rHci3? mn Tl By. m- ^_tf7 ijw Digitized by the Internet Archive in 2010 with funding from Open Knowledge Commons http://www.archive.org/details/epidemicdiphtherOOnews EPIDEMIC DIPHTHERIA A RESEARCH ON THE ORIGIN AND SPREAD OF THE DISEASE FROM AN INTERNATIONAL STANDPOINT ARTHUR NEWSHOLME, M.D.Lond., M.R.C.P Lond. Examiner in State Medicine to the University of London, Medical Officer of Healtli of Brighton SWAN SONNENSCHEIN & CO. LIMD /vvy Butler & Tanner, The Selwood Printing Works, Frome, and London, TABLE OF CONTENTS CHAPTEB . PAGE Introduction 1 I — North and South America 9 Massachusetts, Boston, New York, Brooklyn, Phila- delphia, Baltimore, Pittsburg, Cincinnati, St. Louis, Denver, New Orleans, Chicago, Montreal, Buenos Ayres, Comparison of Eleven American States. TI — England and Wales 29 London, Hastings, Eastbourne, Brighton, Croydon, Portsmouth, Southampton, Bristol, Cardiff, Birming- ham, Wolverhampton, Leicester, Liverpool, St. Helens, Oldham, Manchester, Salford, Sheffield, Leeds, Bradford, Halifax, Huddersfield, Hull, South Shields, Newcastle-upon-Tyne. Ill — Scotland and Ireland . ... . . . 44 Kilmarnock, Glasgow, Paisley, Dundee, Aberdeen, Greenock, Edinburgh, Perth, Dublin, Belfast, Cork, Limerick, Londonderry, Waterford. TV — The History of the Epidemic of Diphtheria in London in the Fifth and Sixth Decades of The Nineteenth Century 49 V — The History of Diphtheria in London, 1888-9G, AND its bearing UPON THE METHOD OF SPREAD OF THE Disease 55 VI — France, Italy, Holland, and Belgium . . 62 Paris, Bordeaux, Lyons, Marseilles, Genoa, Lausanne, Barcelona, Lisbon, Pome, Florence, Turin, Milan, Venice, The Hague, Rotterdam, Amsterdam, Ant- werp, Brussels and suburbs. VII — Germany, Austro-Hungary, and Eussia , . 73 Hamburg, Berlin, Dresden, Leipzig, Breslau, Frank- fort-am-Main, Munich, Trieste, Prague, Buda- Pesth, Vienna, St. Petersburg, Moscow, Bucharest. VIII — Scandinavia 85 Norway, Christiania, Gottenburg, Stockholm, Copen- hagen, Provincial Towns of Denmark, Faroe Islands. IV CONTENTS CHAPTER PAGE IX — Africa and Asia 92 Cairo, Alexandria, Cape Colony, Calcutta, Japan and its Chief Towns. X — Australia and New Zealand .... 96 South. Australia, Victoria, Queensland, Melbourne, Sydney, Adelaide, New Zealand, Nine Districts of North and South Islands. XI — Eelationship between Prevalence of and Mortality from Diphtheria .... 105 Norway, Copenhagen, Berlin, Hamburg, London. XII — Summary of the History of Diphtheria Ante- cedent TO Statistical Records . . . 110 XIII — International Review of the Prevalence of Diphtheria as shown in the preceding diagrams 127 XIV — Are there Indigenous Foci of Diphtheria? . 133 XV — Influence of Improved and Extended Means OF Communication 135 School Attendance, Relative amount in Urban and Rural Districts, etc. XVI — The Conditions determining the Pandemics OF Diphtheria 140 Illustrated by London. Brighton, Croydon, Sheffield and Bradford, Glasgow and Edinburgh, Chris- tiania, Hamburg, Berlin, Frankfort-on-the-Maine, New York, Chicago, Melbourne, Adelaide, Auck- land and Christchurch. XVII — The Relationship between Rainfall and Diphtheria . 157 XVIII — The Influence of Soil upon the Develop- ment OF Diphtheria 168 XIX — Concluding Remarks 189 Index of Names of Persons 193 Index of Names of Places 194 INTRODUCTION. The investigation embodied in tlie following pages has occupied my attention for several years, during which I have been gradually collecting the necessary data from every civilized country. It can scarcely happen that in such a complex task every single death-rate should have been accurately calculated, but the greatest care has been exercised, and I believe that very few errors will be found, and that such as may have occurred will not vitiate any conclusions stated in the body of the book. Some measure of originality is claimed for the method of statement of the facts. The death-rate for each year is plotted out in diagrammatic form on a scale which is constant for all the diagrams throughout the book, with the exception of Figs. 12, 23 and 24. The necessity for tabular statements is thus avoided, as the death-rate for any given year can at once be ascertained by reading the scale attached to each diagram. Furthermore, the relative bulk of diphtherial mortality in different towns and countries can be seen in a manner which is much more impressive than a tabular statement of yearly death-rates, and more accurate than an average death-rate for a series of years. The avoidance of average death-rates for succes- sive series of years obviates a source of error which, in my opinion, is most important and serious. The diagram enables the number of epidemics occurring in the period investigated to be exactly stated; while a comparison of the average death-rates may include periods which contain one, two, or more epidemics, according to the arrangement B 2 INTEODUCTIOX. of the serieS; thus leading one possibly to think that the amount of disease has increased or decreased when the converse is true. A statistical investigation based on official data of varying degree of accuracy is apt to be received with a modicum of incredulity^ the judgment of the reader being prejudiced by erroneous preconceptions. While so many are ready to say that '^ anything can be proved by statis- tics/^ but few realize that without statistics nothing can be proved j and that unless complete ignorance is to persist, even defective statistics must be used to the full extent of their value. Xor is it sufficiently realized that perfectly valid conclusions can be drawn from imperfect statistics, when these are skilfully and honestly employed. In say- ing this, I have no wish — nor is there need — to suggest any distrust of the data on which the present treatise is based. The usual complaints against such statistical data may be classified under three or four heads : Firstly, the changes in nomenclature on which follow necessarily changes in the numbers under a given disease, not due to the varying mortality from this disease. Secondly, the incomplete character of the registration of deaths in a given town or country, which incompleteness gradually becomes diminished, and consequently comparison between recent and more remote years is to that extent vitiated. Thirdly, the fact that deaths from a disease do not neces- sarily form a correct index of its prevalence, as the fatality of diseases like diphtheria varies greatly. Fourthly, differences in the age constitution of the populations under comparison, which disturb any contrast of the actual amount of diphtheria prevalent. 1. The importance of the first of these objections has been minimised by giving, wherever the available data INTEODUCTION. 6 exist, the death-rate from diphtheria in each diagram separately from that from croup. By this means the gradually increasing transference from croup to diphtheria can be seen at a glance ; and the true facts become evident by an inspection of the combined curve for the two diseases. Other transferences than that of croup to diphtheria are of trifling consequence in the periods embraced within the statistics to be given. 2. The second objection may affect any comparison of relative amounts of diphtheria in different communities. It is obvious, however, that any error caused by it will be on the side of undue moderation of statement. Thus the extraordinary amount of diphtheria shown in the American diagrams would almost certainly be increased in many instances had the registration of causes of death been more complete. There is no sudden transition from incomplete to complete registration ; and although the amount in any given year may be understated, the position of this part of the curve in relation to the parts for neighbouring years forms a trustworthy guide as to whether it is a year of epidemic or inter-epidemic pre- valence of diphtheria. The same argument applies to all the diagrams. The amount of diphtheria may in some of the diagrams be understated, but the teaching of the diagrams as to which are the epidemic and which the inter- epidemic years can in each instance be confidently and implicitly accepted. It is on the position rather than on the height or depth of the crests and troughs of the epidemic luaves that the main teaching of this worh is based, and in that respect it is believed to be absolutely accur- ate. •3. That the present work almost throughout deals with deaths from and not with total cases of diphtheria may be regarded by some as a limitation of its utility. After 4 INTRODUCTION. most careful and detailed preliminary investigation, I deliberately adopted this course. In the first place, there were no countries, except the Scandinavian, in which records of a system of general compulsory notification of diphtheria were available for a sufiiciently long series of years to make them valuable in a study of epidemiology. In the next place, the experience of compulsory notifica- tion of diphtheria in this country for the comparatively few years in which it has been commonly adopted, tends to show that cases which a few years ago would have escaped notification are now notified to a somewhat larger extent (see Fig. 50b).* Consequently deaths are in England a more reliable indication of prevalence than notifications. The same remark does not apply in Norway and Copenhagen, as may be seen by a glance at Figs. 48 and 49. The fatality of diphtheria in epidemic and in inter-epidemic periods forms an interesting subsidiary in- vestigation, to which I can only devote one short chapter. 4. Diphtheria is a disease, four-fifths of the deaths from which occur in children under ten years of age. Consequently the comparison between the different towns and countries whose death-rates from diphtheria are graphically represented in the following pages is inexact in so far as the relative proportion of children under ten is not the same in these different populations. It has been quite impracticable to make any correction for varying age-distribution of the populations in different countries and towns, t I must, therefore, content myself * For remarks modifying this statement see page 109. t The extent of correction for variations in age-distributioii of population may be gathered from the following examples : — In 1890, there were in Boston for every 100 persons at all ages 25*2 tmder 15 years of age, in Philadelphia 27"9, in New York 28'8, in Hamburg 32"1, while in London, in 1891, the percentage ages 0-15 was 32"6. It is evident that if correction for age-distribution of INTRODUCTION. O with drawing attention to this source of error^ which is within very small limits, and could furthermore, as pre- viously remarked, only slightly affect the height of any particular curve without altering in the least the relative height of its constituent parts, with which I am almost solely concerned. A glance at the diagrams in the following pages shows that in the majority of instances they relate to urban populations. In the case of London the diagrams have been subdivided (pages 57-9), so as to show the yearly death-rate in different districts of the metropolis ; and more exact truth might have been attained had this been practicable for other great cities. Death-rates for rural districts have not been given, because they are difficult to secure for a sufficiently long series of years, and because with small populations accidental oscillations of yearly curves of mortality become exaggerated. This fact makes the international study of diphtheria here detailed one of diphtheria as affecting chiefly urban communities (see page 138). Having briefly discussed the trustworthiness of the facts set forth in the diagrams which form the statistical basis of this treatise, it remains for me to make a few remarks upon the main conclusions derived from these diagrams. These are two : firstly, that diphtheria spreads from town to town and from country to country, the means of spread being almost certainly personal infection ; and secondly, that certain climatic conditions are necessary for the development of diphtheria on an epidemic, and still more for its development on a pandemic scale. In the population were made, the result would be to increase to an appreciable extent the excess of diphtheria already displayed in the American diagrams. INTRODUCTION. first portion of tHe book, dealing with, the prevalence of dipttheria in different countries, I have, throughout, attempted to discuss the epidemics in a given place as related in time to epidemics in neighbouring towns and countries, and to indicate the gradual progression of the disease. In reading this part of the book, geographical considerations should be borne in mind, and the atlas consulted, as the distances between towns in, for instance, the United States is immensely greater than between towns in England. The teaching as to gradual pro- gression of the disease would have been clearer in certain instances had one constructed curves of monthly or quarterly instead of annual death-rates. Those who are aware of the laborious character and complexity of work of this kind will, however, readily forgive an omission to make the curves by this means more complete but at the same time less easy to follow. In the second part of the book, I have dealt with the climatic conditions, in the absence of which diphtheria never becomes epidemic on a large scale. A superficial perusal of the book might produce the impression that these two factors are not consistent with each other. A more careful study will, I believe, prove that both are necessary. Diphtheria is spread chiefly by personal in- fection, but this infectivity is only operative on a large scale under the influence of the climatic conditions to be hereafter described. This work would have been impracticable, but for the help of numerous distinguished statisticians and medical officers of health in all the countries dealt with in the following pages. Where no other reference is given, it may be assumed that the statistics are obtained from official reports. To many English medical officers of health I am greatly indebted for records embracing long INTRODUCTION. 7 series of years, often obtained witli considerable difficulty. I also tender my cordial thanks to Dr. Bentzen, of Christiania ; Dr. J. Bertillon, of Paris ; Dr. Borthwick, of Adelaide ; Dr. Grimsbaw, Registrar-General of Ireland ; M. Goseti, Venice ; Mr. Goto Sbimpei, Director- General of the Sanitary Bureau Home Department, Japan; Mr. F. L. Hoffmann, F.S.S., Newark, N. J. ; Dr. Janssens, Brussels ; Dr. Von Jurascbetz, Vienna ; Herr Korosi, Buda-Pesth ; Mr. George Leslie, Actuary to the Government of New Zealand ; Dr. Linroth, of Stockholm ; M. Martinez, Buenos Ayres ; Dr. Mantez, Geneva; Dr. Pistor, Berlin ; M. Ravizza, Milan ; Dr. Reincke, Ham- burg; Dr. Sandwith, Cairo; Mr. Symons, F.R.S. ; Dr. Ashburton Thompson, of Sydney ; the Chief Burgomasters, Antwerp and Rotterdam ; the Directors of the Statistical Bureaux of Breslau, Leipzig and Trieste ; the English Consul, Lisbon ; M. Potzniakoff, Moscow ; for help rendered, in some instances involving great trouble. If I have omitted any from this list who should have appeared in it, I trust they will not think I am ungrateful for the generous help which they have rendered in an investi- gation, the complexity of detail of which has sometimes been almost overwhelming. CHAPTER I. NOETH AND SOUTH AMERICA. The vital statistics of the cities included in the following diagrams are not all equally complete^ and some allow- ance has to be made in this respect. Thus in Michigan, according to Dr. Baker (report on the Vital Statistics of Michigan^ 1892), in 1881 the registered mortality from all causes was probably not more than 60 per cent, of the actual mortality, and presumably a similar proportion would apply for diphtheria. It is evident, however, that this incompleteness of registration of deaths will not vary in extent suddenly, and will therefore not materially affect any conclusions that may be drawn from the accom- panying diagrams as to epidemic and non-epidemic years of diphtheria. There can be no reasonable hesitation in accepting the teaching of these diagrams in this respect as sufficiently accurate for all practical purposes. It is the fluctuations observable in the heights of the diagram, rather than the actual height in any given instance, that should be noted. We will commence our consideration of the American returns of diphtheria mortality with those for the New England States, as their statistics are more complete and accurate than those of most other States. The mortality from diphtheria for the whole of Massa- chusetts is shown in Fig. 1 : — 10 DIPHTHERIA. STATE OF MASSACHUSETTS. Death- rate per 1000 of CD 1 CO 1 Popu- lation. 1-80 \ ■60 S Note. — In this and the following diagrams, \. V '. indicates that the statistics for diphtheria and croup are stated in combination (as in Figs. 1 and 2) ; indicates the death-rate from diphtheria alone (as in Fig. 2) ; indicates the death-rate from croup alone (as in Fig. 2). NORTH AND SOUTH AMERICA. 11 It will be seen that for the years 1856-94 inclusive^ the lowest death-rate from diphtheria plus croup was 43 per 100,000 inhabitants in 1858, the highest 196 in 1876, the total endemic prevalence of the disease being very great, as shown by a glance at Fig. 1. It is possible that the minimum year, 1858, had really more diphtheria and croup than is indicated in the diagram, owing to the de- ficient recognition of the disease at that period. After the great epidemic culminating in 1863, a marked remis- sion occurred, the second great epidemic reaching its maximum in 1876-77, after which the amount of the disease remaining endemic was greater than that follow- ing the first epidemic. Boston. — During the 35 years 1861-1895, the death- rate from diphtheria plus croup in this city varied from 35 per 100,000 inhabitants in 1872 to 218 in 1881. The pro- portion between diphtheria and croup is shown for the years 1880-94 in Fig. 2. The general shape of the epidemic curve is the same whether we take diphtheria alone or this in conjunction with croup; and the same remark applies to most of the diagrams to be subse- quently considered. It will be observed that there was a great epidemic of the disease culminating in 1863-4, and that then an interval of comparative freedom from it lasted for ten years. In 1875-6 another great epidemic occurred. With but a short and imperfect ebb of the epidemic tide, a still more fatal epidemic occurred in 1880-1, followed by later epidemics culminating in 1889 and 1894. Between these maximal diphtheria years a large amount of diphtheria was endemically present, the death-rate from 1875-95 inclusive never being lower than 62 per 100,000 (in 1891). Providence. — In Providence during the years 1868-75 the death-rate from diphtheria plus croup varied from 25 12 DTPHTHEBM. per 100,000 in 1868 to 357 in ]877. The high death-rate in 1877 represented the acme of a great epidemic in 1876-79. Following on this were smaller epidemics in 188],, in 1886-90, and in 1895 (Fig. 3). BOSTON. (Population : 1870—250,500 ; 1890—448,400.) Fig. 2. New Yokk. — The death-rate from diphtheria plus croup was at its lowest (69 ^) in 1873, and at its highest (276) in 1877j a year earlier than in Providence, and two to three years earlier than the corresponding maximum in Boston. ^ Where not otherwise stated the deaths are all given in their proportion to 100,000 of the population. The scale on all the diagrams gives the death-rate per 1,000 of the population. NORTH AND SOUTH AMERICA. 13 Death- rate per 1000. 3-60 3-4Q, ^■to 3 00 , a aoi 2gQ g 40 , 2-ao PROVIDENCE. (Population :I1870— 68,904 ; 1890—132,146.) Fig. 3. ]4 DIPHTHERIA. In 1881 a second maximum occurred, closely approaching the first in height, while later epidemics of not much smaller magnitude culminated in 1887 and in 1894 re- spectively. NEW YORK. (Population: 1880—1,111,941; 1890—1,617,997.) Death- rate per 1000. ?.60 O '- w Q d 1 2-40 \ 2 20 ; 200 X r i 1 l-SO \\ i 1 i-eo 1-40 1 1 1" ■ 1 li lil 1 L 1 i' n -L2Q_ : 1 r J l-Ofl i H n M^ •M 1 ^^1 ■GO .40 •20 1 Fi? Beooklyn. — Being close to New York, the curve of its diphtheria mortality is, as might have been anticipated, almost identical with that of the latter citv. The death- NOETH A^D SOUTH AMEPJCA. 15 rate from diphtheria in Brooklyn reached 215 in 1875, and sunk to its minimum of 59 in 1884, as compared with Death- rate per 1000. CO ; 1 CO CO 1 1 S; BROOKLYN. (Population : 1880—564,448 1890—871,852.) igo i.ftrt I.Afl I-90 J;. 100 li 1 L •QQ ■C^O Ln ■A a _^0 ; ■ 1 J Ij .en ■ ll .aa PHILADELPHIA. (Population : 1880—901,380 1890—1,064,277.) ^ J ■■fli H I BH 1 1 Fisf. a maximum for New York for diphtheria of 208 in 1877, and a minimum during the same period for the two cities (1875-95) of 78 in 1890. 16 DIPHTHERIA. Philadelphia. — The death-rate from diphtheria has a considerably smaller range, and is absolutelj lower than in the other cities hitherto described. A portion of this difference (but only a portion) may be caused by a smaller proportion of children under five years of age in its popu- lation; diphtheria being chiefly fatal in childhood. As bearing on this point, it is interesting to note that in 1890 — the census year — the proportion of children under five years to the total population at all ages was 9*9 per cent, in Philadelphia as compared with 11*7 per cent, in New York. The death-rate from diphtheria in Philadelphia ranged from 18 in 1868 to 131 in 1891. A comparative study of the diagrams already given seems to indicate a gradual spread of infection from city to city. Thus taking the period comprised by all the curves from 1870 onwards, it is evident that Boston's first epidemic culminated in 1875-6, Providence's first epidemic in 1876-7, New York's and Brooklyn's first epidemic in 1876-7, and that of Philadelphia, 1875-6. The 1880-81 epidemic in New York and Brooklyn occurs in 1881-3 in Philadelphia ; and the 1886-89 epidemic of these cities culminated in 1891 in Philadelphia. Baltimore. — This city furnishes in its 1881-83 and its 1891-2 epidemics further support of the view that the infection of diphtheria gradually spread from the eastern sea-board of the States westward and southward. The maximum death-rate from diphtheria was 173 in 1882 and the minimum 22 in 1888. Pittsburgh. — The returns for Pittsburgh extend from 1877 to 1894, the maximum death-rate from diphtheria being 340 in 1878, and the minimum 47 in 1894. The great epidemic of 1877-80 evidently corresponds to the 1875-76 epidemic in Boston, and the 1884-87 epi- demic to the 1880-83 epidemic in Boston. BALTIMORE. (Popiilation : 1880—410,000 ; 1890—455,427.) PITTSBURGH, (Population: 1880—156,389 ; 1890—238,617.) Fig. 6. 18 DIPHTHERIA. Cincinnati. — In this city the death-rate from diphtheria 2)lus croup ranged in the period 1868-95 from 165 in 1890 to 50 in 1875. The diagram for this city is remarkable for the comparative smalhiess as well as the lateness of the rise in the decennium 1871-80. Its maximum was in 1878, as compared with 1877 in ISTew York, Similarly in the] next decennium the maximum of diphtheria was reached in 1890, as compared with 1887 in New York. CINCINNATI. (Population: 1870—942,000; 1880—1,206,000.) St. Louis. — During the period 1868-95, the death-rate from diphtheria ^Ziis croup ranged from 262 in 1887 to 26 in 1868. The epidemic in the decennium 1871-80, as judged by the combined curve for diphtheria and croup, reached its acme in 1876^ earlier than in Cincinnati, con- trary to what might have been expected on the assump- tion that the disease had travelled from east to west. In NOETH AND SOUTH AMERICA. 19 the decennium 1881-90 two great epidemics occurred almost commingling, and in 1895 tliere is wliat looks like tlie beo-inning of another large epidemic. ST. LOUIS. ^ I" Population: 1870—310,! 1890- 451,770.) NEW ORLEANS. i'lg. s. 20 DIPHTHERIA. Denver, Colorado. — A return from this city shows that the death-rate from diphtheria ;plus croup, was 28 per 100,000 of the population in 1886; 84 in 1887; 141 in 1888; 114 in 1889; 261 in 1890; 154 in 1891 ; 74 in 1892 ; 85 in 1893 ; 46 in 1894 ; and 28 per 100,000 in 1895. The population had increased during this period from 70,000 to 145,000. The epidemic evidently reached its maxi- mum in 1890, three years later and higher than the maximum in St. Louis. New Orleans. — For this city on the Gulf of Mexico, the death-rates from diphtheria for the period 1890-95 are alone available. They show a death-rate (not in- cluding croup) varying from 22 in 1891 to 58 in 1894. Whether these rates indicate that during the years 1890-95, New Orleans was enjoying an inter-epidemic period, or whether this southern city enjoys a partial immunity from this disease, cannot with the data before us be stated with certainty. Chicago. — The returns for this city are fortunately available from 1859 onwards. They show a large en- demic prevalence of diphtheria and croup, with epidemics at irregular intervals. The highest death-rate from diphtheria pZws croup was 291 in 1880, the lowest 42 in 1874. The only one of the preceding curves for a nearly equally long series of years is that of Boston. The com- parison between Chicago and Boston shows that the earliest epidemic reached its maximum in 1860 in Chicago, not until 1863 in Boston. The curve for Boston it must be observed however only goes back to 1861. The years of incidence of the later epidemics almost coincided in the two cities, maxima being reached for both in 1876, in 1879-81, in 1887-89, and in 1894, the last maximum being much smaller in Chicago than in Boston. NOETH AND SOUTH AMERICA. 21 STATE OF MICHIGAN. (Population : 1870—1,184,282 ; 1890—2,089,792.) CHICAGO. (Population : 1870—298,977 ; 1890—1,046,964.) Fig. 9. 22 DIPHTHEEIA. Chicago being situate on Lake Michigan, though not in the State of that name, the similarity of the curves for the whole of the State of Michigan and for Chicago is interesting. The curve for the State of MONTREAL. (Population: 1872—120,759; 1894—241,748.) O P Michigan represents a fusion of a number of epidemics in its different parts, hence its greater smoothness than when a single city or district is plotted out. XOETH AXD SOUTH AilEEICA. 23 Canada. Only one return has been secured from Canada. Montreal. — The death-rate from diphtheria in this city- ranged during the years 1875-94 from 245 in 1877, and 259 in 1887 to 24 in 1892. There is an obvious resem- blance between the curves for Montreal and for Chicago, as might have been expected in view of their geograph- ical and commercial connections. The 1877-78 maxi- mum in Montreal corresponds to the 1879-80 maximum in Chicago, while the 1884-85 and the 1887-88 maximum in Montreal are represented by the maximum of 1887-88 in Chicago. The total endemic amount of diphtheria in Montreal, like that in the great cities of the United States, is enormous, as can be seen by comparison with the curves of miniature size, though drawn on the same scale, which- represent English experience (see pp.30 to 43). South Ameeica. The only city from which a return is available is Buenos Ayres, and this only embraces the years 1888-95, the municipal office of statistics having been inaugurated in 1887. Buenos Ayees. — The return gives diphtheria and croup together, and shows a death-rate which was 323 in 1888 and steadily fell to 61 in 1895. It is evident that diph- theria has during recent years been very severely prevalent. The death-rate from diphtheria jplus croup in 1888 was higher than that for any American city in the years for which records are extant, with the exception of Providence in 1877, where it was 357. This is a convenient opportunity for summarising the available data as to the epidemiology of diphtheria in the great American continent. It is assumed throughout 24 DIPHTHERIA. that croup is chiefly the laryngeal form o£ diphtheria ; but that the death returns from croup include an un- certain proportion of non-diphtheritic laryngitis. Those BUENOS-AYRES. (Population : 1887—437,875 ; 1896—712,095.) Death- Fig. 11. of the curves which give croup separately from diph- theria — and this has been consistently carried out in all cases in which such separate statistics could be NORTH AND SOUTH AMERICA. 25 obtained — clearly indicate that whether the diphtheria curve or the curve for croup and diphtheria combined is surveyed, the conclusions will in nearly every instance be identical. * The mean death-rate for this city is obtained by summing the anmial death-rates and dividing the sum by the number of years. In all the cases not thus particularised, the mean death-rate has been obtained by the more accurate method of adding together the deaths and the populations respectively for the whole series of years, and then ascertaining the death-rate on this aggregate population. 