CHAPTER XVI

HISTORICAL SKETCH OF OUR KNOWLEDGE ON IMMUNITY

Methods used by savage races for vaccination against snake venom and against bovine pleuropneumonia.—Variolisation and vaccination against small-pox.—Discovery of the attenuation of viruses and of vaccinations with attenuated micro-organisms.—Theory of the exhaustion of the medium as a cause of acquired immunity.—Theory of substances which prevent the multiplication of micro-organisms in the refractory body.—Local theory of immunity.—Theory of the adaptation of the cells of the immunised organism.

Observations on the presence of micro-organisms in the white corpuscles.—History of phagocytosis and of the theory of phagocytes.—Numerous attacks upon this theory.—Theory of the bactericidal property of the body fluids.—Theory of the antitoxic power of the body fluids.—Extracellular destruction of micro-organisms.—Analogy between bacteriolysis and haemolysis.—Theory of side-chains.

Progress of the theory of phagocytes.—Attempts to reconcile it with the humoral theory.—Present phase of the question of immunity.

[Sidenote: [528]]

As protection against disease is one of the most important amongst those questions which are engrossing the attention of humanity, it is natural that very great attention should have been devoted to it from the most remote times. We see primitive races, the ordinary layman, medical men, legislators and even the most subtle thinkers devoting their energies to the solution of the problem of immunity against poisoning and against infections. Historical science will never reveal to us the earliest sources of our knowledge on this question, so remote are their origins. The wide distribution of several methods for protecting man and cattle against certain diseases clearly proves that the origin of this practice dates from a very early period.

[Sidenote: [529]]

The frequency of venomous snakes in many countries has inspired a dread of these reptiles, and this must have led to the search for some method of fighting against the poisoning after the patient had been bitten. Thus, we find that many primitive races make use of various methods of immunising the body against the action of venom. The Portuguese colonel, Serpa Pinto[839], in a letter addressed to d’Abbadie, describes the method by which he was vaccinated by the Vatuas, natives of the east coast of Africa. These savages extract the poison of snakes and prepare from it, by the addition of vegetable substances, a very brown glutinous paste which they introduce into incisions made in the skin. This operation is very painful and is followed by a swelling which lasts for a whole week. The Vatuas assert that this method confers a sure immunity against the venom. Serpa Pinto was never bitten by a snake, but, a short time after he had been vaccinated, he was stung, in the Seychelles Islands, by a scorpion without experiencing any ill effects. This experience confirms the assertion of the Vatuas, because it has been shown that the vaccine against snake venom is also efficacious against the bite of scorpions. The fact that after being stung by another scorpion ten years later Serpa Pinto was so ill that for eight days he believed that he was going to die or at least to lose an arm, shows that he did not enjoy natural immunity, and the innocuousness of the previous bite must therefore be attributed to a vaccination the effect of which had disappeared at the end of ten years.

Another vaccinal method used by primitive races is that against the pleuropneumonia of the Bovidae. De Rochebrune[840] points out that the Moors and the Pouls of Senegambia have “a custom whose origin is lost in the obscurity of antiquity” which consists in the inoculation into their herds of cattle of the virus of the epizootic pleuropneumonia. “The point of a knife of primitive form, or of a dagger, is plunged into the lung of an animal that has died from the disease and an incision, sufficient to allow the virus to penetrate below the skin of the healthy animal, is made into the supranasal region. Experience has demonstrated the success of this protective operation.”

In Europe, the vaccinations of cattle with the virus of pleuropneumonia have certainly been known for more than a century, for, in a pamphlet published at Berne in 1773[841], mention is made of the “inoculation” of Bovidae as a means of preventing the disease in England and in Holland, a disease against which it has been recognised that remedies are powerless.

[Sidenote: [530]]

The inoculation of the variolous virus into the healthy human subject, which comes into the same category as the inoculation of the pleuropneumonic virus into healthy bovine animals, is also a widely extended and very ancient method. The Chinese[842] assert that they have known from the commencement of the 11th century the method of immunising against small-pox. Amongst them, as amongst the Siamese, the matter from the variolous scab is introduced into the nostrils. In Persia variolisation is practised by surgeons and by the staffs of bathing establishments, who introduce the powdered scabs into scratches in the skin. The Ashantis inoculate the variolous virus into seven places on the arms and legs. According to the account of Timoni, a Greek physician practising in Constantinople in the first half of the 18th century, the Circassians and Georgians, intent upon preserving the beauty of their daughters, make punctures at various points in the skin, with needles charged with variolous virus. Everybody is acquainted with the fact that it was from Constantinople that Lady Mary Wortley Montague at the same period (1721), imported into Europe “the Greek method,” which consisted in the inoculation of the contents of small-pox pustules with the object of producing a benign small-pox and of protecting the vaccinated person from severe and dangerous small-pox. This practice was widespread in Europe during the second half of the 18th century, but as it was not unattended by serious drawbacks an attempt was made to avoid them by the employment of all kinds of medicaments. As these, however, were found to be entirely ineffective, the need was felt of replacing variolisation by some more benign method.

[Sidenote: [531]]

It is asserted[843] that in Baluchistan the custom of having cows suffering from cow-pox milked by children who had wounds on their hands has been widespread from time immemorial. This practice conferred upon these children an immunity against small-pox. It cannot be denied that the idea of being able to vaccinate with cow-pox was common knowledge amongst breeders and dairymen in several countries in Europe, especially in England, France, and Germany. It is stated that Edward Jenner learnt from the country people of his native county of Gloucestershire that contact with cow-pox protected against small-pox. Being a man of great understanding and culture, he set himself to verify this opinion experimentally. Having demonstrated by a great number of experiments that the inoculation of variolous virus into persons vaccinated by cow-pox had no ill result, he became the great propagandist of the new method. He worked at this subject for 20 years but only decided to publish his results (in 1798) after he had completely satisfied himself of the great utility of vaccination with the virus of cow-pox. At first Jenner’s discovery met with great opposition, but his method was soon verified in France and several other countries and it was not long before it was generally practised.

[Sidenote: [532]]

When Pasteur set himself to study the infective diseases in their relation to micro-organisms the idea of profiting by the discovery of these pathogenic organisms and of drawing from them a weapon against infections soon arose in his mind. He studied Jenner’s work in order to extract from it any indications capable of putting him into the right path. He induced his collaborators to carry out several series of experiments with the object of immunising the animal organism against infective micro-organisms. During this laborious and original work chance[844] helped in the accomplishment of his task. When, at the conclusion of the holidays in the autumn of 1879, Pasteur and his collaborators Chamberland and Roux wished to resume their experiments on fowl cholera, they found to their great surprise that the micro-organisms of this disease, usually so fatal, had become innocuous. Fowls, that received doses of cultures much more than sufficient to cause death, did not experience any ill effect. Prepared by his previous knowledge and by the continual direction of his thoughts to the prevention of contagious diseases, Pasteur divined at once the great bearing of this check in his inoculations with old cultures, and immediately began to make precise experiments as to the vaccinating power of these micro-organisms which had become innocuous. These researches led him to the discovery of two great principles: that of the attenuation of viruses, and that of the vaccinating property of attenuated micro-organisms. Various memoirs by Pasteur[845] established these laws in a very exact manner; moreover he gave all the information necessary to allow of the principal results being controlled and verified. In France, this great discovery was at once accepted by various investigators, though others found occasion to manifest their scepticism. Abroad this discovery met with very lively opposition and this from the highest authorities, who would not recognise the possibility either of attenuating the virus or of conferring immunity upon animals. The anthrax bacillus can be grown for a very long time on culture media, the potato, for example, without losing its pathogenic power in the slightest degree. Therefore, it was said, this attenuation of virus can have no actual existence. White rats that have resisted one or more inoculations of the anthrax bacillus may die from a later inoculation of the same micro-organism. Therefore there is no acquired immunity, etc. The principles laid down by Pasteur are from every point of view of such prime importance, that very numerous experiments were carried out at once for the purpose of verifying their exactness and the contest was not a long one. In the course of a few years it was universally recognised that the attenuation of viruses, and also the vaccination by attenuated micro-organisms, were realities which henceforth cannot be denied and which must pass into the domain of truths definitely acquired. An attempt was then made to extend these fresh victories to the other infective diseases. Pasteur, Chamberland, and Roux applied themselves to devising a method of vaccinating animals against anthrax and against rabic virus; Pasteur and Thuillier extended their researches on this subject to swine erysipelas. From several other quarters the search for vaccines was instituted. Toussaint made various attempts, at times crowned with success, to immunise animals against anthrax by means of heated anthrax blood. Arloing, Cornevin, and Thomas succeeded in vaccinating the Bovidae against symptomatic anthrax. Loeffler was the first in Germany to demonstrate that rabbits which had recovered from the disease set up by the bacillus of mouse septicaemia acquired an immunity against the attacks of this organism. It is not necessary to cite further examples, so numerous have they become and so unanimously confirmatory.

