278 Duration of Marriage and Nuniber of Offspring 
From these figures the correlation between infantile mortality and birthrate 
was determined. The general results were : 
Mean Birthrate 31-79 
„ Mortality IVIB 
, , ^ . . (Birthrate 4-3328 
standard JJeviation \ ., ,. ^ , 
[Mortality 6-1795 
Correlation -93369 + -00210 
Coefficient of Regression ... ... "6547 
Regression Straight Line (ic = Infantile Mortality, = Birthrate, Origin of x, 
Inf Mort. =0) y = •(i547a; + 2()-5423, the probable error of y being + 1-046. 
The next table shows the number of survivors of age 5 per 1000 births in 
various countries*: 
Country 
Survivors 
Country 
Survivors 
Country 
Survivors 
Country 
Survivors 
New Zealand ... 
889 
Ireland ... 
837 
Belgium . . . 
756 
Italy ... 
632 
New South Wales 
850 
Sweden ... 
783 
Denmark . . . 
755 
Austria 
614 
Victoria 
844 
Scotland ... 
780 
France 
751 
Hungary 
598 
Norway... 
838 
England and Wales 
762 
Switzerland 
748 
Spain . . . 
571 
So great is the mortality generally in high birthrate countries that the 
numerical advantage to the community almost entirely disappears by the end 
of the fifth year, reckoning per 1000 of the total population. Reducing the 
birthrates in the proportions given in the last table, the net birthrate at the end 
of the fifth year is shown in the third columns of the following table : 
Country 
Birth- 
rate 
Surviving 
5th year 
! Country 
Birth- 
rate 
Surviving 
5th year 
Country 
Birth- 
rate 
Surviving 
5th year 
Hungary 
38-9 
23-3 
Denmark ... 
29-9 
22-6 
Sweden 
26-8 
20-9 
Austria 
37-0 
22-7 
England and Wales 
29-2 
22-1 
New Zealand 
25-7 
22-8 
Spain ... 
34-8 
19-9 
Belgium ... 
28-9 
21-9 
Victoria 
25-7 
21-7 
Italy ... 
33 5 
22-2 
! Switzerland 
28-6 
21-4 
France 
22-0 
16-5 
Norway 
30-2 
25-3 
New South Wales 
27-4 
23-3 
It may be urged that whilst the birthrate has fallen in England and Wales 
during the last 15 or 20 years the infantile mortality has practically remained 
unchanged. Even if artificial limitation of families has become more generally 
the practice during that period an investigation of the birthrates and infantile 
mortalities of the various registration districts shows that the association between 
the two rates does exist and that those districts which include large cities and 
towns (where I should imagine the knowledge of restrictive measures would be 
* I am responsible for the New Zealand, New South Wales and Victorian figures; the others have 
been obtained from Mulhall. 