26 DIPHTHEEIA. Certain cautions must be observed in interpreting tbe true significance of the data in the preceding table. Note first, that although for the sake of convenience the years of maximum and minimum mortality from diphtheria and croup have been given and the figures for these years, the course of the epidemics in each city can only be gathered with strict accuracy from the diagrams given in the preceding pages. Note secondly, that the mean death-rates given in the last columns of the table, although they give a rough idea of the relative mortality in each city, do this only with certain reservations, (a) The period of years for which the mean death-rate is taken does not coincide in all instances. (6) Even when an average is struck for the same series of years, there remains the fallacy that one city may have happened to have say three epidemics and another four during the given term of years. It might thus happen, for instance, that if two or three additional years had been included in the series, the place of the two cities would have been reversed as regards their mean death-rate from diphtheria. This fallacy is inherent to all average death-rates for a series of years for any epidemic disease, and is the chief reason why in the present investigation the exact death-rate for eaclh year for each city has been carefully plotted out. By this means the teaching of these death-rates cannot be mis-read. Having made these reservations it is evident that in the American continent, if we may judge from experience of its chief cities, diphtheria is a very fatal endemic disease, which at intervals becomes eindemically and still more fatally prevalent. Some useful indications may be gathered, if the experi- ence of whole States is taken, as may be seen from the accompanying diagram constructed from a table given in NORTH AND SOUTH AMERICA. 27 Dr. Baker's report to the State Board of Health for Michigan, 1892. Slates of ALABAMA. CONNECTICUT. INDIANA. MASSACHUSETTS. MINNESOTA. MICHIGAN. OHIO. NEW JERSEY. NEW YORK rhode;island. province of ontario. .ENGLAND AND WALES. Fig. 12. Fig. 12 has no marginal scale, but even without this, it 28 DIPHTHEEIA. is obvious that the total mass of and the variations in the mortality from diphtheria are enormously greater than in England. When allowance is made for the deficiencies in registration in some of these States, the difference becomes even more remarkable. The curves in the diagram relate to the years 1880-92, that of England being extended to include 1894. The scale in each of the curves is identical : and the striking difference in amount of mortality from diphtheria may be gathered not only by a glance at the curves, but by the fact that, taking Massachusetts and England as examples, the lowest death-rate from diphtheria pZws croup in the former was 53 in 1891, and the highest 134 in 1880; while in England the lowest death-rate from diphtheria plus croup during the same period (1880-92) was 25 in 1880, and the highest 30 in 1887 and 1889, increasing to 39 per 100,000 in 1893. Thus the highest death-rate in England is not much more than half the lowest death-rate in Massachusetts. CHAPTER II. ENGLAND AND WALES. Complete and fairly accurate statistics are available for England and Wales since 1839. In saying this^ tlie fact that a certain but diminishing proportion of deaths from diphtheria have been returned as ulcerated throaty quinsy, croup, laryngitis, or membranous laryngitis, is not ignored. It must be remembered, however, that the increase of accuracy of certification is a gradual one, and thus the resulting transference to diphtheria of a comparatively small number of deaths, will not interfere with the substantial accuracy of any inferences that may be drawn from the English curves now to be considered, tliese inferences being based on the general directions of these curves rather than wpon the magnitude of the rise in any given year. The Registrar General's returns prior to 1860 class diphtheria or angina membranacea with scarlet fever. As, however, for several years prior to this, the former had been distinguished on the ticking sheets in the General Register Office, it is practicable to calculate a death-rate for diphtheria from 1855 onwards in England and Wales, and from 1859 for London, though almost certainly this does not represent all the diphtheria during these years. Judging, however, by the medical writings of the period, there was not, with the possible exception of 1847, a very large admixture of diphtheria and scarlet fever between the commencement of civil registration of deaths in July, 1837j and the great epidemic of diphtheria which invaded 29 30 DIPHTHEEIA. England about 1855-57. The accompanying diagrams for England and for London confirm this view. Croup had been separately classified from 1838 onwards; and so had laryngitis for a considerable portion of the period 1838-95. A glance at Fig. 13 will show that the transference of E"NGLAND AND WALES. 31 laryngitis to diphtheria would not materially affect the character of the curve. Almost the same remarks apply to croup. As diagnosis has improved^ there has been a steady and gradual transference from croup to diphtheria. A comparison of the curve for all England with that for London shows the same general outlines. Diphtheria was epidemically prevalent from 1858 to 1865, and a more recent epidemic reached its culmination in 1893, there having been a steady rise in the death-rate from this disease, especially in London, from 1887 onwards. I had constructed a diagram showing the yearly death- rate from quinsy; but as this does not materially affect the curves, it has been omitted. The average death-rate in England in successive groups of years is as follows -.^ — England and Wales. Annual Death-rate from Diphtheria and Group per million persons living, in groups of years, 1858-94. Period. Diphtheria. Croup. 1 Diphtheria plus Croup. Three years, 1858-60 .... 372-3 274-7 647-0 FiTe years, 1861-65 . 247-6 287-6 535-2 „ „ 1866-70 . 126-8 208-0 334-8 „ „ 1871-75 . 120-8 184-2 305-0 „ „ 1876-80 . 121-8 154-2 276-0 „ ., 1881-85 . 156-2 163-4 319-6 „ 1886-90 . 169-6 125-8 295-4 Pour years, 1891-91 . 251-2 74-0 325-2 It is evident that whether the death-rate from diph- theria alone, or from diphtheria and croup together, be taken as the test, the early epidemic of 1858-65 was more * For remarks on the defects inherent in average death-rates for epidemic diseases, see pp. 1 and 26. 32 DIPHTHBEIA. severe and more fatal than the recent epidemic of 1891-94. Possibly, however, the full brunt of the recent epidemic has not yet been experienced in England as a whole, some districts not having been hitherto reached by the slowly travelling infection. In London, the experience is somewhat different. Death-rate in London from Diphtheria per million living. Period. DipMlieria. Two years, 1859-60 458 Five „ 1861-65 224 1866-70 133 1871-75 114 1876-80 130 1881-85 221 1886-90 298 Year 1891 840 „ 1892 462 „ 1893 760 „ 1894 625 „ 1895 529 „ 1896 606 „ 1897 507 Here it is evident that the later has been more severe and therefore presumably more extensive than the earlier epidemic. There are no actual data enabling one to make a comparison of cases as distinguished from deaths in the two great epidemics. The history of the two great epidemics of diphtheria (1857-65 and 1888 onwards) will be considered later in special chapters (pp. 49 and 55). Place. Years of observation. Diphtheria alone. Death-rate per 100,000. Highest. Lowest. England and Wales. . London 1859-94 / 52 in \ 1859 (76 in 1 1893 9-4 in 1872 8 in 1872 ENGLAND AND WALES. 33 It will be convenient next to take tlie history of diph- theria in various English towns, so far as their records can be obtained, and from these to obtain a bird's-eye view of the progression of the disease in different parts of the country, similar to that attempted for the vaster territories embraced in the United States. SouTHEEN Districts. Hastings and Brighton had a small epidemic of diph- theria in 1883-4-5. In 1889-91 Eastbourne suffered from a severe epidemic, which merged into the earlier epidemic already mentioned. Hastings appeared to suffer, though to a less extent and a year later, from the same outbreak; and Brighton still later and to a less ex- tent. Ckoydon showed an epidemic culminating in 1877, possibly following on the slight rise shown in London in 1875. Both Brighton and Hastings had small epidemics, 1874-6 j Portsmouth suffering to a smaller extent in the same year. The more recent course of diphtheria in Croydon corresponds very closely with that of Lon- don. Portsmouth had no marked excess of diphtheria in 1860-62, and in fact no great epidemic until 1881-2, when it suffered very severely. Smaller outbreaks reached their maxima in 1886 and 1890, and since that year diph- theria has declined. Southampton, during the periods for which records are obtainable (1873-96), has enjoyed a continuous com- parative immunity from diphtheria. i> 2 Death- rate per ; 1000. •80 ' 3 1 1 1 1 1 •60 -i F^ •40 ' 1 1 1 •20 ; 1- ^ Vi ._ l.n . ^ 1 MMgrn 1 2-00 ; 1 '80 : ■ •60 i ; _, •40 ! ; ■20 ; i 1^00 •80 '. ; •60 : ■40 1 •20 : Ji H . r il il •20 -j ; 1 ■° <™ •60 1 •40 •20 k. r 1 L i i HASTINGS. (Population : 1871—28,291 : 1891—52,223.) EASTBOURNE. (Population : 1881—22,014 ; 1891—34,969.) BRIGHTON. (Population : 1871—90,345 ; 1891—115,873.) CROYDON. (Population : 1871—55,652 1891—102,697.) Fiff. 14. ENGLAND AND WALES. 35 Diphtheria plus Diphtheria alone. Mean death- Croup. Death-rate Death-rate per rate for entire Years of observa- per 100,000. 100,000. period. tion. Highest. Lowest. Highest. Lowest. Diph- theria plus Croup. Diph- theria alone. Hastings . . 1874-96 — (60 in "( 1892 7iil in\ 1881/ — 18 Eastbourne . 1883-96 — — /229 in \1890 15 in ) 1894 j" — 42 Brighton . . 1870-96 — — (30 in ( 1893 3 in ) 1879 i" — 13 Portsmouth . 1861-96 — — J 159 in t 1881 4 in ) 1867 / — 23 Southampton 1883-96 ( 37 in '( 1889 6 in 1893 16 in 1888 lin \ 1892 / 18 6 Croydon . . 1875-96 f (95 in "( 1893) (8 in 1882) 83 in 1893 lin \ 1875/ — 29 (See remarks on page 26.) Western Disteicts. Bristol has for a long period suffered very little from diphtheria. The early epidemic appears to have reached it in 1862j and continued, though never producing a high death-rate, for several years. For twenty years there was then very little diphtheria, and the recent epidemic, which commenced in 1891, has hitherto affected the city only to a minor extent (Fig. 15). The curve for Caedipe is most irregular. The popula- tion of this town has quadrupled in the period embraced by the curve; and it appears likely that croup in these returns is not so closely related to or identical with diphtheria as in most other returns. The chief maxima occurred in 1884-87 and in 1892 (Fig. 15). West Central and Midland Districts. Birmingham had a small outbreak of diphtheria in 1871-72. The disease has remained endemic as in other towns of England, there being a steady mortality from it 36 DIPHTHERIA, - ' ' 1 ' •60 i ' 1 J ' 1 •40 ' : : J n K\^ •20 ' M' U [t ■f jntLi ^H^I^^^^^^^^hH^^I F ig. 15. PORTS- MOUTH. (Population : 1861—95,220 ; 1891—159,285.) SOUTH- AMPTON. (Population : 1881—60,051 ; 1801—73,384.) BRISTOL. (Population : 1860—181,799 ; 1896—230,623.) CARDIFF. (Population : 1861—32,054 ; 1891—130,283.) ENGLAND AND WALES. 37 on a small scale. In 1895-96 Birmingham suffered severely from epidemic diphtheria. WoLVEKHAMPTON had an epidemic of diphtheria culmin- ating in 1885, and a much more severe epidemic in 1894-96; which may be connected with the epidemic starting a year later in Birmingham. Death- rate per 1000. 3 Fiff. 16. Leicestee, as will be seen from the curve (Fig. 16) , has suffered very little from diphtheria, the intermittent rises being of very small calibre. There is an appearance of an epidemic on a somewhat larger scale in 1896, the issue of which cannot yet be stated.* * In 1897 the death-rate from diphtlieria had increased from 31 in the preceding year to 36 per 100,000. 38 DIPHTHERIA. Place. Years of observa- tion. Diphtheria plus Croup. Death-rate per 100,000. Diphtheria alone. Death-rate per 100,000. Mean death-rate for the entire period. Highest. Lowest. Highest. Lowest. Diphtheria lihis Croup. Diphtheria alone. Bristol . . . Cardiff . . . Birmingham . Wolverhampton Leicester . . . 1860-96 1865-96 1872-96 1874-96 1856-96 f 122 in 1 1895 12 in 1877 1 26 in ■( 1863 i 64 in 1 1893 ( 37 in ] 1895 98 in 1895 / 31 in i 1896 2 in ) 1880 / in \ 1867 / 9 in ) 1891 / 1 in ) 1882 & 1- 1877 ) 2 in \ 1872 J 35 11 22 16 9 (See remarks on page 26.) Lancashire. This great industrial centre has a large number of towns within a short distance of each other, and a study of the prevalence and order of prevalence of diphtheria in these presents features of considerable interest. Liverpool suffered somewhat severely in the 1860-65 epidemic, about to the same extent as London, more so if the portions of the curves relating to croup are included. Since then there is evidence of numerous small ex- acerbations of the disease, the greatest and most per- sistent being in 1884-87. Liverpool has hitherto suffered but little in the great English epidemic of the present decennium. St. Helens, near Liverpool, has suffered much more than the latter from diphtheria. The excess in Liverpool in 1884-87 is represented by a larger excess in St. Helens in 1888-89; and by the still larger excess in Oldham in 1886-88. "We may next consider the curves for the twin cities of Manchestkr and Salford. It is easy to trace a gradation in time of incidence of the epidemics in Lancashire towns. Thus — MANCHESTER.! (Population : 1871—351,189 ; 1891—505,343.) SALFORD. [(Population : 1861-85,108 ; 1891-198,136.) OLDHAM. (Population : 1881—111,343 ; 1891-131,463.) ST HELENS. (Population : 1871—45,134 ; 1891—71,000.) LIVERPOOL. (Population : 1861^43,938 ; 1891—517,951.) Fig. 17. 39 40 DIPHTHEEIA. (a) The epidemic in Liverpool in 1884-87 caused a death-rate which only reached 25 per 100,000 in the maximum year 1885. {b) The epidemic in Oldham in 1886-88 caused a death- rate of 50 per 100,000 in the maximum year 1887. (c) The epidemic in Manchester in 1888-90 caused a death-rate of 51 per 100,000 in the maximum year 1889. {d) The epidemic in Salford in 1889-90 caused a death- rate of 100 and 105 per 100,000 in these two years re- spectively. There appears to be evidence, therefore, not only of gradual spread of the infection, but also of increasing virulence, as it passed through a population housed under conditions which are frequently the reverse of favourable. The Manchestee epidemic, culminating in 1889;, has been more persistent than the Salford epidemic, which culminated in a very high mortality in 1889-90, and appears to have ''burnt itself out " with great rapidity. Salford, like Liverpool, suffered severely in the epi- demic 1859-64. Place. Years of Observa- tion. ^'P^rS^^''^ Diphtheria alone. Death-?afeper i ^^^*f;;^n^n*' ^'^ 100,000. iuu,uuu. Mean Death- rate for the entire period. Highest. Lowest. Highest. Lowest. Diph- theria plus Croup. Diph- theria alone. Liverpool St. Helens . Oldham . . Manchester . Salford . . 1860-96 1870-95 1877-95 1871-96 1860-95 /87 in \ 1864 f 131 in ■\ 1889 17 in 1893 24 in 1886 26 in 1865 /42in \ 1889 (50 in \ 1887 (51 m \. 1889 106 in 1889 5 in\ 1875/ in) 1871 / 8 in) 1883 & [ 1884 j 4 in\ 1871/ 6 in) 1861 & [ 1873 j 48 67 14 18 17 18 32 (See remarks, page 26). ENGLAND AND WALES. 41 YOEKSHIEE. The Shepeield returns for diphtheria go back as far as 1859, and show a large epidemic of diphtheria in that year. After a remission in I860, a second maximum was reached in 1862. This was followed by a gradual re- mission of the disease, which reached its minimum in 1884. In 1885 an excess of diphtheria on a small scale commenced, which still continues. The comparison of the early part of the Sheffield diagram with the corresponding parts for Liverpool and Salford shows that the remission of the early epidemic which occurred in 1860 in Sheffield, occurred in 1861 in Salford (Fig. 18). Leeds and Beadpord are neighbouring towns in the West Hiding, as are also Halifax and Huddeespield. Hull, on the east coast, is comparatively remote from the other Yorkshire towns. We may contrast the very small epidemic in Hull in 1883-84 with the slightly greater epidemic in Leeds in the same year, the still slightly greater epidemic in Halifax in 1886, the much more severe epidemic in Huddersfield in 1887, and the more protracted epidemic in Sheffield, which culminated in 1889-92. Bradford appears to have almost entirely escaped any epidemic prevalence of diphtheria during 1881-90, or in any other year. Leeds and Hull come next in order of relative exemption, while Sheffield shows a considerable amount of endemic diphtheria. Diphtheria. Years of Death-rate per 100,000. Mean death-rate Place. for the entire period from Diphtheria. Highest. Lowest. Sheffield . . 1859-96 99 in 1859 3 in 1886 15 Leeds . . . 1872-96 19 in 1894 8 in 1887 10* Bradford . . 1871-96 14 in 1878 1 in 1873 7* Halifax . . 1872-96 25 in 1893 3 in 1884 12* Huddersfield 1879-95 46 in 1887 3 in 1893 12* Hull ... 1870-96 23 in 1896 2 in 1896 7* See foot note on p. 25. 42 DIPHTHERIA. HULL. (Population : 1871—122,226; , 1891—219,812.) BRADFORD, (Population : 1871—146,987 ; 1891—216,938.) HUDDERS- FIELD. (Population : 1879—80,248 ; 1894—98,511.) HALIFAX. (Population : 1896—94,755.) LEEDS. (Population : 1873—272,619 ; 1896—347,278.) SHEFFIELD. (Population : 1861—186,375 ; 1891—325,304.) Fiff. 18. NOETHUMBE RLAKD. The curves for NEWCASiLE-rpoif-TYNE and South Shields resemble eacli other in shape and time incidence. There has been no great epidemic ; but it is evident that the amount of diphtheria increased in both towns from ENGLAND AND WALES. 43 the middle of the last decennium onwards, and that these towns now show a declining amount of diphtheria, with- out the amount ever having been alarming. Newcastle has a larger amount endemically present than South Shields. Place. Years of observation. Dipht Death-rate Highest. tieria. per 100,000. Lowest. Mean death-rate per 100,000 for the entire period. Newcastle- upon-Tyne . South Shields 1870-96 1871-95 27 in 1890 29 in 1891 3 in 1879 2 in 1893 12 11 Death- <-, rate per t*^ 1000. S SOUTH SHIELDS. (Population : 1871—45,600 ; 1891—79,067.) NEWCASTLE-UPON- TYNE. (Population : 1871—136,293 ; 1891—187,502.) Fig. 19. CHAPTER III. SCOTLAND AND IRELAND. A GLANCE at the Scottisli returns (Fig. 20 and 21) shows some striking contrasts witli those for English towns. There is a greater mass of diphtheria in the Scotch towns than in the English. The disease is more distinctively and more extensively endemic than in England. The average death-rate from diphtheria for the entire period 1858-95 is much higher than that of most English towns. This is especially true for Glasgow, Greenock, and Edinburgh. All the Scottish cities whose records ex- tend sufficiently far back, appear to have suffered from the early epidemic from 1859 onwards. This reached Aber- deen in 1859, Greenock in 1860, and only reached its maximum in the neighbouring city of Glasgow in 1862-63. Perth suffered to an exceptional extent in 1862; while Edinburgh, like Glasgow, had its maximum incidence in 1863. Edinburgh had an epidemic in 1870-71, while a similar epidemic culminated in Greenock and Glasgow in 1873, in Dundee in 1872-73, in Aberdeen in 1872, and in Perth not until 1876. Most of the Scotch towns have hitherto not suffered, or but little, from the epidemic, which has been so widespread in England from 1888 onwards. The chief exception to this rule is Kilmarnock. SCOTLAXD AND lEELAXD. KILMARNOCK. (Population : 1881—25,864 ; 1391—27,959.) GLASGOW. :'b' (Population: _ 1 1861—397,673 ; ' 1891—567,143.) PAISLEY. (Population : 1881—55,638 ; 1891—66,427.) DUNDEE. (Population : 1881—142,455 ; 1891—155,640.) ABERDEEN. (Population : 1881—105,076 ; 1891—121,905.) Fig. 20. 46 DIPHTHERIA. Death- late per 1000. s ■80 2, o a 1880 i i TMAri/ •60 i. i N ^RL •40 UKLt-iivwrx ■20 L .M •0 i •60 EDINBURGH •40 1 •20 M ^ .i •0 i Ml ■20 l^OO PE RTH ■80 •60 •40 L ▲ . 1 20. 0. A iii y (Population : 1881—69,238 ; 1891—63,498.) (Population : 1881—230,402 ; 1891— 261,*261.) (Population : 1881—29,756 ; 1891—30,760. ) Fig. 21. Note. — The Scottish and Irish curves terminate in angular points instead of in horizontal lines. These diagrams had been completed before the columnar shape of diagrams -^vas finally chosen, but it "vvas not thought necessary to reconstruct the diagrams already comjDleted. SCOTLAIJD AND IRELAND, 47 Place. Years of observation. Diphtheria. Death-rate per 100,000. Highest. Lowest. Kilmarnock Paisley . Glasgow . Greenock Edinburgh Perth Dundee . Aberdeen 1873-94 1859-94 1859-94 1859-94 1859-94 1859-94 1859-94 1859-94 80 in 1892 39 in 1888 & 1898 60 in 1863 87 in 1860 80 in 1871 266 in 1863 85 in 1872 75 in 1864 7 in 1889 2 in 1860 3 in 1860 15 in 1876 6 in 1861 in 1859 9 in 1870 8 in 1877 & 79 DUBLIN. (349,594.) BELFAST. (273,277.) CORK. (7.5,345.) LIMERICK. (37,155.) LONDONDERRY. (33,200.) WATERFORD. (80,852.) Fig. 22. Note. — The populations given within the brackets are those of the census enumeration, 1891. The scale of the diagram is the same as that of the preceding diagrams. Ieeland. Ireland is a sparsely populated country; and on the hypothesis that diphtheria is conveyed chiefly by personal infection, it is not surprising to find that its amount in DIPHTHERIA. Ireland is small. Since 1881, when returns became first available, there has been a small epidemic in Londonderry in 1882, and, in Limerick in 1883. Londonderry had a rather larger epidemic in 1889, followed by a still higher rise, after an interval of two years, in 1883. Dublin has throughout the whole period (1881-95) remained remark- ably free from this disease. The Belfast curve shows a larger mass of the disease than any other Irish town, though less than most Scotch and many English towns. Diphtheria. Place. Years of Death-rate per 100,000. Highest. Lowest. Dublin .... 1881-95 10 in 1885 2 in 1889 Belfast .... 1881-95 24 in 1888 & 1894 6 in 1883 Cork 1881-95 23 in 1888 2 in 1887 & 1895 Limerick . . . 1881-95 29 in 1883 3 several years Londonderry . . 1881-95 51 in 1893 3 in 1883 Water ford . . . 1881-95 15 in 1893 several years CHAPTER IV. THE HISTORY OP THE GREAT EPIDEMIC OP DIPHTHERIA IN LONDON IN THE FIFTH AND SIXTH DECADE OP THE NINE- TEENTH CENTURY. In 1847 the London curve (Fig 13) shows a striking rise in the death-rate from croup, it being 154 per cent, higher than in the preceding year. Does this mean an increase of laryngeal diphtheria, or of some other disease mas- querading under the name of croup ? The curve for the whole of England does not help, as the records for 1847 are absent. The historical summary (page 116) shows that diphtheria was prevalent in London and elsewhere in 1848-49. The sparsity of recorded cases by no means proves that this unrecognised disease was not prevalent in 1847. On the other hand, 1847 was a great influenza year, and it may be that the sudden rise in deaths from croup was really due to inflammatory laryngitis secondary to influenza. Dr. Peacock^s historical account,* however, makes it prob- able that there was considerable true diphtheria. He says : " There was also, during the period of the influenza, a great increase in the number of deaths from the various descriptions of eruptive fever, small-pox, measles, and scarlet fever, as likewise from hooping cough, etc. . . with affections of the fauces and larynx'' (p. 353). On a later page (p. 366) Dr. Peacock says: "From the tables of mortality, it will be seen that the deaths from croup, quinsy and laryngitis were during the epidemic much * " Influenza : an History Survey," Dr. Symes Thompson, 1890, p. 355. 49 ™ 60 DIPHTHEEIA. more numerous than usual." We are, therefore, I think justified in regarding 1847 as a year of epidemicity of diphtheria. It is not without significance that the present epidemic of diphtheria in London has coincided in part with the years of reappearance of epidemic influenza. Until 1859, the metropolitan death-statistics do not tabu- late diphtheria separately from scarlet fever. So far as official returns are concerned, it is therefore impracticable to state the amount of diphtheria, or to form any idea as to the occurrence of epidemic peaks of this disease prior to that year. A careful study of (a) the figures for the whole of England and Wales, and (6) the figures for London for diseases closely allied to, and liable to be con- founded with diphtheria, partially serves to fill up the hiatus in our knowledge. The curve for England and Wales (Fig. 13) appears to give a result more nearly approximating the truth for London, than the metropolitan curve in the same illustra- tion. This curve indicates that in 1858-59 diphtheria was already epidemically present in a most fatal form, although it is fairly certain that the returns for the whole of England do not tell the whole truth for the early years 1855-57. That this is so, is indicated by the following official figures, which serve to show the progress of the English epidemic in the early years, and to give some idea of the extent of transference of deaths by alterations of nomenclature. Year. Deaths in England and Wales from Cynanche Maligna. Diphtheria. Quinsy. Scarlet Fever. 1855 1856 1857 1858 199 374 1,273 1,770 186 229 310 ■ 4,836 371 413 485 623 17,314 14,160 14,229 30,317 The difficulty is increased by the fact that scarlet fever THE EPIDEMIC OF DIPHTHERIA IN LONDON, 1857-65. 51 in a very fatal form was epidemic during tlie same years, and mucli of the diphtheria was regarded by practitioners, in whose experience diphtheria was a totally new disease, as scarlatina anginosa. A careful study of the m'etropolitan figures themselves shows that the epidemic of diphtheria was present before 1859, the iirst year shown on the curve in Fig. 13. In the following table I have calculated the annual death-rate from diphtheria, croup, quinsy, and laryngitis in London from 1839 to li Year. Annual Deatli-rate per 1,000 living from Diphtheria. Croup. •18 Laryngitis. Quinsy. 1839 No entry •01 •05 1840 )) •21 •01 ■04 1841 I) 21 •01 •04 1842 J7 24 •01 •04 1843 :j 11 •01 •03 1844 ',^ 20 •03 •05 1845 ;? 17 •04 •03 1846 77 13 ■10 •04 1847 77 •33 •10 •04 1848 77 13 •08 •04 1849 77 14 •08 •04 1850 7? 13 •08 •03 1851 13 •08 •03 1852 77 14 •08 •02 1853 77 15 •10 •02 1854 77 19 •13 •03 1855 77 21 •11 •04 ■ 1856- 77 •22 •11 •03 1857 77 17 •12 •03 1858 21 •12 •04 1859 •284 15 •10 •03 1860 •174 17 •10 •02 1861 •239 30 ■14 •04 1862 ■255 33 •15 •03 1863 •275 32 •13 •03 1864 •207 30 •13 •03 1865 •144 25 •11 •02 1866 •152 23 •10 •02 1867 •145 24 •11 •01 1868 ■158 23 •11 •01 1869 •107 •28 •13 •01 It will be seen that in 1855 the death-rate from croup 52 DIPHTHERIA. was 53 per cent, and in 1855 it was 60 per cent., above the mean death-rate from this disease in the years 1848-53. A similar increase in laryngitis occurred. Furthermore, the recorded death-rate from diphtheria in 1859 was higher than in any of the epidemic years 1861-64, when the disease had become more generally recognised as such. It is highly improbable, therefore, that the high death-rate of 1859 represented a newly imported disease. We may conclude, with much more probability, that it had been epidemic for a year or two previously, although imperfectly recognised. Collateral evidence points in the same direction. Thus in the Weekly Return of Diseases in the Metropolis for the week ending Saturday, October 31st, 1857, the following remarks occur : * '^ In future numbers of the WeeMy Return a separate column will be allotted to new cases of diphtherite. ... At the present time diphtherite threatens to demand more attention in this country ; for apparently it is to its prevalence, as a cause of death in several districts, that various registrars' notes in the late quarterly report of the Registrar- General refer. . . . There is reason to believe that cases are also occurring within the limits of the metropolis ; sometimes recognised in their true nature, and called either diphtherite or Boulogne fever ; sometimes, perhaps, mistaken for true croup, or for the throat-inflammation of scarlet fever, or for some other disease." Dr. E. Greenhow's work ''On Diphtheria "t contains confirmatory evidence. He states (p. 72) : " Sore throats, which were remarkably frequent during the years 1857, 1858, and 1859, continue to prevail t (March I860)." He * Quoted by Mr. (now Sir John) Simon, in his Second Report to the Privy Council, 1859. See p. 13, vol. ii.. Public Health Reports, by J. Simon, edited by Dr. Seaton. t Jno. W. Parker & Son, Strand, 1860. J i.e. in London. THE EPIDEMIC OP DIPHTHEPJA IN LONPON, 1857-65. 53 also describes a number of cases occurring during these years, whicli were obviously true diphtheria. On page 80 of the same work, he remarks : " In a sporadic form, or in the form of very small groups confined to a limited district, diphtheria has probably never been absent from this country. '' The value of this remark is somewhat diminished by the following remark, with which it is associated : " In this respect it resembles cholera, the other novel epidemic of the present century.^' Mr. Netten Eadcliffe in a paper, " On the Recent Epidemic of Diphtheria" (vol. i., part iii., p. 328, Transactions of the Epidemiological Society), says : " The order of epidemic manifestation of the disease, in point of time, in the different districts of the kingdom, after 1855, was, — In 1856 (1) The West-Midland Counties. (2) The Eastern Counties. (3) The South-Eastern Counties. (4) The North-Midland Counties. la 1857 (5) The South- Midland. (6) The North- Western. (7) Yorkshire. (8) The Metropolis. In 1858 (9) The South- Western Counties. In 1859 (10) The Northern Counties, Monmouth, and Wales." The following further sentences may be quoted from Mr. Ratcliffe's important paper {op. cit. p. 334), as they throw considerable incidental light upon diphtheria in London. "If the mortality returns for the years 1840-59 be examined, it is found that scarlet fever underwent a prodigious increase in 1858, and prevailed in that year to a greater extent than in any previous year of the nineteen. 