THE TAXONOMIC UNIT 
47 
There is no logical reason for stopping short of this, and no law 
to prevent it. It would seem that the tendency here referred to 
is a result of a perverted specialization. One is led to wonder 
whether such a practice is designed to further the ends of science, 
or to furnish an occupation. 
The whole problem of the determination of specific rank has 
been in the tentative stage from the beginning, and there is some 
reason to suspect that it will always be a matter surrounded with 
difficulty and variance of opinion. What is to be gained, there- 
fore, in attempting to establish another hypothetical unit below the 
species ? 
There are two possible conceptions of specific variants. These 
variants may be regarded as due to fluctuating (continuous) var- 
iations, and their relationship should be represented by some 
sort of a radiate pattern, thus : 
Or they may be due to orthogenetic variations, and should be 
represented in a linear system, thus: 
® — (D- — ® — ® 
In the latter case, no matter where we assign the ancestral type 
in the system, it need not have direct continuity with all of the 
other forms. So that, if A, B, C, and D are named forms of 
subspecific rank, and if A is regarded as the prototype, then B 
may be a subspecies, but not so with C and D. C might be called 
a hypersubspecies. 
It might be supposed that D would differ from A to such an 
extent as to justify specific rank; but, according to a principle 
which has grown up in modern systematic zoology, so long as 
intergrading forms exist between A and D, the latter cannot be 
of friedlei was reared artificially, and under such environmental conditions as pro- 
duced also an unusually dark male; but the aberrantly dark male escaped a christen- 
ing. To some this procedure will seem to be a prostitution of the purposes of taxon- 
omy and nomenclature. 
