INTRODUCTION. 



I. CLASSIFICATION OF THE AGNATHOUS SNAILS. 



Most recent systematic writers follow the precedent of Dr. 

 Paul Fischer in grouping all agnathous snails under one 

 family, Testacellida, although Dr. Pfeiffer (1878) and Tryon 

 (1885) had already recognized several family groups, based 

 wholly upon conchological characters. In 1900 (Proc. Acad. 

 Nat. Sci. Phila., p. 564) the writer admitted five families, 

 placed in two super-family groups, thus: Agnathomorpha : 

 Glandinida, Rhytidida, Streptaxida, Circinariidce. Agnatha: 

 Testacellidce. 



Von Moellendorff and Kobelt, in the Systematisches Con- 

 chylien Cabinet (1902-1905) define ten families, which they 

 refer to the single superfamily Agnatha. This view my ana- 

 tomical studies do not support. Their Agnatha is in my 

 opinion composed of three groups of totally independent 

 origin (1) Ditremata (family Rathouisiida) , (2) Agnatha 

 proper, a group related to the Aulacopoda, and (3) Agnatho- 

 morpha, allied to the Holopoda. 



The agnathous family Rathouisiidce Heude seems to be re- 

 lated to the Veronicellida (Vaginulida) , representing a car- 

 nivorous branch of the Ditremata. The anatomy throughout 

 differs fundamentally from other " Agnatha. " The other 

 families under consideration may now be arranged as follows. 



AGNATHA : Testacellidce, Trigonochlamydidtz. 



AGNATHOMORPHA: Aperida, Rhytidida, Oleacinida, Strep- 

 taxidcz, Circinariidce. 



The two superfamily groups Agnatha and Agnathomorpha 

 are not even closely related and are readily separable by ex- 

 ternal characters. In my table of 1900 the Agnathomorpha 

 correctly follow the Holopoda, but the Agnatha should have 

 been bracketed with the next group, Aulacopoda. 



(vii) 



