While Watt’s theory of the working of this principle of substitution is open to criticism on some points of detail, the law itself, as enunciated by him, remains to the present time one of the soundest and most instructive generalities in epidemiology. He based it upon a laborious search of the burial registers, such as no one before him in this country had undertaken. Next he saw correctly that a great rise in the deaths of infants by such a disease as measles could only be accounted for by a great increase in the rate of fatality. Thirdly, he connected the loss from measles with the saving from smallpox. Adopting an old opinion, which may be discovered in Willis[1219], he argued that smallpox, when taken first, served to fortify children so that they passed easily through the measles afterwards; but in the following passage he indicated a better reason why the absence of smallpox gave measles the chance of proving more fatal: “In this point of view we are not to consider the smallpox as so peculiarly fatal in their nature. They perhaps prove so fatal merely by having the start of other diseases. The measles, the chincough, the croup, the scarlet fever, and perhaps many others, would have proved equally fatal had they occurred first.” The principle is true to this extent, that a certain proportion of weakly infants, or children of poor stamina, will succumb to almost any disease—if not to smallpox, then to measles, and if not to measles directly, then to the sequelae of measles. This was perceived in the form of a necessary truth by Haygarth in 1793: “A considerable number of those who now die of the smallpox would die in childhood of other diseases if this distemper were exterminated[1220].” It was commonly believed that smallpox had at length found its real artificial check, not in the inoculation of itself, but in the inoculation of cowpox. At all events it had declined greatly in Glasgow. During the three years before the measles epidemic of 1808, there could hardly have been more than a thousand children attacked by smallpox, or not one in ten of all the children born. During several years the infancy of the city had been spared any great ordeal of infectious disease; the first epidemic that came along happened to be measles, so that it fell to that infection to take off the weaklings. In the economy of nature it is impossible to rear all the young of a species, nor would it be good for the species if it were possible. It is among the birds that the principle of population, or of the survival of the fittest, is seen working in the most admirable way: the annual migration of many species to breed in a remote country brings with it an ordeal for the birds of the year in finding their way to the winter feeding-grounds—an ordeal which only the strongest come through. For some unexplained reason, the young of the human species are peculiarly tried by infectious diseases, which multitudes pass through safely, while many of poor stamina or of ill tending are cut off.

Two remarks remain to be made under the doctrine of displacement. The first is that the substitution of measles for smallpox was one of a series of such changes in the public health of Britain. The great infective scourge of medieval and early modern periods had been plague, which destroyed at times immense numbers of the valuable or mature lives. Its successor was typhus fever, which also cut off the parents more than the children, but did not retard population as the plague had done. The saving of life by the extinction of plague was in great part balanced by the loss from smallpox, which fell, however, more and more upon the earliest years of life until at length it was almost confined to them. The first great decline of smallpox itself corresponded to a great decline of typhus fever during the second half of the French war; but while there was no great infectious disease in those years to thin the ranks of the adults, measles took the place of the more loathsome smallpox in cutting off a certain number of young lives. While the older types of infection have disappeared, the incidence has shifted from mature lives to children, so much so that at the present time enteric fever, and occasional choleras and influenzas, are almost the only infections that correspond to the old plague and to typhus fever in their age-incidence.

The other remark is that the greater prevalence or fatality of measles, as if in lieu of smallpox, meant a good deal more for the bills of mortality than actually appeared under the name of measles. Smallpox was not an infection that did much constitutional damage to those that came through it, although it sometimes destroyed the vision and spoiled the beauty of the face. On the contrary, it was held by many that the general health was better after an attack of smallpox than before; and, if personal experience can justify an opinion, that ought to be my own view of the matter[1221]. But measles is an infection peculiarly apt to leave mischief behind. The bronchial catarrh, which is an integral part of the malady, and is often the cause of death in the second stage of the attack, may so affect weakly children that the respiratory organs are permanently damaged. Tuberculosis of the lungs is apt to follow measles. Some children, again, fall into mesenteric disease after measles, and die tabid, the intestinal catarrh being as dangerous in one way as the bronchial is in another. Another large proportion of the subjects of measles take whooping-cough[1222]. While smallpox did its work summarily, the full effects of measles were longer in being realized. This may in part explain the fact brought out by Watt, that while fewer children died under two years of age, measles being the dominant epidemic disease, there was an increase in the ratio of deaths from all causes between the years of two and five and from five to ten.

