Certain of the hypotheses on acquired immunity are of as ancient origin as are those on natural immunity. For example, it has for long been known that man is constitutionally refractory to certain diseases which are very fatal to cattle. It has also been recognised that after a first attack of a contagious disease, such as small-pox, measles, scarlatina, typhoid fever, etc., man acquires a lasting immunity; and that the same rule applies to domestic animals, for example, cattle that have recovered from cattle plague or sheep that have got better from sheep-pox, become refractory against these diseases.

The discoveries of variolisation and vaccination, as methods of conferring on man a resistance to small-pox, have notably advanced our knowledge upon acquired immunity. The researches on the properties of vaccine had already led to some important results. But it is only since the publication of Pasteur’s investigation, carried out with his collaborators Chamberland and Roux, in the first place, and with Thuillier later, that we have been able to take up the study of acquired immunity in a really scientific manner. The first in this series of discoveries, which have opened up a path so fruitful to science and medical art, was the discovery of the attenuation of micro-organisms. The small cocco-bacillus of fowl cholera after several weeks’ culture in broth was found to have become markedly attenuated in virulence. To Pasteur the idea occurred of testing whether fowls that had resisted the inoculation of these attenuated organisms had acquired any real immunity against virulent fowl cholera. Experiment confirmed his expectation and led to the discovery of the vaccine against this disease. The method was at once applied to other infective epizootic diseases and shortly afterwards Pasteur, Chamberland and Roux found a method of preserving sheep and cattle from anthrax. To attain this end it was found necessary to prevent the bacillus from producing spores (in this they succeeded by cultivating it in broth at a temperature of 42°·5 C.), because these spores fix the virulence and prevent attenuation. Having overcome this main obstacle, Pasteur and his collaborators found that their cultures, thus deprived of spores, become attenuated on exposure to the air and so become transformed into vaccines. They were thus able to prepare their two anthrax vaccines which soon found such wide application in practice. A few years later, Pasteur and Thuillier discovered the vaccines against swine erysipelas and, in collaboration with Roux and Grancher, Pasteur made the first application of his discoveries to the vaccination of man against rabies.

The path thus opened up was traversed by many other investigators and led to many very remarkable discoveries. Vaccination with micro-organisms became a recognised method and in the hands of Arloing, Cornevin and Thomas, soon found its application to symptomatic anthrax. The next step in this onward progress of science was taken when Salmon and Smith, working at hog-cholera, demonstrated the possibility of vaccinating not only with hog-cholera bacilli, but also with culture fluids in which these organisms had developed. These fluids, when completely deprived of micro-organisms by filtration, protected the experimental animals from virulent hog-cholera. This discovery, at first sceptically received, was soon confirmed and extended by the work of other investigators. Beumer and Peiper extended it to the experimental disease set up by the typhoid bacillus in small laboratory animals; Charrin applied it to the disease that he produced by means of the bacillus of blue pus; and Chamberland and Roux prepared vaccines from the soluble products of the septic vibrio and of the bacillus of symptomatic anthrax. And now, as the result of these investigations, vaccinations by microbial products are in everyday use in all research laboratories. The vaccinations now used (anthrax, symptomatic anthrax, swine erysipelas and rabies) are still being carried out by means of the inoculation of living viruses.

The comparative history of acquired immunity is still very incomplete. The facts known concerning the adaptation of unicellular organisms to all kinds of injurious influences of a physical or chemical nature enable us to perceive that acquired immunity is just as general in living beings as is natural immunity; but it is impossible, in the present state of our knowledge, to confirm this hypothesis by exact and experimental data. The reason for this lies in the great difficulty we have in carrying out experiments on the lower animals. The majority of the Invertebrata in captivity do not remain alive long enough and can not be sufficiently often inoculated for us to obtain in them a well marked acquired immunity against micro-organisms. Kowalevsky^[306], the celebrated Russian zoologist, has tried to overcome these various difficulties by making use of Myriapods. He found first that Scolopendrae, when inoculated with anthrax bacilli, die therefrom during the heats of summer, the blood containing a number of anthrax bacilli. But when the temperature does not exceed 17°–18° C., a fairly large number of these myriapods survive. The same survival was observed when Pasteur’s first vaccine was injected. Kowalevsky utilised the Scolopendrae that had resisted the first injection of anthrax bacilli to ascertain whether they had contracted an acquired immunity. The results were not absolutely demonstrative and Kowalevsky sums up his results in the following words: “I cannot say, therefore, that I have succeeded in solving this question of vaccination, but it appears to me very probable” (p. 607).

