The Tanganyika Lake, though situated in the unexplored centre of Intertropical Africa, and until 1858 unvisited by Europeans, has a traditionary history of its own, extending through more than three centuries.

“Accounts of a great sea in the interior of Africa obtained (partially from native travellers) at Congo and Sofala,” reached the Portuguese settlements on both shores of the continent.[6] The details of de Barros (first printed in 1852), whilst affording substantially correct details, such as the length of the Lake—100 leagues—the capability of navigation, and the one large island—Ubwari—are curiously intermingled with the errors of theoretical conclusion. Subsequently Pigafetta (1591), writing upon the authority of Portuguese inquirers, affirms that there is but one lake (the N’yassa) on the confines of Angola and Monomotapa, but that there are two lakes (the Nyassa and the Tanganyika), not lying east and west, as was supposed by Ptolemy of Alexandria, but north and south of each other, and about 400 miles asunder, which give birth to the Nile. From that epoch dates the origin of our modern misconceptions concerning the Lake Region of Central Intertropical Africa. The Nyassa and the Tanganyika were now blended, then separated, according to the theories or the information of the geographer; no explorer ventured to raise from the land of mystery the veil that invested it; and the “Mombas Mission” added the colophon by confounding, with the old confusion, the Nyanza or Ukerewe, a third lake, of which they had heard at Mombasah and elsewhere. It is not wonderful then that Dr. Vincent suspected the existence or the place of the Central Lake, or that the more ignorant popularizers of knowledge confounded the waters of the Nyassa and the Ngami.[7]

The earliest name given by theoretical writers to the hypothetical single lake appears to have been Zembére, Zémbere, Zambre, Zambri, or Zembre, probably a corruption or dialectic variety of Zambesi, that river being supposed, like the Nile, the Zaire, the Manisa, and others, to be derived from it. The word Moravi or Maravi, which still deforms our maps, is the name of a large tribe or a lordly race like the Wahinda, dwelling to the south-east and south-west of the Nyassa. In the seventeenth century Luigi Mariano, a missioner residing at the Rios de Sena, calls the Central Sea the Lake of Hemosura; his description however applies to the Nyassa, Maravi or Kilwa Lake, and the word is probably a corruption of Rusuro or Lusuro, which in the language of Uhiao signifies a river or flowing water. In the ‘Mombas Mission Map’ the lake is called “See von Uniamesi,” a mere misnomer, as it is separated by hundreds of miles from the Land of the Moon: the northern part is termed Ukerewe, by a confusion with the Nyanza Lake and the southern N’hánjá, for Nyassa, the old Maravi water near Kilwa. It is not a little curious, however, that Messrs. Cooley and Macqueen should both have recorded the vernacular name of the northern Lake Tangenyika, so unaccountably omitted from the ‘Mombas Mission Map.’ The words Tanganyenka and Tanganyenko used by Dr. Livingstone, who in places appears to confound the Lake with the Nyanza and the Nyassa, are palpable mispronunciations.

The African name for the central lake is Tanganyika, signifying an anastomosis, or a meeting place (sc. of waters,) from ku tanganyika, the popular word, to join, or meet together: the initial t being changed to ch—ku changanyika for ku tanganyika—in the lingua Franca of Zanzibar doubtless gave rise to Mr. Cooley’s “Zanganyika.” The word Tanganyika is universally used by the Wajiji and other tribes near and upon the Lake. The Arabs and African strangers, when speaking loosely of it, call it indifferently the Bahari or Sea, the Ziwa or Pond, and even the Mtoni or River. The “Sea of Ujiji” would, after the fashion of Easterns, be limited to the waters in the neighbourhood of that principal depot.

