ence. This, very probably, is below the truth; from whence it will follow, that a child born in a country parish or village has at least an expectation of 36 or 37 years; supposing the proportion of country to town inhabitants to be as 3 to 1, which, I think, this ingenious writer's observations prove to be nearly the case in Pomerania, Brandenburg, and some other kingdoms."

By Mr. Milne, in his work on Annuities, and in his article on mortality in the last edition of the Encyclopedia Britannica, by Dr. Bissett Hawkins, and by nearly all statistical writers, the proportions of deaths to the population, and the average ages of death, are treated as equivalent. Dr. Southwood Smith has been misled to adopt the same view. He states in his work on the Philosophy of Health, p. 135, that "There is reason to believe that the mortality at present throughout Europe, taking all countries together, including towns and villages, and combining all classes into one aggregate, is 1 in 36. Susmilch, a celebrated German writer, who 'flourished about the middle of the last century, estimated it at this average at that period. The result of all Mr. Finlaison's investigations is, that the average for the whole of Europe does not materially differ at the present time." "It has been shown that the average mortality at present at Ostend, is 1 in 36, which is the same thing as to assert, that a new-born child at Ostend has an expectation of 35 years of life."

Reference is usually made to the writings of Mr. Milne as the authority on whom the proportions of deaths to the population are taken as equivalents of the ages of death, and as data for the construction of tables to show the expenditure of life.

Mr. Milne's data are thus stated in his chapter "On the construction of Tables of Mortality," in the article, "Mortality," in the "Encyclopædia Britannica:" "Now let us suppose," says he, "the population of a place to have remained invariable for one or two hundred years past (a state of things which it might be difficult to find in any moderate sized market-town for two or three years, much less two centuries), "during which period 10,000 children have been born alive at 10,000 equal intervals of time in each year" (a state of things to which it would be equally difficult to find an approximation at any time or in any place);" also that there having been no migration" (another state of things equally difficult to find), and the law of mortality having been always the same, both the number of the living and that of the annual deaths have remained constant; the whole of the annual deaths at all ages, as well as the number of annual births, having been 10,000." "Then, if the law of mortality, exhibited in the above table, be that which obtains in the place just mentioned, that table will represent the stream of life which flows through it, and fills the vacancies left by those who advance in age, or are carried off by death, their successors incessantly following and being followed in the same course."

Having assumed these data, he reasons upon the assumption (which, for practical purposes, to which such reasonings are proposed to be applied, appears to me to be as misleading as would be reasoning in physics for practical purposes on assumptions of a perfectly calm sea, and a perfectly regular wind or stationary atmosphere, for two centuries), and, by a chain of fifteen more propositions, none of which I shall attempt to controvert, demonstrates that "the number of years in the expectation of life at any age is the same as the number of living persons

at that age and upwards, out of which one dies annually. Thus, for example, the expectation of life at 40 years of age being 25 495 years, the proportion of the living in the place, aged 40 years and upwards, who die annually, is 1 of 25 495, or, which is the same, 1000 out of 25,495."

Having of late had occasion to make rather extensive observations on this subject, it appears to be a pubic duty to state, that in no class of persons, in no district or country, and in no tract of time, has the fact hitherto appeared to be in coincidence with these hypotheses; and also that returns of the proportions of deaths to the population, when taken singly, as the exponents of the average duration of life, are often mischievously misleading, exaggerating those chances of life sometimes to the extent of double the real amount. If Dr. Price, instead of resting satisfied with Susmilch's hypothesis, had taken the actual ages of the dying within the bills of mortality, he would have found only a casual approximation to the hypothesis for the whole metropolis; and if he had taken the worst conditioned districts, that, as applied to them, he would probably have found he was in error full one-half. On Mr. Milne's own data it appears, that the proportions of deaths to the population at Carlisle, instead of coinciding with the ascertained average ages of death (i. e. 38-72,) were in the year 1710, 1 in 35; in 1787, they were one in 43; and in 1801, they were 1 in 44. Having caused an average to be deduced from the actual ages of 5,200,141 deaths which occurred in the Prussian States from 1820 to 1834, it appears that instead of 36 years, the actual average age of deaths was only 28 years and 10 months. The average ages of death in France, as deduced from Douvillard's table, founded on the experience of one million of deaths, instead of being 36 years, was 28 years and 5 months.

The public errors, created and maintained by taking the proportions of deaths as exponents of the average ages of death, or of the chances of life to the population, may be illustrated by reference to the actual experience amongst nearly two millions of the population, or upwards of forty-five thousand deaths in thirty-two districts, equivalent to as many populous towns, which the Registrar-General has obligingly enabled me to examine for the year 1839.

