MALLARD 
15 
resident on Bering Island, though rare in -sdnter (Stejneger, 1885; Bianchi, 1909). It may be well 
to mention here that according to Moulton’s (1914) catalogue there is a specimen in 
the Sarawak Museum in Borneo, coming from that island. This is the only record 
I find for the West Indies. 
Migration 
One approaches this subject with the greatest hesitation, because it is so difficult to write any com- 
prehensive story of the seasonal movements of so cosmopolitan a species. Besides, we have to admit 
that we know almost nothing about the extent of territory which the individual covers; and the more 
records of banded ducks we get, the more complicated does the problem become. I propose, there- 
fore, to outline seasonal movements, and then apply them to the various parts of the Mallard’s 
range. 
This duck undoubtedly becomes resident wherever it can maintain itself throughout the year. 
The Greenland Mallard, which is fairly distinct, forms a group entirely isolated. The Iceland Mallard 
is not so distinct, and is certainly not an absolutely isolated group. Groups, apparently strictly 
resident, occur in many other places, but they have produced no local varieties. 
In Mallards, as in probably all our northern river ducks, the males leave the female during incuba- 
tion, mostly well before the young are hatched. These males may make a considerable excursion 
to some safe retreat, where they congregate and moult. How far these males may travel after they 
have moulted, and before there is any southward movement, we do not know, but we suspect from 
records of banded birds at Bear River in Utah and from other sources, that there may be con- 
siderable summer flights, very likely in a northerly direction. The young birds may even participate 
in such flights. However that may be, there is need of very intensive banding of both young and old 
birds in many different localities, besides observation on banded birds where they may be frequently 
trapped to see how long the individual tarries in one place. In all places where a definite migration 
takes place, the females and young of the year move before the old males. This is borne out by 
many scattered observations such as those of Leopold (1919a) and Hollister (1920a) for the autumn 
migration, and it also applies to the winter distribution. Not only do the males migrate in general 
later, but they stay farther north. Major Allan Brooks made some interesting sex counts of large 
numbers of various ducks at Comox, British Columbia, and he tells me that about the 10th of 
February there were three males to one female Mallard, but the last of the month, with others arriv- 
ing from the south, the sexes were nearly equal. As the same phenomenon w'as seen in various other 
species it no doubt points to some general principle, the value of which to the species it is difficult 
to imagine. Some time during the beginning of migration the sexes must become approximately 
even, for final pairing takes place at that time, and large bodies of ducks can be easily seen to be 
made up of pairs, as soon as they settle and begin to feed. 
In North America, we have a large passage area, that is, an area over which very few birds either 
breed or winter; and this is formed by the north-central and northeastern States with parts of 
Ontario. Colorado, Nebraska, Kansas, Iowa, Missouri, Wisconsin, Illinois, Indiana, Ohio, Michigan, 
Pennsylvania, New Jersey and New A’ork are chiefly passage areas, but more ducks breed there now 
than before spring shooting was stopped. The great bulk of Mallards are bred in the north and west, 
and these winter in large part on the Gulf coast. Many, however, head southeastward instead of 
south, and just touch southern New England en route, distributing themselves all the way from 
Long Island Sound to Florida, becoming more plentiful as one approaches South Carolina. Now the 
Mallards that take this journey do not stay very long on the Atlantic coast, but work their way 
south, so that after January 1st, the species is rare at Currituck, North Carolina, and after February 
1st, practically absent. These ducks must work back to their breeding grounds over a different 
route, for they are not present on the North Atlantic coast in spring. Cook apparently did not 
believe that such elliptical journeys were performed by any of our Anatidce except the Brant, but I 
