It is usually understood that nutritive substances must necessarily be subjected to the influence of the digestive juices in the gastrointestinal canal before they can be utilised for the nutrition of the organism. This is a very old idea. It was based on a well-known experiment by Schiff who injected several animals intravenously with solutions of cane sugar and egg albumen and others with the same substances after they had been artificially digested. In the first case the food substances passed into the urine, in the second they only appeared there when injected in large quantities.

At the recent International Congress of Medicine held in Paris in 1900, the question of extra-buccal nutrition was much discussed^[90]. It has been accepted that fats, when injected into the subcutaneous tissues, are, at least in part, absorbed by the organism, but that carbo-hydrates and albuminoids are never absorbed. This is perhaps true from the point of view of clinical medicine. But, in principle, it must be admitted that food substances of very diverse natures, when introduced into the organism by channels other than the gastrointestinal canal, still undergo profound changes.

When we inject milk, blood serum, or white of egg, that is to say, materials very rich in albuminoid substances, under the skin or into the peritoneal cavity of laboratory animals, we find that after a time they disappear. At the same time they give rise to modifications of the organism which indicate that these injected substances have there undergone profound changes.

After injecting eel’s serum into rabbits, Th. Tchistovitch^[91] found a substance in the blood of the injected animals which gave a precipitate with eel’s serum. Shortly afterwards Bordet^[92] observed that the blood of animals into which he had injected cow’s milk acquired a new property: it gave a precipitate with this milk, a condition never observed in the serum of untreated animals.

The injection of white of egg into rabbits, carried out by Myers^[93] and Uhlenhuth^[94], brought about the same changes in the blood serum. The researches of the latter of these two observers have for our present purpose a special interest. He demonstrated first that the injection of white of egg into the peritoneal cavity of rabbits was followed by the appearance in the blood serum of these animals of a substance which precipitates egg albumen in vitro. Uhlenhuth then obtained this same acquired property of the blood in rabbits which had been made to swallow a considerable quantity of the white of hens’ eggs. Twenty-four days after the commencement of this regimen the serum of the rabbits precipitated white of egg in the test-tube. This example affords a marked analogy between the results of digestion in the alimentary canal and those of resorption into the tissues. Uhlenhuth points out, indeed, that his rabbits which received the injections of white of egg into the peritoneal cavity flourished under this treatment.

A certain number of similar examples are now recognised. They all indicate that various nutritive substances, when introduced into the peritoneal cavity or under the skin of animals, are retained there for a longer or shorter time and are subjected to certain modifying influences on the part of the organism. The proof that these substances are not eliminated intact by the kidneys has been furnished by a large number of experiments. Recently Lindemann^[95] and Néfédieff^[96], working in my laboratory, have established the fact that normal blood serum, when injected under the skin of animals, does not provoke albuminuria at all, or at least produces it in a very insignificant and transitory degree.

The mechanism by which the organism modifies these nutritive substances, introduced by a channel other than the digestive canal, is not as yet sufficiently known; and is therefore not easy to define. But we know, very definitely, that each injection of serum, whether of white of egg, milk or fatty matter, is followed by a rather considerable aseptic inflammation at the point at which these substances are introduced. We might conclude from this that the organism digests the food substances outside the gastro-intestinal canal, by means of an inflammatory reaction. In order to determine more exactly the phenomena that appear under these conditions, it may be useful to consider first, not the fluid substances but the solid elements that are introduced into the tissues and cavities.

Let us begin with the lower animals in which the anatomical organisation and all the functions are of a much more simple character than they are in the Vertebrata. In my Comparative Pathology of Inflammation (Lecture IV) I have directed some attention to the digestion of the Sponges.

The nutritive substances—small organisms—whether they may have entered by the small openings, so numerous on the surface of Sponges, or have been introduced through a rent in the body wall, undergo the same fate. They are seized by vibratile or amoeboid cells which ingest the food and digest it by an intracellular digestion. These two kinds of cells, which come under the category of Phagocytes, have a great resemblance to one another, and we may say that digestion and resorption are two very closely related phenomena.

