468 
Bulletin Xo. 162.—1915. 
of them would die than of males, and therefore there would be a 
correspondingly higher percentage of males at the time of banding. 
In the following pages these points are examined critically on the 
basis of a larger number of data, other age limits being used, however, 
than age at banding. Data to support this contention are furnished, 
and we have therefore examined our records with a view to deter¬ 
mining its validity. 
From our records to December 1, 1914, the age at time of death of 
1,227 birds not including those killed can be determined, and for 
convenience these have been divided into four groups, one of which 
seems to have a natural basis, whereas the limits of the other three 
are more or less artificial. These may be defined as follows: 
Group A. Hatching period. Embryos which were fully devel¬ 
oped but which failed to hatch (“dead in shell”) and birds which 
died within five days after hatching. 
Group B. Nestling period. Squabs dying at from 6 to 28 days, 
inclusive. 
Group C. Immature or juvenile period. Squabs living over 28 
days, but dying before six months of age. 
Group D. Adult period. All birds living to the age of six months 
or more. 
