Bulletin No. ij. 49 



gradually joined together until there was but the one huge flock with few 

 stragglers. 



There were no wet days during the summer, but several thunder 

 storms swept over the region early in the evening at the gathering time 

 of the grackles. The actions of the grackles at such times were different 

 enough from what we have already seen to warrant particular mention. 



On September 5, conditions were favorable for the formation of local 

 thunder storms. Early in the afternoon a storm began to form in the 

 west, which moved eastward as it increased in magnitude. The sun was 

 not obscured until about four o'clock, but the mutterings of thunder gave 

 promise of a considerable storm. At 5:00, the storm cloud had spread 

 well over the heavens, and at 5:04, a light sprinkle began, which gradu- 

 ally increased to a brisk rain. At 5:10, 500 grackles came swiftly in from 

 the south, flying low and settling into the foliage at once. 5:15, steady 

 rain with a light south-west breeze. A flock of about 2000 grackles 

 arrived from the south at several hundred feet elevation ; they went at 

 once to the roost and disappeared in the foliage. 5:16, harder rain with 

 more wind. A flock of about 3500 grackles appeared high up over the 

 roost. They dropped down with half closed wings and at once disap- 

 peared. 5:20, the storm burst with sheets of rain and a westerly gale 

 which tore the leaves from the trees and must have drenched the foliage 

 with the first blast of the storm, which had spent itself in five minutes. 

 During this time there was no sign of the grackles. At intervals of 

 five minutes after the passage of the storm heavy showers followed until 

 5:45. Between showers the birds shifted about some, but did not show 

 themselves. Theirs must have been a wet berth on that night. All 

 arrived noiseless, and uttered scarcely a sound afterwards, even when 

 shifting about. 



The falling rain made it impossible to see the birds until they were 

 nearly ready to descend into the trees, but the direction of their flight 

 seemed to be the same as on other days. The last birds arrived at 5:16, 

 four minutes before the storm burst in earnest, and forty-five minutes 

 earlier than the last arrivals on September 7. Normally the return flight 

 would hardly have begun until after the time that the storm had passed 

 over. The weather after the storm was more suitable for flight than 

 when the birds came in, and there was ample time for them to have 

 reached the roost before dark had they waited for the storm to pass, but 

 they could not risk being kept away from their accustomed sleeping place. 

 Whenever a storm threatened during the aftermoon, unless it passed over 

 before four o'clock, the grackles left their feeding grounds in time to 

 reach the roost before the storm burst upon it. 



