430 Journal of Agricultural Research voi. xvui, No. s 



remark that throughout the entire vegetative development of the fungus 

 on this medium nothing resembling hyphse was observed, the entire 

 process having been exclusively yeastlike. 



Several months after these tests were conducted a fresh supply of 

 this agar was made up, and on it the fungus under consideration de- 

 veloped in a manner in all respects similar to that described above. 



The artificial cultivation of Sorosporella has shown, therefore, that it 

 can be successfully cultivated on a variety of media, and that while its 

 normal method of vegetative development within infected insects is by 

 means of yeastlike budding cells, this habit may be modified by culti- 

 vation to such an extent that a semifilamentous growth may be acquired. 

 The cultures also show that when the resting spores reach maturity be- 

 fore germination — a condition that is never accomplished when the toru- 

 loid or filamentous habit is followed but which does occur as a result of 

 growth of the disassociated blastocysts on certain nutrients — fascicles 

 of conidiophores arise which recall similar fascicular growths of other 

 insect fungi, such as those of the poorly defined genus Isaria. * It should 

 be noted furthermore that no perfect or acigerous stage has been observed 

 in any of the artificial cultures. 



INOCULATION EXPERIMENTS 



Although the morphological characters, particularly of the resting 

 spores, had been studied to some extent prior to the writer's preliminary 

 work with the fungus in question (20) , attempts to inoculate insects arti- 

 ficially, either in the laboratory or in the field, -were rarely recorded, in 

 spite of the fact that in Russia at least the disease was known to occur 

 under natural conditions out of doors in such abundance and in such a 

 manner as to suggest its infectiousness. 



It was deemed advisable, therefore, to perform a series of inoculation 

 experiments to determine, if possible, something of the range of suscepti- 

 ble hosts and the pathogenicity of the organism, as well as to determine 

 a method of infection that could be used in a practical way to inoculate 

 large numbers of insects. 



The successful artificial culture of the fungus greatly facilitated the 

 work of conducting these tests, for a constant supply of viable conidia 

 was at all times available. Since it was determined that insects could be 

 readily infected by the fungus from artificial cultures, the latter were 

 invariably used. 



It has been shown in the foregoing pages that the vegetative develop- 

 ment of the fungus takes place exclusively within the blood of its insect 

 hosts; and it is obvious that to reach the body cavity, the body covering, 

 the tracheae, or the intestine must be penetrated by the conidial germ 

 tubes. As is well known, insects are covered externally by a layer of 

 chitin, which in certain regions, as for example between the body seg- 



