a relatively favorable position among the foreign born in the community as regards both economic and social status. . . . Their larger death rate may be accounted for by their large families and the prevalence of artificial feeding."

On the other hand, the Serbo-Croatians in Johnstown live in congested and badly equipped quarters, where the small proportion of women in their number bear the brunt of the poor housing facilities. How far these various conditions account fully or in part for the death rates in different places and among different races cannot be asserted except after further study. In any particular locality it is evident that special study will be profitable as a basis for planning health programs.

NEED FOR UNIFORM RECORDS

All available morbidity and mortality data classified by race show very definite differences. These differences vary with diseases, places, and ages, but everywhere they become apparent. Whether these and others less noticeable are due to racial tendencies or characteristics, or whether they are due to the various environments in which different race groups find themselves in this country, is a matter on which there is not sufficient evidence to support an opinion.

Whatever the causes, the observed differences are sufficient to be of practical importance to the health officer, the clinical physician, the visiting nurse, and all others concerned with medical and health work. A health officer, for instance, who starts a campaign to reduce morbidity and mortality from respiratory diseases, without having analyzed the race elements

in his community, and without knowing the relative susceptibility to these diseases of the native born, the Irish, the Italian, is likely to waste some effort and misdirect much more.

The most significant conclusion to be drawn from our brief statistical survey is the relative paucity of information. All the investigations which have been cited make only a small contribution to a few points in a large subject. A much larger body of data should be collected before many final conclusions can be stated. Further statistical investigation can be made in two ways: first, by special studies undertaken with particular ends in view, such as comparison of sickness or death rates from particular diseases; second, by including and analyzing the race elements in the masses of vital statistics which are more or less automatically collected by departments of health and many private organizations.

One of the most general difficulties in this work is finding uniform designations for the so-called race groups. The present practice, as shown in the reports of different health departments, hospitals, dispensaries, nursing, and other organizations, has no uniformity. In only a few countries, such as England, are the political and race boundaries the same. person born in Austria-Hungary may be a German, Magyar, Slovak, Bohemian, Jew, Croatian, Rumanian, or even an Italian. The inhabitants of that one country speak many tongues and come to this country with varying heritages.

A

On the other hand, mother tongue alone is not a sufficient designation. A person whose native tongue is French may have been born in Canada. Either of

the two facts without the other is meaningless. Both must be given in order to complete the racial picture. The new political boundary lines drawn by the Great War will render "confusion worse confounded" if both public and private organizations do not quickly adopt some uniform system of recording both country of birth and racial elements. The following chart suggests a classification which includes both the country of birth and the mother tongue.

A SUGGESTED CLASSIFICATION OF THE FOREIGNBORN POPULATION BY MOTHER TONGUE AND COUNTRY OF BIRTH

MAIN HEADINGS REPRESENT MOTHER-TONGUE GROUPS. SUBCLASSES SHOW THE COUNTRY OF BIRTH OF PERSONS USING THE GIVEN MOTHER TONGUE. NO COUNTRIES REPRESENTED IN THE UNITED STATES BY LESS THAN 5,000 PERSONS IN 1910 ARE INCLUDED IN SUBCLASSES