54 DIPHTHERIA. (2) That the mortality from croup advanced year by year from 1854 ; that^ in fact^ the disease was epidemic in 1856, '57, '58, and '59, the epidemic culminating- in 1858. The mortality from the disease also was prodigiously above the average of preceding years ; increasing from 3,660 in 1853 to 6,220 in 1858. (3) That the mortality from thrush also was greatly in excess in 1858 and 1859, though not to the same extent as in 1848 and 1852. (4) That the mortality from quinsy was in excess in 1857 and 1858, in the latter year attaining a higher point than in any previous year. (5) That the mortality from noma had undergone a remarkable increase in 1855, '56, '57, and '59, culminating in 1857. (6) That the mortality from laryngitis had undergone a steady development from 762 in 1847, to no less an extent than 1,439 in 1858. In fact, it is not too much to say that all the affections allied to diphtheria prevailed epidemically contemporaneously with diphtheria." If in the last sentence, we substitute for " affections allied to diphtheria," the alternative words, *^ affections with which diphtheria is liable to be confused," we are justified in the conclusion that in London diphtheria was steadily increasing in amount from 1854 onwards, being widely epidemic early in 1857. The question of transference, since 1881, to diphtheria of deaths which were formerly ascribed to other throat affections, is discussed in an interesting paper by Dr. J. F. J. Sykes.^ * Public Health, vol. vi., p. 331,. CHAPTER V. THE HISTOEY OP THE EPIDEMIC OF DIPHTHERIA IN LONDON, 1888-96, AND ITS BEARING UPON THE METHOD OP SPREAD OF THE DISEASE. The area of London (Registration area) is 121 square miles. At the census of 1891, there were within this area 544,947 inhabited houses, containing an average of 7" 7 persons to a house. Its population at the middle of 1897 is estimated to have increased to 4,463,169, the enumerated population in April, 1891 being 4,211,743. A consideration of the mode of spread of diphtheria in this vast province of houses, as evidenced by official statistics for the several districts composing the metropolis, must throw light on the general problem of the mode of progression of this disease. A glance at Fig. 13 shows that in London as a whole, diphtheria had been increasing fairly steadily, the total volume of the disease (as indicated by deaths) increasing year by year with very slight interruptions until the maximum was reached in the year 1893. The annual summary of the English Registrar-General for 1895 contains, on page 9, a table giving the deaths from diphtheria in the Metropolitan Sanitary Area in the nine years 1887 to 1895, after distribution of deaths occurring in public institutions. From this table, supplemented by later returns for 1896, the death-rates per 100,000 of popu- lation have been calculated and plotted out on the diagrams in Figs. 23 and 24. It must be noted that the census population is assumed in each case to hold good throughout the ten years. This will have produced some exaggeration 55 56 DIPHTHERIA. of the latter half of tlie curves^ but even assuming tliat the exaggeration is not to the same extent in the different diagrams, this is of little consequence, as it is the directions of the curves and their general shape with which we are concerned rather than the actual height in any given year. The western parish of Faddington had a first epidemic in 1888, a second in 1894. In the next parish, Kensington, the first epidemic did not reach its maximum until the epidemic was subsiding'^ in Paddington. Hammersmith., further west, lagged two years behind Paddington and Kensington, and at the end of the curve had not ex- perienced an epidemic '^relapse," while Fulliam, to the south-west,* escaped for several years, did not suffer severely until 1893, still more severely in 1894. Returning eastwards along the north bank of the Thames, Chelsea was a year behind Kensington, two behind Paddington. It had a smart second epidemic in 1896.t St. George's, Hanover Square, abutting on Padding- ton, had its maximum in the same year as the latter, as did also the Westminster parishes, though somewhat irregularly. Proceeding northward, St. Marylehone, the next parish on the east of Paddington, scarcely experienced the iirst J epidemic; but the second J culminated in 1893, a year earlier than in Paddington. * The significance of the curves is better understood if they are stiadied in conjunction with a district map of London. t The rationale of the two epidemics of diphtheria in Chelsea is explained by Dr. Louis Parkes, in an interesting paper {Epi- demiological Society^s Transactions, vol. xvi.). The epidemic of 1890 occurred in its outlying district of Kensal Town, allied geographically to Kensington and Paddington, the late epidemic in the home district. X I am obliged to speak of the "first" and the "second" epi- demics. They are really part of the one great London epidemic, which did not exhaust itself in a single onslaught. THE EPIDEMIC OP DIPHTHEEIA IN LONDON^ -96. 57 PADDINGTON. (117,846.) KENSINGTON. (166,308.) HAMMERSMITH. (97,239.) FULHAM. (91,639.) CHELSEA. (96,253.) ST. GEORGE, HANOVER SQUARE. (78,364.) ST. MARGARET, ST. JOHN, and ST. JAMES, WESTMINSTER. (80,769.) ST. MARYLEBONE. (142,404.) HAMPSTEAD. (68,416.) ST. PANCRAS. (234,379.) ISLINGTON. (319,413.) STOKE NEWINGTON. (229,542.) ST. GILES, ST. MARTIN-IN-THE- FIELDS, and STRAND. (79,615.) Fig. 23. (The figures in brackets are the population of each district or parish at the census in 1891.) 68 DIPHTHERIA. The experience of Hampstead and of 8t. Pancras is identical with that of St. Marylebone^ except that the first epidemic appeared distinctly in their experience. Islington, to to the north-east of St. Pancras, lagged a year behind in its epidemic experience, while Stoke Newington, like St. Pancras, had its maximum in 1893 ; so also did the more central district of the Strand. Coming still nearer the centre of London, and travel- ling thence eastward, we find that Holborn and the Gity escaped the first epidemic, the City maximum occurring in 1892, the Holborn maximum in 1893. Shoreditch, to the north-east of the City, had the same time-incidence of the disease as Holborn and Clerken- well abutting on it. Bethnal Green had its first epidemic, culminating in 1890, and a much larger epidemic "relapse," culminating in 1892-93. Whitechajpel had a somewhat similar ex- perience, as also Limehouse and Poplar, as regards the 1893 maximum, but with a varying incidence of the first epidemic. Crossing the river Thames, the Southivarh parishes and Newington show a first epidemic in 1888 or 1889, and a later epidemic with a maximum in 1893-94. Bermondsey had a similar experience, and so also had Lambeth. In Battersea and Wandsworth there was comparatively little diphtheria before the sudden maximum in 1893. Gamherioell was a year behind Wandsworth; and the outlying districts of Greenivich and Lee, although com- mencing their epidemics in 1893, only reached the maximum in 1895. The experience of Lewisham was very similar to that of Wandsworth, but in both these districts there was a second exacerbation of the disease in 1896. The curves thus briefly described are of great interest. HOLBORN. (99,480.) CITY OF LONDON. (38,320.) SHOREDITCH. (124,009.) BETHNAL GREEN. (129,132.) WHITECHAPEL and ST. GEORGE'S IN THE EAST. (120,257.) MILE END. (164,96i5.) POPLAR. (166,748.) ST. SAVIOUR, ST. GEORGE THE MAR- TYR, and ST. OLAVE, SOUTHWARK. (99,612.) NEWINGTON. (115,804.) BERMONDSEY and ROTHERHITHE. (123,937.) LAMBETH. (275,203.) BATTERSEA and WANDSWORTH. (307,500.) CAMBERWELL. (285,344.) GREENWICH and LEE. (201,516.) LEWISHAM, WOOLWICH, and PLUM- STEAD. (165,556.) Fig. 24. 60 DIPHTHERIA. They clearly show that while one part of the metropolis was suffering severely from diphtheria^ others were relatively exempt ; and that the latter in their turn suffered from the same plague. This is seen over and over again in contiguous parishes, and it is impossible to resist the conclusion that in diphtheria we have to deal with a disease which creeps slowly from place to place, in which months or even several years may elapse before it takes firm root and begins actively to propagate itself. It cannot be imagined that climatic differences between the various parishes of London can account for the erratic time-incidence, of the outbreak of diphtheria in these parishes. The different parts of London differ but slightly in amount of sunshine, rainfall, temperature, or humidity of air. We are hemmed in to the conclu- sion that the differences of time-incidence of the epidemic in the various parishes of London are determined chiefly, if not solely, by the opportunities for free personal inter- communication and infection. This does not exclude the possible operation of factors favouring in certain districts an exceptionally severe development of the disease, or permitting its retention in the district for an exception- ally long period. Nor, on the other hand, does it exclude the operation of some wider causes which, co-operating with personal infection, provoked the great London epidemic from 1888 onwards. A further point is clear from the above curves. In the first outburst of the disease it does not exhaust its local possibilities. A second outbreak is shown in at least fifteen out of the twenty-eight curves contained in Figs. 23 and 24. It is instructive to compare in this respect the curves for large towns remote from other towns, in which such epidemic " relapses " are not so usual. That they should occur in the different THE EPIDEMIC OF DIPHTHERIA IN LONDON, 1888-96. 61 parislies of London is not surprising, when we remember that each parish is a part of the great province of London, and that when the disease has spent itself at a particular focus, there are abundant opportunities for its rejuvenescence by fresh importation from foci in other districts. In the present chapter diphtheria in London has been discussed as though personal infection were the sole factor concerned in its spread. But why should personal infec- tion be so potent in producing an epidemic in certain years, while at other periods, with the same opportunities of personal infection, no epidemic is caused ? The dis- cussion of this wider problem is contained in Chapters XVI and XYII. CHAPTER VI. FEANCB, ITALY, HOLLAND, AND BELGIUM. Paris. — The vital statistics for Paris only extend back to 1865, while those for provincial towns of France are only available since 1886. We find from an historical study of diphtheria, sum- marised on pages 110 to 126, that Paris has during the last two centuries been a favourite home of this disease. It is not surprising, therefore, that the carve for this city since 1865 shows a large amount of diphtheria en- demically present. The French and Italian returns here- after given, and many of the German and other con- tinental returns, do not differentiate between diphtheria and croup, but class them together as one disease. Dr. J. Bertillon, the head of the Municipal Statistical Office in Paris, in a letter to the author, dated January 30th, 1897, says : "Nous, n'avons jamais fait en France de distinction entre la diphtherie et le croup, que nous regardons comme deux formes de la meme maladie." The Parisian curve shows a steadily increasing preva- lence of diphtheria plus croup from 1872 * until the acme is attained in 1877, then occurs a slow and fairly steady decline to a minimum in 1895-96. The contrast with the London curve is sufficiently striking. "When the London curve was at its minimum, that for Paris was very high ; and the London curve rose as that for Paris * There is some doubt as to the accuracy of this portion of the curve, the number living in Paris during the Franco-German war, and particularly during the siege of Paris and the reign of the Commvine, being doubtful. 62 Death- rate per 1000. 1-00 PARIS. (Population : 1895—2,424,725.) BORDEAUX. (Population: 1888—237,073.) VV;-nN LYONS. \^\i\\h- (Population: 1888—400,410.) \ MARSEILLES. (Population : 1888—376,143.) 64 DIPHTHERIA. fell. With regard to the minimum in 1895-96, we have to note the disturbing influence on mortality returns of the new treatment of diphtheria by antitoxic serum ; and on this point I may again quote a remark made by M. Bertillon in the letter above-mentioned : " Vous re- marquerez sans doute combien la diphtheric est devenue rarement mortelle en 1895 et 1896. Cela est du sans aucun doute a I'usage tres repandu de serum de I'Institut Pasteur." I am, however, strongly of opinion that although a portion of the decreased mortality from diphtheria in Paris is almost certainly due to the general employment of the antitoxin treatment, another influence is also at work, viz., the ebb of the tide of epidemic prevalence. The antitoxin treatment was described by M. Roux at the Buda-Pesth International Congress of Hygenie and Demography, in September, 1894, and had not up to that time been used on a very extensive scale. A glance at the Paris curve, however, will show that during the four preceding years the death-rate from diphtheria had very materially declined. The question might be further elucidated by a comparison between the number of cases of diphtheria and the number of deaths from this disease, if this were available. The above remarks as to the disturbing influence of the antitoxin treatment apply also to the statistics of other cities. The three provincial cities, Bordeaux, Lyons, and Marseilles, all show a similar decline in mortality from 1891 onwards. There are, however, some minor differences. Bordeaux had the apex of its curve in 1888, Lyons in 1890, and Mar- seilles in 1890-91, the two latter much behind the Paris curve. FRANCE^ ITALYj HOLLAND, AND BELGIUM. 65 Place. Years of observa- tion. Diphtheria phis Croup. Death-rate per 100,000. Diphtheria plus Croup. Mean death-rate per 100,000 for the entire period. Highest. Lowest. Paris . . . Bordeaux . . Lyons . . . Marseilles . . 1865-96 1886-96 1886-96 1886-96 121 in 1877 78 ., 1888 95 „ 1890 198 „ 1891 17-5 in 1895-6 4 „ 1896 12 „ 1896 28 „ 1895 63* 37 45 113 * Obtained by the less accurate but only available method of adding together the annual death-rates and dividing by the number of years. SWITZE ELAND. The returns from Geneva and Lausanne show a very considerable amount of diphtheria. The first epidemic shown on the curves reached its maximum in 1881 in Geneva, in 1883 in Lausanne. This may be compared with the 1881 maximum in Turin. The more recent epidemics in the two Swiss towns were also not simul- taneous in time. Spain aub Portugal. In Barcelona, in 1881, 122 deaths from diphtheria plus croup were recorded, giving a death-rate of 43 per 100,000. In Lisbon, in 1882 and 1883, 141 and 104 deaths from diphtheria were recorded according to the annual sum- mary of the Hegistrar-Gi-eneral, giving a death-rate of 70 and 51 per 100,000 respectively. From theoflScial weekly records of mortality in Lisbon, for the five years 1892-96, received by the author, it appears that diphtheria plus croup caused 44 deaths in 1892, 50 in 1893, 72 in 1894, 89 in 1895, and 51 in 1896. The census population in 1878 was 242,297; in 1890, 301,206. If we assume a uniform population of 305,000 for the above years, the death-rate p 66 DIPHTHERIA. became 14, 16, 14, 29, and 17 per 100,000 respectively. It is evident, therefore, that in these years Lisbon was enjoying an inter-epidemic period. Two years, 1877-78, missing. GENEVA. (Population : 1876—66,815 ; 1896—80,778.) LAUSANNE. (Population : 1881—30,231 ; 11896—37,805.) Italy. The Italian returns all combine diphtheria and croup together. The return for Florence embraces a period for which the other Italian returns are deficient. It is remarkable for the extraordinarily fatal epidemic of 1871-75, which was followed by a smaller epidemic in 1879-80. In 1871 The record for 1881 is missing, ROME. (Population : 1877—278,099 ; 1895—465,563.) FLORENCE. Fig. 26. 68 DIPHTHEEIA. the death-rate from diphtheria 'plus croup exceeded 4 per 1,000 inhabitants. Turin had an epidemic culminating in 1881. The epidemic at Rome, so far as can be judged from the imperfect returns, culminated in 1877-79. At Venice an epidemic culminated in 1875-76, thus following close on the more virulent epidemic at Florence. Both Rome and Venice show a later maximum in 1891, and Milan a maximum in 1893, which Turin has hitherto escaped. Place. Years of observa- tion. Diphtheria plus Croup. Death-rate per 100,000. DiphtheriapZws Croup. Mean death-rate per 100,000 for the entire period. Highest. Lowest. Turin . . . Milan . . . Venice . . . Florence . . Home . . . 1879-95 1973-96 1870-96 1866-90 1877-95 111 in 1881 267 „ 1874 97 „ 1891 422 „ 1871 122 „ 1879 22 in 1895 47 „ 1887 7 „ 1886 23 „ 1884 8 „ 1895 45 99 33 83 21 Holland and Belgium. The curves for the Hague, Rotterdam, and Amsterdam,, from 1879 onwards, show that all these towns had an epidemic of diphtheria, which culminated in Amsterdam and Rotterdam in 1883-84, and in the Hague in 1885. From this point onwards there are sharp distinctions. At the Hague and Rotterdam, there was a marked remission of the disease, while in Amsterdam, although there was also a remission, the total amount of the disease remaining endemic was much greater than that in the foregoing towns. Furthermore, the next epidemic reached its maximum in 1890 in Amsterdam, and at the Hague and in Rotterdam it only culminated in 1892. The curve for Rotterdam shows two earlier epidemics, 1867 and 1873, with which no comparison for the Hague and Rotterdam is practicable. The record for 1878 is missing. TURIN. (Population : 1879—231,647 ; 1895—344,203.) MILAN. (Population: 1873—271135; 1896—458,405.) VENICE. (Population : 1870—135,221 ; 1896—160,953.) Fig. 27 70 DIPHTHEEIA. From Antiverp returns have been received whicli give tlie deaths from diphtheria and croup together from 1860-66, and again, 3 887-93, while for other years they are stated separately. The curve based on these returns presents a great epidemic, culminating in 1864-65; an excessive prevalence of the disease in 1870-80; an epi- demic, culminating in 1883, and a recent epidemic in 1894. All through the years, although there is evidently some transference between diphtheria and croup, it is plain that diphtheria was endemically present (as also in Amsterdam) to an excessive extent. The returns for the city of Brussels do not distinguish between diphtheria and croup. Those for the city and its suburbs together, dating from 1876, give the two separately. The latter returns deal with a population, in 1896, of 618,387, as compared with 190,313 in the city. There was a great epidemic of the disease, culminating in 1862-65, and it remained very excessively prevalent until the end of the decennium 1861-70, only reaching its minimum point between 1876 and 1881. Then followed an epidemic, the greatest mortality from which occurred in 1885, a year or two later than the corresponding maxima in Antwerp and Amsterdam. The curve for L'Agglomera- tion Bruxelloise follows the course of the city curve, so far as the 1885 maximum is concerned. It would appear that the earlier epidemic in the city did not subside in the suburbs for several years after the minimum was reached in the city. 1000, THE HAGUE. (Population : 1879—110,016 ; 1895—180,455.) ROTTERDAM. (Population : 1867—117,104 ; 1895—276,337.) AMSTERDAM. (Population : 1879-308,592 1895—451,493.) ANTWERP. (Population : 1860—111,709 ; 1896—277,581.) BRUSSELS. (Population : 1862—177,9 1896—190,313.) BRUSSELS and SUBURBS (Agglomeration Bruxelloise). (Population : 1876—390,377 1896—518,387.) 72 DIPHTHEEIA. Years of observa- tion. Death-rate per 100,000. Place. Diphtheria plus Croup. Diphtheria alone. Mean death- rate for the entire period. Max. Min. Max. Min. Diph- theria plus Croup. Diph- theria alone. The Hague . Rotterdam . Amsterdam . Brussels . . AntAverp . . 1879-95 1879-95 1879-95 1862-96 1860-96 / 113 in \ 1865 /217 in \ 1865 15inl881 &1876 17 in 1896 /30in \1892 f 19 in 1 1884 f 72in \ 1883 ) ) ) ) lin ) 1888 J 2 in ] 1887 1 3 in ) 1879 / 47 67 12 7 81 \ \' CHAPTER YII. GERMANY, AUSTRO-HUNGAEY, AND RUSSIA. The great seaport of Haviburg has an enormous amount of diphtheria endemically present. Berlin has even more; and most other GTerman cities show an excessive endemicity of this disease. This can be seen by a glance at Figs. 29 to 34. The returns for Hamburg are exceptionally complete thiough the kindness of Dr. Reincke, the head of the Statistical Bureau of the city. He has obtained the early returns s^^ecially for me, they not having been published before.. Diphtheria and croup are classed together. Even when allowance is made for the probable incom- pleteness of the oarlier returns and for changes of nomenclature, it appears clear that there was no very great epidemic of diphtheria between 1838 and I860, though smaller epidemics culminated in 1838, in 1843-44, and in 1852. Then appear epidemic j)eaks in 1862 and 1869. The first great epidemic visible in the Hamburg curve culminated in 1873-74. It was two years later (1875-76) in Berlin and Breslau. In Leipzig the time incidence was the same as in Hamburg ; while in Frankfort-on-the-Main the epidemic occupied the years 1876-79. The compara- tive lightness of this outbreak in Breslau and Frankfort- on-the-Main as compared with the outbreaks a few years earlier in Hamburg and Leipzig, and with the outbreak about the same time in Berlin, may be noted. The next epidemic occurred earlier in Berlin than in Hamburg, the remission between the Berlin epidemic 74 DIPHTHERIA. culminating in 1875-76 and that culminating in 1883-84 being slight. Leipzig had an epidemic culminating also in 1883-84; at Dresden it was still earlier (1881-83). At Hamburg it only reached its maximum in 1 885-8 7^ at GERMANY, AUSTEO- HUNGARY, AND EUSSIA. /O Breslau, 1887-88. In Frankfort it was still later and more protracted, only commencing materially to decline about 1893. The whole series of G-erman curves is worthy of careful study in its bearing on the hypothesis of gradual progression of the disease from town to town. BERLIN. (Poptdation: 1868—716,088; 1896—1,693,726.) Fig. 30. (Througli an oversight, the part of this diagram representing the death-rate from croup is indicated by ^vhite columns enclosed by black lines, instead of by stippled columns, as in all the other diagrams.) The Frankfort records go back to 1851. In the first ten DRESDEN. (Population : 1878—207.845 ; 1895—324,341.) LEIPZIG. (Population : 1872—110,054 ; 1896—404,947.) GERMANY, AUSTEO-HUNGAEY, AND EUSSIA. 77 BRESLAU. (Population : 1866—168,201 ; 1896—378,089.) Death- g i Si rate per 2 ! 1000. 1-60 Fig. 32. FRANKFORT-AM-MAIN. (Population : 1851—36,396 ; 1895—226,000.) Fig. 33. 78 DIPHTHERIA. years they combine diphtheria and croup ; after this, deaths under the two names are separately classified. The curve shows evidence of excessive prevalence of diphtheria in 1851-60^ and indicates that the next epi- demic (described in previous instances as the " early epidemic") did not culminate at Frankfort until 1865. The Munich curve shows a large endemic prevalence of diphtheria phis croup, with epidemics culminating in 1868-69, in 1880-81, and in 1889-90 respectively. These maxima present interesting variations from the corres- ponding maxima in the other German curves, and in the Austrian curves to be next considered. MUNICH. Return for 1875 missing. (Population : 1868—151,562 ; 1895—396,000.) Death- rate per 1000. Fig. 34. GERMANY, AUSTEO-HimGAET, AND EUSSIA. 79 Years of observa- tion. Diphtheria plus Croup. Death-rate per 100,000. Diphtheria alone. Death-rate per 100,000. Mean Death- rate per 100,000 for the entire period. Highest. Lowest. Highest. ; Lowest. Diph- theria plus Croup. Diph- theria alone. Hamburg . Berlin . . . Breslau . . Dresden . . Leipzig . . Frankfort- \ on-tbe-Main/ Munich . i 1872-95* 1869-96 1866-96 1878-95 1872-96 1851-95 1868-74 1876-95 f 123 in \ 1887 f 242 in 1 1883 f 162 in 1 1887 (253 in 1 1882 (234 in I 1884 f 160 in t 1890 184 in 1869 17 in) 1896 1 34 in 1896 28 in) 1870 ) 52 in i 1895 ! 37 in ) 1896 1 9 in 1860 53 in) 1895 j" 220 in 1883 157 in 1890 31in-\ 1896 J" 7 in\ 1862 / 49 119 78 129 90 64 99 101 53 * The figures for Hamburg relate to a shorter series of years than the curve in Fig. 29. AUSTEO-HUNGAEY. We may commence with. Trieste, which, lies afc the upper end of the Adriatic Sea on its eastern side, Venice being on the western side. The diagram shows three great epidemics, and a large endemic prevalence of the disease. The first epidemic ranges from 1869 to 1874. The first epidemic ranges from 1869 to 1874, culminating in 1873; the second from 1881 to 1886, culminating in 1885; while the third, which began in 1891, culminated in 1894. The returns for Buda-Pesth date back only to 1874. The first epidemic shown on the diagram reached its maximum in 1878. This may be regarded as a sequel to the 1871-75 epidemic in Trieste; or if it be assumed that the infection travelled southwards and to the west, then the Trieste epidemic of 1881-85 would correspond to the Buda-Pesth epidemic of 1871-79. In 1886 another severe Death- g Tate per "^^ 1000. ,00 PRAGUE. BUDA-PESTH. (Population : 1874—296,272 ; 1896—579,275.) TRIESTE. (Population : 1866-116,! 1896—161,886.). GEEMANY, AUSTEO-HUNGAET, AND EUSSIA. 81 epidemic of diphtheria began in Buda-Pesth, which reached its maximum in 1890-92^ and then rapidly- declined, Vienna showed a fairly steady endemic prevalence of diphtheria between 1865 and 1870. In 1876-79 it suffered from a severe epidemic, which was repeated in 1891-94. The time incidence of these epidemics corresponds fairly well with that of Buda Pesth, which is on the banks of the Danube, at a lower point than Vienna. The recent epidemic was, however, later in Vienna than Buda-Pesth, reaching its acme in 1892-94, instead of in 1890-91. The returns for Prague show epidemics reaching their maxima in 1877-78 and in 1887-89 (Fig. 35). VIENNA. (Population : 1365—561,647 ; 1895—1,488,463.) Fig. 36. 82 DIPHTHERIA. Years of observa- tion. Death-rates per 100,000. Place Diphtheria phis Croup. Diphtheria alone. Mean Death- rate. Highest. Lowest. Highest. Lowest. Diph- theria plus Croup. Diph- theria alone. Trieste . . Buda Pesth . Vienna . . Prague . . 1865-96 1874-96 1865-96 1865-95 / 251 in \ 1873 / 194 in \ 1878 (177 in \ 1878 27in\ 1877/ 46 in 1895 42 in 1873 140 in 1891 (144 in 1878)* 110 in 1888 29 in -^ 1884-5/ (19 in ~\ 1884) / lin \ 1876/ 180 108 77 68 _ * The " Diphtheria alone " returns for Vienna refer to the years 1865-89. EuSSIA. St. Petersburg suffered from a severe epidemic of diphtheria in 1881-83, which in 1882 caused a death-rate of 1"52 per 1000 of population. A later epidemic cul- minated in 1894. At Moscow the disease appears to have been continuously present in greater amount than at St. Petersburg, but there were epidemics about the same time as at St. Petersburg. The 1882 maximum corresponds to the 1883 maximum in Berlin. (There is some doubt as to whether the height of the latter part of the cnrve for Moscow is not exaggerated, in consequence of doubt as to its true population. The oscillations alone must, therefore, be regarded, and not the general upward direction of the curve.) In Warsaw, judging by such records as are available, there is a large amount of diphtheria. The death-rate from diphtheria |)Ziis croup in 1885 was 101, and in 1887 was 72 per 100,000 of population. These figures are de- rived from the annual summaries of the English Registrar- GEEMANY, AUSTEO-HUNGAEY^ AND EUSSIA. 83 Greneral. The following figures are derived from a paper in the Proc. Buda-Pesth Congress, tome iv., p. 452. Death-rate from Diphtheria Tear. ]plus Croup per 100,000. 1877 53 1878 100 1880 118 1882 . • 176 1884 161 1886 98 1888 75 1890 63 1892 67 Years of observa- tion. Death-rate per 100,000. Place. Diphtheria plus Croup. Diphtheria alone. Means for the entire period. Max. Min. Max. Mm. Diph- theria plus Croup. Diph- theria alone. St. Peters--) 'burg. . / Moscow . . 1879-90 1892-95 1878-96 1 1 ~ rill in \ 1894 -{ 85 in 1878 154 in 1882 105 in 1893 27 in\ 1893/ 27 in) 1880/ 72 65 64 EOUMANIA. At Bucliarest, in tlie year 1880, 135 deaths from diphtheria were recorded, equal to a death-rate of 67 per 100,000 of the population. No other returns have been secured. 84 DIPHTHEEIA. ST. PETERSBURG. Records for lS91'-92 missing. (Population : 1879—669,741 ; 1893—954,400.) MOSCOW. (Population : 1882—753,469.) Fig. 37. CHAPTEE VIII. SCANDINAVIA. The Scandinavian death returns for diphtheria are of great importance for several reasons. They embrace a long period of years. They can be supplemented and checked by a corresponding record of cases of sickness, based on a national system of notification of certain forms of sickness in the whole of Norway and Denmark. They have a further interest because they relate to sparsely populated countries, in which the influence of epidemics can be easily seen; and in which the inter-epidemic periods might be expected on a_pn'6»ri grounds to be well marked. That this is so is shown very clearly by the curves for Christiania, Grottenburg, Stockholm, and Copenhagen. The curve for Christiania shows a large epidemic, reach- ing its maximum in 1862, but well on its way in 1860, when the record commences. For some years afterwards the amount of diphtheria was excessive, but between 1870 and 1880 there was little of it. In 1884 an epidemic of enormous magnitude commenced, lasting with great force until 1891. The highest death-rates from diphtheria pZws croup were in 1885 and 1887, in which they were 3*28 and 3"29 per 1000 of the population respectively. The curve for the whole of Norway presents identical epidemics, with minor differences (see Fig. 47, p. 104). The earlier portions of the two curves are almost identical in shape ; but the more recent epidemic is somewhat more protracted than in Christiania, and takes a longer time in attaining its maximum than in the capital town, not reach- ing it until 1890. This is what might have been foretold 85 86 DIPHTHERIA. in view of the scattered distribution of the Norwegian population. CHRISTIANIA. Record for 1878 missing. (Population : 1860—53,587 ; 1893—161,157.) Fig. 38. SCANDINAVIA. 