154 Transactions of the Royal Society of South Africa. 



Taylor, who kindly informed me that, to the best of his recollection, the 

 individual which I now figure was that on which his description was based. 

 There is some discrepancy between the photograph and the original en- 

 graving, but the dimensions of the former agree exactly with those given 

 in the description, and there can be no doubt but that my figure correctly 

 represents the actual type. 



The single example found by Cressy is very slightly more obese, and 

 therefore with slightly less acute spire, than the type, and its umbilicus is 

 nearly twice as narrow, but it is hardly varietally separable. 



Conulinus metuloides (Smith), 1899. 

 1899. Buliminus {Conulinus) metuloides Smith, P.Z.S., p. 587, pi. 33, 

 fig. 43. D.F. 



Hab. L. Marques. Mtisherra R. Valley (Cressy). 



Conulinus sordidulus (Mts.), 1897. 

 (Plate IV, fig. 19.) 



1897. Buliminus sordidulus, Mts., D.-O.-A. iv, p. 65, pi. 3, fig. 30. D.F. 



Hah. L. Marques. Wanetsi R., Magude District (Bell Marley). 



This Kenyan species is only known to me from the above reference 

 and it is with considerable diffidence that I assign to it the shells from 

 the Wanetsi River. They are about | mm. narrower in proportion than 

 von Martens' dimensions, and appear to be more faintly angled at the 

 periphery than the specimen he figures ; his illustration, however, may not 

 be quite accurate, since he describes the last whorl as rounded, and the 

 shells agree so closely in all other respects with his description and figure 

 that I do not like to separate them without further knowledge of the 

 species.* 



Conulinus transvaalensis (M. and P.), 1893. 



1893. Buliminus transvaalensis M. and P., A.M.N.H. xii, p. 105, pi. 3, 

 tig. 6. D.F. 



Hab. L. Marques. Makulane, under bark (Junod). 



Conulinus junodi (Conn.), 1922. 



(Plate IV, fig. 18 ; Plate VIII, figs. 5-7.) 



1922. Edouardia junodi Conn., A.M.N.H. x, p. 120. D. 



Hab. L. Marques. Lebombo Mountains (Junod). 



The largest shell in the series was selected as the type ; it contains 5 



* While revising proofs, I have been able to examine an authentic examjile of 

 sordidulus ; it is identical with the Magude shells. 



