We have seen how important in the study of acquired immunity against micro-organisms is the demonstration of the protective power of the body fluids. Without being absolutely general, this power is, nevertheless, widely diffused and is found even in certain examples of acquired immunity against micro-organisms belonging to the animal kingdom. Up to the present I have refrained from doing more than point out the presence, in the fluids of the immunised animal, of this protective property and have studied it exclusively in relation to the animal that produces it. We must now pass to the question: How do the protective substances act in the animal which receives them ready formed? This immunity, which has received from Ehrlich the name of passive immunity against micro-organisms, must now be examined.

The study we now propose to enter upon is rendered much easier from our study of the data acquired on the phenomena exhibited in the animal vaccinated with micro-organisms or their products, data already given in the preceding chapter. There is, indeed, a very striking analogy between the two kinds of immunity, and though we draw a sharp line of distinction between them, this is due to the fact that the immunity conferred by micro-organisms or their products requires some time for its development and endures for a long period, whilst the immunity due to the introduction of specific serums into an animal is set up immediately, but endures for a very short time only.

The diseases of the Invertebrata being seldom due to the micro-organisms that produce infections in the higher animals, the question as to whether the Invertebrata can be immunised by means of protective serums has not yet been studied. Still, we already have certain ideas on the protection of lower vertebrates by specific serums. Gheorghiewsky^[464], in my laboratory, carried out some experiments on this point. He found that the serum of mammals (guinea-pig, goat) immunised against the Bacillus pyocyaneus, was under certain conditions capable of protecting the green frog against a dose of this organism that was always fatal to the control animals. When injected along with the Bacillus pyocyaneus, the serum did not prevent a fatal infection; sometimes this infection developed even more rapidly than in the control frogs. It was only when the protective injection was made 24 or, better still, 48 hours before the inoculation of the bacilli, that the protective action became evident. The serum used in these experiments was not bactericidal for the Bacillus pyocyaneus which grew most luxuriantly; but it agglutinated a large proportion of the bacilli. Gheorghiewsky pointed out, however, that frogs injected with cultures agglutinated by the goat’s serum died just as readily as did the control animals. As the phagocytic reaction was invariably very active in those frogs which resisted the virus, after the injection of protective serum, it is very probable that this fluid exercises a stimulant influence on the phagocytes.

This idea of stimulation by anti-infective serums in cases of temporary immunity conferred by these fluids, has already been set forth in my researches on the immunity of rabbits against the Gentilly cocco-bacillus, induced by the serum of vaccinated rabbits. This view, however, has not been favourably received, especially in view of the discovery of the phenomenon of the transformation of cholera vibrios into granules. Pfeiffer himself noted that this transformation took place not only in the peritoneal cavity of vaccinated guinea-pigs but also in the peritoneal cavity of normal guinea-pigs, into which he had injected small quantities of specific serum. As this latter fluid, in Pfeiffer’s hands, was incapable of transforming the vibrios into granules in vitro, he concluded that the cellular elements of the normal animal were endowed with the power of modifying the inactive substance of the specific serum into bactericidal substance. According to this conception the immunity conferred by this serum was not entirely passive since, in order to prepare the substance which transforms and kills the vibrios, the co-operation of the living cells was necessary.

My demonstration of the possibility of obtaining Pfeiffer’s phenomenon in vitro at once turned the balance in favour of the theory that the immunity induced by the specific serum is due to a direct humoral action upon the micro-organism. Under these conditions such immunity could only be interpreted as being purely passive. This view seemed to be finally established by Bordet’s discovery that a specific serum, inactive by itself, became capable of producing Pfeiffer’s phenomenon, as soon as a small quantity of normal, non-specific serum was added to it. Bordet^[465] thus sums up his theory of the immunity conferred by specific serums: “Passive immunity is due, in part at least, to a chemical action exerted on the vibrios by two pre-formed substances, the one present in the animal before any injection is made, the other found in the serum that is injected; this phenomenon is purely chemical in the sense that it can go on without the aid of a vital reaction, of any new cell secretion: indeed it is found to take place in fluids from which the cells have been entirely removed” (p. 217). These demonstrations led up to the belief that the organism of the animal remained absolutely passive when it was subjected to the action of protective or anti-infective serums, and that the case of the cholera vibrio represented a kind of schema, which was applicable to the whole of the group of phenomena met with in passive immunity.

