With regard to Arabia, it may safely be asserted that the Mosaic table is confirmed by modern research. The Cushitic element has left memorials of its presence in the south in the vast ruins of March and Sana (Re- nan, Hist. Gen. i, 318), as well as in the influence it has exercised on the Himyaritic and Jlahri languages, as compared with the Hebrew. The Joktanid element forms the basis of the Arabian population, the Shemitic character of whose language needs no proof. With re- gard to the Ishmaelite element in the north, we are not aware of any linguistic proof of its existence, but it is confirmed by the traditions of the Arabians themselves.
It remains to be inquired how far the Japhetic stock represents the linguistic characteristics of the Indo- European and Turanian families. Adopting the two- fold division of the former, suggested by the name itself, into the eastern and western ; and subdividing the eastern into the Indian and Iranian, and the west- ern into the Celtic, Hellenic, Illyrian, Italian, Teutonic, Slavonian, and Lithuanian classes, we are able to assign Madai {Media) and Togarmah {Armenia') to the Irani- an class; Javan {Ionian) and Elishah {.Eolian') to the Hellenic; Gomer conjecturally to the Celtic; and Doda- nim, also conjecturally, to the Illyrian. According to the old interpreters, Ashkenaz represents the Teutonic class, while, according to Knobel, the Italian would be represented by Tarshish, whom he identities with the Etruscans; the Slavonian by Magog; and the Lithuani- an possibly by Tiras ( Vdlkert. p. 68, 90, 130). The same writer also identifies Riphath with the Gauls, as distinct from the Cj'mry or Gomer (p. 4 J) ; while Kittim is re- ferred by him not improbably to the Carians, who at one period were predominant on the islands adjacent to Asia Minor (p. 98). The evidence for these identifica- tions varies in strength, but in no instance approaches to demonstration. Beyond the general probability that tlie main branches of the human famil3' would be rep- resented in the jMosaic table, we regard much that has been advanced on this subject as highly precarious. At the same time, it must be conceded that the subject is an open one ; and that as there is no possibilitj^ of prov- ing, so, also, there is none of disproving, the correctness of these conjectures. W^hether the Turanian family is fairly represented in the Mosaic table may be doubted. Those who advocate the Mongolian origin of the Scythians would naturally regard Magog as the representative of this familv; and even those who dissent from the Mon- golian theory may still not unreasonably conceive that the title Magog applied broadly to all the nomad tribes
of Northern Asia, whether Indo-European or Turanian. Tubal and Meshech remain to be considered: Knobel identities these respectively with the Iberians and the Ligurians (p. HI, 119) ; and if the Finnisii character of tlie Basque language were established, he would regard the Iberians as certainly, and the Ligurians as probablv, Turanians — the relics of the first wave of population which is supposed to have once overspread the whole of the European continent, and of which the Films in the north, and the Basques in the south, are the sole surviving representatives. The Turanian character of the two Biblical races above mentioned has been other- wise maintained on the ground of the identity of the names Meshech and Muscovite ( Rawlinson, //e^of/. i, 652).
(111.) Having thus reviewed the ethnic relations of the nations who fell within tlie circle of the Mosaic table, we propose to cast a glance beyond its limits, and inquire how far the present results of ethnological sci- ence support the general idea of the unity of the human race, which underlies the Mosaic system. The chief and in many instances the only instrument at our com- mand for ascertaining the relationship of nations is lan- guage. In its general results this instrument is thor- oughly trustworthy, and in each individual case to which it is applied it furnishes a strong primn facie evidence ; but its evidence, if unsupported by collateral proofs, is not unimpeachable, in consequence of the nu- merous instances of adopted languages which have oc- curred within historical times. This drawback to the value of the evidence of language will not materially affect our present inquiry, inasmuch as we shall confine ourselves as much as possible to the general results.
The nomenclature of modern ethnology is not iden- tical with that of the Bible, partly from the enlarge- ment of the area, and partly from the general adoption of language as the basis of classification. The term Shemitic is indeed retained, not, however, to indicate a descent from Shem, but the use of languages allied to that which was current among the Israelites in histor- ical times, llamitic also finds a place in modern eth- nology, but as subordinate to, or co-ordinate with, Shemitic. Japhetic is superseded mainly by Indo- European or Aryan. The various nations, or families of nations, which find no place under the Biblical titles are classed by certain ethnologists under the broad title of Turanian, while by others they are broken up into divisions more or less numerous.
1. The first branch of our subject will be to trace the extension of the Shemitic family beyond the limits as- signed to it in the Bible. The most marked charac- teristic of this family, as compared with the Indo-Eu- ropean or Turanian, is its inelasticity. Hemmed in both by natural barriers and bj' the superior energy and expansiveness of the Aryan and Turanian races, it retains to the present day the status quo of early times. The only direction in which it has exhibited any ten- dency to expand has been about the shores of the Med- iterranean, and even here its activity was of a sporadic character, limited to a single branch of the family, viz. the Phoenicians, and to a single phase of expansion, viz. commercial colonies. In Asia Minor we find tokens of Shemitic presence in Cilicia, which was connected with Phcenicia both by tradition (Herod, vii, 91) and by language, as attested by existing coins ((iesenius, Mon. Phan. iii, 2); in I'amphylia, Pisidia, and Lycia, parts of which were occupied by the Solymi (Pliny, v, 24; Herod, i, 173), whose name bears a Shemitic char- acter, and who are reported to have spoken a Shemitic tongue (Eusebius, Prcep. Er. ix, 9), a statement con- firmed by the occurrence of other Shemitic names, such as Phoenix and Cabalia, though the subsequent pre- dominance of an Aryan population in these same dis- tricts is attested by the existing Lycian inscriptions; again in Caria, though the evidence arising out of the supposed identity of the names of the gods Osogo and Chrysaoreiis with the Ovomoq and Xpvaiop of Sanchoni-