CENSUS OF SEAL LIFE IMPOSSIBLE. 89 



occupied. Then when the seals were gone we sighted along these 

 stakes to determine the back lines of the rookeries and measured the 

 areas thus determined with a tape hue, using our judgment by observ- 

 ing the nature of the ground to determine the curvature of these areas. 

 We then calculated from our observations three seals to a square yard, 

 and multiplying the yards in the areas measured by three made our 

 estimate. I think the measurements were made as accurately as could 

 be done by the means and instruments employed; however, I am con- 

 vinced that no estimate of any kind, no matter how accurately the 

 measurements are made, would give even approximately the number ot 

 seals on the island, for the animals are constantly in motion, coming 

 and going, and there seems to be almost as many in the water as on 

 land. It is as impossible to estimate them as it is to estimate a swarm 

 of bees. But accurate measurements would show conclusively, if made 

 from year to year, whether or not the seals were increasing or decreas- 

 ing. 



I do not think that the number of seals on the rookeries can be ev*m 

 approximately estimated. No satisfactory meas- 

 urement of the breeding grounds on which to base h. a. GUdden, p. 110. 

 an approximation of the number of seals has ever 



been or can be made. And, even if such measurement could be made, 

 the broken nature of the ground, the inequality of distribution of the 

 seals while on land, and the fact that the females are constantly coming 

 and going, preclude the possibility of any sort of calculation which could 

 be of any value at all. 



Even if these measurements had been correct, which was impossible, 

 I do not believe it is possible to calculate even ap- 

 proximately the number of seals upon the rooker- AUal P. Loud, p. 88. 

 ies, because of the broken nature of the ground 

 and the irregular outlines of the breeding grounds. 



The total number of seals was stated in that report to be " not less 

 than 4,000,000 upon the two islands." I am satisfied 

 that this estimate was too high, and that the more H. E. Mclntyre, p. 48. 

 recent estimates published in the reports of officers 



of the Treasury Department who have been at different times stationed 

 upon the islands, or detailed to report upon the sealeries, have been still 

 more erroneous than my own. My figures were made without any at- 

 tempt at mathematical computation, and were mere guesses at the pos- 

 sible number of seals upon the different rookeries. 



My successors have attempted to measure the ground occupied by 

 the seals, and by multiplying the number upon a given area as ascer- 

 tained by count, by the whole area of the rookeries, to arrive at an ap- 

 proximation to the total number. They added to their computation a 

 large percentage to cover the number supposed to be in the water at 

 the time, but did not subtract for the inaccessible portions of the 

 grounds, vast tracts of which are covered with bowlders and lava rocks, 

 where no seals could lie, or skirted with acclivities they could not as- 

 cend. Thatis, the estimates weremade from measurements necessarily 

 taken after the seals had left the rookeries, and sometimes weeks or 

 months afterward, with only the recollection of the ground they had 

 formerly occupied to guide the observer. Many sections were included 