As in the study of the immunity obtained as the result of vaccinations with micro-organisms or their products, so in “passive immunity” there was seen only the direct chemical action of two substances on the micro-organism, and efforts were made to extend this demonstration to a series of infective diseases.

Pfeiffer and Kolle^[466] having observed that the blood serum of persons convalescent from typhoid fever, as well as that of animals vaccinated with the typhoid bacillus, exhibited a great protective power for the guinea-pig, wished to get some idea of the mechanism of this immunity. They found that in the peritoneal cavity of guinea-pigs, inoculated with the typhoid cocco-bacillus and simultaneously subjected to the action of protective serums, the micro-organisms lose their mobility almost immediately. A little later, they exhibit a degeneration of form, become less refractile and disintegrate. After the injection of large doses of specific serum the bacilli, much as in the case of the cholera vibrio, become transformed into granules. “But,” say these authors, “this last mode of destruction, that is to say the formation of granules at the expense of the injected bacteria, does not occur with such remarkable regularity as it does in Pfeiffer’s phenomenon in the cholera vibrio” (p. 219). Whilst these changes are going on in the peritoneal fluid, the leucocytes begin to come up and to ingest the bacilli and their débris. “Phagocytosis, therefore, undoubtedly plays a part in the destruction of the bacteria. Nevertheless, as most of the injected bacteria die in the fluid of the exudation, phagocytosis can not be regarded as the cause of the protective action of the serum” (p. 220). We see from this description that even in the case of the typhoid cocco-bacillus the direct action of the fluids of the body is perceptibly less marked than in the case of the cholera vibrio. Even in the latter, however, it is necessary to make many reservations. The same laws apply to the immunity against this micro-organism, conferred by the serum of immunised animals, as to the immunity due to vaccinations by the vibrios or their products. To treat this subject fully one would have to repeat almost textually the two preceding chapters, but I will simply recall the fact that this transformation, almost general and very rapid, as we observed in vitro in vibrios placed in contact with fresh specific serum or with the mixture of this serum, heated to 55°–56° C., and normal unheated serum, is only met with in the animal body where phagolysis appears. Pfeiffer first observed the phenomenon which bears his name in the peritoneal cavity, and it is best seen in that situation during the period of the phagolysis of the white corpuscles. Vibrios, mixed with small doses of specific serum which by itself is able to render them motionless and agglutinate them, but which is absolutely unable to transform them into granules, present this transformation immediately they are introduced into the peritoneal cavity of normal guinea-pigs. In this case the vibrios, permeated by the fixative of the specific serum, are affected by the microcytase which has escaped from the injured phagocytes and is found in the peritoneal fluid. The preparation of the peritoneal cavity of normal guinea-pigs by means of an injection of broth or physiological saline solution the day before, prevents the production of Pfeiffer’s phenomenon, in spite of the protective serum, just as in vaccinated guinea-pigs. In both cases the vibrios, without being transformed into granules by the fluid part of the peritoneal exudation, are ingested by the phagocytes en masse and with extraordinary rapidity. This experiment was repeated by Garnier^[467] with the typhoid cocco-bacillus. He first injected into the peritoneal cavity of young guinea-pigs several c.c. of physiological salt solution, of fresh broth or of some other fluid. The next day he introduced into the same situation typhoid cocco-bacilli mixed with blood serum from a donkey that had been for a long time immunised against this organism. A few minutes (2–4) after this latter injection the leucocytes, whose phagolysis had been prevented by the previous day’s preparation, were found crammed with cocco-bacilli. Some of these bacilli, like those still free in the peritoneal fluid, retained their normal form, but a very large number of those ingested by the microphages were already transformed into granules. This experiment affords fresh confirmation of the hypothesis that the substance which transforms the cocco-bacilli or the vibrios into granules is the microcytase. In the phagocytes in their normal condition this microcytase is found in the microphages, but during phagolysis a portion of it escapes into the surrounding fluid. In the control experiments made by Garnier with young normal guinea-pigs not prepared by preliminary injection, the simultaneous injection of typhoid cocco-bacilli and specific donkey’s serum set up this attenuated and not very typical Pfeiffer’s phenomenon described in Pfeiffer and Kolle’s memoir.

