256 
Journal of Agricultural Research 
Vol. II, No. 4 
septation, color, and is further characterized by longevity. Therefore, 
this stage and all conditions favoring its production may be called the 
normal stage. 
The, reliability of the taxonomic work on the Fusarium problem 
depends, therefore, on the reliability of this criterion of the norm on 
which the results are based. If we rely on this test, attention must be 
given to diagnoses based on other criteria of the norm. Many descrip¬ 
tions doubtless include the measurements of swollen and constricted 
stages of conidia and ascospores and consequently show a much higher 
range of variation. The average size, of course, gives a smaller range 
of variation than the absolute size of particular spores. The diffi¬ 
culty is further increased by the fact that the average size can not 
be based on an average of spores produced on agar, or from a single cul¬ 
ture grown on stems, etc., for the reason that constant temperature, 
light, moisture, and atmospheric humidity can not always be produced. 
Even though all these environmental factors are constant, variation in 
the shape and the size of the conidia might result in subcultures by the 
transfer of a different type of spore. Inconstant environmental and other 
factors must be expected both in nature and in culture. If the shape, 
the septation, and the color of the spore are more constant under some 
conditions than under others, this state may be regarded as normal, the 
swelling and constriction of spores as abnormal. 
An example: One conidium of Gibberella was 9-septate and 100 by 6 jn 
in size; another which was 9-septate and with swollen cells measured 
85 by 8//, a third which was 2-septate measured 10 by 2.5/1. These 
would give a range of variation as follows: 2- to 9-septate conidia, 10 to 
100 by 2.5 to 8//. The average of so-called normal spores was 3- to 5- 
septate, 30 to 60 by 4.25 to 5.5/*. 
According to Saccardo (1879, p. 513) 1 , the 5-septate conidia measure 
24 to 40 by 5 fi t while in our Gibberella from wheat grains they average 
45 to 60 by 5 to 5.5/1 and have an absolute fluctuation from 38 to 72 by 
4.5 to 6/z. This fact shows that if Saccardo’s diagnosis is based on the 
same fungus it includes only comparatively dry conidia. Thus, the 
results may be comparable if the influence of such factors as moisture 
is known. In this case comparison was simple because Gibberella is a 
characteristic fungus and can not be confused with any other. In species 
of Fusarium without known ascus stages it is more difficult, since it 
has been proved that most of the species are ubiquitous, or at least are 
not confined to a special host. In order to identify these species, a large 
number of strains must be collected from different hosts and compared 
with the species described from these hosts in the literature, paying care¬ 
ful attention to the conditions under which the species are found and on 
which the description is based. In many cases this was impossible, and in 
1 Bibliographic citations in parentheses refer to '‘Literature cited/’ pp. 384-285. 
