CHAPTER II.

The _Lady Alice_ afloat again—Her mate and crew—Anxiety on her behalf—On hallowed ground—Unwelcome visitors: in the haunts of the Ruga- Ruga—The aborigines’ last retreat—A scene of blood—The robbers of the Lake, and their tutelary spirits—A geological problem: were there once two lakes?—Rising waters—The abode of the genii—A storm—Forest fires—At the mouth of Cameron’s “outlet”—The Lukuga creek.

_June 11._—The saucy English-built boat which had made the acquaintance of all the bays and inlets of the Victoria Nyanza, which had been borne on the shoulders of sturdy men across the plains and through the ravines of Unyoro, had halted on the verge of the cliff rising above Beatrice Gulf, had thrust her bows among the papyrus of the Alexandra Nile, ridden gaily over the dark lakes of Karagwé, and crossed the inundated plains of Usagusi, and the crocodile-haunted river of Uvinza, is at last afloat upon the deep-blue waters of the Tanganika.

She is about to explore the mountain barriers which enfold the lake, for the discovery of some gap which lets out, or is supposed to let out, the surplus water of rivers which, from a dim and remote period, have been pouring into it from all sides.

She has a consort now, a lumbering, heavy, but staunch mate, a canoe cut out from an enormous teak-tree which once grew in some wooded gorge in the Goma mountains. The canoe is called the _Meofu_, and is the property of Muini Kheri, governor of Ujiji, who has kindly lent it to me. As he is my friend, he says he will not charge me anything for the loan. But the governor and I know each other pretty well, and I know that when I return from the voyage, I shall have to make him a present. In Oriental and African lands, remuneration, hire, compensation, guerdon, and present are terms nearly related to one another.

The boat and her consort are ready on the 11th of June, 1876. The boat’s crew have been most carefully selected. They are all young, agile, faithful creatures. Their names and ages are as follows: Uledi, the coxswain, 25 years; Saywa, his cousin, 17; Shumari, his brother, 18; Murabo, 20; Mpwapwa, 22; Marzouk, 23; Akida, 20; Mambu, 20; Wadi Baraka, 24; Zaidi Rufiji, 27; Matiko, 19. Two supernumeraries are the boy gun- bearers, Billali and Mabruki, 17 and 15 years respectively. After eighteen months’ experience with them it has been decided by all that these are the elect of the Expedition for boat-work, though they are by no means the champions of the march. But as they have only light loads, there has never been reason to complain of them.

There is much handshaking, many cries of “Take care of yourselves,” and then both boat and canoe hoist sail, turning their heads along the coast to the south.

Kasimbu, two miles from Ugoy to Ujiji, sends forth her Arab and slave inhabitants to cry out their farewells, and half an hour afterwards we are at the mouth of the Liuché river.

The reason why Arabs, Wajiji, and Wangwana have been more than usually demonstrative is that they will not believe that such a frail structure as our boat will be able to endure the heavy waves of the Tanganika. They declared we should all be drowned, but our Wangwana have ridiculed their fears, and quoted her brilliant exploit round a lake twice the size of the Tanganika, and so at last they had come to be satisfied with a dismal “Well, you shall see!”

_June 12._—During nearly the whole of the next day our voyage to the south is along the forest-clad slopes of Ulambola and the tawny plains of Ukaranga, until we arrive at the mouth of the Malagarazi River. At 3 p.m. we rowed up river, which at the mouth is about 600 yards wide, and sends a turbid brown stream into the lake. When a continuous south-west wind blows, its waters are known to tinge the lake with its colour as far as Ujiji. The river soon narrows to 200 yards, and about five miles up to 150 yards. I sounded twice, and obtained over 50 feet each time. The southern bank is very mountainous, but on the opposite side stretches a plain until the detached ranges of Ukaranga become massed, about five miles from the lake, and, running easterly, form the northern bank of the river.

_June 13._—On the 13th our voyage was along the bold mountain spurs of Kawendi, forming a steep, rock-bound coast, indented at frequent intervals with calm, pool-like bays, and their heights clothed with solemn woods.

At noon we were off Kabogo’s lofty headland, and remembering that Dr. Livingstone had said that he could find no bottom at 300 fathoms, I sounded a mile off shore, and found 109 fathoms. At two miles off I found no depth with 140 fathoms. I then fastened sixty fathoms more, but at 1200 feet obtained no bottom.

About four miles south of dreaded Kabogo, on a narrow strip of sand, we beached our boat and canoe far out of the heavy surf, and then climbed the 2000 feet-high slopes in search of game; but the grass was high, the jungle dense, the slope steep and fatiguing, and we had to return without sighting a single head.

_June 14._—Next day we coasted along land familiar to me from my journey with Livingstone to Unyanyembé, and at 7 p.m. encamped at Urimba, about a mile south-west of the river Luwajeri or Luwegeri.

_June 15._—Having been so successful in January 1872, I sallied out the next day over ground which I looked upon with reverence. The exact place covered by our little tent, only six feet square of land, was hallowed by associations of an intercourse which will never, never be repeated. I recognised the tree above which we hoisted our mighty crimson and white banner to attract the lagging land caravan, the plain where I had dropped the zebra, the exact spot where I shot a fine fat goose for breakfast, the aspiring peak of Kivanga, the weird looking mountains of Tongwé. I knew my road here, and dwelt upon all its features, until the old life seemed renewed, and all things seemed as before.

But I resumed my search. In an hour I am two miles from camp, and in view of a herd of zebra. Billali becomes feverish lest I should miss the game, and, like an honest, faithful servant taking enormous interest in his master’s success, lies down to hug the ground in piteous stillness. I advance a few paces cautiously behind a scraggy acacia, and in a few seconds two of the noble creatures are dead, and the others are sweeping round a clump of hills, whimpering for their lost companions. As we have now enough meat to last us several days, I give them their liberty.

_June 16._—The day is devoted to cutting the meat into long strips and drying it over wooden grates, while each of the forty men composing the lake exploring band seems profoundly impressed with the necessity of forestalling future demands on his digestive organs by consuming injudicious quantities there and then.

In the midst of this most innocent recreation there stepped forth to our view some sinister objects—Ruga-Ruga! As undesirable as wolves in a stern Siberian winter to an unarmed party in a solitary sledge are the Ruga-Ruga to peaceful travellers in an African forest or wilderness. Whatever the accident that brought them, their very presence suggested the possibility and probability of a bloody struggle. They are bandits, wretches devoted to plunder and murder, men whose hands are at all times ready to be imbrued in blood.

They are representatives of that tribe which has desolated and depopulated beautiful Kawendi from the Malagarazi river down to the Rungwa. All alike—whether Arabs, Wajiji, Wangwana, Wanyamwezi, or the aborigines of the land—owe them an unpaid debt of vengeance for the blood they have shed. It was not our special task, however, to undertake the repayment, therefore neither by word nor look did we betray any antipathy.

We gave them gifts of meat at their own request. The tobacco gourd passed round in their polluted, crimeful hands, and we grasped their hands in token of amity—and parted.

_June 17._—On the 17th of June we continued our voyage from Urimba towards Kungwé cape, one of the projecting spurs from the Kungwé mountains, and in the evening camped on Bongo Island, a few miles south- west from Ndereh, the robbers’ village. We were visited in the night by about sixty of them armed with muskets. Though it was an unusual hour, and an unseasonable one for receiving visitors, we avoided trouble, and by parting with cloth and exhausting the powers of suavity, we happily avoided a rupture with the wild and bloody men of Ndereh, and before dawn stole away unperceived on our journey.

