36 BULLETIN OF THE BUREAU OF FISHERIES. 



were present in large numbers. The estimate of 3-year-old males at the beginning of 

 1914, therefore, is 13,880. 



Four-year-old males . — The 4- year-olds living in I9i4were bomin the summer of 1910. 

 There was no count of pups in that year but an estimate was made by the average harem 

 method discussed under the estimate of 3-year-olds (p. 34). Using this method, the 

 average harem on Kitovi, which was 31.7, was multiplied by the total number of harems, 

 giving as a result 43,777. The estimated error in the method requires an addition of 

 21,888 which gives 65,665 as the probable increase for 1910. Reducing this by 50 per 

 cent for the first year's mortality and 15 per cent for the second leaves 27,907 of both 

 sexes or 13,954 2-year-old males in 1912. Although the killings in 1912 were supposed 

 to include a considerable number of 2-year-olds, the records show only 541 yielding skins 

 that weighed less than 5^ pounds and these, therefore, are all that can safely be assumed 

 as 2-year-olds. Deducting these and in addition 10 per cent for natural deaths in the 

 third year, reduces the total to 12,072 3-year-olds at the opening of the season of 1913. 

 Of these, 1,610 were killed having skins weighing from 5^ to 8 pounds, inclusive, and 

 therefore 10,462 were left. The natural mortality in the fourth year is believed to be 

 very small, perhaps negligible, but it may be granted for the sake of conservatism that 

 it is as much as 5 per cent. Therefore the estimated number of 4-year-olds in 1914 is 

 10,462 less 523, or 9,939. Although no exact count was possible, the number of 4-year- 

 olds observed on the hauling grounds in the season of 191 4 was sufficient to make it 

 reasonably certain that this estimate is not beyond the facts. They were seen on all the 

 hauling grounds and in some cases constituted fully 25 per cent of the bachelors present 

 at a given time. 



Five-year-old males. — The 5-year-olds of 191 4 are of the generation of 1909 and were 

 subjected to both land and pelagic sealing as 2 -year-olds. The best method of estimating 

 their present numbers is based on the breeding reserve of 191 2. In that year, before 

 killing began, 2,005 bachelors supposed to be 3-year-olds were given a temporary brand and 

 reserved as breeders, exempt from killing for at least one season. In 1913, they became 

 4-year-olds and if we allow 5 per cent for possible deaths during the winter, they then 

 numbered 1,905. Since killing in 191 3 was restricted largely to 3-year-olds, the re- 

 serve of the previous year was subjected to no more than a slight decrease. According 

 to the records, 247 skins weighing over 8 pounds and under 1 2 pounds were taken in 

 1913, and at least the majority of these were probably 4-year-olds. This leaves 1,658, 

 and since the loss in the fifth year is doubtless too small to be taken into account, it is 

 safe to say there were 1,658 5-year-old males living in 1914. About one-third of them 

 were to be seen on land at any time during the season of 191 4. A count of half bulls 

 made practically simultaneously on all the hauling grounds in 1914 showed 748 present, 

 and of these it can be said that 5-year-olds constituted 80 per cent or more. It is probable 

 also that more than 2,000 escaped in 1912 and 1913, since the killings in those years were 

 confined to a few hauling grounds and since the records of weights show that no very 

 large number of 2-year-olds was killed on land in 191 1. It is evident, therefore, that the 

 estimate of 5-year-olds is well within the facts. Next year, there should be from i ,500 

 to 2,000 lusty 6-year-olds ready if necessary for harem duty. 



Summary of estimates of nonhreeding seals. — The following table shows the number 

 of nonbreeding seals estimated to be present in 1914, and the method of making the 

 estimate : 