Soon after the discovery of Pfeiffer’s phenomenon I^[468] was able to bring forward a proof that it was produced neither in the subcutaneous tissue, in the oedemas set up by the arrest of the circulation, nor in the anterior chamber of the eye of animals when cholera vibrios mixed with anti-infective specific serum were injected into these situations. Here the micro-organisms retain their normal form, remain quite alive and in this condition are ingested by the leucocytes which are brought up to the points invaded. These cells, attracted by the vibrionic products, do not undergo any phagolysis and, untrammelled, fulfil their phagocytic function. Inside them are found vibrios which have kept their elongated form and others which have become transformed into granules. The exudations containing these elements still give cholera cultures on nutrient media, a proof that some at least of the intracellular vibrios are alive. Here we have no destruction of the micro-organisms in the fluids of the body, consequently no direct action of the bactericidal substance. This substance, enclosed in the phagocytes, can only act through the intervention of these elements.

Mesnil^[469] made analogous experiments with the Massowah vibrio, which, unlike the cholera vibrio, is peculiarly virulent when injected subcutaneously into guinea-pigs. In spite of this difference, this micro-organism, when injected along with protective serum into normal animals, behaves much as does the cholera vibrio proper. Mesnil injected the Massowah vibrios at the same time as the anti-infective specific serum, into the subcutaneous tissue of young and adult guinea-pigs and of young rabbits. In every case he observed the same reaction phenomena in the animal organism. The vibrios caused the formation of oedema at the point of inoculation and remained isolated in the fluid. The majority of these micro-organisms became motionless, but a few remained motile. Pfeiffer’s phenomenon was never observed. There was sometimes an aggregation of the vibrios, but this was not comparable with the marked agglutination brought about by the specific serum in vitro. The vibrios retained their power of reproduction, and Mesnil was able to observe them in all phases of division. Some hours (6–8) after inoculation the leucocytes began to come up to the seat of injection and set to work at once to ingest the vibrios. This phagocytosis became more and more marked, and finally there was ingestion of the whole of the micro-organisms. Drops of the exudation containing only intraphagocytic vibrios, when placed in the incubator, gave abundant cultures. The leucocytes died outside the animal body, whilst the vibrios continued to live and grow well under the new conditions. Certain leucocytes became three times their original size, and their contents were seen to be made up of vibrios closely packed together. The subcutaneous exudation, when withdrawn even eight days after the injection of the micro-organisms and sown on nutrient media, still gave colonies of vibrios.

It is evident, therefore, that the direct action of the protective serum on the vibrios was reduced to a mere trifle. It rendered them motionless and brought about a very slight clumping, but it was incapable of transforming the vibrios into granules or of destroying them. We see, then, that even in the case of the vibrios, the part played by Pfeiffer’s phenomenon is very limited. The destruction of the vibrios is effected with certainty, and completely, under the influence of the specific serums, not by a direct action of the two antibacterial substances but through the mediation of the phagocytes. Before the fixative, introduced with the protective serum, can bring about this result, the leucocytes, impressed with a special sensitiveness, must come up to the seat of inoculation, seize the micro-organisms and secrete around them their cytase. It is only as a result of these actions, purely vital, that the chemical or physico-chemical reaction of the substances which intervene in the destruction of the vibrios is brought about.

Under these conditions it can easily be understood that if the vital action of the phagocytes is retarded or depressed the injection of protective serum cannot preserve the life of the animal. Cantacuzène^[470], who had already made a similar demonstration on guinea-pigs vaccinated against the cholera vibrio by these organisms or by their products, carried out numerous experiments on the action of opium on normal guinea-pigs simultaneously inoculated with vibrios and specific serum. Before injecting this mixture Cantacuzène narcotised his animals by means of tincture of opium. The great majority (⅘) of the guinea-pigs so treated died at the end of one or several days. The transformation of the vibrios into granules, under the influence of the serum, took place in the peritoneal cavity, but the leucocytes, on account of the narcotic action of the opium, were tardy in coming up. On their arrival in the peritoneal cavity they were capable of ingesting the granules, but absolutely refused to seize entire vibrios, always fairly numerous in the exudations. In spite of the appearance of a large number of leucocytes, these cells were still too weak to offer any adequate opposition to the vibrios, which increased in number and continued to multiply up to the death of the animal, when the exudation simply swarmed with very motile vibrios. Sometimes the struggle was prolonged. The weakened leucocytes allow the vibrios to develop, but, after a greater or less length of time, they regain their strength and begin to ingest the micro-organisms vigorously. Complete phagolysis follows, but the guinea-pig, attacked by the toxic products of the vibrio, finally succumbs in spite of the absence of free vibrios from its body.

