No. 446.] RFXLASSIFICA TIOX OF THE REPTILIA. 97 



the possession of a single primitive character, namely, the 

 fixed quadrate. 



We find that almost every attempt to classify the reptiles by 

 superficial characters, by external organs, by general externil 

 adaptations, by single internal organs, has proved unnatural. 



True Principles of Classification.. 

 The conclusion is that there are three ruling principles in 

 classification. 



First, as Xo priority, we owe it to our palaeontological forebears 

 not to abandon the lower or higher groups they have proposed 

 except in cases of absolute necessity. In some instances \vc 

 must give a group a higher or lower rank than the author 

 originally assigned to it, or a different position in the system. 

 Priority has no force where a group is based on a false con- 

 ception or on a false grouping of types, as in the definitions 

 of Theromorpha and Archosauria by Cope. 



Second, as to phylogeny, classification is a formal expression 

 of our knowledge of phylogenetic relationships ; it must, there- 

 fore constantly shift and change as new relationships are 

 discovered. The final classification will be the phylogenetic 

 tree. This being the case, it is desired to include within a 

 group its ancestral forms as soon as they have definite y 

 branched off toward it. For example, Hyracothenum should 

 not be placed in the Lophiodontidae, as was done b>- Cope, but 

 in the Equid^. Again, if it should be demonstrated that^^the 

 Protorosauria are ancestral to the dinosaurs and to no ot ei 

 reptiles, they should be placed in the superordcr Dinosauria. ^ 



Third, as to definition, classification, like phylogeny. s lou ( )e 

 based on a number of characters of diff. 

 having different functions, in order to ^ 

 being misled by analogous exolution, o 

 allelism, convergence and homoplasy. 



The neglect of one or other, and in 

 three principles and the loss of the clai 

 Baur have led to great and rapidly increasing 

 arrangement of the Reptilia in recent years, h 

 Broom and von Nopsca, have been working in the right direct 



ish the 



g confusion in the 

 Smith Wo(xlward, 



