POPULATION. 252 POPULATION. year of age. The leportcd number very often is greater, this diirorence between the true and the asserted number being due to inaccuracy in the statements of age made in answer to the in- quiries of the census enumerator. These inac- curacies vary with the degree of education and the economic position of the class reporting. Where they e.xist in hirge numbers they are indi- cated by a disproportionate number of individuals reported with ages at multiples of 5, and espe- cially of 10. Internal evidence of the relative ac- curacy with which ages are reported, and per- haps indirectly of the relative accuracy with which other census questions are answered, may be found by measuring this concentration of re- ported ages on multiples of 5 and 10. The true number of persons in a commimity aged 30, 35, 40, 45, 50, 55, and GO is probably about one-fifth of the total number whose age is reported as be- tween 28 and 02, inclusive. The per cent, by which the reported number of these seven ages exceeds the estimated number affords a measure of the irregularity of the age distribution and so of the probable error in the returns. Measured by this test there were 1,070.000 persons in the United States in 1890 and 1,250,000 in 1900 between 28 and 62 years of age, whose ages were reported erroneously as a multiple of 5. The reported number in 1890 was 38 per cent, and in 1900 23 per cent, in excess of the estimated true number. Perhaps the best single figure indicating the age composition of a population group is the median age. or an age such that halt the members of the group are above and half are below it. For the white population of the United States this median age increased almost steadily during the nineteenth century from 10.0 in 1800 to 23.4 in 1900, the increase being due in part to the in- creased longevity of the adult po]mlation and in part to the steadily decreasing proportion of chil- dren. The median age of the population in the South (South Atlantic and South Central divi- sions) was 19.5 years in 1900, that of the popu- lation in the North (North Atlantic and North Central divisions) was 24.5, a difference of 5 years. The proportion of the population falling with- in various age classes is of importance for itself and as indicating reasons for other statistical differences between population groups. Thus in Indian Territory 15.3, in South Carolina 15.2, and in Utah 15.1 per cent, of the population are under 5, these being the regions of the United States in which the proportion of children is greatest. At the other extreme there are the District of Columbia, with 8.3 per cent., Cali- fornia, with 8.5, and Nevada, with 8.9. Tlie greatest projiortion of elderly persons is found in the northern New England States, Vermont having 8.1, New Hampshire 7.9, and Maine 7.9 per cent, of the population over 05 years of age, M.RIT.-L Co.xDiTiox, The relation of the popii- lation to the social institution of marriace is measured by numbering the classes of the smgle, the married, the widowed, and the divorced. This relation depends largely upon the age composition of the population group. The great majority of those who live to adult years marry, and half of those who marry are widowed, a sniall additional number being divorced. The age at which mar- riage is entered differs greatly accordina to the character of the civilization. 'Thus in countries dominated by Caucasian standards it is usual to assume that marriage does not take place below the age of 15. In India, on the contrary, in 1891 6 per cent, of the boys and 17 per cent, of the girls under 15 years of age were married. In uncivilized and semi-civilized countries marriage of adults is practically universal. Celibacy is regarded as unnatural and almost prohibited by social opinion. A Chinese authority writing for European readers says : "In Europe every one old enough to enter the army receives a military training; in China every one old enough to marry, marries," So in India of the women over 50 less than one per cent., and of the men about three per cent, are still single. In Europe and America among persons who have lived through the reproductive period of life the pro- portion of the single is from 10 to 20 per cent, for women, except in Hungary and the United States, and from 7 to 14 per cent, for men. In the United States the proportion of the popula- tion who are married has been ascertained for 1890 and 1900, with results that differ but little from the above. When the figures are analyzed by age it appears that among the young of both sexes there was a slight increase in the pro- portion of married persons, and at higher ages a slight decrease in the proportion of married, the two about offsetting each other. This change is cont'rary to the general trend in most countries and in one or two American States for wdiich figures covering a longer period are available. It is probably the result of very prosperous con- ditions in the two or three years immediately preceding 1900, marriage figures constituting a sensitive measure of prosperity. Religion. Civilized countries differ in reqiinl to making inquiries into the religious affiliation of the population. In Italy, France, Great Britain, and the United States it is thought un- wise to ask this question through the census. Accordingly, the population, or the enumerated population, of the earth cannot be distributed with accuracy according to religious confession. According to a careful attempt made to secure approximate results in this field, the figures in round numbers are as follows: Estimated num- REHGiON ber of adher- ents in millions Christianity 4ft0 Confucianism 250 Hinduism 200 Mohammedanism 175 Buddhi.sm 150 Taoism 40 Shintoism 16 Judaism 10 Polytheism 120 Total 1,4« The adherents of Christianity are subdivi<led as follows: Estimated num- HELiGioN ber of adher- ents in millions Catholicism 230 Protestantism 145 Greek Church 100 Minor divisions 5 Total MO Language. There are no trustworthy statis- tics for the population of the earth as a wOiole or even the civilized countries, indicating the number of persons using the main languages as their ordinary means of communication. There
