116 



AGE-INCIDENCE OF CAUSES OF INFANT MORTALITY 



several States have been admitted to this area, and its profile 

 may be quite different from that of 1900. A glance at these 

 curves will suggest that such inequalities of age distribution 

 must be reflected in the mortality curves. Death cannot score 

 alike on unlike targets. 



Age distribution of urban populations ; with medians and quartiles. 



By the aid of these profiles we may be relieved of some 

 erroneous notions derived from the age distribution of deaths 

 when plotted on a straight base-line. The base-line of a mor- 

 tality curve should be thought of as conforming to the age dis- 

 tribution curve of that population in which the mortality occur- 

 red; and if, for want of small measurements of population, 

 there are significant errors in one's conceptions of population, 