An analysis of the phenomena observed in the body of an animal treated with antivibrionic serum, demonstrates that, in spite of a certain direct action of the substances contained in this fluid, there still remain a whole series of processes, amongst which the carriers of the cytases, that is to say the phagocytes, fill the most important rôle. Nevertheless, the cholera vibrio with its allied forms is still the most sensitive of all the micro-organisms to the bactericidal action of the fluids of the body. It may, therefore, readily be conceived that the more resistant micro-organisms are even less subject to the direct influence of the specific serums. Thus we have seen that the coccobacillus of typhoid fever presents, in the phagolysed peritoneal fluid, merely an attenuated form of Pfeiffer’s phenomenon. The other representatives of the group of bacilli are still less subject to the direct action of the serums, and Gheorghiewsky^[471], in his studies on the Bacillus pyocyaneus, found that normal guinea-pigs, injected subcutaneously with anti-infective specific serum, and inoculated into the peritoneal cavity with this organism, present the same phenomena as those described in Chapter VIII. He never noticed either lysis of the bacteria in the fluids of the animal or their total transformation into agglutinated masses outside the phagocytes. The resistance offered by the animal was always in direct relation to the rapidity of the appearance and extent of the phagocytic reaction.

In order to determine the relative importance of each of the factors which act in the preservation of animals subjected to the influence of the specific serum, Gheorghiewsky repeated Cantacuzène’s experiments on the effect of narcotisation by tincture of opium. This alkaloid retards diapedesis, but does not affect the tactile sensibility or the motility of the leucocytes. The humoral properties, on the other hand, are not in the least affected by the narcosis. In spite of the fact that in guinea-pigs, narcotised and treated with anti-infective serum, the direct action was not interfered with, the animals always died because the retarded and incomplete phagocytic reaction was insufficient to destroy the bacilli rapidly enough.

Mesnil^[472] studied the action of the specific serum against swine erysipelas on normal animals into which he had injected it some time before inoculation of the corresponding bacillus into the peritoneal cavity. This serum exercises a most marked protective action on the mouse, an animal very susceptible to the pathogenic action of this micro-organism. In mice so prepared complete and rapid phagocytosis takes place. These micro-organisms before being ingested by the phagocytes show no appreciable change; they are always stained very uniformly and intensely by Gram’s method, and they never swell up. The bacilli undergo no agglutination in the body of the mouse, a fact of which we can convince ourselves by examining hanging drops of the exudation. The phenomenon which strikes the observer most is the very pronounced phagocytosis, due principally to the activity of the microphages. Some hours after inoculation these cells are found to be crammed with bacilli, a large number of which no longer stain in the normal fashion. Without being transformed into granules, these micro-organisms undergo intracellular digestion which at the end of a few days is complete. This destruction is more rapid and complete in the microphages, slower in the macrophages. Drops of exudation collected from these mice, at a stage when the ingestion is completed, produce fatal septicaemia in untreated mice. This is proof that at the moment when they were seized by the phagocytes the bacilli still retained their virulence. Mesnil, as the result of his experiments, concludes that “the effect of the serum is to stimulate the phagocytes and especially the polynuclear forms; they ingest more quickly, they digest more quickly. The serum is, therefore, a stimulant of the cells charged with the defence of the animal” (p. 496).

We need not describe the phenomena produced in mice inoculated subcutaneously and treated with protective serum, for even in the peritoneal cavity neither Pfeiffer’s phenomenon nor any extracellular destruction of the bacilli can be observed. The micro-organisms, when subjected to the influence of the specific serum, readily absorb the fixative, as demonstrated by Bordet and Gengou^[473]. This absorption must certainly favour the action of the intraphagocytic cytases. It is not, however, sufficient to explain the protective, anti-infective action of the serum. Such explanation was given by the experiments which Gengou, at my request, was good enough to make. He inoculated mice with the bacilli of swine erysipelas, mixed with specific serum heated to 55° C., to which was added some normal guinea-pig’s serum. The mice so treated resisted the infection but controls died in a few days. Being thus assured of the protective action of the serum, Gengou prepared the same mixtures of swine erysipelas bacilli and of the two serums; but, instead of injecting the whole of the mixture, he removed the bacilli from the serums, after a prolonged contact, and injected the bacilli alone into the mice. The bacilli had become permeated with fixatives, but, in spite of this, they killed the mice almost as quickly as the controls. Consequently, it is not the fixative adherent to the micro-organisms which determines the protective action of the specific serum. This fluid must contain another substance, one that will stimulate the phagocytes.