The Tanganyika occupies the centre of the length of the African continent, which extends from 32° N. to 33° S. latitude, and it lies on the western extremity of the eastern third of the breadth. Its general direction is parallel to the inner African line of volcanic action drawn from Gondar southwards through the regions about Kilima-ngáo (Kilimanjáro) to Mount Njesa, the eastern wall of the Nyassa Lake. The general formation suggests, as in the case of the Dead Sea, the idea of a volcano of depression—not, like the Nyanza or Ukerewe, a vast reservoir formed by the drainage of mountains. Judging from the eye, the walls of this basin rise in an almost continuous curtain, rarely waving and infracted to 2,000 or 3,000 feet above the water-level. The lower slopes are well wooded: upon the higher summits large trees are said not to grow; the deficiency of soil, and the prevalence of high fierce winds would account for the phenomena. The lay is almost due north and south, and the form a long oval, widening in the central portions and contracting systematically at both extremities. The length of the bed was thus calculated: From Ujiji (in S. lat. 4° 55′) to Uvira (in S. lat. 3° 25′), where the narrowing of the breadth evidences approach to the northern head, was found by exploration a direct distance of 1° 30′ = 90 miles, which, allowing for the interval between Uvira and the river Rusizi, that forms the northernmost limit, may be increased to 100 rectilinear geographical miles. According to the Arab voyagers, who have frequently rounded the lake Ujiji in eight stages from the northern, and twelve from the southern, end of the lake, the extent from Ujiji to the Marungu River, therefore, is roughly computed at 150 miles. The total of length, from Uvira, in S. lat. 30° 25′, to Marungu, in S. lat. 7° 20′, would then be somewhat less than 250 rectilinear geographical miles. About Ujiji the water appears to vary in breadth from 30 to 35 miles, but the serpentine form of the banks, with a succession of serrations and indentations of salient and re-entering angles—some jutting far and irregularly into the bed—render the estimate of average difficult. The Arabs agree in correctly stating, that opposite Ujiji the shortest breadth of the lake is about equal to the channel which divides Zanzibar from the mainland, or between 23 and 24 miles. At Uvira the breadth narrows to eight miles. Assuming, therefore, the total length at 250, and the main breadth at 20, geographical miles, the circumference of the Tanganyika would represent, in round numbers, a total of 550 miles; the superficial area, which seems to vary little, covers about 5,000 square miles; and the drainage from the beginning of the great Central African depression in Unyamwezi, in E. long. 33° 58′, numbers from the eastward about 240 miles.

By B. P. thermometer the altitude of the Tanganyika is 1850 feet above the sea-level, and about 2000 feet below the adjacent plateau of Unyamwezi and the Nyanza, or northern lake. This difference of level, even did not high-hill ranges intervene, would preclude the possibility of that connection between the waters which the Arabs, by a conjecture natural to inexpert geographers, have maintained to the confusion of the learned. The topographical situation of the Tanganyika is thus the centre of a deep synclical depression in the continent, a long narrow trough in the southern spurs of Urundi, which, with its mountain-neighbour Karagwah, situated upon the equator, represents the Inner African portion of the Lunar Mountains. It may be observed that the parallel of the northern extremity of the Tanganyika nearly corresponds with the southern creek of the Nyanza, and that they are separated by an arc of the meridian of about 343 miles.

The water of the Tanganyika appears deliciously sweet and pure after the salt and bitter, the putrid and slimy produce of the wells, pits, and pools on the line of march. The people, however, who drink it willingly when afloat, prefer, when on shore, the little springs which bubble from its banks. They complain that it does not satisfy thirst, and contrast it unfavourably with the waters of its rival the Nyanza: it appears moreover, to corrode metal and leather with exceptional power. The colour of the pure and transparent mass has apparently two normal varieties: a dull sea-green—never, however, verdigris-coloured, as in the shoals of the Zanzibar seas, where the reflected blue of the atmosphere blends with the yellow of the sandy bottom; the other, a clear, soft blue—by day rarely deep and dark, like the ultramarine of the Mediterranean, but resembling the light and milky tints of tropical seas. Under a strong wind the waves soon rise in yeasty lines, foaming up from a turbid greenish surface, and the aspect becomes menacing in the extreme.

It was found impracticable to take soundings of the Tanganyika: the Arabs, however, agreed in asserting that with lines of several fathoms they found bottom only near the shores. The shingly sole shelves rapidly, without steps or overfalls, into blue water. Judging from the eye, the bottom is sandy and profusely strewn with worn pebbles. Reefs and washes were observed near the shores; it is impossible to form an idea of their position or extent, as the crews confine themselves to a few well-known lines, from which they cannot be persuaded to diverge. No shoals or shallows were seen at a distance from the coasts, and though islets are not unfrequent upon the margin, only one was observed or heard of near the centre.

