356 
Journal of Agricultural Research 
Vol. XXIV. No. 4 
and in describing new species or in identifying new cultures allow for a 
certain amount of variance? The code quoted just above defines spe¬ 
cies as “connected or coherent groups of individuals.** 
There is little doubt that in some cases physiological strains of species 
must be recognized by some system to be agreed upon, possibly by 
making new species, but until formally adopted by a representative 
body varietal names, especially those based on morphological characters, 
should be avoided. 
SUGGESTION FOR A STANDARD METHOD FOR FUSARIUM STUDY 
An important step in taxonomic work on Fusarium would be to 
standardize the methods for growing Fusarium species in somewhat the 
same way that certain bacteriological methods are standardized. 
In order to do this it would first be necessary to carry on a compre¬ 
hensive preliminary study. Interested workers in different localities 
would grow a large number of species of Fusarium, preferably sub¬ 
cultures from common stocks. The conditions of media composition, 
light, temperature, and humidity should be as nearly uniform as pos¬ 
sible, selecting those suggested by the results of previous workers. 
The method of note taking in the work should be sufficiently uniform 
to facilitate comparative studies of the results. From these compara¬ 
tive studies it could be concluded what conditions proved most satis¬ 
factory in growing species of Fusarium. 
Selecting the most promising method thus obtained as a provisional 
standard, cultural work should again be carried on by a very large 
number of workers and with a very large number of species, and notes 
taken in a uniform manner. A comparison of the various notes on 
single species would determine whether or not the method used could 
be adopted as “a permanent standard.*’ 
By careful study of the various notes taken on all species of Fusarium 
used, the most stable characters could be determined and a really work¬ 
able key made. 
Such a procedure would involve a large expenditure of time and 
• money and much care on the part of the workers. To find enough 
interested workers with the time to devote to such a study might in 
itself be a difficult task. However, until a key based on comprehensive 
data of this kind is made we see little hope for accuracy in identifica¬ 
tion. (See also page 354.) 
results of IDENTIFICATION WORK 
The cultural and microscopic data, acquired as described above, were 
carefully studied and with the aid of Sherbakoff’s {18) key many of 
the cultures imder investigation were identified. Their descriptions 
and identifications follow according to groups. Practically all the 
cultures were found to be included under Wollenweber’s three sections: 
Elegans, Discolor, and Martiella. Notable exceptions to this are No. 
20, 69, and 75. For various reasons some few of the cultures are only 
provisionally identified, and two are not identified at all. The summary 
given in Table I will show which these are and give the reason for indeci¬ 
sion. (See page 362.) 
Since Sherbakoff’s key was used, the descriptions of species given by 
him were taken as the standard in most cases. If questions arose about 
