little among themselves. The most conspicuous divergence from the average is in the department of Mayaguez, where families of from 2 to 4 members, like families of 1 member, are unusually numerous, and where large families (8 to 10 members) are correspondingly few.

The following table shows the proportion of the total population of Porto Rico living in families of specified size, and for purposes of comparison columns have been added giving the same ratios for Cuba and the United States:

In comparing Porto Rico with Cuba and the United States it appears that the average size of family is somewhat greater than in either of the other countries. Families of less than 5 members included in Porto Rico 25.2 per cent of the population, while in Cuba they included 30.1 per cent, and in the United States 31.4 per cent. On the other hand, families of 7, 8, 9, or 10 members were more common in Porto Rico than in either of the other countries. The fewness of small families and the many large families doubtless account for the greater average size of families on this island.

SANITARY CONDITON OF HOUSES.

[See Tables XXXII-XXXIV.]

In the present census all buildings, whether occupied November 10, 1899, or not, were reported by the enumerators, and the facts regarding the provisions in them for supplying water and for disposing of garbage and excreta were ascertained. Before proceeding to a discussion of these topics a brief analysis of this return of buildings may be made.

The total number of buildings in Porto Rico, whether occupied or not, was 178,518, or 5.3 persons to a building. The average number of persons to a building, occupied or unoccupied, may be computed from those tables. The provinces range as follows:

In Bayamon there were 6 persons to a building, while in Mayaguez there were 4.7. In Cuba it was found that the proportion of persons to a building was greater in the cities than in the rural districts. To determine whether the same is true for Porto Rico, the following table has been constructed:

This table shows that in the three cities of Porto Rico there were on an average rather more than 7 persons to a building, while in the rest of Porto Rico there were about 5. That the three cities differ widely among themselves in this regard is shown by the following table:

to

From this it appears that there are about twice as many persons a building in San Juan as in either of the other cities in the island, and more than twice the average number for the whole of Porto Rico. The number in San Juan is also decidedly greater than in Habana city, where there are 8.8 persons to a building. The reason for this crowded state of the population in San Juan is somewhat like that to which a similar condition is due in New York city, namely, location on an island of small area, with the resultant barriers to expansion of the population.

Of the 178,518 buildings, 20,213, or nearly one-ninth, were unoccupied, a proportion slightly less than that in Cuba. The ratio of unoc

cupied buildings in the several departments varies, as is shown in the following table:

The minimum proportion is found in Guayama, where one building in twelve is unoccupied. The following table shows the figures for the three cities and the rest of the island:

From this it appears that the proportion of unoccupied buildings in the three cities of Porto Rico is somewhat, though not much, less than the proportion in the rural districts. The three cities differ considerably among themselves in this regard, Mayaguez having a much higher proportion than the average in the rural districts, while the other two have a lower rate.

Passing to the occupied buildings or dwellings, one may examine the average number of persons to each. Here again only slight differences are found between the departments, Aguadilla having the lowest and Bayamon the highest proportion, as appears from the following table:

The following table shows the number of persons to a dwelling in the three cities and in the rest of Porto Rico:

When the three cities are taken together, the persons to a building in them are more numerous than in the rest of the island, due largely to the influence of San Juan. To show that, the facts for the three cities are given separately in the following table:

It appears that the number of persons to a dwelling in Mayaguez city is not much higher, and in Ponce city no higher, than it is in the remainder of the respective departments, or in the island taken as a whole, while in San Juan the proportion is more than double that for the rest of the department or for all Porto Rico.

DWELLINGS AND FAMILIES.

By comparing the number of dwellings in Table XXXII with the number of families in Table XXXI, one may ascertain the ratio between census families and dwellings. Every dwelling contains at least one family, for, as already explained, one person living alone is for census purposes a family, and an unoccupied place of habitation is not a dwelling. As certain dwellings contain two or more families, the number of census families must exceed the number of dwellings. The figures for Porto Rico, compared with those for Cuba and the United States, are given below:

From these figures it appears that there were more families to 100 dwellings in Porto Rico than in the United States, but less than in Cuba. Still, if the figures for Habana be excluded, the number of persons to 100 dwellings in Cuba outside of Habana is found to be almost the same as the average for Porto Rico.

In the following table the figures are given by departments:

The high rate of Bayamon and the low rate of Aguadilla suggest that in the rural districts each family is more likely to have a dwelling to itself, while in the city we find frequently more than one family to a dwelling. In the following table, therefore, the figures for the three cities and the rest of Porto Rico are given separately:

It appears that the proportion of families to a dwelling in the three cities is much higher than it is in the rest of Porto Rico.

In the following table the facts are given for the three cities separately:

From these figures it appears that in Ponce there is comparatively little crowding of two or more familes into a dwelling; that in Mayaguez there is more; but that in San Juan there are, on an average, about five families to every two dwellings, a proportion even higher than that in Habana city, and further indicative of the serious crowding of population in the capital of Porto Rico.

SOURCE OF WATER SUPPLY.

The original source of water supply in Porto Rico, as elsewhere, is rainfall. This rain may fall on a building and be guided into and stored in a cistern, or may fall on and percolate through the ground either under or upon the surface. Flowing water may be obtained for human use as it comes to the surface either in a natural spring or an artificial well; or it may be obtained as it flows over the surface either in a natural water course or in an artificial water course or aqueduct. Accordingly, the census recognizes four sources of water supply, as follows:

1. Cistern for rain water.

2. Spring or well for ground water.

3. Water from a natural stream.
4. Water from an artificial aqueduct.