observed by Edmondson (1945), and possibly a population of barnacles 
(Balanus balanoides) reported by Hatton (1938) represent three species 
which Deevey found to approach this type of curve (Figure 1). 
Survivorship lines indicating constant mortality rates.— 
When the survivors at each age level are plotted on ordinary graph 
paper, this type of line is concave but not necessarily J-shaped; on 
semilogarithmic paper, however, it has the particular property of 
appearing as a straight line. Such a survival curve is typical of 
the. adult stages for many birds, their mortality averaging about 320 
per thousand per hundred centiles of mean life span (Deevey 19,7). 
It has not been found, however, to hold strictly for all ages of a 
species! life. Indeed many of the species that Deevey grouped under 
this category may, when their complete span of life is finally 
reported, exhibit the convex type of curve described above. 
Anatomy of life tables 
Although life tables have been adequately explained by 
Glover (1921), Hill (1936), Pearl (1941), and by many other writers, 
a few introductory words about them at this point will help readers 
unfamiliar with the literature of vital statistics. 
Life tables contain one or two categories of primary data 
and two or more categories secondarily derived from them. These 
categories have come to be identified with standardized symbols 
(Pearl 1941, p. 220) which are given below. In any extensive study 
of survival, the investigator will find symbols a useful type of 
shorthand in his analyses. In ornithology, the nomenclature of 
survival studies is still so inexact that I have found symbols 
indispensable to clear thinking. 
The skeleton of a life table may be said to consist of four 
or more columns; 
(1) The age, which is best stated as an interval, is often given in 
years but also in other intervals as convenient. This column in a 
life table can scarcely be confused with any other but is often 
symbolized by the letter x. In broad zoological studies, these inter- 
vals can be set up as fractions of a species' mean life span. In 
restricting my own investigation to birds, I have found age intervals 
of one year to be convenient. 
(2) The number living at the beginning of each age interval in column x 
can readily be plotted as a life curve. This series is symbolized by ~ 
i and is synonymous with survival curve, survival series, and survivor- 
ship line. In human statistics, the number alive at the start of age 
O-I is usually taken as 1,000,000 or 100,000. In ornithological 
statistics, it seems best when possible to convert the raw data as 
Deevey (1917) did, and start with a cohort of 1000. As oviparous 
animals, birds lend themselves well to tabulations starting with the 
