The mucn talked-of acclimatization appears doubtful; every fresh attack weakening the system, paves the way for a successor. The consequences of protracted or repeated fevers are severe derangements of the system: these sequelae, which the Arabs, who greatly dread them, call "El Nazlah," or the defluxion of humours, are either visceral or cerebral. The liver, in its efforts to purify the blood, secretes bile in superabundance; hence indigestion and its concomitants, enlargement of the spleen and swellings of the stomach and lower extremities, irritability, hemicrania, liver-coughs, tooth-aches, and painful cramps or spasms, showing entire derangement of the nervous system. The severest cases often end in the loss of one or more senses-idiocy, blindness occasioned by atony of the optic nerve, loss of virility, stiff joints, contracted sinews and a partial paralysis of the extremities, which is sometimes inveterately lingering. The signs of convalescence recognised by the natives are severe ulcerations of the mouth and tongue, feversores and herpetic eruptions, especially upon the lips, sluggish boils, and painful eruptions. Amongst the latter is the nyongo, the pitam of Western India. The attack lasts from five to ten days, during which the patient suffers great pain: his face, hands, and feet swell, and his skin is red and fiery, as if with erysipelas. The Arabs cure it by frequent bathings with cold water, in which copper has been rubbed down.

This typhoid fever appears to defy the usual preventives of sleeping between fires, of wearing gauze-guards over the mouth, and others enumerated by travellers. It is differently treated by all races. The Banyans (Hindus) rely upon starvation and diet; some few use opium. The Indian Kojahs and the Arabs apply their usual variety of simples and nostrums, especially fumigation and steaming, as preventives. Some Europeans at Zanzibar use quinine. before the springs, that is to say, before the new and the full moon -an ancient superstition in the East-and employ cathartics or emetics on all occasions when sudden and severe exercise, following a long rest, induces biliousness. The French mostly affect quinine, which, being taken without due precaution, has caused many deaths. The Delagoans treat the disease with cold affusions and diaphoresis. The Portuguese of the Mozambique, who certainly have profited by length of experience, begin with mild emetics, followed by tonics, principally bark and bitter herbs. They induce profuse perspirations by vapour-baths and draughts of hot kanji, or rice-water. They insist upon the strictest diet: eggs, butter-milk, and stimulants are proscribed till the stomach regains strength to resume its accustomed functions. A little bread and tea or rice-water, and two small slices of "roti," are allowed, much strong meat being considered dangerous as strong drink. The Portuguese and the Arabs account bleeding most pernicious, and

justly so, as the action of the heart cannot be restored. It may be added that in obstinate fevers the Tinctura Warburgii has been found a specific: it was tested at Zanzibar by Lieutenant-Colonel Hamerton, whose doors were sometimes blockaded by the Persian attendants of the Prince clamouring for the "cure of death."

The other diseases of the coast are small-pox brought from the interior, dysentery,-the scourge of caravans as of camps and ships, -abdominal hernia, elephantiasis, pleurisis, and pelagra. The baras, or white leprosy, commonly appears upon the shins and armbones: it commences with violent prurigo, after which the skin changes colour, and, except in rare cases, it does not return to its normal hue. Strangers are liable to this disease. The Arabs speak of juzam, or black leprosy; but it was not observed on the coast. Hydrocele and sarcocele are less general than at Zanzibar. Boils and blains are common and painful, and at Mombasah and other places sores and ulcers, called by the Arabs Kinah, and by the Sawahili Mti, Donda, Kidonda, and Kibata, are but little less terrible than those of Aden and Yemen. They attack the legs, generally the shins near the ankles, swell the limbs into a semblance of elephantiasis, and end by causing distortion of the bone and lameness: the toes are often almost obliterated. Even the Arab traders passing through the country sometimes suffer from them.

The number of rivers on the Zanzibar coast has been greatly exaggerated: creeks or sea-arms, like the "Tuaca or Nash River" of Mombasah, the "Quavi or Cuavo" of Kilwa, and the "Lindi River" near Cape Delgado, have been raised to the rank of large perennial streams. On the other hand, Captain Owen omits all notice of the great Rufiji River, which first appeared in the 'Observations sur la Côte de Zanguebar,' by M. Saulnier de Mondevit, Lieutenant de Vaisseau, under the name of Oufidgy.* The Mombas Mission Map' also altogether ignores the Kingani River, which is constantly navigated for some distance by the Wamrima or coast clans.

