
Following a line of reasoning similar to that used in com- 
puting the size of the pup crop from the commercial kill, the present 
size of the herd is estimated to be: mean size of the herd 1928-32, or 
1,047,037, times 1.5807 equals 1,655,051 or, rounded, 1,660,000. 
It should be pointed out that the herd size in 1928-32 was 
computed on an assumed pregnancy rate of 100 percent, whereas it is now 
estimated to be approximately 70 percent and probably was nearer 80 
percent at that time. Furthermore other errors are introduced by the 
increase in pup mortality in recent years. Both of these factors tend 
to make the estimate determined from the trend of the commercial kill 
too low; hence the estimate is quite in line with the one noted in the 
previous section (1,840,000). 
Estimate of the size of the herd from the 
trend of the harem~bull count 
| In the preceding section we have used the increase (mean 
1928-32 to mean 1947-51) of bachelors available during the harvest 
geason as a rough index of the increase in total size of the herd. The 
increase in harem bulls can be used similarly, since the number of harem 
bulls is directly related to the escapement from the harvest. But the 
number of harem bulls is also related to the number of mature cows on 
land since, by definition, a harem bull is a bull holding one or more 
cows. The extent to which the number of cows affects the standing size 
of the harem-bull class is not known, for the cow-to-bull relation is 
exceedingly complex. For the moment, the assumptions are made (1) that 
uniform sealing practices since 1924 have perpetuated a uniform annual 
percentage escapement and (2) that the harem-bull class has remained 
in approximately the same relation to the cow class during these years. 
In table 5 the smoothed increase in harem bulls from 1924 to 1951 is 
shown. The present size of the herd is estimated to be: mean size of 
herd 1928-32, or 1,047,037, times 1.4757 equals 1,545,113 or, rounded, 
1,550,000. 
Again there are known errors, for which it is impossible to 
correct quantitatively. An error in the assumed pregnancy rate in the 
1928-32 computations would enter here also, while changes in pup 
mortality and harem size (appendix C) would affect the estimate. 
-43- 