87 Gottenburg had. the early epidemic like Christiania^ but it began to decline earlier than in the latter. In 1871-73 it had an epidemic, which is quite unrepresented in Christiania; and the next epidemic is earlier in Gotten- burg than in Christiania, reaching its acme in 1884 instead GOTTENBURG. (Popiaation : 1863—41,200 ; 1895—117,512.) Fig. 39. of 1885-87. It would appear, howerer, that Christiania had accumulated in the longer interval a much more '^ explosive " population, if we may judge from the enormous epidemic among its population which followed close on the heels of the 1882-85 epidemic in Gottenburg. OO DIPHTHERIA. Gottenburg showed tlie commencement of anotlier epidemic in 1894-95. Attention may be directed to tbe regular intermission of its epidemics, an interval of ten or twelve years intervening between the maxima of its respective epidemics. It seems impossible to resist the surmise that Grottenburg is nearer to and more easily influenced by some centre of infection than the more distant Christiania. A comparison with the next curves will confirm this sur- mise, though it must be confessed that the view taken as to the direction in which infection travelled will depend on the point from which one starts. It is probable that no uniformity in this respect exists, and that there is exchange in infection as in the commerce with which it is closely associated. StockJiolm, like Copenhagen and other Danish towns, shows much more diphtheria endemically present than either Grottenburg or Ohi'istiania. They are both more closely connected geographically and commercially with the main portion of the European continent than is- Grottenburg, and still less Christiania, Hence probably they keep up the supply of infection more frequently from Germany and other centres of diphtheria. Stockholm shows an epidemic 1861-64. In 1880 an epidemic began lasting several years, which after only a small remission was followed by a greater epidemic in 1892-93. The returns for Copenhagen and Provincial Towns of Denmark go back to 1845.* They both show a very con- siderable epidemic in 1846-49, of which we have but scant statistical records for the European continent, t * These figures are derived from Dr. Carlsen's paper in Janus, September, 1896. t Compare Fig. 29, Hamburg, and Fig. 33, Frankfort-on-tlie- Main, for records at or near the same period. Death- rate per 1000. 90 DIPHTHERIA. This epidemic culminated in Copenhagen in 1846^ in the provincial towns not until 1850.* The latter, it must be remembered, represents the combined experience of a large number of towns, and indicates, therefore, the gradual progression of the infection. The next epidemic, a large one, reached its maximum in 1865, not so early as at Christiania or Stockholm. The next epidemic was most prevalent in 1877-80. There was a very considerable remission in 1881-86, and then came an epidemic lasting from 1886 to 1894, and culminating in 1891 in Copenhagen, not until 1891-94 in the provincial towns. Years of observation. Death-rate per 100,000. Place. Diphtheria plxis Croup. Diphtheria alone. Mean Death- rate for the entire period. Highest. Lowest. Highest. Lowest. Diph- theria •plus Croup. Diph- theria alone. Christiania . Gottenburg . Stockholm . Copenhagen Provincial ] Towns of 1^ Denmark J Whole of] Norway ^ 1860-93 1861-95 1861-96 1844-95 1844-95 1860-91 Diph. 1867-91 Diph. and Croup (329 in 1 1887 (250 in \ 1873 fl39 in \ 1893 & y 1862 f 164 in \ 1891 ( 203 in \ 1865 ( 123 in 1 1890 17 in 1871 5 in 1891 12 in 1896 15 in 1868 & 1872 15 in 1872 11 in 1870 303 in 1887 137 in 1884 121 in 1893 1 ) 1 \ - 110 in 1890* lin ) 1873 ) Sin 1891 ■ 6 in \ 1878 ; 2in ■) 1866*] 109 58 72 64 84 49* 80 23 45 64 50 * These figures relate to the period 1867-91. The Faroe Islands form a detached part of Denmark, between the Shetland Islands and Iceland, having only * Compare with these records, the great excess of Croup in London in 1847, shewn in Fig. 13, p. 30. SCANDINAVIA. 91 infrequent communication with the continent of Europe. Their statistics therefore have a special interest, and I am glad to be able to reproduce them from a valuable article in Janus, by Dr. J. Carlson (p. 177, September, 1896). Reported Cases. Deaths. •y Diphtheria. 1 Croup. Diphtheria. Croup. 1876 48 5 1 _ 1877 106 6 ? ? 1878 161 32 ? ?. 1879 118 24 6 17 1880 124 27 7 11 1881 69 5 1 2 1882 94 8 11 3 1883 34 2 4 1 1884 31 1 2 3 1885 27 1 4 2 1886 14 5 4 3 1887 7 1 1 1888 6 1 1 1889 — — 1890 — — — 1891 6 — — — 1892 25 9 5 5 1893 10 8 1 8 1894 35 9 1 7 The maximum prevalence in 1878-80 corresponds closely with that in Denmark. The next minimum was several years later in the Islands than in Denmark, and the Danish epidemic from 1886 onwards did not seriously affect the Faroe Islands until 1892. CHAPTER IX. ArRICA AND ASIA. Outside Europe and America, with a few exceptions in Australia, statistical records of the incidence of diphtheria are very scanty. I have only succeeded in obtaining a few scattered records from North and South Africa, from India, and from Australasia. Mortality from Diphtheria phis Groiijj. Alexandria. Cairo. Year. Recorded Approximate Death- Recorded Approximate Death- Deaths. rate per 100,000 Inhabitants. Deaths. rate per 100,000 Inhabitants. 1879 43 20 _ 1880 43 20 — — 1881 33 16 1882 28 13 456 122 1883 154 73 730 195 1884 105 50 657 175 1885 81 39 829 221 1886 90 37 300 80 1887 59 26 121 32 1888 51 22 85 23 1889 61 26 113 30 1890 119 52 98 26 1891 141 61 90 24 1892 128 55 72 19 1893 107 46 112 30 1894 124 54 86 23 1895 100 43 90 24 1896 142 62 135 36 Judging by the available records, the dry and warm climates of Gairo and Alexandria are far from being inimical to the prevalence of a large amount of fatal diphtheria. The preceding returns have been supplied by the kindness of Dr. F. W. Sandwith, of Cairo, who 92 APKICA iLND ASIA. 93 organised the health statistics of Egypt. As there has been no census since 1882, it has been necessary to give the actual deaths in each city. A column has been added for each city giving the death-rate based on the popula- tion figures as stated in the English Registrar-General's Annual Summaries. (For Alexandria these are given as 212,034 in 1879-84, and 231,396 for 1886-91; and for Cairo, as 374,838 for 1886-87.) If these returns are to be trusted, Cairo has suffered much more severely from diphtheria than Alexandria. The former city had a most fatal epidemic in 1882-85, culminating in 1885. In Alexandria there was a smaller epidemic during the same years, reaching its maximum two years earlier than the more inland city. In recent years Alexandria has suffered rather more than Cairo. Cape Colony. A general system of registration having only been adopted on the 13th July, 1894, statistics are not available for earlier years. According to the Cape of Grood Hope Reports on the Public Health, 1894, p. 24, the death- rates from diphtheria plus croup in that year were very high; being 80'4 in Cape Town, 83*0 in King William's Town, 92*5 in Worcester, 162*5 in Malmesbury, and 398'4 in Aberdeen, per 100,000 of population. India. Calcutta. — I am not in a position to state how far the registration of deaths and the causes of death in this city is effective ; but after making a free allowance for incom- pleteness of registration, the figures published in the Annual Summaries of the English Registrar-General show that there is extremely little diphtheria in Calcutta, and that it has not been epidemic in any year since 1879. 94 DIPHTHERIA. The largest number of deaths registered as due to diphtheria since 1879 was 29 in 1882, which is only equal to a death-rate of 6 per 100,000. The lowest number of deaths was 9 in 1879, which is equal to a death-rate of 1 per 100,000. The conclusion that diphtheria is a rare disease in India is confirmed by the Annual Reports of the Sanitary Commissioner with the Qovernment of India. Thus in the report for 1891, the summary table Z states that during the year only two non-fatal cases of diphtheria were admitted in the Bengal European army (strength 40,953), none in the Madras army (strength 13,268), and none in the Bombay army (strength 12,735) ; and that in the native army (143,970), and the jail population of India (strength 101,910), there were no cases during the year. In 1894 there were only two cases (both fatal) in the whole European army in India. In the same year there were only two deaths from diphtheria and eight cases among the children of the European regiments out of an average strength of 5,680. Diphtheria evidently does not thrive in a tropical climate, Japan. The Annual Report for the Home Department of the Imperial Japanese Government for 1892 is stated to be for the twenty-fifth year, but I have been unable to secure information as to deaths from diphtheria further back than 1888, The report for 1892 states, however, that the number of cases in Japan in 1892 was greater than in any other year since 1872. The death-rate in 1892 from diphtheria was 60 ; in 1893, 7; in 1894, 7; in 1S96, 7; and in 1896, 8 per 100,000. The paucity of cases in proportion to these AFRICA AND ASIA. Year. Cases. Deaths. Population. 1888 2,582 1,450 1889 2,669 1,495 — 1890 2,448 1,438 — 1891 3,429 1,974 — 1892 4,359 2,531 41.268,732 1893 5.726 3,205 41,089,940 1894 5,308 2,903 41,388,313 1895 6,100 3,025 41.813,215 1896 8,590 3,279 42,270,620 deaths shows tliat in all probability their notification is very incomplete. The figures for the years 1893-96 have been sent me by the Director General of the Sanitary Bureau, Home Department, Japan, who has kindly added the following figures as to the number of patients and deaths from diphtheria in five principal towns in Japan : — DixMlieria in Japan. Name of 1S94. ' 1895. Towus. No. of No. of Popula- Death-rate ' No. of No. of Popula- Death-rate Patients. Deaths. tion. per 100,000. Patients. Deaths. tion. per 100,000. Tokio . . ' 371 212 1,242,224 17 707 183 1,268,930 14 Kioto . . 111 54 328,411 17 142 43 340,101 13 Osaka . . 129 74 488,937 15 160 55 487,184 11 Yokohama i 49 36 160,439 23 80 40 170.252 24 Kobe . . 1 58 37 158,993 24 62 32 161,130 20 If registration during the above years is fairly complete, the above figures indicate that since 1888 there has been a comparatively small mortality from diphtheria in Japan. They also show that, as in all the countries for which I have been able to secure records, the death-rate from diphtheria is much higher in the towns than in the country at a whole ; from which it follows that for the years to which the above records relate, diphtheria is more an urban than a rural disease. CHAPTER X. AUSTEALIA AND NEW ZEALAND. The most complete statistics available are for South Australia^ (Fig. 43). They show^ as do the other Aus- tralian curves as far as they extend, that diphtheria is a very fatal endemic disease in Australia, interrupted by epidemic exacerbations at intervals of a few years. Curves for South Australia (Fig. 43), and for Victoria (Fig. 41), show that both these neighbouring colonies suffered severely from the " early " epidemic of diph- theria, in 1858-61. In both of them again, after a very short and imperfect remission, a second epidemic occurred, culminating in 1864. With only a single year's interval (1866) a third epidemic culminated in 1867 in South Aus- tralia, a year or two later in Victoria. The record for Victoria is taken from the ReiDort of the Boyal Gam- mission ajjpointed hy His Eminence the Governor to inquire into Diphtheria in Victoria, 1872. The report of this Com- mission states that attention was first drawn to the existence of diphtheria in Victoria, in 1857. It also mentions that cases of the disease were seen in Melbourne over twenty years ago {i.e. presumably before 1852), and that some bad cases of the disease were seen in Castlemaine in 1854-55. This subject is discussed in a paper entitled " Eemarks on the introduction of Diphtheria into Victoria," by Mr. W. Thompson, F.R.C.S. Edin. {Australian Medical Journal, July, 1872). In this paper, it was contended that the * Derived from " A Contribution to the Demography of South. Australia," by T. S. Borthwick, M.D. (1891). 96 AUSTRALIA AND NEW ZEALAND. 97 disease was imported by passengers wlio were landing every week at Melbourne from England, wbere a great epidemic was raging. Tlie first recorded fatal case in Melbourne_, lie states, occurred in October, 1858, and up to the end of that year only six fatal cases were recorded in the colony. Mr. Thomson quotes official statistics Death- g rate per 2 1000. VICTORIA. QUEENSLAND. Fig-. 41. for the colony^ going back to 1853. These] show that, from 1853 onwards, the annual deaths from croup num- bered 32 (six months of 1853), 72, 53, 69, 56, 102, 229, 156, 166, the last number being for the year 1861. Diphtheria first appeared in the returns in 1859. A writer in the Audralian Medical Journal, during 1857, H 98 DIPHTHERIA. states that lie examined tlie larynx in several fatal cases of croup in Melbourne, without finding any false mem- brane. Notwitlistanding tbis statement, I am of opinion MELBOURNE. l(Popiilation : 1861—189,916 ; 1895—443,255. SYDNEY. (Population : 1856—81,581 ; 1898—408,500.) ADELAIDE. (Population : 1875—31,918 ; 1896—40,395.) tbat a certain proportion of the cases of croup before 1858 were diphtberitic, and tbat the great increase of AUSTRALIA A'NB NEW ZEALAND. 99 mortality from diphtheria and croup immediately after- wards did not represent solely deaths from a newly imported disease, though doubtless a large number of cases were imported. According to the above report, diphtheria appears to have been in Tasmania before it was noticed in Victoria. In Queensland (Fig. 41) a considerable epidemic cul- minated in 1873, which lasted with severity until 1877. A second epidemic reached its maximum in 1885. SOUTH AUSTRALIA. Fig. 43. New South Wales, lying between Queensland on the north and Victoria and South Australia on the west, evidently suffered from the " early " epidemic of diph- theria, as shown by the table (top of page 100) quoted from the ninth annual report for 1866 of this colony : — " The first mention of diphtheria occurs in the report for the year 1864, but the disease had occurred to some extent during previous years, and had been tabulated as quinsy >} if; * Quoted from "Age and Sex Incidenc2 Mortality in Michigan, from Diphtheria and from Croup, during 25 years, 1870-94," by C. 100 DIPHTHERIA. Deaths registered in Ne w South Wales, 1856-64. Year. Scarlatina. Quinsy. Diphtheria. Croup. Quinsy, Diph- theria, and Croup.* 1856 31 4 61 65 1857 38 4 38 42 1858 165 6 69 75 1859 120 50 85 135 1860 89 115 52 167 1861 70 153 74 227 1862 103 309 35 344 1863 95 51 239 142 432 1864 350 32 162 85 279 The curve for Sydney, the capital of New South Wales, shows that the early epidemic culminated in 1864, or possibly in 1863, as for the whole colony; but owing to changes made in nosological arrangement of tables, whereby diphtheria was included with other zymotic diseases, figures for 1863 are unobtainable. The acme was reached several years earlier than in Melbourne, in which it was not attained until 1868. The next epidemic was rather later, but more continuous, in Sydney than in Melbourne ; while the most recent epidemic was a year earlier in reaching its maximum in Sydney than in Melbourne. For the valuable and very complete statistical returns, represented in Figs. 44-47, I am indebted to my friend Mr. Greorge Leslie, assistant actuary to the Government of New Zealand, who has given himself enormous trouble to secure them for me. They are particularly full, and L. Wilbur, M.D. {The Journal of American Medical Association, Aug. 15, 1896). *■ The last column, added by me, probably gives the truest conception of the progress of the epidemic, though it may be that some additional transference from the heading " Scarlatina " would be necessary to ensure accuracy. The maximum was reached in 1862-63, as compared with 1860 in South Australia and Victoria. AUSTRALIA AND NEW ZEALAND. 101 liave a special interest in the marked contrasts between their experience and that of Australia. The records begin in 1872. Fig. 44 shows that in that and the two following years diphtheria was epidemic in New Zealand. A study of the details for individual districts shows that the chief brunt of the epidemic fell upon the districts of Auckland in the North Island^ and Canterbury in the South Island, occurring later or remaining entirely in abeyance in the other districts. In some of these dis- tricts diphtheria, as a fatal disease, entirely disappears NEW ZEALAND. (Population : 1871—266,986 ; 1896— 714,16-2.) in the intervals of epidemics. Diphtheria is evidently not a disease of virgin soil, but follows in the wake of aggregated human communities. The contrast between Canterbury and the other districts is further commented upon on page 154. The diagrams, so far as further de- tails are concerned, tell their own story. Note. — Througli an oversight, in Figs. 44-47 the parts of tlie diagrams relating to croup are not stipple:! as in all the pre- ceding diagrams. NEW ZEALAND. (A) North Island. — Districts of — AUCKLAND. (Population: 1871—64,432; 1896—155,940.) TARANAKI. (Population : 1871—4,599 ; 1896, 29,833.) WELLINGTON. (Population : 1871-25,218 ; 1896-123,930.) HAWKE'S BAY. (Population : 1871—6,212 ; 1896—34,709.) Fig. 45. KoTE. — The record begins in 1872. Ko deaths from diphtheria or cronp were registered in Taranaki in 1884, or in Hawke's Bay in 1872- 73 and in 1878. AUSTRALIA AND NEW ZEALAND. 103 NEW ZEALAND. (B) South Island. O I O I 00 o> I 00 I 00 Districts of— CANTERBURY. (Population : 1871—48,807 ; 1896—137,867.) OTAGO and CHAT- HAM ISLANDS. (Population : 1371—73,970 ; 1896—166,407.) Fig. 46. 104 DIPHTHEEIA. NEW ZEALAND. (B) South Island, continued- Districts of — MARLBOROUGH. (Population : 1871—5,385 ; 1896—12,677.) NELSON. (Population : 1871—23,042 ; 1896—36,201.) WESTLAND. (Population;: 1871—15,381 ; 1896—14,598.) Fig. 47. CHAPTER XI. EELATIONSHIP BETWEEN PREVALENCE OF AND MORTALITY FROM DIPHTHERIA. It has been already mentioned that in Norway and Denmark there is a national system of registration of sickness.* From the data based on the records thus ac- cumulated since 1860^ in the case of Norway, and since 1855 for the capital of Denmark, the following curves have been constructed. In order to bring out the exact relationship between sickness-rates and death-rates, these have been calculated in terms of their deviation from the mean rate for the entire period. The mean annual sickness-rate from diphtheria for Norway in the period 1860-91 was 2*25 per 1000 of population; the mean annual death-rate "40 per 1000. In Copenhagen the mean annual sickness-rate for the years 1855-95 was 5-88 per 1000, the death-rate '78 per 1000. The curves for both Norway and Copenhagen show a remarkably close relationship between prevalence and mortality. In the recent epidemic in Norway, the fatality (case mortality) evidently increased during the epidemic years, but this was not so in the earlier epidemic. There is some evidence of greater fatality during the greater part of the inter-epidemic period. In Copenhagen, during the epidemics culminating in * For further details, see a paper by the author, on " A National System of Notification and Registration of Sickness." Journal of the Royal Statistical Society^ vol. lix., part i. (March, 1896). 105 106 DIPHTHERIA. 1865 and in 1879, the fatality of the disease increased; but during the epidemic culminating in 1890, the cases increased in a slightly higher proportion than the deaths. The relationship between prevalence of and mortality Fio-. 4S. NORWAY. — Annual deviation of tlie sickness - rate (o — o — o— o) and deatli-rate (•—•—• — •) caused by diphtheria from the mean rate for the entire period 1860-1891. from diphtheria may also be considered for Berlin, Ham- burg, and London. In Berlin during the years 1885-96, for which the com- parable data are available, the average case-rate from dipbtheria was 3*2 5 per 1000, varying from 2*16 in 1891 EELATION BETWEEN FATALITY AND MOETALITY. 107 to 5-97 in 1885. Fig. 50 shows some increase of fatality in the earlier years of least prevalence of diphtheria in 0681 0881 • - 02,81 0981 Od C 3J tn ,^ S [ 3 £-S S o o O i o 9 00 CO ^ o + + 1 + 1 + 1 1 1 1 1 1 '^ S'" +\ + + 1 1 1 •1 Berlin [i.e. in 1889-93). In the most recent years this is not so^ and it is probable that the disturbing influence of 108 DIPHTHERIA. the treatment of diphtheria by antitoxic serum is to be seen in this change of fatality. Fig. 50 also shows for the years during which notification has been compulsory in London (beginning with the first complete year of noti- fication, 1890) J a close relationship between prevalence and mortality. + 30 + 20 \^1/. [ ^ 1 ^ m -20 '><' J 1 -40 ' / 1 Fig. 50. (A) BERLIN. (B) LONDON.— Annual deviation of the sickness-rate (o — o — o— o) and death-rate (• — • — • — •) caused by diphtheria from the mean rate for the entire period. In the statistical reports for Hamburg the fatality from diphtheria _ph(s croup is given for a long series of years. Beginning with 1872, and ending with 1896, the annual percentage fatahty has been as follows : — 16-0, 17'6, 19-9, 17-8, 17*6, 13-.3, 15-0, 12-5, 14-3, 127, 13-0, 13-8, 15'9, 16-1, 17'0, 16-7, 170, 15-8, 16-4, 14-6, 15-9, 15-2, 15-1, 8-6, 8-4. EELATION BETWEEN FATALITY AND MOETALITY. 109 A study of these figures in connection with the epidemic curve for Hamburg (Fig. 29)^ shows that the fatality was greatest during the years of greatest epidemicity, while in portions of the Berlin and Scandinavian curves the experience is the reverse of this (Figs. 48-50). The years of greatest prevalence of diphtheria at Hamburg are in- dicated by using thicker type for the percentage case- mortality in these years. The preceding figures and diagrams have an incidental utility on which some stress may be laid. They confirm my confidence in the general trustworthiness of the dia- grams of death-rates scattered throughout this book. The more these are examined^ the more, I believe, they will be found to tell the truth in all essential particulars. Note. — On page 4 it is stated that there has been in the last few years in England a tendency to notify as diphtheria, cases which would previously have escaped notification. This remark was based on an impression as to the working of the Infectious Disease (Notification) Act, 1889, which is, I believe, shared by the majority of medical officers of health. Fig. 50 B, however, shows that this, cause of error, if in operation at all, does not appreciably vitiate the statistics of large communities. In 1897, the cases of diph- theria notified in London were 12,811 as compared with 14,224 in 1896, giving a fatality of 17"7 as compared with 18'9 per cent. Thus in 1897, diphtheria in London declined both in amount and in fatality. CHAPTER XII. SUMMARY OP THE HISTOEY OF DIPHTHERIA. The following summary of the history of diptheria is derived from Hirsch's classical work, and from a number of French, English, and American writings, the different outbreaks being placed in chronological order. Owing to the difficulties of nomenclature, there is some difficulty in recognising diphtheria in the early accounts of ^' ulcerous sore throats " or " angina maligna '^ ; but in the majority of the instances here cited, it is probable that diphtheria is the disease which is mentioned. The summary is by no means complete. It would have been easy to double its bulk by including all the refer- ences in old medical literature; but sufficient has been given to indicate the widespread and virulent character of the disease before exact statistical records were kept. Only a few instances of epidemics during the last forty years have been given, and these have been, as a rule, selected from districts in which statistical data are de- ficient. The diagrams already given sufficiently indicate the widespread character of diphtheria during these forty years. A.D. 380. Macrobius speaks of an epidemic at Rome in con- nection with which sacrifices were offered up to a certain goddess — " ut populus Romanus, morbo, qui angina dicitur promisso voto, sit liberatus." 110 SUMMARY OF THE HISTORY OP DIPHTHERIA. Ill 856. Baronius writes {Annal. ecclesiast.) of a '' pestilentia fauciunij qua fluxione gnttur obstructum citam mor- tem inferret/' as having occurred at Rome."^ 1004. A similar epidemic is mentioned; '^ Catarrhus descendens in fauces^ meatus obstruens^ suffocatos miseros homines confestim ruori cogebat.''^ t 1039. Cedremus records an epidemic sickness known as KuvdyKT], wbicb was prevalent in some provinces of the Byzantine Empire^ and was very fatal, j 1337. A fatal epidemic of sore throat occurred in Hol- land. 1389. Short refers to a kind of angina prevalent in England, which was fatal to a large number of chil- dren, § 1515. Herrera described diphtheria of the skin and of wounds, and looked upon the pseudo-membrane he found after death as the essential characteristic of the disease. II 1517. Hecker, in his account of the sweating sickness in England in 1517 {Epidernics, p. 224) describes "a malignant and infectious inflammation of the throat/' which was often fatal within a day. This appeared in Holland, and later in the same year in Basle, " where, within eight months, it destroyed about 2,000 persons." A similar outbreak is described by F. Yon Word in the Rhine districts, in which " men's tongues and throats were covered as with a fungus^ and turned white, etc." * Quoted hj Hirsch, vol. iii., jj. 73-4. + Ibid. I Ibid. § Quoted by AVebster, History of Ejndemic and Pestilential Diseases. Hartford, 1799, i., 143. II Quoted by Jacobi, Treatise on Diphtheria. New York, 1880, p. 2. 112 DIPHTHEEIA. 1557. A similar very fatal epidemic in Holland^ wHcb spread to other parts of Europe. (Petr. Forestus^ obs., lib. vi., de febribus pitblice grassantibus, et lib. xv. obs. 5-11.) 1563-64. A similar epidemic in Naples and in Sicily, which reached Constantinople and Alexandria in 1564. 1565. A pestilential angina described at Dantzig, Co- logne, and Augsburg. 1576. A very malignant form of angina prevailed in Paris. Baillou mentions the finding of the false membrane at an autopsy by a surgeon whose name is not given, and describes its characters. {Ballonii opera omnia, Epid. et Epliemer, lib. xi., 1576. Greneva.) 17th Century. Early in the century an epidemic of an- gina in Spain, known as '' garotillo,^' commenced apparently about 1583, and was still prevalent in 1618. It occurred at Spain later in the century; thus 1630 (Saragossa), 1645-6 (Alaejos), 1666 dif- fused. 1608. Mercado, the Sevillian poet and historian, tells of a child communicating the disease to its father by biting his finger.^ 1610. A outbreak similar to that in Spain occurred in Mantua and Lombardy. 1618. The same disease appeared in Naples, and was known as '' male de canna,^^ disease of the trachea. It raged here more or less for twenty years. A full account of this is given by Carnevale in his treatise De Epidemico 8trangulatione Affectu. Between 1618 and 1642 it was generally diffused throughout Italy. Quoted by Jacobi, op. cit.^ p. 2. SUMMARY OF THE HISTORY OF DIPHTHERIA. 113 1626. Portugal first visited by angina maligna in this year. 1632. Alaymus published a treatise on " Syrian Ulcers/' evidently the same disease as that described by Carnevale. 18th Century. From the middle of the 17th century up to 1740, there is but little mention of the prevalence of malignant angina; but in 1701 it appeared in the Ionian Islands, and in 1715 in Aquilar de Campos (Province of Valencia). 1743-50. In 1743 angina maligna appeared in Paris, where it prevailed until 1750. Chomel described cases of diptheritic paralysis in connection with this epidemic. {Dissert, hist, sur Vasped du mal de gorge gangreneux, etc. Paris, 1749.) A similar epidemic appeared in England and in Cremona, N. Italy, about the same time. The Eng- lish epidemics of angina maligna were described by Eothergill. 1743. An epidemic was noted in Ireland by Molloy. 1746. An epidemic at Bromley, England (Fagge). 1747. Ghisi describes an epidemic appearing in Cremona, which produced death by sufi'ocation. At Orleans, Arnault describes cases of malignant sore throat fatal within twenty-four hours. An epidemic at Greenwich (Fagge). 1749. At Liskeard, in Cornwall, an epidemic of malig- nant ulcerous sore throat appeared, of which Dr. Starr contributed a description to the Philosophical Transactions. He fully described the false mem- brane, including its extension to the larynx ; and, in fact, anticipated Bretonneau's description of diph- theria. About this time appeared Fothergill's An Account of the Sore Throat. (London, 1748.) 114 DIPHTHEEIA. 