34 
Trelease" observed more than tliirt}^ years ag'o that during dry 
weather, and on the drier situations of the field, ants were much more 
abundant and hostile to bollworms than during- rain^^ weather, or on 
the lower and more moist parts of the field. Several observers have 
advanced the theorv that the absence, during- dr}^ seasons, of both 
the boll and cotton worm is largeh^ due to the effectiveness of several 
species of ants in keeping them in check. The importance of this 
theory, in its bearing on the subject under discussion, rests on the 
assumption of the predator}' habits of the ants. From our own obser- 
vations we are not inclined to attach much importance to the work of 
ants, for the reason that frequent and close observations of these 
insects in cotton fields and elsewhere have failed to verify previous 
statements of their habit of voluntarily preying' upon bollworm larvae. 
Under certain conditions, as when provoked, several species of ants 
have been observed to attack and kill small bollworms, but the few 
instances when ants have been observed feeding on larv^ in the fields 
have not been free from the suspicion that the worms had been pre- 
viously injured and more or less disabled, as by one of their fellows. 
At no time during the past two ^ears have any of the native cotton- 
field ants been observed, under natural conditions, to voluntarily 
attack bollworms. 
It remains to mention a fact doubtless of considerable importance, 
namely, that rains produce in the cotton plant a rapid and more succu- 
lent growth, which, by furnishing the larva? an abundance of tender 
food, great!}' favors their development. This and the increased food 
suppl}' for the moths under these conditions, as well as the increased 
percentage which are able to escape from the soil, are reasons which in 
themselves are almost sufficient to account for the greater destructive- 
ness. An illustration of this is to be seen almost every year in the 
known greater destructiveness of bollworms on ^'bottom-land'' cotton 
where the soil is moist and more fertile and the weed growth stronger, 
as compared with the injur}- on the more stunted growth on uplands. 
It is probable that the moths are primarily attracted to the ranker cot- 
ton by reason of the greater nectar supply, and eggs are deposited on 
these plants during the course of their feeding. 
The influence of shade, as during cloudy and rainy weather or by 
reason of the luxuriant growth of cotton in closely planted rows, is 
also apparently favorable to the larv^, but it is to be noted that these 
feed readily without seeming discomfort on the exposed portions of 
the plants, often in the direct rays of the sun. 
In the egg stage the bollworm is subject to parasitism by a minute 
chalcidid fly {Trichogramma jpretiosa Riley), the importance of which 
in destroying bollworm eggs doubtless varies much with the character 
f'Comstock's Kept, on Cotton Insects, Washington (1879), p. 378. 
