INFLUENCE OF INFANTILE MORTALITY ON BIRTHRATE. 245 
the birthrate is sensibly the same for France, with a very 
low birthrate, as it is for the Netherlands and Denmark 
with relatively high birthrate, viz. rates nearly double 
that for France. There may however be other circum- 
stances affecting the relation between infantile mortality 
and birthrate not considered in the fundamental assumption 
which shewed a priori that the linear relation equation 
(9) was justified. 
8. Reverting to equations (1) and (2) the purview may 
be extended to all the countries for which fairly accurate 
Statistics are readily available, viz. for the following 
countries :-— 
Austria, Belgium, Chili, Ceylon, Denmark, Hngland 
and Wales, Hrance, Ireland, Italy, Jamaica, Japan, 
Netherlands, N. 8. Wales, New Zealand, Norway, 
Queensland, Russia, Scotland, S. Australia, Spain, 
Sweden, Switzerland, Tasmania, Victoria, West 
Australia. 
This will reveal the relation for this group of countries, on 
the assumption that, if definite at all, the characteristics 
of the relation do not fluctuate during the whole period 
under review, that is to say, the relation is not a function 
of the elapsed time. 
There is, of course, a fundamental objection of some 
weight to this procedure, viz. that it supposes that the 
groups are comparable in all other respects and are homo- 
‘geneous. Doubtless such a large country as Russia might 
be divided into smaller units in which the range of mortality 
even for a single year would be considerable. And, in 
general, the characteristics of rural and urban populations 
are sensibly different. They nevertheless, fall under one 
grouping by inclusion in the statistical results for the 
whole of a country. These defects minimise the precision 
with which such statistical groupings represent the intel- 
