50 Bulletin No. 15. 



Cloudy days had the effect of retarding the morning departure and 

 hastening the evening arrival. The difference between the time of de- 

 parture and arrival on cloudy and clear days was coincident with the 

 difference in the amount of daylight of two such days. My daily notes 

 of the time of departure and arrival indicate a close correspondence with 

 the varying length of the days from the time the birds were first seen 

 until the roost was deserted, allowance being made, of course, for the 

 effect of cloudy days and thunder storms. 



There was no diminution in the number occupying the roost up to 

 September 21, but not one bird appeared at the old stand on the two suc- 

 ceeding days. On the 24th less than a hundred occupied the trees during 

 the night, and none visited it afterwards. This sudden failure to report 

 promptly was undoubtedly due to the persistent persecution of would-be 

 hunters of tender years, who made it a business to shoot into the flock as 

 it passed over the outskirts of the village. While watching the incoming 

 flock on September 10, a fusillade of shots caused the approaching flock 

 to rise suddenly several hundred feet, turn abruptly back and divide, one 

 part making a wide detour to the west and coming in from that direction; 

 the other going back fully a mile and rising to a thousand feet elevation 

 before venturing back again. This altitude was maintained until the 

 flock was nearly above the roost, when the birds came dropping down 

 like leaves. For about ten days following this serious disturbance the 

 birds came in from a different direction — more southwesterly — avoiding 

 the dangerous region. On September 21 they again came in from the 

 south over the old route for the last time. 



It was feared (hoped by some) that this experience would cause the 

 grackles to abandon the old roost permanently, but on June 3, 1897, the 

 trees were occupied for the first time by about fifty birds. The number 

 has gradually increased since, with fair promise of as large a gathering 

 as usual before the close of the season. At the present writing they have 

 not begun to flock, so the direction of their feeding grounds cannot be 

 determined.* The tendency heretofore to a southern direction may 



*Since the above was written the grackles have formed into considerable flocks, and 

 the direction of their feeding grounds has been determined, which is a little east of 

 north. Their line of flight from the country passes a little east of the east end of 

 Maple street to a large elm tree, where the majority of the birds rest for a few minutes 

 before flying directly to the roost. It is difficult to see them immediately before they 

 reach the roost, because they descend almost to the ground at the intersection of 

 Lorain and North Main streets, then fly upward to the tree tops. It would be inter- 

 esting to know what effect, if any, this change in the direction of the Oberlin birds 

 has upon occupants of the other roosts in the county. That would require the re- 

 mainder of the roosting season to determine. 