The analysis of the mechanism of the immunity termed passive, that is to say, communicated to normal animals by the introduction of an anti-infective specific serum, teaches us that, even when the direct action of the humoral substances is very limited, the protective effect, thanks to the stimulant action which brings about the destruction of the micro-organisms through the mediation of the phagocytic reaction, is still produced. The result at which we have thus arrived is confirmed by the examination of the phenomena observed in animals subjected to the action of anti-anthrax serum. Marchoux^[474] first supplied us with precise details as to the mode of action on the rabbit of the serum of animals treated with anthrax bacilli. He found that, in the peritoneal cavity of rabbits injected the day before with anti-anthrax serum, the inoculated anthrax bacilli almost immediately become the prey of phagocytes. Within a couple of minutes after the introduction of bacilli into the peritoneal cavity, the great majority of them are ingested by the leucocytes; ten minutes later, there are no free bacilli. Not only the ingestion but also the destruction of these micro-organisms takes place with great rapidity, and even a few hours after the injection, the peritoneal exudation, when sown on nutrient media, remains sterile. In the subcutaneous tissue the phagocytic reaction requires a longer time than in the peritoneal cavity, nevertheless, it goes on very rapidly. Thus, when inoculated into the subcutaneous tissue of the ear of rabbits treated with specific serum, the bacilli are in great part ingested at the end of half-an-hour. At the end of an hour phagocytosis is usually complete.

In Marchoux’s experiments the importance of the part played by the phagocytes becomes still more prominent when it impedes their function in any way. Rabbits injected with anti-anthrax blood and 24 hours later inoculated below the skin of the ear with anthrax bacilli always resist infection, exhibiting the well-marked phagocytosis just mentioned. In other rabbits, however, prepared in the same way with the serum, but inoculated the following day into an ecchymosis excited by tapping the ear lightly, a certain number of the bacilli escape the phagocytes and succeed in setting up an abundant oedema followed by a fatal anthrax at the end of a few days. On making a post-mortem examination of these animals the bacilli were not numerous, but they were found in all the organs. The same result was obtained in another experiment in which Marchoux inoculated subcutaneously with anthrax blood which coagulated in situ rabbits prepared with specific serum. The blood clot attracted only the macrophages, as pointed out in Chapter IV. The microphages did not come up until late and then in small numbers. Now, as these are the phagocytes that are chiefly instrumental in destroying the anthrax bacillus, their absence allowed the bacilli to multiply and to set up a fatal anthrax. The rabbits prepared with the same serum but injected with anthrax blood diluted with broth (which prevents the formation of clot) completely resisted infection, thanks to the phagocytic reaction which went on without hindrance.

Sclavo^[475] also, who made numerous investigations on the action of the anti-anthrax serum, is of opinion that this action is not a direct one upon the bacillus but is produced indirectly through the action of the animal organism. He maintains that the serum stimulates the function of the phagocytes and augments the bactericidal action of the body fluids. But since this bactericidal power enters the cytase as a substance destroying the micro-organisms, and this cytase is contained in the phagocytes, we can readily understand what a dominant part in the process these elements play.

Sobernheim^[476], also, has paid much attention to the question now under discussion. As the result of his researches he comes to the conclusion that the anti-anthrax serum “cannot exert any effect on the virus by a direct action of the protective specific substances.” In order that the serum may be effective, the active intervention of the organism of the animal is necessary, otherwise, it is impossible to explain why the serum, used in the same proportion against the same quantity of anthrax bacilli, should protect one species of animals (the rabbit) and allow another (guinea-pig, mouse) to succumb. When Sobernheim tried to apply to anthrax the discovery of the transformation of cholera vibrios into granules, he got only negative results. There was nothing produced comparable to Pfeiffer’s phenomenon and the anthrax bacilli usually underwent no apparent modification. Sobernheim affirms also that the rapid phagocytosis under the influence of the serum, described by Marchoux, “does not appear to be produced under all circumstances” (p. 117). As, however, his researches on this subject were made on guinea-pigs which, in spite of the treatment with specific serum, always ended by succumbing to anthrax, we readily understand that his results cannot be compared with those obtained by Marchoux. I was present at the experiments of this observer and convinced myself of the accuracy of the facts recorded in his memoir.

