The stomach, though capable, through its acid, of preventing the multiplication of certain micro-organisms, protects, very feebly, the rest of the digestive apparatus. As soon as, in the duodenum, the acidity is weakened or neutralised, the various micro-organisms commence to multiply and soon develop very abundantly.

In the animal series the intestine proper presents a very great variability, and even, in closely allied species, exhibits considerable differences. From the particular point of view which interests us these differences are very marked. Alongside insects, such as the silkworm, the larvae of cockchafers and others, whose intestinal canal contains a very rich bacterial vegetation, we have others which contain exceedingly few micro-organisms or, indeed, none at all. This last condition is represented by the caterpillars of small Lepidoptera, and notably by those of several species of clothes-moths. These differences correspond to the variety of the juices and digestive ferments met with in these Invertebrata. As the physiology of digestion in these animals is as yet little understood, it is at present impossible to define clearly the conditions which regulate these phenomena. In any case, it is very probable that the soluble digestive ferments destroy the micro-organisms and prevent them from growing in the intestinal content. Otherwise it is difficult to explain why the larvae of clothes-moths, which live in old dusty textile fabrics, where the germs of bacteria are not wanting, present a digestive canal from which micro-organisms are entirely absent. The digestive juices, adapted to digest wool and even wax, are evidently capable also of digesting the bodies of micro-organisms. In other insects, which feed on vegetables and on substances less difficult to digest, micro-organisms develop in the intestinal content, as in many of the higher animals. Insects often have their intestine lined by a very delicate chitinous membrane which offers no obstacle to the absorption of the products of digestion, but prevents the micro-organisms from reaching the epithelial layer. We have here a defensive apparatus against microbial invasion, which must be the more useful because this membrane is thrown off and renewed at each moult, thus enabling the insect to rid itself at one swoop of a large number of its microscopic inhabitants.

In the Vertebrata the canal of the pancreas and that of the small intestine are always populated by a greater or smaller number of micro-organisms, amongst which bacilli predominate. We know the great difficulty experienced every time we wish to make experiments on the pancreatic digestion outside the animal body. The digestive fluid, alkaline and containing many bacteria, is soon transformed into a microbial purée. We are then obliged to have recourse to antiseptics to arrest this development and to bring into prominence the digestive rôle played by the soluble ferments of the pancreas. This well-known fact may be used as an argument against the existence of any kind of bactericidal power in the small intestine of higher vertebrates. Even in those animals which are distinguished by the remarkable poorness of their intestinal flora, we fail to reveal the presence of bactericidal substances. The Crustacea, e.g. the crayfish, and certain worms, such as the Ascaris, contain few micro-organisms in their intestine. The former feed on putrescent substances, the latter inhabit the small intestine of man and animals, populated by myriads of bacteria. It might be supposed that, under these conditions, the intestinal content must contain a mass of micro-organisms or, if that be not the case, that it must contain some substance which is powerfully bactericidal. In reality, neither one nor the other of these suppositions receives any confirmation. The intestines of the two Invertebrata I have named are very poor in micro-organisms and their contents do not exhibit the slightest bactericidal power. When a little of their contents is placed in tubes and kept at a suitable temperature it is not long before it becomes filled by a great number of bacteria of various kinds.

To explain this poverty of the microbian flora of the intestines in these examples we must postulate some kind of mechanical purification, facilitated by the peristaltic movements of the digestive canal.

Even in animals which have an abundance of micro-organisms in the small intestine, there must be produced some phenomenon which brings about the disappearance of a certain number of them. In mammals the small intestine always contains far fewer micro-organisms than does the large intestine; in birds, the coecum is much richer in bacteria than is the rest of the digestive canal. Schütz^[699] has attempted to demonstrate the disinfecting power of the small intestine in the dog by feeding it on substances to which he had added a large number of Gamaleia’s vibrio (Vibrio metchnikovi). After convincing himself that micro-organisms perish in the digestive canal and are never found in the excrementa, Schütz introduced into his dogs a cannula, one branch of which passed into the pylorus, the other into the duodenum. By means of a small apparatus he could readily interrupt the communication between the stomach and the intestine. The vibrios, mixed with biscuit, and softened with water, introduced directly into the duodenum (whilst the stomach was kept completely isolated), penetrated into the large intestine in small numbers only. The lower part of the colon, the rectum and the excrements gave no cultures of vibrios and did not give rise to any growth except that of the Bacillus coli. In this case the disinfection of the intestine took place without any help from the gastric juice. Further, when Schütz killed dogs, after giving them food in which vibrios were mixed, these organisms were found in the intestine only. The gastric acidity, therefore, is not capable of killing these organisms, or of preventing them from passing into the small intestine, in which alone they were killed. It was only with the aid of purgatives, such as castor-oil or calomel, that Schütz succeeded in preserving the vibrios in the intestines and in finding them in the dejecta. This observer did not carry his investigations further and did not make out the mechanism by which the small intestine destroyed such large numbers of vibrios. He supposes that alongside a mechanical factor, such as the very active peristaltic movement, there exist others, perhaps chemical processes, capable of killing these micro-organisms.

