116 



ALASKA FISHERIES AND FUR INDUSTBIES IN 1917. 



bear this in mind, so that it can be made such use of as conditions 

 may warrant. 



The number of dead pups on those rookeries where they were not 

 counted was arrived at in the same manner as the average harem; 

 that is, the percentage of dead on the areas counted was compared 

 with that of the herd in years past and a percentage assumed in 1917, 

 which was in accordance with these facts. While the result can not 

 be said to be known to be absolutely correct, it must be admitted 

 that it can not be far from the actual conditions which existed. In 

 tabular form the figures show as follows: 



Percentages of Dead Pups in 1914, 1915, 1916, and 1917 on Rookeries on which 

 THE Dead Pups were Counted in 1917. 



For the purposes in hand, the significant portions of this table are 

 the totals, which, properly arranged, appear as follows: 



a From estimates. 



It is thus seen that the percentage of all dead pups in 1917 ust 

 be less than 3.43 on St. Paul Island, more than 1.43 on St. George 

 Island, and less than 3.34 for the entire herd. To produce this 

 result, the figures ^iven in the main table, page 103, have been used 

 and seem very satisfactory for all requirements. 



It might be believed b^ one Httle famihar with seal hfe that when 

 a complete pup count is impracticable a few rookeries could be 

 counted to get the annual increase, and by simply applying this 

 factor to the entire herd a census would be arrived at. Tnis, how- 

 ever, might give very misleading results with the rookeries as they 



