FUR-SEAL INDUSTRY OF THE COMMANDER ISLANDS 
309 
with exactness, and that is the only figure that is of prime importance. With 
that given, the number of the otlier classes can be observed and computed with suffi- 
cient accuracy for all practical purposes. 
In 1897 the writei- declared that a count of the seals such as was undertaken on 
the Pribilofs could not be carried out on the Commander Islands owing to the 
difference in character of the rookery beaches on the two groups of islands. 
With the excessive shrinkage of the number of seals during the next 13 years, 
however, it seems certain from what was seen on North Rookery, Bering Island, in 
1922, that a count of the black pups would have been feasible there in 1910, though 
it would still be impracticable on most of the rookeries on Copper Island. Moreover, 
on account of lack of suitable means of transportation, one person would be utterly 
unable to cover all the rookeries during the short period in which such a count must 
be made. 
As Mr. Suvorof had no previous practical experience and probably had no 
detailed description at hand of the technique of counting the pups as developed on 
the American islands, he was restricted to his own resources and the application 
of some of the methods mentioned above, as definite statistics and not vague 
estimates — which can be of value only when done by persons of wide experience and 
tried judgment - were evidently expected hy the authorities in St. Petersburg, who 
had been accustomed to receive most minute figures from the previous administration 
of the island based on the fantastic "counts" which the native guards, stationed at 
the various rookeries, made at certain intervals during each season. When con- 
fronted with the Bering Island rookeries — Reef and Kishotchnoye — his first 
attempt was directed at ascertaining the number of seals present by the area method. 
Red and white marks and signals were painted all over the two rookeries, and 
calculations of various sorts started. Unfortunately, or fortunately, this work had 
to be left unfinished, as by the middle of July (old style) he had to leave for Copper 
Island. Here that method was plainly impossible. 
By that time he had also discovered, what we others had learned long before, that 
even if one can get near enough to even a depleted rookery to start an autoptic count 
of the seals in sight, "part of the seals always slip away from the counting, and the 
figures obtained invariably prove lower than the actual ones. * * * Part of the 
animals are in motion, some come out of the water, others wend their way to the sea; 
many are concealed by the unevenness of the beach, by jutting rocks and cliffs. 
Many new-born pups, in particular, are concealed in the crevices between the rocks." 
He also observed that the weather, the temperature, and the time of the day made 
a big difference. Nevertheless, he decided that this was the only method possible 
on Copper Island, and he believed "that the error of the per capita count of the 
adults would not exceed 10 per cent and of the new-born pups 20 to 25 per cent," 
adding that "the method of area never gave such exactness" (Komandorskie 
Ostrova, 1912, pp. 181-2). 
During the latter part of July and the beginning of August, 1910 (old style), 
therefore, he "visited all the Copper Island rookeries and carefully counted all the 
black pups lying on the beach and splashing in the near-shore zone." To the 
number thus obtained he "added 141/^ per cent for the females without pups, and 
