FAUNA OF THE AFRICAN LAKES. 599 



animals from difterent groups which constituted Moore's halo- 

 liumic series, none ai'e accepted as peculiarly marine save the 

 Polyzoon Araclmoidea and the medusa. 



But more than this. The endemic animal forms have been 

 described almost without exception as specialised and not primi- 

 tive types. It is true that Mooi'e regarded the remarkable 

 Gastei'opods as essentially primitive in nature, but this view is 

 not shared by other writers (p. 550). If the halolimnic animals 

 are incieed relict forms, they must have been cut off at some 

 i-emote era — though it need not have been the Jurassic period — 

 and ought in consequence to exhibit pi-imitive I'ather than 

 specialised characteristics. 



This review of the zoological evidence makes it clear that on 

 .such grounds it is impossible to justify the contention that 

 Tanganyika was connected with the sea in Jurassic times, or 

 indeed that a connection with the sea ever existed. It is there- 

 fore necessa.ry to inquire what light may be thrown on the subject 

 by the evidence of geology. 



In his book, Moore maintained that resting on the Archaean 

 granites, gneisses, and schists which appear to constitute the 

 basement rocks of the continent, three types of sedimentary rock 

 are to be recognised. The lowest of these consists of beds of 

 sandstone and shale, which are not only well developed in the 

 neighbourhood of the great lakes, but appear to extend over vast 

 areas of the African interior, including a large part of the Congo 

 basin. Then follow the beds discovered by Drummond north- 

 west of Nyasa, and covering these in turn, white shelly deposits 

 (Pleistocene) laid down by the lakes themselves. Drummond's 

 beds being I'egarded as Tiiassic in age and probably estuarine, 

 Moore considered the gi'eat beds of sandstone and conglomei'ate 

 as evidence of an extensive ocean which at some still earlier 

 period covered a great part of the lake regions of Central Africa 

 (137, p. 65 et seq.). 



It is particularly to this last point that exception is taken by 

 other writers on the geology of these regions. Some regard Drum- 

 mond's beds and the great sandstone series as of like age, and 

 on the evidence of the fossils occurring in the former, regard the 

 whole as beds of the lower Karoo (Trias) — or at least as a forma- 

 tion of corresponding age, deposited under similar conditions. A 

 very recent writer on the stratigraphy of this part of the continent 

 (Behrend) speaks of the unfossiliferous conglomerates, quailzites, 

 and sandstones wiiich are particularly well displayed in the 

 neighbourhood of tlie great lakes, as the "Tanganyika System" 

 (14, p. 52). These beds he distinguishes as of different age from 

 similar rocks occuriing near Nyasa and in parts of the Congo 

 basin, assigning them to an earlier pei'iod — Devonian or even 

 prior to that (14, p. 73 and Taf. iii.). While it may be that the 

 relative age of these different strata is by no means conclusively 

 fixed, these I'ecent investigations show that Mooi-e's lowest 

 series — the " Old Afincan sandstones " as he calls them — really 