According to that law of nature which renders the streams in the southern hemisphere inferior in volume to those north of the equator, from Mombasah to Kilwa, the coast, which is everywhere cut by runnels and rivulets, presents but three debouchments that deserve the name of rivers: these are the Pangani, the Kingani, and the Rufiji. They are by no means the sluggish and stagnant streams which infect the air of the western coast. They pour from high inland mountains through deep channels to the sea, and but for the "fungu," or bars of rock and sand which mark their junction with salt water, they might be entered at all times by the larger craft, which must now float in with the high tide. In 1824 the

* See Chap. xi.

Pangani had 2 fathoms of water over the bar; now the fairway passage is not more than 7.50 feet deep, and when a gale blows from the east it is faced with a line of bar-breakers. The coast, moreover, is garnished with "diabolitos" or outliers, little black rocks of a siliceous gravelly conglomerate. Native craft can ride at anchor in these three rivers, and, as in Western Africa, the prevailing winds blowing up the channels enable the mariner to stem the velocity of the current.

The Pangani, which was explored in 1824 at the expense of life by Lieutenant Reitz, attached to Captain Owen's survey, has been so frequently visited, that it has now lost all interest. About 30 miles south of its embouchure, near the town of Saadani, flows a little perennial stream, called the Gama, a mere fillet of water in the dry season, and nearly absorbed by the deep loose sand of the bed. Rising in the highlands of Nguru or Ngu, at a distance of about ten days' march from the coast, and draining the countries between the Pangani and the Kingani Rivers, it bifurcates near the sea, forming a diminutive delta.

The mouth of the Kingani River,† which lies in s. lat 6° 15', is situated north of a point of land projecting seawards beyond the coast-town of Bagamoyo. The estuary is about half a mile broad, and the adjacent mud-banks are flooded by the high tides: the bed narrows after 2 or 3 miles to 100 yards. In appearance the Kingani contrasts strongly with the Pangani River: its low banks, instead of huge palms and massive vegetation, bear only stunted bushes and a few mangroves; its waters, moreover, are of a muddy tawny colour, verging upon red, soft and sweet as if fed by rain, whereas its neighbours appear of a slaty white hue, and have, moreover, the harsh rough taste of rock-streams and snow-water. Concerning the upper bed of the Kingani River, details will appear in a future page.

From its volume and extent, the Rufiji River is the most in

* These details rest upon the authority of Arab information.

+ Mr. Cooley (Memoir on the Geography of Nyassi,' p. 22, which appeared in the Journal of the Royal Geographical Society,' vol. xv., of 1845) asserts that "the Kingani means bar river; the stream, though large, is quite inaccessible to trade." If the latter clause be correct, it has changed of late years. The name is derived by the people from Kingá, a fire brand; they connect it with the legend of a celebrated chief in the olden time. Moreover, Mr. Cooley (in Inner Africa Laid Open,' p. 80) thus confounds the Kingani with the Pangani river :-"We know, on unquestionable evidence, that the mouth of the Ruvú is in lat. 6° 15' s.-fifty nautical miles south of the Pangani." Although Rufu (Ruvu) is to a certain extent a generic term for a river, it is frequently applied to the Pangani, but rarely to the Kingani.

Mr. Cooley (Geography of Ngami,' p. 20) remarks, that "The Sawahili in general say Rivuma and Rufiji. The inland nations and the Arabs substitute L for the initial R." In most dialects of the eastern branch of the great South African family of languages, however, the liquids L and R are interchangeable. In Kisawahili, the Lingua Franca of these regions, Arabs and the more civilized speakers rightly distinguish between the two when there is a difference of sense;

teresting feature in the potamology of the Mrima. It is still involved in some mystery, and nothing beyond a brief description drawn from the accounts of ignorant natives has yet been given to the world. Like the Zambezi and unlike the Nile, this main drain of the Zanzibar coast swells under the pressure of heavy rains in the interior from January or February till May and June. According to the pilots, its delta is cut by eleven or twelve distinct branches, of which one only admits Arab sailing craft, though several can be ascended by canoes. The "Rubbans" point to the low mangrove bank, a breach in the ancient sea-beach opposite, and due west of the Kisimani Mafiyah, or the watering-pits of Monfia island, which lie in s. lat. 7° 56′ 41′′. Boats ascend the stream till the swollen outfall becomes too rapid :-this point is placed at the distance of seven days, and, during the inundation, they row from village to village. The settlements are raised upon piles or poles, beyond the reach of the waters and the crocodiles. The tribes adjoining the Rufiji are barbarous and exclusive: their sultans or chiefs must be conciliated by presents, and, if not in force, the traveller will incur the risk of being plundered. The Rufiji, as is afterwards explained, becomes in its upper course the Rwaha River. The exploration of the delta of this great river is, like the outlet of the Juba on the northern part of the coast, still a desideratum. The

