption in the 

speedily bring about graJua] migrations, which vould be directed townrda 
regions more and more diatant. From tliiit time forward the separation 
of the nation into dijstincl tribes, the greater infrequency of tommuuieation 
and changes in their modes of life, occasioned a certain number of dialects 
to spring forth out of this common language, and to develop* themselves, 
without, however, as yet, becoming detached from their primitive source ; 
and at the same time the original character of the race, becoming modified 
according to circumstances, gave birth to a variety of secondary national 
characteristics, destined, at a later period, to expand, to exhibit their own 
|ieculiar life and to play their part in the great drama of humanity." — 
I'ictet, p. 2. 


AFFErrriEs and deritation of 

tcHAr. n. 

remaining part of the nation also. It would become divided into 
different sections; which would separate from one another and 
establish themselves in different, but probably a«ljacent, countries, 
and would never exhibit so wide a divergence from each other in 
respect of their religion, their institutions, and their genera] 
character, an those earlier emigrant^! who had settled in regions 
at a greater distance. 

The first case wliich I have above hypothetically put is that of 
the Grreeks and Romans," who appear to have broken off at :ui 
early period from the great Arian nation and departed to the 
westward, io quest of new habitations. The distance of the coun- 
tries, viz., Greece, Italy, and the siuTOundiug proviucea, where 
they ultimately settled, firom the cradle of the Arian race, and 
their wide divergence ui religion and language from the eastern 
branches of the same stock, concur to jirove that they separated 
themselves from tlie latter at a very remote era. On the other 
hand the vicinity of the region occupied by the Greeks to that 
inhabited by the Romans, would lead us to 8upi>ose that the an- 
cestors of these two nations migi'ated from the east at about the 
same period, though the differences which we discover between 
the language and religion of the one people as compared with 
those of the other, compel us to assume a subsequent separation 
of the two, and au independent development of each. 

The second case which I liave above supposed, of two branches 
of the oi-iginal Arian stock continuing to live together for a con- 
siderable time after the other brauches ha<i become separated, is 
that of the Perso-Ariaus and the ludo-Arians. Both from the 
closer vicinity to each other of the coimtries in which the Persians 
and the Indians eventually settled, i. e., north-e^i-stem Per«ia 
and north-western India, and from the nearer affinity which we 
perceive between the language and the mythology of these two 





'* For the sake of simplifying the view I give of the questioo, I purposelj 
Jl iDcnlion of the German and olbor branches of tbia great family, uul 
periods at which they migrated westward. 



races than we find to exist between the language and mythology 
of either and those of the Greeks or the Romany, we are led to 
conclude that the ancestors of the Indians and Persians remained 
united ia one community (either in their primeval seats or in 
3me region further to the south) to a much later period than 
the other branches of the Arian race. 

The propositions, then, which I have already proved, or shall 
now attempt to prove are the following : — 

First : That the InJo-Arians, that is, the higher classes of the 
uorthem Indians, or tlie Brahmans, Kshattriyas, and Vaisyas, 
are descended from the same Arian race as the Fersiaus, the 
Greeks, and the Romans. 

Second: That tht- primeval abode of this origiiud Arian race 
was in some country of central Asia, situated out uf, and to the 
uurth-west of, Indifi. 

Third : That different branches gradually separated them- 
selves from this p.arent stock, and migrated to new countries, 
west, south, or east of their early home. 

Fourth : That the ancestors of the Indians and Persians ap- 
pear to have lived together as one nation to a later period than 
the other branches of the Arian race, but at lengtli separated, the 
Indo-Arians migrating into India, while the Perso-Ariaas occu- 
pied the territory of Bactria, and the adjacent provinces. 

I shall not consider it necessary, in the discussion of the sub- 
ject, to handle ea^li of these propositions iu the succession here 
indicated ; but shall rather take up the different topica in the 
order in which the procesti of proof which I shall follow may 
render most conveulcnt 

V Sbct. IV. — TTiiii there U no ohjection arising from jihytiolngical consi' 

H tUratioiu, tu clotting the Iitdiattt among the Indo-Eurupean race*. 

W In proving, as I have already done, that the Greek and Latin 

I languages have a common origin with the Sanskrit, I have 
I adduced the principal portion of the proof which I had to bring 



[COAF. U. 

forward of the common origin of the nations by which those 
several Ifttigiiages have been spoken. The mythologies of the 
Greeks and other western nations have, indeed, been considered 
to present Bome points of contact with that of India, as when the 
Erinnys of the Greeks h&a been identified with the Saranyu of 
the Vedas, tlie Centaurs with tlie Gandharvas, Minos with Iklana, 
Ribhu with Orpheus, Hermes with Sarameya, the Phlegjes with 
the Bhrigns, &c. ; " but it would carry me too far if I were 
to attempt to offer any account of the views which have been 
propounded on this subject. I will now therefore direct my 
atteuttou mainly to exhibiting at greater length the grounds 
which exist for supposing tliat the Persians and the Indians are 
descended from the same common ancestors ; and that, after 
remaining united together, as the constituent parts of one nation, 
for some time subsequent to the migration to the westward of 
the other branches of the same stock, they, too, were at last, broken 
up by the force of circumstances, into two distinct nations, which 
settled in two separate, though axljaceut, regions. I will subse- 
quently pass in review the additional reasons which can be ad- 
duced for supposing that the Indians immigrated into India 
from the north-west. 

Before, however, proceeding to carry out the intention here 
indicated, it will be expedient briefly to show that on phj'sio- 
logical grounds there is no reason for denying that the Indians 
are descended from the same stock as the nations of Europe. 
In their physical characteristics the lirahmanical and other high 
csiste Indians belong, as well as the other nations who have just 
been mentioned, to the so-called Caucasian type. It might, in- 
ileed, at first sight, be sujiposed that the dark-comjilexioned 
Hindus, could not possibly be of the same race as the fair- 
coloured natives of England or Germany. But a closer ex- 



•* See Kubn's Heralikunft dcs Feuers nnd des Gcittcr-trankg. Berlin, 
1059 ; and Miiller's Pupcr on L'oiuparalivc Mjdiulogj, in the Oxford Essays 
fur 1856. 



amination of the diflferent nations to whom, oa philological 
grounds, we are led to assign a common origin, will show that 
they vary in complexion very much accurdiug to the climatic 
influences of the regions in which they ultimately settled, and 
in which they have been resident for a long seriea of ages. If 
we look to the south-eastern and north-western extremities only 
of the vast tract over which the Indo-European races have 
spread, we shall, no doubt, find that there is a complete coutnust 
in point of colour between the occupants of those widely sepa- 
rated countries. But the same wide contrast does not exist 
between the inhabitants of those tracts, (included within the 
same limits,) which are adjacent to each other. The Indians 
do not differ very much in complexion from the Persians, nor 
the Persians from the Greeks, nor the Greeks from the Ita- 
lians, nor the Italiuus from the Germans or the Anglo-Saxons. 
These different nations alter in complexion by almost imper- 
ceptible shades varying nearly according as their respective 
coimtries range successively ftom south-east to north-west. 
While the Indians may be denominated black, the Persians are 
olive-coloured, the Greeks have a still fairer complexion with 
a ruddy tinge, and the Italians approach yet more nearly in 
hue to the Teutonic tribes. It is therefore to tlie varj-iug action 
of different climatic influences, that we have to ascribe the di- 
versity of colour which characterises these several nations. The 
torching rays of an luditui sim, the high temperature of an 
ndian climate, and the peculiai* diet afforded by an Indian soil, 
acting on the Indo-Arians during the long period of 3,000 
years or more since they first settled in Uindusthan, appear 
amply sufficient to account for the various peculiarities of com- 
plexion, of feature, and of corporeal ^■tructure which now distin- 
guish that section of the Indo-European family from the kindred 
branches to the west. In fact the action of these causes is 
Bufficiently conspicuous in India itself. The people of Bengal, 
who are of the same race as the iuhabitants of tiie north- western 
provinces, have, owing to the greater moisture of their climate. 




en* p. rt 

and the want of that bracing temperature which the latter enjoy 
for three or four months of every year, gradually become darker 
in complexion and less robust in their structure. Again it is 
notorious to every one who has lived in northern India, that 
a Brahman from the temperate province of Kashmir is far fairer 
than a Brahman of Mathurii or Benares ; in fact he has quite 
the look of a foreigner. It has also been observed that an 
Indo-Briton, or person partly of European and partly of Indian 
descent, becomes fairer from living in the colder climate of 
Europe : but immediately recovers his ancient complexion on 
being exposed again f o the heat of the tropics. It does not ap- 
pear necessary to enter further into the discussion of this sub- 
ject, as the preceding observations will suffice to remove any 
doubts iis to the common origin of the Indians and the nadons 
of Eiu-ofjc, wiiich may have arisen from their differences of com- 
plexion.'* I will only add that, if the considerations here urged 
have any foundation, the ludo-Arians must have been much 
fairer in complexion at the period of their first arrival in India, 
and while they still continued to occupy t'le north-westerly re- 
gions of the Panjab, than they became at a later period, when they 
had been longer exposed to the fierceness of the Indian sun, and 
when they had penetrated further t^ the south-eaat. And we ac- 
cordingly fiud that this iiupp^jsition tallies with some expressions 
in the Vedic hymns, the oldest of which, pjerhaps, date from that 
eai-ly period. Thus, in the text quoted from the Kig-vedA, iii. 
34. 9. in the First Part of this work p. 43, we find an allusion 
made to the colour of the Ariaii immigranta: "He desti'oyed 
the Dasyus, and protected the Arian colour : " and in R.-V. iL 
12. 4. the same word is applied to designate the Dasyu tribes: 
^ ?[T*i W^fVT ''TT ^'1 "He who swept away 
the base Dasa colour." Tiiough the word vaTna, "colour," 

'^ A full (liiicussioo of tbis subject may be found in Lassen'i Indlsche 
Ahcrthiinifkunde, j. <00— 409. Ste also A. W. von Schclj^el, Essais, pp. 
46G, (T, niid Miiller's " Last R(3«ul(s uf the Turnnian Rtisearckea," in Dunsen'a 

utlinca of llic I'Lil. uf Uuiv. Hist., vol. iii. pp. .')43 — 353, 

•BCT. v.] 


which is here employed, came afterwards to be current as the 
designation of caste, there is every reason to suppose that it 
wa« origiuiilly used tu discriminate the fair-coloured Aryaa from 
tlie dark-complexioned aborigines. But such a term of contrast 
if employed now, would not possess half the force which it did at 
a time when we must suppose the distinction of colour between 
the Arj'as and the savage tril>e8 whom they tncouutercd, to have 
been far more palpable than it is in modern times. 

SsCT. V. — Reason* f<nr Muppoting the Indiana and Pertian* in particular to 
hare a comnum origin, 

I will now proceed to indicate the various grounds which exist 
for concluding that the Indians and the Persians, or Iranians 
are not only descended from the same original stock ; but that 
they continued to form one community even after the other 
kintired tribes had separated from them and migrated to distant 

The first proof is the closer afl&nity which, as we have already 
seen, subsists between the Zend, the language of the ancient Per- 
giana,"* and the Sanskint. From the examples of resemblance 
both in roots and inflections which have been adduced in Section 
II., it is manifest that, upon the whole, the Zend is more nearly 
related to the Sanskrit, than either the Greek or the Latin are. 
It is true that in tlie liat-ti of parallel words wluch have been 
there brought forward, the paralli'l Zend words have been often 
omitted, while the Greek and Latin words have been adduced : 
but this does not arise from the Zend forms having had no 
existence, but either from their not having been discovered in 
any of the extant Zend texts; or from their not being readily 
accessible. But the Zend words which have been brought for- 
ward will be generally foimd to stand in a relation of closer 
resemblance to the Sanskrit than either the corresponding 

** For an account of the various old Iranian dioJectB, see Spiegel in Kuha 
•nd Schleicher's Beitriigo zur vcrg. Sprachf. ii. 6, ff, and App. note 0. 



Greek or Latin words do. I eubjoin some further compara- 
tive lists of Zend and Sanskrit vocables to which the Greek 
and Latin either offer no equivalents in form, or equivalents 
which generally bear a much more distant . resemblance to the 
Sanskrit than the Zend words present. These lists are the 
following : 





f whom 
l{tlftt. pi.) 
jedht, yez5 if 
aozo splendour 

mithwana a pair 






atharTanam iitbraTan^m prieet(acc.) 
adhwinam adhwanSm road (arc.) 




n. TERBS. 








r karoiti 
I karenoiti 

be does 



be is 



he kindles 



he did 

pradeSayam fradaesaem 

I enjoined 



he saw 




T we give 






they were 

fperf. part. 

■an ton) *^ 


/ being 

L (ace. mas.) 

nom. fern.) 




I praise 



he carries 



he kills 



he goea 

ySz { 

to sacrifice 



he is 





they are 



I love, praise 



he gives 

bhavishyanti bnafayainti 

they shall be 



'praise (2 J 
• per«, imp.) 

With the preceding lists should be compared the comparative 
table of Sanskrit and Persian words given above in p. 228, ff,, 
which will contribute to supply their deficiencies. Many Persian 
words will there be found which in form closely resemble the 
Sanskrit terms having the same signification, while on the other 
hand there are in numerous instances no Greek or Latin nouns 
which closely corre.'^ond t<» the same Sanskrit words lK)th in sound 
and in sense. Now if even the modern Persian language, not- 
witlistanding the many modifications it has undergone from di- 
verse influences throughout a long course of centuries, can still 
supply so large a number of words which so closely resemble the 
Sanskrit terms, we may safely conclude that the Zend or early 
Persian, (wViich was the ancient medium through which the 
modem Persian derived all the Arian words which it possesses,) 
must, itself have contained a far larger number of words bearing 
a very much closer resemblance to the Sanskrit. 

These views receive confirmation from the following remarks 

** In Greek, fsan. 

*" In Greek, onto. 



of Professor Miiller in his, " Last Besnlta of the Persian Re- 
searches,'' pp. Ill, 112. : — 

*' It is clear framhis (M. E. Burnoufs) works and from Bopp's 
valuable remarks in his Comparative Grammar, that Zend in 
itfl grammar and dictionary is nearer to Sanskrit than any other 
Indo-European lan^iage. "Many Zend word^ can be re-trans- 
late«i into Sanskrit simply by changing the Zend into their cor- 
responding forms in Sanskrit .... WTiere Sanskrit diflfers in 
words or grammatiwil peculiarities from the northern memben 
of the Arian family, it frequently coincides with Zend. The 
numerals are the same in all these languages up to 100. The 
name for thousand, however, (aahasra) is peculiar to Sanskrit, 
and does not occur in any of the Indo-European dialects except 

in Zend, where it becomes hazanra These facts are 

full of historical meaning; and with regard to Zend and Sans- 
krit, they prove tliat these two languages continued together 
long after they were separated from the common ludu-European 

The second argimient in support of the proposition I have 
undertaken to prove is, that both of the nations in question, 
viz., the Indiana and the Persians, apply to themselves, in their 
earliest written recordfi, the same name of Aryas. 

The Vedas are, as I have already shown, the oldest of all the 
Indian books. They are, therefore, not only the most authentic 
source of information in regard to the earliest language of the 
Indians, but there vs every probability that they woidd preserve 
more distinct and exact traces of their primeval history than we 
find in the other J^astras, which were composed at a later period, 
when tlie most genuine traditions of the origin of the race 
had been obscured anil cornipted. From the Vedic hymns 
accordingly it does, in fact, appear more distinctly than from 
any other of the Indian writings, that the progenitors of the 
Hindus were originally called Aryas. We find this name ap- 
pUed to the forefathers of the higher classes among the Indians 
(in contradistinction to the Dasyus, who appear to have been 



«1CT. v.] 



a people of a different race, and to have been settled in India 
lieforc the Aryas), in such passages of the Vedas as the follow- 
ing : Rig-veda i. 5 1. 8, *' Distinguish between the Aryas and those 
who ai'e Dasyus ; chastising those who observe no sacred rites, 
subject them to the sacrificer." R-V. i. 103. 3, " Indra, thun- 
derer, who art wise, hurl thy shaft against the Dasjni, and aug- 
ment the might and glory of the Arya."' " 

By means of this word Aryj'a, then, we are able to connect 
the early Hindus, with the early Persians, For, first, it appears 
that in ancient times tlte Medes also (who were eventually in- 
cluded in one empire with the Persians) bore the name of Arians. 
This is clear from the following passage of the ancient Greek 
historian Herodotus, who narrated the wars of the Greeks and 
Persians, In the Seventh Book of his history, Sect 62, we have 
"the following statement: — 'Exakiovro Be irdXai irpot Trdmav 
'Apioi' aiTiKOfiivTji Ke Mi/8fi»jy T^y KoX;^i'Sop if 'AOrji'^eov is Toi/s 
*Apiovs rovTOvt, fisre^aXov xal oinoi to ovvofia ' avroi Si vepl 
a-(f)ewv uBe Xtyovai "MrfSoi. "They (the Medes) were formerly 
iCalled Arians by all. But when the Colchian Medea arrived 

dong these Arians from Athens, they also changed their 
name. n>e Medians say these things of themselves." A nation 
or tribe bearing the name of Arians is mentioned by Herodotus 
in Sect. 66 of the same book. 'Apiot hs ro^otat fisv ivKSva- 
KTfjJvoi rfcav MTySmcoto-*, ra he a\Ka fcard nep HaxTpioi • 'Apuov Bi 
ijpXIi "Siiaiifivr}?. ** The Arii were armed with Median bows, 
but in other respects like the Buctrians. The Arii were com- 
manded by Sisamnes." These last mentioned Arians appear to 
have dwelt in the neighbourhood of Herat. (.See Bahr's Herod, 
iii. 93, and viL 62.) llie same tribe is mentioned by the same 
author as paying 300 talents tribute along with the Parthians, 
Chorasmians, and Sogdians : Ildpdoi Be xal Xopdafuoi Kai ^oySot 
Tc Kai "Apeioi Tpuumata idXaina (iii 93 ). The same people are 

** The original psuoges, with many other Rimilnr ones, will be cited 
further on. , 

mentioned by Arrian (iii, 8. 4) as forming part of the ariAy df 
Darius : XaTi^ap^dvris Ss 6 ^Apei'av traTpatrrjs 'Apslovs fpft. TIjc 
Arizanti nre specified, Herod, i. 101^ as one of the seven Median 
tribes. In Herodotus we further find several proper naraes 
which are compounded with the word Ariiis; thus, vii. 67, the 
commander of the Kaspians is called Ariomardus. In the 78tlj 
cliapter of the same book, another person of the same name, aii<l 
son of Darius is mentioned. In other passages of the same writer 
and other ancient authors (viz, Xenopbon, Polybius, Arrian, and 
Quiiitus Curtius), such names aa Ariabignes, Ariaramnes, Arinces^ 
Ariaius, Arimazes and Ariai-athes ( = Aryaratha), are aisigned to 
Persians. The word "Xpiov, which occurs in the ancient Cfreek 
dramatist -iEschylus, Choephoroi, verse 423, (tKo^a KOfLfuiv 
"Apiov, &c., " I have chaunted a Persian dirge,") is interpreted 
by tlie Bchohast on the passage as equivalent to Uepaimv, 
" Persian." 

But, secondly, it is not only in the Greek authors that we 6nd 
the name of Ariana aytjilied to the Medes or Persians ; in tlie 
mo8t ancient books of the Zoroastiian religion also, which are 
c<^m]K)sed in the Zend language, the same word, as a designa- 
tion of the early Persians, is of frequent occurrence. I give, in 
a somewhat abridgwl form. Professor Spiegel's abstract of the 
evidence which exists of the common origin of the Indians 
and Persians, as one of the most recent and complete. (See 
his translation of the Avesta, vol. i. Introduction, pp. 4, flF.) One 
part, of this evidence is their common name of Arya. 

" Ethnograplty, supported by her two handmaids, physiology 
and philology, has in recent times demonstrated that a single 
race (the Indo-Germanic) has spread its branches over the whole 
space from India to the most westerly point of Europe. The 
mast higlily gifted antl civiUsed nations, both of the ancient and 
modern world, are all derived from this stock ; viz., the Indians, 
Persians, Greeks, Romans, Germans, Slavonians, and perhaps 
also the Celts. All these nations are branches of one single 
igiual family, whose abotlea have not yet been certainly de- 

«tcT. r.] 



termined, and perhaps will never be ascertained in a way to pre- 
clude all dispute ; but it is probable that, in the earliest times, all 
theee races dwelt together as one people, on the elevated table- 
land of central Asia. The einigratiou of this people from their 
original scats, and their separation into dift'erent ))rancbe8, are 
eventa which He anterior to all history. Faint indications of the 
degree of culture possessed by these tribes before their separation 
may be derived from the terms for particular conceptions which 
have remained common to them all ; and tlie amount of their 
knowledge is not to be estimated too low. If the state had not 
been organised by them, the family, at least, had been already 
regidattxi, as is proved by the community of the words designating 
relationships. We find names common to them for the different 
kinds of cattle, and for different implements of husbandry. 
Their conceptions of the gods, on the contrary, seem to have con- 
tinued to be of the most general character.*' 

" But in a<ldition to this possession by the whole Indo-Ger- 
iDKnic race of particular words, there exists a closer relationship 
between single meml>ers of this family. This closer relationship 
is to be explained by the fact, that some of these races continued 
to live together even after the others had separated from them. 
Thus, for example, the Greeks and Romans have much that ia 
common to both in their languages and in their ideas, which can- 
not 1)6 explained by their original rtjlationship. But in no instance 
is this affinity more striking or intimate than between the Indiana 
and the Persians. These two bnuiches must have lived long to- 
gether after quitting their common cradle, as is clearly proved by 

« See Kuhn's DisserUtion in Weber'a Ind. Stud. i. 321, ff. The elaborala 
work of M. Adolphe Pictet, above ({uoted (p. 267), has for its object to di«- 
co»cr, by a comparison of the primitive words uommon to »11 tlie Arian 
nations, what was their originiil urid cumiuon country, nnd wlnit the condition 
of tlie parent nation ds regarded its civilisution nnd its iiitkllccluiU and re- 
ligious ctdture before the sepurution of the several branches. The first 
volume, relating to the physical characters of tljc country, has alone aji- 
IMAred aayet 

r t 




cu&r. n. 

linguistic and mythological considerations. The three dialects of 
ancient Persian with which we are acquainted, viz., that of the 
cuneiform inscriptions, that of the second part of the Yasna, and 
the Ituaguage of the reinaining portions of the Avesta, have all 
Buch a close affinity to the ttldest Indian hinguage, the Sanskrit, 
exhibited in the Vedas, that they might almost be all called 
ialecta of one and tlie same language. Other grounds, par 
ticularly of a mythological character, speak no less strongly in 
proof of the two tribes (Indians and Persians) having adhered 
long to each other. It is of especial importance that they both 
eall themselves by the same name. Arya, signifying honourabU, 
in ordinary speech, and derived from ai'ya, which means lord in 
the Vedas, is the most usual, and the most ancient name of the 
Indian people. (R-V. i. 51, & Sama-V L 1. 1. 5.) Among the 
Indians the term Mlechlia, which denotes an impure barbarian, 
is the opposite of Arya, The same is the case among the 
Persians. According to the Persian laws of euphony, 0/7/(1 had 
to be changed to alrya, a name which the Persians long applied 
to themselves, and out of which the more mwiern Iran has 
arisen; a name, too, with which Herodotus had become ac- 
quainted. To this word a'n-ya, another word, aiiairya, or noa- 
irauic, is opposed. 

" It is, however, establisJied tltat this original Arian race, from' 
whicli, at a later period, the Indians and Persians separated 
themselves, cannot have lived as one community either in 
India or in Persia. We must regard it as demonstrated that 
the Indians wljo spoke Sanskrit were not autochthonous in Hin- 
dusthan. The oldest seats of the Indians of which we find any 
mention made, are to lie placed in the Panjab. In the First 
Fargard of the Vendidad, verse 73, a coimtry called Hapta 
Hendu, or India, is mentioned, which, in the cuneiform inscrip- 
tions, is called Hidus, It was not untJerstood for a long time 
what was signified liy Hapta Hendu, Seven-Iudias, but the 
Vedaa have explained this name. In the Vedic hymns we find 
the name Sapt«-Sindhava.s, the seven rivers, still employed to 

SECT, v.] 



designate the country of the Indians.** From the Panjab, the 
Indians, as their later books testify, advanced further towards 
the east ; first, as for as the Sarasvati ; after which, they spread 
themselves over the whole of northern India; and only at a late 
period into the south of the peninsula. The Persian legends 
conduct us with the same distinctness to a primeval country ia 
the north." 

We shall shortly have to return to this question, and inqiiire 
what were the primeval seats of the Arians. In the mean time, 
I revert to the affinities of the Persians and Indians. 

The third proof of this which I have to adduce is, the coinci- 
dences which are discoverable in the ancient mythologies of the 
two nations. On this subject. Professor Spiegel proceeds, pp. 
6, ff. : — " We have already said that the different branches of 
the Ariau family had but few words in common connected with 
theology. The most widely diffused term is the general desig- 
nation of God as the ehining, formed from the ancient root, 
div or di/u, ' to shine.' From this is derived the Sanskrit deva^ 
the Latin deiis, the Lithuanian dievas, the German zio and tyr, 
the Greek Ztvs, and also Jupiter from Diespiier. The old Persian 
daeva belongs to the same root, but has (on what grounds we 
shall presently discover) a somewhat different signification. 
More intimate mythological affinities are only to be found be- 
tween particular branches of the Indo-Germanic family, as 
between the Greeks and Romans, and in particular, between 
the Indians and Persians. A numlier of personages found in 
the Veda correspond in name with others in the Avesta, and must 
originally have been completely identical, though in the course of 
eventfi, it has naturally occurred that this similarity has become 
more or less efface<i One personage whose identity was the 
first to attract attention, is the Yama of the Indians [the son of 
Vivasvat], the Yima of the Persians [who is the son of Vivanghat], 
Id the Vedas and Upauishads we already meet with Yama aa the 
King of the Dead. He inhabits a particular world, where he ha& 

** Spiegel, AveiiU, vol. i. p. 66, uoU- 3. 
V a 

asseniblecl the iraniortals around liim. Among the ancient In- 
dians Ills worli.1 iH not a place of terrors, but its expanses are full 
of light, and the abodes of happiness, pleasure, and rapture.** 
In Iran, Yima is a fortunate monarch, under whose rule there 
was neither deatli nor sickness. Aft€r he haa for some time con- 
tinued to diflfiise happiness and immortality, he is obliged to 
withdraw with his attendants to a more contracted space, on ao- 
count of the calamities which threaten the world. Here lies, 
according to my view, the point of connection between the two 
legends. The Indian regards Yama simply aa the king of the 
dead, or, at least, of the blessed : the Persian limits the number 
of the blessed to a determinate number, who are selected to live 
with Yima, 

" A second renowned personage in the Persian heroic poetry, 
who also occurs iu the \'edjia, is Thraetaono, the descendant of 
Atliwya, the Fredun, or Feridun of a later period, with whom 
the Trita of the Yetla is connected. Trita is the son of Aptya, 
and, according to the Vedic account*, he fights with a seqjent, 
aud smites the three-heatled dragon with seven tails, and liberates 
the cattle. Quite similarly, Thraetaono destroys a pestilent ser- 
pent with three heads, tliree girdles, six tails, and a thousand 

" A third personage, ivlio can be pointed out in both the 
Indian and Persian mytliology, is Suma KeresiLipa, tlie man of 
heroic temper, and the same as the Krisasva of the Indians, who, 
it is tnie, has not yet been discovered iu the Yeda, but who was 
known to the Indian grammarian Piiyini, aud is frequently 
named in the Purauas as a warlike rishi. (Ramay. i. 23. 12 
^hl^. ; L 31. 10, Gorres.) 

** To these three personages may now be added a fourth, Kava 
U^, or Kavya Uwrnas of the Yedas. This is the person called 
KauB at a later period iu tiie Persian legends. Unfortunately, 

** See R.-V. ix. 113. 7—11. quoted by Rotli in the Journal of the 
(lOtiiaii Orient. Society, iv. 4^6, ff The original paasage will be given 
ju iLe Ajip. note C. 

•BCT. v.] 



the Btories of Kava U4 are so few and so brief, that I can 
scarcely venture to indicate their connectioii more in detalL 
(See Weber ^'aja.-8. Spec. II. 68 note.) 

" In addition to this identity of persouEigeB, we find also that 
the Indians and Perdans have some Important ceremonies in 
common. We shall here only mention two, though a closer 
examination of the Persian liturgy will uo duuLt bring others to 
light. The first is the Soma or Homa oflft'ring. (See also vol. 
ii. 69.) In both the Indian and the Persian religions, soma, or 
luLoma, vrhich is identical with it, is the name of a plant, the 
juice of which is pressed out and dmnk, with certain religious 
forms ; and in both religions Soma is also a god.** Soma and 
Uaoma have also a great number of epithets common to then), 
which clearly show how short a period had elapsed since the 
Persian and Indian adherents of this worship had become sepa- 
rated frum each other." 

The Indians and Persians have also some other of their deities 
in common, e.t/., Mittra and Aryaman. 

" In the Veda,'" (says Dr. F. Windischmann, Mithra, p. 54. 
66. and 63), " Slittra occurs as the son of Aditi (boundless 
space), and hence parallel with the sun, and stands almost 
always inseparably associated with Varuna. He appears to 
belong to a race of gods who are already disappearing, and has 
resigned a portion of his fimctions to Indra. In the Veda 
Mitra is the light, while Varuna is to be understood of the 
sky, especially the nocturnal sky. The connection of Mitra and 
Varuna in the Veda is analogous to that of Mithra and Vayu in 
the Zend texts. IVIithra is thus an ancient national god of the 
Arians ; and the chimtcter under which he is represented in the 
Zend Avesta has many points of resemblance to the Vedic 
Mitti-a, though it has also essential diflferences of Zoroastriiui 
origin. Aryaman, who is to be understood of the sun, appears 
in II.-V. i. 36. 4, and elsewhere along with Mitra and Varuna. 

** S«e WindischmuiOi Ueber den Somacultui der Arier; aad App- noteD. 


His name signifies companion or friend, and he also occuth in 
the Zend texts." Spiegel, (in liis note to the 22nd Fargard, 
vol. i. p. 266), says of him, " It ia to be lament«^d that the god 
who is here designated by the name of Airyania occurs but 
seldom, and is but briefly noticed in the Avesta; for he is 
unquestionably the ancient Indo-Germanic deity, who is men- 
tioned in the Vedas under the name of Aryaman.'' 

I proceed with my tiuotation from Professor Spiegel's Intro- 
duction, i. 8 : Secondly, " The reception of neophytes into the 
sacred society is performed among both peoples, the Persians 
and the Indians, by investing them with a girdle or thread. In 
the case of a Brahman the investitiure is to be performed in 
the eighth year after his birth or conception, in the case of a 
Kshattriya in the eleventh, and of a Vaisya in the twelfth. 
Hut the period of investiture for a Brahman has not finally 
expired till his sbcteenth year, for a Kshatriya till his twenty- 
second, or for a Vaisya till his twenty-fourth.*^ After the inves- 
titure, the teacher is to instruct the pupil in reading the Vedas, 
and in the rites of purification. (Manu, ii. 69 ; Yajnavalkya, 
i. 15), Up to his seventh year the Parsee is incapable of doing 
any evil ; and if he does any thing wrong, the blame of it falls 
on his parents. In India he is invested with the Kosti or 
sacred girdle in his seventh year ; among the Parsees who live 
in Kirman, the ceremony is postponed till the tenth year. 
From the seventh to the tenth year, half the blame of the 
ofFt-nceH wliich the child commits, falls upon his parents. With 
his tenth year the boy, according to the view of the Ravacts, 
enters formally into the community of the Paraees ; according 




A^waliijan& G|'ihja-Sulra, i. 20. 


^lAv^ %^^ ^f^^ ^rr^ l"»a[^i ^rr ft^in^ 

i»CT. v.] 



to other books the fifteenth year appears to be that in which he 
is admitted into religious fellowship. 

" All these traces of a common development which we have 
just pointed out between the Indians and the Persians have 
their origin, of course, in a pre-hlstorical period, when both 
nations lived together undivided. Traces can also be discovered 
which lead to the conclusion that the separation of these two 
races was occasioned, in part at least, by religious causes." 
Even if it have been accidental that Aimra, the highest god 
of the Persians, was, under the designation of Asura,*' reckoned 
among evil spirits by the later Indiiins, it can scarcely have 
happened by chance that tlie Devas of the Indiana have, under 
the name of Daevas, been transformed into evil spirits and 
allies of Angra JIainyus ; that Indra, the highest god of the 
earliest Hinduism, is in like manner, banished to hell ; and 
that ^ar>'a occurs as an evil sjnrit, while the Indians have con- 
sidered this name worthy to be a designation of Siva, one 
of the three highest deities of the later fonn of their religion. 
The conjectiu-e is therefore not unnatural that religious differ- 
ences may have been one of the groimds of separation. Still, 
even after their separation, tlie Indians and Persians did not 
remain without some knowle»lge of each other's progress. They 
were not too far separated to render this possible; and the 

** In bis second volume, however, Professor Spiegel adds, on tbis subject, 
the following reservation : — "la ihc first vol. I have alluded to a religious 
alienation ; but loo much importance is not to be ascribed to this view, and 
no adventurous bjpotUescs should be built upon it. Even without the 
assumption of a religious alienation it is quite conceivable bow goila, who were 
held in honour by the one people, should be degraded to the Infernul regions 

bjr the other That which gives probabilitj to the assumption of 

an RCtuai alienation between the Indians and the Iranians on account of their 
religious conceptions, is the ftict that the number of these opposing concep* 
tions is not inconsiderable," pp. cix. ex. 

*• Derived from am = pra/nJ, "wisdom," in the Nighantu. The word anra 
boa also a good sense in Vedic Sanskrit ; it means tarectham prmadak. 
Comp. Siijrana on R.-V. xxxv. 7. 10. 



LCKAF. n. 

Vendidad (L 74) still Khows an acquaintance with India tinder 
the name of Hapta-Heudu, i.e. Sapta Sindhavah, the laud of the 
seven rivers, which was a designation of the Vedic ludia." 

On the eame subject Professer Miiller remarks : " Still more 
striking is the similarity between Persia and India in religion 
and njythology. Gods unknown to any Indo-European nation 
are worshipped under the same names in Sansskrit and Zend ; 
and the change of some of the most sacred expreesions in 
Sanskrit into names of evil spirits in Zend, only serves to 
strengthen the conviction that we have here the usual traces of 
a schism which separated a community that had once been 
muted." (Last Kesults of Persian Researches, p. 112.) 

From the three-fold argument above stated, drawn — (let) 
from the striking BimUarity between the Sanskrit and Zend, 
(2nd) from the common name of Arya, which 1 have shown to 
be applied to themselves by both the Indians and the Iranianfly 
and (3rd) from the coincidences between the religion and m3rtho- 
logy of these two nations, I conceive that a powerful confirmation 
is derived to the conclusiou which I have been endeavouring to 
establish, namely, the common origin of all the nations to which 
the name of Indo-European has been applied. If even from 
philological considerations alone we are entitled to assume the 
descent of the Indians, Iranians, Greeks, and Komaus, from the 
same common ancestors, our general conclusion is very greatly 
strengthened when we can (in the case of two of these nations) 
add to the arguments founded on language, a variety of others 
derivable from community of name, and, to a certain extent, 
of tradition and of mythology. 



Sect. VI. — Was India the Primitive Coimtry of Ote Aryyas or Indo-European 

race* f 

As we have been led by the preceding investigation to conclude 
(1.) that the Sanskrit, the Zend, the Greek, and the Latin lan- 
guages must all have had a common origin ; (2.) that the races 
abo who employed these several languages were all branches of 





one great family; aud (3.) that consequently these difTereut 
braQches rauut at one time have lived together as one na- 
tion in one country : — we have now to determine what that 
country was. First, then, waa India the common cradle of tlie 
Indo-Germauic races, and did the other branches of that great 
family all migrate westward from Ilindiisthan, while the ludo- 
Arians remained in their primeval abodes ? or. Secondly, are 
we to assume some other country as the point from which the 
several branches of the race issued forth in different directions 
to the various countries which they eventually occupied? 

Air. A. Curzon maintains*" the first of these two theories, viz. 
that India was the original coimtry of the Arian family, from 
which its different branches emigrated to the north-west and in 
other directions. 

The opinion that the Ariana are a people of an origin foreign 
to the soil of India, which they are presumed to have invaded 
and conquered, impjosiug their religion and institutions on the 
so-called aborigines, is rejected by him as one founded on very 
insufRcieut data, and as resting on no well-established historical 
grounds. He thinks that it is a course opposed to the evidence 
of facts based on the results of comparative philology to main- 
tain tliat the barljarous aborigiiiai tribes of India, destitute of 
written records, traditional religious .system, or well-defined in- 
stitutions, can be more ancient than tlie Arian -Hindiis, the 
poaeHors of an early civilisation. These rude tribes may, in 
his opinion, have sprimg from some of the barbaric hordes, who, 
under the name of S^akas, niiuas, &c. are mentioned by Sanskrit 
writers as having invaded ludiii, and who, after their defeat, may 
have taken refuge in the hills and forests of Uindusthan. 

Reviewing the different possible suppositious as to the way in 
which the Arians may have entered India, Mr. Curzon infers (1.) 
that they could not have entered from the west, because it is 
dear that the people who lived in that direction were descended 

^0 Jounial Roy. AsiAt Soc. vol. xvi. pp. 172—200. 



from these rery Ariana of India, siich descent being proved by tli« 
fnct that the oldest forma of their language have been derived from 
the Sanskrit (to which they stand in a relation analogous to that in 
which the Pali and Prakrit stand); and by the circumstance that 
a portion of their mythology is borrowed from that of the Indo- 
Arians. Nor (2.) couM the Arians, in his opinion, have entered 
India from the north or north-west, because we have no proof 
from history or philology that there existed any civilised nation 
with a language and religion resembling theirs which could have 
issued from either of those quarters at that early period and have 
created the Indo-Arian civilisation. It was equally impossible 
(3.) that the Arians could have arrived in India from the east, as 
the only people who occupied the countries lying in that dinectioD 
(the Chinese') are quite different in respect of language, religion, 
and customs fruui the Indians, and have no genealogical rela- 
tions with them. In like manner (4.) the Indians could not 
have issued from the table-land of Thibet in the north-east, as 
indeppndently of the great physical barrier of the Himalaya, the 
same ethnical difficulty applies to this hypothesis as to that of 
their Chinese origin. And (5.) the Indians cannot be of Semitic f 
or Egyptian descent, because the Sanskrit contains no words of 
Semitic origin and differs totally iu structiue from the Semitic 
dialects, with which on the contnuy the language of Eg}T>t ap- 
pears, rather, to exhibit an affinity. And (6.) " no monumenis, no 
records, no tradition of the Arians having ever originally occu- 
pied, as Arians, any other seat than the plains to the south-west of 
the Himiilayan chain, bounded by the two seas defined by Mann 
(memorials such as exist in the histories of other nations who 
are known to have migrated from their primitive abodes), can be 
found in India." 

Mr. Curzon (7.) regards as illogical the inference, that because 
the Arians spread at an early period to the south of IndLi, as 
they did also to the west and north-west, they must have 
originally issued from some unknown region to invade and con- 
quer India itself. In the same way, he Urges, it might be 






argued that the Romans invaded Italy from Bome unascertained 
quarter (instead of springing from one region of Italy), because 
they extended their dominion to the south, as well as in other 
directioua. In explanation of their movements, he quotes the 
passage of Manu, ii. 17, ff., (which will be hereafter given at 
length,) and afsumes, in accordance with tlie indications which 
it affords, that the earliest seat of Indian civilisation was in 
Brabmavartta; and that the Arians, as they increased in numbers 
and advanced in social progress, gradually moved forward to the 
central region called Maflhyiidesa, and eventually to Aryavartta, 
the tract between the Himalaya and the Vindhya, extending 
from the eastern to the western sea, Mr. C\u2ou admits the 
existence of a non-Arian people and nationality, viz., the 
Tamulian in the south, which he conceives may have been form- 
ing contemporaneously with the rise of the Arian community in 
the north ; though he thinks that there is nothing to indicate that 
the Tamuliaus, or the hill tribes, or any other indigenous race, 
were ever in possession of Aryavartta (the country north of the 
Vindhya) before its occupation by the Ariims. 

His conclusion (founded ou the assumption that all the lan- 
gtiages of the Arian family have been framed from a Sanskrit 
basis, and are only modified and corrupted forms of what was 
once the original tongue of the Arian race of India) is therefore 
the following, viz., that either (1.) the nations whose speech is 
derived from Sanskrit have spnmg from the gradual dispersion 
of the ancient Ariun race of India, such dispersion being occa- 
sioned by political or religious causes, issuing in the expidsion 
from India of the defeated parties, and their settlement in 
different unoccupied countries chiefly to the westward : or (2.) 
that the Arians invadetl the countries to the west and north- 
west of India, and conquered the various tribes inferior to 
themselves, who were there in possession, im{x>sing upon them 
their own institutions and language. Of these two alternative 
suppositions he conceives the latter to have the greater proba- 
bility in its favour. 



I have stated the opinion of Mr. Curzon on tliia qnestion, to- 
gether with his argtimentB, in considerable detail, as it representi^ 
the view to which the Indian reader will, no doubt, incline as 
the most reasonable ; and it is therefore only fair that a] I that 
can be urged in ita behalf should be fully stated. 

Before discusfdng Mr. Curzon's hypothesis, I shall adduce the 
statement given by Mr. Elphinstone (History of India, voL L 
p. 95, ff. Ist edition) on the same subject. It will be seen tliat 
afler reviewing the arguments on both sides, this distioguLsbed 
author leaves it undecided whether the Hindus sprang from a 
country external to Hindiusthan, or were autochthonous. 

" On looking back to the information collected firom the Code 
[of Manu] we observe the three twice-born classes forming the 
whole commtmity embraced by the law, and the Sudras in a 
servile and degraded condition. Yet it appears that there are 
cities governed by Sudra kings, in which Brahmins are advised 
not to reside (chap. iv. 61), and that there are 'whole terri- 
tories inhabited by Sudras, overwhelmed with atheists, ond 
deprived of Brahmins." (Chap. viii. 22.) The three twice-lwrn 
dasBea are directed invariably to dwell in the country between 
the Himawat and the Vindya Mountains, from the eastern to the 
western ocean. But though the throe chief classes are confined 
to this tract, a Sudiu distressed for subsistence may dwell where 
be chooses. (Chap. ii. 21^ — 24.) It seems impossible not to 
conclude from all this that the twice-bom men were a conquer- 
ing people ; that the servile class were the subdued aborigines ; 
and that the independent Sudra towns were in such of the small 
territories, into which Hindostan waa divided, as still retained 
their independence, while the whole of the tract beyond the 
Vindya Mountains remained as yet untouched by the invaders, 
and unpenetrated by their religion. A doubt however soon 
suggests itself whether the conquerors were a foreign people, or 
a loeal tribe, like the Dorians in Greece; or whether, indeed, 
they were not merely a portion of one of the native ^rtates (a 
reiigioua sect, for instance) which had outs-tripped their fellow- 




citizeiiB in knowledge, and appropriated all the advantages of 
the society to themselves. 

" Tho different appearance of the higher claRses from the 
Sudras, which is so observahle to this day, might incline iw to 
think, them foreigners ; but without entirely denying this argu- 
ment (as for at least as relates to the Brahmina and Oshetriyas) 
we must advert to some considerations which greatly weaken its 

"The clafjs most unlike the Brahmins are the Chondolas, who 
are nevertheless orijjinally the offspring of a Brahmin mother, 
"laid who might have been expected to have preserveil their re- 
semblance to their parent stock, as, from the very lowness of 
their caste, they are prevented mixing with any race but their 
own.*' Difference of habits an<l employments is, of itself, 
sufficient to create as great a dissimilarity as exists between the 
Brahmin and the Sudra ; and the hereditary separation of profes- 
siona in India would contribute to keep up and to ificrease such 
a distinction. 

" It ia opposed to their foreign origin, that neither in the Cmle 
[of Manu], nor, I believe, in the Vethw, nor in any book that is 
certainly older than the Code, is there any allusion to a prior 
residence, or to a knowletlge of more than the name of any 
coiuitry out of India. Even mythology goes no further than 
the Himalaya chain, in which is fiied the habitation of the 

" The common origin of the Sanskrit language with those of 
the West leave-s no doubt that tliere was once a connection be- 
tween the nations by whom they are used ; but it proves nothing 
regarding the place where such a connection subsisted, nor 
about the time, which might have been in so early a stage of 

*' [Se« Part First of this work, p. 176, and Manu x. 12, tliere quoted. It 
>* clear, howevur, that we are not to take tlicsc accounts of llie foniiation of 
the different castes, wrxtten at a time when the Brabmanical system was fully 
developed, and iu the interest of it« defenders, as furnishing the true bisuiry 
of their origin. See Lassen, lud. Ant. i. 407. — J. M.] 




their society as to prevent ita tli rowing any light on the history 
of the individual nations. To say that it spread from a central 
point is a gratuitous assumption and even contrary to analogy; 
for emigration and civilization have not spread in a circle, but 
from east to west. \Vhere, also, could the central point be, 
from which a language could spread over India, Greece, and 
Italy, and yet leave Chaldea, Syria, and Arabia imtouched ? 

" The question, therefore, ia stiU open. There is no reasoa 
whatever for thinking that the Hindus ever inhabited any coun- 
try but their present one ; and as little for denying that thej 
may have done so before the earliest trace of their records or 

Mr. Elphinstone then proceeds to explain how he thiukn castes 
may have originated. 


Sect. VII. — Central At ia the cradle of the Arians : ojiinioua of Schlrgel, 
Lassen, Benfuy^ Miiller, Spiegel, Henatiy and Pictet. 

These views of Mr. Curzon, of which I have given a summary 
in the preceding section, are opposed to the general consent of 
European scholars. A. W. von Schlegel, Lassen, Benfey, Miiller, 
Weber, Roth, Spiegel, Ecuan, and Pictet, however diflFering on 
other points, all concur iu this, that the criidle of the Indians as 
well as of the other branches of the ludo-Gennanic race is to be 
flought for in some country external to India, 

I shall proceed to give some extracts from the writings 3 
these eminent authors : and shall finish with a summary of the 
arguments which seem to carry most weight in favour of the 
conclusion which they have adopted. 

The first authority whom I shall cite is A. W. von Schlegel, 
who in an essay On the origin of the Hindus *' systematically 
discusses the question under consideration, in all its bearingi<. 
He treats of the migratory movements of ancient nations, of 

" De rOHgiiie des tlindous, imbliihetl origiiitjlj in the 2n<l vol. of the 
Transactions of tl>c Koyol Society of Literature, London, I8.'J4 ; ami re- 
printed in his Esuaiii I.ittrrnires et Iliatorif/iies. Bonn, 1842. 



THE lyroisy axd cognate xatioxs. 


the traditions of the Hiudiis regarding their own origin, of the 
diversities of races, of tlie physiological character of the Hindus 
and of the indigenous Indian tribes, of the lieai-iug of compara- 
tive philologj' on the history of nations, on the relations of the 
Arian langiin^es to each other, and finally deduces the results to 
which he is led by the convergence of all these various lines of 
investigation. As I have already treated at length of some of 
these subjects, I shall only cite two passages, the first of which 
furnishes a reply to Sir. Ciirzon's argument against the immi- 
gration of the Hindus from any foreign region, drawn from the 
absence of any national tradition to that effect. It is ail 
follows : — 

" In enquiring into the Viirth-place of any people, and into 
the route by which, and the period at which they have travelled 
to their present abodes, we are naturally tempted first of all to 
interrogate the popular tradition on these points : but if we do 

it may easily happen that either no answer at all, or a false 
one, will be obtained. An illiterate people, ignorant of \vriting, 
which has adopted a stationary life, after a long and arduous 
migration, might, after a few centuries easily lose all recollection 
of its change of habitation : or if certain vestiges of such a 
change were preserved, it might be impossible for a people so 
circunifltanct'd to indicate with precision the point of departure; 
as for this purpose a general knowledge of the shape of continents 
and of seas would be necessary. It has often happened that 
tribes in a barbarous state have emigratetl, either impelled by 
necessity, or to avoid some powerful neighbour. Tlie utmost 
that such tribes could do might be to direct their journey with 
tolerable exactness according to the four cardinal points : shap- 
ing their course so as t« avoid any unexpected diflSculties which 
might arise, they would suffer themselves to be guided by 
chance ; and their only measure of distance would be the fatigue 
and the duration of their march." fEseais, p. 444.) 

The following is the passage in which Schlegel sums up tlie 
results of his researehe* ■ — 

" If we admit (and it is my conviction that the more 
deeply the suliject is investigated the more indubitable 
will the conclusion appear) that the derivation of the [Indo- 
European] languages from one common parent justifies the 
inference that the nations who spoke them also iiisued from 
one common stock; that their ancestors, at a certain epoch, 
belonged to one sole nation, which became divided and sub- 
divided as its expansion proceeded ; — the question naturally 
arises, what was the primeval seat of that parent nation ? It 
is nowise probable that the migrations which have pet)pled so 
large a part of the globe should have commenced at its 
southern extremity, and have been constantly directed from 
that point towards the north-west. On the contrary, every 
thing concurs to persuade us that the colonies set out from a 
central region in divergent diiections. According to this sup- 
position the distances which the colonists would have to tra- 
verse up to the time of their definitive estt»blishmeut, become 
less immense ; the vicissitudes of climate to which tliey wera 
exposed, become less abrupt, and many of the emigrant tribea 
would thus make an advantageous exchange as regards fertility 
of soil, and the temperature of the air. And where is this 
central country to be sought for, if not in the interior of the 
great continent, in the neighbourhood, and to the east, of the 
Caspian Sea? It may perhaps be objected that the country 
in quastion is now occupied by people of a different race : but 
to how many countries has it not happened to undergo a total 
change of their inhabitants? Tlie prolific parent-country of 80 
many swarms of expatriated coloniata might, from that very cir- 
cumstance, be converted into a desert It is probable that, 

since the commencement of history, the nature of this country 
lias changed, and that in former times it was more favourable than 
now to agricidture and to population. According to my hypo- 
thesis, then, the ancestors of the Persians and Hindus must 
have emigrated from their early seats toward the south-west 
and the south-cast ; and the forefathers of the European nations 
\e west and the north I conceive that the tribea 

SECT, vn.] 



which migrated towards Europe followed two great routes ; the 
one along the northern shores of the Black Sea ; while the 
other traversed Asia Minor, and crossed the JEgesm Sea, or the 
Hellespont, Thrace, Illyria, and the Adriatic. It was indubit- 
ably by this latter route that Greece and Italy received their 
colonists." (Essais, p. 514-517.) 

Professor Laasen also decides against the hypothesis that 

dia was the birth-place of the Indo-Emopcan races. He 


" It is, aa we have seen, a result of modern investigation that 
the ancient language of the Indians is so intimately related to 
those of the other Indo-Germanic nations as to establish the 
original unity both of these languages and nations. We are 
therefore driven to the conclusion either; Ist, that the Indians 
migrated to India from some other primeval seat, or, 2nd, that 
all the kindred Indo-Germanic nations had their origin in India. 
The following cou.siderations determine us to decide in favour of 
the former of these alternatives. 

" It would, First, be an improbable supposition that the na- 
tions which are now so widely extended abould have been derived 
from the remotest member of the entire series. Their common 
cradle must be sought, if not in the very centre, at aU events 
in sixch a situation tis to render a diffusion towards the <lifFerent 
regions of the world practicable. This condition is not well ful- 
filled by supposing India to be the point of departure. Secondly, 
none of the phenomena of speech, customs, or ideas observable 
amnng the other cognate nations indicate an Indian origin. 
Of the coiujtries which were anciently occupied by the great 
Indo-Germanic family, India was the most peculiar, and differed 
the most widely from the others ; and it would be very unac- 
countable that no trace of these Indian peculiarities should 
have been preserved by any of the other Indo-Germanic races 
in later times, if they had all originally dwelt in India, .\mong 

** Indiiin Antiquities, i. p. 312, AT. 


the names of plants and animals which arc common to all these 
nationa there is none which is peculiai' to India.** The moat 
widely diffused word for any species of corn {yava) denotes not 
rice, but barley. Thirdly, For a deciaion of this question, the 
manner in which India ia geographically distributed among the 
different nations by which it is occupied, is of great importance. 
Tlie diffusion of the ArianH towards the south, points to the con- 
clusion that they came from the north-west, from the country to 
the north of the Vindhya, probably from the rejfion border- 
ing on the Jumna, and the eastern part of the Punjab. Their 
extension to the east bf^tween the Himalaya and the Vindhyn, 
also indicates the same countiies as their earlier seats. We find 
moreover evident traces of the Arians, in their advance firom the 
north-west, having severed asunder the earlier population of 
Hindustan, and driven one portion of it towards the northern, 
and another portion towards the southern hills. Further, we 
cannot assume that the Arians themselves were the earlier in- 
habitants who were pushed aside : for the inhabitants of the 
Dekhan, like those of the Vindhya rnngo, appear alwaj's as the 
weaker and retiring party who were driven back by the Ariao& 
We cannot ascribe to the non-Arian tribes the power of having 
forced themselves forwaid through the midst of an earlier Ariau 
population to the seats which they eventually occupied in the 
centre of the country : but, on the contrary, everything gpecJn 
in favour of their having been originally settled in those tracts 
where we find them at a later period, and of their having once 
occupied a more extensive territory. These non-Arians were in 
fact feebler nvcen, like the red men of America. The Arians on 
the other hand were a more perfectly organised, enterprising, 
and creative people, and were consequently the more recent; 
just as the earth has at a later period produced the more perfect 

•* [This circumstance, however, migbt be accounted for, as Weber remark* 
(Modern Invest igntions on Ancient ludia, p. 10), by the names being for- 
gotten, from the planta and anim&l» being uuknown in western couatriet. 
See below, p. 318.— J. M.] 



classes of plants and animals. Finally the same thing is shown 
by the political relation of the two branches of the population. 
The iVriaus take up for themselves, i.e. for the three highest 
castes, a position of the most complete contrast to the aboriginal 
tribes, first of all by the name of Arya, and next by their 
prerogatives ; for the name of dvija, " twice-born," with the 
higher rank connected with it, is the exclusive designation 
of the three upper classes. The Ariana in this way mark them- 
selves out as the superior and conquering race. In conErmatiou 
of this we can also adduce an outward mark, that of complexion. 
The word for caste in Sanskrit {carna), originally signified 
colvur. The castes therefore were distinguished by their com- 
plexion. But, as is well known, the Brahmans have a fairer 
colour than the Sudraa and Chaudalas: and the Kshatriyas 
and Vaiiyas who were also Arians must have paj'ticipated in 
the same fair complexion. We are thus led to the conclusion 
that the Arian-Indians were originally diKtinguished from the 
dark aborigines : and this accords with the assumption that 
they aime from a more northern country." 

