ARTIFICIAL INFECTION 77 



c. The characteristic disease should be produced by infection 

 experiments with a pure culture. 



d. The fungus associated with the disease induced should be 

 identified as the one originally separated, ' and any abnormalities of 

 host should likewise correspond. 



The purposes of the infection experiments as above outlined 

 are merely suggestive, and it is evident that a single series of 

 experiments may give all or nearly all of the indications desired. 

 Each subdivision, however, deserves special consideration. 



i. With such obligate parasites as the Peronosporaceae, Exoas- 

 caceae, Erysiphaceae, Ustilaginales, Uredinales, and some others, the 



FIG. 16. CORRECT USE OF BELL GLASSES IN CERTAIN TYPES OF INFECTION 

 WORK. (Photograph by Geo. M. Reed) 



constant association of an organism with a diseased condition would 

 usually be sufficient to denote this organism as the cause of the 

 disease. The conditions required for spore germination are in many 

 cases unknown, and therefore negative results would be of no great 

 value. There is therefore a two-sided opportunity for study. It 

 would, however, be absurd to say that Empusa Muscce is not the 

 cause of the well-known, or commonly observed, fly disease. Yet, 

 so far as the writer is aware, no work has been done which would 

 be counted as successful in the artificial propagation of such dis- 

 eases among insects. With most groups of fungi and with the 

 bacteria, infection experiments must be made if the work in hand 



