Chapter XXI: Appendix: The Ukara or Ukerewe Lake

A DEDUCTION FROM THE REV. MR WAKEFIELD’S ‘ROUTES.’

The following paper was read out on December 11, 1871, before a meeting of the Royal Geographical Society, Major-General Sir Henry Rawlinson, President of the Society, being in the chair.

Much light had been lately thrown upon the dark points of Eastern Africa, especially those which gather round the much-vexed Ethiopic Olympus, Kilima-njaro, by the labours of the Rev. Thomas Wakefield. This gentleman, we are informed (Preface by Mr Samuel S. Barton, General Mission Secretary, to ‘Footprints in Eastern Africa.’ London: Reed, 1866), was one of four missionaries sent out to Mombasah in 1861 by the United Methodist Free Churches under charge of Dr Krapf, who first established the now world-renowned ‘Mombas Mission.’ After a residence of five years he published the interesting series of ‘Notes on a Visit to the Southern Gálas’ above alluded to; and in 1866-7, accompanied by the Rev. C. New, he marched from Mombasah to Upokomo, on the Dana river. He is therefore an African traveller of some experience; and as he has evidently mastered the Kisawahili tongue, he is unusually well qualified to supervise and to correct the ‘Routes of Native Caravans from the Coast to the Interior of Eastern Africa, chiefly from information given by Sádi bin Ahedi, a native of a district near Gázi (Gasi Bandar?) in Udogo, a little north of Zanzibar.’ Especially advocated by my old and tried friend Mr Alexander Findlay, F.R.G.S., this valuable paper was published in the Journal of the Royal Geographical Society (pp. 303-339, No. xi. Vol. xl. of 1870), and I felt somewhat surprised that the extent of its importance has not attracted more attention in England.

I will consider this addition to our scanty knowledge of one of the most interesting regions in tropical Africa under two heads—the philological and the geographical.

Firstly, Mr Wakefield, like Doctors Livingstone and Kirk, all being practical linguists, invariably uses the system of Zangian orthography, adopted by the ‘Mombas Mission,’ and by myself since 1859. He speaks, for instance, of Kilima-njaro and Unyamwezi, not the Monomoezi of Mr Cooley or ‘the authentic word Mueni Muezi,’ translated landlord, or petty chief country (p. 11, The Memoir on the Lake Regions of East Africa reviewed, &c. &c., by W. D. Cooley. London: Stanford, 1864). We find in Mr Wakefield’s Notes (p. 316) ‘_Líma_, a term denoting extraordinary size—Mlíma being the general term for mountain,’ whilst (p. 321) Ki-Mrima is justly applied to the dialect spoken on the Mrima or mainland facing Zanzibar Island. We read also (323) ‘Mtanganyíko,[126] Kisawáhílí, meaning the place of _mingling_ or _mixture_ (rendezvous).’ This is precisely the meaning attached by me to the Lake’s name, yet I was assured in the Memoir above alluded to (p. 7) that nothing can be ‘more ridiculous’ than my explanation of Tanganyika. Even in philological details of the Kisawahili dialect Mr Wakefield agrees with me. He writes, for instance, Udogo (Notes, p. 313), Ugala (Footprints, p. 67, 68), and Ulangulo (Ibid. p. 63). I was assured in the Memoir (p. 9) that U is prefixed to the names of countries only by Dr Krapf and Captain Burton—this, too, after I had for years been talking of Europe as Uzungu, literally, Land of white men. Mr Wakefield speaks of Wasămbá, of Wasawahili (or Wa-Sawahili), and of Wanyamwezi, thus sanctioning the use of Wamrima, continental men, and Wakilima, hill-men. He adopts Kisawahili, Kikwavi, Kimasai, and so forth, prefixing an adjectival particle ‘Ki’ to the root, and denoting chiefly dialect, yet I was assured by Mr Cooley (Memoir, p. 9) that ‘Ki’ has never an adjectival form. I may now invite the author of Inner Africa Laid Open to revise the verdict (Memoir, p. 7) which pronounces me ‘totally ignorant’ of the language of which I affect to be master.

