Jan. 15, 1920 Further Studies of Sorosporella uvella 437 



instances until the small bit of leaf was entirely consumed. It is there- 

 fore possible to say that the fungus did not come in contact with any 

 external portion of the body except the mouth parts, and the first and 

 second pair of legs, which grasped the leaf, and held it edgewise to facili- 

 tate feeding. All conditions under which the experiments were con- 

 ducted were carried out in such a manner as to preclude the possibility 

 of infection from other sources ; and since the fungus at no time came in 

 contact with the body except in the manner stated above, it is obvious 

 that infection was brought about through the ingestion of food. 



During the summer of 191 8 a number of inoculations were made in 

 this manner. The hosts used were largely cutworms of various species, 

 chiefly Prodenia eridania Cram. ; and a death rate was obtained in all 

 instances that was as high as in those inoculations which were made 

 by the direct method. A detailed account of one experiment may be 

 given. Twenty-five army worms were fed on August 7 with clover 

 leaves which had previously been smeared with conidia from an agar 

 culture. As a control, 15 larvae of the same lot were fed with fresh, 

 unsmeared leaves only. On August 10, 3 of the infected specimens 

 and on August 12, 2 of the control larvae died from unknown causes. 

 On August 16, however, 10 of the infected army worms were dead, in 

 all of which the fungus was recognized; and 3 were dead from other 

 causes. On August 15, 2 and on August 20, 3 larvae died in the control 

 dishes, but apparently not from the Sorosporella disease. On August 

 17, 4 and on August 19, 2 of the infected insects died from the disease, 

 and I died from which the fungus was not recovered. The experiment 

 was closed on September 3, when 2 larvae of the infected lot and 8 of the 

 control lot were alive. In spite of the fact, however, that the conidial 

 germ tubes must obviously pass through the intestinal wall, it has been 

 impossible actually to observe them in prepared stained sections. 



SUMMARY 



(i) The presence of Sorosporella uvella, an entomogenous fungus, is 

 recorded for the first time in America. 



(2) The previous association of Sorosporella with the Entomophtho- 

 rales is shown to be erroneous, and the proper position of the organism, 

 among the verticilliacious Hyphomycetes, is designated. 



(3) The reproductive bodies are thick-walled resting spores or chlamy- 

 dospores and thin-walled conidia, the latter being herein definitely 

 associated with the life history of the organism for the first time. 



(4) It is shown that yeastlike vegetative cells, existing within the 

 blood of infected insects, are ontogenetically related to other phases in 

 the development of the organism. 



(5) There is an ingestion of these vegetative cells by certain of the 

 blood corpuscles (phagocytosis), the process being apparently followed by 



