4 BULLETIN 18 7, UNITED STATES NATIONAL MUSEUM 



others; however, because of paucity of specimens from many areas 

 we have had to use our judgment as to whether intergradation is in- 

 dicated; that probably we have been inconsistent in this respect is 

 consistent with all similar arbitrary procedures that must be based 

 upon insufficient data. 



We take as a species any population isolated reproductively, geo- 

 graphically, morphologically (lacking other data), or (rare cases) by 

 total ensemble of characters, that has attained an evolutionary stage 

 in which at least one recognizable character always distinguishes that 

 population from all others. With one or two exceptions a subspecies 

 is the same as a species ; the exception may be only that reproductive 

 isolation is incomplete, or only that a lesser percent (as little as 75) 

 of its population is recognizably different; subspecies of the latter 

 type may be either completely or incompletely isolated from the near- 

 est related form. 



The attainment of a certain degree of recognizability (i. e., number 

 of recognizable differences), involving a variable interpretation in 

 different groups of animals (including genera), does not as a rule 

 figure in our definition of species and subspecies; this is consistent 

 with the functional concepts of these categories, interpreted as expres- 

 sions of the existence of differences, not degree of difference. Thus 

 two species may well be distinguished by fewer characters than two 

 subspecies. 



The expression of degree of difference — a measure of similarity — is 

 reserved for the genus. Between the generic and specific categories 

 may and usually do occur wide possible ranges in degree of distinct- 

 ness; some species (or subspecies) may show a minimum distinctness^ 

 while others show many differences from the nearest relatives. How 

 many differences between species of a single genus can be accepted as 

 commensurate with the definition of the latter category must vary 

 with the inclination of the investigator, the trend of the times, and 

 the nature of the individual problem and of the group ; no standardiza- 

 tion has yet become evident. We have therefore not been consistent 

 in our interpretation of the genus, nor do we believe consistency is to 

 be expected under present conditions. 



The term "form" we use only as a convenient and brief substitute 

 for either of the terms species or subspecies, or for both. Reference 

 to the "forms" of a genus or species involves all the lesser, taxonomi- 

 cally recognizable entities, whether species or subspecies; if any of 

 the latter are recognized, each is considered a separate "form," as is 

 each species lacking subspecies. 



The term "race" we use only as an exact substitute for subspecies. 



A "group," as we understand it, is a term including two or more 



related forms ; it may be used in reference to, for example, an assem- 



