title the various inflections of nouns and verbs) has undergone a total alteration. How essentially different from that affinity and analogy described by Mr. Brown between the Sanscrit and the Greek; or exhibited in the resemblance of the inflections of the verbs in these two languages, in the passage already referred to from Bopp!2 From all the accounts that have fallen in my way, I am led to suspect, that the number of Sanscrit words which can be traced to a Greek root, bears no proportion to that of the words which, in the Romanic tongues, are evidently of Latin origin. Upon the hypothesis which I have proposed, all this is not only explicable, but must necessarily have happened.

It was upon these grounds that I remarked, in a former publication, that "the affinities and filiations of different tongues, as evinced in their corresponding roots and other coincidences, are incomparably more easy in the explanation, than the systematical analogy which is said to exist between the Sanscrit and the Greek in the conjugations and flexions of their verbs, and in many other particulars of their mechanism."3

If such a scholar as Dr. Bentley or Dr. Parr should ever make a serious object of studying Sanscrit, he would be able, I should think, without much difficulty, to ascertain, from internal evidence, which of the two languages was the primitive, and which the derivative dialect. He would also be enabled to decide, whether the mechanism of the Sanscrit affords any satisfactory evidence of its being manufactured by such a deliberate and systematical process as I have conjectured. It seems to be in this way alone that these points can be settled beyond contro

versy.

To all this we may add, that it appears difficult, if not impossible, to conceive how a tongue which was once spoken over regions of such vast extent, should have ceased to be a living language. It is by means of the most overwhelming and un

See page 79 of this volume.

2 "See Edinburgh Review, vol. xxxiii. p. 431. G. Mid. Voice, σεβομαι, σέβεσαι, σεβεται, σεβομένα, σεβεσθε, σεβονται. Sanscrit Mid. Voice, Sebe, sebase, sebate, sebamahe, sebadhiva, sebante. The root

Seb has the same signification in Greek and in Sanscrit."

Dissertation prefixed to the Supplement to the Encyclopædia Britannica, [supra, Woks, vol. i. p. 426.]

sparing foreign conquests, that languages have been generally changed or destroyed; and that no causes of this sort have operated in the countries where Sanscrit is alleged to have once prevailed, is demonstrated by the religious and political institutions, (more especially by the division of people into castes,) which remain unaltered in the very same countries, from the most remote periods of authentic history. It seems at least

equally inconceivable how a language, so very perfect, should have grown up, contrary to the analogy of every one else known, from popular and casual modes of speech.

The same objection seems to me to apply with still greater force to an hypothesis proposed in the Edinburgh Review, by a gentleman whose authority is deservedly high in all matters connected with Indian Literature. In the opinion of this writer, "it is no improbable hypothesis, that the Bramins entered India as conquerors, bringing with them their language, religion, and civil institutions. The Purana," continues the same writer, seem even to point out the conqueror in the person of Parusaramo, who, at the head of an army of Bramins, extirpated the military tribes, and overthrew all the existing monarchies. But the period of this event is before the æra of historical record."2

"Whatever be its antiquity," says Sir William Jones, "it is of a wonderful structure, more perfect than the Greek, more copious than the Latin, and more exquisitely refined than either, yet bearing to both of them a stronger affinity, both

from both those tongues, as Arabic differs from Persian, or German from Greek. Now, the general effect of conquest is to leave the current language of the conquered people unchanged, or very little altered in its ground-work, but to blend with it a considerable number of exotic names, both for things and for actions; as it has happened in every country that I can recollect, where the conquerors have not preserved their own tongue unmixed with that of the natives, like the Turks in Greece, and the Saxons in Britain; and this analogy might induce us to believe, that the pure Hindì, whether of Tartarean or Chaldean origin, was primæval in Upper India, into which the Sanscrit was introduced by conquerors from other kingdoms in some very remote age; for we cannot doubt, that the language of the Védas was used in the great extent of country which has before been delineated, as long as the

religion of Brahma has prevailed in it.” -Asiatic Researches, vol. i. pp. 421,

422.

1 According to Mr. Bentley, the Purana, in point of antiquity, are not older than 700 years; and Mr. Pinkerton thinks he has been successful in demonstrating his assertion.-See his Geography, vol. i. p. 718.

