152 JOURNAL OF ECONOMIC ENTOMOLOGY [Vol. 9 



taken to the laboratory, and isolated from one another in autoclaved 

 pasteboard or sterile tin boxes. Thus by isolation, all animals that 

 harbored the disease and died before being used for experimental 

 purposes were prevented from infecting others. This precaution was 

 found necessary for the reason that it is impossible to determine dur- 

 ing the early stages of the disease, whether or not a caterpillar is free 

 from wilt infection. Of course Escherich's blood test method for 

 eliminating chronic cases might have been used, but Glaser and 

 Chapman (1913) and Glaser (1915) have shown that the blood test 

 is impracticable for large experimental series. 



Caterpillars dead from wilt were emulsified with sterile water. 

 This material (amounting to about 100 c.c.) was filtered through 

 cotton and later through paper. The filtrate thus obtained was 

 passed through a Berkefeld Grade "N" filter by the use of a vacuum 

 of approximately 28 inches. This Berkefeld filtrate was used for the 

 infection experiments. Of course, the filtrate was tested for its bac- 

 terial sterility by plating on ordinary nutrient media. If the proper 

 technical precautions are taken the filtrate obtained will be free from 

 bacteria and polyhedral bodies. 



The caterpillars were infected by holding the animals with their 

 ventral side up. By means of a sterile eye dropper a drop of the fil- 

 trate was then placed directly over the mandibles. By exercising 

 sufficient patience the animals can be made to drink the drop or even 

 two or three large drops. It must be borne in mind that caterpillars 

 will not drink prior to moulting, so it is important to obtain them after 

 moulting or two or three days previous to this act. Moreover, cater- 

 pillars which have recently fed are unwilling to drink and therefore 

 it is best to starve them for 24 hours before attempting an infection. 



Every caterpillar in our experiments was infected with the same 

 amount of the virus. Suffice it to say, that we prepared the virus 

 in the same way for each experiment, and further attempted to have 

 conditions as constant as possible. 



Table I gives the results obtained in the first experiment. Twenty- 

 five caterpillars were infected with the Berkefeld "N" filtrate and 25 

 controls were infected with the same filtrate sterilized by autoclaving 

 for 20 minutes. Eighteen caterpillars died of wilt in the experiment 

 and 19 in the controls. One individual escaped; one died of another 

 cause (possibly bacterial infection), and 11 moths were obtained. 

 Table II is self-explanatory. Tables III and IV represent similar 

 experiments with the exception that the material was passed through 

 the finer Grade "W" Berkefeld filter. By an examination of Tables 

 I and II it will be seen that the deaths due to wilt in the controls ex- 

 ceed those in the experiments. In Table III the deaths in the experi- 



