It will be noticed that between the ages of 25 and 55 the proportion of persons married was uniformly about double the proportion of persons living together without marriage, but in the younger and older age periods the proportion of persons living together to those married was higher. The larger proportion between 15 and 25 suggests that unions of this sort are entered upon at a somewhat younger age than ceremonial marriage. This may best be tested by finding what proportion of the total number of each class were under 25. Among the married about one-ninth (11.9 per cent) were under 25, but of the persons living together without a marriage ceremony about one-seventh (14.6 per cent) were under 25.

The large proportion of persons over 65 who were living together by mutual consent is probably connected with the presence in Cuba of many aged negroes born in Africa and imported before slavery was abolished or the slave trade effectually suppressed. It is probable that such persons before living together seldom go through a formal ceremony of marriage. There were also many Chinese males in Cuba and their median age was over 53 years. The same remark would hold true of them. The decreasing proportion of colored to the total population of Cuba during the last forty years is also a factor to be considered in explaining the difference.

Classification by age and sex.-As the age during which marriage, lawful or unlawful, occurs varies widely with the sex, it is important to supplement the age analysis already given by one in which the sex difference is also included. That is done in the following table, in which the proportion of persons living together at each period and for each sex is given. The proportions of persons married of the same sex and the same age are added in parallel columns, and as these two together make up the total of persons living in any sort of marriage relationship in Cuba, a third column gives the total, while a fourth column gives the proportion of persons of the same sex and age married in the United States in 1890. In each column the maximum ratio is marked by a + prefixed.

In the preceding table it will be noticed that the maximum proportion of persons married was reached in Cuba for each sex ten years earlier than in the United States. It may be that this is a remote result of the ten years' war, 1868-1878. It is a familiar fact that marriages decrease during periods of war or serious economic calamities. This fact is conspicuously illustrated by the vital statistics of Cuba during the last ten years, discussed elsewhere in the present volume (Appendix XVIII). It is probable, therefore, that the number of marriages in Cuba during the ten years 1868-1878 was materially reduced and that the number of marriages celebrated after the capitulation of El Zanjon was above the normal. Many men over 45 years of age may have been prevented from marrying by the disturbances. during the years of their early manhood, and, on the contrary, men between 35 and 44 would have been at the threshold of the age at which marriage is most common, when peace returned to Cuba in 1878. This hypothesis may also explain the proportion of children in Cuba between 10 and 20, which was shown in the discussion of age (p. 85) to be larger than in the United States or Porto Rico. Such children, aside from the few immigrants, must have been born in Cuba between 1879 and 1888.

The preceding table shows that for every 100 married men over 15 years of age there were 52 living together by mutual consent (126÷240=52.5 per cent), and for every 100 married women over 15 years of age there were 54 living together by mutual consent. The difference is due to the fact that the married men in Cuba outnumbered by 3,783 the married women. The ratio of those living together by mutual consent to the married was below the average for males 25 to 54 years of age and for females 20 to 44 years of age. The proportion of persons living together by mutual consent was therefore excessive in both sexes during the earlier and later years of life.

Classification by race.-The following table gives the facts by race

Legal marriage has already (p. 125) been shown to be more than three times as common among whites as among colored. The present table shows that unions by mutual consent were more than three times as prevalent among colored as among whites. The comparison may be made more clear by the following table:

Of the total unions among whites 81 per cent were lawful marriages. Of the total unions among colored 28 per cent were lawful marriages. In the following table the facts are given by race for the several provinces:

This table brings out noteworthy differences between the several provinces and shows that the differences do not run parallel for the two races. Among whites the smallest proportion of consentual unions to lawful marriages was in Habana province outside the city, where only about 1 union in 9 was merely consentual. At the opposite extreme comes Santiago, where among whites more than 1 union in

3 was merely consentual. Next to Santiago, but at a long remove, comes the province at the other end of the island, Pinar del Rio, where about 1 union in 5 was by consent only. Among colored, the province having the fewest merely consentual unions, Puerto Principe, had about 1 in 3, or rather more of such unions than the province of Santiago had among whites. Next to Puerto Principe at a long interval comes Pinar del Rio, where there were 2 unions by consent among colored for each lawful marriage. Matanzas stands out conspicuously in the column for colored, with nearly 8 consentual unions for 1 legal marriage, a proportion about thrice as great as in any other province. It is noteworthy that the provinces in which this form of married life was least common among colored, Puerto Principe and Pinar del Rio, are those in which the colored formed the smallest proportion of the population, and the province in which consentual unions were most common among whites, Santiago, is the one in which the whites are but little more than half the population.

In the following table the facts are given separately for urban and rural Cuba, and as the conditions in Habana city are often widely different from the average conditions in other Cuban cities, urban Cuba has been subdivided into Habana and the remaining 13 cities separately reported:

This table shows that among both races consentual unions were most common in the rural districts, but that for the white race the minimum of such unions was found in Habana city, while for the colored race the minimum was in the other 13 cities, and that the proportion of consentual unions among colored, both in the rural districts and in Habana, is double the average for the other cities. It will be of interest to see whether the same relation holds when both consentual unions and lawful marriages are compared with the population. This comparison is made in the following table:

From the preceding table the following percentages are computed:

This table confirms the preceding in showing that for both races consentual unions were most common in the rural districts. An appar

ent difference of result between the two methods is that the former indicated that among whites consentual unions were least common in Habana city while this table fixes the minimum of such unions in the 13 other cities. The two may be reconciled by noticing that the proportion of married to population among whites in Habana was decidedly greater than in the other cities. Hence when the consentual unions are compared with the numerous legal marriages as in the first table, they appear fewer than they do when compared with the population. These secondary cities had the smallest proportion of consentual unions for each race, but by an interesting anomaly they had the largest proportion of married among the colored. It may be that the social standards or economic situation of the colored in these. cities is somewhat higher than elsewhere, or it may be that the cities lie mainly in the center of the island and reflect the average conditions in their immediate vicinity. The last possibility may be tested by the following table:

The following table shows the same facts in the form of percentages: