ii8 



Journal of Agricultural Research 



Vol. IV. No. 2 



in a dark field, except minute dancing granules, which may be identical 

 with those mentioned under ' ' Pathology of wilt ' ' as having been observed 

 in diseased tissue. 



In the first experiment lo caterpillars were fed with the i to 25 dilution 

 passed through a Berkefeld filter, and 10 were fed with the i to 50 dilution 

 passed through another Berkefeld filter. Twenty controls accompanied 

 this series. 



Ten caterpillars were fed simply with the unfiltered i to 25 dilution, 

 while 10 were fed with the i to 50 unfiltered dilution. Fifteen controls 

 accompanied this series. 



The controls were treated in exactly the same manner as the other 

 individuals, except that they were fed wnth material that had been 

 sterilized by autoclaving. The infectious material was administered to 

 the caterpillars in two ways : Either some of the liquid 

 was sucked up in a sterile eye dropper, by means of 

 which a drop was placed over the caterpillar's mouth, 

 or the oak leaves were submerged in the material 

 before being placed in the trays. The first method is 

 advantageous, for the reason that a large quantity of 

 the virus can be administered. The caterpillar is 

 held upside down and a drop is placed directly over 

 its mandibles. These and the maxillae soon begin to 

 move, and gradually the drop disappears. If cater- 

 pillars are to be infected in this manner, it is neces- 

 sary to avoid the period directly preceding a molt, as 

 they will not drink until this process has been com- 

 pleted. Furthermore, if caterpillars have eaten shortly 

 before one wishes to infect them, it is difficult to make 

 them drink; consequently the animals were usually 

 starv^ed for a few hours before infection. The second 

 method is also good, and has the advantage of not 

 consuming as much time as the first. One can be 

 certain whether or not they have eaten by examining the old, infected 

 leaves before replacing them with fresh foliage. In the first experiment 

 both methods were used, and since none died up to June 25, the insects 

 were infected again on that date in exactly the same manner as before. 

 In order to determine the effect of temperature on the incubation 

 period,^ it was recorded daily at noon. 



Figure 2 shows the mortality of the 20 caterpillars fed with the 

 Berkefeld filtrate. The ordinate represents the range of temperature 

 between June 18, when the experiment was begun, and July 27, when it 

 was concluded. In order to avoid unnecessary space, the abscissa rep- 

 resents the day when the experiment was begun, the day on which 

 caterpillars died, and the day on which the experiment was concluded. 



Fig. 2. — Curve showing 

 the mortaUty among 

 20 gipsy-moth cater- 

 pillars fed with the 

 wilt virus after filtra- 

 tion through the 

 Berkefeld filter. 



' The term "incubation period" is used here as the time elapsing between infection and death. 