2

Edinburgh Review, vol. xiii. p. 369. After all, is it not possible that the excellencies of Sanscrit may be somewhat overrated by Sir William Jones, from the same bias which has led him to overrate so immensely the merits of those ancient compositions, of which he has enabled the public to judge by the translations with which he has favoured us from that language? Mr. Mill has justly observed, that "languages, on which equal eulogies are bestowed to any which can be lavished on Sanscrit, are the languages confessedly of igno

in the roots of verbs, and in the forms of grammar, than could possibly have been produced by accident; so strong, indeed, that no philosopher could examine them all three, without believing them to have sprung from some common source which perhaps no longer exists." The only possible supposition, I apprehend, on which all this can be explained, is, that Sanscrit was a language formed by the Bramins, and always confined to their order; and that the Greek tongue not only served as a model for its syntax and system of inflections, but supplied the materials of its vocabulary on abstract and scientific subjects. Difficulties, I am aware, may be started in opposition to this conjecture, and, in particular, it may be objected, that there are various other tongues (the Pehlavi, for instance, or ancient Persian) between which, and the Sanscrit, a close affinity has been remarked.2 But it deserves consideration whether these objections apply exclusively to the hypothesis I have proposed, and whether they are not equally unaccountable upon the other theories which are in general currency. All of these, too, (it must be remembered,) are encumbered with this additional difficulty, that they are forced

rant and uncivilized men. Molina in-
forms us,
that, of the language of the
Chilians, the grammar is as perfect as
that of the Greek or Latin; that of no
language does the formation and struc-
ture display greater ingenuity and feli-
city. The language of the Malays is
described by Mr. Marsden, as remark-
ably sweet and well adapted to poetry.
Clavigero knows not where to set a
limit to his admiration of the Mexican
tongue, it is so copious, polished, and
expressive, that it has been esteemed
by many superior to the Latin and even
to the Greek."-(Mill's India, vol. i.
p. 392.) I myself recollect, at the time
when it was as fashionable to extol with
enthusiasm the Poems of Ossian, as it is
now for the same class of critics to de-
ride them, to have heard many of our
Celtic scholars talk of the Gaelic in a
like extravagant strain. Macpherson's

VOL. IV.

translation they allowed to be as good as an English version could be; but they insisted (and who could contradict them?) that there was a richness and force in the original to which no known language but the Greek could do justice.

1 Works of Sir William Jones, vol. i. p. 26.

2 I have now lying before me a book entitled, "Tableaux Synoptiques des Mots similaires qui se trouvent dans les langues Persane, Samskrite, Greque, Latine, Maesogothique, Islandoise, Suéo - Gothique, Suedoise, Danoise, Anglo-Saxone, Celto-Bretone ou Armorique, Angloise, Alémanique ou Francique, Haut-Allemande, et Bas-Allemande," par H. A. Le Pileur, &c. &c. &c.-Paris and Amsterdam. (No date; but evidently published under the Imperial Government.)

G

to have recourse to supposed events which may have happened prior to the date of our historical records. In favour of our hypothesis, it may, on the other hand, be observed, that it has the advantage of assuming no imaginary event, while the difficulties with which it is attended admit of an easy and probable solution. Whatever other foreign idioms were at different times introduced into India, may have supplied words which have been incorporated with Sanscrit as well as with the vulgar languages, or rather which have been naturally incorporated with the former through the medium of the latter. It deserves also to be considered, whether those vulgar tongues in India, of which Sanscrit is supposed to be the basis, may not have furnished, at different times, to that sacred language, those words which are commonly referred to it as their original source.

The most formidable objection, however, is suggested by this consideration, that the Sanscrit is represented by some as bearing much more resemblance to the Latin than to the Greek. Mr. Halhed's words are these:-"Let me here cursorily observe, that as the Latin is an earlier dialect than the Greek, as we now have it, so it bears much more resemblance to the Sanscrit, both in words, inflections, and terminations."

In another passage he tells us, that, "in the Sanscrit language, as in the Greek, there are forms of infinitives and of participles comprehensive of time; there are also other branches of the verb that seem to resemble the gerunds and supines of the Latin."2

Sir William Jones, also, in the preface to his translation of Sacontala, bears a strong testimony to the close affinity between the Sanscrit and the Latin:-"I began with translating it verbally into Latin, which bears so great resemblance to the Sanscrit, that it is more convenient than any other language for a scrupulous interlineary version. I then turned it into English."

1 Grammar of the Bengal Language, p. 137.

2 Ibid. p. 138.

3 I must be allowed here to remark,

that these statements of the resemblance of the Sanscrit to the Latin, strong as they are, do not amount to any thing so full and precise as that of Mr. Brown,