80 THE FUR SEALS OF THE PRIBILOF ISLANDS. 
fur-seal rookeries. The surveys of the rookeries themselves can not be verified, for 
the conditions have changed with the reduction of the herd, and no permanent land- 
marks were left. Not even of the survey of 1890 is there left a single recognizable. 
stake or stone to show that it ever existed. All that is left of either survey is 
the unsatisfactory estimate of the seals based upon it. These surveys should 
have formed the basis for subsequent comparisons of the condition of the rookeries. 
As such they would have been extremely valuable, but all traces of them have 
disappeared. 
THE SURVEYS CAN NOT BE VERIFIED. 
It is therefore not possible for us to verify Mr. Elliott’s surveys of the rookeries, 
but his maps giving the shore lime of the islands are available as a measure of’ his 
work as a surveyor. Of these maps Captain Moser, in his hydrographic report! on 
the islands in 1896, made certain tests. Of Mr. Elliott’s shore line he says: “It was 
a bad misfit * * * and rarely stood the test of an instrumental angle.” He 
further says of the topography of the maps that ‘it is so vague and indefinite that 
it is next thing to impossible to do anything with them; I should call them sketches.” 
If this is true of the fixed and permanent shore line, it is not to be supposed that 
the changing rookery margins, which were necessarily noted from a distance in the 
summer and measured in winter, after they had melted away, were more correctly 
located. 
THE EFFECT OF INACCURATE SURVEYS. 
The correctness of the survey of the rookeries is of vital importance to the 
accuracy of this enumeration. This importance does not lie in ascertaining the 
mere length of a given rookery. This can be easily obtained, and’ in any event a 
mistake of a few feet or of a hundred feet in the length is comparatively insignificant : 
but the width of the rookery is another matter. To each one of seven of the ten 
rookeries of St. Paul Island, Mr. Elliott ascribes an even average width of 150 feet. 
Two of the remaining breeding grounds have a width of 100 feet each, and the third 
40 feet. Therefore, for the 40,000 feet of rookery shore line on this island, 35,000 have 
an average width of 150 feet.2, Suppose there is an error of but 1 foot in this average 
width, it is multiplied throughout the entire distance. According to the method of 
the computation involved this would mean the addition or subtraction of 17,500 
animals, depending upon the side upon which the error falls. Again, suppose the 
average width was 140 or 160 feet, this would mean a difference of 175,000 seals one 
way or the other, as the case might be. 
AN INADEQUATE UNIT OF SPACE. 
But aside from the question of accuracy in the surveys themselves, Mr. Elliott 
has assigued an impossible space to each individual seal. His unit of space is 2 
square feet to each animal, young or old, or 4 square feet for the cows, ignoring the 
! Hydrographic Notes, Captain Moser, Part III. 
® Whatever the average width of each rookery may have been, it is certain that it was not the 
same for all. Neither now nor at any past time have Tolstoi, Polovina, Vostochni, the Reef, Kitovi, 
Lukanin, and Zapadni had the same ‘average width.” The 150 feet is a guess, and that only. 