When we examine somewhat higher Invertebrata, such as the Medusae or certain other Coelenterates, we can still trace a close analogy between the true digestion of the food that goes on within the epithelial cells of the entoderm and the resorption of certain foreign bodies which make their way by an extra-buccal channel into the intermediary tissue. Here these bodies are surrounded by amoeboid cells which fulfil their function as phagocytes by ingesting and digesting the substances that have come from outside.

It is, here, unnecessary to go over the whole gamut of the perfecting of the organisation of the Invertebrata, in its relation to the resorption of foreign bodies, especially as it has already been treated in my Lectures on Inflammation. Let us choose merely some of the more common and better-known representatives of the Invertebrata and dwell for a few moments on the phenomena manifested in their organism, into the midst of which have been introduced a few nucleated red blood corpuscles^[97].

If a small drop of defibrinated blood from a goose be injected beneath the skin of a snail and another under the skin of a cockchafer larva, the red corpuscles are disseminated in the blood fluid which, of itself, is incapable of modifying them, but at the end of a few hours the leucocytes of the two invertebrates that we have chosen for the experiment will have ingested a certain number of the injected red blood corpuscles. The next day red blood corpuscles are still to be found intact in the blood plasma, but the great majority have been devoured by the leucocytes (Fig. 13). Inside these cells the red corpuscles undergo constant and marked changes. In the snail they become round and their walls permeable. In the vacuoles that are produced around the ingested red corpuscles dissolved haemoglobin is found (Fig. 14); a portion of this colouring matter passes into the nucleus of the red corpuscles, so that it also has undergone a profound change (Fig. 14). Many of the nuclei become emptied, only the peripheral layer remaining. This layer and the membrane of the red corpuscle are the parts that resist the action of the leucocytes longest and they are found for some time after their ingestion. The white corpuscles of the snail, having devoured one or more red corpuscles, may themselves become the prey of their fellows.

In the “ver blanc” (French popular name for the larva of the cockchafer) the phenomena of resorption of the red corpuscles of the goose resemble those just described. The blood plasma leaves intact the red corpuscles which undergo no change until they have been ingested by the leucocytes. The haemoglobin diffuses into the leucocyte, whilst the nucleus and the membrane persist for a very considerable period (Fig. 15), though they lose their normal aspect, shrivel, and become transformed into an irregular mass of brown pigment which may remain in the substance of the leucocyte (Fig. 15, p) for weeks.

Having once injected goose’s blood into snails and “vers blancs,” if we repeat the injection several times, the phenomena observed are invariably the same. The red corpuscles are unacted upon by the plasma and undergo the same changes within the leucocytes. These changes are in fact comparable to those described in the preceding chapter in discussing the intracellular digestion of the red corpuscles by the intestinal cells of the Planarians. In both cases the red corpuscles are seized by amoeboid cells and subjected to the influence of their contents. In the intestinal phagocytes of the Planarian, as in the phagocytes of the blood (leucocytes) of the snail and “ver blanc,” the haemoglobin diffuses through the wall of the red corpuscle, whose most resistant parts are the nucleus and the membrane. These resistant residual fragments, impregnated with haemoglobin, become brown in the Planarian, in the “ver blanc,” and also, but in a less degree, in the snail. The most appreciable difference consists in the formation of excretory vacuoles, containing concretions, in the Planarian, and the absence of these vacuoles in the blood phagocytes of the other Invertebrata. We have, however, less right to attribute a fundamental importance to this difference, in that the phenomena in the Actinians, which ingest the red blood corpuscles by the amoeboid cells of their entoderm, are in all respects (with the exception of the presence of these special excretory vacuoles) comparable to the phenomena observed in the Planarians. From the fact that in these two examples we have to do with a true intracellular digestion, it must be admitted that the modifications of the red blood corpuscles within the phagocytes of the blood in the snail and in the larva of the cockchafer, must also be placed in the same category of phenomena.

