18 . 



this year the sealers entered Bering Sea. It is probable that with the doubling of 

 the Northwest catch in 1881, which, in addition to the loss of mothers it involved, 

 would produce an increased secondary loss in 1.S83 by the failure of a certain num- 

 ber of the i>ups of 1881 to return as' breecb-rs, the decline of the herd began. In 

 1886 the Northwest catch fell Irom 21,000 to 13,000, but the Bering Sea catch rose 

 suddenly to 14,000. 



DECLINE OF THE HERD. 



From this tiuu- on the decline was more rapid and has been continuous, though 

 there is evidence that the modus vivendi of 1892-93, by which Bering Sea was 

 closed to the sealing deet, has produceil tor I8O.0 and 1896 a slight check of the 

 diminution. 'I'he reason for this is that in addition to the saving of mothers, no 

 pups were starved to death in 1892 or 1S93, and those which might have been starved 

 have returned as Inceders or as lullable seals in 1895 and 1^91!. The cessation of 

 laud killing which took place at the same time has produced only harm to the herd 

 by the needless auguientation of tiie number of 6 and 7 year old bulls which now 

 struggle in vain to enter the rookeries at the height of the season. There has never 

 been a time since 1870 at least when the cessation of laud killing could have been 

 helpful to the herd on the Pribilofs. 



KILLABLE .SEALS ONLY NOTICED 1$Y MOST OBSEMVEL'S. 



The facts and dates aliove given can only be ajiproxiniately stated. In past times 

 the interest of all persons concerned has centered in the nunilier of killable seals, 

 while the comlitiou and nuniliers of the l)reeding females have received little atten- 

 tion. Almost (^very account of the diminution of numbers is based on the fewness 

 of bachelors or on the deserted condition of tlie grasps grown hauling grounds. 

 Decrease in numbers of killable seals is not a canse in itself. It does not appear till 

 the cause has passed. For example, the killable seals in 1890 are largely the pupsof 

 1893, and their number depends on the conditions surrounding the females of that 

 year. If these pups were killed as 2-year-olds in 1895, the hauling grounds would 

 be empty in 1^96, even though there were no deciease of females. The number of 

 killable seals was ])rol)abIy fewer in 1890 than at any iteriod before or since. This is 

 due in part to the losses from pelagic sealing, but primarily to the fact that the nat- 

 ural bachelor (juota of tliat year had been taken in 18>!9 or even in 1888. But this 

 fact in itself did not indicate diminution of the breeding herd — only the premature 

 gathering of its marketable increment. 



We have exi)licit records of the nnud)er (sf b.achelors killed year by year, and of 

 the dates at which it was ])ossible eacli year to till the (|uota. But of the numbers 

 of females and pups we have until 1895 no exact records whatever. 



DISTlilBUTION OF SEALS ON KOOKEBIES. 



In general, tlie greater the number of females the more extended are the bounds of 

 the rookeries in the height of the season. This general rule can not be usctl for exact 

 computation of numbers, because the rookeries often grow sparse as the seals dimin- 

 ish in numbers without material change in dimensions. ^lorcover, in all cases, at 

 least at the jirescnt time, * the seals are very unevenly distributed, their arrangenumt 

 being as unecjual as that of trees in a forest. On some rookeries, as on Tolstoi sands 

 at the height of the season, the seals lie thick as swarms of bees, almost all the sur- 

 face being covered and each adult seal being restricted to about 12sc|uare feet (11,775 

 cows estimated on about 140,000 S(iuare feet of surface). On other rookeries, as the 

 Lagoon, detached harems si)rawl over the rocks and each seal has upward of 35 t 

 Sfpiare feet. Betwei-u these every intermediate condition may be found, and the 

 attempt to establish an average acreage of breeding seals is the continuous multipli- 

 cation of estimates by assumptions. In some cases errors may be made to balance 

 errors. Where errors do not balance each other comes the curious result that the 

 most accurate estimate is that farthest from the truth. 



ELLIOTT'S CENSUS OF FUR SEALS. 



The first serious attempt to estimate the number of seals on the Pribilof Islands 

 is that of Mr. Henry W. Elliott in 1874. Mr. Elliott recognized first the important 

 fact that "the mother seals are coming and going," while the pups, remaining near 



*" According to Mr. Elliott, they were everywhere as close together as they could 

 lie in 1874. 



t Sixty-five srjuare feet, according to Messrs. True and Townsend, but their count 

 was made when about half the females Avere absent. 