The affluents of this lake are neither sufficiently numerous nor considerable to alter by sedimentary deposit the depth or the shape of the bed. The borders are generally low: a thick fringe of rush and reed, obviating erosion by the element, conceals the watery margin. Where the currents beat, they cut out a short and narrow strip of quartzose sand, profusely strewn with large shingle, gravel, comminuted shells, and marine exuviæ, with a fringe of drift formed by the joint action of wind and wave. Beyond this is a shelving plain—the principal locality for cultivation and settlements. In some parts it is a hard clay conglomerate; in others, a rich red loam, apparently stained with oxide of iron; and in others sandy, but everywhere coated with the thickest vegetation extending up to the background of mountains. The coast is here and there bluff, with miniature cliffs and headlands, whose formation is of sandstone strata tilted, broken, and distorted, or small blocks imbedded in indurated reddish earth. From the water appeared piles of a dark stone resembling angular basalt, and amongst the rock-crevices the people find the float-clay, or mountain meal, with which they decorate their persons and the sterns of their canoes. The uncultivated hill summits produce various cactaceæ; the sides are clothed with giant trees, the mvule, the tamarind, and the bauhinia. On the declines, more precipitous than the Swiss terraces, manioc and cereals grow luxuriantly, whilst the lowest levels are dark with groves of plantains and Guinea-palms.

A careful investigation and comparison of statements leads to the belief that the Tanganyika receives and absorbs the whole river-system—the net-work of streams, nullahs, and torrents—of that portion of the Central African depression whose water-shed converges towards the great reservoir. Geographers will doubt that such a mass, situated at so considerable an altitude, can maintain its level without an effluent. Moreover, the freshness of the water would, under normal circumstances, argue the escape of saline matter washed down by the influents from the area of drainage. But may not the Tanganyika, situated, like the Dead Sea, as a reservoir for supplying with humidity the winds which have parted with their moisture in the barren and arid regions of the south, maintain its general level by the exact balance of supply and evaporation? And may not the saline particles deposited in its waters be wanting in some constituent which renders them evident to the taste? One point concerning the versant has been proved by these pages, namely, that the Tanganyika cannot be drained eastward by rents in a subtending mountain ridge, as was supposed by Dr. Livingstone from an indiscriminately applied analogy with the ancient head-basin of the Zambezi. Dr. Livingstone (chap. xxiv. xxvi. et passim) informs his readers, from report of the Arabs, that the Tanganyika is a large shallow body of water; in fact, the residuum of a mass anciently much more extensive. This, however, is not and cannot be the case. In theorising upon the eastern versant and drainage of the Tanganyika, Dr. Livingstone seems to have been misled by having observed that the vast inland sea of geological ages, of which Lake Ngami and its neighbour Kumadau are now the principal remains, had been desiccated by cracks and fissures, caused in the subtending soils by earthquakes and sudden upheavals, which thus opened for the waters an exit into the Indian Ocean. This may have happened to the Nyassa, or Southern Lake; it must not, however, be generalized and extended to the Nyanza and the Tanganyika.

As in Zanzibar, there is little variety of temperature upon the Tanganyika. The violent easterly gales, which, pouring down from the cold heights of Usagara, acquire impetus sufficient to carry the current over Ugogo, Unyamwezi, and Uvinza, are here less distinctly defined. The periodical winds over the Lake—regular, but not permanent—are the south-east and the south-west, which also bring up the foulest weather. The land and sea breezes are felt almost as distinctly as upon the shores of the Indian Ocean. The breath of the morning, called by the Arabs El Barad, or the zephyr, sets in from the north. During the day are light variable breezes, which often subside, when the weather is not stormy, into calms. In the evenings a gentle afflatus comes up from the waters. Throughout the dry season the Lake becomes a wind-trap, and a heavy ground sea rolls towards the shore. In the rains there is less sea, but accidents occur from sudden and violent storms. The mountainous breakers of Arab and African informants were not seen; in fact, with a depth of three feet from ridge to dell, a wave would swamp the largest laden canoe. Wind-currents are common. Within a few hours a stream will be traversed, setting strongly to the east, and crossed by a southerly or south-westerly current. High gales, in certain localities where the waves set upon a flush, flat shore, drive the waters fifteen to twenty feet beyond the usual mark. This circumstance may partly explain the Arab’s belief in a regular Madd wa Jarr—ebb and flow—which Eastern travellers always declare to have observed upon the Tanganyika and Nyassa Lakes, and which Mr. Anderson believes to exist in the little Ngami. A mass of water so large must be, to a certain extent, subject to tidal influences; but the narrowness of the bed from east to west would render their effect almost unobservable. Mr. Francis Galton referred me for the explanation of this phenomenon to a paper ‘On the Seiches of Lakes,’ by Colonel J. R. Jackson, F.R.G.S., published in the ‘Journal of the R. G. S.,’ vol. iii. of 1833, in which the learned author refers the ebb and flow of the waters of Lake Leman, or of Geneva (and of the lakes of Zurich, Annecy, and Constance), to “an unequal pressure of the atmosphere on different parts of the lake at the same time; that is, to the simultaneous effect of columns of air of different weight or different elasticity, arising from temporary variations of temperature, or from mechanical causes.”

