to the conclusion, from figures given in the following table, that there was an increasing tendency to early marriage among the young people of this country, though since 1872 the proportion of marriages under age as thus measured has probably decreased (see table, page 49).

Number of Marriages per 1,000 persons living under 21.

Fecundity of Marriage.-Marriage being the great institution by which the population is chiefly regulated, it becomes necessary to consider the conditions regulating the fertility of marriage. The two most important of these are the duration of married life, and the age at which marriage is contracted by women.

Dr. Farr in his census report for 1871 stated that 8 out of every 10 brides and bridegrooms at their first marriage are between 20 and 30 years of age, the mean of both sexes at their first marriage being 25, that of the bridegroom 25.65 years, and of the bride 243 years. The mean time that a couple of such lives survive is, as shown by calculations, 27 years, which represents the duration of married life, and approximately the time during which the issue of the marriage enjoy the joint protection of both parents. The further mean lifetime of the surviving parent is in the case of the father 9.44 years, of the mother 11:31 years.

Late marriages, by shortening the probable duration of married life, increase the proportional number of orphans.

The age at marriage is the chief factor controlling the proportion of children to a marriage, the age of the wife being the most important element, because of the fact that childbearing is limited practically between the 16th and 45th years of life. The fathers and mothers of nearly half the children

born are, according to Dr. Farr, under 30 years of age. If all the women at 30 should marry, and none before, the births would be reduced to about two-thirds of their present number; and if the marriage age were postponed to 35, the births would fall to one-third of their present number, and the population would rapidly decline. For not only (a) would the number of births in each generation diminish, but also (b) the interval between the births of successive generations would lengthen, the length of life remaining the same according to the hypothesis.

It has been doubted whether early marriages are really more fruitful than later ones; and this is certain, that women who have married at about 17 or 18 frequently cease child-bearing before they are 30, the reproductive powers being exhausted. But even supposing the fertility of early and late marriages to be identical, the number of children in the latter is less because of (a) the greater length of the generation, and (b) the fact that many who would have been parents have died before reaching the later age of marriage.

The following table, from the Registrar-General's 40th Annual Report, gives the comparative fecundity of various European States. During the three years 1876–79 the average number of births to a marriage in England was 4:57.

Assuming that the interval between the mean age of mar

riage and the mean age of mothers in England is six years, then the legitimate births in 1874, divided by the average number of marriages in the three years 1867-9, will give the average number of births to a marriage, which was 4:57. In 1864 it was 4:30, but the apparent increase is probably owing to improved registration of births.

According to the last census in France, 20 per cent. of the families in that country have no children, 24 per cent. have one child, 22 per cent. 2 children, 15 per cent. 3 children, 9 per cent. 4 children, 5 per cent. 5 children, 3 per cent. 6 children, and 2 per cent. have 7 or more children; while there are on an average 259 children to 100 fertile marriages.

Proportion of Population Unmarried.-The census report of 1851 shows that of the population at ages of 20 and upwards, about 1 in 10 men and 1 in 8 women may be referred to the category of those never marrying. It is evident, therefore, that celibacy as well as marriage without children is the natural state of a portion of the population.

The relative number of married and unmarried women during the decennium 1851-60 is shown in the following table:

Married and Unmarried Women aged 15-55 in Years 1851-60.

Thus in England 52 in 100 women of the age 15-55 were wives, while in Scotland only 44 were such.

The following table, arranged by Dr. Longstaff from the 1881 census report, gives the numbers and conjugal condition of both sexes at the last census:

Excess of Males or Females at Different Groups of Ages, together with their Conjugal Condition.

Signatures in Marriage Register.-These throw an interesting sidelight on the state of elementary education in this country, as indicated by the ability to sign the marriage register. Proportion Unable to sign Marriage Register.

In the country as a whole, the number of men unable to write is considerably less than of women, though the difference between the two sexes, as shown in the above table, is rapidly diminishing. Taking the counties separately there is great discrepancy, the general rule being, according to the RegistrarGeneral, that in agricultural counties the male, and in mining and industrial counties the female, is the worse educated sex.

CHAPTER V.

BIRTHS.

Estimation of Birth-rate.-History of Registration of Births.-Birth statistics of Counties of England and Wales.-Influence of Social Position and National Prosperity.-Influence of Nationality.—StillBirths.-Proportion of Males and Females at Birth.-Illegitimacy.-Influence of Early Marriage and of Education on Illegitimacy.-The Malthusian hypothesis.-Laws of Increase of Population and of Subsistence.-Influence of Birth-rate on Age-Constitution of Population.— Natural Increase of Population.

HE consideration in the last chapter of marriages and their average fecundity naturally opens the way to a discussion of birth statistics. Such statistics are of value in giving information as to the rate of increase of the population, and the age and sex-distribution of the population, in addition to their great social interest, especially when the statistics of illegitimacy are included.

Estimation of Birth-rate. The birth-rate may be estimated by the following plans, of which the second is the more accurate, though not so easily available for ordinary purposes. (1) It is reckoned as a rate per 1,000 of the population living at all ages, in the middle of the year. This may be described as a crude birth-rate. It is satisfactory when used for the same community in a series of years, or in comparing the birthrates of communities whose populations are known to be nearly if not quite identical in their age and sex-composition. If, however, the number living at child-bearing years differs in two populations, the birth-rate per 1,000 of the total population would vary from this cause, apart altogether from varying