46 
THE GRAPE ROOT-WORM. 
on a large scale, so that variations would be minimized and the final 
averages would represent approximately normal conditions. The 
numerous separate experiments have involved the handling of a lar°-e 
bulk of rearing material, which, together with the simulation of nor- 
mal conditions, has to some extent necessitated special rearing 
devices and methods of handling. The experiments have been con- 
Fig. 20.— Earthen pot with glass cylinder used in rearing the grape root-worm. (Original.) 
ducted either in the field or under an open breeding shelter, a portion 
of which is shown in figure 18. This consisted of a temporary struc- 
ture of light wooden framework covered with waterproof canvas. 
Most of the rearing material was obtained in the spring, some time 
previous to the transformation of the larvae. During the past two 
years of the investigation the insects were to some extent reared 
