112 



ALASKA FISHERIES AND FUE INDUSTRIES IN 1917. 



have the means of getting this unit from actual counts on known 

 areas. He appears to have been unduly influenced by large level 

 massed areas and did not allow for much more enormous spaces, 

 fully occupied, but less dense, owing to topographical features. At 

 any rate, his results differed widely from conditions as they exist 

 to-day. 



In 1917 careful counts were made of harems on all rookeries and 

 entire pup counts on 11 rookeries. Then the areas of all breeding 

 grounds were ascertained. Thus there was secured the unit, as wefl 

 as it was possible to get it, on 11 rookeries. These take in aU types 

 of topography found here. Some were small rookeries, but on the 

 major ones the unit runs from 7 to 9 s(juare feet per breeding seal. 

 The following table gives this information concretely: 



Areas of Pribilop Islands Rookeries in 1917 and Seal Units. 



o Rookeries marked with an asterisk are those upon which complete counts of pups were rot made and 

 these were estimated by the average harem method as explained on page 108. South rookery, St. George 

 Island, on which all pups were counted, is omitted from ihis tabulation. It had but 24 pups in 1917. 



The average area per seal is seen to be about 6 square feet. The 

 area per bull was obtained by dividing the areas by the number of 

 harems which were found; and the distance each bull was apart is 

 the square root of the area each occupied. 



Such data as these are very valuable for use in conjunction with 



{)artial counts for arriving at a complete census after the herd is too 

 arge for all pups to be counted^ It is to be regretted they have not 

 been obtained since 1912 along with the complete counts. We 

 would now have had concrete knowledge of the seal unit of area such 

 as is obtainable only when the herd is small. 