In order to make a more thorough study of this intracellular digestion in the phagocytes of the blood, we must direct our attention to larger and more highly organised animals than the snail and the “ver blanc.” Let us take, first, an example among the inferior cold-blooded Vertebrata. The red blood corpuscles of a few drops (0·25 c.c.) of the blood of a guinea-pig injected into the peritoneal cavity of a gold-fish (Cyprinus auratus) are not appreciably changed by the peritoneal fluid itself; but the numerous leucocytes that are found in the peritoneal fluid seize them and ingest them, just as do the phagocytes of the blood of Invertebrata, or the intestinal phagocytes in the Planarians and Actinians in the case of the red blood corpuscles of the goose. Each leucocyte of the Cyprinus ingests several red blood corpuscles and subjects them to intracellular digestion. The stroma of the red corpuscles becomes permeable; the haemoglobin diffuses into the nutritive vacuoles and at the end of a shorter or longer period the whole is dissolved and decolorised (Fig. 16). Here no brown pigment is produced and the red corpuscles are completely digested, leaving no “remains”; in this respect differing from the process in the Invertebrata mentioned.

This result depends, probably, partly upon the more feeble resistance offered by the non-nucleated red corpuscles of Mammals, and partly upon the more active digestive power of the leucocytes of Fishes.

As the result of several injections of guinea-pig’s blood into the peritoneal cavity of Cyprinus, the peritoneal fluid acquires new properties^[98]. If, a fortnight after the first injection, a little of the peritoneal exudation in the gold-fish be withdrawn, it is found that a drop of the serum which floats on the surface produces, almost immediately, well-marked agglutination of the red corpuscles of the guinea-pig, this being soon followed by the rapid solution of these red blood corpuscles in the fluid. This new property, which does not exist in the untreated fish, also makes its appearance in the blood serum of Cyprini treated with guinea-pig’s blood. The experiment is very successful at a temperature of 18°–19° C.

As the solution or lysis of the red blood corpuscles in the serum is exactly like that which takes place within the leucocytes of Cyprinus, we are justified in assuming that, in both cases, it is produced by the same substance. And, since the solvent or haemolytic power of the serum is only acquired as the result of the intracellular digestion of the red blood corpuscles by the leucocytes, it is probable that the solvent substance represents the intracellular ferment derived from the leucocytes.

The subject we have just broached is of fundamental importance in connection with the study of resorption and of the phenomena of immunity dependent upon it. It is necessary, therefore, that we should go more fully into its analysis. With this object we must first review the processes that go on during resorption in the higher animals and continue our examination of the changes that injected or extravasated blood undergoes in various positions of the organism.

This study is rendered comparatively easy for us by the numerous researches that have been carried out by pathological anatomists for the purpose of ascertaining the fate of effusions or extravasations of blood so frequently met with in disease. It has long been known that in subcutaneous, cerebral and other haemorrhages, or in hepatised lungs, there are found in the escaped blood a great number of cells containing red corpuscles. As was mentioned in the preceding chapter, these cells were evidently amoeboid cells that had ingested red blood corpuscles. To Langhans^[99] especially we owe a detailed study of the phenomena that follow extravasation of blood produced artificially in the subcutaneous tissue of the pigeon, rabbit and guinea-pig. In all these animals the haemorrhage is early followed by exudative inflammation, during which the leucocytes come up in great numbers and ingest the red blood corpuscles which are modified in the interior of the leucocytes. There is a formation or deposition of pigment and finally all traces of the red corpuscles disappear. In Mammals the pigment is brown or brownish, just as it is in the Planarians and in the “ver blanc”; in the pigeon it is green and resembles that found in the Actinians. In short there is a great analogy between the resorption of red corpuscles and the true intracellular digestion of the red blood corpuscles that goes on in the intestinal cells of the Invertebrata.

But what is the nature of these amoeboid elements that intervene in the resorption of the extravasated blood? At the period when Langhans carried out his investigation, we were unable to differentiate the cells at all satisfactorily. It is only since the publication of Ehrlich’s classic researches on the white corpuscles that we have been able to bring more order into this question. Thanks to the use of various aniline stains, Ehrlich was able to arrange the leucocytes found in the Vertebrata into several definite groups.

The question has already been touched upon in our eighth lecture on inflammation; it is therefore unnecessary to treat it here at length. We must, however, before entering on the analysis of the essential phenomena in the resorption of cells, as we now understand them, give a rapid survey of the different varieties of amoeboid cells that are found in the Vertebrata.

