9 BULLETIN 731, U. S. DEPARTMENT OF AGRICULTURE. 
bracts surrounding the cotton square and on to the bud itself. In addi- 
tion, a considerable portion of the experiments included principally 
treatments very early in the season before the squares had appeared. 
At this stage of growth the prime objective was to drive the poison into 
the terminal bud which forms the tip of the young cotton plant and 
which is the favorite weevil food before squares are formed. As has 
been mentioned, however, the results secured from all these attempts 
to poison the weevil were, generally speaking, very dubious and gave 
no apparent reason for hope for the successful poisoning of the weevil. 
THE WATER-DRINKING HABIT OF THE WEEVIL AND ITS BEARING ON 
POISONING. 
During the seasons of 1913 and 1914 the writer conducted a numper 
of biological investigations on the boll weevil from which an intima- 
tion was obtained of the possibility of utilizing certain newly dis- 
covered features in an attempt to poison the weevil. It seemed 
quite obvious that there was no hope of successfully poisoning 
weevils if full dependence were placed on the amount of poison they 
would secure in the course of their feeding. These studies, however, 
demonstrated very clearly the importance of water to the continued 
existence of the weevil. It was found that under cage conditions 
they drank water very regularly and it seemed reasonable to assume 
that under field conditions they would secure this apparently essen- 
tial moisture by drinking from the rain or dew collected in droplets 
on the leaves of the cotton plants. With this idea in view, the 
writer decided to conduct experiments in which the attempt would 
be made to poison the water which the weevils would drink. 
THE POISONING TESTS OF 1915. 
The first tests of this nature were instituted in 1915 and have been 
continued to date by the various agents of the Delta Laboratory of 
the United States Bureau of Entomology under the direction of the 
writer. The majority of these tests were located in the vicinity of 
Tallulah, La., which is in typical delta territory and normally subject 
to an exceedingly heavy degree of weevil injury due to the great 
humidity and excessive rains which prevail. For the purpose of these 
experiments of 1915, triplicate series of plats were surveyed on three 
different plantations near Tallulah. Each of these series consisted 
of five plats of about one acre each, as nearly uniform as possible in 
all conditions affecting their production of cotton. In each case the 
two end plats were given four, five, and six applications of poison, 
respectively. It was soon evident that this poisoning was exerting 
a very decided control on the weevils, as the weevil infestation was 
considerably reduced in the poisoned plats and these plats continued 
blooming much later in the season than did the adjoining unpoisoned 