The scenery and the navigation of the Tanganyika have been illustrated in the last chapter. Remains only a succinct account of the physical and ethnological features of its Periplus, carefully collected from authorities on the spot.

According to the Wajiji, from their country to the Runangwa or Marungu River, which enters the Lake at the southern point, there are twelve stages; this Periplus numbers 120 khambi or stations, at most of which, however, provisions are not procurable. An extended list of fifty-three principal points was given by the guides; it is omitted, as it contains nothing beyond mere names. There are, however, sixteen tribes and districts which claim attention: of these, Ukaranga and Ujiji have already been described.

The kingdom of Urundi, which lies north of Ujiji, has a sea-face of about fifty miles; a low strip of exceeding fertility, backed at short distances by a band of high green hill. This region, rising from the Lake in a north-easterly direction, culminates into the equatorial mass of highlands which, under the name of Karagwah, forms the western spinal prolongation of the Lunar Mountains. The residence of the Mwami, or chief sultan Mwezi, is near the headstream of the Kitangure (Kitangule), or River of Karagwah, which rises at a place distant six days’ march (sixty miles), and bearing north-east from, the Tanganyika. His settlement, according to the Arabs, is of considerable extent; the huts are built of rattan, and lions abound in the vicinity.

Urundi differs from the lake regions generally in being a strictly monarchical country, locally governed by Watware or headmen, who transmit the customs and collections at stated periods to their suzerain. The Mwame, it is said, can gather in a short time a large host of warriors who are the terror of the neighbouring tribes. The Warundi are evidently natives of a high cold country; they are probably the “white people resembling Abyssinians,” and dwelling near the Lake, of whom European geographers have heard from Zanzibar. The complexion varies from a tawny yellow, the colour of the women, to a clear dark brown, which is so brightened by the daily use of ochre mixed with palm-oil, that in few cases the real tint is discernible. The men tattoo with circles and lines like cupping-cuts; some burn up alti rilievi of large shining lumps an inch in diameter, a decoration not a little resembling large boils; others chip the fore teeth like the Wanyamwezi. Their limbs are stout and well proportioned, many stand upwards of six feet high, and they bear the appearance of a manly and martial race. Their dress is the mbugu, worn in the loosest way; their arms are heavy spears, sime, and unusually strong arrows; their ornaments are beads, brass wire, and streaks of a carmine-coloured substance, like the red farinaceous powder called in India gulal, drawn across the head and forehead. The Waganga, or priests of Urundi, wear a curious hood, a thatch of long white grass or fibre, cut away at the face and allowed to depend behind over the shoulders; their half-naked figures, occasionally rattling wooden clappers, and capering causelessly like madmen, present a savage and horrid appearance. Honourable women wear long tobes of American domestics from below the arms to the ankles; they are followed by hosts of female slaves, and preserve an exceptionally modest and decorous demeanour. Their features are of the rounded African type of beauty. Their necks and bosoms support a profusion of sofi and other various-coloured beads; their foreheads are bound with frontlets, fillet-like bands of white and coral porcelain, about three fingers deep, a highly becoming ornament probably derived from Karagwah; and those who were seen by the Expedition invariably walked about with thin staves five or six feet long, pointed and knobbed as the walking-sticks of ancient Egypt.

