POISONING COTTON-BOLL WEEVILS. in 
bination of seasonal conditions and resulted in an almost complete 
absence of weevil damage in all of the cotton in which tests had been 
planned. In other words, the yield of cotton in these cuts was just 
the same asif no weevils had been present. Of course weevil- 
control experiments could not be conducted under such conditions 
and the majority of the Tallulah experiments had to be given up for 
the season. Nevertheless, a dozen or more cuts were located imme- 
diately adjoining timber where a somewhat heavier degree of infesta- 
tion was experienced, and in these cases the rule prevailed that the 
heavier the infestation, the greater the gain due to poisoning. These 
results, of course, fully confirmed those of the preceding two years, 
but the necessary postponement of many experiments still left many 
Fic. 4.—Typical view of opened cotton in check plat on Algodon Cut No. 2, second picking, October 12, 
1916, Tallulah, La. For comparison with figure 3. 
gaps in the information essential to outlining a general procedure for 
weevil poisoning. 7 
Figure 5 illustrates something of the results secured in one of the 
more heavily infested cuts near Tallulah. This photograph was taken 
to show the difference in the amount of top cotton produced by the 
poisoned and unpoisoned plats, and shows only the second picking. 
A considerable gain had already been secured at the first picking, 
and in the total a gain of something over 50 per cent was secured. 
EXPERIMENTS IN ARKANSAS AND MISSISSIPPI. 
In addition to the experiments just detailed, a number were con- 
ducted in the North Delta, in Chicot County, near Lake Village, Ark., 
and in Washington County, near Scott, Miss. At both of these 
