Abel — A Winter Rorin Roost 171 



From this time observations were made only on the North- 

 east quadrant, and there seemed to be a very rapid decrease 

 in numbers. Since a complete census was not attempted 

 later it will probably be best not to give any incomplete fig- 

 ures. Suffice it to say that the observations were made every 

 three or four days until the middle of October. After Sept. 

 27 there were only scattered individuals; and on Oct. 16 none 

 flew in from any direction. 



It seems to be a regular habit of Robins to roost in im- 

 mense flocks in their winter homes in the south, and there 

 are numerous published accounts of the wanton slaughter of 

 them under such circumstances. But if it is their habit to 

 roost thus at the close of their breeding in the north, before 

 the fall migration, it does not seem to be very generally rec- 

 cognized. 



Mr. Wm. Brewster publishel nearly twenty-five years ago,^ 

 a most admirable account of several " Summer Robin Roosts " 

 near Cambridge. His observations on the behavior of the 

 birds were so complete that little can now be said in addition. 

 The present paper must be largely in the nature of a confirm- 

 ation. 



Our observations agree in that the flight was not equal in 

 all directions, and that various other species of birds, in 

 small num.bers, often became associated with the Robins in 

 the roost. 



Mr. Brewster also discusses the matter of the composition 

 of the roost as to the sex, and gives some reasons for think- 

 ing that the summer roosts are made up of the males and 

 young of the first broods of the yea-r. Fisher - also expresses 

 his belief that the fall roosts of the Barn and Bank swallows 

 consisted of males. I have been unable to make any positive 

 observations on this point ; but, in as far as the lighter color 

 of the breast is indicative of sex, I should incline to tbe opin- 

 ion that the roost under my observation contained a fair pro- 

 portion of females. 



The manner of flight to the roost seems to be distinctive in 

 ^Tbe Auk, 7, 1S90, pp. 360-373. 

 = The Observer, 7, 1896, pp. 382-384. 



