40 The Wilson Bulletin — No. G6. 



departure, which was later than that of most of the other spar- 

 rows. Associated with them during the greater part of their 

 stay were three Fox Sparrows ; at times also numerous White- 

 throated, Song, Swamp, Field and Tree Sparrows, Juncos, a 

 half-dozen or more Lincoln and two or three Clay-colored 

 Sparrows. Somewhat apart from this company was a small 

 flock of Purple Finches, the same birds it is believed were seen 

 every day and remained nearly three weeks. During the day 

 the Harris Sparrows were generally to be found in one of four 

 places, either in a brush-pile, a thicket composed of dwarf plum 

 trees and raspberry bushes, a weed patch, or in willow trees 

 that overhung a favorite bathing place for the birds ; all four 

 of these places can be described within a circle having for its 

 radiance one hundred yards. 



While driving ten or twelve miles over prairie roads during 

 migration days, one soon comes to look for Juncos and their 

 congeners chiefly in the vicinity of the farm-houses, which, 

 more often than not, are built on hill-tops and provided with 

 wind-breaks of evergreen trees. On the other hand, a drive of 

 equal length through the neighboring woodlands with farm- 

 houses in clearings reveals the sparrow hosts, not near the 

 farm buildings, but where thickets are growing in sheltered 

 places. These observations may lead one to think that spar- 

 rows in their migration halts are influenced to tarry in certain 

 places quite as much by sheltered roosting-places as by good 

 food supplies. 



;\N EXERCISE IN P.IRD STUDY. 



I!V W. F. COl'ELAND. 



During the summer school at Ohio University, the class 

 in l)ird Study did some work that I wish to report. Students 

 were asked to select a nest containing young birds, make ob- 

 servations for one day, and report same to the class. Such 

 studies have been made by others and have always been highly 

 recommended by them. Others have made the criticism that 

 there is a probable error because the birds are more or less dis- 

 turbed by the presence of the observer. In the reports here 



