Nothing can be more erroneous than the following sentence: ‘But the Mohammedan natives of the Eastern Coasts of Africa, who are comprehended under the name of Sawáhili, do not pronounce the hard _h_ of the Arabs; the vowels, therefore, between which it stands in their name, unite to form a diphthong, like the Italian _ai_ or the English i in wile; and Sawáhili is pronounced Sawïli’ (Inner Africa Laid Open, p. 88). The Wasawahil merely change the hard Arabic h (ح) into the softer guttural (ه). Footnote 121: I regret to read such statements as the following in the Journal of the Anthropological Society: ‘It may be asserted, without fear of exaggeration, that it is to this demand for slaves that are to be attributed the desultory and bloody wars which are waged in Central Africa.’ (On the Negro Slaves in Turkey, No. 29, April, 1870.) Footnote 122: Concerning Kilwa further details will be found in Vol. II. CHAPTER XII. PREPARATIONS FOR DEPARTURE. ‘The port of Zanzibar has little or no trade; that to Bombay consists of a little gum and ivory, brought from the mainland, with a few cloves, the only produce of the island; and the import trade is chiefly dates, and cloth from Muscat, to make turbans. These things are sent in small country vessels, which make only one voyage a year. The trade is consequently very trifling.’—CAPT. HART, Commanding H. M. S. Imogene, 1834. The dry season—and uncommonly dry it had been—was judged by all old hands unfit for travel, and I was strongly advised to defer exploration of the interior till after learning something of the coast. The Rev. Mr Erhardt’s Memoir on the Chart of Eastern and Central Africa, which threw into a huge uninterrupted Caspian half-a-dozen central lakes, and called it in the south Niandsha (Nyassa), in the north Ukerewe, and on the coast Niasa and Bahari ya Uniamési, had proposed a choice of three several routes. The first was through Dshaga (Chaga), and the lands of the Wamasai. It numbered 59 days, over level land, though studded with many isolated hills and mountains, and it traded to the Wanyamwezi, ‘of the race of Wazambiro,’ probably the Wafioma of the Usambíro district, near Karagwah. The middle caravan was reported to start ‘from Bagamoyo and Mboamaji to Uniamési.’ The general features of the country, the distances, and even the position of Ujiji, were remarkably well laid down. The ‘Stadt Ujiji,’ inhabited partly by Arabians, partly by Wahas (Wahhas), of course, did not exist: the saline stream of the Wapogo[123] and the people, whose teeth became yellow by drinking of another water, were evidently the creations of some lively negro’s fancy. The ‘third or southern caravan line,’ set out from Kilwa, and after 200 miles struck the ‘Niasa or Niandsha’ Lake. In this section the distances were miscalculated, and except the Wafipa and the Wabembe, the tribes were incorrectly named and placed.