_June 18._—The peaks of Kungwé are probably from 2500 to 3000 feet above the lake. They are not only interesting from their singular appearance, but also as being a refuge for the last remaining families of the aborigines of Kawendi. On the topmost and most inaccessible heights dwells the remnant of the once powerful nation which in old times—so tradition relates—overran Uhha and Uvinza, and were a terror to the Wakalaganza. They cultivate the slopes of their strongholds, which amply repay them for their labour. Fuel is found in the gorges between the peaks, and means of defence are at hand in the huge rocks which they have piled up ready to repel the daring intruder. Their elders retain the traditions of the race whence they sprang; and in their charge are the Lares and Penates of old Kawendi—the Muzimu. In the home of the eagles they find a precarious existence, as a seed to reproduce another nation, or as a short respite before complete extermination.

[Illustration:

KUNGWÉ PEAKS. (_From a sketch near the entrance to the Luwulungu torrent bed._) ]

The best view of this interesting clump of mountain heights is to be had off the mouth of the torrent Luwulungu.

From Cape Kungwé south, the coast as far as Ulambula consists of a lofty mountain front, pierced by several deep and most picturesque inlets, gorges, ravines, and rifts. Into these pours the Luwulungu, rushing along a steep, stony bed, from the chasms and defiles overshadowed by the tall peaks of Kungwé—and the Lubugwé, emptying its waters into a pretty cove penetrating to the very heart of the mountain wall. At an angle of 45° this mountain wall rises up to the height of 2000 feet, clothed from base to summit with the verdure of cane, wild grasses, and tall straight trees with silvery stems. Then next comes the Kasuma inlet, and here, straight before our eyes, is seen a river dropping, in a succession of falls, from the lofty summit, into shadowy depths of branching tamarinds, acacias, and teak. All is silent in the deep- bosomed cove except the rhythmic waterfall, and the trees stand still, as though fascinated by the music, and the grim heights frown a silent approval; the pale blue arm of the lake rests expectant for the moment when it shall receive the impetuous child of the mountains which it sees leaping down to it from above, and flashing so brightly at every great leap. Along the glorious, green, steep headlands we wind in and out, cast a glance in at Numbi’s pretty cove, and encamp for the night near the bold cape of Ulambula.

_June 19._—We resumed the voyage on the 19th, and, shortly after leaving our camp in the neighbourhood of the cape, saw a point of land connected by a narrow neck with the mainland, under which were two natural arches, spanning two channels. From the cape the mountain range gradually recedes from the lake, until, near the Rugufu river, it again approaches and finally forms the headlands of Buyramembé.

A little south of the cape the crest of a small and lately submerged island was also seen. At noon I took observations for latitude, at the north end of Kabogo, an island lying parallel to the mainland at a distance of from 300 to 500 yards. On the shores, both of mainland and island, flourishes the borassus palm. Kabogo was once densely peopled, but the bandits of Ndereh, the scourge of Kawendi, have caused them to emigrate to other districts to crave protection from chiefs more powerful than their own.

About 2 P.M. we came in view of Kiwesa, which appeared from the lake to be a very large village. But as we approached its shores under sail, we were struck with the silence which reigned around, and the sight of a large herd of buffalo grazing near the village still more astonished us.

The guides declared that only five weeks before they had stopped in it and traded with Ponda, the chief, and they could give no reason why—as two boats under sail would most likely attract the attention of the natives—the people of Ponda did not appear on the shore.

We resolved to venture in to discover the cause. There was a deathly silence around. Numbers of earthenware pots, whole, and apparently but little used, were strewn about the beach and among the reeds flanking the path which led to the village, besides stools, staffs, hand-brooms, gourds, etc., etc. This was ominous. There was probably a trap or a snare of some kind laid for us. We retreated therefore hastily to the boat and canoe, and thirty men were armed. Thus, better prepared against the wiles of savage men, we advanced again cautiously towards the village.

As we were surmounting the high ground on which the village was built, we saw a sight which froze the blood—the body of a poor old man, in a decomposed state, with a broad spear-wound in the back, and near it much dried blood. He had probably been dead five or six days.

A few yards farther, we saw the decapitated corpse of another man, and 10 feet from it, in a furrow or water channel, the bodies of three men and a woman, one of them dislimbed.

We arrived before the village. The defences were broken down and burnt. About fifty huts still stood unharmed by fire, but all the others were consumed. A few scorched banana stalks stood as a witness of the fury of the conflagration. But despite the black ruin and the charred embers which so plentifully strewed the ground, evidences were clear that flight had been hasty and compulsory, for all the articles that constitute the furniture of African families lay scattered in such numbers around us that an African museum might have been completely stocked. Stools, mats, spears, drinking-vessels, cooking-pots of all sizes, walking-staffs, war-clubs, baskets, trenchers, wooden basins, scoops, etc. There were also abundant proofs that this ruin was recent; a few wood-rails still smoked, the hearths were still warm, the dead were not putrefied. A coal-black cat made a dash from one of the houses yet standing, and the sudden motion startled us all in this place of death and vengeance.

Ponda, the chief of this village, had no doubt given some provocation to this unknown enemy. Para thought the enemy must have been the robbers of Ndereh, for the condition of the village bore signs of superior energy in the attack. And yet it had been constructed with a view to secure immunity from the fate which generally overtakes weak communities in Africa in the neighbourhood of ferocious and war-loving tribes. A wide ditch—in some parts 10 feet deep—and a strong palisade, with an earthwork, surrounded the settlement. The lake was close by to supply water, the country near it was open, and the sharpshooters’ nest-like towers commanded a wide view. From some thirty bleached skulls arranged before Ponda’s own house we argued that he did not himself fail to proceed to the same extremes which his enemies had now adopted to his utter ruin. It is the same story throughout Africa.

We resumed our voyage for the mouth of the Rugufu river. The shore between Kiwesa and the river is comparatively low. The waves have so beaten and shaken the low red bluffs and soft ferruginous clay that numerous landslips are constantly occurring. The débris are then vigorously pounded and crushed by the surf, and at length they are spread out into a narrow line of beach at their base, over which the spluttering waves surge up continuously.

The impression received at Ujiji, that the Tanganika is rising, was confirmed whenever we neared low shores. Especially was it the case at the Rugufu river, and Para the guide, as we entered it, stood up and exclaimed:—

“Oh, mother, mother, mother, see ye now! When I was with that other white man here, we camped on a strip of land which is now buried in the water! The Tanganika is indeed eating the land!”

The Rugufu oozes out from the midst of a broad bed of papyrus and reeds between precipitous banks.

On leaving the river we coasted along the bluff, steeply rising slopes of a mountain range which trends south-south-east as far as the settlements of Ruhinga, Kafisya, Katavi, and Kantamba.

Between the Rugufu and Buyramembé Point a stratum of a very dark hornblende slate is visible, resting upon gneiss in undulating, vertical, or diagonal lines; farther on we come to a stratified quartzose and greenstone rock. On the crest of this part of the range and its projecting spur is a thin forest of poor trees. The soil too is poor, and much mixed with shaly débris.

The mouth of the Gezeh river is a frequent haunt of herds of buffaloes, and also, being a fine haven, of trading canoes. Among the stories related of this place is one of a wonderful escape of a party of Wajiji traders from the bandits of Ndereh. The robbers stole into the camp while the Wajiji were all asleep, but some of the canoemen, awaking, punted their boats out of reach, and shouted to their comrades, who sprang into the water to avoid the fate that would otherwise have certainly overtaken them all.

The settlements of Kafisya, and the others just mentioned, have such an ill repute that I cannot imagine any necessity inducing a traveller to cultivate the acquaintance of the evil-conditioned people, unless, of course, he is so rich in cloth and followers that waste of them is of little consequence to him.

