3l6 O. F. COOK 



appear to explain parallel variation under parallel conditions, 

 but the recognition of the kinetic principle enables us to under- 

 stand parallel variation even under different conditions. 



Inside specific lines descent is a completely connected fabric, 

 but superspecific descent, the phylogeny of genera, families, and 

 orders, is not reticular at all. For lack of adequate evidence 

 we may be unable to decide which is the nearest relative of a 

 given group, but when we represent our groups as having com- 

 plex interrelationships we are merely making graphic represen- 

 tations of alternative solutions of unsolved problems. The com- 

 mon possession of an ancestral character affords, in itself, no 

 assurance of closer relationship, nor do the separate acquisitions 

 of similar characters. Each character must be placed, as it 

 were, in its true chronological position before its phylogenetic 

 significance can be appreciated. Without careful regard for 

 sequences, phylogeny becomes as hopeless as history without 

 dates. 



If the parallelism of variation be accentuated by selective in- 

 fluences there occur wonderful approximations in the characters 

 of different and unrelated organisms living under diverse nat- 

 ural conditions in remote and isolated regions. The facts have 

 been effectively summarized by Professor Osborn and made the 

 basis of what is called " The Law of Adaptive Radiation." ^ In 

 each continental area and geological period there have arisen 

 among the mammals specialized groups adapted by their teeth 

 to all the different kinds of food available. There are always 

 some with slender skeletons and long legs adapted to escape by 

 running, and others stout-footed and heavy- limbed, able, in all 

 probability, to protect themselves by sheer strength and ferocity 

 or by defensive armor. 



Adaptive radiation is inconsistent with both of the current 

 ideas, that evolution is caused by the environment or by a pre- 

 determining hereditary mechanism. The conditions are too di- 

 verse to cause such similarity of results, but at the same time 

 the results are too diverse to warrant the inference of predeter- 

 mination. The trutli lies, obviously, between the two extremes. 

 The environment dors not cause evolution, but neither is evolu- 



' American Naturalist, 36 : 353, 1902. 



