156 JOURNAL OF ECONOMIC ENTOMOLOGY [\'ol. 9 



from which the animals were derived, Ij per cent died spontaneously 

 of wilt and 78 per cent of "other causes." 



Tables VII and VIII represent experiments similar to the one pre- 

 sented on Table VI with the exceptions that fourth stage American 

 caterpillars were used in experiment VII and Japanese third stage 

 animals in experiment VIII. 



Table IX represents an experiment performed with American fourth 

 stage caterpillars. Five animals were used in the experiments and 

 five in the controls. Not a single animal died from wilt. 



Tables X and XI represent passage infections. A caterpillar which 

 died in one of the previous infection experiments was ground up and 

 diluted to 10 c.c. with sterile water. This material was strained 

 through cotton, filtered through paper in a Buchner filter, and fed to 

 foreign fifth stage individuals represented on Table X. Twelve indi- 

 viduals were fed with the Buchner filtrate; 12 were fed with this 

 filtrate autoclaved and 33 untreated caterpillars accompanied the 

 series. The results can be gathered by consulting the Table. 



Table XI represents the next passage. A caterpillar that died in 

 the previous experiment was used. The material was treated in the 

 same way, i. e., ground up, diluted to 10 c.c. and filtered. Fifth stage 

 foreign caterpillars were used. 



Discussion of the Infection Experiments 

 The following data derived from a comparative study of the mor- 

 tality tables are worthy of discussion: The number of caterpillars 

 which died of wilt in the experiments and checks; the relation between 

 the percentage of wilt mortality in the stock cultures and the percent- 

 age of wilt in the checks; the number of caterpillars which died of 

 "other causes" in the experiments and checks; the relation between the 

 percentage of "other cause" mortality in the stock cultures and the 

 percentage of deaths due to that cause in the experiments and checks; 

 the number of moths obtained in the experiments. 



In the tables the wilt mortality of the experiments equals 68 per 

 cent; in the checks 4 per cent. How can we explain the 4 per cent wilt 

 in the checks? By comparing the wilt mortality in the checks with the 

 wilt mortality in the stock culture in each table, we find that the con- 

 dition of the stock usually explains the wilt mortality in the checks. 

 Thus in Table VI none of the checks died of wilt and the wilt mortality 

 ill the stock was very low (1| per cent). In Table VII we have two 

 cases of wilt in the checks and the mortality in the stock is rather 

 high (9 per cent). In Table VIII one check died of wilt and the mor- 

 tality in the stock is high (11 per cent). In Table IX no wilt is re- 

 corded and none in the stock. In Table X none is recorded in the 



