the place to which it is entitled in the councils of the nations which employ this classification; but, as a matter of fact, very much was accomplished.

Dr. Foster, as chairman of the American Medical Association's Committee, and Dr. Batt, as chairman of the Committee of the American Public Health Association, have been engaged for over a year in co-operation with the Bureau of the Census and with committees appointed by many national medical organizations devoted to special branches of medicine, upon the question of the proper classification and nomenclature of diseases with special reference to the improvements to be made in the International Classification at the recent revision. Meetings were held by these committees at Philadelphia, New York, and Washington, and important suggestions were formulated which were duly transmitted to the Secretary-General, Dr. Bertillon, and were laid before the commission in the special book prepared for its use. A very considerable proportion of these were adopted by the International Commission. Perhaps the most important of all the measures especially recommended by the United States was the improvement in the principle of the statistical classification of deaths from violence. Dr. Bertillon, Dr. Livi, of Italy, and myself were appointed a special committee of the Commission to adjust this portion of the Classification. I believe that the list as revised will be much more acceptable to American registrars, and that it will give the information in regard to the industrial causes of mortality in a more satisfactory way than any classification previously prepared. Of course, all of the recommendations of the American delegates could not be adopted. There were twenty-three countries represented in the International Commission, and conservatism is a characteristic of European officials, and especially of European statisticians. It is perfectly right that this should be so, because it creates endless confusion when many changes are made in an established system of compilation. Furthermore, it is hardly to be expected that a country like the United States, whose registration officials had never before joined in an international congress and whose

statistics relate to only about one-half of the population of the United States, should be able to prevail against the established views of the representatives of countries where complete and comprehensive vital statistics have been published for a long series of years. Nevertheless, the fullest consideration was given to the American propositions, and the utmost courtesy and harmony prevailed. I feel that the American delegates owe profound thanks to the Secretary-General, Dr. Bertillon, and to the French Government, and that American registrars should loyally abide by the recommendations adopted, and use the International Classification without any modifications or changes, except such as are entirely permissible under its constitution, for the next ten years.

We start out now at the beginning of a new census decade with the revised classification of causes of death, in which American registration officials and American physicians have had their say; a revised standard certificate of death, which will be adopted by the American Public Health Association at Richmond next month, I feel entirely confident, and put into effect for all of the registration area, beginning Jan. 1, 1910; and with new rules and instructions recently formulated by the Director of the Census, and promulgated to all reporting offices for the purpose of obtaining more complete and correct transcripts of the deaths now registered.

In regard to the last I may now say a word. The inspiration for this effort came from my observation of the methods employed in the national registration offices of France and England. The Bureau of General Statistics under M. Lucien March receives individual returns of deaths from all parts of France in a manner similar to that in which we receive transcripts of deaths from the registration area of the United States. Prior to 1903 such returns were not made to the central bureau, but the statistics of France were based upon compilations made in the departments. At the present time in France, as in the United States since 1900, the national mortality statistics are based upon compilations made at the central office, and it was therefore of special interest to see the manner in

which these returns were received from the local registrars and handled at Paris. We have yet very much to learn with respect to care and precision in this matter in the United States. Our difficulties arise partly from the fact that the returns made to the Bureau of the Census are entirely for statistical purposes, and are not official legal records such as pertain to a national registration office. Nevertheless, we can make great improvements, and I believe that the circular of instructions recently issued by the Director and the monthly accounting for all returns, with prompt correction of discrepancies between the numbers reported by the States and cities and the transcripts received by the Bureau of the Census, will tend to eliminate, I hope completely, the differences that have occurred between the Federal and local compilations and form a basis of standard tables for the calendar year corresponding to the census of 1910, and for subsequent years.

The same extreme care in the reception and correction of returns, and especially the correction of personal particulars relating to the individual, was apparent in the RegistrarGeneral's office of England, through which I was shown, with the fullest possible explanation of all details of administration and classification, by Dr. John Tatham, who has very lately retired as the Medical Superintendent. Here, again, may I note the immense value of personal contact with the leading registration officials of other countries. Dr. Tatham supplied to me, for the use of the Bureau of the Census, a copy of his personal instructions in regard to the compilation of joint causes of death, something which has never heretofore been permitted to go out of the Registrar-General's office. As in the reforms in this country, which, I think, have been possible only through the meeting together of registration officials in the American Public Health Association, so, I believe, future reforms in international statistics, and the accession of Germany, England, and other countries which are not yet users of the International Classification, may be brought about through the personal acquaintance and confidence that result from international meetings. I am especially hopeful that it will be

possible for the Third Revision of the International Classification to be called by the American Government to meet in Washington in 1919. As the beginning to this desirable end, I hope that the American Statistical Association will unite with the American Public Health Association in a joint meeting at Washington next year at the time of the International Congress on Hygiene and Demography, when the most eminent sanitarians and vital statisticians of the world will be here.

A STUDY OF NEW ENGLAND MORTALITY.

BY HARRY A. RICHARDS.

The study of mortality tables has always disclosed interesting facts as to the life of a community. William Farr, the eminent English statistician, has set forth in many places in his reports the advantages of their use, and has constructed many valuable tables. Such tables and deductions from them have been made, indeed, in the other leading European countries. But in the United States adequate statistics on which to base them have been, until very recent years at least, wholly lacking. Some records of deaths and births, and of population, are to be found, however, in New England, extending back to the earliest times. I have attempted to make some use of these data, and to criticise my own results in such a way as to be sure of not proving too much. It is not without utility to prove that nothing can be proved, if no better result can be reached.

The prevalent notion is that conditions favoring longevity are increasing, and this is probably correct in respect of recent times. That it was so uniformly, from the settlement of New England, is certainly open to doubt. Professor Bailey, in his "Modern Social Conditions," says, after giving Wigglesworth's life-table for Massachusetts constructed in 1793 or thereabouts (see Table IV) and the Board of Health Table for Massachusetts constructed in 1897 (see Table II), "We see from this table that in the course of a century the expectation of life at birth has increased sixteen years for males and eighteen for females, assuming that the figures for males and females are identical at the earlier date." This sort of comparison is of doubtful value, for it does not take into account the fact that the two tables were constructed on entirely different principles. Wigglesworth himself knew the limitations of his work very well, and in fact made a correction for the