December 27, 1912] 



SCIENCE 



895 



the results about the same. Mr. Elliot found 

 a total of 3,193,000 breeding seals and young, 

 Y2,000 less than Captain Bryant. He esti- 

 mated the non-breeding seals also, finding a 

 number sufficient to bring the grand total for 

 the herd up to 4,700,000 animals of all classes. 



In 1890 Mr. Elliott duplicated his census of 

 1872-Y4. It was a greatly reduced herd he 

 found at that date. The breeding area he 

 estimated at 1,900,000 square feet, and ap- 

 plying to this the same space unit, found 

 950,000 breeding seals and young. 



The next serious attempt to estimate the 

 herd was made in 1895 by Mr. Frederick W. 

 True, o:^ the Smithsonian Institution, and Mr. 

 Charles H. Townsend, then connected with 

 the United States Bureau of Fisheries. The 

 herd had suffered still further decline through 

 the ravages of pelagic sealing, an indiscrimi- 

 nate form of hunting in the open sea partic- 

 ularly destructive to the breeding females. 

 Messrs. True and Townsend were able to count 

 the individual animals on certain breeding 

 areas, 7,479 in all. From charts of the rook- 

 eries, on which the areas had been traced at 

 the height of the season, the extent of the 

 counted area was obtained and hence an in- 

 dividual unit of space. Each animal was 

 found to occupy a space of 46 square feet on 

 scattered breeding grounds and half this space 

 on massed grounds. Completing the measure 

 of breeding space for all the rookeries, from 

 the charts, and applying to it the units of 

 space, a total of 131,833 breeding seals and 

 young was found, with non-breeding seals 

 enough to bring the total for all classes up to 

 155,977. 



Coincident with the above estimate was one 

 made by Colonel Joseph Murray, a govern- 

 ment agent on the islands. He estimated and 

 counted the breeding families, 5,000, and as- 

 signed arbitrarily an average of 40 cows to 

 each, thus reaching a total of 405,000 breed- 

 ing seals and young. 



In 1896 a new investigation of the fur seal 

 herd was begun by a commission under the 

 leadership of President Jordan, of Stanford 

 University. Both the above estimates were be- 

 fore this commission. The method of obtain- 



ing the unit of space used by Messrs. True 

 and Townsend commended itself as worthy of 

 imitation, but on test the rookery charts were 

 found unreliable and a new basis of estimate 

 was sought. The areas on which individual 

 cows had been counted in 1895 were re- 

 counted, and enough additional space to bring 

 the total up to 16,679 individual cows, in 

 1,245 families, an average of 13 cows to a 

 harem. A complete count of harems was 

 then made with the intention of applying the 

 average to it after the manner of Colonel 

 Murray. 



While this census of 1896 was in progress, 

 however, it was discovered that there were 

 more pups on the counted areas than the 

 number of cows previously counted war- 

 ranted, and a full count of pups showed them 

 to outnumber the cows two to one. In all 

 previous estimates it had been assumed that 

 at the period of development in rookery pop- 

 ulation known as the height of the season all 

 or practically all the cows were present. The 

 count of pups proved this to be an erroneous 

 assumption, that in fact when most of the 

 cows were present half at least of them were 

 at sea feeding. The average harem obtained 

 from the count of cows was therefore aban- 

 doned and one obtained from the count of 

 pups substituted. This gave a total of 157,- 

 405 tjows, with a like number of pups, in 4,932 

 harems, or a total of 319,742 breeding seals 

 and young. The estimate for non-breeding 

 seals in 1896 brought the total for all classes 

 up to 450,000 animals. 



In 1909 the writer duplicated this census 

 of 1896, finding 50,626 cows in 1,387 harems, 

 or, adding a like number of pups, a total of 

 102,639 breeding seals and young, with non- 

 breeding seals sufficient to bring the total for 

 the herd up to 158,520 animals. 



The method of enumeration thus estab- 

 lished in 1896 has been continued each sea- 

 son since with slight variation. The method 

 of estimate was not held to be exact. It was 

 recognized that exact results could only be ob- 

 tained by a full count of pups and this was 

 considered in 1896 to be physically impossible. 

 The chief importance of the enumeration. 



