Page:The Zoologist, 4th series, vol 4 (1900).djvu/168

140 the roosts will hardly conceive it possible that they are natives only of the limited area surrounding the roost, and we can only conclude that the true natives of the district receive additional numbers that migrate thither from outside districts or even from the Continent.

I will conclude with a statement which may perhaps be received with incredulity. Nearly all ornithological writers say that Starlings breed two, and occasionally three times a year. Careful observation has convinced me that a very large proportion of Starlings—perhaps one-half—rear only one brood in the year; many rear two broods. I have never known three broods, and there are some Starlings that do not breed at all. Mr. G.W. Murdoch, Natural History Editor of the 'Yorkshire Weekly Post,' writes in reply to a query on this point:—"In Hants Starlings very often breed twice a year; I never knew them do so in Scotland. I am of opinion that a good many Starlings never breed at all, but for what cause I know not. That is also the opinion of my friend the Rev. H.A. Macpherson, M.A., author of 'The Fauna of Lakeland.'" Mr. E.S. Cobbold writes in April, 1899:—"Why are the Starlings congregating in flocks now? Hundreds fly over Stretton from the south-east at 5 or 6 p.m. to roost, I think, in a Scotch fir-plantation near; I have seen them three nights in succession, and on April 3rd I saw rather large flocks down by Craven Arms feeding together." He adds that at the very time when he saw the flocks flying overhead, the Starlings that were breeding about his house were busy looking after their nests or young, and did not take any notice of the others, much less offer to fly off after them. "Is it," he concludes, "the autumn habit not yet abandoned? I am inclined to think not, for I remember in previous years noticing them early in summer, when I supposed they had done breeding."

A simple and probable explanation of the phenomenon is this:—When Starlings rear two broods in the year, the second brood would be younger than the first by some two months or so. When the breeding season comes round in the following spring, the second broods are not sufficiently adult to breed, and, not having any duties to call them elsewhere, they naturally continue to resort at night to the old familiar roost. This