tribes.”(1311) So amongst the Sakais, an aboriginal race of the Malay Peninsula, a man goes to a considerable distance for a wife, generally to a tribe who speak quite a different dialect.(1312) The Indian tribes of French Guiana have each their own dialect and would hardly be able to understand each other, were it not that almost every person marries a wife or a husband of a different tribe, and thus the newcomers serve as interpreters between the tribe in which they live and that in which they were born and brought up.(1313) It is well known that the Carib women spoke a language which differed in some respects from that of the men, and the explanation generally given of the difference is that the women preserved the language of a race of whom the men had been exterminated and the women married by the Caribs. This explanation is not, as some seem to suppose, a mere hypothesis of the learned, devised to clear up a curious discrepancy; it was a tradition current among the Caribs themselves in the seventeenth century,(1314) and as such it deserves serious attention. But there are other facts which seem to point to a different explanation.(1315) Among the Carayahis, a tribe of Brazilian Indians on the Rio Grande or Araguaya River, the dialect of the women differs from that of the men. For the most part the differences are limited to the form and sound of the words; only a few words seem to be quite distinct in the two dialects. The speech of the women appears to preserve older and fuller forms than that of the men: for instance, “girl” is _yadokoma_ in the female speech but _yadôma_ in the male; “nail” is _desika_ in the mouth of a woman but _desia_ in the mouth of a man.(1316) However such remarkable differences are to be explained, a little reflection will probably convince us that a mere intermixture of races speaking different tongues could scarcely account for the phenomena of language under consideration. For the reluctance to mention the names or even syllables of the names of persons connected with the speaker by marriage can hardly be separated from the reluctance evinced by so many people to utter their own names or the names of the dead or of chiefs and kings; and if the reticence as to these latter names springs mainly from superstition, we may infer that the reticence as to the former has no better foundation. That the savage’s unwillingness to mention his own name is based, at least in part, on a superstitious fear of the ill use that might be made of it by his foes, whether human or spiritual, has already been shewn. It remains to examine the similar usage in regard to the names of the dead and of royal personages. § 3. Names of the Dead tabooed.