It may be deemed trivial to dwell upon these philological minutiæ, but, firstly, nothing is unimportant when it affects the accuracy of a traveller, especially of an explorer, in the smallest matters of detail. Secondly, without an exact nomenclature all topographical literature must be imperfect and of scant value. And, finally, as Mr Cooley and I have been differing upon these points for the last ten years, it is well that the portion of the public which takes an interest in the subject should see who is right and who is not. I have no personal feeling in the matter; and if the ‘Geographer of N’yassi’ will bring, as I have done, independent testimony to bear upon the points in question, and not evolve his learning out of the depths of his self-consciousness, I am at all times ready and willing to own myself wrong.

Far more important and generally interesting, however, is the geographical knowledge brought home or rather confirmed to us by Mr Wakefield’s ‘Routes.’ We now know that the block whose apices are Mounts Kilima-njaro and Kenia (alias Doenyo Ebor, Mont Blanc) to be a great upland, bounded on the South by the Panga-ni river in S. lat. 5°, and on the N. West by a lacustrine region in S. lat. 6°; whilst it may possibly anastomose to the North, as was suggested by my friend Dr Beke, with the Highlands of Harar and of Moslem Abyssinia, lying upon the same meridian. The breadth between N. West and S. East will be included between East long. (G.) 37° and 39°. Assuming, therefore, roughly the bounding lines to measure 240 by 120 direct geographical miles—we obtain a superficies of 28,800 square geographical miles, more than a fourth of the area assigned to the British Islands. We can now safely believe, with Dr Krapf and Mr Rebmann, the explorers, that the block is a high volcanic country, separating the watershed of the Nilotic Basin from that of the Indian Ocean; sending off, like the Highlands of Abyssinia, its own tributary or tributaries to the White river, and corresponding with the Camarones or Theōn Ochēma in West Africa (N. lat. 4°); that it is a land of high plains and thickly forested hills, rising to summits capped, not with delomite and quartz, but with glaciers and eternal snows; and that it abounds in the lakes and swamps, sweet and salt, necessary to feed the inland ‘smoke mountains’ or volcanoes,[127] whose existence before appeared so problematical. And now the two mighty summits, Kilima-njaro, explored by the late lamented Baron von der Decken and Doenyo Ebor, reported to Dr Krapf under the alias Kenia or Kirenia, and unexpectedly confirmed by fresh evidence, have obtained local habitations as well as names.

But the interest of Mr Wakefield’s Routes culminates in the fact that they show even to a certainty the existence of a lake before unknown, and they lead to the conclusion that the area of 29,900 square geographical miles, assigned to the so-called Victoria Nyanza, contains at least four and probably a greater number of separate waters; that it is, in fact, not a Lake, but a Lake Region.

Mr Keith Johnston observes (p. 333), ‘It is remarkable that not one single name of a district, people, or place (with the exception of that of the Wamasai, a general name for the people of the white region west of the Lake)[128] given in these new routes has any such remote resemblance to names reported by Speke and Burton as to warrant any identification with any one of them.’ The reason will presently appear in the fact that we are speaking of different waters. The annotator further observes (p. 333) that ‘the arguments which Captain Burton used in recommending a division of the Nyanza had not a sufficient basis of proof to give them moment, as is shown by the acceptance of the Lake as one sheet by the whole geographical world.’

The mapper will readily understand that it is much more sightly and convenient to have a basin neatly outlined, and margined sky-blue, like the Damascus swamps, than to split it into fragments as I did. A volume published by the late Mr Macqueen and myself (The Nile Basin. London: Tinsleys, 1864), offered a sketch of what was actually seen by the second expedition, and the aspect of disjecta membra was not inviting. Afterwards, however (p. 334), Mr Johnston remarks, ‘Captain Burton’s recommendation would seem to receive some slight support from the new information obtained by Mr Wakefield.’ To this I would add that his language might have been less hesitating, as these ‘Routes,’ so important to the geography of Eastern Africa, at once establish the existence of two lakes wholly independent of the so-called Victoria Nyanza.