Most of the examples here studied justify fully the hypothesis of the stimulant action of protective serums, a view that I formulated as the result of my researches on the immunity of rabbits against the Gentilly cocco-bacillus^[477]. In this the first case of anti-infective immunity, due to the serum elaborated by an immunised animal, I could not find either a bactericidal action, however slight, or any agglutinative or attenuating property of the fluids of the body. As, on the other hand, this serum had no antitoxic power, everything indicated that we must look for its action, which was nil or very slight on the micro-organism, as being exerted on the organism of the animal into which it was injected for protective purposes. A comparative examination of the course of the phenomena in the subcutaneous tissue of the ear in rabbits, some of which received an injection of the specific serum into the veins whilst others were kept as controls, at once showed how widely different were the two cases. In the control animals, the cocco-bacilli immediately began to multiply without meeting with any opposition on the part of the organism of the animal; on the other hand, in the rabbits treated with serum, the serum became rich in leucocytes which at once set to work to ingest the micro-organisms. In course of time the latter gradually diminished in numbers, whilst the leucocytes went on increasing. The phagocytosis, also, became more and more marked. This struggle was continued for more than 24 hours, after which the purulent exudation, containing masses of leucocytes, no longer included any cocco-bacilli visible under the microscope either outside or inside cells. Nevertheless, this pus was still capable of producing a fatal septicaemia in untreated rabbits, clearly proving that it still contained some living and virulent micro-organisms. These cocco-bacilli persisted for a long time inside the phagocytes; their presence being demonstrated by injecting the exudation into unprotected rabbits and thus setting up a fatal infection. Finally, however, they disappear completely. On consideration of such facts as these I considered that I was justified in formulating the following conclusion at the end of my memoir: “From the facts I have described, taken collectively, we may draw the conclusion that the preservation of unvaccinated rabbits treated with serum is due to a superactivity of the phagocytic defence; and it is allowable to express the opinion that the protective serum of hog cholera acts in rabbits by stimulating the phagocytes, rendering them less sensitive to the toxins, and by stimulating them in their struggle against the bacteria” (p. 310). The facts since collected by various observers fully justify this hypothesis. Amongst the other micro-organisms against which a rapid immunisation has been obtained by means of serum, we must cite the cocco-bacillus of bubonic plague. Numerous experiments, carried out on several species of animals, have shown that antiplague serum markedly augments the phagocytic reaction.

In the group of the cocci, the streptococci have been specially fully studied from the point of view now under discussion. As already stated in another chapter, success has been attained not only in thoroughly immunising several species of animals against this dreaded micro-organism but active serums have been obtained capable of conferring distinct and certain immunity. The protective action of Marmorek’s serum, prepared at the Pasteur Institute, has been specially carefully studied. This serum is obtained from horses that have received numerous injections of various races of streptococci pathogenic for animals and for man^[478]. At Louvain, Denys and his pupils prepared several other antistreptococcic serums and studied their protective effect on laboratory animals.

In collaboration with Leclef, Denys^[479] began by vaccinating rabbits against streptococci and studied the mechanism of the immunity obtained in these animals. A summary of their researches will be found in the eighth chapter. Denys and Leclef considered that the serum of vaccinated rabbits intervenes in two ways, first by directly hindering the multiplication of the streptococcus and then by exalting the activity of the leucocytes. They applied these results to the case in which immunity is conferred upon normal rabbits by the intervention of the serum of the vaccinated rabbit, but they were unable to furnish any data bearing directly on this immunity. Somewhat later, Denys^[480], in collaboration with Marchand, published another memoir in which he describes experiments on the mechanism of the immunity conferred on rabbits by injections of the blood-serum of vaccinated horses. From these experiments they draw the conclusion that “the serum of the horse immunised against the streptococcus possesses no bactericidal properties, properly so called, against this micro-organism; it does not affect it directly; but it contains a substance which renders the phagocytic power of the leucocytes extremely active. Even in the presence of small quantities of this serum, the white corpuscles rapidly ingest the streptococci and are capable of stopping all development so long as they retain their amoeboid movements.” “The action of the serum upon the leucocyte in its conflict with the streptococcus, is really derived from the horse immunised against this organism. It exists neither in the ordinary horse nor in the horse vaccinated against diphtheria” (p. 15). Against these experiments of Denys and Marchand we might bring the same objection that we raised against the analogous experiments of Denys and Leclef, because, in both cases, these writers lay too much stress on the presence or absence of the phenomena of phagocytosis in preparations kept outside the body of the animal. Under these conditions phagocytosis is effected in a fashion too artificial to be capable of furnishing exact information.

