It has been pointed out that the distribution of recoveries, as mapped 
and tabulated, may or may not represent the true behavior of the popula- 
tion. Some of the more important factors causing distortions in the 
data are variations in hunting pressure, the time of hunting, habitat 
distribution, climate, whether both wintering and migratory populations 
are involved, and the method of trapping. There are relatively few 
data available that will give us a measure of the impact of these in- 
fluences on populations or on the resulting recovery of shooting records. 
An important behavior pattern of the species which has a 
bearing on whether a representative series of a migratory population 
is being banded and shot, is the time and extent of migration. With 
some bandings the time of migration is rather difficult to determine. 
However, with certain of the northern stations, in which a substantial 
number of direct recoveries from early fall bandings are available, we 
can derive considerable information. The northeastern coastal flights 
lend themselves well to this analysis since the population moves south 
along a restricted course from state to state and does not spread 
widely over the country. 
The accompanying charts (figs. 23-31) present the data more 
lucidly perhaps than involved discussions. The explanation of the 
chart is as follows: 
(a) Across the top the fall season is divided into periods 
of approximately 10 days. 
(b) Down the left-hand margin the provinces and states are 
listed, beginning with the northernmost and following consecutively to 
_ the southernmost in which reccveries were taken. 
‘e) Recoveries are shown under the 10-day period and the 
state or province in which they were taken. The number of black units 
represents the number of recoveries in each period. 
(d) The heavy line indicates the overall span of the hunting 
season in each state or province during the years in which banding 
took place. 
By arranging the data in this manner the approximate arrival 
date for some flights can be determined. Banding data and hunter-bag 
checks indicate that in recent years a third to a half of the kill is 
made the first 3 or 4 days of the season so that if banded birds are 
in an area the probability is high that some of these birds will be 
taken during the opening period of the season. It is likely, too, that 
new migrants are more easily taken immediately after arrival than later 
when the birds have learned to avoid the gun. 
a 
