688 REVIEWS 



merous amphibia. But the hst of features in which Mesosaurus differs 

 from the cotylosaurs, as cited by von Huene, seems insufficient evidence 

 on which to revive the idea of a multiple origin of the reptiles. 



A. S. ROMER 



Die Familien der Reptilien. By Franz Baron Nopcsa. Fort- 

 schritte der Geologie und Palaeontologie. Heft 2. Berlin: 

 Gebriider Borntraeger, 1923. Pp. 210, pis. 6. 



The rapid strides made in our knowledge of fossil reptiles during the 

 past two decades have led to a number of attempts at a reclassification 

 of the group. Nopcsa's arrangement contains many interesting features 

 but, like its predecessors, must be regarded as provisional in many 

 respects.. 



Some twenty-five "types" are selected, around which are centered 

 the discussions of the groups to which they belong. More general 

 questions of relationships are treated in a later section. Twenty-one 

 ordinal groups are established, arranged in ten superorders. A discus- 

 sion of reptilian footprints is a novel feature. 



From the nature of the subject, many of the conclusions are, of 

 course, highly debatable. The group "Rhizosauria," root-reptiles, is 

 established for the reception of Datheosurus, Eosauravus, and Sauravus. 

 But the former is incompletely known, and the two latter are probably 

 lepospondylous amphibians. The comparison of expanded cotylosaur 

 ribs or rib plates with Eunotosaurus and chelonian ribs and plates is 

 untenable, as a consideration of the muscular relations of the forms shows. 



The use of "Pelycosauria" for all the American Permo-carboniferous 

 forms with a lateral temporal opening is conservative and perhaps justi- 

 fied. But the superfamilial separation of Ophiacodon from Theropleura 

 and Diopaeus, which are practically indistinguishable from it except in 

 the temporal region, seems unnatural. The inclusion of the Thalat- 

 tosauria in the petycosaurs seems somewhat rash. 



The treatment of the mammal-like reptiles is radically different 

 from that now generally accepted. Cope's term Theromorpha is used 

 exclusively for South African forms, in sharp contrast to the usage of 

 Williston and other writers. The many resemblances noted by Broom, 

 Watson, and even many earlier authors between pelycosaurs and the 

 South African forms are briefly dismissed ; the homology of the temporal 

 openings in the two groups is denied. The dicynodonts and dromasaurs 

 are grouped together as " Chainosauria" on not very obvious grounds. 