[Sidenote: [533]]

After the first steps had been taken along this new path Pasteur and his collaborators began to apply the knowledge they had gained to the preparation of vaccines capable of giving practical results. The two anti-anthrax vaccines and the two vaccines against swine erysipelas were the fruit of these attempts. Here, again, numerous objections were raised against these discoveries. Sheep which had received enormous quantities of the bacillus may die from anthrax in spite of the two Pasteurian vaccines and from that it was wished to conclude that these vaccines should not be employed in practice to protect sheep against the anthrax fever. The results of experiments made on a large scale in various parts of the globe have demonstrated the inadequacy of these objections and these questions are now regarded as definitely settled.

[Sidenote: [534]]

So large a number of investigations, in response to the most urgent and immediate needs, was not favourable to minute researches on the mechanism of this immunity which had been revealed in so marvellous a fashion. In spite of this, Pasteur applied himself to the solution of this problem so far as this was possible under the conditions in which he carried on his investigations. He thought that acquired immunity was the result of the impossibility of the growth of a pathogenic micro-organism in a medium in which it had previously been cultivated. When the micro-organism of fowl cholera sets up in certain individuals a disease which though grave is not fatal, or when the attenuated micro-organism produces a simple, transient discomfort, it lives in both cases in the fluids and tissues of the animal. This existence is possible in consequence of the absorption of certain nutrient substances. Once these substances are consumed they are not easily renewed, and in consequence the vaccinated organism becomes incapable of nourishing the special micro-organism a second or a third time. To support this brilliant hypothesis by precise facts Pasteur made experiments on the conditions met with in the development of the micro-organism of fowl cholera _in vitro_. He filtered a broth culture of this micro-organism after it had grown luxuriantly for several days, and into the fluid, which had now become clear and transparent, he sowed afresh the same micro-organism. No growth took place and the fluid remained quite clear. This absence of development might be explained either by the presence in the fluid of some excremental substance thrown off during the first culture or by the absence of some substance indispensable for the nutrition of the micro-organism. Pasteur excluded the first hypothesis by an experiment which demonstrated that it is sufficient to add to the filtered fluid a small quantity of fresh nutritive substances to enable the micro-organism again to develop abundantly. It is therefore to the absence of some element essential to the existence of the micro-organism that we must attribute the immunity enjoyed by animals which have been vaccinated or which have undergone spontaneous cure. This is how Pasteur[846] expressed himself on this point: “the muscle which has been much affected has, even after healing and repair, become in some way incapable of supporting the growth of the micro-organism, as if the latter, by a previous culture, had eliminated from the muscle some principle that life does not bring back and whose absence prevents the development of the small organism. There is no doubt that this explanation, to which the plainest facts at the moment lead us, will become general and applicable to all the virulent diseases.”

This explanation appeared to be a reasonable one to several observers, amongst whom I may cite Chauveau[847], the distinguished author of important works on viruses. “In all probability this seductive theory,” says Chauveau, “based on one of the most interesting of those clear and decisive experiments for which Pasteur is famous, applies to the majority of cases of immunity acquired by protective inoculation.” But Chauveau thinks that it does not explain natural immunity, especially that of the Algerian sheep, against anthrax, an example that he had studied on several occasions. When he inoculated into these animals large quantities of anthrax bacilli, not going beyond certain limits, the sheep resisted perfectly; but injections of enormous doses were nearly always capable of overcoming this natural immunity of the Algerian sheep and of inducing in them a fatal anthrax. Chauveau thinks that this fact is best explained by the presence of an inhibitory substance in the blood plasma, whose action becomes exhausted when distributed over a very large number of bacilli. This opinion was not, however, shared by Pasteur[848], who raises the objection that natural immunity can really be produced and maintained without the presence of this inhibitory substance from the fact that fowls, which exhibit such marked resistance against anthrax, readily contract the disease when the temperature of their bodies is lowered. Under these conditions it is unimaginable that an inhibitory substance has disappeared under the influence of cold.

[Sidenote: [535]]

The controversy existent from the birth of theories on immunity shows us that from the very commencement the problem was found to be a very complex one, and that to attack it in a satisfactory way we must as far as possible multiply and deepen our study of the phenomena which accompany the resistance of the animal against pathogenic micro-organisms. Thus, Chauveau[849] was not long before he undertook experiments having for their object the determination of the fate of anthrax bacilli when injected into the blood vessels of Algerian sheep. He found that these organisms disappeared from the blood at the end of a few hours, but they were then to be found accumulated in the lung, spleen, and certain other viscera. In these positions the bacilli become incapable of reproducing themselves and in refractory individuals soon disappear, being opposed by the inhibitory substances of the blood plasma.

The two theories just sketched have this point in common, that they both attribute the natural or acquired immunity to humoral and purely passive properties. According to one theory it is the impoverishment of the fluids of the animals which prevents the development of the pathogenic organism, whilst according to the other it is the presence of some bacterial poison which brings about the same result. To give experimental support to his theory Pasteur brought forward his attempts at sowing micro-organisms in culture media exhausted by a previous development of the same organism, eliminating, so to say, the active influence of the animal organism. It is true that, in order to explain natural immunity, it was necessary to ascribe a rôle to the “constitution” and to the “vital resistance,” interpreting this, as Naegeli had already done, in the sense of a competition for the oxygen and the nutritive substances between the parasites and the cells of the body.

[Sidenote: [536]]

Adopting this point of view, Hans Buchner[850], a pupil of Naegeli, attempted to gain a more precise idea of the conditions under which acquired immunity against infective diseases is set up. He developed his theory in various publications; this theory consists, briefly, in the property of the animal organism to reinforce the local resistance of the organs by means of an inflammatory reaction. The starting-point of this local theory is the thesis that each pathogenic micro-organism can only manifest its pathogenic action when it enters the particular organ in which it is capable of living and maintaining itself. Thus, the pneumonococcus can live in the lungs only, the cholera vibrio in the intestines only, and so on. Every time that a pathogenic micro-organism becomes localised in its special organ, an inflammatory action is set up which results in the reinforcement of the living elements of the organ in question. Inflammation, therefore, is regarded by Buchner as a salutary reaction, which acts, not directly on the exciting morbific cause, but through the mediation of the specific cells of the organs. This theory of immunity led Buchner to propose arsenical treatment as a remedy against microbial disease, because arsenic is, of all drugs, the one capable of setting up the greatest inflammatory reaction.

Another German observer, Grawitz[851], proposes a theory of acquired immunity, according to which a first attack of an infective disease sets up “the adaptation of the cells to the power of energetic assimilation of the fungi.” This reinforced adaptation is transmitted to the descendants of the cells which have acquired it, and for that reason the immunity may persist for months, and even years. Grawitz attempted to base his views on experiments on the acquired immunity against the fungus of the lily of the valley, but Loeffler[852] soon demonstrated that this thesis could not be maintained, and that the immunity assumed by Grawitz did not, in reality, exist.

It will be seen that all the theories summarised above are marked by their vague character and want of precision; this is not at all astonishing when we take into consideration the very imperfect knowledge of the phenomena of immunity. It is evident that if we wish to gain a satisfactory idea of the mechanism of the resistance of the animal body against pathogenic micro-organisms, we must inform ourselves as to the modifications which take place in the organs and tissues at the time of the acquisition of the immunity, and also find out what becomes of the micro-organisms in a refractory animal.

We have seen that Chauveau demonstrated that anthrax bacilli when injected into the vessels of Algerian sheep disappear, but he was unable to say anything as to the way in which this disappearance was brought about in nature. Buchner accepted the reinforced resistance of inflamed organs without being able to describe the phenomena which manifest themselves during the inflammation of tissues invaded by the pathogenic micro-organisms.

[Sidenote: [537]]

Independently of these theoretical and rather speculative views on immunity, there has been an addition to our scientific assets of fairly exact data on the relation of certain pathogenic organisms to the organs and tissues of susceptible or refractory animals. When, as a result of the labours of Davaine and Obermeyer, the attention of pathologists, especially of those working at pathological histology, was drawn to the part played by micro-organisms in infective diseases, a diligent search was instituted for these organisms in sections of the organs of persons who had died from various diseases. Masses of cocci especially were found in the organs of individuals who had died from diphtheria, puerperal fever, and various forms of pyaemia. In the course of these investigations attention was drawn fairly frequently to the presence of micro-organisms inside the white corpuscles of pus and of other morbid products. Amongst the first to make this observation I may cite Hayem[853] in France, and Birch-Hirschfeld[854], Klebs, Rindfleisch, von Recklinghausen, and Waldeyer in Germany. Klebs[855] speaks of the presence of micro-organisms in infected wounds, in the interior of contractile white corpuscles, and attributes to these cells the principal rôle in the transport of these parasites in the lymphatic tissue. Waldeyer[856] cites a case of puerperal fever in which the corpuscles of the peritoneal pus were filled with bacteria. Similar observations were by no means rare; and they led to a general conclusion that micro-organisms meet with such favourable conditions inside the leucocytes that they would contribute to their dissemination through the body. This opinion had become so general that when Koch[857], in frogs inoculated with anthrax bacilli, made the discovery of round cells containing large numbers of these micro-organisms he did not hesitate to conclude that the bacilli found a favourable medium in the substance of these elements. Now the frog, under ordinary conditions, is refractory to anthrax.

