454 



SCIENCE. 



[N. S. Vol. XI. No. 273. 



describers have all been governed by the 

 same principles, or that all who have de- 

 scribed subspecies have had very well-de- 

 fined ideas as to what a subspecies really is. 

 I am sure that I myself have given trinom- 

 inal names to new forms with rather hazy 

 ideas upon the relation of species and sub- 

 species. 



The present practice of most systematists 

 in this country seems to be to regard any 

 two given forms as distinct species, unless 

 they are known to possess morphological 

 characters which intergrade. If characters 

 showing intergradation are present, the one 

 later described is regarded as a subspecies 

 of the other. The intergrading may be of 

 two kinds of categories : 



1. It may be associated with the known 

 joining of the two respective habitats. In 

 this case the individuals possessing the in- 

 termediate characters would come from the 

 region where the two forms, or the habitats 

 of the two forms, join or overlap. If E. 

 iowce were known to intergrade with the 

 Etheostoma of Aubeenaubee Creek, the con- 

 necting forms, if of this category, would be 

 found in the mouth of the creek, or in the 

 lake near the mouth of the creek. This, it 

 seems to me, would be a good example of 

 what we mean by a subspecies. 



2. In the other case, the two habitats do 

 not join, but an examination of a series of 

 specimens from each will show an inter- 

 gradation of characters. Though the sum 

 total of the characters is different, individ- 

 uals will be found in one series which pos- 

 sess, in varying degree, all the characters 

 shown by those of the other. The individ- 

 uals showing the intergradation come from 

 the same localities from which have been 

 obtained the typical individuals. They do 

 not come from intervening localities be- 

 cause the habitats do not join. 



Is it correct to regard these two forms as 

 sustaining the relation of species and sub- 

 species ? 



Trinominal names have, in many in- 

 stances, been given because the differences 

 separating the form under consideration 

 from some previously-known form are 

 slight. But now, the almost uniform prac- 

 tice seems to be to regard constant differ- 

 ences, however slight, in the absence of 

 known intergrading forms, as of specific 

 value ; and differences, however great, if 

 known to intergrade, to be only of sub- 

 specific value. 



If this view be correct, there are many 

 trinominals in current fauna! and floral lists 

 which are there without sufficient warrant. 

 In the latest systematic work on American 

 fishes 3255 species and subspecies are rec- 

 ognized. Of these, 125 are ranked as sub- 

 species. But an examination of the facts 

 regarding each shows that very few of them 

 should stand as subspecies, but as species. 

 They have, in most cases, been called sub- 

 species simply because thej' differed but 

 slightly from the most closely related 

 species. I doubt if intergradation is known 

 to exist in 25 per cent, of the cases. 



I have been told that many of the trino- 

 minals in current use in ornithology rest 

 upon the same insufficient evidence. 



It is the practice of some, I believe, to 

 describe as subspecies forms which, though 

 not knoivn at the time to intergrade, will in 

 all probabilitj', be found to do so. The de- 

 scriber's knowledge of the group to which 

 they belong, the principles of geographic 

 distribution, and the geography of the coun- 

 ti'y in which they are found, justify him in 

 anticipating the evidence of actual inter- 

 gradation. 



Personally, I doubt if this is the best 

 course to pursue. Would it not be better, 

 either to wait until the evidence is in hand, 

 or describe the new form as a species ? 



We sometimes hear the remark that sys- 

 tematists often go too far, and describe as new 

 species or subspecies forms which differ but 

 slightly from known forms ; that they give 