Von Lingelsheim^[481] met Denys and Marchand with the fact that, in their researches, the serum of the horse immunised against the streptococcus was only feebly bactericidal. After a prolonged contact (6–12 hours) with a specific serum, the streptococci, when transferred to rabbit’s blood, showed retarded development as compared with streptococci subjected to the influence of the antidiphtheritic and antitetanic horse serum. Von Lingelsheim himself, however, points out that the bactericidal action of the antistreptococcic serum was feeble and transient, and required the intervention of the reaction of the animal cells within the body.

The researches carried out by Bordet^[482] in my laboratory are not open to the objections that we were justified in putting forward against the experiments made by Denys and Marchand, since he carefully watched the phenomena of immunity as they developed in the body of the animal subjected to the action of antistreptococcic horse serum. Bordet began by studying the properties of this serum and accepted Denys’ and Marchand’s statement that bactericidal power, however small, was absent. The streptococcus grows as well in this serum as it does in that of the untreated horse. In the specific serum, however, markedly longer chains are produced than in normal serum. This difference is found only in the earliest period of the growth. The agglutinative power of the antistreptococcic serum is but feebly marked. The injection of a large quantity of this serum into a normal rabbit confers no bactericidal power upon the serum of this animal. “The serum extracted 24 hours after injection is quite as suitable for use as a culture medium as that furnished by the blood before the introduction of the serum. In both the micro-organism grows rapidly and vigorously” (p. 195). Consequently, in the antistreptococcic serum there is nothing comparable to what we obtain so readily with antivibrionic serum: nothing which recalls Pfeiffer’s phenomenon, even of an attenuated nature. We have already noted the result obtained by Bordet, according to which the streptococci, developed in the specific horse serum, were found to be endowed with their normal high virulence. The antistreptococcic serum, injected into the peritoneal cavity of the rabbit the day previous to the microbial inoculation, protects the animal absolutely, provided that the micro-organisms be not too numerous or the quantity of serum not too small. Under these conditions the virus is ingested pretty rapidly and, so far as we know at present, completely. The micro-organism is thus prevented from developing and the animal remains in good health, whilst the control animal, which has received no serum, dies in a very short time.

When the number of streptococci is increased the effort of the animal organism to get rid of them becomes, in spite of the protective serum, more severe and much more prolonged. Some of the micro-organisms certainly become the prey of phagocytes, but a sufficient number remain free in the peritoneal cavity to multiply rapidly. When the number of streptococci has become sufficiently great a phenomenon, to which Bordet gives the name of “phagocytic crisis,” is suddenly observed. In the peritoneal exudation, which has become thick and has taken on the aspect of a homogeneous and white pus, a most rapid phagocytosis is evidently set up. In a short time the whole of the streptococci, which were swarming outside the cells, are ingested by the leucocytes. “The essential condition for recovery is always this complete ingestion” (p. 203). If the ingestion is not general, the rabbit may die, although much later than the control animal.

The phases of the struggle between the animal organism, when subjected to the influence of the protective serum, and the streptococcus, recall Salimbeni’s experiments on immunised horses. The rabbit, in which phagocytosis could not take place at once owing to the presence of too large a number of micro-organisms, exhibits first a stage of free development of the streptococci, after which the phagocytes begin to fulfil their antibacterial function. Here it is especially the macrophages which act; the microphages, although present in fairly large numbers, are entirely inactive. This first stage of phagocytic reaction is insufficient. It is followed by a period when the streptococcus appears to gain the upper hand. Many small chains, having escaped the phagocytes, multiply and give birth to quite a new generation of micro-organisms. If a fresh impulse to phagocytosis does not take place the animal dies from infection. When, however, the protective serum has been of sufficient strength, a new army of leucocytes arrives on the scene and these become masters of the situation. Phagocytosis becomes complete and microphages as well as macrophages devour a large number of streptococci.