[Sidenote: [538]]

As early as 1874, however, Panum[858] had given expression to the view, in a vague fashion it is true, that leucocytes might assist in the destruction of micro-organisms. In his memoir on putrefactive poisons we find a note wherein occurs the following reflection: “For the solution of the question as to how and in what situations the ordinary bacteria of putrefaction disappear, an interesting communication made by Birch-Hirschfeld seems to me to furnish an indication. According to this observer the micrococci, introduced into the circulation, are deposited in the lymphatic glands and in the spleen, after having, for the most part, entered into the blood corpuscles. That the ordinary bacilli of putrefaction really die in the body is proved, not only by the circumstance that they remain inactive after the acute paroxysm of putrid intoxication has been happily surmounted, but also by the important observations made by Eberth on the innocuousness of the inoculation of ordinary bacteria into the cornea.” These lines contain the indication that the corpuscles of the blood (in this case undoubtedly leucocytes) ingest the bacteria introduced in the blood current and destroy them.

Some years later, in 1877, Grawitz[859], in connection with his researches on the parasite of the lily of the valley, made the remark that the fungi, when introduced into the blood of mammals, are seized by the white corpuscles and thus “withdrawn from contact with the assimilable fluid.” Gaule[860] who, as we know, sought to demonstrate that the _Drepanidium_ of the frog’s blood is nothing but the fragments of cell nuclei transformed into ‘Würmchen,’ has described the structure of these organisms in the amoeboid cells of the spleen. “I happened on one occasion,” he writes, “to observe an amoebocyte of the spleen of the frog which in a short time ingested three ‘Würmchen,’ and then went away briskly without leaving any trace of where it had been. Following its movements I was able at the first to make out within the contents of the amoebocyte the refractile body of the ‘Würmchen.’ But this body became paler, and half-an-hour later it had been completely assimilated.” Undoubtedly these “Würmchen” were nothing but parasites (_Drepanidium_), and have no connection with the cell nuclei of frogs. Their ingestion, followed by destruction, was, therefore, a defensive act on the part of the body manifested by the amoeboid cells of the splenic pulp.

[Sidenote: [539]]

[Sidenote: [540]]

In the same year, 1881, in which this observation by Gaule was published, Roser[861], assistant in surgery at Marburg, published a small pamphlet on the lower animals. In this pamphlet the possibility of growing certain unicellular organisms in urine and milk and the adaptation of these organisms to saline solutions received special mention. At the end of one of his paragraphs Roser expresses his views on immunity, although this subject was not discussed at all in his pamphlet. He expresses himself thus: “The immunity of animals and plants in complete health depends in my opinion: (1) on the relative quantity of salt contained in their fluids, and (2) on the property of their contractile cells of ingesting the enemy which enters the animal body” (p. 18). As these statements have been put forth without receiving any further development, in the midst of all kinds of other speculations, it is not astonishing that the words I have just quoted, as well as Roser’s pamphlet itself, should not have attracted the attention of either zoologists or medical men. In the reviews for these two sciences (Schmidt’s _Jahrbücher_ and the _Zoologischer Jahresbericht_ of the Zoological Station at Naples) it is not even mentioned. It appears that not only did other biologists and medical men attach no importance to Roser’s speculations, but that the author himself did not claim any great value for them. I draw this conclusion from the fact that five years after his first pamphlet he published a second on inflammation and healing[862] in which he does not apply his theory of immunity to explain these two phenomena. This new work is of an even more speculative character than was the first, and instead of attempting to show any relation between the anti-infective part played by the leucocytes and their migration during inflammation, Roser insists on the fundamental independence of this phenomenon of healing. For him the inflammation, accompanied by diapedesis, must not be looked upon as a healthy reaction of the body, but as a manifestation of disease. The heat which is observed under these conditions must be attributed in part at least to the production of heat by infective micro-organisms. I must confess that Roser’s two pamphlets were unknown to me for many years, and it was Hueppe who drew my attention to them by his mention of them in the fourth edition of his work on bacteriological methods[863] which appeared in 1889. I had then, independently of the Marburg surgeon and by a totally different path, arrived at my conclusions as to the part played by the amoeboid cells. At the commencement of my researches on healing and immunity the passages cited above from the publications of Panum, Gaule, and Grawitz were also unknown to me. Having long studied the problem of the germinal layers in the animal series, I sought to gain some idea of their origin and significance. The part played by the ectoderm and the entoderm appeared quite clear, and the former might quite reasonably be regarded as the cutaneous investment of primitive multicellular animals, whilst the latter might be regarded as their organ of digestion. The discovery of intracellular digestion in many of the lower animals led me to regard this phenomenon as characteristic of those ancestral animals from which might be derived all the known types of the animal kingdom (excepting, of course, the Protozoa). The origin and the part played by the mesoderm appeared the most obscure. Thus, certain embryologists supposed that this layer corresponded to the reproductive organs of primitive animals: others regarded it as the prototype of the organs of locomotion. My embryological and physiological studies on sponges led me to the conclusion that the mesoderm must function in the hypothetically primitive animals as a mass of digestive cells, in all points similar to those of the entoderm. This hypothesis necessarily attracted my attention to the power of seizing foreign corpuscles possessed by the mesodermic cells. This fact has long been recognised. It was known that the white corpuscles of the Vertebrata often contained various kinds of cells, especially red and white blood corpuscles. It was known, also, that the amoeboid cells were capable of ingesting granules of coloured substances. When making an injection of indigo into the vessels of _Thetys_, Haeckel[864] in 1858 was surprised to find the blue granules inside the amoeboid blood corpuscles of this beautiful gasteropod mollusk. This fact has since been confirmed by many observers, and the capacity of the amoeboid cells to take up foreign bodies became recognised as a general phenomenon. Nevertheless this phenomenon was not regarded as being analogous to digestion. Thus Haeckel[865] himself, in his researches on the calcareous sponges, advocated the view that the foreign bodies penetrated into the interior of the viscous protoplasm in a purely passive fashion.

[Sidenote: [541]]

Observations that I made on sponges and on certain pelagic animals, transparent and of simple organisation, convinced me that the presence of foreign corpuscles in the amoeboid cells of the mesoderm must be attributed to an active ingestion by these cells which, in every respect, might be compared to the phenomena of intracellular digestion in the epithelial cells of the digestive canal of many of the lower animals. In order to demonstrate this fact clearly it was necessary to bring forward exact experimental proof. I set myself, therefore, during my stay at Messina in 1882 and 1883, to study the rôle of the amoeboid cells of the mesoderm from the point of view of intracellular digestion. I found it an easy matter to demonstrate that these elements seized foreign bodies of very varied nature by means of their living processes, and that certain of these bodies underwent a true digestion within the amoeboid cells. My principal thesis, that is to say the idea of the intimate relations between the entoderm and the mesoderm, was thus fully confirmed.

Pondering over these results, which were quite new at the time, the idea suggested itself to me that the digestive function, so profoundly rooted in the mesodermic elements, must play a part in many of the vital phenomena of animals. Starting from this standpoint, I succeeded in demonstrating that, during the very complicated metamorphoses of Echinoderms, such as the _Synaptae_, the amoeboid cells of the mesoderm fulfil a function in the atrophy of numerous larval organs. I have never prosecuted any medical studies; but some time before my departure for Messina I listened to the reading of Cohnheim’s treatise on General Pathology, and I was struck by his description of the facts and of his theory of inflammation. The former, especially his description of the diapedesis of the white corpuscles through the vessel wall, seemed to be of momentous interest. His theory, on the other hand, appeared to be extremely vague and nebulous. It occurred to me that a comparative study of inflammation in lower animals of simple organisation would certainly throw light on the very complex pathological phenomena in the Vertebrata, even in the frog which had served as the starting-point for Cohnheim’s remarkable experiments.

[Sidenote: [542]]

Since, in the atrophy of the larval organs of the _Synaptae_, the essential rôle is accomplished by the amoeboid cells of the mesoderm which accumulate and unite into masses, the richness of inflammatory exudations in white corpuscles may perhaps signify that these corpuscles have a very important function to fulfil. This reflection led me to make the following experiment: to wound and introduce spines beneath the skin of very transparent marine animals; if my hypothesis should be well founded this should bring about an accumulation of amoeboid cells at the injured spot. I selected for this purpose the large Bipinnaria larvae of starfish, so abundant at Messina, and inserted prickles of the rose into their bodies. Very shortly these prickles were found to be surrounded by a mass of amoeboid cells such as we see in human exudation as the result of the introduction of a spine or other foreign body. The whole process took place under my eyes in a transparent animal possessing neither blood nor other vessels, nor a nervous system. The first point was settled. The inflammatory exudation must be considered as a reaction against all kinds of lesions, the exudation being a more primitive and more ancient phenomenon in inflammation than are the functions of the nervous system or of the vessels.