In view of this doubt, I asked Mesnil to make a fresh attempt, making use of Scolopendrae and inoculating them with anthrax bacilli. These creatures, however, were so delicate and so little capable of remaining alive under the artificial conditions of their captivity, that the attempt soon had to be abandoned. I tried to obtain better results with the larvae of Oryctes nasicornis; here again the difficulties were too great. These insects exhibit a perfect natural immunity against certain micro-organisms, but for others they showed an insurmountable susceptibility. It is very evident, then, that it is not an easy matter to set up an acquired immunity in the Invertebrata.

It was necessary, therefore, to go higher up the animal scale and have recourse to cold-blooded vertebrates. The choice naturally fell on the frog. I asked Dr Gheorghiewski^[307], who was working in my laboratory, to try to vaccinate the Batrachians against pyocyanic disease. I ought first to state that the bacillus of blue pus is pathogenic for the frog, which it kills both at the ordinary laboratory temperature, and at that of the incubator, 30°–37° C. In the first case the fatal dose is much greater than in the second, but it is always easy to induce a fatal infection. In this respect, therefore, the Bacillus pyocyaneus is much better adapted for study than the anthrax bacillus or many other micro-organisms. Gheorghiewski vaccinated green frogs (Rana esculenta), which had been accustomed to the incubator temperature, 30° C., by injecting every 4 to 7 days considerable doses of cultures of Bacillus pyocyaneus heated to 80° C. in order to kill all the micro-organisms. Some (3–4) weeks afterwards the prepared frogs became more resistant to the Bacillus pyocyaneus than were the controls placed under the same conditions. The frogs, inoculated with a fatal dose of the bacilli, clearly exhibited a certain, though slight, degree of acquired immunity. They withstood a dose that was always fatal to the controls or even a dose and a half, but died when injected with double the fatal dose. The lymphatic fluid of the vaccinated frogs feebly agglutinated (1 : 20–1 : 30) the Bacillus pyocyaneus although it still formed an excellent culture medium for this organism. Gheorghiewski satisfied himself that the agglutination was insufficient to assure immunity to the frog. The bacilli agglutinated into clumps were very virulent.

A detailed examination of the phenomena observed in the immunised frogs revealed the following facts. During the earliest stage the bacilli, injected into the dorsal lymphatic sac, were found free in the fluid, retained their form and were not transformed into granules. The bacilli carried by the lymphatic current spread rapidly throughout the body. Very shortly after inoculation, however, some of the leucocytes began to ingest the bacilli which became transformed into spherules within these cells. Later, the phagocytic reaction increased and at the end of 15 to 20 hours all the bacilli were found inside leucocytes. It was easy to demonstrate that these bacilli had been ingested in a living condition. Forty-eight hours after inoculation, no bacilli were to be found in the lymph of the dorsal sac, either inside or outside the cells. But this fluid when sown on nutrient media gave colonies of the Bacillus pyocyaneus up to 15 and even 18 days after inoculation.

We may conclude from these facts that the cold-blooded vertebrata are capable of acquiring immunity to a slight degree and that, in this acquired immunity, a marked phagocytosis may be observed, but no bactericidal action of the fluids.

In order to gain a more complete idea of the mechanism of acquired immunity, it is necessary to observe it in higher vertebrates in which a well developed immunity of this type is readily obtained. Here we must have recourse to mammals and pass in review an ample number of examples, before we attempt to give to our readers a general summary of the question.

For long, researches on acquired immunity were confined almost exclusively to the analysis of the facts observed in animals submitted to anti-anthrax vaccinations by means of the two vaccines of Pasteur. A large number of important facts were thus collected, the more weighty of which must be presented to the reader. But, before entering on the subject, a general orientation on acquired immunity in laboratory animals against vibrios is indispensable as this example dominates, so to speak, the whole of the chapter on acquired immunity against micro-organisms.