Beside mobile amoeboid cells, represented by several forms of white corpuscles, we must distinguish fixed amoeboid cells. These are permanently fixed in certain situations in the body; this, however, in no way prevents them from throwing out amoeboid processes in various directions and seizing foreign bodies or certain elements of the same organism. The nerve cells, the large cells of the splenic pulp and of the lymphatic glands, certain endothelial cells, the cells of the neuroglia, and perhaps some connective tissue cells, belong to the category of fixed amoeboid cells. All these elements, under certain conditions, are able to ingest solid bodies; consequently, they act as phagocytes. With the exception of the cells of the nerve centres, all these fixed phagocytes are of mesoblastic origin. It has been much discussed whether certain processes of the nerve cells may not really serve to seize foreign bodies and carry them into the cell contents. It appears to us that sometimes they undoubtedly do fulfil this function. For example, it is only by means of such amoeboid movements that leprosy bacilli can be introduced into the interior of ganglion cells and cells of the spinal cord^[100]. We must not dwell on this question, as the phagocytic property of the nerve elements plays no part in the resorption of cells. On the other hand, the neuroglia cells contribute largely to this process and their phagocytic function is now admitted by many observers^[101].

For long the large “dust” cells of the respiratory channels were looked upon as being epithelial cells which were capable of ingesting carbon particles, micro-organisms and other foreign bodies. The researches of N. Tchistovitch, carried out in my laboratory more than twelve years ago, made it evident that these elements are nothing more than white corpuscles that have immigrated into the alveoli and bronchi.

It is probable that the same is the case as regards the stellate cells of the liver, known as Kupffer’s cells. First described by Kupffer as cells of a nervous type, having long processes, they were later recognised by several observers as belonging to the endothelial tissue of the blood vessels of the liver. Kupffer^[102] himself has accepted this view and in his recently published monograph on these stellate cells, he describes them as endothelial cells that have retained their independence. Some researches on the resorption of blood, of which I shall speak shortly, have led me to think that these cells are nothing but white corpuscles that have been arrested in the hepatic capillaries. I have asked Mesnil, head of my laboratory, to study this question for me. His investigation is not yet concluded, but the demonstration already made that the livers of guinea-pig embryos and new-born rabbits do not possess any Kupffer’s cells is an argument in favour of my hypothesis.

Certain white corpuscles have undoubtedly been often mistaken for epithelial or connective tissue cells. We must not conclude from this, however, that these elements are never capable of sending out amoeboid processes and of ingesting foreign bodies. It would, however, be useful to collect new and incontestable proofs of the accuracy of this thesis. In spite of this uncertainty, it may be accepted as fully demonstrated, that certain fixed amoeboid cells, such as the large elements of the splenic pulp, of the lymphatic glands, and of the omentum, play an important part in the resorption of cells. It is there that elements filled with red corpuscles and white corpuscles in process of being destroyed are so often found.

Just as certain fixed cells do not function as true phagocytes, so also in some leucocytes this function is undoubtedly absent. The suggestion has been made several times that any cell element, provided it be young, is capable of ingesting foreign bodies. The examination of white corpuscles proves exactly the contrary. The smaller white corpuscles found in fairly large numbers in the blood and the lymph, and which are commonly known as lymphocytes or small lymphocytes, are simply leucocytes with very little protoplasm which in this state never fulfil phagocytic functions. It is only when it becomes older, when its nucleus, single and rich in chromatin, becomes surrounded by an ample layer of protoplasm, that the lymphocyte becomes capable of ingesting and resorbing foreign bodies. Several authors, with Ehrlich at their head, still assign to these larger cells the same name—lymphocytes. Others, however, give them the name of large mononuclear cells. Confusion is thus possible, especially as Ehrlich includes under the same term the large mononucleated leucocyte, a very rare form of cell in human blood, which is distinguished by the greater staining capacity of its nucleus. To avoid this inconvenience I propose to designate the large lymphocytes by the name of blood macrophages and lymph macrophages (haemomacrophages, lymphomacrophages). This term is preferable to that of mononuclear leucocytes, especially as in exudations we frequently meet with macrophages with two and even several sharply separated nuclei. Giant cells, moreover, are nothing but polynucleated macrophages. On the other hand, the leucocytes so often designated by the name of polynuclear in reality contain but a single nucleus. Even Ehrlich, who introduced this term, acknowledged its imperfection but he retained it for some time because it was already very extensively used and could, he thought, give rise to no misunderstanding. In his excellent work on anaemia, published jointly with Lazarus^[103], he now agrees that the name of “cells with polymorphous nuclei” would be more exact.