I know quite well that, at the period when I made my researches (1882), pathologists regarded inflammation as the consequence, if not always, at least in the majority of cases, of the penetration of micro-organisms. From this followed the conclusion that the diapedesis and accumulation of white corpuscles in inflammatory diseases must be regarded as modes of defence of the organism against micro-organisms, the leucocytes in this struggle devouring and destroying the parasites. According to this hypothesis the significance of inflammation at once became simple and clear. With the object of verifying my hypothesis I began to make experiments on the lower animals, so abundant in the Straits of Messina, and to make myself acquainted with the results that had been obtained in general pathology and in pathological histology. A perusal of Ziegler’s treatise on Pathological Anatomy made it clear to me that in these branches of medical science there had long been accumulated a great number of observations fitted to facilitate the acceptation of the new hypothesis on inflammation and healing. Numerous and well-established facts on the absorption of extravasated blood, on the fate of the coloured corpuscles in the body, on the presence of micro-organisms inside leucocytes, etc., confirmed me in my view.

[Sidenote: [543]]

When I had got together certain information and a number of facts in support of my hypothesis I communicated the results to my lamented friend, Kleinenberg, at that time Professor in the University of Messina. Both medical man and zoologist, he was well qualified to offer a judgment upon the matter; this judgment was favourable. Sometime later I had the great pleasure of meeting the celebrated Professor Virchow at Messina. I imparted to him my ideas and he was kind enough to come with me to examine my preparations of Bipinnaria larvae and other lower animals in which I had set up the phenomena of inflammation without the assistance of nervous or vascular systems. This eminent observer greatly encouraged me to continue my investigations. When I explained to him my view that the inflammatory reaction on the part of the amoeboid cells could only be understood by accepting the hypothesis that the white corpuscles gave chase to the micro-organisms and destroyed them, Virchow replied that in pathology just the opposite was invariably taught. The general opinion was that micro-organisms were certainly found inside the leucocytes and that they made use of these cells as a means of transport and of dissemination through the body.

[Sidenote: [544]]

During my stay at Messina my researches were limited to the lower animals, but later I began to study inflammation and the phenomena of infection in the Vertebrata. It was not until eight months after I had commenced my researches in this direction that I decided to publish my results. I first set them forth in an address given at Odessa before the Congress of Naturalists and Medical Men in 1883. Later, they were published in a special article inserted in Claus’s _Arbeiten_ at Vienna[866], and in a small work which appeared in the _Biologisches Centralblatt_[867]. I sought especially to develop the idea that the intracellular digestion of unicellular organisms and of many Invertebrata had been hereditarily transmitted to the higher animals and retained in them by the amoeboid cells of mesodermic origin. These cells, being capable of ingesting and digesting all kinds of histological elements, may apply the same power to the destruction of micro-organisms. In order to support this conclusion I introduced various kinds of bacteria into the bodies of some of the lower animals and I demonstrated that they were ingested and destroyed by the amoeboid cells. It was evident, however, that this proof was not sufficient. I then set myself to study the diseases of small Invertebrata sufficiently transparent to be observed directly under the microscope. The _Daphniae_, those small crustacea so numerous and so frequent in fresh-water, furnished me with a favourable medium in which to study a real struggle which takes place between their leucocytes and the spores of a vegetable parasite belonging to the group of the Blastomycetes. In many cases the amoeboid cells guarantee the integrity of the animal by devouring a large number of these spores and transforming them into an inert detritus. In other cases, on the contrary, the fungi get the upper hand in the struggle; they succeed in germinating and in overcoming the resistance of the leucocytes by reproducing themselves rapidly and by killing these cells with their poisons. The history of this disease and of this struggle was published in _Virchow’s Archiv_[868].

Some time afterwards I published in the same journal my work on the anthrax bacillus[869], in which I attempted to demonstrate that in the Vertebrata also the invasion of pathogenic micro-organisms sets up a desperate struggle between them and the amoeboid cells.

In these four works I made use of the term “phagocytes” to designate the amoeboid cells capable of seizing and digesting the micro-organisms and other formed elements. To the theory based on this property of the defensive cells I gave the name of “theory of phagocytes.”

I thought, as already mentioned above, that the observations on absorption and leucocytes, which had been accumulating for years in pathological histology, had sufficiently paved the way for a favourable reception to the idea that the amoeboid cells are defensive elements of the body capable of guaranteeing to it immunity and cure. In this I was mistaken. It was precisely the specialists in this branch of science who from the first manifested the most lively opposition to this theory.

However, in the Presidential Address delivered before the 66th meeting of the British Association held at Liverpool in 1896, Lord Lister said[870]: “If ever there was a romantic chapter in pathology, it has surely been that of the story of phagocytosis.” These words encourage me to put before the reader the essential features of this story.

[Sidenote: [545]]

My first two memoirs published in 1883 did not in any way attract the attention of the medical public. These investigations had a character that was too zoological to be noticed by pathologists. But the two following publications, in which I treated of the _Daphnia_ disease and especially of bacterial anthrax, immediately roused severe criticism. Baumgarten[871], the well-known pathologist, opened the battle by the publication of a review of my researches on phagocytosis. He attempted to sap the basis of my theory, and not contented with _à priori_ arguments, he set his pupils to make a series of researches on the fate of micro-organisms in the refractory animal. These researches resulted in several theses for the doctor’s degree which sought to demolish every point of the theory of phagocytosis.

Later, Baumgarten[872] published a long and above all admirably written analytical article entitled: “Zur Kritik der Metschnikoff’schen Phagocytentheorie,” in which, with much talent and wit, he attempted to demolish the bases and conclusions of the phagocytic theory.

Baumgarten regards the precise observations which I had been accumulating for some years as incorrect and refuted by the observations and experiments of his pupils. The arguments that I give to justify my theory are, according to the same critic, contrary to logic and to truth. If the phagocytes are really elements destined to guarantee the integrity of the animal organism how is it, asks Baumgarten, that just at the moment of greatest danger, when the blood and the tissues are invaded by the micro-organisms, the leucocytes are conspicuous by their absence? The answer that there is no predestination in the phagocytosis, and that the danger is the greater the more feeble the phagocytic reaction—a fact which is in perfect harmony with the law of causes and with the principles of the evolution of species according to Darwin’s theory—did not satisfy my critic. He says: “If the interpretation which Metschnikoff gives of the activity of the leucocytes appears to be rather the product of a rich imagination than the result of the objective observation of the seeker, it matters little that his account of the development of the leucocyte in what he wishes to see in it should be in conformity with the principles of the theory of evolution” (p. 4).

[Sidenote: [546]]

I was able by numerous researches[873] to refute point by point the objections based on the work of Baumgarten’s pupils, but that did not prevent him from persisting in his negation. Only, commencing by writing long articles, he contented himself, later, with denying the theory of phagocytosis in small annual notes, appearing in his reviews of works on bacteriology, which were unsupported either by argument or by any facts mentioned in his abstracts.

Baumgarten’s example was followed by many other pathologists. Ziegler, the well-known author of a text-book on pathological anatomy that has certainly had a wider circulation than any other work, vigorously attacked the theory of phagocytosis. As it was precisely from this treatise that I had acquired my knowledge of the large number of facts that had accumulated in pathological literature on the part played by leucocytes in resorption, I was persuaded that Ziegler, who had collected these statements, would be one of the first to recognise the importance of phagocytosis in inflammation, healing, and immunity. But this distinguished pathologist, in several of his publications[874], expressed himself very vigorously against the phagocytic theory. The intervention of these cells, according to him, must be purely accidental and their rôle in the defence of the body against the micro-organisms very insignificant. The better to demonstrate this thesis he caused his pupils to undertake investigations on several infective diseases, and these young observers all arrived at the same result, that phagocytosis has nothing to do with the struggle of the animal against the anthrax bacillus or against the bacillus of symptomatic anthrax. It is the less necessary to enter into these details now because I have, in the preceding chapters, given sufficient proofs of the incorrectness of the objections advanced by Ziegler’s school. It has been demonstrated most conclusively (by Lubarsch’s researches, as well as by many other works) that in anthrax in man phagocytosis, denied by one of Ziegler’s pupils, is most marked. It is likewise well known from the researches of Ruffer, Leclainche and Vallée, as well as from my own observations, that in symptomatic anthrax, in which the phagocytic reaction is denied by another of Ziegler’s pupils, it is a very important and highly developed feature.

[Sidenote: [547]]

The opposition emanating from another eminent pathologist, Weigert[875], particularly impressed me, because this investigator is known not only to be an observer of great accuracy but to possess a mind of great imagination and generalising power. In several papers he put forward his utmost ingenuity to demolish the phagocytic theory root and branch. He would recognise neither the importance of phagocytosis in healing and immunity, nor the defensive function of the giant cells. Weigert, however, contented himself with formulating theoretical objections, and no works directed specially against the doctrine of phagocytosis have issued from his laboratory. It must be stated, however, that although there has been such opposition on the part of certain of our most eminent pathologists, others amongst them have, from the beginning, expressed themselves in more favourable terms. Thus, Virchow[876], in an introductory article in the 101st volume of his _Archiv_, continued his friendly attitude with regard to the works on phagocytic defence and spoke of them as opening up a new field of research. Ribbert[877], in a series of publications, maintained the importance of the phagocytes in the resistance offered by the animal to the aggression of micro-organisms, and pointed out, especially in connection with the diseases set up by the staphylococci, the frequency of the ingestion of these parasites by the leucocytes. He insists specially on a modification of the phagocytic reaction, which consists in the accumulation of white corpuscles around the centre of microbial infection. In these cases, without the occurrence of any real ingestion of the micro-organisms into the substance of the phagocytes, these organisms may have their morbific manifestation hindered by the assemblage of the white corpuscles. It is needless to insist that this act, which I referred to in my first work in 1883, constitutes the prelude to a true phagocytosis and is closely bound up with this defensive phenomenon. Another pathologist, Hess[878], supports the theory of phagocytosis by confirmatory researches of great value.

