well-controlled experiment was carried out at Milton, Massachusetts, in 1809, with the same result.

The statistics of the early part of the nineteenth century furnish the first evidence of the effect of vaccination as applied upon a large scale. In Sweden, for example, the average annual smallpox rate per million was 1,914 from 1792 to 1801, 623 from 1802 to 1811, and 133 from 1812 to 1821. In Berlin the actual deaths from the disease amounted to 4,453 for the ten years 1782-91, 4,999 for the next decade, 2,955 for 1802-11, and 555 for 1812-22. The facts are brought out in a still more striking manner when the figures are plotted graphically, as was done for London from 1650 to 1900 by Dr. Newsholme.* Wallace published a similar diagram of the Swedish death rates which is alone enough to convince a candid student that something remarkably affected smallpox mortality about 1800; but he closed his eyes to its obvious teaching, and maintained that inasmuch as the curve fell off sharply from 1800 to 1803 before vaccination had become general, the decrease was due not to vaccination, but to 'sanitation.' It is certainly true that the deaths from smallpox decreased in the two or three years after 1800 without reference to vaccination, just as they had decreased periodically after every epidemic in the eighteenth century. But after every such previous decrease the mortality had risen again within five or ten years to another maximum. Why, after the decrease in 1803, did the death rate in Sweden remain at a minimum, never having risen since 1809 over 1,000 per million, and but four times over 500, while in 1801 it was 2,566, in 1800 5,126, in 1799 1,609, in 1796 1,963 and in 1795 2,956? There is not the

'The Epidemiology of Smallpox in the Nineteenth Century,' British Medical Journal, July 5, 1902.

In

smallest shred of evidence that 'sanitation' received any great and sudden impetus at exactly this time, unless sanitation be used to cover all the arts which tend toward 'the prevention of premature death.' this wholly legitimate sense sanitation includes a number of prophylactic measures, each adapted to the diminution of a specific disease. When sanitation covered only isolation and quarantine it could control plague to a certain extent, but not smallpox, not typhoid fever, not diphtheria, not measles. When vaccination became a sanitary measure, sanitation conquered smallpox; but typhoid fever was not restricted until the day of water supplies and sewerage systems; diphtheria, not until the introduction of antitoxin. A fairly steady decrease in the general death rate has, indeed, occurred, due to a complex of factors not easily analyzed, but a sudden collapse such as that which affected the smallpox death rate after 1800 has never been manifest without a definite and tangible cause. That 'sanitation' has not affected the other zymotics to the same degree as smallpox has been graphically shown by A. F. Burridge in a recent publication.*

During the first quarter of the nineteenth century it was thought that a single vaccination in infancy would give indefinite protection against smallpox; but about 1830 this view began to lose ground. An adult population now existed, protected, not, as in other times, by previous attacks of smallpox, but only by the less potent effect of vaccination. Smallpox began, therefore, to recur, but modified in two notable respects. In the first place, its age incidence had shifted; whereas of 1,252 cases in three Prussian towns before vaccination began, 94.5 per cent. were under ten and not one over twenty years; of 1,677 cases in Wür

* Vaccination and the Act of 1898,' Journal of the Institute of Actuaries, October, 1902.

temberg after vaccination began, 18.4 per cent. only were under 10, and 42 per cent. over twenty years. So it is shown by Dr. Creighton in his article on 'Vaccination' in the Encyclopedia Britannica,' that, in England and Wales, about 1847, three fourths of the deaths occurred under five years, while in the eighties less than a quarter of the decedents were of this age. In the second place, beside this shifting of incidence, smallpox among the vaccinated proved much less fatal, even when it was contracted, than among the unvaccinated.

