[256] Ferrier’s Institutes of Metaphysics, p. 13.
[257] Adr. Balbi, Atlas ethnographique. Disc. prélim. lxxv-lxxix.
[258] Garnett’s Philol. Essays, p. 85, &c., where the supposed instances are examined. Most of them are, as might have been expected, simple onomatopœias of the most obvious kind. See Renan, Hist. des Langues Sém. p. 450 seqq. Nothing requires more care than an inquiry of this kind;—often two words which have identically the same letters have no connection with each other, while two others derived from a common source have not one letter in common. As an instance of the former case, take the French souris “a smile,” and souris “a mouse,” (from subridere and sorex respectively); as an instance of the latter, take the word cousin, derived from soror through consobrinus.
[259] Outlines, i. 476.
[260] Outlines, i. 143, 165 seqq.
[261] A very curious instance of this is the word שווין shoes, found in a Syro-Chaldaic Lectionarium in the Vatican. We may here remark that Dr. Young’s celebrated calculation—that, if eight words are identical in two languages, the chances of a direct relation between the languages are 100,000 to one—is very exceptionable. See Dr. Latham, in the Encycl. Brit. Art. Language. The greatest care is necessary to distinguish between words really cognate, and accidental isolated resemblances. See Pictet, Orig. Ind. p. 13, 17.
[262] Survey of Lang. p. 11.
[263] Renan, p. 216.
[264] Hist. des Langues Sém. p. 84 seqq.
[265] Renan quotes Mövers, Die Phœnizien, i. 33.