A. O. POWYS 
263 
in longevity, the mean duration being 21 years — an advance of over 2 years — 
and there can be no doubt that the increase is a I'eal one, whatever may be the 
cause. It of course may be urged, and with much force, that amongst those 
following hazardous occupations, such as the industrial classes, marriage is generally 
Fig. 10. Age at Marriage and Duration of Life beyond Age 46 by Survivors of that Age. 
Males. 
26| \ 1 1 r 1 1 1 ] 1 1 1 [— 1 1 \ 1 1 1 1 1 1 1 1 
25 — <? — 
13 
12 
11 
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 
Age at Marriage. 
Equation to Eegression Straight Line : 
aa' (Origin at 19 years of Age. Unit of .i- = l year) 
?/ = 19-2847 + -10642.r. 
contracted early in life, whilst amongst those following the least hazardous — 
the professional — marriage is contracted much later. This is true, as will be 
seen from Section E, though the difference in age at marriage does not exceed 
1| years in the two classes. The adults' death rates are also shown in the same 
section, that of the professional class being 15-73 and that of the industrial 22 72 
per 1000 exposed to risk. But notwithstanding this great difference in the 
mortality in view of the slight difference in the age at marriage, it is doubtful 
whether if all allowances be made the difference in the duration of life can 
be thus accounted for. Some portion is probably due to the struggle to maintain 
the large family generally connected with early marriage*. 
[* It will be seen from Biometrika, Vol. i. p. 46 et seq., that a man's prime is about 28 in the same 
sense as a woman's is about 24. May not therefore the lessened expectation of life in the case of man 
as in the case of woman be clue to exercise of sex-functions before the prime ? Further is not the 
association of extreme fertility with shorter duration of life in women due in part to the fact that very 
large families mean as a rule early marriages ? K. P.] 
