SYSTEMATIC ZOOLOGY 271 



character or group of characters which may be relied on for systematic 

 purposes in all groups of animals. Certain characters of remarkable 

 constancy or obvious significance in some groups are quite unreliable 

 in others. Color is a conspicuous illustration of this important fact, 

 for in certain groups it is very constant, or has some very evident sig- 

 nificance, while in others it is so inconstant and diversified that it is 

 of no use to the systematist. And this remarkable difference in the 

 value of color may occur within the limits of a single genus. Only by 

 extended knowledge of a group can it be determined what characters 

 are of value for either the differentiation or the grouping of species. 

 It is perfectly proper to speak of questioning the validity of a species 

 and its validity must ultimately be determined by the constancy of the 

 character or group of characters supposed to distinguish it. But it is 

 not correct to speak of questioning the " validity " of a genus. Its de- 

 sirability and the accuracy of its definition may either or both be ques- 

 tioned. One genus is as " valid " as another but the desirability of 

 naming any group is a matter of opinion. And this brings us to the 

 last of the principles I wish to formulate. 



In all systematic work, the line between facts of nature and opinions 

 of the worker should be sharply drawn; the value of the work often 

 depends on the clearness of this line. 



One of America's greatest zoologists was wont to repeat over and 

 over again to his students these words : " The assertion that outstrips 

 the evidence is a crime." Like most aphorisms of the kind this sentence 

 needs some qualifications, but zoologists will hardly question the modi- 

 fied form that " the assertion as a fact, of what is really only an opinion, 

 is a crime." The more unqualified the assertion, or the hazier the 

 opinion, the greater is the crime. The opinions of a writer, particularly 

 if based on careful observations and long experience, may be as valuable 

 as his facts, but it ought never to be possible to confuse the two. It 

 should be the aim of every systematic zoologist to set forth his facts so 

 distinctly and so unmixed with opinions that any qualified worker may 

 form accurate opinions for himself, and to so express his opinions that 

 the justification for them in the facts of nature may be clear to all. 



Having thus set forth what seem to me five of the most important 

 principles of systematic zoology and realizing the possibility of varied 

 shortcomings of my own with reference to them, I can only add in con- 

 clusion, as indicative of my repentance, Peccavi. 



