No. 468.] THE. WILLOW CONE GALL. 865 
two gnats emerged about the middle of February. From those 
of March 20.several gnats emerged early in April. From the 
April specimens they began to appear within a week and con- 
tinued to emerge for over three weeks. Because of the difficulty 
in rearing gall gnats my data are not complete. But it is evi- 
dent that the insect exists for a long time in the larval state and 
that thé pupal period is comparatively short. 
Walsh found it difficult to devise an artificial mode of treat- 
ment for rearing the gall gnats in the house. My results with 




2 MO M Le iE 

Fi. 3.— Gall gnat (Rabdophaga strobiloides O. S.) enlarged. 
the earlier lots of galls were not very successful. At first I 
threw them in jars and boxes but found that in a short time 
larvae and pupz dried up in many of the galls. Then I tried 
keeping them with the stems in water but with scarcely: better 
success. With the last lot I used a different plan. 1 tied the 
galls in squares of cheese cloth, opening the bundles and sprink- 
ling them every day or two. This plan was very successful. 
Of course it is possible that the season had something to do 
with it, for it was nearly the regular time for the emerging of 
the gnats. But that cannot be the full explanation, for, in pack- 
ages which were neglected, the results were less satisfactory. 
The sprinkling seems to have taken the place of the spring 
showers. | ; 
Cecidomyiids offer considerable difficulty in the matter o 