That the Arians were not autochthonous in India, but came 
from some country to the north, is also the opinion of Professor 
Max Miiller. 

" At the first dawn of traditional history we see these Arian 
tribes migrating across the snow of the Himalaya southward 
toward the " Seven Rivers " (the Indus, the five rivers of the 
Panj.^b, and the Sarasvati), and ever since India has been called 
their home. That before this time they had been living iu 
more northern regions, within the same precincts with the 
ancestors of the Greeks, the Italians, Slavonians, Germans and 
Celts, is a fact as firmly established as that the Normans of 
William the Conqueror were the Northmen of Scandinavia. 
The evidence of language is irrefragable, and it is the only 
evidence worth listening to with regard to ante-historical periods 
, . , . "\Miile most of the members of the Arian family followed 
this glorious path *' [/. e, to tlie north-west], " the southern 

X 3 

txibes were slowly migrating to the mountains wliich gird 
the north of India, After crossing the narrow passes of the 
Uindu-kush or the Himalaya, they conquered or drove before 
them, as it seems without much effort, the original inliabitiwts 
of the Trans-Himalayan countries. They took for their guldoi 
the principal rivers of northern India, and were led by them 
to new homes in their beautiful and fertile valleys." (Last 
Results of Sanskrit Researches, in Bunsen's Out. of Phil, of 
Un. HiBt, vol. i. pp. 129, and 131 j andAnc. Sansk. Lit. pp. 12, 
13, 1,5.) 

Again in the Last Results of the Turanian Researches, ibid., 
p. 340, the same able writer remarks: "It is now gener»lly 
imitted that this holy-land of the Bralimans, eyen within 
its eailiest and narrowest limits, between the Sarasvati and 
Drishadvati, was not the birth-place of the sons of Manu. The 
Arians were strangers in the land of the Indus and Uie Ganges, 
but no one can now determine the exact spot whence they 
came, and where they had been previously settled. Traditions, 
current among the Brahmans as to the northern regioni', con- 
sidered the seats of the blessed, may be construed into some- 
thing like a recollection of their northern immigration — holy 
places along the rivers of northern India, where even in later 
times Brahmans went to learn the purest Sanskrit, may mark 
the stations of their onward course — the principal capitals of 
thtse ancient kingdoms may prove the slow but steady progress 
towai-d the mouths of the principal rivers of India — but with 
the sources of those rivers, the homes of the Arian strangers 
vanish from our sight, even after we have reached the highest 
points of view accessible on Indian ground." 

Professor Benfey exjjrcsses an equally confident opinion that 
India was not the original country of the Hindus. His reasons 
are as follows. After giving some account of the various tribe« 
by whom southern and central India are occupied, he proceeds : 
"We thus find the whole of the Dekhan covered with the 
remains of a nation of which it is highly probable that the 


»KCT. VII.] 



several parts were connected by affinity. But we know with 
certainty that the Sanskrit-speaking people did not establish 
themselves in the Dekhan till a later period, and as coloniste, 
who apparently began their occupation by making themselves 
masters of the coasts. . . . Now it is hardly probable that 
thoHe barbarous tribes could have pushed themselves forward 
into the midst of the Arian-Indians at a period when the latter 
had attained to the height of their social and political develop- 
ment ; and yet it is at this very period that we already find 
mention made of several of these barbarous races. We are 
therefore compelled to recognize the latter as being the earlier 
inhabitants of the Dekhan, who were reduced to subjection by 
the arms of the Sanskrit-speaking race, and either incorporated 
into their community as a servile caste, or driven back into the 
lesses of the mountains." Indien, *** p. 9. In p, 12, the same 
.thor proceeds : " From the foregoing sections it appears that 
the Sanskrit-speaking people, who caUed themselves Aryas and 
Vises, can be shown to have immigrated from foreign regions 
into their new abodes. It can be jwsitively demonstrated that 
they once formed one nation, spoke one speech, and possessed 
the same civilisation, with the races who are allied to them 
by language, viz., the Aryas properly so called (ie. the 
Iranians), the Greeks, Latins, &C. It is scarcely to be doubted 
that the theatre of this early union was one of the countries of 
Asia: but the time is ko far antecedent to the dawn of history, 
and so many commotions, migrations and so forth, must have 
swept over the region which they formerly occxipied, that every 
trace which tlie Sanskrit-speaking race might have left of their 
residence there, has been obliterated. Argument* of a general 
nature render it not improbable that Tartar}', which even in 
historical ages has sent forth so many children of the steppes 
to the most various regions of the world, may, at one time, have 
embraced this race also in its vast wildernesses WLen 

>* * In £r<cli and Gruber's Encyclopaedia (German). 

X 4 



CHJLP. n. 

the once united nation became broken up, the Indo-Ariaiu 
remained still for a long time united with the Perso-Arian*, 
as the language of both distinctly proves. The abodes which 
they occupied at that time may be determined with tolerable 
probability by the aid of a widely diJQfused sacred legend." 
Professor Benfey then refers to the Indian legends re^rding 
the sacred lakes Manosarovara, and Ravana-hrada, and the 
descent of the Granges in the same Himalayan region, and its 
subsequent division into separate streams. He regards the 
Ganges as a personification of the original fountain of waters, 
and identifies it with the apdm nnpat, or "son of waters^*' 
of the Veda. This again he considers to be identical with 
the nnpdt apa-ivm of the Zend ^vritings, which in like manner 
he interprets as a mythical representation of the primevnl 
source of earthly waters, issuing from a mountain. He then 
proceeds : " Here then we have in the Zend writings a repe- 
tition of the Indian legends regarding the original Bouroe of 
waters, along with the Indian locality: and the only qnestion 
which arises is whether the knowledge of this locality was 
acquired by the Persians in later times, or brought with them 
from their earlier abodes. When we reflect how great is the 
general agreement between the Indians and Persians in respect 
of religion, of pf>litical institutions, and other things of the 
same kind, it seems to me extremely probable that, as they 
boar in common the name of Arya (in Zend Airya), or • the 
honourable,' — a name which implies a high sense of their own 
importance, — they must also have attained in common a con- 
siderable degree of civilisation. This of course, pre-RuppoA68 
that they dwelt together in the same country for a t^olerably 
long period ; and where can we more reasonably seek for this 
locality than in the quarter to which the recollections of both 
nations point as to their earliest atmdes ? For what but recol- 
lections of this sort could have given to this region the appella- 
tion of a holy land ? For it lies geographically too far distant 
from the principal seats of later Hinduism as well as of the 



Zend religion, to admitr of its coming, in any other way, to be 
regarded by both nations as a sacred region, at a period wbeD 
they had become settleii so far from it, as well as from each 
other. We can Bcarcely go wrong if we identify this region, 
(a» the quarter in which the Indians placed the abode of the 
blessed), with the renowned Uttarakiiru, which according also 
to the cl.'ussical authorities of tlie west is to l>e fixed here. We 
shall not, however, spend more time in confirming (hy grounds 
which we cannot render perfectly convincing) our opinion that 
the Sanskrit and Zend peoples once dwelt together in the 
country which we c^iil Little Thibet, as a nation which had 
attained a certain st.ige of civilisation. At the period when the 
Aiyas settled in this region, a division of the great Indo-Gothic 
race appears to have already taken place. After this separa- 
tion, the other members of the same great family who subse- 
quently came into notice in the weatem world, remained for a 
time by themselves in their earlier abodes somewhat more to 
the north." 

The following remarks of Professor Spiegel (Introduction to 
Avesta, vol. ii. pp. cvi. flF.) will serve as an answer to Mr. 
Curzou's allegation that the language and mythology of tlie Per- 
RJans ai-e derived fi-om thoKC of India: " Though it is universally 
admitted that a primeval country is to be assumed, where the 
Arians lived in pre-historical times as one people, and from 
which they gradually migrated ; and although it is allowed 
that the Indians and Iranians must have dwelt together for a 
length of time in this, or in some other atljacent country, even 
after the separation of the other branches, still it is by no 
means clear what should be regarded as that primeval country. 
Agreeably to Mr. Curzon'a ajssumption India was the fatherlan<l 
of the Indo-Germanic races. From that country the individual 
branches of that stock migrated westwards, and last of all the 
Iranians, who continued to dwell in the immediate vicinity of 
tlieir original country, which henceforward remained in solo 
possession of a single race, the Indians. According to thia 

assumption the relation of Iran to India admits of a very aionple 
adjustment; India is the cradle, the Indian language (/.«. the 
Vedic Sanskrit) is the mother-tongiie of all the Indo-Ger- 
manic nations. If accordingly, an important affinity id discern- 
ible both in language and in ideas between the Indians and 
Iranians, the reason of it is simply this that the Iramans 
emigrated last from India, and thus carried with them the 
largest share of Indian characteristics. On this view the older 
monuments of Iranian literature would stand in the same 
relation to the Vedic literature, that the Pali and Prakrit stand 
to the later Sanskrit, Lassen "' hati however previously declare*! 
himself against this ass\imptioa that India was the cradle <A 
the Indo-G ermanic races; and his arguments have not been 
invalidated by Mr, Curzon. And as regards the relation of the 
old Iranian dialect to the Sanskrit of the Vedas, I boldly assert 
that we cannot possibly suppose the former to stand in any 
such relation of dependance to the latter as the Pali or the 
Prakrit stands in to the later Sanskrit; and no one who impar- 
tially examines the question, will do otherwise than support 
my view. 

" We may therefore at once set aside the supposition that 
India was the cradle of the Indo-Germanic race. We prefer to 
assume with Lassen that their original abode is to be sovight in 
the extreme east of the Iranian country, in the tract where 
the Oxus and Jaxartes take their rise. 

** But the second question in regard to langxiage is not thus 
determined. For it might still he imagined possible that not 
only the Indiaun but also the Iranians along with them, had 
migrated to the countries on the Indus; and that the Iranians, 
perhaps ovNing to religious differences, had retraced their steps 
to the westward. The great aflBuity between the Sanskrit, and 
the ancient Bactrian, languages, and the resemblances between 
the mythologies, of the Vedas on the one hand, and the Avesta 

»» Ind. AiJi. i. p. 512. See above p. 307, S. 


on the other, would then admit of the same explanation ; viz., 
that the Iraaians had speut the Vedic period, or at least a great 
part of it, in conjunction with the Indians ; and hence the 
close affinity between their ideas. This is in fact the view of 
a scholar who is very familiar with this branch of study. Pro- 
fessor Max jMiiller.*" 

" I cannot agree with this view, as I am quite unable to dis- 
cover that there is any historical reminiscence by which it can 
be established.**' The facta which I have above collected regard- 
ing Zoroaster and his religion, certainly do not point to the 
conclusion that he was a Bactrian, much less that (he rehgion 
of the Bactrians came from India ; on the contrary, these 
accounts seem to lead us to believe that their religion came first 
from Media. .... But if there be no historical recollection, 
what eke is there to favour the opinion in question? Surely it 
oannot be the similarity of structure between the languages of 
India and Persia ! We esteem the Sanskrit so highly, not 
because it was the original speech of the Indo-Germanic race, 
bat because it stands the nearest to that original language. 

*• " Lnst Results of the Persian Reworchw," p. 113. " If regarded from 
aVaidik point of view . . . the gods of tbc Zoroastrians come out once more 
as mere reflections of the printitive and authentic gods of the Vedas. It can 
now be proved, even hy geographical evidence, that the Zoroastrians hud 
been 8eltlc<l in India before tbcj immigrated into Persia. I say the 
Zoroastrians, for we have no evidence to bear us out in making the same 
assertion of the nations of Media and Persia in general. That tbe Zoroas* 
trianti and their ancestors started from India during the Yaidiii period can 
be proved as distinctly as that the inhabitants of Mus^ilia started from 
Greece, The geographical traditions in the first Fargard of the Vendidnd 
do not interfere with this opinion. If ancient and genuine, they would 
embody a remembrance preserved by the Zoroastrians, but forgotten by the 
Vaidik poets — a remembrance of times previous to their first common 
descent into tlie country of the Seven Rivers. If of later origin, and this 
is more likely, they may represent a geographical conception of the Zoroas- 
trians after they had become acquainted with b larger sphere of countries 
and nations, subsequent to their craig;ration from India." 

*** See, however, App. note E. 



[CBAT. n. 


Now it cauDot surprise us that another language of the same 
family, as the aucieiit Bactrian is, should have remained on a 
nearly similar level. It is not in the least at variance with ih'w 
view that the last named language is far yoimger tliaa the 
Vedic Sanskrit, for it is well known that external circumstances 
frequently occasion the speedy corruption of one langtiage, while 
another can long preserve its ancient level. And so in thia 
caae, hoth languages issued in a nearly similar form from one 
common parent form of speech, and were then developed inde- 
pendently of eacli other. And as the phenomena of the two 
languages do not necessitate the assumption that the ancient 
Bactrian language has passed thrmigh the Vedic Sanakrit, so 
neither is this view forced upon us by the contents of the Avesta. 
Beference has, indeed, been made to the points of contact 
between the legends, and even between the manners and cus- 
toms exhibited in the Veda and the Avesta. But the few par- 
ticulars which recur in the Vedas cannot be set against the 
far larger number of which there is no trace there. Similar 
common legends have been discovered in the Greek mythology, 
and yet it has never been imagined by any one that the ancient 
Greeks must have believed in the Vedas. We are, therefore, 
warranted in supposing that in the old Bactrian language and 
literature we possess the monuments of a people, who certainly 
lived together with the Indians longer than any of the other 
kindre<l races, and have therefore a certain number of religious 
and other conceptions common t^ it with the former. But 
these common elements are so insignificant when compared I 
with those which are of peculiarly Iranian growth, that we are 
justified in regarding the language and literature as independent 
Iranian productions. How, and by what causes the separation 
of tlie Iranians from the Indians was occasioned, is a point 
which, owing to our want of information on that early period, 

can no longer be certainly determined Among the 

grounds of it I have (in the 1st Vol p. 9.) referred to a religious 
alienation between the two nations, but too great importance 



slioiild not be assigned to this view. Even without assuming 
any such alienation, it is conceivable that goda who were 
honoured by the one people, might be <legraded to hell by the 
other."** .... That which gi\ es probiibility to the idea of au 
actual alienation between the Indians and Iranians on religious 
grounds, is the number of such opposing conceptions. 

" We must accordingly maintain that the Indians and Iranians 
have each gone through their own proper development apart 
from the others. Any points of coincidence between the two 
must thus be referred to the early pre-Vedic period, not to 
the era of the special development of either of the two peoples. 
None of the common features which I have referred to in Vol. I. 
(see above, p. 290, ff.) are of such a character as to make it at 
all necessary for us to suppose the country bordering on the 
Indus to have been the scene of their origination. An origin 
in that locality might, with most probability, be ascribed to the 
legend of Vrittrahan, as Indra is designated, as the slayer of 
Vrittra, who withholds the clouds and the necessary rain. The 
■word recurs again in the old-Bactrian vevetraya ' victorious, 
(the deity, Verethragna I regard as being certainly of far later 
origin). From the circumstance that no special sense is assigned 
to the word in the ancient Bactriau language, I do not conclude, 
as is coramouly done, that in the Avesta it has lost its special 
meaning ; but, on the contrary, I assume that the Indian limita- 
tion of the word to Indra did not take place till after the sepa- 
ration of the two peoples, and that the word had originally a 
more general meaning." 

The following is the opinion of Professor Weber on the same 
genera] question- In his tract, entitled " Modern Investigations 
on Ancient India," p. 10, after sketching the physical and intel- 
lectual condition of the early Arj'aa, as deducible from the words 
common to all the Indo-European languages, he proceeds 
thus: — 

*^' See App, note F. 



CoiAr. u. 

** In tbe picture just now drawn, positive signs are after all 
almost entirely wanting, by which we could recogni«e tbe 
country in which our forefathers dwelt and had their common 
home. That it was situated in Asia is an old historical axiom : 
the want of all animals specifically Asiatic in our enumeration 
above seems to tell against this, hut can he explained simply by 
the fact of these animals not existing in Europe, which oo- 
casioned their names to he forgotten, or at least caused tliem to 
be applied to other similar animals ; it seems, however, on the 
whole, that the climate of that country was rather temperate than 
tropical, most probably mild, and not so much unlike that of 
Europe ; from which we are led to seek for it in the highlands of 
central Asia, which latter has been regarded from time imme- 
morial as the cradle of the human race." 

My next quotation is from the recent work of M. Pictet, " Les 
Origines Indo-Europcennes," in which he endeavours, by an ex- 
amination of all the accessible data, geographical, philological, 
and ethnograjihical, as well as by a survey and comparison of all 
the terms common to the Arian languages, which refer to climate, 
to topography, and to natural history, to determine what that 
country was, which the common ancestors of tbe Indo-European 
nations originally inhabite<l." 

I shall not attempt to follow the course of M. Pictefa multi- 
farious investigations and reasonings, or to pass any judgment 
on liis particular deductions ; but shall content myself with ex- 
tracting his account of the general results to which he has been 

" By consulting successively national appellations, traditions, 
geography, philology, and ethnography, we have arrived at the 
fiillnwing conclusions: — The Arian people, as they called them- 
selves in opposition to the barbarian, must have occupied a 

"M, PJctet's second Tolumc, whiclj has not yet nppearei], is to treat of the 
state of civilisation and intellcctiinl culture which this primilire people had 
Attained btiforc it was bruken tip into dltTerent nations. 


region, of which Bactria may be regarded as the centre. This is 
the conchision to which we are at once led by merely comparing 
the directions followed by the swarms of men who issued from 
this centre, and which all radiate from it as a point of departure. 
The geographical configuration of this portion of Asia completely 
confirms this first induction ; for the only jxissible oiitleta throiigh 
which the population could issue occur at the very points where 
the principal curreutss of emigration have actually flowed, if we 
may judge by the ultimate positions of the Arian people, and 
the scattered traditions which they have preserved of their 
origin." . . " . 

" We may presume (1,), from the order and direction of tlie 
migrations which determined the ultimate positions of the Aiian 
races, (2.), from tlie traces of their ancient names, left by the 
several nations along the routes which they must have followed, 
and (3.) from the more special affinities which connect toge- 
ther the different groups of Arian languages, that the primitive 
Ariana, at the period of ita gi'catest extension, must have em- 
braced nearly the whole of the region situated between theHindu- 
kush, Belurtagh, the Oxus, and the Caspian sea; and, perhaps, 
extended a good way into Sogdiana, towards the sources of the 
OxuB and the Jaxartea. I do not meau that Ariana then formed 
one strongly constituted state. It is much more probable that it 
was at that time partitioned among distinct tribes, unite<l solely 
by the general bond of race, by similarity of manners and 
language, by a common stock of beliefs and traditions, and by 
a sentiment of natural brotherhood. This is to be inferred 
both from the topographical character of the country, and from 
the successive emigrations which must have followed each other 
at considerable intervals. I have attempted in Chapter HI. to fix, 
by approximation, the relative positions of the diflerent branches 
of the i-ace before their dispersion." 

[I introduce here, from p. 51 of M. Pictet's work, the subatanoe 
of the passage referred to, so far as it relates to the Iranians, 
Indians, Greeks, and Latiujs : — 




" Assuming Bactria to have been the centre of the region 
peopled by the primitive Aryas, the Iranians must h*«re 
possessed its north-east corner, bordering on Sogdiana, towards 
Behirtagh, and have at first spread towards the east, as fajr as the 
high mountain vallies, from which they afterwards descended Uj 
colonize Iran. Alongside of them, to the south-eairt, prolwibly in 
the fertile regions of Badakhshan, dwelt the Indo-Ariaus, occu- 
pying the slopes of the Hutdu-kush, which tliey had afterwards 
to cross, or to roimd, in order to arrive in Cabul, and penetrate 
thence into northern India, To the aonth-west, towards the 
sources of the Artamis and the Bactrus, we should place the 
Pelasgo-Arians (the Greeks and Latins), who must have advanced 
thence in the direction of Herat, and continued their migration, 
by Khorasan and Mazenderan to Asia Minor and the Hellespont.*] 

" Though nothing more than a hypothesis, the preceding distri- 
bution appears to account better than any other for the entire fjacta 
of the case. But it can be shown, in a more precise manner, 
that the Aryas must have be«";n originally divided into two 
groups, the one eastern and the other western, from which, on 
the one side, the Aryas of Persia and India issued, and ou the 
other the European nations. The principal arguments in sup- 
port of this statement cannot, however, be unfolded till I come 

to the sequel of my work In regard to the 

period when the Arian emigrations took place, I may say, by an- 
ticipation, that, in all probability, the earliest of them cannot be 
placed at less than three thoTisand years before the Chrisrtian era, 
and that perhaps tlioy go back to a still remoter period-" — 
(Pictet, Les Aryas Primitifs, pp. 536, ff.) 

I slmll now attempt briefly to sum up the arguments in 
favour of the conchi.sinn, that the Indn-Arians were not autoch- 
thonous, but immigrated into Hiudusthau from Central Asia. 

Mr. Curzon entertains, as we have seen, a different opinion, 
whicli he grounds on the assumption that the languages, aa well 
as the mythologies, of the Persians, as well as of the Greok«! and 
Latins, are derived from India. We have already seen (p. 269, ff.) 



how imt«tiable the aotioa is that the Greek and Latin languages 
colli i have been derived from Sanskrit; and the jjoiuts of coin- 
cidence between the Greek, the ItaJian, and the Indian mytho- 
logies are ^x» few and too remote to justify the idea of their 
derivation from the Indo-Ariaus, at any period nearly so recent 
SiS the Ijypttthefiis would retjuire. I am not prepared to pro- 
nounce it iiltogether inconceivable that the Greek and Latin 
races could have emigrated from India within any periotl short 
of 1500 years B.C., without distinct traces of this migration 
being discoverable in their own literature, or in that of other 
nations : for, as we have already seen { page 305) the traces of 
such movements may soon disappear from the traditions of an 
illiterate people. But if the languages and religions of Greece 
and Italy be not deriveil from those of India, there is no ground 
for this hypothesis. And any emigration from India at an 
earlier period than that indicated appears to be improbable. 
Fifteen hundred years before the Christian era, India, if (as 
it no doubt was) alrea<ly occupied by the Indo-Arians, must, 
at oil events, have been very thinly peopled. The Aryas had 
not, at that period, extended themselves beyond the north- 
west quarter of India.^ Large tracta both to the east and 
west of the Ganges, and to the north of the Vindhya range, 
must then have been still uncultivated. While such facilitiea 
remainotl for the occvipation of new territory for the purposes of 
agriculture or pasturage, in their own immediate neighbourhood, 
it does not appear what sufficient motive could have existed to 
impel any branch of the small Arian population to desert the 
fertile plains and the warm temperature of India (to which, by 
the hypothesis, they must have been long accustomed) for the 
rugged mountains and the barren and less genial regions which 
lay to the north-west and west of the Indus. 

As regards the deiivation of the Iranian language and my- 
thology from the Indian (which may he asserted with more show 

** Tbis will be made evident by the deUib which I »baU shortly add uco, 
relative to their difiuiioa in liiudusthui. 




ctuv. n> 

of probability than iu the case of the Greeks and Latins), I may 
place the authority and the arguments, just quoted, of Professor 
Spiegel in opposition to those of Mr. Curzon. 

Mr. Elphiustone, as we have seen, does not decide in favonr of 
either theory, but leaves it in doubt whether the Hindus were 
an autochthonous or an immigrant nation. As a justificatioa 
of his doubt, he refers to the circumstance that all other kuown 
migrations of ancient date have proceeded from east to west 
and have not radiated from a common centre. But this reason- 
ing cannot claim to offer more than a limited presumption, and 
cannot be set against the stronger probabilities which, in this 
case, are suggested by the subsequent history of the different 
Arian nations iu favour of a radiation from one common centre. 

The common origin of the Arian tongues implies, as we 
have seen, the anterior existence of one parent language from 
which they all issued, and conducts us necessarily to the condu- H 
siou that the several nations, who spoke those separate dialects 
were all descended from the. same common ancestors, who em- 
ployed the parent language in question, and formed one Arian 
nation inhabiting the same country. As the question where this H 
country was situated cannot be decided by history, we are 
thrown back upon inference ; and we are therefore led to enquire 
what that region was which by its position was most likely to 
Lave formed the point of departure from which nations situated 
in the opposite quarters ultimately occupied by the Indians, the 
Iranians, the Greeks, the Romans, the Germans and the Slavon- 
ians must have issued iu order to reach their several abodes by 
the most easy and natural routes. The point of departure which 
b&st satisfies this condition, is in the opinion of the eminent M 
writers whom I have cited, some region of central Asia, lying to 
the north-west of India. We may therefore place the cradle of 
the Arians in or near Bactria. 



Sect. ViLI. — On the National Traditiont of the Indiant regarding their own 

Originid Country. 

I shall now inqtiire whether there are auy data to be found 
among the traditions of the Indians or the Peraians, fi-om which 
we can derive any confirmation of the conclusion to which we have 
l)een led by other considerations. I must, however, begin with 
a candid admission that so far aa I know, none of the Sanskrit 
l»ooka, not even the most ancient, contain any distinct reference 
or allusion to the foreign origin of tho Indians. This does not, 
perhaps, afford any just ground of surprise. The Vedic hymns 
themselves do not carry us back to the first ages of the nation, 
but contain allusions to personages and events of a still earlier 
date. The past history of their race is very liable to be forgotten 
by an unlettered people, as is justly remarked by Schlegel, inthe 
'pase&ge quoted above, p. 305 ; imd any traditions which may at 
one time have existed of the early Arian migrations might very 
easily liave been overgrown and effaced by the luxuriant harvest 
of l^endary inventions for which India has been remarkable from 
tlie earliest ages. This process of obscuration is distinctly trace- 
able in other parts of Indian history, aud it has beeu frequently 
remarked how greatly the myths and even the allusions of the 
Vedas have been amplified and distorted by more recent mytho- 
logist^ I shall, however, proceed to quote such passages as may 
apjK?ar in any way to imply the tradition of a foreign origin. 

First- In the Rig-veda,*' an expression occurs from which we 
might infer that the Indians still retained some recollection of 
their having at one time occupied a colder coimtry. It is this : 
Ii.-V. i. 64. U. ^V?F Ti^l? TR^ im f%^:i "May we 
cherish [such] a sou and grandson a hundreti winU'rs " I The 
expression q^T| IT^^: 1[^ ^Ff^^ "STTT' "STTT^l "May 
we see — may we live — a hundred autunms" also occurs in the 

** WilaoD, lalrod. to Rig-vedo, vol. 

T 2 

II » 



[caxr. ti. 


R-V. vii. 66. 16. See also E.-V. x. 18. 4. This may perhapB 
be a more recent form of the phrase, dating from a period when 
the recollection of tlie colder regions from which they had 
migrateti, was Ijecoming forgotten by the Aryas. 

Second. In tbe Satapatba-brahmana, one of a class uf 
works which, as we bave shown, form the oldest remains of 
Hindu literature next to the Vedic hymns, we find the follow- 
ing Itgend, which may possibly be based upon an old tradi- 
tion of the Hindus having originally crossed, at some point, the 
great mountains to the north or north-west, and entered into 
India by that route : — *^* 

1IT MKf<j*y*A(Hi '^'Ei x^: ^: i^rr U^<i \ <17T- 

^^Tt ^NJjM**Ulm*ll^5>l W ^ft^ ^0<^ *1N*<IM^ I ■ 
ff ^ ^TT^ ^M«*J(«j^'^l fT^ 35^ ^TT^: tjni -Rf^WTRII 

"•This legend is Iranslutcl lij Weber, In J, Stud. i. ICl, ff. ; nnd 
lUUer, Anc. Saiu. Lit. p. 4i5, d'. 



TRfj: ^s^^i inTt^f%ffT Tfti ^i«<<ri4^f?n ^ 
iRirrsnrmi fit ^ »?^^t^ ^rr^tfni rR ^fw rf^i 

** They brought to Manu in the morning water for washing, as 
they are in the habit of bringing water to waeh with the hands. 
As he was \ising the water, there came into hia hands a fish, which 
said to him, 'Preserve me and I will save thee.' [Manu inquired] 
* From what wilt thou save me ? * [The fish replied] ' A flood 
shall sweep away all these creatnres: I will rescue thee from 
it.' [Manu asked] 'How is thy protection [to be effected]?' 

1 5 




cv^r. n. 

The fiflh answered * So long aa we are small, we axe in great peril, 
and even fish devours fish : preserve me first in a jar. When I 
grow too large for the jar, dig a trench, and preserve me in it. 
TNTien I become too great for that, carry me to the ocean ; I shall 
then he beyond the reach of danger.* Straightway it became 
a great fish : for it grew exceedingly. [The fish then said,] * In 
so many years the flood will come : make a ship therefore, and 
worship me; and when the flood rises, embark on the ship, 
and I shall deliver thee.' Accordingly Maim preserved the fish, 
an<i brought it to the ocean j and in the same year w^hich the 
fish had declared, he built a ship and worshipped [the fish.] 
Wlien the fluod ascended, he entered the ship, and the fish 
swam near him : and he fastened the cable of the ship to the 
fish's horn. By this means he pafised over this northern 
raoimtain. The fish then said, ' I have delivered thee, fasten 
the ship to a tree.' But lest the water should abandon 
thee when thou art upon the mountain, as fast as the water 
subsides, so fast shalt thou descend along with it. Accord- 
ingly he descended as the water subsided. Hence, this was 
* Manu's descent ' from the northern moimtain. The flood 
had swept away all creatures: Manu alone was left. Being 
desirous of offspring he laboriously performed a religions 
rite. And there, too, he sacrificed witli the pdka sacrifice. He 
cast clarified butter, thicken e<l milk, whey, and curds, as an 
oblation into the waters. After a year a female was produced, 
who rose unctuous from the waters, with clarified butter under 
her feet. Mitra and Varuua met her ; and said to her * \Mio 
art thou ? ' ' Manu's daughter,' she replied. They rejoined, 
' Say that thou art our daughter.' She answered, * No : I am 
the daughter of him who Ijegot me.' Then they demanded a 
share in her. She promised, and slie did not promise ; but 
passed on and came to Manu. Manu asked her 'Who art 
thou?' ' Thy daughter,' she replied. 'How, thou divine .one, 
art thou my daughter?' he inquired. She replied, 'Thou 
begotten me from those oblations which thou didst cast 




into the •waters. I am a benediction. Introduce me at the 
sacrifice. If thou shalt do so, thou shalt increase in offspring 
and cattle. WLatever boon thou ahalt supplicate through me, 
shall accrue to thee.' He accordingly introduced her in the 
middle of the sacrifice : for that is the middle which stands 
between the introtiuctory and concluding prayers. He lived 
with her worshipping and toiling, desirous of oflTspring. By her 
he begot this offspring, which is the offspring of Manu." 

It is true that in this legend Manu ia said to have crossed 
the northern range, and to have descended into India, and 
there sacrificed for the purpose of obtaining offspring; and that 
a daughter was born to him in that country. But if the pro- 
genitor of the Hindus was deemed to have been once an 
inhabitant of some country to the north, we may understand 
the legend as intimating that their ancestors generally lived at 
one time out of India. 

Manu, as we have already seen, (Part I. pp. 15, 16, 25, 41, 
&c), is declared in some passages of the LiBtitutes of Hindu 
polity which bear liia name, and in the IVIahabharata, to be the 
progenitor of tlie immau race. One of the Bnihmanas says in like 
manner, ^T^ f% TJSIT TfW f^ WT^IW^l *" " Creatures 
are descended from Manu." In the Ilig-veda also he appears to 
be regarded as the forefather or representative of mankind, or 
of the Arian race at least. (Weber, lud. Stud. i. 165, 194, 
195; Burnouf, Introd. to Bhag. Purana, lix. ff; Wilson, Rig- 
veda, vol. iL p. 292, note; N^ve, Essai sur le mythe des 
Kibhavas, p. 68, ff.) I shall quote some of the passages of 
the R.-V., in which Manu is mentione<l. Thus, i. 31. 4, 
«^^ TR^ <yi*i4lXl<^: M^<^^ ^?i^ ^4«Tl<: t " Thou, 
Agni, hast made known the sky to Manu ; thou bast been a 
greater benefactor to Puriiravas the performer of pious works ; " 
i. 36. 19, f% rrr»m T^^ ^^VIt! ^^nV ipg^l " Manu 
has established thee, Agni, as a light to all mankind," (see 

•* 4>r»;rtn« on R.-V. i. 08. 4- 
T 4 




also, verse 10); i. 68. 4, 'ftrTT f^TrfV iT^HTW TaiTf^l 
" [Thou, Agni,] hast taken up thy abode as a priest among the 
race of Manu;" i. 96. 2, ^ qf?IT UUi^l *<*lr\ l ii \ U.M \: 
■Jf^ ^!r5TTi(«*l«JH 1*1^1 " In consequence of Ayu's primeval 
bymn, he [Agni] proJuoerl tlieae children of the Mauus," which 
Sayana explains, 71«f«(| '^7^: .^5T^ ^l^<(1: V^l Tf^ 
^^'iUti I "Being hymneJ by Manu, he produced all the 
progeny of Manu." ^(h^sT Si^.if*!'*! «^TT^rr*(^| i. 112. 
18, "Whereby ye preserved the hero Manu with food.**' 
ir^ ^9 ^rg 4{3<lii^ fW ^rU[fi. t i. 114, 20, " The 
prosperity ami security which oManu our father prociu-ed." 
Wf% ^TT^trftrn f^ T: «,rtjlf<l ii. 33. 13, « Thoso 
[medicines] which our father Manu chose," &c. ^ u^j 

M^l'fl W^: ^rjfHTJ^ 'rer ^TTJ ^l fW ^"^ 
fV^ '^ST^ *" f^: I viii, 52. I, « He whose gates the an- 
cient priest (or beloved) of the great ones, adorned [or made 
known ?] by sacrifices : our father Manu revealed among the gods 
(priests?) the prayers of the sky?" ^al>«^ H^Uti ^IMR^^I 
\W[ rfTT^ ^T? ^nr^l X. 46. 9, "\Miom [Agni] Mata- 
riswan and the goJs formed first as an object of praise and 
worship for Slanu." ^ tJ^^ jjw\^ Wl^TT'^ ^^ ^^M} 
SW^^ ^rrTT^I X. 73. 7, "Thou hast formed for Manu 
beautiful paths leading speedil)' to the gods." ^' 

As Manu thus appears to be the progenitor and representa- 
tive of the Indo-Arians, we might (as I have said) undeistaod 

" Siiyana explains food in tliis passage by tlie words ^J^y^jf TH*T 
T^Tf^'VT»*4%M*l.H VJ^I I i-e- food coiuitling of wheat and other grain 
Khich the earth had concealed. In his annotation on verse 16, Sayana calls 
Manu, a, or the, royal rkhi of that natne. 

" SeeR.-V. i. 102. 1. 

" See, however, Benfey'a rendering in bis glossary to the SiLma-Teda, 
undor the word natnucki', see, also, the passages, R.-V. i. 130. 6, it. 26. 1, 
if. 20. 7, vii. Id. 3. 




the interesting legend which I have quoted regarding his pre- 
servation from the deluge, his passage across the northern 
mountain, and his descent from its summits down into the 
region at its southern base, to intimate generally that the an- 
cestors of the Hindus originally came from some region to 
the north of the Himalaya. I would not, however, be under- 
stood as laying very much stress upon this inference. 

The story of the deluge is narrated in the Vana Parva of the 
Mababharata, verses 12,746 — 12,804, at much greater length, 
and with considerable variations from the preceding legend 
of the Jsatapatha-brahmana. Every tradition of the deluge 
has an intrinsic interej^t of its own, quite independently of its 
bearing on the point which I am ftow concerned to illustrate. 
Though, therefore, the legend as recordetl in tlie Mahabharata 
contains nothing directly connected with the subject of thus sec- 
tion, I will quote it both on account of its own importance, and 
aa an instance of the modifications which ancient Indian legends 

nerally undergo in tlie hands of the later mythologists.*'* 
Jiiahabharata begins with an account of the strenuous 
ities of Manu. While he was thus engaged, a fish 
came to him on the banks of the Chirini," soliciting de- 
liverance. He cjust the fish into a jar. When it became too 
large for the jar, he threw it into a pond. When it grew too 
vast for the pond, it thus addressed Mann (verse 12,764): 

^^Errf^ "^VJ 3rr rTTrr ^RT^II "Bring me, O divine and holy 
aiige, to Ganga, the ocean's beloved queen ; in her I shall dwell : 
or do whatever else thou jjleasest," Manu accordingly cast 
the fish into the Ganga.** Finally, when the fish became too 

as an v 


••* Tlie text of this epi»o<le was long since publiabeJ by Bopp. 
•♦Verse 12,751; ' ^R^RfOi^^H i m 'T^ T'J'T^r5(ft«\ll 
« Verse 12.766 : 1?^^^ ^T^T^I^R^ »T»T^T'|^ ?lftl 

'Rff wf^ rnr %^ ^ T(Tf^M<^7T: n 



IcHAF. U. 

great for the Ganga, Manu brought it to the occaa. \Miea 
thrown in there, the fish annnmicetl to Manu the approaching 
universal dehige; and then proceeded (venjcs 12,776 — 8): 

■fffft^^ '^Tmf^ll " And tliou shalt cause a strong ship to be 
built, with a cable attached, in which thou must embark with 
the seven rishis. And take with thee all manner of seeds, as 
anciently described by the Bralimans, severally well preserved : 
and then await my arrival." Manu did as be was comniand<?<l, 
verses 12,782, ff.: 41<»1MKI<* ^^ifw W\7\i: V[^ ff^l 

<ri*<ll 15»nn- ^ TT^^^W^ ^r<'<*lll "Manu then 
taking all the seeds floated on the billowy sea in the beau- 
tiful ship." He then thought on the fish, which speedily 
arrived ; and the cable of the ship was bound to its vaat 
horn. Then we have the following description which, in some 
places, is not devoid of poetical power, verses 12,786, flf. : 

vTrrlHi ^4»«ni wiiiwt ^-^■'m^m ^1 Ti^ ^^\ 




^ Ar^m ^^: g^T 3?^ f%TRifRr^ I TTW 'Tt^'vi 

*' The PMsh being attached by the c&ble, drew the ship with 
great rapidity over the briny deep ; and transported ita crew 
across the ocean, which seemed to dance with its waves, and 
thunder with its waters. The ship, tossed by the mighty winds, 
-whirled around like an unsteady intoxicated woman. Neither 
Mith nor the eight quarters of the world appeared : everything 
■was water, and firmament and sky. Amid this perturbation of 
the universe, the seven rishis, Manu, and the Fish were perceived. 
In this manner, the Fish, unwearied, drew along the ship for 
many periods of years amid the mass of waters ; and at length 
brought it to the highest peak of the Himavat. Then Rpake 
the Fiah, gently smiling, to the rishis : * Bind the ship without 
delay to this peak of the Himavat' They accordingly fastened 
the ship there according to the command of the Fish. That 
loftiest peak of the Himalaya is even to this day known by the 
appellation of * Naubaudhaua ' [the binding of the ship]." The 
I<^h then reveals himself to the rishis as Brahma, the supreme 
Lord of Creatures ; and commands Manu to create all living 
beings, gods, Asuras and men, all worlds, and all tilings move- 
able and immoveable:*® — a command which Manu fulfilled. 

It will be observed that this legend differs in several of its 
details from thut of the J5atapatha-lirahmana. In the latter, 
the original abode of ISIauu is undefined ; but as he is said to 
have crossed the northern raoimtain (which the commeutator 
erplains of the Himalaya) and as we must suppose it to have 
Ijeen the southern slope which formed the scene of " Manu's 
descent," it may be presumed that the Brahmana intends to 

•• Vcr«. 12,798 : 'M^^J ^ TTSfT'. ^: Wi 

*i\i^\ : *i4mi*i^ t|^ ^i^ HffTni 


APFiyrrres asd DERrrATros" of 

LCBAr. n. 

represent him as having come from the northern side. In 
the Mahabharata, on the other hand, the scene is laid on the 
banks of the Chirini. Though the position of that stream is 
undetermined, there is no doubt that it must have been in 
northern India, as the Ganges is shortly after named as one of the 
receptacles into which the Fish was thrown- Mauu therefore is 
overtaken by the flood in northern India ; and after being carried 
about in the ship for many years, is landed on the highest 
peak of the Himalaya. As no mention is made of his having 
crossed to the northern, we must suppose that he continued on 
the southern, side. If, therefore, the legend, as narrated in 
the Briibmana, contains any reminiscence of the immigration 
of the Aryas from the north into India, it is clear that this 
feature of it has been lost in the epic poem. 

In the Pm-anas the story undergoes still further transform- 
ations. The Bhagavata places the scene in the south of 
India, in Dravida, instead of in the north ; declares it to have 
been Vishnu who became incarnate in the fish, (instead of 
Brahma, who is represented in the Mahabharata as the de- 
liverer) ; and omits all mention of the ship's descent on tho 
peak of the Himiilaya. (See M. Burnours preface to the third 
vol. of the Rbfigavata-purana, pp. xxiii., ff.) 

lliird. In the allusions made to the Uttara (or northern) 
Kiirtis in the Indian books, there may be some reminiscence of 
an early connection with the countries to the north of the 
Himalaya. The ful lowing passage from the Aitareya-brahmana 
viii. 14. (quoted by Weber, ludische Studien, i. 218), con- 
tains the oldest reference to this people of which I am aware, 

nt<lfdLrtJH|*t^'^jfi^l(jM»tM^^|| « Wherefore in this northern 
region, all the people who dwell beyond the Himavat^the Uttara 
Kums and the Uttara Madras, are consecrated to separate rule 
{vairdjya.) Those who are consecrated are called vivdL" 



r. Tin.] THE INDIi3T 



The following quotation from another part of the Aitareya- 
trahmana, viii. 23, will, however, show that even at the early 
period when that work was composed, the countiy of the Uttara 
Kunis had come to be regarded as belonging to the domain of 
mythology: T^ ^ ^ ^ ^TTTf^^^ =llf*l8: ^T3IT^ 

^4fT: if^ irf\ m 'm^i w Trrr^ninjf^ 
^M^f^ 4^ srr^iw *^ti<^*^^*(<<h. '^i^ '^^ ^ 

•rffMH. ^ttt|#qi f^:i5^rJT^ ^rf^T-^dM^j: ijfww: ^^ 

^ISfl ^rmT II *^ " Satyahavya of the race of V&sishtha de- 
clared this great inauguration, similar to Indra's, to Atyarati 
son of Janantapa ; and in consequence Atyarati, who was not a 
king, by [that] knowledge traversed the whole earth round, 
reducing it to subjection. Satyahavya said to him, * Thou hast 
subdued the whole earth round : exalt me now to greatness.' 
Atyarati replied, 'TIMieD, Brabraan, I conquer the Uttara 
Kurus, then thou shalt be king of the earth, and I will be only 
thy general.* Satyahavya rejoined, ' That is the holy laud of the 
gods; no mortal may conquer it: thou hast acted injuriouHly 
towards me ; resign, therefore, that [which I have bestowed]." 
In consequence of this the foe-destroying Sushmina, the son 
of i§ivi, slew Atyaniti, sou of Janantapa, who hail [thus] become 
bereft of his vigour, and destitute of strength." (See Colebrooke's 
Misc. Ess. 1. 43). 

^* I am indebted to ProfeMor M . Milller for copjing for mc ihiia pas^gc 
from K MS. of the Altareja-brihrnana. 




In the First Part of this work I have already cited the first 
of the preceding passages and eeTeral others, relating to the 
northern Kurus. 

I shall make some further extracts from the passage in the 
Rivmayana there noticed, and adduce Bome other texta on the 
same subject. In the " description of the northern r^on," 
Ramayana iv. 44. 82, ff. we have the following account : 

Wf^ ♦1^1*11*11^^ P(f4|riaR^ *IH*«(<I»1^I T «T^ if^- 

'^ Tift »<i^<:ii 

" Go, most excellent of monkeys, to those illustrioua Uttara 
Kurus, who ai'e liberal, prosperous, perpetually happy, and 
unilecaying. In their coimtry there is neither cold nor 
heat, nor decrepitude, nor disease, nor grief, nor fear, nor lain, 
nor sun." A great deal more follows in the same hyperlxilical 
strain and then it is added (verse 117): ^<^4^|W 414lfri4i^ 

^TtT^ tr^ f^:\ rT^ ^flf^rfK ^ f^<uj4<i^>4ry 

4{MI«t,l antl in verses 121, 122: ^ ^^^^ 4 1 ??(«*) ^^WT- 

f% ^»lfjlR 4r ^«II«H*|!^ ^^' I " Beyond the 
Kurua to the north lies the ocean ; and there the vast Soma 
mountain is situated, resembling a mass of gold." "You must 
not travel to the north of the Kurus. That region is tmtrodden 
by the steps of otlier living beings also. For that Soma- 
mountain is difficult of access even to the gods themselves." 

In the same way when Arjuna, in the course of his conquests, 
aa described in the Ligvijaya Parva of the Mahabharata, comes 
to the country of the Uttara Kurus in Harivarsha,. he is thus 
addressed by the guards at the gate of the city : (Sabha Parva 

verses 1045, ff.): qr^ %^ JH^n l^ VT^ ^?^ ^TST^I 







^ ^J: ^: -Rf^^ "3^ W 5T W^ 'It: 


'i^ ^v^ ^ TTT^^ii irRrr^f^ f^ ^^^ ^^ 

" This city, king, cannot be eubdued by thee. . . . He who 
enters this city must be more than mortal. . . . There is nothing 
to be beheld here, Arjuna, which thou mayest conquer. 
Here are the Uttara Kiu-us, against whom no one attempts to 
combat. And even if thou shouldst enter, thou couldHt behold 
nothing ; for here no one with a mortal body can see." 

On this passage (part of which is a mere repetition of the Ait. 
Brah. viii. 23.), Professor Lassen remarks (in the Zeitschrift 
liir die Kunde des Morgenlandes, ii. 62): "At the furthest 
acces-sible extremity of the earth, appears Harivarsha, with the 
northern Kurus. Tlie region of Hari or Vishnu belongs to 
the system of mythical geography; but the case is different 
Tvith the Uttara Kurus. Hero there is a real basis of geogra- 
phical fact; of which fable has only taken advantage, without 
creating it. The Uttara Kurus were formerly quite independent 
of the mythical system of Ji'f^Jws, though they were included 
in it at an early date." Again the same writer says at p. 65 ; 
*' That the conception of the Uttara Kurus is based upon an 
actual country and not on mere invention, is proved (1.) by the 
way in which they are mentioned in the Vedas" [the Aitareya- 
brahmana, aa just quoted p. 332] ; " (2.) by the existence of 
Uttara Kuru in historical times as a real country ; and (3.) by 
the way in which the legend makes mention of that region as 
the home of primitive customs. To begin with the lart point, 
the Mahabharata speaks as follows of the freer mode of life 
which women led in the early world, Book 1, verses 47 19 — 22: 

^•ii^(ii: ^rw ^ f%^ '^IT¥\ <<<MHi ^nr^rd^- 




CHAP. n> 

^^^ li * Women were formerly miconiined, and roved about at f 
their pleasure, independent Though in their youthful innocence, 
they abandoned their huabanda, they were guilty of no offence; 
for such was tha ride in. early times. Tins ancient custom is 
even now the law for creatures born as brutes> which are free 
from lust and anger. This custom is supported by authority 
and is observed by great rishis, and it is atill practiecd among 
the northeni Kurua.^ *** 

" The idea which is here conveyed is that of the continuance 
in one part of t!ie world of that original blessedness which pre- 
vailed in the golden age. To afford a conception of the happy 
condition of the southern Kunis it is said in another place," 

(M.-Bh. i, 4346 : ^JtIK: ^f^: ^^ ^f^WTI ^T^^f^I 

f^*J^^*1HI ^Tft^mr t'^fHrr^: l) " ' The southern Kimia 
vied in happiness with the northern Kurus, and with the divine 
rishis and bards.' " 

Professor Lassen goes on to say : "Ptolemy (vi. IS.),"* is also 



*' [I am myself responsible for the translation of these lines. The pntC' 
lice of promiscuous intercourse was, according to the legend, aViolisked by 
dvetaketu, son of the risbi TJdJalaka, who wua incensed at seeing his mother ■ 
led Bway by a stranfje Brahnitin. His fatlier told him there was no reason to 

beangry, as; ■^HT^rlT f% ^W ^^^TTT^T^^TflT ^1 ^TOT 

TR: R^ldlHIM ^ ^ ^ Wm 3^1:11 "The women of all 
cutes on earth are unconfincd: just oa cattle are situated, so are human H 

beJn^, too, within their respective castes." 6»etttkctu, however, could not 
endure this custom, and established the rule that henceforward wives sliould 
remain faithful to their husbands, and husbands to their wives. Maha- 
bharats, i. verses 4724-.13 — J, M.j 
*** The origiual passage will be given in the App. note G. 



acqtiaijited with Utlara Kurti. He speaka of a mouataiu, u 
people, and a city called Ottorokonxi. Most of the other 
ancient authors who elsewhere mention this name, have it from 
hira. It Is a part of the country wliioh he calls Serica ; accord- 
ingly to him the city lies twelve degrees west from the metro- 
polis of Sera and the inuijiitain extends from thence far to the 
eikstwai'd. As Ptolemy has misplaced the whole of eaatern Asia 
eyond the Ganges, the reltitive position which he assigns will 
guide U8 better than the cthaolute one, which removes Ottoiv- 
kori'a BO far to the east that a correction is inevitable. 

" According to my opinion the Ottorokorra of Ptolemy must 
be sought for to the east of Kashgbar." 

Lassen also thinks that Mcgasthenes had the Uttara Kurus in 
view when he referred to the Hyperboreans, who were faltled by 
Indian writers to live a thousiiud years.*' In his Indian Anti- 
quitieS) (Ind. Alterthumskunde, L 51 1, 312, and note,) the same 
writer concludes that though the passages above cited relative 
to the Uttfira Kiiras indicate a belief in tlie existence of a really 
existing country of that name in the far north, yet that the 
descriptions there given are to be taken as pictures of au ideal 
paradise, and not as founded on any recollections of the northern 
origin of the Kunis. It is proliable, he thinks, that some such 
reminiscences originally existed, and still survived in the Vedic 
era, though there is no trace of their existence in later times. 

The sanctity of Kashmir is thus celebrated in the "\''iina Parva 
of the Mahabhiirata, verses 10,545—46: ^ | ajn<44Hj^ -§rTfT^ 

^rn?nT^ "^ ^IT^TII "And this is the region of Kfismira, 
-holy, and inhabited by great riahis : behold it, along with 

*" iSeltsohrift, as above, ii. 67. and Schwanbeck, Megulhenu lodica, pp. 
70, 117. Wtfti a Tiv xiXitTwl' 'Yllfpfl<>fiuv ui irvra Xipiv TCifiufiSy tai XUtrinit^ 


thy brothers. It w 
northern risbia witl 
Kityapa, occurred." 

Fourth. In the Atharva-vetla, v. 4. 8. the salutary plant 
*'kusbtha" ia spoken of as growiug on the other side of the 
Himalaya: ^3^^ "SfTrfT i^^T^: MI^I *f1<*»l 'SRlf^^l 
" Produced to the north of the Himavat, thou ait carried to 
the people in the east." This reference may perhaps be 
lield to imply that the coutemporariea of the author of this 
mantra had some acquaintance with the country on the other 
side of the great chain. 

Fifth. In a passage of the SankhayanA or Kaushitaki-brah- 
mana vii. 6 (cited by Weber, Ind, Stud. i. 153. note, and 
alluded to by Miilier, " Last Results of the Turanian Researchc*,* 
p. 340) it is reported that the north was resorted to at an early 
period for the purpose of studying language, as it was best 
known in that region : r^^m ^^\<{\^ f^ «|i>IMI^ 

f?['^ M^MTII "Pathya Svasti (a goddess) knew the northern 
region. Now Pathya Svasti is Vach [the goddess of speech.] 
Hence in the northern region speech is better known and better 
spoken : and it is to the north that men go to leam speech : it 
is said that men listen ta the instructions of any one who conies 
from that quarter : for that is renowned as the region of speech." 
On this the commentator Yinnyaka Bhatta remarks (Weber, as 
above): H'illHHil TRT^I ^rn?Pr^ ^T^Tft ^ft^l 

WT% TTf«l^*H^ W ^4«!ri*: II "'Language is better under- 
stood ftiid spoken : ' for Saiiusvati is spoken of [as having her 






abode] in Kashmir, and in the hermitage of Badarika [Batl- 
arinath in the Himalaya, apparently], the sound of the 
Vedaa is heard. ' Men go to the north to learn language ' : 
to obtain the favour of Sanlavati ; and ' he who comes thence,' 
having obtained her favour, * is listened to with attention,' as 
every one knows, and repeats." 

There may lie in this pa.s.sage some faint reminiscence of an 
early connection with the north. 

S«CT. IX. — Ancient Penian Tradition of the Earlieit Abodes of the Arian 


I shall now proceed to quote at some length the First Fargard 
of the Vendidad, descriptive of the creation of various coniitriea 
by Ahura-mazda, w^liich has been held, not without probability, 
to contain a reference to the earliest regions known to the 
Iranians. Being imacquainted with Zeml, I shall borrow the 
abstract which I give of this section from the versions of Pro- 
fessor Spiegel,'" and Pr. Hang." 

" Ahura-mazda spake to the holy Zarathustra : • I formed 
into an agreeable region that which before was nowhere 
habitable. Had I not done this, all living tilings would have 
departed to Alri/ana-vaejuJ''" 

* I, Ahura-mazda, created as the first, best region, Airyana- 
t'a«;o, in a state of excellence. Then Angra-mainyus, the 
destroyer, formed in opposition to it, a great serpent, and 
winter [or snow] the creation of the daivae. There are there 
ten months of winter, and two of summer.' 

''*' Aveita : Die Ileiligeii Scbriflen der Farscn (Avcata: The Sncred 
Writings of the Parsis), vol. i. pp. 61, S. 

Das Erste Kapitel des Vendiddd (The First Chapter of iLe ycndidiid)^ 
18, ff. 
^ The purport of this is, Dr. Haug remarks, that Airyana-Tsejo was 
originallj the only cultivated coutilrjr, and that all other countries Werd 
waste. As it was to be feared that the iuhabilanis of the waste would over- 
run Airjana-vaejo, other countries also wen made habitable by Ahura- 

I 3 




• I, Ahura-inazda, created as the second, best region, Gau, 
in which Sughdha is situated.' 

[Here, and in most of the following cases, I omit the counter- 
creations of Angra-mainyus.] 