Measles in the Period of Statistics.

The history of measles for nearly a generation after the great epidemics of 1808 and 1811-12 is little known. No one in Glasgow continued Watt’s laborious tabulation of the causes of deaths in the numerous burial registers[1223]; nor was any regular account kept elsewhere except by the Parish Clerks of London. The following deaths by measles in their bills from 1813 to 1837, when the modern registration began, were probably no more than from a third to a half of the deaths in all London:

The inadequacy of these figures to the whole of London will appear from the fact that the registration returns under the new Act gave for the last six months of 1837 the measles deaths at 1354, while the bills of the Parish Clerks gave them at 577 for the whole year. But the old bills enable us to compare the deaths from different diseases within the same area and under the same system of collection, and to compare the deaths “within the bills” in a series of years since the last of the new parishes were taken in about the middle of the 18th century. Using the bills so far legitimately, we find that measles at length came to be of equal importance with smallpox itself as a cause of death in childhood, and that it had become a larger and steadier total from year to year.

So far as concerns Glasgow, the high mortality from 1807 to 1812, making 10·76 on an annual average of the deaths from all causes, was not maintained. When the tabulation of the causes of death was resumed from 1835, the annual average of measles for the five years ending 1839 was found to be only 6 per cent. of the deaths from all causes, the average of smallpox having come back to 5·3 per cent. During that unwholesome period, in which there was much distress among the working class and a great epidemic of typhus, measles and smallpox were dividing the infectious mortality of childhood somewhat equally, the age-incidence of measles being only a little lower than that of smallpox:

Ages of the Fatal Cases of Measles in Glasgow, 1835-39[1224].

In Limerick, which may stand for a typically unhealthy Irish city in the worst period of over-population, there were many more deaths from smallpox among children than from measles, the age-incidence being nearly the same, according to the following dispensary statistics for a number of years before 1840[1225]:

Limerick Dispensary Deaths.

Although it is impossible to prove it, yet the indications all point to measles having kept for a whole generation after 1808 the leading place among infantile causes of death which it then for the first time definitely took[1226]. Almost the only direct references to the subject were made by way of controverting the doctrine of Watt; but these are too meagre, or too general in their terms, to be of any use[1227]. The epidemics of measles seem to have travelled then, as they do now, from county to county in successive years. Thus in 1818, while most parts of England were or had recently been suffering from smallpox, the Eastern counties were suffering from measles “very frequent and fatal.” Smallpox at length reached Norwich in 1819, and became the reigning epidemic in the place of measles, which was “hardly met with” so long as the enormous mortality of the other disease proceeded[1228]. At Exeter in the spring of 1824 measles became epidemic after a long interval; many susceptible children had accumulated, and of these few escaped. The mortality was very great, and was caused by severe pulmonary inflammation, the catarrhal symptoms being mild. In one day seventeen children were buried in one of the five parish churchyards of the city; but that high mortality, according to the parochial surgeon, did not on an average stand for more than four deaths in one hundred cases[1229].

When the curtain rises, in the summer of 1837, upon the prevalence and distribution of diseases in England, as ascertained by the new system of registration of the causes of death, measles is found in the first place among the infectious maladies of childhood, thereafter yielding its place to smallpox for a year or more, and taking the lead again until it was passed by scarlatina.

Deaths by Measles and Smallpox in London, 1837-39.

The epidemic of smallpox hardly touched the Eastern counties until 1839; so that while the home counties in that year had far more deaths by measles than by smallpox, Norfolk had only 72 deaths by the former against 820 deaths by the latter. In the same year measles took the lead in four out of six great English towns, scarlatina being the dominant infection in one (Sheffield), and smallpox in one (Bradford):

Deaths in 1839 by the three chief infections of Childhood.

