100 SEAL LIFE ON THE PRIBILOF ISLANDS. 
not in the water or at the water’s edge or in positions where they could 
be supposed to have been cast up by storms. 
On August 2 I counted 23 dead pups on Ketavie rookery. These 
were all on the higher parts of the rookery, and there were none at the 
water’s edge. A few dead pups were found on the highest parts of sev- 
eral of the rookeries, and these, as Mr. Townsend suggested, were such 
as had been worried to death by the bachelors which hang about the 
rear of the rookeries. On several occasions I saw young bachelor seals 
Snatch pups away from the pods on the higher parts of the rookeries, 
and badger them about as if they had been cows. 
On Ketavie rookery, on July 27, I saw a youngish bull at the back of 
the rookery seize a pup out of one of the upper pods with his teeth and 
earry if up the hillside. He dropped it several times, but picked it up 
again. The pup presently bit at his assailant, and when the attention 
of the latter was distracted, ran away down the hill again. Another 
pup, apparently brought up the slope by this same bull, had crawled 
in under a rock, so that it could not be reached. 
On August 5, on Ketavie, | saw a bachelor attempting to copulate 
with one of these pups. The instinct of reproduction seems to spring 
up at an early age in the seals, and hence these unnatural assaults on 
the pups. I agree with Mr. Townsend in thinking that a portion, at 
least, of the dead pups found high up on the rookeries are such as have 
been worried to death in the manner indicated. Some of these pups, 
however, are doubtless such as have wandered away from their mothers, 
beyond the confines of the rookery, and have died of starvation. 
ESTIMATES OF NUMBER OF SEALS. 
When Mr. Townsend arrived on St. Paul in July we endeavored to 
obtain data for an estimate of the number of seals present. With this 
in view we went to Ketavie, Lukannon, Tolstoi, and the Lagoon rook- 
eries and made an actual count as far as the conditions would permit. 
As we proceeded it soon became apparent that to count all of the 
several classes of seals was impracticable. The first difficulty that we 
encountered lay in the fact that the roughness of the ground prevented 
us from seeing some of the harems distinctly, and we were unable to go 
down among the seals without stampeding them. Later we came upon 
the triangles or wedges of breeding seals massed as close together as it 
was possible for them to lie. It was necessary to use a field glass in 
investigating these groups, and as there were no salient points which 
could be picked out, as the count proceeded the eye soon became con- 
fused and the count lost. The larger groups of pups presented the 
same difficulty. This class was omitted later on, because on areas which 
had to be viewed from a distance so large a proportion of pups were 
concealed by the bowlders that the count could not be at all relied upon. 
We came eventually to count only the cows. At the Lagoon we counted 
by passing in front of the rookery in a boat, using a low-power glass. 
The harems were separated here by considerable intervals, and as the 
whole rookery was in plain view there was no obstacle to counting. 
On Ketavie we counted without a glass for the most part, going down 
among the seals as best we could, and the same course was pursued at 
Lukannon and Tolstoi. 
It can not be claimed that our count is mathematically accurate, but 
I believe that a sufficiently large area was covered to render the inac- 
curacy of comparatively little consequence. An actual census could 
be made only by driving all the seals inland and passing them in 
