10 BULLETIN 1363, U. S. DEPARTMENT OE AGRICULTURE 



Table 3. — Host list of Compsilura eoncinnata, etc. — Continued 



Hosts 



May- 



June 



July 



Aug. 



Sept, 



Oct. 



Nov. 



Dec. 



Jan. 



Feb. 



Mar. 



Apr^ 



Pyrophila pyramidoides 







































Samia cecropia Linne" 



Schizura concinna Smith 









- 











Schizura unicornis Smith 

















Scoliopteryx libatrix Linne. 

 Sphecodina abbotii Swain- 



















I 

















| 









Sphinx gordius Stoll 



Stilpnotia salicis Linne 





- | 





















_ 

















































I 

















There is a great variety of insects among the summer hosts of 

 Compsilura. Three orders are represented and in the Lepidoptera 

 18 families are included. The species attacked are widely divergent 

 in habit as well as in external appearance. Apparently there are 

 no distinguishing characters peculiar to the host insect. Just what 

 means Compsilura uses to select its host is not known. It is possible 

 that it has no sense of discrimination. Experiments positively show 

 that year after year it will attack hibernating brown-tail moth 

 caterpillars without the least chance of survival. The same behavior 

 has been observed in its relation to Ol-ene basiflava. During the 

 winter of 1922 C. F. W. Muesebeck of the gipsy-moth laboratory, 

 while dissecting some of the overwintering larvae of 0. basiflava, 

 found a single hibernating larva of Compsilura. It seems certain 

 that the parasite would here meet the same fate as it does in its 

 attack upon the brown-tail moth, but since the writers have been 

 unable successfully to rear many of the overwintering larvae of the 

 host there are insufficient data to settle this point. It is probable 

 that in most cases the death of the parasite is due to the lack of cor- 

 relation between the host and the parasite ; that is, in the spring, the 

 development of the parasite exceeds that of its host, thereby result- 

 ing in the death of both (7, p. 220). Culver (gipsy-moth laboratory 

 records) found in his laboratory experiments that, although Comp- 

 silura would in many instances larviposit upon the last-stage cater- 

 pillars, the results were seldom successful. He found that the silk- 

 worm {Bombyx mori L.). when attacked in the advanced stages, 

 would often complete its cocoon and the parasites reaching the adult 

 stage would be unable to escape therefrom. He also noted larviposi- 

 tion on a chrysalid of Pontia ranae which achieved no results. There 

 are other records of attempted larviposition upon lepidopterous 

 pupae, all of which resulted in failure. In the laboratory experi- 

 ments, larvae of several species have been attacked, but no progeny 

 obtained. It is these facts which lead one to believe that Compsilura 

 uses but little discrimination in its choice of hosts and that it will 

 waste much effort in futile attack upon an unsuitable one. 



