By weight analyses Moore and Pearson (191) have show that 
very young birds constitute a sizable fraction of the mourning doves 
taken by hunters in Alabama. The September-October kill in that State 
is thus directed heavily upon the highly vulnerable late-hatched young. 
If we can say with certainty what fraction of the dove production comes 
off in each month of the season, it may be eventually possible to as-~ 
certain the relative proportions that early and late production contri- 
butes to the hunter's bag. At the present time, the life tables derived 
from banding data do not demonstrate--on a continental scale at least— 
a marked vulnerability of young birds to the gun. This may be later 
clarified when we know what fraction of the dove population is taken 
by hunters. 
Mean number of broods raised.-—-It seems impossible with the 
data at hand to settie the interesting question of the mean number of 
broods raised by mourning doves that start each nesting season. There 
are virtually no published observations on marked adults observed through- 
out the breeding season. Highly circumstantial evidence is frequently 
published on the maximum number of broods raised in one year by (unmarked) 
pairs that consecutively use the same nest site. Records of a few pairs 
that raise four such broods do not imply, however, that all pairs raise 
four broods. ; 
Thus if all the mourning doves in a region made four nesting 
attempts, and a 50 per cent nesting_success occurred at random through 
out the population, it can be shown? that 6.25 per cent of the birds 
would tend to raise four broods, 25 per cent three broods, 37.5 per 
cent two broods, 25 per cent only one brood, and 6.25 per cent no 
young at all. These calculations do not take into account adult mor- 
tality; they do yield a mean of two broods per pair. 
Similarly, if all the doves in a region made five nesting 
attempts, and a 50 per cent nesting success occurred at random through- 
out the population, 3.1 per cent would tend to raise five broods, 15.6 
per cent four broods, 31.2 per cent three broods, 31.2 per cent two 
brocds, 15.6 per cent one brood, and 3.1 per cent no young at all. 
These calculations again do not consider adult mortality; they yield 
a mean of 2.5 broods per pair alive at the start of the nesting season. 
- Adult mortality should lower the mean number of broods raised 
by pairs that start the breeding season; its net effect should be to 
decrease the number of brcods raised each month (exactly as Monk's brood 
data illustrate). Monthly variations in nesting success should, at 
least for some months, constitute an even more important variable, but 
cf the operation of thas factor, we know virtually nothing at this time. 

lpy expanding the binomial (a + b)4 where a = nesting success and b = 
nesting failure . 
106 