It is said that when they see the Wajiji trading canoes pass by them, the robbers pray to the Muzimu of Katavi to induce the Msaga—the tempestuous sea—to drive them ashore. The Muzimu of Katavi is one of the most powerful spirits along the shores of the Tanganika, according to legendary lore. Though he is capable of much mischief, he takes freaks of charity into his whimsical head, such as gratuitously killing buffaloes and then informing the inhabitants where the meat may be found; he is also said to have a relentless animosity towards the bandits of Ndereh, and frequently entraps them to their destruction. A conical knoll is called, after this spirit, Katavi’s Hill.

Before the mouth of the Mkombé river there lies a low submerged island, just west of south from Katavi. Only a few shrubs and the heads of some tall cane were visible above the surface of the water.

_June 20._—From Gezeh to Igangwé Cove was a good day’s journey against the south-easter which blew strongly against us. Igangwé Cove penetrates about a mile deep into the mountain folds. Though it was the season when the grasses are becoming sere, and some of the trees lose that vivid greenness of foliage which is their glory during the rainy season, the aspect of the slopes had still a freshness and beauty which, with the placid mirror-like cove, made a picture worth preserving.

_June 21._—A day’s journey south by east took us to the village of Karema—the chief being Massi-Kamba, a sub-chief of Kapufi, king of Fipa. It is situated in the angle of a bay which begins at Igangwé Point and terminates in the weird rock-piles of Mpimbwé Cape.

All that mass of looming upland from Igangwé to within a few miles of Karema is included in the country of Kawendi or, as it is sometimes called, Tongwé. South of that line begins Fipa.

Arabs are beginning to establish themselves at Karema for trade, the Wa- fipa being more amenable to reason than the “scattered and peeled” tribes of Tongwé.

Between Karema and Mpimbwé Cape lies a fine country studded with coves and hills, square-topped and round. Game is abundant and easy of approach. A buffalo and a small red antelope were obtained by me here, the two shots supplying the crews with abundance of meat.

Proceeding some eight miles south-west towards Mpimbwé, we come to a narrow ridge rising about 600 feet above the lake. Its shore is deeply indented, and the wash of waves has bared enormous masses of granite.

At the south-western corner of this bay there is a neck of low land which all but connects Mpimbwé ridge with the mainland; only half a mile’s breadth of low land prevents Mpimbwé being an island. Near Kipendi Point, which is halfway between Mpimbwé ridge and Karema, there is a tree in the lake which was pointed out to me as being not many years ago on dry land. There is now 9 feet of water around it!

Mpimbwé Cape offers a view similar to the rocks of Wezi, only of a still more gigantic size and a ruder grandeur.

Their appearance betrays the effects of great waves which have at one time swept over them, pouring their waters into their recesses, cleansing by force every cranny and flaw of their vegetable mould, and washing out of them every particle of soil, until, one day, by some sudden convulsion of nature, the lake subsided, leaving, a hundred feet above its surface, these grey and naked masses of granite.

Any one who has seen on a rock-bound coast the war of sea-wave against granite, basalt, and sandstone, will at once recognize the effects visible at Mpimbwé. There lie piled up rocks, hundreds of tons in weight, some of them appearing to rest in such precarious positions that it seems as if the finger of an infant would suffice to push them into the deep blue lake. These, however, are not scored and grooved as are those exposed to the eroding influences of ceaseless ocean waves; they are cleanly fractured, with their external angles only exhibiting a rough polish or roundness which I take to be clear signs that at some remote time they were exposed to waves of great power. Besides, the condition of the rocks at the water-line confirms this theory.

Still, it is strange that the lake should have been rising steadily ever since living men can remember, and that those rocks of Mpimbwé should bear witness to the lake’s subsidence.

_June 25._—On the 25th of June, after coasting the western side of this extraordinary range and proceeding south some fifteen or sixteen miles, we arrived at Mkerengi Island, in the bay of Kirando; the large island of Makokomo lying to west of us a mile off. The natives of these parts are very amiable, though extremely superstitious. At the north-west end of Makokomo there is another lately submerged island. Close to the south-west end is a group of inhabited islands, Kankamba and Funeh being the largest and most fertile.

Kirando is situated among other large villages in what appears to be a plain, hemmed in on the east by the continuation of the mountain ridges which we lost sight of when we left Karema Bay. The truncated cone of Chakavola terminates the ridge of Mpimbwé, and lies north-east from Mkerengi.

Continuing our voyage southerly along the coast between the isles of Kankamba and its brothers and the shore of the mainland, and passing a couple of creeks running deep inshore, we came to Cape Muntuwa. From this cape to Msamba Island, where we encamped, the lake is edged by one successive series of gigantic blocks and crags of granite. Rock rises above rock, and fragment above fragment. Here towers a colossal mass the size of a two-storied house, bearing upon it a similar mass perhaps entire, but more probably split with a singularly clean and fresh fracture, and there springs up from the surrounding chaos a columnar block like a closed hand with outstretched fore-finger—but everywhere there is the same huge disarray, ruin, and confusion.

We had need to be cautious in sailing along the coast, because for several hundred yards into the lake the rocky masses, which the uneasy billows only exposed in glimpses, rose nearly to the surface.

A suspicion flashed into my mind as these new features revealed themselves that in remote times this part of the lake—from Mpimbwé south—was a separate lake, and that Mpimbwé ridge was connected with some portion of the western coast—probably the southern portion of Goma—for while coasting from the extreme north end of lake Tanganika down to Mpimbwé, I saw nothing resembling in character this portion of the coast. In no part of all the eastern coast down to Mpimbwé is there anything to lead me to suppose that the lake was ever higher than at present; but from Mpimbwé to Msamba I see numerous traces that the lake has been many yards higher than it is at present. All this dreary ruin of wave-dismantled and polished rock was at one time covered with water.

_June 26._—On the 26th we camped at Mtosi, where Livingstone, who calls it Motoshi, camped on the 23rd of October, 1872. The chief’s name is Kokira. A beautiful little bay leads from the lake to the miserable village where he lives.

_June 27._—We rested on Msamba Island for the evening of the 27th. The islanders told us that there was a cave about 60 yards in length on the mainland opposite, where they sometimes hid in time of danger. For a small island Msamba is densely populated, and every inch seems cultivated. The islanders are clever manufacturers of a strong, coarse, cotton cloth: cotton being abundant in Fipa. The Rukugu river empties into Msamba Bay.

The irregular ridge which follows the coast between Msamba Island and Wanpembé, our next journey, is remarkable for a solitary columnar rock rising from 50 to 80 feet high and about half a mile from Column Point. Rounding Kantentieh Point, we have a view of three columnar rocks, the central one being singularly like a mutilated Memnonium. These columns are visible from a considerable distance north or south.

Before reaching Wanpembé, Para, the guide, gathered a peculiar kind of berry called _owindi_, from a low scrubby tree, whose appearance was anything but promising for such a fragrant production as he now showed to us. The odour was not unlike that of lavender, and its strength was such that all in the boat near him were benefited by its exquisite perfume.

In the little cove close to Wanpembé, on the north side of the point on which it is situate, the boat floated over the submerged fence of a village, and her keel was 3 feet above it.

We obtained abundance of provisions at this large village, but as there were some Watuta strangers within the palisades, our visit was not an agreeable one. However, despite their insolence, the peace was not broken.

Minza, a neighbouring village, is also very large, and possesses a strong stockade, the base of it being embanked with the earth excavated from the ditch.

There appears to be no diminution in the altitude of the mountain ranges which lie along the entire east coast of the Tanganika, or on the western side, as we have had the west mountains plainly in view since leaving Mpimbwé. Now and then we saw small streams issue into the lake, but met no river of any importance, until we came to the Zinga, or Mui- Zinga, as the Wajiji call it, which separates Fipa from Urungu.