The first is that which we named from hearsay Bahari ya Ngo, contracted to Bahari Ngo, sea or water of Ngo(-land). In the atlas of Mercator (Gerhard Kauffmann) we find it written Barcena for Barenca or Barenga. Mr Wakefield prefers (324) Baringo, meaning a ‘canoe,’ and ‘possibly so called from its form.’ I shall follow his example, at the same time observing that African negroes rarely adopt such general and comprehensive views of larger features or venture upon such comparisons unless they can command a birds-eye glance at the prospect. Route No. 5, from ‘Lake Nyanza’ to Lake Baringo, conclusively proves that the latter is not ‘a sort-of backwater’ connected with the former ‘by a strait, at the same distance from the East of Ripon Falls as the Katenga river is to the West.’ Nor is it a ‘vast salt marsh’ without effluent: the saline water has evidently been confused with the lately reported Lake Naïrvasha or Balibali lying S. West of Doenyo Ebor. Native description supplies the Baringo with the Northern Nyarus—the southern effluent of the same name being clearly an influent. Nyarus thus corresponds with the old Thumbiri, Tubirih, and Meri, afterwards called Achua, Usua, and Asua, words probably corrupted from Nyarus.

The map of 1864, printed by Mr now Sir Samuel Baker in the Proceedings of the Royal Geographical Society, affirms the Asua to have been a dry channel 150 yards wide when he crossed it, in Jan. 9, 1864, but rolling 15 feet deep in the wet season. He can hardly be speaking of the drain from the Baringo Lake, which must be large and perennial, and which therefore must be sought farther north, unless it anastomoses with some other stream. M. d’Arnaud, the French engineer sent in 1840-1841 by Mohammed Ali Pasha to explore the Upper Nile, reported (Journal Royal Geographical Society, vol. xviii. p. 73) that about 30 leagues south of where the expedition was stopped by shallow water in N. lat. 4° 42′ 42″, and therefore in N. lat. 3° 12′, the several branches unite, the chief one flowing from the east.

The Baringo Palus must act reservoir to the whole N. Western declivities of Doenyo Ebor, whose snows have given it a name. Ptolemy (iv. 8) distinctly mentions the χίονας, or (melted) snows which feed the Nile; and though he places them in S. lat. 12° 30′, he is correct as to the existence of snowy feeders. Some years ago a Swiss traveller drew my attention to the fact that glacier-water would explain the term White river as opposed to Blue river. The quantity of melted snow or glacier- water which finds its way to the true Nile may be comparatively inconsiderable, but that little may perhaps modify the colourless complexion of rain-water when its suspended matter has been deposited, and distinguish it from the pure azure of a stream issuing from a Lake Geneva. In 1857 Captain Speke, an experienced Himalayan, easily detected, when drinking from the Pangar-ni or Rúfu river, the rough flavour of snow water.

More important, however, than Baringo is the new Lake announced to us by Mr Wakefield’s African Pandit, Sádi bin Ahedi. The latter ignoring Nyanza, calls it Nyanja, possibly a dialectic variety, and therefore a difference neither to be dwelt upon too much nor wholly to be neglected. Of greater value is the name Bahari ya Pili, or Second Sea, not called so, we are expressly informed, because inland of the First Sea—Indian Ocean—but evidently because leading to a neighbouring water on the west. Most suggestive of all, and therefore adopted by me, is the term ‘Bahari ya Ukara,’ or Sea of Ukara, the latter being the region on the Eastern shore. Here we detect the true origin of the ancient Garava, and of Captain Speke’s Ukewere, which he applied to a peninsula projecting from the Eastern shore, and which the Wanyamwezi, translating ‘island water,’ gave to the Oriental portion of the so-called Victoria Nyanza.

Respecting the length of the Ukara Lake, Sádi was informed that it could be crossed by canoes in 6 full days, paddling from sunrise to sunset; but if the men went on night and day, the voyage is to be accomplished in three days. Now the native craft used upon these dangerous plateau- waters never dare to cross them: the voyager may rush over the narrow parts of the Tanganyika Lake, but of course he would not attempt the physical impossibility of navigating without chart or compass beyond sight of land. It is impossible to believe in a canoe-cruise of 6 days across the lake: it is evident that a coasting-cruise is meant. The total of hours, allowing the day to be 12, and without halts, would be 72. Upon the Tanganyika I estimated the rate at a little more than 2 knots an hour. Thus, in round numbers, we have 145 miles, which probably require reduction: an estimate of the mean amount of error distributed on the whole of Mr Wakefield’s ‘Routes’ gives, according to the annotator, an exaggeration of 1.24 : 1.0; and of course, when estimating the length of these distant and dangerous navigations, exaggeration would be excessive. We may, therefore, fairly assume the semi- circumference of the Ukara Lake at 120 miles, and the total circumference at 240. Sádi, we are told (p. 309), made Bahari-ni on the Eastern shore the terminus of his long journey from Tanga Bandar to the ‘Lake Nyanza’ (Nyanja?). Let us protract the full 145 miles as the exceptional rate of 3 knots an hour upon Captain Speke’s last map, without allowing anything for the sinuosities of the coast, and the end would strike the entrance of ‘Jordan Nullah’ off the ‘Bengal Archipelago,’ about half the width of the so-called ‘Victoria Nyanza.’

