Table 39.—-Published Mortality Rates for Herring Gulls 
Each rate is here expressed as per cent per year. Marshall (1947) 
and Paludan (1951) used a life table starting as of September 1. 
Paynter (19,7) claimed that his first table started as of July 1. 
Subsequently he reported that it and a second series of mortality 
rates hold for years starting "sometime" in early August (Paynter 
1949, p. 161). 
a Cc 
Age Marshall Paynter Paynter Paludan 
Interval (1947) (1947) (19))9) (1951) 
O-1 60 2 50 62 
1-2 38 31 Tel 22 
2+3 29 2k 28 19 
3-h, 29 21 23 16 
y=S 39 29 35 ly 
5-6 48 35 27 17 
6-7 39 kl 17 13 
7-8 28 32 19 17 
8-9 v, 58 28 16 
9-10 ¥, 100 4 16 
10-11 / ~ 100 9 
Wadults" 2/ 35 30 32 17 


1/ not calculated (small sample); last bird died at age 16-17 
2/ excluding age 0-1; my calculations 
last paper appeared too late for consideration in this report. It 
is a model study, and I believe the low mortality rates cited point 
up the inferior bands used here in America. 
Are these discrepancies in the American results principally 
due to differences in computation and analysis; or do they represent 
real differences between (a) a generalized sample of 3806 herring 
gulls banded at various places in North America from 1921 to 1939 
inclisive (= Marshall's sample), (b) a gpecific sample of 1252 banded 
in New Brunswick from 193) through 1943~ (= Paynter's first sample), 
and (c) 362 birds banded in New Brunswick in 1936 and recovered up - 
to June 30, 1946 (Paynter's second sample)? 
(a) Starting dates.--Marshall (197) started each age 
interval as of Soptonter 1. Paynter (1947) claimed that his first 
table started as of July 1, the date when banding of nestlings 

11943 is not specifically mentioned by Paynter as the last year of 
banding work he used. The use of any subsequent work would create 
an important bias in his analysis. 
92 