1752. An epidemic was noted in New York by Father Middleton. A severe outbreak in Simmentkal (Swit- zerland) in tke same year. 1753. An epidemic at Zuricli (Langhaus). 3 757. Huxbam described an Engbsb epidemic, closely associated with scarlet fever. This appears to have prevailed more or less from 1751-53. 1758. Angina maligna seen again in Paris; also in 1759 and 1762. 1761. Rosen described an epidemic of cynanche maligna in Sweden, apparently beginning in 1755. 1765. Dr. Home's treatise on croup was published at Edinburgh. He showed that the formation of a false membrane in the larynx and trachea is essential to the disease. 1769-70. Kittel observed an epidemic in Utrecht associ- ated with croup. 1771. Dr. Bard described an epidemic of suffocative angina appearing in New York in this year. In the same year Crawford, in Scotland, brought out a dis- sertation. {De Cynanche Stridnla. Edinburgh, 1771.) 1774 and 1787. Angina maligna prevalent in some Nor- man towns and at Poictiers. 19th Century. At the beginning of the 19th century angina maligna retired into the background as an epidemic disease; so much so that in 1855-60, when it again became widely prevalent in Europe and North America, experienced observers regarded it as a new disease. France was an exception to this rule. 1801. Epidemics of angina maligna at Marienwerder and several places in East Prussia. 1805. Epidemic of angina maligna at Padua. 1816. An epidemic of the same in Crete. SUMilAEY OF THE HISTOEY OF DIPHTHERIA. 115 1817—25. An epidemic of malignant sore throat in Kent in 1817, at Glasgow in 1819 and 1825, and at Kelso in 1825 * 1821. Diphtlieria appeared for tlie first time at Lima; afterwards to a limited extent in 1850 and 1855, and again broke out in 1858. The black races proved as resistant to diphtheria as they were to yellow fever, t 1834. Epidemic of angina maligna at Skien, in Xorway. 1835. An epidemic of malignant sore throat at Lisbon. 1810-11. Angina maligna was prevalent at Lyons. 1818-21. The well-known epidemic described by Bre- tonneau occurred at Tours. 1824-25. Epidemic at La Ferriere. 1826. Epidemic prevalent in Geneva and Canton Vaud. 1825-36. The disease became widely prevalent in France^ particularly in the north-west provinces. 1818-21 and 1825-26. Term " diphtherite '' originally used by M. Bretonneau in his treatise on the sub- ject,, which appeared in 1826. His treatise was based chiefly on epidemics of malignant sore throat in Tours and its neighbourhood in 1818-21, and again in 1825-26. The report of the Lancet Sanitary Commissioners, 1859, states that, "while in the year 1825, a year remarkable for its extreme dryness, the communes * The Kent outbreak occurred at Asliford. It is mentioaed by Dr. Burdon Sanderson as "an epidemic of malignant sore throat, which proved extensively fatal to children, -was never accompanied with any eruption, and differed from scarlatina." (Thome, Xatural History of Dijphtheria, p. 15.) t Quoted in Dr. Babington's Presidential Address {Transactions Epidemiological Society^ vol. i., part i., p. 9), who refers to Hays's. American Journal, vol. xxiv. p. 521, for a statement as to the gradual spread of the disease to Lima from the north. 116 DIPHTHERIA. north, of Orleans were laid waste by diphtlieria, it made as many victims in tlie damp and warm year, 1828, in the country south of Orleans." 1838 and 1849. Dr. Hoskins, of Guernsey, stated: "It is my firm conviction that the scarlatina epidemics of 1838 and 1849 had a diphtheric character, although the throat was not so decidedly affected as it was in 1820" (when he states that scarlet fever was very fatal in the island, " being accompanied by that peculiar inflammation of the throat, called by Breton- neau diphtherite"). 1841. Becquerel described an epidemic of diphtheria in the Children's Hospital at Paris. (Gazette Medicate de Paris, 1843.) 1848-49. Cases of diphtheria observed in St. Thomas's Hospital, London; in Herefordshire, in Staffordshire, and near Yarmouth."^ 1849. According to Netten Radcliffe, the first epidemic in this century in England was in 1849-50, in Pem- brokeshire. {" On the recent Epidemic of Diph- theritis," Lancet, 1861.) 1850. In October of this year, Dr. Bennett read a paper before the Medical Society of London, detailing cases of undoubted diphtheria in his practice during that year (? in London) .t Diphtheria was epidemic in Norway. J 1851. Dr. F. J. Brown, of Chatham, described cases of diphtheria at Rochester in 1851 ; again in 1852, '53, and '54. § 1853. Diphtheria appeared in Moscow, and was epidemic * Quoted by Netten Eadcliffe {Transactions Epidemiological Society, vol. i., part iii., p. 328). t Op. cit., p. 35. I Op. cit., vol. i., part iii., p. 333. § Transactions Epidemiological Society, vol. i., part i., p. 33. SUMMAEY OF THE HISTOEY OP DIPHTHERIA. 117 m 1855. At the same time it was observed as the most prevalent throat affection of the French army in the Crimea. An epidemic of diphtheria among the military at Avignon. 1854. Three cases occurring in an isolated house ten miles from Colchester were described.^ 1854. Diphtheria broke out in Canton Zurich. 1855. A very virulent epidemic of diphtheria in Paris. 1855-56. The second epidemic in England began (Rad- cliffe) in Cornwall. At Launceston it began 30th September, 1855, coming to a height in August, 1856. In 1856 more numerous epidemics occurred. Several deaths near Spalding, Lincolnshire, in 1858. The disease appeared in Kent, November, 1856. 1855-57. From early in 1855, to March, 1857, a very virulent epidemic in Boulogne, killing 366 persons, including many English. 1856. Diphtheria first appeared in Iceland, breaking out first in Reykjavik, the point of communication with Europe, and thence spreading northward and west- ward. 1857. The relation of albuminuria to diphtheria was first observed by Dr. Wade, of Birmingham {Midland Quarterly Journal of Medical Sciences, 1857), and shortly afterwards in Paris, by M.M. Bouchut and Empis. 1858. An epidemic of diphtheria occurred in the rural parish of Hertingfordbury, Herts, beginning in Octo- ber and suddenly ceasing at the end of the year. There were 53 attacks (14 fatal) in a population of 750. The outbreak is described by Dr. Burden * Ojy. cit., p. 35. 118 • DIPHTHERIA. Sanderson {Epidemiological Society's Transactions, vol. i., part i.^ p. 61), and appears to have arisen under circumstances rendering importation of the infection improbable. 1859. In 1859 diphtheria had become general through- out England. 1861. Diphtheria made its first appearance at Phthlistes (Hellas). 1862. An epidemic in the first quarter of this year in the Commune de Ceyret (Puy-de-Dome) . 1863. A severe epidemic in the arrondissement de Mont- beliard (out of 77 patients 17 died). A very fatal epidemic, October, 1863, to end of 1865, at Louhaus (Saone-et-Loire). Of 2,500 cases 397 died. 1864-67. Leopold Graf {Deutsche Klinih, 1868) gives statistics of 24 cases at Munich, of which 7 were fatal. 1865. An epidemic ravaged the Communes of Fabreges and Saussau (mid-France) from September, 1865, to February, 1866. In the same year an epidemic raged in the arron- dissement of Blaye, and it was noted that the first patient was a sick stranger who had brought diph- theria with him, while the second was a girl in the same house (quoted in the BictionnairG Encyclop. des Sciences Mediccdes, tome xxix., p. 639). 1865. Diphtheria is said to have first appeared in Smyrna. 1865-66. A small epidemic at Schorndorf in Wurtemburg [QQ patients, 23 deaths). The deaths at Pekin in this winter from diphtheria were estimated at 25,000. SUMMARY OF THE HISTORY OF DIPHTHERIA. 119 1866. Becker described an epidemic in Hanover, attack- ing 155 inhabitants out of 487, and killing 29. Similar outbreaks in Kiel, etc. 1866. Lange (in Journ. fur Kinderhraiikheiten, 1869) de- scribed tlie increasingly epidemic character of diph- theria in Denmark. Thus in — 1861 550 cases. 1862 1,220 „ 1863 2,304 „ 1864 5,987 „ 1865 121,826 „ (see also Fig. 40). 1866. A wide epidemic in Kaffraria, described by Lewson {Transactions Epidemiological Society, 1869). 1868. Diphtheria is said to have first appeared in Eou- mania, spreading to Turkey, etc. 1869-70. At Bucharest, Professor Felix describes an epidemic, in which 200 out of 415 patients died. It is noted that the Jewish population of 1,400 almost entirely escaped, probably owing to their comparative isolation in their quarters. 1871. A grave epidemic at Thoury (Loir- et- Cher). Of 21 patients, 16 died. Similar epidemics at Saint Laurent de la Free (Charente-Inferieure), at Vienna (Isere), etc. 1872. Several epidemics in different parts of France. 1875. Dr. Sainton describes an epidemic in the three communes of Bar-sur- Seine, Celles-sur-Ource, and Mussy-sur- Seine (Aube), lasting from 20th November, 1874 to the end of 1875. Out of a total of 5,203 inhabitants, 628 were attacked, and 80 died; the Note. — An excellent account of the History of Diphtheria in Denmark and Germany is given by Dr. J. Carlsen, of Copenhagen, in Janus, i. pp. 48, 161, and ii. p. 1. 120 DIPHTHERIA. mortality among boys was 4-^ among girls ^, and among adults -^. Several other outbreaks, 1874-75, in France (described in JDldionnaire Encydoi?. des Sciences Medicales, p. 642, tome xxix.). 1875-76. An epidemic at Veroli, province of Frosinone (Italie) . In the same year an epidemic in Tunis. 1876. Yeats describes (^'On an Outbreak of Diphtheria in Auchtergaveny, Perthshire, with Eemarks," in Edinburgh Medical Journal, July, 1876) a severe epidemic lasting from March, 1875, to end of June. Out of 1,500 inhabitants, 183 were attacked. Several months previously there had been several fatal cases of croup in a village ten miles to the south-west. 1877. Diphtheria became epidemic at Yokohama. 1879. M. Droumoff {These de Paris, 1879) describes an epidemic in Roumania, in the district of Braila. 1881. Epidemics of diphtheria at Sourabayaand Batavia, in Java. 1881-82. Diphtheria again epidemic at Yokohama. Special Histoey of Diphtheeia m Ajieeica. 1736, Dr. Douglas, of Boston, published an account of the first appearance of a " sore throat distemper " in this country.* Dr. Fothergill, in the first vol. of Medical Observa- tions and Inquiries, vol. i., p. 211 (London, 1771), gives an extract from a letter from Mr. C. Golden, dated Coldenham, New York, October 1, 1753, in which he states that the throat distemper first appeared at Kingston, an inland town of New England, about * Dr. Douglas' treatise is entitled The Practical History of a Neio Epidemical Erujytive Miliary Fever tcitli an Angina Ulciiscu- losa, tfhicJi prevailed in Boston, New England, in 1735 and 1736. Probably there were mixed cases of scarlet fever and dii^htlieria. SCJIMAEY OF THE HISTORY OF DIPHTHEEIA. 121 1735; gradually spreading westward, and not reaching Hudson's Eiver till nearly two years afterwards. "It continued on the east side of Hudson's Eiver before it passed to the westward, and appeared first in those places in which the people of New England resorted for trade, and in the places through which they travelled." 1752. An epidemic of malignant sore throat noted in New York by Father Middleton. 1771. Dr. Bard described an epidemic of sore throat in New York, characterized by false membrane, etc. He emphasizes the difiiculty in detaching the false membrane found post-mortem in the trachea.^ He also contrasted the sudden onset of scarlet fever with the drooping in diphtheria, and speaks of a "few white specks on the tonsils, which in some increased so as to cover them over with one general slough." 1809. Epidemic of malignant sore throat at Philadel- phia. 1826. Epidemic of malignant sore throat at New York, Salem, and Danville. 1831. An epidemic in Philadelphia, in 1831, similar to the one described by Dr. Bard. 1836. An epidemic of malignant sore throat at Orizaba, in Mexico. 1855. Dr. Wynn mentions epidemics at Lima, in 1855 and 1858. (A Paper on Diphtheria. New York, 1861.) 1856. A terrible epidemic occurred in San Francisco, and in other towns of California ; on which Dr. J. V. Fourgeand published a monograph. * Dr. Bard's treatise on Angina Suffocation was piiblished in 1771. In 1810 it was republished in Paris, and was known to Bretonneau. -who quoted from it. J 22 DIPHTHERIA. 1858. Diphtheria epidemic in Albany, N.Y., in 1858. There were 167 deaths in a population of 60,000. The following figures relating to Philadelphia are quoted by Mr. Thomson (see page 96) from the treatise by Professors Meigs and Pepper: — Philadelphia. Number of Deaths from Year. Scarlet Fever. Diphtheria. Croup. 1855 163 _ 265 1856 992 — 268 1857 704 — 256 1858 241 — 292 1859 232 — 312 1860 206 807 354 1861 329 502 304 1862 461 325 258 1863 275 484 444 1864 349 357 455 This is quoted by Mr. Thomson as an indication that in Philadelphia, as in Melbourne, diphtheria was a newly imported disease in 1858-59. It cannot be accepted as such, in view of the preceding historical facts, but it is valuable as indicating the date of maximum of the "early" epidemic ("early" so far as statistical data are concerned) in Philadelphia. Special History op Diphtheria in Norway. For the facts of the following summary, dealing with the history of diphtheria in Christiania in this century prior to 1865, I am indebted to the great kindness of Dr. Bentzen, the Medical Officer of Health of Christiania. He tells me that it was compiled many years ago from unprinted medical reports by the Medical Officer of Health. The facts are given almost in the exact words of the English translation made by Dr. Bentzen's assistant. SUMMARY OP THE HISTORY OP DIPHTHERIA. 123 This exact record is necessary^ in view of its important bearing on the climatic question (see page 148.) Angina Membranacea [Gronjj) in Ghristlania. 1833. Angina trachealis^ a comparatively rare disease in Christiania^ killed several children this year. 1834. At the beginning and end of this year a larger number of children were carried off by this illness than in the last two years. 1845. In the months of July and August some children were attacked. 1846. Specially in January, and in the latter half of the year, several even older children were attacked, usually with a fatal issue. 1847. Angina Memhranacea Diphtheria. {Grou]p). Appeared occasionally. Continued from last year specially during the autumn among both younger and months, older children, but seemed to decrease as measles be- came more frequent, though it appeared now and then throughout the year, more frequently at its end, al- though probably not as fatal as before. 1848. Attacked some children Some persons, adults and during the year, but seemed children, were attacked in not to be so fatal as before, the first and last part of the year. 124 DIPHTHERIA. Cheistiania. 1849. Angina Memhranacea {Group). Biphtheria . Reports are missing. Reports are missing. 1850. Both diphtheria and croup attacked several, both adults and children, but were not very fatal. (See also note on page 116.) 1851. Appeared sporadically, Attacked several adults and claimed some victims. and children, specially dur- ing the autumn and in the first months of the year. 1852. Appeared sporadically. Appeared now and then. Some cases died. but seldom and only spo- radically, in adults as well as children ; was not very dangerous. 1853. Carried off some children, Angina faucium was com- especially about the end of mon in the first months of the year. the year, during the epi- demic of cholera, and was often of an exudative nature. Before this was most fre- quently of an ordinary cha- racter. 1854. Appeared only quite spo- Inflammations of the radically. throat with exudates ap- peared now and then. SUMMAEY OP THE HISTORY OF DIPHTHERIA. 125 Christiania. 1855. Angina Memhranacea [Croup). Diphtheria. Was rare ; three cases Now and tlien all the are stated to have died of year round, but only spo- this illness. radically, inflammations of the throat of an exudative nature appeared. They were but seldom fatal. 1856. Appeared in single cases Appeared singly, not fre- in the autumn, coincident quently nor malignant, with catarrhal affections. 1857. Was in this year more Inflammation of the phar- frequent than in the preced- ing, and is notified in all months of the year except February, July, and August. Most of the cases occurred in December. The total number of cases of which account exists was 10 in the city and 15 in the suburbs. Of these, 16 died. 1858. ynx with exudates occurred all through the year, prin- cipally in the months of March and May. Five per- sons are stated to have died of diphtheria. Within one family, all the eight children were attacked in the course of two months. Appeared frequently dur- ing the last four months of the year. In the National Hospital 17 were treated, of which 10 died. To the Medical Society 32 cases, with 24 deaths, were re- ported. Appeared through the whole year, but principally about its end. To the Medi- cal Society were reported 69 cases, of which 31 occurred in November and De- cember. None of them died. One doctor has mentioned three deaths. 126 DIPHTHERIA. The following additional data are furnislied by Dr. Bentzen : — Cheistiania. No. of Notified Cases of Notified Cases of Diphtheria Year. Ano'ina Membvan- acea (Croup). Diplithei-ia. plus Croup per 1,000 of Population. 1859 62 119 3-42 1860 53 124 3-30 1861 32 84 245 1862 23 285 5-65 1863 31 193 4-07 1864 29 163 3-43 1865 27 126 2-71 1866 27 74 1-83 1867 17 88 1-78 1868 21 82- 1-67 1869 20 53 1-15 1870 17 81 1-51 1871 10 62 1-08 1872 10 48 0-85 1873 6 46 0-74 Notwithstanding the fact that in the early years the notification of cases was not so complete as it subsequently became, it is evident that in 1859, the year in which the record begins, a large epidemic of diphtheria was preva- lent in Christiania, the number of cases in proportion to the population being larger in that year than in any of the subsequent years in the above table, with the exception of 18G2, 1863, and 1864. CHAPTER XIII. INTERNATIONAL REVIEW OP THE PREVALENCE OP DIPHTHERIA AS SHOWN IN THE PRECEDING DIAGRAMS. This review naturally divides itself into three parts — (a) For the American Continent. {h) For Europe, (c) For Australasia. In discussing the source and mode of spread of epidemics^ the greatest difficulties necessarily arise in connection with cities like the American cities and the cities on the European continent^ in which there is a large mass of diphtheria constantly endemic. The epidemics occurring under such conditions may be taken simply to represent intermittent exacerbations of the local infec- tion^ from an accumulation of susceptible persons, or more probably from an increased infectivity of the disease at intervals. Or they may be regarded as due to the importation of a new " strain " of disease ; and there are not wanting, as we have already seen, indica- tions of gradual spread from city to city and from dis- trict to district. The following tabular statement of years of maximum epidemic prevalence in the chief cities in America from which records have been obtained makes this clear : — 128 DIPHTHERIA. -^ -# O CO CO CO O tH tH Cvl CO CO ■TV QD CD GO CO S CO ■ CO 00 6 Ci I 00 CO S cii ^ t^ I T H CO lo ub g CO CO o Ct) i CO CO 00 CO ■ 00 c- c- c^ I ' I CD "^ o CO CO 00 CO CO 00 CO • OD 'X X . CO j t— C- CO C- C- L— -i-H 1 CO o GO O O CO -i 1 T-H 1 UO XC L~- • t- co CO • 1—1 - tH oo CD • CO ^ -^ ■ ck CO . 00 T-H ^ !^ r:=^ ^ CO CO p! c5 IB i^ ® ;:; P t» rt "III - fe 2 '^ +^ -2 & « Hi o cs Co 4^ ri .a PI n:) r-| ~ < m O PI ^Fqp^pqp^pq(l^b^fiO^Fq EEVIEW OP THE PREVALENCE OP DIPHTHERIA. 129 o g CO CM (M CO CO CO CO I CO I .. GO lO CO lO I rk CO 52 CO CO I CO CO tH CO CO (M (M (M C5 O I CO 00 i-H i 00 CM r-l I I I CO o o I I CO 00 I GO »9 in ^ L CO .1 o o Ci Oi I I 02 Ci ! CO CO 00 I 1-1 l> I O xO CO CO CO CD cog °? ^ S5 "? CO -r-i CO t CM CO CO I 00 CM GO CO -^ CO lO I ce 6JD '■+3 1=1 03 03 -p -ta -iiS O o o 03 O © S Ec c3 ce "V W s -tJ ® -f^i O H P^ ll OT 1 CO CO 00 T— 1 I 00 i:~- CO co iH CO CO ::3 rt ^ -^ 2 P^ M 2 .S 03 Pi o 03 r^ '" d K 130 DIPHTHEEIA. The years of greatest epidemic prevalence in European towns and cities are shown in the tabular statement on the preceding page. The chief epidemic years in Australasia are similarly tabulated below : — Maximum Epidemic Yeaes in Australasia. Melbourne 1858-1861 . 1867-8 . 1873-4 . . . 1889-90 Sydney .... 1863^ . . . 1875 .... 1889 Adelaide 1876-78 .... 1890-91 New Zealand 1872-4 1882.1888 . 1892 The evidence derivable from the diagrams_, already fully discussed, and from the preceding imperfect tabular summary of their contents, abundantly justifies several important conclusions. (a) Diphtheria tends to spread from place to place by the ordinary channels of communication. It is impossible in every instance to track the course of the pestilence. This would probably have been easier had it been prac- ticable to plot out the quarterly instead of the annual mortality from the disease for each town. It is almost certain that its spread is not regular, that it radiates in different directions, and that the same city may at the same time in relation to other cities be both giving and receiving infection. {b) There are certain years in which over whole countries, or even over more than one continent, diph- theria is at the same time epidemic. The possible causes determining these pandemics will be considered shortly. (c) Eemarkable differences will have been observed in the amount of diphtheria endemically present in different countries and cities. This amount is smallest in Ireland, next smallest in England and in New Zealand, then comes Scotland. In determining- the relative amount in towns REVIEW OP THE PREVALENCE OP DIPHTHERIA. 131 on the continent of Europe, there is the difficulty that many of them combine diphtheria and croup in their returns. It would appear, however, that Paris and Berlin occupy a supreme position as regards the amount of endemic diphtheria. Munich and Hamburg are not far behind, while in Brussels it is somewhat less fatal. Italian towns also suffer very severely. In Scandinavian cities there is a very great mortality from this disease ; but it prevails chiefly in an epidemic form. Turning to the American returns, it seems probable that in Boston, New York, Chicago, and Montreal, there is even more endemic diphtheria than in Berlin or Paris. A safer and more trustworthy impression of the actual amount can be obtained by a comparative survey of the diagrams than by a study of the average death-rates given in the tables on preceding pages. [d) In places in which there is no large amount of endemic diphtheria the preceding diagrams show a dis- tinctly cyclic character in the recurrence of the epidemics and pandemics. (e) The intervals between these cycles vary greatly. They may be only three or four years. The Gottenburg curve is a good instance of an almost regular cyclic appearance of diphtheria at intervals of ten or twelve years. In Christiania the interval is much longer. In London, when diphtheria became very prevalent in 1858, it was thought to be a new disease. It did not become epidemic again until 1875, and then feebly; and it was not until 1893 that the third epidemic, of which statis- tical records are extant, culminated. New York shows epidemics culminating at intervals of four, six, and seven years respectively; Chicago shows epidemics culminating at intervals of five, fifteen, and six years. Other in- stances can be gathered from the diagrams. 132 DIPHTHERIA. (/) The cycles themselves extend over periods varying from a single year to six or even ten years. In large cities it is usually five or more years before the epidemic has worn itself out. In smaller towns the epidemic mor^ quickly dies down. The recent epidemic in London is a striking instance of protracted duration of a diphtheritic cycle ; and the " true inwardness " of this prolonged duration is understood when we refer to the detailed diagrams for London during this period, showing the relative prevalence of diphtheria in each separate district of London. From these it is clear that diphtheria gradu- ally spreads from one district to another, and that in the course of the one great epidemic the same district in a great city may suffer epidemically more than once (see Figs. 23 and 24). CHAPTER Xiy. ARE THERE INDIGENOUS FOCI OF DIPHTHERIA ? The series of diagrams clearly show that, although the amount of endemic diphtheria varies greatly in different countries, in no town from which records have been obtained is there a complete absence of the disease in a single year since the records commenced * In all, there- fore, diphtheria is an endemic disease. In several in- stances, as in 1873 in Christiania, 1891 in Gottenburg, 1876 in Prague, 1880 in Hastings, and 1867 in Perth, diphtheria almost disappeared. But as a rule a material amount of diphtheria remains endemic in all centres of population. It is perfectly legitimate to suppose that either (a) this local diphtheria is the source from which the epidemic diphtheria is derived, or that (b) the epi- demic diphtheria is imported from the centres in which it is more extensively endemic. That there are favourite endemic foci for the disease is evident. Some of the great cities of America come under this head ; so also do Berlin, Paris, and other cities on the European continent. So far as the last forty years are concerned, we are not, therefore, in a position to solve the problem as to whether epidemic diphtheria originates in certain indigenous foci, analogous to those for yellow fever and for cholera, or whether it represents the blazing up of the embers of diphtheria endemically present in nearly all the cities and towns from which records have been secured. * Some districts of New Zealand (embracing towns) show a com- plete inter-epidemic absence of deaths from diphtheria. See charts, pages 102 and 104, 133 134 DIPHTHERIA. There are reasons for tliinking that in some of its present endemic centres diphtheria was formerly ]3reva- lent only in epidemics at wide intervals. Thus in Eng- land prior to 1856-7 diphtheria appears to have been almost unknown for a generation. Probably it would be more correct to say unrecognised, for commingled with scar- latina anginosa, ulcerated sore throat, quinsy, croup, and cynanche tonsillaris, there must, I think, have been a considerable share of diphtheria. In France, according to Mons. A. Sanne (art. "Diphtheria," Encycl. des 8c. Medicales. Paris, 1884), diphtheria confined itself for a very long time to certain departments of the centre; it then became epidemic in Paris in 1842-3 ; and after this there was a long interval, the disease reappearing in 1855, and finally ended in becoming endemic, with frequent recru- descences. CHAPTER XV. INFLUENCE OP IMPEOVED AND EXTENDED MEANS OP COMMUNICATION. The increased facility and rapidity with whicli in recent years cholera has become pandemic has been the subject of frequent comment. The epidemics in Europe of 1832 and 1849 spread but slowly ; while those of 1855 and 1866 spread with a rapidity previously unknown. The ^^ early '^ epidemic of diphtheria (1855-60) coincided with a considerable development of the means of locomo- tion in the whole of Europe ; and it is not unlikely that this increased the rapidity and the extent of its spread. The same reason may explain why in so many cities diphtheria now remains constantly endemic to a very considerable extent. The influence of attendance of children at school may receive a brief notice in this connection. Most physicians have experienced instances in which personal infection in connection with school attendance has caused out- breaks of diphtheria. Occasionally these are of consider- able magnitude. Mr. Shirley Murphy^ the Medical Officer of Health of the Administrative County of London^ has shown that when allowance is made for the period of incubation of the disease, there is a marked fall in the number of cases of diphtheria notified for the three or four weeks embracing the August-September summer school holiday, the efi'ect being most marked for cases at ages 3-13, the years of school attendance. It isj how- ever, one thing to admit that school attendance in common 136 DIPHTHERIA. with all other means of personal communication between susceptible persons (of whom scholars at elementary schools form a vast majority) is an important factor in determining- the extent of the spreading of diphtheria, and another to attempt to explain, as some have done,, the recent epidemic of diphtheria in England as if it were caused chiefly by the increased enforcement of com- pulsory school attendance.* Mr. Murphy is satisfied to show that school attendance is a means of spread of diph- theria. With that all must agree, and must further agree as to the urgent necessity for additional precautionary measures in connection with the close aggregation of children in our elementary day schools. Others, less judiciously, attempt to explain the entire or the greater part of the recent increase of diphtheria in England as due to school attendance. Such a notion will be corrected by looking at diphtheria from an international view-point. Notwithstanding its enforcement of com- pulsory school attendance since 1871, London even in its maximal epidemic year (1893) presents no more diph- theria in proportion to their respective populations than is constantly endemic in Paris or Berlin. The writer has made personal enquiries of the respec- tive public health oflScials of Christiania, Copenhagen, Stockholm, Berlin, and Hamburg, as to the degree and * In the] review of The Annies Medicus, 1S91 (Lancet, December 25tli, 1897), it is stated : " No one . . . can have watched the recent weekly returns of the Eegistrar-G-eneral as regards the metropolis without feeling that the waste of child-life which is going on in London is a matter calling for the most careful in- quiry. Nearly every one who is conversant with the subject is convinced that the increase has taken place not only synchronously with, but consequent on, the increased aggregation of children at the most susceptible age for diphtheria in our elementarj^ schools." IMPEOVED MEANS OP COMMUNICATION. 