In all England and Wales during fully half-a-century of registration, measles has fluctuated somewhat from year to year but has not experienced a notable decline among the causes of infantile mortality (see the table at p. 614). In the decennial period 1871-80, its annual average death-rate was 377 per million living; in the next decennium it rose to 441, the previously high rates of scarlatina having fallen greatly. Among the highest rates for the ten years 1871-80, were those of Plymouth, 1·13 per 1000, East Stonehouse 1·79, and Devonport 1·19 (owing to a great epidemic in 1879-80), Exeter, 0·82, Liverpool ·91, Bedwelty (Tredegar and Aberystruth collieries) 0·88, Wigan 0·74, Whitehaven 0·71, Alverstoke 0·81. In the most recent period there have been some very high death-rates; thus at Jarrow the annual rate, which was only ·27 per 1000 from 1871 to 1880, rose in the nine years 1881 to 1889 to an annual average of ·94, having been made up almost wholly by great epidemics every other year—in 1883 (2·9), 1885 (2·4), 1887 (1·4), and 1889 (·9)[1230]. In the year 1888, an epidemic at Stoke-on-Trent, Hanley, &c. with 342 deaths, made a rate of 2·8 for the year; in Wolstanton, Burslem, &c., 221 deaths were equivalent to a rate of 2·6.

The latest reports of the Registrar-General have traced a progression of the epidemic of measles from county to county or from district to district in successive years, such as was remarked, both for smallpox and measles, by some of the 18th century epidemiologists in England, Scotland and Ireland.

Measles has no such decided preference for a season of the year as scarlatina and enteric fever have for autumn or infantile diarrhoea has for summer. But it often happens that most deaths are recorded from May to July, owing, doubtless, to the greater number of attacks in summer and not to any excessive fatality of that season. In London and the great industrial towns the deaths are spread somewhat uniformly over the year; or, in the language of statisticians, the maxima do not rise far above the mean of the year. In a tabulation of the weekly deaths in London from 1845 to 1874[1232], it appears that they touch a higher point in mid-winter (Nov.-Jan.) than in summer, a fact which may be readily accounted for by the injurious effects of the London air in winter upon a disease which is largely a trouble of the respiratory organs. In the great industrial populations of Lancashire, which resemble London in their high death-rate from measles, the rise of the deaths in mid-winter is almost the same as the summer increase[1233].

Most of the deaths from measles fall at present upon the ages from six months to three years, just as they did when the deaths were comparatively few, as at Manchester from 1768 to 1774. Deaths of adults, which were not altogether rare in the first great epidemic of modern times in 1808, are seldom heard of at present, for the same reason that adult deaths used to be uncommon in smallpox, namely, that the disease is passed by almost everyone in infancy or childhood. Although the deaths from measles sometimes reach large totals—in London during the spring of 1894 they were in some weeks as high as one hundred and fifty—yet it is the common experience of practitioners that a strong or healthy child rarely dies of measles, that the fatalities occur among the infants of weakly constitution, and especially in the numerous families of the working class in the most populous centres of mining, manufactures and shipping.

To bring these various characteristics of measles together in a concrete instance, I shall give briefly the facts of a recent epidemic in a town in Scotland of some twelve thousand inhabitants. There had been only five deaths from measles for two years. There had not been a case of smallpox for at least ten years. The measles epidemic, when its triennial opportunity came, reached a height in July, on a certain day of which month there were seven or eight burials from measles or its direct sequelae. Nearly all the children in the place who had not been through the measles in the corresponding epidemics of 1889 or 1887 suffered from it on this occasion, excepting the class of very young infants. The deaths in the whole epidemic numbered about fifty, which would not all be registered, however, as from measles. Yet this high mortality was not due to any unusual malignancy of the disease, but to the feeble stamina of a certain number of infants, or to the indifferent housing and tending of the poorer class. One did not hear of a death in the well-to-do families (probably there was none), although they had their full share of attacks. The frequency of the burials for a short time, and the effects of the epidemic on the mortality from first to last, must have been very nearly the same as in an epidemic of smallpox a century before, when the population was only a third or fourth part as large. But in the period when smallpox was in the ascendant, having few rivals among the infective causes of death in childhood, the general conditions of health in this town were altogether different. One or two specimens of the thatched huts of the poorer class had been left standing into the era of photography, so that we could compare past with present, in externals at least; also, of the houses of the richer class some still remained, perhaps turned into tenement-houses, with small windows, low doorways, and crow steps on their gables; and it was on record by the parish minister at the end of the 18th century, that within the memory of that generation there had been peat stacks and dunghills before the doors on the High Street of the burgh.

