362 MICHIGAN SURVEY, 1908. 



The floor of the building was strewn with pellets consisting prin- 

 cipally of fishbones, skeletons of small mice, and some insect remains. 

 Under some of the rafters this had accumulated to such an extent that 

 the deposit was four or five inches in depth. In places it was weathered 

 so badly that it apjK^ared simply as a, mass of brownish earth. 



54. Cor mis hracJii/rhi/nohos (488). American Crow. 



Range : Xorth America from the Fur Countries to the southern border 

 of the United l^tates. Locally distributed to the west. 



Stations: Rock Harbor, Beach, I, 1; Spruce and Balsam Forest, I, 

 2-3; clearing along Benson Brook, II, 1; IV, 9; III, 3. 

 Washington Harbor, clearing, I, '04; Forest, VI, '04. 



Resident : Throughout our stay. 



The Crow was not common at any locality visited. It was occasion- 

 ally met along the shore of Rock Harbor (I, 1) where it fed on the 

 cleanings thrown out by the returning fishermen. Several were ob- 

 served in the birch forest at the top of Greenstone Range (IV, 9) and 

 also at the Bulrush Zone and Delta at the western end of Rock Harbor 

 (III, 3), but the natural and artificial clearings proved to be the most 

 favorable for these birds throughout the island. They were quite rare 

 at Siskowit Bay, but proved to be quite comihon at times at Wash- 

 ington Harbor. Here they resorted to the clearings and roads where 

 they fed greedily upon the swarms of grasshoppers. Xo nests were 

 found, but young of the year were more abundant than adults at Wash- 

 ington Harbor. The residents reported that these birds leave the island 

 the last of October. 



55. Acjelaius phoeniceus fortis (498). Thick-billed Redwing. 



Range: Breeding range, Mackenzie River, Athabasca, and other in- 

 terior districts of British America. During migrations, the Great Plains, 

 from eastern base of Rocky Mountains to Manitoba (Red River settle- 

 ment), Iowa (Burlington, October), Indian Territory (Beaver Creek, 

 November) western Illinois (Henderson County, Morgan County, March) 

 northern Kentucky (Mason Co., December) and southward through 

 more southern Rocky Mountains to Arizona (Fort Verdi, December, 

 February; Big Chino Valley, March), and western Texas (El Paso, 

 February). 



Stations : Washington Harbor, clearing, I, '04 ; forest along river., 

 II, '04. 



Migration, August 19 to Sept. 20. 



By all odds the most abundant black bird on the island. On July 

 11 a Red-winged Blackbird was seen in a marsh at Rock Harbor, but 

 was not procured, so identification, where so slight a difference exists 

 as between the species and subspecies, was impossible. It is my opinion 

 that this was the common Redwing {Agelaiiis lihoenivcus phoeniceus) 

 and not the northern form. On August 19 large flocks of the latter 

 form, came to the clearing at Washington Harbor (I, '04). On the 

 day previous several sjieciniens of phoeniceus were taken, but none were 

 seen after this. The keeper at the club-house said none of these black- 

 birds had been seen before this date, and as we found none on any 

 other part of the island (unless the one previously mentioned should 

 have been foi~tis) it seems probable that this form does not breed on 

 the island, and only appears here during the spring and fall migra- 