[Sidenote: [548]]

The pathologists who were adversaries of the phagocytic theory combined their efforts to demolish it, without troubling themselves to replace it by any other theory of defence on the part of the body which might more easily be made to accord with their principles and their statements. Baumgarten certainly tried to prove that micro-organisms perish in cases where immunity is produced or healing occurs, not as the result of the phagocytic reaction or of any other manifestation on the part of the menaced animal, but simply “of themselves” (von selbst), that is to say, they have simply accomplished the normal cycle of their existence and die a natural death, this bringing about healing and immunity. As may be readily understood he was unable to bring forward the slightest evidence of the correctness of this hypothesis, which, I believe, has never been accepted by anyone, nor even been defended by its author. In this respect the attacks directed against the theory of phagocytosis by bacteriologists have been of a very different character. Not content with overturning this hypothesis, these observers have sought to build upon its ruins new theories capable of offering a better explanation of the phenomena of immunity. I must here confess at the outset that these attacks have been much more important than those coming from the pathologists and pathological anatomists, and have led to discoveries of the greatest value.

One of Fodor’s experiments[879], one not altogether new, served as the point of departure for much work and for a large series of objections directed against the phagocytic theory. The Hungarian investigator found that the defibrinated blood of the rabbit was capable of destroying _in vitro_ a great number of anthrax bacilli. From this it was concluded that the fluids of the living body possessed a bactericidal power sufficient to explain the immunity against infective micro-organisms. The destruction of the anthrax bacillus by defibrinated blood was confirmed by a young American investigator of great talent, Nuttall[880], who carried out an important work on this subject in the laboratory and under the direction of Flügge at Breslau. He was able to follow step by step, by the observation of anthrax bacilli on the warm stage, their degeneration under the action of the defibrinated blood. This destruction of the bacilli took place outside the phagocytes. The same phenomenon could be shown by the method of gelatine plate cultures. The bacilli, subjected to the influence of the defibrinated blood of rabbits and other vertebrates, usually died or were markedly injured. The blood when heated to 55° C. completely lost its bactericidal power.

[Sidenote: [549]]

These observations, perfectly exact in every detail, gave Flügge[881] and his assistant Bitter[882] the opportunity to criticise vigorously the theory of phagocytosis. The cells were said to be incapable of ingesting living micro-organisms; these latter must be previously destroyed by the bactericidal action of the body fluids, and it was only their dead bodies which were devoured by the phagocytes.

Flügge based his criticism upon considerations of a general character and upon observations made mainly by Nuttall. “There is no necessary point of analogy,” says the learned Breslau hygienist, “between the ingestion of food and the struggle against infective micro-organisms, nor between nutritive substances and living micro-organisms” (p. 225). “From Nuttall’s results it must evidently be accepted as possible that the phagocytes can ingest dead bacteria only and that they have not the power of ridding the body of the living infective agents” (p. 226). The following passage is especially significant. “When we examine, with an open mind, a series of preparations which show the relations between the phagocytes and the bacteria in various infective diseases, the phagocytes sometimes present themselves as the victims of the bacteria, which continue their triumphal march; sometimes they produce the impression of tombstones lying in large numbers behind the line of battle and after the end of the struggle. On the other hand, they in no way force themselves upon our notice as instruments of slaughter which the attacked organism makes use of to defend itself” (p. 227).

These arguments have been regarded by many investigators in all countries as perfectly sufficient to overthrow the phagocytic theory. The bactericidal power of the body fluids became the rallying cry of a great number of works always directed to the same object: to replace the rôle of phagocytosis by that of a bactericidal power of the body fluids. It is quite unnecessary to weary the reader with a list of the very numerous publications that have appeared on this subject in every European language. But it is not possible to pass over in silence the work of some of the principal partisans of the humoral theory of immunity.

[Sidenote: [550]]

The first place amongst these works certainly belongs to von Behring’s memoir[883] on the natural immunity of white rats against anthrax. As already stated in Chapter VI of this work, von Behring discovered the very remarkable power possessed by the rat’s blood of destroying anthrax bacilli with very great rapidity. This investigator did not hesitate to conclude therefrom that this bactericidal property of the blood must, in the rat, bring about a great resistance against anthrax. We should have in this case, then, an example in which the immunity did not depend in any way upon phagocytosis, but would be bound up entirely in a purely humoral property.

With the object of deciding whether the bactericidal property of the blood is really the general and essential cause of natural or acquired immunity, von Behring, in collaboration with Nissen[884], carried out a long series of experiments, the results of which, however, did not confirm their expectations. They found that in animals well vaccinated against certain bacteria (notably Gamaleia’s vibrio or _V. metschnikovi_), the blood plasma undoubtedly acquires a high specific bactericidal power, but at the same time they satisfied themselves that the blood, even of well immunised animals, was generally incapable of killing the micro-organisms. The bactericidal property, then, according to their researches, presented itself not as a general character but as one of limited importance. These facts even led von Behring to abandon the theory of the bactericidal power of the body fluids as an explanation of immunity.

This theory found many warm partisans, especially at Munich. Emmerich had already announced at the International Congress of Hygiene, held at Vienna in 1887, that in the blood of rabbits vaccinated against the bacillus of swine erysipelas an antiseptic substance of remarkable activity is produced. To this, exclusively, in this instance, and not to the phagocytes, he attributed the acquired immunity. Later, Emmerich[885] in an investigation carried out in collaboration with di Mattei developed this view. We may refrain from giving any account of the contents of their memoir as well as from criticising their conclusions, as this has already been done in Chapter IX. Let us content ourselves with stating that our own experiments, as well as those made later by Mesnil, have demonstrated the inaccuracy of Emmerich’s statements.

[Sidenote: [551]]

Another Munich bacteriologist, H. Buchner, at first expressed himself[886] very favourably on the theory of phagocytosis. He regarded it as more capable of explaining most of the phenomena of immunity than was his own older local theory. But little by little he declared himself in formal opposition to the cellular theory of immunity and went over to the camp of his sometime adversaries. He adopted[887] the humoral theory of the bactericidal action of the body fluids, upon which subject he carried out several important investigations. He was able without difficulty to confirm Nuttall’s discovery of the disappearance of the microbicidal power when the defibrinated blood was heated to 55° C., and he added to this fundamental fact many others of great value. He demonstrated the part played by the salts in the exercise of this bactericidal power, and laid great stress on the fact that this power depends on the presence of a special substance of albuminoid nature, to which he gave the name of _alexin_. Buchner[888] combatted with success the idea that I had expressed, according to which the bactericidal power of the body fluids is reduced in great part to a plasmolytic action of the blood serum upon certain micro-organisms. It cannot be denied that my hypothesis is only very partially applicable, and that the larger share in the bactericidal action of the body fluids belongs to the alexins. Buchner also made the study of this action more easy by the demonstration that the red blood corpuscles of a foreign species undergo, under the action of the blood and of the serums, a globulicidal action comparable to that which occurs in the case of micro-organisms.

Whilst Flügge, von Behring and many others of the old partisans of the bactericidal theory of the body fluids abandoned it more or less completely as an explanation of immunity, Buchner remained faithful to it and tried, aided by the collaboration of his pupils, as far as possible to defend it.

[Sidenote: [552]]

In France this humoral theory was adopted chiefly by Bouchard[889] and his pupils, amongst whom I must cite more particularly Charrin and Roger. They sought to confirm it by personal researches, the greater part of which were carried out upon the bacillus of blue pus. These investigators studied it especially in relation to acquired immunity. A comparison of the mode of development of the pyocyanic bacillus in the serum of susceptible animals and of vaccinated animals of the same species, convinced them of the great importance of the action of the body fluids. In cases where these fluids were found to be incapable of killing the micro-organisms they exerted over them an injurious influence, either by attenuating their virulence, or by producing more or less important modifications in their forms and functions. The essential cause of natural or acquired immunity was always attributed by Bouchard’s school to the property of the body fluids. The phagocytes were said to intervene only secondarily, either to carry off the dead bodies of the micro-organisms, or to ingest the bacteria, rendered inoffensive by the humoral action.

The humoral theory of immunity, with some slight modifications, spread very generally into every country, and many investigators accepted it without reserve. But certain observers ventured to run counter to the general current and raised objections of principle against the theory of the bactericidal power of the fluids of the body. After the principal facts established by the partisans of this theory had been confirmed, it was asked whether the phenomena of the destruction of micro-organisms observed _in vitro_ are really equivalent to those produced in the refractory animal. A glance at the data brought together with so much zeal was sufficient to demonstrate that this parallelism does not exist. The blood of animals susceptible to certain micro-organisms was found to be bactericidal for these organisms, whilst that of refractory animals was incapable of destroying them. It is useless to cite examples, so numerous are they. On the other hand, the bactericidal power of the body fluids, so marked for certain pathogenic organisms such as the anthrax bacillus and especially the cholera vibrio and the typhoid coccobacillus, is insignificant or _nil_ as regards many bacteria against which refractory animals are not wanting.

