A. O. POWYS 
247 
per cent, were born elsewhere. Of the 6457 who died at age 65 or over these 
figures are 665 or ]0'3 per cent., 5545 or 85 9 per cent., and 247 or 3 8 per cent, 
respectively. 
From the results of this section of the investigation we are amply justified in 
concluding : 
(a) That in females extreme fecundity is unfavourable to extreme longevity. 
(b) That mothers of moderate sized families of about six live on the average 
longer than those with smaller or larger families. 
(c) That although in males increased fertility is apparently associated with 
increased longevity throughout the whole of life, such association in part at least 
is due to second and subsequent marriages. 
(d) That the married of both sexes have a greater expectation of life than 
the unmarried, and that this superiority is not all due to physical infirmities which 
prevented marriage. 
B. Distribution of Fertility. 
Professor Pearson, in his essay "Reproductive Selection*," has dealt with 
this subject rather exhaustively, using the data derived from certain sections of 
the Anglo-Saxon community and from the upper and middle classes of the city 
of Copenhagen tabulated by Messrs Rubin and Westergaard. The statistics now 
to be presented are for the whole state of New South Wales — urban and rural — 
and will exhibit the distribution of fertility under varying conditions of ages 
of women and durations of marriage. Taking first of all the women used in 
the last investigation, i.e. those who died beyond the reproductive period irre- 
spective of the age at marriage or of the duration of marriage, the distribution 
is as shown in Table XIII. 
From this table we see that nearly 11 per cent, of the women were sterile, 
and it would appear from Fig. 5 (broken lines) that this is unduly high. The 
excess might at first sight be attributed to absolute prevention of family, but 
as these statistics relate to married women dying beyond the reproductive period, 
they comprise widows — whether widowed before or after the reproductive period — 
and of women who may have married late in that period and whose fertility had 
but little opportunity to display itselff, and of women who may have married 
beyond that critical period and consequently no opportunity whatever. Therefore 
although this percentage of barren women agrees fairly closely with that obtained 
by Rubin and Westergaard for marriages of duration of 15 years and upwards 
with both husband and wife alive, viz. 12'87, this ratio is distinctly too high 
for New South Wales. For marriages under the last-mentioned conditions, as 
we shall subsequently see, the barren marriages are but 5 563 per cent., which 
* Chances of Death, pp. 63 et seq. 
t Vide Biometrika, Part i. Vol. i. p. 33, see curve showing initial fertility in deferred marriages. 
