354 mr. L. f. spath ox the [Dec. 19T4, 



stress on the presence of dependent auxiliaries in some psilocerates, 

 nor in relying on the costation. Psiloceras cannot, however, yet 

 be attached to any particular form of the Monophyllitida?. 



But if, on the evidence of the suture-development, we favour 

 the recognition of an early-neanic Mottophyllites stage 1 and a late- 

 neanic Psiloceras stage, we must remember that the evolute and 

 rounded- whorled Tragophylloceras ancestor was constricted at a 

 comparatively late date, if we are to argue phylogenetically from 

 the persistence of the constrictions in Tr. numismalewpto a diameter 

 of about 20 mm. Constrictions occur in a number of genera 

 belonging to the Phylloceraticlse and Lytoceratidse of the Alpine- 

 Mediterranean Lower Lias, but unfortunately the ontogeny of these 

 forms is unknown. Involute forms are referred to ' Phylloceras ' 

 and most arbitrarily distributed among the groups of Ph. lietero- 

 phyllum, Ph. capitcmei, and Ph. partschi. Rliacopliyllites is 

 distinguished chiefly by evolute whorls (obviously a most unsatis- 

 factory character) with a correspondingly smaller number of 

 auxiliaries, but has, like Phylloceras, typically diphyllic saddles. 

 Fig. 3, PL L. shows the development of the suture of Rhaco- 

 phyllites stella (Sow.) from a diameter of 4 mm. The monophyllic 

 second lateral saddle may point to the derivation (by way of 

 Discopliyllites) from the Monophyllitidae. The shape of the conch 

 of Rliacopliyllites stella at diameters of 4 and 6 mm. respectively 

 (PI. L, tigs. 1 & 2) is the morphic equivalent of the hypothetical 

 Tra g opli yll oeeras ancestor ; and, if the evidence of the suture-line 

 were not so decisively against it, one might be tempted to accept 

 (lever's opinion of the relationship of these two genera. 



The undue importance given to whorl-shape in these phyllocerates 

 is responsible for the extraordinary confusion which prevails in the 

 grouping of the 120 forms of the Lower and Middle Lias. The 

 forms referred to Gey er oeeras 2 and to the group of Phylloceras 

 persanense Herbich, show the variability in section. Some of these 

 forms are, by their sutures, allied to Phylloceras, while others are 

 clearly rhacophyllitic. although excluded from Rliacopliyllites 

 on account of their involution. Again, Phylloceras glaberrimiini 

 Neumayr, and Ph. psilomorplium jSJeumavr. which, according to 

 Oeyer (1886), probably belong to Rliacopliyllites, are referred (like 

 Ph. Immense Meneghini) by most authors to the group of Phyllo- 

 ceras heterophylhtm, also because they are involute. Their external 

 saddles are diphyllic, as is also the case in Ph. togatum Mojs., 

 although this form, 3 on account of its constrictions, is referred by 



1 The nepionic stage, which must agree with the larval development of that 

 Triassic genus, will be dealt with later. 



2 A specimen of Geyerocem* cylindriexvm (Sow.), which I dissected by 

 kind permission of the Keeper of the Geological Department of the British 

 Museum, was unfortunately, owing to its preservation in brittle iron-oxide, 

 not reducible to a sufficiently small diameter, and consequently the develop- 

 ment of its suture-line could not be traced. 



:! According to Wanner (on. cit. pt. viii, 1898, p. 175) this is identical with 

 Rliacopln/llites stella Neumayr. 



