The Decame the Larue is bere exhibited is Eat mi Thin with ST Micens

one more example the many furthwa mamei Longinal nace baring abandoned Such relin yashment of the mother-tongue

Ma 20% ni tas throughout Hindustan Proper and the Vetem 1 1 1s throughout the whole of the Tan mat fenominated the Tara, not excinding la valley of Assam, there are but a few exceptions * t. a tue generi state of the case; whilst in the Central Hima, a tea onnal tunges are daily giving way before te Koda language, which, though originally and still traceably een yet more altered by Aman infuences than The very significant even the cultivated Dravirian tongues. cause of the phenomenon it will be our business to explain In the meanwhile the fact is well deserving of

this passing notice, with reference to the erroneous impression abroad as to the relative amounts of Arian and non-Arian elements in the population of India-an impression deepened and propagated by the further fact, still demonstrable among many of these altered aborigines, of the abandonment of their creed and customs, as well as tongue, for those of the Arians. We thence learn the value, in all ethnological researches, of physiological evidence, which, in regard to all these altered tribes, is sufficient to decide their non-Arian lineage, and to link them, past doubt, with the Himálayan and Indo-Chinese conterminous tribes on the east and north. It should be added, however, that, in a sheerly philological point of view, it becomes much more difficult to determine who are the borrowers and who the borrowed from, when both are non-Arians, than when one is Arian and the other non-Arian; and that, for instance, and in reference to the present vocabularies, we can decide at once that the Kondh numerals (save the two first) are borrowed from the Arian vernaculars, whereas it is by no means so certain that the Gadada and Yerukala numerals are borrowed from the Telugu and Karnata respectively, merely because they coincide; and so also of the pronouns where the same coincidence recurs. All such questions, however, are subordinate and secondary; and if we succeed in determining with precision-by physiological, lingual, and other helps-the entire Turánian element of our population, we shall then be able to advance another step and show the respective special affinities of the several cultivated and uncultivated Turánian tribes of India to each other and to certain of the tribes lying beyond India towards Burmah and Tibet, with at least an approximation to the relative antiquity of the successive immigrations into India

A word in defence of these vocabularies, of which the utility has been impugned, and impugned by special comparison with brief grammatical outlines.

When I commenced this series of vocabularies I expressed as strongly as any one could do the opinion that their utility must be circumscribed; and that the ethnology of India would only then be done complete justice to when every branch of

VOL. IL

ABORIGINES OF THE EASTERN GHATS.

To the Secretary of the Bengal Asiatic Society.

SIR,-Pursuant to my purpose of submitting to the Society, upon a uniform plan and in successive series, samples of all the languages of the non-Arian races of India and of the adjacent countries, I have now the honour to transmit six more vocabularies, for which I am indebted to Mr. H. Newill, of the Madras Civil Service, at present employed in Vizagapatam. These six comprise the Kondh, Sávara, Gadaba, Yerukala, and Chentsu tongues. In forwarding them to me, Mr. Newill, a very good Telugu scholar, has noted by an annexed asterical mark such words of these tongues, and particularly of Yerukala, as coincide with Telugu. He has also remarked that many of the Chentsu vocables resemble the U'rdu.

Having, as you are aware, a purpose of submitting to the Society an analytical dissection of the whole of the vocabularies collected by me, I shall be sparing of remarks on the present occasion. But I may add to M. Newill's brief notes a few words, as follows:

The Chentsu tribe, whose language, as here exhibited, is almost entirely corrupt Hindi and U'rdu, with a few additions from Bengali, affords one more example to the many forthcoming of an uncultivated aboriginal race having abandoned their own tongue. Such relinquishment of the mother-tongue has been so general that throughout Hindustan Proper and the Western Himalaya, as well as throughout the whole of the vast Sub-Himalayan tract denominated the Tarai, not excluding the contiguous valley of Assam, there are but a few exceptions to this the general state of the case; whilst in the Central Himalaya the aboriginal tongues are daily giving way before the Khas language, which, though originally and still traceably Tartaric, has been yet more altered by Arian influences than even the cultivated Dravirian tongues. The very significant cause of this phenomenon it will be our business to explain by and by. In the meanwhile the fact is well deserving of

this passing notice, with reference to the erroneous impression. abroad as to the relative amounts of Arian and non-Arian elements in the population of India-an impression deepened and propagated by the further fact, still demonstrable among many of these altered aborigines, of the abandonment of their creed and customs, as well as tongue, for those of the Arians. We thence learn the value, in all ethnological researches, of physiological evidence, which, in regard to all these altered tribes, is sufficient to decide their non-Arian lineage, and to link them, past doubt, with the Himálayan and Indo-Chinese conterminous tribes on the east and north. It should be added, however, that, in a sheerly philological point of view, it becomes much more difficult to determine who are the borrowers and who the borrowed from, when both are non-Arians, than when one is Arian and the other non-Arian; and that, for instance, and in reference to the present vocabularies, we can decide at once that the Kondh numerals (save the two first) are borrowed from the Arian vernaculars, whereas it is by no means so certain that the Gadada and Yerukala numerals are borrowed from the Telugu and Karnata respectively, merely because they coincide; and so also of the pronouns where the same coincidence recurs. All such questions, however, are subordinate and secondary; and if we succeed in determining with precision-by physiological, lingual, and other helps-the entire Turánian element of our population, we shall then be able to advance another step and show the respective special affinities of the several cultivated and uncultivated Turánian tribes of India to each other and to certain of the tribes lying beyond India towards Burmah and Tibet, with at least an approximation to the relative antiquity of the successive immigrations into India.

A word in defence of these vocabularies, of which the utility has been impugned, and impugned by special comparison with brief grammatical outlines.

When I commenced this series of vocabularies I expressed as strongly as any one could do the opinion that their utility must be circumscribed; and that the ethnology of India would only then be done complete justice to when every branch of

the subject should be carefully and simultaneously studied, upon the plan exemplified in my work on the Kóch, Bódó, and Dhimál. Much and toilsome labour has, however, since then, convinced me that inquiries confined wholly to India and its immediate vicinity would yield results far less satisfactory than such as should be greatly more extended even if they were less complete; whilst these continued labours have more and more satisfied me that limited grammatical comparisons are much more apt to give rise to error than limited glossarial ones. Perhaps the fascination of such extended inquiry may have somewhat biassed my judgment; but I am still decidedly of the opinion that the true relations of the most shifting and erratic, the most anciently and widely dispersed, branch of the human family cannot be reasonably investigated upon a contracted scale, while the subject is so vast that one must needs seek for some feasible means of grasping it in sufficient amplitude to comprehend its normal character (a thing rather of surface than of depth), at the same time that one neglects not more complete and searching investigation of certain actual or supposed characteristic samples. Such is the course I have been pursuing for some time past. I have examined, and am still examining, the complete grammatical structure of several of the Himalayan tongues; and I have at the same time submitted the whole of my vocabularies to the alembic of extended comparative analysis. I hope soon to be able to present the results to the Society. Those of the analysis have been fruitful beyond my hopes, owing to the extraordinary analogy pervading the Tartaric tongues in regard to the laws which govern the construction of all their vocables, save, of course, the monosyllabic ones, which, however, are very rare. Even a superficial examination of the vocabularies suffices to indicate this prevalence of common constructive principles; and to such persons as have neither time nor skill to trace and demonstrate those principles, the mere collocation of the terms as they stand, if done on a sufficiently ample scale, will afford such evidence of general relationship and family union between the whole of the Indian aborigines and the populations of Indo-China, Sífán, Tibet, and Himálaya, aye, and of China also, as philological