1318 U.S. NATIONAL MUSEUM BULLETIN 23 7 part 3 



On September 3 of the same year I tried an identical experiment with an adult 

 female which was just starting to molt. I trapped and released her at the same 

 points as I had the male, and nine days later found her back where I had trapped 

 her. By this time she was molting heavily on the wings, breast, and belly. 

 Her tail feathers had been renewed but were just beginning to grow out. 



The significant point of these experiments is not that the birds returned to 

 their breeding areas, but that they did so at a time when activity is normally 

 reduced to the minimum and when the pairs into whose area they may have 

 wandered would almost certainly have made no effort to drive them out. 



Banding data for the young birds indicate that during most of fall 

 and winter they too are probably limited to a restricted area, within 

 a few hundred yards of which they attempt to establish territories 

 the next spring. Such wandering as they do probably occurs before 

 they undergo the postjuvenile molt. Depending on when they are 

 fledged, therefore, the period of wandering could last from May 

 through most of August for young of a first brood, or about a month 

 for young fledged in July. 



On August 26, 1934, I trapped within 400 yards of its birthplace, a young bird 

 beginning its postjuvenile molt; it had been banded as a nestling the previous 

 May. Through the following winter this bird (male xi) stayed with a flock of 

 unbanded immatures within an area of 1.7 acres which included the spot where 

 I trapped him and at the edge of which he settled the next spring. In October 

 and November of 1935 I banded three immatures, two of which stayed near by 

 tbe rest of the winter and bred the next spring within 50 yards of where trapped. 

 One of these was retaken the following July less than 200 yards away. These 

 histories, together with the fact that I had seen groups of immatures foraging 

 day after day in the same spots, made me suspect that they, like the adults, 

 might limit themselves to small areas in fall and winter. 



To substantiate this, I color-banded fourteen immatures in August and 

 September of 1936. All but one were seen again that winter: twelve, from one to 

 fifteen times within 200 yards of where trapped, and one, once, within 400 yards. 

 Eight settled the next spring on territories from 100 to 400 yards from the trap 

 sites. None was ever found farther than 400 yards away. 



* * * The young birds begin to sing, though rarely, in mid-July, while still in 

 juvenile plumage. This song may take the form of either a weak or abbreviated 

 version of the immature song or a prolonged trill. As in most passerine birds, 

 the immature song is much more variable than that of the adult. The young 

 bird sings weakly, hesitatingly, often repeating notes of the first phrase before 

 uttering the trill, which is sung so slowly that the separate notes can be heard. 

 Every conceivable variation is used: the first three notes may be sung and the 

 trill omitted, or one or more of the first three notes may be left out, and there are 

 innumerable variations in the trill itself. 



By mid-September — when most individuals have finished the 

 molt — song and other evidences of renewed territorial interest may 

 recur. One male I watched in 1936 on the Berkeley campus sang 

 from a conspicuous perch in early September and throughout October. 

 Twice in the first week of October he sang loudly every few seconds; 

 his song was indistinguishable in force and frequency from that of 



