26 BULLETIN 828, U. S. DEPARTMENT OF AGRICULTURE. 
hundred have been carried continuously in culture for considerable 
periods — in some cases for nearly four years. 
It was early noted by Dr. Erwin F. Smith 1 that his isolations of 
Bacillus trachei'philus from squash would always infect cucumber, 
but that his isolations from cucumber usually gave no infection at all 
on squash and at most only a slight local wilting near the point of 
inoculation. It seemed possible from these results that there might 
be at least two true biological strains of the organism. 
With a large number of isolations from different hosts and localities 
at hand it seemed well worth while to investigate this question 
further. For this purpose relative virulence tests were made of 
large numbers of isolations grown on the same medium and in cultures 
of the same age inoculated in as nearly an identical manner as pos- 
sible into one of the host varieties as an indicator. Chicago Pickling 
cucumbers were selected as being a representative variety of one of 
the most susceptible host species; and in each series of inoculation 
tests plants 12 to 18 inches high and of like age and similar vigor 
were inoculated simultaneously with cultures of the different isola- 
tions. From two to eight series of these tests were made, and in 
each series two or more (usually two) plants were inoculated with 
each isolation under study. Experience showed that within a 
variety a fairly constant reaction was given to any one isolation of the 
wilt organism, provided the plants were of comparable age and vigor. 
Occasionally an individual plant would show a wide variation from 
its fellows in reaction to a particular isolation, but since considerable 
numbers of plants were used these comparatively rare aberrant cases 
did not alter the end results of the study. 
In each series of tests careful observations were made of the abso- 
lute and relative progress of the disease, special attention being 
given to three cardinal points — (1) the number of days after inocula- 
tion when incipient wilt was noted, (2) the total number of days after 
inoculation required for actual wilting of the larger part of the inocu- 
lated leaf, and (3) the total time after inoculation for the wilting of 
the entire plant. By ' ' incipient wilt ' ■ is meant either a slight loss of 
turgor without actual wilting or a slight indication of wilt immediately 
around the point of inoculation. For purposes of ready comparison a 
" virulence index' ' was worked out for expressing the degree of 
relative virulence of each isolation to the indicator used. This 
virulence index was secured in each instance by adding together the 
three cardinal points above mentioned. These figures as they stand 
express the relative virulence in inverse ratio, the smaller figure 
expressing the greater virulence. In order to show the relation 
more clearly, expressing the greater virulence by the larger number, 
the numerator "1" was placed over each of the figures so obtained 
lOp. cit., p. 284. 