_June 30._—On the 30th we were coasting along the base of the mountain ranges of Urungu, and passing by Kalavera Point came to a bay before which were two small grass-covered islets. On a point of the mainland, nearly opposite these, stands Kakungu village. This point is formed of a grey shaly rock supporting a white clay, out of which the Wajiji on their return homeward paint the bows of their canoes. The scenery just beyond is bold and imposing.

Kirungwé Point consists of perpendicular walls—from 50 to 200 feet high above the lake—of a fine reddish sandstone with horizontal strata. Their peculiar appearance may be imagined when the boat’s crew cried out:—

“Oh, mother, this is a fort! See, there are the windows, and here is one of the gates.”

Kirungwé Point appears to be a lofty swelling ridge, cut straight through to an unknown depth. There seems ground for believing that this ridge was once a prolongation of the plateau of Marungu, as the rocks are of the same material, and both sides of the lake show similar results of a sudden subsidence without disturbance of the strata.

South of Kirungwé, or Castle Point, there lies what we may almost call an island, which the guides said a few years ago was connected with the mainland. It is almost entirely separated now. A village which once nestled comfortably in the hollow between the rising ridges is now half buried in water. The huts appeared ready to collapse, for the water had already flooded them. This village was called Ma-Zombé.

In the evening, as we prepared to encamp, four canoes, of Ujiji, loaded with women and children, to the number of sixty-four—slaves from the Rufuvu river and from Muriro’s—passed by our camp.

The bay of Kawa, which we passed through next day, is very picturesque; woods clothe the slopes and heights, and huts for the accommodation of the Muzimus, or spirits, have been erected in several of the bends. Kawa river empties into this bay.

_July 2, 3._—During the 2nd and 3rd of July we rowed close to the uninhabited shore, and at noon of the 3rd arrived at the extreme south end of the lake—which we ascertained to be 8° 47′ south latitude—in the district of Ukituta. The little stream Kapata issues into the lake at this end through a dense and dark grove, the dead trees standing in front of the grove bearing witness to the destructive action of the rising waters.

[Illustration: THE EXTREME SOUTHERN REACH OF LAKE TANGANIKA.]

I scoured the country in search of game, but, though tracks of buffalo were numerous, I failed to obtain a glimpse of a single head. Safeni, who was the coxswain of the _Meofu_ canoe, accompanied me, and pointed out various points of interest in connection with Livingstone, as we followed the road which he had travelled over in his last fatal journey to Lake Bemba. The myombo and borassus palm flourished on the higher terraces.

_July 4._—On the 4th, after rounding a point of a ridge—three miles from the Kapata—which lay north, we turned westward between Ntondwé Island and the mainland, and then passing by Murikwa Island, we reached, in two hours, the southern termination of the western shore of the Tanganika, whence may be seen the Wezi river tumbling down from the plateau of Urungu.

The village of Mwangala, where we camped, was at first hidden from our view by a dense line of water-cane, which sheltered its small fishing- canoes from the storms of the lake. One glance at the village fence informed us that here also was evidence that the lake was rising. We asked the natives if they did not think the water was gaining on them.

“Can you not see?” said they. “Another rain and we shall have to break away from here, and build anew.”

“Where does the water of the lake go to?”

“It goes north, then it seems to come back upon us stronger than ever.”

“But is there no river about here that goes towards the west?”

“We never heard of any.”

That part of the western coast which extends from Mbeté or Mombeté to the south, as far as the Rufuvu river, is sacred ground in the lore of the ancients of Urungu. Each crag and grove, each awful mountain brow and echoing gorge, has its solemn associations of spirits. Vague and indescribable beings, engendered by fear and intense superstition, govern the scene. Any accident that may befall, any untoward event or tragedy that may occur, before the sanctuaries of these unreal powers, is carefully treasured in the memories of the people with increased awe and dread of the Spirits of the Rocks.

Such associations cling to the strange tabular mounts or natural towers, called Mtombwa. The height of these is about 1200 feet above the lake. They once formed parts of the plateau of Urungu, though now separated from it by the same agency which created the fathomless gulf of the Tanganika.

Within a distance of two miles are three separate mounts, which bear a resemblance to one another. The first is called Mtombwa, the next Kateye, the third Kapembwa. Their three spirits are also closely akin to one another, for they all rule the wave and the wind, and dwell on summits. Kateye is, I believe, the son of Kapembwa, the Jupiter, and Mtombwa, the Juno, of Tanganika tradition.

[Illustration:

THE “HIGH PLACES” OF THE SPIRIT MTOMBWA: VIEW OF MTOMBWA URUNGU. ]

As we row past close to their base, we look up to admire the cliffy heights rising in terraces one above another; each terrace-ledge is marked by a thin line of scrubby bush. Beyond Kateye, the grey front of the paternal Kapembwa looms up with an extraordinary height and massive grandeur.

From Kapembwa to Polombwé Cape the plateau merges in a cliff-crowned wall 1500 feet above the lake. The cliff itself is probably 200 feet high, rising above a slope bristling with great rocks half hidden by the verdure of trees, bush, and grass. Yet the natives cultivate some part, and their fields were seen far up the steep ascending slopes.

[Illustration: KAPEMBWA.]

_July 6._—On the 6th we left the neighbourhood of Polombwé, at a place called Umisepa, and rowed round into the Rufuvu river.

This river is about 400 yards wide, and retains that width for about three miles—flowing with a current of a knot an hour, and between lofty wood-clothed mountains—and then broadens out into a lake-like expanse nearly a mile wide. From the right or south bank of the river, a plain slopes gradually to the grand cliffy walls of Kapembwa. From here to Liendé village, where we camped, our course lay east-south-east.

Here, as elsewhere, the water has encroached upon the soil, and has flooded a large portion of land formerly devoted to tillage. It is a populous spot, indeed the most populous we had seen since leaving Ujiji. Our reception by the people was most cordial, and I was not sorry to become acquainted with such gentle creatures. Not one angry word or insolent look was exchanged, but they visited us with the greatest confidence, and a lively interest for barter. We obtained such abundance of provisions here that we might have cruised for a month without having to call at any native port.

The chief, Kiumeh, or Chiuma-Nanga, who lives at Mkigusa village, was visited with all due ceremony, and proved a most kindly old man. I gladly rewarded him for his small presents of food, and separated from him with feelings of attachment.

Livingstone, who was here in May 1867, writes of this plain and river as follows:—

“We came to a village about 2′ west of the confluence. The village has a meadow about four miles wide, in which buffaloes disport themselves, but they are very wild, and hide in the gigantic grasses. The Lofu—or Lofubu (Rufuvu)—is a quarter of a mile wide, but higher up 300 yards.”

Between the 6th of July, 1876, and May 1867, that is, in nine years’ time, the river Rufuvu has encroached upon the “meadow” which Livingstone saw by over a thousand yards!

It is true the plain or meadow is very low, and that 2 feet or more of a rise would extend the river over half a mile more of ground, but the proofs are gathering that the lake has been steadily rising. What was meadow land in the days when Livingstone made the acquaintance of the people of Liendé is now clear water half covered with growths of pale- blue lotus. The depth of the river in mid-channel is 21 feet.

I should estimate the population of the plain from Polombwé westward to where the river narrows between the hills, a district of about eight square miles, as about 2000 souls. We heard of Wangwana and some Arabs camping at a village called Kungwé higher up the river on the left bank, but as we had no occasion for their acquaintance we did not deem it necessary to go through the form of visiting them.

