BACTERIAL WILT OF CUCURBITS. 29 
the same virulence relations as the originals. The relative virulence 
indexes of these four isolations as deduced from many series of 
cucumber inoculations are, respectively, 0.0384 and 0.0366 for R 230 
and En 1J. and 0.0218 and 0.0229 for R 235 and En 15. 
The least virulent isolation {En 160 B) of the series which at any 
time caused complete wilting in cucumbers had a relative virulence 
index of 0.0149. This was one of four isolations made from slightly 
wilted cucumber vines collected at Goldsboro, N. C, in June, 1916. 
One of these cucumber isolations {R 317) made in the field caused wilt- 
ing of only a few cucumber leaves, never of entire plants. The other 
three isolations {En 160 A, B, C) were from cucumber material col- 
lected in this same locality and sent to the laboratory at Washington, 
D. C. } for isolation. These three isolations sometimes caused wilting 
of entire plants after a comparatively long time and in other cases 
caused only partial wilting. The bacterial wilt was rare in the 
vicinity of Goldsboro, and only after careful search were these few 
cases found. With the exception of this Goldsboro locality all the 
southern isolations possessed a comparatively high degree of 
virulence. 
Leaving out the one very exceptional case at Goldsboro, N. C, the 
average indexes of relative virulence for all cucumber isolations 
grouped according to geographical source were as follows : Ten isolations 
from Norfolk,Va., southward, 0.042; 15 from the District of Columbia, 
0.041 ; 17from Long Island, 0.040; 7 from middle New York to Canada, 
0.036; and 9 from Michigan, Wisconsin, and northern Iowa, 0.032. 
As will be readily noticed, the average virulence is highest in the 
South and gradually decreases northward. A study of the figures 
making up these averages shows that the southern isolations as a 
whole tended to higher virulence, while the northern isolations con- 
tained examples of both high and low virulence instead of decreasing 
as a unit. The higher virulence of the southern isolations is shown 
also by the fact that out of seven isolations from cucumbers obtained 
in North Carolina, South Carolina, Georgia, and Alabama, six of 
them {R 318, 320, 321, 323, 324, and 325) were found capable of caus- 
ing wilt in White Bush Scallop squash. Of the northern isolations 
from cucumber, the larger number obtained have caused no apparent 
infection when inoculated into squash. 
The isolations from squash and cantaloupe when grouped according 
to geographical source have shown this same general tendency, but 
not in quite so marked a degree, since all isolations from these hosts 
have fallen within the more highly virulent group. 
Viewing the general results of these relative virulence tests, it 
seems possible that instead of having only two distinct and well- 
differentiated biological strains of Bacillus tracheiphilus we may be 
dealing with a long series of " races" or "pure lines" of the organism 