At the northern extremity of the Urundi sea-face, and at the head of the Tanganyika, lies the land of Uzige; it is rarely visited except by the Lakist traders. This people, who, like their neighbours, cannot exist without some form of traffic, have, it is said, pursued the dows of the earlier Arab explorers with a flotilla of small canoes; it is probable that negro traders would be better received. In their country, according to the guides, six rivers fall into the Tanganyika in due order from the east: the Kuryamavenge, the Molongwe, the Karindira, the Kariba, the Kibaiba, and westernmost the Rusizi or Lusizi. The latter is the main drain of the northern countries, and the best authorities, that is to say those nearest the spot, unanimously assert that it is an influent.

The races adjoining Uzige, namely, the Wavira on the north-western head of the Tanganyika, and their southern neighbours, the Wabembe cannibals, have already been mentioned. The Wasenze inhabit the hills within or westwards of the Wabembe. Further southwards and opposite Kawele in Ujiji are the Wagoma highlanders. The lower maritime lands belonging to the Wagoma supply the gigantic mvule trees required for the largest canoes. These patriarchs of the forest are felled and shaped with little axes on the spot; when finished they are pushed and dragged down the slopes by the workmen, and are launched and paddled over to the shores of Ujiji.

South of the Wagoma are the Waguhha, who have been mentioned as the proprietors of the islets south-west of Ujiji. In their lands, according to the Arabs, is a lake or large water called Mikiziwá, whence the tribe upon its banks derives its name Wamikiziwá. Through the country of the Waguhha lies the route to Uruwwa, at present the western terminus of the Zanzibar trade. The merchant crossing the sea-arm which separates Kasenge from the mainland of the Tanganyika, strikes towards Uruwwa; the line runs over low levels shelving towards the lake, cut by a reticulation of streams unfordable after rain, and varied by hilly and rolling ground. Provisions are everywhere procurable, but the people, like the Wavinza, are considered dangerous. At Uruwwa the khete, or string of beads, is half the size of that current in other countries. The price of ivory per frasilah is 15 miranga, or 150 large khete of white, small-blue, and coarse-red porcelain beads, the latter called Lungenga; besides which a string of sungomaji (pigeon-egg beads), and a few sámesáme, or coral-beads, are thrown in. The route numbers nine long or sixteen short stages; the general direction is south-westerly. Kiyombo, the sultan of Uruwwa, is at present friendly with the Arabs; he trades in ivory, slaves, and a little copper from Katata or Katanga, a district distant fifteen marches north-west of Usenda, the now well-known capital of the great chief Kazembe. The grandfather of the present Kazembe, the “viceroy” of the country lying south-west of the Tanganyika, and feudatory to Mwátá yá Nvo, the sovereign of “Uropua,” was first visited by Dr. Lacerda, governor of the Rios de Sena, in 1798-99. The traveller died, however, after being nine months in the country, without recording the name and position of the African capital; the former was supplied by the expedition sent under Major Monteiro and Captain Gamitto in 1831-32; it is variously pronounced Lucenda, Luenda, and by the Arabs Usenda, the difference being caused probably by dialect or inflexion. According to the Arabs, the Kazembe visited by the Portuguese expedition in 1831, died about 1837, and was succeeded by his son the present chief. He is described as a man of middle age, of light-coloured complexion, handsomely dressed in a Surat cap, silk coat, and embroidered loin cloth; he is rich in copper, ivory, and slaves, cloth and furniture, muskets and gunpowder. Many Arabs, probably half-castes, are said to be living with him in high esteem, and the medium of intercourse is the Kisawahili. Though he has many wives, he allows his subjects but one each, puts both adulterer and adulteress to death, and generally punishes by gouging out one or both eyes.

On the Uruwwa route caravans are composed wholly of private slaves; the races of the Tanganyika will not carry loads, and the Wanyamwezi, unmaritime savages like the Kafirs, who have a mortal dread and abhorrence of water, refuse to advance beyond Ujiji. On account of its dangers, the thriving merchants have hitherto abandoned this line to debtors and desperate men.

South of Uguhha lies the unimportant tribe of Wat’hembwe, whose possessions are within sight of Kawele in Ujiji. The race adjoining them is the Wakatete or Wakadete, and the country is called by the Arabs Awwal Marungu, on the northern frontier of Marungu. Marungu is one of the most important divisions of the lands about the Tanganyika. Amayr bin Said el Shaksi, a sturdy old merchant from Oman, who, wrecked about twelve years ago on that part of the coast, had spent five months with the people, living on roots and grasses, divides the region generically termed Marungu into three distinct provinces—Marungu to the north, Karungu in the centre, and Urungu on the south. Others mention a western Marungu, divided from the eastern by the Runangwa River, and they call the former in contradistinction Marungu Tafuna, from its sultan.

