Vol. 8, 1922 
STA TISTICS: PEARL AND LEBLANC 
301 
that even with the coarse grouping the index is reflecting the main dif- 
ferences in the form of the age distributions. 
The present note reports a test of this sort. The 1915 census of lowa^ 
gives in Table I the age distribution of the inhabitants of each of the 99 
counties of the state in such form as to permit the formation of the follow- 
/OO 
0, 60 - 
Spot diagram showing the correlation between 61 and ^2. Each dot represents one 
county. 
ing classes: 0-4, 5-9, 10-17, 18-20, 21-44, 45 and over, both figures being 
inclusive in each case. For each of the 99 counties 4> was computed on 
the basis of this grouping of the ages. Call these values <^i. Then for 
the same 99 counties the population was combined by addition into the 3 
age groups 0-9, 10-20 and 21 and over. A new set of <^'s was then calcu- 
