ALASKA FISHERY AND FUR-SEAL INDUSTRIES^ 1920. 113 



This leads to interesting speculation upon the comparative mor- 

 tality of the sexes. But it will be sufficient here to dismiss the subject 

 with the observation that no record is known to me of any individual 

 bull having returned to the rookeries for more than three successive 

 years. It is believed, however, that in the majority of cases they do 

 return for at least five years. 



Undoubtedly a mistake was made once in the branding in sorting 

 the sexes, and a male pup got the mark across his back. He developed 

 and occupied the same position on East Cliffs Rookery for three suc- 

 cessive 3'ears, 1913, 1914, and 1915 In his last year he had to be at 

 least 12 years old. 



BULLS. 

 HAREM AND IDLE BULLS, 



At the height of the breeding season the bulls which had harems 

 and those usually termed idle were counted with the greatest possible 

 accuracy. Steps have already been taken to adopt certain rookery 

 improvements whereby this important count can be continued in the 

 future, regardless of how large the herd becomes. Experiments with 

 a stable skeleton tripod in 1920 demonstrated the superiority of such 

 a structure over any form of elevating apparatus jireviously tried. 

 The manner in which this was used is shown in figure 3. 



Bv placing such a tripod near the beach line and having a skeleton 

 walkway leading to it over the seals, a cheap and very effective 

 means of counting large flat areas will be provided. It is expected 

 that this will be tried out thoroughly in 1921. A few permanent 

 markers on the breeding areas to divide them into sections should 

 enable the enumerator to have absolute confidence in his result. 



On account of the absence of some trustworthy manner to get up 

 high enough to view the rookeries the harem counts since 1917, wlien 

 bulls Ijecame very abundant, have not l)een as good on the large 

 rookeries as desired. Many exjiedients have been used to help, such 

 as counting from a boat, erecting markers of driftwood, climbing a 

 long ladder, as well as natural elevations, which are sometimes 

 reached with great danger; but all were more or less makeshifts, 

 Avhich it is planned shall be abandoned. 



Although there doubtless is an error in the number of harems given 

 for the larger rookeries, it is reallv verv small and is conservatively 

 estimated to be not greater than one-hulf of 1 per cent. The desir- 

 al>ility of having even tliis small variation corrected is due to the fact 

 that this is the Ijasis of the census. A small error here will multiply 

 throughout the computations. 



The decrease in the number of bulls from 1919 was apparent not 

 only in the records; on tlie rookeries it was very much safer to ap- 

 proach counting points, and disturbance of even the rear harems 

 rarely resulted. There appeared to be a general " loosening up " of 

 the compact areas maintained when the bulls were stationed closer 

 together, so that the increase of area over 1919 api^eared to be more 

 than the figures indicated had been the increase in cows. 



If it is possible to do so in the future, the counting dates on St. 

 <^ieorge Island should be set back to correspond with the beginning of 

 the count on St. Paul Island. It has heretofore been customary to 



54777°— 21 8 



