fon, why the Jabadios, or Yavadwipa, of ProLEMY was rendered in the old Latin version the ifle of Barley; but we must admire the inquifitive spirit and patient labour of the Greeks and Romans, whom nothing obfervable feems to have escaped: Yava means barley in Sanferit; and, though that word, or its regular derivative, be now applied folely to Java, yet the great French geographer adduces very strong reasons for believing, that the ancients applied it to Sumatra. In whatever way the name of the last mentioned island may be written by Europeans, it is clearly an Indian word, implying abundance or excellence; but we cannot help wondering, that neither the natives of it, nor the best informed of our Pandits, know it by any fuch appellation; especially as it ftill exhibits vifible traces of a primeval connexion with India: from the very accurate and interefting account of it by a learned and ingenious member of our own body, we difcover, without recourfe to etymological conjecture, that multitudes of pure Sanferit words occur in the principal dialects of the Sumatrans; that, among their laws, two pofitive rules concerning fureties and intereft appear to be taken word for word from the Indian legislators NA'RED and HA'RITA; and, what is yet more obfervable, that the fyftem of letters, ufed by the people of Rejang and Lampún, has

any

the fame artificial order with the Devanagari; but in every feries one letter is omitted, because it is never found in the languages of those iflanders. If Mr. MARSDEN has proved (as he firmly believes, and as we, from our knowledge of his accuracy, may fairly prefume) that clear veftiges of one ancient language are discernible in all the infular dialects of the fouthern feas from Madagascar to the Philippines and even to the remoteft islands lately difcovered, we may infer from the specimens in his account of Sumatra, that the parent of them all was no other than the Sanferit; and with this obfervation, having nothing of confequence to add on the Chinefe ifles or on those of Japan, I leave the fartheft eastern verge of this continent, and turn to the countries, now under the government of China, between the northern limits of India, and the extenfive domain of those Tartars, who are ftill independent.

That the people of Pótyid or Tibet were Hindus, who engrafted the herefies of BUDDHA on their old mythological religion, we know from the researches of CASSIANO, who long had refided among them; and whofe difquifitions on their language and letters, their tenets and forms of worship, are inferted by GIORGI in his curious but prolix compilation, which I have had the patience to read from the first to the laft of nine

hundred rugged pages: their characters are apparently Indian, but their language has now the disadvantage of being written with more letters than are ever pronounced; for, although it was anciently Sanfcrit and polyfyllabick, it seems at prefent, from the influence of Chinese manners, to confift of monofyllables, to form which, with fome regard to grammatical derivation, it has become neceffary to fupprefs in common difcourse many letters, which we fee in their books; and thus we are enabled to trace in their writing a number of Sanfcrit words and phrases, which in their spoken dialect are quite undistinguishable, The two engravings in GIORGI's book, from fketches by a Tibetian painter, exhibit a fyftem of Egyptian and Indian mythology; and a complete explanation of them would have done the learned author more credit than his fanciful etymologies, which are always ridiculous, and often grofsly erroneous.

The Tartars having been wholly unlettered, as they freely confefs, before their converfion to the religion of Arabia, we cannot but fufpect, that the natives of Eighúr, Tancút, and Khatà, who had fyftems of letters and are even faid to have cultivated liberal arts, were not of the Tartarian, but of the Indian, family; and I apply the fame remark to the nation, whom we call Barmas, but who are known to the Pandits by

the name of Brahmachinas, and feem to have been the Brachmani of PTOLEMY: they were probably rambling Hindus, who, defcending from the northern parts of the eaftern peninfula, carried with them the letters now used in Ava, which are no more than a round Nágari derived from the fquare characters, in which the Páli, or facred language of BUDDHA's priests in that country, was anciently written; a language, by the way, very nearly allied to the Sanferit, if we can depend on the teftimony of M. DE LA LOUBERE; who, though always an acute observer, and in general a faithful reporter, of facts, is charged by CARPANIUS with having mistaken the Barma for the Páli letters; and when, on his authority, I spoke of the Bali writing to a young chief of Aracan, who read with facility the books of the Barmas, he corrected me with politenefs, and affured me, that the Páli language was written by the priests in a much older character.

Let us now return eastward to the fartheft Afiatick dominions of Ruffia, and, rounding them on the northeast, pafs directly to the Hyperboreans; who, from all that can be learned of their old religion and manners, appear like the Maffageta, and fome other nations usually confidered as Tartars, to have been really of the Gothick, that is of the Hindu, race; for I con

fidently affume, that the Goths and the Hindus had originally the fame language, gave the fame appellations to the stars and planets, adored the fame falfe deities, performed the fame bloody facrifices, and profeffed the fame notions of rewards and punishments after death. I would not infist with M. BAILLY, that the people of Finland were Goths, merely because they have the word hip in their language; while the reft of it appears wholly diftinct from any of the Gothick idioms: the publishers of the Lord's Prayer in many languages reprefent the Finnish and Lapponian as nearly alike, and the Hungarian as totally different from them; but this must be an errour, if it be true, that a Ruffian author has lately traced the Hungarian from its primitive feat between the Caspian and the Euxine, as far as Lapland itself; and, fince the Huns were confeffedly Tartars, we may conclude, that all the northern languages, except the Gothick, had a Tartarian origin, like that univerfally ascribed to the various branches of Sclavonian.

On the Armenian, which I never ftudied, becaufe I could not hear of any original compositions in it, I can offer nothing decifive; but am convinced, from the best information procurable in Bengal, that its basis was ancient Perfian of the fame Indian ftock with the Zend, and that