The Carlisle table is taken as the standard for the duration of life, to measure the loss of life in the several districts, as that table gives the probability of life from infancy, well ascertained for one town, and nearly coincides with the experience of the annuity offices, on the select class of lives insured by them, and with the results which I have obtained from the mortuary registries, showing the average age of death in the county of Hereford. Each of the recognized insurance tables may, however, be used. If the Carlisle table be taken, the chances of life at infancy would be 38.72; by the Chester table it would be 36.70; by the Northampton, 25 18; by the Montpellier table, 25-36; by the last Swedish table, 39.39; by the experience of Geneva, 40 18. After the attainment of twenty years of age, these several tables give the chances of life as follows: by the Carlisle table it would be 41 46; by the Chester table, 36 48; by the Northampton table, 33 43; by the Montpellier table, 37 99; by the Swedish table, 39·98; by the Geneva experience, 37·67; and by the experience of the Equitable Society, 41-67. For civic pur poses in this country, the most important period for considering the chances of life is after coming of age, or after the attainment of twenty

one years; the average ages of all who die above that age, in each district of the metropolis, are therefore given to illustrate the extent of loss of life to each class of adults, which is the more important to be observed, as it has been hastily supposed that the pressure of the more common and removeable causes of disease is almost exclusively upon the infant population.

In illustration of the errors occasioned by taking the proportions of deaths as the exponent of the duration of life; if we take the proportions of deaths in the district of Islington, with its population of 55,720, we find the number of deaths for the year only 1 to every 55 of the population, which would appear to be 16 years beyond the chances given by the chief insurance table, deemed a healthy standard; whereas, when we examine the average death of all that population who have died during that year, we find it to be only 29 years; in other words, we find that the average duration of the period of existence has even in that district been shortened by at least nine years below the period assigned by the Carlisle table. If we examine the pressure of the causes of death upon each class of the community, in the same district, we find that the class of artisans, instead of attaining 39 years, have, on the average, been cut off at 19 years; and hence that children and adults, and on the average ail those of the labouring classes who have died, have been deprived of 20 years of the natural expectation of life; and that even the class of adults who have died have been deprived of 15 years of working ability, involving extensive orphanage and premature widowhood. If we take such a district as Bethnal Green, inhabited by weavers and a badly conditioned population, the returns of the proportionate number of deaths to the population (1 in 41) would lead to the supposition of an average vitality of nearly double the real amount, which appears from this year's return to be only 22 years for the whole population. For the lowest classes in that district it is no more than 18 years. In the parish of St. Margaret, Leicester, which has a population of 22,000, almost all of whom are artisans engaged in the manufacture of stockings, where the average age of death in the whole parish was, during the year 1840, 18 years, I succeeded in obtaining the ages of death in the different streets, when it appeared that this average was made up as follows:-Average age of deaths in the streets that were drained, (and that by no means perfectly,) 23 years; in the streets that were partially drained, 17 years; in the streets that were entirely undrained, 13 years. Though the defective drainage and cleansing was the main cause, it was doubtless not the only cause of this variation. That, however, was a year of a heavy mortality, and the average age of death in that and another district, during the years 1840, 1841, and 1842, was in the streets drained, 25 years; in those partly drained, 21, and those not drained 17 years. The general average was 21 years. The proportions of death to the population in Leicester were, during the same period, 1 in 36.

So far as estimates of the number of the people, before a census was taken, may be depended upon, it appears that the proportionate numbers of deaths in the metropolis were, at the commencement of the last century, 1 to 20. At the time the first census was taken (1801) the proportion of deaths to the population within the bills of mortality appeared to be 1 to 39. At the present time it appears to be 1 to 40. Having had the average ages of death within the bills of mortality in the metropolis, cal

culated from the earliest to the later returns published, they appear to be, as far as they can be made out from the returns, which are only given in quinquennial and decennial periods, as follows:Of all returned as having died during the

The average Age was:

