FUR-SEAL CENSUS, PRIBILOF ISLANDS, 1918. 117 
able information that this is about as low as it can go. As long as 
the percentage of nonbreeding bulls to breeding bulls remains as 
high as it has been in these two years there is no probability that the 
average harem will materially change for the herd. 
There is a general stability or sameness developing on the rookeries 
due to the excess of bulls, and the deviation of the average harem on 
certain rookeries from the general average seems now to be due 
largely to topography of the breeding areas. This is indicated by 
the similarity of the average harems on rookeries counted the past 
two seasons. 
Since pups could not be counted on all rookeries, the average harem 
had to be computed for many of them. On account of the closeness 
of the 1917 and 1918 figures on those areas counted it was placed at 
the same as the 1917 figure in most cases. The only exceptions to 
this rule were in those cases in which field observations were con- 
vineing proof that the 1917 condition was abnormal and had 
disappeared. 
Since so many factors relating to the pups are directly dependent 
upon the average harem, this has formed the basis of most of the com- 
putations found in the following tables. Thus it is well known that 
the percentage of dead pups on a given rookery is dependent on the 
topography of the rookeries, the number of harem bulls, and the 
average harem. With these factors constant, the percentage must 
be constant when pelagic sealing is nonexistent and no epidemic 
occurs, no evidence of either of which was observed in 1918. There- 
fore it seems justifiable to place the percentage of dead on rookeries 
not actually counted at the same as was carefully computed for 1917. 
This brings it to 3 per cent for the herd. 
DISTRIBUTION OF Pups IN 1918. 


Living Dead Total | Percent- 
Rookery. Date of counts. pups. pups. pups. |age dead. 
—_——————q— | —_qe | m | uu 
ST. PAUL ISLAND, 





2,502 38 | 2,540 1.49 
2)174 61| 2/235 2.72 
9) 154 332 | a9) 486 3.49 
778 14] 792 1.76 
18, 890 508] 19,398 2.61 
6, 167 158 | 6,325 2.49 
493 19 512 3.71 
16,349 376 | 216,725 2.24 
11, 156 304 | @ 11, 460 2.65 
8) 795 294 | a9/089 3.23 
517 19 536 3.54 
5, 191 152| 5,343 2. 84 
1, 824 58 | @ 1/882 3.08 
le F 1473 18 | 1,491 1.20 
MORO WIN Ase eee sa Soa n ee Ge bo nioe oe See 3, 224 lll @ 3,335 3.32 
Wonumbiriassn os rem hoe 30,057 | 1,411 | #31, 468 4.48 
AS EE Ne See Ee ae re 118,744 | 3,873 | 122,617 3.15 
ST. GEORGE ISLAND. 
cn SS SS eee? 0 ee re 6, 837 174| «7,011 2. 48 
Se SE SE aE) oe Sea 5,332 108 | 45,440 1.98 
jen Eri hs 2 SP oe ee eee se See eS Bmp 14. STIS 796 10 1.24 
Pee Seite a re 5d05.o ee ("| Ee ae 1 [ee 
pubdaest’ gpeteeus |<) Ogee (eT SRI 2,092 28| 2,120 1.31 
PORN. NEC DELS SO RUE OF EC RE eo 4751 109 | @4)860 2.24 
ets. eens: ree ito 19, 869 429 | 20,298 2.11 
Utae, (anak elena ee tT 138,613 | 4,302 | 142, 915 3.01 


2 Based on estimated average harem, 
