have a limited series (39) from bandings during the fall, winter and 
spring. Here again only the indirect recoveries (27) are of much use 
in showing distribution. Indirect recoveries from all seasons of band- 
ing are combined to show the distribution pattern. It is obvious that 
bandings during the fall and spring may show a somewhat different dis- 
tribution than winter bandings, but until we have more records with 
which to work it is necessary to combine all the data. 
From the northern banding data we would expect the Delaware 
River population to be affiliated more closely with the central north- 
ern division, with the northeastern showing a diminishing and the 
northwestern an increasing importance. 
Breaking down the recoveries from the northern Delaware and 
western New Jersey stations into the same three categories as above 
we find: 
Northeastern ~ 9 percent (Long Island 2 percent) 
Central - 16 percent 
Northwestern - 9 percent 
This tabulation indicates that nearly half of the Delaware 
River birds are derived from the country directly north in Quebec, 
Vermont, and the Hudson Valley, and that the northeastern and the 
northwestern flights may be about balanced. However, as mentioned 
previously, we are by no means justified in assuming that these per- 
centages show other than trends. The mapped recoveries indicate that 
the spring-banded population, which shows subsequent recoveries into 
North Carolina, probably influences the totals to some extent in favor 
of the northwestern division. The northwestern group has one far- 
western record, Minnesota. 
. Maryland 
On the Eastern Shore of the Chesapeake Bay, most of the band- 
ing was done on the Blackwater Refuge primarily during the fall and 
spring. There are very few records of direct recoveries beyond the 50- 
mile radius. Of the 25 recoveries from bandings before the shooting 
season, 23 (92 percent) were killed locally. Even with the indirect 
recoveries amounting to 53, 62 percent were local recoveries. This 
leaves only 20 recoveries spread over a wide range to the north. A 
breakdown of the recoveries shows: 
Northeastern - 2 percent (Long Island, 0) 
Central - 6 percent 
Northwestern - 11 percent 
The indicated distribution pattern is the same as that shown 
by northern bandings. The ratio between regions is probably far from 
accurate but nevertheless one can assume that for the eastern Chesapeake 
225. 