These polymorpho-nuclear leucocytes are very numerous in the blood and in many exudations and are distinguished by the greater selective affinity of their nucleus for basic aniline dyes and by a certain tendency of the protoplasm to become stained by acid aniline colours, such as eosin. The true macrophages are without granulations, but the “polymorpho-nuclear” leucocytes contain many. These granulations are sometimes “eosinophile,” “pseudo-eosinophile” (or “amphophile”) or even “neutrophile” (as in man and the horse).

These two main groups of leucocytes are generally distributed in the Vertebrata; and we already meet with them in one of the lowest vertebrate forms—the Ammocoetes (the larva of the lamprey). The macrophages of this fish present all the principal characters of the group to which they belong (protoplasm without granules, easily stained with methylene blue, large nucleus rich in nuclear juice). In the “polynuclear” forms in this lower vertebrate the protoplasm does not stain with methylene blue, but assumes a faint rosy tint with eosin; the single nucleus is divided into several lobes. In Vertebrates which are much higher in the scale these characters change. Thus in the cayman (Alligator mississipiensis), according to the researches of Madame Podwyssotsky, carried out in my laboratory, the two great varieties of leucocytes are readily found in the blood, lymph and exudations. The macrophages, however, especially in the exudations, are very often furnished with two or several nuclei, whilst the small leucocytes possess only a single nucleus, which is not divided into lobes. In spite of this peculiarity the two groups are readily distinguished. The staining reactions of the macrophages are identical with those of the corresponding corpuscles in all the other Vertebrata; whilst the small leucocytes, in spite of the absence of a polymorphous nucleus, are easily recognised by their eosinophile granulations and by the special affinity of the nucleus for basic aniline dyes. Under these circumstances it would be quite inappropriate to designate those leucocytes, which are really polynuclear, that is to say, possessing two or several nuclei, by the name of “mononuclear,” and to reserve the name of “polynuclear” for the small corpuscles which possess only a single nucleus undivided into lobes. For this reason it is much more rational to retain for these so-called polynuclear cells my proposed name of microphages. Moreover, the microphages are true phagocytes. It was formerly thought that the eosinophile leucocytes, such as the “‘overfed’ cells (Mastzellen)” of Ehrlich, which are identical with the clasmatocytes of Ranvier, never ingested foreign bodies. But, (especially after the researches of Mesnil^[104]), we have been compelled to change our opinion on this point. The true eosinophile cells are able to devour foreign bodies, especially micro-organisms, and must therefore be regarded as phagocytes belonging to the group of microphages.

It is the peculiar merit of Ehrlich and of his school that they have thoroughly established the fact that, in Mammals at any rate, the two principal groups of white cells are distinguished, amongst other characters, by the diversity of their origin. The lymphocytes and the mononuclear cells are developed in the spleen and lymphatic glands, whilst the “polynuclear” cells arise from the granular mononucleated myelocytes of the bone marrow. This is now generally accepted as applicable in the great majority of cases. In Ammocoetes, however, the two chief varieties of leucocytes arise from one and the same organ, regarded by several observers as a kind of primitive spleen, which runs along and in part surrounds the intestine. Mesnil has been good enough to make sections of this primitive organ in which it may be demonstrated that the macrophages and the microphages in the larva of the lamprey have the same seat of origin. Frog tadpoles and Cartilaginous Fishes also possess microphages which do not arise from the bone marrow, since in them this tissue is completely absent. But even in Mammals, at least in certain pathological conditions, Dominici^[105], in a research executed with much care and a perfect technique, has demonstrated the myelogenous transformation going on in the spleen. Thus in the adult rabbit affected with septicaemia by the typhoid bacillus, he found in the spleen developmental centres of amoeboid elements which, normally, appear to develop in the bone marrow only, i.e. the megacaryocytes, or large cells with budding nuclei, the neutrophile myelocytes (amphophiles), basophiles and eosinophiles.

