The average commencement of ten periods after the third week in May was 
less than one-half hour after sunset. The duration of the post-sunset 
period is rarely less than one hour; when it is shorter, inclement 
weather seems to have influenced it, and if it extends to more than an 
hour and a half either moonlight or clear cloudless conditions appear to 
cause the extension. In each post-sunset period there also seems to be 
a period which may be termed the "peak interval" when the maximum number 
of individuals are performing simultaneously. Counts made during this 
peak interval show great possibilities of being an index of population. 
The pre-dawn period seems to be much more definite in its commence- 
ment. Data on seven such periods show that three of them began at a.m. 
exactly, and three others at 3: a.mey 3:45 a.m, and 3:46 a.m, respec- 
tively. The other was only three minutes off the a.m. starting point. 
If any conclusions could be based on such scanty data, I would suggest 
that the rather consistent temperatures during the night might have some 
bearing on the consistency of the commencement of these periods. The pre- 
dawn period also shows a peak of maximum performance, but it appears to be 
much shorter and of less dependability than the post-sunset peak interval. 
There is remarkable consistency, under similar conditions, in the lengths 
of both post-sunset and pre~dawn periods. 
When four or more birds are involved, activity during the pre-dawn 
peak interval is a less reliable indicator of population, Table 1 
illustrates this, 
Table 1.--Comparative percentages of periods of activity 
Postesunset period Pre-dawn 
One bird 100% 100% 
Two birds 82 5% 83% 
Three birds 66% 63% 
Four birds 8% 37% 
Five birds 30% 17% 
Table 1 gives percentages of periods during which different numbers 
of birds were active, averaged for six post-sunset periods and three pre- 
dawn periods followed through from May 5 to May 1. An examination of the 
accompanying winnowing charts (Figs. 1, 2, and 3) also indicates that the 
pre-dawn period, although comparatively equal in length and more depend- 
able in its commencement has not the stability of the post-sunset period. 
This may be because there is a more dynamic change in light intensity 
during the post-sunset period, i.e. a change from maximum light to dark- 
ness in a minimum period. On May 12, at Tompkin's, for example, the post- 
sunset period was preceded by fifteen hours of light, whereas the predawn 
period was preceded by less than nine hours of darkness. 
It is not certain that the number of snipe performing even during 
peak intervals is an indication of the number of males actually in the 
area, although the data obtained under nearly constant conditions at 
Tompkin's (Fig. 3) suggest that such is the case. 
Ly 
