112 
Psyche 
[December 
should, theoretically, attack any larva regardless of what it 
had eaten. 
v. Roesel v. Rosenhof (1749) reported that Chariclea 
delphinii L. frequently ate larvse of its own species. 
Calyminia trapexzina (L.) and Agrotis ypsilon (Rott.) have 
been reported to be cannibalistic by Berg (1875, 1892). 
Lederer (1932) reports the case of a noctuid, Scopelosoma 
satellitia L., attacking and eating the posterior end of its 
own body. It is known that some larvae when injured in this 
region of the body will devour themselves. Schultz (1935) 
reports that the following larvse ate chrysalids of their own 
species: Acidalia herbariata F., Eupithecia castigata Hb., 
and Dianthcecia capsincola Hb. He further states that 
Arctiidse commonly nibble at chrysalids and that Caradrina 
exidua is cannibalistic. 
Mr. C. M. Williams (unpublished data) noticed that when 
Htemorrhagia thysbe Fab. larvse were crowded some would 
attempt to eat others. He also noticed when many Telea 
polyphemus Cramer larvse were crowded in confinement 
with an ample supply of their food plant that certain indi- 
viduals attacked others and succeeded in breaking through 
the integument. In neither case, however, was the attack 
carried beyond this point. 
Another interesting case is that of a larva of Epizeuxis 
lubricalis Geyer which I observed devouring a considerable 
portion of the wings of a dried Colias philodice Godart con- 
tained in the same collecting box. The caterpillar then pro- 
ceeded to build a cocoon with what remained of the butterfly. 
While this last case should really be classed as saprophagy, 
it should be noted that the remaining cases of cannibalism 
occurred under laboratory conditions rather than in nature. 
The reversion to a meat diet in many cases seemed to follow 
conditions of crowding or of lack of sufficient food. The 
following cases, especially the first, are offered to show that 
similar conditions can and undoubtedly do prevail in nature. 
From October eleventh to November third 1937 I ob- 
served (Dedham, Mass.) an immense swarm of the larvse of 
Isia Isabella Smith & Abbot which was estimated at not less 
than one hundred thousand. One could collect five hundred 
of these larvse in three minutes without visibly decreasing 
the hordes covering the ground. The area thus overrun was 