As regards the breadth of the Ukara Lake, we read (p. 310), ‘Standing on the Eastern shore, Sádi said he could descry nothing of land in a western direction, except the very faint outline of a mountain summit, far, far away on the horizon.’ This passage is again suggestive. The sandy and level Eastern shore of the Nyanja (i. e. water) or Ukara Lake about Bahari-ni, whence Sádi sighted, it is probably in E. long. (G.) 35° 15′. The easternmost, that is, the nearest, point of the Karagwah, or, as Captain Speke writes it, the Karague Highlands, is in E. long. (G.) 32° 30′. Thus the minimum width is 165 miles, whilst man’s vision under such circumstances would hardly cover a dozen. Here, again, we have room for a First as well as for a Second Sea. Mr Johnston suggests that the mountain-summit in question might be an island rising high in the midst of the Lake; but, he adds, such a feature could not well have been missed entirely by Captain Speke. Here I join issue with him for reasons which can be deduced from these pages—my companion and second in command never saw or heard of the Ukara Lake. But it is highly improbable that those who could tell Sádi the number of days required to cross or to coast along the Lake would not have known whether the summit was that of a mountain on terra firma or of a lacustrine islet. The latter feature is not unfamiliar to Mr Wakefield’s informant: he does not fail to mention (p. 324) the small conical hill in the southern waters of the Baringo Lake.

When Sádi declared that ‘he travelled 60 days (marches?) along the shore without perceiving any signs of its termination,’ he evidently spoke wildly, as Africans will. His assertion that the natives with whom he conversed were unable to give him any information about its northern or southern limit, simply means that in this part of the African interior neither caravans nor individuals trust themselves in strange lands, especially with the prospect of meeting such dangerous plunderers as the Wasuku. Similarly a ‘two months’ journey’ and ‘going to Egypt,’ asserted by ‘all authorities without exception, African and Arab,’ signify nothing but the total ignorance of the informant concerning the country a few leagues beyond his home. A lake 120 miles in length, that is to say, even a little smaller than the Baringo is supposed to be, will amply satisfy all requirements in this matter.

Finally, we have Sádi’s report that 8 or 9 years ago (before 1867?) the Ukara Lake was navigated by Europeans. Certain very white men, we are told, who bought only short ivories (Serivellos), refusing long tusks, and who purchased large quantities of eggs—Africans have learnt by some curious process to connect Europeans with oöphagy—came up in a large vessel, carrying three masts and another in front (bowsprit?), with many white cloths (sails). The event took place only a month and a half before he reached the Lake, and it is described with an exactness of detail which seems to vouch for its truth. If this be a fact, it is clear that the Nyanja cannot be Captain Speke’s Nyanza, and that the visitors could not have made it viâ his ‘White Nile,’ with its immense and manifold obstructions. But it may be that of which he heard (Journal, p. 333) from the ‘Kidi officers,’ who reported a high mountain to rise behind the Asua (Nyarus?) river, and a lake navigated by the Gallas in very large vessels. We now understand why the ‘King’ Mtesa (Ibid. p. 294) offered to send the traveller home (to Zanzibar) in one month by a frequented route, doubtless through the Wamasai and other tribes living between the Nyanja and the Nyanza. Thus Irungu of Uganda (Ibid. p. 187) expressed his surprise that Captain Speke had come all the way round to that country, when he could have taken the short and safe direct route up the mid-length of his own lake—viâ Umasai and Usoga, by which an Arab caravan had travelled.

The Ukara Lake will be found laid down (A.D. 1712) in the Africa of John Senex, F.R.S. (quoted by the late Mr John Hogg, F.R.S., ‘On some old maps of Africa, in which the central equatorial lakes are laid down nearly in their true positions’). It is evidently the Garava of Mercator (A.D. 1623), whose atlas supplies it with a northern effluent draining to the Nile. The ‘Couir’ of D’Anville’s folio atlas (A.D. 1749), and placed where the Lake No and the Bahr el Ghazal actually exist, may be a confusion with the equatorial Lake Kura Kawar, given by Ja’afar Mohammed bin Musa el Khwarazmi (A.D. 833) in the Rasm el Arzi, published in Lelawel’s Géographie du Moyen Age (Brussels, 1850), and, like Garava, both may be derived from Ukara.