This question of the defensive action in the small intestine is, consequently, far from being settled. The data collected indicate merely that the problem is a very complex one. It has been shown, however, that very virulent bacteria may pass through the digestive canal not only without injuring the animal but even meeting their own death in this organ. The anthrax bacillus, so fatal to mice and guinea-pigs, may be swallowed by these animals without the slightest danger to them. It may then be found in the small intestine, but not in the large intestine, this proving that the gastric acidity is incapable of destroying them outright. To produce generalised anthrax by way of the intestine, it was necessary that the animals should swallow the spores of anthrax along with spiny plants, as in the experiments of Pasteur and his collaborators^[700], or along with sand or powdered glass. In these cases the intestinal lesions served as the port of entry for the bacillus, the intact mucous membrane of the intestine preventing their penetration. Mitchell, in an unpublished work, undertaken in my laboratory, succeeded in giving fatal anthrax to guinea-pigs, even when he fed them with spores mixed with the “crumb” of bread soaked in milk. During the whole period of the experiment the animals took no food capable of producing lesions of the wall of the intestine. But examples of infection under these conditions are altogether exceptional. In the great majority of instances the animals were not attacked. The same rule applies also to many other micro-organisms, which can be ingested with impunity although their inoculation into the blood and tissues sets up infections which are inevitably fatal. Many animals may, without running the least risk, swallow large numbers of bacteria which in man produce grave intestinal disease. Thus, it has never been possible to produce typhoid fever regularly and with certainty in any of the species of animals to which masses of typhoid coccobacilli were given by ingestion. We may recall the difficulties which so many investigators have met with in inducing intestinal cholera in laboratory animals, which are so refractory to Koch’s vibrio. Only very young animals, especially unweaned rabbits, are capable of contracting fatal intestinal cholera, but these animals may contract it not only from the true cholera vibrio but also from Gamaleia’s vibrio. As soon as rabbits begin to feed on vegetables they acquire an immunity which is insuperable.

It is most assuredly not the digestive ferments of the intestine that protect the animal against infection through the intestine. The contents of every part of the small intestine of the Vertebrata permit an abundant development of all sorts of bacteria, and in solutions of trypsin not only do pathogenic and resistant micro-organisms grow luxuriantly, but also saprophytes and the most inoffensive bacteria. Weigert^[701] influenced by this fact even saw in it an objection to the theory that the destruction of micro-organisms in the animal, notably that which is effected by the phagocytes, is to be regarded as an act of digestion. It is a remarkable fact that whilst trypsin is so powerless against micro-organisms the intracellular ferments, and especially microcytase, whose kinship with the group of trypsins is undeniable, are able to bring about their digestion so completely.

It was thought that among the digestive fluids the bile more especially exerts a definite antiseptic power. It is undeniable that this fluid is not indifferent for certain bacteria. Talma affirms that it is bactericidal for several micro-organisms, especially the diphtheria bacillus. In many of his experiments, however, the bile proved to be incapable of killing micro-organisms introduced directly into the gall-bladder. According to the researches of Gilbert and Dominici^[702] the bile does not prevent the abundant development of micro-organisms capable of setting up diseases of the biliary passages, such as the Bacillus coli. I have tried to prevent the multiplication of the cholera vibrio by the addition of bile, but my results were entirely negative. If the bile in an undiluted state has such a slight action upon so many kinds of bacteria, it is evident that we cannot count upon its antiseptic action when it passes into the small intestine, where it is mixed with all sorts of other substances.