[Sidenote: [553]]

All these facts throw doubt on the predominating part played in immunity by the bactericidal power of the body fluids. Lubarsch[890] attacked the humoral theory, showing by a great number of experiments that animals whose fluids are very bactericidal _in vitro_ are very susceptible to a much smaller quantity of bacteria of the same species introduced into the body. Thus, the defibrinated blood and the blood serum of rabbits destroy a large number of bacteria in a very short time, whilst the rabbits themselves contract fatal anthrax after the introduction of a small number of these micro-organisms into the blood vessels. This contradiction cannot be explained except by the profound changes which the blood must undergo outside the body. Facts of the same nature have been shown for the anthrax of rats by Hankin, Roux, and ourselves, as described in Chapter VI.

[Sidenote: [554]]

The International Congress of Medicine, assembled at Berlin in 1890, was the first occasion on which I spoke publicly of the new theories of immunity. In the addresses given at the general meetings, leaders of medical science in several countries summed up their opinion on this question. Koch[891], in his memorable report, declared that the new acquisitions had destroyed the basis of the theory of phagocytes, and that consequently it must give place to the humoral theory of immunity. Bouchard took up a more conciliatory position, but, according to him, the bactericidal power of the fluids of the body was the primary and essential cause of immunity. The phagocytes only intervened later, in order to finish the work begun without their assistance. Lord Lister expressed himself[892], on the other hand, much more favourably on the subject of the theory of phagocytosis. This observer, who is not only a great surgeon, but is perhaps even more remarkable for his great powers of generalisation, has paid special attention to the problem of immunity. With the object of clearing up this very complicated and at the same time important question, Lord Lister seized the occasion of the meeting of the International Congress of Hygiene in London in 1891, to bring about an exchange of views between the partisans of the various theories of immunity. Under his presidency he devoted an entire sitting of the Section of Bacteriology to the discussion of this question. Buchner presented a report[893] drawn up exclusively from the point of view of the humoral theory and devoted to the demonstration of the slight importance of phagocytosis, and also to the preponderant part played by the alexins dissolved in the body fluids and circulating in the plasma of the blood. He attempted to harmonise the facts on the bactericidal power of serums observed _in vitro_ with the special conditions to be met with in the animal body. He specially insisted on the point that, in the blood and the organs, the alexins cannot act with the same rapidity that they can in test tubes containing serum. In this way he recognised that between the bactericidal action _in vitro_ and that in the body of the animal, there exists a marked difference, but he would not consent to attribute it in the latter case to the intervention of the phagocytes.

Roux[894] also made a report on immunity at the same sederunt, speaking very distinctly in favour of the cellular theory. A chemist by inclination, he was sympathetic at first to the humoral theories of immunity. Working with Pasteur, and side by side with him, Roux, from the beginning of the new era of medical science, had made numerous experiments on the part played by the body fluids in immunity. But as the results were not sufficiently precise and demonstrative they were soon abandoned. The attachment of Roux, however, to the humoral theories was manifested in his work, carried out in part with Chamberland[895], on the subject of vaccination by means of microbial products. Later, having obtained a deeper knowledge of various facts concerning natural and acquired immunity, he rallied to the cellular conception and developed it in his report presented to the above Congress in London. Several microbiologists took part in the discussion, and I myself[896] was able to communicate certain facts concerning the immunity of guinea-pigs, acquired as the result of vaccination against Gamaleia’s vibrio. I chose this example because it presented, according to von Behring and Nissen, the clearest case of a bactericidal property developed during the course of immunisation. I was able to furnish the proof that, in the vaccinated animal, the micro-organism in question, in spite of the great bactericidal power of the blood serum _in vitro_, remains alive in the animal body for a long time, and that its destruction is effected by the phagocytes, which ingested it alive. In this example I showed that the leucocytes of the exudation, that have ingested vibrios, may still furnish cultures of this organism if they are taken from the body and transferred in hanging drop to the incubator.

[Sidenote: [555]]

The fact that, even in the case which appeared most to favour the humoral conception of acquired immunity, phagocytes play the principal part, must to many members of the Congress have appeared sufficiently significant. Indeed, several observers who were present at the debates, received the impression that the phagocytic theory had not been overturned by its adversaries. At this period the question of the importance of antitoxins from the point of view of immunity had scarcely been raised. The great discovery made by von Behring and Kitasato was already accepted by everyone; but there was no ground for attributing to it any general importance. In fact, though proved for tetanus and diphtheria, and extended by Ehrlich’s beautiful experiments to the vegetable toxins (ricin, abrin, and robin), the antitoxic property of the fluids of the body presented itself rather as a special than as a general phenomenon. It is in this sense that Roux had assigned to it its place in the chapter of immunity. The two diseases, against which antitoxic serums had been discovered, are certainly distinguished from the great majority of infections by the localisation of the micro-organisms and the abundant secretion of their toxins.

It was only after the London Congress that this question came prominently forward. Von Behring thought that the antitoxic power of the body fluids is generally distributed in all cases of acquired immunity, and that micro-organisms, introduced into the animal possessing this power, become incapable of any pathogenic manifestation. Certain facts, brought together in Bouchard’s laboratory, tell against the hypothesis I have just mentioned. With the object of throwing light on this question I began, immediately after the close of the Congress, to study the acquired immunity of rabbits against the micro-organism of the pneumo-enteritis of pigs. I was able to demonstrate[897] that in this case the resistance of the animal against the micro-organisms does not depend on the acquisition of any antitoxic property by the body fluids; such a property is completely absent. At the same time I showed that the serum of vaccinated rabbits possesses a very marked protective power against infection by the coccobacillus of pneumo-enteritis. It was for the first time proved that independently of the antitoxic and bactericidal properties of serums, there exists another special property, the anti-infective property. This I conceived to be of the nature of a stimulant action on the part of the phagocytes.

[Sidenote: [556]]

It has already been stated in an earlier chapter that before the discovery of antitoxins Richet and Héricourt[898] had observed an immunising action of the serum of animals refractory to staphylococci. These observers were content with this demonstration and did not seek to penetrate more deeply into the mechanism of the action of their serum. For this reason when von Behring and Kitasato announced their discovery of antitoxic serums it was generally thought that the antistaphylococci serums were also antitoxic serums. The immunity against the micro-organism of the pneumo-enteritis of pigs taught us that here we might have to deal with quite a different matter. It was soon demonstrated that the serum from the immunised animal might in fact, without being antitoxic, present the same anti-infective property as in the case of pneumo-enteritis. That was first proved in the case of the experimental disease set up by Koch’s cholera vibrio.

[Sidenote: [557]]

The reappearance of cholera in Europe in 1892 drew the attention of bacteriologists to this disease, and was the occasion of many new researches on immunity against the cholera vibrio. Several important works on this question were published by Pfeiffer[899], at this period director of the scientific staff of the Koch Institute at Berlin. He obtained, in animals well immunised against the cholera vibrio, a serum endowed with a high anti-infective power but entirely without any antitoxic property. The guinea-pigs themselves, very resistant to the cholera peritonitis, were found, on the other hand, to be very susceptible to the minimum lethal dose of the cholera poison. The absence of antitoxic power in the fluids of the body taken in connection with a well-marked phagocytic reaction in a large number of cases of immunity, natural and acquired, has turned the scale in favour of the cellular theory. The impossibility on the part of those who maintain the purely bactericidal theory of the body fluids, to reply to the objections above mentioned has accentuated this favourable movement. Just at this moment, when the theory of phagocytes might be regarded to have obtained the rights of citizenship, a discovery was made which appeared to overturn it completely. I have mentioned more than once that the attempts of the partisans of the bactericidal theory of the body fluids have failed whenever it was necessary to give evidence of their action in the refractory animal. Instead of a destruction of the micro-organisms in these fluids, it was always found that they perished inside the phagocytes. These facts have even led to the manifestation of a desire to harmonise the humoral theory with the theory of phagocytosis. Denys, with certain of his collaborators, and Buchner and his pupils came to the conclusion that the alexins are merely leucocytic products. As regards the theory of phagocytosis we have this section, who attribute an important function in healing and immunity to the emigration of the leucocytes towards, and their accumulation at the menaced spot. They admit that the leucocytes really represent the healing elements of the animal body; it is not, however, they say, their phagocytic functions which confer upon them this rôle but their power of secreting alexin. These bactericidal substances act outside the phagocytes—in the plasmas of the blood and of the exudations—and phagocytosis only intervenes at a later period and secondarily.

This new modification of the bactericidal theory of the body fluids has often been termed by Buchner a connecting bridge between the humoral theory and the cellular theory of immunity.

In the midst of this movement of conciliation, Pfeiffer[900] in 1894 published a work on the immunity of the guinea-pig against experimental cholera peritonitis. He maintains that here the destruction of the vibrios takes place without any co-operation on the part of the phagocytes and exclusively by means of the body fluids. The vibrios, before their complete destruction and solution in the fluids of the body, are transformed into granules, presenting the transformation to which we have given the name of Pfeiffer’s phenomenon.