Von Behring and Nissen^[308], in their researches on the bactericidal power of serums, examined, amongst others, several specimens of serums coming from animals that had been vaccinated against various micro-organisms. In the majority of the examples given by them the acquired immunity produced no increase in this power, but the blood serum of guinea-pigs that had been immunised against Gamaleia’s vibrio (Vibrio metchnikovi) was found to be much more bactericidal as regards this micro-organism than the serum of normal susceptible guinea-pigs. These authors came to the conclusion that in acquired immunity, at least as regards the vibrio mentioned, the chief part is played by a bactericidal substance which is developed in the fluids of the vaccinated animals. They were content with the mere demonstration of this fact without making any attempt to follow the course of events in the destruction of the vibrios as it occurs in the organism of the vaccinated guinea-pig. R. Pfeiffer^[309] in collaboration with Issaeff sought to fill this gap. But, instead of taking Gamaleia’s vibrio, these observers concentrated their attention on the study of the acquired immunity of guinea-pigs against the cholera vibrio. As this vibrio is as a rule less virulent than Gamaleia’s vibrio, it was necessary, in order to obtain a fatal infection, to inject it, not into the subcutaneous tissue but into the peritoneal cavity. We have already seen (Chapter VI) that the cholera vibrio when inoculated into the peritoneal cavity of the guinea-pig, there meets with a vigorous resistance on the part of the leucocytes which seize the living and virulent vibrios and digesting them rid the animal of their presence. But when the dose of the vibrios is increased, they multiply in spite of the phagocytic reaction; they are found swarming in the peritoneal cavity, whence they invade the lymphatic and blood vessels and cause the death of the animal. It is easy, then, to induce a fatal infection of the guinea-pig with the cholera vibrio. But it is also easy to vaccinate these animals against this experimental disease. We have only to inoculate them with a non-fatal quantity of living cholera vibrios, or to inject into them a culture in which the vibrios have been killed by heat, or some of the culture fluid from which the vibrios have been removed by filtration. All these methods soon produce an acquired immunity in guinea-pigs. If, when this has been brought about, a little blood is withdrawn and to the serum a small quantity of cholera vibrios is added, in vitro, we can readily demonstrate their disappearance, under the influence of the bactericidal substance dissolved in the fluid. In this respect there is, then, a marked analogy with the fact established by v. Behring and Nissen as regards Gamaleia’s vibrio.

When into the peritoneal cavity of vaccinated guinea-pigs a certain quantity of cholera culture containing virulent and very motile vibrios is injected, we find that in the peritoneal fluid drawn off by means of a fine pipette, the vibrios have undergone profound changes in the refractory organism. Even a few minutes after the injection of the vibrios, the leucocytes disappear almost completely from the peritoneal fluid; and only a few small lymphocytes and a large number of vibrios, the majority of which are already transformed into granules, are found (fig. 39); and there is presented a most typical case of Pfeiffer’s phenomenon. Alongside the round granules may be seen swollen vibrios, and others which have kept their normal form, but all are absolutely motionless. Some of these granules are gathered into small clumps, others remain isolated in the fluid. When to the hanging drop containing these transformed vibrios a small quantity of a dilute aqueous solution of methylene blue is added, we observe that certain granules stain very deeply, whilst others take on merely a very pale tint, scarcely visible. Many of these granules are still alive, because it is easy to watch them develop outside the animal and elongate into new vibrios. A large number of the granules, however, no longer exhibit any sign of life and are evidently dead. R. Pfeiffer and certain other observers affirm that the granules may be completely dissolved in the peritoneal fluid just as a piece of sugar dissolves in water. We have repeatedly sought for this disappearance of the granules in hanging drops of the peritoneal fluid, without being able to find any diminution in the number of these transformed vibrios, even after several days; nor have we been able to observe the phenomenon of the solution of the granules. It is at any rate indisputable that this granular transformation is a manifestation of very profound lesions undergone by the cholera vibrios under the influence of the peritoneal fluid of the immunised animal.

An attempt has been made to define the mechanism of Pfeiffer’s phenomenon more exactly, and Fischer^[310] has sought to refer it to osmotic action, exercised by the salts of the fluids in which the vibrios are suspended. These vibrios, under the action of media richer or poorer in salts than is the fluid in which they had developed, are said to present an increase of their internal pressure, in consequence of which the vibrios swell up or allow a spherical droplet of protoplasm to escape at one of their poles. This explanation was inadequately supported by its author and cannot be regarded as proved. On the other hand, one is compelled to the conclusion that the granular transformation is due, as we shall see later, to a fermentative action of the peritoneal exudation.

Whilst the vibrios are undergoing this transformation in the peritoneal cavity of an immunised guinea-pig, the animal recovers from a malaise that is quite transitory and continues to live, whilst normal unvaccinated guinea-pigs die, an enormous quantity of vibrios swarming in the peritoneal exudation. The difference between the two animals is most striking, and we can readily understand that Pfeiffer was so impressed by it that he was led to attribute the acquired immunity of his guinea-pigs solely to the granular transformation set up by a bactericidal substance contained in the fluids of the immunised animals.

