306 The Philippine Journal of Science 1922 
borne in mind that variation in conditions doubtless is in some 
degree responsible for the results here presented. 
Concretely, in the Philippine Islands, of necessity, the dwell- 
ings are of a construction far different from that found in Amer- 
ica or Europe. Whereas in occidental countries warmth, to- 
gether with protection, must be considered in the construction 
of homes, in the Philippines, since there exists no need of heat- 
ing the houses, a loose construction is common that affords per- 
fect ventilation. Thus we find a fundamental difference. 
The size of the family per housing unit is generally taken as 
a criterion of “crowding.” In America the evil of this is mostly 
felt in centers of population, and it is especially among the 
poorer class and foreign element that this condition flourishes. 
In the Philippines it must be acknowledged that the same num- 
ber of people could live under a given roof, the ratio of which 
would be considered “crowding” in America, and not have the 
same injurious effect on the health of the Filipinos, due to the 
difference in climate, ventilation, etc. In fact, with the poorer 
class, which constitutes the bulk of the population, it is a com- 
mon occurrence to find several families not only under the same 
roof but also living in the same room. Consequently, it would 
appear that, with such close relationships, the possibility of 
interdetrimentalness would be greater than in the United States ; 
but the neutralizing effect of the facts that dwellings cannot be 
closed and that they have at all times perfect ventilation must 
be considered. Therefore, though the figures may show the 
index of “crowding” in the United States to check numerically 
that of the Philippines, yet upon application of the correlation 
method to death rates the resulting effects may be found differ- 
ent, due to other factors. Such was actually found to be the 
case. 5 
FACTORS CORRELATED 
In the development of this idea, by the method of correlation, 
certain Philippine data were analyzed. For simplicity and with 
a view to further work along this line, the diseases selected as 
best suited for this problem are malaria, smallpox, Asiatic chol- 
era, dysentery, leprosy, beriberi, and pulmonary tuberculosis; 
and, lastly, all causes of death combined were considered. The 
diseases named are those adopted in the standard classification 
of the International Nomenclature of Causes of Death. Also, 
this classification is strictly adhered to in the annual reports of 
the Philippine Health Service, and was furthermore adopted in 
the Philippine Census of 1918. The death rates from these dis- 
