In the face of these facts, well established and confirmed many times, it is impossible to accept Thiltges’^[201] denial of the ingestion of these bacteria by the phagocytes of the fowl. Some fault of technique, which I am not at the moment in a position to indicate exactly, has evidently slipped into this author’s work. The positive data, however, on phagocytosis in the fowl, obtained by Hess, Wagner, and Trapeznikoff, data confirmed by myself, render unnecessary any fresh researches for the purpose of explaining the negative results obtained by Thiltges. As regards his experiments on the bactericidal action of defibrinated blood and of blood serum of fowls against the bacillus and its spores, experiments whose results are opposed by those of Wagner, the contradiction may be explained pretty easily, at least in part. Thiltges mentions several times that the bacilli, when sown in the blood serum of the fowl, were aggregated in clumps. Nevertheless, he has failed to guard against this source of error and has attributed the diminution in number of the colonies on plates to the destruction and not to the agglutination of the bacilli. Thiltges gives so few particulars of the conditions under which his experiments were performed that we do not even know at what temperature he kept his tubes containing blood and serum sown with bacilli. As Wagner kept his at 42°–43° C., a temperature which corresponds to that of the body of the fowl, I asked M. Gengou to make a series of experiments on the bactericidal power of the plasma and blood serum of fowls on the anthrax bacillus, keeping his tubes at 37° C. The result of his experiments was in complete accord with those of Wagner. Under the conditions that I have just stated the fluids of the fowl are no more bactericidal than they are under the conditions maintained in Wagner’s experiments.

In summing up these data on the natural immunity of fowls against anthrax, we are certainly justified in concluding that it is due to the phagocytosis and not to any bactericidal property of the “humours.”

The pigeon is more susceptible than the fowl to the action of the anthrax bacillus, still it manifests a certain degree of resistance against the microbe. After what we have said on the subject of the fowl we need make but few remarks on the pigeon, in spite of the very animated discussions that have taken place on the mechanism of its immunity. When Baumgarten was offering a systematic opposition to the part played by phagocytic reaction in immunity, he set his pupil Czaplewski^[202] to investigate the resistance of pigeons against anthrax. The results of this investigation were absolutely negative as regards phagocytosis. The latter was said to have no importance in the defence of the organism, which resisted simply because it was impossible for the bacillus to live in the body of the pigeon. I then set myself to study this question^[203], and I was able to demonstrate that the anthrax bacillus is quite capable of keeping alive in the pigeon, that it can develop in its fluids, but that it is unable to defend itself against the aggression of the phagocytes which ingest it and completely digest it. By isolating the phagocytes that had ingested the bacilli injected into the body of the pigeon, I was able to prove that a number of these bacilli were still alive. The enfeeblement and death of the phagocytes when outside the body allowed the anthrax bacilli again to get the upper hand in this struggle, to develop and to give virulent cultures. The part played by phagocytes in this example of natural immunity was thus placed beyond doubt.

Later, Czaplewski^[204] himself became convinced that his previous negative results would not stand criticism, and Thiltges, in his work already mentioned, when discussing the fowl, was able to confirm the importance of phagocytosis in the defence of the organism of the pigeon against anthrax. He was struck by the difference between these two species of birds. In the pigeon it was easy for him to prove that in the individuals that succumb to anthrax the phagocytic reaction is very feeble, whilst in those which ultimately resist the bacillus it is very pronounced. Thiltges likewise observed that the blood and blood serum of pigeons when sown in vitro with the anthrax bacillus, manifest only an insignificant bactericidal power, a fact that further warrants him in attributing great importance to phagocytosis in the maintenance of the natural immunity of the pigeon. It is remarkable that, in presence of these facts, it did not occur to the author to ask whether this fundamental difference in the mechanism of the resistance, which he thought possible in two birds so closely allied as are the pigeon and the fowl, really did exist in nature. I infer that his experiments on the fowl were made before those on the pigeon, and that the difference in his results depended specially on the fact that he had acquired greater skill in executing his later experiments.

Having observed that frogs die readily when inoculated with an anthrax bacillus that was adapted to develop at a low temperature, Dieudonné (l.c.) endeavoured to suppress the immunity of the pigeon by using bacilli adapted to a high temperature. But the inoculation of a second generation of the anthrax bacillus, cultivated at 42° C., was borne by five pigeons without inconvenience. Even bacilli that were rendered capable, by cultivation through sixteen generations, of developing at this temperature were not in a condition to kill more than five pigeons out of thirteen inoculated. These attempts to explain immunity as due to the properties of the bacilli rather than to those of the organism of the pigeon, have therefore led to a result very different from that anticipated by Dieudonné.

The pigeon is further of interest to us because of its natural immunity against the bacillus of human tuberculosis. It resists considerable doses of this bacillus, so virulent for man and for the majority of mammals, and even for some birds (canaries and parrots). Dembinski^[205], studying the mechanism of this immunity, was able to prove that the bacilli of human tuberculosis encounter in the organism of the pigeon a very great resistance from the phagocytes, especially from the macrophages. These cells fuse together around masses of bacilli and imprison them within real giant cells or polynucleated macrophages (Fig. 21). The microphages in this struggle play only a secondary part, but the resistance offered by the macrophages is a most effective one. Incapable of completely destroying the bacilli, these phagocytes exercise over them an unfavourable influence and prevent them from multiplying and exhibiting their noxious action. The importance of the defence by the macrophages comes out still more clearly when compared with what takes place if, instead of the bacillus of human tuberculosis, we inoculate into pigeons the bacillus of avian tuberculosis. In the latter case the microphages certainly promptly seize the bacilli, but being powerless against them they perish, whilst the macrophages only intervene later on and in small numbers. The result is that in the pigeon the avian bacillus becomes generalised in the organism and sets up a fatal tuberculosis.

It must be admitted, then, that the immunity of the pigeon against the bacillus of human tuberculosis is due to the defence by the macrophages. This conclusion is corroborated by the fact that in the fowl—equally refractory against the same bacillus—there is also observed a very strong macrophagic reaction.

Nocard^[206], who for several years has been carrying on studies on the relations between the bacilli of human and avian tuberculosis, conceived the idea of adapting the former to the organism of the fowl. With this object he enclosed a culture of the bacillus of human tuberculosis in a sac of collodion which he then introduced into the peritoneal cavity of fowls. Under these conditions the bacillus, protected against the aggression of phagocytes, continued to live inside the sac through whose walls the fluid part of the peritoneal lymph could diffuse. After several passages from sac to sac the human bacillus becomes acclimatised to the body of the fowl and is transformed into a variety quite comparable to the bacillus of avian tuberculosis. This experiment has definitely settled the question so long under discussion of the specific difference between the two tubercle bacilli. It has resolved it in the sense of affirming their unity; the avian bacillus is only a modified race of the same bacillus which sets up tuberculosis in man and other mammals.

In spite of the great difference between the anthrax bacillus and that of human tuberculosis, the immunity against these two bacteria, which is shown in birds, depends in every case upon the reaction of the phagocytic system.

Having rapidly glanced at natural immunity as we ascend the scale of the animal series we now come to it as it presents itself in the highest class, Mammals, a question on which it is necessary to dwell at greater length because of its great importance, and also because of the fuller study that has been given to it.

As the immunity of the Invertebrata and of the lower Vertebrata against the anthrax bacillus has furnished us with several important indications we will first endeavour to throw light on the mechanism of the resistance offered to anthrax by certain mammals. The representatives of this class being, however, for the most part extremely susceptible to this disease, examples of true natural immunity are very rare. The first place among resistant mammals is occupied by the dog. Although young dogs, as demonstrated by Strauss^[207], readily take fatal anthrax, the canine species may nevertheless be regarded as possessing a real immunity, as adult dogs withstand, without inconvenience, the inoculation of large quantities of bacilli. When introduced beneath the skin these bacilli excite a local inflammation, accompanied by a very marked diapedesis of white corpuscles which at once begin to devour the bacilli. This phagocytosis has already been observed by Hess^[208], Malm^[209], myself, and several other investigators, so that its existence cannot be doubted. Recently, Martel^[210] has demonstrated a very distinct phagocytic reaction in all those cases where he has had to deal with dogs that were refractory or not very susceptible. This reaction is shown by the ingestion of the bacteria and by the large accumulation of leucocytes at the seat of inoculation. His researches are of special interest by reason of the counter-test that he was able to make upon dogs that were susceptible to anthrax. It was demonstrated some years ago that the natural immunity of the dog against the bacillus, although very real, is, nevertheless, relative and limited. Thus Bardach^[211] established the fact that dogs from whom the spleen, an organ full of phagocytes, had been removed, became susceptible to anthrax. Even dogs into whose veins he injected fine wood-charcoal powder suspended in water, with the purpose of “diverting” the phagocytosis, readily succumbed to anthrax.

Martel endeavoured to suspend the natural immunity of dogs by injecting into them phloridzin or pyrogallic acid. But he obtained much more constant results by inoculating the bacillus into rabid dogs. The organism, weakened by this terrible disease, became very susceptible to anthrax, and the rabid animal succumbed to anthrax before the rabies had completed its evolution. By its passage through the rabid dog the anthrax virus is so augmented in virulence that it becomes fatal for normal dogs. Martel succeeded also in reinforcing the bacillus isolated from a cow affected with anthrax. In all these cases where the reinforced bacilli set up a severe and rapidly fatal infection, Martel could demonstrate only a feeble phagocytic reaction.

Researches on the phagocytosis of dogs, inoculated with the anthrax bacillus, have always demonstrated a regular and constant relation between this reaction and the resistance of the organism. On the other hand, experiments undertaken for the purpose of establishing the part played by the body fluids in this immunity, have always given negative results.

As the dog, of all mammals, exhibits the greatest natural immunity from anthrax, it is very natural that in the bactericidal property of its blood the key to the enigma has been sought. Thus Nuttall^[212] concludes from his experiments that the anthrax bacillus is readily destroyed by defibrinated dog’s blood. But, as this result was not in accord with my observations^[213] that the bacillus is easily cultivated in dog’s blood, and as several observers, especially Lubarsch^[214], had arrived at conclusions opposed to those of Nuttall, systematic researches were made for the purpose of solving this complicated problem. Denys and Kaisin^[215] sought to remove the objections formulated against the explanation of the immunity of the dog as due to the bactericidal property of its blood by affirming that this power, which is absent in the inoculated dog, develops whilst the animal is under the influence of the bacillus. Immunity is reduced, then, in this case to the establishment of a new property in the fluids during the course of the struggle of the organism against the inoculated bacillus. None of the observers, however, who have repeated these experiments, e.g. Lubarsch^[216] and Bail^[217] were able to confirm the results of the Belgian observers. Denys himself, indeed, having resumed this study with Havet^[218], had to reject the conclusions of his former work executed in collaboration with Kaisin. He is persuaded that their error was due to the fact that in their experiments in vitro, the living leucocytes ingested the bacilli and prevented their development. As the result of these new researches Denys and Havet have come to the conclusion “that the main, the predominating part of the bactericidal power of the dog’s blood must be ascribed to the leucocytes acting as phagocytic elements” (loc. cit. p. 15).

As a result of the investigations I have summarised the conclusion is forced upon us that the natural immunity of the dog from anthrax is a function of the phagocytes. In presence of this uniformity of the experimental results it becomes very important to make a more profound study of the phenomena that manifest themselves during the destruction of the bacilli by the phagocytes of the dog. What are the phagocytic elements which play the principal part in this struggle, and by what means do they attain this result? Gengou^[219] undertook a detailed investigation in my laboratory to answer these questions. He was able to convince himself, in agreement with the statements of his predecessors, that not only was the serum of dog’s blood not bactericidal for the anthrax bacillus, but that the plasma of the blood is no more so. The fluid of the aseptic pleural exudation obtained after injection of gluten-casein, was likewise incapable of killing the anthrax bacillus. When Gengou, by means of centrifugalisation, isolated the leucocytes from these exudations, washed them in physiological salt solution, froze them, and then macerated them in broth, he obtained suspensions of white corpuscles, to which he added bacilli. He was able to demonstrate that when the exudations contained macrophages principally, as is observed in exudations taken at the end of two or three days, the bactericidal power of the suspensions was nil or insignificant. When, on the other hand, the leucocytes came from exudations only twenty-four hours old and were composed almost exclusively of microphages, the destructive action on the bacilli of the extract of the microphages in broth was most marked. Now it is fully demonstrated that in the exudation set up in the refractory dog by the injection of anthrax bacilli, it is especially the microphages which exhibit the phagocytic reaction against this bacillus.

This is how the question of the immunity of the dog from anthrax stands at present. The natural immunity of this species, which although not unlimited, is very real, depends on the activity of phagocytes. These elements, under the stimulus of the bacillus and its products, exhibit a positive chemiotaxis of the most marked character, they approach the bacilli, ingest them by a physiological act, and destroy them by means of a substance which is not found in either the plasma or the blood serum, but which can be demonstrated in an extract of the microphages.

In spite of the uniformity and precision of these data, it is impossible to rest satisfied with describing, as an example of natural immunity from anthrax, the single case of the dog. If the resistance of the rat against this disease was merely of historical interest because of the large number of works devoted to this question, we might relegate it to the chapter reserved for the history of our knowledge on immunity. But it is not so. The anthrax of rats is a subject full of very valuable instruction, and von Behring was quite justified in saying that whoever wished to get a true conception of natural immunity from a virus should pay special attention to this example.

As a matter of fact, it may be stated that the grey rat (Mus decumanus), the black rat (Mus rattus), and white rats are far from enjoying a true immunity from anthrax. They, nevertheless, exhibit a more or less marked resistance against this disease and are always less susceptible than are the other laboratory rodents: mice, guinea-pigs and rabbits. Rats resist attenuated bacilli (anthrax vaccines) better than do these three species, and in order to induce in them fatal anthrax it is necessary to inoculate a much larger number of virulent bacilli. On the other hand, rats are distinguished by a great irregularity in the resistance they offer to the bacillus. At times they resist very virulent bacilli; at others they contract a fatal disease after an injection of very attenuated bacilli (Pasteur’s first vaccine).

In my first memoir on anthrax^[220] I noted the fact that in rats the phagocytosis against the bacillus when injected subcutaneously was more marked than after the same inoculation into the rabbit and guinea-pig. Later, this fact was disputed by several observers, who refused to accept the extent and importance of the phagocytic reaction in the rat. This opposition was strengthened by a very interesting discovery made by von Behring^[221], namely, that the blood serum of the rat possessed a remarkably destructive power for the anthrax bacillus. When this observer added a certain quantity of anthrax bacilli to some blood serum of the rat, instead of elongating into filaments and dividing they underwent a change, lost their normal refraction and took on staining reagents very imperfectly. The membranes alone remained of the bacilli thus treated. Von Behring considered that this bactericidal action of the serum depends on the presence of an organic base dissolved in the blood fluid. He had merely to neutralise the serum by means of an acid, and there was at once a very abundant development of the bacillus. From these researches von Behring came to the conclusion that the natural immunity of the rat from anthrax can be reduced to terms of the chemical action of the blood on the bacillus.

In one of his most recent publications this author^[222] returns to the question of anthrax in rats and sums up his present point of view as follows. He regards the immunity of these rodents as being relative, not absolute. “The anthrax bacilli”—he says—“die in rat’s serum in vitro; and in the cases where the inoculation of these animals with the anthrax virus is not fatal, it is at least reasonable to assume that the blood fluid likewise produces this protection in the organism of the living rat. Now, an immunity that manifests itself without the aid of any activity of the cell must undoubtedly be regarded as being of a humoral character” (loc. cit. p. 202).