The ease with which we can gain an idea of the change of form in the vibrios under the influence of the fluids of the body, greatly aids the study of the bactericidal substance. Before passing to the question of the part played by this substance in acquired immunity we must consider for a moment the principal properties of this acquired immunity. Very manifest in the peritoneal fluid, the power of causing Pfeiffer’s phenomenon is equally evident in the blood serum of immunised guinea-pigs, as has been demonstrated by Bordet. A drop of this serum, when quite fresh, readily and rapidly transforms a number of vibrios into granules. When the serum is kept for several days or has been heated to 55° C. for an hour, the total disappearance of the substance which produces Pfeiffer’s phenomenon is brought about. This at once betrays the presence of microcytase in the fluids of guinea-pigs that have acquired immunity against the cholera vibrio. Yet the blood serum and the peritoneal fluid of these animals, having been deprived of their microcytase by heating to 55° or 56° C., still retain a remarkable power over the vibrios. These organisms no longer undergo granular transformation, under the influence of the heated body fluids, but they are deprived of all power of motion, agglutinate into clumps and acquire a special susceptibility to the action of cytase. Soon after the discovery of Pfeiffer’s phenomenon, I^[311] was able to demonstrate that this granular transformation can be obtained in vitro under the following conditions. Prepare a hanging drop with the blood serum of a guinea-pig vaccinated against the cholera vibrio, a serum which has lost the power of transforming, by itself, the vibrios into granules. Add to it a drop of the peritoneal lymph of a normal unvaccinated guinea-pig; this lymph contains dead or living leucocytes and is, by itself, also incapable of producing Pfeiffer’s phenomenon. When, to the mixture of these two fluids, which are inactive when they are employed separately, a few cholera vibrios are added, they are quickly transformed into granules. This transformation, obtained in vitro, is remarkably like that produced in the peritoneal cavity of the vaccinated animal.

Jules Bordet^[312], working in my laboratory, made a very complete investigation of Pfeiffer’s phenomenon outside the animal body and found that, in my experiment, the peritoneal lymph can be replaced by the blood serum of the vaccinated guinea-pig without thereby in the least hindering the granular transformation. After making a specially thorough study of the question he has come to the conclusion that Pfeiffer’s phenomenon is the result of the action of two substances. One of these is found in the blood serum and in the peritoneal fluid of guinea-pigs vaccinated against cholera, heated to 55°–56° C. or deprived by some other means of their individual power of transforming vibrios into granules. This substance resists this temperature and only loses its activity on being heated to 68°–70° C. The second of the two substances, that found in the peritoneal lymph or in the blood serum of the normal guinea-pig, is, on the other hand, destroyed at 55°–56° C. and is nothing but the ordinary cytase (or alexine) present in the fluids of normal animals.

The facts we have described with regard to Pfeiffer’s phenomenon in the body fluids of immunised animals must, then, be interpreted as follows. The fresh peritoneal exudation or blood serum of these animals readily produces the granular transformation, because in these fluids both the two necessary substances are found. But as microcytase is a very unstable substance which, under the influence of time or heating to 55°–56° C., is destroyed, the fluids of immunised animals are very readily deprived of it. The blood serum then, after being some time outside the body, becomes incapable of transforming vibrios into granules; but when to it is added a small quantity of the cytase, found in the blood serum or in the peritoneal lymph of the normal guinea-pig, the transformation takes place with great rapidity. To the serum of the immunised animal, which has become inactive, is restored its property of setting up Pfeiffer’s phenomenon. This interpretation, formulated by Bordet, corresponds to the whole of the known data and is now generally accepted.

As the fluids of immunised animals, that have become incapable of transforming vibrios into granules, still retain their power of rendering these organisms motionless and of uniting them into clumps, it might be asked whether this agglutinative substance might not be the substance, stable under heat, which is necessary for the production of Pfeiffer’s phenomenon. For some time, indeed, it was believed that this phenomenon is due to the microcytase acting on vibrios which have first been modified by the agglutinative substance. This latter substance resists heating to 55°–56° C., is only destroyed at higher temperatures, and is retained in the blood serum long after the cytase has entirely disappeared. The analogy between the agglutinative substance of the fluids of animals that have acquired immunity and the substance in the same fluids that is stable under heat is undeniable, and yet these two substances are not identical. A whole series of observations, which we shall presently describe, demonstrate this thesis clearly. A serum may be highly agglutinative without being capable of bringing about the transformation of vibrios into granules; the converse also holds good. The substance which sets up Pfeiffer’s phenomenon, and which is found in the fluids of immunised guinea-pigs, is a “fixative substance” analogous to those we have already met with in the serums of animals so adapted that they are able to resorb the various cell elements. As in the resorption of cells, so also in the destruction of micro-organisms, the fixatives are specific. The substance which aids the transformation into granules is not only distinct from the fixatives which sensibilise red blood corpuscles or spermatozoa, but also from the fixatives which sensibilise bacteria. This specificity has been demonstrated and carefully studied by Pfeiffer, who has shown that it may even serve to distinguish species of bacteria. The serum of a guinea-pig which has been immunised against the cholera vibrio, will render sensitive these vibrios, and these only, to the action of the microcytase. Even allied vibrios, such as various water vibrios, for example, are not sensitive to the fixative of anticholera serum. On the other hand, the serums obtained after the inoculation of these aquatic vibrios are incapable of producing granular transformation in the cholera vibrio.