• I, &c., created as the third, best region, Mom-u, the mighty, 
the holy.' 

• I, &.C., created as the fourth, best region, the fortunate 
Bakhdhi, with the lofty banner.' 

« I, &c., created as the fifths best region, Nisai.' ■ 

' I, &c., created as the sixth, best region, Haroyu, abounding 
in houses [or water].' 

' I, &c.y created as the seventh, best region, Vaekereta, where 
Dujak is sitimted. In opposition to it, Angra-niainyus, the de- 
stroyer, created the Pairika Khnathaiti, who clung to Kere^pa.' 

• I, &c., created as the eighth, best region, Urva, full of 

• I, &c., created as the ninth, best region, Khnenta, in which 
Vehrkana lies.' 

• I, &c., created as the tenth, best region, the fortunate Ilara- 

' I, &c., created as the eleventh, best region, Haeturaat, the 
rich and shining.' 

' I, «S:c., created as the twelfth, best region, Ragha, with tlixi 
fortresses [or races].' 

• I, &C., created as the thirteenth, best region, Chakhra, the 

' I &c., created as the fourteenth, best region, V'arena, with 
four corners; to which was born Thraetaono, who slew the 
serpent Dahaka.' ■ 

• I, &c., created as the fifteenth, best country, Hapta-hendii 
[from the eastern to the western Hendu "']. In opposition, Angra- 
maiuyns created untimely evils, and pernicious heat [or fever].' 

• I, lie, created as the sixteenth, and best, the people whi 
live without a ruler on the sea-shore.* 



" Siiiegcl omits the words wilbin brai-kets. 




• There are besides other countries, fortunato, renowned, lofty, 
prosperous and splendid.' " 

I shall now adduce the most important comments of different 
authors on this ciuious passage. 

Haug observes (p. 9), that " the winter of ten months dura- 
tion asaigned to Airyana-vaejo, points to a position far to the 
north, at a great distance beytjnd the Jaxartes ; but the situation 
cannot, in the absence of any precise accounts, be more spe- 
cifically fixed. Only so much is undeniable that the Iranians 
came from the distant north. The same thing results from the 
icond fiu-gard of the Vendidad, where the years of Yima are 

umorated by winters, and the evils of winter are depicted in 
lively colouns." The same writer further remarks (j>p. 23, 24,); 
" By Airyana-vaejo we are to understand the original country 
of the Arian.s, and paradise of the Iranians. Its ruler was 
King Yima, the ^euo^vned Jemshed of Iranian legends, who is 
hence called 4ndo Alrv/cne-^'aejahi, • famous in Airyana-vaejo ' 
(fargard ii.). In this region Ahura-mazda and Zanithustra 
adore the water of the celestial spring {Ardvi dum andhltd, 
Yasht, 5. 17. 104); and here, too, Zarathustra supplicates 
Drva^pa, and Ashi. Thus, Airj'ana-vaejo had become an 
entirely mj'thical region, the abode of gods and heroes, free 
from sickness, death, frost and heat, as is said of Yima's realm. 
We can, however, discover a historical substratum in the 
chapter before us. In Airyana-vaejo the winter laatfl for ten 
months; but winter being a calamity inflicted by Angra- 
mainyus, wa« not compatible with the idea of a paradise, the 
abode of joy and blessedness. This long duration of winter is 
however perfectly characteristic of regions lying far to the 
north, and is a primitive reminiscence of the real cradle of the 
Iranians. In the legend of Airyana-vaejo an actual historical 
recollection of this earliest home has thus become blended 
with the conception of a primeval abode of mankind in para- 
dise, such as is represented in so many popular traditions." 

" Airyana-vaejo," says Spiegel, " is to be placed in the fiirthest 

s 3 


AFFiNirrEg and derivation op 

{cBxr. a* 1 


east of the Iranian plateau, in the region where the Oxns and 
Jaxartes take their rise." 

The second country is Sogdiana; the third, Merv (the ancient 
Margiana) ; the fourth, Balkb (the ancient Bactria) ; tlie fifth, 
Nisa (the ancient NLssea) ; the sixth, Herat (the ancient Aria) ; 
the Beventh is Kabul, according to Spiegel, and Sejesbui ac- 
cording to Bumoiif, Lassen, and Haug; the eighth is Kabul 
according to Haug ; the ninth is Gurgan according to Spiegel, 
and Kandahar according to Haug; the tenth is the Arachosia 
of the ancients ; the eleventh is the valley of the Hilmend 
river ; the twelfth is Rei in Media ; the thirteenth may be M 
Chilirem in Iran (Spiegel) or a city in Khorasan (Haug); the " 
fourteenth is variously placed ; the fifteenth is the country of 
the seven rivers (saptasmdhavas), or the Panjab; and the 
sixteenth maj', Haug thinks, be sought on the shores of the 
Caspian Sea- 

In regard to the age of the section under review. Dr. Hau|^ 
remarks (p. 6) : " The original document itself [as distinguished 
from certain adiiitions which appear to have been interpolated M 
in it] is certainly of higli antiquity, and is undoubtedly one of ■ 
the oldest of the pieces which compose the existing Vendidad. 
But in the form in which it lies before us (even after striking 
out the late interpolfitionFi) it is decidedly subsequent to Zara- 
thustra; and later than the so-called Gdthds, in which, for the 
most part., the genuine sayings and doctrinee of Zarathustra 
have been handed do\vn. The chief reason for this conclusion 
is, that the passage under review exhibits the Persian doctrine 
in a far more developed shape than the songs of Zarathustra." 
And again in p. 7, '• Though, there is thus no doubt that this 
fargard only dates from the period after Zarathustra, we do 
not thereby mean to say that it is of modem origin ; on the 
contrary, its whole contents show that it must be very ancient. 
"We can scarcely derive from it any fixed historical data. From 
the names of the countries, however, we can gather not only 
that the geographical knowledge of its author was very limited, 






but olao that the region actually occupied by the Ariaus was 
rinuch more contracted than we afterwards find it." 

Professor Spiegel remarks on the same fargard as follows 
(L 59): " The great importance of this first chapter for the pre- 
bistorical age of the Indo-Germanic race in general, and of the 
Persian nation iu particular, has been fully allowed by inves- 
tigators of the mythology and history of the ancient world. 
Ilecren, Rhode, Lassen, and others, have recognised in these 
accounts of the Vendidad a half- historical, half-mythical frag- 
ment, which reveals to um the state of geographical knowledge 
among the followers of the Avesta at the time when it was com- 
posed. Perhaps, we may also, with Rhode, discover in it the 
history of the gradual diffusion of the Iranian race, regarding 
the first mentioned country as their primeval abode, and those 
which follow as the regions which were peopled at a later date. 
The order in which the countries are arranged appears to tell in 
favour of this hypothesis." 

In his second vol. p. cix., however, Professor Spiegel retracts 
his qualified adhesion to the view of Rhode. He says: "I 
cannot coincide in the attempt to discover in the first chapter 
of the Vendidad an account of the gradual migration of the 
Iranians. It has been said that tliat list of countries is a 
continuous history of their attempts at colonisation, beginning 
with their northern home, and ending with Hapta-Hendu 
or India. But the list nowhere speaka of any such migra- 
tion Hence, I see in this chapter nothing but a specifi- 
cation of the countries known to the Iranians at a particular 
time. This period however can not be a very recent one, as 
the name Hapta-Hendu is connected with the Vedic period. 
This name however may have been prescrve<l in Persia after it 
had distvppeared in India, and we cannot conclude from it that 
this fargard was composed contemporaneously with the Yedas." 

M. Pictet, on the other hand, makes the following observa- 
tions : " Tliese names [of countj-ies] enable us to follow step by 
step the extension of the Iranians over the vast domain which 

S 4 


thoy have ever since occupied. The thing which interesU us 
the most in this enuineratiou is the point of departure, and the 
general direction of the inoveraent. The first perfect alwdo 
which Ormuzd create*! is called Aii-yana-vaejo. ... As Rittw 
and Lassen remark, the ten montliB of winter and only two (if 
summer can only apply to the highest valliea of Bchirtagh and 
Mnatagh at the north-east comer of (he Iranian table-land. 
But it is difficult to conceive that an ' excellent ' abode could 
ever have existed there, unless we assume a verj' improbable 
alteration of climate. We are as little ahle to imagine how 
a country so savage and so poor could have been the cradle of a 
race 80 prolific as the Aryas. I believe, then, that we must 
separate, in this tradition, the mythical element from the hls- 
triricjil data. Airjjatm-vuejn, the primeval paradise, was pro- 
liably nothing more than a very cotifu8e<l reminiscence of the 
country originally inhabited by the Arj'as, At their dispersion, 
the Perso-Ariau branch, driven back perhaps by the gradual 
increase of the Ariaii riopulation, may have directed their steps 
tnwards the east as fur as the high vallies of Belurtagh and 
Mustagh, where their further progress would be arrested- At a 
later period, when the emigration of the other Arian tribes bad 
left the fif Id clear, they descended from these improfitable regions 
towardb tlie more favoureil countries of which they had pre- 
served some recollection, as we learn from the myth in the 
Veudidad." — Origiues Indo-Europcennes, pp. 36, 37. 

Professor Miiller's views on the first fargard of the Vendidad 
will be found ahove, in note 56, p. 315.'^' 


Sect. X. — What mu the Jiovte hy which the Aryaa penetrattd inta 

India f 

\Vc have already seen (pp. 304-322) that according to the most 
numerous authorities, Bactrifi, or its neighbourhood, was the 
country which the different branches of the Indo-European nice 

'»• See App. nott H. 

r. I.] 



occupied in common before their separation. Professor Benfey, 
wlu), apparently, differs to some extent from other scholars iii 
designating that primeval country as Tartary, is of opinion that 
the Indian and Persian branches of this family may, after their 
separation from the others, have dwelt together, more to the 
soutli, in Little Thiliet, tlie country near the sources of the Indus. 
In regard to the route by which the Indo-Arians immigi-ated into 
Ilindusthan, he makes the following remarks (ludien, pp, 14, ff.): 
" If, then, as we assume, the Arians (/. e. the Indians and 
Persians) originally dwelt together in the region of Little Tliibet, 
the question arises how they came into the separate seats which 
we find them occupying in historical times. It is by no means 
impossible that the entire mass of Arians may have first of all 
taken a western route towards Bactria, &c., and have then spread 
themselves through the passes of the Hindu-kush into Oaitul 
and Affghanistan, that great region which formed as it were tlie 
bridge between the eastern Arians or Sanskrit-speakers and tlie 
wefitem, or proper Arians, and in which, in historical time^*, both 
branches encountered each other. The Sanskrit-speaking Arians 
would then liave penetrated into their new abodes by the same 
route across the Indus by which foreign nations have generally 
entered. Their immigration by this side has been assumed by 
Wilson, A. W. von Schlegel, I^assen and others. I do not ven- 
ture directly to deny that this was the route they took, 
especially as I admit, in limive, that historical certainty is not 
to be expected for periods of so great antiquity : but many con- 
siderations appear to me to tell in some measure against it. The 
Indians regard Brahmavartta, the tract of land between the 
Sarasvati and Drishadvati, as the peculiar centre point of 
Indian civilisation. Bordering on this tract are Kurukshetra, 
(in the region of the modern Delhi) the Alatsyas, (on the 
Jumna) the Panchalas in the vicinity of the modern Canoj, and 
tbe SiJrasenas (in the district of Slathura) called by Arrian 
(after Megastheues) 1,ovp(un]vai. This is called the land of the 
Brahmarshis. Modhyada^ or the central laud, is next named. 



[cnAr. a. 

which embraces the two preceding tracts, is bounded to the sotith 
by the ^^indhya, on the north by the Himalaya, and stretches 
from Vinasana in the east to Prayaga in the west. Finally 
Aryavartta embraces all the foregoing tracta and reaches from 
sea to sea. We see here the narrow limits to which the coimtry 
regarded as the most sacred, is confined ; and how this division 
appears, as it were, to indicate the grailual extension of the 
Sanskrit-si>eak.iug people from that point, in continually widen- 
ing circles. This Rmall tract is bordered by the Sarasvati and it 
is remarkable that at the point where the countrj' of the western 
Arians (the Persians) begins, the same name meets us in its 
Zend form HaraqaitL In the holy land (to the north) the 
Sarovara (the best of waters) played the principal part : are we 
then to suppose that the two Arian nations, independent of each 
other, yet united by the spiritual bond of water-worship, in re- 
collection of the sacred Saras (' lake') recognised again in their 
new abodes the Sara»vati ? Although this point alone ap}>ears 
to me to indicate that the Sanskrit-speaking people fonnd their 
first Indian abodes on the banks of the Sarasvati, the fact of the 
Ganges being their holiest river tells still more in favour of the 
Fame conclusion. If they came to India with the religious neces- 
sity of paying reverence to a great river, why (if they arrived from 
the north-west) were they not contented with the Indus (which 
they first encountered) ; and how was it that they only found 
the satisfaction of this want in the Ganges, wliicb was so far 
distant'*? Finally the oldest city which according to the legend 
was fownded in India is that which the Pandavas (the piUe, 
white men in contro-st to the 'black population) built in the 
Khandava forest on the Jiuuna — a situation which was far re- 
moved from the Indus and close to the holiest region. These 
Pandavas, according to another passage of the ]\Ialmbharata, 
were educated in the Himalaya. 



'* [But tbey seem ot first to liave paid very little regard to the Gang«f, ta 
WG sball see furtlier on. — J. M.J 

flSCT. X.] 



*♦ )^Tiile, however, through the force of all these considerations, 
we incline to the opinion that the tract between the Jumna and 
the Saraavati was at once the lutliest and the earliest seat of the 
Aritms in India, we arc not thereby compelled to abandon the 
view that they might have penetrated by the route of the Indus, 
though this view thus loses much of its prob;ibility. If in the 
description of the several tracts given in Manu's Institutes, we 
are to recognise a sort of history of the extension of the Aryaa 
in northern India, we may also deduce from it with tolerable cer- 
tainty that the mogt ancient colonists immigrated in small num- 
bers. This view is further supported by the circumstance that, 
though apparently 3000 years have elapsed from that time to 
the present, still so many fragments of the aboriginal population 
have survived, and have never been entirely subdued by the 
Arians, even in their most flourishing era, and in the tr&cts of 
which they were most completely masters. If, however, the 
Arynfl immigrated only in small numbers, they could not 
possibly have traversed the regions lying between the Indus and 
the Sarasvati (which Alexander the Great in his invasion never 
reached), without ever entertaining the thought of settling till 
they arrived at the tract which, it appears to us, must be 
regarded as their first fixed abode. I conjecture from this that 
they crossed over from their ancient seats beyond, and in the 
northern rallies of, the Himalaya, into the southern plains, 
rather as peaceable colonists than as martial conquerors. The 
Bs over which the roatl lies are, it is true, difficult, but by no 
means insuperable, and are available for traffic and every sort of 
intercourse, though difficult for warlike operations. By these 
routes the first Aryas who settled in India, partly following the 
various branches of the Ganges, might have found their way, 
through Kemaon, Garhwal, or Sirmur, to the plains situated to 
the south of the Himalaya. Here they founded Indraprastba, 
and thence spread themselves around, subduing the feeble 
Mlechhas, and grmlually conciuexiog all the parts of India which 
were not too difficult of access." 

A. W. von Schlegel, od the otber hand, thinka that the ludo- 
Arians must have penetrated into India from the west. After 
describing the difficulties of the sea routes leaiUng to India from 
the south, and of the land route over the Himalaya from tlie 
north, he goes on to say : " The western side of India appears 
to be more open, as from Kashmir to the delta of the Indus 
the boundaries are not otherwise marked than by that river 
itself. But in its upper course the Indus is not navigable, 
owing to its rapidity and its cataracts: and in additioa its 
right Imnk is flankod by nio\mtaiu8. Towards the sea it spreads 
out into, or is surrounded by, marshes : more in the interior, 
and even above the contiueuce of the five rivers, it is bounded 
by eandy deserts. From that point to the place where it enters 
the plains neai* Attock, a tract intervenes where the passage 
iiijiy be more easily cilected. Accordingly it is on this side that 
India has always been entered by foreign conquerors, by Semira- 
mis, if her Indian expedition is authentic, ... by Alexander 
the Great, Seleucus, <aud the Greek kings of Bactria, by the 
Indo-Soythiaus, or nomad riices, wlio invaded certain provinces 
during the century preceding our era: by Mahmud of Ghazni, 
by the Afghans, tlie Moguls, and the Persians under Njidir 
Shah. Thus all probabilities are unitt-d in favour of the suppo- 
sition that the ancestors of the Hindus came from the same 
side; a supposition which we find to be confirmed by argiunent«5 
of another kind. The Panjab would consequently be the first 
country occupied by the colonists. Tradition does not, how- 
ever, celebrate this as a classic region. On the contrary, 
in a passage of the Mahabharuta, published and commented on 
by La.sseu, its inhabitants are described as less piu-e and correct 
in their customs tlian the real Aryas, as perhaps they had been 
corrupted by the vicinity of barbarians. This leads us to be- 
lieve tliat it was only iifter the colonists had spread themselves 
over the plains of the Ganges, that thoir form of worship, and 
the social order dependant upon it, cuuld have .issmned a per- 
manent form." — Essais littdraires et historiques, pp. 455-457. 

fBCT. X.J 

TOE i?n)rA?r A?rD cogitate natioxs. 

The same view is taken by Lassea (Indian Antiquities, 
i. 511). 

"The Indians, like moat other nations of the ancient work], 
believe themselves to be aut<^)chthonoufi : their sacred legends 
represent India itself as the scene of creation, as the abode of 
the patriarchH, and the theatre of their deeds ; and they have 
no recollection of having sprung from any country out of India, 
or of having ever lived beyond the bounds of their own Bharata- 
varsha. (See however above, p. 323, ff.) 

" It is true that we might be tempted to discover in the supe- 
rior sacredness which they ascribe to the north, a reference. 
unintelUgible to themselves, to a closer connection which thej 
had formerly had with the northern countries ; for the abodes of 
most of the gods are placed towards the north in and beyond 
the Hiinalaya, and the holy and wonderful mountain Meru is 
flitiiated in the remoteat regions in the same direction. A more 
exact examination will, however, lead to the conviction that the 
conception to which we have referred, hna been developed in 
India itself, and is to be derived from the pecvdiar character of 
the northern mountain-range. The daily prospect of the snowy 
summits of the Himalaya glittering far and wide over the plains, 
and in the strictest sense insurmountable, and the knowledge 
whicli they had of the entirely diflferent character of the table- 
land beyond, with its extensive and tranquil domains, its clear 
and cloudless sky and peculiar natiu-al productions, woidd neces- 
irily designate the north as the al)ode of the grnls ami the 
theatre of wonders ; while its holiness is explicable from the 
irresistible impression pro<luced upon the mind by surrounding 
nature. Uttara Kuru, the Elysium in the remotest north, may 
be most properly regarded as an ideal picture created by the 
imagination, of a life of tranquil felicity, and not as a recollec- 
tion of any early residence of the Kurus in the north. Such at 
leant is true of the representation which we have of this country 
in the epic poems. It is, however, prolwible that originally, and 
as late ua the Vedic era, a recollection of tliis sort attache<l 





itself to that country, though in later times no trace of it has 
heen preserved." 

After stating the reasons (already detailed above, pp. 307 — 
309), which lead us to conclude that the Indians could out 
have been autochthonous, Lassen proceeds as follows (p. 515) : 
" There is only one route by which -we can imagine the Arian- 
Indians to have iinmiyrated into India; they must have come 
through the Panjab, and they must have reached the Panjab 
through western Kabulistan. The roads leading from the 
country on the Oxus into eastern Kabulistan and the valley of 
the Paujkora, or into the upper valley of the Indus down upon 
Gilgit, and from thence either down the course of the Indus, or 
from Gilgit over the lofty plateau of Deotsu down on Kashmir, 
are now known to us as the roughest and most difficult that 
exist, and do not appear t-o have been ever much or frequently 
used as lines of communication. We can only imagine the 
small tribes of the Darodas to have come by the second route 
from the northern Hide of the Jlindukush into their elevated 
vallies, but we cannot suppose the mass of the Arians to have 
reached India by this road. All the important expeditions 
of nations or anniea which are known to us have proceeded 
through the western passes of the Hindukusb, and if we sup- 
pase the Arian-Iutlians to have come into I;idia from Bactria, 
this is the only route by wliich we can assume them to have 
arrived." It is true that the Hindus attach no idea of sanctity 
to the Panjab; on the contrary, " the Sarasvati is the western 
boundary of the pure laud, govei-ned by Brahmauical law. 
iTiere are, indeed, Indians dweUing further to the west, but 
they do nut observe the Bralimanical ordinances in all their 
integrity. But tliis mode of regarding the western tribes can 
only have arisen after the Indian institutions had been deve- 
loped, and a marked difference had become observable between 
the people living east, of the Sarasvati, and those on the western 
border. The people of the Panjab always appear as descended 
from tiie same stock, and in spite of the aversion in question, 




the epic legends recnunt to us frequent relations between the 
kings of the pure portion of India, and the tribes to the west- 
ward. There is no break in tlie chain of Indian races towards 
the west*" 

M. Burnouf briefly indicates his opinion on the question with 
which we are now occ\xpied, by speaking of " the movement 
which from the earliest agea had carried the Arian race from 
the Indus to the Ganges, and from the Ganges into the Dekhan," 
&c. Preface to Bhag. Pur., vol. iii. p. xxix. 

I am not aware whetlier Professor Roth lias ever expressed 
an opinion as to the precise route by which the Arians entered 
India; but in bis work on the Lit and Hist, of the Veda (1846), 
p. 136, he writes as follows: "It is more than probable that the 
bulk of the tiibes which we may designate as the ^''edic people 
dwelt nearer to the Indus thiui the Jumna, and that the battle 
which is described in the hymn before us was one of those con- 
flicts in which the northern tribes pressed upon the southern, 
on their way towards the regions which they were eventually 
to occupy. The Indus is well known and fre«]uent]y celebrated 
in the hymns of the Rig-veda, while at this moment I know of 
only one passage in which the Ganges is mentioned, and that 
fldily in a way which assigns to it an inferior rank." 
" The same writer in his article on " Brahma and the Brah- 
mans," in the Journal of the German Oriental Society for 1847, 
p. 81, again expresses himself thus: "When the Vedic people, 
expelled liy some shock — and that at a period more recent than 
the majority of the hymns of the Veda — relinquished their seats 
in the Panjab and on the Indus, advanced further and further 
to the south, drove the aborigines into the hills, and occupied 
the broad tracts lying between the Ganges, the Jumna, and the 
Vindhya range, the time had arrived when the division of 
power, the relations of king and priest, could become trans* 
fnrmed in the most rapid and comprehensive manner." 

Professor Weber also speaks of the Arians as at one time 
dwelling beyond the ludua. In his Hist, uf Ind. Lit. (1852 y. 




pp. 2 and 3, he writes : " In the oldest ports of the Rig-Tt-da 
the Indian people appear to us as settled on the north-we-steni 
borders of India, in the Panjab, and even beyond the Panjah, 
on the borders of the KubUii river, the Kanf>T}v in Kabul." Tbe 
gradual diffusion of this people from this point towards the eawt, 
beyond the Sarasvati, and over Hindustan as far as the Gang«i, 
can be traced almost step by step in the later portions of thft 
Vedic writings." 

In his " Recent Investigations on Ancient India," the same 
writer similarly remarks : " The oldest hymns of the Veiia show 
us the Arian people still dwelling beyond, or at least, only on 
the north-western frontiers of India ; viz., in the tract between 
the Cabul river and the Indus, as well as in the Panjab. Their 
advance from this point, and extension over India Ciiu be tmce<l 
step by step in their literature. Their road lay to the north 
of the great desert of Marwar, frova the Satadni (the modern 
Sutlej) to the Sarasvati, a river (esteemed at a later period aa 
of the highest sanctity) which loses itself in the sands of the 
desert. This must have been a point where they made a halt 
of long continuance, as may be concluded from the great sacred- 
neas ascribed in later times to this region. At that period it H 
formed tht- Lovuulary line between the Brahmauical organization 
which was being now formed in Hindustan, and those Arian 
races of the west which retained the free manner of life inherited 
from their forefathers." — Indian Sketches, pp. 13, 14. 

■" In his Indische Studien, vol. i. p. 165 (publlsLed 1849, 1850), Weber 
speaks of llie " Arian Indians being driven by a deluge from their home (sec 
above, p. 324, IT.), and com'wg/rom the north, not from the west {as LoMtn 
i. 515, wtll have it) into India; first of all to Kashuir and tiiC Panjab; as 
it is only in this way that wc can explain the northern Kurus and the 
northern Madras wiih wLum tbe conception of the golden age became after* 
wards associated." Asi, however, in tlie passages <]ui)tcd in the text, which 
were written at a later tlalo, Weber supposes the Arians to have dwelt on 
the Kabul river, they must, in order to arrive there, have either arrived by 
the route which Lassen assigns, or have afterwards spread themsdves to the 


• ECT. X.] 



M. Langlois in the Preface to his French Translation of the 
R.-V., speaks to the same effect, pp. ix. x. : " The hymns of the 
Rig-veda were composed for tribes which liad come from the 
banks of the Indus, and were liviug in the plains watered by the 
Ganges. This people seems to have belonged to that great 
branch of the human race known under the name of the 
Aryas. They brought with them a raild and simple civiliza- 
tion, patriarchal manners, a ptilished language. • . . These 
Aryas, as they established themselves in India, drove back before 
them the ancient jwpulations, which then proceeded to occupy 
the forests and mountains, and which, on account of their savage 
customs and mmdercnia depredations, became, for the Aryas, the 
types of those evil spirits which they have depicted in their books. 
At the head of the first colony there must have been a prince of 
the Ari.<ui nation called Mann, whom the traditions represent as 
the father of mankind.'^ 

In another place, in a note to R.-V. i. 33. 3 (p. 264, vol. i. of 
his work, note 2), the same author writes still more explicitly 
as regards the point under consideration : " It is my opinion 
that the Indian colony conducted by Mann, which established 
itself in Aryavartta, came from the countries which lie to the 
west of the Indus, and of which the general name was Aria, 
Ariana, Hiran.'^ 

Professor Miiller does not, afi far as I am aware, any where 
determine the route by which the Arians arrived in India, 
more precisely than is done in the following passages: "At 
the first dawn of traditional history we see these Arian tribis 
migrating across the snow of the Himalaya, southward towarcU 
the 'seven rivers' (the Indus, the five rivers of the Panjab, 
and the Sarasvati), and ever since India has been called their 
home." — I^ast Results of Sanskrit Researches, p, 129. And 
D, at p. 131, he writes: "After crossing the narrow passes 
of the Hindu-kush or the Himalaya, they [the southern Arians] 
conquere<l, or drove before them .... the aboriginal inhabi- 
tants of the Trans-Himalayan countries." In bis "Last Results 

A A 





of the Turanian Researclies,"' p. 340, he says: "The JVriaas 
were strangers in the laud of the Indus and Ganges, but no one 
can now determine the exact spot whence the}' came, and where 
they had been previously settled. Traditions current among 
tlie Brahmans as to the northern r^ons, considered, the seats 
of the blessed, may be construed into a recollection of their 
noi-thern immigration — ^holy places along the rivers of northern 
India, where even in later times Brahmans went to learu the 
purest Sanskrit, may mark the stations of their onward course — 
the principal capitals of their ancient kingdoms may prove the 
slow but steady progress towards the mouths of the principal 
rivern of India, but with the sources of those rivers the homes 
of the Arian strangers vanish from our sight, even after we 
have reached the highest points of view accessible on Indian 
ground." (See above, pp. 309, .310.) 

Whatever other and minor differences of view may exist 
between the several authorities whom I have last cited, they are 
all of one accord at least in regard to this one point, that India 
is not the original country of the Hindus. 


BacT. XI. The immigrotion of the Indo-Ariatu/rom the Xorth-weti 
proltuhle lif the tenor of the Veclic Hymnt, 

The immigration of the Arians, the progenitors of the Brah 
manical Indians, into India from the north-west, is further 
rendered probable by the fact that the WTiters of the Vedic hymBB 
appear to be moi^t familiar with tlie countries lying in that 
direction, i.e., with the north-western parts of India itself, as well 
as with the countries bordering on, or beyond the Indus, and with 
the rivers which flow through those regions ; while the coimtries 
and rivers in the central and eastern parts of India are more 
rarely mentioned ; and no allusion whatever is made to the 
regions of the south. On this subject I borrow the following 
remarks from Professor Roth's work on the Lit. and Hiat, of 
the Veda, p, 136: "The Siudhu (Indus) is well known and 



frequently celebrated in the hymns of the Rig- Veda, while at 
present I know of only one hjonn in which the Ganges is men- 
tioned, and that only in a subordinate capacity. This passage 
occurs in one of the hymns ascribed to Sindhukshit, son of 
Priyamedha, which h addressed to the Sindhu, ' the most 
copious of streams,' (apasdm apaatamn). The other rivers are 
solicited to regard graciously the praises of the poet, which 
are dedicated to the Sindhiu^** The passage is, after Ynaka 
(Nirokta, ix. 26), to be explained thus: * Ganga, Yamuna, 
Saraavati, Sutudri, with the Parushni, receive graciously my 
hymn. Jlarudvridha, hear with the Asikni, the Vitasta; 
Arjikiya, hear with the Sushoma.' '''' 

Another pas«ige in which the Indus is mentioned is the follow- 
ing, R.-V. L 126. 1. : ^W^T^ ^V^^ "RH^ ^pftTT 

«. R.-V. jc. 75. 6: ■?:Tf ^ Tif If^ ^T^f^ l^?5f|- ^^ 

^^TTT ^f^^^l ^f^WiJT FTfV f%fTWm#^ a?^T 

■g^it^T^ II 

" Part of Yaaka'g note (Nirultta, ix, 26) ia os follows : X^ ^ T^ 

^9^ «j<^f?T ij^fsT ^^^f^ ^VwH. ^n^w^^ ^f«^ 
^ w^ jf^^ fl<T^^ ■'^mNft^ ^ni5^ ^fm^ 
^f?i ^57?^r^M • • • TTRffT m^Hinpyiss:! • • • 
^jf?r^ji5iRTf%^t . . . ?T^fVT: w#T T^ Jf^ Tprr 

^^^rf%l • • • ^[iif<*l<<i ^Mlf^rtJIS:! (See Parti, p. 116, 
end of note 23.) " The entire sense is, ' Receive this hymn, O Gangi, 
Yamana, Saraswiiti, l^utudri, Purushni, and Marudrridhii along with the 
Asikni, and Arjikira along with the Vitustii and SuslioiniL' . . . Pa- 
rushni is a name of the Iravati. . . . Asikni wenna 'black.' . . . All rivers 
[maj be called] KUrudvridhn, because they are swollen bj the Maruta. . . . 
Arjiklyi is a niune of the Vipiii." See RolL's remarks on tkeae rivers, in 
hia Lit. and Hist, of the Vedv pp. 136—140. 

A A 2 


AFn>TrrES and derttatiox or 


CHAT. !>. 

^arr^^ "^(fTTT ^^^ l^*1l«i:i "With my intellect, I pro- 
duce energetic encoiuiiims upon Svanaya, the son of Bhavja, 
who dwells on the Siudhii; the invincible prince, vrho, desirous 
of renown, has performed tlirough me a thousand oblation*.'* 
In the 7th verse of the same hymn we find a reference which 
iudiciitea familiai-ity with the country of the Gaiidliaris and 

its sheep: ^ ^T^fw 0*^*11 <l«-m0^ l f*l<< l Q* l i 

*' I am all downy, like a ewe of the Gandharis." Gaudhara is 
placed by Lassen (in the map of ancient India in VoL IT. 
of his Indian Antiquities) to the west of the Indus, and to th© 
south of the Cophon or Kabul river, the same position to which 
the Gaudaritis of the ancients is referred." The word Sindhti 
also occura in the following passages of the Rig-veda, vii, 
1 94. 16; i. 122. 6; ii. 15. 6; iv. 30. 12; v. 53. 9; 
vii. 33. 3; viii. 20. 25; x. 64. 9. It is, however, difficult 
to gay whether the ludua be always meant. The last of these 
passages (which occurs in a hymn to the Visve devas) is as fol- 
lows, E.-V. X. 64. 9: WT^rft w^: fwf^^^f» iffV 

"^(TSm^ Wr W«=T^ ^T ^^flH "I^et the Sarasvati, the 

Sarajni, the SitiJbu, with their waves; let the great rivers come 
with their succour. Divine waters, niuthera, flowing, impart (?) 
to ua your waters with butter and houey.'^ 

The verse which has been cited above fix)m the Rig-veda, 
X. 15. 6, in the extract from Professor Roth's work, is followed 
ty another '* in which the names of several other rivei-s are men- 

" Tiie Ganilarii are mentioneJ by Ilcroilotus, vii. 6(5, along with the 
Portliiaiia, CLoroatuians, SiigJians, and Uadika:, aa forming port of the annr 
of Xerxes. See the Asiuiic Researches, vol. xv. 103, ff. 

,• R.-V. X. 75. 7 : WSJTJ^ 1?^ ^Hl^ W^ ^Wr^T J:V7n 

^ajT 7QJ I -^ fw^ ^v[m wnr^ ^ ^^^\ w^ 



MCT. Xl.] 



tioned, viz., the Trisbtama, the Rasa,'* the Sveti, the Kubha, 
the Gomati, and the Krurau. In Roth and Bohtlinsrk's Lexicon 
the last three Btreams are set down as being affluents of tlie 
Indus.^"' That they were rejilly so is rendered proUible by 
their being mentioned in conjunction with that river. In the 
case of the Kubha, the probabihty is strengthened by its name, 
which has a close resemblance to that of the Kuf'^v, or Kabul river, 
which falls into the Indus, a little above Attock (see the passage 
from Weber'a Ind. Liter., above p. 352). This river is men- 
tioned again in R.-V. v. 53. 9 : HT ^ T^Sf'T'^^^ ^^TT 

gftfwt ^raH T<T ■g^TT^ ^: 11 '« Let not, Marnts, 
the Rasa, the Anitabha, the Kubha, the Krurau or the Sindlm 
arrest you: let not the watery Sarajni stop you: let the joy 
you impart come to us." Another of the rivers named in the 
verse previously cited (R.-V. x. 75. 7), and declared by Roth 
to be an affluent of the Indus, is the Gomati. It is not 
necessary tliat we should identify this river with the Gomati 
(Goouitee), which rises to the north-west of Oude and flows 
past Lukhnow. A river of the same name is mentioned again 
in R.-V. viii. 24. 30: T^ ^qfjifdl ^T ^ l ^ H dl^HW 
f^^fffll "This powerful man dwells afar on the [l>auks of 
the] Guuiati." It is quite possible that the name of the river in 

'° The Rb&u is consMerfid bv Dr. Aufreclil, in his explanation of R.-V. x. 
108, to denote there nnd elsewhere the " niill»y way." See Journal of the 
German Oriental Society, vol. xiii. p. 498. Yaaka merely espluins it as 
meaning a river : '^^^ •T^ [| Nir. xi. 25. 

"*• In his Elucidation* (Erlauterungen) of the Nirukta, p. 34, note. 
Professor Both remarltg: "Tho Kophcn is the Kubha of the Veda, 
in«Titione<l in R.-V. t. 53. 9, and x. 75. 7. If we identify the Kruniii and 
Gomati of this last text, with the Kurum nnd Gonial which flow into the 
Indus from the west (aa Lassen proposes in a letter), we m»y regard the 
rivers whose names precede [the Tri^h^iima, Rasa, SvelT, and Anilnbhii] as 
being affluents of the Indus furclier to the north iban the Kophen." 

A A 3 



Oude may have been borrowed from some stream further west,* 
Another river, the Sijvastu, which may be an affluent of the M 
In.lua, is mentioned in R.-V. viiL 19. 37 ; ^^T^T ^rfv 
i^Jcff^ II These words are quoted in Niniktii, iv. 15, aud 
explained thus: ^Tpg^^l 7^ ^^9 W^f^ll "Suvasta is 
a river; txigma means a holy place." On this passage Both 
observes, Erliiutcrungen, p. 43 : " The bard Sobhari is recount- 
ing the presents which he received from Trasadasyu, son of j 
Piirukut«i, on the Lanks of the Suvastu. In the Mahabbarata, 
vi. 333," the Suviwtn is connected with the GaitrL Now, 
according to Arrian, IndicA, 4. 11," the Soastos and Garoiss 
flow into the Kciphen. From comparing these two passages, ■ 
it results with tolerable certainty that the Suvastu is the | 
same as the modem Suwad, a stream which flows into the 
Kabul river from the north, after first joining the Panjkora." 

Returning now to R,-V. x. 75. 6, and taking first the most 
westerly streams there specified, we come (1.) to the Vitasta or 
Behat, (2.) the Asikni or Chenab (Akesines), (3.) the Parushnf, 
Iravati, or Kavec, (4.) the Arjikiya, Vipas, or Beeas, and (5.) the 
Sutudri, or Sutlej. Yaaka, as we have seen, identifies the Pa- 1 
rusbni with the Iravati, and the Arjikiya with the Vipafi ; Pro- ' 
fessor Roth considers the Asikni to be the same as the Chenab 
or Akesines ; and there is no doubt that the Vitasta is the 
Hydaspes, and that the Sutudri is the Sutlej. We have, conse- 


I. ^ 

*° There is a stream called GomXti in Kemaon, wliich must be distil 
from the river in Oude, as the laUer rises in the plains. 

" In the list of rivers in the description of Jarabukhnrifjii, The wordj 

»re: ^T^ ^^T^ ^TTK!^ ^R^J^ ^S^^V^B^^H " IT^e 
Tutu, the Suvastu, the Gauri, the Kampana, and the Uiranvati." 

rai>peinv, ««;«rui »t roir IfCor. " The Xophen unites with the Indus in Peuk&- j 
Ixctis, bringing with it the Malantus, the Soastus, and the Garraas." Pro-l 
feasor Wilson, (Ariana Ant. pp. 18,1, 190, 194) thinks these two last nanietj 
really denote one and the same ri%-cr. See App. note I. 

sacT. XI.] 



quently, in this passage on enumeration of the rivers of the 
Panjab. The Astkni is again mentioned in R-V, viii. 20. 25 ; 
the Parushui in R.-V. vii. 18. 8, 9, and viii. 63. 15; the 
i^utudri in iii. 33. 1 ; and the Vipaa in iii. 33. 1, 3. and iv. 
30. 11. 

The other rivers named in the passage so often referred to, 
TL-Y. X. 75. 6, are the Sarasvati, the Ganga, and the Yamuna, 
The following are some of the most remarkable passages in 
which the Sarasvati i.s celebrated. In iii. 23. 4, it ia thus men- 
tioned along with the Drishadvati, (with which Mauu, ii. 17, 
also associates it) and the Apayfi : f^f (^J <[^ ^^ "^ "jf^T^T 

TacT^n^ ^f^^ 'viMi*K} "l^^^t »rr^ ^iM<ji<<i 

V<4!4(^( <^<lj f<<^P|tl "O" an au^spicious day I place 
thee on the most sacred spot of Ilsl, the earth. Shine, Agni, 
wealth-bestowing, in the assembly of men on the banks of the 
Drishadvati, the Apaya, the Sarasvati.' In R.-V. vi. 61. 2, the 
tame river is thus magnified : ^ IJTjf^T f^^WT 4.<=IH5.<i1r^ 

^K^fllH^ ^9T f^<^l*i*f T|tf7Tf»T: II " By her force, and her im- 
petuous waves she has broken down the sides of the mountains, 
like a man digging lotus fibres. For succour let us, with 
and ceremonies, invoke Sarasvati who sweeps away her banks." ** 
In verse 13 of the same hymn the same epithet apasdm apa- 
etamd, " most copious of streams," which is applied to the 
Sindhu in R.-V. x. 75. 7. (see above p. 355), is also assigned 
to the Sarasvati. 

*• In reference to lUis verse, Yaska observes, ii. 23 : rj^ ♦Jl^'ft" 
♦I* (e|rt "There are texts which speak of Sarasvati both as a river and 

as a goddess. In the following she is referred to as a river." 
quotes the verse before us. 

A A 4 

lie thea 


[cm A p. It. 

Hymns 95 and 96 of the seventh book of the Rig-veda are de- 
voted to the praises of the Sarasvati and her male correlative 
the Sarasvat, The first and part of the second verse of the 
former hymn are as follows : T( s^l^^ll \rRIWT W^ JPSfT 

H<4sifn v^wiT^ isim+ft: ^:i "R ^^VPTT T^ 'nfr 

•!<'^«ll 35f^ ^$rft f^lf^WI ^l*J*J5lr^H " This Sarasvati has 
flowed on with a protecting current, a support, an iron de- 
fence. This stream [or the Sindhu] rushes on as if [driven by] 
a charioteer, by her greatness outrunning all other rivers." 
Sarasvati is known as the one river, flowing on pure from 
the mouutains to the sea." " 

Tlie Jumna is mentioned in two other passages of the Rig- 
veda besides x. 75. 5. In v. 52. 17, reference is made to 
property in cows and horses on the banks of the Yamuna ; *• 
and in vii. 18. 19 it is said that the '' Yamuna gladdened ludra." '^ 
I have found a reference to the Ganga in one other passage be- 
sides X. 75. 5, viz., in vi. 45. 31,'* where the adjective gangya^ 

'* Sec the Iranalation of this verse in Benfejr's Glossary to the Satna-redo, 
p. 157 under the word ra/Ai. 

^ Langlois, rol. iii. p. 241, note IS, thinks that SArasvaii in this hjmn 
Rlnnds, not for a river, but for " the goddess of sacrifice," with her libations. 
" These llbaliuns form a rircr, which flows from the mountains, where the 
sacriGcc \a perfurmed, and where the soma plant Is collected. This river 
riowR into the tamudra (sea), whiuh is the vessel destined to receive the 

MR.-V.v. 52.17: ^IWm^^^rfW ^fT^ ^ TTVV »rai 

f^ f% TTWT TT^ ^^11 
" R.-V. vii. 18. 19: "^IT^f^ Jf^ T^Tf^ll 
" See Roth, Litt. und Gesch. des Weda, p. 136 ; and above p. 354, 

The words arc: '^3^: ^tqI 7f 4|[ JJJ : || Roth, tub voce kakafaa, nja, 

the sense of the word kahnha is uncertain. Lanj;loia does not translate it. 

AVilson misapprehends Sajana's explanation. 





154. I 




"belonging to the Ganga," occurs. But the Rig-veda con- 
tains no hymn devoted to the celebration of the Ganga, such as 
we find appropriated to the Siudhu and Sarasvati. 

The Sarayu is also referred to in three passages in the R.-V. iv. 
30, 1 8, V. 53. 9. and x. 64. 9. The first of these texts runs thus : 

^37T ^T ^^ ^rr^T «j<^R*i^ ^ny\:\ ^'^if^Trm 

^Sl^^Jt* I " Thou hast straightway slain these two Arya.s, Arna, 
and Chitraratha on the other side of the Sarayu." The 
second and third have been already quoted in pp. 356, 357. 
The Sarayii named in these passagesj particularly the two, 
may be different from the river of the same name which now 
flows along the north-eastern frontier of Oude, as it is men- 
tioned in connection with rivers all of which appear to be in 
the Paujab. But it is not necessary to suppose this, as we 
shall presently see that one of the Vedic rishis was acquainted with 
Kikata or Behar. In the Rig-veda we have no mention made 
of the rivers of the south, which have in later ages become so 
renowned in Hiudusthan for their sanctity, the Narraada, the 
Godaveri, and the Kaveri. 

We have already seen (p. 337) that the Himalaya mountains 
are mentioned in the Atharva-veda, In a fine hymn, the 12l8t 
of the 10th mandala of the R,-V., also, we have the following 
verse : x. 121. 4 : i!J^ f^lRitiV *ff^?m "^m ^^T <^«l 
4j-^|f;: II " He whose greatness these snowy mountsiins, and the 
sea with the river declare," ™ «&c. But no allusion to the Vindh^'a 
range, which runs across the central parts of India, is to be 
found in the Rig-veda. 

The following text from the R.-V. shows that the author of 
the hymn (said to be Vi^wamitra) knew something of the coim- 

** See Miiller's translalioo in Bunscn** Gott in der Gcschlchtc, Part II. 
p. 107. The IlimSlnya i.i also mentloin'J, A.-V, xii. 1. 11 : fl|<^€J(^ 
^lirTT f?^R^ST^ ^ yM^ ^^TfT^II " Ma, thy mountains 
be iaovj, O earth, nml thy wilderness benutiful."* 

a bint in Toska : 


tries to the eastward as far as Kikata or Behar : R.-V. iii. 53. 14, 

vm^i ^ ^ »n: twt'^^ "^^ %^ixt^ »to^ 

^T'tf^T m II " WTiat are thy cows doing among the Kikatas? 
They jield no niilk for ohlations; and they heat no fire. Bring 
us the wealth of Pramaganda [or the usurer] ; and subdue to ub, 
O Maghavat, < Indra), the degraded man (naichasakha)." Yaaka 
explains Kikata as "a country ijihabitfd by people wlio were not 
Aryya.s ;" Nirukta, vi. 32 : <l?|«*d| ifT^ $lftS«1l *4f^=ll*i: tl 
The word Kikata is given in the vocabulary called Trikanda- 
^eoha, as equivalent to Magadlia. In Bohtlingk and Roth'a 
Dictionary, the following lines are quoted from the Bhagavata 
Purana i. 3. 24.: rTrT: ^^ ^3?"?% WTTre ^^^^TT^I 
5^ 5mrr^T^: 3^Krg Hf^f?TI -Then, when the 

•* SSyapi g!?e« an altei-nallve explanation of kikal/ih^ borrowed from 

^^ f^<< i r* T^ ^<i<M^ i *i<j i ^<infi T: f* 
r: 3RgrT fqRiT ^r^iTra^ wmt m 

^ MT Tin ^^mt Sfrf^^T: 4f1*ii:H "Or tUc KUco(5s are 
atheists, who, being destitute of faitti, saj, ' what fruit will result from sacri- 
fices, alms, or oblations ? rather eat and drink, for there is no other world 
but this.' " In Saysna'a introduction to the Rig-veda (Miiller's edit. vol. i. 
p. 7), an aphorism of the MimariEa, with a comment, is quoted, in which an 
objector demurs to the eternity of the Veda, because otjccta and per»niu 
who existed in time are mentioned in it. In the objector's statenienf, Nai- 

chaiiikiia is spoken of as a city, and Pramaganda OS a king : f^ ^ ^QcffWl 

^rnsniTr* ll " in t'>e verse, * what do thy cows among the Kikataa, &c.,' a 
country named Kikata is recorded, together with a city called NaichaiakLa 
and a king called Pramaganda; all which are non-eternal objects." 



•ICT. XI ] 



Kali age bas begun, a person named Buddha, son of Anjana, 
will be born among the Kikatos, in order to delude the enemies 
of the gods (the Asuras)." The commentator on the Bhsig. 
Put. explains the Kikatas by If^ 4| ^ I M ^ X( I " '" ^^^ 'dis- 
trict of Gaya." Again, Bhag. Pur. vii. 10. 18, it ia said : 

■^TTT^ U*<«TlSf^ ^4di: II "Iq everyplace where those 
who are devoted to iiie, wlio are calm, who regard all things as 
alike, who are holy and virtuous are bom, the men [of that 
country] are purified, even if they be Kikatas." Weber, how- 
ever, in his Ind. Stud. i. 186, states hi.s opinion that the 
Kikatas were not (as Yaska tells us) a non-Arian tribe, but a 
people, who, like the Vratyas, were of Ariau origin, though 
they did not observe Arian rites ; and they may, be thinks, 
have been Buddhists, or the forenmners of Buddhism.'' 

From these passages there seems to be uo doubt that the 
Kikatas were a people who liveil in Magadiia or Bebar. 

The following verses from one of the mantras of the Atharva- 
Teda, V. 22, quoted and explained by Professor Roth in his Lit. 
and Hist of the Veda, pp. 37 — 42, may tend to show what were 
the limits of the country occupied by the Aryas at the date of 
its composition. These liraita coincide with those indicated 
in the preceding passage from the Rig-veda, in which the Ki- 
katas are mentioned. This mantra contains an invocation to 
Takmau, apparently a perdonified cutaneous disease, who is 
supplicated to withdraw to certain other tribes, whose names are 
specified, and whom we may therefore with probability conclude 
to have been regarded as without the Aiian pale. A.-V. v. 
22, verses 5. 7. 8. 12. 14 : ^StT^ ^Sm *^^<i'r\ '^^T 


K5!^?rr ^«- ^f^ qrwi ^mt^ rr^^ ^ TRj%^Tf% ^ 
^mii^ii 7pw\ >rr^ 'iHiIj'T ^^^ ^Tfij*<^i ^r^i wm 
'Hi<^^^< 's^ <i^i4H7?^ ^pm^ii^^ii Ts^fcirft?^ ^ara- 

T^lfw ll\8ll " Hia (Takman's) abode are the Mijjavate, his 
abode the Jlahfivrishas. As soon as thou art born, Taktnrui, 
thou sojournest (?) among the Bablikas. Go, Takman, to the 
Mujavats, or far away to the Bablikas. Choose the female 
Sudra for food: and sliake her. Passing by oxir friends(?), devour 
the Mabavrishas and the Miijavats, We point out to Takiuan 
these or those foreign regions. Takman, along with thy 
brother Balasa, and with thy sister Kasika (cough), and with 
thy nephew Paman, depart to that foreign people. We transfer 
Takujan as a wervant, and as a treasure, to the Gandharins, the 
Mujavata, the Angas, and the Magatihas.^ 

The Mnjavats are again mentioned in the Vajasaneyi-sanhita^ 
3. 61, as follows: T^TTf^^ ^ ^^^ ^ ^TTt" ^^SRrft 

Sf^i ^raTTff^^T [M-ii4i^*j: lif^^r^T ^^ 5f: 

fljlTSiftf^ll "This, Rudra, is thy food; with it depart 

•' Miijavat is explained by tlie commentator on the V.-S , as the name 
of a mountain, llic place of Kudra's abode, TT^sl'^ l'<4141 *f^<1 V^dT 
fJ^Z ^ ^lUI^MH II The commcnintor on tlie J^atapotba-braliman* 
says it is the " Northern Mountain," \j«^-^: ^^<f I W The Satopatha- 
brahmnna (2.6.2. 17.) thus comments on the text of the Taj.-S. after 
quoting it: TR^tFI ^ ^[l^ ^ff^l r^^ UI«i4J^<<|- 

TfTTI f^Miqi^^^'T^ Tn!\l Wf^ f% ^ ^rg^ 

>BCT. Zl 




beyond the Jliijavata. With thy bow unbent, and concealed 
from view, and clad in a skin, pass beyond, umnjuring us and 

The !Mujavatfl being mentioned along with the Bahlikas, a 
Bactrian race, and with the Gandharis (see above, p. 356,) may, 
as Roth thinks, be a hill tribe in the north-west of India ; and 
the Mahavrishas may belong to the same region.'^ 

The Augas and Magadhaa mentioned in verse 13, are on the 
contrary, tribes living in south Behar, and the country border- 
ing on it to the east. We have t!ius in tliat verse two nations 
situated to the north-west, and two to the south-east, whom we 
may suppose, from the maledictions pronoimced on them, to 
have been hostile, or alien tribes, who lived on the borders of 
Rralimanical India, and to have been beyond it;3 boundaries at 
the time this incanfcvtion waa composed. (Roth, Lit. and Hist, 
of the Veda, p. 42.) 

It does not, however, follow that the tribes who, in the 
AtharvA-veda are spoken of as if they were hostile, or alien, 
were really of a uon-Arian origin. 

Thus, the Arians appear in later times to have been in com- 
mimication with the Gaudharas. In the i^atapatha-bralimana 
allusion is made to a royal sage called Svarjit, son of Nagnajil, 

f^'lf^l fT^T?Cr? ?lf%WTVr Tf?lH "Men go on their way 
with provttion. He therefore sends him (lludra) off with proviaion, wher- 
ever be has to go. Here his journej ia beyond the Mujavats ; hence he says 
'pass beyond the Mujavats;' 'with bow unbent and concealed,' ' uninjuring 
us and propitious, pass beyond.* lie adds ' clad in a skin.' This lull:! him 
to sleep; for while sleeping he injures nobody. Wherefore he says 'clad 
in % skin.'" A derivative of the word Mujnviit occurs also in the K.-V. x. 

84. 1 : 4^*<^c< W^^^fT^ %r^: II " Like a draught of the soma 
produced on MQjiwat, or nniony the Mujavats." Yaskn, Nlr. 9. 8. expl.iius 

the word thus : ^ftaT^<fl" ^^ ^^TWt lJ5f^T\ ^^l 
" ^ Maujavatak'' means produced on Mujavat: Mujavat is a mountain." 

** On the Baiilkas and Diiblikas, tee Lassen, ZeitKh. 1840, p. 194, and 
for 1B39, p. 52, ff. 



Lour. tt. 

the Gandhara, who had expressed an opinion on the natore of 
breath or life; and although his view was not regarded as 
autlioritative, still the very fact of its being quoted, and its 
author mentioned as a Kajanya, proves his Arian origin- This 
is the passage: Sat- -Br. viii 1. 4. 10. y^ ^ W^ 

^T^^^^^f^ ^^ ffS^t^ II ** Further Svarjit, son of 
Nagnajit said. Now Nagnajit was a Gandhara. .... This 
which he said, he sjxake as a mere Rajanya," Nagnajit the 
Uaudbara, is also mentioned in the Ait.-Br. vii. 34, as one 
of the persons who received instruction regarding a parti- 
cular rite from Parvata and Narada.'* He is also mentioned ** 
in the following passage of the Maha-Bh. L 2439 — 41. 

" Nagnajit Subala was the disciple of Prahrada, Owing to the 
wratlj of the gods the offspring bom to him I)ecame the enemies 
of righteousness. Two children were bom to the king of 
Gandhara (Nagnajit 8ubala), i?akuni Saubala, and the mother 
of Dnryudliana, who were both intelligent." Duryodhana was a 
Kuru prince, and one of the heroes of the Mahnbharata. 