The mesoblastic phagocytes of the Vertebrata are divided, then, into fixed phagocytes—the macrophages of the spleen, endothelia, connective tissue, neuroglia, and muscle fibres—and free phagocytes. These latter are sometimes haemo- or lympho-macrophages, sometimes microphages. The fixed macrophages and the free macrophages resemble one another so greatly that it is very often extremely difficult, if not impossible, to differentiate them. For this reason it is often very useful, when the exact origin of a large phagocyte is not known, simply to name it “macrophage.”

The two principal groups of phagocytes—(1) fixed and free macrophages, (2) microphages—are distinguished not only by their morphological characters; they also give evidence of very marked physiological differences. All phagocytes are endowed with amoeboid movement which allows them either to move about freely or merely to put out protoplasmic processes. These movements are regulated by a very great sensitiveness, often different in the two groups. Besides a tactile sense, the phagocytes possess a kind of sense of taste or chemiotaxis which enables them to distinguish the chemical composition of the substances with which they come in contact. The existence of this chemiotaxis could be anticipated from the moment that an important part in the life of the organism began to be ascribed to the amoeboid cells. Leber^[106], Massart and Charles Bordet^[107] have, however, demonstrated it by rigorous experiment. Following the method used by Pfeffer to demonstrate the chemiotaxis of the vegetable spermatozoids and of Bacteria, these investigators introduced into the bodies of higher (rabbits and guinea-pigs) and lower (frogs) Vertebrates small glass tubes filled with different solutions (peptone, broth, salts, bacterial products, etc.). The leucocytes, guided by their positive chemiotaxis, made their way into the tubes and there formed plugs which were often very voluminous; when, on the other hand, the chemical composition of the solutions excited their negative chemiotaxis, the leucocytes avoided the tubes.

Having acquired information as to the chief characters of the leucocytes, we may ask, To which group do those amoeboid cells, which, according to the observations of Langhans and many other investigators, bring about the resorption of the red corpuscles of the blood, belong? This resorption goes on more rapidly and is observed much better if, instead of introducing blood of the same species into any part, we inject defibrinated blood, or red blood corpuscles from which the serum has been removed by washing, from another species of Vertebrate. It will be found best to inject the nucleated red corpuscles of lower Vertebrates into Mammals, or (as already described above) to introduce the non-nucleated red blood corpuscles of Mammals into lower Vertebrates. In all these cases the injection of such blood or corpuscles sets up an aseptic inflammation which attracts a large number of free phagocytes to the seat of injection. In subcutaneous, peritoneal or intraocular exudations produced under these conditions, we find, in addition to a number of microphages, many macrophages. Whilst the former ingest the injected red corpuscles merely in isolated cases, the positive chemiotaxis of the macrophages manifests itself much more actively. In the resorption of the red blood corpuscles the more important part is played by the macrophage. To get a clear idea of the phenomena that accompany this resorption, let us take a concrete example. Inject defibrinated goose’s blood into the peritoneal cavity of guinea-pigs^[108]. During the first few hours after injection the oval nucleated red corpuscles are found intact in the fluid of the peritoneal lymph. The plasma, by itself, exercises no destructive or solvent action on the red corpuscles of the goose.

