MALLARD 
31 
Although pairing is usually initiated by the drake, the females end by actually 
singling out and pursuing certain drakes, so that during pairing the male is at first 
active and later plays a more passive role. The drake thus singled out usually retires 
from the female in a very ungallant fashion, perhaps in order to separate from the 
fiock, and at first pays no attention to her. Other drakes follow but are frequently 
driven away by the female, though serious fights never occur. Attachment of the 
pairs, though at first weak, becomes later on strong and permanent. Brock concludes, 
and I think rightly, that there is no selection on the part of the drake, but a definite 
choice, which may even amount to a strong factor in natural selection, is practiced 
by the female. The influences governing the choice of the female appear to take the 
form of certain stimuli provided by the display of the male, the external attributes 
of the male, or the cumulative effect of both. The whole subject of the relationship 
between display and choice of mates well deserves further study. 
The mating act always takes place on the water, and is followed immediately by 
a postlude in which the male swims hurriedly in a semicircle around the female, 
as in position nmnber five, with neck stretched low over the water. A moment later 
both birds go through rapid bathing actions, in order to shake the water from their 
feathers. The remarkable nuptial flight, common also to other species, is seen when 
the breeding grounds are reached. The male of one couple flies after the female of 
another couple, which has probably encroached upon the former’s nesting ground. 
A mad pursuit begins, both the female’s mate and the strange male going through 
various evolutions at a great height. Finally the strange male becomes wearied, or 
at any rate breaks away and returns to his own mate. The significance of these 
flights is not known. Wetmore (1920) evidently considers them part of the dis- 
play, while Heinroth (1911) who has described them from semi-wild birds, offers no 
explanation at all. In my opinion these flights are not to be regarded as part of the 
display, but are rather to be connected with the competition between mated pairs 
on the actual breeding grounds. Nowhere else, so far as I know, are they seen. But 
it may also be argued that these flights result from imfaithful males, whose mates 
are incubating and whose sexual instincts are not yet satisfied. Francis Harper who 
spent the summer of 1920 in the Athabasca Lake region, and who paid particular 
attention to ducks, tells me that he saw pursuit flights, mostly around the middle 
of May, and the last one on June 15. The number of males taking part in these 
flights varied from two to six. 
The nesting period is moderately early, and varies from March to July, accord- 
ing to the latitude and elevation. In temperate regions the Mallard begins to lay 
about the middle of April, though late clutches are found in June. In California the 
height of the breeding season is reached in late April (Grinnell, Bryant and Storer, 
1918). The earliest record for eggs is for March 25. In England the nesting season 
is very early, from the end of March to the middle of April. In Greenland eggs have 
