PAPILIONmAE: IPIIICLIDES AJAX. 1279 



self, that mounts upon it and from a puncture extracts the contents. I 

 frequently met the shells so despoiled before I discovered the cause, and 

 have since observed the marauder in its operations. I have also lost in a 

 single night, owing as I supposed to crickets, numbers of eggs laid in con- 

 finement." Speaking of the caterpillar he says : "I have . . . seen spiders 

 feeding upon them, attacking even the head, and they have other enemies 

 among the insects. They are very little troubled liv ichneumon flies in 

 this valley, and I have rarely lost a chrysalis from that cause. Conse- 

 quently no [swallow-tail] is so abundant here throughout the season. I 

 find on breeding them that a considerable percentage of the eggs do not 

 hatch, and that more or less of the larvae die at every moult, as well as in 

 the effort to change the chrvsalids. Multitudes of chrysalids must be 

 destroyed in the winter by birds and mice, as they are but imperfectly con- 

 cealed under stones and I'oots, or even among the stems of the grasses, so 

 that of the tens of thousands of eggs that are annually deposited, but a 

 very small proportion produce butterflies." The caterpillar is, however, 

 sometimes attacked by Trogus exesorius BruUe (88: 3) the imago of which 

 escapes from the chrysalis by cutting a circular opening, usually through 

 one of the wings ; also by Exochilum mundum, according to Mundt ; and 

 by Pimpla annulipes, as discovered by J. B. Smith, all three hymenop- 

 tera of pretty large size. No dipterous parasites are known. 



Desiderata. Although so carefully studied by Mr. Edwards, there 

 are still some points in the history of ajax which require investigation. 

 The distribution of the insect in the west and north should be more defi- 

 nitely determined. The season of the apparition of the different varieties 

 in the extreme southern states, and of the different broods of ajax every- 

 where, is still unknown and will require careful study : but perhaps the 

 most interesting and fruitfid investigation will be to follow still further the 

 line of Mr. Edwards's experiments, and study the proportion of chrysalids 

 of each brood which retain their inmates imtil spring : noting every instance 

 of the partial retention of the chrysalis, to discover to what extent pupae, 

 apparently destined to hibernate, disclose the butterfly the same season ; 

 and, further, to determine whether both marcellus and telamonides are in- 

 differently produced from any of the broods of the previous year. Mr. 

 Meldola's studies would lead us to conjecture that marcellus is generally 

 produced from the later broods of ajax, and telamonides from the earlier 

 broods of the same, and from telamonides and marcellus ; but Mr. 

 Edwards's experiments show that this is not invariably the case. The pos- 

 tures of the butterfly have not been described. 



