CONTROL OF BACTERIAL WILT 9) 
hence it is desirable to arrange the cropping system so that corn 
precedes tobacco. 
OTHER CONTROL MEASURES STUDIED 
Progress has been made toward the development of more effective 
control measures. Breeding stocks are on hand that have a high 
degree of resistance, and much progress has been made in the devel- 
opment of resistant strains of good quality. Excellent results were 
obtained by treating the soil with urea to control the disease and 
then growing corn for one season to deplete the soil of excessive 
nitrogen prior to planting tobacco. 
RESULTS WITH CROP ROTATION 
ExpERIMENT ON SMALL PLOTS 
The major objective of the experiment on small plots was to test several rota- 
tions under conditions that provided maximum protection against the spread of 
the wilt parasite between plots. For this purpose, plots 12 by 24 feet were estab- 
lished on soil demonstrated by tobacco plantings in 1935 to be severely and uni- 
formly infested. Protection was provided from the two chief methods of spread 
in the following manner: (1) To keep out surface water, side walls were made of 
pine boards treated with creosote and so constructed that a watertight enclosure 
extended 8 to 10 inches below and at least 6 inches above the soil surface. (2) To 
prevent spread of the disease by infested soil, ali cultivation was carried out with 
hand tools and was preceded by sterilization in formaldehyde solution of imple- 
ments and shoes of workmen. The plots were separated by walkways 4 feet wide, 
which were kept free of weeds by frequent cultivation, and the entire experimenta! 
area was enclosed by a woven-wire fence. 
Prevailing cultural practices were followed for the commonly grown crop plants. 
An almcst pure stand. of crabgrass. (Digitaria sanguinalis (L.) Scop.) was obtained 
from natural seeding by the method of Crafton and Wells (2), in which the soil 
was cultivated about May 1 and lteft undisturbed the rest of the season. Bare 
fallow plots were cultivated frequently to keep the soil free of all vegetation. A 
winter cover crop of rye was grown each year on all plots except those in bare 
fallow rotations. 
Corn, soybeans, and bare fallow were more extensively tested than other rote- 
tions because they represented extreme types. Corn provided a rotation of an 
immune crop. Soybeans provided a rotation of a resistant crop, that is, a plant 
that was successfully inoculated by artificial means but immune to the disease 
when grown on badly infested soil in the field (6). Bare fallow provided a rotation 
that was free of crop effects and was expected to give maximum starvation of the 
parasite. In addition, rotations of crabgrass, lespedeza, cowpeas, Crotalaria specta- 
bilis, sorghum, and cotton were tested. Three plots were planted in tobacco each 
year as controls. Counts of wilted plants made August | were closely correlated 
with yield and were used as a measure of wilt severity. Yield data were not 
reported for this experiment, because the plots were too small for accurate detei- 
minations of quality and value. 
The results obtained from the various rotations are given in table 1. ‘Tobacco 
grown after tobacco was completely destroyed (fig. 1, A), all plots showing more 
than 90 percent infection on August 1 of each vear. Bare fallow, the rotation 
expected to provide the maximum starvation effect, gave measurable control, 
but less than rotations of most crop plants. A rotation with corn was found to 
be more effective than with other crops after a cropping period of 1 year, but in 
longer rotations there was no consistent difference between corn and soybeans 
(fig. 1, B and C). Reductions in wilt severity were obtained with all crops 
tested, but the results were not uniform. For example, tobacco grown after 2 
years of corn, crabgrass, and sorghum showed 2, 13, and 71 percent wilted plants, 
respectively, on August 1, 1938. These three crops are immune to the wilt disease, 
but were not equally effective in controlling it when used in rotations. 