Let us begin by analysing the facts as presented in rats into whose subcutaneous tissue we have injected anthrax virus. A certain number of them resist, without exhibiting any lesion other than a certain exudative inflammation at the seat of inoculation. The exudation is, in this case, very rich in leucocytes which quickly exert their phagocytic function and destroy the ingested bacilli. In this reaction it is the microphages that play the chief part, the macrophages intervening later and in a much less pronounced fashion. Usually, however, the inoculated rats exhibit a more serious illness: the bacilli multiply at the point of inoculation and excite the formation of an extensive oedema, rich in serous fluid, transparent, and very poor in leucocytes. It is only later that these cells intervene in any considerable number. The exudation becomes thicker and turbid, the numerous white corpuscles devour the bacilli and cause their disappearance. Under the influence of this marked reaction the animals in most cases recover, as has already been established by Frank^[223]. But even in those individuals which succumb to anthrax death occurs more or less tardily, an examination of the internal organs then revealing a considerable phagocytic reaction. The spleen, often of enormous size, contains numerous macrophages which are filled with normal or more or less altered bacilli. In the liver these macrophages, which have devoured several microphages and some bacteria, are also found (Figs. 22 and 23).

When instead of bacteria in the condition of rods, anthrax spores are inoculated subcutaneously or into the anterior chamber of the eye, we can observe their germination. There is developed a whole generation of bacilli which behave like those we have already described, that is to say, they excite an exudation and are ultimately digested within the phagocytes (Figs. 24 and 25). All these phenomena of phagocytosis I described in detail more than ten years ago in my memoir on the anthrax of rats^[224]. Since then not a single fact has been brought forward to invalidate the results there set forth.

How is this paradoxical fact to be explained, that anthrax which grows in the body of the rat, there setting up a disease more or less grave and sometimes fatal, is so readily destroyed by the serum and blood when removed from the organism? From numerous experiments, carried out by Hankin^[225] and by Roux and myself^[226], it has been demonstrated that the bactericidal power of the fluids of the rat cannot be invoked as the cause of the animal’s resistance to anthrax. Those rats which show themselves very susceptible to this disease and die from anthrax infection, furnish, nevertheless, a serum that will prevent anthrax in other rats, and which will protect even mice into which the bacilli have been injected. Rats into which we inoculate on one side of the body a little anthrax culture, and on the other side the same quantity of bacilli mixed with blood serum from the same animal, manifest oedema at the former place only. It is from this latter point that the general infection takes place, the side where the anthrax bacilli mixed with serum was introduced remaining unaffected. Sawtchenko^[227], who has investigated the immunity of the rat in my laboratory, has to the facts just mentioned added the observation that when the injection of bacilli causes haemorrhage the rat survives. When, on the contrary, the injection is made with a fine needle and without effusion of blood, the rat contracts a fatal anthrax.

It follows from these facts that the blood, immediately it has escaped from the vessels, undergoes a change in its composition and becomes bactericidal for the anthrax bacillus, whilst, when it is circulating in the organism, it exhibits no such power. Sawtchenko has studied the substance in the serum which kills the bacilli and has demonstrated that it will resist heating to 56° C.; even when heated to 61° C. the serum still exercises a certain amount of bactericidal power for very attenuated bacilli (Pasteur’s first vaccine). Researches on the distribution of this bactericidal power in the living rat have convinced Sawtchenko that none of it passes into the fluid of the passive oedema set up by the slowing of the circulation, nor into that of the active oedema developed as the result of the inoculation of anthrax bacilli. He observed that even the bacillus of Pasteur’s first vaccine grows abundantly in the oedematous fluid produced by the injection of virulent bacilli. The peritoneal lymph, however, exerts a very marked bactericidal action on the bacilli. Having demonstrated this fact Sawtchenko put to himself the question: May not the great difference between the action of these fluids depend on the fact that the lymph is rich in leucocytes, whilst in the fluid of the oedema they are almost absent? Pursuing this question, Sawtchenko made a comparative study of the bactericidal power of the serum, prepared outside the body, and of the blood plasma obtained by means of an extract of the heads of leeches, and he concluded from his researches that the bactericidal substance circulates in the plasma of the living rat and that it is not derived from the microphages, but must be looked upon rather as a secretion of the macrophages in the blood and of endothelial cells. This result was not confirmed by Gengou^[228], who also took up the study of this important question in my laboratory. Instead of preparing the plasma by means of the addition of an extract of leeches he made use of a method much more perfect and free from sources of error. He introduced no foreign substance capable of affecting the results of his experiments. Collecting the rat’s blood in paraffined tubes, and centrifugalising it in similar tubes, he obtained a fluid which approaches much more closely the plasma of circulating blood than does serum. This fluid, however, will coagulate at the end of a fairly long interval, which proves that it cannot be looked upon as blood plasma. Gengou examined the bactericidal power of the fluid portion of the “plasma,” obtained by the process just described, on the anthrax bacillus, and also that of serum prepared in tubes in the ordinary way. The difference between the two fluids is very marked; whilst the serum destroys the bacilli sown in it very rapidly and dissolves their contents, the fluid of the “plasma” has no similar action. These results, confirmed several times, demonstrate very definitely that the plasma of the circulating blood does not contain any bactericidal substance. This, during the life of the animal, is found inside leucocytes and only escapes from them when the cells burst or undergo profound lesions, this taking place when the clot is formed and when the serum is prepared outside the organism, or in the effused and coagulated blood, or again in the peritoneal lymph during phagolysis. This phagolysis is inevitably produced as a result of rapid injection of foreign fluids into the peritoneal cavity, e.g. of broth or of physiological salt solution, containing bacteria in suspension.