137 duration of enforced school attendance, with the follow- ing results : — At Copenhagen school attendance is stated to have been *' always '' compulsory. There has been no increase of its stringency during the last fifty years. At OhrisUania school attendance has been compulsory since the passing of the Act of 12th July, 1848 ; probably in a large measure earlier. Attendance is compulsory from the seventh year to confirmation, which is not allowed before fourteen years old. The age of com- pulsory attendance has never been altered. At Stockholm school attendance has been compulsory for at least fifty years, and there has been no increased stringency of enforcement of attendance in recent years. At Berlin school attendance is compulsory from six to fourteen, and has been so for about 100 years. At Hamburg it has been compulsory for forty to fifty years. Comparing these data with the incidence of diphtheria in the above cities, we find that at Christiania, Stockholm, and Copenhagen, there has been a very great increase of diphtheria since 1880, not associated with any in- creased enforcement of school attendance. The Berlin and Hamburg curves show an increasing amount of fatal diphtheria in the earlier years of the curves, followed by a large decline during the present decennium. Thus the steady enforcement of school attendance has been associated with a declining amount of diphtheria in Berlin and Hamburg in recent years; in the Scandinavian cities the same steady enforcement of school attendance has been accompanied by an increase of diphtheria ; while in London the enforcement (possibly incomplete in the early years) of school attendance since 1871, for thirteen years was not followed by any material 138 DIPHTHERIA. increase of diplitTieria, but subsequently a slow increase of mortality from diplitberia occurred, culminating in 1893 in a death-rate still well below tbe average diph- theria death-rate of most European cities. The only safe conclusion is that school-infection is only a minor cause of the spread of this disease : it forms but one incident in a battle, which by no means determines the issue of the entire campaign. In saying this, I am anxious to make it clear that I entertain very strong views as to the necessity for increased precautions in connection with the return to school of children who have been absent on account of illness of an ill-defined character. The strenuous efforts made by school-teachers and school- attendance officers cause the return to school of children who are undoubtedly a source of danger to their school- fellows, although they may not have had a recognisable attack of diphtheria. The machinery with respect to return to school of absentees requires radical change ; and as this would imply additional medical help, and the free use of bacteriological diagnosis, a considerable in- crease of expense will have to be faced. Assuming that, as will be contended later, wider pandemic influences are at intervals of years in operation, over which we have little control, we are not justified in abstaining from every effort to minimise the action of these wider causes by preventing personal infection. The relative amount of diphtheria in urban and rural districts has a bearing upon the question of the influence of freer inter-communication. When a sufficiently long series of years is taken, as in many of the preceding diagrams, it appears clear that there is more diphtheria in urban than, in rural communities. This might have been expected, diphtheria being an infectious disease, and consequently increased by improved and extended means of personal IMPKOVED MEANS 0¥ COMMUNICATION. 139 commanication. The validity of this conclusion can be tested by reference to the diagrams. Thus the whole of Michigan, which has a large proportion of rural population, has much less diphtheria than the neighbouring city of Chicago; the whole of Massachusetts has less diphtheria than Boston or New York ; the whole of England less than London ; the whole of Japan less than its great towns ; the whole of South Australia less than Adelaide (comparing corresponding periods 1875-96 in Figs. 42 and 43). CHAPTER XVI. THE CONDITIONS DETERMINING THE PANDEMICS OF DIPHTHERIA. Do the accumulation of a susceptible population of children and the accidental introdution of infection from without embrace the entire etiology of an epidemic of diphtheria ? As regards susceptibility, any population containing an average proportion of children is a suscep- tible population; and it is doubtful whether a previous attack of diphtheria very greatly diminishes this suscepti- bility. It is not unlikely that one attack predisposes to another, as in rheumatic fever. Whether this is so or not, in view of the fact that in all European and American cities diphtheria has been constantly endemic to a certain extent during the last forty years, some further cause than the importation of infection is necessary, in order to explain why in certain years diphtheria becomes epidemic: and still more to explain why in certain years diphtheria becomes pandemic. This further cause might conceivably be found in some alteration in the personal condition of a large proportion of the total population, rendering them more susceptible to the disease. For this explanation to suffice, it must be supposed to act over large continents in pandemic years, a supposition which breaks down under the weight thus imposed upon it. The more probable explanation is some change in the external conditions of life, and our search must be directed towards the most likely of these, which are meteorological and telluric. 140 DETBEMmiNG CONDITIONS. 141 The influence of these conditions in different counti'ies will now be tested. It has been impracticable to insert more than a few diagrams illustrating these external con- ditions in relation to diphtheria ; but although I have tested a considerable number of instances not here given, I have not found one which would necessitate a conclusion, so far as that particular place is concerned, contrary to the totally unexpected one stated in the following pages. In London, there is evidence of an epidemic in 1847, a second epidemic in 1857-59 and 1861-64, a very slight epidemic in 1874-76, an increasing amount of diphtheria from 1884, culminating in an epidemic which obtained a firm hold of the metropolis in 1888-89, and after a slight remission for two years, became more extensive in 1892-96 (see Fig. 13, page 30). Compare these facts with the rainfall for a long series of years at Grreenwich. During the 55 years, 1841-95, the average annual rainfall at Greenwich was 24*3 inches. In Fig. 51, the percentage deviation of each year's rainfall is graphically displayed. The years of great epidemic prevalence are shown in black, of minor epidemic pre- valence in columns enclosing small circles, while non- epidemic years are shown by the stippled columns. It is not always easy to distinguish between epidemic and inter-epidemic years ; but the accuracy of the distinction can be tested in each instance by consulting the diagrams of yearly incidence of diphtheria. It is plain that of the eleven years, 1855-64, in which diphtheria was more or less epidemic in England and in London, in only three, viz. 1859-60 and 1862, was the annual rainfall above the average line for 55 years. In the five years 1854-58, the rainfall was extremely light, and it was towards the end of these five years that the epidemic of diphtheria in London reached its maximum. 142 DIPHTHEEIA. The diphtheria curve for London (p. 30) shows a dis- tinct remission in I860, corresponding to the increased rainfall of that year, rising again with the deficient rain- fall of 1861 and 1863-64. The deficient rainfall of 1873-74 was followed by a slight increase of diphtheria in 1874-75. In the thirteen years 1883 to 1895, the rainfall was above the mean line for 55 years, only on three occasions, viz. in 1888, in 1891, and in 1894. During this dry period an epidemic of unex- ampled magnitude prevailed in London. It is evident from the above that there is in London no exception to the rule that diphtheria only becomes epidemic in years in which the rainfall is deficient, though in some years the excess of diphtheria in dry years is but small. There are no instances of a succession of wet years in which diphtheria was epidemic. In London the number of deaths from croup were above the average from 1854 onwards (see Fig. 13, p. 30) ; and in view of this and other facts (see p. 51) it is probable that we must date the London epidemic from the year 1854, though its ravages did not become excessive, and it was not generally recognised before 1857-58. This failure to recognise the disease is in part explicable by the fact that scarlet fever was in the same years prevalent in a fatal form, and doubtless there was some commingling of the two diseases. The above facts bring out more clearly the connection between the deficient rainfall of 1854-58 and the epidemic of the same years. In both the years, 1858 and 1859, in which the curve for England shows diphtheria at its absolute maximum, the river Thames ran so low in summer as to give out a stench,* ^' which was thought to forebode much fever." * Quoted by Creighton from British Medical Journal, l5'ovember 9th, 1861, p. 485. DETEEMINING CONDITIONS. 143 The true significance of the relationship between rainfall and prevalence of diphtheria may be further elucidated by plotting out the yearly winter (October to March) and summer (April to September) rainfall. This is done in Fig. 52. It will be noticed that most commonly in dry centage RAINFALL AT GREENWICH. *^t1oif' Percentage deviation of each year's rainfall from the mean for 55 years (=24-3 inches), from mean O; "^ ^ ^: ^'- S' Si rain- 2 ' faU. + 40 ; Fig. 51. Note.— In this and the subsequent diagrams of rainfall, etc., the magnitude of the epidemic in any given year is not mea- siired by the lengtli of the black columns or of the columns en- closing circles in them. For the actual amount of diiahtheria the diagram of yearly incidence of diphtheria for each place must be consulted. An attempt has been made, however, to distinguish between severe and less severe epidemics by using black columns for the former, and columns enclosing small circles for the latter, though the relative m.agnitude of these can only be ascertained by reference to the previous diagrams. years the summer and the winter rainfall were both deficient in amount^ though this is not uniformly so. The Brighton returns tell the same tale. The average 144 DIPHTHERIA. rainfall in Brig-liton in the years 1868-96 was 29'3 inclies. In 1874 diphtlieria prevailed excessively, the rainfall in that and the foregoing year being below 25 inches. In 1884-85 diphtheria was excessive, with an annual rainfall RAINFALL IN HERTFORDSHIRE. Deviation of rainfall from the mean amount for the years 1842-92, in Hertfordshire, near London, for each summer (N.B. _| i I I f \_) and winter half-year (jV.B. /\/^/), viz., from April to September, and from October to March respectively. The mean -ninter rainfall, 1842-4-3 to 1891-92, was 13'09 inches ; the mean summer rain- fall, 1843-91, was 13-24 inches. Fig. 52. of 26 inches; in 1887-90 the rainfall was deficient every year (the average amount being under 26 inches), most so in 1890, when diphtheria was greatest. In 1893 there DETERMINING CONDITIONS. 145 was again deficient rainfall with, excessive diphtheria (see Fig. 14). At Ceoydon the average rainfall during 1867-94 was 26*3 inches. An epidemic maximum occurred in 1877 with a rainfall of 32"2 inches. This followed on four dry- years, 1873-76, in which the rainfall was 257, 24-1, 26-9 and 26"8 inches respectively. It would appear, therefore, that in this, as in some other instances, dry seasons are provocative of diphtheria, especially when there is a series of them, and that the epidemic may continue and even reach its maximum in the wet year which follows the series of dry years (see Fig. 14). In 1886 the rainfall in Croydon was 27'3 inches; in 1887 it was 22" 7 inches, with an epidemic maximum. In 1890 the rainfall was 22*9 inches, with an epidemic peak. In 1892-93 there was excessive diphtheria with a rainfall of only 24*6 and 19"5 inches respectively. Of the two Yorkshire towns, Sheffield and Bradford, the former shows some relationship between diphtheria and deficient rainfall ; in the latter there has been so little diphtheria that the two cannot be compared. For Sheffield the mean annual rainfall during the twenty-two years, 1872-93, was 30*95 inches. During the first fifteen years of this period the rainfall was only below the average in two consecutive years, viz., 30 per cent, below in 1873 and 22 per cent, in 1874, without an appreciable rise in the death-rate from diphtheria. The next series of dry years was 1887 (40 per cent.) 1888 (15 per cent), 1889 (15 per cent.), and 1890 (17 per cent. below the average). In 1885 the rainfall was average in amount, and in 1884 it was + 16 per cent. Fig. 18 shows that some epidemic prevalence of diphtheria commenced in 1885, gradually increasing until 1892, in which year the rainfall was —4 per cent., while in 1891 it was +6 L 146 DIPHTHERIA. per cent. The epidemic continued to a less extent in 1893j in whicli year tlie rainfall was +21 per cent.* In Beadpord the mean annual rainfall for the twenty- seven years, 1866-92, was 39*4 inches. In 1871, when the diphtheria record begins (Fig. 18), the rainfall was — 24 per cent., being the fifth year in which it was below the average. From that year it was above the average each year except 1879 ( — 12 per cent.), 1884 ( — 6 per cent.), and 1885 ( — 7 per cent.). In 1887 and the consecu- tive years it was — 33, — 12, — 20, — 9 per cent., without any epidemic. What is the inhibitory condition of soil or climate causing Bradford and, to a less extent, other Yorkshire towns, to resist the development of epidemics of diphtheria under conditions which give rise to them elsewhere, I am unable to say. The average annual rainfall at Aberdeen for the twenty- nine years, 1866-94, was 30*7 inches. In 1870 the rain- fall was -16, in 1871 it was -10, and in 1872 +43 per cent., with an epidemic peak in the last year. In 1878, with —9 per cent, rainfall, another epidemic peak occurred after three years of rainfall above the average. In 1884-86 diphtheria was excessive, with a rainfall of — 13, — 9, and —11 per cent.; but there was less diphtheria in the three next years, in which the rainfall was — 14, —7, and —8 per cent. In 1893-95 excessive diph- theria followed a rainfall in 1890-94 of +6, - 7, - 2, -3, and —6 per cent, respectively. At Glasgow the average annual rainfall was 39*45 for the twenty-seven years, 1865-93 (omitting 1884 and 1887, * Since the above was written, I have ascertained the anniial rainfall in Sheffield (Broomhall Park) from 1866-71. In all these years it was at or above the average for the years 1872-93, except in 1870, when it was 26'01 inches ; and during the same period diphtheria mortality was on the decline. DETERMINING CONDITIONS, 147 for whicli years tlie records are defective). From 1866 the rainfall was excessive until 1869-71, wlien it was average in amount^ - 10, and + 4 per cent, respectively. From 1881 onwards it was below the mean line in every year for which records exist, but there are no great epidemic peaks, though evidence of excessive prevalence appears in 1882, in 1886-88, and in 1895. Diphtheria varies comparatively little in amount from year to year in Glasgow, being constantly endemic on a fairly large scale, and climatic conditions appear to play a minor part in its production and continued prevalence. The comparison of Edinbuegh with Glasgow is interest- ing. The former in the fifteen years, 1866-80, had a mean annual rainfall of 28'32, the latter in the same period of 43 '08 inches. During these years Edinburgh had an epidemic peak in 1871, which was the third year in succession of sub-normal rainfall. Glasgow had three years of normal and sub-normal rainfall, 1869-71, and a small peak was reached in 1872. The epidemics of diphtheria are of much greater magnitude in Edinburgh than in Glasgow, but the amount endemically present is greater in the latter than in the former city. In Cheistiania the relationship between deficient rain- fall and excessive diphtheria is very evident. In 1869-70-71 the rainfall was low, and the records show very little excess of diphtheria. In 1875-76 the rainfall was again deficient, without excess of diphtheria. From 1884-1889 inclusive the yearly rainfall was low, and coincident with this, Christiania had the greatest epi- demic of diphtheria from which it has ever suffered. On the other hand, the epidemic maximum in 1862 coincided with a year of rather heavy rainfall, following on two previous years of excessive rainfall ; while there is no statistical record for the years 1851-59, in which the 148 DIPHTHERIA. rainfall was almost uniformly below tlie average, and in most years greatly so. When this volume was preparing for press, it was necessary to send Fig. 53 to be executed, although it apparently embodied an important exception to the rule previously stated, of coincidence between RAINFALL AT CHRISTIANIA. Percentage deviation of each year's rainfall from the mean for the 45 years, 1851-95 (=23-8 inches). Fig. 53. Note. — The years 1857-65 ought all to appear as black columns (see evidence on pages 122-6). epidemic diphtheria and dry seasons. But while writing (Dec. 24th), a letter from Dr. Bentzen, the Medical Officer of Health of Christiania, has entirely cleared up the apparent discrepancy. The exact details furnished by him are stated on pages 122-6. From these it is clear that DETBEMINING CONDITIONS. 149 diphtheria was already slightly prevalent, 1851-54, and that in 1857-58 it became widely prevalent, due allowance being made for the fact that it was only imperfectly recognised as such. In 1859 the number of cases was higher than in any of the subsequent years shown in the table on page 126^ with the exception of 1862-64. There was a slight lull in 1861, a second year of excessive rain- fall j but the epidemic increased in the next wet year and persisted to the end of the next dry period in 1865. The persistence of the epidemic during the comparatively wet years, 1860-64, represents the continuance of the evil produced in the preceding dry years. The epidemic owed its genesis to the conditions prevailing in the long series of dry years, 1852-59, and all the great epidemics, of which we have exact knowledge, appear to have a similar genesis."^ In Germany the same relationship between epidemic diphtheria and deiicient rainfall is visible. For Hamburg I have only returns of rainfall from 1877-1 895j the average annual rainfall for which period was 726*6 millimetres. In 1877-80 the rainfall was much above the average with a relatively small amount of diphtheria. For the next three years the rainfall was small, while in 1884 it was slightly over the average. During these years the amount of diphtheria was increas- ing. In 1885-6-7 the rainfall was exceptionally small, and diphtheria reached its absolute maximum. In 1888-91, with excessive rainfall, diphtheria fell to its * Hirsch (Syd. Sac. Trans., vol. iii. p. 86), says, "Diphtheria began to epidemic in Norway about the same time, first in 1845 in Trondhjem, the year after at Thoten, 1847 at Levanger and Skogn, after that in the district of Namdal, ivhile the general diffusion over the country took place in the year 1855 {Cold, Ugeskrift for Laeger, 1867, Nr. 28). 150 DIPHTHERIA. minimum, rising to a fresh peak in the second of the two dry years, 1892-93. In Beelin the same relationship is visible, though, as in Hamburg to a less extent, the large total amount of diphtheria endemically present makes it difficult to estab- lish it satisfactorily. The years 1875, 1882, 1884, and 1893 were dry years and years of epidemic maxima. At FEANKFORT-ON-THE-MAmE I have only been able to find a record of rainfall from 1855-63. This, however, is interesting as showing that the 1857-8 epidemic peak (Fig. 33) occurred in years in which the rainfall was only 16*65 and 18*43 Parisian lines respectively, after a rainfall of 25-06 in 1855 and of 32*36 in 1856. In 1863 the rain- fall was 19*64 lines, after three years in which it had been 30*15, 25*59 and 26*60 respectively. This was followed by a large epidemic peak in 1864-5. I have no rainfall records for these latter years. Among American records the same difficulty arises as in the curves for European towns. The amount of endemic diphtheria is so great that it is only by considering the apices of the diagrams that the influence of wet or dry seasons can be judged. The New Yoek diagram (embracing the twenty-eight years, 1868-95, with an average annual rainfall of 44*0 inches) shows deficient rainfall in the three years 1875-77, accompanied by a great epidemic in these and in the next following year, in which the rainfall was above the average. In the second of the next three dry years, 1879-81, excess of diphtheria showed itself, reach- ing its maximum in the third dry year, and continuing to a less extent into the following year, in which the rain- fall was a little above the average. A similar experience is repeated with almost mechanical uniformity in con- nection with the two next epidemic peaks. After two DETERMINING CONDITIONS. 151 years of excessive rainfall, the rainfall was below the average in 1884-85, at the average in 1886, while diphtheria increased from its minimum in 1883 to a great epidemic culminating in 1886-88, and therefore continuing Fig. 54. NEW YORK. Percentage deviation of an- nual rainfall from the mean annual rainfall for the 29 years, 1868-95 (=44-0 inches). Note.— The rainfall in 1886 being of the average amount, the large epidemic in this year is indicated by an arrow. CHICAGO, Percentage deviation of an- nual rainfall from the mean annual rainfall for the 25 years 1871-94 (=34-34 inches). into the comparatively wet years 1887-89. In 1893-95 very dry years were associated with a large epidemic of diphtheria. The rapid recurrence of the epidemics in New York lends itself to the supposition that the climatic factor is 152 DIPHTHERIA. here at least of greater importance than the accumulation of susceptible children. It is doubtful if a previous attack confers the slightest permanent immunity from a second attack of diphtheria, and whether the corn- Fig. 55. parative rareness of second attacks in the same person is not due to the diminished liability to attack associated with advancing age. DETEEMINING CONDITIONS. 153 The record of rainfall in Chicago embraces the twenty- five years, 1871-94, during which it averaged 34'34 inches. The epidemic peaks are relatively smaller, while the amount of diphtheria constantly endemic is greater than in New York. So far as the influence of rainfall is evident, there was an epidemic peak in 1876 unassociated with, and not immediately preceded by, a year of deficient rainfall. With this exception the usual association be- tween epidemic peaks and deficient rainfall is seen. The Australian returns also show the relationship between deficient rainfall and excessive diphtheria. Thus, if the diagrams of diphtheria in Melbourne and Adelaide (Fig. 42) be combined with those for the colonies of Yictoria and South Australia (Figs. 41 and 43), of which they are the capital towns, the result is as shown on the accompanying diagram of rainfall (Fig. 55). The great epidemics are plainly associated with years of deficient rainfall. The only apparent instances of an epidemic beginning in years of rainfall above the average are in Adelaide in 1863 (judging by the dia- gram for the whole of South Australia, the diagram for Adelaide only dating from the year 1875); in 1872, where the same remark applies; and in 1890. In the first two of these instances we almost certainly have to deal with epidemics owing their genesis to previous years of pro- tracted deficiency of rainfall^ and continuing into the comparatively wet years. It is not without significance that of the three chief Australian towns, Sydney, with an average annual rain- fall (for 1840-95) of 49" 72 inches, has, as may be seen by a glance at Fig. 42, least diphtheria ; Melbourne, with an average rainfall (for 1857-95) of 26*48 inches, consider- ably more diphtheria ; and Adelaide, with an average rainfall (for 1839-94) of 21*44 inches^ has the largest 154 DIPHTHERIA. amoant of diphtheria, if we compare the period 1875-95 common to the three diagrams (Fig. 42). Dr. Borthwick, of Adelaide, has supplied me with a valuable table of rainfall of Adelaide, with, remarks upon it. From these remarks I make the following extracts as to the three exceptional years named above : — " 1863, wet October " (rainfall only 11 per cent, above mean). '' 1872, good rains in May, June, July " (rainfall only 6 per cent, above mean). " 1890, dry summer, followed by wet winter, wet October to November, very dry December " (rainfall 21 per cent, above mean). The New Zealand diagram (Fig. 44) shows much less diphtheria than that of any part of Australia. In judging of the relationship between diphtheria and rainfall in New Zealand one has to remember that the diphtheria statistics do not deal with single towns, but with sparsely popu- lated and extensive districts. Allowance being made for this, the epidemics of diphtheria show a close relationship with years of deficient rainfall. The records of rainfall for 1878 are missing. Fig. 56 compares the deviations of rainfall from the mean amount in Auckland in the Auckland district. North Island, and the epidemics of diphtheria, with the same data for Christchurck (1867-80) and Lincoln (1881-95), towns within a few miles of each other in the district of Canterbury, South Island. There is no complete record for a single town in Canterbury for the entire period. In Auckland (Fig. 45) the epidemic peaks are but small, but the chief one began in 1873-74, which were average or dry years, and extended into the two sub- sequent wet years. If we take the croup part of the diagram, a smaller epidemic, 1884-85 corresponded to these two dry years, and in 1891 occurred another very slight peak with deficient rainfall. DETEEMJNING CONDITIONS. 155 Canterbury shows mach more diphtlieria than any other district of New Zealand. In the dry years^ 1872-74^ a great epidemic occurred in years two out of three of AUCKLAND. Percentage de-^iation of annual rainfall from the mean annual rainfaU for 1867-95 (=41-23 inches). CHRISTCHURCH AND LINCOLN. Percentage deviation of annual rainfaU from the mean annual rainfaU for 1867-80 (=24-78 inches) and 1881-95 (=27-08 inches). Fig. 56. which were dry, the epidemic being mitigated in the next wet year. A further peak was reached in 1879-80^ 156 DIPHTHERIA. two dry years, following upon four preceding dry years. The year 1887 appears to be an exception to the general rule, and no epidemic followed the three dry years, 1889-91. Mr. Leslie, in sending the valuable figures for New Zealand, which I have used somewhat copiously (Figs. 45-47) remarks : " Why the Canterbury district has so relatively high a death-rate from diphtheria I am quite at a loss to explain." He adds : " The chief charac- teristic of the district is that it is less hilly and broken up than the other provincial districts, the Canterbury plains being far famed for grass, mutton, and wheat. All along the east coast ^ the land is flat, with the exception of Banks Peninsula, and the land rises gradually to the west till the line of the Southern Alps is reached, which forms the western boundary of the district. It is well watered, all the rivers of any consequence taking their rise in the snow-clad Alps, the chief of which is Mt. Cook; but I like to think of it under its Maori name, Aorangi, or rather Anoa-rangi, ''the pioneer of the clouds of heaven." In many places, especially about the city of Christchurch, Artesian wells are so numerous that I am not aware of any other domestic water-supply for the inhabitants." If we add to the above, what might be anticipated from Mr. Leslie's graphic description, that the mean rainfall at Christchurch in Canterbury for the years 1867-80 was 24" 78 inches, at Lincoln, a neighbouring town, for the years 1881-95 was 27*08 inches, while the average at Auckland for the years 1867-95 was 41-23 inches; at WellingtoD, 39-82 inches; and at Dunedin, 36-86 inches; we are prepared to associate the relative excess of diph- theria in Canterbury with its relatively deficient rainfall. * The map of New Zealand should be consulted. CHAPTER XVII. THE EELATIONSHIP BETWEEN RAINFALL AND DIPHTHERIA The towns and districts in wliicli tlie coincidence of excessive annual rainfall and scanty diphtheria^ and of scanty rainfall and excessive diphtheria^ occurs might be multiplied ; but those given in the preceding chapter will suffice to bring out a completely unexpected conclusion. I anticipated results which would agree with the generally accepted belief in the etiological connection hetween damp houses and damp soil and the origin of diphtheria. Although Oertel ^ says " the disease is not affected by either heat or cold, drought or rain," the majority of observers have assumed and believed that wet weather favours the origin and spread of diphtheria. There has, so far as I know, been no attempt to dis- tinguish between the effects o£ wet weather occurring in years which have an average rainfall or are excessively wet throughout, and wet weather occurring during or following upon periods of protracted deficiency of rainfall, as I shall endeavour hereafter to do. It was therefore a great surprise to me to find that the great pandemics of diphtheria, so far as my information extended, had ail occurred during periods of exceptional drought. The following general induction is, I think, warranted by the instances already adduced : — I Diphtheria only becomes epidemic in years in which the rainfall is deficient, and the epidemics are on the * Vol. i. p. 583 of Von Ziemssen's Cyclopcedia of Medicine (English, translation, 1875). 