In this examination of the published data on common terns, 
the trapping efficiency of Austin and his co-workers has not yet been 
mentioned. This is a very difficult statistic to calculate, and 
Austin (1949) has discussed its complexity in detail. Upon this 
point rests the precise calculation of a mortality rate from the 
retrap data; this is a task for the original bander, not a reviewer. 
In claiming that Austin's (192) data suggest a fluctuating adult 
mortality rate varying from "178 to 636 per thousand even between 
the fourth and eighth years, when the numbers are large," Deevey 
(1947) was led sadly astray. These statistics await a future report 
by Austin on the records of banded common terns found dead by the 
general public. A comparison of such data with productivity stat- 
istics and returns obtained by retrapping should make a picture of 
population dynamics in this species finally possible. The great 
merit of Austin's work lies not only in the size of his samples 
and the continuity of his banding over a long period but also in 
his use of a retrapping program to replace bands on the older 
birds. 
Summary 
Caspian terns have been found to have a mean of about 
1.8 eggs per nest and (in a sample of 37) 1.5 young per nest (Miller 
1943). Survival studies of banded birds are handicapped by the 
extremely low recovery rate (about 2.1 per cent). A preliminary 
life table based on 106 birds yielded a mortality rate of 4 per 
cent for the first year of life and a mean of 18 per cent for all 
years past age 1-2; band loss after age 8-9 is suspected as a poten- 
tial source biasing these statistics. The age at which this species 
breeds seems to be unreported, and the survival results on banded 
birds cannot be verified at this time. 
Marshall's (197) work on herring gulls was in part re- 
peated. A 60 per cent first-year mortality rate was found to be 
identical to one he reported; a man adult mortality rate of 27 
per cent per year differed from a 35 per cent mean indicated by 
his data. Band loss after age 10-11 has apparently made both 
these statistics too high. Some critical objections are raised to 
Paynter's (1947, 1949) work on the same species. 
In a brief review of Austin's (1942, 1945, 1947a) valuable 
work on the common tern, his time-specific tables are alleged to 
misrepresent somewhat the age composition of the populations he has 
been studying, the proportion of very old birds being apparently 
higher than he has indicated. 
100 
