I returned home with the conviction that the Tanganyika is a still lake. This view, however, appeared a strange hydrographical puzzle to geographers, who were not slow to combat it. Messrs Vaux and Galton, and my kind friend Mr Findlay, who has never ceased to impress the public with what he holds to be the true state of the case, doubted that an immense reservoir 250 miles long, situated at a considerable altitude in the African zone of almost constant rain, whose potable waters are free of saline substances washed down by its tributaries from the area of drainage, and which shows no marks of great accession of level, can maintain these conditions without efflux. The most natural explanation was to make the Marungu, Luapula, or Runangwa river, at the southern extremity of the Tanganyika, act outlet, and drain it to the Nyassa or Kilwa Lake, bearing S. 55° East, and distant 340 to 350 miles. The universal testimony of the natives to its being an influent formed in the mind of my companion (Journal, p. 90) six years afterwards ‘the most conclusive argument that it does run out of the lake.’ It did not appear equally conclusive to others.

The absence of all connection, however, between the Tanganyika and the Nyassa Lakes was proved by Dr Livingstone’s second expedition and by the excellent paper ‘on the probable ultimate sources of the Nile’ (Mr Alexander Geo. Findlay, F.R.G.S., read June 3, 1867). The latter showed that no considerable stream draining an area of at least 3000 square British miles, or a country as large as England and France combined, enters Nyassa from the north. Since that time Dr Livingstone has placed (Letters to Dr Kirk, July 8, 1868, and to the Earl of Clarendon, July, 1868) the Nile sources between S. lat. 10° and 12°, north of the great Serra Muxinga of the Portuguese travellers Lacerda, Monteiro, and Gamitto, nearly in the position assigned to them by Ptolemy uncorrected for latitude.[[55]] About 400 miles south of the southernmost extremity of the Nyanza or Northern Lake, he finds ‘not one source, but upwards of 20 of them,’ and he is under the impression that he had stood on the water-shed between the Zambeze and either the Congo or the Nile. Mr Keith Johnston jun.’s excellent paper[[56]] shows that the Serra Muxinga, of which more presently, may represent that portion of the Rocky Mountains which send forth the Missouri, the Columbia, and the Colorado. And he apparently would drain the northern fall to the Congo river, whereas in the Mittheilungen it takes the direction of the Albert Nyassa, and the labours of Capt. George, R.N., would throw the water into the Tanganyika.

Thus the theory of the southern effluent lost favour, and that which made the Rusizi a northern influent soon shared the same fate. In 1863 Capt. Speke converted it into a lake or a ‘broad,’ of which he had heard the year before, lying between the Tanganyika and the Luta Nzige, Mwutan or Albert Nyassa. Presently Sir Samuel Baker (1864) caused the southern extremity of the Luta Nzige, which he placed 2200 feet above sea-level, to over-lap the Rusizi. ‘I therefore claim,’ concludes Mr Findlay, ‘for Lake Tanganyika the honour of being the Southernmost Reservoir of the Nile until some more positive evidence, by actual observation, shall otherwise determine it.’

To this view the geographical public offered two objections. The first was that the northern end of the Tanganyika is encircled by the ‘concave of the Mountains of the Moon.’ This was easily removed, as the reader of these pages will see, by a collation of the several maps forwarded by the Expedition from the interior. The first, bearing date May, 1858, was sent from Kazeh on July 2, 1858: it showed the results of our discovery (in February, 1858) and of the information supplied to me by narratio obliqua through the Arabs and Africans of Unyanyembe. Having no theory to support, it laid down, what we saw or thought we saw, an open longitudinal valley running northwards from the Tanganyika Lake. But that which my companion brought home in June, 1859, bore signs of great change, especially in a confused mass of mountains completely investing the northern third of the long narrow crevasse: this by degrees resolved itself into a huge horseshoe, which was incontinently dubbed the ‘Mountains of the Moon, about 6000 feet.’ In his second expedition (Journal, p. 263) Capt. Speke declares that the range had been laid down ‘solely on scientific geographical reasons,’ in fact, out of the depths of his self-consciousness, and he supplemented it with a Lake Rusizi. I saw it growing up under his hands, as copy followed copy: I repeatedly objected to it, yet it managed to deform the maps of Central Africa for years afterwards. It threw us once more back into the romantic geography of the Arabs, who wove into one line Jebel Kumri, and transferred north of the equator the scattered ranges which Ptolemy (iv. 9) disposed at the antarctic end of his habitable Africa. These, going from east to west, are represented by Barditon Oros (S. lat. 16°) Meskhe or Ineskki, the Region of Agysimba (S. lat. 13°), Xipha or Ziphar (S. lat. 8° 20′ 5″), Daukhis Oros (S. lat. 13°), and Ion, the mountain of the Hesperian Æthiopians (S. lat. 8° 20′ 5″).

