gari, to the confonants; with which they form a fyllabic fyftem extremely clear and convenient, but difpofed in a lefs artificial order than the system of letters now exhibited in the Sanfcreet grammars; whence he conceived it may justly be inferred, that the order contrived by Panini or his difciples is comparatively modern; and he entertained no doubt, from a curfory examination of many old inscriptions on pillars and in caves, which had been fent to him from all parts of India, that the Nagari and Ethiopian letters had at first a fimilar form.

On this fubject it is impoffible to omit, though before partially cited, the additional evidence of Mr. Halhed, who, in the preface to his Grammar of the Hindoftani Language, afferts the Sanfcreet, or ancient language of India, generally spoken before the invasion of Alexander, to be a language of the most venerable and profound antiquity; the grand fource as well as facred repofitory of Indian literature, and the parent of almoft every dialect, from the Perfian Gulph to the China Sea. He is of opinion, that the Sanfcreet was, in ancient periods, current not only over ALL INDIA, confidered in its largeft extent, but

over ALL THE ORIENTAL WORLD, and that traces of its original and general diffufion may ftill be discovered in almost every region of Afia. He was astonished to find "the fimilitude of Sanfcreet words with thofe of Perfian and Arabic, and even of Latin and Greek ; and that not in technical and metaphorical terms, which the mutuation of refined arts and improved manners might have occafionally introduced, but in the ground-work of language, in monofyllables, in the name of numbers, and the appellations of fuch things as would be firft difcriminated on the immediate dawn of civilization. The resemblance which may be obferved in the characters on the medals and fignets of various districts of Afia, the light which they reciprocally reflect upon each other, and the general analogy which they all bear to the fame grand prototype, afford another ample field for curiofity. The coins of Affam, Nepaul, Cashmeere, and many other kingdoms, are all ftamped with Sanfcreet characters, and moftly contain allufions to the old Sanfcreet mythology. The fame conformity I have observed on the impreffion of feals from Bootan and Thibet. A collateral inference may likewise VOL. VII.

P

be

be deduced from the peculiar arrangement of the Sanfcreet alphabet, so very different from that of any other quarter of the world. This extraordinary mode of combination still exifts in the greatest part of the East, from the Indus to Pegu, in dialects now apparently unconnected, and in characters completely diffimilar; and it affords a forcible argument that they are all derived from the fame fource. Another field for fpeculation prefents itself in the names of perfons and places, of titles and dignities, which are open to general notice, and in which, to the fartheft limits of Afia, may be found manifeft traces of the Sanfcreet."*

Thus, deduced from various fources, and flowing through various channels, the ftream of argument carries us back to the central point whence we originally fet out; that of a primeval language, univerfally prevalent among the early branches of the family of Noah, and diffused with the first colonies through the habitable world; but, in the course of ages, as new events arofe, as new governments were formed, and as new ideas

ལ།

* Hilhed's Grammar of the Bengal Language, p. 3.

poured

poured in upon them, undergoing fuch material alterations and modifications, as fcarcely to leave any veftige of its origin remaining, except the radices of fome principal words every dialect of it, by which the relation of the fecondary to its primary tongue may be faintly recognized.

in

I have hitherto attended folely to alphabetic writing; to letters as the fymbols of found, not to thofe of the hieroglyphic kind, which are properly the fymbols of ideas and objects. The latter appear to have no connection with the Indian alphabet, whatever they may have with that of the old Egyptians or that of the present Chinese; the only people, befides the Japanese and Mexicans, who now make use of so complicated a system of conveying their ideas and perpetuating their fentiments. On the confideration, therefore, of that particular clafs of alphabet, there is no occafion that I should enter in any detail, yet, upon this subject, I cannot avoid remarking, that, if the hieroglyphic and fymbolic character, ufed by the Chinese, be no proof of their descent from the Egyptians, upon which ground M. de Guignes founded his arguments for fuch defcent,

P 2

defcent, by the fame line of reasoning, the fyftem of the Brahmins, adopted by Sir William Jones himself, who infifts upon their having originally fprung from a Hindoo ftock, is, I fpeak with refpectful fubmiffion to his genius and learning, — very confiderably weakened, if not wholly fubverted; for, is it poffible, that, during the gradual migration of their tribes eastward, and at that early period, when the Sanfcreet flourished in its full vigour, that they should have lost all remembrance of their native tongue, either the vulgar Sanfcreet dialect, or the elegant and polifhed Devanagari; and should have adopted, in the room of an alphabet already elaborately formed, and justly distinguished for its comprehenfive utility, its refined correctness, and the beauty of its arrangement, vague and prolix fyftem of fymbolic characters compofed of the forms or detached parts of the forms of animals and objects, inconceivably intricate in their combinations, and infinitely diverfified in their number and fignification? I am aware that the fancy of fome learned men has endeavoured to fix a hieroglyphic ftamp on many letters of the Hebrew, Arabian, and other Eastern alpha

a

bets;