19 



from the mean in both directions. The main purpose of the 

 experiments, however, was not to make mathematically exact 

 calculations, but to find out in what manner and to what degree 

 the gypsy moth may be infected by the wilt. For these pur- 

 poses the calculations made should be sufficient, and the figures 

 given will not be far from the real average. 



As to the control experiment, the average of eggs in a normal 

 cluster was found to be 437. I am not aware that the eggs of 

 a large series of normal clusters have ever been counted in order 

 to establish an accurate average. In published reports the aver- 

 age of eggs in a normal cluster is stated as " 400 to 500." 



Control Experiment, Concord, Mass. 

 Five clusters were taken in the western part of this locality, 

 where there certainly did not occur any disease among the gypsy 

 moth caterpillars. These clusters, therefore, must be regarded 

 as normal. The microscopic examination of the eggs of these 

 clusters gave the following results : — 



Unfertilized eggs, . 

 Eggs with dead embryos, 

 Eggs apparently alive. 



Total, 



Unfertilized eggs, . 

 Eggs with dead embryos. 

 Eggs apparently alive, . 



Total, 



Unfertilized eggs, . 

 Eggs with dead embryos, 

 Eggs apparently alive, . 



Total, 



Unfertilized eggs, . 

 Eggs with dead embryos, 

 Eggs apparently aUve, . 



Total, 



Cluster 1. 



o\— -73 per cent., or about 1 per cent. 

 643 =99.27 per cent., or about 99 per cent. 



547 eggs. 

 Cluster 2. 



„ > = .47 per cent., or about .5 per cent. 

 424 =99.53 per cent., or about 99.5 per cent. 



426 eggs. 

 Cluster S. 



= 1.74 per cent., or about 2 per cent. 

 395 =98.26 per cent., or about 98 per cent. 



402 eggs. 



1 



Cluster 4- 



= .93 per cent., or about 1 per cent. 

 424 =99.07 per cent., or about 99 per cent. 



428 eggs. 



