Dec, 1908 1 Wheeler: Studies on Myrmecophiles. 207 



the flat superficial chambers there were also several of the earthen 

 cases of Coscinoptera dominicana containing pupae. Part of the schau- 

 fussi colony was confined in a Fielde nest together with the Microdons 

 and the Coscinoptera. The ants paid no attention to either of these 

 insects. By April 22 all the Microdon larva; had attached themselves 

 to the glass floor of the nest and had pupated. May 15 and 16 ten of 

 them hatched, and eight of them hatched on May 20 ; the six remain- 

 ing failed to develop. The flies emerged by breaking off the cover 

 formed by the tripartite anterior third of the dorsal portion of the 

 puparium and were at first of a dull drab color. Their bodies were 

 moist and their covering of pile was glued to the chitinous surface. 

 While in this stage they walked about among the ants without being 

 noticed or, at any rate, without being molested. After a few hours, 

 or sometimes much sooner, their wings had expanded and were folded 

 over each other on their backs, their integument had acquired the 

 adult metallic coloration and the pile had dried and become con- 

 spicuous. They remained very quiet except when rudely touched. 

 Then they emitted a loud humming sound, but without moving their 

 wings. This peculiar behavior has also been noticed by Bignell 

 (1891) in M. mutabilis. The humming attracted the ants, which at 

 once seized the defenseless creatures by the legs or wings and drenched 

 them with formic acid. This killed them almost instantly. After 

 seven had been dispatched in this manner, I removed the others from 

 the nest in order to save them. These observations show very clearly 

 that the adult M. tristis is treated with pronounced animosity even by 

 ants among which for many months it has been tolerated with indif- 

 ference as a larva and as a pupa. It is probable that the migration of 

 the full-grown larvae to the entrances of the nest and their pupation in 

 the superficial galleries are an adaptation to making it as easy as pos- 

 sible for the adult flies to escape to the open fields without being 

 noticed by the ants. All of my flies hatched at night and I believe 

 that this must be the time of emergence also out of doors. The 

 nights in May are apt to be cool, so that the ants would be in a torpid 

 condition or huddled together in the lower chambers of the nest, and 

 would not molest the flies while they were expanding their wings 

 and perhaps even mating and ovipositing in the nest. In several in- 

 stances, as in the case of the F. consocians nest mentioned above, I 

 have found empty puparia in nests containing partially grown Microdon 

 larvae. This would indicate that the flies are apt to oviposit in the 



