TABLE 39.- Infant Mortality of Cities, Ten Years, 1881-90.

Passing on to the next years of childhood, and employing the method advised by Dr. Chalmers for the whole period of forty years, the death-rates are as follow:

TABLE 40. Death-rate per 1,000 Living at Same Ages.

From these figures it appears that certain interesting changes have taken place in the mortality of children in Massachusetts during the forty years under consideration. The death-rate of infants under one year is greater now than it was in the first half of the period, but the death-rate of children from one to five years is less than it was then.

The probability that an infant under one year old will complete its first year of life is less at the present day than it was in the first half of the period, but, having survived his first year of life, his chance of completing the next four years is considerably greater than it then was. A discussion of some of the conditions contributing to these changes will be found on a later page, under the title "The Balance of Mortality."

The infantile death-rate differs much in different countries, since in some communities it amounts to as much as one-third of the births and in others it is not more than one-tenth.

The following table presents the infantile death-rate, as stated by Dr. Eross in a recent paper in Koch's "Zeitschrift f. Hygiene" (Vol. 19, 1896, page 371): —

TABLE 41.- Infant Mortality, Deaths under One per 1,000 Births.

Dr. Schlossmann states the importance of this portion of the subject very tersely as follows: "The death of every child constitutes a definite loss of national wealth" ("Zeitschrift f. Hygiene," Vol. 24, page 19).

For the remaining age periods (five to ten years and upward) the following table may be consulted (page 756). In this table the column published in the registration report of 1890, page 294, "under one," is omitted in consequence of the serious defects in the census for children of that age. (See report of State Board of Health of 1894, page liv.) For the succeeding years up to age forty there has been a decrease in the death-rate at each period of life, as follows, comparing the figures of 1865 with those of 1895:

TABLE 42.- Death-rates per 1,000 Living at Certain Age Periods, 1865 and 1895.

On the other hand, the later ages of life present in each instance an increase in the death-rate, as follows:

TABLE 43. — Death-rates per 1,000 Living at Certain Age Periods, 1865 and 1895.

These figures show that while no very decided changes have taken place in the general death-rate during forty years (that of the first twenty years being 19.49 and that of the last twenty 19.51) very marked changes have taken place in the death-rates at different ages. The causes and conditions under which these changes have taken place will be more fully considered under the head of "The Balance of Mortality," following the discussion of the "Causes of Death."

No. 34.]

VITAL STATISTICS.

NOTE. -The table on page 756 cannot be considered as an accurate table, from which precise conclusions can be drawn. Its chief merit lies in the fact that the figures selected for presentation are those of census years and hence more accurate than those which are estimated in the intervening years. The serious defect of the table consists in the fact that single years may at any time present unusual departures from the mean of a series. Thus in this table, 1865 was a year of unusually high death-rate, being much above the mean of the forty years 1856-95, and the death-rate of A much better method is 1895 was also considerably below the mean. that of comparing continuous series of years, as in tables 8, 19 and 31.

CAUSES OF DEATH.

It has been the custom in earlier summaries of the mortality of different countries to attach considerable importance to the broad and general groupings or classes of disease and causes of death; but in later years, and especially in many of the best foreign reports upon vital statistics, greater stress has been laid upon the presentation of diseases separately, without regard to their relation to each other. In the recent reports of the registration officer of Michigan a provisional classification has been adopted for temporary use, subject to such modification as medical progress may demand.

These changes have become necessary in consequence of the rapid advances in medicine, and in the knowledge of the causes and natural history of disease. As evidence of this may be cited the introdiphtheria" in 1858 (a disease well known duction of the term " and observed before under other complex names), the separation of typhus from typhoid fever, the disuse of such meaningless names as nephria, metria, noma, etc., the gradual disuse of the term "zymotic," and the transfer of tuberculosis to a place among infectious diseases.

These considerations have more weight in the present summary (which embraces the statistics of forty years) than in the ordinary digest of the statistics of a single year.

In proof of the decided changes which have taken place in the mortality from different causes during the forty years under consideration the following table is presented, in which is shown the comparative order of fatality of the ten most destructive causes and groups of causes of death for the first and last ten years of the whole period. These two series of years are selected for the purpose of presenting the conditions prevailing at the beginning and close of