A BULLETIN 731, U. S. DEPARTMENT OF AGRICULTURE. 
poisoned cotton daily. When the cotton matured a remarkable 
sight was presented. The unpoisoned cotton showed practically no 
open bolls, while the poisoned plat to the very last row was practi- 
cally white with open cotton. This is illustrated in figures 1 and 2, 
which show different views of the dividing line between the poisoned 
and unpoisoned cotton shortly after the first killing frost. When the 
cotton was picked it was found that the unpoisoned plats yielded 45 
and 65 pounds of seed cotton per acre, respectively, while the poisoned 
plat yielded about 500 pounds per acre. This, of course, did not 
constitute a good yield of cotton but was very remarkable in view 
of the opportunity which this cotton had had to produce a crop. 
Fic. 1.—View showing dividing line between poisoned and unpoisoned cotton on iecaen Cut No. 1 
October 30, 1916, Tallulah, La. Beginning of additional unpoisoned cotton can be distinguished at 
upper right-hand corner of view. 
This test was particularly interesting because of the severe conditions 
which prevented the unpoisoned cotton from making any production, 
because the plants were so small, and because 1t was possible to poison 
the last row of the treated plat effectively without allowing the 
poison to drift on to the adjoining check row. For this reason the 
line of demarcation between the plats was much more pronounced 
than usually was possible. 
Additional tests during the same season located in bee cotton 
frequently gave larger gains in production per acre, but of course the 
percentage of gain was not so large. In one case of poisoning during 
the month of July a gain of about 600 pounds of seed cotton per acre 
was secured. Figures 3 and 4 show a comparison of the typical 
