Page:Bird-lore Vol 01.djvu/192

17^ Bird -Lore Every evening they came at dusk by thousands and tens of thousands, winging their way in long lines from all points of the compass, and settling down on the reed-covered islands in a solid black phalanx. This winter roosting habit of the Crows is well- known, and many roosts have been located, but the habit seems still to lack a satisfactory explanation. Why should these birds fly back and forth every day over miles and miles of country to roost in some definite spot which, so far as we can judge, is no better suited for roosting purposes than hundreds of other places which they pass by ? And why should they gather together every night in such numbers as to attract general attention and invite slaughter by thoughtless gunners, when, by roosting in small numbers wherever they happen to be feeding, they would escape notice ? These are questions I shall not attempt to solve. Estimates placed the number of Crows in these two island roosts at half a million, and they held possession of the islands undis- turbed until about the time of the establishment of Fort Delaware. They did not relish this intrusion, and determined to desert the ancestral Pea-patch roost ; being also influenced, no doubt, by a storm which flooded the island at night and drowned thousands of the unfortunate birds. The Reedy Island roost continued in use until the establishment of the Quarantine Station, at a much later day ; then it, too, was de- serted, and the famous island roosts were no more. I have long been interested in the winter gatherings of the Crows, and made inquiry of the light-keeper at Reedy Island to as- certain whether any Crows at all remained there at the present time. I was informed that they came across from Delaware as of old in long flights from the west, northwest and southwest, but all passed over the island into New Jersey, where he judged they had estab- lished new winter quarters. The location of this new roost at once became a matter of in- terest. By further inquiry I learned that Crows at Salem, N. J., nearly opposite the Pea-patch, flew southwards at evening, and by plotting this flight line with those given by the light-house keeper,. on a map, I found that they joined some four or five miles below Salem, and here I felt sure the roost was to be found. I had little trouble in impressing an ornithological friend, who resided at Salem, with the importance of locating this roost, and one cold afternoon in January found us driving off in the direction taken by the Salem Crow flight. When we neared the point at which we thought the roost ought to be, we noticed a scattered line of Crows coming up from the