ON THE BORDERERS, MOUNTAINEERS, &c. 163

races of borderers, who have long been established on the limits of Arabia, Perfia, India, China, and Tartary; over the wild tribes residing in the mountainous parts of those extensive regions; and the more civilized inhabitants of the islands annexed by geographers to their Afiatick divifion of this globe.

Let us take our departure from Idume near the gulf of Elanitis, and, having encircled Afia, with fuch deviations from our courfe as the fubject may require, let us return to the point, from which we began; endeavouring, if we are able, to find a nation, who may clearly be shown, by juft reasoning from their language, religion, and manners, to be neither Indians, Arabs, nor Tartars, pure or mixed; but always remembering, that any small family detached in an early age from their parent ftock, without letters, with few ideas beyond objects of the firft neceffity, and confequently with few words, and fixing their abode on a range of mountains, in an island, or even in a wide region before uninhabited, might in four or five centuries people their new country, and would neceffarily form a new language with no perceptible traces, perhaps, of that spoken by their ancestors. Edom or Idume, and Erythra or Phanice, had originally, as many believe, a fimilar meaning, and were derived from words denoting a red colour;

but, whatever be their derivation, it seems indubitable, that a race of men were anciently fettled in Idume and in Median, whom the oldest and beft Greek authors call Erythreans; who were very diftinct from the Arabs; and whom, from the concurrence of many strong testimonies, we may safely refer to the Indian ftem. M. D'HERBELOT mentions a tradition (which he treats, indeed, as a fable), that a colony of those Idumeans had migrated from the northern shores of the Erythrean fea, and failed across the Mediterranean to Europe, at the time fixed by Chronologers for the paffage of EVANDER with his Arcadians into Italy, and that both Greeks and Romans were the progeny of those emigrants. It is not on vague and suspected traditions, that we must build our belief of such events; but NEWTON, who advanced nothing in fcience without demonftration, and nothing in history without fuch evidence as he thought conclufive, afferts from authorities, which he had carefully examined, that the Idumean voyagers "carried with them both arts and fciences,

among which were their astronomy, navigation, and letters; for in Idume, fays he, they "had letters, and names for conftellations, before "the days of JOB, who mentions them." JOB, indeed, or the author of the book, which takes its name from him, was of the Arabian stock,

as the language of that fublime work incontestably proves; but the invention and propagation of letters and aftronomy are by all so justly afcribed to the Indian family, that, if STRABO and HERODOTUS were not grofsly deceived, the adventurous Idumeans, who first gave names to the stars, and hazarded long voyages in fhips of their own conftruction, could be no other than a branch of the Hindu race: in all events, there is no ground for believing them of a fourth distinct lineage; and we need fay no more of them, till we meet them again, on our return, under the name of Phenicians.

As we pass down the formidable sea, which rolls over its coral bed between the coaft of the Arabs, or thofe, who speak the pure language of ISMAÏL, and that of the Ajams, or those, who mutter it barbarously, we find no certain traces, on the Arabian fide, of any people, who were not originally Arabs of the genuine or mixed breed: anciently, perhaps, there were Troglodytes in part of the peninfula, but they seem to have been long fupplanted by the Nomades, or wandering herdsmen; and who those Troglodytes were, we shall fee very clearly, if we deviate a few moments from our intended path, and make a fhort excurfion into countries very lately explored on the Western, or African, fide of the Red Sea.

That the written Abyffinian language, which we call Ethiopick, is a dialect of old Chaldean, and a fister of Arabick and Hebrew, we know with certainty, not only from the great multitude of identical words, but (which is a far ftronger proof) from the fimilar grammatical arrangement of the feveral idioms: we know at the fame time, that it is written, like all the Indian characters, from the left hand to the right, and that the vowels are annexed, as in Devanagari, to the confonants; with which they form a fyllabick fyftem extremely clear and convenient, but difposed in a less artificial order than the system of letters now exhibited in the Sanferit grammars; whence it may juftly be inferred, that the order contrived by PA'NINI or his difciples is comparatively modern; and I have no doubt, from a curfory examination of many old infcriptions on pillars and in caves, which have obligingly been sent to me from all parts of India, that the Nágari and Ethiopian letters had at first a similar form. It has long been my opinion, that the Abyffinians of the Arabian stock, having no fymbols of their own to represent articulate founds, borrowed thofe of the black pagans, whom the Greeks call Troglodytes, from their primeval habitations in natural caverns, or in mountains excavated by their own labour: they were probably the first inhabitants of Africa, where they be

came in time the builders of magnificent cities, the founders of feminaries for the advancement of fcience and philofophy, and the inventors (if they were not rather the importers) of fymbolical characters. I believe on the whole, that the Ethiops of Meroë were the fame people with the firft Egyptians, and confequently, as it might easily be shown, with the original Hindus. To the ardent and intrepid Mr. BRUCE, whose travels are to my tafte uniformly agreeable and fatisfactory, though he thinks very differently from me on the language and genius of the Arabs, we åre indebted for more important, and, I believe, more accurate, information concerning the nations established near the Nile from its fountains to its mouths, than all Europe united could before have supplied; but, since he has not been at the pains to compare the feven languages, of which he has exhibited a specimen, and fince I have not leifure to make the comparison, I must be fatisfied with obferving, on his authority, that the dialects of the Gafots and the Gallas, the Agows of both races, and the Falafbas, who muft originally have used a Chaldean idiom, were never preferved in writing, and the Ambarick only in modern times: they muft, therefore, have been for ages in fluctuation, and can lead, perhaps, to no certain conclufion as to the origin of the feveral tribes, who an