514 mi. W. A. CUXNINGTOX 0.\ THE 



occasions and at dill'ei'ent places only vary between 74"8" and 

 74"1". Stappei'S gives the temperature of tlie surface water as 

 oscillating between 25° and 28° C. (77°-82-4° F.) (176). Mar- 

 quardsen, however, quotes exactly 25*0° 0. (77° F.) as a mean 

 of 70 surface readings made by Jacobs. The latter was also 

 able to carry out deep water observations, to which Marquardsen 

 refers as follows : — " Unter 400 m. bis in die grossten Tiefen 

 liegt dann eine fast homotherme Wasserschicht von 23"15° bis 

 23^2° " (73-67°-73-61° F.) (Ill, p. U7). A less extended series of 

 water temperature observations was njade in Nyasa* and also in 

 Victoria Nyanza during the Third Tanganyika Expedition, but 

 the figures are sufficient to show that there are no important 

 difi'erences from the results obtained in Tanganyika.. It is 

 obvious that the amount of water lost by evaporation must be 

 very great in all these cases. 



As fai- as I am aware, thei'e is I'elatively little known con- 

 cerning the nature of the water in these African lakes. They 

 are considered in general terms to be " fresh " f, but, except in 

 certain instances, no analyses have been made, and the nature 

 and quantity of the salts dissolved in the water are unknown. 

 The information which is to hand regarding Tanganyika and 

 Kivu makes it clear that they differ from normal fresh waters, 

 and this difference appears to be reflected (as will be shown in 

 due course) in the nature of the organisms they contain. Tan- 

 ganyika water, while containing a relatively small amount of 

 calcium salts, is iinusually rich in salts of magnesium (176). 

 Lake Kivu is an extreme case, where the salinity is veiy liigh. 

 In this instance, salts of calcium appear to have been almost 

 entirely I'eplaced by considerable amounts of sodium and 

 magnesium salts, the latter being quite excessive in quantity 

 (103). 



It is possible that the water of Tanganyika has somewhat 

 changed in nature within recent years, for most of the early 

 explorers agree that the water was brackish, or at least *' pecu- 

 liar " in taste. If the lather vague statements are to be relied 

 upon, this altei-ation in salinity is probably to be associated with 

 the question of the outflow of the lake. A per-iod during which 

 the lake possessed no outlet would result in an accumulation of 

 salts producing brackish water, while a subsequent period with 

 a regular outflow would diminish the quantity of saline matter 

 in solution. An examination of the available records suggests 

 that something of this kind has taken place, and it thus becomes 

 necessary to consider the whole subject of the outflow of Tan- 

 ganyika. 



This is not the place to discuss in detail all the evidence 



* Certain observations concerning Nj'asa were made bj' Fiilleborn (83, p. 334 

 et seg.). 



■f Grogan, referring to Lalce Edward, says: — "Altliongli llie water was salt, it 

 was drinkable" (91, p. 191). In the same connection Bright remarlis : — "The 

 water of Albert is drinkable, but not good " (55, p, 138). 



