BUREAU OF ENTOMOLOGY AND PLANT QUARANTINE 71 
the terminal buds of the plants and also cause shedding of large num- 
bers of squares and small bolls. Further studies on seasonal abun- 
dance and host-plant relationships confirmed previous observations 
that alfalfa, sugar beets grown for seed, grain sorghum, and desert 
vegetation are the principal sources of the populations that move to 
cotton during the latter part of June and in July. Accurate deter- 
mination of the populations, as a basis for beginning control measures, 
is difficult, and no satisfactory method has been developed. 
Major attention was given in 1938 to field tests on tin* relative 
efficiency of the insecticide which had previously given the best results, 
the number of applications, and intervals between dusting for the 
most economical gains. Tests were conducted with sulfur dust and 
with dust mixtures of paris green and sulfur (1 to 12) and calcium 
arsenate and sulfur (1 to 4) in three sections of the State. The gains 
from the control experiments, while not so large as in 1937, resulted 
in profitable increases in yields and better grades of cotton when 5 to 
9 applications of 15 to 18 pounds of inseel icides were applied at 7-day 
intervals. Applications at 7-day intervals gave better results than 
those at 14-day intervals. 
At Mesa, where the infestation was very light, the gains were as 
follows: From the paris green-sulfur mixture. 375 pounds of seed 
cotton per acre, or an increase of 14.4 percent ; from calcium arsenate 
and sulfur. 168 pounds, or 6.5 percent ; from sulfur, 157 pounds, or 6 
percent. 
At Buckeye, where the infestation was moderate, the gains were as 
follows: From sulfur, 1,620 pounds, or 102 percent; from calcium 
arsenate and sulfur, 791 pounds, or 41 percent; from paris green and 
sulfur, 392 pounds, or 31 percent. 
In the Yuma Valley, where the infestation was heavy, the final 
yields from the experiments were unfortunately not obtained. How- 
ever, in the first picking the best gain was from the calcium arsenate- 
sulfur mixture, where the increase was 549 pounds of seed cotton per 
acre, or 28 percent over the checks. 
An entire farm of 3,133 acres of cotton near Tucson was dusted 
by airplane under the general supervision of the Bureau, and while 
no checks were left, the management was so well pleased with the 
results that the entire acreage will be dusted again in 1939 if 
necessary. 
BOLLWORM 
The bollworm infestation at College Station, Tex., was again late, 
and the control experiments were limited to the second brood of 
worms, which appeared on late-planted cotton during August. No 
rain occurred from July 8 to October 17, and cotton began wilting 
and shedding squares; consequently a heavy bollworm infestation did 
not develop. The insecticides tested in 1938 were calcium arsenate, 
barium fluosilicate, and synthetic cryolites containing from 20 to 90 
percent of sodium fluoaluminate. Three applications of each were, 
made at 5- to 7-day intervals beginning on August 7, when an 
average of 28 bollworm q^ per 100 terminal buds were present. The 
control from the calcium arsenate and barium fluosilicate was equal, 
resulting in an increase of 366 and 371 pounds of seed cotton per 
acre, or 41 percent. The gains from the cryolite containing 20 per- 
cent and 85 percent of sodium fluoaluminate were likewise approxi- 