These passages are amply sufficient to prove that the Gan- 
dliaras were a people with whom the Arians of India w^ere in 
the habit of holding intercourse, and contracting aflSnities, and 
from this intercourse we may reasonably infer a community of 
origin and language. On this subject Lassen remarks (Zeitsch. 
fiirdie Kunde des Morgenl., iii. 206): "Though in individual 
pas.sagc.s uf the ilahabharata, hatred and contempt are expressed 
in reference to the tribes living on the Indus and its five great 

»' Rotli, Lit. and Hist of ttic Veda, pp. 41, 42. 
•* Sec Weber, Ind. Stml. i. 218—220. 



tributaries, yet there is no trace of these tribes being ever 
regarded as of non-Indian origin. That there was no essential 
difference in their language, is proved, as regards a later period, 
by the testimony of Paniui, which I liave already cited." The 
previous passage here referred to is from the same article, 
p. 194, where it is said : "The word Btihika is used not only 
ia the Mahabharata, but also in Pfinini "* as a general designa- 
tion for the tribes of the Paujab. The use of this appellation 
is thus fully certified; and if the grammariau found it ne- 
cessary to give special rules for forming the names of the 
villages in the Babika country, we may hence conclude that 
the Biihikas spoke Sanskrit, though they applied particular 
affijces differently from the other Indians." 

The same writer elsewhere'* remarks; "The Indians dis- 
tinguish, not expressly, but by implication, the nations dwelling 
between the Sarasvati, and the Hiudu-kusb, info two classes : 
first, those to the eastward of the Indus, and some of those 
immediately to the westward of that river, as the Gandharas, 
are in their estimation still Indians ; . . . . but with the 
exception of the Ka-sbmiras, and some less known races, these 
Indiana are not of the genuine sort: the greater freetlom of 

»* The aphorisms here referred to are ili. 3. 78, and iv. 2. 1 17, 1 18. Tlie 
two latter, with the comments, areas foUowa: '<H I "^^i ^ I ^ H{ ^ || '^'^^ |t 

^r I ^ 4 II *l) ^ I II "117. The affixes tAafi and fii<A are eniplojcd ir. 
wnrds tnkinj; cj-idtUii, which denote villages of tlie Yuliikaa ; aa ^kaliki 
^akaliki. 118. The oflixcs thafi ami nit/> are oplioTially employed in words 
taking criitJhi, which denotu Vahika villages in the country of the Uiinaras; 
OS Suiuim ianUil, Satulariantha ; or with ihcuhhax alBx, SaudaHanii/ii. 
*° Zeitschrill, ii. 58. See also Aaiat. lies. xr. 108 ; and App. note J. 



tlieir customs is regarded as a lawless condition." And Webcr 
similarly remarks'^ : "The north-western tribes retained their 
ancient customs, which the other tribes who migrated to the 
east had at one time shared. The former kept themselves 
free from the influences of the hierarchy and of cast^, which 
arose among the latter as a consequence of their residence 
among people of alien origin (the aborigineii). But the later 
ortlmdox feelings of the more eastern Arians obliterated the 
recollection of their own earlier freedom ; and caused them to 
detent the kiudreil tribes to the westward as renegades, instead 
of Irioking on themselves as men who had abandoned their own 
original institutions.^ 

There are other races also, who, although in the later Sanskrit 
literatuie they are spoken of as being now aliens from the 
Brahmanical communion, are yet declared to have once belonged 
to the Kshattriya caste ; and to hare lost their position in it I 
from neglect of sacred ritea*' (See Part First of this Work, 
pp. 94, and 177 — 183.) In addition to this tradition, however, 
we have yet fui-t!ier proof of the Ariau origin of some at least of 
these tribes. Thus, it appears from the following passage of 
the NLrukta (already quote<l above, p. 161), that the Kambojaa 
spoke an Arian language : Nirukta, ii. 2, TT^JTf^ 3j gj(^t; 

irl^ »TT^r% f^»<i*< T?:^ I inf?T '^fri**?! ^trts^^ 

"^n^^ TT^^^t%gll "Among some (tribes) the original 
forms are used, among others the derivatives. Savuti for the 
•act of going' is used only among the Kambojas, while its 
derivative Sava is used among the Aryas. Duii is employed 
by the eastern people in the sense of • cutting,' while the wonl 



»' Ind. Stud. i. 220. 

•• This tradition is, liowover, erroneouilj extended to some of the eastern 
•nd southern tribes, the Pundras, Odras and Dravidas, who, u we thoU after- 
wards see, could not have been of Arian origin. 



tldtram 'sickle' (only) is used by the men of the north." If, 
therefore, the testimony of Yaska in regard to the language 
used by the Kambujas is to be trusted, it is clear that they 
spoke a Sanskrit dialect. It ia implied in the remarks he has 
made, that a close affinity existed between the languages of the 
Aryas and Kambojas ; that the substance of both was the same, 
though in some respects it was variously modified and applied. 
For it ia only where Buch a general identity existti, that the 
differences existing between any two dialects can excite any 
attention. Had the two languages had but little in common, 
no such comparison of minor variations could have suggested 
itself to the grammarians. Now the country of the Kambojas waa 
situated to the north-west of India, on the other side of the 
Indus. It is clear, therefore, that Sauijkrit was spoken at some 
distance to the west of that river. 

Professor Roth is even of opinion that this passage proves 
Sanskrit g^rammar to have been studied among the Kambr»jas. 
In hia Lit and Hist, of the Veda, p. 67, he observes: "The 
multitude of grammarians whose opinions are cit^l in the 
Pratiiakhyas, proves how widely grammatical studies were 
piu'sued ; and Yaska (Nirukta, ii. 2.) confirms this in a remark- 
able passage, according to which verbal forms were variously 
employed by the grammarians of four different provinces. 
These four tribes were the Kambi")jas antl Aryas, together with 
the Prachyas and Udichyas (or eastern and northern peoples). 
It is thus irrefragably proved that the Kambtyas were originally 
not only an Indian people, but also a people possessed of Indian 
cultiue ; and consequently that in Yaska's time this culture 
extended as far as the Iliudu-kush. At a later i>eriod, as the 
well-known passage iu Mann's Institutes (x. 43.) shows, the 
Kambujas were reck<uictl among the barbarians, because, their 
customs differed from those of the Indians. .... Tlie same 
change of relation has thus, in a smaller degree, taken plact; 

It B 


[cnAT. n. 

between the Karnbnjas and the Indians, as occurred, in a remote 
antiquit}', between tlie latter anil the ancient Persians."'' 

Now, the fact that Sanskrit was spoken by the tribes to the 
west of the Indus, proves that that tract of country was inha- 
bited by races of Arian origin, and of common descent with the 
Indians;'"" and affords au additional argument in support of the 
position that the Indo-Arians immigrated into India from that 

It may, however, perhaps, be objected that this passage not 
only proves that Sanskrit was spoken by the Kamlxijas, to the 

^ In bis later work, the edition of the Nirulta, Rotk mupects for certain 
rcnsons, that so much of the passage before ua as refers to the Kambojas may 
be iiit'CrpuIated. lie adds, however, that " It h tii so far valuable, as it iibow* 
that the ancient Indians iiiingined the Kmnbojns also to be students of Sans- 
krit Grmumar," Erliiut., pp. 17, 18. In the Journal of the G«niua 
Oriental Socictj, vit, 373 — 377, Professor Mlillcr makes some remarka 
on the same passage. He alludes to the fact thut a similar passage occurs 
in the Alahubhushva ; and observes that "though this circumstance ap- 
pears partly to confirm Roth's conjecture regarding the spuriousneac of 
portions of the passage, it may also be possible that the Muhabbnshya has 
borrijwed it Irum the Nirukta, or that both the Kirukta and the MahS- 
bhu»hya may have taken it from the common source of ancient grammatical 
tradition." In any case, this reference to a distant race like the Kambojas, 
looks as if it must have been borrowed from some ancient source. Tbe 
p.nssagc of the Miihiibhrishj-a Is as fljllows, p. 62 of Dr. Ballantyne's edition : 

Tfy\^ ^r4r MT^^ im Tfhi ^^mf^: ^ttt^ iy^' 

3rr^r5 TT'^'^TV^TJ H " •^"p"''. "^ » verb of going, is employed 
imly by the Kambojas; the Aryas use it in the sense of change, for a corpse. 
The Suriish^ras use hammati, the central and eastern tribes ranhati, but the 
Aryns only garni in the sense of 'going.' Ddli occurs among the eastern 
tribes as the verb for 'cutting;' ddtra, a 'sickle,' alone is used by the 
people of the north." 
"» See App. note K, 




DOrth-west, but by the men of the east also. Now, as we may 
presume that Yaaka lived on the banks of the Soi-asvati, or of 
the Yaraiind, or of the GangJi, the people wliom lie designates 
Prachyos, or " men of the eaat," must have been the Kikatas, 
or the Magadlms, or the Angas, or the Vangas. But since it ia 
evident from this passage that these tribes also spoke Sanskrit, 
it might in like manner be argued from this circumstance that 
the Aryas must have penetrated into India from the eastward. 
To this I reply, that we can prove from other passages, such as 
that in the J^atapatha-brahmana, i. 4, i. 10 — 18 (which will be 
quote<l further on), that the Arian civilisation travelled from 
tlie west to the east; and that therefore we may reasonably 
suppose that these Prachya tribes did not originally live in the 
eastern coimtry, but formed part of the population which had 
migrated from the west, or at least did not begin to speak 
Sanskrit till they had learnt it from the Arians coming from the 
west. Ami besides, this passage which I have quoted from 
Yadca does not stand alone ; it is only auxiliary to the other 
arguments which have been adduced to show tliat the Indo- 
Arians came from the north-west. 

This fact, that tribes speaking dialects of Sanskrit, lived to 
the north-west of India, might, it is true, be also explained 
on Mr. Curzon's liypotliesis, that these tribes had eraiffrated 
/rom Imliit. But this hypothesis is opposed, as we have already 
1, p. 304, ff., to the other circumstances of the case. 

The argument then, which I derive from the facts just 
detailed, when briefly stated, is this: We find the north- 
west of India to bo occupied by various tribes, who spoke 
the same language as the Arian Indians. On the other 
hand, we find, (as will be shown at length in the next 
iapter,) that different parts (the eastern and southern 
well as the north-western) of Hindusthan it^self, were 
inhabited by a variety of tribes speaking languages funda- 
mentally distinct firom those of the Arian race. From this I 

D B 3 


draw the conclusion that the Arian Indians must have come 
from without, from the same side which we find to be occupied 
from the earliest period by tribes speaking the same language ; 
and have driven before them to the east and south the non- 
Arian races, to whom, on penetrating into India, they found 
themselves opposed. This subject, however, will be handled at 
length in the following chapter. 

caxr. in.] 






Is the preceding chapter 1 have endeavoured, by a variety of 
arguments derivefl from comparative philology, and from general 
history, as weU as from the most ancient wi-itten records of the 
Indiana and the Iranians, to prove — First, that the dominant 
race, which we find established in Hindusthan at the dawn of 
history was not au^?chthonous, but immigrated into that country 
from Central Asia; and Sej:^ndl>f, that the route by which this 
people penetrated was from the north-west through Kabul, and 
across the Indus. I shall, for the fiiture, assume that both of 
these two propositions have been substiintiiited; and shall pro- 
ceed to trace the history of the Indo-Arian tribes after they had 
eotered the Paajtib, and had commenced their advance to the 
th and east. We have already discovered (see above, pp. 354, 
) from an examination of the oldest Indian records, the hymns 
of the Rig-veda, that the country on both sides of the Indus 
was the earliest seat of the Indo-Ariana in India. We shall now 
see (as has also been already intimated, pp. 288, 289) that, in 
these same hj-mns the ancient bards designated the men of their 
own tribes by the name of Aiyas, and distinguished them ex- 
pressly from another class of people called Daayus, who, we have 
every reason to suppose, were a race of distinct origin from tha 
Aryas, and different from them in colour (see above, p. 284), in 
language, in religion, and in customs, who had been in occupa- 
tion of India before it was entered by the Indo-Arians from the 
north-west. I shall afterwards afiduce various passages from th« 
Brahmanas and post-Vedic writings, illustrative of the progress 

n D 3 




cB*r. in. 

of the Indo-Arians aa they advanced to the east and south, driv- 
ing the indigenous tribes before them into the hill* and foreatjs, 
and taking possession of the territory which the latter had pre- 
viously occupied. I shall suUsequently furnish some illustrations 
of the fundamental differences which exist between the Sanskrit 
and the languages of the south of India — differences which de- 
monstrate tliat the tribes among which the latter dialecta were 
origiually vei'nacular, must have been of a different rac« from 
the Indo-Arians. And finally, I shall indicate the modes in 
which these various classes of facts support the concljision to 
which we have been already led, that the Indo-Arians were not 
autochthonous in India, but immigrated into that country from 
the north-west. 


Bect. I. — Distinction drawn betieeen the Aryat and DatyuM in the Big-Ptda. 

J proceed, then, first, to show that the authors of the Vedic 
hymns made a distinction between the members of their own 
communities, and certain tribes whom they designated a« Da»- 
yiis. This will appear from the following texts. R.-V. L 51. 8, 9 : 

-*l^*4M«f l*[^: II " Distinguish between the Aryyas, and those 
who are Dasyus : chastising those who observe no sacred rites, 
BuVyect them to the sacrificer. Be a strong supporter of him 
who sacrifices. I desire (to celebrate ?) all these thy (deeds) at 
the festivals. Imlra subjects tlie impious to the pious, and de- 
Btroy.s the irreligious by the religious."' X. 86. 19: "V^^H^ 
t^^gr^rST^ t%f%^5t^ ^T¥J?p4^II "Here I come" (says 
Indra) " perceiving and distinguishing the Dasa and the Aryy^" 

' Thia text, oa well as It-V. i. 103. 3, given below, is quoted by Professor 
Miiller, " Languages of the Seat of War," flrst edition, p. 28, note. The 
word tadhamdda, here rendered " festivals," occurs also, K.-V. x. 14. 10. 


•VCT. I.] 



I. 103. 3 : ^ ^T^:*mT ^^^^tfT^ "VTSil ^ f ^'<<9i^< ^ 

f% ^T¥t: I t^T^ ?f^^^^i ^f?»»rei ^4 ^ ^>hn- ^- 

flpj^ll " Armed with the lightning, and trusting in his strength, 
he (Indra) moved about shattering the cities of the Dasyus. 
Indra, thunderer, who art wise, hurl thy shaft against the Dasyu, 
and increase the might and glory of tlie Aryya." I. 117. 21 : 

«<c5<Ull yfVf^m ^3^ ^f^^y*^<l*4l*4ll "O beautiful 
A«wins, sowing barley with the plough, drawing forth {lit. milk- 
ing) food for man, and sweeping [or blowing] away the Dasyu 
•with the thunderbolt; ye have created a great light for the 
Aryya."* I. 130. 8 : T^: ^TJ^ ^^*fM*J^ ^fV^ TTT^ 

f^^ ■airmfFnjf^ ^nSf^^ ^ifaitji ij^k iim^ "^rsr- 

cfT*! <«r^ SEWTT^^^J^Iff II " Indra, who in a hundred ways 
protects in all battles^ in heaven-conferring battles, has preserved 
in the fray the sacrificing Arj'ya. Chastising the neglectors of 
religious rites, he subjected the black skin to Manu " (or the Ar- 
ian man).' III. 34. 8, 9: 4j«JH ^: ifmff ^l*<<1*il^ 

' {?Sjana interprets the " great liglit," either of the glory acquireJ by the 
A^wins : *st<^M ^^ ^TT^T^ ^^W^'. I or of the sun : ^^4 

4^l4l^ «^Tf^: l( " F'X" '' '» tl'e living man who beholds the sun :" 

5lj^<4»l f% '4^4^ H«^f^ II Roth considers this verse to refer to some 
forgotten legend. Nirukt«. Evlaiit. p. 92. The two followinj» passages also 
similarly speak of light i B.-V. ix. 92. 5 ; ^^if^ T^T ^^ ''S^I^T 

'^T^ TfT^ ^ ^^^ ^ ^SH^^T^II "When he (Soma) 
gave light to the day and uflbrJed space, he dtliverod Manu, nml arrested 
the Dasyu." R.-V. x. 43. 4 : . . . f^^rf^ ^^ Jpf^ ^Vii^l- 
ir?? II "He (Indrn) gave to Monu blessedness (and) n glorious light." 

* This piwsage is translated in a review of the First Part of this work, con- 
tained in the " Tiinea " of 12th April, 1858. The " black skin," is there ia. 



[cuAr. tw 


^^•T 1J ^srn^ t^l|f^ | 4f( ^ II " The wise gladden Indra, vho 
bestowed the earth and this tirmament. Indra gave lioi«e8, he 
gave the sun, he gave the many-nourishing cow, he gave 
golden wealth. Slayiiig the Dasyu, he protected the Aryyan M 
colour." IV. 26. 1, 2: ^ TRT*^ ^7^ ^ ^T^fW 

H^ qi?frTT mil'i;!! ^ ^jf^m^TT^IP^T^ ^ ^t TT- 

^rl'TI^«t,4l 5^ II ** I," says Indra, " was Manu, and I the sun; 

I aiu the wise rishi Kakshivat. I auhdue Kiitaa the son of Ar- 
juiii. I am the sage Usanas : behold me. I gave the earth to 
the Aryya, and rain to the sacrificer. I have led the sounding 
waters. The gods have followed my will."* IV. 30. 18 : The 


terprcted of the dark colour of the Dasyus. The next passage is aim putlj 
quoted in the same article. 

■* Suyona connects the wor<l iirya ns an epithet with Manu understood. 
Professor Weber, Ind. Stud. I. 195, note, thinks that Monu means in this 
ge, the moon. (In pp. 194, 5, he has a disscrtotion on the word Manu). 
The speaker in these verses appears to be Indra. (See Biithl, and Roth's 
Diciionary, mb voce, Unanas.) The Atiukrantnnl, as quoted bj SSjrana, 

■^1<W M*t^H " Jn 'lie first three verses the i-ishi celebrates bimself ■ 
as if under the character of Indra : or Indra celebrates himself." Kohn 
(Ilerabkunfl dcs feuers, p. 143,) conjectures that VamAdeTa maj per- 
haps have been an ancient epithet of Indra. In R,-V. x. 48. I, lodra M 
says, Bimilarly: ^ i^\l^ f^V^^^:f^ ^frsR^I ** I distri- 
bute food to the sacriCccr," &c. The pantheistic author of the Vrihad _ 
Aranyakn Upanishad, thinks that the Rishi Vamadeva is speaking of him> I 
self in these wonU, (Bibliolheca Indies, pp. 215,216): ff^ ^ ^ 



Sanskrit text of the following is given above p. 361 : " Thou, O 
Indra, hast speedily slain those two Aryas, Arna and Chitraratha, 
on the opposite bank of the Sani}!!" (river). VI. 25. 2, 3: 

^^m fw^ wf% ^ -^^mrf^ w^f% ^rr^in "By 

these (succours) subdue to the Aryya all the hostile Dnsa 
people everywhere. Indra, whether it be kinsmen or strangers 
who have approached and injuriously asaailefi us, do thou destroy 
their feeble power, and put tliern to flight" VI. 33. 3: 7^ 

^VlR rti I f4^ '! " Do thou, heroic ludra, destioy both these our 
foes, (our) Dasaand our Arj-ja enemies," &c. VI. 60. 6 : ^^ 

** Do ye, lords nf the virtuous, slay oiu- Aryya enemies, sluy 
our Dasa enemies, destroy all those who hate us." ^^I. 83. 1 : 

S^THT II "Slay both the Daaa enemies and the Aryya; pro- 
tect Sudas with your succour, Indra and Varuna." X. 38. 3 : 

" U much lauded Indra, whatever ungodly person, Diiiai or 
Arjya, de.'^igufi to fight against us; let these enemies be easily 
sulxlued by us. May we destroy them in the conflict." X. 
49. 3 : ^ 15^»!r^ -^m Wfi^ T ^ T^ ^^-4 Tm 
^^^ II " I, the slaj'er of Sushna, have restrained the holt, — I 

'U tj^yM II " AVTiosoerer of guils, r isliis, or inen, niiilersloiMl TA"t, ho be- 
came Th/it. Perceiving tliis, the Rishi Vuinadeva obtninctl this text, • 1 was 
Miuiu, I the Bun, &c.*' Usatuu is coonCL-tucl with lodm iu R.-V. vi. 20. 11. 




CHAP. 111. 

who do not abandon the Aryyan name to the Datyu." X- 65. 

S'lijfV ■^f'l II " These bountiful ones " (the gods named in the 
precetliug verse) " have generated prayer, the cow, the horse, 
plants, trees, the earth, the moimtains, the waters; — caasiag 
the sun to aacend the sky, and spreading Aryyan rites over the 
earth."* X. 83. 1 : WHUTfi i,\MMl*M Wm ^WT ^ W%- 

^r)«( W^Wl M-i^rM || " May we " (0 Manyu), " associated 
with thee, the mighty one, overcome both Dasa and Aryya 
through (thy) effectual energy." X. 102. 3 : ^ST^ 4^ fsf^- 

^e|<Jl IV'^^II "Restrain, O Indra, the bolt of the mxu-derous ; 
remove far away the weapon of our assailant, be lie Dasa or 
Aryya." X. 138. 3: f% ^^^ »r^S^^ ^^ f^^ <mi<| 
Mfd^l*l^[ ^* II ** The sun has launched hia car in midheaven : 
the Aryya has paid back arecompense to the Dasyn.'' VIII. 24. 27: 

<{m<j< ^f^^'^^ 'ffTO'l "Wlio delivered [us] from the 
destroyer, from calamity ; wlio, powerful [god], didst avert 
the bolt of the Dasa from the Arya in [the land of] the seven 

The above cited texts seem to leave no doubt that the Rig- 
veda recognises a distinction between the tribe to whiclj the 
authors of the hymns belonged, and a hostile people who ob- 
eervefl different rites, and were regarded with contempt and 
hatred by the superior race. This appears from the constant 

* Compare R.-V. vii. 99. 4 : Urum yajhdyu chakrathur u loham jtatayatitd 

surynm tishaxa magnim. Dd$asya chid vrishaiiprati/a tnCiya jaghnalhur nara 
prUandjyefhu : " Yu, (Indra and Vihlinu,) Iiave provided abundant room for 
the sacrifice, creating the sun, the dnwn, and fire. Ye, heroes, have de- 
stroyed the illusions of the bull-nosed Diss," 


SECT. I.] 



antithetic juxtaposition of the two names Aryya and Da^ju, in 
mofltt of these texts; and from the specification in others of 
enemies, both Aiyya and Da83Ti. If human enemies are desig- 
nated in these latter texts by the word Aryya, we may reasonably 
suppose the same class of foes to be commonly denoted by the word 
Dasifii. It is not, of course, to be expected that we should find 
the Indian commentators confirming this view of the matter more 
than partially ; as they had never dreamt of the modern critical 
view of the origin of the Aryyas and their relation to the barbarous 
aboriginal tribes. Yaska (Nirukta vi. 26), explains the terra 
Aryya, by the words " son of a [or, of the] lord."" The word 
DnsyiL is interpreted by him etymological ly, thus : " Dasyu 
comes from the root dae, to destroy ; in him moistiue is con- 
sumed, and he destroys (religious) ceremonies."^ 

Sayana interprets the word Aryya, by " wise performers of 
rites*;" "wise worshippers';" "wise'";" "one to whom all 
sliould resort" ;" " the most excellent race [coloiu-] consisting of 
the three [highest] castes'*;" " practising ceremonies'^;" "most 
excellent through performance of ceremonies'*;" and in two 
places, i. 117. 21, and iv. 26. 2, he regards it as an epithet of 
Manu. The same commentator interprets the word Dasyu of 

• Nir, vi. 26 : ^Sn^ t?5TT^' H 

• f^^t^: ^?nT:i on 1. 103.3. 

"* fW^I oni. 117.21. 

^<Un«J ^^^fRH^ I 0" '■ '30. 8. 

" **5*jmif«n ouvi.22. 10. 

*i^H8ra"^T ^^rf^l onYi.83. 3. 


the "robber Vrittra'*;" " enemies who destroy the observers of 
Vedic rites'* ;" " the Aeuras, Pisachas, &c., who destroy "' ;" " the 
vexmg Asiiraa'* ;" "all the people who destroy religious rites";" 
" Vala and the other Asiiras who destroy religious rites ***;" "ene- 
mies devoid of religious worship."*' From these quotations it 
will be seen that Sayana mostly understands the Dasyos of su- 
perhuman beings, demons, or Titans, rather than of human ene- 
mies. In hie note on i. 100. 8. he speaks of them as " destroy- 
ing enemies living on the earth'*;" and in another place he 
explains the Dasa vama, as being either " the Sudras and other 
inferior tribes, or the vile destroying Asura."*' 

There ia no doubt, that in many passages of the R.-V., to 
which I shall presently refer, the words Dasyu and Ddsa are 
applied to demons of different orders, or goblins {AauraSj Hdkeha- 
808, &c.) ; but it is tolerably erident from the nature of the 
case, that in all, or most of the texts which have been hitherto 
adduced, we are to imderstand the barbarous aboriginal tribes of 
India as intended by these terms. This is yet more clearly ee- 
tablished by the sense in which the word Dasyu is used (i. p. fur 
men and not for demons) in the Aitareya-brahmana, in ^lauu, 

" ^ T^l on i. 33. 4. 

" ^HBld lllT*ri^Mf4iflK: "ST'^WM oni. 31.8; and i. 103. 3. 

" ^M^*l*lRy.l*Hy< fMl{Mir^**i^l oni. 117.21. 

" ^nr^rpriTT^ oniii.34.9. 

•» *45uii4jy^Mr«i-l^f^'^: ^ f^: ttwt:! onvi.25.2. 

" **5^1«ii: iT"^: I '« >•'• 60- 6. 

i. 100. 8. 





and in the Mahabharata, Thus the author of the Aitareya-brah- 
mana, after making Visvamitra say to his fifty sons : " * Your de- 
scendants shall possess the extremities [of the land],'" adds, 
" These are the Andhras, Pundras, Sabaras, Pulindas, Mutibas, 
and other numerous frontier tribes. Most of the Dasyus are 
deflcended from Vi^vamitra,"'^* And in the authoritative defini- 
tion already quoted (Part I. p. 178 — 180) Manu tells us: 
•* Those tribes in the world which are without the pale of the 
castes sprung from the mouth, arms, thighs, and feet [of Brahma], 
whether they speak the language of the Mlechhas, or of the Aryyas, 
are all called Dasyus.^ '" 

The Mahabharata thus speaks of the same people : iL 26. 
1025: ftr^ ^fv f^faJrtj ^^ w4d«ilf^:i 1WT\ 

^S^g^^^ffffl^ ^Sl^H^ W^ m^mn "Having vanquishetl 
the Paiirava in battle, the Pandava conquered the Utaavasan- 
ketas, seven tribes of Dasyus inhabiting the moimtains." And 
again: Ibid, ver. 1031 — 2: ^|^'^1«( ^Hf ^|*-^-S^^5T«|d 

xnmrref'T:! «i*iTi<i f^ v "^ -^Hrtjif^irfj "^^1^:1 

^4> I ^41 ^1*1 ^f^^i'Mtl <.l*tfMlt " Faka^asani conquered the 
Dar:ulas, with the Kanibojas, and the Dasyus who dwell in the 
north-east region, as well as: all the inhabitanta of the forest, \vith 
the Lohas, the Parama-Kambojas, and the northern fiishikas." 
And once more : " Saineya (Krishna's charioteer) made the 
iutiful earth a mass of mud with the flesh and blood of thou- 
ids of Kambojtis, ^akas, Savaras, Iviratas, Varvaras, dei^troying 
thy host The earth was covered with the hebnets and shaven 
and bearded heads of the Dasyus, as with birds bereft of their 
wings." ** Here it is evident that the word Dasyu, in the latter 

** Aitareya-briihniBna vii. 18 (already quoted in Port First, p. 84). 
•» See above p. 160. 

*« M.-Bh. Drona Parvo. Sect. 119, ver. 4747, ff. : lOrcady quoted, Puri 
First, p. 179. 


[caAP. m. 

part of the sentence, is a generic term denoting the whole of 
the tribes who had been previously mentioned, the Kauibojaa, 
Sakas, &c. 

Another passage occurs in the Santi Parva of the ]M.-Bb. sect. 
65, VBTBes 2429, ff., (quoted in the First Part of this work, p. 
179), where the tribes there enumerated are said to live after 
the fashion of the Dctsyus; ami where the duties to be observed 
by the Dasyiis are described. The Dasyiis therefore cannot 
have been regarded by the author of the M.-Bh. as demons. 

If any further illustratiiiu of this point be re<^uired, it may be 
found in the folloNving story (from tlie Mahdbharata, Santi P« 
sect. 168. verses 6293, ff.) about the sage Gautama living 
among the Dasyue : '^^ ^cji^j | -^^ ?( ^'rlf^QiS'ffVrfTTTr^ 

f<trr»lt ^^Hn Iff: ^r^^U Bhlshma is the speaker : " I 
will tell thee an ancient story about what happened in the northern 
region among the Mlechh.ns. A certain Brahman of the central 
country, perceiving a particular village, which was destitute of 
Brahmans, to be in a prosperous condition, entered it to solicit 
alms. Tliere lived there a wealthy Dasyn, who was ac<]ttaint4yl 
with the distinctions of all the castes, religious, truthful, and 
liberal. Approaching his house, the Brahman asked alms," and 

a house " From proximity with the Dasyus. Gautama 

[the Brahman iu t[U€f^tiou] became like them. While he was 
thus dwelling happily iu a village of Dasyus," another Brahman 


net, 1.] 




arrived, who demandeei of him : " What is thia that thou art 
foolishly doiDg ? Thou art a Brahman of good family, well-knowii 
in the ceutral region : how is it that thou hast Bunk into the 
condition of a Dasyu ?" 

From the evidence aflbrded by these passages of Manu and 
the Mahabharata, it is probable that the word Dasyu, when 
occurring in the Veda, is frequently to be understood of men, 
and, consequently, of the wild aboriginal tribes, whom the Arion- 
Indians encountered on their occupation of Hindusthan, It is 
true that, by the later authorities whom I have quoted, the 
*psyu8 are regarded as degraded Ariaus, " (though Manu 
s that some of them spoke Mlechha dialects), and that tribes 
unquestionably Arian, as the Kambojas, (see above, p. 368, ff.) 
are included among them. But though it is true that some of 
the Ariau triljes who hail not adopte<l Brabmanicol institutions 
were so designated in after times, the term Dasyu could not well 
have been so applied in the earlier Vedic era. At that time the 
BrahraauicAl institutions had not arrived at maturity; and the 
tribes who are stigmatised by the Vedic poets as persons of a 
different reUgion must therefore, probably, have been such as 
had never Ijefore been brought into contact with the Arians, and 
were in fact, of an origin totally distinct. 

It thus appears, that by the Dasyus who are mentioned in the 
Uig-veda we must, in many passages, though not in all, under- 
stand the barbarous aboriginal tribes with whom the Aryyas, on 
their settlement in the north-west of India, were brought into con- 
tact and conflict. Before we proceed further, however, it will be 
interesting to review some of the other principal texts of the R.-V. 
in which the Aryyas and Dasyus are mentioned- I should be glad, 
if, while doing so, I could hope to arrive in each case at a definite 
result in regard to the application maile of the words Dasi^u 
and Ddsa, and to detcrukine precisely the relations which sub- 
sisted between the tril>es so designated, and the Aryyas. But 

*' Sec LasMi), ZiiUchrKV, ii. 49, (T. 


>IA : 



the sense of the texts is often so obscure, that I cannot alvraja 
expect to fix their interpretation, or, consequently, to deduce 
from them any certain cou elusions As, however, I have collected 
and arnuiged a considerable number of passages bearing on this 
subject, I think it best to present them to the reader, with auch 
illustrations as I am able to supply, in the hope that a fuller 
elucidation may sooner or later be supplied by the mature re- 
searches of some more competent scholar. 


Sect. IL — Additional Vedic texts bearing on the relationt of the Aryai 


First : In the following passages, or some of them, reference 
appears to be matle to the earth or territory lK?ing lieatnwcd on 
the Aryyas: i. 100. 18: ^^5^^ f^P^"^ ^T^fT V^ "^^^T 

^9^\\ ^^ H'ii,^: ^^:ll "(I«Jj"a) the much invoked, 
having smitten on the ciu-th the Dasyus and Simyus [or 
destroyers] by his flying hosts, destroyed them with his thun- 
derbolt. The thunderer bestowed on, [or with] his white 
friends the fields, bestowed the sun, bestowed the waters."* 

" Several points are obscure in this passage. Is the word ^iinju U>e name 
of a tribe, (03 Professor Wilson renders it,) or does it merely mean a «le- 
slroyer? In R.-V., vii. 18. 5, we huve the words inrdhantam iiinyum, which 
I'rofeaaor Uotli (Lit and Ilist. of the Veda, p. 94) renders by " defiant 
wrong-doer." Sayana explains the word 3[J^{J(J«J aTV^IlT'jT TT" 
"^H I ^•T II "Subdiiers i. e. slajers, Raksb:isa.«, &c." ; aiiJ again: 

WTl«r' Ht*^' tl " The verb iam designates one who contemns every ona ^ 
else, ^imyn therefore = Rak»ha$a, &c." Then, wlio are the "white | 
friends " of Indra, in the second clause Y The Maruts P or the fair complex* 
ioned Aryyns ? In verse 2 of this hymn, we find the words lahhibhih geelhir 
evaih, "his flying friends," which Saynna interprets of the ilaruts. He 



ir. 20. 7: ^ -^^TfT T^: ^f^O^T^: Jj<*^0 <i41<< - 

^ f% I ^3H4I<^ f^ ^mM^rillf^ I " Indra, the slayer 
of Vritra, and destroyer of cities, scattered the servile (hogts) of 
black descent. He created the earth and waters for Manu."" 

explains verse 18 thus: f^'^RTt 


^414il '^fl It "Along willi Lis wliiJe-coloureJ, (i- «. whose limbs were 
shining with ornaments,) friontls, ihe Maruls, he divided the territory be- 
longing to his enemies." On the other hand, we have, in verse 6 of ihis 
hjcnn, the worabi])pcrs themselves spoken of us tlie persona with whom the 

sun was shared. The words there are : ^Sf^T^ff>T ^f*?! i^<( 4J«1ri I 
which Sajana renders : ^<^<f)t| ^^l ^^l -^if T^;^Wni ^'TT^ 

" Let him divide the light of the sun with our men, und involve our enemies 
in darkness which shnll obstruct their view." The same words are rendered 
by Rosen : Nottratihwi viri» solem eoncedal, " Let him beafow the sun on our 
countrymen," where the words in the instrumental cnse have the sense of the 
dative assigned to them. If they bear that sense in verse (5, they may equnlly 
Lave it in the 18th also. The ueanin>,' would then be, ** tie bestowed the 
land, the sun, the waters, on his fair friends." On the last words, ProA-ssor 
Wilson remarks, Rig-ve<la, i. p. 260, note : " These, according to the scholiast, 
are the winds, or Maruls; but why they should have a share of the enemy's 
country {ilalrunam hhiimim) seems doubtful. Allusion is more jirnbably in- 
tended to earthly friends or worshippers of Iitdra, who were white (iwitni/a) 
in comparison with the darker tribes of the conijuercd country." The wor- 
shipper's friendship with Irvdra is mentioned in many passages of the R.-V., 
as, i. 101. 1 ; iv. 16. 10; vi. 18. 5; vi. 21. 5 and 8; vi. 45.7. Rosen renderi 
the poAsage : Erpugimvit terram tociis tui* nitentibut" " He conquered the 
earth for, or with, his struggling companions;" thus giving another sense to 

fH|9iJHT^ I I" ^^o other hymns, vii. 99. 3, and x. 6o. II, (quoted abovo 
p. 377, 378), we find mention made of the sun in a somewhat similar man* 
nfr, as in the verse under review. 

Sayann explains the words krithwyoiAk, &c., thus : hfUhnatfShir nihrink- 
tfiitir, dusir upakshapayatrir aturih tenuk, " the destructive armies of the 
Liunis, of degraded rank." 

C C 




ca&r. »t. 

The passages iii. 34. 9, and iv. 26. 2, which have been already 
quoted above (p. 375, 6), should be again referred to here. VI. 

18. 3: 1^ ^ ^ m^ •*i<*<<J^ <4^*: ftyi<«j«ri<i<uji«(i 

"Thou" (Iiidra) "hast alone subdued the Dtisyns : thou ha^ 
given petiple tu the Aryya.""* VI. 61.3: "^ff f^t^WTS^^- 
7^f^«<^ : I « And thou (Sarasvati) hafst obtained lands for men."" 
VII. 19. 3: qV^f?^ ^«i<<^*lM: ^^«JMI "T^^^ ^- 
^7f i "Thou hast preserved Traaadaayu, «on of Purukut«a, 
and Puru, in fi<^hts fur the acquisition of land." VII. 100. 4: 

Vishnu traversed this earth, to give it for a domain to Mano." 
It is possible that in these passages, or in some of them, aJlosion 
may be mado to the occupution of the plains of India, and the 
subjugation of the aboriginal tribes by the Aryyas, on their im- 
migration from the north-west ; but it must be confessed, that 
the explanation ia uncertain. In R.-V. x, 65. 11, quoted above 
(p. 378), there maj be a reference to the spread of Aryan insti- 

Second: In two of the passages already quoted (i. 51. 8, 9; 
i. 130. 8), the epithets dvrdtti and apavrata, " devoid of, or op- 
posed to, religious rites," will have been noticed aa applied to the 
Dasyus. I proceed to cite some further passages in which the 
character and condition of the Dasyus (whoever they may be) 
are specified. 

They are (!.") described a.s a degraded race : i. 101. 5 : 'ITST^ 

^ T^Tv^t '^■=tifd<^ »r^^ m^\n ^^T^i "We 

invoke to be our friend, Indra, attended by the Maruts ; who de- 

* Sajana explains Mfithtili, " people," by pulradutSdin, " children, sUvGa," 

" Sfiyana explains nvanih, by Atwrair apahritii bhumiJf, " lands taken twnjr 
hj the Asiiras." Ruth (Diet.) assigns also to the word the sense of " streams ;" 
which it might seem to be the function of Sarasvati to give rather than 



stroyed the base Dasytis."" 11. 11. 18: tv^T IR! "JIX ^ 

(Jctjrl* ^if^ ^^f^^l "Maintain, hero, that strength by 
which thou hast broken down Vrittra, Danu, Aurnavabha- Thou 
haat revealed light to the Aryya, and the Dasyu has been set on 
thy left hand."'» The text of the following, R.-V. ii. 12. 4, 
has been already given in p. 284 : " He who swept away the low 
Dasa colour."" IV. 28. 4 : fin j ^ ti lft^ Ft^nmif^^ T^^^ 
f^lft 4l*nf*^<.M1^i: I " Indra, thou hast made these Dae- 
yua lower than all, and the servile people without renown." 
They are described (2.) as having either no religious worship, or 
rites different from those of the Aryyas. I. 33. 4, 5 : ^{JsnfMI 

WT^: vfw^fr^ri tnr f%*\ ^ft^ {^^-^ T^ Vii^\^ 

V^f^l m^m^X'. I *' Tlie unsacrificing Sanakas perished. 
Contending witli the sacrificers, the non-sacrificers fled, Imb-a, 
with averted faces."'" I. 131. 4: ITTW^rf'T^ T^JHTSaf^raTT- 
f^ I " Thou, Indra, hast chaatised the mortal who sacrifices 

** Sayana cxpluins this of making the Asnras vile und slaying them : 
Amrun adhardti nikritkiuH kfitea ; but the words will also bear the sense I 
have put upon them. 

" Sajana ezplttios the word Dasyu in this verse of the mythical personage 
Yf iltra. The words ni -f 'u'/i, making together niskuiii, may have suggested, 
OP have been sug^rested by, the word NUhiula. 

•* Roth (Diet.) gives the sense of " removing, putting away," to gvh& 
AoA. Sayana explains it of hiding in a carem. The word varna, colour, 
race, whif^h ia applied to the Aryyas, iii, 34. 9, is here made use of in speak innr 
of the Dasyoa. Siiyana explains the latter, either of the ^udra caste, or of 
the Asuras. 

** Siiyana describes the Sanakas as followers of Vrittra: eUmnamaki 
Vfittramtcharuh. I cannot say who may be meant by the Sanakas here. 
They may have been heretical Aryya* and not Daiyu*. Sanaka was a mind- 
born son of BrahmiL Wilson, Vish. Pur. p. 38, note 13. Weber, Iiid. Stud, 
i. SfiS, note, quotes a text of the M.-Bh. xii. 13,078, where be is mentioned 
•8 a sage. 

c c a 


not."" 1.132.4: ^'^^\ <*-W*<[ ^y<*<H*<^ ^U J Kjni 
f^^^TTfTliT I " Subject txj those who offer libations the irreligious 
[maij], the wmthful irreligious [man]." IV. 16, 9 : f^ fnVT- 
Wr»T ^Sra^ 4<!H<t1 I " Tlie deceitful, priestless Dfusyu has 
perished.' y. 42. U : ^mWrTT\ ^(^^ ^TSVTfT^^ f^' 
f^^: "^T^TT ^«4<|4(4I "Remove from the sun the irreligious, 
the haters of the priest, [or of sacred rites,] who increase in pro- 
geny." VIII. 59. 10, 11 : ^ ^ ^7^ ^7T^?Wrf%fV f^ 

^^?r?7!i;^^ ^«M«*<^ ^ra^sjT'i^ "<j<c(^^i '^r ^: ^r^ 

^V^t<T ^^'. ^rrn? ^^ ^TT: ll " Thou, Indra, lovest our 
religious rites; thou satiatest those that revile thee; thou, most 
excellent and vigorous hero, hast smitteu the Dosa in the centre 

*• In 5. 100. 15, the word inarttn^, "men," is opposed to deeah, "godi." 
Tlie word martya, " morlui," is usually applied to men. But from the fol- 
lowing pasMge of the ^tapathu-bruLuana it appears that the Ajuras alio 
arc roj^urded as mortQl, and that the gixls too were formerly so. II. 2, 2. 8, 
(V. : DeViUchfi r<'t Aturiiiichii ubhai/e prujupatyuh panpridhire. TV ubhat/e eva 
tiHi'itiutiiui tiMuh, vtartiju hyatuh : anatriiti hi martyah. Teshu ubhayetku mar- 
tytshu Agnir cva amrita asa. Tarn ha sma ubhaye anifUam upajieaHti . . . . 
Tuto tlrvwilitHii/6it»a tea pariiiiMire. Te orchantah irainyantai cheruh, Ubt 
Aaun'tn tapainun marUji'in abhihhovema iti te tUvi awp(am ugmjiXdhcyam da- 
driiuh. Tf ha uvhvh : hantu id>tm amfitam antarutiimnn iidadhanuihai, TV 
idam amritnm anUtralmaiia udh/iya amrita bhutra astaryya hhutca ttaryan aa- 
patniin murtyun ubhibhiivithyumu iti. " The gods and Asuros, both the ofl* 
sprliij,' of Prujiipati, gtrove togetlier. They were both soul-les^, for tliey 
were mortal ; for he who is aout-ltaa is mortal. While they were both inor- 
tnl, Agna alone was imniortitl ; and they both derived life froni him, the im< 
mortnl .... Then the gods were left as the inferior. Thej continued to 
practise devotion and austerity, and (while seeking to) overcome their foe?, 
the mortal Asuras, they beheld thia inimortfd consecrated fire. Thej then 
said 'Come let us place this immortal (fire) in our inmost soul. Having 
done so, and having [thus] become immortal and invincible, we shall over- 
come our mortal and conijuerable enemies.' " The gods accordingly placed 
the sacred fire in their hearts, and by this means overcame the Asuras. 





of his thighu. May our friend Parvata, may Parvata with a vi- 
gorous stroke, strike down from heaven [?] the Dasyri who ob- 
serves different rites, who i.s inhuman, who does not sacrifice, nor 
regard the gods."^^ IX. 41. 2: ^Jftrfjiit T^HT^tn? I "Sub- 
duing the irreligious Dasyu."'* X. 22. 7, 8 : '^ ^ ^^ T^T 

\f X ^ ^ '■J nJ 

«^ fTWrfif^T'l, W^TO^ TTH^H " Thou, Indra, receiv- 
est [?] our iiplift<jd pmyer. ^\'e implore of thee that succour 
whereby thou didst strike the inhuman Sushna. The Dasyu, 
irreligious, fooliah, observing other rites, and inhuman is against 
U.S : do thou, slayer of our foe-s, subdue the strength of this 

Another epithet which is frequently applied to the adversaries 
of the Vedic bards, or of their deities (whether those adversaries 
may have been Aryas, Daayiis, or demonsi), Ls Anindra, " des- 
pisers of Indra." It occurs in the follnwing texts : R-V. i. 133, 

^T; I " By sacrifice I purify both worlds, I consume the hostile 
realms which regard not Indra." R-V. iv. 23. 7 : T^ fal'^ltJ*^ 
^^T^r^ "^it^^ ^f^ frPWr fj^ ^-fl*! I " Seeking to slay 
the injurious anil destructive [Rakshasi ?] who regards not Indra, 

'^ The epithets of the Dn«yu in the lait verse seem to be those of a mor- 
though the mention of huarcn may seem to point to an aerial foe. In 
se 7, of the same hymn a luortnl enemy is referred to : na ilm adeva apad 
\am (tirghiiyo martyah : " O long-lived god, let not a godless mortal obtain 
prosperity." In his comment on R-V. v. 20. 2 : Suyana explains the word 
anyuvrata thus : caidikml vibhahtain rraiam kurma yatya tasyn Aturnsya, " the 
Aaura whose rites are diOercnt from those of the Veda." See Guldstiieker, 
Diet, smb eoee ^^ anyavrala." 

** Benfey in a note to his translation of Sama-vedo, ii. 243 (p. 251), un- 
derstands DaxjfKm avratOM of Vfittra, or the Evil Spirit in general. 

c c 3 

he makes his sharp weap^^os sharper for her deetniction." B.-V. 
v,2.3: rT{<i^4nf Ijf^^^ "^KM^ i^T^-qia^ ^T^ 

<4^<14^T! I " From an adjacent spot, while offering to him an 
imperishable, widely diffused [? oblation], I beheld [Agni] the 
golden-toothed, the bright-coloured, fashioning his weapons: 
what can those who regard not Indra, and recite no hymns, do 
to me?" R.-V. vii. 18. 16 : ^ ft^^ ' UdMIH . ^f^Ti' ^^ 
ira*^ '\*i\ ^sfif '^\*i I " Indra hurled to the ground the 
half of the struggling heroes, drinkers of the oblation, and dis- 
regarders of Indra."'* In R.-V. x. 27. 6, Indra says : ^]{«f il 

v^r 34riMi«t^ '^'!n"*l. wry^: Tn::^ ^FJimTT^i -^ ^ 

^ f^f^: H<^l«<H. ''tV ^g TTS ^?^ fk^:\ " I behold 
here those who drink the oblation, aud regard not Indra, who 
are strong of arm, and grasp at the thunderbolt [?] : may the 
thunderbolts fall on those who revile the energetic friend." **• In 
E.-V. X, 48. 7; Indra again speaks: ^!r^4*i^ y,<**<<*T ^V^W 

H^ ftf TT fH"^f^ II'^sfT "^if^^K \ •' I alone vanquish this 
one enemy ; I vanquish two ; what can even three do ? [In V>attle] 
I destioy numerous foes like sheaves of com [?] on the threshing- 
floor. ^^^y do the enemies who regard not Indra revile me ? '^'^ 
The following text speaks of men who are destitute of hymns 
and priests: x. 105. 8 : ^Sj^ ^ ^CfsfWT fspftf^ "^^ ^^ 

"^[^'A SfTflUT ^irST "^VT^ ^Y^fW ^1 "Take away 


'^' See Ruth's interpretation of this verse in his Lit. and Hist, of the Veda, 
pp. 08, 99. Some at least of the tribes who are here designated as amindrm, 
appear from Roth's account to hare been of Arion descent. See Ibid., 
pp. 132-135. 

** See Benfey's Glossary to S.-V. p. 121, sub voce '^paoi." 

*' See Nirukta, iii. 10, and Roth, Erlaiiler., p. 39. 


r. II.] 



edamities. By sacred verses may we overcome those who em- 
ploy no holy hymns. Thou takest no great pleasure la a sacri- 
fice without a priest [or without prayers]." 

Aa we have seen above, (p. 284,) allosioa appears to be made 
in the Veda (3.) to tiie distinction of complexion which existed 
between the Aryas and the aborigines. On this fiubject I quote 
the following remarks, made liy an able writer, (whom it is not 
difficult to identify,) in a review of the First Part ttf this work, 
which appeared in the "Times" newspaper of 10th and 12th 
April, 1858. "At the time," he says, "when this name of 
* varna ' was first used in the sense of caste, there were but two 
. castes, the Aryos and the uon-Aryas, the bright and the dork 
race. This dark race is sometimes called by the poets of the 
Veda ' the lihick skin.' Rig-veda, i. 1 30, 8 : * Indra protect*H.l 
in battle the Aryan worshipper, he subdued the lawless for 
jVlanu, he conquered the bhick skin.'""'* (This passage has been 
already quoted, p. 375.) 

• This phrase " the black skin," occurs also in R.-V. iz. 41. 1, a text 
which reappears in the Sama-veda, i. 491, and iL 242. The words are Pra 
ye gavo na hhurnayat tatthii aydio akramuh, ghnaiUah hri$hnam apa tra- 
eham; which are thus rendered by Professor Benfuy : "The lluuiing, the 
teinpc8tuoua [gods] approach like furious bulls, and chase away the bltick 
tkin." In n note he adds the explanation : " The Maruts (winds) chase the 
clouds." In his glossary to the S.-V., the same author explains the phrase 
" black skin," bj "night." A similiar expression, Iracham anikmm, occurs 
in R.-V. \x. 73. S : Indradvishlfim apa dhamanti muyaya iracham airikniin 
Mumamd divas ptri. ** By their skill they [I cannot say who are here re- 
ferred to] sweep away from the sky the black tkin of the earth, hated of in- 
dra." So Benfey translates this line in his glossary, tub voce " asikni." But 
possibly the words " black skin " should not be construed with the word 
♦• earth." 

I omitted to state, when R.-Y. i. 130. 8. was first quoted in p. 375. 
that it is ascribed to the ^ishi Panichbepa ; and that the compositions 
of this poet are distinguished by the repetition of some of the preceding 
words at the close of the lines. This peculiarity is thus noticed in the Ni- 
rukta, X. 42. Abhyiite bhUyansam artham tnanyante yathti, " aho darfatiiya, 
aho darianiya." Tat PaructkhepOitgO ftioM : Paruehhepa rithih, " Some con- 
sider that greater force is added to a sentiment by repetition, as tn the ex- 

c c 4 

392 THE ABIAXS IN INDIA : [cur. 

Some other paaaoges in which black-coloured enemies are 
mentioned may also possibly be referable to the dark al>ori- 
gines ; such as R.-V. ii. 20. 7, already quoted (p. 385) : " ludra, 
the akiyer of Vrittra and destroyer of cities, scattered the servUe 
{^dd»ih) [hosts] of black deaceniJ" But Professor Roth (in his 
Lexicon), explains thi« last expression, kr'^hnayotiih, as well 
as hriahnagai'bhdh, in R.-V. i. lOl. 1, as descriptive of the 
bUick clvuda. The latter of these two phrases, is aiiuilarly un- 
derstood by M. Regiiier in his Etude sur I'ldiome des Vedas, p. 
154. In R.-V. iv. 16. 13, mention is made of 50,000 black ly- 
ings (explained by the commentator as Rakshasas) being de- 
stroyed by Indra. These also, axe perhaps to be regarded as 
aerial foes.*** See also Professor Benfey's explanation of R.-V. 
viii, 85. 15, in his trausliitiou of the Sama-veda, i. 323, p. 228. 
In the following text allusion is ina<le to bla<:k tribe<« ; II. -\'. 
vii. 5. 3 : i^f^^ f^ ^n"^J^lf%^tT^^RT ^npfNM-ilPlI 

fear of thee the black tribes fled, scattered, relinquishing their 
possessions (or food), when thou, Agui Vaisvanara, gleaming 
behalf of Piiru, didst tear and bum the cities." Professor Koth, 


;in ■ 


presRion, * O beautiful, O beautiful.' This was Paruchhepa's tarn of nunJ. 
lie was a riitbL." Here Ydska, tbe author of the Nirukla, speaks of ■ pftr> 
ticular mode of composition as peculiar to Purucbhcpa, one of the Vedio 
riabig. But if tbe form of the couipositiun was the rosult of the rishi's own 
particular genius («■«"/«»«), he must have done more than "see" the hymn 
(above, p. 205) ; he must have dctcrniined its particular form. The hjmn 
could not therefore have existed eternallj, expressed in its present words. 
Yfij-kii, therefore, appears to be inconsistent with himself, when he states this 
doctrine in other pauage^ as x. 10, 46, where he says, risher driiihtorOuuiia\ 
pruir rMi/thuumijfukta : " Here the rishi, after he had beheld tbe contcnU | 
[of a hymn], expresses his pleasure in narration." If, indeed, we are to un- 
derstand by the word artha, " contents," that the fnalttr of the hymn only, ] 
without the worrls, was revealed, there will be no inconiistency. 

**• See, however, Wilaon, Trans, of Big-veda, toI. iii. Introd. pp. ▼iii, ix. 
sir. and xv. 


•BCT. IJ.] 



however (Diet sub voce "asikni^, explains the words blxick irihes, 
as meaning " spirits of darkness." A similar phrase occurs in 
K.-V. viii. 62. 18 : J?: ?f v^T ^TT^Ifl W^^SPiJT ^rfv<ff f^lTTl 
" impetuous [god], break down as it were a city, being ha- 
rassed by the black race." It is not clear who is the deity here 
apostrophised ; hut as it is he, and not his worshipper, who was 
assailed l>y dark-coloured enemies, we cannot suppose the latter 
to have been terrestrial. 

There are (4.) some passives in which the epithet mfidhramck 
is applied \o the speech of the Dasyus. If it were certain that 
the aboriginal tribes were alluded to in all or in any of these 
texta, and that the adjective in question had reference to any 
peculiarity in their language, the fact wouJd be one of the highest 
inter&st ; hut unfortunately both points appear to be doubtful. 
The following are the passages alluded to: — R.-V. i, 174. 2: 

T^T f^ T^ ^V^T^: ^TT ^ 5?:: ^F i[\<{\ t^J 

" When thou, O Indra, didst for our welfare [?] destroy the seven 
autumnal cities, thou didst subdue the people of imperfect 
utterance. Thou, blameless one, ha*t impcdled the flowing 
waters : thou hast subjected Vrittra to the youtlifid Purukutsiu" 
E.-V. v. 29. 10 : IT ^SRT^^rTi^: ^^ ^WT^ ^I^ 

^ftrr ^mris^: i trtwY ^t^, "^f^^ 4v^ f^ ^;^^ 

^9TW^ 4j t j<^ | H| ; I "Thou didst detach one wheel of Surya: 
the other thou gavest to Kutsa, that ho might acquire wealth. 
Thou hast with thy weapon smitten the mouthless [or noseless] 
Dasyus; in the battle thou hai^t pierced the imperfect wpi-uking 
people." V. 32. 8 : TIT f^ "VtX^ Ijvfi l[^nWf^^'^ WS[ Ji^ 

v^|-^4J I "Tlie fierce [Indra] seized that huge, restless 
[Vrittra], the holder of the waters, reclining, insatiable, the 
overspreading ; and destroyed in battle with his great weapon. 




that footless. 


and imperfectly-speaking [demon}" 

VII. 6. 3 : f^ yiffTs:^ n\^^ ipf^r^: mh^rs^ ^rpH 

<3U«4 I " The senseleas, false, iinperfectly-speakiug, unbeliev- 
ing, unpraising, unworshipping Panis ; these Dasyus Agni removed 
far off. It was he who first made the irreligioufl degraded." 

