14 Bulletin American Museum of Natural History. [Vol. XXIII,. 



very similar in shape to tho.se of Ph. dentata above described. This phthiser- 

 gate had neither hatched nor changed its appearance by iVugust 29 — 15 

 days later. It eventually disappeared (eaten by the workers?). The 

 colony was kept for several weeks but as no second brood of Orasema 

 appeared, it was abandoned. August 20 I found near Beaver Ranch, south- 

 east of Colorado Springs, a third infested vinelandica colony. This two was 

 kept in an artificial nest, but the observations revealed nothing that has not 

 been recorded for Ph. instahUis and 0. virldis. According to Forel, Ph. 

 vinelandica occasionally produces intermediates between the typical soldiers 

 and workers, but although I have collected many colonies of this ant in 

 different parts of the United States, I have seen only the typical soldiers 

 and workers. 



The foregoing observations prove that 0. coloradensis, like 0. viridis, 

 does not confine its depredations to a single host ant. It may be said to be 

 even more catholic in its habits, since it not only infests a species of Pheidole 

 but also attacks ants belonging to the very different genus Solenopsis. The 

 number of colonies observed was not sufficient to show which of these genera 

 represents the original and more frequent host, but, judging from analogy 

 with 0. viridis and the next species to be considered, there can be little 

 doubt that this is Ph. vinelandica. 0. coloradensis probably occurs also in 

 the nests of Ph. pilifera var. coloradensis and Ph. ceres, both common ants 

 in the neighborhood of Colorado Springs and other localities of about the 

 same altitude. 



C. Orasema wheeleri Ashmead. 



A third species of Orasema, which Dr. x\shmead will describe as 0. 

 wheeleri, was found July 14, 1902, on Keesy's Ranch at Fort Davis, Texas, 

 in the nest of Pheidole ceres Wheeler. This ant is a small, dark colored 

 species which makes obscure nests in the shade of the cotton-woods {Pop- 

 ulus fremonti) and garners the woolly seeds of these trees. The Ora- 

 sema was seen in only one of the nests, and though about a dozen pupte 

 were found among the ant brood, only a single mature specimen, which 

 happened to be a female, could be secured. It measures 2.7 mm. and 

 closely resembles 0. viridis, especially in the shape of the scutellum and 

 epinotum, but the thorax is broader and stouter. The last pupal envelope, 

 like that of viridis, is without pustules, though it has strong intersegmental 

 welts in the abdominal region. 



I have examined many colonies of Ph. ceres near Fort Davis and in 

 various Colorado localities without finding any intermediates between the 

 soldiers and workers. In the single colony infested with 0. wheeleri, how- 



