lects in India, prevailing in as many different civilized nations, who occupied the provinces of Hindoostan and the Deckan.

The Saraswati was a people which occupied the banks of the river Seraswattee, and the Bramins of that nation now inhabit the Panjab. Their language may have prevailed over the southern and western parts of Hindoostan Proper, and is probably the idiom called Prácrit. It is a cultivated language, and great part of most dramas, and many poems, are

written in it.

The Canyacubjas possessed a great empire, the capital of which was Cannoge. Their language seems to be the groundwork of the modern Hindustani or Hindwi, of which there are two dialects, in the most refined of which there are numerous poems, and both abound in songs, or rather ballads, and odes, Well educated people in Hindustan and the Deckan, use this language, and there is scarcely a village where some of the inhabitants do not understand it; which I beg you to observe, is the reason I particularly advised you to study it, that you may not be among those speakers of jargon, whom one hears violating all the rules of grammar and good sense, at our settlements in India, till they have actually produced a tongue

that I am persuaded no Hindû, fresh from the interior, would understand.

The Gaura, or Bengali, is spoken in the provinces of which the ancient city of Gaur was once the capital, and of which nothing remains but widely spread ruins. The language contains some original poems, besides many translations from the Sanscrit: it appeared to me, when I heard it spoken, to be a soft agreeable language, though less pleasing to the ear than the Hindustani.

The Maithila, or Tirhuctya, is used in the Circar of Tirhut and the adjoining districts, and appears not to have been much cultivated.

The language and alphabet called Uriga are used in the Suba of Orissa, whose ancient names are Utcala and Odradesa.

These five countries are called the five Gaurs, and occupy the northern and eastern parts of India, though Orissa seems more properly to belong to the five Draviras which occupy the Peninsula as far as Cape Comorin; and Guzerat, which is sometimes reckoned among the Draviras, would find a more natural place among the Gaurs.

The language of Guzerat or Gurjera is nearly allied to the Hindwi, and, like it, is commonly written in an imperfect form of the

Devanagari character, in which the Sanscrit is expressed.

Dravira is the southern extremity of India, and extends from Cape Comorin to twelve or thirteen degrees of north latitude, The language is the Tamel, called by the Europeans Malabars. I have seen translations from some Tamel songs, both of love and of war and one I recollect of a humourous description, purporting to be the quarrel between a man's two wives, one of whom was a Tamel and the other a Tailinga lady; but as it appeared that one was much younger and handsomer than the other, the quarrel was naturally enough decided in her favour; though I own that, to me, the other seemed to have the right side of the argument.

The Maharashtra, or Mahratta, is a nation which has in the two last centuries greatly enlarged its boundaries; but it anciently comprehended only a mountainous district south of the Nermada, and extending to the, Cocan. The language boasts of some treatises of logic and philosophy, besides many original poems, chiefly in honour of Rama and Crishna, and some translations from the Sanscrit.

Carnata, or Canara, is the ancient language of Carnataca, a province which has given names to districts on both coasts of the Peninsula; the

C

dialect still prevails in the intermediate mountainous tract.

Tailangana must formerly have comprehended not only the province of that name, but those on the banks of the Crishna and Godavery. Its language (Telinga) has been cultivated by poets, if not by prose writers.

Besides these ten polished dialects, there are some others, derived, like them, from the Sanscrit, and, like them, written in a character more or less corrupted from the Deva Nagari. There are also some spoken by the mountaineers, who are probably the aborigines of India, and which have certainly no affinity with the Sanscrit.

Some of these tongues are divided into local and provincial dialects, and many beautiful pastorals are written in the two most remarkable-the Panjabi, spoken in the Panjab or country of the five rivers, and the Vraja Bhasha, spoken in the neighbourhood of Mathura, which derives its name from the cow-pens, Vraja, of the forests of Vrindha.

Translations of at least part of two Sanscrit Grammars appeared in English in the year 1808; the first from Saraswata, by Mr. Colebrooke, and the second, by Mr. Carey, is partly a translation, partly original, from the Grammars used in Bengal, where the teachers have

unfortunately accommodated the sacred language to the vernacular idiom and pronunciation. A Sanscrit Grammar, by Mr. Wilkins, appeared in the same year, which has the character, among the learned, of accuracy, preciseness, and perspicuity, notwithstanding its great length, which the multitude of rules and exceptions in the language has swelled to 656 pages.

The author of the able article upon this Grammar, in the thirteenth volume of the Edinburgh Review, has given a very interesting table of the analogy of the Sanscrit with some other languages, which certainly goes far to confirm the opinion of Mr. Colebrooke and of Sir William Jones, concerning the primeval tongue from which these languages may have been derived, and which I quoted in the early part of this Letter. The first part of the analogy consists of words expressing the names of different parts of the body, and the relations of consanguinity, thus

In this last word, there is an example of the manner in which the Sanscrit double letters are