The digestive fluids of the small intestine, either those that are non-bactericidal, the pancreatic juice, or those that are not very active, the bile, are, nevertheless, capable of producing a marked influence on certain poisons, and amongst others on certain microbial toxins. According to the experiments of Nencki and of Mmes Sieber and Schoumow-Simanowski (l.c.), trypsin is much more antitoxic against the diphtheria poison than is pepsin. Thus, the pancreatic juice of both the rabbit and the guinea-pig destroys this toxin much more actively than does the gastric juice. The pancreatic juice of the dog exerts a very powerful action on the same toxin. A gramme of this fluid neutralises ten thousand lethal doses of the toxin. Wehrmann, also, found that trypsin inhibits the poisonous action of snake venom. Bile also exerts an action upon certain poisons. Mixed with diphtheria and tetanus toxins it prevents their pathogenic effect. It also neutralises the venom of snakes, as has been observed by Fraser^[703], Phisalix^[704] and Calmette^[705]. All the venoms, when placed in contact with fresh bile for 24 hours, induce no injurious effect when the mixture is injected into normal animals. Bile, heated to 100° C., and even to 120° C., is still, though more feebly, active. To obtain these results, however, it is indispensable to prepare, beforehand, a mixture of the two fluids. When injected separately, whether at the same time as, before, or after, the venom, the bile does not prevent poisoning. The venom when injected directly into the gall-bladder of the rabbit sets up fatal intoxication to the same degree as does the same dose of venom introduced subcutaneously. Calmette, who made this experiment, explains this negative result as due to the too rapid absorption of the venom, which has not had time to be affected by the destructive action of the bile.

A protective action of the bile has been determined with regard to two viruses, the micro-organisms producing which are not, as yet, known. Koch^[706] succeeded in vaccinating Bovidae with the bile of animals that had died from rinderpest, and Frantzius^[707] prevented animals from contracting rabies when he inoculated into them rabic virus mixed with the bile of rabbits that had succumbed to that disease. Vallée^[708] points out, however, that the bile of the normal rabbit produces exactly the same effect. Here, then, we have to do with a preventive action of the bile, as such, against the rabic virus. In the present state of our knowledge it is impossible to say whether this influence of the bile is directed against the toxin or against the unknown micro-organism. Analogy would lead us to accept the former of these two suppositions.

The bile, active against certain poisons, does not, however, prevent poisoning by cholera toxin nor by that of botulism, two most typical intestinal intoxications.

Since diastases and the digestive juices are incapable of affecting micro-organisms and since certain of these latter perish in the intestines we must seek some other cause for their destruction. It is probable that the vital competition among the micro-organisms, whose rôle could be foreseen in the buccal cavity, is of still greater importance in relation to the phenomena of pathogenic action or of the innocuousness of infective bacteria in the intestinal canal^[709]. This complex and difficult chapter, up to the present, has been studied in a very imperfect fashion. In our observations on cholera we have remarked that under certain conditions the cholera vibrios do not develop on gelatine plates, except in the neighbourhood of certain adjuvant micro-organisms such as the Torulae and the Sarcinae. Guided by this fact we have succeeded in producing intestinal cholera in suckling rabbits, with races of vibrios which, when ingested alone by these animals, remain innocuous or set up the disease only occasionally. We have convinced ourselves of the helpful action of certain representatives of the gastro-intestinal flora upon true cholera^[710]. Following on these observations, it was quite natural to suppose that this flora might also contain micro-organisms capable of hindering the development and toxic action of the cholera vibrio. We have even advanced the hypothesis that these “hindering” micro-organisms in the flora of the digestive canal may explain the immunity of animals, of many human individuals, and even of the population of unattacked towns, to intestinal cholera. We should have, then, in the intestinal contents, inhabited by a number of micro-organisms and deprived of bactericidal juices, an important factor which in many cases guarantees a refractory condition. It must be stated, however, that prolonged studies, carried out with the object of demonstrating in suckling rabbits the precise part played by these micro-organisms which prevent cholera, have not given any satisfactory results. This we attribute to our very imperfect knowledge of the microbial population of the digestive organs.

If the destruction by representatives of the normal intestinal flora of the micro-organisms which penetrate into the intestines has not as yet been satisfactorily demonstrated, the power of these latter to destroy microbial toxins cannot be doubted. We^[711] have shown that a great number of micro-organisms develop well in broth cultures of the tetanus bacillus which contain a quantity of specific toxin. This toxin is destroyed under the influence of this microbial vegetation, but the production of antitoxin never results. Charrin and Mangin^[712] have observed similar facts.