Immediately after the injection the lymph of the peritoneal cavity begins to show important changes. The white corpuscles which, in the normal condition, are fairly abundant, disappear almost completely; some small lymphocytes presenting their ordinary aspect may indeed be found, but the few macrophages and the microphages that remain show signs of very grave lesions. They lose their mobility, run together into clumps and become incapable of ingesting foreign bodies. At this moment the phagocytes undergo a critical change which we have designated by the name of phagolysis. This condition lasts for about an hour, sometimes it continues longer, according to case and circumstance, but after this the peritoneal fluid becomes filled with leucocytes that have newly come on to the scene. These cells make their way, by diapedesis, through the walls of the congested vessels of the peritoneum. A true aseptic inflammation is produced which induces an exudation of a large number of white corpuscles, amongst which are found microphages and still more numerous macrophages. The latter show a very pronounced positive chemiotaxis towards the injected red corpuscles of the goose. Soon after their appearance, that is to say two or three hours after the injection of the blood, the macrophages send out very small protoplasmic processes and affix them to the surface of the red corpuscles. There follows an aggregation of the macrophages of the guinea-pig with the red corpuscles of the goose and characteristic masses, in which can be recognised both kinds of cells, are produced. This union with the very small pseudopodia is the first stage in the ingestion of the red corpuscles by the macrophages (Fig. 17). The red corpuscle, seized by amoeboid processes, passes into the interior of the macrophage. This macrophage seldom rests contented with ingesting a single red corpuscle. Usually it devours a large number and sometimes enormous macrophages may be seen filled with a score of red corpuscles.

If the quantity of goose’s blood injected into a guinea-pig is large (5–7 c.c.), the ingestion of red corpuscles by the macrophages continues for a considerable period—often for three to four days. During the whole of this time a certain number of the red corpuscles remain free in the peritoneal plasma, but, in spite of this prolonged stay, none of them undergo extracellular solution.

The red blood corpuscles, anchored by the amoeboid processes of the macrophages, at first present a normal appearance. Later their membrane begins to wrinkle, but as soon as they have passed within the phagocytes the wrinkles disappear and the corpuscles regain their normal aspect. If a little neutral red solution be added to a drop of peritoneal exudation (Fig. 18) we observe that the nucleus of the ingested red corpuscle and even its contents are stained red, whilst the red corpuscles adherent to the surface of the phagocytes retain their normal yellow colour. This reaction enables us to see that the red corpuscles are seized by the macrophages whilst still in their normal condition, but that they undergo a change immediately after they have been ingested. Little by little the devoured corpuscles are digested within the phagocytes. The haemoglobin diffuses into the contents of the macrophage through the stroma, which has become permeable; the nucleus of the ingested red corpuscle also becomes stained by the haemoglobin. Part of this colouring matter is excreted by the phagocyte. The body of the red corpuscle is pretty soon digested, but the nucleus, impregnated with haemoglobin, persists for a much longer period. It divides into several fragments, recognisable by their yellow colour, and in certain cases these remnants of red corpuscles may be met with for weeks in the interior of the macrophages. These macrophages do not remain permanently in the peritoneal fluid. Some (3–4) days after injection the lymph of the peritoneum contains only leucocytes that have newly come up and which contain neither red corpuscles nor their remains. We must open the guinea-pig to find any macrophages that have devoured red corpuscles. They are to be met with in large numbers in the glandular portion of the omentum, in the mesenteric glands, in the liver and in the spleen. They are fairly easily recognised by the characteristic aspect of the débris of the red blood corpuscles. Having devoured the red corpuscles the macrophages leave the peritoneal fluid and the digestion is completed in the positions just mentioned. In the liver they are seen as large mononuclear cells often with highly developed processes. In this condition they remind one of Kupffer’s stellate cells—a fact that suggested to me the idea that these elements are nothing but white corpuscles which have immigrated into the vessels of the liver.

Following up the fate of the macrophages that have resorbed the red blood corpuscles, we find them in the large hepatic vessels, in the vena cava and even in the blood of the heart. But in these latter situations they contain merely a few scarcely recognisable traces of their prey. These phagocytes, which left the blood during the inflammation that followed the injection of red corpuscles of the goose, re-enter it, having fulfilled their function, during the final period of the resorption. This resorption must undoubtedly be regarded as an intracellular digestion. When we compare the essential phenomena taking place inside the macrophages containing red blood corpuscles with those we have described in the intestinal phagocytes of the Planarians or Actinians after a meal, the analogy between the two becomes very apparent. In both cases the red blood corpuscles undergo a marked change which results in a diffusion of the haemoglobin. The membrane and nucleus of the red blood corpuscles persist longer but they also are ultimately digested. The excretion of haemoglobin from the phagocytes, just mentioned in the case of the macrophages of the guinea-pig, is also observed in the Actinians, whose coelenteric cavity is tinted by a rose-coloured solution.

