434 Journal of Agricultural Research voLxvni, no.s 



to crawl over the fungus thallus for several hours, the conidia may have 

 been ingested in some numbers; and in such instances the ingested 

 conidia may have produced the disease; but on the other hand, in those 

 instances in which the fungus was applied to the external parts of the 

 body only, it seems necessary to believe that infection was produced by 

 penetration of the body wall by the conidial germ tubes. 



The larvae of all species of cutworms were quite well developed when 

 inoculated, pupating in several instances in a short time. It is worthy 

 of note in this connection that a number of the army worms employed 

 were in the pupal stage at the time of death, and that a few imagos 

 emerged, lived a day or two, and then died of the disease. The occur- 

 rence, therefore, of typical resting spores in imagos of the army worm is 

 significant and indicates that the organism, though unable to kill the 

 larvae and pupae, passed through the various metamorphic changes of 

 the host and finally caused death after it emerged. The presence of the 

 fungus in the flying adults is furthermore significant, since it suggests the 

 possibility of dissemination of the organism. It should be noted, how- 

 ever, that the wings of the four adults from which the fungus was 

 recovered were imperfectly formed, and other organs were malformed 

 or missing. An antenna of one, for example, was entirely lacking, and 

 a portion of the leg of another was absent, indicating that during the 

 metamorphosis certain of the imaginal tissues were destroyed. 



During the summers of 191 7 and 191 8, 20 tests were performed by the 

 direct contact method, in which many insect species were used; and 

 while it is not desirable to discuss all of them, a few may be considered. 



On April 9, 191 7, 26 larvae of Feltia jaculifera were inoculated by 

 allowing them to remain in the culture flasks for 12 hours. Thirteen 

 died of Sorosporella on April 20, 8 on April 23, and 2 on April 24. On 

 May I when the experiment was closed 3 were alive. There were 12 

 larvae in the control dishes, i of which died from Sorosporella on April 20. 

 The others were alive on May i . This larva is the only one that died of 

 the disease in control dishes during the course of all of the experiments. 



On August 3, 100 army worms were inoculated by being allowed 

 to crawl over the fungus for two minutes or less. On August 6, 8 

 larvae were dead and on August 10, 9 more, from some other cause 

 than Sorosporella. On August 13, however, the fungus was recovered 

 from 9 insects. Metarhizium had killed 2, and 14 had died from unknown 

 causes. On August 15, Sorosporella was found in 20, and 2 were dead 

 from other causes. On August 16, 13 more were dead from Sorosporella; 

 on August 20, 7 emerged, and there were 5 dead from which Sorosporella 

 was not recovered. Finally on the date the test was closed, August 23, 

 4 adults were found infected with the fungus, as well as 7 pupae, all of 

 which were filled with resting spores. Of the 25 control larvae, 5 died 

 from unknown causes and 20 emerged. 



In another test, May 9, 15 larvae of Feltia sp. were inserted in as 

 many salve boxes in which were placed moist sand and a bit of fungus 



