nearly offset by corresponding differences in the proportion of those living together by mutual consent, which range from 5.4 in Aguadilla, where the proportion of married was greatest, up to 12.3 in Humacao, where the proportion of married was least, making the figures in the fourth column, which express the proportion of those living together under the legal sanction of marriage, together with those living together by mutual consent, very nearly the same in all the departments. Furthermore, it will be noted that those departments in which the proportion of all the married was the smallest, namely, Bayamon and Ponce, were those containing the only two large cities of the island, San Juan and Ponce. This difference between the urban and rural parts of the island is, furthermore, brought out in the statement that, while the proportion of married in the three cities of San Juan, Ponce, and Mayaguez, together, is 14.6, in the rest of the island, which may be regarded as a rural region, the corresponding proportion was 16.8, while the proportion of those living together by mutual consent was the same in both the cities and the rural districts.

Classified by race, it appears that the proportion of those legally married was greater, and of those living in consensual unions was less, among the whites than among the colored, while the proportion of the two classes together was somewhat larger among the whites than among the colored, as appears in the following table:

It appears, furthermore, from the above that the proportions of married and of those living together among the colored were very nearly equal.

Classifying the married and those living in consensual unions by sex, as well as by race, shows the following results, expressed in percentages of the total number of each class:

Thus it appears that the proportion of the married among the white females was somewhat larger than among the males, while those in consensual unions were about equal. Among the colored there was no decided difference in either class.

Classified by age, as well as race and sex, the following results appear:

Thus it appears that among the native white males the maximum proportion of married occurred in the age group between 55 and 64 years, where it reached three-fifths of the whole number, while among those living together by mutual consent the maximum occurred about the age of 35 years. Among the native white females the maximum proportion of married occurred between the ages of 35 and 44 years, where one-half of this class were married. The subsequent diminution in the proportion of this class is explained by the increase in the proportion of widowed, as will be shown later. The maximum proportion of those living together by mutual consent occurred between the ages of 35 and 44 years, where it reached one-fifth of the total number. Among colored males the maximum proportion occurred, as among the white males, between the ages of 55 and 64, and the maximum among those living together by mutual consent between 35 and 44 years. Among the colored females the maximum among the married was found between 35 and 44 years, and of those living together by mutual consent at about the age of 30.

The map opposite this page shows the proportion which those in consensual unions bore to the married in various parts of Porto Rico. A study of it shows that the consensual unions were relatively most numerous in the eastern and southeastern parts of the island, especially on

the eastern half of the south coast. Of the total number of municipal districts in the island, 35 contained a larger proportion of consensual unions than the average of the island, and in 9 municipal districts consensual unions were more numerous than lawful marriages.

THE WIDOWED.

Of the inhabitants of Porto Rico, 4.9 per cent, or nearly one in twenty, were widowed. The proportion of widowed among native whites (5.3 per cent) was greater than the average of the island, and that of the colored (4.1 per cent) considerably less. This is due to the fact, as shown above, that consensual unions were more numerous relatively among the colored, and that upon the dissolution of such union by the death of one of the members the other became classed with single rather than with widowed.

The widowed native white males formed 2.8 per cent only of the population, and the widowed native white females not less than 7.7 per cent. The corresponding figures for the colored were, for males 4.1, and for females 6 per cent.

The following table shows the distribution of the widowed by race, sex, and age:

As will be seen, the proportion of widowed increased steadily with advancing age among all classes. It reached in the native white males 30.6 per cent at the oldest age, and among the colored males 23.8 per cent, while among females it reached 65.7 per cent among the native whites and 48.5 per cent among the colored. There was here, therefore, as elsewhere, a much larger proportion of widowed among females than among males. This excess is in part due to the fact that the death rate is greater as an average, among married couples, with men than with women, on account of the greater average age of the husband, and in part to the fact that more men remarry than women, and, to a considerable extent at least, select single women rather than widows.

EDUCATION.

As far as history throws any light on the subject, it would appear that prior to the year 1799 there were no schools in Porto Rico, outside of the cities of San Juan and San German, which had free schools for girls, where needlework and the catechism of the Catholic Church were