Bordet, who, through his previous investigations, was well acquainted with the direct action of the protective serum on vibrios, could find nothing resembling it taking any part in the struggle of the organism of the animal treated with antistreptococcic serum against the streptococcus. The most that he could find was that the streptococci which again begin to swarm in the exudation are smaller in size than the normal streptococcus. It must be accepted, as indicated by the most recent researches, that this micro-organism becomes permeated by the fixative substance of the specific serum. We know already, however, that this fixation cannot deprive the micro-organisms of their virulence. In any case, then, a large share in the process must be attributed to the action of the phagocytes, stimulated by the protective serum, in the struggle of the animal against the streptococcus.

Having considered this series of examples of immunity against bacteria conferred by specific serums, we are in a position to form some idea of the mechanism of this immunity. Before we come to any general conclusion, it may be useful to glance at an example of this so-called passive immunity against a micro-organism belonging to the animal kingdom. Such examples are not numerous, as, in the majority of cases of acquired immunity against Protozoan parasites, the serum is inactive and incapable of communicating immunity to normal individuals. We have only a single example, the Trypanosoma of rats, against which Dr Lydia Rabinowitch and Dr Kempner^[483] have demonstrated the possibility of immunisation by the blood serum of vaccinated white rats. The mechanism of this immunity has been studied by Laveran and Mesnil^[484], who found that it was like that described (Chap. VIII) in connection with the immunity in white rats, conferred by the inoculation of living Trypanosomata. The specific serum does not affect these infusoria except that it brings about slight agglutination. Trypanosomata placed in contact with it retain their pristine vitality and motility. This fact led Mme Rabinowitch and Dr Kempner to advance the hypothesis that the protective action of the serum must depend upon its antitoxic power. Since, however, in the infection of rats by the Trypanosomata, the toxic action is very feeble if not nil, it is very difficult to accept this view. It certainly appears to be much more probable that the serum acts in this case, as in many others, by stimulating the phagocytic reaction. The rapidity with which the living Trypanosomata are ingested by the phagocytes has been shown by Laveran and Mesnil.

Reviewing the whole of the data on immunity produced under the influence of anti-infective or protective serums, it is evident that they fall under two main categories. On the one hand there is a direct action of these serums on the micro-organisms, an action that is either microbicidal properly so called, agglutinative, or fixative. On the other hand, a stimulation of the phagocytic defence which leads to the final destruction of the micro-organisms is set up. This last factor is general; even in the case where the direct action is most marked (vibrios in the phagolysed peritoneal cavity), its importance is considerable. The micro-organisms which can be deeply injured by the direct action of the specific serum are few in number. In most cases this action is a feeble one and needs, for its completion, effective co-operation on the part of the phagocytes. In this respect micro-organisms present a whole gamut which begins with the cholera vibrio, the micro-organism most sensitive to the action of the body fluids, and ends with the Trypanosoma of the rat, a flagellated Infusorian which cannot have even its motility affected by the direct action of the fluid elements. In all these cases, of course, the immunity conferred by the serums is due to the final destruction of the micro-organisms which invariably resolves itself into the same fundamental act—digestion by the cytases, a phenomenon which can only be produced at all quickly by the action of cytases contained in the protective serums or that have escaped from the phagocytes during phagolysis. The digestion by the cytases may also, and this is usually the case, be effected only after the manifestation of a regular series of vital phenomena on the part of the defensive elements of the body. As this factor fills such an important rôle, it is readily understood that we can not accept the term passive immunity by which to designate the immunity conferred by the specific serums. The action of the cytases, which is necessary to bring about the final result in this immunity, depends too much on the activity of the cells which contain the bactericidal ferment. For this reason, when the functional activity of the phagocytes is in abeyance or is retarded, the animal succumbs, in spite of the presence in its organism of a more than sufficient quantity of cytases. In this connection Wassermann’s^[485] suggestion of adding normal serums rich in cytases to the specific serums must be regarded as very apposite. When protective serums poor in cytases or which have lost them as the result of heating, of the use of antiseptics, or simply from the influence of time, are injected, no immunising effect is ever obtained, simply because of the inactivity of the phagocytes, the cells in which the cytases are found. If at the same time normal serum rich in cytases ready prepared be injected, better results should be obtained. We may recall here an analogous example—the anthrax of the rat. Although possessing a large quantity of cytase, very effective against the bacillus, the organism of the rat can make no use of it, because the phagocytes which contain it do not manifest a sufficient activity. But the injection into a rat of blood serum from the same species containing a certain amount of cytase that has escaped during the formation of the clot, is sufficient to preserve the animal against a fatal infection.