254 HUGH WATSON. 



cinidse arose in the Brazil- African continent.^ All this is 

 highl_y problematical ; but the uncertainty which surrounds 

 the precise origin of the Oleacinidfe does not affect the 

 general conclusion that the family is probably more nearly 

 related to the Achatinidae and their allies than to any of 

 the other carnivorous forms excepting Testacella. 



It is not necessary for me to discuss at length the phylogeny 

 of the Rhytididas; this family is evidently of very ancient 

 origin — possibly it is the oldest of all the carnivorous families — 

 and I do not suppose that anyone would maintain that it had 

 arisen from the Oleacinidas. I hope to deal in greater 

 detail with the affinities of the EhytididEe and Strep- 

 taxi dte when treating of the South African members of 

 these families ; and I have already said enough to show that in 

 all probability the carnivorous characters of the Rhytididfe 

 and Oleacinid^, and therefore of Apera and Testacella, 

 have been acquired independently. The remarkable resem- 

 blance between Apera and Testacella seems only to be 

 another example of convergence due to the common acquisition 

 of carnivorous habits. 



One thing is quite clear from the preceding argument : the 

 tribe A gnat ha is not a natural group, and should therefore 

 find no place in the classification of the Pulmonata. The 

 Rathouisiida3 should be placed with the Veronicellidse 

 among the Ditremata, as some authors have already done. 

 The Trigonochlamina3, Plutoniinas, and Daude- 

 bardiinae should be placed among the Aulacopoda or 

 Oxygnatha, next to the Parmacellinge, Vitrininas, and 

 Zonitinge. The Oleacinid^e and TestacellidaB should 

 form a group by themselves, near the Achatinid^e and 

 Megaspiridfe, if my views are correct; and to this small 

 group Morch's term Agnatha may well be applied, for his 

 original description is simply: "Agnatha. Ohne Kiefer: 

 Oleacina, Testacella. "^ The Rhytidid^e, Aperidae, and 

 probably the Streptaxidte, should be placed in another 



^ ' Man. of Concli.,' vol. xix, p. xiv. 

 - ' Mai. Blatt..' 1859, vol. iv, p. 109. 



