PREDACEOUS HABITS OF BOLL-WORM. 303 



Such uniformity depending upon brood, or diversity from diversity 

 of brood or food-plant, can by no means be laid down as a rule. The 

 early brood, however, seems to consist almost entirely of green individ- 

 uals, and those feeding upon other plants than corn and cotton are more 

 usually green also. The pink individuals are more common upon cotton 

 and the roasting-ears of corn. As Mrs. Treat has stated, a green worm 

 may turn brownish after the later molts, but ha (/"-grown brown worms 

 are very abundant in the bolls of cotton. In this connection, Mr. Glover 

 states : 



These variations of color are not easily accounted for, as several caterpillars changed 

 color without any apparent cause, being fed upon the same food and in the same box 

 with others. Several planters assert that in the earlier part of the season the green 

 worms are found in the greatest number, while the dark brown varieties are seen 

 later in the autumn, as we know is also the case with the caterpillars of the cotton- 

 worm. 



As already noted in Chapter VI, of Part I, the larva of Heliotliis has 

 one redeeming character in its occasional cannibalistic and predaceous 

 turn of mind. Boll-worms, when in confinement, have the habit, in 

 common with other lepidopterous larvae, of devouring one another. All 

 through the past summer larvae were being sent to the department from 

 the South, but whenever more than one boll-worm were mailed in the 

 same box, one only would reach us alive, all the others having been 

 destroyed. This was the case even when the box was filled with cotton- 

 leaves and bolls or corn-leaves. It might, however, be said that the 

 food dried up on the journey, and that hence they were driven to destroy 

 one another; but the fact is that even when confined inbreeding cages, 

 where fresh food was always at hand and where the conditions were 

 made as natural as possible, they seemed as hungry as ever for their 

 companions, and it was impossible to rear more than one in the same 

 box or cage. 



Still more conclusive, however, and of extreme interest, is the fact 

 that Mr. Trelease actually saw, upon several occasions, on the plant 

 and undisturbed, large boll-worms catch smaller ones, which they de- 

 voured " hoof and hide," or simply pierced the skin with their mandi- 

 bles so that the juice could be sucked, the refuse being dropped. 



In addition to this we have the fact fully established during the past 

 season that the boll- worm, in a state of nature, preys more or less fre- 

 quently upon the chrysalis of the cotton- worm. We have already quoted, 

 in an earlier part of the report, Mr. Trelease's observations upon this 

 point, but it seems eminently proper that they should be repeated here: 



Owing to its tough integument, the pupa of Aletia seems to be freer from insect 

 attack than the larva is, yet even its hard skin does not always save it. About the 

 middle of August I first noticed what appeared to be an anomalous preparation for 

 pupation in the boll-worm (Heliothis armigera), for I found several full-grown larvae of 

 this species with leaves closely webbed around them, precisely as Aletia webs up be- 

 fore changing to a pupa. An examination of one of these leaves, however, showed 

 uie that the' boll- worms had not webbed them about themselves, but had insinuated 