CHAPTER VI.

It is singular that a malady so distinctively marked as whooping-cough is should figure so little in the records of disease from former times. Astruc could find no traces of it in the medical writings of antiquity or of the Arabian period. In modern times the first known account of an epidemic of it is under the year 1578, when Baillou of Paris included a prevalent convulsive cough as part of the epidemic constitution of that year, remarking in the same context that he knew of no author who had hitherto written of the malady[1234]. Yet, if whooping-cough had been as common in former times as it has been in quite recent times, it deserved a high place among the causes of infantile mortality. Doubtless it occurred in former times in the same circumstances in which it occurs now. Baillou in 1578 speaks of it as a familiar thing; and it can be shown from an English prescription-book of the medieval period that remedies were in request for a malady called “the kink,” a name which survives in Scotland (like other obsolete English words of the 15th century) in the form of “kink host[1235].”

In Phaer’s Booke of Children (1553) chincough is not named. It is perhaps more singular that the disease should be omitted from the list in Sir Thomas Elyot’s Castel of Health (1541), of maladies proper to three periods of childhood; for that list has every appearance of being an exhaustive enumeration[1236]. Still, it would be erroneous to suppose that the convulsive cough of children which is so common an epidemic incident in our time, and in some impressionable subjects is the almost necessary sequel of a coryza or catarrh, did not then occur in the same circumstances as now. When Willis, in his Pharmaceutice Rationalis of 1674, remarks that pertussis was left to the management of old women and empirics, he suggests the real reason why so little is said of it in the medical compends. Sydenham mentions it twice, and on both occasions in a significant context. Under the name of pertussis, “quem nostrates vocant Hooping Cough,” he brings it in at the end of his account of the measles epidemic of 1670, without actually saying that it was a sequel of the measles. His other reference to it, under the name of the convulsive cough of children, comes in his account of the influenza of 1679. In both contexts it is adduced as an instance of a malady much more amenable to bloodletting than to pectoral remedies, the depletion being a sure means of cutting short an attack that was else very apt to be protracted, if not altogether uncontrollable[1237]. One glimpse of it we get among the children of a squire’s family in Rutlandshire in the summer of 1661. On the 26th of May the mother of the children writes to her husband then on a visit to London[1238]:

In Dr Walter Harris’s Acute Diseases of Infants[1239], the convulsive or suffocative coughs are mentioned in one place without being identified as chincough, while in two or three other places the malady is briefly referred to under its name. Thus, “corpulent and fat infants troubled with defluxions, and having an open mould, are most subject to the rickets, chincough, king’s evil, and almost incurable thrushes.” Again, chincough of infants is one of the inflammatory diseases that are “not altogether free from contagion”; and again: “Albeit that any notable translation of the subject matter of the fever into the lungs, and chincoughs, do advise bloodletting for the youngest infants, yet it is most evident that it is not a remedy naturally convenient for them.... And therefore its help is not to be invoked for all the diseases of infants except in the chincough or any other coughs that do attend and are concomitants of fevers that do suddenly begin”—showing his deference to Sydenham, his master.

Probably the “any other coughs” are those that he thus describes in another place (p. 26):

These were cases of whooping-cough, although they are not so called. Among his eleven cases, Harris gives two in infants of the Marquis of Worcester; one had been “very often troubled with an acute fever,” and was found to be much weakened by a chincough when the physician was called to him; the other, an infant of eleven months, had at the same time an acute fever “and a cough almost convulsive.”