Western Marungu extends according to the Arabs in depth from Ut’hembwe to the Wabisa, a tribe holding extensive lands westward of the Nyassa Lake. Travellers from Unyamwezi to K’hokoro meet, near Ufipa, caravans of the northern Wabisa en route to Kilwa. Between Marungu and Usenda, the capital of the Kazembe, the road lies through the district of Kavvire, distant seven marches; thence nine stages conduct them to the end of the journey. There is an upper land route through Uruwwa for those travelling from Ujiji to Usenda, and many caravans have passed from Unyanyembe direct through K’hokoro and Ufipa, to the country of the Kazembe. Mr. Cooley (“Geography of N’yassi,” p. 7) conjectures that the Ambios or Imbies, Zimbas or Muzimbas, celebrated by the old Portuguese historians of Africa on account of an irruption, in 1570, from the north as far as the Zambezi River, “were no other than the M’Biza, or Moviza, as they are called by the Portuguese who still occupy its (the Nyassa’s) south-western banks.” The proper name of this well-known tribe is Wábísá (in the sing. Mbisá), not Wábíshá, as it is pronounced at Zanzibar, where every merchant knows “Bisha ivory.” The Wábísá extend according to the Arabs from the west of the Nyassa or Kilwa Lake towards the south of the Tanganyika. They dress in bark-cloth, carry down their fine ivory to Tete and Kilimani (Quillimane); and every four or five years a caravan appears at Kilwa, where, confounding their hosts with the Portuguese, they call every Arab “muzungu,” or white man. They are a semi-pastoral tribe, fond of commerce, and said to be civil and hospitable to strangers. It must be observed that those geographers are in error who connect the Wabisa with the Wanyamwezi; they are distinct in manners and appearance, habits and language. Mr. Cooley has, for instance, asserted that “the ‘Moviza’ and the ‘Monomoezi’ are similar in physical character and national marks.” The only mark known to the Wabisa is the kishshah, or crest of hair; not, as Khamisi Wa Tani asserted to Mr. Cooley (“Inner Africa laid Open,” p. 61), a dotted line on the nose and forehead; whereas, the Wanyamwezi, as has been seen, puncture the skin. Thus Lacerda calls the “Moviza” a frizzled and periwigged people. The Arabs deny the assertion of Pereira, recorded by Bowdich, that the Moviza, like the Wahiao, file their teeth.

Marungu is described by the Arabs as a hilly country like Ujiji and Uvira: the precincts of the lake, however, are here less bold than the opposite shore. Off the coast lie four or five islands, two of which, according to the Arabs, are of considerable size; the only name given is Ukungwe, which appears, however, to be rather the name of the farthest point visible from Kasenge, and bearing S. 58° E. On the north-western frontier of Marungu, and about three marches from the lake, is the district called Utumbara, from Mtumbara its sultan. This Utumbara, which must not be confounded with the district of the same name in Northern Unyamwezi, is said by the Arabs to be fifteen to twenty days’ march from Usenda.

Marungu, though considered dangerous, has often been visited by Arab merchants. After touching at Kasenge they coast along Uguhha for four days, not daring to land there in consequence of an event that happened about 1841-42. A large Arab caravan of 200 armed slaves, led by Mohammed bin Salih and Sulayman bin Nasir, and with four coadjutors, Abd el Al and Ibn Habib, Shiahs of Bahrayn, Nasir and Rashid bin Salim el Harisi (who soon afterwards died at Marungu) took boat to Marungu, and in due time arrived at Usenda. They completed their cargo, and were returning in a single boat, when they were persuaded by the Sultan Mtumbara to land, and to assist him in annihilating a neighbour, Sámá or Kipyoká, living at about one day’s march from the Lake. The Arabs, aided by Africans, attacked a boma, or palisade, where, bursting in, they found Sámá’s brother sitting upon pombe, with his wife. The villagers poured in a shower of arrows, to which the Arabs replied by shooting down the happy couple over their cups. Sámá’s people fled, but presently returning they massacred the slaves of the Arabs, who were obliged to take refuge in the grass till aid was afforded by their employer Mtumbara. Sámá, thus victorious, burned the Arab boat, and, compelling the merchants to return to Usenda, seized the first opportunity of slaying his rival. The Arabs have found means of sending letters to their friends, but they appear unable to leave the country. Their correspondence declares them to be living in favour with the Kazembe, who has presented them with large rice-shambas, that they have collected ivory and copper in large quantities, but are unable to find porters. This being highly improbable in a land where in 1807 a slave cost five, and a tusk of ivory six or seven squares of Indian piece-goods, and as, moreover, several merchants, deluded by exaggerated accounts of the Kazembe’s wealth and liberality, intrusted these men with considerable ventures, of which no tidings have as yet reached the creditors’ ears, the more acute Arabs suspect that their countrymen are living from hand to mouth about Usenda, and are cultivating the land with scant prospect of quitting it.

