of deaths per bed (the bed death-rate) is given as an estimate of the relative mortality and sanitary condition of different hospitals and infirmaries. Of course the chances of death are increased in proportion to the number of times a house or a bed is occupied during the year. In a less degree, the same remarks may apply to a town subject to frequent ingress of visitors, the chances of death increasing with the frequency that the population of the town undergoes change. The following remarks from a local report of my own give a good example of the difficulties in estimating the true deathrate in a town where the visitors at certain times of the year increase the population of 143,956 by as much as 50,000 persons: "The fact that the population of Brighton is, during the greater part of the year, swollen by the ingress of thousands of visitors, of whose number we have no trustworthy estimate, and who yet furnish a quota to its mortality, makes a true calculation of the death-rate per thousand of the population a difficult matter. My predecessor, Dr. Taaffe, estimated that an addition of 10,000 to the resident population would represent the average number of visitors in Brighton. It should be remembered, however, that the census enumerations in April, 1871 and 1881 (which form the basis of calculation of the population for the current year), necessarily included all the visitors present at those dates, and that Brighton benefits by the very slightly lowered death-rate which these additions to its population involve. It must be added, however, that there is probably no time of year in which the number of visitors in Brighton is at so low an ebb as in April (when the national census is taken). "So far as population is concerned, the course theoretically least open to objection is to exclude any estimated population of visitors and all deaths of visitors in calculating the true death-rate of Brighton. "I append a comparative table of death-rates for recent years; those in the first column being the crude death-rates as given by the Registrar-General, and those in the second column being the death-rates calculated on the number of deaths excluding deaths of visitors, and the population excluding all visitors except the small number included in the resident April population, which latter I can suggest no means of eliminating. The death-rates in the second column are consequently too small, though this deficiency cannot exceed 0.5 per 1,000 inhabitants." Effect of Public Institutions.-The consideration of the effect public institutions exert on local bills of mortality naturally follows on a consideration of the effect of migrations, as the inequality arising from such institutions is due to migrations into them from outlying districts. The rule in dealing with a public institution is to deduct the deaths of those inmates derived from outside the district concerning which the calculation is made, at the same time including the deaths of inhabitants of the said district which may have occurred in other institutions outside the district. Thus, Workhouses and Asylums are often situated outside the district from which they receive inmates; and in London and other great towns the large public hospitals receive patients from outlying districts. The deaths in all these cases must be relegated to their respective districts. Thus, during the three months ending Sept. 30th, 1888, there were 71 deaths in the public institutions of Brighton (hospitals and workhouse). The total number of deaths in Brighton during the same period was 416-i.e., 345 in private houses, and 71 in public institutions. G Of the deaths in private houses 7 were of visitors. Now the quarterly population of the municipal borough of Brighton=29,996. Therefore the annual death-rate : 397 x 1000 29,996 =13.23. A more difficult case is that of the Wandsworth sub-district of the Wandsworth District (London, S.W.). This sub-district has within its borders the County Lunatic Asylum, the Hospital for Incurables, St. Peter's Hospital, Wandsworth Prison, and the Royal Patriotic Asylum for Girls; and it sends its sick poor to the Wandsworth and Clapham Workhouse Infirmary, which is outside its borders, while some of its inhabitants die in the large metropolitan hospitals. Total deaths in Wandsworth in 1885 were . 628 In addition there occurred in the Union Infirmary and other outlying Institutions 78 deaths. The mean population of Wandsworth in 1885 was 31,497. (1) The death-rate, uncorrected for internal institutions, and 628 × 1,000 =19.83 per not including outlying institutions, is thousand. 31,497 (2) In order to ascertain the death-rate corrected for internal institutions without including the deaths in outlying institutions, we must ascertain the population of its institutions as well as their deaths. By the census of 1881 this was 1,482, which may be taken as nearly correct for 1885. Corrected population=31,497-1,482=30,015. (3) The death-rate excluding the population and deaths of internal institutions, and including the deaths (but not the population, because this is unknown) of Wandsworth parishioners in outlying institutions. The last death-rate more nearly than any other represents the true mortality of the Wandsworth sub-district. For it excludes the population and deaths in its numerous internal institutions (to which Wandsworth contributes only an inappreciable quota); and it allows for the deaths of Wandsworth parishioners in external institutions. The only fallacy is, that no allowance is made in the population for those inhabitants of Wandsworth who are in the outlying institutions without their illness proving fatal. This number must, however, be comparatively small, and cannot very materially affect the result. Each Sanitary Authority in London is supplied quarterly from the Registrar-General's office with particulars of deaths of their inhabitants in outlying institutions, so that the necessary correction can be made. In provincial towns, however, there is reason to believe that such corrections are commonly avoided. A lunatic asylum or workhouse outside the borders of the town receives patients from it, but its mortality is steadily ignored. On the other hand, the mortality of many rural districts is lowered by the fact that many poorer people when ill become inmates of the neighbouring town hospital and there die. No just reason exists why these deaths should not be fairly stated, and we may hope that public opinion will in time demand this. Influence of Birth-rate on Death-rate.-The influence of the birth-rate on the increase of the population and on the ageconstitution of the population has been already considered. The age-constitution of the population has a most important bearing on the death-rate, which it may considerably modify; the birth-rate affecting the death-rate only in so far as it alters the age-constitution of the population. Great fallacies are involved in the question of the relation between the birth-rate and death-rate. The two are assumed to be connected in two different ways. (1) A high birth-rate causes a high death-rate, and a low birth-rate causes a low death-rate. (2) The exact converse of this holds good, the death-rate controlling the birth-rate. The interpretation of the true relationship between the birthrate and death-rate depends on an appreciation of what has been said (page 70) concerning the influence of the birth-rate on the age-constitution of a population. It is evident that if, owing to a high birth-rate, there is a larger proportion of children in one community than in another, and the relative sanitary conditions of the two are equal, there will be more deaths of children in the former; and inasmuch as the rate of mortality of young children is higher than that of all others except the aged, the general death-rate will be raised. But if the high birth-rate be continued, there will not only be a large proportion of children, but of others between 10 and 40 years of age, at which ages a low rate of mortality holds; and this factor counterbalances the other, and makes a continued high birth-rate produce a low death-rate. Thus, speaking generally, the mortality of a population in which there is an excess of births over deaths is lower than that of a stationary population, in which the births and deaths are equal in number; the reason being on reflection obvious. In the latter case there is a larger proportion of old people than in the former. The rule therefore is, that the lower the average age of a population (mean age of living) the lower ought to be its death-rate. The late Dr. Letheby may be regarded as the spokesman of those who take a view diametrically opposed to that just stated. He held that "the birth-rate is the controlling element of the death-rate"; that "an increase in the rate of mortality is often a sign of prosperity, for a high death-rate means a high birthrate, and a high birth-rate is the invariable concomitant of prosperity." According to this view, a high birth-rate is a direct cause of a high death-rate, owing to the great mortality among infants We have, however, following the lines of Dr. Farr's argument, shown the fallacy of this theory, ignoring as it does the important fact that a continuously high birth-rate |