18 BULLETIN 1374, U. S. DEPARTMENT OE AGRICULTURE 
shows an increase of 32.42 per cent in shedding of bolls of the infested 
blooms above that of the uninfested. 
This does not seem to check with the data on which Figure 7 is 
based. In that case, as has been pointed out, the percentage of 
infested blooms on the plants (fig. 6) is equally as great as the per- 
centage of infested fallen forms (fig. 7). These data, however, do 
not offer a basis for direct comparison. In the first place, the ex- 
periments were carried out in two different fields in which the rate 
of natural shed may have been quite different at the time. Then 
again, Figure 7 represents all shed forms, not indicating what propor- 
tion of them were blooms. And, lastly, it is possible that the feeding 
of the larva in the bloom may cause bolls to shed even though the 
larva does not touch the young boll itself, in which case an examina- 
tion of the shed boll only, after the bloom has dropped off, would 
not reveal the fact that its shedding was due to the pink bollworm. 
The figure, 32.42 per cent increase in the shedding of the bolls of 
infested blooms, or an increase of 53.1 per cent above that of the 
uninfested blooms, compares with Loftin's figures 10 of a difference 
of 26.8 per cent in shed due to the pink bollworm, or an increase in 
shedding of bolls of infested blooms of 65.7 per cent above that of 
bolls of uninfested blooms. 
DAMAGE TO MATURE BOLLS 
The damage done by the pink bollworm to bolls that reach maturity 
has been separated into damage to picked cotton and cotton render- 
ed unpickable. The latter is the cotton which is left in the field 
by the pickers, because it is too greatly damaged to be worth picking. 
DAMAGE TO PICKED COTTON 
In 1921 a number of samples of cotton were picked from different 
fields and experimental plats to obtain samples of lint and seed and 
to determine the extent to which they were damaged. These 
samples were obtained by selecting average stalks and picking all 
pickable bolls from each of these stalks. One hundred bolls were 
picked for each sample. To obtain a check sample, a number of 
locks corresponding to 100 bolls and apparently not injured by the 
pink bollworm were selected, and picked lock by lock. All of these 
samples were then ginned on a 10-saw hand gin. The damage to 
the seed was determined by examination at Tlahualilo. After 
repeated disinfection, the lint samples were sent to the Bureau of 
Markets at Washington for classification and testing. 
DAMAGE TO SEED 
From the seed of every one of the above samples a certain volume, 
averaging about 1,000 seed, was taken. These samples were carefully 
examined, the damaged seed was separated from the sound seed, each 
part was counted and weighed, and from these figures the damage was 
calculated, expressed as percentage of reduction in weight of the 
samples due to pink bollworm feeding. This method should give 
approximately the damage to the seed of the picked cotton. The 
results of this calculation are given in the column under " Percentage 
reduction in weight" in Table 16. This gives the loss in quantity 
of seed only. There was in addition a loss in quality, but what this 
amounted to was not determined. 
,fl U. C. Loftin, K. B. McKinney, and W. K. Hanson. Op. c't. 