The facts we have brought together on the subject of anthrax in rats form a whole whose several parts are in complete harmony. The phagocytes of this species of rodent contain a bactericidal ferment, a kind of cytase, which resists temperatures approaching 60° C. This cytase is very active against the bacilli, but in the living animal it can only act within the phagocytes, or, in a transitory and incomplete fashion, outside these cells, when phagolysis is taking place in the peritoneal cavity. The resistance offered by the rat to anthrax depends, then, on this phagocytic activity. For its manifestation it is necessary, first of all, that the phagocytes should manifest a positive chemiotaxis for the bacilli, and then that they should seize and ingest these organisms. These are the vital acts that decide the result of the struggle. When the phagocytes show themselves inactive the bacilli multiply in the oedematous fluid which contains no bactericidal cytase, and pass into the plasmas of the lymph and of the blood, which also are incapable of killing these bacteria. The animal may, then, die of anthrax, in spite of the presence in its body of a large quantity of bactericidal cytase which is to be found in situations to which the bacilli have not penetrated. In those cases, on the other hand, where the phagocytes accomplish their function, where they rush up to the menaced point and devour the inoculated bacteria, these bacilli are placed in contact with the intracellular cytase and undergo complete digestion. The organism in this case gets rid of its enemies and victoriously resists infection.

Anthrax in rats, then, presents one of the most instructive examples of natural immunity. But the detailed analysis of the mechanism of this resistance demonstrates very clearly the great part played by the phagocytes in this process. In this respect the organism of the rat presents, in a general fashion, a great analogy to the natural immunity of the dog, of birds, and of other representatives of the animal kingdom that we have examined. Under these conditions it is useless to insist at any length on other examples of resistance against anthrax which, moreover, have relation much more often to a natural immunity against attenuated bacilli than to one against true anthrax virus. Rabbits and guinea-pigs, so sensitive to this virus, often resist the inoculation of Pasteur’s vaccines. The rabbit is, in general, refractory to the first anthrax vaccine; it may even resist the second vaccine. The guinea-pig, a more sensitive animal, does not exhibit any natural immunity except against the first vaccine. In all these cases the mechanism is similar to that which the rat and the dog oppose to virulent anthrax. The bacilli, into whatever part of the body they are injected, set up an exudative inflammation which brings up a large number of leucocytes to the point menaced. These cells readily exert their phagocytic function and rid the organism of the introduced bacteria. In order to obtain a complete grasp of the part played by this reaction it will be found useful to inject beneath the skin of one ear of a rabbit a little anthrax vaccine and beneath the skin of the other the same quantity of virulent bacilli. The difference between the reaction in the two cases is very striking. The ear inoculated with vaccine soon becomes the seat of a circumscribed inflammation with a purulent exudation, all the bacilli in which have been devoured by the leucocytes. The other ear, on the contrary, presents, around the injected virus, only a serous or blood-tinged exudation containing no, or few, leucocytes; the bacilli are found free in the liquid and multiply without let or hindrance. Meeting with no opposition the virus becomes generalised throughout the organism and brings on death by anthrax septicaemia. Rabbits, into which anthrax vaccines only are introduced, oppose to the invasion of the bacilli a leucocytic barrier which arrests their extension. The natural immunity of the sheep, rabbit and guinea-pig is also a phagocytic immunity, but it is only capable of being exercised against bacilli previously attenuated in virulence. The researches of Mme Metchnikoff^[229] on the reaction of the phagocytes of these animals to the bacilli of Pasteur’s two anthrax vaccines have demonstrated the importance of the destruction of these bacilli by the leucocytes. All the other examples of natural immunity against anthrax are also merely relative. The fowl that resists an anthrax virus strong enough to kill an ox or a horse, succumbs to a special variety of anthrax cultivated by Levin^[230]. The dog, as we have seen, in spite of its pronounced natural immunity against anthrax, is killed by the special anthrax bacillus prepared by Martel.

