200 CALIFORNIA ACADEMY OF SCIENCES. [Proc. 3D Ser. 



(Puzosia), on Plate XI, fig. 5, a-k. On Plate XIII, fig. 5, 

 a-n, Branco has figured the early stages of Cymhites, which 

 show all the essential characters observed in the larvse of 

 the other genera mentioned, although less accelerated, and 

 necessarily more primitive. 



The systematic position of HopNtes, according to the 

 investigations of the writer, and of Sarasin (with whom the 

 writer substantially agrees), is not with the Stephanocerat- 

 id£e, but with the Perisphinctinas. Des^noceras, on the other 

 hand, undoubtedly belongs to the Stephanoceratidse, along 

 with Sonneratia, for adolescent stages of both genera 

 resemble perfectly Stephanoceras^ and ITolcostephanus, and 

 these, in turn, sprang from Cceloceras. The eegoceran an- 

 cestor of Cceloceras and of the perisphinctoid group is the 

 same, and can be traced back to a Cyinhitcs-Vike form, and 

 this must have originated in a genus with the essential char 

 acters of Nannites, with the possible intermedium of some 

 primitive unknown member of the Phylloceratidae. The 

 immediate ancestor of Hoplites is not to be sought in a per- 

 fected Cosmoceras, nor did this genus spring from any 

 highly specialized Perisphinctes, which, according to Buck- 

 man, is a degenerate group of the Stephanoceratid^; nor 

 was any fully differentiated eegoceran form the parent of 

 this perisphinctoid ancestor. We must rather seek, in each 

 case, the radicle in the primitive, unspecialized beginnings 

 of each group. And if all these genera come from a phyllo- 

 ceran stock, they certainl}^ do not come from Psiloceras, 

 Phylloceras, nor Lytoceras, but from some primitive member 

 of the PhylloceratidiE, as yet unknown, or at least not recog- 

 nized as belonging to that group. At any rate, the stage 

 between that resembling Nannites, and that suggesting 

 Cymbites, is too short, and too little differentiated in any 

 species of Hoplites yet investigated for a probable reference 

 to Monophyllites. 



1 The writer has recently worked out a remarkably perfect series of Sonneratia stan- 

 toni Andersou MS., Desmoccras Jiaydeni Gabb, D. breweri Gabb, and D. hoffvianjii Gabb, on 

 which series the above statements are based. 



