note 25-6 to 21.7 per 1,000; the mortality from typhoid fever was reduced 10-75 per cent., and from phthisis 11-49 per cent.

The facts stated in the preceding table may be compared with Dr. Kelly's figures (page 120). There can be no question as to the great diminution of phthisis in this country, and that this is due to sanitary improvements. It is worthy of note that Mr. (now Sir John) Simon, commenting on Dr. Buchanan's results in his report for 1867 to the Privy Council, while thinking that there is no possibility of questioning the conclusion that dampness of soil is an important cause of phthisis to the population living on the soil, is careful to state that this conclusion does not purport to be more than a contribution to the etiology of phthisis. He points out that the disease has an industrial aspect, that its hereditary relations are important, and that the infective nature of the morbid process accompanying it requires elucidation.

Mortality at School Ages.-The influence of the present educational system on the health of children in this country is so important a question that no apology is necessary for quoting the following statement from the Registrar-General's Annual Report for 1882.

"We have no means of giving the death-rates of children of the precise ages 3-13, as requested, for this age-period is not one that is used by us in abstracting deaths. We can, however, give the death-rate of children of from 5 to 15 years of age, and this age-period will probably be a better one than 3-13 to test the possible effect of school-work on children's mortality; for, though the children of three and four years of age are omitted, school can scarcely be supposed to affect them in the way of injury from overwork; and, though boys and girls of 14 and 15 are included, it must be remembered that any injurious effects of overwork at school that may exist will be carried on after the actual school attendance is finished. The effect, if there be any, of overwork would be in the latter years of school attendance; viz., the 12th and 13th years, and

the injury would cause increased mortality, not only in those years, but at least for some few years afterwards. The ageperiod, then, which we take is the decade from 5 to 15 years of age, and the periods we take for comparison are the decennia 1861-70 and 1871-80.

"The death-rate of children (5-15) in 1861-70 was 6.3 per 1,000. It fell in 1871-80 to 5.1 per 1,000; a decline of 19.05 per cent. The main part in this fall was due to diminished mortality from the chief zymotic diseases; namely, small-pox, measles, scarlatina, whooping-cough, diphtheria, fever, and diarrhoea. These diseases caused a mortality of 2.9 per 1,000 in the first decennial period, but only of 2.1 per 1,000 in the second.

"But after excluding these diseases from the account as irrelative to the question at issue, there was still a considerable decline in mortality. In the first period (1861-70) the deathrate from all causes other than zymotic was 3.4 per 1,000; in the second it was only 3.0.

"But inasmuch as school work, if it be injurious to health, would probably be so by affecting the brain and, generally, the nervous system, it would be well to split up these death-rates from causes other than zymotic, into death-rates from diseases of the nervous system and death-rates from other causes. When this is done, we find that the entire fall was due to diminished mortality from other causes. The rate from these fell from 2.9 to 2.5 per 1,000, whereas the death-rate from nervous affections remained unaffected. Indeed, if a second place of decimals were taken, it would appear that nervous diseases had slightly, very slightly, increased.

"It would appear, therefore, that, while the mortality of children from all causes and from zymotic causes has considerably diminished, their mortality from diseases of the nervous system has exceptionally remained stationary. The general improvement has not affected this class of diseases.

Annual Death-Rate per 1,000 living of Children aged 5-15 in England and Wales in the two decades 1861-70 and 1871-80.

"To the above answer to the question addressed to this office may now be added the supplementary fact that the figures for 1881 and 1882 also showed a further and considerable decline in the deaths of children of the ages in question from all causes together, and from the chief zymotic diseases, while their deathrate from nervous diseases remained in each of the two years at exactly the same point as in the two preceding decades."

"It may also be well to note that tubercular meningitis or acute hydrocephalus is not included in the foregoing table among diseases of the nervous system; and that the mortality ascribed to this cause was somewhat higher in 1881 and 1882 than in 1861-70. Both in 1881 and 1882 the death-rate among children of school age from this cause was 0.22 per 1,000, whereas in 1861-70 it averaged 0.20."

These statistics undoubtedly appear to give countenance to the view that the strain of education produces in a certain proportion of children injurious effects on the nervous system. It would, however, be unsafe to draw certain conclusions from these figures. The medical certificates of death relating to children are often ambiguous in statement, and the absence of any decrease in diseases of the nervous system may be due to accidental causes of variation and not to the influence of schoollife. The whole question of "overpressure" is discussed in nry manual of School Hygiene.

CHAPTER IX.

DENSITY OF POPULATION AND MORTALITY.

Estimation of Degree of Aggregation.-Relation between Density and Mortality.-Urban and Rural Population.-Urban and Rural Mortality. -Apparent and Real Differences.-Effect of Density on Mortality of Young.-Range of Action of Density on Mortality.-Effect of Higher Degrees of Density on Mortality.-Size of House and Mortality.— Artisans' Block Buildings.-Back to Back Houses.-Causes of High Mortality from Increased Density.-Impure Air.-Infectious Diseases. -Phthisis.-Poverty and Drunkenness.

DR

R. FARR first called attention to the influence exerted by density of population on mortality in the fifth Report of the Registrar-General (1843), and since that time the relationship between density and mortality has been further elucidated.

Method of Estimating Degree of Aggregation of Population. -Two methods are commonly adopted. (1) The number of persons living to each square mile of area is stated. (2) The average number of acres occupied by each person in the population is stated.

The following table gives an example of both these methods:

The unit of space in the first method is taken to be a mile; in the second (so that whole numbers may be secured), an

acre.

Relation between Density and Mortality.-Dr. Farr found that the mortality increases with the density of the population, but not in direct proportion to their densities, but as their 6th root.

Thus if d and d'=density of population in two places, and m and m'=mortality

m' then =

m

and m': m:: ď: vā

In his report for 1843, Dr. Farr gives as examples seven groups of districts the death-rate of which, calculated from their densities, approximated very closely to the observed death-rates. Thus

Death-Rates

Calculated:
Observed:

18.90-19-16-20·87-25·02-28.08-37·70—38·74.

16.75-19 16-21.88-24·90-28.08-32-49-38.62.

The formula was subsequently altered by Dr. Farr (Supplement to 35th Annual Report, page clviii.), 0-11998 being substituted for in the above formula; or, more exactly

Thus in 1861-70, in the 345 districts, which had a mortality of 19.2, the density was 186 persons to a square mile; in the 9 districts with a density of 4,499, what was the mortality? It was happily not expressed by the proportion of the two densities; i.e., 186: 4,499: 19-2: ; but by this proportion nearly

(186)012 (4,499)012 :: 19-2x=28·1.