When we inject into one and the same animal several species of vibrios we obtain a serum or a peritoneal fluid which produces Pfeiffer’s phenomenon with the vibrios of all the species which have been used to make the inoculations. This antivibrio serum contains only a single cytase for the vibrios, but contains as many different fixatives as there were species inoculated.

The transformation of vibrios into granules, when produced in a high degree against virulent vibrios, under the influence of the body fluids of immunised animals, affords a very valuable indication of the simultaneous presence of cytase and of specific fixative. As we have already stated, at the commencement of this account of the acquired immunity of guinea-pigs against the cholera vibrio, Pfeiffer’s phenomenon is manifested in the peritoneal fluid of these animals in a very short time (5 to 20 minutes) after the inoculation of the vibrios. This proves that in this fluid the two substances really occur together, and that the fixative and the cytase are in solution in the plasma of the exudation. Is it the same in every part of the body of immunised guinea-pigs? If, instead of introducing the cholera vibrio into the peritoneal cavity, we inject it into the subcutaneous tissue or into the anterior chamber of the eye of these animals, Pfeiffer’s phenomenon does not make its appearance. The vibrios, isolated or collected into small clumps, do not undergo granular transformation; they keep their normal vibrio form and remain alive much longer than in the peritoneal cavity. Some of them may be found still living 24 hours after subcutaneous injection and several (4–6) days in the anterior chamber of the eye. Nor can Pfeiffer’s phenomenon be observed when the cholera vibrio is introduced into the oedema of the foot, produced in consequence of the slowing of the circulation, the vibrios remaining alive for a fairly long time. These facts clearly indicate that in the fluid thrown out in passive oedema, just as in the aqueous humour of the eye or in the subcutaneous tissue, the two substances necessary to set up the granular transformation are not present. Are both of them absent or only one? This question is easily answered on adding to the fluids mentioned normal guinea-pig’s serum, a serum which, by itself, is incapable of producing Pfeiffer’s phenomenon. Bordet^[313] has made these experiments and found that when to the fluid of the passive oedema of the immunised guinea-pig normal serum is added, this fluid transforms the cholera vibrio into granules, but does so in less degree than does the serum of the same immunised guinea-pig, when heated to 55°–56° C., to which normal serum has likewise been added. There is, then, reason to conclude that the fluid of the oedema does not contain any cytase, but contains a certain quantity of cholera fixative, less, however, than that which is found in the blood serum. As to the aqueous humour of the eye of immunised animals, analogous experiments have demonstrated that it contains neither of the two substances necessary for the production of Pfeiffer’s phenomenon.

With the help of the facts I have here summarised, we arrive at the following conclusion. In the animal that is immunised against the cholera vibrio, microcytase is found in the peritoneal exudation; it does not pass, however, either into the fluid of the passive oedema or into the aqueous humour of the eye; the cholera fixative is found in the peritoneal fluid and passes into the oedema, but does not penetrate into the fluid of the eye. This indicates that microcytase is found in fluids rich in leucocytes, but is absent from those which contain very few or none of these cells.

The introduction of vibrios into the peritoneal cavity of immunised guinea-pigs at once produces Pfeiffer’s phenomenon, and at the same time causes the disappearance of the majority of the leucocytes from the peritoneal lymph. We have already had occasion, several times, to speak of this phagolysis, because it is produced as a sequel to the injection into the peritoneal cavity of blood, spermatic fluid, and all kinds of fluids. The greater the quantity of fluid injected and the greater the difference of the temperature between it and the contents of the normal peritoneum the more vigorous is phagolysis.

