cent more numerous than those reported on the population schedule. In the following table the results are brought together by departments:

In five departments more children were reported on the school schedule, and in two, more were reported on the population schedule. The general results from the two independent sources of information, however, are not very widely divergent and serve rather to corroborate than to impeach each other.

The number of school teachers reported on the school schedules was 623, while the total number of teachers reported on the population schedule (Table XXV) was 809. This suggests that there were either many teachers on the island who were not school teachers or many from whom no returns were obtained regarding their schools. The deficiency was almost entirely among male teachers, for the occupation returns show 246 female teachers, while the educational returns show 248 female teachers. On the other hand, the occupation tables show 563 male teachers, while the returns on the educational schedule show only 375 male school teachers.

Aside from the conclusions already drawn from the tables for education, school attendance, and literacy, the following inferences from Tables XXI and XXII seem warranted:

1. The reported seating capacity (29,164) was about 8 per cent greater than the entire number of pupils (27,118).

2. The average attendance was only about four-sevenths of the pupils enrolled.

3. Of the schools, about seven-eighths were public and one-eighth private or religious; a condition widely different from that in Cuba, where only about one-half were public.

4. In five departments the number of school buildings was equal to the reported number of schools; in the other two departments the school buildings slightly outnumbered the schools.

5. Of the school-teachers, about three-fifths were male and twofifths female. In Cuba about two-fifths were male.

6. Of the pupils, about nine-tenths were free pupils and one-tenth pay pupils. In Cuba three-fourths were free pupils and one-fourth

pay pupils. But in Cuba, outside of Habana province, the proportions were more like those of Porto Rico.

7. The following table gives the proportions of each class of population reported as in school:

1 Seemingly 100 pupils in Guayama were not returned by race. Compare the figures for this department in Tables XXI and XXII.

It is surprising that the proportion of negroes in school should be higher than that of the whites. Some light may be thrown upon this fact by a study of the proportion in school of each race in the several departments. That analysis follows:

Here again it appears that in all but one of the seven departments the proportion of negroes attending school is greater than that of the whites, while the proportion of mixed is lowest of the three. These ratios are so contrary to the usual fact in such cases that one is compelled to cast about for an explanation. It is perhaps fair to assume that a certain number of school children have been reported on the school schedule by their teachers as negro, while the same children were reported on the population schedule by their parents as mixed. If so, the true number of negro school children would be exaggerated or the number of negroes in the total population understated.

In confirmation of this hypothesis the following evidence may be offered. By the present census there were 363,817 returned as colored, substantially all being negroes or mixed, the Chinese constituting less than 100 of the total. Of these colored only 59,390, or 16.3 per cent, were returned as negroes. Among the children attending school, as returned on the school schedule, there were 9,144 colored, and of these 2,427, or 26.5 per cent, were returned as negroes. There seems little reason to suppose that the proportion of negroes among colored school children is greater than the proportion of negroes among the colored

population of all ages. The presumption apparently would be in the opposite direction. Assuming, however, that the true proportion was the same, it would seem that while only one-sixth of the colored population were reported at their homes as negro, more than one-fourth of the colored children attending school were reported as negroes by the teachers giving information concerning them. This tends to discredit the testimony obtained by the census from the population of the country regarding the degree of intermixture between white and colored, and to strengthen the distrust of the figures for negro and mixed already aroused by the surprising change in the reported proportion of mixed between 1897 and 1899. (See p. 58.) Whether the line separating the white population from that of mixed blood has been drawn in the census with greater accuracy is open to question, but upon that no internal evidence has been found.

OCCUPATIONS.

[See Tables XXIII-XXX.]

The instructions issued in Spanish to Porto Rican enumerators with reference to filling this part of the schedule may be translated as follows:

This inquiry (column 11) applies to every person 10 years of age and over having a gainful occupation, and calls for the profession, trade, or branch of work upon which each person depends chiefly for support, or in which he is engaged ordinarily during the larger part of the time. In reporting occupations avoid general or indefinite terms which do not indicate the kind of work done. You need not give a person's occupation just as he expresses it. If he can not tell intelligibly what he is, find out what he does, and describe his occupation accordingly. For wives and daughters at home, engaged in the duties of the household only, write "at home" (en casa). For children not actually at work, write "at school” (estudiante) or “at home" (en casa), as the case may be. Spell out the name of the occupation and do not abbreviate in any case.1

The number of Porto Ricans reported as having gainful occupations was 316,365, or 33.1 per cent of the total population. The figures for Porto Rico are compared with those for Cuba and the United States in the following table:

It is desirable that some brief terms should be introduced to describe persons covered by the preceding instructions, and accurately but clumsily described as persons engaged in gainful occupations and persons not engaged in gainful occupations. In the following discussion the terms “breadwinners" or "persons at work" will sometimes be used for one class, and "dependents" for the other. Any term must be understood in accordance with these instructions rather than with its usual and popular meaning.

From these figures it appears that Porto Rico has only one-third of its population engaged in some gainful occupation, while in Cuba the proportion is about two-fifths, and in the United States about midway between the two. Some reasons for the difference will appear as the subject is probed farther.

The absolute and relative number of persons engaged in gainful occupations in the several departments is as follows:

Here, as elsewhere, one finds little difference between the departments of Porto Rico in comparison with that between the provinces of Cuba or the states of the American Union. The range in Porto Rico between the highest and the lowest divisions was 3.4 per cent, while for the provinces of Cuba it was 12.6, and for the states of the United States 25.4 per cent. There is no correlation traceable in the preceding table between the proportion of persons in gainful occupations and the proportion of urban population. Yet it is generally found that the relative number of breadwinners1 is greater in cities than in the rural districts. To determine whether this general statistical fact holds for Porto Rico, the following table has been prepared for the three large cities and the remainder of the departments in which those cities lie:

From the above it will be seen that in Mayaguez city the proportion of breadwinners is nearly three-tenths greater than in Mayaguez department outside the city, and that in Ponce and San Juan cities. the proportion is over two-fifths greater than in the remainder of the departments in which they lie. The difference between city and country in Porto Rico as thus measured, is more than one-half greater

See page 86, footnote.

than the average difference in the United States, and fully four times greater than the average difference in Cuba.

BREADWINNERS CLASSIFIED BY SEX.

The following table gives the absolute and relative number of males and of females reported as engaged in gainful occupations in Porto Rico. For comparative purposes the figures for Cuba and the United States have been included.

From this table it appears that the small proportion of breadwinners in Porto Rico, as compared with Cuba, holds only of the males, the proportion of females at work in Porto Rico being somewhat greater than in Cuba, although decidedly less than in the United States. The differences between the departments in the number of male breadwinners are slight, the range between Bayamon and Ponce, the departments with the lowest and highest proportions, respectively, being less than 3 per cent. Among females, the range between Arecibo, with the lowest proportion, and Mayaguez, with the highest, is nearly 5 per cent, a range somewhat greater than that obtaining with males. As the department with fewest males at work is not that with fewest females at work, and as the department with most males at work is not that with most females, it would seem that the conditions affecting breadwinning on the part of the two sexes are not identical. The three departments with a high ratio of female breadwinners are the three containing large cities. This suggests that breadwinning on the part of women may be more common in the cities. To test this, the percentage of breadwinners in the urban and rural population of these three departments is given with distinction of sex.

Per cent of breadwinners in urban and rural population, by sex.

From this table it appears that while gainful occupations are more common with both sexes in the cities, yet the difference is far greater among females than among males. Female breadwinners in the cities