8 CONTROL OF THE BOLL WEEVIL. 
weevils of their food at early dates in the autumn. In fact, the 
experiments showed a practically uniform increase in the number 
of weevils surviving as the dates of the destruction of the plants 
became later. For instance, in all of the experiments performed in 
Texas it was found that September destruction resulted in a sur- 
vival of only 0.2 per cent; destruction two weeks later showed a 
survival of 2.3 per cent; destruction during the last half of October, 
5.6 per cent; and during the first half of November, 15.4 per cent. 
The results of the Louisiana experiments were similar. Destruc- 
tion in September showed a survival of 0.3 per cent; destruction in 
the first half of October, 2 per cent; in the last half of October, 8 per 
cent. 
In addition to the experiments in which the weevils have been 
placed in cages at different times in the fall the Bureau of Ento- 
mology has conducted considerable field work to show the benefits 
of fall destruction. The most striking experiment was performed 
in Calhoun County, Tex., in 1906. In this experiment an isolated 
area of over 400 acres of cotton was utilized. There was no other 
cotton within a distance of 15 miles. By contracts entered into by 
the department, the farmers uprooted and burned all of the stalks 
during the first 10 days in October, 1906, and provision was made 
to prevent the growing of sprout cotton. As a check against this 
area, cotton lands about 30 miles away were used. Here the stalks 
were not destroyed in the fall, and the interpretation of the results 
of the experiment was based upon a comparison of the number of 
weevils present during the following season in the two localities. 
In May, 1907, following the destruction of the plants, careful search 
revealed only one weevil in the experimental area. In the check, 
however, the weevils were so numerous at this time that practically 
all of the squares had been destroyed. Examinations made later 
showed similar advantage in regard to freedom from the boll weevil 
of the area where the stalks had been destroyed the preceding Octo- 
ber. The last examination was made on August 20, 1907. At this 
time there were 10 sound bolls per plant on the experimental area 
and only 3 per plant in the check area. The difference in yield 
between the two areas was about 600 pounds of seed cotton per 
acre. The work, therefore, resulted in an advantage amounting to 
about $18 per acre. 
Newell and Dougherty ! have described a very satisfactory device 
for cutting the cotton stalks in the fall. It consists of a triangular, 
V-shaped, wooden framework designed to pass between the rows 
and cut two at the same time. In the process of cutting the ma- 
chine windrows the stalks from two rows in the middle near where 
1 Circular 30, Louisiana Crop Pest Commission. 
