300 
STATISTICS: PEARL AND LEBLANC Proc. N. A. S. 
8 Chambers, R., J. Gen. Physiol., 2, 1919 (49-68). 
9 Heilbrunn, I. F., /. Exper. Zodl., 30, 1920 (211-237). 
^° Fresenius, G., Beitrage zur Kenntniss Mikroskopischer Organismen. 1858. 
" Dallinger, W. H., and Drysdale, J., Mon. Microsc. J., 13, 1875 (185-197). 
^2 Stein, F. W., Der Organismus der Flagellaten. 1878. 
13 Klebs, G., Zs. Wissensch. Zodl., 55, 1893 (265-445). 
1" Alexeieff, A., Zodl. Anz., 44, 1914 (193-213). 
A FURTHER NOTE ON THE AGE INDEX OF A POPULATION"- 
In an earlier paper in these Proceedings, PearP proposed the function 
as a single numerical index of the age distribution of a population. In 
this expression 5 indicates summation, for all age groups given in the origi- 
nal statistics, of the expression in brackets; A is the deviation, for each 
age group, of the percentage of the actual population in each age group 
from the percentage of the same group in the standard population of 
Glovers^ Life Table, denoted by P; M = mean age of living population; 
Mp = mean age of persons in a stationary population unaffected by 
migration. 
In connection with certain problems now under investigation in this 
laboratory it becomes a matter of importance to know how sensitive this 
age index is to change in the size of the age groups of the original sta- 
tistics. Since its original proposal the index has been used in many stud- 
ies in this laboratory, and the larger experience has strengthened our 
confidence in its reliability as an index of significant variations in the age 
constitutions of populations. But it has always been used hitherto with 
at least 6 to 8 age groups covering the life span. Suppose the original 
statistics furnish only 3 age classes for the entire life span. Will this age 
index <f) then give a reliable picture of the significant variations in age dis- 
tribution, as we pass from city to city, or county to county ? 
To test this point, the obvious thing to do is to determine the correlation 
between the age indices for n communities, on the basis of say 3 divisions 
of the life span, with the age indices for the same communities on the basis 
of say 6 divisions of the life span. If the correlation is high it will mean 
By Raymond Peari^ and T. J. LeBlanc 
ScHooi. OF Hygiene; and Public Health, Johns Hopkins 
Communicated September 7, 1922 