where such is not the case they prefer the R. Thus Major Gamitto ('O Muata Cazembe. Introducçao,' p. xxii) says, "Adverterei porém que os Muizas e os Cazembes ou Lundas nao pronuntia a letra R, em cujo logar usam do L." The slaves, on the other hand, and the barbarians of the interior, convert R into L, and appear indeed so fond of the latter letter, that they will prefix and infix it ad libitum. As, however, they have no standard, and the Arabs have, the civilized pronunciation will invariably be retained in these pages, though it is not contended that it is the more correct. To this general rule there are, of course, exceptions: the Arabs, for instance, name their ancient emporium Kilwa, whence the Portuguese Quiloa; the Africans, on the other hand, call it Kirwá. It is impossible not to remark the recurrence of the syllable Ru or Lu as an initial in the names of East African rivers; for two instances of many, the Rusizi and the Ruguru; by addition of a syllable it becomes a modification of Rufu or Lufu, as in Rufiji, Rufuma (Livuma), Rufuta, and Rufita. A glance at the map will show that the same formation extends to Western Africa, always bearing in mind that the interior barbarians prefer Lu and Lufu to Ru and Rufu. The Arabs explain the fundamental idea of the word to be that of destruction, scil., by water. In the Kisawahili, and most of its Zangian cognates, Kú Fá, or by a normal increment Ku Kufá, signifies to die, a word which brings to mind the , (he died) of the

Arabic.

* According to Dr. Livingstone (chap. 26), the Leeambaye, which proved to be the upper stream of the Zambezi, floods in July and August, a little before the period of inundation in the various streams which unite to form the Niger. On the other hand, the principal African rivers of the southern hemisphere, the Zambezi and the Rufiji, inundate in February, March, and April; on the course of the Upper Nile the rains last from March till November. Thus it appears certain that the same cause, namely, the northing and southing of the sun, which attracts the mass of vapour derived from the ocean reservoirs around, produces opposite results, modified by local features of ground, according to the position of the several streams with respect to the equator.

barbarous tribes conciliated, this river might be made one of the great gates for commerce into Eastern Africa.

The Mrima is no exception to the general rule of the country; it contains many settlements, but not a single town. "Oppidulis præcingitur." A chain of little hamlets, which, when near-neighbours, are comprised under a single comprehensive name, although each is distinguished by its own appellation, girds the broken line of point, inlet, and estuary. Between Bagamoyo and Kaole, a distance of 3 miles, there are nearly a dozen. The traveller wonders that men do not combine to build a city which might insure safety, comfort, and society. The unconstructive African, however, loves his hut, and has a superstitious horror of stone walls; moreover, the exigencies of commerce, as will presently be explained, tend to disperse the population.

*

The principal settlements, in their order from Pangani southward, are the following:-At the distance of a few miles lies the roadstead of Kipumbui; here the approach is rendered perilous by the "diabolitos," which are 2 miles distant from the coast, and are steep-to, giving no soundings at 60 fathoms. Beyond Kipumbui, and bearing north-west from Zanzibar town, lies Saadani, the principal port of the ancient "kingdom of Atondo." This is the Portuguese corruption of Utondwe, a point or headland bounding the bay, and still showing vestiges of habitations; moreover the people of Saadani are still called Watondwe.† Saadani was lately burnt to the ground by Mohammed, the headman of Marumbi, a petty village distant about 3 miles, during the absence of his cousin "Bori," who is considered the bravest and is respected as the most powerful diwan or chief of the Mrima. South of Saadani lies Whinde, a settlement whose well-armed inhabitants have earned for themselves an infamous celebrity as kidnappers; assisted by Kisabengo, a Mzegura robber-chief of Ukami, an inner district, they can raise from 300 to 400 muskets, and they have wasted with fire and sword the fairest provinces of Usagara. Bagamoyo (in s. lat. 6° 17′) is one of the great points of departure for the caravans trading to Unyamwezi ; it is garrisoned, as well as its neighbour Kaole, by a small body of Baloch. South of Bagamoyo lies Konduchi, in s. lat. 6° 40′ 24′′: numerous small settlements, of which the principal are Msasani, Mzizima, Magogoni, and Mbezi, prolong the line of copal depôts to the great centre of the Mrima traffic, Mbuamaji, commonly called Boromaji, in s. lat. 6° 51′ 49′′. From Mbuamaji to the

This is given upon the authority of the late Lieut.-Col. Hamerton. Kipumbui was not visited by the Expedition.

+ According to Mr. Cooley (Geography of N'yassi,' p. 22), the banks in front of the Kingani River are called Watondui, or the picking-grounds; i.c., the banks for gathering shell fish. This derivation is not confirmed by the people of the country; moreover, the form “Watondui" would be a personal plural, not a locative noun.