DELAWARE VALLEY ORNITHOLOGICAL CLUB. 37 



females visit the roost and towards the last of the month young 

 birds begin to come in in company with the males. 



July. — Most of the birds come to the roost as family groups 

 during July. At this time of year they do not follow any 

 regular lines of flight but come in from every direction, mostly 

 flying low and often stopping to feed along the way. 



August. — All this is changed soon after August 1st. The birds 

 have for the most part completed their domestic cares and family 

 groups are rapidly consolidated into large flocks which come to 

 the roost from considerable distances. The numbers are very 

 greatly increased and the birds in flying to and from the roost 

 follow much more closely a regular well defined route. 



September and October. — During September and October the 

 greatest numbers are reached and the birds come in at night in 

 great flights, one flock following another so closely as to give the 

 impression of a single long-drawn-out flock. The flight begins 

 about 5:30 p. m. and lasts for about twenty or twenty-five min- 

 utes, but scattered birds and small flocks continue to come in 

 until dark. I believe that from fifty to seventy-five thousand 

 birds visit the roost every night during these two months. 



November. — The birds come regularly to the roost until they 

 begin to go south about the middle of the month where two or 

 three nights are sufficient to finish the large flights. 



December. — Quite a number of birds remain over into Decem- 

 ber, probably double the number to be found in January but 

 upon very cold nights they are quite likely to be entirely absent, 

 returning again as soon as the weather moderates. 



Robins use the roost to the number of one thousand or more, 

 their numbers being hard to judge with any degree of accuracy 

 on account of the way they mix with the Grackles. 



On a visit paid to the roost on September 17th, I was sur- 

 prised at the large number of birds which still showed marked 

 evidence of the molt. While fully one-fifth of the birds lacked 

 one or more flight feathers, this number was insignificant com- 

 pared to the number showing faulty tails. Not over one-tenth 

 of the birds observed had tails which could be called even ap- 

 proximately perfect. 



Birds whose tails were fully developed seemed to have the 



