of the Local Government Board, as to "the danger of breathing AIR which is foul with effluvia from the same sorts of impurity". -a danger whose source obviously lies in the atmospheric transportation of disease-germs. I have left myself but little space for the discussion of the second part of my subject the bearing of the natural history view of zymotic diseases upon the question of their origin and mutual relations. It is, doubtless, needful for the purposes of pathological study, that these diseases should be defined as "specific types," just as the naturalist defines "species" of plants or animals; and as, in our pre-evolution days, it was held that every true species was separated from every other by constant characters genetically transmitted from parent to offspring, so it has been generally believed that the poisons, not only of small-pox, scarlatina, and measles, but of a large number of different forms of fever, as well as of other maladies propagated by contagia, are to be ranked as specifically different. " The species-making naturalist of the past generation laid greater stress on points of minute difference than on those of general agreement, disregarded the modifying influence of "environments,' and selected the strongly-characterized examples for description, neglecting the intermediate forms by which these are often gradationally connected. But in the light of the modern doctrine of evolution, the scientific naturalist makes it his aim to ascertain how the different races of plants and animals have come to divaricate from each other; and studies their respective "variations," as affording the best clew to the origin of their larger and more constant specific" differences. And those who have most carefully studied the tribe of saprophytes" to which disease-germs belong, have long since come to the con. clusion that there are no forms of vegetation whose range of variation'' under differences of environment" is so wide; it being yet uncertain, indeed, that we know the entire life-history of any one of them. Now, it has been too much the habit of pathologists, in scientifically defining specific types of disease, to follow exact ly the same course as the species-makers among naturalists-insisting on minute differences rather than on points of agreement, and assuming that these differences are constant. Every practitioner of medicine, on the other hand, who has had opportunities of observing the same diseases in different localities, at different seasons, and in different epidemics, well knows how greatly their characters vary; "hybrid forms' and "sub-varieties" presenting themselves from time to time, which receive passing notice and then die out. Thus, although no eruptive fevers are more clearly differentiated, when occurring in their characteristic forms, than measles and scarlatina, yet cases every now and then occur, in which their symptoms are so mingled as to puzzle the most experienced doctors. I even remember such a hybrid disease to have been epidemic some thirty years ago in the East of London; and as Sydenham, one of the most sagacious medical observers that ever lived, did not separate the two, I cannot but think it probable that this "hybrid'' was the disease prevalent in his time. Again the small-pox epidemic of 1871 and subsequent years has been characterized by the re-appearance of the malignant" type of that disease, which had not previously shown itself in Europe, except in a few isolated cases, during the present century. The whole course of that hæmorrhagic' type, when presented in its most characteristic form (in which death occurs before the appearance of the eruption), is so entirely different from that of ordinary small-pox, whether “confluent' "discrete," that the two diseases might be well accounted specifically different, if it were not certain than they originate in the same contagium. So, again, some of those who have had largest experience of the severest forms of malarious disease, are satisfied of the unity of causation that underlies variety of manifestation. Thus, says Dr. Haspel, the author of a very able work on the "Diseases of Algeria" (Paris, 1850), or " fevers, dysentery, and diseases of the liver constitute an indivisible whole under the dominion of a single cause; and those who deny this truth are either misled by theoretical prejudices, or will not make use of their eyes. It is a significant fact, rightly insisted on by Dr. Maclean, that exactly in proportion as we have banished malaria from the soil of the British Islands, so have we got rid not only of ague, but of dysentery and of suppurative inflammation of the liver, as endemic diseases. I have already adverted to changes in the type of fever from "non-infective" to "infective'' of which there seems to me adequate evidence; and I might adduce a number of other instances such as the difficulty that often occurs in India in discriminating between cholera and enteric fever -in support of my position, that even the best-marked types of zymotic disease are not distinguishable by constant and invariable characters, but that, just as higher plants are modified by cultivation, so may