48 Bird - Lore feathers, notably of Wild Ducks, ornamented with a red catkin, probably of the alder. One could do far worse than to spend the entire season at the Magda- lens to study bird-life, and remain long enough to take in the fall migration, especially of the shore-birds, now so scarce on the coasts of the New Eng- land and Middle States. In such an event, prudence would require taking along a good supply of canned provisions, unless one wants to live on the soda-biscuit, crackers, cheese and ' tea ' of the poor but hospitable islanders. A Familiar Sparrow Hawk By NATHAN CLIFFORD BROWN TOWARD five o'clock of an afternoon about the beginning of Jan- uary, 1906, I saw a Sparrow Hawk, apparently a male, fly under the piazza roof of the Kirkwood, a winter hotel which had just been opened for the season, at Camden, South Carolina. He immediately flew out again, but soon returned. Within a few minutes he went in and out several times. Finally he alighted upon an electric light wire running along the rafters of the piazza, then, after a moment, descended to the capital of a pillar at the corner, whence he mounted almost instantly between two rafters and disappeared. I found him perched upon an irregular piece of board, perhaps twelve by eight inches along its longest sides, which had been nailed between the rafters so that three electric-light wires might be easily brought about the corner. With the rafter running obliquely and forming the ridge in the roof at the corner, and with two intersecting rafters and the roof, it made a sort of box closed all about except upon one side, but with a small triangu- lar hole at one corner in the bottom. This hole was about three inches by three by four. The bird's wings and tail protruded through it. The box was about five inches high. As I stood beneath it, I suppose it was about seven feet above my head. Until January 20 I went to this place several times every day, and never failed to find the bird there about five o'clock in the afternoon; I never found him at other hours except before he left in the morning. He did not, however, go to roost regularly. One very fine afternoon, when the sun was fully an hour high, I found him already established for the night at twenty minutes to five, eastern time. Two days later, the weather being damp, and the sky heavily overcast, he had not retired at ten minutes past five. At a quarter to six he was there. On January 13 I found him in his box at twenty-eight minutes past four. This was a dark afternoon, with a drizzling rain falling and the thermometer at forty-six. On January 15 he went to roost at a quarter past five, on the departure of a carpenter who