The word mfidhravdch, which I have rendered " imperfectly- 
speaking," IS explained by Sayana (in his conuuents on the lAst 
three passages) as meaning " one whose organs of speech are 
destroyed" {Idiisitavdyiiidriya, or himitavachaska). The same 
term is rendered by Wilson, in his translation of the second and 
third passages, by "speech-bereft," or "speechless." Both, in 
his " Illustrations of the Nirukta," p, 97, rejects the explanation 
of Yaska, (who (Nir. vi. 31) renders mfitlftravdchali by mridtlr- 
vdchah, " softly-speaking,") and considers that it means " epeak- 
iug injuriously." Dr. Kuhn, again (Herabkunft des feuers, 
p. 60), is of opinion that the epithet in question means " a stut- 
terer ;" and thinks that, in R.-V, v. 29. 10, reference is made to 
the gradual dying away of the distant thunder, which is regarded 
as the voice of the vanquished demon. In treating of the alwri- 
ginal races of India, Professor Miiller (Last Results of Turanian 
Researches, p. 346) remarkft, that "the 'anasas* enemies, whom 
Indra killed with Ids weapon (R.-V. v. 29. 10), are probably meant 
for noseless (a-nilsas), not, as the commentator supposes, for 
faceless (an-asas) people." He must therefore either regard the 
Dasyus, who are here referred to, as human beings, or conceive 
that the epithets applied to the demons by the Vedic rishis were 
borrowed from the features of the aboriginal tribes. Professor 
Wilson, on the other hand, remarks (R.-V. vol. iii. p. 276, note) : 
" andaoj Sayana says, means (IsyarafiUdn devoid of, or deprived 
of, words ; ds^ya, face or mouth, being put by metonymy for sabdOf 
the sound that coraes from the mouth, articulate speech, alluding 
possibly to the uncultivated dialects of the barbarous tribes, bar- 
barism and uncultivated speech being identical, in the opinion of 






the Hind 118, as in the familiar term for a barbarian, mleckfiOy 
wbicb ijs derived from the root, inlechk, to apeak rudely ; " and 
adds, in reference to Professor Miiller'a proposed interpretation 
of aiidsa: "The proposal is ingenious, but it seems more 
likely that Sayona is right, as we have the Da»tfus presently 
called also mj'idhravdchas, .... having defective organs of 

There are only two of the four preceding passages containing 
the word viridhravdch, in which the Daayua are named ; and in 
the last of these two texts (R.-V. vii. 6. 3) this word is a])plied to 
the Panis, the mythical beings who stole the cows of the gods or 
the Angirases, and hid them in a cave. (See Wilson's R.-V. vol. 
i, pp. 16, 17, note.) In any case, the sense of the word mfi- 
dhravdch is too imcertaia to admit of our referring it with con- 
fidence to any peculiarity in the speech of tlie aborigines. 

In the R.-V. frequent mention is made (5.) of the cities of the 
Dasyus, or of the Asuras. One of these passages, i. 103. 3, has 
been already quoted in p, 375. The following are adtlttional in- 
stances: — R.-V. i. 51. 5 : rSf f^^\ Ah^I TfT^^l ^l 3T '^- 
fbl^yi*! ^W^rZl^lf^ij I •' Benevolent to men, thou hast bro- 
ken the cities of Pipru, and protected Rijiswan in his battles 
with the Daayus." R.-V. i. 63. 7 : ^ ^ (fff^nT ^TT ^^ZT'f 
^rt" <<nj*t . M<v^<^>*< T^M "Thou, ludra, thimderer, 
fighting for Purukutsa, didst destroy those seven cities," I. 
174.8: fifV^ ^ ^ f»T^S^<'<ft ^Tjfr ^NT^^^ 
i^t^TTt I " He has broken the destroying and godless cities : 
thou hast bowed dovra the weapon of the godless destroyer." HI. 

12. 6: %^^[^ liR^ jfV TT^trSftT^^fT^l «!*«**» 
qr^l^ I "Indraand Agni, by one effort together, have shat- 
tered ninety cities belonging to the Dasyus.'" FV. 26. 3 : 

Hym^ "km ^iriTfU fi4\i^\HMU?*ii^ ^^^^^^\ " Ejthila- 
rated, I have destroyed at once the ninety-nine cities uf feam- 



LCSAp, tn. 

bam: the hondredtli I gave be to Lnliabited, when I protected 
DiTodasa Afithigva at the sacrifice."*' ^^. 61. 4 : t^ 3(<^l^4 
"SF^T^ 5^ a i M^^mfflf*! JW^: I " Thou hast <le8tn._v...l 
bimdreds of aneqttalled cities of tlie Daeyu Sambara.** 

Iron cities are spoken of in the following paaaage: ii. 20. 8: 

*• NMien they placed the thimderbolt in his (Indra's) hand, he 
slew the Dasrus, and overthrew their iron cities.'" ** 

In the following texts " autumnal cities'' are spoken of, 

L 131. 4 : — f^^ ^^rej ^*^*n ^T^: Y^ <<f<*i? mT- f 
<)<^iri<:i wT^rnrr ^<<ird<:i ^rr^^rffTr? ^R7Q"JT?r55 

9(^V<jiJdll " M^Q know this heroism of thine that thou hast 
overthrown the autumnal cities, violently overthrown them. Lord 
of power, thou hast chastised the mortal who sacrifices not."** 
iSee also R.-V. L 174. 2, which has been quoted in p. 393. \1. 

20. 10 : ^TT ^ 3T: ^ ^\<{\ ^^'^^ T^' ^^^W^ 
faji^iff I " Because thou didst break down the seven autumnal 
cities with thy thunderbolt, slaying the Dasa (people), and giving 
(wealth?) to Purukutsa."«« 

** See Kuhn'a Herabkunft des fenen, p. 140, and note. 

** Mention is also made of " iron cities" in the following texts: R.^ 
iv. 27. 1 ; rii. 3. 7 ; vii. 15. 14; vii. 95. 1 ; viii.89. 8; and x. 101.8; but 
in cotincction with the Dasrus. 

*' Do the "autumnal" cities mean huts of branches and leaves, or of 
straw, hastily constructed in the rainy seascn ? Or do they mean the 
brilliant battlcmented cloud-castles, which are so oClea visible in the Indian 
sky, at the same period of the year? Suyana in loco explains the term tLiu: 
\,4aradih samvatianuambandhinih tamvaUaraparyantain prahara'parikhndAhir 
'dridhikrituk purah iairvnam purih. " The enemies' annuai cities, forti6ed for 
B year with ramparts, ditches, &c."; but see next note. 

«6 Sayana in his note on this verse, explains the word iJra</iA difTerentJy, 
as, iarannilmnah asuratya mmbandhinih. "Belonging to an Asura called 
darad." Say ana renders the word iarma in this passage by ** with thy 
thunderbolt." In his note on R.-V. i. 174. 2, he had previously rendered 
it by " for our happiness." 


•■CT. W.] 



" Ancient" cities are spoken of in ii. 14. 6 : ff. J[^ l^h^i^m 
^Tt f^^<^\m•^'i ^^*\ "^Vlioshatteredthe hundred ancient 
cities of Jsamhora, as with a rock," &c. "Eternal" cities are 
spoken of iu the following texts, viii. 17. 13: i ^ ii"T?rr ^TJ 
IT^fftTrf^n^t ifsH'^t ^^1 " The impetuous destroyer of 
the eternal^' cities, ludra, is the friend of sages.'' VIII. 87. 6 : 

^fprfifwi I " Thou, ludra, are the destroyer of the eternal 
cities, the slayer of the Dasyus, the benefactor of Manu, the lord 
of the sky.""** Cities of stone are mentioned in one passage, 
iv. 30. 20 : IPT^ 'S3H5a?|^t J^rf^T^ ^JT^Te^l f^^V^- 

Hm ^ I 'J^CJ I " Indra has thrown down a hundred cities built 
of atone for his worshipper Divodasa."*' In R.-V. viii. 1. 28, 
mention is made of a " moving" city : T^ XiJ^ ^fr^ WV('. 
1^^!PS <|fi-mi l '»J " Thou hast shattered with thy bolts the 
luoving city of iSui.hna."'*' 

The cities referred to in these Vedic hymns were, in later times 
at least, understood of cities of the Asuras; and the following 
legend was invented to explain what they were. In the Com- 
mentary on the Vajasaneyi-Sanhita of the Ynjur-veda, the fol- 
lowing passage occurs: — ^-sj^^^ll^lfil^d ^f^ I ^q! 

** Aiipat has, however, according to the Nighantu, also the sense of 

** In R.-V. viii. 84, .1, we find the snnic epithet applied to persons : 
Twam hi MvaRnam patih raja vUdm ati. ** Thou art the lord, the king of 
the eternal peoples." 

** SSjana interprets aimcmmaytnam hj pathanair nirTniVaadm, "built of 
Btooe," and says they were the cities of Sanibam. 

^ It appears as if the moveable cloud-cities were here meant. 





rKf^M<<il<J 'Hi: ll " On this test [Vaj. Sanh. 5. 8.] the fol- 
lowing story is cited : — The Asuras having been vanquished liy 
the gode, performed penance, and built three cities in the uni- 
verse, — one of iron on the earth, one of silver in the atmosphere, 
and one of gold in the sky. Then Agni was supplicated by the 
gods to bum these cities with the upasiui fire. In ooosequetioe^ 
Agni in the form of the upasad deity entered these cities, and 
burned them. Then these cities became tlie bodies of AgnL It is 
I to this that the Mantra (text) has reference."*' The ^tapath»> 
[brdbmana (iiu 4. 4. 3, ff.) has the following passage on the same 
ibject:— ^cTT^ ^ ^q<T"g ^?H^ MlsilMrtJi: ^H^f^l 

3irfi?^'^^ T^T^^ ^sft^TT^ 3rr3w\i ?t^tt^T5^tw^ ^ 

" The gods and Asuras, who were both descended from Prajapati, 
contended together. Then the Asuras constructed cities in these 
worlds, one of irou in tliis world, one of silver in the atmosphere, 
and one of gold in the sky. The gods were envious of this. They 
sat near (iipa-asidanyvnth these upasads: and from their thus 
sitting the name of upusiul originated. They smote the cities, and 
conquered these worlds. Hence the saying that men conquer 
a city with an upasiul. Because they sit near, they conquer 



'* The reference here ia to the text of the V.-S. 5. S, wliich contikina 
the words ya le Agne uyahiaya taituh ; yd le agve ■ntjahiaya tanuli ; ffa te agnt 
hariiat/a lanuh : "The body of thine, Agni, which reposes in iron; which 
reposes in silver j which reposes in gold." The upttsad was a fire which wu 
kept burning for several days. See Boht. and lloth's Lexicon. 





this city of mortals. By these upasadg the gods smote the 
cities, aiid conquered these worlds," (See Weber's Ind. Stud. ii. 

In several texts Dasyus are mentioned in connection with 
mountains. Thus in R.-V, ii. 12. 11, we have the words 

?t: TT^yrTT^^f^^ -«f »q iR » j^ i ^<<ij*<f^*^fi ^i "Who 

discovered Sambara living in the mountains in the fortieth 
autumn ;" and in i. 130. 7, f^T^n^ TIXV T^fTTf^T^ ^T^ 

■^tsWT f^^l V*fr5^V5f¥T II " Indra, dancing, thou didst 
shatter by thy bolt uiuety-iiine cities for Puru, fur tlie great sa- 
crificer DivodJisa; — dancing, — for the sacrificer. Fierce, he 
hurled down Sambara from the mountain for the sake of Atithi- 
gva, bestowing great wealth by bis power, all wealth by bis 
power.'' rV. 30. 14: '^rl ^ ^f^fTT 'ZT^f' M4fil<f^i 
^SlT^f^T^ ai#-«j<*J 1 " Thou, Indra, didst cast down the Daea 
Sambara, son of Kulitara, from the great mountain." VI. 26, 5 : 
^R f*l^<T^ 'H^^X "^^ TTRt f4<ri<m*(^l " Having hurled 
down the Datia Sambara from the mountain, thou didst preserve 

The wealth or property of the Dasyus or Asuras is spoken of 
in various places. Thus in L 33. 4 ; ^\f^f^ ^]^ vf'PT ^If 
U.ii^i'^M llI*t*lf<*S 1 " Thou, O Indra, advancing singly, 
though (*upport<?d by powerful [allies], hast slain the wealthy 
Dasyu with thy destructive weapvon,"** I. 176. 4 : ^^^•ri^T? 

*' Savana reronrks on this verse: " The Vijananejins dlslincUy fecord 
the wealth [of Vrittra] in these words; ^ Vrittraiydntith tarve devah 
tarcnicha nidyiih narvuni kaviHshi cha wtan : la Vfittni were ConlAined all 
the gods, all the sciences, and ull oblations.' " 


ZcBAf. ill. 

^ l^jrl I " Kill all those who make no oblations, though di£B- 
cult to destroy, and who cause thee no gladness ; give us their 
wealth: the worshipper erpects it." II. 15. 4 : ^ lj4^ad|pi 

'^i 4^^<^ T^f^I I " Encountering those (Asuras) wlio carried 
away Dabiiiti, he burnefl all their weapons in the blazing HrCj and 
presented [?] Dabhiti with their cows, horses, and chariots." FV'. 
30. 13:^rT ?J4<4IH^ "t^-^^ T^^ ^ifi? "^^^^i jft 

^^^ ^f?(HJl* I " Thou, iinpetuoua god, didst collect the 
wealth of Sushna, when thou didst overthrow his cities." VIII. 
40. 6: ^ "rg M<I1N< ^T^^fr^ ifqifl ^TsfV 

" Koot up, like ail aucieut tree [?] overgrown by a creeping- 
plaut, subdue the might of the Dasa; may we share with 
Indra (or divide by means of Indra) his collected wealth." 
X. G9. 6. : ^^ ^I^[JT qwr ^^ ^T^ ^^^Tf^ ^U^ 
fs|J|^| "Thou hast conquered the property, whether situated 
in the plains or hills, (thou hast conquered) the Da£a and the 
Aryya enemies." 

In the follomng and numerous other texts (as well as in 
some of the prece<ling), various Dasyiis, or at least adversa- 
ries, are specified l>y name, together with the persons who were 
delivered from them. R.-V. i, 51. 6: ^ SiH < te^'^ru^tlPl^V 

^sTT'^wr s^f?T f^ ^ I «j ir'-«J<*ii Tfpff f^^^ 'ii^M): 

V[^ V^I^e< <*i<l5rtJT^ ^f%^l "Thou hast preserved 
Kutsa in his fights with f^ushna ; hast subdued i^mbara to Ati- 
thigva ; thou hast trodden under foot the great Arbuda ; of old thou been born to destroy the Da.syus,"*' I. 63. 3 : ?^ 55"! 

T3i^ ?% ^rnrr ^ ^<ui*j ^^ ^^ ^jif^i " Thou 

** This verse is followed almost iimnediatclj by the text, i. 51. 8, 
quoted above, p. 374. " Distinguish between the- Aryjas," &o. 






bast in the close and deadly fight slain Sushna, in aid of the 
young aud brilliant Kutsa." VII, 19. 2: 1^ f? WfT^ 

^(4)4{M: a^^M*llU!*tI^T ¥^l jvi ^ 35^ ^f^t% 

^rorr ^<«'trJIt "SSTT^^^TO far'^'! I "THou art he, O Indra, 
who didst deliver Kutsa in the fray, interposing with thy body, 
wheii, bestowing favour on tliat son of Arjuni, thou didst subdue 
to him the Dfisa t>us!itia, and Kiiyava."^' 1.53.8,9,10: 

Tt^ w^(^ fTTrf?Tf>Tww ^^T^Tf^rfTT f3;^^w^i «^?rw 

^^'TfTff^n^^T^ ^ Tra ^ 'V<«'y«fl«i:iJ\"»l| "Thou 

hast sljiin ICaranja and Parnaya with the glittering spear of 
Atithigva, Unyielding, tliou hast broken down the huniired 
cities of Vangrida, wldch had been blockaded by Rijisvan. 
Thou, renowned Indra, ha«t with thy swift chariot-wheels, 
repelled these twenty kings of men, who assailed the unaided 
Sii^'avas, [aud their] sixty thousand and ninety-nine [fol- 
lowers]. Tliou hast by thy aids protected Susravas, and by thy 
help Turvayaniv. To this mighty youthful king thou lia« sub- 
jected Knt«a, Atithigva, and Ayu."** II. 30. 8: 9^-^lf?f 

*' In R.-V. iv. 26. 1, (quoted above, p. 376,) also, Kutsa la called the 
son of Arjuni. Kuhn considers tliAt Kutsa is a personification of the lightning, 
a view which he considers to be Confirtnecl by his patronymic of Arjuni, 
Arjitva, being on epithet of Indra, and of the thunderbolt. Sec Ilerabkunft 
des feuers, pp. 57-62, 65, 140, 176. Kujova is also mentioned in i. 103. 3: 
Kthirena tnatah Kuyavasija yoahe, hale te syiitum prarane Siphtiydh. "The 
two wives of Kuynva bnthc with water; may they be drowned in the stream 
of the f^iphn.'' 

" The youthful king alluded to in the last verse, is said by Roth 
(Diet, under the word "Atithigva") to be TurvaySna. These names occur 

Kin itk vi. 18. l;i. 

!► D 

402 THE ARIANS IN IXDIA : [cHAr. to. 

7ff^ef)<<*HHim^ -^y^T ^f^ |TnT irr'Ji*l«ll*<^l "Sank- 
awati, do thou, impetuous?, attended by the Mamta, protect u*, ftad 
conquer uur enemies. ludra destroys the chief of the Sandikiu, 

arrogant, and making a display of his Btrength."'* IV. 30. 15 : 

" Thou ha'Jt slain a thousand and five hundred followers of f 
the Daea Varchin like fellies of a wheel."" IV. 30. 21 : ^i(<IM<I^ 

by iiis wistloui, put to sleep with liis weart<iiis tliirfy tliou.sanJ 
Biwiw for Dabhiti." V. 30. 7, 9 : ^SR i^ l Mm STO^I flpfV 

^^ TI^W ^ ^^ <<^f*<-[^:il ""^Mien desiring hftppincw 
fur Manu, thou didst overthrow the head of the Diwa N»- 
muchi." The Dasa took his wives for allies in battle. Wli*t 
will his feeble hosts do to me ? He concealed his two fiur 
ones: and then Imlra went forth t'l fight with the Das}'u." 
Vlir. 32. 26: ^r?1^ |^4j^H« ^^PHT^tlJ^ f^' 
^Tlf^'^^^^^H.' " Iw^lrft slew Vrittra, Aurnavabha, Ahi^tnra: 
with frost he pierced Arbuda,'MT[II. 40. 10: ^|di4JI J^yg^yj- 
WtI^ ^TTfJ^ ^^^ ^^ffK^.'l "By hia might he crushes 



** Sn^nnn says tbat Sandumarkuc Aturapurohitau : " Sand* and Amarka, 
are the priests of the Asuras." In thi.' ^iitapatlia-brahmnna, i. 1. 4. 14, 
(p. 9, of Weber's edition), Kilata and Akuli are declai-eil to be tbe pri«U 
of the Asuras : KilataktUi Hi tia Atura brahmtie Amtuh. Z. D. M. G. for 1850, 
{». 302. In Ind. Siud. i. 32, Weber quotes the fullowing word« firom Uie 
Panvhvin^a-brnbniana, 13.11: Guupavandndm eai tatram annanam Kird- ] 
tdhitlj/i'ic asuramai/e, &c. " While the Gaupiivanafi were yeateil at a ucrifio^J 
Kirata and Akuli. &c." See alw Ind. Stud. i. p. 186, 195, ii. 'S43. 

»' See also R.-V. Tii. 99. 4. 

^* There is a legend about Indra and the Ajura Namuohi ia tb» Sctaa. i 
Brfih., V. 4. 1. 8. p. 43!>. 

ncT. It.] 



the eggs of Sushna ; he conquered the celestial waters." 

X. 54. 1: irrwV <tfi*^^ ^ifriO 4i*J*riaf: inrT^ «^ 

Tj^[^ir^:| "Thou [InJm] hast protected the gods [priests?], 
thou hast overcome the might of the DasyiiP, when thou didst 
bestow [l>oous] on this people." X. 73. 7: <«T^ ^«^^ *mf% 
«T^ JVi ^WT^' "^^ f^rora^l " Thou haat shun 
Namuchi, desirous of the sacrifice, making the Dasa devoid 
of magic for the rirthi."^* X. 95. 7 : ^frf rfr M^<<4) <U ![ eH - 
^^l^n^ ^^T^T^rn? ^TT! I " ^\'hen, Puriiravas, the gods 
strengthened thee for the conflict with the Dasyus," 

I have gone over the names of the Dasyus or Asuras men- 
tioned in the R.-V. with the view of discovering whether any 
of them could be regarded as of non-Arian or indigenous 
origin ; but I have not observcjd any that appear to be of thi8 
character. But we should recollect that the Arians would 
not unnaturally designate the aboriginal leaders (if they specified 
any of them in their sacred hymns) by names of Arian origin, 
or at softened into an Arian form. The Greeks introduced 
Greek modifications into Persian and other proper names, and the 
Clialdeans gave Chaldean appellations to the Jews. 

In some pas-sagcs tlie Da.'fyuH are spoken of its monsiters. 
Thus, ii. 14. 4 : tl nJ;^ ^"^Tf «TW ^I|"N •!<^fr|^ *ll^«t^l 

^ ^4<^«t 'fl'^r spfT^ 1 " V^TiO slew Urana, displaying 
ninety-nine arms; who struck down Arbuda," &c. X. 99. 6: 

f%^ ^ '^•ST^ Y'^m fwm TriT»raTTT«rT V\l "This 
lord tumbled and subjugated the roaring Dasa, with six eyes 
and three heiwls.*" Trita, increasing in strength, struck this 
boar with his iron-tipped finger." The enemies of Indra are 

" See Benfej, Gloss, to S.-V. umler tlie word Namuchi. 
^ •" In d. P. Br. p. 57, a son of Twashtri with three beads anU six eyes 
is mentioned u having bis three heads cut off by Indra. 



[ciur. lu. 

spoken of iu a few passages as scaling the sky: thu», ii. 12. 12: 

" men, he who, armed with the thunder, slew Raubina as 
he was scaling the heaven, is India." VIII. 14. 14: 'JTRTTf^^^- 

f^<iHi« T^ ^rr^TT^^^fT: I ^ra t4^^¥3^' ' "TI^ju, iu- 

dra, hast hurletl down the Dasyiis, who, liy their magical powers, 
were mounting upwards, tind seeking to scale heavaa."** In u. 
11. 2, the Dasyus are said to regard themselves as immortal : 

•^ w^f^ VT '^ft^: MRfHHi '^ifw 7i:T ^:i 

hast, O heruie ludra, let louse the gretit primeval waters, which 
thou aiignientedst when they were stopped hy Ahi. Gaining vi- 
gour by hymns, he shattered the Daan, who regarded himself 
a.s immortal." In v. 7. 10, tlie Da-sj-us are conjoined with men : 

^«T I " Hereupon, Aj,Tii, may Atri overcome the irreligious 
JJasyus, may he overcome hostile men." 

I have thus brought under review in this section a variety of 
passages which bear, or might be conceived to bear, some re- 
ference to the conquest of territory by the Aryas, and to the 
Condition, colour, speech, religious rites and cities of the Da»i/us. 
The meaning of many of these texts is, however, as we have 
seen, extremely doubtful ; and some of them are clearly of a 
mythological import. Such, for instance, are those which de- 
scribe the contests of Indra with Vrittra, the demon of the clouds, 
who withholds rain ; where we are, no doubt, to understand both 
the god and his advei-sary as personifications of atmo.spherical 
phenomena. In the same way, iSambara, iSualina, and Namnchi, 
are to be regarded n.s mjthieal personages, of a kindred chara^'ter 
with Vrittra. And yet there are many passages in which tJie 
word vrittra has the signification of enemy in general (as R.-V. 

'* In i. 7S. 4, the cxjircwion yo dam/unr ava dhuntuhe recurs. 

«ECT. II.] 



vi. 33. 3; vi. 60. 6; vii. 83. 1, p. 377); aod Professor Spiegel, 
as we have seen (p. 317), is of opinion that the words vrittrahm 
and lYitlrafjhua had originally nothing hnt a geneml significa- 
tion, and that it was only at a later period that they canie to be 
epithets of Indra. The word Sambara, again, aa Beufey (Glos- 
sary to Sama-veda, p. 181) remarks, is given in the Nigha^tuas 
synonymous with vieyhn, a cloud (i. 10), with udahi, water (i. 
12), and with 6«/a, force (ii. 9); while the mythical narrations 
generally identify him with Vrittra. In regard to this word 
Professor Roth remarks as follows (Lit. and Hist, of the Veda, 
p. 116): "In the passages which speak of Divodusa, mention is 
niade of his deliverance, by the aid of the gods, from tlie opprt\s.sor 
^mliara, c y., K.-V. i. 112. 14, ix. 3. 1, 2. It is true that 
{Sambara is employed at a later period to designate an enemy in 
general, and in particidar the enemy of Indra, Vrittni; but it is 
not improbable that this may be the transference of the more 
ancient recollection of a dreAded enemy to the greatest of all 
enemies, the demon of the clouds," 

Professor MiiUer (Liist Results of the Turanian Researches, 
pp. 344, £F.) remarks as follows on the use of the words Daayu, 
Rakshas, Yatudhana, &c. : *' Dasyu simply means enemy ; for 
instance, when Indi-a is praised because ' he destroyed the 
Dasj'ua and protected the Arian colour.'" The * Dasyus,' in the 
Veda, may mean non-Arian races in many hymns ; yet the mere 
fact of tribes being willed the enemies of certain kings or priests 
can hai'dly be said to prove their barbarian origin. Vasishtha 
himstilf, the very type of the Arian Brahman, when in feud 
with Visvamitra, is called not only an enemy but a 'YStudhana, 
and other names which in common parlance are only bestowed 
on barbarian savages and evil spirits." (.See the First P«u-t of 
this work, p. 132, where the original passage, and the traus- 

^ [Tbi« piutiage, iii. 34. 9, nppean to me to be, rather, one of those iu 
-which the contrast i« most strongly draim between the Arjru and Utc 
aborigines. S*e Bbove, pp. SB-l, and 387. — J.M.] 

D n 3 

lation are given.) " In other paesages [of the R.-V.] the word 
. . * devil (rakshas) is clearly applied to barbarous nationa. 
Origiually ral'shas meant etrong and powerftil, but it soon took 
the sense of giaut and barbarian,"' and in this sense it occurs 
in the Veda, together with Yatiidhaua, Another Vaidik epithet 
applied, as it seems, to wild tribes infesting the seats of the 
Aryas, is 'anagnitra,' 'they who do not keep the fire.' Thus we 
read, *Agui, drive away from us the enemies — tribes wlio keep 
no sacred fires came to attack us. Come again to the earth, 
sacred god, with oil the immortals, come to our libation.'*' The 
same races are called * kravyad,' or flesh-eaters. In a famoiu 
hymn of Va*islitha we read: 'Indraand Soma, burn the R ak s h aH, 
ileetroy them, throw them down, ye two Bulls, the people that 
grow in darkness. Hew down the madmen, suffocate them, kill 
them, hurl them awny, and slay the voracious. Indra and Soma, 
up together against the cursing demon ! may he biun and hiss 
like an oblation in the fire I Put your everlasting hatred on the 
villain, who hates the Rrahraan, who eats flesh, and who0e 
look is abominable.' ^ . . , Ivj-avyad, flesh-eater, means people 

"• The Rev. Dr. Wilson (TndJa Three Thousand Years Ago, p. 20) ihinki 
that the word;) Riikshnnti, FiSuchn, nnd Asura, were originalljr names of 
tribes ; but he ii<lduce,» no proof of this «iid 1 have found none. 

*' R.-V. i, 189. 3. Agne tvam asnuid yttyodhi amwa anagiiiira ahhyamanta 
hrithlhih I Punar u»mabhyam tiwitaya Deva ktham vUvebhir amritehkir 
yajatm \ | 

"* [li.-V. vii. 104. 1, 2. Tnilra-Sotna tapatam rahtha ubjotam ni arpoj/atam 
rrishanu tamovridhah | Paruifinitam achito ni oshatam katam nudetham « 
iiiUam alrinah \ Indru-Sumd sum aghaiaiisam nbfiyagfuim tapwr yuytutu 
charur agnivan iva \ Brahmadvishe krapyade ghorachak»haae dveaho dhat- 
tam avavCiynm kimullne | In a strain, Viivfimitra, tlie rival of 
Viwisbtlia, snj-g, in R.-V. iii. 30. 15 — 17, Indra drihya yamakofd abhimnt 
yajhaya iUinha grinale tahkibhyak \ Durmuyavo durevii martyuao nithangino 
riparo hantvumk \ sam ghoshah srinve uvamair amitrair jaiti ui ethu aianim 
tapixkthum | Vfifcha \m adhaxtnd vi rxijii sakatoa Juki rakiho Magkaran 
randknyanpii \ Udrriha rakxkah snhamuliim Indra rriicku madkyam prati 
iiurnm iriiiihi I A hientah saltiliikam ckakartka hrakmadruke taptukim ketim 






who eat raw meat, . . . and they are aUo called drnddus, . , . 
or raw-eaters, for the cookitig of meat was a difitingxibliiug 
feature of civilizetl natioiLs, aud frequently invested with a sacri- 
ficial chai"acter. Agni, who in the Vedas ia the type of the 
sacrifice, and with it of civiliRatiun and oocial virtues, takes an 
entirely different character in his capacity of ' kravyad,' or flesh- 
eat«r. He is represented under a form as hideous as the beings 
he is invoked to devour. He sharpens bis two iron tusks, puts 
his enemies into his mouth aud swallows them. [R.-V. x. 87. 
2, ff.] He heats the edges of his shaf^, and sends them inf« the 
hearts of the Raktihas. He tears their skin, minces their 
memhers, and throws them before the wolves to be eaten by 
them, or by the shrieking vultures. These Raksbas are them- 
selves called ' acbitas,' mad, and ' mfiradeva^i,' ** worshippers of 
mad gods. Nay, they are even taunted with eating human flesh, 
and are called 'asutripas,' as enjoying the life of other men. In the 
Rig-veda we read [x. 87. 16], 'the Yatudhanas who gloat on the 
bloody flesh of men or horses, and steal the milk of the cow, O 

asya \ " Indra, be strong ; [tbe Rakshases ?] have stopped tbe rood to ibe 
ucrifice : bestow favour on thy worshipper and his friemlg : let our mortal 
foes, bearing quivers, discharging weapona, and assailing us, be destroyed. A 
sound has been heard by our nearest foes ; hurl upon ihem ihy hottest 
bolt, cut them up from beneath, shatter them, overpower them; kill and 
subdue the Rukshas, O Mnghavau ! Tear up the Kukahai by the ruoM, 
Indra, cut him in the midst, destroy him at the extremities. Haw long 
dost thou delay ? Hurl thy burning shaft a<;ainst the enemy of the priest." 
See Roth, lllustr. of Nir., vi. 3, p. 72.— J. M.] 

" [The far more opprobrious epithet of iiiiWrpn is applied, apparently to 
the same class of people, in It.-V. vii. 21. 5; and x. 99. 3. The former of 
these texts ends thus: ma iUnadftS apt gvr ritrtm nnh : "let not the lascivi- 
ous wreicbes approach our sacred rite." Yaska (Nir. iv. 19) explains tUtui' 
deva by ubmhmacharyya ; Uoth, (Illustrations to Nir., p. 47,) thinks the wonl 
is a scornful appellation for priapic, or sensual demons. The lust line of the 
second text (x. 99. 3, apparently sjioken of Indra), is as fidlows: Aimrra yat 
iiiladunuya cedo ghruin iiinadevdn ahhi varpatu hhui. "When, smiting the 
wealth of the [city] with a hundred portals, the irresistible [god] overcame 
the lasciriotts wretches." — J. M-] 

Agni, cut off their heads with thy fiery sword.' All these epi- 
thets seem to apply to hostile, and most likely, aboriginal raoes, 
but they are too general to allow us the inference of any ethno- 
logical conclusions. The Vaidik rishis certainly distingniafa 
between Arian and non-Arian enemies. The gods are praised 
for destroying enemies, Arian as well aa barbarian {d<isd cha 
i-riUrd hatavi drydni cha), and we frequently find the expression 
< kill our Arian enemies, and the Dasa enemies; yea, kill all oar 
enemies.' But there is no allusion to any distinct physical 
features, such &a we find in later wiitera. The only expression 
that might be interpreted in this way is that of * susipra,' as 
•applied to Arian gtxls. It means ' with a beautiful nose.' As 
people are fain to transfer the qualities which they ore mort 
proud of in themselves to thek gods, and as they do not become 
aware of tlieir own good qualities, except by way of contrast, we 
might conclude that the beautiful nose of Indra was suggested 
by the flat noses of the aboriginal races. Tribe* with flat, or 
even no noses at all, are mentioned by Alexander's companions 
in India, and in the hpnns of tlie Kig-veda, Mauu is said to have 
conquered Yi-sisipra (Pada-text, vi^i-sipra), which nmy be trans- 
late<:l by 'noseless.' The Daaa, or barbai'ian, is also called 
' vrishasipra ' in the Veda, which seems to mean ' goat or bull- 
nosed,' and the ' anasas,' enemies whom Indra killed with his 
weapon (R.-V. v. 29. 9, lU), are probaljly meant for noseleas 
(a-nasas), not, as the commentator supposes, for faceless (an-esas) 
people." (See above, p. 394.) 

Professor Miiller then proceeds to remark that tlie phj'sical 
features of the aboriginal tribes are more distinctly desci-ibed in 
the Piirauas. (See tlie First Part of this work, p. 62.) 

W'ti rany perhaps be better able to luidcrstand many of the 
expressions and allasions in the hymns, and the manner in which 
some particular phrases and epithets are applied, (as it would 
appear, indiscriminately, to the different classes of beings, 
human, etherial, or demoniacal,) if we can first of all obtain a 
clear idea of the position in which the Aryas, on their settle- 



ment in India, would find themselves placed in reference to the 
aboriginal tribes; and if we, secondly, consider that the hymns 
in which these phrases are recorded, were composed at various 
date«, ranging over several centuries ; that the same words and 
phrases are perpetually recurring in the different hymns ; and 
that expressions employed in one sense in the earlier hymns may 
Jiave been transferred, in the compositions of a later date, to 
• different class of beings. We have further to recollect, that 
the hymns may not always have been handed down in a complete 
state, and that portions of different compositions, which had ori- 
ginally a different subject and purpose, may have been erroneously 
thrown together by compilers in after ages. I shall say a few 
words on each of these topics. 

First , then, we may conceive the Aryas advancing from the 
Indtis in a south-easterly direction into a countiy probably 
covered with forest, and occupied by savage tribes, who lived 
in rude huts, and subsisted on the spontaneous products of the 
woods, or on the produce of the cbace, and of fishing ; or perhaps 
by some rude attempts at agriculture. These btulmrians were of 
dark complexion, perhaps also of uncouth appearance, spoke a 
language fundamentally distinct from that of the Aryas, differed 
entirely from them in their religious worship, which no doubt 
would partiike of the most degraded fetishism, and (we am 
easily suppose) regarded with intense hostility the more civilised 
invaders who were gradually driving them from their ancient 
fastnesses. The Aryas, meanwhile, as they advanced, and 
gradually established themselves in the forests, fields, and 
villages of the aborigines, would not be able all at once to 
secure their position, but would be exposed to consttiut reprisals 
on the part of their enemies, who would " avail themselves of 
every opportunity to assail them, to carry off their cattle, 
disturb their rites, and impede their progress." *• The black 
complexion, ferocious aspect, barbarous habits, nide speech, 

« Wilson, R.-V. vol. i. Introd. p. xlii. 



jcmtr. m. ' 

and savage yells of the Dasyua, and the sudden and furtifie 
attacks which, under cover of the impenetrable woodti," nod 
the obscurity of night, they would make on the encampnientB of 
the Aryas, might naturally lead the latter to speak of tJiem, in 
the highly iigiu'ative language of an imaginative people in \h& 
first stage of civilisation, as ghosts or demons ; ^ or even to j 
conceive of their hidden assailants as po&sessed of magical Aod 
Buperhuman powers, or as headed by devils. The belief in 
ghosts is not obsolete (as every one knows) even in modem 
times and among Christian nations. In the case of noctximal 
attacks, the retiu-u of day would admonish the assailants to 
withdraw, and would restore the bewildered and harassed Arras 
to security; and, therefore, the rising of the sun in the east 
would Ije spoken of aa it is in one of the Brahmanas, as driving 
away or destroying the devils.™ In a similar way the author 
of the Rauiayana, si>eiakB, as we shall shortly see, of the barltarian 
tribes encountered by Kama in the Dekhan as Rakshafias and 
monkeys.^** ITiis state of things might last for some time. The 

"^ In the Rig-veda, there is a h)'ran (x. 146) of six versM, addressed to 
Aranyani the goddess of forests, which we can conceive to have he«n composed 
at a period such as that dcjcribc-d above, by a rishi accustomed to liveaaidst 
vast woods, and to the terrors incident to wandering through ibeir siilitud«s. 
The iJKt and last stanzas of this hymn are as follows : Arant/dni Aranyami 
atau yu preva naiynsi \ hatha grumam na prichhaii na tva bhlr ica rinda- 

tim\ Anjatuigandhim ntrabhim bahvaniium ahrithicalSm \ praham mrigamim 

mataram Aranyunim aiantisham. " Aranyaiii, Aranjani, thou who almost 
losest thyself, how is it that thou seekesl not the hamlet ? Doth not fear posuM 

thee? I have celebrated Arnnyani, the unctuous-sccnt«d. the fragraoti 

•bounding in food, destitute of tillage, the mother of wild beasts." Sec 
Both, Illustr. ofNIr., p. 132. 

'" In R.-Y. viti. 18. 13, human enemies are spoken of as acting like Ra- 
kshases : yo nak kaichid ririkshati RahnhcuUvena martyah rNiui sa evaik rfnrVA- 
Uhta: "May the man who seeks, with Kakshas-like atrocity, to injure us, 
perish by his own misconduct." 

" Quoted by Sayana on R.-V. i. 33. 8. adityu hyevodyan pttrattAI ra- 
kthdnty apaltanti. 

'" And in our own experience the Chinese speak of Europeans as •' foreign 




SBCT. n.] 



Arjaa, after advancing eome way, loight halt, to occupy, to 
clear and t«» cultivate the territory they had acquired ; and the 
aborigines might continue in possession of the adjacent tracts, 
sometimes at peace, and sometimes at war with their invaders. 
At length the further advance of the Aryas would either drive the 
Dasyiis into the remotest corners of the country, or lead to their 
partial incorporation with the conquerors as the lowest grade in 
their community. When this stage was reached, the Aryas 
would liave uo longer any occasion to compose prayers to the 
go«ls for protection against the aboriginal trills ; but their 
superstitious drea<l of the evil spirits, with which the popular 
mind in all ages has been prone to people the night, would still 

Secondly. Throughout the whole period, (which we may 
presume to have extended over several centuries,) during 
which the state of things just described continued, the com- 
position of the Vedic hymns was proceeding. These hymns were 
(as we have supposed, pp. 206, 209) preserved by the 
tiesceudanta of the several bards, who, on their part again, 
were constantly adding to the collection other new composi- 
tions of their own. The authors of these new effusions would 
naturally incorporate in them many thoughts and phrases 
borrowed from tiie older hymns which were preserved in their 
recollection,^' and which were now, perhaps, beginning to be 
invested with a certain sanctity. As circumstances changed, 
the allusions and references in the older hymns might be for- 
gotten; and it might happen that some of the expressions 
occurring in them would no longer be distinctly understood, 
and might in this way be applied to circumgrtances and events 
to which they had originally no reference. The same thing 
might also happen by way of accommodation : phrases or epithets 
referring to one class of enemies might be transferred to another, 
as Professor Roth supposes to have been the case with the word 

f> Compare Renan's Ilistoire dcs langues Semitiquea, 2d ed. p. 120, note 1. 




^ambara. Wlieu, iu fact, we see tbat hinte and alluisiuns in the 
Vedas have been often developed in Ibe Puriinas into legends ofl 
an entirely different cbfiracter and tendency; that the functio 
and attributes of the Yedic gods were quite changed in later 
and that even in the Brahmanas the true meaning of many of the 
Vedic testa has been misunderstood, it becomes quite admissible to 
suppoae, that, even iu theagewben the later hymns were composed, 
the process of misapplication may have commenced, and that their] 
authors may, in various instances, have employed tlie words of 
the earlier hymns in a different manner from that in which thej 
were at first applied. Iu thiiii way it is conceivable tliAt whi 
was originally said of the dark complexioued, degraded, ac 
savage aborigines, of their cities, and of their conflicts with th« 
Aryai!, may have been at a later period transferred to the fou 
sprites of darkness, to the hostile demons of the clouds:, and 
the conflicts of the gods with the Asuras and the Daityas. Or it i4 
perhaps a more probable supposition that, in the artless style of 
early poetry, the eai'thly enemies of the worshipper were men- 
tioned in the hymns along side of the malicious spirits of dark- 
neas, (with whom, as we have just seen p. 409, tJiey might be 
supposed to have some affinity or alUance,) or the aerial foes 
of Indra, This conjunction or confusion of different kinds of ^ 
enemies becomes the more intelligible if the word Dasyu, ag ijf 
supposed by different scholars, originally bore the generic sense 
of destroyer. It would thus come to be applied to all kincls 
of enemies, as the mention of one description of foes would 
naturally suggest a reference to the others, and to the epithets 
applicable to them. We oiu-selves apply to the devil the appel- 
lation of the foul fiend, a word which means enemy {/eiuil) in i 
German ; and when employing such prayers as *' deliver us] 
from the fear of the enemy," we naturally include all disturl 
of uur peace outward or inwai'd, physical or ghostly. Epithets] 
like anyatfrata " observing different rites," avinta " mthoufe) 
rites," ayaji/u " not sacrificrng,'' adeva " without gods," whichj 
were originally applicable only to men, might thus, in the ] 
of a fantastic mythology, be afterwards transferred to demons. 


. 11.] 



That this is not merely a presumption, but that the process 
in (question actually took place in India, may be illustrated by 
the following |xissHj^'e from the Chhando^'va Upanishad, p. 585 : 

r1*«l<fM ^^^ ^<<H*(^ '^Jii^yHH. ^<4i»l*1MH, ^TT^' 

<lfiO 3Rf^| ^*i<mii ^lTTf%^ ^?R1 IJ^K f^^^ 

" Hence even at the present day a person who is destitute of 
liberality and faith, and who does not sacrifice^ is contemptuously 
nfldressed as one of the Asura race. This is the sacred doctrine 
of the Asnras : the)' adorn the bodies of the dead with gifts, 
with raiment, and jewels, and imagine that by this means they 
shall attain the world to come." ^* 

The following ptissage of the Satap. Br. iii. 2. 1. 22 and 23, 
(p. 235, Weber's ed.), may also serve to show the connection 
between the aborigines and Asuras : Te Aaurd dttavachaso lie 
alava he alava iti vadantuk pardbubhuvuh. Tatra etdm api 
vdcham udiir ujiajijMsifdm, ect rnl^chhae, Ta^mdil na 
brdhmano vdechheil, Asnryd ha esbd vdk: " The Asimis, 
impaired in speech, and crying /«; alavah (*0 enemies,' incor- 
rectly, insteatl of he arayah) were defeated. Here they spoke 
this doubtful expression. This is incorrect language [or one 
who speaks so is a mlechha]. Therefore let no Brahman speak 
incorrectly; for this is the language of the Asuras."" 

'" See Weber's Ind. Stud. i. 271, 2, and note. 

" In the BrnhntDTUis numerous my thieal talcs occur of bnttles between t!ic 
Devat (gods) and Asura*, wliich Weber (Ind. Stud. i. 186, and ii. 2*3,) 
tliinks arc oflcn to be undcr«too<l of contests between the Arians and (lie 
aborigines. This lie con.<<i<lors to be proved hy the passage about Kilala 
U([whO!!C mime nearly corresponds wiiL tbat of the Kirtitas, an aboriginal race) 
and Akuli, prieiits of the .Vsuras, quoted above, p. 401 . note, from the S. P. Br. 
and hy the legend of Ravnna. It may be nUo worthy ofnote, that the word krivi, 
when occurring in the R.-V. (as in ii. 17. 6, and elsewhere), though taken by 
Both, in his Lexicon, to mean generally a clowl, is nnderstooil by the Com- 
mentator as the nnme of an Asura ; while this «amo word is slated in the 
S. P. Ur, xili. .'>. 4. 7, to be an old name of the Panchiilas ; Krivaya Hi ka vai 
fmra Panchidan uchaluhate. 

8jKrr. IIL — Tie Aricau on the Sarateati, and Aeir diffiuiom 

toulhward from that point. 

In the preoediDg sections we have seen thrit the Atyas on 
penetrating in Hindiisthan from the north-west, and advancing fl 
across the Panjrib from the Indus towards the SurasA-ati, found 
themselves in contact and conflict with a race of people, appa* j 
rently aboriginal, who are designated in the Vedic hymns by the I 
appellation of Dasyas. We shall shortly find that at a Uter 
period, on their southward progress from the Doab towards and 
across the Vindhya range, the Arians again began to press npoa 
the aborigines and drove them farther and further into the] 
Dekhan. We shall also find in the epithets applied in thaj 
KrLmnyana to the barbarous tribes of the Dekhan, a ooixfinnatioaj 
of the opinion that in the hymns of the Kig-veda the same class] 
of people are designated by such terms as Rakshas, &c. Before 
proceeding with this investigation, however, we miLst first ponsel 
for a moment with the Brahmanical Indians in the holy land on 
the banks of the Sarasvati, and then by the aid of some texts 
from their ancient writers, trace their advance from that point 
to the eastward, and their diffusion over northern India gene- 

We shall now, therefore, suppose that the Aryas, after tra- 
versing the country of the five rivers, have arrived on the banks 
of the Sarasvati, and have even extended themselves as fiu* as 
the Jumna and the Ganges. It would appear that the narrow 
tract called Brahniavartta between the Sarasvati and the Dri- 
shatlvati, alluded to in the classical passage of Manu, iu 17 — 24, 
must have been for a considerable period the seat of some of the 
most distinguished Indian priests and sages, that there the 
Bralimanical institutions must have been developed and ixuir 
tured, and perhaps the collection of the Vedic hymns complet«»d 
and the canon closed. (See above, pp. 301, 345. 348.) It is 
not easy to account in any other way for the sacrecl character 




attached to this small tract of country J* On this subject Lassen 
remarkfl as follows (in his Zeits<,'hnft, iii. p. 201): — " The sacred- 
nesjs [of the Sarasvati] must also rest on historical grounds, and 
be referred to an age when the contrast had become strongly 
marked between the inhabitants of inner India, whose institu- 
tions were fmrae<l according to rigid sacerdotfl.! principles, and 
the occupants of the Panjab, by whom such rules were but im- 
perfectly observed, Tliis couti-ast, however, was not only applic- 
able to the people of the west: towards the south also the 
country which was rfg;\i]ated by institutions of a strictly Indian 
character, terminated witli the Sarasvati : the place where that 
river disappeared was the door of the Nishada country : and she 
disappeared in order that she might not come into contact with 
that impure race."^* Of the same locality Professor Wilson 
(Vishnu-Puraiia, Preface, p. Ixvii.) remarks: "Various ail ven- 
tures of the first princes and most famous sages occur in this 

'* Some texts of the Veda relating to the Sarasvati h«ve been quoted 
above, pp. 3-59, 360. Weber (Inil. StuJ. ii. 311) ((uotes two passages from 
the l^aiikbayana and Ailarcya Bralimanos, about the rishfs holding a 
sacrifice on the banks of this river. The commencement of the legend ia 
the Ait. Br., ii. 19, is as fullow.s : Riihayo vai Sariuvatyam satrum atata, te 
Kupanham AUu$ham tomad anayan: datyah putrah kiUivo hrahmanah katham 
no iiuidhyc dlktkishteii, &c. : "ITie rishls attended nt a sacrifice on the [b.inks 
of] the Sarasvati. They rcmnvc<l Kavaeha Ailusha from the soma, saying 
'Thou art the son of a bondmaid, a spurious Brahmnn, how hast thou joined 
with us in the sacred rite?'" They were however at length compelled to 
admit him to their fellowship. In the M.-Bh. iii. ^074, quoted by Lassen 
(ZeitAchr., iii. *iOO), it is said : Dahxhinena Sarasvatyii Driihadratyuttarena cka, 
ye vatanti Kurukghetre te ratanti tripithtape : "Tho-te who dwell in Kuru- 
kshetra south of the Sarasvati, and north of the Dfishadvati, dwell in 
heaven." See also Part Firet of this work, pp. 203, 204. 

'* M.-Bh. iii. 10,535*. Etad Vinaianamnama Samsrutya rUwnpate : dvdram 
Niahadartuhtratya yesham doshut Sarasvati, Prarighttl jtrithivim vira ma 
Niihudd hi mam cidtth. "This is the place called the VinaJlana (disappearance) 
of the Sarasvati, the gate of the country of the Nishudas, to whose impurity 
it was due that the Sarssvati sank into the earth, lest, [as she satd,] the 
NisbSdas should become acquninted with her." 



[chat. m. 

vicinity; and the Asramas, or religious domiciles, of Mreral of 
tlie latter are placed on the banks of the Sarasvati. According 
to some authorities, it was the abode of Vyasa, the compiler of 
the Vedas and Puranas ; and agreeably to another, when on one 
occasion the Vcdas had fallen into disuse, and been foigottec, 
the Brahmans were again instnicted in them by SanevAt^ the 
son of Sarasvati.^^ One of the most distinguished of ibe tzibes 
of the Brahmans is known as the Sarasvata. .... Tfae 
river itself receives its appellation from Sarasvatl the goddev of 
learning, imder whose auspices the sacred literature of the 
Hindus assumed shape and authority/' [May we not with as much 
probability suppose the converse to have been the case, and thai 
the goddess derived her name from the river on whose banks she 
may be said to have been bom ?] " These indications render it 
certain that, whatever seeds were imported from without, it was 
in the country adjacent to the Sarasvati river that they were first 
planted and cultivated and reared in Hindusthau." See also 
La£sen, Zeitschrift, iii. p. 202. 

The high degree of sanctity ascribed by the Indiana, at the 
doee of the Vedic era, to the country between the Sarasrad and 
Prishadvati, is further proved by tlxe following panage of Mana, 
already repeatedly reforre<i to : 

Mann, ii. 17— 24: 4(<fs<fft^ M^rtfl <=t*1tri<5< tI <*<^l <T 

^C«ir*iri ^ M^l<«-H M-^^dtt fT%\ ^ ^ SMI< : 

** Aa abctnnt of the kgead here tdencd to it gtm bj riuftaioi' 
WBhb, at p. SM of his vocfc, aote 9l The Jitamgi ocean in the ix* or 
dalft Parra of the hL-Bh. vene 9960. A twdve yrm' drot^t oocMrrad 
dorii^ wktck ih« gre«t p*^ trtreUed hither aad llutWr in ae«rch of taoi, 
oad thtt* hkrt the Vciu («mAmi kdndi pmiUmmm MdM wtdm Tmiirtiw) 
The Mmai Sinsrau vu thoat to deport •bo^ hot ww Ji— ailtd hf iho 

fmm^ oa^iaAto do AUroai.- "Go ao* heaee. oqr md : I viH ghrv 
tbco OMtflMit fahlbr food.' He oocordii^jr if wa eit . "fgM tr i iu g hb 
I tho VoJhJ* fanMM Mrf^^b A^mmI- m 



I to «ht othar i«An at ihar 


SECT, ni.] 




}^^ ^>TT^; I w if^T ^fwr ^rr f^^^tt^^: ^t; ii 

^fW\ ^ f^Rlf^ ^frl^rtSflMI "Tbe tract, fnsliioued by 
the go»la, which liea betweea tlie two divine rivers Stirasvati 
and Drisliadvati, is called Brahmavartta, The usage relatinfj 
t«j castes and mixed castes which has been traditioiuilly received 
in that country, is called the pure usage. The country of 
Ktiriikshetra, and of the Matsyas, Pauciialas and l§ura.senas, 
■wliich adjoina Brahmavartt'v is the land of brahmarshis (divine 
rishis). From a Brahman born in tliat region let all men upon 
earth learn their respective ditties. The tract situated betweea 
the Himavat and the Viudhya ranges to the east of Vinasana, 
and to the west of Prayaga, is known as the Madhyadesa 
(central region). The wise know as Aryavartta the country 
which lies between the same two ranges, and extends from 
the eastern to the western ocean. The land where the black 
antelope naturally grazes is to be regarded as the proper region of 
sacrifice; beyond that limit lies the country of the Mlechhas, 
Let twice-born men be careful to remain witliiu these (specified) 
countries. But a Sudra may dwell anywhere, when compelled 
to seek subsistence." 

From this passage it appears that at tlie period in question 
the Brahmans had not ordinarily penetrated to the south of 
the Vindhya range, though adventurers might have visited, or 

E E 



[«mtr^ ^ 

hermits might have nettled, in those regions. And erea to tlie 
north of the Vindhya we find the countxy <listrihut€d into aevenJ 
tracts more or less holy, according to their distance from tbe 
hallowed spot in the north lying on the banJu of tbe Saimsrtti. 
First, then, we have this small region itself, firahmararttJL 
Tliis name may signify (1.) either the region of Brahma, the 
creator, in which case it may have been r^arded as in some pe- 
culiar sense the abode of this god, and possibly the scene uf tL« 
creation : or (2.) the region of devotion or the Vedas {brattma), 
and then it will rather denote the country which was sanctified l»y 
the performance of holy rites, and tbe study of sacred literstme. 
Next in order we have the land of the BrahmaTshis, comprising 
Kiirukshetra (the country west of the Jimma and stretching from 
the Sarasvati on the north towards Vrindarana and Mathari)^'* 
witli the country of the Matsyas, Panchalas and SiuBflenaa 
The Panchalas are said by Kulluka Bhatta, the commentator 
on Manu, to have occupied the couDtry about Kanyakabj* or 
Kanauj ; and the $uraseuas to have lived in the neighboorbood 
of Mathjira." The third tract called Madhvade^ embraees a 
wider area, and stretches, north and south, from the Uimolay* 
to the Vindhya range, and north-west and south-east, from Vina- 
lana where the Sarasvati disappeared in the desert (see abore, 
p. 415) to Prayaga or Allahabad. The fourth region, AryS^ 
varttay or tbe abode of the Aiyas, is yet more extensive than 
the last, and extends within the same limits of latitude, from 
the sea at the mouth of the Indus to the bay of Bengal (tlie 
caaton and western oceans). 