In this immunity against anthrax we have to deal with a bacillus capable of living and reproducing itself in extremely varied media. Hence the reason, it may be said, that the bactericidal influence of the fluids is so little pronounced in this case. To bring it into relief we must, therefore, choose a bacterium less capable of adapting itself to the chemical composition of various culture media. In this matter we cannot do better than select pathogenic spirilla of extremely delicate nature and analyse the mechanism of the natural immunity of certain species of animals with respect to them. It must not be forgotten, however, that here we are making use of representatives of an infinitely small minority of pathogenic bacteria, the majority resembling the anthrax bacillus in the facility with which they can be cultivated in all sorts of nutritive media.

The spirillum of recurrent fever of man (Spirochaete obermeyeri) was the first pathogenic microbe found in an infective disease distinctly human. Discovered a third of a century ago, it has passed through the hands of the most skilful bacteriologists, who have tried all possible methods of cultivating it outside the body. Koch himself tried to solve the problem, but, in spite of his incomparable skill, did not succeed. Later, Sakharoff^[231], at Tiflis, discovered a spirillum very similar in appearance which produced a fatal septicaemia in the goose. He, also, tried to cultivate it, but in vain. His successors have not been more fortunate in this respect. Here, then, are two micro-organisms, against which natural immunity should be easily obtainable and in a fashion quite other than that against anthrax. Nothing, indeed, is more frequent than examples of very stable natural immunity against the spirilla of Obermeyer and of Sakharoff. As I wished to obtain a clear idea of the mechanism by which the guinea-pig resists injections of the spirillum of goose septicaemia (Spirochaete anserina) I made injections of goose’s blood, containing a quantity of these organisms, into the peritoneal cavity of guinea-pigs. This injection, as usual, causes the disappearance of most of the leucocytes, as the result of a very marked phagolysis. We know that, under these conditions, the damaged leucocytes allow a certain quantity of the bactericidal cytase to escape. In spite of this the spirilla remain intact and exhibit very active movements in the peritoneal exudation. This exudation, after a period of phagolysis, which lasts for two or three hours, begins to be stocked again with leucocytes which come up in increasing numbers, a fact that does not prevent the spirilla moving about with great rapidity. Even seven hours after the injection of goose’s blood we still find many extremely active spirilla among a large number of recently migrated leucocytes, some of which even at this stage contain red corpuscles of the blood of the goose. It is not until later that the ingestion of these spirilla by the leucocytes commences, the leucocytes at last damaging and completely destroying them. This act of phagocytosis may be readily observed in hanging drops of the peritoneal exudation of inoculated guinea-pigs. The attention of the observer is drawn to certain macrophage leucocytes which throw out one or two conical-looking processes (Figs. 26–28). These pseudopodia attach themselves to spirilla which exhibit very violent movements as though wishing to extricate themselves from the grasp of the leucocyte. Sometimes the spirillum succeeds in escaping, but usually it becomes surrounded by the protoplasm and sinks more and more deeply into the substance of the leucocyte. Even when almost surrounded the free part of the spirillum still continues to move (Figs. 29–31). These movements cease only after the complete ingestion of the spirillum. Once inside the phagocyte the spirillum is digested and soon becomes unrecognisable.

