ARCESTID.E. 129 



Shumardites, Waarienoceras, Cifdolohus, and HyaUoccras have phylloicl 

 development of the septa, and are unhke the Popanoceratidu' in <»-eneral 

 characters. K. A. von Zittel has chissed some of tliese nnder the group 

 Cyclolobidfe, which term is liere retained as a family name for the Paleo- 

 zoic Arcestida? with phylloid septa. The writer luis observed in Shumar- 

 dites a transition from Schistoceras; and in Schistoceras a development from 

 Gastrioceras, through Famlegoceras. This family is, therefore, undoubtedly 

 derived from the gastrioceran branch of the Glyphioceratid;T>. 



It is not possible, at present, to determine which of these families gave 

 rise to the typical Arcestidpe of the Trias, but the writer is inclined to the 

 opinion that they came from the Cyclolobid*. The young stages of some 

 Arcestes, however, show a strong resemblance to Adrianifes, and both 

 branches may be represented among them. Popanoceras, at any rate, still 

 persisted until the middle Trias, with its characters little changed, and may 

 verv well have given rise to a number of the genera commonly assigned to 

 Arcestes; this genus then would prove to be polyphyletic, and the sub 

 genera into which it is divided would be given full generic rank. 



The oldest of the Arcestidae, Popanoceras, has been found in the Mid- 

 dle Coal Measures, while AgatJiiceras, from which it is supposed to have 

 been derived, is not known below the Upper Coal Measures. Agathir.cras 

 probably came from either Paralegoceras or Schistoceras, although the latter 

 genus is not known at present to antedate it. But in Gastrioceras, Parale- 

 goceras, Schistoceras, Shumardites, Waagenoceras we have a line of descent in 

 which the geologic sequence, the transitions of the adult forms, and the 

 individual ontogeny are all in perfect accord. 



Since these Arcestidfe are so well represented in the American Coal 

 Measures, and are unknown elsewhere in the world below the Permian, 

 they may be considered as having originated in the American region, and 

 to have reached the rest of the world afterwards by migi-ation. 



Suess" years ago formulnted the hypothesis that the appearance of 

 ammonites in the Mediterranean region at the beginning of Permian time 

 marked an invasion of tropical forms from a supposed southern region. 

 But the evidence brought forward in this paper shows clearly that the 

 Arcestidae need not have come from this supposed southern region, espe- 

 cially as we do not know anywhere in southern Paleozoic deposits members 



« Antlitz der Erde, Vol. II. p. 316. 

 MON XLII — 02 9 