The manner in which theee seTeral cotmtries are h&e sue* 

oeGRTcIy introduced £ie«ms to intimate that the Airas proceeded 

gradually from tbe banks of tbe Sansrati (where, as we have 

^•uppoMd, ^y had eGtaUidied t&ansdves after trayetsing tbe 

BM^)abk) towards the east and south; and that tbe coontnea 


8m tU aait oTaMMBt lai», is LaaxB, lad. AM. v«L SL 


. in.] 



forthost to tLe eouth and eAAt, ae well as to the eouth-weet, of 
the Sarasvati were tLoBe with which they had become last 
act|uaiiited, and were least familiar. Another proof that the 
Ariona had not yet penetrated to the south of the Viudhyn, or 
thoroughly occupied the eastern provincee of northern India, at 
the time to which we refer, may be found in Manu, x. 43, 44 
(eee Part First of (his work, pp. 1V7 and 182), wliere we are 
told that the Pauijijrakaa, ()(Jra8 and Dra\'i(Jas wiio had formerly 
been Kshatriyaa, had from neglect of religious rites and tlje 
abeence of Brahmans, sunk to the state of Vnshalas, or Sudnw. 
?Vom this it is clear that Bome at least of tJie people of those 
countries, *. e, of the inhahit^mts of Bengal proper and of 
northern Orissa, as well aa those of the Coromandel coast,^ were 
then living without Brahinanical institutions; and consequently 
that the Brahmans had not yet taken complete possession of all 
those provinces ; though, if the definition alwve given by Manu 
of the word Arydvartta may be relied on ae proof, the Arj'as 
must have carried their conquesta as far eastward aft the eastern 
ooean, or Bay of Bengal. 

In the Satapatha-brfihraana we find the following remarkable 
legend, to which attention was first drawn by Weber (in his Ind. 
Stud., i. 1 70, S.), regarding the advance of the Brahmans, and 
the spread of theii- religious rites in an easterly direction from 
the banks of tlie Sarasvati (S.-P.-Br., i. 4. 1. 10, fif.) : — 

'* Sec Wilson's Visbnu-Puruna, pp. 190, 192, and notes. 

K X 3 




^rrw ?f?i^t^t ft 'fm^r^ ?jipmV fli^^rg ♦ii^^r 
T^fw ! ffff t^ttN ■Rrft^ «<tj<ri ^ti[Wt: i jt^ ^ '^- 

%rTf^ %^d<f*<^ stimuli ^ f% ^;5TO^ Wd%^T\l 
^rf^ 'JlVlJii %^T^ T^fiA^C ^FT^raf^ TTT^ ^^trrrSTfrKAJl 

ijfw ^Hii*r<iii|i H iV^r^ f^^^ m^R: wt^ v^rf^ 

" Jlatliava the Videglkci** bore Agni Vaisvauara in his mouth, i 
The Riwhi Gotama Rahugana'* was his priest {purohita).i 
Though addressed by him he (Mathava) did not answer, * leefr) 
(he said) Agni (Fire) should escape from my mouths' The] 
priest began to invoke Agni with verses of the Rik ; ' We kindle I 
thee at the sacrifice, O wise Agni, the sacrificer, the luminons, 
the mighty, O Videgha.' (R.-V. v. 26. 3.) He made no answer. 
[Tl»e priest then repeated,] • Thy bright, brilliant, flaming) 
beams and rays moimt upwards, Agni, O Videgha ' (R.-V. viiL 
44. 1 6). [Still] he made no reply. [The priest then recited :] 
* Thee, dropper of butter, we invoke,' «S:c. (R.-V. v. 26. 2). 
S«) f:ir ho littered ; when immediately on the mention of butter j 
{i/hrita), Agni Vaisvanara flashed fort-h from his mouth : 


* Afterwards prakritized to TldeLa ? 
" See R.-V. i. 78. 5. 

S£CT. III.] 



could not restrain him, 8o he issued from his mouth, and fell 
ilovm to this earth. The Videgha Mathava was then on [or 
in] the Sarasvati. [Agni] then traversed this earth, Inirning 
towards tiie cast. Gotama liahugana and the ^'idegha Mathava 
followed after him as he burned onward. He burnt across all 
these rivers; but he did not bum across the Sadanii-a, which 
descends from tlie northern moimtain [the Himalaya]. The 
Brahraans formerly did not use to cross this river, because it had 
not been hurat across by Agni Vai^vanara. But now many 
Brahmans [Hve] to the east of it. It iised to be uninhabitable, 
and swampy, being untasted )»y Agni Vaisvanara. It is now, 
however, habitable ; for Brahmans have caused it to be tasted by 
sacrifices. In the end of summer this river is, as it were, incensed, 
being still cold, not having been burnt across by Agni Vaisvanara. 
The Videgha Mathava spake: ♦ AiMiere shall I abide?' [Agni] 
replied, ' Thy abode [shall be] to the east of tliis (river).' This 
stream is even now the boundary of the Koalas and Videh:is ; 
for they are the descendants of Mathava," 

In this passage the gradual advance of the Aryas, with their 
Brahmanical worship, from the banks of the Sarasvati eastward 
to those of the Sadanlra, and afterwards beyond tijat stream, is, 
as I have said, distinctly indicated. At the time when the {^ta- 
patha-brahmana was composed the Brahmans had already (jis 
Weber remarks) dwelt for a long time, in the country beyond 
the Sadanira, and it had then become a principal seat of the 
Brahmanical religion; as Janaka, the king of Videha, appears in 
that work with the title of sami'dt, or emperor. Thus (."^atap.- 
Br. xi. 3. 1, 2.) we have the following notice : " Janaka the 
Vaideha put this question to Yajnavalkya. * Dost thou, O Ya- 
jnavalkya, know the Agnlhotra sacrifice ?' '0 monarch,' he replied, 
* I know it.' "" Professor Weber makes the following remarks 

** d. F. Br. p. 846 : Tml ha etaj Janako Vaideha Yajhavalkyam pnprachha 
vetiha ahnihotram YtijAuBaUtj/d iti veda tamrml iti. See also 6. P. Br. xi, 
G. 2. 1, p. 87'i, where Janakn is addressed by tho lirahmaus hy tLe same title. 




IcnAT. m.] 

on the legefld I have just ({uoted : *• Under the name of Agni 
Vaiivanara (</m5 Fire which fcicnw for cdl men) the sacrificial 
worship of the lirahmans appears to Ihj iatended- The part 
which in the legend the priest plays in reference to the kixig, ia 
mysterious ; but I understand it to mean that he compelled him 
to propagate the Arian worship towards the east. The Sadiairi 
presents on obstacle, not from ita magnitude or the difficulty of 
oroasing it (which the Ganges and Jiimna must have previoudy 
done in a greater degree), but from the inhospitable character of 
the territory beyond ; for the word ardvilaram, * somewhat 
flowing,' designates tlie nature of the tract as an inundafeed 
Bwarap." Even after Videgba M^thava had a<ivanced mtmm 
the river, the firahmans (as the Arians are here called) appear 
to have remained for a long time on its western bank, said to 
have only crossed in greater numbers aft*jr the king with his 
people had cultivated the country ; until, in the time of the 
i§atap.-6r., it had attained suck a flourishing oonditiou, that the 
tradition of its oceanic origin was only indistinctly preserved." 
lud. Stud. i. pp. 178, 179. 

In the Vocabularies of Amara Siuha, i. 2. 3. 33, and Hema^ 
chaudra, 4. 130, Sadanir^ is given as a synonym of Karatojra, 
a river in the north of Bengal Proper. But as the Sadauira is 
in this passage described as forming the boimdary between the 
Koi^alas and Videha.s, or the countries of Oudli and North Behar, 
it seems that the river at present called the Gtmdak must be 
meant. — Weber, as above, p. 181. 

Lassen (Zeitschrift for 1839, p. 22), quotes, for another pur- 
pose, the following passage from the M.-Bh., in which tiie name 
of the Sadanira occurs ; but it throws little light on its position. 
" Departing from the Kurus (from Indraprastha) they 

"' In illustration of thia a line of the M.-6b. ii. 1078, is referred to i 
by Weber, which stales that BLlmasena, in the course of his conquest 
oC the eastern country came to the territory bordering on the HimavBt, j 
which was of aiiueoas origin : iaio Uimacaiah parivmn tamabhtfetya jtAtd'X 


8SCT. IV.] 



through the middle of Kurujangala, and came to the lovely 
Padma lake. Theu passing Kalakuto, they crossed successively 
on one moiintain (or In Ekaparvataka?) the rivers Garulaki, 
Mahaiooa, and Sadanira. Having theu crossed the beautiful 
Sarayu, and seen the eastern Ko^la, they crossed the river Mi'da 
Charmauvati, and came to Mithila."*^ In this passage (if any 
order has been preserved) it will be noticed that the Sadauira is 
placed between the Gaud;iJci and the Sarayu, and so to the west 
of the latter river. Ita position does not, therefore, seem to be 
well defined in the ancient Indian authorities. This, however, is 
«)f little consequence for our present, as any uncertainty 
in regard to the precise loc^Uity of tlio river Uoia not obscure the 
plain and express purport of the legend, viz., that the Brahmana 
with their worship advanced from the Sarasvati eastward to 
Behar and BengaL « 

Skct. TV.— Advance of M« Ariam from tkt Doab aero** the Viudhya 
Mountaiiu ; and their conJUcts with the aboriginal tribe* of the Dekhaii. 

It is not essential for the object which I have in view to attempt 
to trace with any precision the different stages in the progress of 
the Aryas to the east and south, which a review and comparisou 
of the data supplied by the Brahmanas, the Kamayanaand the 
MiUifibhai-ata may enable the careful investigator to determine^ 
and to refer to particular periods. 

As it ifl only necessary for my argument to prove that they did 
advance from the north-west to tlie east and to the south, juid 
that in so doing they came into contact with alwriginal races who 
had been in previous occupation of the country, it will suffice for 
this purpose if (after the foregoing notice of their progress to the 
eastward) I now pass on to that great southward movement, 

M M..Bh. ii. 793. Kundthjiuh pnuthitM te tn madhy«na KungSmgalam | 
Rtmvotn PatimoMtro gatva Kulakulam aUtyaeha j Gottdahlncha Muhu' 
tonam Sttdanlram tulhaioa cha \ Ehaparvatahe nadyah kramennili/acrujanta 
te I UUiryit Simitjum ramyam dfithted purcahcha Kohlim | AtUya jagmur 
MUhilam MOldm Churmanvaiim nadim || 

E B 4 




KUAr. uu 


of which we can discern the indistinct outlines in the poetic 
and hyperbolical narrative of the Ramayana. 

*' The Riimayana," remarks Professor Lassen (Ind. Ant- i. 534 ), 
*' in the action of the poem, designates, for the most part, only 
the north of Hindusthon as Arian. It represents Mithila and 
Anga in the east as Arian countries ; and regards the Kekajas 
in the west, thuugli dwelling beyond the Sarasvati, as a pore 
Arian race; aiul to this tribe one of the wives of king Da^iaratha 
belongs. Among the persons who were to be invited to tb« 
sacrifice of that monarch are the following : " Ramayana, 
(Schlegel) i. 12. 20, flF,, filfinirTt^RtH "S^ ^PT^ X^f^ 

^iTTT^^t^ tirfi^iiT'l^i ^rf^wnjT^ -ii^tfjig ^wT^n^^ 

TT f^T^II " [Bring] Janaka, the heroic king of Mithila, of 
Btubljoru vtdour, versed in all the Saatra*, and in the Vedas. . . . 
Bring also the aged and very religious king of the Kekayafi, the 
father-in-law of the lion-like king, together with his son ; and 
Lomapada, the devout and god-like king of the Angas, paying 
him all honour. And bring speedily all the eastern, the Siudhu- 
sauvira, the Surashtra, and the southern monarchs." 

The word "southern kings" may, Lassen says, be em- 
ployed here in a restricted sense, for from other parts of the 
poem it appears that the country to the south of the Viudbya 
was still unoccupied by the Aryas. Even the banks of the 
Ganges are repr&sented as occupied by a savage race, the 
Nishadas. Thus Ram. ii 50. 18, ff. (Schlegel's edition) (ii. M 

At n tr :_ n :.'^ -j:i.: \. » _■ . . — ■ _^^ 


47. 9, ff. in Gorresio's edition) : Tf^ 



«<icW*i*<: ^WTi f^^T<rwrafT ^^RT^ ^^"Sf?! f^^:M 





^Tf?ffH^liyil*l*(ft^ll " There [there was] a king called Guha 
of the race of the Nishadas, an intimate friend of Rama, and re- 
nowned aa a powerful chief. He, hearing that the eminent 
Rama had come to his country, approached him attended by his 
aged ministers and relations.'* This chieftain provided a boat to 
ferry Rama with his wife and brother acrosa the Ganges (Ram. 
ii, 52, w. 4 — 7 and 71, fF.) : and afterwards attended on his 
other brother, Bharata, when he also passed the same way. 
(Ram. ii 83. 20, and 84. 1, 10, &c., &c.) 

In the same poetical narrative^ the Dandaka forest is repre- 
sented as beginning immeiliately to the south of the Jummt. The 
whole country from this point to the Godavaxi is described as a 
wilderness, over which separate hermitages are scattered,** while 
wild beasts and Rakshasas everywhere abound. ** The Rama- 
yana," says Lassen (i. p. 535), " contains the narrative of the first 
attempt of the Arians to extend themselves to the south by con- 
quest ; but it presupposes the peaceable extension of Brahmauical 
missions in the same direction, as having taken place still earlier. 
Raraa, when he arrives on the south of the Vindhya range, finds 
there the sage Agastya, by whom the southern regions had been 
rendered safe and accessible. Agastya appears as the adviser and 
guide of Rama, and aa the head of the hermits settled in the 
south. In this legend we cannot but recognise the recollection 
that the south was originally a vast forest, which was first 
brought into cultivation by Brahmanical missions. The Ra- 
ksbatias who ore represented as disturbing the sacrificfes and de- 
vouring the priests, signify here, as often elsewhere, merely the 

** Kam. iii. 6. 1. (Gorresio) : PravUan ta maharanyain Dandakaranyam 
vttnmam \ Dadiiria Ramo durdhariham tiipaiairama-mandnlam \ Rama 
alludes to the vastness of the forest, iii. 15. 33: Na tu junnmi tarn deiam 
vanasywiya muhatlnyA\ Yatrairama-padam pnnyam nuiharthes tatya dhi- 
matah | " From the vastnessof the rorest, I cannot discorer the spot where 
the sacred hermitage of the great and wise rishi exists. 





vngi tribes vhich placed themselrea in hostile oppoiiliiBlij 
the Biahmanical institutions. The ooly other actor9 wbo ippwl 
in the legend in addition to these inhabitant^:, are the ma 
which sdly themselves to Rama, and render him asE 
Tliis can only mean that when the Arian K^hatriyas fint 
hostile incursions into the south, they were aided by i 
portion of the indigenous tribes. Ratua reinstates in 
of his ancestral kingdom a monkey-king who had been i 
and in return receives his assistance." 

The following are some of the passages of the HamaTsna it ^ 
which the proceedings of the Rakshasas are described. TTw 
idea of the monstrous characteristics which are assigned to Ui« 
gigantic demons may very well have been borrowed from tfce 
barbarous tribes whom the Brahmauical anchorites found in oocn- 
pation of the forests, and from whom they would no doubt wffpr 
continual molestation and cruelty. These savages with wfaca 
as we have alrejuly seen, p. 409, ff., the Arian Indians had been 
femiliar in the regions further north, had, even in the Vedic era, 
been magnified into demons and giants by the poetical anil 
superstitious imaginations of their early bards. The hermits 
in the neighbomrhood of Chitrakuta, thus represented to Ram» 
the sufferings to which they were exposed ; Ram. ifi. i. 1.5, ff: 


^I<l^ T^f% ^*i*M^ ^^K*lH:il %^<lriHt><<lQg- 

" Men-devouring Rakshasas of vorioua shapes, and wild beasta 
£or serpents] which feed on blood, dwell in this vast forest. They 
harass the devotees who reside iu the settlements, and slay them 
in the forest: repress them, Kaghaviu . . . These shapeless and 
ill-looking monsters testify their abominable character by varioiis 
cruel and terrific displays. Tliese base-born (andi'ya) wretches 
implicate the hermits in impure practices, and perpetrate the 
greatest outrages. Changing their shapes, and hiding in the 
thickets adjoining the hermitages, these frightful beings delight 
in terrifying the devotees. They cast away the sacrificial ladles 
and vessels, they pollute the cooked oblations, and utterly defile 
the offerings with blood. These faitldesa creatures inject fright- 
ful Boimds into the ears of the faithful and austere eremitea 
At the time of sacrifice they snatch away the jars, the flowers, 
the fuel, aad the sacred grass, of these sober-minded men. 
Seeing that the hermitage is infested by these wicked monsters, 
the devotees take counsel with thee, desiring to depart else- 
where. .... It is not expedient for thee to tarry here alone 
with thy spouse, iu the neighbourhood of these cruel Rakshasas. 
Thou art indeed able to destroy them ; but be not too confident^ 
for they are a treacherous race." 

Proceeding on liis journey through the forest, Rama encounters 
Viradha, a Rtlkshasa, who is thus described, Ram. iii. 7. .'5, tf. : 

^'Miwf^T^^i Ath*{\M f^Mi^ <Hto ri^<ri<<*<^i 

f^ arsi^ 'g flifr ?f^i ^^r^ ^ >iirw wfii 

^fvfVf^^t "^T^ *j4^riMi ^lTlH-irii^|Tj**i^|| "He 
beheld the Rakshasa of terrible aspect, like a mountain peak, 
with long legs, a huge body, a slayer of wild beasts and serpenta, 
with a crooked nose, hideous eyes, a long face, a pendent belly, 
bearing on the point of a spear eight lions dripping with blood, 
and the huge bead and tusks of an elephant smeared with fat, 
clad in the bloody skin of a tiger with the feet attached, an 
object of terror to all creatures, like Death with open moutL** 
This demon, who was slain by Rama, turned out to be a 
Gandharva, who by a curse had been transformed into a 
Rdkshasa; but now on his death, regains his primeval form. 
He, nevertheless, requests to be buried after the manner of the 
Kakshasw; Ham. iii. 8. 19: ig^ -^[Tf^ ^ JJJf Tff^qjR 

^ The Nisliadiu also are described in the Puranas as very black and a^H, 
but differ from theliiikshasas in being vcrjr short. Sec Part First of this vorki 
pp. 62-64 ; and Wilson's Vishnu-Pur. p. 100. The Bbag.-Pur. it. 14. 43-46, 
thus describes thetn : — " Viniichityaicam rithayo vipannasya mahipateh 
mamunihur urum tarata tatriUld hiihuko varah \ kuhakfithno 'tikratvdngo 
hnuvabahur mahdhanuh | Ilragvapad nimnanutagro roktSkshtu tamnoKSr' 
dhajah I Tarn lu te 'vanatam dinam Aim kdromtti vadiiiam \ nitkidety ahnaant 
tata ta nighadas tato 'bhavat \ Tatya vaniastu naUhudd giri-kanana-gochardk | 
"The rlshia having thus resolved, hastily rubbed the thigh of the defunct 
king (Vena), when there issued from it a servile man, black as a crow, very 
ehort in limb, with short arms, large jaws, short feet, pendent nose, red 
eyes, and copper-coloured hair. This man, humble and bowed down, asked 
them what he should do. They answered, " Sit down " {nishida), and he, 
in consequence, became a Nithdda. His descendants are the Naishadas, 
who dwell in hills and forests." We are informed by Prof. Wilson that the 
Padma Parana (Bhu.-Kh.) ''has a similar description, adding to the 
dwarfish stature and black complexion, a wide mouth, large ears, and a 



M<.*HM! II " And, Rama, cast tliis body of mine into a trench ; 
for such is the immemorial custom in regard to deceased Ea- 
kshosas; such of them as are so interred, attain to worlds of 

happiness. Accordingly, Saumitri (Lakehmana) raised 

up the body of Viradha, resembling a mountain, and diig a 
deep trench, in which he buried it."*' This may allude (aa 
Weljer remarks, Ind. Stud. i. 272, note) to a difference between 
the funeral rites of the Brahmanical Indians and the aboriginal 

The following are two further passages in which the Ra- 
kshasaii and their oppression of the anchorites are described. 
The sufferers, it appears, assert that they possess the power 
of ridding themselves of their enemies by their superhuman 
facilities t but these faculties they do not choose to exert for 
the reason assigned. Ram. iii. 10, 16, ff.: wVs^ ^^ff^- 

^TfT^t TT^ T^tfH ^^TT WSVJ ^11 q*-Mli^^|(%- 

^ l »<t<l*jj Tp^Tf%'ft*lftH R^|i^il«!l*ll«1l^ fwrq^ ^»^ 

Tf^rf II " This large company of hennits, principally Brahmans, 
being oppressed by the Rilkshasas, has resortt-d to thee for 
defence. Come, Rama, and behold numerous bodies of tlie 
meditative nmnis, lying slain by the Rakshasas in many parts 
of the forest, A great slaughter is being perpetrated of the 
dwellers on the Pampa, and the Mandakini,*' and the residents 

*" Id the Bcqiiel, bowerer, Riima encounters another Rokshasi, whom be 
burns on a funeml pile. — Ram. iii.. sec. 75, verses 45, 50, IT. 

"* Here it will be observed that the name of a river in the IliqiiilBya (a 
br.auch of the Ganges, sec Part First of this work, p. 187, note) is iipplicU 
to a river in the centre of India. ITiis illustrates what has been said above 





on Chitxakuta,'* Hanj. i"- H. 12, ff. : ^«{«|^ S fil ^l ^ l iwi 

•^ f^^HfH*if% i^Y^ xrr»TT »Tt?r:n ^rnt 7r:iwtit 

^T^f^ ^^m^w ^liM riM^-^' ^^ %^ ?T«rai ^ 
■5rr4 1 »Hi(44V H^mwrg Tr^:ii "At the time of 

offering tbe ajn'tfiotixL sacrifice and on festivals, the fierce, fleeh- 
devouring Bakshasas ineiilt ue. Thus liaropsed, the devotees 
find, on consideration, Uiat they have no resource but in your 
assistance. It is true that by the power of our austerities we 
could slay these goblins ; but we are unwilling to nulUfy the 
merit which has been earned by long exertion. The acqnisi- 
tiou of such merit is ai-duous and attended with many obgtade« : 
it is on that account that, though exposed to be devoured, we 
abstain from launching curses against our oppressors." The 
utterance of a curse, it appears from this passage, was an act 
prejudicial to the sanctity of bim who pronounced it. Sita, 
however, thinks that her husband Ruma has no right to pro- 
tect tbe devotees by slaying the RaksbasaB who wer* not in a 
state of hostility witli him, and had done him no injury." 

It does not appear, however, why the aid of Itama should 
have been so earnestly invoked, as the sage Agastya appears to 
have been perfectly successfiil in keeping the Kakshasas imder 
rotitiriiut. His prowess is tbiu? described; Earn, lii, 17. 17, ff.: 

^Pii^ Tf?T ^: WT^ ^rti y^3f T^r^\ ^\^M] jus^ 

(pp. 357-3^8 and 361) about tlic BpplicAtion of the same name to diCerent 

''" Iiuin. iii. 13. 22. — }l€h«hn«uiii!m rfna tairam ladho tiramt ]/ujyaU\ 
AjMTuilhud file nupi hanlaryu Rakihatiit ttuya 


r. IT.] 



?l^r1<l^*<M< TWT^T^ V^ TT^:i f^W jfWOT 
7TT7T T^ 'TttT»pif^|| <(<l«lff% -^itftMl f^fM 

^■$r^T ^<*4f>f:il S^tVT<l^ 'T?^ ^i^TfT'T, HT^iT^ 

liermitage of Agastya, renowned in the world by hie holy acts, 
(that hermitage) which offers relief to the wearied, is now in 
[This is tlie sage] who has restrained death by the 


power of his austerities, and who, through his benevolence to 
mankind, has rendered the southern regiuiis perfectly secure 
(see above, p. 425). This is the liermitage of that saint by 
■whose might it ia effected that this southern region is only 
gazed upon, and not possessed, by the Rakshasas. Ever fsince 
that holy man has visited this region, all the goblins have be- 
come subject to him. Through the name of this saint this 
Rfmthem country has become prosperous, and renowned in 
the three worlds, as secure against the gaze of the cruel. The 
Vindhya range, which in its wrath had grown to a great height, 
vieing with the mountain of the sun (Mem), now, submissive to 
this sage's command, increases no farther. He too swallowed 
the ocean with all its monsters, when he had been propitiated 
by the gods with Indra at their head, to destroy the Dana- 
vas. .... No liar, or] cruel, fiendish, impure, oppressive, or 
wicked man may dwell here." 

In the preceding sect. 16. 13, ff., the destruction of two 
Asuras called Vatapi and Ilvala, by this sage is described : 

" Formerly the cruel Vatapi and Ilvala, two brothers, who were 
great Asuras and slayers of Brahmans, lived together here. The 
ruthless Hvala assuming the shape of a Brahman, and speaking 
SanKkrit, invited the Brahmans to a funeral oeremony." Urala 
served up to them his brother Vatapi who had been transformed 
into a ram ; but after they had eaten him, he called him lack, 
wlien he issued forth, rending asunder the Brahmans' bodie& 
Thousands of Brahmans were killed by them in this way. 
Agastya however came to the place, and devoured Vitiipi 
according to his brother's invitation, but would not allow him 
to issue forth again; and burnt up Ilvala by the flash of his 

Agastya is again spoken of (Ram, vi. 100. 15, 16) as the 
comiueror of the south :«' firf^rTT^ jp^lT W^ HMIHfl l ^ - 

irfwri ^TT^ ^TTV^T ^5^^ Tf%^^ f^ll " Thou 
hast," said R;inia to Sita, "been conquered by me from the 
hand of the enemy, as the inviolable southern region was by the 
invincible Agastya." 

Vibhishana, the brother of Havana, is representetl by his sister 
iSiirpanakha, in her interview with Kama, as having abandoned 
the practices of the Rakshasas." Can this allude to some of 
the southern tribes or chiefs, who allied themselves to Rama, 
having adopted Brahmanical usages? Vibhishana eventually 
desexts his brother and is kindly received and embraced a^ a 


** An explanation of this legend is suggested by Weber, Ind. Suid. i. 475. 
He thinks it mfty parlljr have taken iu rise in the rpmonil)raDce of sonie 
cannibals living in the Dekhan. 

•' On Agastya see the quotation from Lassen above, p. OS ; and 
ColdwelPs Druvidian Ciraminar. 

•• Ram, iii. 23, 38. — Vibhiskmaieha dharmatma Bakihatachuratarjtlah \\ 

CT. ivj 



friend by Ramn,.'* In the Ramopakliyana in the III. Book 
of the ]VL-Bh., verses lo,913 — 18, while Eavana asks Brahma to 
make him invincible by Huperhuman beings, Vibhishana, on the 
other hand, prays, " that even in the greatest calamity, he may 
never incline to imrighteousneas, and that the Brahnianical 
weapon may appear to him a thing he had never learned to 
wield."®* He thus indicates hia submiijsive disposition towards 
the Brahmans. 

The Rakshasas are described by Khara, one of their chiefs. 
Ram. iii. 28. 18, as being " of fearful swiftness, unyielding in 
battle, in colour like a dark blue cloud."" 

Khara himself is characterised by Rama as the " perpetual 
enemy of the Brahmans," *' as " cruel, hated of the Brahmans, 
ilevc»id of righteousness, and wicked." ^^ Ravana is stigmatised as 
A ** destroyer of religious duties, and the ravisher of the wives of 
others ; " *** as " having frequently at the sacrifices and obla- 
tions polluted the Soma which the Brahmans had offered with 
hymns ; " and as a "destroyer of holy sacrifices, a killer of Brah- 
mans, and a being of wicked life."'' Nevertheless, to inspire 
confidence, RAvana approaches Sita pronouncing the Vedas, 
Ram. iiL 52. 20. "» 

•* Bdm. ▼. 91. 20: Tuficha Hamah samutthapya paruhvajya cha IlukMha- 
Itm I Uvticha madhuram vukyam inkha mama hhavCtn iti \ \ 

*• M.-Bb. iii. 13,918: Puramiipadgaiasyupi ruuOiarme me tnatir bhavel] 

HithiUificha hhagavan brahmuslrum pralibhatu me \ lu Tersic 15,897 Vi- 
bhLshiinais itj\cd dharmagoptu kriydnitih, "a protector of riglileoaraeffi, and 
devoted to religious rites." 

*^ liakthuum hhlmavegunam tatnurethvanivarltinam] NUajimuta-varnii' 

flHW, &C. 

^ Ruin. iii. 3.7. 68 and 100: ^uioad hrakmanakanlaha. 

" Ibid, verse 70 : Kruriitman hrahmavidvishUjt tyahtudhannu tupapakrit. 

** Ibid. 36. 11: Uehhetarahcha dharmuniim paraduriibhimardaaam. 

** Ibid. 36. 11, ff. : ManlrairnbhihtUum ptirvam adhvarethu dvijatibhih \ 
Ilarirdatieshu yah Soinain dharshaydtndta naikaiah | . . Punyayajnakuttam 
kriiram brahmaghnam duxhtachiirinam \ 

^^*> Bruhmu-ghotham udiniyan j In the M.-Bb. iii. 15,981 ibctonB ofVaiSra- 
vnna i. r. Ilavnna and bis brolbers, are said to have been originally tarre 

¥ F 




xr. m. 

Under the deBignation of moakejs, again, which plaj M 
important a part in the I^amayana, we may have another dan 
cif the aborigines, who allied themselves to the Brahmsuu, and 
embraced theLr form of religious worship. In Earn. iii. 75. 66, 
it is said that " Sugriva, though a monkey, is not to be despised, 
as be is gratefiil, can change his form at ^ill, and is actire io 
aiding his friends." '°^ And we ai'e told that at the inaugunUdoa 
of this same Sugriva, who was reinstated by Kama io liis kingdom, 
from whioli he had been expelled by Bali, "the monkeys 
gratified the Brahmans in due form, and by proper distribution, 
with gifts of jewels, clothes, and food : after which these men 
skilled in the Vedic hymns, poured clarified butter, consecratai 
by sacred texts, upon the kindled fire, which had been placed 
on kusa grass." "** M 

The iiioiikej-s are de8cril)ed ns living in a cavern, (Kam. iv. 33. 
1, (F.,) which Lakahmana is represented as entering to convey* 
message of remonstrance to Sugriva for his tardiness in aiding 
Riiuia. The cavern, however, is a cave only in name, a«, in the m 
usual style of later Indian poetiy, it is depicted as filled with 
gardens, woods, flowery tliickets, palaces, templt-s of the god«, 
(devatdndm niketdtiMta) ponds, a mountain stream, &c. This 
feature of monkey-life (their occupation of a cavern) may be 
either purely poetical, and intended to be in keeping with their 
other characteristics, or it may have reference to the rude habits 
of the aboriginal inhabitants of the southern forests. 

It appears to be more probable that the extravagant descrip- | 
tions of the gigantic and sylvan inhabitants of the Dekban 
which I have just quoted, should have originated in some actual 
and hostile contact with the savages who occupied the then un- 

wdttvidah Urah tartc micharitavratak, " ail of tbem learned ia tLe Yedss, ^ 
bolii, and atleotivG to religious rites," H 

'^' ya tu te M 'mmnntacynh Stigrivo vt'tnaro 'pi »au \ hfitiijiiuk kumarupi 
eha tahuydrthe cha krityavan 1 See above, p. 166. 

"* Ram. iv. 25, 27, 28 : Tatatle vanaruirenhth'i yathahhugam yathaoidki 
Tlatsiair tattraifcha bhalighyaii cha loshayitvd drijiirnhiibhfm | Tatak kuiO' j 
ptiristimnm sumiddhaw j<iUtvt<lasam | Afanlrnj'filciia havithn hntvu munfravuial 
' ^h if 




cleoxed forests of that region, than that tb^ should be the 
simple offspring of the poet's imagination. 

It is certain that the description given of the RakBhasos in 
the Ramayana corresponds in many respects with the epithets 
applied to the same class of beings (whether we take them for 
men or for demons), who are so often alluded to in the Rig- 
vedn. The Riimayana, aa we have seen, depicts them as in- 
fcetiug the hermitages or settlements of the Arious, as obstructing 
their sacred rites, "" as enemies of the Brahmans, as eaters of 
men,""* aa horrible in aspect, as changing their shape at will, 
Ac, &c. In the same way the Rig-veda (see above, pp. 386, ft". 
and 403, ff. ) speaks of the Da«yu«, Rakshasas, or Yatudhan.ns 
is being "destitute of, or averse to religious ceremonies^ 
{akarman, avrata, apavrata, ayajyu, ayajvan), as "practising 
different rites'' {anyai>rata),Qa "godless" {adeva, adevayu), 
" worsliippers of mad gods " (^muradeva), " haters of Brahmans, 
or prie.'rte " {brahviadvish), as "inhuman " (^amdnusha), " fero- 
cions looking, or with fierce eyes" (ffhora-ckakahas), as " flesh- 
eaters" {kixti'ydd), "devourersof life," or " insatiable " {itsntrip), 
M "eaters of human and of horse-flesh," (R.-V. x. 87. 16 : Yafi, 
jHiurutheyena kraviafui aamankte yo a^'yena 'pcuSund yaiu- 
dhdnah)', as monstrous in form, and poeaessed of magical or 
superhuman powers. It i^ quite possible that the author of the 

"* In the M.-Bh. xiv. 2-172-74, the same hostile act which is so often 
assigned in the Rilina^'ana to Rakshasas, is attriliiited to a Nisbada. Arjuna 
is there said to have arrived in the course of his process to the south, in 
the country of Eksluvyo, king of the Nishadas ; and to have vanquished 
that king's son, who hod come to obstruct a sacrifice (ifitji'tavighuurtAam 

""' In the story of Gautama, already partially quoted, in p. 382, from 
the M.-Bh., the very saine epithet of " man-eater " {purufhiida) which the 
RSmaynna applies to the Rikahasits, is emplnycd to characterise the DIlsyu^ 
who arc regarded in the M.-Bk. meroly^as a tribe of savages, and not as demons. 
The Brahman who reproaches Gautama with sinking into the condition of a 
Dasyu, is said to have seen him "coming home with a bow in his hand, his 
limbs besmeared with blood, and in appearance like a maH-etiter," &c. (. . . 
dhahtukpunim dhriltiynilham \ Rudhirctiata*ibtiiiigmn grihadvtiram upiga- 
tum I 7Vi» dfithlvii pwruthiidabhtttn apadhvatUan k*kaj/iigatam, &c.) 

r F i 




Ramnyana may have borrowed many of the traits which be 
aacribe» to his Rakshasaii from the hymns of the Rig-veda, 

The last editor and translator of the Rainayaoa, Signer 
Gorresio, writes as follows in regard to the fabulous races with 
whom that work has peopled the Dekhan ; (Notes to voL vL ppc 
401,402): "The woodland inhabitants of India south of the 
"N'indhya range are called in the Ramayana monkeys, in con- 
tentpt, I conceive, of their savage condition, and also, perhaps, 
because they were little known at that time. In the same 
way Homer related fabulous stories about the races who, in hi» 
a^'e, were unknown to the Greeks. The occupants of the Dekhaa 
difft'red from the Sanskrit-speaking Indiana in origin, worship, 
and language." And in regard to the Rakshasas he obserres, 
p. 402 : " The author of the Ramayana has no doubt, in mythicRl 
allegory, applied the hated name of Raksbasas to a barborout, 
people who were hostile to the Sanskrit-speaking Indians, and 
differed from them in civilisation and religion. Tliese Rakahaau 
were, I say, robbers or pirates who occupied the southern coaata 
of India and the island of Ceylon." In his preface to the lost 
Volume (the xth) of the Ramayana (pp. i.-ix.), Sig. Gorresio 
turns to tliis subject; and, after remarking that the Arian tri' 
on their immigration from Northern Asia into the Panjab, had 
to encounter indigenous races of a different origin,"** whom they 
partly drove before them, and partly reduced to servitude, h« 
proceeds to make a distinction between the savage tribes occu- 
pying the Viudhya and its neighbom-hood, and those further 
south. The first, whom the Ramayana styles Vanaras or inoD-»j 
keys, though they differed from the Aryaa in race, lauguagev! 
colour and features, must, he thinks, have shown a disposition id 
receive the Arian civilisation ; since they entered into league' 
with Rama, and joined in his expedition against the black tri 
further south. The greater part of the tribes south of tlie 
Vindhya also submitted to the institutions of the Aiyas; but 

"^ The same thing, Le remarks, hnppencd to the Semitic raoei aim, wh 
ranic inio contact with the Ilaniitic or Cii^hitic tribes come of them nearl 
furagc, na llie UepUivim nnd \W Ztt.wi.uu\\v\vw., ^twi. ii. 20. 




towardfi tbe extremity of the peninsula and in Ceylon, there was 
(Gorresio believes), a ferociuus black race, opposed to their wor- 
ship. To this race the Arians applied tbe name of Rokshasas, 
an appellation which, in the Veda, is assigned to hostile, savage, 
and hated beings. It is against tliis race that the expedition of 
Rama, celebratt-d in the Rfimayana, was directed. The Arian 
tradition undoubtedly altered the attributes of these tribes, 
transforming them into a race of giants, deformed, terrific, 
truculent, and able to change their form at will. But notwith- 
standing th&se exaggerations, tbe Rjimayana has ( Gorrt-sio thinks) 
preserved here and there certain traits and peculiarities of the 
race in question which reveal its real chaiacter. It represents 
these people as black, and compai'es them sometimes to a black 
cloud, sometimes to black collyrium ; attributes to them crisp 
and woolly hair, and thick lips; and describes them as wearing 
gold eai'rings, necklaces, turbans, and all those brilliant orna- 
ments in which that race has always delighted. These people 
are also represented as hostile to the religion of the Arya-s, aud 
as distiubers of their sacrifices. The god whom they prefer to 
all others, and specially honour by sacrifices, is the terrible 
Rudra or Siva, whom Gorresio believes to be of Hamitic origin. 
Their emblems and devices are serpents and dragons, symbols 
employed also by tbe Hamites.'"* Sig. Gorresio considers the 
story of Riima's expedition against the Rakshasas to be historical 
in its foimdatiou, though exaggerated by mythical em1>ellLsli- 
ments; and he observes that the Arian tradition has even pre- 
served the memory of an earlier struggle between the same tw(j 
races, as some Purauic legends relate that Karttavirya, of the 
Yadava family, a contemporary of Paraiurama, and somewhat 
anterior to the hero of tbe Ramayana, invaded Lanka (Ceylon), 
and made Ravana prisoner (Wilson, Vishnu Purana pp. 402, 

">* As Signer Gorresio has not supjilie<I any references to the pnuagcs 
inwhicli these various clinractcristics of the Uakshosai arc <lt.'scribc<l, I aiu 
unable to Tcrify his details. 

'"* Tlie btory is thus toUl m il)e Vishnu Puninti, iv. II. i : Aluhith- 

1 1 1S 


toi»»P- in. 

In regard to Signor Gorresio's views as above expoundad, 1 
will only observe here, that the aborigines of southern India are 
not generally regarded as of Hamitic origin ; but, as we shall 
Bee iu a eubset^uent Section, are considered by other philoiogLrta 
to be of Turanian extraction. 

Professor Weber is of opinion (Hist of Ind. Lit. p. 181), thi 
the principal chai-acters who figure in the Ramayana, are not 
historical personages at all, but mere personifications of certain 
events and circumstances. Sita (the furrow), he remarks»oooura 
both in the Big-veda,'" and in the Grihya ritual, as an ob- 
ject of worship, and represents the Arian agricultxire ; while he 
regards Rama aa the ploughman personified. The Ramayana 
has only, he thinks, a historical character iu so iar as it refen to 
an actual occiurence, the diSiisiou of Arian civilisatiou towkrdft 
the south of the peninsula. 

nriL ■ 

Sect. Y. — Indian traditions regarding the tribes in the sooth o/ the Ptwiuitb 

Having furnished some account of the advance of the Aryat 
into southern India, and of the races whom they there encoun- 
tered, according to the fabulous narrative of the Ramayana, I 
have now to enquire whether the other Hindu traditions offer us 
any more probable explanation of the origin and affinities of the 
tribes who occupied the Dekhau before ita colonization by the 

mall/dm digvijayabhyugato Narmada-jalavagahana'hfidd-nlpSnaniadiiknUwiy 
[ch ?'}ulemiica tena aicsha-fleva-duityd-gttntJhitrvem-jayutiUiula-madueaUpo 
'jn Ravannh paiuriva baddhnh tva-ttagnraihiiute BlhtipiUtk, " When, in 
the course of his campaign of conquest, Ravana came to Mahi^hmati (thd 
capital of Karttavirjn), there he who bad become filled with pride from his 
vifloric's over all the devas, duityas, nnil the chief of the Gaodharva*. w«j 
captured hy Kiirttavlrya (who was excited hy balhing and sporting in the 
Narmadii, and by drinking wine), and was confined like % wild beast in s 
comer of his city." Prof. Wilson (p. 417, note), states that, according to ^ 
the VSyu Pur., Kfirttivviryfl invaded Lanka, and there took Havana prisoner ; H 
but that the circumatnnces are more generally related as in the Vi*hnu 
"" Bi^-veda, iv 37. 6 and 7. 



Among the Dn»>fu txibes which, according to the Aitareya-brah- 
mann, vii. 18,'°* were descended from the RiBhi ViBvaniitra, are 
meutiuued the Andhras. And Manu, x. 43, 44,"" specifies tlie 
Dravidas among the tribes which had once been Kshattriyas, but 
bad sunk into the condition of Vrishalas (or f^udraa) from the 
extinction of sacred rites, and the absence of Brabraana. In like 
manner the Cholas and Kerahia are stated in the Harivaniki to 
have once been Kshattriyas, but to have beeu deprived of their 
Hocial and religious position by King Sagara."" lu the same way 
it appears that several of the Puranas, the Vnyu, Mataya, Agni, 
and Brahma, claim an Arian descent for the southern races, 
by making their progenitors, or eponyms, Pandyaj Karnata, 
Chola, and Kerala, to be descendants of Turvasu, a prince of 
the lunar line of Kshattriyas. (See the First Part, p. 53),"* 
Tiurvasu, the Puranas say, was appointed by his father to rule 
over the south-east. Thus the Harivauda relates: "Yayati, 
eon of Nabusha, having conquered the earth with its seven 
contiuent.s and oceans, divided it into five ]>ortions for lus 
sons. This wise monarch placed Turvasu over the south-east 
region." '" 

According to the legend, Turvasu, in common with most 
other of Yayati's sons, had Jeetined to accede to his fathers 
request that he should exchange his condition of youthful 

'*• Quoted in the First Port of this worV, pp. 84 and 178. 

"* Already quoted In the First Part, pp, 177-182, together with other 
panillel testa from the M.-Bh. 

"0 Seethe First Part, p. 182. 

'" The Ilarivnn.^n, sect. 32, verse 1836, substitutes Kola for Karna^H: 
Knntthumad atk' Akrl^ni chottSra* la»i/a ch" atmajdh | Pan^saicha Ktra- 
liiichoira Kolii Chiiliiicha parlhimh \ TmhSm jtuui/Miduh tphUiih PCmifijiii 
CholAh i^ahrraliih \ " From Kurutliama aprong Akrifja, who had four soii.i, 
PiTidva, Kerala, Kola, and Chola, who were the kings of the rich oountritis 
of Pfinijy*. Chola, and Kerala." 

"* Ibid. sect. 30, verses 1G16, ff. : Saptadvipam Yayutittu jitva pfilhrim «a-. 
tagaram\ vyabhajnt panchadbu nij'inptUrdm'im Nahiuiiai tadd ] Diii dukthinn- 
purvasyim Turvatam matimun probhnh j . . , . nyayojaynt \ \ 

I r 4 

vigour for his father's decrepitude, and was, in oonseqneoce, 
cursed by the old maa. The M.-Bh. i. 3478, ff., j^nves the 
following particulars of the curse : " Since thou, though Iwrn 
from withiu me, dost not give me up thy youth, therefore thy 
offspring shall be cut off. Thou, fool, shalt be king over tho«e 
degraded men who live like the mixed castes, who many in the 
inverse order of the classes, and who eat flesh : thou shalt rule 
over those wicked Jllechhas who commit adultery with their 
preceptor's wives, perpetrate nameless ofifences, aud follow the 
practices of brutes." "* 

The Andhras, Dravidas, Cholas, and Keralas, who hare been 
mentioned in the preceding passages as degra<le<l Kahattriyas. 
or as descendants of Turvasu, were the inhabitants of Telingana, 
of the centra] aud southern parts of the C'oromandel coast (or 
the Tamil country), and of Malabar respectively. It is evident 
that the legendary notices which I have just quoted do not 
throw any light on their origin. That these tribes could not 
have been of Ariau descent, I shall proceed to show in the next 
Section by the most satisfactory of all proofs, that derived from 
the language of their modern descendants. 

Sect. VL — Languaget of the SoiUh nf India, and tkeir fundamental diffrmiet 

from Sunskrit. 

As I have already intimated in the earlier parts of this 
volume, there exist in the vernacular dialects of northern Intlia 
many distinct remains of older languages, distinct from San- 
skrit, which are supposed to have been spoken by non-Arian 
tribes settled in that portion of the peninsida before the immi- 
gration of the Aryas ; and I have also alluded to the existeocts 

'-' Yatloam me hfidatfuj jaio xxtyah nam na prayaehhan \ tasmatpraja m- 
muehheditm Tureato tava yanyati \ Sanhirndchara-dharnushu pratilomachare- 
thu cha I PiiiUWxhn ch' untijeshu mi'idha raja bhacithyati \ Gurudam-pra- 
nakteshu tiryagyonigaieiihu cha \ Paiudharmi»hn pdpenhu Mkchheshu ttytm 
bhat>ishi/asi \\ In verse 353.1 Turvasu is said to be the progenitor of tlie 
Yavftiins, Turraitor Yiivnm'ih »mfitdJi j| 



of a class of languages in the south of India, viz., the Telugu, the 
Tamil, the Malayulim, and the Canarese, which are fundament- 
ally diflferent from the Sanskrit."* I shall now proceed to 
ntablish in detail the assertions I have made regarding these 
southern languages. 

Various savage trihes are still to be found among the hilly 
tracts in central India, such as the Gonds, Kols, «&c., whose 
language is quite distinct from any of the ancient or modern 
Prakrit dialects derived from the Sanskrit. It is not, however, 
necessary that I should enter into any details regarding the 
speech of these wild races. It will suflBce for the piu-poses of 
my argument if I show that the same remark applies equally to 
the far more numerous, and more cultivated tribes who occupy 
the Dekhan ; and that the various languages which are current 
in the different provinces of the south, while they have a close 
affinity to each other and a common origin, are, in their entire 
character, essentially distinct from Sanskrit and its derivatives. 
In regard to these Inuguagea, information of the most conclusive 
character may ht< obtaine<l from the preface to Mr. A. D. Camp- 
bell's Telugu Grammar, (including the note by INIr. Ellis,) as 
well as from the Rev. Dr. Caldwell's Comparative Granimar of 
the Dravidian languages. From the last named work I abstract 
the following details : — " There are four principal languages 
current in the different provinces of southern India, Tamil, 
Telugu, Canarese, and Malayalim, spoken collectively by up- 
wards of thirty-one miiliuns of people, besides five minor dia- 
lects, spoken by 650,000 persona. These forms of speech are 
not, in the proper sense of the word, dialects of one language, 
AS no one of them is so nearly related to any of the others, as 
that two pcreons using different members of the group, the one, 
for instance, Tamil and the other Telugu, would be mutually 
intelligible. The Tamil and the Malayalim have the most 
affinity to each other, and yet it is only the simplest sentences 

««« Se« ttbore pp. 07-W. 

442 THE ASIANS IN lyDTA : [cnAP. ni. 

in one of these languages tlmt would be understood by a pererMt 
who spoke ouly the other. The Tamil and the Telugu on tiie 
other hand are the furthest removed from each other of tha 
four languages; and though the great majority of roots ia 
both are identical, yet they are so disguised by inflectioQ and 
dialectic changes that persons speaking each only one of th<*e 
two languages would be scarcely at all understood by ench other. 
The various Dravidian idioms therefore, though sprung from a 
common stock, must be regarded as distinct languages. 

"The northern Pandits chutsify tlie vernacular dialects of 
India"* in two sets of five, the five Gauras and the five Dra- 
vidas. In the latter, they include the Maliratlia and Gurjan, 
as well as the Telinga, the Karnataka, and the Dran«la or 
Tamil. The first two languages are, however, erroneoiuJy 
coupled with the last three; as, though the Mahratha and 
Gurjara (Guzeratee) possess certain features of resemblance to 
the languages of the south, they yet differ from the latter so 
widely and radically and are so closely allied with the northern 
group, Hindi, Bengali, iSic, tliat they must be placed in the 
same class with the latter. The Dravida proper or Tamil, tlie 
Telinga or Telttgri, and the Karnataka, or Canarese, are &ot> as 
the northern Pandits suppose, derived from the Sanskrit, like 
the northern dialects, but, as regards their original and fimda- 
nieutal portion, are qwte independent of Sanskrit. The dif- 
ference Ijetween the northern and southern dialects conasts io 
this, that though the former contain a small proportion uf 
aboriginal or non-Sanskrit wurds, they are mainly composed of 
words derived by corruption from the Sanskrit,"' while the 
Tatnil, Telugu, and other southern languages, on the oontruj, 
though they contain a certain amount of Sanskrit words, are yet 
buth as regards tlic great bulk of their vocabulary, and their 
whole genius and spirit, totally distinct from the dasscal apeech 
of the Arians." 

"* S<« Colcbrooke'* Misc. Essays, toI u. pp. SI, tt. 
•*• Sto iiburv. p. 41. C 






On this subject: I aball introduce here some quotations from 
a note by Mr. F. W. Ellis, appended to the preface to Camp- 
bell's Telugu Grammar : " In arrangement the two latter, 
[the Canmta and Teliugaua alphabets] which are nearly the 
same, certainly follow the Ndgari, but in the form of the letters, 
mode of combination, and other particulars, there is no resem- 
blance ; and the Tamil is tohiUi/ di^fercHt, rejecting all aspirates 
and having many sounds which cannot be expressed by any 
alphabet in wLich the Sanscrit is written. . . . Neither the 
Tamil, Telugu, nor any of their cognate dialects are deriviitions 
from the Sanscrit ; the latter, however it may contribute to their 
jjolish, is not necessary for their existence ; and they form a dis- 
tinct family of languages, with which the Sanscrit has, in latter 
times especially, intermixed, but with which it has no nuiical 
connexion." — (p. 2.) . . . "The Telugu, to which attention is 
here more specially directed, is formed from it« own roots, which, 
in general, have no connexion with the Sanscrit, nor with those 
of any other language, the cognate dialects of Southern India, 
the Tamil, Canuadi, &c., excepted, with which, allowing for the 
occasional variation of con-similar sounds, they generally agree ; 
the actual difference in the three dialects here mentionetl is in 
fact to be found only in the affixes used in the formation of 
words from the roots; the root« themselves are not similar 
merely, but the same." — (p. 3.) 

" To show that no radical connexion exists between the San- 
scrit and Telugu, ten roots in alphabetic order, under the letters 
A, C, P, and V, have been tiikeu frtun the common Dhatumfda, 
or list of roots, and with them have been compared the Telugu 
roots under the same letters taken from a Telugu Dhatumaln. 
. . . These will be found in the following lists, the mere inspec- 
tion of which will show, that among the forty Tehigu roots not one 
agrees with any Sanskrit root." These lists I will copy here: — 


(lr, t)) marh, moee, move torbwiuly. 
Ag, h move, moee torltwUMli/. 

Akkalii, to contract the abdvminnl mut- 



IcDAr. ttb 

I lo ntark. 
Anga, J 

Agh, to move, detpUe, begin, mone 

Agha, to tin. 
Acli, to honour, aeree. 
Ancb, to move, speak uninteUigibly, 

»peak intelligiblt/. 
Aj, to throw, move, ihine. 

** [ lo move. 
A|b, J 

Ad, to occupy, undertake. 

Kak, to hint desire, go. 

Kukk, laugh. 

Kakl), laugh. 

Kakkh, lat^h. 

Kag, to move. 

Kach, to tie, shine. 

Kaj, to hiccup. 

Knt, lo move, skreen, rain. 

Ka^h, to/ear, recollect anxioruly, 

Kii^i to eat, rejoice, divide, preserve. 

F»ch, to cook, explain, stretch, 

Fa4i to shine, move. 

Patli, to tpcak. 

Pan, to traffic, praise. 

Fat, to rule, move. 

Path, to move. 

Piiil, Ui move, be fixed. 

Pan, to praise. 

Pamb, to move. 

Poi'bb, to move. 

Vtik, to he cooked, move. 
Vug, lo be lame. 

Agalu, to separate, break. 

Aggu, to worship 

Aggala, to be insttfferabU, exeetuve. 

Ats. to give by comptdtio*, to incar 

Antu, to touch, adhere, anoint Ike head. 
A^aogu, to be destroyed, submit, be 

Adaru, to shine, shoot aL 
Adulu, to weep bitterly. 
Adu, to slap. 

Kakku, to vomit. 

Kilts, to play dice, chess. 

Krals, to tcant. 

Ka^t") '" '''< build, become pregnaa 

Kiiiliigu, to wash. 

Kadangu,j ^^^j^^ 




K adaru, to call aloud. 

Kadalu, to move or shake. 

Kadi, to approach, obtain. 

> to lick as a dog. 

" ' f '" break, make forked, 
Pangalvi, J 

Panchu, to divide, send away, appoinL 

Piit(u, to seize, touch, begin, knead the 
limbs, understand, unite intimately. 

Padu, to suffer, fall. 

