56 BULLETIN OF THE BUREAU OF FISHERIES. 



The average price received for land-taken skins in 1909 was $34.68, so in general 

 terms it may be said that wigs in that year brought at least two-thirds as much as 

 2 and 3 year olds. In view of these figures, it is obvious that prime skins of 5 and 6 

 year old fur seals would always command a price that would compensate for a con- 

 siderable proportion of the revenue which might have accrued if they had been taken 

 as 3-year-olds. Even if no figures were available, it would still be evident that the 

 skins of young bulls must have a substantial value. The obser\-er, seeing these fine 

 young animals in prime condition, and knowing the growing demand for fur of all kinds 

 and the wide range of uses to which it may be put, can not fail to conclude that the skins 

 of young bulls would find a ready market and bring profitable prices. Therefore, while 

 failure to kill at the proper time causes a financial loss, it does not cause a total loss. 

 In emergencies, such as will exist in 19 15 when nearly 10,000 5-year-olds will be 

 present, it would be possible to derive at least a moderate revenue from seals of this class. 



THE AVERAGE HAREM. 

 VALUE OF THE AVERAGE HAE^M. 



The average harem, as commonly considered, is the average number of old cows 

 held in harems by old bulls in a given season. It is calculated by dividing the total 

 number of pups (equivalent to cows) by the total number of harems, and may be obtained 

 for a single rookery or for the whole herd. During pelagic sealing, when full counts of 

 pups were not feasible, the average harem was used as a means of estimating the total 

 number of cows and pups by counting pups on one or several rookeries and applying 

 the average thus obtained to all the rookeries. Estimates made in this way, as shown 

 by recent full counts of pups, were apparently from 30 per cent to 40 per cent too small 

 (see this report, p. 34), but their comparative value is nevertheless very great. For 

 this reason it is highly important that this information be obtained whenever it is not 

 possible to make a full count of pups. 



The average harem is of value also as an indication of the relative strength of the 



breeding males. Thus in 1914 an increase in harem buUs without a proportionate 



increase in bearing cows caused a reduction of the size of the average harem. Further 



reduction may be expected with further increment of bulls and regulation of the size 



of the average harem may be accomplished by increasing or decreasing the number of 



bulls. 



VARIATION OF AVERAGE HAREMS. 



Variation of average harems is of two kinds, that of seasons, from one to another, 

 and that of different rookeries in the same season. From season to season the average 

 harem of the total herd has varied in recent years within relatively small limits. Thus 

 in 1912 it was 60.4, in 1913 it rose to 65.8, and in 1914 dropped to 59.8. During the 

 same period it showed a net decline for St. Paul Island and a continuous rise for St. 

 George Island. Variation on the different rookeries from year to year keeps within 

 moderate limits in most cases, and this may be taken as some indication that, unless 

 unduly influenced, the same set of animals repairs each year to the same rookery. In 

 other cases, the average harem shows sudden and pronounced ups or downs which are 

 sometimes obviously due to well-known abnormal conditions and are sometimes wholly 

 inexplicable. A summary of average harem data follows: 



