302 BIRDS. 



served first on August 10, when a male was seen in the jack pines 

 on the sand dunes between Rush Lake and Saginaw Bay, and four 

 more were noted between this date and August 18. Four of these 

 were immature birds and had evidently been bred nearby. On 

 Sand Point it was seen first (at the west end) on the morning of 

 August 15. This bird (an adult male) was feeding in a small jack 

 pine. On August 17, at the same place, a large flock of warblers 

 that included several of this species was seen. Still another (the 

 last one observed) was seen in a thicket near the same place on Aug- 

 ust 24. 



108. Seiurus aurocapillus. Oven-bird.— Michigan: Abundant 

 summer resident, breeding throughout the state. 



Rather a common breeder in the sand region, and observed in the 

 clay country near Rush Lake. Not seen on Stony Island. A nest 

 with three eggs was found on Sand Point, at the edge of an old fallen 

 tree top on the edge of a wooded swamp. (PI. XVIII b.) This nest 

 was built in, and partly covered by, dry oak leaves. Xo more eggs 

 were laid in this nest, and the young were soon hatched and gone. 

 The young birds were often seen in the thickets, between July 15 and 

 25, and the species was still common on the Point when we left, on 

 August 27. 



109. Seiurus noveboracensis notabilis. Grinnell's water thrush. — 

 Michigan : Not uncommon as a migrant ; may breed on Isle Royale. 



This species was not an uncommon migrant in the sand region, 

 and was found in dry willow thickets and at the edge of wet swamps 

 on Sand Point and in the region of Rush Lake. It was observed as 

 early as July 30, at Rush Lake, but the first records for Sand Point 

 were on August 5, 6 and 7, when several were seen. The same habi- 

 tat was thoroughly worked again on August 8, but not a bird could 

 be found, and no more were seen until August 24, when a single bird 

 was observed in a clamp willow-poplar thicket. Nearly all of the 



' individuals found were on the ground not far from the water. Only 

 three were seen at Rush Lake, between July 30 and August 13, and 

 all in low wet thickets. 



The Michigan water-thrushes have usually been referred to the 



■ typical form, 'and Ridgway* gives the breeding ground of Grinneirs 

 as 'X'orth and west of Lake Superior." If this is the true breeding 

 range these records show a migration to the southeast. The whole 

 subject of Michigan water-thrushes is in need of careful examination 

 to determine the status of the two forms in the state. The probabili- 

 ties seem to be that the typical form (noveboracensis) does not enter 

 our limits, although one of the Isle Royale specimens has been so 



*Birds of North and Middle America. Vol. II, p. 640. 