[Sidenote: [558]]

Several of Pfeiffer’s pupils have confirmed his view in connection with the cholera vibrio, and have extended it to several other micro-organisms such as the typhoid coccobacillus. The destruction of the micro-organisms in these cases is brought about, according to Pfeiffer and his collaborators, not by the alexins of Buchner, but by a separate substance. The protective anti-infective serum contains it in an inactive state only; but immediately this serum is introduced into the body of a normal animal, the bactericidal substance is acted upon by the endothelial cells and becomes “active,” capable of destroying a large number of vibrios. Pfeiffer has developed this theory more especially in an article published in 1896, entitled “Ein neues Grundgesetz der Immunität[901].” Pfeiffer’s observation and his theory built upon it gave a new lease of life to the humoral theory and for some time many observers believed that the theory of phagocytosis was now finally overturned. Fränkel[902] announced, in a public address, that science in its progressive march has “discovered the methods of defence employed by the animal organism against its most dreaded enemies, methods which have nothing in common with phagocytosis, which act quite independently of the phagocytes and manifest an action so energetic that we may calmly eliminate all other factors.” This view is based on the discovery of antitoxins and the bactericidal substance studied by Pfeiffer.

It will be readily understood that as soon as I learnt of the existence of a real extracellular destruction of micro-organisms I at once began to study it in order to find out its real importance amongst the phenomena of immunity. First of all, I examined Pfeiffer’s phenomenon in connection with the cholera vibrio[903], and I was able to show that it was produced only under special conditions. The pre-existent phagocytes must be greatly injured before the cholera vibrios can be transformed into granules. Phagolysis (so I termed this transitory damage to the phagocytes) is indispensable for the manifestation of Pfeiffer’s phenomenon in the peritoneal fluid. When it is suppressed, by preparing the phagocytes by means of injections of various fluids, we find that, instead of Pfeiffer’s phenomenon, phagocytosis is almost instantaneously produced. In positions where very few or no leucocytes are pre-existent, as in the subcutaneous tissue, Pfeiffer’s phenomenon is never observed.

[Sidenote: [559]]

Even in the case of the cholera vibrio the extracellular destruction is observed, therefore, only in special cases. Most of the other pathogenic micro-organisms do not undergo this destructive process at all under conditions in which the cholera vibrio exhibits Pfeiffer’s phenomenon in a marked degree. These facts appeared to justify me in the conclusion that the destruction of micro-organisms takes place in the animal body by means of soluble ferments, the result of phagocytic digestion. These ferments are found under the normal condition within these phagocytes and escape from them when they are destroyed or receive some transient injury. This conclusion was in flat contradiction to the theory and statements of Pfeiffer, who attributed an important function to the endothelial secretions. To settle this controversy I tried to obtain Pfeiffer’s phenomenon outside the body, that is to say independently of any co-operation from the peritoneal endothelium. It is sufficient to add a little peritoneal lymph, rich in leucocytes, to the inactive anti-infective serum, to obtain in hanging drops the transformation of the cholera vibrios into granules.

Bordet[904], in my laboratory, repeated this experiment with the object of determining its essential mechanism. He succeeded in obtaining Pfeiffer’s phenomenon _in vitro_, not only by adding peritoneal lymph from a normal guinea-pig to the specific serum, but also by adding to it a drop of fresh blood serum from the same animal. The analysis of the phenomena which take place under these conditions led Bordet to the following hypothesis. The destruction of micro-organisms in vaccinated animals takes place by the co-operation of two substances. One of these is Buchner’s alexin which is found normally in the phagocytes; it sets up bacteriolysis properly so-called when it is enclosed within the leucocytes or after it has escaped from them at the time of phagolysis. To attain this end, however, the alexin needs the co-operation of another substance. This is the protective or sensibilising substance of Bordet. It circulates in the plasmas and carries a specific character which is absent from the alexin. I need not here insist at any length on this theory, because it has already been sufficiently explained during the course of this work.

[Sidenote: [560]]

The data on the restricted part played by Pfeiffer’s phenomenon and on its mechanism, above summarised, have been attacked by Pfeiffer and by several other observers, but they have received general confirmation, so that their accuracy can no longer be in doubt. Objections were also raised to Bordet’s view of the mechanism of bacteriolysis. Thus, Abel has criticised it in the following argument[905]: “In spite of the soundness and the boldness of the majority of Bordet’s statements on the importance of the various factors, and especially of the leucocytes in immunity, it cannot be doubted that later researches will modify and correct his interpretations which we, in Germany, do not accept in their full extension. Up to the present, the victory in the various rounds has always been with Pfeiffer, whose researches, solid and exempt from bias, have made him, to use a sporting expression, the ‘favourite’ with all those who follow attentively the international contest in the arena of the problem of immunity.” Abel is certainly a highly esteemed bacteriologist, but he is not a good prophet, and he assumes a mistaken attitude in looking at the subject from a “national” point of view[906]. In Germany much interest is taken in scientific movements and, very naturally, original and new theories are there criticised and discussed. But that does not justify one in putting forward against an opinion the statement that it is not accepted in Germany. In this country, so rich in scientific work, we find partisans of the most opposite views. In any case, in the conflict between Pfeiffer on the one hand, and Bordet and myself on the other, things have not turned out as Abel predicted. The two substances which act in the destruction of the micro-organisms are now accepted by the whole world. The intimate relations between the alexins and the leucocytes are equally recognised by very many observers. The fact that the alexins are confined within phagocytes has been confirmed by several observers, and has received a very convincing proof from Gengou’s experiments on the comparative action of the serum and blood plasma against micro-organisms. The existence of phagolysis, denied at first by some observers, has been verified by others and can now no longer be doubted.

[Sidenote: [561]]

The relations between the sensibilising substance and the phagocytes are less easily grasped than are those between the alexins and the leucocytes. Nevertheless, the experiments made by Pfeiffer and Marx[907], have led these observers to recognise that the former arise from the spleen, the lymphatic glands, and the bone marrow, that is to say, organs which are pre-eminently phagocytic. This result has been confirmed by Deutsch and must be regarded as definitely settled. All the data collected in recent years have, therefore, confirmed the view that the destruction of micro-organisms in the refractory animal presents itself as a special example of their absorption by formed elements. This truth was so fully recognised in our laboratory that the analogy between bacteriolysis and the destruction of animal cells was looked upon as quite natural and evident. Bordet had for some years past observed that the blood serum of certain animals presented a marked analogy in its agglutinative property in regard to micro-organisms and in that against red blood corpuscles. In 1898, studying the fate of the spirilla of the goose in the peritoneal cavity of guinea-pigs (see Chapter VI), I observed that these micro-organisms underwent the same changes both within and outside the phagocyte; this fact appeared to me to be in perfect harmony with the whole of our knowledge concerning the absorption of formed elements and on intracellular digestion.

Bordet[908], prepared by his preceding researches on the agglutination of the red blood corpuscles, set himself to study the fate of the red corpuscles in the animal body. He easily established a close relationship between the development of the bacteriolytic property and the haemolytic power of the serum of animals prepared by repeated injections of bacteria and of blood. His results were soon (January, 1899) confirmed by Ehrlich and Morgenroth[909], who supplemented them with the important statement that Bordet’s sensibilising substance, or intermediary substance (E. and M.), has the property of attaching or fixing itself to the red blood corpuscles.

[Sidenote: [562]]

The works on haemolysis, carried out during the last three years by Ehrlich and Morgenroth on the one hand, and by Bordet on the other, have allowed us to extend our study of the mechanism of the action of the two substances on micro-organisms and on animal cells. Ehrlich has extended his ingenious theory of antitoxins to the bacteriolytic substances, which he regards as side-chains detached from the cells and capable of absorbing the toxins. In a series of remarkable investigations, most of them carried out in collaboration with Morgenroth, Ehrlich has developed his theory which attempts to offer an account of the essential mechanism which presides over the destruction of micro-organisms and over the neutralisation of their poisons. This theory is at present in full swing of development. Some of his points contradict several of the conclusions in Bordet’s works. Whilst the latter maintains that the sensibilising substance becomes fixed as a mordant, Ehrlich regards it as entering into chemical combination with the molecular group of the micro-organisms and of the animal cells. According to Bordet, the alexin of the same species of animal is always the same substance. Ehrlich energetically maintains the plurality of the alexins, to which he gives the name of complements.

This controversy has caused a most interesting exchange of views and has led to experiments which are remarkably ingenious, but it must be admitted that as yet all the points in dispute are not definitely settled. It is evident that we have here a new line of research which promises most fruitful results for science.

We have described in various chapters of this work the fundamental elements of Ehrlich’s theory. Many think that this theory is, in principle, antagonistic to the theory of phagocytosis, but we have already observed that this view cannot be accepted. It is true that Ehrlich maintains that the bacteriolytic and cytotoxic ferments which we have called _cytases_ (alexins or complements) circulate in a state of solution in the blood plasma, whilst, according to the theory of phagocytosis, they are found under normal conditions inside phagocytes. But this view has nothing to do with the basis of the theory of receptors, or of Ehrlich’s side-chain theory, according to which the antitoxin and certain other antibodies (intermediary substance) are regarded as products detached from cells having an affinity for the toxins and the microbial products.

