82 PEOCEEDINGS OF THE MALACOLOGICAL SOCIETr. 



Tlphohia with some of the recent Purpurm. Besides being quite of 

 thin texture, the Tanganyikan shell has a peculiarly prolonged rostrum, 

 which is hollowed out or grooved on the inner side ; moreover, the 

 spines at the angle of the whorls are hollow and delicate, whereas in 

 Purpuroidea the nodules in the same position are shorter and solid. 

 The anterior canal in the latter genus is somewhat like that of 

 Purpura, being shallow, broad, and short ; in fact, I may say it bears 

 no resemblance to the grooved rostrum of Tipliohia. 



Thus we come to the conclusion of the consideration of these 

 so-called ' halolimnic ' forms with reference to their supposed Jurassic 

 prototypes, and with what result ? Of the eight genera discussed, 

 one only, in my opinion, can be regarded as satisfactorily agreeing with 

 tlie fossil form, namely, Faramelania with Putpiirina, but this is the 

 one instance, I have shown, in which the genus in question has 

 a representative in a later period, namely, Pyrgulifera of Upper 

 Cretaceous times. 



With regard to the rest of the genera, I think it has been clearly 

 demonstrated that, in my opinion, they do not correspond to the 

 Jurassic forms from which they are said to be indistinguishable. In 

 some cases they exhibit a general resemblance of outline, and that 

 is all, but when we come to take into consideration their other 

 characters, especially with regard to the aperture, we find so much 

 difference, that it cannot be said that any one of them is absolutely 

 identical with the Jurassic type. 



That these Tanganyikan shells have had a marine ancestry, the 

 same as other fresh-water shells, of course must be recognised, and 

 that they may have retained a more thalassoid facies than others is 

 possible, but that it can be said that they are indistinguishable from 

 certain Jurassic types I cannot admit. Other lakes besides Tanganyika 

 have their special faunas, including forms which are found nowhere 

 else, for example, Lake Baikal and the Caspian and Aral Seas, and 

 in all three we find living together both fresh-water and marine types. 



Fresh-water molluscs do not, of course, form a natural class of them- 

 selves. The different families have their relationships with various 

 marine groups, and this connection may be more or less intimate. 

 Although they may have had common ancestors in remote ages, yet 

 the divergence of characters existing between them at the present 

 day precludes the possibility of affirming definitely their common 

 origin. All fresh-water molluscs have had their position in the system 

 of classification assigned to them, and in this connection I may refer 

 to a few examples. 



The genera Clea and Canidia appear to be closely allied to the 

 marine Nassas and Buccinums ; the Melanias, according to Bouvier, 

 show a near relationship to the Cerithia ; Vivipara should be ranged 

 near the Turbinidae and Trochidse, AmpuUaria is considered to come 

 near the Viviparas, Bithinia has an afiinity with Littorina, and Valvata 

 with BiOiinia and the Ilissoidae. 



I merely refer to these relationships, in passing, to call attention to 

 the fact that, whether thalassoid in aspect or not, the relationship of 

 fresh-water shells with marine forms is a recognised fact. It is not, 