Although minor epidemics began to recur, smallpox in vaccinated countries was insignificant in amount until 1870–5, when a 'pandemic' swept over Europe which recalled the normal conditions of the prevaccination period. Considering the varying virulence of disease at different periods, and the fact that the importance of revaccination was not at all realized, such an epidemic was to be expected. The statistics for the early seventies have been used most dishonestly by the anti-vaccinationists in comparison with selected years of low mortality immediately after the introduction of vaccination in the attempt to show that no progress has been made. The worst year of this period in England and Sweden, however, showed a death rate about half the average yearly rate for the last quarter of the eighteenth century.

A comparison of the incidence of smallpox in this pandemic of 1870-5 upon different countries introduces the second class of evidence as to the value of vaccination. Thus Dr. Edwardes shows that for four countries having compulsory vaccination the average yearly smallpox death rate per million inhabitants was as follows: England, 361: Scotland, 314; Bavaria, 346; Sweden, 333. On the other hand, the rate for the same period was 953 in Prussia, 1,360 in Austria, 1,293 in Belgium and 958 in the Netherlands. All these countries

had at this time no compulsory vaccination. The reverse has been affirmed in the case of Prussia, and Creighton, in the 'Encyclopedia Britannica,' states that revaccination has been more or less the law in Prussia since 1835,' and that 'Prussia was the best revaccinated country in Europe' in 1871. Dr. Edwardes discusses this question in some detail and quotes the official documents, which show explicitly that there existed in Prussia 'kein gesetzlicher Zwang zur Impfung.' Furthermore, the actual ratio of vaccination to births in Berlin is on record, and the percentage ranged from 29 to 58 between 1865 and 1870. In this city there were 6,326 smallpox deaths per million living in 1871!

The great pandemic' taught the lesson that both vaccination and revaccination were essential. In 1874 Germany enacted a law providing for compulsory vaccination within the second year and revaccination within the twelfth year. In Prussia the death rate, which had ranged from 95 to 2,624 per million from 1866 to 1874, dropped to 36 in 1875 and has been under 10 since 1885. For the empire as a whole, statistics, available only since 1886, show a rate of 4.2 in that year, decreasing to .5 in 1895 since which there has been annually less than one death per million. A comparison with the statistics of Austria graphically made by Dr. Edwardes furnishes as striking a proof that vaccination is the only kind of sanitation which affects smallpox as could well be desired. Before 1870 the two countries had about the same amount of smallpox; since 1875 that in Austria has increased and that in Prussia has practically disappeared. The only difference in conditions lies in the law of 1874.

Army statistics furnish striking confirmatory evidence. Thus Burridge* compares

* Loc. cit.

the Prussian army, in which revaccination on entrance has been compulsory since 1834, with the French army, where it has only been thoroughly carried out since 1888, and with the Austrian army, where there was no revaccination prior to 1886. The attack rate per 100,000 in 1875-85 was 4.7 in the Prussian army, 133.6 in the French army and 333.7 in the Austrian army. In the twenty-five years 1875-99 there were only two deaths from smallpox in the Prussian army, one in 1884 and one in 1898. The main point to notice is that these extraordinary results have been attained by a general revaccination of the whole population. Revaccination of only a single class in the community can not prevent the occurrence of occasional cases in that class, because in a large body of men there must always be some vaccinations which have not been successful. Thus the smallpox death rate in the English army with revaccination has ranged from zero to twenty-nine during the last forty years. Wallace in 'Vaccination a Delusion' made these figures look larger by raising them to rates per million (the basis of calculation being about 200,000 men), and then compared them with the rates for Ireland at the age period 15 to 45, which were only slightly higher from 1864 to 1894 (58 for the army, 65.8 for Ireland). Later he showed that the rate for 1873-94 was 37 in the army, 36.8 in the navy and 14.4 in the city of Leicester, and concluded that 'all the statements by which the public has been gulled for so many years as to the almost complete immunity of the revaccinated army and navy are absolutely false.' "There is no immunity. They have no protection.' That is, Mr. Wallace selects one island in Europe where, largely from its isolation, smallpox happens not to have been serious, and one town in England where there has been almost no smallpox, and because these two places have had

extraordinarily low death rates he maintains that the low army death rate, indicates no protection! Yet the figures were before him which showed that the average of the annual death rates in the navy, which was less than 32 from 1873 to 1899, had been 257 from 1860 to 1873; in 1873 an order was issued which provided for the vaccination of all recruits on joining.