the germs of these diseases develop themselves in a great variety of modes, giving rise to very different maladies, according to the conditions, whether local or individual, under which their development takes place." * But the same analogy carries us further, and suggests that the peculiar morbific activity possessed by each specific type of disease-germ may be derived from the operation of particular "environments" on ordinary saprophytes through a long succession of generations, just as among plants of higher types. And this view is borne out by the remarkable influence of artificial "culture" upon some of those which have been most carefully studied in this matter. It is a fact of great significance, that the malignant Bacillus anthracis of charbon" does not differ morphologically in any important character from the innocent Bacillus subtilis of hay infusions; and although it has not yet been certainly shown that any method of treatment can give to the latter the potency of the former, yet it seems not improbable that such will prove to be the case. With Dr. William Roberts, "I see no more difficulty in believing that the Bacillus anthracis is a 'sport' from the Bacillus subtilis, than in believing, as all botanists tell us, that the bitter almond is a 'sport' from the sweet almond-the one a bland, innocuous fruit, and the other containing the elements of a deadly poison.' So, as it seems to me, there is nothing inconsistent with our recognition of cholera and typhoid as specific types of disease, in the admission that under some possible conditions they may originate de novo from saprophytic germs not ordinarily capable of engendering such maladies in the human system. Among my earliest professional recollections, going back to the year 1829, is that of the occurrence of a very remarkable outbreak of a severe malady in a school at Clapham, of a type then quite unknown to practitioners in this country, but which an old Indian doctor, who was asked to see the patients, declared to have the characters of the cholera of India, which was then (as it subsequently appeared) on its way toward us, but whose advent no one at that time regarded as probable. Having lately referred to the Medical Gazette of August 22d in that year, I have found this recollection fully confirmed by the record of the "Fatal Cholera at Clapham" published at the time; and cannot hesitate in the belief that if the outbreak (affecting twenty out of twentytwo boys at the school, and the two children of the master, of whom one died after only eleven hours' illness) had occurred during a cholera epidemic, the patients would have been regarded as suffering under that disease. A few days previously, a cesspool had been opened to let off from the playground stagnant water accumulated by the recent heavy rains, and its contents had been distributed over the garden adjoining the boys' playground. Whether true Asiatic cholera or not, this sudden simultaneous outbreak can scarcely be regarded as a mere result of putrescent emanations; it had every character of a specific disease implanted by germs; and the probability seems strong that these germs were those either of some other type of zymotic disease, or of ordinary saprophytes, to which some special conditions had imparted a choleraic potency. Although from the time when Sir William Jenner pointed out the marked distinctions between typhus and typhoid (or enteric) fevers, their distinctness has been generally recognized, and any difficulty in diagnosing a case has been commonly set down to ignorance or imperfect observation, yet I have the high authority of the late Sir Robert Christison for stating that these diseases are not at all times, or in all places, so definitely distinguishable. Not long before his death, the Nestor of the medical profession in Scotland emphatically assured me, that "looking at this class of diseases from the natural history point of view, he had been led by an experience of half a century to regard them, not as uniformly marked out, one from another, by well-defined boundaries, but as shading off gradationally one into another.' Being specially anxious that those who are laboring to build up the noble Science of Preventive Medicine, should work no unsound material into the fabric they are constructing, I will earnestly press upon them to avoid all exclusive theories, and to take nature alone as their guide. The broader and deeper the foundation they lay, the more solid and durable will be the edifice that rests upon it. [SINCE the above was in type, the French Commission which was sent to Egypt to investigate the recent epidemic. of cholera has reported, as the result of its inquiries, that this epidemic was not imported, but was born as well as bred in the country itself; especial stress being laid on the recent prevalence of a cattle-plague, and on the practice of throwing into the rivers and canals the bodies of animals that had died of it. It was, moreover, the opinion of the Commission that the disease was not pure Indian cholera; but that in some of its symptoms it rather resembled plague. These conclusions are entirely in harmony with the views advocated in the latter part of this paper. A small treatise has been recently published on the Evolution of Morbid Germs," by Mr. Kenneth M. Millican, which contains a body of additional evidence, derived from clinical experience, of the variability in the types of zymotic diseases propagated by the same contagia; that of the intercommunicability of scarlatina and diphtheria (under certain conditions) being peculiarly cogent. -W. B. C.]