- 11 - 
raised to 13,500 adult group animals. 
There is little actual evidence that the tagged seals are distri¬ 
buted at random within the atoll. However, if tagged seals do not mix 
randomly through the population, that is if each individual tends toward 
exclusive use of a given area, then an excessively high percentage of 
tagged individuals would be expected. As this was not the case, the 
estimate tends to be conservative. If the tagging experience caused 
the animal to leave the atoll the estimate would be biased in the op¬ 
posite direction. Repeat observations of all but nine tagged animals 
rule out this factor. In spite of the fact that movement is poorly 
documented it must be remembered that observations on the uninhabited 
islands of Laysan, Lisianski, and Pearl and Hermes are extremely scarce. 
The documentation of any inter-atoll movement must then be largely due 
to chance, and one should probably expect more than has been reported. 
The mean percentage of new adult group animals observed for 22 months 
is 47.2 (SD 21.6). This means that anywhere from 25 to 69 percent of the 
adult group animals observed in a given month were not seen during the 
• previous one. The 140 adult group seals which were observed in more than 
one month at Kure had a collective total of 310 absences from 1 to 12 
months in length. The mode of these absences is one month and the median 
two. 
The limited movement recorded, the percentage of new individuals, 
and the length of absences suggest a slow steady turnover in the composi- 
tion of the adult group utilizing a given island. This may be a cycling 
through all utilized Leeward islands or primarily exchange between two or 
