240 
AMERICAN JOURNAL OF BOTANY 
[Vol. io, 
lead to the deduction of entirely erroneous conclusions regarding the 
pathogenicity of the parasite, especially when the experiments are being 
made by investigators who are a little over-zealous in detecting evidence of 
inconstancy in pathogenes. Serious mistakes of interpretation often are 
made on account of the use of supposedly pure, but actually impure, host 
material. It is not sufficient for the careful investigator of wheat diseases, 
especially when fundamental relations are being sought, to know which 
variety he is using, but he also must be sure that he is using a single pure 
line of that variety. 1 
If it is true that the greatest caution is necessary to assure the use of 
pure lines of host plants, it certainly is true that still greater precautions 
must be taken to use “pure lines” of the pathogene, because it usually is 
more difficult to detect genetic differences in the lower organisms. Does 
the practice under the present species concept give the pathologist the 
necessary assurance regarding the specific purity of the pathogene? 
It will be agreed, at least by most pathologists, and I believe by many 
mycologists, that pathologists often have derived but little aid and comfort 
from published descriptions of pathogenic fungi. Possibly this has been 
due not so much to the lack of a proper species concept as to the failure 
properly to apply that concept. Until recently species have been delimited 
primarily on the basis of morphology. But the pathologist can not be 
content to know what fungi look like; he must know also what they can do, 
because that is his primary concern. He is compelled to study fungous 
behaviorism, and, in so far as morphological descriptions aid him in this 
study, they are extremely valuable. It probably is superfluous, however, 
to state that morphology alone no longer can be considered a sufficiently 
accurate basis for determining the specific purity of many organisms. 
It would scarcely be profitable, even if it were possible, to define species. 
In general, however, as applied to fungi, the concept seems to be based on 
the general underlying ideas, (i) that all the individuals comprising a 
species are sufficiently alike morphologically to make it possible to differ¬ 
entiate them from individuals of other species by means of morphological 
characters, and (2) that the characters are relatively stable through succes¬ 
sive generations—“a perennial succession of like individuals,” according 
to Farlow. 
The grouping of individuals into species, then, is an attempt to make 
it possible quickly to recognize closely related plants and to call them by 
name readily. The morphology of the fungus has been used as a criterion 
of essential similarity of individuals. And unfortunately it often also has 
been misused. The careless, premature, and rather reckless naming of new 
species of fungi has been a tremendous handicap to the pathologist. Every¬ 
one is so familiar with the unfortunate consequences of this tendency, now 
happily disappearing, that it is scarcely necessary to cite specific examples. 
Too often temporary modifications, which are not inherited at all, have 