The second objection was the elevation of the Tanganyika Lake. Its low level in the great central plateau proved, however, to be a mere mistake: only one observation was made, and that gave but 1844 feet above the sea. But presently Mr Findlay found a pencil memorandum by Capt. Speke, showing that when he again reached the coast our thermometer, a common bath instrument, used because all the others had been broken, boiled at 214° (F.) instead of 212° (F.). Moreover, the observations of Sir Samuel Baker, carefully compared with those of the second expedition, decisively proved that 1000 feet must be added, placing the Tanganyika and the Nyanza on nearly the same level. Again, Dr Livingstone reports from Bangweolo (July, 1868) of the Liemba Lake, that he would have set it down as an arm of the Tanganyika, but that its surface is 2800 feet above sea-level, ‘while Speke makes it 1844 only.’ Finally, the great African traveller, who has now been long resident in the regions west of the Tanganyika Lake, always writes of it as if he considered the connection between it and the Luta Nzige established. Thus the altitude of Lake Tanganyika was raised to 2800 feet, which would easily carry its waters to the Nile. ‘It may appear strange,’ as Mr Galton has remarked, ‘that there should be an error of a thousand feet of altitude suspected in the observations of an explorer, but the method of operating in uncivilized countries is quite different from that employed at home.’ Evidently Capt. Speke allowed the altitude of the lake to lie uncorrected for the same reason which made him raise his ‘Lunæ montes.’ This will also answer M. Parthey (June 2, 1864, Royal Academy of Sciences, Berlin).

A few words concerning the Moon Mountain, in which are the Ptolemeian Nile sources. It is placed in S. lat. 12° 30′, which, by applying the reduction as before proposed, we should convert to S. lat. 6° 30′, and between East long. 57° and 67°, which, if taken from S. Antonio (Antão), as the late Mr Hogg suggested, would be = E. long. G. 30° to 40.° Its northern slope drains to the lake under the parallels of S. lat. 6° and 7°, and separated by about 8° of meridional distance. In many maps is added a third, or equatorial lake, which may be the Baharingo, or Baringo, and indeed in chap. xvii. (lib. i.) we find a plural form τὰς λίμνας, possibly showing a knowledge of two large and sundry smaller features. The great Unyamwezi Upland, using the name at its fullest extent, is bounded both north and south by huge latitudinal blocks and chains of mountains. The equatorial is the Highland of Karagwah, extending eastward to the Æthiopic Olympus Kilima-njaro. The southern, corresponding with Ptolemy’s parallels, is the great chain and plateau, whose apex is the Serra Muxinga or Muchingwe, named by the explorer Dr de Lacerda Cordelheira Antonina, in honour of his prince. Lying in about S. lat. 12°, this feature, ranging from 3000 to 6000 feet high, may evidently be the divide of the Nile, the Congo, and Zambezean basins; whilst the north-eastern projection feeds with four considerable streams the Lake Liemba, discovered by Dr Livingstone on April 2, 1867, and supposed to connect with the Tanganyika by the River Marungu. The altitude of the Serra was estimated in 1831 by Messrs Monteiro and Gamitto at a Portuguese league (= about 19,700 feet) above sea-level, palpably exaggerated, as in winter (August 10) neither ice nor snow was found upon it. They describe the head as nearly always enveloped in clouds, and as by far the loftiest summit in that part of Africa; the profile rises steeply and abruptly from the table-land, commanding an extensive prospect northward, and the ridge is broken by terrible and dangerous precipices. Snow in this part of the continent may be alluded to by João de Barros, who declares that in the Matouca country, though situated between the equator and the tropic of Capricorn, the natives die of cold. Later Portuguese historians declare the Lupata to be a snowy range, probably referring, not to the gorge of that name, but to the great block with which it is connected. Dr Livingstone represents this, his latest discovery, to cover a space south of the Tanganyika some 350 miles square, dotted with lakes, and traversed on the eastern side by the River Chambeze, which was first mentioned by Dr de Lacerda, and which has hitherto been confounded with the Zambeze. The Greek term ‘Mountain of the Moon’ may, I have already suggested, be derived from ‘Unyamwezi,’ an empire whose position between the Tanganyika and the Nyanza group is laid down in the map of Duarte Lopez (A. D. 1578–1587). The name of this extensive region is still contracted upon the coast to Mwezi, meaning the Moon, and thus we might translate Ptolemy, Mountain of Unyamwezi. Similarly, the ancients derived the Erythrean Sea from the Sea of Edom and of Himyar, both signifying Red: Diascorias was a corruption of Dwipa Sokotra, and, to quote no more, Dr Beke has shown how the Ptolemeian Iabadíou (Java-dwipa) became Barley Island without growing barley.[[57]]

Finally, if we reject Unyamwezi and Muxinga as the original Lunar Mountains, we must seek the latter with Dr Beke in the icy peaks of the Æthiopic Olympus, prolonged to the Highlands of Karagwah.

A longer delay at Uvira than we had intended greatly improved my health: the state of our finances, however, compelled us to set out without delay from Ujiji to Kazeh. The rains had ceased on May 15, and the return (June 11) by a straighter and more southerly road, was far less unpleasant than the up-march. After a short interval for repose, and for recovering his sight and hearing, Capt. Speke volunteered to explore a lake reported to lie north, and known to the Arabs as ‘Ukerewe,’ or Island-land. I had heard of it in Zanzibar Island as a water called ‘Karagoa,’ parallel with and one month west of the Sea of Ujiji. A signal disappointment at the ‘Ziwa’[[58]] of Ugogo, which proved to be a mere pond, made me suspect the informants: yet Snay bin Amir and Musa Mzuri had both visited the mountain regions to its west, and their observations were represented in the sketch map, 1858, which, I repeat, is far less incorrect than the exaggerated growth of 1859.[[59]] I was, however, delighted with the prospect of a month’s leisure for inquiry amongst the intelligent Arabs. It was also necessary to copy out notes, which ill health had left in confusion, and to learn something about the southern as well as the northern regions. Moreover, if truth must be told, I sighed for the

‘Beata Solitudo

Sola Beatitude’