550 DR. W. A. CUNNIN6T0N ON THE 



discussion on this topic at the British Association Meeting in 

 1906 (138), Moore expressed less extreme views, while Pelseneer, 

 speaking with a wider knowledge of the anatomical characters 

 of these mollascs, asserted positively that there was no special 

 resemblance between them and marine types and that they were 

 no more primitive than other well-known fresh-water genera 

 (144). Pelseneer added that the halolimnic Gasteropods were all 

 to be regarded as belonging to the Melaniidse (Tiaridae) or families 

 nearly akin, in which case they belong to an essentially fresh- 

 water group. If this be indesd a fact, Moore's whole hypothesis, 

 as far as it I'ested on these molluscan forms, falls to the ground. 

 It no longer becomes necessary to speculate as to how and when 

 Tanganyika received animal types from an ancient sea, it is suffi- 

 cient to invoke prolonged isolation to accou.ut for endemic genera 

 and species — and this is precisely what has been done in other 

 groups. 



It is true that no satisfactory explanation of the marine aspect 

 of these Gasteropod shells is forthcoming, although certain 

 guesses may he hazarded. Without attributing it to a direct 

 community of descent with certain salt-water types, it might be 

 regarded as simply due to convergence. It might be urged that 

 the resemblance is purely accidental, or that it results from the 

 quasi-oceanic conditions which prevail in Tanganyika.. It might 

 be due to the magnesium salts in the water, or, perhaps, to the 

 greater degree of salinity which is believed to have formerly 

 existed. Beyond such suggestions it is impossible to go, but it 

 must not be overlooked that a number of fresh-water shells from 

 other parts of the world exhibit the same characteristically marine 

 aspect, though there may be no other case with so extensive a 

 series of forms. Indeed, Bourguignat and others have not hesi- 

 tated to claim a thalassoid aspect for certain Melaniida^ from 

 Nyasa and even from the River Congo (cf. 48, p. 40 : 87, p. 564). 



A discussion as to the source from which the striking group of 

 thalassoid Gasteropods has been derived might follow here, but is 

 deferred for consideration at a later stage. It may, however, be 

 pointed out that forms possibly akin to those in Tanganyika 

 which have aroused so much interest, formerly occurred in the 

 Balkan Peninsula [cf. Brusina, 58 : 59), from which neighbourhood 

 distribution through the valley of the Jordan and the Great Rift 

 Valle}'- may have been effected. 



Lamellibranchiata. 



The Lamellibranchs of the African lakes fall far short of the 

 Gasteropods in point of general interest, since there do not exist 

 in Tan^-anyika or elsewhere any of those types which have been 

 described as thalassoid. The species belong, in consequence, with 

 but few exceptions, to well-known fresh-water genera such as 

 Corbiciola, Unio, f\,\\^ Mutela. The list of species admitted reaches 

 a much smaller total (53) than in the case of the Gasteropoda, 



