hope of gaining, by peaceful ways, what could only be gained by war at a vast loss; the conviction that most wars have been unnecessary, if only some amicable settlement and impartial judge could have been found;-these influences, which are worth many fifteen and a half millions of dollars, will, as we. with many others hope, help the spread of peace over the world. At all events, the world will honor the two great and proud nations for the course which they have taken.

ARTICLE V.-ON THE LAW OF MORTALITY THAT HAS PREVAILED AMONG THE FORMER MEMBERS OF THE DIVINITY SCHOOL OF YALE COLLEGE.

A CATALOGUE and biographical record of the persons who have been members of the Divinity School of Yale College, during the fifty years of its existence, from 1822 to 1872, has just been prepared and published by Prof. George E. Day.* The materials for the record have been collected with very unusual care, being in almost every case from autograph letters, or other authentic documents. This fact, as well as various other peculiarities of these statistics, make it worth while to look at them with special reference to the vital experience of the persons named in the catalogue.

By the law of mortality of a community, or of a large selected group of persons, is usually understood the varying probability of death in that community, or group, for the different ages of the members. Thus out of a thousand persons aged thirty, we find that in one year there will be in the mean 8 or 10 deaths; of a thousand aged fifty there will be from 14 to 19 deaths; of a thousand aged seventy, there will be 60 to 70 deaths. These numbers of deaths for each age of life, accurately found for a community and put in a table, express the law of mortality for that community.

The law of mortality is not the same for all groups of men. It varies with residence, with occupation, with sex, and with other causes. The law for assured lives is different from that of annuitants; that of scholars differs from that of laborers; that of the clergy from that of lawyers, or doctors. The law deduced from observation in one year, might be expected to vary from that of another year, if the prevailing diseases differ. The law for American lives must differ to some extent from that for English lives, and both differ from that for Ger

* A general catalogue of the Divinity School of Yale College; a brief biographical record of its members in the first half century of its existence as a distinct department. Published by the Alumni. 8vo, 164 pp. New Haven, 1873.

man, or for French lives. Such differences, of course, though real, are often quite small.

The graduates of this Divinity School form a class in some respects peculiar to themselves. Most of them are preachers, and the larger part of the rest are teachers, or are at other work in which the risks of disease and death are like those in the ministry. The effect of care and forethought in prolonging life is nearly the same for them all. The medical selection of assured lives has given a peculiar character to all tables deduced from the experience of insurance companies, but from such an effect the statistics of this catalogue are nearly free. Although, therefore, the number of years of experience is not large, being less than 20,000, and the number of deaths less than 200, it seems worth while to see what evidence they furnish about the law of mortality among such lives in our country.

Some readers who have kindly followed us thus far may care only for the result of the discussion. Others will ask for processes, or may wish evidence of care in the treatment of the facts, or perhaps will wish suggestions for the use of similar data from other sources. I beg the former class to skip over the next two pages, which are not intended for them at all.

In the treatment of the data I have proceeded as follows: Each person who has belonged to the school is regarded as coming under view on the day of the Yale College Commencement of the year in which his class finished its course. The risk (to use a technical term) begins on that day, and his name continues on the list until the anniversary of that Commencement next following his death, if he is dead, or till that day in 1872 if he was then living. By this rule the students who die while in the seminary, that is, before the Commencement of the year of graduation of their class, are thrown out of our account. Against this, of course, are thrown out also the years of favorable experience of the students who survive. Each name belongs in the group an exact number of years.

It was assumed also that the Commencement day named is to each person his birthday, and that the completed year of his age on that day is the age attained upon the nearest birthday, before or after. That age was written opposite each man's name in a copy of the catalogue. Along with this was also

written the age just attained on the corresponding day of 1872, if he was then living; and if dead, the number of the unfinished year of age, the birthday being assumed as before.

If the statistics had been complete this process would give us the means of getting completely the vital experience. How shall we treat the cases where the facts are not all given? We cannot throw out such names, for these records are in many cases incomplete because of the death of the persons, and in other cases we know about a man because some one has written an obituary notice of him. It is better to retain all cases in which we can assume without serious error the dates and ages needed to make the data complete. Thus for three or four persons the year of birth without the day is given. For thirteen. persons the age is obtained from the matriculation records of Yale College. Six persons, who are doubtless living, were last heard from in 1870 and 1871. The risk upon their lives. ceases with those years. One person is reported dead, but the year is unknown, and I assume that he lived half of the time from graduation till 1872. Five deaths are reported in given years, but the month and day are not told. For them, the first of January is made the assumed birthday, and fractions therefore appear in the columns of the deduced tables. In one important case the death occurred in the year of graduation, whether before or after Commencement is not known. I have added a fraction (0.36) in the years of risk, and a like fraction in the column of deaths.

Of 41 persons the birth-year is not known. The average of 771 known ages, at the time of finishing seminary studies, is 27.7 years. The remaining 41 persons are without doubt older, as many of the persons are reluctant to tell their ages. I believe the error will not be large if I assume them to have been on an average 29 years old.

There are seventeen persons of whom we have no recent satisfactory information. Six of them have been heard from at an average of more then 10 years after graduation. This group, about two per cent of the members in the catalogue, I have left out entirely. Perhaps we have not heard from them because they are dead; perhaps because they are not dead. To balance any greater chances of the former kind may be set off

their sixty-three years of known favorable experience. If the experience of these seventeen persons can be learned, and if it shall prove to have been more or less favorable than that of the rest of the graduates, the following conclusions will need correction by about two per cent of the difference.

In Prof. Day's catalogue there are 854 names. Eight died before finishing their studies, seventeen are not heard from, and seventeen were of the class of 1872. This leaves 812 persons of whom 771 were of known ages, and 41 of an assumed age. Opposite these 771 names I wrote in the catalogue the ages at the beginning, and also at the end of the years of risk, as has been stated above. By simply counting the times that each number occurs in the first set of ages, we get the following table. It shows how many persons enter our selected group, at each age of life.

Table showing the number of students finishing their seminary course of study at each year of age:

By counting in like manner how many times each age occurs in the second set of ages, we form a table showing how many persons leave our selected group at each age, either by surviv ing in 1872, or by having died. Now adding the numbers of each table down to any given age, inclusive, and taking the difference, gives us at once the number of persons who enter upon the specified year of life. The results are in the second. column of the table upon the next page. In the third column. of the same table are given the numbers of persons dying in each year of age, the age being however now the completed years of life. The reason for fractions in these columns has been given above.

The next two columns state the corresponding facts for the 41 persons whose ages are not known, but are assumed to have been 29 upon finishing seminary studies. I give them separately, that those who wish may use another assumed age. The