Whatever the origin and reason of this roosting habit, I have no doubt that it is universal. Middlesex County birds cannot be in any respect peculiar. Whoever will keep a close eye upon the robins in his neighborhood, in July and August, will find them at sunset flocking to some general sleeping-place.
It would be interesting to know how far they travel at such times. The fact that so many hundreds were to be seen at a point more than a mile away from the Belmont roost is significant; but I am not aware that any one has yet made a study of this part of the subject. My own birds seemed to come, as a rule, by easy stages. In the long narrow valley east of the roost, where I oftenest watched their approach, they followed habitually—not invariably—a zigzag route, crossing the meadow diagonally, and for the most part alighting for a little upon a certain wooded hill, whence they took a final flight to their nightly haven, perhaps a quarter of a mile beyond. Farther down the valley, a mile or more from the roost, birds were to be seen flying toward it, but I found no place at which a general movement could be observed and large numbers counted.
As to the size of these nightly gatherings, it seems wisest not to guess; though, treating the subject in this narrative manner, I have not scrupled to mention, simply as a part of the story, some of my temporary surmises. What I am told of the Belmont wood is true also of the one in Melrose: its shape and situation are such as to make an accurate census impossible, no matter how many "enumerators" might be employed. It could be surrounded easily enough, but it would be out of the question to divide the space among the different men so that no two of them should count the same birds. At present it can only be said that the robins are numbered by thousands; in some cases, perhaps, by tens of thousands.
FOOTNOTES:
[14] Mr. William Brewster has been aware of this habit for twenty-five years, but, like myself, has never seen it mentioned in print. He devotes to it a paper in The Auk for October, 1890, to which I am happy to refer readers who may wish a more thorough discussion of the matter than I have been able to give. My own paper was printed at the same time, in The Atlantic Monthly, and had been accepted by the editor before I knew of Mr. Brewster's intention to write. References to a roost in Belmont, Mass., discovered by Mr. Brewster six years before, are frequent in the following pages.
[15] This roost was discovered by Mr. William Brewster, in August, 1884, as already mentioned.
[16] A week later, my correspondent reported a similar state of things at the Belmont roost. "A very large proportion of the birds are spotted-breasted young of the year, but occasionally I have detected an adult male." He examined the birds at near range, and at rest, after they had come into the roost in the earlier part of the evening.