Recently, Sawtchenko^[232] took advantage of an epidemic of recurrent fever at Kazan to make similar investigations on the natural immunity of the guinea-pig against Obermeyer’s spirillum. He observed that these organisms, when injected into the peritoneal cavity, remained there, alive, for 24 and even 30 hours, whilst these same spirilla, when kept at 37° C. outside the organism in their natural medium, died at the end of some (4–7) hours. The injection of human serum containing spirilla into the peritoneal cavity of guinea-pigs set up a phagolysis succeeded by a considerable afflux of leucocytes. In spite, however, of the arrival of quite an army of these cells, the spirilla continued to move rapidly; for a long time they evaded the phagocytes which, however, in the end always ingested them. But it is only the macrophages which fulfil their phagocytic function (Figs. 32 and 33); the microphages obstinately exhibit an absolutely negative chemiotaxis. Now, as the macrophages do not make their way into the peritoneal cavity until after the microphages have appeared, it is easy to understand that phagocytosis can only take place at a late period. Sawtchenko came to the conclusion that “in the peritoneal cavity of animals naturally refractory, the spirochaetes perish as the result of a slow phagocytosis and not from the action of the bactericidal substances of the fluids.” In conformity with this result this observer has often noted the ingestion of living spirilla by the macrophages, in hanging drops of the peritoneal exudation of inoculated guinea-pigs. The phenomenon corresponds exactly to that described in connection with the spirillum of the goose.

In spite of the great difference between the spirillum and the anthrax bacillus from the point of view of their adaptation to surrounding media, the general result is the same with both these microbes: animals endowed with natural immunity get rid of them through the agency of their phagocytes.

It would be impossible and even useless here to pass in review all the cases of natural immunity against infective micro-organisms. We must consequently limit ourselves to several examples which may have an interesting bearing on the study of the problem as a whole. The spirilla, whose history we have just recorded, remain in the peritoneal fluid, without change of form, up to the moment when they are captured by the macrophages. Let us see by what mechanism the natural immunity against micro-organisms, characterised by a very special sensitiveness to external influences and by a considerable change of shape, is produced. The cholera vibrio and its allies best satisfy this postulate. When they find themselves placed under unfavourable conditions, these vibrios immediately become transformed into small spherical bodies which are much more like cocci than vibrios. The cholera vibrio is pathogenic for the laboratory rodents, especially for the guinea-pig, when a fairly large quantity of a culture is injected into the peritoneal cavity. Against smaller doses, however, the natural immunity is a most marked one. If we take a race of the cholera vibrio of medium virulence, and inject into the peritoneal cavity of guinea-pigs a sublethal dose of a culture, the following phenomena may be observed^[233]. The inoculated vibrios move actively in the peritoneal fluid, from which almost all the leucocytes have disappeared. There remain only a few lymphocytes which appear to be indifferent to the influences that set up a real phagolysis. But, little by little, fresh leucocytes come into the exudation and engage in a struggle with the vibrios which, so long as they are free, retain their curved form and complete motility. The microphages, especially, swarm into the peritoneal cavity. Some of them begin to ingest vibrios, but this phagocytosis is at first slight. Later it becomes much more active. The microphages and macrophages seize vibrios that are evidently living and uninjured, which, sometimes, may be observed inside the vacuoles of the leucocytic contents exhibiting very lively movements. Once ingested, however, many of the vibrios become transformed into round granules. This change of shape is constant when inside microphages, but is completely absent when inside macrophages (Figs. 34 and 35). Finally, the phagocytosis becomes complete, and the organism gets rid of the vibrios solely by means of this reaction. Even seven hours after injection of the vibrios, when the peritoneal fluid, crammed with leucocytes, has become thick and turbid, there still remain a few scattered vibrios which always retain their shape and their normal activity. A drop of this exudation, maintained at 38° C. outside the organism, gives, in a few hours, an abundant culture of very active vibrios. It must, therefore, be concluded that the fluid part of the exudation was powerless to destroy the vibrios or even to render them motionless, whilst the living leucocytes have shown themselves capable of ingesting and digesting them. The peritoneal exudation, withdrawn at a period when it no longer contains any free vibrios, still gives cultures of the organism for some time. Soon, however, there comes a period when the inoculated exudation remains sterile, this proving that the vibrios, ingested in a living state by the phagocytes, have at length been killed by the microphages and macrophages.