Pierallini^[314], working in my laboratory, studying phagolysis in the peritoneal cavity of the guinea-pig, has obtained several results worthy of attention. Of all the fluids used by him, such as water, broth, filtered cultures of micro-organisms and physiological saline solution, the last of these caused the least intense phagolysis, yet one sufficiently well marked. Immediately after the injection of any of the above fluids the number of leucocytes in the peritoneal lymph diminishes very considerably, the cells being found collected in clumps on the omentum. Many of them exhibit signs of enfeeblement and of partial destruction. Alongside the leucocytes are found fibrinous masses, this affording evidence that some of the leucocytes have been greatly damaged and have given up the fibrin-ferment which induces coagulation of the fibrin. When Pierallini injected fluids containing coloured powders in suspension, such as Indian ink and vermilion, he observed that these substances accumulated on the greater omentum, which became stained black or red. Microscopical examination revealed the existence of a not very intense phagocytosis and a number of free coloured granules in the midst of filaments of fibrin.

The leucocytes which, during this phagolysis, allowed the fibrin-ferment to escape might also give up a certain amount of their microcytase. This microcytase would pass into the peritoneal fluid and give rise to Pfeiffer’s phenomenon. If this hypothesis be correct, the suppression of phagolysis would result in the absence of the transformation of the vibrios into granules. It is not a difficult matter to verify this hypothesis as we have a means of preventing phagolysis or at least of reducing it very considerably. Issaeff^[315], in an investigation carried out in Pfeiffer’s laboratory, demonstrated that an intraperitoneal injection of physiological salt solution, broth, urine, etc., reinforces the leucocytes and brings them up in large numbers into the peritoneal cavity. It is easy to foresee that such an injection would serve to diminish the intensity of the phagolysis. In fact, if we first inject a few cubic centimetres of physiological salt solution or of fresh broth into the peritoneal cavity of a guinea-pig, and if, on the following day, we repeat the same operation, we shall find that after the second injection phagolysis is much less powerful than after the first. Pierallini, who repeated these experiments, observed that the phagocytosis of the coloured granules is much more complete in the guinea-pigs that were treated by a preliminary injection into the peritoneal cavity. The amount of fibrin on the omentum is in this case much less, and the phenomena as a whole show that in these guinea-pigs the damage to the leucocytes is very considerably attenuated.

We have been able to demonstrate that in the case where phagolysis is thus diminished, Pfeiffer’s phenomenon is not produced or is manifested in a very feeble degree. If the experiment succeeds, the fluid taken from the peritoneal cavity of a guinea-pig prepared the day before and then injected with a culture of cholera, is opaque and thick, like pus. It contains a mass of leucocytes in good condition, a large number of which gorge themselves in a few minutes with a number of vibrios. The plasma of this exudation contains few vibrios, and these retain their normal form and do not exhibit, save exceptionally, a granular change. A little later there remain no free vibrios; they are all contained within leucocytes. Pfeiffer^[316] declared himself against the facts I have just summarised, because he was never able to prevent the granular transformation of the vibrios, in spite of the preparatory injection of sodium chloride. Abel^[317], who repeated the experiments, expressed an intermediate view: in guinea-pigs prepared by injections the day before, he observed that one portion of the vibrios became transformed into granules, whilst another became the prey of the leucocytes. The fact is, the suppression of phagolysis demands special conditions: the broth that is injected must be freshly prepared, and before its introduction into the peritoneal cavity it must be heated to 37°–39° C. Even when these precautions are taken it sometimes happens that the experiment is not very successful. In making the experiment we must be guided by the appearance of the peritoneal fluid withdrawn into the small glass pipettes. If the fluid which enters the tube is clear or scarcely clouded, it indicates that phagolysis has taken place, in spite of the preparatory injection. The experiment is successful in those cases where the peritoneal exudation is very cloudy and resembles pus.