The third water is evidently the Nyanza of which I first heard at Kazah of Unyamwezi, whence Captain Speke was despatched on a reconnoitre between July 29 and August 25, 1858. After returning, he reported that this lake being nearly flush with the surface of the level country to the south, bears signs of overflowing for some 13 miles during the rains. The second expedition found no traces of flood on the marshy lands to the North and the N. West of the so-called ‘Victoria Nyanza.’ This fact, combined with a difference of level amounting to 400 feet in the surface of the two waters, speaks for itself. We are justified in suspecting a fourth lake, or broadening of a river along whose banks Captain Speke and Grant travelled northward to Uganda, and there must be more than one, if all his effluents be correctly laid down.

Briefly to resume: Mr Wakefield’s very valuable ‘Routes’ teach us these novelties:

1. That the Baringo is a Lake distinct from the so-called Victoria Nyanza; that it has a northern effluent, the Nyarus, and consequently that its waters are sweet.

2. That the Nyanja, Ukara, Ukerewe, Garava or Bahari ya Pili, is a long narrow formation like the Baringo, perhaps 20 miles broad, with 240 of circumference, and possibly drained to the White River or true Nile by a navigable channel.

And I have long ago come to the following conclusions:

1. That the 30,000 square miles representing upon our maps the area of the so-called Victoria Nyanza represent not a lake, but a Lake Region.

2. That the Victoria Nyanza Proper is a water—possibly a swamp—distinct from the two mentioned above, flooding the lands to the south, showing no sign of depth and swelling during the hot season of the Nile, and vice versâ.

3. That the Northern and N. Western portions of the so-called ‘Victoria Nyanza’ must be divided into sundry independent broads or lakes, one of them marshy, reed-margined, and probably shallow, in order to account for three large effluents within a little more than 60 miles.

I cannot finish these lines without expressing my gratitude to Mr Wakefield for the interesting information with which he supplied us. He has returned to his labours at Mombasah, amongst the Wasawahili and the Wanyika, and as he has, I am assured by my friend Captain George, R. N., qualified himself to take astronomical observations, we may rest assured that with his aid the ‘Mombas Mission’ will lose nothing of its well-won fame for linguistic study and African exploration.

END OF VOL. I.

---------------------

JOHN CHILDS AND SON, PRINTERS.

------------------------------------------------------------------------

Transcriber’s Note

The word ‘seaboard’ appears both hyphenated and unhyphenated.

The word ‘Kipiní’ on p. 387 is likely to be an error for ‘Kipini’.

On page 397, the printed Arabic word محکمّه) does not match the Romanization. A reading as محكّمه does.

My thanks to Olive and Manny for providing assistance with the rendering of the Arabic words and phrases.

Errors deemed most likely to be the printer’s have been corrected, and are noted here. The references are to the page and line in the original.

6.30 [‘]And see the twain Added.

45.15 the Gulf of Cambay.[’] Added.

58.6 a subscrip[t]ion for exploring Inserted.

63.25 (Map of the Lake Region of Eastern Africa[)] Added.

126.20 to the [‘]Roman Port’ Inserted.

127.13 ‘Shirazían[’] dynasty Added
(probably).

129.5 (S. lat. 10° 41′ 2”[)] Added.

129.25 [‘]men (des hommes):’ Added.

131.21 Zeramu, and Gogo.[’] Added.

131.35 with other words.[’] Added.

174.7 and in 1856 [i/o]n August 26. Replaced.

178.13 the inevitable ‘seasoning fever[’] Added.

197.21 (Viverra civetta, and V. genetta)[./,] one Replaced.
small,

248.26 and ran a[.] vein of silver Removed.

279.4 after touching at Makdish[ú/u] Replaced.

300.26 It began, ‘Yá Dayyus! yá Mal’ún[’], Added.

387.14 with leaden ‘Kipin[í/i’] (in the plural Replaced.
‘Vipini’)

393.27 for [t]he occasion Restored.

397.26 (محکمّه>) perhaps
محكّمه