Jan. IS. 1920 Further Studies of Sorosporella uvella 435 



thallus. On May 14, i larva died from some unknown cause, but on 

 May 21, 12 larvae died of the fungus. On the two succeeding days 2 

 other larvae died of the same cause. In the control dish in which 10 

 larvae were inclosed, 9 were alive on May 23 and i had died from an 

 unknown cause. 



As an example of those inoculations in which the larval bodies were 

 rubbed with a bit of the fungus, an experiment begun on June 29 with 

 23 larvae of Noctua c~nigrum may be cited. Thirteen of these were 

 inoculated and 10 were used as controls. On July 6, 3 larvae died in 

 which the fungus could not be detected. On July 13, 14, and 15, how- 

 ever, 5, 3, and 2 insects, respectively, had died from the disease; and at 

 the close of the test on July 22, 7 larvae were alive in the control dish, 3 

 having died before July 1 5 from unrecognized causes. 



In addition to a large number of similar tests that were conducted 

 with noctuid larvae of various species the direct contact method was also 

 used in inoculating house-fly larvae, white grubs, wireworms, nymphs 

 and adults of grasshoppers, and white ants, as well as the larvae of 

 Hyphantria textor, Bombyx mori, and Phlegethoniius sexia. The insects 

 in all cases were placed in culture tubes or flasks of the fungus and allowed 

 to remain therein for from 3 to 4 hours, after which they were removed 

 and placed in boxes with suitable food. 



The house-fly larvae, for example, when treated in this manner were 

 found to be covered with conidia when they were removed from the 

 cultures, yet after they had burrowed in the dung of the rearing boxes 

 for a few hours no conidia could be detected with the low power of the 

 microscope. While several specimens, particularly of Hyphantria textor 

 and house-fly larvae, died from unknown causes, the fungus was not 

 recovered from a single inoculated insect of any of the species enumerated 

 above. 



The experiments with the susceptible hosts show, however, that under 

 laboratory conditions a high percentage of mortality may be realized 

 and, furthermore, that the death rate is not appreciably higher when 

 larvae are kept in contact with the fungus for a long time than when 

 they are subjected to infection for a minute or two. 



The length of time necessary for the fungus to kill nearly mature 

 cutworms is rarely less than 10 days; and in certain instances it was longer 

 than 10 days, a possible explanation for which is given on page 417. 

 This last summer, however, smaller larvae in the second and third instars 

 were inoculated for sectioning purposes, and it was found that such 

 insects often succumbed to the disease in 6 or 7 days. 



SPRAY METHOD 



Ten cc. of sterile water were poured into a culture flask, which was 

 then shaken until a quantity of conidia were in suspension. The liquid 

 was then sprayed upon healthy larvae by means of a small hand atomizer, 



