ghain, which is pronounced by the Berbers and Negroes somewhat like the Northumberland burr, hence usually transliterated by rh in non-Semitic words.
[172] It should be noticed that these terms are throughout used as strictly defined in Eth. Ch. I.
[173] Barth's account of Wulu (IV. p. 299), "inhabited by Tawárek slaves, who are trilingues, speaking Temáshight as well as Songhay and Fulfulde," is at present generally applicable, mutatis mutandis, to most of the Songhai settlements.
[174] As so much has been made of Barth's authority in this connection, it may be well to quote his exact words: "It would seem as if they (the Sonrhay) had received, in more ancient times, several institutions from the Egyptians, with whom, I have no doubt, they maintained an intercourse by means of the energetic inhabitants of Aujila from a relatively ancient period" (IV. p. 426). Barth, therefore, does not bring the people themselves, or their language, from Egypt, but only some of their institutions, and that indirectly through the Aujila Oasis in Cyrenaica, and it may be added that this intercourse with Aujila appears to date only from about 1150 A.D. (IV. p. 585).
[175] Hacquard et Dupuis, Manuel de la langue Soñgay, parlée de Tombouctou à Say, dans la boucle du Niger, 1897, passim.
[176] "A Survey of the Ethnography of Africa," Journ. Roy. Anthr. Soc. XLIII. 1913, p. 386.
[177] Barth, IV. pp. 593-4.
[178] The Ischia of Leo Africanus, who tells us that in his time the "linguaggio detto Sungai" was current even in the provinces of Walata and Jinni (VI. ch. 2). This statement, however, like others made by Leo at second hand, must be received with caution. In these districts Songhai may have been spoken by the officials and some of the upper classes, but scarcely by the people generally, who were of Mandingan speech.
[179] Barth, IV. p. 414.