1893.] SHELLS FROM BRITISH CEXTRAL AFRICA. 635 



7. AciLVTrffA sp. inc. 



Hah. Angoui-Land (^-1. Whi/te). 



A number of specimens of two species of AcJiatina occur in the 

 collection ; they are all, however, in too bad a condition for identi- 

 fication. One is very hke J., tavaresiana, Morelet, from Angola ; 

 the other is considerably larger and rather like the form of Reeve's 

 figure of A.fulka (Conch. Icon. pi. 2. f. 8). 



8. Ampullaria ovata, Olivier. 



Hah. Sumbu, Itavva, S.W. side of Lake Tanganyika (-B.CrajysZia?/). 

 This Nilotic species has ah'eadv been recorded from the lake 

 (P. Z. S. 1880, p. 348). 



9. Lanistes solidus, Smith. 



Hah. Karonga, west shore of Lake Nyasa {R. Crawsliay). 

 One dead specimen. 



10. L.vjs-iSTES AFFi^ris, Smith. 



Hah. "With the preceding (Oraicshaj/) ; Angoni-Land (WJujte). 

 In dead condition fi-om Karonga. The large number of speci- 

 mens fi'om Angoni-Land are all small or only half -grown. 



11. La^-'istes ntassanus, Dohrn. 



Hah. Angoni-Land, south-west end of Lake Nyasa (A. Wliyte). 

 The eight specimens of this fine species are all in a worn con- 

 dition and \\ere evidently picked up dead upon the shore. 



12. L.AtfiSTES ovmr (Peters MS., Troschel). 



Hah. Kabwiri, Lake Mweru {11. Cravjsliay). 



This species was originally described from Mozambique, and 

 although quoted from Lake Nyasa I have not yet seen specimens 

 from there myself. 



13. ViVIPARUS TANGANYICENSIS (Smith). 



Hah. Sumbu, Itawa, S.W. end of Lake Tanganyika {R. Craiv- 

 sJiay). 



Some of the specimens from this locality are remarkably tabu- 

 lated, as in Bourguignat's figure (Ann. Sci. Nat. 1890, vol. x. 

 pi. iii. fig, 1), and the body-whorl is even more strongly keeled at 

 the periphery. The eight so-called species of Neotliauma figui'ed 

 by Bourguignat must be regarded as varieties of this variable 

 form. 



The nomenclature of this genus has been fuUy discussed of late 

 by Mr. Dall\ but his conclusions, to ray mind, are open to 

 objection. In the first place, I would point out that Martini \\a3 

 not a binoinial author, excepting in occasional or chance instances, 

 and therefore that such chance names as he may have given can- 

 not be accepted. 



' Traus. Wagner Free Inst. Sci. Pliilail, 1892, vol. iii. pt. 2, pp. 332-335. 



