[29]. Foreigners—Arab, Persian, and Indian,—call them Sawáhili. They call themselves Msawahíli in the singular, and Wasawahíli in the plural, always accenting the penultimate syllable. In the Zangian tongues a prefixed M is evidently an abbreviation of Mti, a tree, e. g. Názi, a cocoa-nut, Mnázi, a cocoa-nut tree, or of Mtu, a man. Before a vowel it is euphoniously exchanged to Mu, e. g. Muarábu, an Arab. The plural form is Wa, a contraction of Watu, men. ‘Wá’ also is the sign of the personal, or rather of the rational animate plural opposed to ‘Má,’ and must not be confounded with the possessive pronoun ‘Wá,’ of. Mr Cooley (Memoir of the Lake Regions, &c., Reviewed, Stanford, 1864), asserts that ‘Wa mtu,’ ‘of a man,’ becomes by rejection of the singular prefix, ‘Watu,’ ‘men[‘men] (des hommes):’ consequently it is an error to call the coast people Wamrima and the mountaineers Wakilima. If so, it is an error made by every Kisawahili-speaking man. There are, however, tribes, for instance the Rabai and the Doruma, that do not prefix the normal ‘Wá,’ to form a plural. A prefixed ‘Ki,’ possibly contracted from ‘kitu,’ a thing, denotes the language, e. g. Kisawahili: it also acts diminutive, e. g. Kigito, a little mto, or river; and it appears to have at times an adjectival sense. Opposed to it is ‘Ji,’ an augmentative form, e. g. Jito, a big mto. U, possibly derived from an obsolete root which survives in the Kinyika ‘Uatu’ (a place), denotes the country, e. g. Uzaramo, Usagara, and Uzungu—Europe the land of the Wazungu. Some names arbitrarily refuse this locative, for instance, Khutu, Karagwah, Sanga, Bondei, and others: we never hear Ukhutu, and so forth. ‘U’ is also a sign of abstract words, e. g. Mzuri, a handsome man; Uzuri, beauty; Mtajíri, a merchant; Utajíri, merchandise; Refu, long; Urefu, length. I may here remark that Captain Speke’s analysis of Uzaramo and Usagara into U-za-ramo and U-sagara, the country of Ramo and Gara, making them ‘obviously triple words,’ is wholly inadmissible. The root of national and tribal names, whatever it may be, is used only exceptionally amongst the Zangian races. Upon this point I shall presently offer a few observations.
[30]. Captain Guillain (vol. iii. p. 107, et passim) is correct upon the subject of the word ‘Mrima.’ Mr Cooley (Memoir on the Lake Regions, &c., p. 8) informs us that ‘Wamrima’ (the mainland people) signifies ‘of the mainland; for it is a mistake to suppose that Mrima is but a dialectic variation of Mlimá (read, Mlíma) hill, in its primary sense, cultivable ground; it is, in truth, a corruption of the Arabic word Marâ’im, signifying the land to the west, or under the setting sun. When the early Portuguese navigators told us that the Querimba Islands were peopled by the Morimos, we must understand by this name the people of the mainland.’
This is an excellent illustration of how dangerous a thing is a smattering of philology. The ‘Arabic word Marâ’im’ is absolutely unknown to the Arabs of Zanzibar. It is evidently coined out of the dictionary from ‘رَعَمَ observavit occidentem solem.’ I would also ask how ‘Comazinghi is Arabic?’ (Geography, art. 15). Similarly, we find (Journal Royal Geographical Society, xix. 190) the Somali ‘Aber’ (error for Habr) derived from the Arabic (Hebrew?) Bar, and explained by Benú (sons), when it really signifies mother or old woman.
It may be noted that in the Kisawahili of Zanzibar, Mríma is applied to the coast generally, especially between Mtangata and the Rufiji River, and it is mostly synonymous with the Arabs’ ‘Bar el Moli,’ whereas Mlíma means a mountain. From the latter comes the diminutive Kilima, a hillock, also synonymous in composition with the French mont. It enters into many East African proper names, e. g. Kilima-njaro, Kilima-ni, &c.
I cannot agree with Messrs Norris and Beke, despite their authority as linguists, in stripping the national and racial names of their inflections, e. g. Sagara for Usagara, Zaramo for Uzaramo. Mr Cooley is equally wrong in stating that the ‘Sawáhily and the Arabs write Nika, Zeramu, and Gogo.’[Gogo.’] The Arabs may, the Wasawahíli do not, thus blunder. Captain Guillain, I have remarked, is no authority. He confounds (vol. i. p. 231) the land of Wak-wak (the Semitic Gallas) with the South African Wamakua; and, worse still, with the ‘Vatouahs.’ And (vol. i. p. 281) he writes the well-known ‘Abban’ of the Somal, ‘Hebban.’ He also unduly neglects the peculiar initial quiescent consonant M, e. g. (i. p. 456) ‘Foumo’ for ‘Mfumo.’ The bare root-word, I repeat, is never used by the people, who always qualify it by a prepositive. This, in our language Brit or Brut may be the monosyllable upon which Briton and British are built, but it is evidently barbarous to employ it without suffix. In the Zangian tongues the prefixes are clearly primitive words; nouns, not as the Rev. J. L. Doehne explains them in his Zulu-Kafir Dictionary (Cape Town, 1857), ‘pronouns, in the present state of the language, used as nominal forms compounded with other words.’[words.’]
[31]. ‘What Booken (Bukini) means I do not know.’ Wake on the Madecasses. Journal, Anthrop. Soc. No. 28, xxxi., Dr Krapf (Kisuáheli Grammar, p. 106) uses Bukini as Madagascar generally.
[32]. Captain Guillain (vol. i. 121—139, et passim) contends, and with much show of reason, that the Periplus was written after the days of Ptolemy (A.D. 139 and A.D. 161). ‘Tant de lacunes dans l’œuvre du grand géographe grec, ne semblent-elles pas assigner à son travail une place toute naturelle entre les écrits de Marin de Tyr et le Périple?’
[33]. The daily run (مَجرا) native craft varies from 40 to 50 knots per diem, and 50 may be assumed as an average. Captain Guillain estimates it higher, from 48 to 60. Abulfeda gives the Majrá or δρόμος νυχθημέρος, 100 Hashemi miles = 170 of our geographical miles, here too high a rate unless aided by currents. Other Arab authors propose 100,000 paces = 100 Roman or Arab miles = 80 geographical miles. The pilot Theophilus (Ptol. i. 9) rated the day and night run in these seas at 1000 stadia = 100 miles, or two Ptolemeian degrees; the Pelusian geographer having, I have said, reduced the degree to 500 instead of 600 stadia.
[34]. See Part II. chap. 11.
[35]. Dr Ruschenberger, I know not on what authority, says that the island is undulated and crossed by three principal ridges, whose most elevated points are 500 feet high. My information, derived from hearsay, however, not from actual inspection, assures me that the waves of ground are disposed north and south.