546 Bulletin American Museum of Natural History [Vol. LITT 
icance for them. Only in the extreme south and west is there some 
peculiarity. The extensive Orange River system appears to possess the 
negative feature of lacking Ampullariide, Viviparide, Mutelide, and 
Etheriide. The Melaniide are represented by a few Melanozdes only and 
the Unionide by Unio (subgenus Cafferza). In the extreme west, in 
Upper Guinea, Saulea, Afropomus, and Potamopyrgus inhabit isolated 
Liberian streams and are not found elsewhere. Rhinomelania, Pachy- 
melania, Pseudogibbula, and the very distinct bivalve genera Hg:ria and 
Iphigenia are known only from the western coastal rivers and estuaries. 
Potadoma is practically restricted to the Congo basin and Lower Guinea. 
It is also of interest that the many genera evolved in Lake Tanganyika 
have not spread from that center, though affluents have brought the 
generally distributed Ethiopian genera into the lake. 
The general uniformity of the African fresh-water fauna is best 
accounted for by assuming that frequent and easy migrations have 
been possible between the various river systems. In the case of the Lym- 
neeidee, Planorbide and Spheriide mechanical transportation by ‘water 
birds may suffice to explain their wide distribution. But, it is extremely 
doubtful whether such a purely passive means of dispersal could have 
brought about the same result for the large Melaniidz, Unionidz, Muteli- 
de, and Etheriide, especially for the viviparous species. It seems more 
reasonable to admit that they migrated actively to various parts of the 
continent. 
There is fortunately plenty of geologic evidence to show (a) that 
during the Pleistocene there have been periods of extreme humid condi- 
tions in Africa, when the amount of rainfall, and consequently of super- 
ficial water, was considerably larger than nowadays; (b) that Africa has 
been undergoing a steady desiccation ever since the beginning of the 
modern epoch; and (c) that toward the end of the Pleistocene the limits 
of the hydrographic basins and watersheds were quite different from 
those prevailing at present. More in particular, there existed about that 
time a series of large lakes—quite apart from those that are now found in 
the depressions of the East African rift valleys, some of which may or 
may not have existed then, at least in their present shape. 
We have attempted to represent on Map 14 hydrographic condi- 
tions as they may have existed at some very moist epoch toward the 
close of Pleistocene times. Although such a map is, of course, largely 
athens Ion nO Oued ot io temaas Doik Pee eee (1913), oog, Bund. 
sp pore Cetpaes HO Cer a tnt an Gp at an) fhe akon which we 
existence of recent alluvial deposits—usually with Ereuhiwuter ila oe eee 
over large CONTINUOUS:areas, and next the topographical features of the countr i 
yo y. It could easily b 
shown that the lakes drawn by the above-mentioned authors do not agree with these two series of a 
