172 
ZOOLOGY: G. H. PARKER 
instead of 12.5 and in 1914, 7.8 instead of 1.1 and the general average would 
be 9.2 instead of 9.4. Thus for the last five years the annual increase of pups 
has averaged a little over 9% and is probably nearer 9.5 than 9. Hence both 
Clark's figures and Hanna's are obviously too high for the present state of the 
herd. In fact Clark's assumption of 13% is quite unwarranted for he gives no 
reason for accepting this number rather than the other chosen by him, 12.5. 
The average rate of increase indicated in table 2, 9.4%, would bring 
about a doubling of the number of pups in approximately eight years. If no 
untoward conditions arise, it is fair to expect that in 1920 about twice as 
many pups will be born as were born in 1912. 
Beside the annual increase in the number of pups born, the Alaskan fur 
seal herd has given evidence of growth in the increase of its harem bulls. 
These bulls, which are the sexually mature males that have succeeded in 
having associated with them one or more breeding females, are perhaps the 
most accurately counted element in the whole herd. The determination of 
their numbers between 1912 and 1917 was made by direct count. The re- 
sults of this census are given in the column in table 1 marked 'harem bulls' 
where it will be seen that they represent an increasing series running from 
1358 in 1912 to 4850 in 1917. The distribution of their rates of increase de- 
scribes a curve (Graph 2, A) not unlike that seen for the pups. An extra- 
polation on the basis of this curve leads to the prediction that the number 
of harem bulls in 1918 will be somewhat over 6000. 
The average harem for a given season is the average number of females 
associated with the harem bull of that season. It is a derived number found 
by dividing the number of females as indicated by their pups for a given 
season by the number of harem bulls for that season (table 1). If the males 
of the herd have been considerably reduced in numbers, as might result 
from excessive killing, the number in proportion to that of the females would 
be naturally small and conseauently the average harem would be large. 
Such a condition, if excessive, would be an unfavorable sign in the herd 
and improvement would be marked by a decrease in the size of the average 
harem. The course of events in this particular between 1912 and 1917 is 
represented in table 1 the details of which can be better appreciated by ref- 
erence to Graph 3. In this it will be seen that the average harem presented 
its most unfavorable condition in 1913 after which there was a steady im- 
provement to 1917 in that the average number of females to each harem bull 
fell from 65.8 to 26.4. It is to be remembered, however, that the extremely un- 
favorable point in the curve, 1913, is dependent upon the anomalous pup 
count of that year. Had this count been in line with the others, the average 
harem for 1913 would have been very near those of 1912 and 1914. The 
subsequent change, however, would not have been affected and this reduc- 
tion can not be looked upon as anything but a favorable sign. From the 
extrapolation in Graph 3, there is good reason to believe that the average 
harem will be even smaller in 1918 than it was in 1917. 
