1913] Nichols,— Notes on Connecticut Mosses,— IV 7 
and most comprehensive sense — may vary considerably in different 
regions, the divergence frequently becoming so pronounced and appar- 
ently so constant as to make it convenient from a taxonomic stand- 
point to recognize more than a single species. From the foregoing 
remarks it can readily be seen that the difficulty in drawing specific 
lines in this genus is very great, and it is also obvious that not all of 
the various described species possess equal specific worth. While 
there is no question regarding the specific value of many of them, 
others represent merely variants due to regional or environmental 
factors.. The method of treatment adopted by Cardot takes cogni- 
zance of these conditions and his scheme of classification is outlined 
briefly in the next paragraph. 
On a basis of the vegetative characters of the various species Cardot 
first subdivides the genus into six sections. Each of these sections 
comprises a natural group of closely related species which may be con- 
sidered as in all probability having been derived from a recent common 
ancestor. He further classifies the respective species of each section 
into four ranks according to their relative importance or distinctness 
as species. Species of the first rank — there may be more than one 
such species in a given section — exhibit marked structural differences 
from one another. Species of the three lower ranks are separated 
from one another and from those of the first rank by fewer and less 
definite points of distinction — they are therefore subordinated to 
species of the first rank. Among themselves first rank species are as 
nearly as possible of equivalent value as regards the sum total of their 
distinctive characters, and the same observation might apply equally 
well to species of the second, third, and fourth ranks respectively. 
Transitions do not exist between species of the first rank, but species 
of the lower ranks are not always clearly delimited either among them- 
selves or from the higher members of the same series. Pari passu 
there are sometimes encountered forms which it is impossible to name 
with any degree of certainty. Species of the third and fourth ranks 
generally occur as local or regional races, in most cases replacing or 
excluding in a given region the species from which they have probably 
been derived. Thus, for example, while the typical European form 
of F. antipyretica is very rare in this country it is paralleled, so to 
speak, by F. antipyretica, var. gigantea, F. neomexicana, etc.— forms 
which are not present in Europe. 
1 Species of the latter sort correspond variously to the elementary species of DeVries, 
the ''kleine Arten” of K. Müller, and the subspecies of many authors. 
