If the principal portion of these deaths (those embraced in the term "cancer" without specification as to location) be considered separately, the increase in the death-rate of this portion is found to be comparatively small.

It is plain that the increase in the death-rate from cancer is partially due to more accurate methods of diagnosis. For example, the deaths from cancer in 1876 were 659, of which 565 were from cancer (part not stated), 92 from cancer of the stomach and 2 from epithelioma, while in 1895 the deaths from cancer were 1,749, which were specified as 903 from cancer (part not stated), 201 from cancer of the liver, 1 from cancer of the pancreas, 305 from cancer of the stomach, 193 of the uterus, 108 of the breast, 19 of the kidneys, 2 of the bowels, 2 of the spleen and 14 from epithelioma.

The increase in the deaths registered as cancer simply, without stating its location, during the twenty years 1876-95, was only from 3.4 to 3.6 per 10,000 of the population, while the increase in the deaths from cancer in which the part or organ affected was stated was nearly six-fold, or from .56 to 3.4 per 10,000 living. It is needless to add that the changes due to improved diagnosis would naturally become manifest mainly in this latter class, although a transfer of some deaths from the indefinite to the definite class might also occur.

Sex. The following table shows the death-rates of the two sexes from cancer for the two twenty-year periods.

The deaths of males in the first period were 2,627 and of females 5,592, and the death-rates 2.02 and 4.00 per 10,000; and in the latter period the deaths were respectively 7,256 and 15,665, and the death-rates 3.70 and 7.47; the increase of males being at the rate of 2.41 per cent. and that of females 2.85 per cent.

TABLE 63.-Death-rates from Cancer per 10,000 Living, Massachusetts,

The following figures show a similar steady increase in the death-rate from cancer in England and Wales:

Deaths from Cancer per 10,000 of the Population in England and Wales.*

But little account has been taken in the registration reports of the changes which have occurred in the mortality from the diseases and incidents of child-birth. Hence a column has been introduced in Table 46 in which these data are presented.

It is desirable to state clearly at the outset what deaths should be included under this title, since several terms are employed, and also what shall be the standard of comparison in stating a death-rate. It was therefore decided to include all deaths classed as from childbirth (under this term are embraced all deaths specified as from abortion, child-birth, miscarriage and puerperal convulsions), puerperal fever, metritis, metria, † puerperal septicemia, and the excess of female deaths from septicemia over those of males.

With reference to the standard of comparison these deaths are those of one sex only, and mostly at ages from fifteen to forty-five, hence it seemed best not to employ the living population as a standard, as is done in the other columns of the table, nor should the living births be used as a standard, since still-births subject the mothers to the same puerperal dangers as living births. The standard employed is, therefore, the number of cases of births occurring in each year, obtained by taking the sum of cases of living and of still births.‡ Each case of plural birth is also reckoned as a single case. The relative figures represent deaths per 10,000 cases of birth, and show a considerable decrease comparing the two periods. The whole number of cases of birth in the forty years was 1,820,756, and the deaths by child-birth, etc., 12,075, or 66.3 per 10,000. In the first period (1856-75), with 745,461 cases of birth, there were 5,584 deaths in

* Registrar-General's Decennial Supplement, vol. 1, London, 1895, page 54.

+ An obsolete term adopted by Dr. Farr.

This method was adopted in order to make the figures comparable with those of other countries.

child-birth, or 74.9 per 10,000, and in the second period (1876–95) 1,075,295 cases of birth produced 6,491 deaths, or 60.4 per 10,000. The death-rates by five-year periods were as follows: 68.6, 68.1, 76.2, 83.8, 73.3, 72.9, 54.4, 48.2.

There appears to have been an increase up to the fourth period, and then a greater decrease to the close.

In the following table and diagram are presented the figures for Massachusetts and those of a similar table published by Brennecke for Prussia, in each of which a rise appears during the first half of the period and then a greater fall to the close of the table:

100

20

1855

In commenting upon that portion of the foregoing table which relates to the mortality of child-bearing women in Prussia, Brennecke

* Vierteljahrsschrift für öff. Gesundheitspflege, 1897, page 81, vol. 29.

DEATH RATE FROM CHILDBIRTH

PER 10,000 BIRTHS.

MASSACHUSETTS
PRUSSIA

The column of figures for Prussia in this table is taken from a paper by Dr. H. B. Brennecke of Magdeburg, entitled "Errichtung von Heimstätten für Wochnerinnen," and extends over a period of fifty years (1844-93). It contains the statistics of 296,867 deaths from child-birth and its incidents out of 44,095,659 cases of labor.

Death-rates from Child-birth, Massachusetts and Prussia, 1856-95.

1860

1865

-1870

·1875

·1880

1885

-1890

1895

says: "Hegar in 1892, at the gynecological congress in Bonn, properly called attention to the fact that the sudden and surprising fall of the puerperal mortality curve in 1874 was not due to the influence of antiseptics (which were not employed at that time in the country districts), but may be attributed solely to the fact that in 1874 the administration of the official registration passed out of the hands of the clergy into those of the civil authorities. I have no doubt," says Hegar, "that the collection of the statistics was conducted more accurately by the clergy than by the civil authorities, since in the country districts the clergy are especially well informed as to the condition of the families under their care. They are summoned in cases of child-birth, and know when death in childbed occurs. The civil authorities are not so well informed."

Whether the foregoing comments in regard to the cause of the changes in the death-rate from the incidents of child-birth in Germany are correct is a question. They certainly do not apply to the similar column of statistics for Massachusetts, since while the Massachusetts column shows precisely the same sudden fall three years later, it is not true that any change in the methods or the authorities collecting the statistics have here taken place, since they have from the very outset, in 1842, been collected by the city registrars and the town clerks throughout the State.

Certain Minor Causes of Death.

Cerebro-spinal Meningitis. - This term does not appear as a cause of disease in the published registration returns until 1873, in which year 747 deaths from this cause were registered, a number more than three times as great as that of any succeeding year from the same cause. During the few preceding years all deaths returned as from this cause were classed under the vague and indefinite term cephalitis."

66

In the preceding year, 1872, Dr. Derby had called attention to the prevalence of cerebro-spinal meningitis, and had published a brief statement relative to the disease in the registration report of that year.

In the three years 1873, 1874 and 1875, the last years of the first twenty-year period under consideration, 1,159 deaths from this cause were registered, and when compared with the mean population of the three years the death-rate was 2.4 per 10,000 living.