On the average in the five cities almost exactly one-third of the children 5-17 attended school during the year preceding the census. If the figures for these cities be subtracted from those for all Cuba, the following results are reached:

The proportion attending school in the large cities was almost three times that in the rest of the island.

In the following table the figures are given for the six provinces after the five large cities have been excluded:

Both in its capital city and in the rest of the province Matanzas had a larger proportion of children attending school than any other city or province. It will be noticed that the rank of the provinces in school attendance is often at variance with the rank in regard to the proportion able to read. Thus Puerto Principe ranks next to Habana in literacy, and yet the proportion of persons at school in Puerto Principe was lower than in any other province except Pinar del Rio. The anomaly may be explained by assuming that in the sparsely settled districts children are often taught at the home rather than in a school. In that case the figures regarding school attendance lose much of their significance.

Sex.-The following table gives the facts for Cuba by sex:

The two sexes attended school in about equal proportions.

Race. The following table gives the facts regarding school attendance for Cuba by race:

The school attendance of the whites was somewhat higher than that of the colored, but the difference was not very great.

Age. In the following table the proportion of school attendants in the several age classes is given:

The maximum proportion of school attendants was between 10 and 14, but the preceding five-year period shows almost as high a proportion.

LITERACY AMONG PERSONS OVER 10 YEARS OF AGE.

The majority of persons able to read probably learn to do so in early childhood. Hence it is the usual practice for a census in gathering information on this topic to disregard all children under a certain age. This has not been done in censuses of Spain or the Spanish colonies, but in American census practice all children under 10 are omitted from the illiteracy tables. This classification is made in the present census of Cuba and will be regarded in the following discussion. For reasons already explained, only two classes will be considered those who are and those who are not able to read. The following table gives the facts for all Cuba:

Rather more than two-fifths of the population of Cuba, excluding young children, were able to read, a proportion rather greater than that of New Mexico in 1880 (35 per cent) and less than that of South Carolina in the same year (44.6 per cent), but decidedly less than the pro

portion in any American state in 1890, owing to the rapid development of the American school system in the last score of years. Sex.-In the following table the facts are given for Cuba by sex.

The corresponding per cents for the United States are males, 87.6; females, 85.6; so that in both countries, and indeed in most countries where the information is obtainable, the ability to read is somewhat more general among men than among women.

Race. In the following table the number of persons able to read is given for Cuba with distinction of race and sex:

From this table it appears that one-half of the whites and rather more than one-fourth of the colored were able to read. But among the colored the illiteracy was greater among the males, thus reversing the usual rule. Two reasons for this anomaly may be suggested. Colored men work more largely in the country and less largely in the cities of Cuba than colored women do. In the five cities of over 25,000 were found 27.6 per cent of all colored females over 10 years of age, but only 22.2 per cent of all colored males over 10 years of age. It has been shown that school attendance was larger and illiteracy smaller in the cities than in the rural districts. Hence the sex which is most numerous in cities has better facilities for learning to read and probably city life tends to strengthen the desire for this attainment. Then, too, the colored females outnumber the colored males at every age period from 15 to 50 and the males outnumber the females between 50 and 80. This is probably due in part to the survival in Cuba of some thousands of negroes born in Africa, two-fifths of whom are males, and to almost 15,000 Chinese, nearly all of whom are males. Its effect is to make the median age of colored males over 10 fully two years higher than the median age of colored females. And as illiteracy in Cuba is greater among elderly people than it is among those in middle life this higher median age of the colored males would tend to accentuate the illiteracy of that sex.