2 54 



S. W. WILLISTON 



are required for geological correlations, or for minor problems in 

 evolution, and their naming should, so far as possible, be deferred 

 until much more material is available for comparison. As I have 

 said elsewhere, the genus is practically our unit for most of the 

 older vertebrates. In very few of the Permian vertebrates do we 

 yet know what specific differences really are; we have few measur- 

 ing sticks yet to measure them by; and it is only the morpho- 

 logical characters of the genus and higher divisions that concern 

 us much in paleontology. Perhaps this protest is a sort of belated 



Fig. 6. — T rimer orhachis. A, left femur, from behind; B, tibia; C, humerus, 

 radial side; D, another humerus, dorsal side; E, coracoid ( ?); F, the same, end view; 

 G, right ilium, outer side; //, anterior intercentrum, from below; /, the same, from 

 behind. All figures three-fourths natural size. 



repentance on the part of one who has himself contributed hundreds 

 of specific names to zoological and paleontological literature, but 

 the many hours he has spent in the endeavor to identify organisms 

 from vague and imperfect descriptions at least give him the right 

 to urge temperance in the making of more or less worthless types! 

 About twenty-five specific names were given to the mosasaurs of 

 Kansas. The most diligent search of abundant material has 

 resulted in the recognition of scarcely a half-dozen. The same fate 

 probably awaits the great majority of the specific names which 

 have been given to the Permian vertebrates. 