-Nineteenth Century. EFFECT OF MARRIAGE ON LIFE. * IN the year 1867 a statement was made by Dr. Stark, Registrar-General for Scotland, which attracted a good deal of notice. He announced, as the result of his investigations into the relative death-rates of married and unmarried men, that the mortality is very much greater among the latter than among the former. Since then several years statistics have been published by Drs. Stark, Drysdale, and others, in this and other countries, and they appear to corroborate the doctrine that marriage may be regarded as a sort of life-insur ance. It is not with any desire to invalidate the truth of this doctrine that we propose here to point out the great *This article, excepting a few words relating to its more extended application, and bring. ing it down to date, was written in the summer of 1867-our readers may probably guess by whom.-SUB-ED. uncertainty of such statistics as these. There are enough arguments in favor of matrimony without introducing false ones. An old proverb tells us of the advisability that no cask should rest on an alien basis, and this is especially the case where the basis of our figurative cask is already wide enough to secure stability. The advocates of marriage will therefore, we trust, look upon us rather as an ally than as a foe if we exhibit, as we think we shall be able to do, the unsubstantial nature of the argument based on such statistics as we have referred to above. Let us in the first place see what the evidence is on which the argument is founded. A single case will suffice. Take the earliest by combining the results of two years' observations. Dr. Stark thus compares the mortality per thousand of married and unmarried men : After At first sight, it might seem that nothing could be clearer or more satisfactory than this evidence. We see that between the ages of 20 and 25, the date rate of the unmarried men is nearly twice that of the married men. this the ratio gradually diminishes, so that when we come to the quinquennial period between 45 and 50, the ratio, instead of being 2 to 1, is only 21 to 17, but still it is a ratio of excess and so, up to the last recorded period, we find the same evidence in favor of the married men's prospects. And again, let us take another view of the matter. It is easy to determine the mean age of the married men and of the bachelors at death. We find that the former age exceeds the latter by fully 19 years! Here, then, we seem to have the most striking evidence in favor of matrimony as an agent in producing longevity. It would seem almost that all we need fear would be the undue extension of the argument. If one wife does so much to prolong a man's life, what effect, it might be argued, should two, three nay, a dozen wives, for that matter not produce? Passing over this view, as a manifest invention of that enemy of social happiness, the confirmed old bachelor, let us seriously inquire what force there really is in the evidence adduced; for the evidence is not wholly without force, only it has been asked to bear rather more than it is capable of doing. There are two most important rules in the application of statistics, for want of attending to which many have fallen into serious error. First, we must assure ourselves that there is nothing in the examples collected which savors of selection and secondly, we must have a sufficient number of examples. As respects the second rule, we do not think there is any reason to complain of the evidence. For although the period over which the results extend is not a very long one, yet the wide range of country included in the registering is fully sufficient to make up for the defect in point of time. In fact, the close accordance observed by Dr. Stark between the results of the first two yearly periods dealt with was quite sufficient to prove that a more extended series of observations was not needed. Results of this sort only repeat themselves when they are severally founded on a sufficiently wide range of statistical inquiry. But, as respects the first rule, we think there is very strong reason for suspecting the evidence before us. We must note, in the first place, that it is one of the most difficult things in the world to free results from selection" in some form or other. Take the simple instance of tossing up a halfpenny; is the chance perfectly equal that head or tail will turn up? It seems so, but it is not necessarily so. As the halfpenny turns over and over in the air, there may be an irregularity- imperceptible to the sense-due to the unequal distribution of the metal on the two faces. here we see the importance of the second rule mentioned above. Any irregularity in the figure of the coin will she