8 BULLETIN 731, U. S. DEPARTMENT OF AGRICULTURE. 
points a heavier infestation was produced by different seasonal con- 
ditions, and pronounced results were secured from the poison. In 
every case the experiments were conducted on comparatively small 
areas, subject to a continual influx of weevils from surrounding un- 
treated cotton, but in spite_of this a very definite weevil control 
resulted from poisoning. The open cotton in every case showed a 
definite gain to the last row of treatment. The gains per acre ranged 
_ from 250 pounds to 1,007 pounds of seed cotton. Views of two of 
these tests are shown in figures 6, 7, and 8. 
It is, of course, impossible to estimate how much larger these 
gains would have been if entire cuts or entire plantations had been 
Fig. 5.—View looking across rear end of poisoned cotton on Mound Plantation Cut No. 2, with poisoned 
eotton on right of view and unpoisoned on leit; second picking only, October 23, 1917, Tallulah, La. 
treated, and thus the inflow of weevils from unpoisoned cotton pre- 
vented, but it is clear that these gains secured on small plats were 
very conservative. In fact, this was brought out well by one large- _ 
scale treatment described below. 
A LARGE-SCALE TREATMENT. 
About the middle of August the writer was requested to attempt 
the control of the weevil on a large section of an Arkansas plantation. 
This cotton was on very fair land, but had not been planted until well 
along in May. Weather conditions then retarded it greatly and it did 
not start setting a crop until about the latter part of July. About the 
middle of August a fair crop of bolls was present, but the plants were 
