296 THE AMERICAN NATURALIST. [Vor. XXXIII. 
this, they are but confessing that species are for them descrip- 
tions in books, not aggregates of similar individuals in nature. 
From the biological standpoint there could scarcely be more 
useless and unproductive labor than this of matching descrip- 
tions. The merits of a proposed species, that is, its normal 
segregation in nature, are not to be inferred from the formal 
description of a few individuals, and all so-called “ species ” 
established on such.a basis are merely tentative propositions — 
suggestions for study. The question is not whether the de- 
scription is different from all other descriptions, but whether 
the type specimen is, in reality, a member of an independent 
series of individuals, a distinct branch of life, a separate island 
of existence. The systematist is in no way responsible for the 
conditions; it is his business merely to recognize and record 
them. The difficulties vary greatly in different parts of nature, 
depending upon facts in the biology of the various groups. 
Thus the well-nigh inextricable confusion of the genus Sphag- 
num is undoubtedly connected with the fact that all the species 
have almost exactly the same position in the economy of nature, 
and affect the same habitat, while segregation has been further 
hindered or at least obscured by the absence of any natural 
period to the life of the individual plants. Remotely ancestral 
and all succeeding forms may still exist simultaneously and 
contiguously, affording a rare complication of difficulties. But 
in this case, as in others, we can best gain knowledge of the 
lines of divergence and learn something of the present tenden- 
cies of evolution by locating the breaks in the series of forms. 
The use of the term “species ” may bea matter of indifference, 
but confusion in this formal regard should not be allowed to 
obscure the interest which attaches to the history and status 
of so isolated a group of plants. 
Having once located and delimited our specific islands, the 
work of studying their internal topography is comparatively 
simple. The limitation of subspecific groups is necessarily 
arbitrary, but it need not be on that account artificial, the object 
of such subdivision being the recognition of tendencies toward 
segregation, rather than the formation of groups of uniform 
size. We are concerned, in other words, with the natural 
