COMPARATIVE ANATOMY AND GENERAL BIOLOGY 333 



which, while it belongs mainly to the province of comparative 

 anatomy and embryology, laps farther over into the field of devel- 

 opmental mechanics than any of the others noticed. I may say, 

 too, that I have found less suggestion of it in the writings of other 

 zoologists than I have of the preceding ones. I state it as follows: 

 The reliability of an organ or part as evidence of genetic relation- 

 ship is directly proportional to the unlikelihood of its having arisen 

 independently within the limits of the groups of animals being com- 

 pared; and the test of unlikelihood of di- or polyphyletic origin of 

 a part is the number of more or less distinct elements that enter into 

 its composition; or, what amounts to about the same thing, the com- 

 plexity of the part. To illustrate: Were .paleontology to discover 

 structures in Silurian strata, one kind of which would be allowed 

 by all to resemble closely mammalian hair, and another kind as 

 closely to resemble avian feathers; and should there be an entire 

 absence of direct evidence as to the creatures these structures 

 belonged to, the feather-like structures would furnish stronger 

 presumption of the existence of birds in Silurian times than would 

 the hair-like structures of the existence of mammals in the same 

 epoch; and this on the strength of the evidence itself, and without 

 appealing to any collateral considerations, like, for example, the 

 fact that the general course of evolution makes it probable that 

 birds originated earlier than mammals. The stronger probability 

 attaching to feathers would be due to the much greater complexity 

 of feathers than of hairs, this making it less likely that feathers 

 should have arisen more than once. Or, again, the complexity of 

 the ambulacral system of echinoderms diminishes the probability 

 that such a system would have been elaborated more than once, 

 and consequently its presence in any sort of an animal, however 

 generally unlike any known echinoderm, would still be strong 

 evidence of kinship with the echinoderms. 



Inadequate as has been my treatment of the role of comparative 

 anatomy in the investigation of problems of affinity, still less 

 adequately am I able to deal with its significance for other groups 

 of problems. Its relation to the various aspects of experimental 

 zoology, for instance, is intimate and vital. The best thing I can 

 say in this connection is that, were it my office to prescribe the 

 qualification that should be exacted of all who would go into experi- 

 mental morphology, for one thing I should insist upon thorough 

 familiarity with Roux's Problems, Methods, and Scope of Develop- 

 mental Mechanics, which introduced his Archiv to the biological 

 world; and further, I would exact unqualified accord to that portion, 

 at least, of the essay that sets forth the author's views concerning 

 the relation of developmental mechanics to the several older bio- 

 logical disciplines. 