_July 7._—On the 7th, soon after quitting the Rufuvu river, we had a rough experience of the worst Ma’anda “south-wester”—Para or Ruango, our guides, had ever been in. The _Meofu_ was soon disabled, for its rudder was swept away, but being towed behind by a rope, it was fortunately not lost, while our boat flew with double-reefed lug over the wild waves like a seagull. The tempest sang in our ears, and the waters hissed as they flew by us with great high curling crests. But Kasawa Cape was still before us, and no shelter could be obtained until we had rounded it. We shook a reef out, lest we might be swamped, and the increased force swept us over the topmost crests at such a speed as made Para and Ruango set their teeth. The canoe was out of sight; along the rock-bound shore thundered the surf; the wind was rising into a hurricane, but Kasawa was getting nearer to view, and we held on with all sail. In fifteen minutes we were safe behind the grey bluffs of the headland, in a little creek amid a heap of driftwood, and the haunt of hippopotamus and crocodile. I sent a land party back to hunt up news of the missing canoe, and by night received the glad news that soon after they were disabled they had managed to beach their boat without injury.

Between Kasawa and Kipimpi capes there are deep bays, which I have taken the liberty of calling the Cameron Bays.[2] A sterile and bleached country stretches on every hand around these bays, and the general appearance of their sterility is somewhat increased by the chalky character of some of the low cliffs.

Footnote 2:

So called after Verney Lovett Cameron, Commander R.N., the first to navigate the southern half of Lake Tanganika.

North of the river Rufuvu extends Uemba. Ruemba—the country of the lake L’iemba—in the language of the great Bisa tribe, which all speak, with slight differences of dialect, in this region, signifies “lake.” Mapota River separates Uemba from Marungu.

Between Kipimpi Cape and Kalambwé Cape, King Muriro, or “Fire,” an immigrant from Unyamwezi, has, with the aid of a colony of restless spirits, established a formidable village called Akalunga, close to the lake. It is a resort for slavers, for Muriro has numbers of slaves on hand to exchange for powder and guns, and his people are always roving about on the look-out for more.

From Kalambwé Cape northward the mountains loom higher and steeper, the shore is indented with many narrow inlets, vertical strata of greenstone being thus exposed, with thin forests crowning the loose soil which covers them. The depressions between the hilltops are numerous and shallow, consequently the drainage is quickly carried away in small rills.

Beyond the Mapota the scenery becomes still bolder, and the more imposing woods impart with their varied hues of foliage and waving crowns a picturesqueness that since leaving Fipa has been wanting in the landscape of the stupendous and upspringing, terraced plateau wall of Western Urungu, or in the uniform contours of Eastern Urungu.

_July 11._—At a camp near an inlet north of Kalambwé Cape we set fire to some grass to have a more open view of our surroundings. In an hour it had ascended the steep slope, and was raging triumphantly on the summit. Three nights after we saw it still burning about fifteen miles north of the locality whence it had first started, like a crown of glory on a mountain-top.

Observation of this fire, and many others, explains why, in the midst of African uplands nourishing a dense forest, we suddenly come across narrow, far-penetrating plains covered with grass. They are, no doubt, so many tonguelike extensions from some broader, grass-covered expanses caused by fierce fires. Wherever the ground retains an excessive quantity of moisture, grasses, with stalks as thick as cane, spring up during three months’ rain to a height of eight to ten, sometimes fifteen, feet. In May these grasses become sere, and by June are as dry as tinder. The smallest spark suffices to set them in a flame, and the noise of two brigades of infantry fighting would hardly exceed the terrible rack, crackle, and explosions made by the onrush of the wind- swept element. It devours everything that stands before it, and roasts the surface of the ground, leaving it parched, blackened, and fissured.

Though the mountains of Marungu are steep, rugged, and craggy, the district is surprisingly populous. Through the chasms and great cañons with which the mountains are sometimes cleft, we saw the summits of other high mountains, fully 2500 feet above the lake, occupied by villages, the inhabitants of which, from the inaccessibility of the position they had selected, were evidently harassed by some more powerful tribes to the westward.

The neighbourhood of Zongweh Cape is specially distinguished for its lofty cones and great mountain masses. Mount Murumbi, 2000 feet above the lake, near Muri-Kiassi Cape, is a striking feature of the coast of Marungu.

The wooded slopes and dense forest growths which fill the gorges are haunts of what the Wangwana call “Soko,” a distinctive title they have given either to gorillas or chimpanzees. I heard the voices of several at Lunangwa river, but as they were at a considerable distance from me, I could not distinguish any great difference between the noise they created and that which a number of villagers might make while quarrelling.

The Rubuko, or Lofuko, a considerable river, divides Marungu from Tembwé. On the south side of the river is Mompara, or, for short, Para, remarkable as being the place whereat Livingstone embarked in canoes, February 1869, to proceed to Ujiji for the first time (‘Last Journals,’ vol. ii.):—

“_14th February, 1869._—Arrived at Tanganika. Parra is the name of the land at the confluence of the river Lofuko.”

The chief of the Para village is patronised by Jumah Merikani, who, while he is absent in Rua collecting slaves and ivory, entrusts his canoes to the chief’s charge, from which it appears that the latter is a trustworthy man. Formerly Sheikh Sayid bin Habib honoured him with the same confidence.

Four hours’ sail brought us to the wooded headlands of Tembwé, the most projecting of which is about twenty-five miles from Makokomo Island, in the Bay of Kirando, on the east coast of the lake.

Near this point is seen a lofty range, rising a few miles from the lake in an irregular line of peaks, which, as it is depressed towards the north, presents a more arid appearance, and presently forms a range of a much lower altitude than the mountains of Tongwé, Fipa, Urungu, or Marungu. This continues—with gaps here and there for rivers to escape through to the lake—until a little north of Kasengé Island, where it rises again into the mountains of Goma, the highest of all round Lake Tanganika.

At Kankindwa, which is in a little cove near Tembwé Cape, a native told us that the Lukuga flowed _out_ of the lake to Rua; another denied that flatly, and a third said that the Lukuga flowed out of the lake towards Rua, but that, meeting another river descending towards the Tanganika, it was stopped, and the two rivers formed a lake.

_July 14._—As we sailed north from Tembwé and observed the comparatively low altitude of the Uguha range, I began to feel that it was, of all the countries we had seen, the most likely to have a gap for the escape of the waters. It reminded me of some parts of Usukuma, on Lake Victoria. We explored the mouths of the Ruanda, Kasenga, Ruwye-ya, Rutuku, and Kahanda rivers, and then from Mirembwé Cape sailed for the Lukuga—the river that formed now the most interesting object of our exploration.

[Illustration: MOUNT MURUMBI.]

_July 15._—On the evening of the 15th of July we made the acquaintance of Kawe-Niangeh, the chief of the district on the south bank of the Lukuga Creek near the mouth. He remembered Cameron distinctly, described his person and dress, and informed me that he had accompanied him to the reeds which he said blocked the head of the creek. At the time of his visit, he said—pointing at the same time to a long line of breakers which marked three-fourths of the broad entrance from the lake to the creek—there were two spits of sand running from either side of the mouth, and there was a small fishing settlement on the one which projected from our side. As they were now covered up, he entertained a suspicion that Cameron had dropped some powerful medicine, which had brought on this destruction! If one white man had brought so much mischief, what might not two white men do? “Why,” said he, “the whole country will be inundated, and nothing will be left except the tops of the great mountains!”