Years,

Months,

Thus, whilst it would appear from the proportionate number of deaths to the population, that the average duration of life in the metropolis has doubled during the last century, it appears from the returns of the average ages themselves, that it has only increased four years and nine months, or about one-fifth. The district of the old bills of mortality comprehends little more than one-half of the metropolis. The average age of death for the year 1839, for the whole metropolis, it will have been seen, is only 27 years. So far as an average for that year for the old district can be made out from the several recent returns, it would appear to be no more than 26 years. But the earlier mortuary registration was known to be extremely defective, especially in the registration of deaths in the poorer districts, and the recent lower averages are ascribable to the closer registration of the infantile mortality in those districts. The earlier returns are only to be regarded in so far as the errors from period to period are likely to have compensated each other; they are only adduced as indicating the degree of proportionate progression, correspondent with the general physical improvements of the population. But the slow general improvement, made up by the great improvements of particular classes, is consistent with the positive deterioration of others. The average age of death of the whole of the working classes we have seen is still no more than 22 years in the whole of the metropolis. In large sub-districts, if we could distinguish accurately the classes of deaths, the average would be found to be not more than half that period: a rate of mortality ascribable to increased over-crowding and stationary accommodation, greatly below anything that probably existed at the commencement of the century. The chief errors in the ordinary statistical returns are errors which cause the extent of the evils which depress the sanatory condition of the population, and the mortality consequent on those evils to be under-estimated.

The erroneous conclusions as to the ages of the populations from the proportions of deaths, have perhaps arisen from assumptions of the existence of states of things rarely, if ever, found, namely, perfectly stationary populations, and perfectly stationary causes of death. I have been asked, "If 1 out of 40 die yearly, must not the average age of all who die be 40 years?" The answer, by actual experience, as we have seen, is, that it is often not 30 years; and perhaps the reason why it is not so will be most conveniently illustrated by hypothetical cases. For example, let it be assumed that in any given year 40 persons die out of 1600, which is in the proportion of 1 to 40, and in consequence of an unusual prevalence of measles, or some disease to which children are subject, the greater number of deaths occur amongst the infant portion of the population, and hence, out of the 40 deaths, 20 occur at 5 years of age, 10 at

25, and 10 at 60. Then the total existence bad, would have been (20 × 5) + (10 × 25) + (10 × 60) 100+250 + 600 = 950 years, and this divided by 40, the number who died, would give 250 24 years, nearly as the average duration of life to each of the 40 who died.

On the other hand, suppose a severe winter, in which the peculiar causes of mortality may have pressed unusually heavy upon the older lives, and let the numbers who died have been 20 at 60 years of age; 10 at 40; and 10 at 5; in such case, the total existence enjoyed would have been (20 × 60) + (10 × 40) + (10 × 5) = 1200 + 40 + 50 = 1650 years, which, divided by 40, would give 150 41 years, as the average duration of life to each.

And again, where, in fact, the proportion of death in one year may be represented as 1 death out of 20 of the population; the average existence enjoyed may be greater than when 1 in 40 died for the reason given in the former case. As for example, in the year when 1 in 20 died, it may have happened that the deaths were among the older lives, and that, taking one with another, the average age of all who died might be 50; while in the other case the mortality might have been amongst the infant population, when the average age might have been 20. If the proportion of 1 in 40, or 1 in 20, were to obtain each year continuously, taking one life with another, the average duration to a population just born, of whom 1 in 40 died, and whose place should be supplied each year by a new birth, would be about 20 years to each life, or one-half; and of a similar population, of whom 1 out of 20 died annually, the average duration of life to each would be about 10 years, or one-half the period at the expiration of which all the lives would have expired.

When these examples are considered, it will be understood that the average age of death may remain stationary, or may go on increasing, whilst the proportions of death remain the same, or vary. The actual mortality of most districts is found to be coincident chiefly with its physical condition, and is most accurately measured by the years of vitality which have been enjoyed, i. e. by the average age of death, or the total numbers of years which every individual who has died has lived, divided by the numbers who have died. The numbers of deaths increase or diminish considerably, and frequently create erroneous impressions, whilst the average ages of death are found to maintain a comparatively steady course, always nearest to the actual condition of the population, and give the most sure indications.

The chief test of the pressure of the causes of mortality is then the duration of life in years; and whatever age may be taken as the standard of the natural age, or the average age of the individual in any community may be taken to judge what are the standard numbers of death in that same community. For example, in the returns of the St. George's Hanover Square district, it appears that in 1839, the proportions of death was 1 to 50 of the population; but the average numbers of years which 1325 individuals, who died during that year, had lived, was only 31 years, or 8 years below the average period of life in Carlisle. There was then in that district during that year a total loss of 10,600 years of life, which, at 39 years, may be considered as equal to an excess of deaths of 272 persons, and in a healthy state the proportions of deaths should have been 1 in 63, instead of 1 in 50 of the population.

The effect of migration or of emigration, in disturbing the results of re