Bulletin No. ij. 51 



be at least suggested by mention of three other roosts in the county. 



The Vermillion roost, which is fourteen miles north-westward, con- 

 sisted of about 3000 birds, nearly all of which came from the east or 

 south-east in 1896. It is situated on the shore of lake Erie, thus making 

 a northerly feeding ground impossible. Birds belonging to this roost 

 were traced as far east as directly north of Oberlin, and five miles inland. 



The Elyria roost, nine miles north-eastward, is apparently a larger one 

 than the Oberlin roost, the birds of which feed in a south-easterly direc- 

 tion. A small proportion range northward and westward. 



The Wellington roost, eight miles directly south, is said to be fully as 

 large as the one on the Oberlin campus. I am told that the most of the 

 birds inhabiting this one feed southward. For information relating to 

 this and the Elyria roost I am largely indebted to residents of the two 

 villages. 



The reason why the birds of the Oberlin roost foraged in a south- 

 westerly direction is thus made clear. The foraging grounds in other 

 directions were occupied by the inhabitants of the other roosts. Several 

 times it was noticed that when a strong wind blew from the east or north 

 more birds than usual came from that direction ; and vise versa, when the 

 wind blew strong form the south or west, almost no birds came from the 

 opposite direction. This suggests that when the wind is favorable for 

 them to do so, such individuals of the other roosts as wander away in 

 the direction of this one, come to it rather than breast the strong wind to 

 return to their own. And it is undoubtedly true that an unfavorable 

 wind caused some birds belonging to this roost to seek the shelter of one 

 of the others. Thus there would be a slight variation from day to day 

 in the number at each of the roosts. Strong north winds were of rare 

 occurrence, but when they did blow, there was no apparent effect upon 

 the incoming host from the south. It was too large and too well organized 

 to be turned back by weather. 



After the grackle host had formed into one huge flock, the form of that 

 flock, as it passed over the fields and woods on its return to the roost or to 

 its feeding grounds, was an interesting study. It could not be seen to ad- 

 vantage from Spear Library or near the roost, because the birds were 

 headed that way ; but at the proper time of day almost any station in the 

 fields south of Oberlin was sure to give one an excellent side view of the 

 passing flock. The accompanying representation of the varied appear- 

 ance of the flock on different days as well as at different stages of the 

 flight on the same day, will give a fairly accurate idea of the average forms 

 assumed. At the beginning of the return or departing flight a form similar 

 to the upper figure was assumed, and the end of it, either at the feeding 



