TREE SPARROW UGl 



flooding the States. They normally reach Oklahoma, near the south- 

 ern part of their range, by early November and remain until March. 



Traveling, apparently, in family and neighborhood groups, the 

 flocks upon arrival at Ithaca, N.Y., were composed of almost equal 

 proportions of males and females, with a IG percent dominance of 

 males (Baumgartner 1937a). First-year birds (with skulls incom- 

 pletely ossified) exceeded old birds 3 to 2 between mid-October and 

 mid-December. During the non-migi*atory period between January 

 and Alarch, the proportion dropped to almost equal numbers of 

 young and old birds, indicating a higher mortality of first-year birds. 

 Females, traveling farther south, dropped to a 3 to 1 proportion until 

 near the end of March. 



The extent of mnter territory and flocking habits has been the 

 subject of several intensive studies, foUowing different techniques 

 and with a variety of conclusions. At Ithaca, N.Y., in 1933, 1934, 

 and 1935, I marked individuals at foiu' stations by gluing different 

 colored feathers to their tails (Baumgartner, 1938b). By following 

 these marked individuals and by retrapping at other stations, I found 

 that the normal feeding range in winter was 500 to 1,700 feet in diam- 

 eter, average 1,000 feet. Extensions from 3,000 to 6,800 feet, recorded 

 for 40 percent of the birds, were associated with faU settling, spring 

 excitement, or heavy snowfalls. A few individuals appeared to be 

 regiJar wanderers. 



Winter flocks appeared to be loosely defmed units of varying num- 

 bers and individuals, with a fairly definite flock territory, from which, 

 however, individuals strayed at will. Flocks inhabiting open country 

 ranged more widely than those in denser cover such as marshes. In 

 severe weather, flocks split into smaller groups and wandered more 

 widely. 



The histories of individually marked birds suggest several explana- 

 tions for low return ratios at banding stations. Besides migrants, 

 which are unlikely to make the same stop-overs a 2nd year, this study 

 recorded several of each of the following types: 



(1) Trap-shy individuals that were caught only once or twice a 

 season, though observed regularly in the vicinity. 



(2) Wanderers that covered so large an area they missed their 

 original station another season. 



(3) Regular residents of another area that meandered only once 

 into the trap where banded. 



(4) Emigrants that shifted from year to year. 



A project at Amherst, Mass., during the winter 1957-58 (Sargent, 

 1959) used colored bands at five stations. Sargent likewise con- 

 sidered flock structure unstable, with no fixed associations within a 

 flock, or regular times of day for visitations to the traps. He also 