Pandu, to reprove., product, he doum. 

Fadayu, to obtain. 

Fantangu, to vow, 

Padafu, to act precipitately, speuk 
nonsense, threaten, 

Founu, to join steers to a plough, pre- 

Panatsu, to send, employ. 

Vaga, \ lo grieve, pretend grief, eon* 


a, 1/0 
11, J 



Vncb, to tpeak, order. 
Vaj, to move, renew, or repair. 
Vat, to turrountl, share, gpeak, 
Vii^a, to sttrround, share, 
Vanta, to thare. 
Yn\h, to go alone, be able. 
Vad, to shine, tur round. 
Vnn, to sound. 

Voglr, to speak deceit/uHy, bark as a 

Yongu, to stoop. 
Vata, to come. 

Vantsu, to bind, pour out water. 
Vrats, to dioide. 
Vatu, to become lean. 
Vattu, to dry up, 
Yattru, to shine, 
Yad^a, to serve food. 

I Mr, EUia then (p. 7) adduces a list of fifteen roots Teliigu, Ca«- 

I narese and Tamil, taken in alphal»etical order, " to show that an 
f intimate radiad connection exists between the Teliigu and other 
I dialects of Southern India." As the affinity between these lan- 
guages is admitted by all competent scholars, I do not consider 
it necessary to quote this comparative list. Mr. Ellis then pro- 
ceeds (p. 11) to prove by further details that these three lan- 
guages are not only radically connected, but have also an inti- 
mate relation to each other "as regards terms used for the 
expression of ideas." With this view he first quotes a native 
writer, Maniidi Vencaya : 

" ilamidi Vencaya, the author of the Andhra Dipika, an 
excellent dictionary of the Telugu, has, in the preface to this 
work, introduced a concise analysis of the language, the sub- 
stance of which ... is translated in the following paragraph. 

" ' The modes of derivation in the Andhra [Telugu] language 
are four; they are Tatsamam, Tadbhavam, Deiyam, and Gram- 
yam. Tatsamam consists of Sanscrit tenns, pure as spoken in 
heaven, the Telugu terminations being substituted for those of 
the original language.' " 

Of these the following are examples ' " : — 













'" [A few examples only are selected under iwo heads.— J. M.J 



CHAP m. 

"*TadbhaTam consists of terms formed, either from the Sao- 
Bcrit direct, or through one of the six Pracrits, varied by the 
interposition of syllables, and hy the substitution, increment, 
and decrement of letters. . . . The several modes of derivation 
... are exemplified in the following lists :'" — 







Separate lists follow of Tadbhava terms introduced from 
Sanskrit into Telut^u through the Mahanlshtri, the i^uraaeni, 
the Msigadhi, the I'aisiiclii, (said to be spoken in the couutnes 
of Pilndya and Kekaya), the Chulika-Paiaachi, (8}>oken in 
Gandhara, Nepala, and Kuntala), and the Apabbrau^ spoken 
iu the country of Abbira, and on the coast of the we^em 

Mr. Ellis proceeds, p. 15, with his extracts from Mamidi 
Vencaya : '* ' Desyain, in utlier words Andhra or Telugu, is of 
two kinds: the language which originated in the country of 
Telingaua, and Anya-desyara, or the language of foreign coun- 
tries intermixed with it>' " Previously to showing what part of 
the language originated in Trilingam, the native author quotes 
from the " Adharavana Vyacaranain," a description of the 
country to which this name applies."* Air. Ellis givea the 

'" This passage, as quoted in the AndLrukauniudT, is given by Mr. Camp- 
licit in tbc Introiluclion to Iils Grammflr, p. ii. note. I am indebted lo ihe 
late Prof. H. II. 'W'ilson, for transcribing it for me from the Telugu inio Ro- 
man charaelcrs : SrUaila-Bhima-kaleia'Mahetidra-giri-tanyutam | PrakoT' 
antu mahat hfitca trXni dcuruni ch' akarot | Trilochano vuiheias^ triiiloHcha 
kure Bahan] Trill ngarupl nyavamltri-dtarethugaHairTritah\Andhm- Viaknuh 
Surayvto Danujena Nishurnhhuna \ Yuddhea trayodaia yugan hatva turn 
Ruk»haiottamam\ Avatat tatra rishibhir yuto Godavari-iate] Tatkala-prabhfiti 
hihetram Trilingam Hi viinitam\ I trnniilale this anew OS follows: — "He 
[the Andhriaii Vishnu before mentioned], having conslructcd n vast wall 
connecting $ri:^uila, Bhinic^rora, Kille^vsra, and tbc Ihlahendra hills, formed 
in it three gatua. There in the form of three Lingas, with three ejeg, 
bearing in his band the trident of Mohe^a (Siva), he dwelt in the three 
gates surrounded by bis bosts. The Aadliriao Vishnu, attended bj tbc 





author's definition of the native Telugu, as the language which 
arose within the boundaries of Trilinga, as follows : " As it is 
here said, iu the couutiy between ^risailara, the station of Bhi- 
meswai-a at Drachaianiani, the greater Kaleswarain, and, as the 
fourth, the mountain of Alahendra, in these holy places were three 
lingains, and the language which originated iu the country kuowu 
by the name of the Trilinga-desam, is tliat now under considera- 
tion ; this is the At«u or pure Telugu, anil is thus described in the 
Appacaviyam (verse) : ' AU tfiose words whi-ch are in iLse among 
ihii several races who are- aboriffines of the country of Andhra, 
which are p&tfcctly clear and free from all obscurity, these 
shine forth to the worUl as the p'Ui'e native speech of Andhra 
{Suddha-Andhra-Dp*!yamy ^ The following are some of the 
examples given, viz., pdlu, milk, pentfju, curdled milk, ney, 
clarified butter, pudami, the earth, padaluka, a woman, 
koduku, a mm, tala, the head, tw/o, the ntoon, viadi, a field, 
pulif a tiger, magavan^'o, a man. Alamidi Vencaya then pro- 
ceeds to the t^rnis introduced into Telugu from foreign countries. 
"The following verse is from the Appacaviyam : ' K&sava, the 
natives of AndJira, fiaving resided in various countries, by 
iising Telugu terms conjointly with tfwse of other countries, 
these have become A ndhra terms offoreiffn oriffinj' " 

This is what Mamidi Vencaya has to say about the Gramyam 
terms: "Terms which cannot be subjected to the rules of 
mmar, and in which an irregular increment or decrement of 
etters occiut^, are called (framyam ; tbey are corruptions, and 
are described in the following verse from the Appacaviyam 
(verse) : ' Such Ttnugu woi'da as are comvwnly used by rustic 
folk are known as Oi-dmyam terms: these lose some of tJieir 
re{/ular letiei'S and are tivt found in poelnf, unless, as in 
abusive language, the nse of them cannot be avoided.' " 

Suras, fanving slain the illustrious Rnksliasa l^isbambbu, the son of Danii, 
nfter a conllict lasting for thirteen ^lUgiis, resided there with the fuhh, on 
the banks of the Godavari. Siuce that time ihis sacred territory hii» iKcn 
culled Trilinga." 

" In the preceding extracts," (Mr. Ellis proceeds,) " tbs 
author, eupport«d by due authority, teaches that, rejectinsj 
direct and iudii-ect derivatives from the Sanscrit, and words 
borrowed from foreign languages, what remains is the pure 
"native language of the land: this constitutes the great body 
of the tongue, and is capable of expresaiug every mental and 
Ijodily operation, every pousible relation and existing thing; 
for, with the exception of some religious and technical terms, 
no word of Sfinscrit derivation is necessai'y to the Telugu. This 
pitre native language of the land, allowing for dialtfctic differ- 
ences and variations of termination, is, with the Telugii, commoa 
to the Tamil, Cannadi {L e. Canarese), and the other dialects of 
southern India: this may be demonstrated by comparing the 
Dr-flyam terms contained in the list taken by Veucaya from the 
Ajipacaviyam with the terms expressive of the same ideas in 
Tamil and Cannsidi. It haa been already shown that the radi- 
cals of these languages mutatid mutandis are the same, and 
this comparison will show that the native terms in geaerul xat 
in e<u;h, also, correspond." 

A comparative list of Telugu, Canarese, and Tamil words is 
then annexed, pp. 19 — 21, which I omit, Mr. Ellis then goes 
on (p. 21): "From the preceding extracts and remarks on 
the composition of the Telugii language, as respects terms, it 
results that the language may be divided into four branches of 
whirli the following is the natural order. Deeyam, or Atsu- 
Telugii, ^jH re tiativa ^errxs, constituting the basis of this lan- 
guage, and, generally also, of the other dialects of southern 
India : Anyadesyara, terms borrowed from other counirits, 
chiefly of the same derivation as the preceding: Tatsamam, 
jjure Sanscrit tei-nis, the Telugu aflixes being substituted for 
those of the original language : Tadbhavam, Sanser'it lieriva- 
lives, received into the Telugu direct, or through one of the 
sLx Pracrits, and in all instances more or less corruptetl. The 
Gramyam (literally the rustic dialect, from Grrhnam Sans, 
a village), is not a constituent portion of the language, but j| 




formed from the Atau-Telugu by contraction, or by some per- 
mutation of the letters not authorised by the rules of grammar. 
The proportion of Atsu-Tt;higu terniH to those derived from 
every other source is one half; of Anya-desyara terms one 
tenth ; of Tatsamam terms in general use three twentieths ; and 
of Tadbhavam terms one qtuiiier, 

'• With little variation, the composition of Tamil aiid Cannadi 
are the same as the Telugii, and the same distinctions, conse- 
quently, are made by their grammatical writers. The Telugu 
and Cannadi both admit of a freer adoption of Tatsjuiiara terms 
than the Tamil : iu the two fonner, in fact, the discretion of the 
writer is the only limit of their use; in the higli dialect of the 
latter those only can be used which have been admitted into the 
dictionaries by which the language has long been fixed, or for 
which classical authority can be adduced ; in the low dialect the 
use of tliera is more general ; by the Brahmans they are pro- 
fusely employed, more sparingly by the Sudra tribes. The Can- 
nadi has a greater, and the Tamil a less, proportion of Tadbhavam 
tea-ms than the other dialects; but in the latter all Sanscrit words 
are liable to greater variation than is produced by the mere 
difference of termination, for, as the alphabet of this language 
rejects all aspirates, expresses the first and third consonant of 
eacb regular series by the same character, and admits of no 
other combination of consonants than the duplication of mutes 
or the junction of a nasal and a mute, it is obviously incapable 
of expressing correctly any but the simplest terms of the San- 
scrit. All such, however, in this tongue are accounted Tatsamam 
wlien the alteration is regular and produced only by the deficien- 
cies of the alphabet, 

" But though the derivation and general terms may be the 
same in cognate dialects, a difference of idiom may exist so 
^eat that in the acquisition of one no assistance in this respect 
can be derived from a knowledge of the other. As regards tlie 
dialects of southern India this is by no means the case : in col- 
location of words, iu syntactical govermnent, in phrase, and 

U G 



[chat. nx. 

iodeed in all that is comprehended under the term idiom, tbey 
are not similar only but the same. To demonstrate this, and to 
show how far they agree with, or differ from, the Saoscdt," 
Mr. Ellis proceeds to give a series of comparative renderings of 
sentences in Sanskrit^ and in the Tamil, Telugu, and Cauareae. 
A.S however it would lengthen this Section too much to cit« 
these deti\il8, I must refer the reader who is desirous of pursuing 
the subject further, to Mr. Ellis's "Note" itself. 

Fropi Mr. Campbeirs Introduction to hi£i Grammar, pp. tIU 
viii. ff., I supply some furtlier particulars regarding the earljr 
cultivation of Telugu and the belief of the native grammanana 
as to the origin of their language : — " The most ancient Teloo- 
goo grammarian of whom mention is made in the native books is 
the sage Kuuva, who is said to have been the first that composed 
a treatise on the principles of the language. It"^ la stated that 
he executed this work by command of a king of Andhra, named 
Andbra Royoodoo, aon^** of Soochundra. . . . The worka of 
Kiuiva, of Aiidharvan Achar}' (sic), and of several other andeot 
graiuni(u-ians are not now to be foimd. All the treatises oo 
Telougou grammar at present extant, consist of Sanscrit com- 
ineuttuies on a series of concise apophthegms written in Sanscrit 
by a IJramin named Naunapa, or Nimniah Bhutt." 

" It has been very generally asserted, (says Mr. Campbell, 
p. XV. S.,) and indeed believed, that the Teloogoo has its origin 


<>» " Kunva said * Be who tptaki irrtttrtnlXjf of wiff grammar, eampomd 
hjf the commami of Atulkra VuknoOy AaO be coktidertdai guUtjf qf im vtrrmee 
to Am prittt' Andhra Cowmudi." The original i« as follows: KmmmIIi 
jroMu oAa A»dkrO'tith*cranHJM-ifitafya luaffryaJkarawuya drvAi gu r m «frwH8 f . 
'*** In Rfcard to this king Mr. Campbell quotes tbe fuUowiog tiwigf 
l«hich precedes that dtod ia mj former note, p. 446 : AmAraniUkn MakA' 
\nHr yitkamhhm-dbmfi^U \ Ptura ScagoKMimt Mamok hSb Kalqpga 
iignii I Kikmlf ryit w yaayq S^etmdra^i« laiiiMkmvttJk \ Abkmnt mtroa tie' 
rtt k Utt Ma^ijfUak \ ■* Fonnerlj, in the time of Matia SraTauultLu, 
tk* KaH I4P^ HttV Ike kird uf Andhra, the great Visbnn, ti>e elajer of the 
I>aafev« Kiabuabktt, wtt b«ra ia Kaivla as tfaesoa oTtheHMaardiSiidiaitdn, 
I ami WW aUwidwl Kv ■*''~?^^ii^ «s wdl as tvtenaeei bj ili ■ankjnd" 


MCT. VI.] 



in the language of the Vedams, ... I venture publicly to state 
my enquiries to hare led me to a contrary conclusion ; but I do 
so with the less hesitation as I find myself supported by the 
concurrent evidence of all native authors who have ever written 
on the subject of the Teloogoo language." 

*♦ In common with every other tongue now spoken in India, 
,piodem Teloogoo abounds with Sanscrit words; . . . neverthe- 
less there is reason to believe that the origin of the two languages 
is altogether distinct" " In speakiug the Teloogoo the Soodras 
use very few Sanscrit words : among the superior classes of Vysyas, 
and pretenders to the Rajah caste, Sanscrit terms are used only 
in proportion to their greater intimacy with the Bramins, and 
their books ; and when we find even such Sanscrit words as these 
classes do adopt, pronounced by them in so improper and rude 
a manner as to be a common jest to the Bramins, who, at the 
same time, never question their pronunciation of pure Teloogoo 
words, I think we may fairly infer it to be probable at least that 
these Sanscrit terms were originally foreign to the language 
spoken by the great lK)dy of the people." 

'• Some native grammarians maintain that before the king 
Andhra lioyadoo'" established his residence on the banks of the 
Godavery, the only Teloogoo words were those peculiar to what 
is emphatically termed the pure Teloogoo, now generally named 
the lniign(i{/e of tlie land, which they consider coeval with the 
people^ or, as they express it, * created by the god Brimha.' The 
fnllowcTB of this prince, say they, for the first time began to 
adopt Sanscrit terms with Teloogoo terminations, and by degrees 
corruptions from the Sanscrit crept into the language, from the 
ignorance of the people respecting the proper pronunciation of 
the original words."' This would imply that the nation still 

i«i u -j^ig jg (In; prince vrho is now worslii]ipc<l as a divinity at Sieearollum 
on the river Kriahnn, and who was the pnlron of Kunva, the first TcloogoQ 

'" The rollowhig is the passage rofcrrcil to, and it foUowa the one (luoted 

c a 2 






retain some faint remembrance of those times in which their 
language still existed independent of the Sanscrit; and it is 
certain that every Teloogoo grammarian, from the days of 
Nunnia Bbutt to the present period, considers the two languages 
as derived from sources entirely distinct; for each commenoe« 
his work by classing the words of the language under four 
separate heads, which they distinguish by the respective names 
of Deshyurnoo, lavgiuu/e of the land; TutmimuTnoo, Sanacrit 
derivatives ; Tudbhuvumoo, Sanscrit corrwpticna^ and Gram- 
yumoo, provincial tei'ms. [Compare the Grammar, p. 37.3 '^'^ 
these, later authors have added Aiiya-ileahnuiiioo, fi>reiffn. words." 

**The words included iu the first class, which I have denominated 
the language of the land, are .... the most numerous in the 
language, aud the model by which those included in the other 
ohi.sses are miHlified and altered from the different lauguagcfi to 
wLieli they (irigiiiully belonged. The name by which they are de- ^ 
signated iniiilies ' thut which belongs to tite couni^'y w land; it ^ 
marks the words in question, not as merely * current in the 
country,' Ijut ius the gr^n^'th and priKluce of the laud." 

" In the course of this work it will be obvious to the Sanscrit 
scholar that the declension of the noun by particles or words 
added to it, — the use of a plural pronoun applicjible to the first 
and second j)erson8 conjointly — the conjugation of the affirma- 

in the note, p. 446 : TntratySt tattamutapn* latkrltitin Hnrer bhaluh | ialena 
inafuitd sarvom tuhanmm xralpa-huddhihhih \ AiudithothcharynmaMam tat 
tttdhhawiiicheti summatam | Vihar»ku'Vyatyttyubhyancha padiirdkokCi ri'ie- 
shatah I Tuhhavam Hi kulhyurUe halena inalmta tamuh \ Drnhmanu nirmita 
vuchah piirvam Andhreiitur llareh \ Achchu i/i cha kalhyante tiip-ktld-dhStu- 
gaimmvitiih | "The adherents of Ilari who dweU there (in Trilingn, on the 
banks of the Godavari) at that lime, spoke toUatna words. In proctrss of 
lime these tolsutmi wonJg lM>)»!in to be incorrectly pronounced by ignorant 
pcrs-ons, and were rcgiink'tl as ladhhava, Tutgama words were denominate*! 
tadhhata from loss or siilislitution [of letters], or from being contracted it 
fourth or a half. Words, consisting of nouDA, verbale, and roots, which 
were fashioned by Brnhron before the time of Ilnri, the lord of Andhra, are 
called achcha (pure}." 

ncT. VI.] 



live verb — the existence of a negative aorist, a negative impe- 
rative, and other negative forms of the verb — the union of the 
neuter and feminine gentlera in the singular, and of the mascu- 
line and feminine genders in the plural, of the pronouns and 
verbs — and the whole body of the syntax, are entirely uncon- 
nected with the Sanscrit; while the Tanail and Kamataca 
scholar will at once recognize their radical connexion with each 
of these languages. The reader will find all words denoting tlie 
diflferent parts of the human frame, the various sorts of food or 
utensils in common xise among the natives, the several parts of 
their dress, the compartments of their dwellings, the degrees of 
affinity and consanguinity peculiar to tliem, in short rdl tL-rnis 
expressive of primitive ideas or of things necessarily named in 
the earlier stages of society, to belong to the pure Teloogoo 
or lai>f/nage of the. land. It is true (so mixeil have llie two 
languages now become) that Savscrit derivatives or coivuptiona 
may without impropriety, be occasionally used to denote some 
of tliese. This, however, is not common : the great Ixwly of 
Sanscrit words ailmittt'd into the language consists of abstract 
terms, and of words connected witli science, religion, or law, as 
is the case, in a great degree, with the Greek and Latin words 
incorporated with our own tongue: but even such Sanscrit words 
as are thus introduced into Teloogoo are not allowed to retain 
their original forms; they undergo changes and assume termi- 
nations and inflections unknown to the Sanscrit, and, except as 
foreign quotations, are never admitted into Teloogoo irntil they 
appear in the dress peculiar to the himjuage of the land." 

At the rifik of some repetition, I shall add a few further ob- 
servations, abstracted from Dr. Caldwell's grammar, pp. 29, £f., & 
56, in proof of the radical ditfereucea between the Sanskrit and 
the southern languages : — "No person," he remarks, " who is 
ao<]uaint«d with comparative philology, and who has compared 
the primitive and essential words, and the grammatical structure 
of the Dravidian langtiages with those of the Sanskrit, can ima- 
gine that the former have Ijeen derived from the latter by any 

a o 3 

known process of corruption or decomposition. We shall first 
advert to the Sanskrit element which has been intnxluoed into 
these langiiaf^es, and then revert to their non-SanKkrit or 
tial basis. First, the most recent infusion of Sanskrit wordu 
into the Tamil, Mr. Caldwell states, (p. 56,) waa effected by the 
great religious schools of t^aukara Acharyya anti Raman lija, from 
about the tenth to the hfteenth century A. d. The words tlicn 
introduced (excepting a few points wherein change wa» unavoid- 
able) are pure, unchanged Sanskrit. Secondly, at a period 
partly preceding and partly contemporaneous with the abore, 
from the eighth to the twelfth or thirteenth century A. n., the 
Jainas introduced the largest proportion of the Sanskrit deriva- 
tives that are to be found in Tamil. This period of Jaina in> 
tellectual predominance was the .Augustan age of Tamil literature, 
a period when the celebrateii college of Madura flourishwl, and 
the CuraJ, the Chintfuuani, and the classical vocjilmlaries and 
graranmrs were written. The Tamilian writers of this period, 
from national feeling, and their jealousy of Brahminical influence, 
niddifu'd the Sanskrit words which they emjiloyed so as to accord 
with the euphonic mles of Tamil. Thus lokut * world,' beciMnes 
ularju in Tamil ; rdjd, ' king,' becomes aroifu ; and rA, • night*' 
(from rdtri) becomes in-avu. Nearly the whole of the Sanskrit 
words found in the Telugu, Canaiese, an<l Malayalim belong to 
these two periotls, or correspond mainly with the Sanskrit deri- 
vatives found in the Tamil of those two periods, especially the 
more recent. These derivatives are divided into the two claases 
of TdiSitma, words identical or nearly so with pure Sanskrit, and 
Tu(Hi/utva, words which are Iwrrowed from Sanskrit or the 
northern Prakrits, but have been to some degree modified in 
form. Thirdly, the Tamil contains many derivatives, belonging 
to tlie very earUest period of the literary cultivation of tliat lan- 
guage, wliich were probably inti-oduced before Sanskrit words 
had be^m to be imported Into the other southern dialects. The 
Sanskrit of tills {x^riod is more corrupted than tliat of the Jaina 
period, and the corruptions are of a different character. The 




Jainas altered the Sanskrit words in accordance with the euphonic 
rules of Tamil, whereas the words introduced in the earliest 
periofi have been changed in defiance of all rules ; as the San- 
skrit jfrT, 'sacred,' into Hru. While, however, a certiiin propor- 
tion of Sanskrit words have been introduced into the Dravidian 
tongtxes in the w^ays just described, — it would be quite a mis- 
take to suppose that these languages are derived from the San- 
ckrit in the same manner as the Hindi, Mahratti, and other 
Gauda dialects. For (1,) the non-Snuskrit portion of the Dra- 
vidian languages exceeds the Sanskrit portion nearly aa much as 
in the North-Indiun dialet'ts the Sanskrit element exceeds the 
indigenous or non-Sanskrit element. (2.) The pronorma and 
munerals of the Dravi<lian languages, their mode of inflecting 
verbs and nouns, tixe syntactic arrangement of their wowia — 
everj-thing, in fact, which constitutes the essential structure of a 
language, are radically different from those of the Sanskrit. 
The contrary is the case with the vernacular dialects of the 
north, in which the pronouns, the numerals, and a large propor- 
tion of nouns and verbs, have been derived by adoption or 
gradual transformation from the older Prakrits and ultimately 
from the Sanskrit. (3.) Tlie true Dravidian words, which form 
the great majority in the southern vocabularies, are placed by 
the native grammarians in a different class from the Sanskrit 
derivatives, and are honoured with the epithets ' national words' 
and ' pure words.' " In support of this Dr. Cahiwell refers to 
the passage already quoted in p. 452 ; and gives it as his opinion 
that Andhrarjlya probably lived several centuries before the 
Christian era. " (4.) lu the uncultivated languages of the Dra- 
vidian stock, Sanskrit words are not at all, or very rarely, 
employed. And further, some of the cultivated Dravidian lan- 
giiages which do make itse of Sanskrit derivatives, are able to 
dispense with these altogether. This indee<i is not the case with 
Telugu, Canai'ese, or Malayalim : but Tamil, the most highly 
cultivated, as regards its original stnicture, of all the Dnividian 
idioms, is not dependent on Sanskrit, for the full expression of 

a o 4 


Lca<ie. ttt. 

thought. In fact the ancient or classical dialect of this language, 
the Shen-Tamilj in which nearly all the literature has been 
written, contains very little Sanskrit ; and even diflfers chiefly 
from the coUoqixial dialect by the jealous care with which it re- 
jects derivatives from Sanskrit and restricts it«elf to pure Dra- 
vidian elementa. So much is this the case that a Tamil compo- 
sition is regarded as refined and classical, not in proportion to 
the amount of Sanskrit it contains, but in proportion to the 
absence of Sanskrit. It ia also worthy of remark that though 
the principal Telugu writers and grammarians have been Brab- 
mans, iu Tamil, on the contrary, few Brahnians have written 
any works of distinction, while the Tamilian Sudras have culti- 
vated and developed their language with great ardour and 
success ; and the finest compositions iu the Tamil language, the 
Cural and the Chintamaui, aie not only independent of the 
Sanskrit, but original iu design and execution." 

A few more specimens of Tamil words derived from Dr. Cald- 
well's book, passim, may be added to show how perfectly dis- 
tinct they are from the Sanskrit, and North-Indian vernacular, 
words having the same sense, with which I shall presume 
readier to be acquainted. 



UNS, &c. 




the eje 


the moon 




the nose 








a grove 






a sou 






a daughter 


the 8ca 



11 Ian 

a husband 

111 till 111 





a wife 


a, bowel 




a washerman 




of iron 


a washerwo- 


a cock 


a wall 


ODC [man 


the ground 






an ox 






a shwp 


a well 


four ^^ 


a monkey 


the liver 


five 4 


a tlnj 


a fright 


six m 

SECT. Vll.] 











a bunJred 




three biin- 
sixly [tJrcd 

Tamil declension of manei, a house. 

singular. Plural. 

NoM, manei mancigal 

Ace. inaneiyei maiieigalui 

IxsT. maneiyal niuneigulal 

CoMj. manciyodu mancij^alodu 

Dat. maneikku lUBni'ignluicku 

Abl. maneiyilirundu mancigaHrunilu 

Gbn. maneiyin maneigalin 

Loc, maDeiyidnttil mancignti<lattil 

Voc. Dianeiyc inaiiuij^nlu 


imikkiradu it is tulir to .'prout 

perugugirudu it increases pugar to praise 

a(lanp;u to be cuiituiucd magir to rejoice 

adukku to contain sural to whirl 

figu to become kuyil to sound 

akku to make tuval to bernl 

niiigu to quit urul to roll 

nikku to put away karbikku to suSfcr pain 

nirunibu to be fuU tara to give 

iiirappu to fill Tara to come 

vnjar to grow 

" (5) The grammatical structure of the Dravidian languages is 
radically different from that of the Sanskrit; and proves that 
they are quite independent of that language." For further illus- 
trations of this fact. I must refer to Dr. Caldwell's Grammar, 
pp. 34, ff., and to the suhsequent details given in that work, passim. 

Sect. VII. — Resvltt deducibU from the preceding Sections. 
In the last section I have supplied abundant evidence, derived 
from the be.'?t authorities, of the radical ditferences which exist- 
between the languages of the south of India, and the Sanskrit. 
The evidence which I have adduced is not (as will have been 
noticed) confined to the fact of those ilissiniUarities of roots 
and of structure which are sufficient to convince the comparar- 

THE AKtASs IS nn)u t 

[ciiAF. rrr. 

tive philologist that the DravidiAn dialects have no original aflUnrty 
with the Indo-Germanic tongues. We have also the teetimony 
of the native gi-arnmariaus of the south to the same effect (as we 
have sefsn, pp. 445 — 447,451, fT.). The Telugu authors hold 
that the words of which their language is composed are of four 
clafises, Df^ya or Atsu (or aboriginal), Tatsama (pure Sanskrit), 
Tadbhava (modified Sanskrit), and Grdmya (or rustic); and 
they consider tliat tlie first class, the De^ya or Atsu Telugu 
words, constitutetl the primeval basis of the language before the 
introduction of Tatsama words in the time of King Andhra- 
raya,'*' and were created, with a complete grammatical structure 
of their own, by the go<l Brahma- I am not in a position to cite any 
similar testimony on the part of the Tamil grammarians, !mt Mr. 
Ellis informs us (see p. 449) that the same distinctions are made 
by them as by the Telugu writers, and their idea of the relation 
of perfect independence in wliieh their language stiiuds to the 
Sanskrit, is siiflSciently shown by the fact that they regard that 
Tamil as the most pure and classical, in which there is the 
smallest atlraixturf of Sanskrit. It is therefore a fact, established 
beyond all doubt, that the Dravidian or South-Indian languages 
have, as regards their original and fundamental portion, no 
affinity with the Indo-Germanic languages; and could nort, 
by any means known to comparative pliilologists, have been 
derived from any member of that family. There are certain 
processes and modes of mutation which are always discoverable 
when one language springs out of another. The words of the 
derivative tongue are always recognisable, (even if considerably 
mo<lified,) in the new forms which they have assumed ; and the 
steps of their transformation can be either exactly traced, or at 
least divined with certainty. But the primitive words and 

»»» We h«ve alreadjr seen p. 455, ihnt Dr. Culdwelt considers this monarcli 
to have llourishuil several centuries b. c. Fiom llic Vishnu Puriina, !▼. 24. 
it appears ibat an Andhra-bhritya dynnsiy of kings reigned iu Ma gpH^^ 
whose accession Wilson (V. P. p. 474.) calculates to have dated from IS 
yean b.c. See also Lassen, Ind. Ant. ii. 755. 934. 


ncT. vti.] 



forms of the South-Indian dialects could not have issued from 
the Sanskrit Ly any known law of modification. 

But if the Dravidian languages be of a stock altogether dis- 
tinct from the Sanskrit, it follows inevitably (see above, p, 277) 
that the races which oriffinalbj spoke these two classes of lan- 
guages must also have been distinct from one another in their 
descent, and could never have belonged to tlte same Jirancli of 
the human fainilj', Htid the Dravidian nations been of Arian 
lineage, the whole of their languages, the Tamil, Telugu, 
Canarese, and Malayalim, must have resembled either the 
older Prakritfl (deiwribed in the early part of this volume) or 
the later Hindi, Mahratti, and Bengali, all of which have 
evidently arisen, in great part, from the decomposition of San- 
skrit. But the very contraiy is (as we have seen) the case in 
regard to those southern dialects. 

And as the Dravidians now make use of languages whicli 
are radically distinct from Sanskrit, we cannot possibly Bup{K>se 
that the aboriginal part of the nation ever, at any former time, 
Hp(»ke a language which had any affinity to Sanskrit. Such a 
supposition would be at variance with the traditions preserved 
by the Telugu grammarians. And no race of mankind has ever 
been known which (except under the pressure of external 
influence) has lost, or abandoned, the language which it had 
derived from its forefathers, and of itself adopted a form of 
speech fimdamentally different. But as we have no proof of 
any such external influence which could have led the Dravidians, 
to exchange their original language for another, we must con- 
clude that they have derived their existing dialects from their 
forefathers : and these their forefathers, as their speech was dis- 
tinct from that of the ancestors of the Arians, must have been 
distinct in lineage also, from the latter. As, therefore, the ori- 
giual Dravidian Indians of the south of India are of a different 
race from the Arian Indians, they could not, as Maim and the 
MahabhiTrata assert, (see above, p. 439,) have been degradeti 
Kshattriya<!. And this conclusion is not in the slightest degree 

affected by the fact that a conBiderable portion of the existing 
Dravidiaa cominuuities, though speaking the language of the 
south, belongs, or claims to belong, to the higher Arian castes. 
For if the southern Bi'ahmans, and some of the other castes, be (aa 
in all probability they are,) of Arian descent, this dora not prove 
that the same is the case in regard to the great mass of the Dravi- 
dian population ; for there is every reason to believe that those 
southern communities esistetl before the Ariana had spread tbem- 
selves to the south of the Vindhya mounttuns, and that the Brah- 
mans emigrated at a comparatively recent period from uortljern 
to southern India. On their arrival in the south, these Bi'ahmans 
no doubt spoke Sanskrit or rather one of its derivative Pi-akrits. 
But though, from their superior civilisation and energy they soon 
succeeded in placing themselves at the head of the Dravidian 
communities, and in tutroduciug among them the Brahnianical 
religion and institutions, they nuist have been so inferior in 
numbers to the Dravidian inhabitants as to render it impracti- 
cable for them to disltHlge the primitive speech of the coimtrj', 
and to replace it by their own language. They would there- 
fore be compelled to acquire the Dravidian dialect of the pro- 
vince in which they settled ; and, in a generation or two, the 
majority of them would lose the vernacular use of the Prakrit 
dialects which they had brought with thera. This, however, 
■would not prevent their retaining in use a great many words of 
Sanskrit origin. And as many of these Brahmans were learned 
men, and as their religious books were composed in Sanskrit^ 
they would necessarily preserve their acquaintance with that 
sacred tongue, and with it« literature ; and would no doubt 
from time to time introduce fresh Sanskrit words into the local 
vernacular,'" just as wo see that English is continually enriched 

'*• I may take this opportunity of adverting ngnin to the probubilitj 
niready alluded to above, in note 41. p. 43, that Sanskrit has not only in- 
fluenced the aboriginal tongues both of northtrn and gouthcm India, but 
has also received some inlluenco from one or from both of them in retant. 



by the addition of new Greek and Latin words. The fact that 
many of the present inhabitants of the south of India are of 
Axian extraction affords, therefore, no reason for doubting that 
the primitive language of those provinces was totally distinct 
from the Sanskrit, and that the population by whom that lan- 
guage was originally used was totally unconnected with the Arian 
race. For even the existence of the limited proportion of non- 
Sanskrit words which we can discover (see above, pp. 35 — 43) in 
the Hindi, Mahratti, and other northern dialects, is sufficient to 
prove that there originally existed in northern India one or 

Mr. E. Noiris observes, (Journ. Roy. As. Soc. vol, xf. p. 19) : " I will here 
express my conviction that the sounds called cerebral are peculiar to the 
Tartar or FJTiniab olass of laiin;iiaj;cs; that the renlly Indian [i. e. the abori- 
ginal, or non-Ariaii — J. M.] languages are all of Tartar origin, or, at leiist, 
that their phonetic and grammatical .iffinities arc Tartar; and that the 
writers of Sanskrit adopted the sound from iheir Indian neighbours." And 
Professor Benfey says (Coniplete Sanslcrit Gnimoinr, p. 20): "The mute 
cerebrals have probably been introduced from the phonetic system of the 
Indian aborigines into Sanskrit, in which, however, they have become firmly 
established." And at p. 73, of the same work, he thus writes : " Sanskrit is 
a language of great antiquity and of wide diffusion. Long after it had 
ceased to be vernacularly S[K)ken, it coutiniied to be employed as the organ 
of culture and religion, and in this capacity it prevailed over extensive regions 
where there existed alongside of it, nut merely a variety of dialects which 
had been developed out of it, but also several popular dialects which were 
originally quite distinct from it From these circumstances it has resulted, 
nut only that forms which have been admitted into the Prakrit dialects 
have been afterwards adopted into Sanskrit, but, further, that words which 
were originally quite foreign to the Sanskrit hare been included in its 
vocabulary. To separate these foreign words will only become p<;ssib]e 
when an accurate knowledge of the dialects which have no aflinity with San- 
skrit shall have been attained. But it is almost as difBctdr. to distinguish 
those irregular forms which have originated in the dialects derived from 
Sanskrit and have bt^n afterwards received into Sanskrit, from those forms 
which have arisen in Sanskrit itself; because, on the one hand, Sanskrit litera- 
ture and its history are as yet but little known, and, on the other hand, those 
phonetic changes, which attained their full power in the Prakrits, had olready 
begun to woik in Sauskrit itself." See also above, p. 152. 

4 02 


[cuAr. m. 

more races of non-Arian inhabitants who occupieii tJie countiy 
before the immigration into Uiadustban of the Sanskrit-«peak> 
ing Aryas. 

It now remains for me to inquire how this important fact 
that the great bulk of the population of the Dekbaa is non-Arian 
in its descent, coupled with the other conclusions of the fore- 
going sections, affects the results at which I had previously 
arrived in regard to the trans-Himalayan origin of the Ariaos, 
and their immigration into India from the north-west. 

In the preceding chapter, we were led by a variety of con- 
siderations, all poiiiting to the same result, to conclude that the 
Aryas had penetrated into India from the north-west. The 
facts which have been substantiated in the foregoing sections of 
the present chapter are in perfect harmony with that conclusion. 
These facts are (I.) that the Aryas, when living in the Panjab, 
found themselves in conflict with a class of enemies whom, in 
contrast to the men of their own race, they called Dasj'us : (2.) 
that the Aryas, after occupying the north-west of India, from 
the Indus to the Sarasvati, began, at length, to move forward to 
the east and to the south : (3.) that, still later, they crossed the 
Viiidliya range, and commenced to colonize the Dekban, which 
had been previously occupied exclusively by savage tribes ; 
and (4.) that the nations who at the present day inhabit the dif- 
ferent provinces of the Deklian, and who (with the exception 
of such part of the population as is descended from the later 
Arian immigrants) ai*e the direct descendants of the aboriginal 
tril)es, — speak a class of languages which are nuiically die^ 
tinct from the Sanskrit It may be expedient, however, to show 
somewhat more in detail the manner in which these circiu«- 
stances corroborate, or at least harmonize with, the theory that 
the Arians are not autochthonous, but of trans- Hi mtUayan 
origin, and that they immigrated into Hindusthan from the 
north-west First, then, the fact that at the dawn of Indian 
hiskiry, the earliest Vedic period, we find the Arian Indians 
inhabiting the Panjab; then advancing gradually eajstwsird 


along the southern border of the Himalaya from the Sa^ 
rasvati to the Sarlauira, and spreadiug, simultaneously, no 
doubt, over the southern parts of Doab, into Behar ; and at 
length crossing the Vindhya mountains into the Dekhan, — 
affords the Btrongeat presumption that they penetrated into 
ludia from some quarter closely adjoining the north-western 
comer of that country, which was the starting-point of their 
onward course of conquest and colonization. Secondly: the 
indubitable fact that the Arians found, on penetniting into the 
Dekhan, a people speaking a language radiailly different from 
their own, who had been in earlier occupation of the country ; 
and the almost equally certain fact that they had previously en- 
countered similar alien tribes in the Panjab and in the Doab, 
place it nearly l)eyond a doubt that they (the Arians) could 
not have been the race by whom ludia was originally peopled. 
For, we must either suppose that both of these two races, 
the Arian and the non-Arian, grew up together in India, 
where we find them in contact from the earliest period, or 
that one or both of them have immigrated into that country 
from witliout. But it seems unlikely that two races differing 
HO essentially, as these Arians luid non- Arians appear to have 
done both physically and intellectually, could have sprung up 
naturally ia the etime country, under the same climatic influ- 
ences, or, if they did so, that their language, religion, and in- 
stitutions, should have been so different^ It is much more likely 
that one or both of them should have been foreign. Tho fact 
is that both have prubably imiaigrated into India from the 
north-west;'** but the evidence in favour of this supposition 
is far stronger in the case of the Arian, than in that of 
the noa-Arian tribes. For, besides the proofs derived from 
their language, which clearly connects them with the nations to 
the west of the Indus, we have the evidence of their complexion, 

'^ In the App. note L, I sliall quote the views of the Rev. Dr. Cnldwcll 
and othor writers, regarding the origin and relations of the difTurenl non- 
Arinn tribes. 

which in the present day is far fairer than that of the aborigines, 
and in earlier times, was probably still more clearly distiuguiitb- 
able from the dark colour of the latter, (see pp. 2H4, 309.) But if 
neither of these two races was indigenous in India, and if they did 
not at first occupy any portion of that country contemporaneotuly 
with each other, which of them is most likeJy to have been the 
first possessor ? We must no doubt conclude that the Daeyus 
or barbarous races were the earliest occupants For as Lassen 
observes (see above p. 308), we perceive evident traces of the 
Arians having severed asunder an earlier population, and 
driven one portion of it towards the northern and another to- 
wards the southern hills ; and the inhabitants of the Vindbya 
range, and of the Dekban, appear always as the weaker and 
retiring party who were driven back by the Arians. And w« 
cannot ascribe to the non-Arian tribes the power of forcing them- 
selves forward through the midst of an earlier Arian population 
to tlie seats which they eventimlly occupied in the centre and 
south of the penia&u!a : for the Arians were from the beginning 
a more powerful and civilised people than their adversaries, and 
from a very early period have held them in subjection. It is 
indeed objected by Mr. Curzon (see above p. 299), that these rude 
so-called aboriginal tiibes may have been descended from some 
of the barbaric hordeB, who imder the name of l^iakas, Hv'mas, 
&c, are mentioned by Sanskrit iHTiters as having invaded India, 
and some of whom, after their defeat, may have taken refuge in 
the liills and forests of Hindusthan. But I apprehend that this 
exphmatiun will not meet the facts of the case. We can have no 
assurance, that such legends as that regarding the {§akas which is 
tliiote<l ill the First Part of this work (p. 177, ff.), have any histo- 
riail foundation. And the period at which the Indo-Scythians, 
who were repelled by Vikramaditya, made themselves masters, 
and retained possession, of the western frontiers of India, can- 
not be placed much earlier than the commencement of the 
Christiau era. (See Lassen's lud. Antiq. vol. ii. 365, ff., 398. 
408, 409.) But the traces which we discover in Indian lit^^rature 





of the existence of the Dasyus are (as we have seen from the 
Tarious Vedic texts cited above) much older than this period. 

In conclusion, I return to the point from which I started at 
the commencement of this volume ; and, as the result of tiie pre- 
ceding investigations, repeat the following propositions : First, 
that the Hindus of the superior castes are sprung from the same 
race with the Indo-European nations of the west : Secondly, that 
as the parent race had its origin in Central Asia, the Indian branch 
of it could not have been indigenous in Hindusthan, but must 
have immigrated into that coimtry from the north-west. 

I must, however, reserve for a future Volume, the historical 
evidence of my assertion, that the fourfold division into castes 
was unknown to the earliest progenitors of the Hindus, and was 
only gradually developed after the settlement of the Indo-Arians 
in Northern India, and coincidently with numerous modifications 
in their original mythology and worship. 



NOTE A, p. 267. 

" The conronnities [between the languages of tlie same family^ are 
BBtonisliing ; nnd especially so, because tliej enter into the minutest 
detniln, and even into the nnoin.'ilics. It is a curimis phcnumoiioii to 
discover such an inconftjivublc tenacity in idiotns wbicli mi-ilit ainiear 
to be nothing more than pnsaing capriccsi. The most volatile poriinn 
of languages, I mean their pronunciation, hns evinced its stability : in 
the midst of niulntiuna of letters, which ari', nevertheless, subji'ct to 
certain rulejs, voweln, long or short, have often preserved their quan- 
tity." "On the oihir hand tlicdispnrity isgrciit: tlie distHneea which 
the languages have traversed in their individual development are 
immense. After we have exhausted all the analogies, even the most 
secret, there remains in eiiRh of tlase languages a portion which is 
no longer susceptible of comparison with the other hm;riiages of the 
ae family. We must therefore admit as the causes of that partial 
incommensurableness, two opposite principles^ viz., oblivion ami i«- 
vtntion. The oblivion of forms and words formerly in use is but too 
manifest in the languages with whose history we are most intimately 
acquainted; and it has frequently injured tlieir richness and beauty. 
Such oblivion must always fyUotv a retrograde movement in civili- 
eation : in proportion as the intellectual sphere is contracted, a gene- 
ration which lias relapsed into ignorance and barbarism, abandons 
expressions which have now become superfluous. And as regards 
invention, I find no difficulty in that either, since in order to com- 
prehend the absolute origin of language, we have no choice between 
having recourse to a miracle, and conceding to mankind an in- 
stinctive power of inventing language." — A. W von Schlegid, d« 
Torigine dea Ilindous, Kssais ; and in the Truosactions of the Royal 
Society of Literature of the United Kingdom, vol. ii. p. 433. 



NOTE B, p. 285. 

'* Sirabo telb as tfaat the tribes of the Persians, Medes, Biictrian 
and Sogdians, spoke nearly the same language. Wo can have 
difficultj in supposing that this similarity of speech which existed i| 
Strabo's age, existed also in earlier times. I1ie old Iranian dialecL 
or which the monHmcnt^ have been still preserved to us, justify iLii' 
assumption. Of these there arc four, (1.) the speech of the earlier 
AchsmenidiE, (2.) that of the later Achxmenido:, (3.) the dialect < 
the Gathos' (4.) the old Bactrian, the ordinary language of 
Avesta. The two last dialects tnight perhaps also be emt 
under the designation of Arestic. The first two of these 
belong to western, the lost two to eastern, Iran." — Spiegel, in EuL 
and Schleicher's Beitrjige zur Vergl. Spracbf. ii. G. I must refer 
the original paper for further details regarding these dialects. I wt 
only quote one or two remarks. In bis account of the old Persian 
earlier Achsmenidan dialect, Spiegel observes (p. 7), that, " we find 
in it all the classes of the Sanskrit alphabet represented, excepting 
the cerebrals, which have a purely local origin," (See above, p. 
note.) At p. 13, he remarks: "We have pointed out iu the ent 
grammar of the old Persian so much that is identical with the Sail 
skrit, that it may now be time to notice the differences which stat 
it as a distinct langu.-ige from the old Indian. Not a few such pecu 
linricies are to be found in all the departments of grammar. In 
a phonetic aspect, there is this important deviation that the olil 
Persian has tlie letter 2, wliich is unknown to the Sanskrit, an4 
tliat it, like the Greek, changes the Indian t into A.'' — (See above^ 
pp. 314, and 316.) 


NOTE C, p. 294. 

Rig-veda ix. 113. 7 — 11. Vatra jt/otir ajasram yasmin lake srar^ 
hitam \ Tatmin mam dhehi pavamana amjUe lake akshUe \ Vatr 
riija Vaivasvalo yatravarodhauam divah \ i'atramur j/ahvat 

' [It U scarcely oeccsuTy to my that this has no connociioQ with the InJii 
Gniliu (liukcL d«KTiLca above, pp. 124, ff. .T.M.] 





apns latra mam amritam kridhi \ Vatranithamam charanam 
irirtfiAe tridive divah | Lokn tjatra jijotishmantas tnlra mam, &c. | 
Ultra kaina nikamakcha yatra bradhnasya vishtnpnm \ Svadha 
yatra (fiptiicha tatra mam, &c | Yatrananda&cha modascha 
mudfth pramuda asate | Kamasya yatriiptah kattias latra mam, &c, | 
^^' Place me, O purifier (Sornii), in that divine unchanging region, 
^^wbere perpetual light and eunshine abide. Make me immortal in the 
^^Vorld where king Vaivaivata (Yamn) reigns, wiiere the sanctuary 
^^W" (he eky is, and the great waters are. Make mo immortal in 
the third heaven, where action is unrestrained, nlicre the shining 
regions exist. Make me immortal in the world where all enjoy- 
■ents exist, in the realm of the sun, where celestial food and satis- 
ction are found. Make me immortal in the world, where there 
are manifold pleasures and joys, and wliere the objects of desire aro 
Ittained." Benfcy, Gloss, to Sama-veda, under the word nikama, 
iders svadha and Ij^iptih by " nectar and ambrosia." 


NOTE D, p. 295. 

I slinll here translate or abstract, the most important parts of 

r. Windischmann's Dissertation, " On the Soma-worship of the 
Arinus." Dr. Windisehmann begins with the following remarks : 

If we advert to the striking contrast which exists between the 
uclrine ol Zaratliustr.i, and the Briihnianical system, and to the fact 
that the former must be looked on as the work of a reformer seeking 
to preserve the old nature-worship from the mythological transform- 
ations with which it was threatened, it must api>ear as a matter 
uf the greater importance to throw light upon those points in which 
the two religions agree. For as regards those conceptions which 
existed before the two systems had developed their opposing princi- 
ples, wc may reasonably assume that they were possessed in common 
long before the separation of the Ariun race into the Indian nnd 
Iranian branches, that they formed part of the (already existing, 
and distinguisliablc) religions of the Zendavesta and tlie Veda, and 
that they had been inherited from the most primitive tradition. Such 
traditions arc, indeed, comparatively few; but the concurrence of 

ose which have been preserved, is so much the more striking ; as, 

H n 3 




finr example, Lassett (Id«L Aot. L 517) luu shown in TCgmrd to tlio 
Iranlui kgead of king Tima, coo of Vivalgliftt, « ho coitc9|>od*1b \o 
tbe Indiaa Turn, son of Yirtsrat. Yimn, howcrer, i:; regnrUccl lir 
the Medo-IVrsiaiu as the first king, lawgiver, and founder of iIm 
Iranian worship, while Yanta is looked on by tbe Brabmans as lord of 
Hades, (IL-V. L 3o. 6.), and jodge of tbi: dead, and it [a his brotkxr 
llano who pbirs tbe came part as Yiraa." " But by far the ino>l 
remarkable analogy is that whkh exists between tbe Ilaoma of the 
Zendavesta, and the Soma of the most ancient Bnibmanical book*, 
■n analog which b aot eoafined to sone few features of the legend, 
but extends to the entire Soma- worship of the early Arian race. 

'^Bmama and Soma are names etymologtcally ideniienL Both 
CMM frtMB the root sw, in Zend im; whidi significfi, ' to brget,* and 
also, bat especially in the Vedie dialect, to ' drop,' or * to preaa out 
joiee.' In later lodtao mythology Soma means the moon sad its 
doty : but in the ZendaTesta and the Vedas it signifies a oeiebnted 
plant, and its juice. This is tbe 8$clepiiis adda, or sarcoctema Tiniinalts, 
the expressed joiee of which prodoces a pecultsrly astri^ent, oareotk 
and intoxicating effect. The plant,' plucked up by tlie roots, is col- 
lected by moonlight on tbe moantains ; stripped of its leares : carried 
on a car drawn by two goats to the pbce of sacrifice (wbet« a gf(A 
covered with grass and twigs is prepared) ; crashed between atoms 
by the priests ; and is then tlirowo, stalks as well as juice (sprinkled 
with water) into a siere, whence, aAer tbe whole has been further 
fuvssed by the hand of the Brahtnans, the juice trickles into a Tessd 
(called drvna) which is placed beneath. Tbe fioid is then mixed with 
I clarified butter, wheaten and other floar, and brought into a state of 
fermentation ; it is then offered up thrice a day, and partaken of by 
the Bmbmanft The S&ma-veda is almost entirely made up of soosis 
to accompwDj ibis ceremony ; and tbe Rig-Tcda, too, contains nume- 
roos passages which have reference to it. It was nnquestiooaUy 
the greatest and the holiest offering of the ancient Indian worship. 
Tbe sound of the trickling juice is regarded as a sacred hyma. The 
gods drink the offere^l beverage; they long for it (as it docs for 
them) ; they are nourished by it, and thrown into a joyous intoxica- 
tion : this is tbe case with Indra, (who performs his great deeds 

* Compare StCTeoton's Tnadatioa of the Sbaa-vcda, 
ir|«*icdly rcferrett to ia the scqneL 




undur ita influence), with the A^vins, the Murutfl, and Agni. The 
•leverage is divine, it purifies, it inspires greater joy than alcohol^ 
it intoxicates Sukrn, it is a water of life, protects nnd nourish(.-s, 
gives h(.-alth and immortality, prepares the way to heaven, destroys 
eniMnies, Jto, Tlio Sima-veda distinguishes two kinds of Soma, the 
gri>en and tlio yellow : hut it is its golden colour which is for the 
most part celebrated. 

" If we compare all this with what the Persians say of the Haomn 
plant, we find the most surprising agreement. Haoma is the first 
of the trees planted by Ahura Mazda in the fountain of life. He 
who drinks of its juice never dies. According to the Bundtdicsh, 
the Gogard or Gokcren tree bears the Haoma, which gives health 
and generative power, and imparts life at the resurrection. The 
Haoma plant docs not decay, bears no fruit, resembles the vine, is 
knotty, nnd has leaves like jessamin ; it is yellow and white. Its 
juice is prepared and offered with aacrcd rite?, and is called Para- 
haoma. Thus in Ynvna, iii. 5, it is said haomencha para-Aaomen- 
c/ia ayiii, 'I reverence the Haoma and the Para-Iinoma.' 

"The fact that the Magians offered up a plant was known to 
Plutarch *, but what ttiis plant was is not certain The plant 

' Tlic pamgrai'h in wbii^b U)i» iufomiation is fuund (or which Wiaditchmann 
cit«» un\y a few words) is u follows: — 

Plutarch ik- IsiJ. ct Osir. 46. tiotu^ovai -y/kf »} /lif Sfoll tlfoi t6o iraOcbr«p imri. 
Xrovt, rhr ^)r a>A0(»r, t&i< 81 ipai\u» Stifuovpyif • U Si rhy nir d^rfravn dtbr, rin> 3t 
'frt/Mr ialfiova, voAoiVif ■ i(mi> Zmipiain fit d/iiyot, $r ireiToJtiiTX'^'''" *'«'" ■»'<•»' Tp«i- 
Koiv ytyorfrat, iarupovatr. O&Tol oJi' ixiSu T^f/Ur 'n^o>ui^i)r, Tinr 3i 'Afittfiiirtar' Ktl 
wpoaaiti^edrrTO rlv ftir 4eiKtrai ^onl liiXiara riv oicrfrijTii', rbv it tfitraJuir axir^ Kal 
a-jfoia niaov Si i^alv rii- Vll6prir tlrarlih Kui yiifprii/ n/^<ra4 rbr fitolriiy iroiii^ovaar • 
ttlttii* mJ' t^ tiicrcuadviir nol XV"""^*"*. *^V " i-rorpSxaua ttol ff«w9poFTii, TliuryAp 
TiKttKiJirTojrTti IfUDtu «oAou^VrT)K «> !<X^4), rhy'/iiv afoxnAuviTQl Koi T^** VKirruV (FTa 
fdlayrti a^«Ti XDioi; a^a.yiino%, f/j i6wov d>^AiOK iK^ipovai Kal pirrovai. Kol ykp 
Tar ;ftvTir rofiti^oua-i ra fii» tou iyaeav d*ou, rk Si rau wvcau ialfuiros (fiw * wal 
riir l^uuy, iairtp Kuyat koI 6priiat <ccil x'P'^'"^ /x'*'»''*< "^'^ iyoBov • tou Si ^v\ou 
Ttlit irviptut tlrcu, Sth (col riv uriiVoiToirAfiirroi;! tiiSatitoylfovat. 