157 158 DIPHTHERIA. largest scale luhen three or more years of deficient rainfall immediately follow each other. Occasionally dry years are unassociated with epidemic diphtheria^ though usually in these instances there is evidence of some rise in the curve of diphtheritic death-rate. Con- versely, diphtheria is nearly always at a very low ebb duriug years of excessive rainfall, and is only epidemic during such years when the disease in the immediately preceding dry years has obtained a firm hold of the community, and continues to spread presumably by personal infection. Apart from the exact records of rainfall and of incidence of diphtheria, upon a comparison of which the preceding general proposition has been based, certain general con- siderations strongly confirm the same induction. The records for India show that diphtheria is a com- paratively rare disease in tropical climates, the main characteristics of which are not only excessive heat, but also excessive humidity of air and excessive rainfall. Furthermore, the fact that diphtheria may be described as relatively a continental rather than an insular disease is in favour of the induction. The curves given in Figs. 1-47 show that diphtheria is immensely more rife in the great cities of the European and American continents than in British and Irish towns, and than in the different districts of New Zealand. The essential diff'erence between continental and insular climates consists in the greater variations in temperature and in rainfall characterising the former ; and the rainfall may be taken as the index of these variations. Dr. F. Waldo remarks * : ^' In fact, there is found to be the general law that the intensity of the oscillations of rain- * Modern Meteorology (Walter Scott, 1893), p. 409. EELATIONSHIP BETWEEN EAINFALL AND DIPHTHERIA. 159 fall increases with the continentality of the region." The following table is extracted from Dr. Waldo^s work (p. 409). It expresses the average maximum and minimum rainfall and the ratio between these two according to the continental distribution, based on the experience of the years 1830-80. Rainfall in Millimetres. Average Average Ratio of Min. Max. Max. to Min. Eastern part of England 599 744 148 North Germany 573 705 1-23 Soutli West Russia 447 570 1-26 South East Eussia 273 384 1-40 TJ. S. America, West Coast .... 379 517 1-38 far Western interior . 483 684 1-42 ,, Eastern interior . . 890 1059 1-20 Thus for every 142 mm. rainfall occurring in the wet periods there are 100 mm. in the dry periods in Western America, while in England the difference is only as 118 : 100. The iyicreased oscillation in amount of rainfall is not the only difference in respect of rainfall between insular and continental climates. There is a difference in the actual amount, whether we take extremes or averages. Thus the average minimum and maximum rainfall is much higher in England, with its comparatively trifling amount of diphtheria, than in most of the continental regions, in which diphtheria is so fatally endemic and epidemic. The induction is further favoured by the fact that where such comparison is practicable, in a given country the amount of diphtheria is greatest in the parts having the smallest rainfall. Compare on this point Bristol with London, or Bradford, Huddersfield, or Halifax with Sussex or Norfolk. Speak- ing generally, in this country towns on the west coast 160 DIPHTHERIA. have mucli less diphtheria than those inland or to the east of the ranges of hills which determine the rainfall in inland towns. The small amount of diphtheria in rainy Ireland confirms the same result; and although Glasgow appears to be a partial exception to this rule, when contrasted with Edinburgh, it is so solely because diphtheria in Glasgow is chiefly an endemic as contra- distinguished from an epidemic disease. The comparison between Australia and New Zealand brings out the same point. The climate of Australia is continental in character, long periods of drought inter- rupted by rainfall of tropical or sub-tropical amount being common. The case of New Zealand is of special interest, as may be gathered from the number of dia- grams which have been constructed for it (Figs. 44 to 47). The contrast between the district of Canterbury and the other districts of New Zealand has been sufiiciently emphasized on page 156. So also has the contrast be- tween Melbourne, Sydney, and Adelaide (p. 153). Denver, in Colorado, is an instance of an extremely dry place, with excessive diphtheria. It may be asked why the consideration of the relation- ship between deficient rainfall and epidemic diphtheria has been confined almost entirely to urban communities. There are good reasons for this. It is almost impossible to obtain statistics of diphtheria for rural districts, stretch- ing over a sufficiently long range of years. Even if they could be obtained, the curve constructed from them would be subject to the violent accidental oscillations necessarily associated with statistics based on sparse num- bers. If wider areas, as counties, were taken, another source of fallacy would be introduced. In a slowly spread- ing disease like diphtheria, largely spread by personal EELATIONSHIP BETWEEN RAINFALL AND DIPHTHERIA. 161 infection, the different districts of a country are not, as a rule, simultaneously affected, and tlie mortality curve constructed on the basis of the combined experience of these districts would be smoothed out, and not represent the actual experience of any particular district. This objection holds in great towns to a much smaller extent. Hence the experience of towns has been selected deliberately, as furnishing more exact data, on a suffici- ently large scale to avoid accidental causes of variation, and enabling the fundamental causes of pandemics to be examined under favourable conditions. If the coincidence between periods of deficient rainfall and epidemics of diphtheria be a constant one, there can be no hesitation in regarding the relationship as causative in character. To test this conclusion, we must inquire as to the occurrence of (1) Periods of exceptional deficiency of rain without the occurrence of epidemics of diphtheria ; and of (2) Wet periods in which such epidemics began. In 1873-74 there was deficient rainfall in London, with only a slight excess of diphtheria. But, as already pointed out, this was a much less protracted deficiency of rain than that characterising the great epidemics of 1857 onwards and 1887 onwards. The same remark applies in a less degree to the 1873-76 deficiency of rainfall at Christiania, and to similar occurrences elsewhere. Instances may be quoted, as at Bradford and probably Leicester and other towns, in which there has been a succession of comparatively dry seasons without epidemics of diphtheria, notwithstanding that a small amount of diphtheria is endemic in these towns, ready, presumably, to produce a destructive epidemic when the conditions necessary for this arise. I am unable to explain what are the local influences in these instances which protect them 162 DIPHTHBEIA. from epidemic diphtheria, or whether, in fact, their com- parative immunity to the present time is not due to accidental causes. Wet periods associated with epidemic diphtheria present greater difficulties than dry periods in which it does not appear. The chief instances of epidemic peaks associated with excessive rainfall are in Chicago in 1876, Adelaide in 1863 (doubtful, see p. 153), and in 1872 and 1890, and Canterbury, New Zealand, in 1887. Assuming the accuracy of the data respecting these exceptional years, it will be noted that the epidemics occurring in them were not of the exceptional magnitude characterising epidemics associated with a series of dry years. It must be remem- bered also that it is not claimed for dry seasons that they are the sole and sufficing cause of diphtheria. The essen- tial causa causans of diphtheria is the invasion of sus- ceptible persons by the Klebs-Loeffler bacillus. This does not, however, exhaust the causation of diphtheria. If we find this bacillus in a house or town, an epidemic of diph- theria therein is not sufficiently explained. The bacillus may constantly be found in the throats of sporadic cases of diph- theria during inter-epidemic periods, and yet no epidemic arises in these intervals, which are sometimes protracted over long years. It may also be found in the throats of attendants upon diphtheritic patients, without any illness on their part. To exhaust the etiology of an infectious disease like diphtheria, we require, — adopting the phrase- ology of the old scholastic logicians, — to know not only the causa causans of the disease (that is, the specific micro- organism) but also the causa efficiens and the causa movens, which must be present in order that the disease may laecome widely prevalent. To use a simple illustration. A drunken man is thrown by a restive horse, and dies from a broken and dislocated neck. The cause of death EELATIONSHIP BETWEEN EAINEALL AND DIPHTHEPJA. 163 may with accuracy be ascribed to the man's drunken- ness, to the restiveness of the horse, or to the fractured and dislocated cervical vertebra. Properly it is ascribed to the conjunction of all these circumstances. A neglect to consider any one of these factors will give a distorted and imperfect view of the cause of the accident and its fatal issue. So it is with infectious diseases, and especi- ally with diphtheria, in which the influence of personal predisposition and of environment is so marked. The bacteriological study of diphtheria, far from being the negation of traditional etiology, is but the logical outcome of it-; and secondary causes still maintain their traditional value as the indispensable concomitants of the Klebs- Loeffler bacillus. The causa efficiens and the causa movens consist of the combined favouring influence of personal constitution and conditions of environment, which Colin has named "le milieu epidemique." * Mr. Justice Fry says : '''' There is, so far as I know, no physical or logical distinction between principal and minor causes or between cause and conditions in the case of two or more constitu- ent parts of a cause, each of which is necessary, and none of which is by itself sufficient." t We have seen that nurses and others may receive the causa causans upon the mucous membrane of their throats, without suffering from the attack of diphtheria which would follow in susceptible subjects. It need not there- fore be the subject of surprise that occasionally when the external conditions are those known to be favourable to the development of diphtheria, an epidemic does not occur. The occurrence on the rare occasions noted above * Article on Epidemie : Dictionnaire Encydopedique des Sciences Medicates. t Nature, Nov. 1st, 1894. 164 ^ DIPHTHERIA. of epidemics when these external conditions are known to be unfavourable, is also explicable on similar grounds. In the illustration previously given, the man might have broken his neck, even if he had not been drunk, or if his horse had not been restive. And yet, in the instance given, these were essential contributory causes of the accident. That they were so is indicated also by the frequency of this particular accident when they are present and its rarity in their absence. Similarly with diphtheria. An occasional outbreak may be caused by personal infection in an overcrowded elementary day school or in a densely populated and insanitary district, where the circumstances favour the acquirement of the property of increased infectivity, even though it be in the midst of wet seasons ; but such instances are very rare. This does not controvert or disprove the statement here advanced that the chief efficient and motive cause of epidemics, and still more of pandemics, of diphtheria is the occurrence of a series of dry years, giving rise to conditions which favour the multiplication and probably at the same time increase the virulence of the diphtheria- bacillus. In advancing this statement, I am aware that the experience on which it is based is somewhat limited. It is, however, as extensive as available data, collected with incalculable trouble, can make it. It is of the essential nature of a cause that one, being aware of its existence, should be able to predict the corresponding result, or conversely be able to state that this result will cease to appear when the cause is removed. The difficulty of doing this in respect of diphtheria is the occasional occurrence of minor epidemics in the absence of the favouring climatic conditions. But these are not great EELATIONSHIP BETWEEN EAIIS^EALL AND DIPHTHERIA. 165 epidemics or pandemics.'^ We may, notwithstanding these, apply the preceding test to epidemic diphtheria, and prophesy that almost certainly the present great epidemic of diphtheria in London will subside, and diphtheria de- cline towards its ^^ normal" endemic amount, when we have completed the present cycle of dry years, and pass into a wet period. A remarkable instance of a similar investi- gation with a similar forecast is given by Dr. Waldo, t quoting from an elaborate research by Professor Briickner of Berne. Bruckner has made use of observations of rain- fall at 321 points on the earth's surface, distributed as follows : Africa, 6 ; Australia, 12 ; Central and South America, 16; North America, 50; Asia, 39; Europe, 198. For most of the stations the data are for the years 1830-85. For the averages five-yearly periods are taken. The results are classified in the following table (see p. 166). * The only instance of a very severe epidemic of diphtheria occiirring in comparatively wet years which I have been able to discover was in 1881-82 in Portsmouth. I have no record of the rainfall before 1880, but in the epidemic years, 1881-82, the annual rainfall was 32'0 and 29"0 inches respectively, while in the imme- diately succeeding years, 1882-85, after the epidemic had disap- peared, it was 29-3, 27'6, 21'6, and 27"5 inches respectively. The annual reports of the medical officer of health leave little doubt that school infection was extensively at work, the defective con- dition of the main sewerage and house drainage, and the unsatis- factory character of the soil on which a large number of new houses had been built, being also adduced as co-operating causative influences. Dr. Sykes' annual report for 1881 contains the follow- ing remarks : " To the congregating of large numbers of children in schools I attribute the wide spread of this epidemic beyond all other epidemics of the same disease. ... A marked diminu- tion of those attacked was shown at the close of the summer holidays." t Op. cit, p. 407 et seq. 166 DIPHTHEEIA. Periods of Rainfall. Deficiency. Excess. Deficiency. Excess. Defici- ency. Europe Asia Australia .... North America . . Central and South America. . . . 1831-40 1831-40 -45 1831-40 1881-45 1841-55 1841-55 1846-55 1841-55 1846-60 1856-70 1856-70 1856-65 1856-65 (71-75) 1861-75 1871-85 1871-85 1866-75 1866-70 (76-85) 1876-85 1876-85 Taken all together 1831^0 1846-55 1861-65 1876-85 — Dr. Waldo, after summarising the facts as to rainfall in the table, adds, "^ It would appear probable that we are now entering upon another period of very low rainfall during 1890-95," "^ This forecast has proved correct, the dry period being still (January, 1898) with us in England. It is impossible exactly to compare such average results with the diphtheria diagrams for particular towns in different countries, which have not been constructed by a similar system of averaging. But with this proviso the preceding table is most suggestive. The period of de- ficiency of rain, 1856-70, evidently coincided in Europe Australia and North America with a period of pandemic diphtheria. No records of diphtheria exist for the cor- responding period in Central and South America and in Asia, but we know that diphtheria does not thrive in the tropical parts of Asia. The next period of excess of rainfall was one charac- terised by local epidemics ; but a rapid survey of all the diphtheria diagrams indicates that there was on the whole a lull in the prevalence of diphtheria during this period. * Op. cit.y p. 411. RELATIONSHIP BETWEEN RAINFALL AND DIPHTHERIA. 167 At the time when Briickner constructed his diagram the next period of deficient rainfall had only arrived in North America (1876-85), and a glance at the American diagrams shows that this was the period par excellence in which diphtheria was fatally prevalent in the States. Since then Europe has had its dry period. It still per- sists in England, and with it an epidemic of diphtheria, which is only equalled by that of 1856-65. CHAPTER XVIII. THE INFLUENCE OP SOIL UPON THE DEVELOPMENT OP DIPHTHERIA. It lias become established almost as an axiom in medical writings that tbere is a close etiological connection be- tween damp bouses and tbe origin of diphtheria. Many years ago Dr. Thursfield showed {Layicet, August^ 1878) that diphtheria hung about damp houses. This observa- tion has been confirmed by others, and it may be that we have in it a partial explanation of the reason for the excessive endemicity of diphtheria in North Wales, where houses are so commonly found built against hillsides, without damp-proof courses, and often without through ventilation. The fact that epidemic diphtheria arises chiefly in exceptionally dry years does not necessarily invalidate this observation. Damp houses mean cold, a great vital depressant, favouring sore throat and catarrh, and thus preparing the way for the inroads of the diphtheria-bacillus. This remark naturally leads to a statement of the distinction which I regard as most important, between rain occurring during a dry year and rain forming part of a wet year. The latter gives the diphtheria-bacillus an effectual quietus, especially when the excessive rain- fall is continued in a second or third consecutive year ; the former produces outbreaks of the disease due in all probability to the diphtheria-bacillus being driven from its normal habitat in the soil, and obliged to take refuge, if we may use the language of volition, in a parasitic life. INFLUENCE OF SOIL UPON DIPHTHEEIA. 169 I have plotted out tlie dates of onset during the last two years of all notified cases of diphtheria and other notifiable diseases in my own district, on a chart giving the daily variations of the chief meteorological data. The association has been so regular that I am now in the habit of anticipating that whenever a rapid fall in the barometer occurs, especially if this is associated with rainfall, a sudden increase in the notifications of infectious diseases, especially of diphtheria and scarlet fever, will occur, due allowance being made for periods of incuba- tion.^ Similar observations, with a similar result, have been made in respect of rheumatic fever, although, owing to the fact that this disease is not compulsorily notifiable, the data are on a scanty basis. Hence the association between wet weather and limited outbreaks of diphtheria, so frequently observed, is one which I can quite confirm. This remark, however, applies only to the sporadic cases which cannot be traced to any source of personal infection. The outbreaks due to personal infection at school or elsewhere or to infected milk, occur equally in dry or wet weather. We may next pass on to the influence of soil. The present consensus of opinion favours the view that a damp cold soil is most favourable to the production of diphtheria. The following extracts will make this clear. Dr. Copeman t says : " Many districts, which, although usually dry, are liable to occasional floods, are remarkably free from the disease, so that it appears that a persistent impregnation of the soil with moisture is of more import- * Dr. M. A. Adams, the Medical Officer of Health of Maidstone, drew attention to this association in his annual report for 1889. t Article on " Tiie Influence of Soil on Health " in Stevenson and Murphy's Treatise on Hygiene and Public Health, vol. i. p. 339, 1892. 170 DIPHTHERIA. ance than fluctuations in the height of the ground water, particularly if these have any considerable range." These remarks appear to imply an antithesis between dry soils and diphtheria, as well as a causative relationship between persistently moist soils and diphtheria, on both of whicli issues I must give an exactly opposite opinion. Dr. (now Sir Richard) Thorne Thorne, while agreeing that " the broad geological features of a district have no known influence on the development or the diffusion of diphtheria," states that he can hardly agree with Hirscb that " the assumption that conditions of soil have some influence in the development of diphtheria, or on its epidemic diffusion, is one that has no warrant." The essence of Dr. Thorne Thome's teaching is contained in the following paragraph : " Soil, and especially surface soil, when considered in connection with relative altitude, slope, • aspect, and prevailing rainfall, has, I believe, concern in the maintenance and diffusion of diphtheria, and has very possibly some relation with its beginnings. Speaking generally, I think that the experience of careful investigations extending over a number of years is to the effect that where a surface soil is, by reason of its physi- cal constitution and topographical relations, such as to facilitate the retention of moisture and of organic refuse ; and where a site of this character is, in addition, exposed to the influence of cold wet winds, there you have condi- tions which do tend to the fostering and fatality of diphtheria, and which also go to determine the specific quality of local sore throat." ^ Dr. C. Kelly, Medical Officer of Health of the Combined District of West Sussex, covering an area of about 524 square miles, with a population in 1889 of 105,520, has shown that when the diphtheria statistics of a series of * Natural History of Diphtheria, p. 17 et seq. INFLUENCE OF SOIL UPON DIPHTHERIA. 171 years are combined, the death-rate per million among the population living on pervious soils (as green-sands and chalk) is 127; among those living on moderately pervious soils (chalk covered with loam and brick- earth), 216; among those living on retentive soils (weald clay and other clayey beds), 454.* The method employed in this and all similar investigations into the geographical distribution of epidemic diseases appears to me to be fundamentally open to objection, though it is, I admit, difl&cult to suggest a substitute entirely free from objection. The diagrammatic method employed in the preceding pages, giving pictorially the death-rate for each single year of a long series, enables one to see almost at a glance the relative amount of diphtheria in large com- munities, concerning which a comparison is desired. It wouldj however, give violent accidental oscillations if applied to rural districts with sparse populations. If this method were applied to the varying districts in West Sussex, it would, I have no doubt, be found that the annual death-rate from diphtheria among populations on either pervious or impervious soils was high in dry years, especially if more than two consecutive dry years occur, and low in wet years. It is difficult to say how much of the excess of diphtheria in populations on pervious soils would remain after correction for age distribution. I do not suggest that this factor explains more than a portion of the difference. It appears to me that the materials are not yet available for an exact comparison of the quantity of diphtheria in different districts, except on a scale sufficiently great to eliminate accidental variations due to relative facility for the introduction of infection, and so on; as when one compares an English with a German or an American city. It is only possible at present accu- * See Elements of Vital Statistics, p. 121, 172 DIPHTHERIA. rately to compare eacli district with, itself, determining tlie periods of maximum and minimum prevalence of dipli- fheria — comparing district with district, or city with city, rather from the point of view of times of incidence of the maxima and minima, than with a desire to weigh or measure the actual amount of the disease. This point leads to a consideration of Dr. Longstaif's investigations, based on the decennial death-rates caused by diphtheria in the different counties of England. His statistics have been continued up to the end of 1895 by Dr. W. E. Smith, the Medical Officer to the London School Board, and the maps accompanying the latter's report furnish a ready means of comparison of each county from 1861 to 1895. But decennial and quinquennial death-rates from infectious diseases do not allow for epidemic influence, and must therefore, as I have already urged, be received with, reserve. Waiving this point, a careful study of the maps shows that between 1861 and 1895 there has, on the whole, been much more diphtheria in the eastern and south-eastern than in the western and south-western districts of England. That is, there has on the whole been more diphtheria in the districts characterized by relatively deficient rainfall. There are exceptions to this rule, especially that of North Wales. Thus : Dijplitlieria : Mean Annual Death-rate per 1000. 1861-70 1871-80 1881-90 1891-95 England and Wales ■18 •12 •16 •25 North Wales •29 •20 •17 •19 Soutli Wales •18 •15 •10 •20 Cornwall •11 •13 •16 •18 In the first two periods, the North Wales death-rate exceeded that of England, in the third it was practically equal to, and in the fourth it was lower than that of INrLUENCE OF SOIL UPON DIPHTHERIA. 173 England. The maps and the statistics on wliich. they are based present many puzzles as to the distribution of diph- theria. Some may represent real facts ; many, I believe, are artificial in character, and can only be unravelled by a careful study of annual in addition to decennial or quinquennial death-rates. If it be agreed that the chief epidemics and all the pandemics of diphtheria of which statistical records are extant occur in exceptionally dry periods of years, there can be little hesitation in believing that the de- ficiency of rain in some way favours the origin of these epidemics. What is the nature of this unexpected association between epidemic diphtheria and years of deficient rainfall ? The analogy of other infective diseases may throw light on this problem. It has been shown independently by Drs. Gresswell and LongstaflF that the yearly mortality from scarlet fever is inversely to the amount of rainfall. Dr. Gresswell has further suggested * that " not only the rainfall of the year, but also that for prior years, has influence on scarlatina.''^ Dr. Longstaff showed t that the chief increases in the death-rates from scarlet fever, erysipelas, puerperal fever, and rheumatism in England and Wales occurred in years of deficient rain- fall. In the Milroy Lectures for 1895, J I showed by elaborate mortality and sickness statistics derived from the general mortality experience of different European countries, from the general notification experience of Scandinavian countries, and from the experience of large general hospitals in England and other countries, that rheumatic fever is an epidemic disease, of which wide- * A Contribution to the JSlatural History of Scarlatina (Claren- don Press, 1890), p. 192. t Studies in Statistics (Stanford, 1891). X Lancet, May 9tli and IGth, 1895. 