94 THE AMERICAN NATURALIST. [Vol. XXXV 1 II. 



Lamarck-Darwin theory of the law of descent which involves 

 a branching or phylogenetic scheme of relationships. Toward 

 this we are slowly progressing. The final classification will be 

 a formal or tabular statement of the tree of descent, in which 

 only so much pre-Darwinian classification will survive as was 

 based upon the perception of real phylogenetic characters. 



The evolution philosophy held out a strong temptation to 

 rapid generalization in phylogeny. It is a striking fact that the 

 evolutionists, Huxley, Cope, Haeckel, perhaps because they 

 attempted to generalize too rapidly, have proved less fortunate 

 in their arrangement of the Reptilia than Owen, whose pre- 

 Darwinian systems of 1839 and 1859 have best stood the 

 test of time and of discovery. 



Both Cope's and Huxley's systems are largely wrecks today ; 

 Huxley's because while entirely logical in method it outran the 

 state of knowledge and discovery. Cope was less logical ; his 

 fatal error was over reliance on single characters without dis- 

 criminating whether they were primitive or adaptive. Marsh 

 was gifted with unerring taxonomic judgment as to real phylo- 

 genetic relationships ; the chief defect of his system was that he 

 partly or wholly ignored the rules of priority, renaming and 

 redefining groups which had previously been defined with suf- 

 ficient clearness to be recognized. It is with real regret that 

 I feel compelled, as a matter of historic justice, to revive some 

 the older names for certain groups of which our knowledge is 

 a niost entirely due to the fundamental contributions of Marsh. 



Pala^ontological discovery is constantly swelling and 

 expanding the groups of fossils discovered long ago ; it is con- 

 sequently necessary either to abandon these groups or to raise 

 7' 7' ^^"^ e.xample, Owen's ^'family" Cyno- 



;ontia has become a suborder, his "families " Dicynodontia and 



an!n')in!"aurr^^^^^^^^ .'''''^''T''' ^'J' " ^"^^"lodontia 



i>'>sUiM„ oriuinally assigned to it by its author in a verv limited 



