106 U. S. BUREAU OF FISHERIES. 



Increase or Decrease in Number of Pups in 1920 from 1919. 



'Pups counted. 



It will be noted that upon some of the eight rookeries counted there 

 has been no gain, whereas on others it has been considerable, and 

 again that the gain on the lot has been only 3.77 per cent. This latter 

 fact indicates very strongly that the gain of breeding cows has not 

 been the 9 per cent which has been considered to be the normal rate. 



It is true that the smaller rookeries do not increase as rapidly as 

 the large ones, but there sliouhl not be this mucli discrepancy even 

 though it is possible. It is entirely possible for the rookeries counted 

 to have lost and the herd as a whole gain the normal amount, but it 

 is not probable. The abnormal gain on the other rookeries would 

 have been a noteworthy feature to a person who had had them in 

 view during successive years. After a consideration of all of the 

 factors v,hich enter into the problem, the gains on the several rook- 

 eries have been estimated as shown in the above table. This, of 

 course, is made up largely from the data on average harems (p. 116). 

 It shows the gain of pups (and breeding cows) for the entire herd 

 to be 6.59 per cent. 



A noteworthy feature of the rookeries on which counts were made 

 is the actual decrease in the number as well as the percentage of pups 

 found dead in spite of an actual increase of total pups. This is entirely 

 in accordance Avith predictions as to what would liappen when the 

 size of the harem would be increa.sed. By using the figures of dead 

 pups found on the various rookeries the numbers have been esti- 

 mated for those which were not counted. This is shown in the table 



