4 BULLETIN 1080, U. S. DEPAETMEXT OF AGEICULTUKE. 
of 1915-16. Ten clusters were collected at each point on the first of 
each month, beginning in September and ending in April. Subse- 
quently they were handled exactly as described. 
As the sets of weather instruments were located at three of the 
collection points, the records obtained from these are the most valu- 
able. The other three, however, were kept as collection points for 
purposes of comparison. 
MONTH'S EXPOSURE. 
As controls on the naturally produced egg clusters obtained in 
"monthl}^ collections," a series of experiments were conducted with 
clusters produced at the laboratory. To obtain these, vigorous males 
and females were mated in a small cage, and later the females were 
confined in small tin boxes, Avhere they deposited their eggs upon 
pieces of thin wood. 
Ten such clusters were tacked to trees at each point where there 
were weather instruments and allowed to remain in the open just 
one month. The first set was out from September 1 to October 1, 
the second from October 1 to November 1, and so on. The last set 
placed out April 1 was brought in just before hatching time. After 
exposure the clusters were handled as in all the other experiments. 
NATURAL PROTECTION. 
Notes taken in connection with the " observation-point " investiga- 
tions showed that in certain localities there was considerable differ- 
ence in hatching on different parts of the trees. Clusters found un- 
der roots or in cavities close to the ground, as well as those fully 
exposed close to the ground, appeared to hatch completely, while 
those higher up failed to do so. 
Investigation at a number of chronic nonhatch points proved that 
in general it would be rather difficult to find enough clusters naturally 
deposited where they were wanted. Therefore, at each of the six 
points from which monthly collections were obtained 100 new clusters 
were cut from the trees and tacked back in the following positions : 
Twenty-five clusters were tacked under roots or in cavities close 
to the ground. Usually these were covered with leaves. 
Twenty-five at the base of a tree close to the ground but entirely in 
the open. 
Twenty-five on the trunk well up from the ground, usually about 
5 feet. 
The remaining 25 were brought in to serve as controls. 
Collection of these clusters was made just before hatching time 
each spring. 
ARTIFICIAL PROTECTION. 
Originally these experiments were planned when the fungus found 
in the nonhatch eggs was looked upon as a possible cause. Various 
means were tried of protecting the clusters against infection, but 
most of these were abandoned when the true nature of the fungus 
became apparent. 
Two experiments in artificial protection were continued. The 
first, by means of a wire cage suspended between trees, sought to 
protect the clusters contained therein from any influence the tree 