The theory of phagocytosis seeks to establish the part played by these cells in the destruction of micro-organisms. It maintains that the vital manifestation of the phagocytes, irritability, mobility, and voracity, constitutes an essential factor in ridding the animal of micro-organisms, because the true bactericidal ferment is contained within the phagocytes, except in cases of phagolysis. The destruction of the micro-organisms follows the laws which govern the absorption of formed elements in general. This absorption, finally, is the work of two soluble digestive ferments, one of which (fixative) is readily excreted by the phagocyte into the plasmas of the blood and exudations. The theory of phagocytosis seeks to establish these principles with the greatest possible exactness, but it has not yet ventured to penetrate more deeply into the phenomena of intracellular digestion which are confounded with the action of soluble ferments in general. This problem is still far from being satisfactorily solved.

[Sidenote: [563]]

In spite of very numerous objections, of which the principal ones have already been mentioned, the theory of phagocytosis, within the limits indicated, so far from being overturned, has become more and more consolidated, thanks to the numerous observations made since its foundation. It is for this reason that the opposition has calmed down of late years and that in many works the opinions expressed have become more favourable to the rôle of phagocytosis in immunity.

Soon after the Congress of Hygiene in 1891, the Pathological Society of London devoted several meetings to a discussion of the question of immunity. Many eminent observers took part in these debates, which were, in general, favourable to this theory of phagocytosis[910].

At the International Congress of Hygiene, held at Budapest in 1894, the question of immunity was again discussed. Buchner[911] made a report in which he specially insisted on the leucocytic origin of the alexins, regarding this fact as particularly capable of reconciling the bactericidal property of the body fluids with the theory of phagocytosis. The alexins, however, secreted by the leucocytes, must, it was assumed, carry out their principal function in the plasmas of the blood and exudations. Phagocytosis would only intervene secondarily for the purpose of ingesting the micro-organisms which had been already killed or seriously injured by the alexins of the body fluids.

[Sidenote: [564]]

In his last summary of the question, presented to the International Congress of Medicine at Paris in 1900, Buchner[912] maintains his theory of leucocytic secretions. But he already takes one step more towards the theory of phagocytosis, at least as regards natural immunity. He consents to accept the fact “that phagocytic activity is in many cases of decisive importance in overcoming the infective processes, especially in those cases in which the secreted alexins were unable to bring about more than a temporary attenuation of the vital functions of the bacteria. Under these conditions the bacteria could only be modified in so far as their chemical functions were transformed into a latent state, from which they would be ready to regain their full vital activity should it happen that the phagocytes were not there to prevent them from doing so.” In any case this view is widely removed from the old theory, according to which phagocytes were regarded as capable of ingesting dead and inoffensive bacteria only.

A second adversary of the theory of phagocytosis, von Behring[913], gives a place to this theory not only in certain examples of natural immunity but even in some cases of acquired immunity, e.g. in the immunity of sheep vaccinated against anthrax, an example I have already cited in Chapter VIII (cf. _supra_, p. 242).

It would take too long to describe the change of opinion on the theories of immunity that has taken place during recent years. I will content myself with citing certain examples which shall be taken from the works of declared adversaries of the theory of phagocytosis. Thus, Flügge, who early declared against the cellular theory completely and categorically and at the same time argued strongly in favour of the humoral theory, has been gradually led to depart from his first position. We may follow the steps of his conversion in the different editions of his _Outlines of Hygiene_. In the first edition published in 1889 he expresses himself in the following manner[914]: “Recent researches indicate the probability, however, that the phagocytes in by far the greater majority of cases seize the infective agents which, already dead, are not in a condition suitable for the performance of a defensive function. On the other hand, it is proved that the blood and blood plasma of warm-blooded animals possess the property of destroying, very quickly, enormous numbers of pathogenic bacteria,” ... etc. In the fourth edition of the same work, published in 1897, we find at the corresponding place the following passage[915]: “Recent researches indicate the probability, however, that the theory of Metschnikoff ... is not in a position to offer a complete explanation of the process of immunity.” This passage is followed by a somewhat conciliatory and eclectic development of the theory.

[Sidenote: [565]]

Let us take as a second example Günther’s _Introduction to the Study of Bacteriology_, widely read both in the original and in translations. In the first edition published in 1890[916] the theory of phagocytosis is curtly dismissed as “being incapable of withstanding criticism.” In the fifth edition of the same work, however[917], published in 1898, this theory is no longer treated thus summarily. It is given a place amongst the theories of immunity and an attempt, similar to that made by Buchner, is made to reconcile it with the humoral theory.

A change in the same direction may also be observed in Charrin’s view. In the first edition of his _Pathologie générale infectieuse_, this observer[918] had already taken an eclectic view on this question of the theories of immunity. But the function which he assigns to the phagocytes is subsidiary and secondary, whilst to that of the humoral properties is assigned a position of primary importance. In the second edition of the same work, which appeared seven years later[919], the importance of phagocytosis is recognised in a much larger measure, as may be gathered from the following passages: “For my part, I have always accepted phagocytosis: at the same time I have always accepted the existence of special humoral properties. As early as 1888 I showed, _in vivo_, that the germs are modified outside the cells; but I did not know from what groups of anatomical elements these properties were derived, I exaggerated their importance and it is the decision of this origin and this importance that renders it possible to reconcile the two theories” (p. 250). “After all, the defence rests upon these two great processes or cellular activities, phagocytosis in the first line, and then humoral influences, some of them bactericidal and injurious to the living germ, others antitoxic and injurious to their secretions” (p. 253).

[Sidenote: [566]]

Whilst the theory of phagocytosis has been consolidated by the demonstration: (1) that the phagocytes, in cases of immunity, ingest and destroy the living and virulent micro-organisms without the latter needing to be previously deprived of their toxins; (2) that the phagocytes absorb toxic substances; (3) that the phagocytes contain bactericidal cytases and produce fixatives; the humoral theories, in spite of all the efforts made to defend them, could never be developed as theories that were in the slightest degree of general application. Certain observers who from the first were very sympathetic to the humoral theories have attempted to give a complete summary of these properties. Thus, Stern[920] and later Frank[921] have published reports drawn up with great care and in a very impartial spirit on the works treating of the properties of the body fluids and the part they play in immunity. This is how they sum up the question. Stern came to the conclusion that it is impossible “to demonstrate at all regularly the existence of relations between the bactericidal action of the blood and immunity in all the infective diseases. In some cases, however, these relations are so marked that, for these examples, a causal bond between the two factors is extremely probable.” Frank expresses himself in the following manner: “It follows most clearly that the immunity of an animal—immunity innate or acquired—corresponds with the bactericidal property of the blood in certain exceptional cases only. The only animal, absolutely susceptible to anthrax and whose blood is entirely without any bactericidal power, that it is at present possible to cite, is the mouse.” “The bactericidal action of the blood serum is undoubtedly a fact of great biological importance; but equally certainly it cannot be the general cause of immunity, whether innate or acquired.”

An attempt was made to give fresh life to the humoral theory, either by assuming that the bactericidal substance is nothing but the eosinophile or pseudo-eosinophile secretion of the leucocytes (Kanthack), or by supposing that for the destruction of micro-organisms in the animal body the intervention of the agglutinative substance dissolved and distributed in the body fluids is essential (Max Gruber). These two views were put forward in a tentative form and as preliminary communications only; there is no possibility of raising them to the dignity of theories, and of late years they have not been upheld.

It cannot be denied that not one of the humoral theories has been able to retain its position or to stand against the numerous facts that have been accumulated during recent years.

This extraordinary discrepancy between the bactericidal power of the body fluids and immunity is explained by the circumstance that the microbicidal substances exist in the living animal within phagocytes and only escape from them when these cells have been injured. The fact, so well demonstrated by Gengou, that the blood plasma is without any bactericidal power has given the final blow to the microbicidal theory of the body fluids and it can no longer be maintained.

[Sidenote: [567]]

The humoral theories, based on the antitoxic and protective power of the body fluids, can claim only a very restricted application. These properties are met with in acquired immunity only, and even there are not constant. Many cases of acquired immunity against micro-organisms are unaccompanied by any antitoxic power, and in several examples of this immunity the body fluids do not exhibit any protective power.

There is only one constant element in immunity, whether innate or acquired, and that is phagocytosis. The extension and importance of this factor can no longer be denied.

It is clearly proved that phagocytes are susceptible cells which react against morbific agents, whether organised or not. These cells ingest micro-organisms and absorb soluble substances. They seize microbes whilst these are still living and capable of exercising their noxious effect and bring them under the action of their cellular contents, which are capable of killing and digesting the micro-organisms or of inhibiting their pathogenic action. Phagocytes act because they possess vital properties and a faculty of exerting a fermentative action on morbific agents. The mechanism of this action is not yet definitely settled, and we can foresee that for future researches there will be a vast and fertile field to be reached by pursuing this path.

The present phase of the question of immunity constitutes one stage only in the development of biological science and one which is capable of many improvements.