" For sonic ihink ilmt tlioro are two k'>*1*> "* '' **''* opposed in their fnnctions, 
the oiii; the framcr of good olijccta, llio other of bad. Some call the moro excellent 
licing God, and ihc other Demon; iis Zoroaster, the Mogian, whoj* related to 
Imvo lircd 5000 year* before tho Trojnn war. lie called the one Oromaae«, tho 
otliLT ArimunitiB, and declartd that the former rcsomblcd light mod of all icnsiMo 
things, and the latter darkness and igaoranco. He also said that Mithras was in- 
tcrmtdiatc Itetwccn thcin. This is the rci«son why the Tirsians call Milhrn* tlia 
mediator, lie iau(;ht Ihcm to sacrifice votive and ihaak-ofTvrings to iho one tOro* 

u It 4 



seems to Bave changed with the locality : and tho somn-plant of 
the Indiane does not appenr to be thti same as the hnoma of t!>e 
PersianB; at leiist the latter ntHrin timt their sacrifidal plant does 
not grow in India. Abura Mazda causes the white haoma to grow 

among the numerous kinds of trees. A constant npitellatioo 

of tlie iiaoma is the goKl-coloured {znirigaono) jur<t as in tlie Veda. 

" But these are not the onlj points of resemblance between the 
Soma worship of the Indians nnd Persians. There is one other very 
important particular in which they both agree. In the Vedu, 
Soma is not merely a sncred sacrificial bevernge, but also a god. 
This is proved by numerous passages of the Veda (Stevenson, 
p. 98) i and in particular by the splendid hymn to Soma, Rig-vedn, 
i. 91. Precisely in the same manner, Iiaoma is, in the Zeodareata, 
not a plant only but also a powerful deity ; nnd in both vrorks the 
conceptions of the god and the sacred juice blend wonderfully witli 
each other. The roost important passages regarding this peri«mfied 
llnoma are to be foun<l in the 9th and lOih sections of the Tains, 
which are explained by striking analogies in the hymn of the Veda 
just referred to. Tlie 9th section begins thus: * In early morning 
Ilaoroa came to Zurathustra, who was consecrating his sacred fire, 
and repeating prayers. Zaralhustra asked liira " What man art 
thon, whom I see to be the most excellent in the whole existing 
world on account of his immortal life ? *' Hereupon Haoma, the 
pure, the remover of sickness, answered me, " I am, O Ziirathuslni. 
the ptire, the remover of sickness. Invoke me, holy man, pour ran 
forth to drink, celebrate me with praisCj as formerly the holy men 
used to do." Then Zarathustra said, "Bevercncc to Haoma.*"* 



nifU!c»), and to the other gloomy oUalions to avert bis wrath. For after ponnding 
a certain Ijorb callcJ limomi in a niortnr, ilicj- invoke Pluto and darkness: and when 
they liaTe mixed it «itli (he blood of n slaughtered wolf, thcj carry it to a !Bnlr<« 
spi.t nnd cast it away- They also regard certain [ilants as belonging to the good 
deity, and others to the evil demon ; and some animals, as dogs, and bird«, uid 
Ledgfhoi^, to tho foruier, (and others as) sca-orchins, to the latter; and iliey feli- 
citate those who have kilted the greatest nnnihcr of these last." 

* Compare Spiegel's translation ufihe same iias^age, and it« continnation, AveMs 
ii. 68, tr. lo note 4 he remarks; " Haoma, like various other deities of ibe Avesta, 
is regarded as at onoc a personal god, and as the thing on aceouni of which this god 
was imagined. Hnonm is at once a Yaxata and ■ drink. Tlie original identiirof 
the Indian Soma with the Hiiuma of the A vesta has been excellently shown in F. 
Windisihrnunn's disscrliUion. Among Iwth nations the healing power of the 
Uftoma is promineDlly notiecd, but .^mong the Piursis it is particularly the wbit« 





laoma is nerc callea * remover ot beat, or sickness,' and in 
Mimo vray Soma is said in Rlg-vedo, i. 91. 12, to bo amlvaha ' tlie 
destroyer of suffering/ This passoge of the Yaina clearly shows 
how, as I have before mentioned, the separate ideas of the god and 
of the juice are blended. Haoma desires that he himself shall be 
prepared for sacrifice. 

" This passage is followed by a specification of the four original 
worshippers of Haoma. The first was Vivaiihat, who prepared the 
ctlestittl beverage hunuta, and in consequence obtained a blessing, 
and the fullitmcnt of his wish that a son should be born to him. 
This was King Yinia, the most glorious of men, in whose rcnlra 
men and animals never died, wnti-T and trees never dried up, food 
was superabundant, and cold, heat, disease, death und devilish envy 
were unknown. 

'' What lias before been said of Yiina shows the importance of this 
passage. The worship of Ilaomu is placed anterior to Yima, t. e., 
to the commencement of civilization ; and in fact is declared 
to be the cause of that Imppy period. The Rig-veda also refers to 
this high antiquity of the Soma worship, when (i. 91. 1.) it says of 
Soma: 'By thy guidance, brilliant (Soma), our courageous fathers 
have obtained treasures among the gods.' Like Vivanliat, the two 
next worshippers of Haomn, viz, Alhwya and Thrita, also obtained 
offspring, — ThraC'taonO and other sons who destroyed the Ahrimaniaii 
monster. The heroic age of the conflict of Light is thus referrtd 
back to Haonia, whilst in the Rig-veda (i. 91. 8), Soma is invoked to 
' deliver from destruction, to sufTiT none of bis friends to perish \' and 
(in verse 15) to protect from incantations and fiom sin; and in the 
Suma-veda (Stevenson, p. 259), he is said to drive away the Ra« 

'*It is interesting to remark, ihnt while Thra'tnonri is said here to 
have been bestowed by Haoma, tlie SHiua-veda names a Kislii Trita 
aa an oflvrer of Soma. 

" The fourth worshipjter of Ilaouia is Pourusailpa, the father of 
Zaratliustrn : his reward was the birth of this illustrious son, tho 
promulgator of tho anti-demonic doctrine. Here also the ancient 

la which imparts tmmoitality. The lodinn plnnt is the nsclcpias acidat 
^Punian is not (lct«rinincd. Both nations notice that the plant grew on uiuun- 
tiiiiu, anJ urigiHiUly, ax Ic.ut, it must liavc been the auinc plunt which both cm- 




leg«tHl confirms the priority of the Haoma worsliip to the ZoMMttka 

" ^Vhen ZnralLiidtra lias thus loamt that ho owes his own cxtstrnee 
to Ilaomn, he celebrates his praises ; and the epithets whicb h« iMre 
applies to the go<l, agree in n remarkable way with those of 
Veda. Some of these parallel epithets are hvareia, ZeDd,3=«ctfr 
Sanskrit (R.-V. i. 91. 21), ' giring light ;' veretArnJao, Xei>d,=^rri- 
tra/ia, Sanskrit (R.-V. i. 91. 5), * destroyer of enemioa ;" btikkra 
Zend,= suAratuh, Sanskrit, (R.-V. i. 91. 2), ' offering good merit 
or ' wise,' or ' strong.' The blessings supplicated by Zarathnstn 
frum Hnomn also agree in many points with those which the Vedie 
poet nsks from Soma." 

It is not necessary, however, to pursue the subject farther. I 
refer the reader, who wishes further details of this sort to Dr. Win- 
dischmann's dissertation itself. 

I copy the fullowiog remarks on the Soma worsliip from Mr. 
Whitney's "' Main Results of the later Vedic Rcsenrclies in Ger- 
many." (Journal of the Amer. Orient. Society, iii. 299. 300.) The 
*' hymn5, one hun>]red and fourteen in number [of the 9th Book of 
the Rig-vidn^ ore, witliout exception, addressed to the Soma, and 
bein^ intended to be sung while that drink was expressed from the 
plant that afforded it, and was clariRed, are called pdcamauyas 

' purificalional.' The word soma means simply '«xtnel,* 

(from the root»< to expres.«, extract), and is the name of a bcT«r«g6 
prepared from a certain herb, the asclepios acida, which grows 
abundantly upon the mountains of India and Persia. This plantc 
which by its name should be akin to our common milk-weed, 
fornisht's like the latter, an abundant milky juice, which, when 
ierraenled, possesses intoxicating qualities. In this circumstance, 
it is believed, lies the explanation of the whole matter. The simple- 
minded Arian people, whose whole religion was a worship of the 
wonderful powers and phenomena of nature, had no sooner perceived 
thnt this liquid had power to elevate the spirits, and produce a temporary 
phrcnzy, under the influence of which the individual was prompted to, 
and capable of, deeds beyond his natural powers, than they found 
in it something divine ; it was, to their apprehension, a god, endowing 
those into whom it entered with god-like powers ; the plant which 
afforded it became to ihem the king of plants ; the process of pre- 
piiring hfdy sacrifice ; the instruments used therefor irer 


sacred. Tlie liigl* anti<iiiity of this cultus is attested by the re- 
feri'Dci'8 to it found occurring in the Persinn Arcsta ; it scorns, 
liowuvcr, to liavc received a new innnjlse on Indian territory, ns the 
pavamani/a hymns of the Vedo exhibit it in a truly reinnrkiible 
stale of development. Soma id there addre-^sed as a goil in tiio 
iiiglipst strains of n<Iulntinn and veneration ; all powers belong to 
him ; all blessings are besought of him, as his to bestow. And not 
only do such liynins comjiose. one whole book of tlie Rik, uiid occur 
scattered here and there through other portions of it, but the most 
numerous singSc passages Hnd relerences everywhere appearing, 
show how closely it had intertwined itself with the wliolo ritual 
of the Yedic religion." 

I^asAen remarks in reference to the atBnities of the Iranians and 
Indians (Ind. Ant. i. 516) : "It sliotild first be recollected that the 
Zendavcsta shows us the [Irani.nn] doctrine not in its original, 
but in a reformed shape ; a distinction id made between the men of 
the old and of the new law ; and wo may conclude that the points 

b wherein the Bralimanical Indians and the followers of Zoroaster 
^incide, belong to the old, and those in which they ditTer, to the 
Hew. system. Of the beings who are the objects of veneration in 
the Avesta, it is the seven highest, i. e. Ahura 3Iazda and the 
Amesha ^pentas, who are peculiarly Iranian ; their names are 
unknown to the Brahmans ; the Vcdas recognize no class of seven 
divinities of the highest rank who are of the same character. On 
the other hand there is no trace of Brahma among the Iranians. 
The fundamental principle of the Zend doctrine, the dualistic 
Bipnralion of the good and evil principles in, in like manner, 
foreign to Bralmiatrism. But there are, nevertheless, other deilief:, 
who are equally venerated in the Zendavesta and the Veda, viz., fire, 
the 8un,»tlie moon, the earth, and water; a fact which indicates 
that both religions have a common foundation." 

Lassen also treats of the legend of Yima, and of other points of 
connection between the Indian and Iranian religions, pp. oJ7— o2G; 
and then observca: "The common reminiscences of the Eiistern 
Iranians, and the Arian Indians, cannot be explained from any 
communication?, such as neighbouring nations might make to one 
another. On the contrary, we perceive sometimes a varying, some- 
times a contradictory, conception of important traditions and 
appelhitiona, which is only intelligible if wc presuppojc an earlier 
ogrucment, which had, in part, become lost and modilied in the 

course of time, aHer the separation of the two nations-, &nd in 
part liad become converted into a contradiction by a division in 
their opinions. Even this contradiclioa indicates a closer coanectioa 
between the two nations at an earlier period." 

See also Professor R. Roth's articles in the Journal of the German 
Orieutal Societj, for 1848 (pp. 216, ff.), 18.50 (pp. 417, ff.), and 
1852 (pp. 67, ff.)t on the legenda of Feridiln and Jemshid, and on the 
"highest gods of the Ariun nations-," and also his paper on 
Nabanazdista, at p. 243 of the last named rolume. In the vol. for 1848, 
p. 216, he proposes to show by an example, " how the Veda and the 
Avesta dow from one fountain, like two streams, the one of wbicli. 
the Vedic, has continued fuller, purer, and truer to its original 
character; while the other has become in many ways polluted, has 
changed its original course, and consequently cannot always be 
followed back with equal certainty to its sources." See also Professor 
Jluller'a "Last Results of the Persian Researches." 

NOTE E, p. 315. 

" The question regarding the lime and place of the separation is of 
yet greater importance than that concerning its cause. For our 
present inquiry, it is of less consequence to determine the place, tlian 
the time, of that separation. As regards the region where the 
Indians and Iranians dwelt together, several suppositions may bo 
made. The Iranians may have immigrated into the Panjiib along 
with the Indians, nnd have turned thence in a westerly direction. 
But, on the other hand, the Indians might have separated themselves 
from the Iranians, and travelled towards the east. Thirdly, it might 
be conjectured that tlio two races had parted from each otEer before 
they migrated towards India and Iran. Be this as it may, though 
we ore unable to assign any date to the period of the .separation, we 
must decidedly hold it to have occurred before the Vedic era. No 
Buch relation exists between the two races as would justify us in 
assuming that the Iranians formed one community with the Indians 
during the Vedic period. The great majority of the Vedic gods nnd 
of the Vedic conceptions are as little known to the Iranians, us the 
Iranian conceptions are to the Indians. The ideas which are common 



to both nations may bo most easily and satisfactorily explained by 
supposing thorn to have been developed in the ante-Vedic period." 
Spiegel, in Kuhn and SthleicUtir'a Bi^itiiige zur vergl. Sprnchf. 
vol. ii. pp. 3 and 4, 

NOTE F, p. 317. 

" It is the common view that it was religious grounds which 
occasioned tlie separation of the Indians and Iranians. This opinion ia 
supported by the fact that the names of several divinities which have 
a good signification among the one people, are used in a bad sense 
by the other, and vice versa. Thus the Indian deva (god), has become 
a demon among the Iranians under the form of daeva ; and Indra as 
Aiidra has experienced a .similar degradation. It must not be denied 
that these differences of conception maif have had their foundation 
in a religious schism between the two nations; but this opinion 
should not be regarded as more tliun a probable conjecture, or held 
to be a historical fuct, which follows from the linguistic data with the 
same certainty as the proposition that the Indian and Iranian nations 
bad originally the same common ancestors. Other possible modes 
may be conceived, in which this opposition may have arisen ; such 
as the internal development of the Iranian people itself. We have 
only to reflect on the case of the Gorman religions, and their ancient 
gods, who, in presence of Christianity, came to be regarded as 
evil spirits. Dualism, with its rigorous consequences, was a power 
which operated in Iran in precisely the same manner as Christianity 
did in Germany. This dualism, which was a result of the particular 
development of the Iranian people, was compelled to make room in itfl 
system, in the best way it could, for those forms of religious belief 
which it found already in existence, and did not feel itself strong 
enough to discard. Many beings formerly regarded as gods, may 
thus havo been transformed into evil spirits, because they stood in 
too strong a contrast to the new moral system. It appears to me 
that the opposition between the religious conceptions of the Indians 
and the Iranians grew up gradually, and not all at once, in consequence 
of a reform of Zorathustra, as some have assumed." Spiegel, as 
ftbove, p. 3. 




NOTE G, p. 336. 

l*toleiny, Geogr. tL 16. has the following notice of Ottorocorra : — 

'Opi| ci iti^iKiynfy Zifpm})', rn rt caXov/icfa 'Ayrt^a, tr. r. X. ** TliC 
country of Serica is surrounded by mountain ranges," vie., the AnnjJ 
Lion, the Auxncian, the Aamtrseao, the Casian, the Thagurinn, and thrt* 
of Emodus : 

Kai TO KaXovptyov 'Orropok-uppa^, o5 ra iripaTa l-ri-^ii ^oipa^ p^O Xr 
tra'i por X9. "[Another of these ranges] is that called OttoroeorrU|. 
the Hmita of which extend from 169° 36' to 176° 39' east long.'' 

Ta fiiy ovy apkTiKUTtpa rfjt ^ijptinjc KaTavifiovrat idri] ' KyBpttTco^yir, 
*' The northern parts of Serica are inhabited by the tribes of the 
Anthropophagi " (men-eaters). The Annibi, Sizyges, &c, follow : 

Kai fit<rrifi€piywTaroi wapit rd 'H/i<i>2d n'aS Si;pi«:d opij OrropoKu^pat, 
*' And soutlicrnmost of ail, near the Emodian and Surican tnountaii 
dwell the Ottorocorra." 

Among the cities of Serica is mentioned Ottorocorra, in east 
long. 165*' 37' 15". 

Ottorocora is again alluded to by Ptolemy in book viii., in his r<!> 
marks on the eighth map of Asia : — 

'II 'OrropoKopa ri)v fityi'trrrfy ifftipay i^ti oipCiy iJyo tyy^Ta' eat 
lii<TTr)t:tv 'AXilayiptius TTpiis tv vpaic iirra, "The greatest length of 
the day in Ottorocora is nearly 14§ hours. It is distant from Alex- 
andria seven hours towards the east." 

See, for an account of Ptolemy's geographical 8y8tem,L.a«)en's Ind. 
Ant. iii. 94, ff. ; and for the position of Ottorocorm, the map at the 
end of the same volume. 

KOTE H, p 344. 

In regard to Airyanem Vaejo, Lussen observes, (Ind. Ant. i., 
226, IT. ) : " If we assume that the Arian Indians and the Iranians 
had originally the snmc common abodes, out of India, we should 
expect to find a tradition on the subject among the latter people rather 
than among the former. We have already said that the Indians have 
no longer any legend of this sort, though they imagine a sacred 
region nnd the seals of the gods to exist to the north of India.' 


The Iranians, on the contrary, clearly designate Airyanetn Vaejo oa 
the first created country : this they place in the extreme east of 
the Iranian lughlands, in the region where the Oxus and Jaxartes 
take their rise. Tiiia country was afflicted with winter by Aliriman, 
and had only two mouths of summer, as if the tradition of a de- 
crease in the earth's temperature still existed. We must suppose 
the cold highlands on the western slopes of Belurtag and Must«g to 
be meant," &c. [The next paragraph will be quoted in Note 
K]. The following remarks are added : " It suffices to have made it 
probable that the earliest abodes of the Indians and Iranians are to be 
sought in the extreme east of the Iranian higlilands ; but we may 
assert it to be more than probable that the Indians were derived from 

$ome part of tiio Iranian country The means of arriving at 

a conclusion on this subject are uncertain ; we can only form con- 
jectures from a review of the later geographical positions occupied by 
these nations ; and we fire thus led to fix on the country lying be- 
tween the Caspian sea and the highlands before mentioned, as having 
been most po}bably their ancient scats." . 

See also Ariiina Antiqua, p. 134, quoted in Note K. 

Baron von Bunsen also treflts of the first Fargnrd of the Vendidad 
in one of the Appendices to his Bibel-werk, vol. v. pp. 315,316. 

'Laaccn'g idea, qnoteJ in p. 349, thot the "'daily prosptct of the snowjr snm- 
mica of the Himalaya, glittering fiir and wide over the plains," iind the know. 
Icdgo the Indians had ot the "tuble-Isnd beyond, witit iis extensive and tranquil 
domains, its dear niid cloudless ftky," &&, would point oat the " north o« tho 
abcido of the gods, and the theatre of wonders," is coniirmedby IIomcr'«do«cription 
of Olympns, Odyw. vi. 42, (!'.: 

OCKi>fLit6vV , 304 ^>eui\ dcwr iSoj ia^aKkt eu'c) 
^T-fifAtvou' oth* dvr ftoiat TiydtrtrfTcu, otrrc iror* Ai^Sp^ 
Afi'ifrai, q(t* x^y iiri.irt\yaTtu- dA\k fid^' nWp>j 

" Olympus, where they My the blcsied goda 
Repose forever in secure abodes: 
No Ftormy blusts athwart those summits sweep, 
No showers or snows bcdcw the sacred steep ; 
But cloudless skies serene above nre spread, 
And golden radiance plays around its head." 

ThU, however, is the ideal Olympus. The monniain is styled iyimn/^, 
"snowy," in Iliad i. 420, where the toholifl.<it explains the discn^piincy by saying 
that the epithet "snowy " applies only to the parts below the cIuikU, tho summits 
U'ing above the clouds and exempt from rain or snow. 



I abstract the following remarks : — " The s&cred books of Zoroutn'ii 
foUowerH begin with a descriptioD of the gradual diffusion of tin 
Arian races of Bactria, as far sis the Penjab. The account of theme 
migrations of the Bactrian Arians is preceded bj a remarkable re- 
ference to the primeval country in the north-east, from which their 
forefathers removed to their present abodes, in consequence of a 
great natural convulsion. It appears that that once perfect primcTml 
country, Airjano, liad originally a very mild climate, until tiiA 
hostile deity created a powerful serpent, and snow ; so that only two 
months of summer remained, while winter prevailed during ten. 
The country next occupied was' Sogdiana ; and the third Bactria. 
The progress of tlie Ariana with their civilisation is, as it were, tlw 
march of Ahura Mazda, the lord of spirits. This advance has a 
historical import, for all the countries which are specified form • 
continuous scries, extending towards the south and west, and in all 
of tiiem the Arian culture is discoverable, and even now (io part ex- 
clusively) predominant. The first named country can be no other 
than that where the Oxus and Jaxartes take their ris« ; the taU«> 
laud of Pamer, and Khokand. Assuming the genuineness and sn- 
tiquity of the Bactrian tradition, we liare here a testimony deserriag 
of the highest consideration, to the historical character of the 
Biblical tradition regarding the interruption of the life of the 
Asiatic population by a great natural convulsion confined to tlut 
locality. The country lying between the highlands just mentioned 
to the east, and the mountains of Caucasus and Ararat to the west, with 
the Caspian Sea in its centre, is regarded by scientific geologuts. such 
OS Humboldt and Murchison, as the very region where the most recent 
convulsions of nature have occurred. The snow and the pro- 
longed winter alluded to in the oldest Arian tradition must have 
been the result of an upheaving of the land into mountains." 


NOTE I, p. 358. 

" Now there c*n be no doubt that by the Kophen is to be understood 
the Kabul River ; for^rrian says, that having received the Malamin- 
tus, Siiastus, and Garo:us, it mixes with the Indus, in the country 
of Peukelaotis ; and the latter part of Alexander's operations west of 

the Indus, shortly before he crosses that river, ore can-icd on in the 
BAme district along the Indus and the Cophen." — Wilson, Ariann 
Ant, p. 183. "The united stream [of the Punjkora and Sewat] is 
called either the Punjkora or Sewat River ; and tliis may explain 
why Arrian, in his Indica, speaks erroneously of a iSuastus as well 
as a Garcsus, whilst in Ptolemy wc have no other river than the 
^Dastus descrilied." — Ibid. p. IfX). «' Alexander crossed, according 
to Arrian's narrative, four rivers before he reached the Indus ; and 
these, the Kophcn, Khoes, Eunspla, and Garoeus, we have still in 
the Punjsliir, Alislitmj, Khonar, and Punjkora. . . . Thus even 
Arrian is a better authority as an historian than as a geographer, for 
he descrihos in the latter character, the Kophen as bringing with it 
to the Indus, the Malamantus, Sunstus, and Garccus ; two of which 
he does not name at all in his narrative, and of which the third is 
probably the same with the second." — Ibid. p. 194. La.ssen, on the 
contrary, holds that Ptolemy is in error. " It must surprise u?," lie 
remarks (Ind. Ant., iii. 129), " that, of the rivers of Eastern Kabul, 
Plolomy mentions only the Suastos, and passes over the Garoias'* in 
silence, though this river must have been known to him from the 
accounts of the writers of the Macedonian age, who, however, are 
tvrong in making the Suastos to unite not with it, but with thtf 
Kophen. This is the more surprising, as Ptolemy is acquainted 
with the region called Goryaia after that river- . . . Ptolemy is 
thus misled into making the Suastos rise too far to the north.'' See 
I also Lassen's Ind. Ant. ii. 668, 669. 
I In any case, the existence of a river in the Kabul country, called 
I Suastus at the date of Alexander's expedition, is undoubted. 

^P NOTE J, p. 367. 

In the Asiatic Researches, vol xv. 108, Professor H. H. Wilson 
translates parts of a long passage in the Karna Parva, or viii"> 
book of the M.-Bh., verses 2025, ff., in which the manners of tlie 
BiLhikas, Madras, Gandliaras, Araltas, and other tribes of the Panj- 
iib are stigmatised as disgraceful. The same text is quoted nnd 
translated in the appendix to M. Troyer's Rajataranglul, vol. ii. 
pp. 549, ff. I will cite a few specimens from this passage. The 

' The Ancient name vrta G«uti; the present is Piuijkora. 
I I 



country irlicrc the Bttliikas dwell is tlius defined (versea 2029, flf.) : 
Vtihishhrita Himaviitn Gangat/a cha vahinhhr'tlnU \ Svr Yii 

mamiya Jsuruhfliftrcrfa chapi ]/c | PancJiitnam SmtJ/i' lam 

nadinam yeantaraipiiih | Tan dhumtaVfiliyaa aiuckm Bahikam 
parivajjayel. "L«t every one avoid those impure Bs,Likas, who axt 
outcuDit^ from righteousness, who are iJtul out by the UioiaTut, 
the Gangtt, the Sarasvati, the Yamuna, and Kurukshetrn. and trho 
dwell between tiie five rivers which are associated with the Siudbu 
(Indus), as the sixtii." 

Their women are thus described {v. 2035): Gayanty athacia 
nrityanli tlriyo mattah rivasasafy, Naguragaravapmhu vafiir mat' 
ytmuleptuiah, &c. "The women drunk and undressed, we«rin|} 
garlands, and perfumed with unguents, sing and dance in publie 
places, and on tlie ramparts of the town," &c. ; with much more to 
the same effect. 

Agnin (v. 2063, C); Panchanadf/o vahanty eta yaira 
parvatat, Aratfa naiiia Dak'ika na ttshv Aryo dvyhnm roMf 
(v. 2068, ff.) AratVi noma te dcia Biihlkam nama tajjuliun ; Brak- 
manupasnda yatra tulyakalah Prajapatek. I'edo rut tetham roi* 
dyancha yrtj'io yajanam eva cha, VratyaMam daiatniyanatn annam 
dtvn na bhuvjate- Prasthala Madra-Gatidliara Araftd ndmutcify 
Kha-iah, Vasati-Shidhunauvira Uiprayo'tikutfitah. '* In the region 
wliere these five rivers flow after issuing from the mount&ioa, dwell 
the BtihikaB, called Araf (as ; let no Arya dwell there eveo for two 
days. . . The name of the country is Aratta ; the water of it is called 
Saliika. There dwell degraded Brahmans, contemporary with Pn^a- 
pati ^^?). They have no Veda, no Vedic ceremony, nor any sacrifice. 
The gods do not eat the food olTyred by Vratyas and servile people. 
The Prasthalas, Madras, Gandharas, Arattas, Khaaas, Yasatis, and Sin- 
dhusauvTras are nearly all very contemptible." Again it is said of the 
same country (v. 2076, if, ) Tafm rai lirtihmaito Lliutva tato bhavati 
Kfhaltriyah, Vaiii/ah Sudra&cha liahikas tato bharati tiajiitak, 
Kfipitaicha tato bhJ/ttfi pttnar bhavati Brahmanuh. Dpijo MUtcA 
chn tntraiva pnnar diixo 'bkijayate. Bfiavalyekulk huh Viprak 
prairUhtah kamachiiruiah, Gandhara Mndrakatchaiva BakikA- 
iehiUpnchetasnh. " Tliere a Baliika bom a Brahman, becomcu after- 
wards a Kjhattriyn, a Vaiiya, or a Sudra, and eventually a barber. 

' These expressions ."dluirmO'MA^un" and " raAi*A.kriuii " »eem to contain 
* "'— on ilid nnnio of the Bakikat. This tribe is mentioned iii the S. P, 
boTC, p. 213. 



And again the barber becomes a Brahman. And once again the 
Brahman becomes a slave. One Brahman alone is born in a family 
among the senseless Gilndharas, Madras, and Balitkas ; the [other 
brothers] act as they will without restraint." 

In the Rjijatarangini, i. 307, ff., the Gfindliiira Brahmans ore thus 
characterised: — Agrafiarnn jagrihirc Gandhara-Lrfihrnattas tatah, 
samanaillai tasi/aiva dliruvan te 'pi dvijadhamiih, Bhaginivar- 
ga'Setmhhoga-nirlajja Mtechhavaniajah, Snuiha-sangati-sakdicha 
daradih santi papinah. Vustubltavais tal/ta bhadya (bftatya?) 
bharr/ija-vihrayakarinah, &o. " Then the Gandliiira Brahmans 
seised upon rent-free lands; for these most degraded of priests were 
of the same disposition as that tyrannical prince. These sinners 
sprung from Mlechhns, are so shameless oa to corrupt iheir own 
sisters and dtiughters-in-law, and to offer their wires to others, 
hiring and selling tliem, like commodities, for money." 

M. Troyer remarks (vol. ii. 317) that *' the inhabitants of tho 
Panjab are in this passage of the M.-Bh. named generally Bahikas 
and Arnftas, wliile the Gandhilraa are associated with the different 
tribes into which these inhabitants are subdivided, such as the Pra- 
sthulas and Madras, in such a way that it can scarcely be doubted that 
tlie former (the Gaiidhiiras) lived in their neighbourhood, diffused like 
them between the six rivers of that country. . . The Sindhu-Gan- 
dharas mentioned, Raj. i, 66, lived on the Indus." 

And Wilson sajs, (As. Res. xv. lOo): "According to the M.-Bh. 
the Gandhari are not only met with upon crossing tho Sctlej and 
proceeding towards the Airavati (Ravi), or where Strabo places 
Gandaris, but they are scattered along with other tribes throughout 
the Panjab, as far as to the Indus, when we approach Gandaritis. 
According also to our text (Raj. i. 66) one body of the Gandhari 
oppear to occupy a division of their own on the last river, which is 
named after that very circumstance, Sindhu-Gandhar, and these may 
> liare extended westward as far as the modern Candnhor." In liis 
Vishnu-Pur., p. 191, note 83, the same writer says of tho Gan- 
dliiiras : " These arc also a people of the north-west, found both on 
the west of the Indus, and in the Punjab, and well known to classical 
authors as the Gandarii and Gandaridso." See also Rawlinson's Hero- 
dotus, iv. pp. 216, 217. 

The delinilion given of the Gandharn country in p. 356 may be 
modified accordingly. 

I I 2 



NOTE K, p. 370. 

Lassen, Ind. Ant. i. 527, remarks as follows : " The oitmion 
the original scats of these [the ludian and Iranian] nations &re to 
be Bought here in [the extreme east of the Iranian highlands], re- 
ceives great coniirmation from the fact, that we find branches of 
these nations ou both sides of this lofty range ; for the nncieot in- 
habitants of Caaghar, Yarkhand, Khoten, Aksu, Turfan, and 
Kliainil nre Tfijlks and speak Persian ; it is from this point only 
that they arc di^'used towards the interior of upland A^ia ; so that 
their most powerful germ seems to have been planted on this rsoge." 

And Professor II. II. Wilson says: " Without extending the limits 
of India, however, too far to the north, there is uo reason to doubt 
that the valleys of the Indian Caucasus were properly included 
within tbera, and that their inhabitants, as far as to the Puner 
mountains and Badakhshan were Indians, who may have been at 
first tributary to Persia, and afterwards subjects of some branched of 
the Greek rnce of Bactrian kings." — Ariana Antiqna, p. 134. 

Badakhsliiin is the country on the banks of the Oxus, near it» 
sources, situated between lat. 36° and 38° north, and lying cAstwunl 
from Balkh. Pamer lies in the same direction. See the map in 
Ariana Ant., p. 211, or that of Ancient India in Lassen's Ind. Ant., 
vol. ii. 

NOTE L, p, 463. 

A question of considerable interest here presents itscli, on which 
it may be desirable to make a few remarks, viz., whether the in- 
digenous or non-Arian races, who now speak Tamil, and the other 
languages of the southern group, are of the same family as tho60 
tribes who were brought into contact with the Aryos on their Hr&t 
arrival in India, and the remains of whose languages have survived 
in the vernacular dialects of northern Ilindusthan. The lat« 
Rev. Dr. J. Stevenson appears to have been of opinion that the 
non-Sanskrit element in the northern oud southern TernaculAr 
dinlccts was originally to a great extent tlie same, and that tba 




people w1)o spoke tbem also belonged to one race. ITe remarks 
(Art. vii, Journ. Bombay Branch Roy. As. Soc, No. XII. for 1849^, 
"It is usually taken also for granted that between the non-Siinskrit 
parts of the northern and southern families of languages there is 
no bond of union, and that the only connecting link between the 
two is their Sanskrit element. It is to this last proposition that the 
writer of this paper demurs." He afterwards proceeds : " The 
theory which has suggested itself to the writer as (be most probable 
is, that on the entrance of the tribes which now form the hight'St 
CAStes, those of the Brahmans, Kshattriyas and Waisyas, into India, 
they found a rude aboriginal population, speaking a different language, 
having a different religion, and different customs and niaiinern ; 
that by arms and policy the original inhabitants were all subdued, 
and in great numbers expelled from the northern regions, those 
that remained mixing with the now population, and being first their 
slaves, and then forming the Sudra caste. The langunge of theso 
aborigines is supposed to have belonged to the southern family of 
languages, tlie most perfect remaining type of which family is the 
Tuniil." The fundamental aiBuiites of tlie norlhcrn and southern 
languages are then discussed by Dr. Stevenson in various papers in 
(he same journal, which Appeared in the years 1851 and 1852. 
Dr. Caldwell, however, has expressed his dissent from Dr. Stevenson, 
both in regard to the affinities between the pre-Arjan races them- 
flelves of the north and of the south, and their original languages. 
(See pp. 38, ff. and 69, ff. of his Dravidiiin Grammar.) In regard 1o 
the languages he remarks (p. 39, ff.) that the hypothesis of their 
affinity does not appear tu him to have been established ; as (hough 
various analogies iu grammatical structure eecro to connect the non- 
Sanskrit element in the north-Indian idioms with the Scythian or 
Tartor tongues, yet that no «;>ffcio/ relationship of the former to thu 
Dravidian languages has yet been proved to exist. If the non-Sanskrit 
element in the northern vernaculars (p. 40) had been Dravidian, we 
might have expected to find in tlicir vocabularies a few primary 
Dravidian roots such as the words for head, hand, foot, eye, ear ; 
whereas Dr. Caldwell has been unable to discover any trustworthy 
analogy in words belonging to this clnss. Fnrthir research, he adds 
(p. 42), may possibly disclose the existence in the northern vernaculars 
of distinctively Dravidian forms and roots, but their presence does not 
yet appear to be proved ; and be therefore concludes that the non- 

I I 3 



Sanskrit portion of the northera languages emnot flafclj b 
in tbe aat daas with the eouthero, except, perhaps in t! 
of both beiof Sojthian rather than Arian. The sanM is the opinion of 
the Rer. Dr. J. WiUon, who in his " Notes on the Marathi Lnngwce," 
in th« prefiim to the aecood editicm of liolc»worth's Mantiii 
Dictioojuy (p. xxii.), thos writes : — " The Scjtbian words in tiw 
Usrathi are, in general, like those of the other Turanian toogoH^ 
more in their /brms than in their toundt, Tliej difler verj naek 
iram the rocahles of the Turanian langnagoi in the aootfa of hniut, 
(the Canarese, Telngn, Tamul, and MakyaianX the eomparison of the 
dictionaries and graaiBan of which throws tot littk light on thu 
Haratli] ; and though they maj be classed in the same tribe of 
languages, the/ eTidentI j belong to a different familjr, to • diflcreat 
Turanian immigration into India, jet to be explored bj th« eoabiaed 
labours of the philologist and the ethnographer." Begatdii^ the 
qoestioo whether the non-Anan tribes of the north and ike ioalh 
are themselves of the same stock. Dr. Caldwell remarks (pw'(2) thai 
the Dravidians maj be confidently regarded as the earliest inhahi* 
tants of India, or at least as the earliest that entered froa the 
north-west, but it is not so easy to determine whether they ars the 
people whom the Arians found in possession, or whether tbej had 
been alremdy expelled from the north by the irruption of another 
Scythian race. Without deciding this point positively. Dr. Caldwell 
is led by the apparent differences between the Draridian langaagea 
and the aboriginal element in the northern Temacularsi, !• iacliae to 
the supposition that the Draridian idioms belong to an eHiet stage 
of Scythian speech ; and if this view be correct, it aeena to follow 
that the ancestors of the Scythian or non-Arion portioB of the 
north-Indian population must have immigrated into Iikdift at a later 
period than the Dravidians, and nttist have expelled the Dravidians 
from the greater portion of north-India before they were themaelvrs 
subjugated by a new race of Arian invaders from, the iu>rih-west 
In any case Dr- Caldwell is petsttaded that it was not by the Arians 
that the Dravidians were expelled from northern India, and thali 
•s no reference occurs either in Sanskrit or Dravidian traditioa to 
any hostilities between these two races, their primitive relations' 
cnuld never have been otherwise than amicable. The pre-Arian 
Scythians, by whom Dr. Caldwell supposes that the Dravidians may 
I iiave been expelled from the nortliern proviocee, are not, he coo* 






i to be conFounded willi the Kolas, Santhals, BUill?, Doms, and 
oQier aboriginal tribes of the oorth, who, he supposes, nmy linve 
retired into the forests before the Dravidiana, or, like the Bhotan 
tribes^ have entered into India from the north-east. The languages 
of these forest tribes Dr. Caldwell conceives to exhibit no affiiiily 
irith the aboriginal element in the north- Indian vernaculars. We 
have therefore, according to the views just summarily expounded, 
four separate strata, so to speak, of population in India : 

First and earliest, the forest-tribes, such as the Kolas, Santhals, 
fihills, 8(c., &c. who may have entered India from the north-east. 

Second. Ttie Dravidians, who entered India from the north-wesf, 
and either advanced voluntarily towards tlicir ultimate seats in the 
south of the peninsula, or were driven by the pressure of subsequent 
hordes, following them from the same direction. 

Third. We have the race (alluded to at the end of the preceding 
head, No. 2) of Scythian or non-Arian immigrants from the north- 
west, whose language afterwards united with the Sanskrit to form 
the Prakrit dialects of northern India. 

Fourth. The Arian invaders who (after separating first from the 
other branches of the Indo-Germanic stock, and last of all from the 
Persian branch of that family) advanced into India, drove before 
them the non-Arian tribes who were previously in possession of the 
Punjilb and other parts of the north-west provinces of India, and 
aAer organising Brahmanical communities, and founding Brahman- 
ical institutions in the north, gradually diffused themselves to the 
east and aoutli, and eventually extended their discipline, and to 
some degree their sacred tangunge, to the remotest parts of the 

To whatever degree the details of this theory mny be capable of 
proof, tiic general conclusion, at least, seems to be undentabh', vii,, 
that the ancestors both of the Dravidian nations, and of other non- 
Sanskritic tribes now occupying different parts of India, were in 
occupation of that country before the immigration of the Arians; 
and that the former could not (as is erroneously intimated in various 
Puranic and other traditions) have been descended from the latter. 
If the Dravidian Cholas, Keralas, he, were originally Ksbattriya4) 
who fell away from Brahinanism, they must have been reconverted 
to that system ; a double process of which tbeio is no historical 

I I 4 

.-y ' 4S»^ 

• ti^!**. 



(P. 82, line 7.) 

ThQ oceorrenee of the wordyaAa*r-tMi for peUhS-i'.tmtim Ab^UUt 
^ the L«lit» Vistata (see p< 130), maj be heU to JiwKkirte^hwwwr, 
ibatthe introdnctioa of ^e enphomo cooMHiadti^ peeafiarto lUi^ 
had been hogau in northffi'n India. 

(P. ;61, line 27.), 
For "ayoBxnaya* read "ayaamaja." 

(P. 165, line 17.) 
Seo also Sajana's commentary on the verse R.-V. i. 164. 45. He 
there defines thus the words vyavahariki vak : " Bhogavishayi 
* gam anaya ' ityadirupa vyavahariki." " The common language is 
that which refers to objects of enjoyment, such as, gam anaya 'bring 
the cow ?' " These words are of course Sanskrit. 

(P. 179, lines 4, 19, 34.) 

Compare Benfey's remarks on the Vedic scholiasts, (in the Intro- 
duction to his Sama-veda, p. Ixv. Ixvi.) where he observes : " How 
high soever may be the antiquity assigned to the oldest grammatical 
and hermeneutical treatises on the Vedas, a long period appears to 
have intervened between these and the composition of the greater 
portion of the hymns, during which very much that was peculiar to 
the Vedas was forgotten. Their interpretations rest essentially (as is 
shewn not merely by the commentaries which have been alluded to, 
but also by Yaska's Nirukta) on etymology, on conclusions drawn 
from the context, and the comparison of similar passages. The oldest 


attempts at Intarpretation seem to bo contained in BrUhmanas, in 
collections of passages {nigama), in collections of worda (niffhantii) 
and in explanations (uiru/tla), of which last, two are mentioned by 
Sayana, (R.-V. vol, i, p. 45, lines 16 and 18) viz., one by 6aliapQni and 
another by Sthaulashthivi, in addition to that of Yiiska." 


(P. 203, line 27.) 

In his Hist, of lad. Lit, p. 140, Weber tells us that "in the Anukra- 
manl of Kutyuyana to the Vajasaneyi-sanhita of the Yajur-vedaj tho 
authors (rishis) assigned to the particular verses (rich) usually coin- 
cide with the authors assigned to the same verses in the Anukramnul 
of the Rig-veda j but that there are many exceptions to this rcmaric. 
Id particular (aa happens abo in the Rig-anukramani) the name of the 
author appears often to be borrowed from some word occurring in 
the verse. And in the case (a very frequent one) of a verse being 
repeated in another part of tho Vaj.-San., it is often assigned to an 
author different from the one to whom it had previously been ascribed. 
Many of tho rishis here referred to do not occur among those of the 
Rig-veda, and belong to a later stage than the latter ; and among these 
rishis peculiar to the Vajoaancyi-sanhita there are several who are 
named in the ^atapatha-briihmana as teachers." 

(P. 220, line 1.) 

In his illustrations of the Nirukta, p. 85, Roth observes in regard to 
the ftfth hymn of the Fourth Book of the R.-V. : " The author of tho 
hymn, Viimadeva, himself professes to make known a mysterioue 
and recondite wisdom, which had been revealed to him by Agni 
(verses 3 and 6)." The third verse is as follows : Soma dvi- 
barha mahi ligma/>firis/itih sahasrarcta rrishahhai tuvishman \ J'a- 
dam na gor apagijlham vividvan Agnir maht/am predu pochttd 
maHtahum || " Agni, occupying two positions, the fierce-flaming, the 
prolific, thoshowercr of benefits, the opulent, who knows the vener- 
able hymn, mysterious as the track of a [missing] cow, hath 
declared to mc its knowledge." 

(P. 225, line 10.) 
" A comparison of the grammatical structure of the Sanskrit, 




Celtic, Greek, Latin, Qerman, Letto-SIavonian, and Persian, he., 
tonches us that all these languages have a corotnon basis, or in other 
words tliat they are derived from one common original speech ; and lb* 
gradation of sounds and forms points to the Sansicrit as the lang 
wliich in general still preserves the most original form, and has < 
parted least from the original tongue. This existence of one com* 
inon original language necessarily leads us to conclude that at the 
period when it was still a living and spoken tongue, the people also 
which employed it formed one nation ; and it results that the iodi- 
vidual nations as well aa their languages were formed bj a gradual 
eeparation from the Indo-European people, and its language. And, 
moreover, the greater or less similarity of the several lasgoAgei 
among each other, and particularly in reference to the Saaskrit, 
enables us to conclude wiietlier the separation from the original stock 
took place in each case at an curlier or a Inter period." — Weber, ladiaa 
Sketches, p. 7. 

(P. 226, line 23.) 

Tiie want of resemblance between the Semitic and lndo-(jei^ 
tnauic languages in rc*pcct of tlicir ruots is too strongly asserted in 
the text. The subject is ably treated by Renan, " Histoire des Lan- 
Rucs Sfimitiques," 2nd ed. p. 434, ff. He observes that the criterion 
of the distinctness of families in languages is to be found in the im- 
possibility of deriving one from another. Thas, he says, it ia qaile 
intelligil>le how, notwithstanding their differences, all the lodo- 
Kuropean tongues may be related to the same type, and have 
sprung from the same primitive idiom ; while it is impossible to ex- 
plain how, by any series of corruptions, the Zend or the Saaskrit 
could have become Hobrew, or how the Hebrew could have L<.<otne 
changed to Sanskrit or Chinese (p. 434). It is generally recognised 
llwt there is a wide distinction between the gi'ammatical system of 
the Semitic languages and that of the Indo-European tongues, and 
that the one system could not be derived from the other by any pro- 
cedure known to comparative philology. If we except the principles 
common to all, or to most, languages, (which are nothing else than 
nn rxpresiion of the laws of the human mind,) there ia scarcely 
any grammatical nu>ch.nnism of importance which ia common to the 
two familicjs ^p« 444). But in the clas^idcation of hingoag^s^ gram- 






maticnl nremucU more important than lexicographical consiJeratioDs 
£t. e. the structure of a language ia of muclt iiioru cousequence thnii 
the words of which it is connwsed]. Many languages could be 
quoted whicli have enriched or renewed their vocabulary, but very 
few which have corrected their grammar. Grammar i8, therefore, 
the essential form of language, that which constitutes its iudi- 
viduality (pp. 447, 448). On the other Land M. Renan admits 
that the Semitic and Indo-European languages have a considerable 
number of roots which are eonimon to both, independently of 

:ch as they have borrowed from each other within the historical 

riod. But he doubts whether this circumatance is sufficient to 
prove the primitive unity of the two families, and scarcely ventures 
to hope that a demonstrative result will ever be attained on this point. 

he greater part of the roots common to the two families owe their 
imilarity, he considers, to natural causes, as they belong to the class 
of biliteral and monosyllabic ouomatopceias, which reappear in the 
trilitural radicals actually existing, and in which original sensations 
appear to have left tlieir traces. Is it at all strange, he asks, that 
in order to express outward action, the primitive man, still sympa- 
thising so closely with nature, and scarcely separated from lier, 
should have sought to imitate her, and that the same objects should 

ve been universally imitated by the same sounds? (pp. 449, 450.) 
W. Renan Illustrates these remarks by a number of instances, but 
admits tiiut, among the roots which appear to be common to the 
Semitic with the Indo-European languages, there arc a certain number 
in which the reason of the onomntopooia is more difficult to seize 
(p. 452). He concludes that in tiie present stnJe of philological 

lience, a sound method of theorising requires us to regard the Semitic 
and Indo-European families of language as distinct (p. 457); while 
nt the same time he remarks, that nothing wliich he has adduced in- 

alidatea the hypothesis of a primordial affinity between the races 
Ijy whom the Semitic and the Indo-European languages respectively 
were spoken (p. 431). For details I must refer to his work itself. 

I (P. 274, note 28.) 

See also Benfeya Complete Sanskrit Grammar, p. 20, wb«re !( is 
aid : " ^ appears never to be original in Sanskrit, but to liavo 
arisen from the weak aspirates ^ V ^ ■ This derivation can bo 





illustrated by many examples from the Vedas, or from tlie klnili 
languages. Compare llie Vedio dughaita from duh ; tamUyka 
dihi sadJta for suha ; grabk for grah^ 

(P. 296» line 6.) 

Professor Spiegel Las, however, subsequently retracted the opinifl 
here expressed of the identity of the Iranian Airyama with the Indian 
Aryaman. In Kuhn and Schleicher's Beitriige zor VergL Spraehf. 
i. 131, ff, he says : " I have in my note on Vend. xxiL 23 (p. 266), 
regarded the Airyama of the last chapter as the Vedic Aryaman. This 
comparison is only in part correct. It is true that, letter for letter, 
Airyama is the Sanskrit Aryaman, and therefore the phonetic affinity 
cannot be doubted. It does not, however, follow that the significaiion 
most therefore be the same. If, as is supposed by many, the Iraniaos 
had issued from the bosom of the Indian people, if the entire culture 
of the Indians, as exhibited in the Vedas, had been the basis of 
theirs, this assumption would be less questionable. Bat acoor^ng 
to my view such is not the case, but the separation of the two nations 
took place before (though, perhaps, not long before) the Vedic 
period. The question thus arises whether, — supposing both nations 
to hare already had the word Aryaman, — we are to assame that : 
conception of die god Aryaman had been already formed. The we 
occurs in sereral places in the second part of the Yaina, where bow-^ 
erer, the context does not justify us in explaining it as a proper 
name.* Spiegel then goes on to state his opinion that in the last 
chapter of the Vendidad, Airyama is not to be understood of a god, 
bat merely as denoting a particular prayer in which that word occarii, 
and which Ahura Maada discovers to be more efficacioaa in healing 
sickness than another sacred text to which be had first had r ec wi rwfc 

(P. 321, line 5.) 

I find, on recurring to Mr. Coraoo's paper (pi. IS3]^ that he is of 
opinion that " it was safaseqnently to their eztCBsioo orer this terri- 
tory [the Dekhanj and ita occopation, which may be regarded as the 
third era in their history, when the Aryans bad attained an adTaaeed 
state of civUiaalioB, when the Yedas had been *«— p~-<T, and a 




national system of religion established ; when the Bralimanical hier- 
archy had been formed, the Aryan tongue cultivatedj and codes of 
law compiled ; when tribes had separated under particular princes, 
and founded diEFerent governments in various parts of the country ; 
when religious schisms had begun to arise, anti-Brahmanical sects had 
increased, political dissensions and civil war had spread their eifccts 
— that the migrations in a westerly and north-westerly direction 
which terminated in the extension of the Aryan tongue over the 
geographical zone," [including Ariana, Persia, Armenia, Phrygia, 
Greece, Italy, Germany, &c., &c.,] which he had "pointed out, took 
place." Thus explained, his theory becomes far less probable than in 
the form in which I have stated it in tlie text (p. 321). If the Arians, 
or rather (in tlial case) the already Brahmanized Indians, had invaded 
and conquered the countries lying to the west of the Indus at a period 
subsequent to their occupation of the Dekhan and to the fuUdevelop- 
Qient of tlieir civilisation and their peculiar institutions, it is scarcely 
conceivable that no trace of this sweeping invasion should havu 
remained either in their own literature, or in that of any of the western 
nations, and that no specifically Brahmanical influences should have 
been discoverable in the religious or political syBtems of Persia, 
Greece, Rome, or Germany; for the period at which such a supposed 
extension of the Brahmanical Indians took place could not have been 
an "ante-IIellenic" era (p. 187); nor, consequently is it imaginable 
that all record of it should have disappeared in a presumed " age of 
darkness" (p. 186). The "ante-Hellenic" period terminated at least 
1000 years b.c., and the Brahmanical institutions could not hare been 
developed very long before tluit time. 

(P. 321, line 31.) 

I may add as a further answer to the reasoning which I have 
combated in this paragraph, that if, as is there supposed, the 
Aryas had been indigenous in Aryavartta, and had sent out colonizing 
or conquering expeditious from their aboriginal cradle to the west of 
the Indus, it luiglit have been expected that they would also have 
colonized the south of India at the same early period at which these 
presumed expeditions must have been made. But wc have no record 
of any such early Arian occupation of the Dekhan ; fur the era of 
Agastya and Ruma is comparatively recent. 



(P. 323, line 2d.) 
In R-V. ix. 74. 8, wc find the words, Kahshivnte iutahimHyn, 

" tw ICukshivat wlio bns lived a hundred winters." 

(P. 342, lino 5.) 

In ft paper " On the Geographical Arrangement of the Aruin 
Countries mentioned in the First Fargard of the Vcndidad," pulilished 
in the Transactions of the Berlin Academy tor 1856, pp. 621 — 647, 
Dr. Eicpert contests the conclusions of Dr. Uaug and others in 
regard to the position of some of the countries. Dr. Haug defcuds 
his own views in a paper in the Journ. of the Germ. Or. Society, 
vol. xi. pp. 526—533. 

(P. 357, note 79.) 

In his trnuslatton of Sutna-vedB, ii. 247 ( = n.-V. ix. 41. G). B«if<7 
translates rasa by ' ocean.' In his Glos.-ary he explains It o( " a par- 
ticulur river which fleparntes the world of Indra from that of tbe 
Punis (?) "; refenin!,' to R.-V. x. 108. In R..V. i. 112. 12 li« ex- 
plains it of the river Rasii. 

(P. 383, line 18.) 

After the word " maturity ; " add " the distinction hetwcen ihoae 
who observed them strictly and those who observed them laxly 
could not have arisen ; " 

(P. 413, line 3.) 

Compare Manu, xi. 20. Yad d/ianam j/njnaSilanam deva-tvam 
tad ridurbudhiih | Ayajvanantu yad vittctm asiira-tvam tad uchffatt\\ 
" The wealth of tliose wIjo practise sacrifice is regarded by the wise as 
tlie property of the gods; but the wealth of those whenever sacrifice 
13 called the property of the Asuras." See also Satapathtt-bralimatia, 
13. 8. 1. 5. and Wubei's Ind, Stud. i. 189 ; as well as the text from the 
Taitliriya-briiiunnna cited in the First Part of this work, p. 14. 

(P. 413, note, line 7 from the bottom.) 
For p. 401 read p. 402, 

(P. 425, line 15.) 

Lt a Inter period, after the coininencemcnt of the Mahomcdan 
'^ into Ilindusthan, southern India caiuo to be regarded as the 


sanctuary of the Brahmanical religion and learning. Thus in the 
verse of VySsa cited by Weber (Hist, of Ind. Lit. p. 247, note), it is 
said : SamprapU tu kalau Kale Vindhyadrer uttare sthitah | 
Brahma^ yajharahita ji/oti^-iastra-parangmukhah \ " In the 
Kali age the Kpdimans living to the north of the Yindhya are desti- 
tute of sacrifice and averse to astronomy ;" while another law book 
quoted by the same writer, says : Vindhyasya dakshine bhage yatra 
Godavari sthita | tatra Vedaicha yajhaicha bhavuhyanti Kalau 
yuge { "£i the Kali age the Vedas and sacrifices will be found to the 
south of the Yindhya, on the banks of the Godavari." 

(P. 435, line 13.) 

In R.-Y. iv. 4. IS. another epithet, viz. akas, " one who does not 
praise [the gods]," is applied to the Bakshases. Daha akaso 
Eakshasah pahi asman druho nido mitramaho avadyat : " Thou 
who art to be revered by thy friends, burn the Bakshases who offer 
no praise ; deliver us from the reproach of the oppressor and the 





5 01^ TIT 2DS 




(650) 723-9201 

All books are subjecF lo recall.