12 BULLETIN 875, U. S. DEPARTMENT OF AGRICULTURE. 
injury is frequently distributed over the plants very unevenly, as the 
weevils display quite a pronounced tendency to concentrate on the 
upper portions of the plants, and percentage counts should be based 
on an examination of all squares on the plants rather than on 
squares examined at random on the upper portions. Naturally the 
percentage of injury which can be permitted varies with the stage 
of the cotton plant, but it has been found that usually, if 10 to 20 
per cent of the squares are punctured early in the season, practically 
no harm results. The number of punctured squares gradually in- 
creases until later in the season as high as 60 per cent or more can 
be punctured without any reduction in crop yield. To put cif pois- 
oning until there is such a high percentage is not advisable, however, 
owing to the rapidity with which weevils multiply after they reach 
this point, and thus exhaust the square supply and start attacking 
the young bolls. The majority of the poisoning operations in the past 
have been planned so as to start when about 15 to 20 per cent of the 
squares were punctured and then to repeat often enough to prevent ‘ 
the infestation from getting above about 25 per cent until the crop is 
set and the bolls are safe from weevil puncturing. In some cases 
where it is particularly desirable to confine the weevils to a certain 
cut and prevent any chance of migration, it is well to start at a 
somewhat lower percentage and continue even later, but where the 
only object in view is the benefit to the particular cut poisoned, there 
is apparently little to be gained from starting applications before at 
least 15 per cent of the squares are punctured. 
TIME INTERVAL BETWEEN APPLICATIONS. 
Lhe question of the time interval between applications is very 1m- 
portant and one on which only conditional advice can be given, since - 
it varies under different local conditions. In the past, once the ap- > 
plications were started they were generally made once a week as 
long as this seemed advisable or necessary. In reality, however, this 
selection of a time interval of one week was purely arbitrary and 
more recent results seem to indicate that in the majority of cases 
much better results can be secured by shortening this time interval. 
The effectiveness of a single application of calcium arsenate is de- 
cidedly limited in its duration. Its persistence on the plants natur- 
ally depends to a considerable extent on conditions prevailing at the 
time of application and immediately thereafter. In fact, while a 
very high percentage of control is secured during the first day of the 
application, this decreases the second day, and by the fourth day 
there is generally little or no effect. This short interval of effective- 
ness is due to two factors: (1) The poison is either washed or blown 
from the plants, and (2) new foliage is developed so rapidly that 
€ 