This inclusion, under the generic name of cough, of cases that had all the signs of whooping-cough, namely, the paroxysmal seizures, choking fits, and blackness of the face, is found also in the London bills of mortality. Although “coughs” are entered as the cause of a not very large number of deaths in the earlier annual bills, with an occasional special mention of whooping-cough among them, it is not until 1701 that “hooping cough and chincough” becomes a separate item, with six deaths in the year; next year the entry is “hooping cough” alone, with a single death, and so on for a number of years in which the deaths are counted by units; in 1716 they rise to eleven, and continue to be counted by tens until 1730, when 152 deaths are set down to “cough, chincough, and whooping-cough.” It would be a mistake to suppose that these figures during the first thirty years of the 18th century are anything like a correct measure of the number of infants in London who suffered from whooping-cough, or are at all near the number who might have reasonably been returned as dying from it. It was in that generation that the entries of the Parish Clerks became most indefinite as to the causes of death in infants, five-sixths of the enormous total of deaths under two years being entered under the generic head of “convulsions” and “teeth,” while the item “chrysoms” received the deaths under one month old.

The increase of whooping-cough in the following table, from units to tens, from tens to hundreds, and thereafter to a somewhat steady total of hundreds year after year, can hardly be explained except on the hypothesis of more exact classification of infantile deaths, corresponding to the actual decline of the article “convulsions” in the second half of the century.

(Continued in the table of measles deaths, p. 655)

It is not without significance that the vital statistics of Sweden were the first to give whooping-cough something like its rightful place among infantile causes of death: from 1749 to 1764 the deaths set down to that cause were 42,393, or an annual average of 2600, the epidemic year 1755 having 5832. In this we should find merely the influence of systematic nomenclature. Nosology, or the scientific classification of diseases, may be said to have begun under Linnaeus, who was for many years professor of medicine at Upsala before he became professor of botany, and was teaching a somewhat rudimentary nosology to the Swedish students of medicine before the great work of his friend and correspondent Sauvages made classifications general.

Concerning the year 1751, which has 275 deaths from whooping-cough in the London bills, Fothergill writes in May: “Great numbers of children had the hooping cough, both in London and several adjacent villages, in a violent degree. Strong, sanguine, healthy children seemed to suffer most by it; and to some of them it proved fatal where it was neglected or improperly managed”—the deaths having become more numerous towards the end of the year[1240]. At Edinburgh, during the second year of high mortalities in the famine-period 1740-41, whooping-cough has 101 deaths to 112 from measles, having had only a fourth part as many the year before (see p. 523). In the Kilmarnock register from 1728 to 1763, “kinkhost” is credited with a total of 116 deaths, about 3 on an annual average, measles having a total of 93 during the same thirty-six years. In Holy Cross parish, a suburb of Shrewsbury, chincough has 9 deaths in the ten years 1750-60, and 6 in the next ten years, measles having 4 and 15 in the respective periods, and convulsions 9 and 31. In Ackworth parish, chincough has no deaths in the ten years 1747-57, and 2 in the next ten years, “infancy” having 13 in each decade, “convulsions” and measles none in the first, 6 and 2 respectively in the second. Warrington, in the disastrous smallpox year, 1773, had 16 deaths from chincough and 34 from convulsions. In the two years 1772 and 1773, Chester had 33 and 10 deaths from chincough, 70 and 69 from convulsions, 17 and 13 from “weakness of infancy.”

Watt’s researches in the registers of all the Glasgow burial-grounds brought out the fact that whooping-cough during a period of thirty years, 1783 to 1812, had been a common and somewhat steady cause of death among infants, having made 4·51 per cent. of the annual total of deaths at all ages in the first six years of the period, and 5·57 per cent. in the last six years[1241]. This was a higher annual average ratio than in the London bills for the same period (see the tables at p. 647 and p. 655), and was probably the maximum in Britain, inasmuch as the Glasgow death-rate of infants was the worst from all causes.

Whooping-Cough in Modern Times.

When the causes of death began to be registered, in July, 1837, whooping-cough was found to have the following relative place among the principal maladies of children during the latter six months of the year in London and in all England and Wales.

Mortality by diseases of Children, last six months of 1837.

Throughout the whole registration period, whooping-cough has kept its place steadily among the chief causes of infant mortality, neither decreasing nor increasing notably in the successive periods from 1837 to the present time. Its mortality has varied a good deal from year to year, owing to occasional great epidemic years such as 1866 and 1878; but on the mean annual average of decennial periods, it has varied little:

Annual Deaths by Whooping-cough per million living at all ages.