We laughed at this, and, eventually joking him out of these ideas, succeeded in obtaining his guidance to explore the creek, and in eliciting the following items, which I jotted down in my journal the same evening:—

“_July 15._—Opinion at the mouth of the Lukuga is much divided respecting this river, or creek, or inlet, or whatever it may be. The information, when compared with Cameron’s statement, is altogether incomprehensible. The old men and chiefs say that formerly the Luwegeri met the Lukuga, and that the meeting of the waters formed the lake. The result of this marriage of the Lukuga from the west, and the Luwegeri from the east, is the Tanganika, and a cordial understanding between the waters has been kept up until lately, when it appears that the Lukuga has begun to be restive and wayward, for it sometimes flows west, and sometimes east; or, in other words, the Lukuga during the rainy season flows into the Tanganika, bearing with it an immense amount of water, grass, wood, and other matter, but during the dry season, when the south-east monsoon prevails, the Lukuga is borne west, lifting its head clear of the dry ground and mud-banks, and flows down to the Kamalondo, near Kalumbi’s country, under the name of Ruindi or Luindi. Until this rainy season, or say March of this year, 1876, there stood a low bank of earth or mud, several hundred yards long, between the Luindi and Lukuga, but this year’s rainfall has united the two rivers, the Lukuga flowing over this by Miketo’s country into Rua. Kamalondo is a river, and not a lake, being another name for the Lualaba.

“When Cameron was here in 1874, there was a spit of dry sand lined with grass or cane, projecting from the south side, and a similar one from the north side, the two being separated by a narrow channel, but to-day both spits are covered with a line of wild breakers. The spot where Cameron camped is no longer tenable, but is exposed to the billows of the Tanganika, which at this season are driven in by the south-east monsoon.

“Take it any way you please, such conflict of opinions among people who ought to know what an outlet or an outflowing river is—many of them having seen the Luapula flow out from Bemba lake, others having seen the Lualaba plunge down from Mweru lake—makes it clear that there is either a crisis approaching in nature or that it has lately taken place, or is occurring—one cannot say which until the Lukuga is explored, and this work I propose to begin to-morrow.”

Cameron says, on p. 304, in vol. i. of his ‘Across Africa’:—

“Its entrance was more than a mile across, but closed by a _grass- grown sandbank_ with the exception of a channel 300 or 400 yards wide, and _across the channel there is a sill_ where the surf breaks heavily at times, although there is more than a fathom of water at its most shallow part.”

An inflowing river meeting the billows of the Tanganika might be supposed to form a “surf,” or a sandy sill, it being only natural that there should be a conflict between the opposing forces. To this struggle then must be attributed the formation of the “sill of sand” which Cameron said ran across the channel.

_July 16._—On the 16th we sailed up the creek.

The mouth of the Lukuga, which was about 2500 yards wide, narrowed after a mile to 800 yards, and after another mile to 400 or 500 yards. Upon rounding the point of land on which Mkampemba stands, and where there is a considerable tract under tillage, I observed that the water changed its colour to a reddish brown, owing to the ferruginous conglomerate of which the low bluffs on either side are composed. This was also a proof to me that there was no outflowing river here. Clear water outflowing from the Tanganika, only two miles from the lake, ought never to be so deeply discoloured.

As we proceeded on, the chief told us to stop, and threw a stick into the water, asking us to note how, despite the ripple and wind from lakeward, the stick and the water-bubbles persisted in struggling against them towards the lake. His face was triumphant as he thought he had completely proved one part of his statement, that water came into the lake. It only remained now, as he thought, to prove that water flowed out towards the west.

Wherever there were indentations in the bluffs that banked it in, or a dip in the low grass-covered débris beneath, a growth of Mateté or water-cane and papyrus filled up these bits of still water, but mid- channel was clear and maintained a breadth of open white water ranging from 90 to 450 yards.

Within an hour we arrived at the extremity of the open water, which had gradually been narrowed in width, by the increasing abundance of papyrus, from 250 yards to 40 yards. We ceased rowing, and gently glided up to the barrier of papyrus, which had now completely closed up the creek from bank to bank, like a luxuriant field of tall Indian corn. We sounded at the base of these reeds along a breadth of 40 yards, and obtained from 7 to 11 feet of water! With a portable level I attempted to ascertain a current; the level indicated none! Into a little pool, completely sheltered by the broadside of the boat, we threw a chip or two, and some sticks. In five minutes the chips had moved towards the reeds about a foot! We then crushed our way through about twenty yards of the papyrus, and came to impassable mudbanks, black as pitch, and seething with animal life. Returning to the boat, I asked four men to stand close together, and, mounting their shoulders with an oar for support, I endeavoured with a glass to obtain a general view. I saw a broad belt some 250 or 300 yards wide of a papyrus-grown depression, lying east and west between gently sloping banks, thinly covered with scrubby acacia. Here and there were pools of open water, and beyond were a few trees growing, as it seemed to me, right in the bed. I caused some of my men to attempt to cross from one bank to the other, but the muddy ooze was not sufficiently firm to bear the weight of a man.

I then cut a disc of wood a foot in diameter, drove a nail in, and folded cotton under its head. I then rove a cord 5 feet in length through this, suspending to one end an earthenware pot, with which I tried an experiment. Along the hedge of papyrus I measured 1000 feet with a tape line, both ends of the track marked by a broad riband of sheeting tied to a papyrus reed. Then, proceeding to the eastern or lake end of the track, I dropped the earthenware pot, which, after filling, sank and drew the wooden disc level with the water. I noted the chronometer instantly, while the boat was rowed away from the scene. The wind from the lake blew strong at the time.

The board floated from lakeward towards the papyrus 822 feet in one hour and forty seconds.

In the afternoon, wind calm and water tranquil, the disc floated in the opposite direction, or towards the lake, 159 feet in nineteen minutes and thirty seconds, which is at the rate of about 600 feet in the hour.

This was of itself conclusive proof that there was _no_ current at this date (July 16, 1876). Still I was curious to see the river flowing out. The next day, therefore, accompanied by the chief and fifteen men of the Expedition, we started overland along the banks of this rush- and mud- choked depression for three or four miles. The trend of the several streams we passed was from north-west to south-east—that is, towards the lake. At Elwani village we came to the road from Monyi’s, which is used by people proceeding to Unguvwa, Luwelezi, or Marungu on the other side of the Lukuga. Two men from the village accompanied us to the Lukuga ford. When we reached the foot of the hill, we first came to the dry bed of the Kibamba. In the rainy season this stream drains the eastern slopes of the Ki-yanja ridge with a south-east trend. The grass-stalks, still lying down from the force of the water, lay with their tops pointing lakeward.

From the dry mud-bed of the Kibamba to the cane-grass-choked bed of Lukuga was but a step. During the wet season the Kibamba evidently overflowed broadly, and made its way among the mateté of the Lukuga.

We tramped on along a path leading over prostrate reeds and cane, and came at length to where the ground began to be moist. The reeds on either side of it rose to the height of 10 or 12 feet, their tops interlacing, and the stalks, therefore, forming the sides of a narrow tunnel. The path sank here and there into ditch-like hollows filled with cool water from 9 inches to 3 feet deep, with transverse dykes of mud raised above it at intervals.

Finally, after proceeding some two hundred yards, we came to the centre of this reed-covered depression—called by the natives “Mitwanzi”--and here the chief, trampling a wider space among the reeds, pointed out in triumph water indisputably flowing westward!

The water felt cold, but it was only 68° Fahr., or 7° cooler than the Lukuga. I crossed over to the opposite or southern bank, on the shoulders of two of my men. The bed was uneven; sometimes the men rose until the water was barely over their ankles, then again they sank to their hips. The trees I had noticed from the open creek stood on a point projecting from the southern bank across the Mitwanzi, but they were now dead, as the former dry tract had become quaggy. The name Lukuga clings to the bed until a few miles west of Miketo’s, when it becomes known as the Luindi, Ruindi, or Luimbi.

The Mitwanzi is still daily traversed without trouble by men, women, and children.

