Jone, 1893. CANNIBALISM AMONG INSECTS. 445 
All the caterpillars belonging to a particular group or family 
showed a preference for the flesh of their own relations. They 
devoured one another in considerable numbers, rarely feeding on the 
plants suitable for their nourishment. The caterpillars of the family 
of the Bombyces devoured their kin with skin and hair, and they 
went as far as to break through the cocoons in order to finish with 
the chrysalis within. 
In a similar manner, the caterpillars of the Noctuids behaved 
with those of their own kind and with those of the Bombyces. 
The caterpillars of these latter attacked also those of the former 
without any exception. The most voracious of the Noctuids was the 
caterpillar of Heliothis armigey; a single one of these consumed, in 
twenty-four hours, six or seven other caterpillars. 
The caterpillar of the well-known ‘butterfly, Pyvameis carye, is 
also a carnivorous cannibal, but with moderation, preferring always 
fresh plants to meat, while the others, and chiefly those of the 
Noctuine, after having tasted the flesh of their own kin, would no 
longer touch vegetable food. 
I explain this particular character of the Patagonian caterpillars 
in the following manner. During the principal part of the summer in 
Patagonia there is considerable heat and dryness which causes the 
vegetation quickly to be parched up and scarce. When this happens 
the caterpillars lose their means of subsistence, but, in order that 
some may survive, the struggle for existence has taught them another 
means of subsistence, that is, the flesh of their own kind. Once this 
instinct has been inherited, the descendants will use it whenever an 
occasion presents itself, and, in many instances, even when there is no 
lack of vegetable food. In some cases it is necessity which produces the 
habit ; in others, inheritance which leads up to it ; thus new biological 
characters are formed. 
Of other herbivorous insects cannibalism has been noticed in 
crickets in confinement. The first notice of this strange taste we owe 
to Sir W. Brodie (Canad. Entom. vol. xxill., p. 137, 1891), whose 
observation has been recently confirmed by Mr. Philip Laurent 
(Entom. News, vol. ii., p. 180, 1891). 
In an assemblage of many crickets kept for certain observations 
in a rearing drawer, or box (caja de herborizacion), the numbers 
diminished from day to day; at last only one—not a little fattened— 
remained by the side of the remains of his former companions. 
Hitherto cannibalism among the crickets has been noticed only 
among captives, but I am now enabled to state that, under certain 
conditions, cannibalism is present among some Orthoptera in the 
free state, at all events among the locusts. 
In the summer of 1883, in which the excessive heat and drought 
had brought about the nearly entire disappearance of the vegetation 
in a good part of the country, and more particularly in the broken 
country of the Banda Oriental, I had occasion to make a journey 