As the demonstration of the suppression of Pfeiffer’s phenomenon as well as that of phagolysis is of fundamental importance, I asked M. Garnier^[318] to carry out further experiments with the object of setting the question at rest. Using a whole series of fluids for the preparatory injection he found that fresh broth gives the best results. In guinea-pigs in which the phagolysis had been reduced to a minimum, phagocytosis commenced immediately after the injection of the vibrios. In from two to five minutes many vibrios are found inside the leucocytes, the free vibrios now being few in number and not exhibiting Pfeiffer’s phenomenon. Garnier in his memoir gives photographic reproductions of leucocytes crammed with vibrios; these should convince the veriest sceptic. Since the publication of this paper no objection has been advanced, and this question of the suppression of the granular transformation of vibrios may now be regarded as definitely settled. I have since demonstrated this feature to many observers, all of whom have assured themselves of its accuracy. It must, then, be accepted that Pfeiffer’s phenomenon is not produced in the peritoneal cavity except when there is phagolysis. As this fact renders it very probable that the microcytase, which is necessary for the transformation of the vibrios, escapes from the injured leucocytes, it becomes necessary to verify this hypothesis by a series of other experiments. If this hypothesis be well founded, Pfeiffer’s phenomenon should not be observed in those situations in the body where there are no, or almost no, leucocytes already present. These conditions can be realised by injecting cholera vibrios into the subcutaneous tissue or into the anterior chamber of the eye of guinea-pigs that are well vaccinated against the cholera vibrio. Under these conditions, as I have demonstrated in my work on the extracellular destruction of cholera vibrios, the vibrios retain their normal form and are never transformed into granules. Pfeiffer has questioned this result, stating that beneath the skin of vaccinated guinea-pigs the granular transformation is always produced, though in a more feeble fashion and after more delay than in the peritoneal cavity. The contradiction between Pfeiffer’s experiments and my own can, however, be explained. When inoculating the vibrios into the subcutaneous tissue, or during the withdrawal of the exudation formed at the point of infection, small haemorrhages are sometimes produced and a certain amount of microcytase is set free from the leucocytes found in the effusion of blood; these cells also give up to the extravasated blood a portion of their fibrin-ferment. When the experiment is successful, that is to say when no haemorrhage is produced during the operations involved, the subcutaneous exudation contains normal vibrios only, without the appearance of any trace of Pfeiffer’s phenomenon in the fluid.

If the extracellular transformation of the vibrios into granules were the real cause of the acquired immunity, the absence of this phenomenon in the subcutaneous tissue of the vaccinated guinea-pig should lead to the death of the animal. As a matter of fact this does not take place and the animal resists the inoculation of the vibrios. This conclusion is open to one serious objection. As the cholera vibrio in the great majority of cases is incapable of producing a fatal infection when inoculated subcutaneously, even in normal unvaccinated guinea-pigs, this example of immunity must be placed in the category of natural immunity, a kind of immunity which may depend on causes other than those on which acquired immunity depends. To answer this objection it was necessary to select a race of vibrios capable, when injected subcutaneously, of causing death. Mesnil^[319], chief of my laboratory staff, undertook to carry out experiments with the Massowah vibrio, which is regarded by some authors as belonging to the true cholera species. When inoculated subcutaneously into unprotected guinea-pigs, it induces local oedema, in which the vibrios swarm; the infection rapidly becomes generalised and causes the death of the animal in 24 hours. Yet this vibrio, when injected into the subcutaneous tissue of well vaccinated guinea-pigs, is completely resisted by these animals and not the least trace of Pfeiffer’s phenomenon is produced. Under these conditions, a certain number of the vibrios are at first united into masses, but a considerable number remain isolated and motile. Some hours after inoculation the number of clumps diminishes and the isolated vibrios become more numerous, a fact which indicates a certain amount of adaptation of the vibrio to the medium in which it now finds itself. But never, so long as the vibrios remain free in the subcutaneous exudation, do they become transformed into granules.

Salimbeni^[320], in an investigation carried out in my laboratory, endeavoured to satisfy himself whether or no Pfeiffer’s phenomenon is produced in the subcutaneous tissue of a horse that had been hyperimmunised against the cholera vibrio. This animal had, during a period of 14 months, received considerable quantities of these microorganisms, and the serum of its blood transformed the vibrios into granules with great rapidity and intensity. In spite of such favourable conditions for the manifestation of Pfeiffer’s phenomenon, it was never produced beneath the skin of this horse. The vibrios when injected in this position became completely motionless in a very short time, but they kept their vibrio form and remained alive for a number of hours. The exudation drawn off up to 24 hours after inoculation still gave growths of the cholera vibrio.

As it is more easy to introduce, without effusion of blood, the cholera vibrio into the anterior chamber of the eye than beneath the skin, and as the aqueous humour contains no fixative, the absence of the granular transformation in the first of these two situations has been observed even by Pfeiffer himself. The demonstration of this fact presents no difficulty, and for a considerable period we may observe free and perfectly motile vibrios moving about in the aqueous humour. The exudation from the eye contains many of these living organisms, which when sown on culture media made their appearance as colonies even when the fluid has been withdrawn from the eye several days after inoculation.

