HATCHIXG OF GIPSY-MOTH EGGS. 7 
Collections made at the nonhatch points up to February 1, 1917. 
the only month havin<r severe cold that winter, hatched completely, 
while those obtained March 1 and afterwards were all nonhatch. 
Those laboratory-bred clusters exposed during the month of Febru- 
ary were all killed, but those exposed every other month hatched 
completely. The next winter severe cold came in December, Janu- 
ary, and Februar}^ Collections made up to December 1 gave com- 
j^lete hatch, while neither those collected after that date nor the 
month's exposure clusters for the three cold months hatched. The 
next winter. 1918-19. was mild, and all clusters for all months 
hatched completely. The winter of 1919-20 was another ver}^ cold 
one, with the extreme cold coming in January and February, result- 
ing in hatch records identical with those of the other cold periods 
for both monthly collections and month's exposure. 
During the entire four winters only tAvo drops greater than 
— 10° F. were recorded from the point which had never shown non- 
hatch. The first, in December, 1917, —15°, resulted in about 50 per 
cent of the eggs in the monthly collections and month's exposure 
clusters being killed. The second drop, — 16°, came in February, 
1920, with the same effect on the eggs. 
The method employed in determining the extent of injury to the 
eggs in the various experiments ma}^ need explanation. In the 
first place, care was taken to provide plenty of control clusters for 
all the experiments and these were treated exactly like those for 
which they were controls, except that they were not exposed to con- 
ditions at the various points. Clusters Avhich did not hatch at all 
could be listed as complete nonhatch. However, in figuring per- 
centage of hatch in partially hatched clusters, the number of infer- 
tile eggs was deducted first, as they could not possibly have hatched. 
A careful comparison of the temperature and nonhatch records 
has led to the conclusion that —20° F. is about the highest point 
at which all the eggs in a cluster will be killed and a further drop 
will make it more certain that all exposed eggs will be killed. There 
is reason to believe that the resistance of the eggs depends to some 
extent upon their vitality, as would be only natural. Evidence on 
this point was secured from the sets of clusters which were only 
exposed to a single severe drop in temperature. One set exposed 
to —22° F. was entirelv killed, but a verv few eggs in a set exposed 
to —23° F. hatched. 
There are no records of any eggs hatching after being exposed 
to a temperature of —25° F. or lower, whether they Avere subjected 
to one such exposure in the " single drop experiments "' or several, 
as sometimes happened with the '' month's exposure " sets. On the 
other hand, exposures to —15° and —16° F. killed half the eggs in 
the clusters, but no temperature above —20° F. killed an entire 
cluster. Various records of exposure to temperatures higher than 
— 15° F. showed no injury at all. 
The conclusion, therefore, is that the vital point for complete 
killing lies between —20° and —25° F., with an absolute certainty 
that all eggs exposed to any further drop will be killed. 
The artificial protection experiments may properly be considered 
after the monthly collections and month's exposure, for the data 
secured from the first in a wav serve to substantiate those ob- 