The evidence derived from a comparison of the same country before and after the introduction of vaccination, and that based on the contrast at the same period between countries having different degrees of vaccination, have now been briefly considered. The third class of facts includes the 'direct evidence' of the incidence of smallpox upon persons in the same community protected and unprotected by vaccination. At Chemnitz in 1870-1, a special census was made to determine the condition of the population as regards vaccination, and it appeared that among those protected by vaccination or previous smallpox the death rate was 1.2 per 10,000, while in the unprotected it was 442.9. Similarly at Sheffield in 1887-8* the deaths per 10,000 were 7.5 among the vaccinated and 347.9 among the unvaccinated. An objection to statistics of this sort, made with some plausibility, is that the unvaccinated class includes a large proportion of children and of persons in feeble health or living under poor sanitary conditions. Regarding the first point, the Sheffield figures are conclusive. They are divided according to age periods and show that the rates per 10,000 living between fifteen and twenty years were, 7.0 in the vaccinated and 1,355.5 in the unvaccinated. Here no age difference comes in question. The second contention is met by the statistics collected by Körösi with reference to 14,678 persons dying from various causes in some Hungarian hospitals in 1886. The unvaccinated constituted 14. *Reviewed by Burridge, loc. cit.

per cent. of those who died from other diseases than smallpox and 81 per cent. of those who died from smallpox. Obviously it was the lack of vaccination which was at fault here, not feeble health nor unsanitary conditions. In opposition to these figures the anti-vaccinationists quote the experience of the city of Leicester, where since 1882 the number of vaccinations has steadily decreased, falling to less than two per cent. of the births in some recent years. Smallpox has been introduced a number of times (38 cases in 1892, 308 in 1893, 8 in 1894, 4 in 1895, 4 in 1901), but has not spread extensively, and the death rate has remained very low. The opponents of vaccination also quote, by way of contrast, statistics showing that an increasingly large proportion of hospital cases of smallpox occur among the vaccinated,* and that in epidemics the attack of an unvaccinated person is often not recorded for some time.† Facts of the last two classes have, of course, no special significance except to indicate the need for revaccination. No one now supposes that a single vaccination affords absolute permanent protection, and with the increase of vaccination there must naturally come an increase of cases among the yaccinated. The experience of Leicester, on the other hand, is certainly of interest. It shows that under certain conditions the dangers of neglected vaccination may for a time be braved with impunity by a considerable portion of the community. This has been so far accomplished by prompt reporting and strict isolation of cases, and, according to the chairman of the public health committee of the town by the fact that 'a handful of the population, including the medical men, sanitary staff,

* London smallpox hospital, 40 per cent. in 1838, 94/10 per cent. in 1879-Wallace.

The first unvaccinated case was the 174th at Cologne in 1870, the 42d at Bonn in the same year, and the 225th at Liegnitz in 1871-Creighton.

smallpox nurses, etc., are as well vaccinated in Leicester as in any other town, so that a cordon of protected persons can at once be drawn around any case of smallpox which may occur. It should be remembered, however, that the population of Leicester is still to some extent protected by the vaccinations carried out prior to the anti-vaccinationist agitation. Thus of the 358 persons attacked in 1892-5, 198 were returned as having been at some time vaccinated. The experience of Gloucester is ominous for the future of the 'Leicester experiment.' Prior to 1892-3, according to Dr. Edwardes, 'vaccination had been. almost in abeyance, in Gloucester, and the inhabitants lived in a fools' paradise.' The result was an epidemic of 1,979 cases, with 434 deaths in a population of about 40,000, giving a death rate of 10,000 per million!

