A. O. POWYS 
279 
best known and the practice reach its zenith) still maintain a high birthrate 
associated with a high infantile mortality, whilst the essentially rural districts have 
generally a low birthrate and a low mortality. 
These rates I have computed from the data in the Registrar-General's Report 
for 1902, and are shown in the following table : 
Kegistration District 
Birth- 
rate 
Inrantue 
Mortality 
Registration 
District 
Birth- 
rate 
Infantile 
MortaHty 
Bep;istration 
District 
Birth- 
rate 
Infantile 
Mortality 
Durham 
35-9 
14-1 
London 
28-6 
14-0 
Cambridge 
24-2 
11-0 
Monmouth 
35-5 
13-5 
Cumberland 
28-0 
11-0 
Huntingdon 
24-1 
9-4 
Stafford 
33-7 
14-3 
Worcester 
27-8 
11-7 
Berkshire 
23-9 
10-1 
South Wales 
33-1 
14-6 
Cheshire 
27-5 
12-7 
Surrey 
23-7 
11-2 
Northumberland 
32-7 
13-1 
Lincolnshire 
26-7 
12-3 
Hertford 
23-6 
9-2 
Nottingham 
31-6 
14-7 
North Wales 
26-6 
12-3 
Oxford 
23-5 
10-9 
Warwick 
30-3 
14-1 
Shropshire 
26-6 
10-4 
Devonshire 
23-4 
12-4 
Derbyshire 
30-3 
12-7 
Suffolk 
25-8 
10-7 
Hereford 
23-4 
10-8 
Essex 
29-9 
11-9 
Northampton 
25-7 
11-2 
Bedford 
23-3 
10-7 
Lancashire 
29-8 
15-0 
Buckingham 
25-6 
10-4 
Cornwall 
23-2 
13-9 
North Riding, Yorkshire 
29-5 
13-5 
Gloucester 
25-6 
12-1 
Somerset 
23-0 
10-0 
East Riding „ 
29 '2 
12-9 
Norfolk 
25-5 
12-8 
Dorset 
22-9 
8-7 
Middlesex 
28-9 
12-4 
Kent 
25-4 
11-6 
Sussex 
22-1 
9-8 
Leicester 
28-7 
13-6 
Hampshire 
24-7 
11-4 
Westmoreland 
21-9 
8-2 
West Riding, Yorkshire 
28-5 
14-4 
Wiltshire 
24-3 
9-8 
Rutland 
21-9 
7-1 
Thus we see that notwitlistanding a few irregularities birthrates and infantile 
mortalities are closely associated in the registration districts of England and 
Wales — the irregularities being probably due to the differing ages, conjugal habits, 
sex proportions and other conditions. Where these factors are alike I think 
the infantile mortality becomes the chief determinant of the birthrate. As we 
have seen from a previous table, although natural selection is responsible for a 
larger birthrate its further operation tends to restore the balance. 
It might be inferred from the table showing the survivors at 5 years of age in 
various countries that by the operation of natural selection in countries where the 
mortality was heavy during that period the survivors could be regarded as specially 
selected lives, and that their subsequent expectation of life would be greater than 
in those countries where the mortality was comparatively light. But this is not 
always the case, as will be seen from the following table, which is extracted from 
life tables given by Mulhall from the experiences of the censuses of 1880 and 1881. 
I have added the results of New South Wales, which I have computed for the 
census of 1891. 
Survivors at various ages of 1000 &orm. 
Age 
England 
France 
Prussia 
Austria 
Italy 
Spain 
Sweden 
Norway 
Belgium 
Switzerland 
New South 
Wales 
20 
706 
685 
616 
532 
554 
496 
711 
760 
699 
689 
798 
40 
597 
566 
514 
423 
462 
412 
621 
644 
589 
578 
696 
60 
405 
408 
351 
267 
328 
292 
473 
494 
419 
393 
500 
80 
96 
104 
65 
44 
65 
69 
139 
161 
101 
69 
116 