These carefully established facts show very clearly that the microcytase is only met with in the fluids of the living animal in those situations in which there are many pre-existing leucocytes and under conditions in which the cells undergo a more or less marked phagolysis. This may be corroborated by a decisive experiment. When we inject a suspension of the cholera vibrio directly into the veins of a guinea-pig, well vaccinated against these organisms, and whose serum produces in vitro Pfeiffer’s phenomenon with great rapidity, this phenomenon is not manifested. This experiment has been performed and described by Bordet^[321]. Having injected a suspension of this vibrio into the jugular vein of a guinea-pig vaccinated against the cholera vibrio, he killed the animal an hour later and found, in the blood of the heart, vibrios that had kept intact their form and their property of staining with methylene blue. Cultivation of the blood of the heart, liver and spleen gave growths of vibrios. In another guinea-pig, hypervaccinated against the same organism and inoculated by the same method, the blood drawn off shortly (4–15 minutes) afterwards showed, in preparations treated with methylene blue, well-stained vibrios, of normal form and quite intact. This is the most direct proof of the absence of Pfeiffer’s phenomenon in the blood fluid of a living animal that enjoys a very marked acquired immunity. The intact vibrios were lodged inside the leucocytes.

Levaditi^[322] repeated these experiments in my laboratory and varied the conditions under which the vibrios were injected into the blood vessels. He was sometimes able to observe phagolysis of the leucocytes of the blood and their almost complete disappearance from the peripheral circulation. In these cases the injured leucocytes accumulated in the pulmonary capillaries and masses of them were seen surrounding groups of vibrios that were transformed into granules. It was, however, easy to exclude phagolysis by preparing the animals by means of injections of physiological saline solution or broth. Under these conditions the leucocytes remained in the blood current and very soon ingested the vibrios. Whilst the vibrios that were still free in the blood plasma retained their form and staining power intact, those found inside microphages were already, in great part, transformed into granules. The rapidity with which these phagocytes ingest the vibrios and set up the changes in them is really extraordinary.

In this case, which affords a typical example of the reaction of the animal organism in acquired immunity, we see a very marked and immediate phagocytosis. It is this same process that has already been described as occurring in the peritoneal cavity of vaccinated guinea-pigs in which phagolysis was absent as the result of preparatory injection. In the subcutaneous tissue and in the anterior chamber of the eye, where Pfeiffer’s phenomenon is regularly absent, the phagocytosis follows its ordinary course and causes the destruction of the vibrios. This result has been confirmed repeatedly—see works by Bordet, Mesnil and Salimbeni already quoted.

We need only compare the extension of Pfeiffer’s phenomenon and that of phagocytosis in animals that are immunised against the cholera vibrio, to satisfy ourselves that the former phenomenon is a limited one whilst the latter is general. There might be advanced against the latter conclusion the fact of the absence of any ingestion of the vibrios in the peritoneal fluid of guinea-pigs that are immunised but are not preserved against phagolysis. When a little of the peritoneal fluid is drawn off with small tubes shortly after the injection of vibrios into the peritoneal cavity, as a matter of fact, only a very intense Pfeiffer’s phenomenon is seen, phagocytosis being completely or almost entirely absent. But this procedure is insufficient. If we are to get an idea of what really takes place in the abdominal cavity, the animal must be killed and the peritoneum and especially the omentum very carefully examined. As first demonstrated by Max Gruber^[323] and later by Cantacuzène^[324], the greater omentum is, in these cases, covered with a thick layer which contains a large number of leucocytes, of which some are filled with vibrios; further, this layer contains a mass of vibrios, in part transformed into granules, in part agglutinated or isolated and retaining their vibrionic form intact. As time goes on, the phagocytosis becomes more and more marked, and it is impossible to deny its existence or to attribute to it merely a secondary part.

We have seen that the suppression of Pfeiffer’s phenomenon in the peritoneal cavity and in the blood, or its total absence in the anterior chamber of the eye, does not in the least deprive the vaccinated guinea-pig of its acquired immunity. The animal resists the vibrios perfectly, without these requiring to be transformed into granules in the body fluids. This transformation does take place undoubtedly, but only inside the phagocytes. As already stated in the discussion on natural immunity (Chaps. VI, VII) the vibrios, after being ingested by the microphages, almost immediately undergo within these cells a change in form, very similar to that observed in the real Pfeiffer’s phenomenon. The microphages are often full of a quantity of granules, derived from the ingested vibrios, which in a short time are completely digested. This fact, of such constant occurrence in the phagocytosis of the vibrios, furnishes us with still another proof of the microphagic origin of microcytase.