We travelled another three miles along the Mitwanzi, until we came to the southern end of the Ki-yanja ridge, for it is through the gap between this and the Kihunga ridge, which terminates on the south bank, that the Lukuga flows toward the west. Even here it was but a trivial stream, oozing and trickling through a cane-grass grove.

The most interesting object here was the rounded end of the Ki-yanja ridge, sloping at an angle of 30°. As the highest point is probably between 600 and 1000 feet, there has been some agency at work to wear down this gap through the ferruginous conglomerate and soft sandstone—and some agency stronger than this trivial stream smothered in reeds, for it has no force or power.

We got back to Lumba Creek, where we had left our boat and canoe, late at night. The next day was devoted to sounding the creek from the Mitwanzi to the outer bar.

_July 17._—The next morning I took a trip to the top of the conical hill behind Mkampemba, a village of Kawe-Niangeh, to lay out and take bearings.

I am of the opinion, after taking all things into consideration, that Kahangwa Cape was, at a remote period, connected with Kungwé Cape on the east coast—that the Lukuga was the effluent of the lake as it stood then, that the lake was at that period at a much higher altitude than it is at present, that the northern half of the lake is of a later formation, and that, owing to the subsidence of that portion, and the collapsing of the barrier or the Kahangwa Cape and Kungwé Cape ridge, the waters south emptied into that of the deep gulf north, and left the channel of the Lukuga to be employed as the bed of the affluents Kibamba and Lumba, or the eastern slope of the Ki-yanja ridge, to feed the lake. But now that the extension of the profound bed—created by some great earthquake, which fractured and disparted the plateau of Uhha, Urundi, Ubembé, Goma, &c.—is on the eve of being filled up, the ancient affluent is about to resume its old duties of conveying the surplus waters of the Tanganika down into the valley of the Livingstone, and thence, along its majestic winding course, to the Atlantic Ocean.

I say this after having circumnavigated the lake and examined it most thoroughly. Underground caverns are myths, the fables of Wangwana and superstitious natives. The great deep lengthy cañon occupied by the fathomless lake is not closed in by rocks of such a nature as to admit of the theory of underground passages. It is rimmed by mountains and hills—the least altitude is 600 feet, the highest 4000 feet, above the lake. But to those seeking an elucidation of the fact that an enormous fresh-water lake is without an out-flowing river, are presented as rational solutions the stream-worn gap in the conglomerate of the ridges Kihunga and Liyanja, the wave-washed rocks and boulders of Mpembwé and all along the eastern coast down to Urungu, the bare headlands of Tembwé, and the naked steeps and cliffs of Kungwé and Karinzi. It is an undeniable fact that if the evaporation from a body of water be greater than the supply, that water must necessarily become saline from the particles washed into it from salt-beds and _salinas_.

It is also as undeniable that, if the supply to a body of water be greater than its evaporation, the quantity of the water must be increased until the receptacle—whether pool, pond, or lake—overflows and obtains an outlet.

[Illustration: UBUJWÉ AND UGUHA HEAD-DRESS.]

[Illustration: UGUHA HEAD-DRESS.]

In the instance of the Tanganika we have a fresh-water lake, which—according to the evidence of native Arabs and the observation of several travellers—is steadily rising. We have also seen in the Lukuga the first symptoms of that overflowing which must come. At present there are only a few inches of mud-banks and a frail barrier of papyrus and reeds to interpose between the waters of the lake and its destiny, which it is now, year by year, steadily approaching. When the Tanganika has risen 3 feet higher, there will be no surf at the mouth of the Lukuga, no sill of sand, no oozing mud-banks, no rush-covered old river course, but the accumulated waters of over a hundred rivers will sweep through the ancient gap with the force of a cataclysm, bearing away on its flood all the deposits of organic débris at present in the Lukuga Creek, down the steep incline to swell the tribute due to the mighty Livingstone.

_July 21._—On the 21st of July we sailed from the mouth of the future outlet Lukuga by Cape Kahangwa, to the Arab crossing-place near Kasengé Island.

The Waguha, along whose country we had voyaged south since leaving Tembwé, are an unusually ceremonious people. They are the first specimens of those nations among whom we are destined to travel in our exploration of the western regions.

[Illustration: A WOMAN OF UGUHA.]

The art of the coiffeur is better known here than in any portion of Africa east of Lake Tanganika. The “waterfall” and “back-hair” styles are superb, and the constructions are fastened with carved wooden or iron pins. Full dress includes a semicircle of finely plaited hair over the forehead painted red, ears well ochred, the rest of the hair drawn up taut at the back of the head, overlaid and secured by a cross-shaped flat board, or with a skeleton crown of iron; the head is then covered with a neatly tasselled and plaited grass cloth, like a lady’s breakfast-cap, to protect it from dust. In order to protect such an elaborate construction from being disordered, they carry a small head- rest of wood stuck in the girdle.

[Illustration: UHYEYA HEAD-DRESS.]

Their mode of salutation is as follows:—

A man appears before a party seated: he bends, takes up a handful of earth or sand with his right hand, and throws a little into his left—the left hand rubs the sand or earth over the right elbow and the right side of the stomach, while the right hand performs the same operation for the left parts of the body, the mouth meanwhile uttering rapidly words of salutation. To his inferiors, however, the new-comer slaps his hand several times, and after each slap lightly taps the region of his heart.

Kasengé Island is a small island with a grassy cone rising from its centre. It is well-cultivated, and grows papaws, pomegranates, lemons, and sweet limes, having been favoured for a long period by Arabs, when their intercourse with the western regions was but beginning.

Between the lately severed promontory of Katenga, in Goma, which is now a large island, and Mtowa, the southern end of the bay, there is quite a cluster of islets, of which the largest are Kirindi, Kivizi, and Kavala.

When we have passed the northern point of Katenga Island we behold the Goma mountains in an apparently unbroken range of vast height and excessive steepness, and lengths of steep and cliffy slope. But as we sail on to the northward, we observe that from Katenga we saw only the profile or the shoulders of great lofty spurs. Behind almost all of these are beautiful secluded inlets and bays, overshadowed by black- bearded mountains, which give birth to myriads of clear crystal streams. Deep chasms in their huge fronts are filled with forests of enormous trees, out of which the famous Goma canoes are cut. Through every gap in the range roars and tumbles a clear cold stream, and piled up behind are the loftiest alps of Goma. The eye cannot fail to be struck with the contrast between the serene blue of the sky, the gloom of the chasm, and the dark tops of the tree-crested ranges. The margins of these calm havens are lined with green water-cane and eschinomenæ, to which hundreds of yellow-breasted birds have suspended their nests, where the industrious little creatures may be seen in flocks together, clinging belly upward, or flying up and down, ever chirping their wheedling, persuasive song. On a firm bough extended over the wave sits the glossy and sleek diver, contented, sated with his finny prey; or, perched upon the tall branch of some towering sycamore or teak, may be seen the white-collared fish-eagle, uttering at intervals his weird shrill call to his mate—a despairing, wailing cry. Presently, from some distant tree, at a commanding height, is heard the response, in the same doleful strain.

But from Katenga, as far as the Bald Mount, near Mugolwé, the crests of the ridges are tawny and treeless. From Tanga to Mdanga Cape, gaps and chasms, inlets and bays, like those above described are numerous, and between Kabogo River and Missossi Mount there is a bay with five separate streams, descending from heights of 2000 feet in long silvery threads to the lake. The mountains seem to be dissolving in tears, for through every ravine or cleft or gap, chasm or rift, streams roll with impetuous course to the lake. Wherever foothold is obtained on a square- browed hill, terrace, or slope, cultivated fields and villages are seen, while on either side of them the cliffs drop sheer to profound depths.