No other epidemic malady has shown the same excess of female deaths in proportion to the numbers of the sex living, diphtheria being the only other that shows an excess at all.

The excess of deaths by whooping-cough among female infants was roughly shown by Watt in 1813, viz. 975 females to 842 males in the registers of the Glasgow High Church, College Church and the North-Western Cemetery, the relative numbers of the sexes living at the respective ages being then unknown. In all Scotland in 1889 the ratio was 1043 male deaths to 1225 female. The singular difference between the sexes in this respect is almost certainly related to the corresponding differences in the formation and development of the larynx, the organ which gives character, at least, to the convulsive cough of children. The expansion of the larynx in boys, which becomes so obvious at puberty and remains so distinctive of the male sex, is one of those secondary sexual characters which begin to differentiate quite early in life, and are probably congenital to some extent. It is not known whether female children are more often attacked than males; but it is probable that they are predisposed both to acquire coughs of the convulsive suffocative kind and to have their lives shattered by the attack—for the same anatomical and physiological reasons, namely, the imperfect development of the posterior space of the glottis with the spasmodic closure by reflex action[1242]. The deaths have been nearly all under the age of five.

Deaths by Whooping-cough per million living at the respective age-periods.

These proportions are almost the same as those given by Watt in 1813 from three of the Glasgow registers.

Most of the deaths are in the first year, and in a rapidly declining ratio until the fifth, according to the following rates per million of male children living at each age-period (these figures are for a single year, 1882):

The mortality from whooping-cough falls very unequally on town and country. Thus, in Scotland in 1889, it caused 2268 deaths, being 3·13 per cent. of the deaths from all causes, and equivalent to a rate of ·58 per 1000 living. The death-rate varied as follows: ·91 in the eight principal towns, ·46 in the group of large towns, ·45 in the group of small towns, ·25 in the mainland rural districts, and ·08 in the insular rural districts. In England, the capital has more than its share of deaths from whooping-cough, Lancashire coming next, while the death-rates of Monmouthshire, Cornwall and Warwickshire are also a good deal above the mean of the whole country. The lowest death-rates are found in the purely agricultural counties.

During the last half-century there has been a decline in the death-rate from all causes, including the infectious diseases as a group; but it can hardly be said that whooping-cough has had a due share in this decline. Notably in Ireland, where the decline of infectious disease has been most marked, it has been, as it were, pushed to the front of its class by the shrinkage of the other items. In Scotland it is now decidedly at the head of the list, and in England it has shared the first place with measles since the great diminution of scarlatina deaths.

Annual average Death-rates per 100,000 living.

There is a small decrease in the death-rate of whooping-cough within the last decennial period, whereas in that of measles there is a slight increase (except in Ireland). The comparative steadiness of whooping-cough among the causes of death is doubtless owing to the fact that the bulk of its fatalities are among infants, and that there appears to be an irreducible minimum of the deaths from all causes at that age-period.

Whooping-Cough as a Sequel of other Maladies.

Although it is convenient to group whooping-cough among the infectious diseases, and although it is a clear case of a malady that comes in epidemics, yet its pathology is peculiar. It seems to be more a sequel of other diseases than an independent or primary affection. The whoop of the breath, from which it is named, is really proper to any convulsive cough of some infants or children. Adults, having undergone the change in the form and relative size of the larynx at puberty, have the convulsive cough usually without the whoop if they have it at all. After the successive influenzas of recent years (1889-92), many adults suffered from convulsive paroxysmal cough which was whooping-cough in all respects but the whoop, the choking fits, the blackness of the face, and the vomiting being, of course, all kept in subjection by the greater control of adults over their reflex actions.

It has been often remarked that the ordinary whooping-cough of children has followed epidemics of influenza, or widely prevalent catarrhs. Thus, Hillary records in July, 1753, an epidemic of whooping-cough, or “the fertussis,” all over the island of Barbados following the epidemic catarrh which was at a height in January of the same year. Whooping-cough had not been known in the island for many years past, “neither could I find by the strictest inquiry that I could make that any child or elder person did bring it hither[1243].” Willan, in his corresponding records of the succession of diseases at the Carey Street Dispensary, London, from 1796 to 1800, has the following: