420 Journal of Agricultural Research voi. xviu, No. s 



smears. However, although the conidiahad disappeared, no evidence of 

 their incorporation by leucocytes was obtained in spite of the fact that 

 these bodies were quite abundant. A few insects so treated were not 

 subsequently punctured to get blood smears but were left unmolested, 

 and after a period of two weeks they were still alive. 



It is not possible to say whether the conidia were destroyed by the 

 blood antibodies, or whether they were taken up by fixed cells, which of 

 course would not be seen in blood smears; but in any event their disap- 

 pearance is suggestive of some such action, and the fact that certain 

 control insects similarly treated survived further indicates that the fun- 

 gus cells were in some way rendered impotent. 



On the other hand, when a large dose of the fluid was injected into silk- 

 worms and white grubs, first conidia and then blastocysts could be 

 detected in the blood in due time after injection. The blastocysts ap- 

 peared within one or two days after injection; and so far as could be 

 determined they reproduced in a way identical with that in susceptible 

 hosts. Similarly the act of phagocytosis was recognized; but, as in 

 susceptible hosts, the phagocytes seemed unable to cope with the fungus, 

 the vegetative bodies of which were formed in enormous numbers, 

 furthermore the blastocysts rounded up and ultimately formed typical 

 resting-spore aggregations. 



In order to determine whether or not there is at first an active phago- 

 cytosis in infected susceptible hosts, conidia of Sorosporella were in- 

 jected into specimens of the semitropical army worm (Prodenia erida- 

 nia Cram.) in small and large doses in a manner similar to that de- 

 scribed above. 



Blood smears were made from such insects at periods varying from 

 two hours to three days after injection. While it was possible to detect 

 both the fungus cells and the phagocytes in all stained blood smears 

 whether or not such smears were made from insects into which a larger 

 or smaller amount of the fluid was injected, phagocytosis was not observed 

 within two days after inoculation although careful search for infected 

 leucocytes was made. On the other hand, smears made two or three 

 days after injection showed blastocysts incorporated in the phagocytes, 

 although no signs of disintegration of the former could be detected. 

 Furthermore, it is certain that the fungus cells were reproducing rapidly 

 in the blood, for specimens in the 2-celled stage of division, such as are 

 shown in Plate 52, I, were in great abundance. 



A detailed discussion of Metchnikoff's discovery of phagocytosis in 

 • Daphnia, a crustacean, and his subsequent formulation of the theory of 

 phagocytosis as an explanation of immunity in man and other animals is 

 not within the scope of the present paper, but it seems advisable to 

 consider such phases of this and other theories as may have a bearing 

 on the problem under consideration, 

 \ 



\ 



\ 



