BY H. L. KESTEVEN. 63 



genus. Dr. Dall was led by this to examine the 

 species of Cossmann's genus, probably from figures only, 

 and was satisfied that they formed a natural group, and he 

 postulated from their adult similarity an apex for all, similar 

 to that of T. radialis. It is to be regretted that he did not 

 examine my group with more care, for, of the six species it 

 contains, five are the Australian Tertiary fossils referred to 

 by Cossmann ; and I showed that, among these, there are at 

 least four forms of protoconch, one of which is undeniably 

 similar to that of the single recent species included. 



Another example of the unreliability of the Xerforonr/t as 

 a taxonomic feature, is provided by the Sinusigera apex. In 

 one of my first papers, I was led to expect that this apex 

 would prove to be confined to species of the genus Purpura, 

 but doubted whether it would be present in all the species of 

 the genus (24). Mr. Iredale (22) has recently re-written 

 "Kesteven's dictum thus: 'I do expect the embryo of every 

 Purpura (or any other sectional name) (group ?) to be of the 

 Smusigera type." He overlooked the fact that Baker (1) had 

 described the apex of P. lajnllus, in 1897, and that Pelseneer 

 has studied the whole ontogeny, and that neither of these 

 authors describes a Sinusigera apex for the species. In the 

 same communication, Iredale records the presence of this 

 apex in Alectrion (Nassa) and Coralliophila, whilst Mr. Hed- 

 ley has described it in Coralliophila and Bittium (18). Here 

 then IS an exceedingly characteristic Protoconch, such as one 

 was justified in expecting to prove taxonomically significant, 

 but not only is it not present in all the members of any one 

 genus, but occurs sporadically in at least four genera distri- 

 buted over three families, one of which falls within the 

 Tcfiiioglossa, and two in the Rarhiglossa, and of these latter 

 one belongs to the subdivision Glossophora and one to the 

 Aglossa. 



It may be used neither to correlate nor to segregate. 



The unreliability of the Gastropod Phylenihryo, as repre- 

 sented by its Nectoconch, for taxonomic purposes just demon- 



