140 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 56. 



the females after being examined. A larva of Naemia seriata, how- 

 ever, that had been rejected on July 26 was attacked and parasitized 

 on the following day, after it had been kept over night m a vial 

 with larvae of Ooccinella o-notata, m the meantime evidently having 

 absorbed some of the more attractive odor' of the latter species. 

 Another Naemia larva was rejected again, however, after being kept 

 with larvae of this Coccinella. 



Although the larvae of Coccinella o-notata and difficilis were imme- 

 diately attacked when exposed to the Hornalotylus, the results of 

 parastism on the whole suggested the great probability that neither 

 of these species was a customary host. Thus, out of the four larvae of 

 5-notata which were parasitized only one produced Homalotylus, two 

 transformed into beetles, and one died in the pupal stage and showed 

 no sign of parasitism when dissected. Only two larvae of dijfidlis 

 were used in the experiments, one dymg in the larval stage, the other 

 in the pupal stage, without signs of parasitism m either case. 

 On the other hand, the single larva of Naemia which was parasitized 

 produced Homalotylus in due time, but Naemia does not occur in 

 Utah and the species succumbed more because of weak resistance 

 perhaps than because of adaptation to it on the part of the parasite. 

 A similar case was observed at Whittier, California, m the summer of 

 1912, when the same species of Homalotylus was reared from Cheilo- 

 menes sexmaculatus (Fabricius), an Indian species of Coccmellid, 

 which the writer was attempting to establish m California at that 

 time. Two parasitized larvae of this beetle M'^ere found on a vrahmt 

 tree where the Coccinellids had been liberated previously and on 

 which the larvae of Olla ahdominalis v/ere also abundant, the latter 

 species probably serving as the usual host for the parasite. 



These observations and experiments have led the writer to believe 

 in the existence of physiological races as noted above. Homxilotylus 

 has been reared also from Chrysomelid larvae on at least two different 

 occasions, once in Europe as recorded by Francis Walker, and once 

 in North ^imerica by George Dimmock. In the light of the above 

 experiments it does not seem probable that a Homalotylus commg 

 from a Coccinellid host would attack or at least successfully parasitize 

 a Chrysomehd larva, when even closely related species of CoccineUids 

 may be rejected or found resistant to parasitism. The forms parasit- 

 izing Chrysomelids therefore are probably extreme physiological 

 races, although in other respects they appear to be identical with the 

 ordinary forms. 



Because of the difficulty in decidmg about the identity of colora- 

 tional forms the writer resorted to a study of the male genitaha m the 

 hope of finding useful contributary characters, but without satisfac- 

 tory results. Although mmute differences could be detected they 

 did not seem to present enough constancy among themselves nor any 



