4 CIRCULAR 692, U. S. DEPARTMENT OF AGRICULTURE 
disease. Following this lne of thought, bare fallow rotations could 
be expected to give complete control. Corn and certain other crops 
could be expected to give control approaching that of the bare fallow 
rotations, because they are immune to the disease. Soybean or 
cowpea rotations could be expected to give moderate control, because 
they are susceptible under conditions of stem inoculation and hence 
some carry-over might be anticipated. 
The data from carefully conducted experiments failed to support 
this reasoning. Bare fallow rotations gave some reduction in wilt 
severity, but they were less effective than rotations with most plants. 
Soybean, which is susceptible to stem inoculation and was once con- 
sidered a host plant, was one of the most effective control crops when 
used in the longer rotations. In addition, there was considerable 
variability in the control obtained with such immune crops as corn, 
sorghum, crabgrass, lespedeza, and crotalaria. It is unlikely that 
these differences can be attributed to susceptible weeds or reinfesta- 
tion, because careful provision was made to exclude these factors from 
the rotation experiments, and other tests also showed that reinfesta- 
tion and the presence of susceptible weeds were of minor importance 
in determining the degree of control. 
It appears, therefore, that two separate factors were operating to 
give wilt control by crop rotation: (1) The starvation effect asso- 
ciated with an absence of host plants, as illustrated by the slight 
reduction in extent of wilt obtained with bare fallow rotations; and 
(2) the crop effect associated with the use of certain plants in the 
rotation, as illustrated by corn. A comparison of wilt severity on 
tobacco grown in rotation with bare fallow and with corn showed an 
average difference in all four trials of 39.4 percent in favor of tobacco 
grown after corn, proving that the crop effect was considerably more 
important than the starvation effect. The nature of the crop effect 
was not investigated, but it is probable that the biological situation 
in the soil was a factor. 
The degree of control obtained from crop rotation was variable. 
The different experiments provided an opportunity to study several 
factors thought to be involved. These studies included degree of 
resistance of the crop plants grown, presence of susceptible weeds in 
the immune crops, recontamination of the soil during the rotation 
period, effectiveness of the same rotation in different seasons, length 
of rotation, and effectiveness of rotations on different areas of the 
same field. Even though considerable information has been obtained 
on the control to be expected from rotations, an adequate explana- 
tion was not found for the variability observed. Factors of an in- 
direct type, not under positive control, appeared to determine the 
effectiveness of the rotations. 
CROP ROTATION RECOMMENDED 
Crop rotation should be recommended until more effective methods 
are available, even though not found uniformly effective. A rotation 
period of 3 years or more between tobacco crops is desirable because 
the longer rotations were more effective during seasons of unusual 
wilt severity. During seasons when wilt was only moderately severe, 
however, 1 or 2 years of other crops gave striking reductions in wilt 
damage. Corn, soybeans, and redtop were the most effective crops 
tested. Corn tended to give better results than other crops, and 