1 74 DIPHTHBEIA. spread epidemics occur at intervals of a few years, thougli in the intervals it is never entirely absent from most com- munities. I also drew attention to pandemics of rheu- matic fever, particularly those of 1868, of 1874-75, and of 1884. I also showed that in England the epidemic preva- lence of rheumatic fever has always occurred in years of exceptional scarcity of rainfall. In the same lectures I stated : " In the instances where records were available it has been found that when deficient rainfall was, owing* to its seasonal distribution, not accompanied or followed by exceptional lowness of ground water, there was no epi- demic prevalence of rheumatic fever. It is probable that mere lowness of ground water is not the only factor con- cerned in favouring rheumatic fever, but this along with some hitherto unknown factor of temperature of soil or rate of flow of ground water. Whether this is correct or not, it is certain that dryness of soil is favourable to the occurrence of rheumatic fever to an epidemic extent." I must draw attention to the following further remark : " The effect of deficient rainfall is not produced imme- diately. It takes time to develop ; and it is warrantable to assume that the influence of deficient rainfall is exerted as the residt of its effect on the subsoil, this effect usually show- ing itself by a marTced lowering of the ground water." It will save repetition if I quote the following further remarks from the same lectures : " Low ground water must be regarded as leading to excessive rheumatic fever not by any essential causative relationship between the two. The low ground water is an indication of certain con- ditions of dryness and temperature of the subsoil which greatly favour the growth of the telluric contagium of rheumatic fever. This being so, it is conceivable that low ground water, when through collateral circumstances it is unassociated with the required conditions of temperature, mFLUEI^CE OP SOIL UPON DIPHTHERIA. 175 etc., may not be accompanied by an increase of rheumatic fever, though conversely we never find a high ground water accompanying excessive prevalence of rheumatic fever. ■'^ The preceding facts as to diseases which are close con- geners to diphtheria, throw considerable light upon the latter disease in its relation to weather and soil. The hypothesis which appears most fully to meet the require- ments of the case is that a portion of the life-history of the specific micro-organisms of the above specific febrile diseases is spent in the soil, in a saprophytic stage of existence. It is not surprising, if this be so, that ex- ceptional warmth and freedom from excessive moisture of the soil should lead to an increased multiplication of these micro-organisms, and that they should subsequently become displaced from the soil by rainfall, or become aspirated from the soil, when apart from rainfall the barometer falls, or the interior of a house is at a much higher temperature than the subsoil underlying it. The idea that micro-organisms may be displaced from con- taminated soil by means of rain may occasion momentary surprise, as the first thought would be that they would be washed into the deeper subsoil. But this notion ignores the fact that in towns, and even in villages, the area on which direct percolation of rainfall from the ground level can occur is, in these days of impervious pavements and roads, becoming rapidly diminished ; while at the same time the entire ground under dwelling-houses is only in a small minority of instances covered with impermeable material, or with this material in sufficient thickness to exclude the ground-air. Consequently when rain falls there is great lateral as well as upward displacement of ground-air, and the rainfall, in conjunction with the simultaneous aspirating effect of a 176 DIPHTHERIA. lowered barometric pressure^ ensures the entry of ground- air into the house. The only observations in this country as to the con- nection between ground-water and diphtheria are, so far as I know, those of Mr. M. A. Adams, Medical Officer of Health of Maidstone. His researches, and the conclusions based upon them, are given in a paper which should be consulted in full, as Mr. Adams' able argument can only be fully understood by a study of the ingenious and elaborate diagram accompanying his paper.^ His in- vestigation deals with the nine years 1885-93. I am able to comment on it, and on the additional figures derived from his annual report for 1896. The main conclusion is stated as follows: '^Broadly speaking, right through the nine years, from beginning to end, a strict concordance may be traced between soil dampness and diphtheria on the one hand, and absence of diphtheria and soil dry- ness on the other hand." This is evidently in accordance with the generally accepted views as to diphtheria. Stress is laid also on the seasonal incidence of the damp- ness and dryness of the soil. " As long as the soil is well washed by the winter's high tide, and afterwards dried and aerated during the summer's low tide, all goes well, diphtheria is kept in abeyance ; but so soon as these salutary movements are arrested, or their order disturbed, diphtheria gets the mastery, reaching its acme of violence when stagnation is most complete ; and I wish to lay particular stress upon the fact that the virulence of the disorder increases with the stagnation of the soil air." The following table gives some of the data for the years in question : — * Public Health, vol. vii. p. 2 et seq. INFLUENCE OP SOIL UPON DIPHTHERIA. 177 Total deaths from No. of Subsoil Water Levels. notified No. of Year. cases of Diphtheria inches of Rainfall. Diphtheria. Croup. and Croup. Highest. Lowest.* 1879 _ 3 _ i' 27-2 1880 2 10 — 29-0 — — 1881 4 8 — 24-0 — — 1882 7 14 — 27-9 — — 1883 5 5 — 24-5 — — 1884 1 3 19-6 — — 1885 6 9 — 24-6 187-9 212-4 1886 5 9 — 23-0 188-6 211-7 1887 1 2 21-7 186-7 208-6 1888 12 5 21-4 189-8 2033 1889 29 4 — 22-8 193-7 204-0 1890 14 1 24-4 192-2 206-5 1891 14 33 24-9 188-3 201-6 1892 21 — 106 27-3 182-2 204-0 1893 24 — 163 23-3 181-2 203-8 1894 5 3 72 30-8 176-9 199-0 1895 9 72 i ? 186-0 201-2 1896 16 — 54 26-4 179-4 203-9 * In inclies beneath tlie surface of Dr. Adams' laboratory. The estimated population of Maidstone increased from 31,211 in 1885 to 33,555 in 1896. It is evident that the epidemic of diphtheria reached its maximum in 1889; but after two years of slight dimi- nution, a second lower maximum was reached in 1892-93, and a third still lower in 1896. The mean rainfall for the years 1879-96 (omitting 1895) was 24*8 inches. From 1883 the rainfall was at or below this level until 1892. There was no evidence of any considerable excess of diphtheria until 1888. The maximum of diphtheria as estimated by the number of deaths, occurred in the next year ; but the epidemic continued for several years longer, with a marked recrudescence in the next dry year, 1893. Thus the epidemic conforms to the type already frequently described. It began, and probably owed its genesis, to the succession of dry years ; and it continued to a certain extent during several succeeding wetter years. Turn- N 178 DIPHTHERIA, ing to the subsoil water levels, we see that tlie lowest minimum level was in 1885, while the lowest maximum level was in 1 889 ; after this the level of ground water gradually rose. Consequently the epidemic had its genesis in years of deficient rainfall and of low ground water, i.e. of comparatively dry subsoil. The meaning of the ground- water levels, can, however, only be fully understood by a careful examination of the diagram in Public Health, vol. vii. page 3. Dr. Adams very properly lays stress on the maleficent influence (re diphtheria) exercised by an absence or a deficiency of what we may describe as the " spring-cleaning " of the subsoil, which is due to deficient rainfall in the preceding winter months. This failure in the " spring-cleaning " of the subsoil implies a drier soil in the earlier months of the year than occurs in more normal years; and thus presumably morbific germs in the soil, which would ordinarily be killed off, are enabled to survive. Owing to imperfect data, it will be impracticable to consider in detail the circumstances as to ground-water in other places than Maidstone. A curve of monthly level of ground-water in Hamburg from 1883 to 1895 is given in an article by Dr. J. J. Reincke in the Deutsche Vierteljahrsclirift filr offentliche Gesundheits- pflege (Band xxviii. Heft 3, 1896). Comparing this with Fig. 29, it is evident that the average level of the ground- water gradually sank from 1883 to the end of 1887, and that in 1886, and still more in 1887, the usual ^^spring- cleaning" of the soil failed almost entirely. These were the two years in which diphtheria was more fatally epidemic than at any other time. At Berlin a curve of the monthly ground-water levels from 1873 onwards (not reproduced) shows a minimum level in 1873-75. Fig. 30 shows an epidemic peak in INFLUENCE OF SOIL UPON DIPHTHERIA. 179 1874-75. In 1877 the ground-water fell from the level of the foregoing year, and diphtheria rose in amount. In 1881-82 the ground-water was lower than in any of the preceding years, the spring rise entirely failing to appear in 1882, The years 1882-3-4 had a higher death-rate from diphtheria than any other year. In 1889-92 the ground-water was again very low, with only small peaks of epidemic diphtheria in 1890 and 1893. In England very few ground-water observations are made, and but few instances can therefore be cited. At Croydon the level of the water in a well in the chalk at Wickham Court has been kept for a long series of years. The very remarkable dip in the level of the ground- water in the three years, 1874-5-6, was associated with an epidemic of diphtheria, which began in 1876, and lasted through the next two years, the ground-water during this time rapidly rising. After 1884 the level of the water in this well steadily fell, and the conditions are probably not due to natural conditions, but to over- pumping from the chalk. At Brighton a chart of monthly levels of the water in the chalk subsoil shows the same relationship between excessive diphtheria and low ground-water, in 1873-76, in 1887-88 and 1890, and in 1893. In most districts of London there is but little gravel overlying its clay subsoil. Consequently there are no continuous ground-water observations. There are, how- ever, observations of the monthly flow of the Rivers Thames and Lea, which derive their water in part directly from surface-water and in part from springs. These observations show the same relationship between dry seasons and excessive diphtheria, as before. In 1858 and 1859 we have already noted that the Thames was so low that the mud gave out a stench. From this came, not the prophesied " fever,'' but diphtheria. 180 DIPHTHERIA. Fig. 5". Mean annual discharge over Feilde's Weir on the River Lea. Percentage de\'iation from the mean annual discharge for the entire period, omitting the years 1870-71-72 and 1878-79, for which there are no records. Note. — All the years in which the discharge was below the mean are indicated by black columns. Most of these were epidemic years, but Fig. 13b should be consulted for exact details. INFLUENCE OP SOIL UPON DIPHTHERIA. 181 In the following diagram the discharge over Feilde's Weir in the Eiver Lea, at a point in Herts a little above where a large portion of the Metropolitan water supply- is derived from it, is shown. The close connection between years of deficient dis- charge of water at Feilde's Weir and the occurrence of epidemic diphtheria is very evident. The longest period of deficient water in the river Lea is 1854-5-6-7- 8-9. This corresponds to a great epidemic of diphtheria in London. In 1860 the amount discharged over the weir on the Lea was much above the average. In that year diphtheria declined in London (see Diagram, p. 12). Then during the four years, 1861-2-3-4, the dis- charge of water was again deficient, and diphtheria once more was rampant. In 1868-69 there was some deficiency of discharge of water, but no repetition of the epidemic. In 1874 a slight epidemic coincided with deficient flow of water; and the deficient flow in most years since 1884 has coincided with an enormous epidemic of diphtheria. This can be followed out in greater detail in Fig. 58, dealing with the monthly flow over Feilde's Weir from 1880 onwards. It will be observed that from 1887 the average dis- charge of water did not exceed 200 million gallons in any month, except in December, 1891, and in February and December, 1892; while before 1887, with the exception of the two dry years, 1884-85, it was regularly larger in amount. We have here an indication of a similar failure of the subsoil to receive its spring-cleansing, which was noted in Maidstone. Similar data derived from two different sets of observations (the corresponding diagrams being on dif- ferent scales) are available for the river Thames. The first (Fig, 59) deals with the monthly flow of the Thames 5 1? !:X) S SS S o" ^ to ft o <1) ^ a CD p ,^ f> ffi) >> a2 rt a: 3 ^^ S J) 184 DIPHTHERIA. at Thames Ditton, 1853-73. The exceptional lowness of the river from 1853 to 1859 inclusive is plainly seen, being greatest in 1854-55 and in 1858-59. The lull in the great epidemic of these years corresponded in 1860-61 to a very great increase of flow of water, the epidemic resuming its course as the flow of water diminished in the succeeding years. The small flow of water 1870-71 was not sufiiciently protracted to produce epidemic diph- theria. The second set of observations is derived from Ted- dington Weir on the Thames, and is complete from 1883-97. An increasing amount of diphtheria coincided with diminished flow over this weir in 1884-86, while the diminished flow from the middle of 1887 to near the end of 1891 was associated with epidemic diphtheria, as was still more the scanty flow of water from 1893 onwards. The occurrence of floods renders river-gaugings a somewhat inaccurate test of the true level of ground- water, and it is probable that floods account for some of the sudden rises in Fig. 60, as, for instance, in February, 1893, a year in which diphtheria reached its absolute maximum. "With the imperfect data available it is impracticable to pursue the relation of ground-water levels to diph- theria further. There is great need for the general establishment of stations for observation of ground-water levels and ground temperatures in this country. With each additional year the value of such observations rapidly increases ; and a systematic series of such ob- servations will in a few years serve to fill in the gaps of the evidence here set forth. We may summarize our observations on the relation between rainfall and ground-water and the rise of epidemic diphtheria in the following propositions : — INFLUENCE OF SOIL UPON DIPHTHEEIA. 185 1. An epidemic of diphtheria never originates, in the Tt^r-rff^^^Siti: „_ v~~— — -V — .- o CO CO CO c- CO •* T m Zil d (^ o ^ o ^ o ^ u ^ tr 1—1 rt s o r^ w oil >1 ci -P i^ o rr >1 fl. +J o ^ r-i rt I-: ou eg i-H 186 DIPHTHERIA. towns and countries in whicli I liave been able to collect facts, when tbere has been a series of years in which each yearns rainfall is above the average amount. 2. An epidemic of diphtheria never originates or con- tinues in a wet year [i.e. a year in which the total annual rainfall is materially above the average amount), unless this wet year follows on two or more dry years immedi- ately preceding it. 3. The epidemics of diphtheria, for which accurate data are available, have all originated in dry years [i.e. years in which the total annual rainfall is materially below the average amount). 4. The greatest and most extensive epidemics of diph- theria have occurred when there have been four or five consecutive dry years, the epidemic sometimes starting near the beginning of this series, at other times not until near its end. 5. Dry years imply low ground-water, and we find therefore in the years of epidemic diphtheria that the ground-water is exceptionally low. The exact variations in the ground-water which most favour epidemic diph- theria cannot with the data to hand as yet be stated ; but it is probable that when this is cleared up, it will become clear why in exceptional years which have a deficient rainfall epidemic diphtheria is either absent or but slight. The preceding propositions enable us to formulate a worhing hypothesis of the causation of diphtheria. The specific micro-organism of this disease has a double cycle of existence, as have the specific micro-organisms of enteric fever, erysipelas, scarlet fever, rheumatic fever, etc. One phase is passed in the soil, another in the human organism. One is saprophytic, the other parasitic* * It is not contended that there is a regular alternation of saprophytic and parasitic generations ; but that such alternations do occur. INFLUENCE OP SOIL UPON DIPHTHERIA, 187 It is not strange, therefore, that the epidemic prevalence o£ all the above diseases is favoured by deficient rainfall, if this is sufficiently long continued. This deficient rain- fall implies a low subsoil water, and a subsoil above the level of this water, which is relatively dry and warm, probably the optimum conditions for the saprophytic life of the above pathogenic micro-organisms. The causes of the transition of the diphtheria-bacillus from the sapro- phytic to the parasitic phase of life may be surmised both as regards (a) season and (6) years of special epidemic prevalence. Diphtheria is most prevalent in autumn and in the early winter months, when the optimum tempera- ture and the optimum degree of humidity of the soil are rapidly disappearing or have departed. It is also most prevalent after the wet weather occurring in or immedi- ately following exceptionally dry years. Both these con- ditions tend to raise the ground-water and to drive out any pathogenic micro-organisms from the soil. The preceding working hypothesis may appear to give undue importance to climatic conditions as contrasted with personal infection. It must be remembered, however, that one of the main objects of this book has been to show, from exact official statistics, the gradual spread of diph- theria from town to town, and from country to country. This has been indicated in repeated instances. There is no reasonable doubt that 'personal infection is the chief means by which diphtheria is thus spread. Personal infection does not, however, explain why in some years diphtheria, although present in a district in an endemic form, does not spread ; while in another year, in which only the same opportunities of personal infection occur, it becomes extensively epidemic. Still less does it explain the occurrence of widely scattered epidemics and even pandemics in certain years. To explain these the 188 DIPHTHERIA. operation of wider general causes must be pre-supposed. It might be tliat the susceptibility of entire populations to the infection of diphtheria increases at times, though this is improbable ; or it might be that the diphtheria-bacillus under certain conditions becomes more actively virulent and infective — more remote from its saprophytic phase of life ; and that, thus, persons who can resist the ingress of the feebler, fall victims to the more powerful micro- organism. The latter is probably the correct hypothesis ; and the evidence already given clearly points to the con- clusion that of the external cultural conditions leading to increased virulence of the diphtheria-bacillus and greater readiness for assuming a parasitic life, exceptional de- ficiency of rainfall and consequent exceptional deficiency of moisture in and exceptional warmth of the subsoil form an essential part. CHAPTEE XIX. CONCLUDmG EEMAEKS. It appears desirable^ in conclusion, to state in a few words what degree of confidence is claimed for the facts and arguments set forth in the preceding pages. A care- ful distinction must be drawn between facts and argu- ments, as they evidently stand on a different footing. The degree of accuracy claimed for the statistical facts pictorially represented in the sixty figures scattered throuffhout the text has been already stated. After spending many laborious hours for several years past in the critical study of these figures, I am confident that they tell the essential truth. The changes produced by alterations of nomenclature and other causes are of minor importance; and the peaks and valleys of the preceding diagrams represent real epidemics and real inter-epidemic periods. Absolute accuracy is impossible ; and it may be that here and there errors have crept in. Such errors will not, I believe, in any single instance alter the state- ments made in the previous chapters as to the years of epidemics and of intervals between them. Although absolute accuracy has been aimed at, approximate accuracy amply suffices to bring out the essential character and meaning of the great facts set forth in the preceding sixty diagrams, and such approximate accuracy has, at the least, been attained. The arguments based upon the statistical premises of the preceding chapters and the general law, — an at- tempt to establish which has been made, — invite criticism, and they are submitted with that object. An attempt 190 DIPHTHERIA. has been made to examine epidemic dipHthena from a view-point unprejudiced and unbiassed by tbe accepted teaching upon the subject. Not many years ago the- mention of the word " diphtheria," or even of the less exact word, " sore throat/' created a presumption in the minds of the majority of medical men that drain-poison- ing was operating. (Possibly it was, in a minor sense, by increasing the vulnerability of the patient.) Now there has grown up an equally erroneous general im- pression that the occurrence of diphtheria — apart from personal infection, which all agree is a most potent factor — is intimately associated in its origin with damp- ness of soil, and that ceteris ijarihus diphtheria may be expected to be more rampant, for instance, on clayey or water-logged soils than on soils that are pervious and dry. I have given my reasons for disagreeing toto coelo with this conclusion. These reasons must in their turn receive criticism, and as the result of further accumulation of confirmatory or contradictory instances, it will be proved or disproved that an essential factor in the production of epidemic and still more of pandemic diphtheria is the occurrence of a succession of years of deficient rainfall, associated as they must be with abnormally dry conditions of the soil and subsoil. Every general truth in science is hedged round with exceptions and modifications. Instances have been quoted in the preceding chapters which seem to prove the reverse of what the general law states, or to make the statement of it appear inaccurate. Such exceptions do not show that the general law is untrue, or that for practical pur- poses we cannot forecast the future of diphtheria on a national and international scale from this general law. It is a safe rule that winter is colder than summer. If the mean temperature during the next summer should be COXCLTJDma EEIIARKS. 191 lower than that of the present exceptionally mild winter^ the rule will remain true in the majority of cases. So with diphtheria. In the majority of instances epi- demics occur under the climatic conditions previously detailed; and the occurrence of a marginal fringe of doubtful or even occasionally of minor contradictory instances does not diminish the value of this general law as applicable to the majority of epidemics. The main conclusion stated in the preceding pages may on the first blush produce a depressing sense of comparative helplessness^ such as one feels in connection with the return of epidemic influenza since 1889. If the occurrence of pandemics of diphtheria is governed largely, if not chiefly, by meteorological conditions over which we have no control, what scope is there for the intervention of preventive medicine ? Before answering this question, let it first be clearly stated that we must have the truth at any cost. If the statement of the truth by implication means that our preventive measures are but Canute-like attempts to stop the inflowing tide, still it is well that the truth should be known. But this is not a correct view of the case. Diphtheria is spread chiefly by personal infection. This personal infection is immensely more potent in epidemic that in inter-epidemic years — a fact which should lead to redoubled efforts to prevent personal infection during such epidemic periods, rather than to a fatalistic inertia. Similarly redoubled efforts are re- quired to prevent ground-air from gaining admission into houses, and to render more wholesome the soilin districts in which diphtheria has become endemic. How this can be done in towns, how the soil can, without more open spaces than are obtainable in most of our great cities, be made to resume its virgin salubrity and purity cannot be 192 DIPHTHERIA. stated here. It is one of tlie greatest problems of public health. But to assume that because we do not yet know how to exterminate diphtheria, or because we cannot hope in our day to be entirely successful in preventing its spread, it is therefore useless to attempt anything, would be as unwise as it would be for a city council to dismiss their fire brigade staff and dispose of their fire-preventing apparatus, because the staff had not been successful in at once extinguishing every fire, or because the city council were impressed with the fact that the present appliances for extinguishing fire are of a very imperfect character. INDEX OF NAMES OF PERSONS Adams, M.A Alaymus Arnaiilt Baillou Baker . Bard . Baronius Becker Becquerel Bennett Bentzen Bertillon Bortliwick Bouchiit Bretonneau Brown, F. J Bruckner . Carlsen Carnevale Cedremus Chomel Copeman Crawford Douglas Droumoff Fagge Forestus Fothergill Fourgeaud Ghizi Greenhow Gresswell Grimsliaw Goseti Hecker Herrera Hirscii . 176 Hoffmann . . 113 Home . 113 Hoskins Huxham . 112 . 27 Janssens 114, 121 Juraschetz . . Ill . 119 Kelly . . . 116 Kittel . . 116 Korosi 6,122 . 6,62 Lange . 6,154 Leslie, G. . . 117 Lewson 113, 115 Linroth . 116 Longstaff . . 165 Macrobius . 91, 119 Mantez . 112 Martinez . . HI Mercado . 113 Middleton . . 169 Murpty, S. . . 114 Oertel . . 120 . 120 Peacock Pistor . . 113 Potzniakoff. . 112 113, 120 Eatcliffe, N. . 121 Ravizza Eeincke . 113 Eosen . . 52 Eoux . . 173 6 Sanderson, B 6 Sandwith. . Seaton . Ill Simon . . Ill Smith, W. E . 110 Starr . 6 . 114 . 116 . 114 6 6 . 170 . 114 6 . 119 5, 100, 156 . 119 7 . 172 . 110 7 7 . 112 . 121 . 135 . 157 . 49 7 7 53, 116 • . 7 7, 73, 178 114 63 117 7,92 52 52 172 113 193 194 INDEX OP NAMES OP PLACES Sykes, J. F. J. Symons Thompson, A. Thomson, "W. Thorne-Thorne Thursfield . . 54 7 7 96, 122 . 170 . 168 Von Word Wade . Waldo, F. Wilbur Wynn . Yeats . PAGE 111 117 159 100 121 120 INDEX OF NAMES OF PLACES Aberdeen . 44, 146 Cape Colony .93 Adelaide . 98, 153 Cardiff . 35 Africa 92 Chicago 20, 153 Alabama . 27 Christiania. 86 , 122, 137, 147 Albany 122 Cincinnati . . 18 Alexandria . : 92, 112 Colchester . . 117 America^ North . 9 Cologne . 112 America, South . . 23 Connecticut . 27 Amsterdam . 68 Constantinople . . 112 Antwerp 70 Copenhagen 88, 105, 137 Auckland . 101 Cork . . 47 Avigsburg . , 112 Cornwall . . 113, 117 Avignon 117 Crete . . 114 Australia . '. 96, 153 Crimea . 116 Australia, South. , 96 Croydon 33, 145, 179 Aiistro- Hungary 79 Dantzig . 112 Baltimore . , , 16 Denmark . . 88, 119 Barcelona . 65 Denver . 20 Belfast 47 Dresden . 74 Belgium 68 Dublin . 47 Berlin . 73, 106, 137, 150, 178 Dundee . 44 Birmingham 35 Bordeaux . 63 Edinburgh . . 44, 147 Boston . 11 England ... 27 Boulogne . 117 Bradford *• 41, 145 Faroe Islands . . 90 Breslau 73 Florence . 66 Brighton . 33, 141, 179 France . 62, 115 Bristol 35 Frankfort-am-Mai] ae . 73, 150 Bromley 113 Brooklyn . 14 Geneva . 65, 116 Brussels 70 Germany . . 73 Bucharest . '. 83, 119 Glasgow 44, 115, 146 Buda-Pesth 79 Gottenburg . . 87 Buenos Ayres • 23 Greenock . Greenwich . . 44 . 113, 141 Cairo . . 92 Guernsey . . 116 Calcutta 93 Canada , , 23 Hague, The . 68 Canterbury 101 Halifax . 41 INDEX 01" NAMES OE PLACES 195 Hamburg 72 Hastings Hawke's Bay Holland Huddersfield Hull . Iceland India Indiana Ireland Italy . Japan . Java . Kaffraria . Kelso . Kent . KiliTiariiock Kioto . Kobe . La Ferriere. Lancashire . Launceston Lausanne Leeds . Leicester Leipzig Lima . Limerick Lincolnshiri Lisbon. Liskeard Liverpool Lornbardy London 30,49,55,108, London, Districts of Londonderry Lyons . Maidstone . Manchester Mantua Mariedwerder Marlborough Marseilles . Massachusetts Melbourne . Mexico Michigan . Milan 37, 149, PAGE 178 , 33 102 68, 111 112 41 • 41 117 93 27 47 113 66 94 • 120 119 115 115, 117 44 95 • 95 115 38 117 65 41 37 73 '. 116, 121 47 117 . 65, 115 113 38 112 16, 141 179 56 47 '. 63, 115 177 38 112 , , 114 , , 104 , , 63 , , 9 . 97, 153 121 21 68 Minnesota Montreal Moscow Munich . 27 . 23 82, 116 78, 118 Naples ... 112 Nelson .... 104 Newcastle-uiDon-Tyne . 42 New Jersey ... 27 New Orleans ... 20 New South Wales . . 99 New York . 12, 27, 114, 121, 150 New Zealand . . 100, 154 Northumberland . . 42 Norivay 85, 105, 115, 116, 122 Ohio . Oldham Ontario Osaka . Otago . Padua . Paisley Paris 62, 112, 113, 114, 116 Pekin . Pembrokeshire Perth . Philadelphia . 16, 121 Pittsburg . Portsmouth Portugal Prague Pi^ovidence . Queensland Peykjavik . Rhode Island Rochester Pome . Rotterdam Eoumania Russia Salford San Francisco Scandinavia Scotland Sheffield Sicily . Skien . Smyrna 27 38 27 95 103 114 45 117 118 116 44 122 . ' 16 33 65, 113 . 81 . 11 . 99 . 117 . 27 . 116 68, 110 . 68 119, 120 . 82 . 38 . 121 . 86 44,114 41, 146 . 112 . 115 . 118 196 INDEX OP NAMES OP PLACES Southampton South Shields Spain . St. Helens . St. Louis St. Petersburg- Stockholm . Siveden Switzerland Sydney Taranaki . Tokio . Tours . Trieste Tunis . Turin . . 33 . 42 65, 112 . 38 . 18 . 82 88, 137 . 114 . 65 99, 163 102 96 115 79 120 68 Utrecht Venice Victoria Vienna Warsaw Waterford Wellington Westland "Wolverhampton Wurtemburg Yarmouth . Yokohama Yorkshire Zurich PAGE . 114 . 68 . 96 81, 119 . 82 . 47 . 102 . 104 . 37 . 118 . 116 95, 120 . 41 114, 117 Butler & Tanner, The Selwood Printing Works, Frome, and London. DATE DUE ■W 'L^Wi- T!T 134 m DEMCO 38-296 ]iJ47 IIC153 .©*■ *i -._ o (i,.