With regard to the smallpox occurring in persons once vaccinated, there are two points to notice. In the first place, the ratio of deaths to cases is far lower than among the unvaccinated. Thus at the Leipsic city hospital in 1870-2 99 died among 139 unvaccinated cases, 116 died among 1,504 vaccinated cases, and none among 13 revaccinated cases. Creighton and Wallace object to these statistics on the ground that the death rate thus apparent among the unvaccinated is obviously too high, because 'in pre-vaccination times the death rate (18.8 per cent.) was almost the same as it is now in the vaccinated and unvaccinated together' (Creighton. Now it is quite impossible to fix any such general fatality rate; the ratio of deaths to cases has varied within wide limits both in the eighteenth century and recently. In the second place, it has been claimed that the 'unvaccinated' death rate is swollen by the inclusion in that class, of children who escaped vaccination on account of feeble

Windley, Leicester and Smallpox,' Journal of State Medicine, January, 1903, p. 21.

health. In the case of Gloucester, where vaccination has been so generally neglected, this objection can hardly apply. Yet at Gloucester in 1892-3, there were, under ten years of age, 26 attacks among the vaccinated with one death, and 680 attacks among the unvaccinated with 279 deaths. Statistics for six towns collected by the English Royal Commission of 1889 showed fatality rates of 35.4 among the unvaccinated and 5.2 among the vaccinated. The third objection made to the hospital statistics, namely, that the deaths of the unvaccinated class are unfairly increased by the inclusion of doubtful cases and those who have been vaccinated but show no scars, can scarcely apply to the commission's analyses. It will not, at any rate, have much weight, except with those who, like Mr. Wallace, believe that "in this matter of official and compulsory vaccination both doctors and government officials, however highly placed, however eminent, however honorable, are yet utterly untrustworthy."

A second important characteristic of the cases of smallpox in a once vaccinated population is that they are not only comparatively light, but that they affect the later periods of life; and this represents an important gain in the life capital of the community. During the epidemic of 1870-3, Bavaria, with compulsory vaccination, had 851 deaths under, and 3,520 deaths over, twenty years, while the Netherlands without compulsory vaccination had 14,048 deaths under twenty and 6,524 at higher ages. In the same great epidemic 71 per cent. of the deaths at Leicester and 64 per cent. of the death at Gloucester occurred under ten years. In London the percentage falling in this age class was 37, and in Warrington, with still more thorough vaccination, it was 22.5.

A single vaccination then greatly reduces the probability of an attack of smallpox,

postpones it to a later period of life and renders it less dangerous if it does ensue. To ensure absolute protection revaccination. is required; and its efficacy is well indicated by the experience of the Prussian army. In addition, one single bit of evidence may be adduced which is more striking, perhaps, than all the rest, the statistics of nurses in smallpox hospitals. These figures are of special interest because we have here a fairly large class of persons whose condition as to vaccination is accurately known, and who are uniformly exposed to the contagion of the disease; and the experience of two such communities is quoted by Dr. Edwardes. "During the epidemic of 1871 there were 110 persons engaged in the Homerton Fever Hospital, in attendance on the smallpox sick; all these, with two exceptions, were revaccinated, and all but these two escaped smallpox." "Of 734 nurses and attendants in the Metropolitan Asylums Board Hospitals, 79 were survivors from smallpox attack-they escaped infection; 645 were revaccinated on entrance-they all escaped; 10 were not revaccinated, and the whole 10 took smallpox."

If statistics ever proved anything those quoted above prove the protective influence of vaccination. If any fact in science is. certain, it is certain that a successful vaccination absolutely prevents smallpox for a period of some seven to ten years, that after that period it renders the disease less fatal, and that its complete protective effect may be renewed by revaccination. The conclusion is obvious, not only that the state should oblige primary vaccination, but, in the words of a minority of the British Royal Commission, that 'a second vaccination, at the age of twelve, ought to be made compulsory.'

C.-E. A. WINSLOW.

BIOLOGICAL DEPARTMENT,

MASSACHUSETTS INSTITUTE OF TECHNOLOGY.