We have seen that in the Actinians intracellular digestion takes place in a distinctly acid medium, whilst in the intestinal cells of the Planarians it takes place in one that is only weakly acid. The macrophages of the guinea-pig, during the resorption of red blood corpuscles of the goose, carry on the digestive process in a medium which shows a still weaker acidity. When made to ingest granules of blue litmus there is no change of colour. Nor does alizarin sulpho-acid give any reaction, probably owing to the fact that it exerts a toxic action on the protoplasm of the macrophages. If, however, we add to a drop of the peritoneal exudation of a guinea-pig, containing macrophages filled with red blood corpuscles of the goose, a little of Ehrlich’s 1% solution of neutral red, the red brick tint at once makes its appearance in the content of these phagocytes. This coloration is identical with that described in the Amoebae which digest Bacteria or in the intestinal phagocytes of the Planarians. It may, then, be regarded as an indication of weak acidity. This coloration is maintained for some hours, after which it gives place to complete decoloration, a phenomenon that must be attributed, as in many other cases, to the neutralisation of the acid by the alkaline protoplasm that has been macerated in the fluid after the death of the macrophages.

The example we have chosen—the destruction of red blood corpuscles of the goose by the macrophages of the guinea-pig—may serve as a prototype of the resorption of formed elements in general. If, instead of red blood corpuscles of the goose, we inject into the guinea-pig’s peritoneal cavity pigeon’s or fowl’s blood, the essential phenomena will be the same. The red blood corpuscles will always induce positive chemiotaxis, especially of the macrophages, which in turn will ingest the nucleated red corpuscles. It may be that in certain cases, when fowl’s blood containing red corpuscles that are not very resistant is injected, a certain number of the corpuscles immediately undergo a partial solution in the peritoneal fluid^[109]. Here also the stromas and the nuclei of all the red blood corpuscles, as well as many of the corpuscles unacted upon by the plasma of the phagolysed exudation, undergo digestion inside the macrophages.

When, instead of blood, we inject white corpuscles from the bone marrow, spleen or lymphatic glands of animals into the peritoneal cavity, we may still observe their final disappearance in the macrophages. The spermatozoa of man or of various mammals (bull, rabbit, guinea-pig, etc.), when injected into the peritoneal cavity of the guinea-pig or rabbit, are well adapted for this line of investigation. Here again the immediate result of injection is the very marked phagolysis of the leucocytes. This phenomenon gives place to an exudative inflammation which brings into the peritoneal cavity a number of phagocytes. These, especially the macrophages and in a much smaller degree the microphages, devour the spermatozoa which in no case are dissolved, even partially, in the plasma of the exudation. The macrophage seizes the spermatozoa which sometimes, by the active movements of their flagella, exhibit great vitality. At the end of several hours all the spermatozoa are found inside phagocytes where they are completely destroyed. The flagellum is digested first, but the head and medial portion soon suffer the same fate. Neutral red reveals the feebly acid reaction, perhaps with even more distinctness than in the case of the red blood corpuscles.

The résumé of Langhans’ investigation given in this chapter would lead us to expect that resorption in the subcutaneous tissue will follow the same rules as that going on in the peritoneal cavity. As a matter of fact, blood injected at this position sets up a diapedesis of phagocytes which ingest the red blood corpuscles. In some cases only is there a partial solution of these corpuscles in the fluid of the subcutaneous exudation. It is for this reason that goose’s blood, injected under the skin of a guinea-pig, gives rise to a fluid exudation coloured a bright rose red by the dissolved haemoglobin. This haemoglobin is derived from red blood corpuscles which are damaged by the goose’s blood serum that was added to the plasma of the exudation. The stroma and nuclei of the red blood corpuscles cannot, however, be dissolved in this fluid. They undergo the same fate as the red corpuscles that have remained intact, that is to say they are ingested by the macrophages which immigrate into the subcutaneous tissue and which finally digest all these elements. The cells, less fragile than certain red corpuscles, are, in the subcutaneous tissue, as in the peritoneal cavity, destroyed solely in the interior of the phagocytes.