1 'iS BiCKNELL ou the Singing of Birds. [April 



October. Extreme dates are August 24 and October 7 ; more 

 usual limits would fall in the first and fourth weeks of Septem- 

 ber. In several years I have noted imperfect songs in the fourth 

 week of August, about midway between the two seasons of 

 song. These seemingly misplaced songs I have usually consid- 

 ered as appertaining to the song-period from which they were 

 separated by the least interval of time. But they may be wholly 

 aberrant ; or the two song-periods may sometimes be connected ; 

 or perhaps in some years the first song-period is prolonged and 

 the second does not occur ; for in more than one instance I have 

 noticed that an undue extension of the first song-period seems to 

 be at the expense of the second. Either of these suppositions 

 could be supported by my records pf certain years, but recalling 

 the likelihood of the bird to be overlooked in the autumn, we 

 find ourselves justified in no conclusions without more extended 

 data. 



How far birds of the year enter into the subject of change of 

 song in the autumn with this sjDecies cannot at present be said ; 

 but a male bird shot while singing on September i, 1S80, was in 

 fine plumage and bore every indication of being fully adult. I 

 find this species in the autumn without much fat, and with 

 feather-growth continuing slightly into October. 



Anorthura troglodytes hiemalis. Winter Wren. 



A silent migrant wdth respect to song, though often amply 

 noisy with its ordinary notes, the movement of which fitly corres- 

 ponds with the excited bobbing of the little brown-plumaged body 

 from which they proceed. Once only have I heard its song in this 

 latitude — on November 21, 1S80, a cold and wintry although 

 still morning. The song was three times repeated, and though 

 brief was sufficiently perfect to bring to mind the summer home 

 of its author in mountain forests northward. In winter I have 

 found dark yellow fat encasing its small body. 



Telmatodytes palustris. Long-billed Marsh Wren. 



There seems to be an irregularity about the singing of this bird 

 in the late summer and fall which requires for full explanation 

 more complete data than my records afibrd. The first song- 

 period normally ends early in August, dates of final songs in six 