itself in a sufficient number of trials as certainly as by the most accurate mesurement and the most careful examination And We had lately a remarkable instance in our own experience of the difficulty of removing all trace of selection. points, We wished, for a particular purpose, to distribute a number of dots perfectly at random, over a sq face. This may seem a simp ] but we did not find it so. suggested-" Take a handful and throw over the surface at then mark the place of eac fact, however, that the grains a handful, and were spread show itself. Another methoc suggested: Prick a number without directing the motion o by the will." But how are this very thing? If we close we shall naturally make for th parts of the surface, for fear c are sur matter may be grains andom; The id form ut, will may be of holes he hand to do ur eyes central missing the surface altogether, and here at once is "selection ;" and if we open our eyes it is absolutely impossible not to aim each stroke with some object, however much we may persuade ourselves that we are striking quite at random. The method we finally adopted was this. We divided each side of the square into 100 parts, which we numbered in order, and drawing lines through the points of division we divided the square into 10,000 small squares. We then took a book full of figures (in fact that inviting work, a table of loga rithms), and opening at random placed the point of a pencil at random on the page. The figure nearest to the point we marked down, and we took out in this way 4000 figures. We now took the first four figures-8, o, 1, 7 say, and did thus with them-the first two gave the number 80, the next two the number 17, and we accordingly marked a dot on the eightieth row of squares, in the seventeenth square of the row. Thus we had 1000 points distributed as we thought quite at random. But on a closer inspection we suspected the influence of selection; and where does the reader suppose we detected it? In the shapes of the figures used to represent numbers. In taking the number nearest to the pencil point we had omitted to notice (when the point seemed half-way between two figures) that the 1's, the 4's, and the 7's, do not cover quite as much space as the other figures. And in fact, when we came to count over our list of numbers we found there was a marked deficiency of these, and a marked excess of 8's, 5's, and 2's. This excess showed itself in the arrangement of the dots over the square surface. We seem to have wandered a long way from our bachelors and married men, but if we have succeeded in showing how subtle an influence selection, conscious or unconscious, is capable of exerting, it will be found that our digression is in reality very much to the point. It may be asked, "If Dr. Stark took the mortality of the whole population, how can there have been any selection?" We answer by another question, Is there nothing in the state of bachelorhood itself which affords suspicion of selection?'' In answering this question we wish to avoid possible misconception. In dealing with averages, individuals are not to be considered. And therefore, if we say anything of bachelors, as a class, which may seem disparaging, individual bachelors are not to be on that account offended; though, perhaps, many of our readers would not be greatly troubled even if offence were given to given to a few of the single-minded. Well, then, it appears to us that if we look on bachelors as a class, we shall see evidence that they are not on a par with married men. A It will not be denied that many men are prevented from marrying by illhealth or a weakly constitution. This may happen in more ways than one. man may either, through ill-health, suffer by comparison with the hearty and stalwart, or he may feel that he is unfit to struggle with other difficulties than those he has as a single man—that he is unable, perhaps, to provide for wife and children, or that the cares and anxieties which married life necessarily brings with it would be more than he could bear. For instance, the very thought of a crying child disturbing his night's rest would shake the nerves of an invalid. Or again, if a man is consumptive, or suffers under any other ailment which is apt to repeat itself in successive generations, he may well and wisely eschew the thought of marriage, fearing lest he should become the parent of unhealthy children. In these and in many other ways, unhealthy or weaklyconstitutioned men fall into the list of bachelors. No one can fail to recognize the influence of this form of selection' on the comparative mortality of the two classes we are considering. Again, it cannot be doubted that very indigent persons and the members of unhealthy trades are, on the whole, kept somewhat from the lists of marrying men. Of course, hundreds of these marry in any given country; but comparing them as classes with other classes, there is, undoubtedly, such a tendency as we have mentioned. The influence of this cause, again, cannot be doubted, since the longevity of the classes we have named is undoubtedly inferior to the average longevity of the population. Here, then, we see two causes (and |