1903] WHEELER:— CERAPACHYS AUGUSTAE 207 



Ponerinae (Leptogenys, Pachycondyla) , but without any definite orientation in a 

 single subspherical mass. The first eggs did not hatch till May 14th, showing 

 that the incubation period must exceed eight days. They hatched rather slowly, a 

 few at a time. 



The larvae (Figure, b) were extremely slender, not twice as broad behind as 

 at the anterior end, with well-marked segmental constrictions. The head is pro- 

 portionately large, with strong, acute mandibles projecting beyond the clypeal and 

 labial regions. The maxillae are furnished with a pair of prominent sensory 

 papillae and the labium with a well-developed duct to the spinning glands. The 

 dorsal surface of the head as well as the whole surface of the body is covered uni- 

 formly with short, slightly curved hairs. There are no traces of tubercles of any 

 description. Attempts to observe the method employed by the ants in feeding 

 their larvae were unsuccessful. Once, on placing a number of eggs and young 

 larvae of Canipojwtus festinatus in the nest, I saw the young Cerapachys larvae 

 feeding on the former after they had been carried under the slide by the workers. 

 It was apparent also that the ants and their older larvae soon began to feed on the 

 unhatched eggs and younger larvae of their own species, for the number of progeny 

 decreased rapidly from day to day. In order to reduce the size of the colony and 

 thus save as many of the young as possible from destruction, I killed and removed 

 twelve of the workers May 17th. I was not successful, however, in stopping the 

 infanticides, so on the following day I removed six more workers. By this time, 

 however, though I had provided the nest daily with fresh termite food (the ants 

 would not eat sweets, larvae of other ants, or miscellaneous insects) the Cerapachys 

 became so demoralized that by May igth no eggs and only five half-grown larvae 

 remained. These larvae were carried by the ants after the manner of Eciton and 

 Leptogenys, i. e. by the neck, with the long slender body extending back between 

 the legs of the worker. The ants were quite as careful of their larvae as of their 

 eggs. 



To my intense disappointment, it soon became manifest that I should be 

 unable to rear the few remaining larvae to the pupal stage. I therefore killed the 

 surviving workers and the three larvae still uneaten May 20th. Thus I was unable 

 to settle two important questions : first, the method of feeding the larvae, whether 

 by regurgitation or with pieces of insect food ; and, second, the character of the 

 pupa, whether naked or covered with a cocoon. The powerful development of the 

 larval jaws would seem to indicate that the young are fed with pieces of insect 

 food, and from the fact that the larval spinning glands seem to be well developed, 

 one may infer that the pupa is enclosed in a cocoon. Whoever is so fortunate as 

 to happen on a colony of these ants during the middle or latter part of June will 

 probably be able to determine the pupal characters without difficulty, as the pupae 

 should at that time be found in the nests. 



