HATCHIXG OF GIPSY-MOTH EGGS. 9 
between the new level and the old. A new storm might then come and 
cause the level to rise again. So it went during the entire winter, and 
clusters on the border line of snow protection were alternately ex- 
posed and protected. Observation made on some such clusters proved 
conclusively that the snow is the medium of protection. 
There remains only one series of experiments to be discussed, 
namely, temperature resistance, which was conducted for two years 
only. As it developed, the first year selected for this type was, very 
opportunel}^, the mild winter 1918-19. The only killing of eggs for 
that year that was recorded in our experiments developed in some of 
these sets of clusters exposed at the northernmost limits of the gipsy- 
moth area, in the only towns from which records anywhere near 
—20° F. were taken. After the next winter, as was to be expected, 
none of the clusters in this series hatched. 
RESULTS OF SUMMER SURVEY OF HATCHING. 
A few observations made in the field after the end of the hatching 
season which followed the extremely cold winter of 1917-18 disclosed 
the fact that there was an unusual amount of nonhatch. This gave a 
very good opportunity for a careful study under natural conditions. 
Accordingly, extensive plans were made to extend observations over 
as large a portion of the gipsy-moth area as possible. Kesults ob- 
tained from experiments had pointed strongly toward low tempera- 
ture as the causative factor of nonhatch. To a certain extent, how- 
ever, the experiments had an element of artificiality, and it was de- 
sired to obtain as much information as possible under purely natural 
conditions. 
The plans called for observations on hatch of egg clusters on trees, 
undergrowth, stumps, debris on the ground, boulders, stone walls, 
and other objects. Many factors were taken into consideration, such 
as nearness to the ocean and bodies of fresh water, height of land, 
degree of exposure to prevailing winds, f avorability or unf avorability 
of food plants and the abundance or scarcity of these, the degree of 
gipsy-moth infestation, and any other points which might present 
themselves. These observations were taken during the three sum- 
mers 1918, 1919, and 1920. 
The observations taken in the summer of 1918 disclosed a complete 
nonhatch x)f all clusters above the snow-protection line in Maine and 
New Hampshire, with the exception of a section along the seacoast. 
This section showed only a partial hatch. All clusters found on the 
ground or on other objects close to the ground where they could be 
snow-protected hatched completely. Hatching above the snow-pro- 
tection line in Massachusetts, with the exception of a coastal section 
which included all of Cape Cod. was poor with a considerable number 
of whole clusters killed, particularly in the northern and northeastern 
parts. Hatching of low clusters in all this area was uniformly good. 
In Khode Island and along the coast as far north as the Massachu- 
setts-Xew Hampshire line the hatch was almost complete. 
Apparently the modifying influence of the ocean overcame the 
severe cold at least enough to prevent the temperature from dropping 
to the killing point all along the coast of Ehode Island and Massa- 