The topmost height of Mount Missossi is about 3000 feet above the lake. As the lake is very wide between Goma and Ujiji—about forty miles—the waves rise very suddenly and drive in long billowy ridges against the massive and firm base of the mount, and when the south-easters prevail, the gale has command of sixty miles of clear water from Kabogo Cape. Navigation in canoes, while the wind is rising, is very dangerous.

We left Kabogo River’s safe haven about 7 P.M., and at nine were pulling by Missossi Mount, exposed to a rising gale of great power about half a mile off a lee shore. To avoid being swept on the rocks, which were all afoam, we had to row direct eastward, and to handle both boat and canoe very delicately to avoid foundering. For two hours we laboured hard to get a mile to windward, and then, hoisting sail, we flew northward, just grazing the dreadful rocks of Mdanga Cape.

Nature, as already seen, has been in most frantic moods along the western coast of the Tanganika, but in Goma, where she has been most wanton, she has veiled herself with a graceful luxuriance of vegetation. Where the mountains are steepest and highest, and where their springs have channelled deepest, there the pillared mvulé and meofu flourish most and attain their greatest height, and in loving fellowship they spread themselves up opposing slopes and follow the course of the stream in broad belts one either side down to the edge of the lake. Underneath their umbrageous foliage grows a tropical density of bush and plant, meshed and tangled, and of such variety that to class or specify them would require the labour and lifetime of an accomplished botanist.

As we look towards the lofty heights of Northern Goma we observe that they have a grassy pastoral aspect. We turn our eyes south to catch a farewell glimpse of those refreshing views which we had admired, and we see that distance has already transformed them into a long blue hazy outline.

We sailed all day within a stone’s-throw of the shore of Goma, and in the evening put in at Kaganza, just north of Kiringi Point.

_July 25._—On the 25th, on leaving Kaganza, we bade farewell to Goma, whose bare majestic front, as we continued north, was terminated by the low rounded hills of Kavunweh, and then, steering north-east, we skirted a low grassy land whose highest ridge was only 200 feet above the lake. This is the isthmus which connects the promontory of Ubwari and Karamba with the mainland. It is seven miles across to the gulf which separates Ubwari promontory from Ubembé and Usansi.

Burton describes Ubwari thus:—

“It is the only island near the centre of the Tanganika, a long narrow lump of rock, twenty to twenty-five geographical miles long, by four or five of extreme breadth.”

Livingstone calls it in his ‘Last Journals’ the islet Mozima, and in ‘How I Found Livingstone’ I called it the island Muzimu.

The end of the isthmus is distinguished by two or three palms, which served us as a landmark when we had voyaged round into the gulf of the western side. It is also indented with two or three deep bays.

Near Karamba Cape, south latitude 4° 29′, the land again rises into a ridge about 1500 feet above the lake, and runs north from the southern cape to Panza Point, a distance of twenty-seven miles. Some very fine mountain scenes are presented here also, but after stupendous Goma they appear almost tame and commonplace.

Near the little round island of Muzimu, or the Spirit, we made a very comfortable camp near a fine gravel beach. The photograph of the Spirit Island given opposite suffices for description.

The Wabwari are by no means a handsome race: nor indeed are the Wavira, Wagoma, or Wabembé (cannibals); but they are all industrious tribes, and the Wabwari, though somewhat ready to take offence, are very much liked by all. They cultivate an enormous quantity of cassava, or manioc, and at this season the flat rocks were strewn with the sliced root. Dried whitebait is another article of commerce, and bags of millet are exchanged with the Warundi on the other side for palm-oil and butter, and with the Wajiji for cloth and beads.

_July 27._—On the 27th we rounded Panza Point, and skirted the much- indented western side of Ubwari, until we reached the extreme southern reach of Burton Gulf.[3] At evening we camped in a tiny creek, near a grassy ridge, undisturbed. In the morning I ascended the ridge, and took bearings of Missossi Mount, Kiringi Cape, Karamba Cape, and by aid of the palms on the isthmus was able to identify the position. We rested until noon, and obtained south latitude 4° 22′. As Panza Point, the north end of Ubwari, is in south latitude 4° 2′, the length of Burton Gulf is twenty miles, by from five to seven miles in width.

Footnote 3:

So named after Captain Richard Francis Burton, the commander of the Burton and Speke Expedition, which first discovered Lake Tanganika.

[Illustration: THE SPIRIT ISLAND.]

Then coasting along the south end of Burton Gulf, we came to Masansi, which begins on the west side, and near each large village lowered our sail and inquired the names of the various rivers, villages, points, and countries. On coming near a village on the west bank of the Kasansagara river, we were forewarned of a rude reception. Approaching nearer, we were warned away by the Wabembé, who are most inimical to strangers. Wishing to test how far this hostile spirit would proceed, we continued to advance upon the shore. From wild gesture, such as striking the ground with their spears, beating the water, and frantic hopping up and down, they took to throwing stones of such large size as might well be termed dangerous missiles. Motioning a halt, we calmly surveyed the natives, watched the rocks fly through the air, and making deep pits in the water, as though we were simply looking on at an entertainment specially got up for our amusement. Not a word, gesture, or movement on our part indicated either resentment or pleasure, until the natives ceased their furious demonstrations. Para was then told to inform them that we would have nothing to say to such wild people, who at sight of strangers showed such foolish fury.

We turned away without another word, resumed our journey, and in an hour were abreast of Kiunyu, the village of the chief Mahonga. We spoke to them: they mocked us. We asked them if they would sell us some grain, but they replied that they were not our slaves, and that they had not sowed the land with grain to sell it to us. We pulled away from them without another word. The silly people cried out that we were running away, and at once launched about a dozen canoes and followed us. Encouraged by the infuriates and mockers on the shore, as also by our pacific behaviour, they became excited to a dangerous state, and gesticulated with their arrows and spears. Owing to the ferocious spirit of the people, we had to seek a camp among the reeds and papyrus in the delta of the Mtambara river, where, though troubled with mosquitoes, we slept undisturbed by the insensate ferocity of the Wabembé cannibals.

_July 28._—On the 28th we skirted the low land which lies at the foot of the western mountains, and by noon had arrived at the little cove in Masansi, near the Rubumba or the Luvumba river, at which Livingstone and I terminated our exploration of the northern shores of Lake Tanganika in 1871. I had thus circumnavigated Lake Tanganika from Ujiji up the eastern coast, along the northern head, and down the western coast as far as Rubumba river in 1871, and in June-July 1876 had sailed south from Ujiji along the eastern coast to the extreme south end of the lake, round each inlet of the south, and up the western coast to Panza Point, in Ubwari, round the shores of Burton Gulf, and to Rubumba river. The north end of the lake was located by Livingstone in south latitude 3° 18′; the extreme south end I discovered to be in south latitude 8° 47′, which gives it a length of 329 geographical miles. Its breadth varies from ten to forty-five miles, averaging about twenty-eight miles, and its superficial area covers a space of 9240 square miles.

_July 29._—On the 29th we crossed over from our haven near Muzimu Island, on the east side of Ubwari, to Kioga, in Urundi, where we were welcomed by our old friend Kinoza, the chief.

In mid-lake, I sounded, using a 3½-lb. sounding-lead with 1280 feet of cord, and found no bottom. I devoted an hour to this work, and tried a second time a mile nearer the Urundi coast, with the same results—no bottom. The strain at such a great depth on the whip-cord was enormous, but we met with no accident.

_July 31._—On the 31st we arrived at Ujiji, after an absence of fifty- one days, during which time we had sailed without disaster or illness a distance of over 810 miles. The entire coast line of the Tanganika is about 930 miles.