The people of Marungu are called Wámbozwá by the Arabs; they are subject to no king, but live under local rulers, and are ever at war with their neighbours. They are a dark and plain, a wild and uncomely race. Amongst these people is observed a custom which connects them with the Wangindo, Wahiao, and the slave races dwelling inland from Kilwa. They pierce the upper lip and gradually enlarge the aperture till the end projects in a kind of bill beyond the nose and chin, giving to the countenance a peculiar duck-like appearance. The Arabs, who abhor this hideous vagary of fashion, scarify the sides of the hole and attempt to make the flesh grow by the application of rock-salt. The people of Marungu, however, are little valued as slaves; they are surly and stubborn, exceedingly depraved, and addicted to desertion.

Crossing the Runangwa or Marungu River, which, draining the southern countries towards the Tanganyika, is represented to equal the Malagarazi in volume, the traveller passes through the districts of Marungu Tafuna, Ubeyya, and Iwemba. Thence, turning to the north, he enters the country of the Wapoka, between whom and the Lake lie the Wasowwa and the Wafipa. This coast is divided from the opposite shore by a voyage of fourteen hours; it is a hilly expanse divided by low plains, where men swarm according to the natives like ants. At a short distance from the shore lies the Mvuma group, seven rocks or islets, three of which are considerable in size, and the largest, shaped like a cone, breeds goats in plenty, whilst the sea around is rich in fish. There are other islets in the neighbourhood, but none are of importance.

Ufipa is an extensive district fertilised by many rivers. It produces grain in abundance, and the wild rice is of excellent flavour. Cattle abounded there before the Watuta, who held part of the country, began a system of plunder and waste, which ended in their emigration to the north of Uvinza; cows, formerly purchased for a few strings of cheap white beads, are now rare and dear. The Wafipa are a wild but kindly people, who seldom carry arms: they have ever welcomed the merchants that visited them for slaves and ivory, and they are subject to four or five principal chiefs. The servile specimens seen at Unyanyembe were more like the jungle races of the Deccan than Africans—small and short, sooty and shrunken men, so timid, ignorant, and suspicious, that it was found impossible to obtain from them the simplest specimen of their dialect. Some of them, like the Wanyoro, had extracted all the lower incisors.

North of the Wafipa, according to the Arabs, lies another tribe, called Wat’hembe (?), an offshoot from the people on the opposite side of the Tanganyika. Here the lake receives a small river called the Murunguru (?). The circuit of the Tanganyika concludes with the Wat’hongwe, called from their sultan or their founder Wat’hongwe Kapana. In clear weather their long promontory is the furthest point visible from Kawele in Ujiji; and their lands extend northwards to Ukaranga and the Malagarazi River.

Such are the most important details culled from a mass of Arab oral geography: they are offered however to the reader without any guarantee of correctness. The principal authorities are the Shaykh Snay bin Amir el Harsi and Amayr bin Said el Shaksi; the latter was an eye-witness. All the vague accounts noted down from casual informants were submitted to them for an imprimatur. Their knowledge and experience surpassing those of others, it was judged better to record information upon trust from them only, rather than to heap together reliable and unreliable details, and as some travellers do, by striking out a medium, inevitably to confuse fact with fiction. Yet it is the explorer’s unpleasant duty throughout these lands to doubt everything that has not been subjected to his own eyes. The boldest might look at the “Mombas Mission Map” and tremble.