As the destruction of bacterial toxins by micro-organisms takes place with great constancy and rapidity, it is quite natural to suppose that the same phenomenon occurs also in the intestinal canal of living animals in which pathogenic micro-organisms have succeeded in secreting their toxic products.

The liver having long been recognised as the purifying organ of the products resulting from digestion, it has been asked if it might not also play a part in the destruction of microbial poisons. Certain facts point to its inhibiting influence on the action of nicotine, atropin, and of certain other alkaloids, and we have other facts which demonstrate the power of the liver to transform into urea the ammoniacal substances arising from the activity of the digestive glands. When Nencki, Pawloff, and their collaborators^[713] succeeded in making the portal vein communicate with the vena cava, and thus were able to suppress the purifying function of the liver, they found that their dogs became poisoned in consequence of the accumulation of ammonia in the animal organism.

Guided by these data as to the protective rôle played by the liver an attempt was made to apply them to the action of this organ on bacterial toxins such as the diphtheria poison. The numerous attempts undertaken in this direction have given negative results: the liver was not found to be capable of destroying this toxin. Bouchard, Charrin and Ruffer have studied the action of the liver on the pyocyanic toxin. They thought that they could make out a certain antitoxic action of this organ, but, later, Charrin^[714] convinced himself that the bacterial secretions are only “moderately modified” under these conditions, and that it is more especially the parts soluble in alcohol which undergo modification in the liver. Now, the true bacterial toxins, as is well known, are distinguished by their insolubility in alcohol. Moreover in the numerous experiments made by Roux and Vaillard and so many other observers on the tetanus and diphtheria toxins there has never been any evidence of any kind of antitoxic action of the liver.

The digestive organs are furnished throughout with a defensive apparatus against micro-organisms; this consists in an accumulation of lymphoid tissue in the form of patches or groups of solitary glands:—the tonsils, Peyer’s patches, and the solitary glands of the intestine. These organs produce a large number of phagocytes which are able to come into close contact with the micro-organisms. Ribbert^[715] and Bizzozero^[716] have, independently or almost simultaneously, described glandular masses in the coecum of the rabbit in which they recognise the presence of many micro-organisms derived from the intestinal content. They noted that the greater number of these bacteria were within cells, and regarded this as an example of phagocytic reaction. Manfredi^[717] was able to confirm this interpretation by the demonstration that the ingested micro-organisms were dead. Later, Ruffer^[718] studied this question in my laboratory. He observed intestinal phagocytosis in Peyer’s patches in several species of animals, and showed that the lymphoid tissue contained large macrophages filled with bacteria and microphages in process of intracellular digestion. Amongst these latter he recognised leucocytes, which in turn contained micro-organisms. The accumulation of phagocytes in the lymphoid organs of the digestive canal constitutes, so to speak, the last act of a struggle which is spread over a very wide field.

Some years ago Stöhr demonstrated^[719] that the wall of the intestine, and especially the tonsils and other lymphoid organs, are traversed by an enormous number of leucocytes which execute a kind of migration towards the cavities containing micro-organisms. This continual and normal condition is often termed Stöhr’s phenomenon. It is evident that we have here a process of phagocytic defence in which the leucocytes, disseminated through the digestive canal, give chase to the micro-organisms that are nearest to the living portions of this organ. When we remove a particle of mucus from the surface of the tonsils of a person in good health we always find that it contains leucocytes, especially microphages, filled with micro-organisms of all kinds.

The protection of the digestive mucous membrane is a more complicated process than that of other mucous membranes, and many of the points concerned therein are still obscure and need to be elucidated by further research. It might be thought that the phenomena, associated with the defence of the mucous membrane of the genital organs, being much more simple and yet of similar nature, should be much more easily made out, and that these would throw light on several aspects of the problem of the general defence of the animal. Obstetricians and gynaecologists have certainly given much attention to this question as regards the female genital organs, but we are still far from possessing a satisfactory knowledge of this subject. There already exists quite a literature on the question, dominated by the work in two volumes published by Menge and Krönig^[720], but a satisfactory solution has still to be obtained.

At birth the vulva and the vagina are free from micro-organisms, but they soon become inhabited and a fairly abundant flora, in which may be recognised certain predominant species, such as the bacillus of Doederlein, is developed. Micro-organisms, therefore, can exist in the vulva and the vagina, and yet, when we introduce into these organs cultures of various bacteria, saprophytic or pathogenic, they soon disappear. We have the phenomenon to which Menge has given the name of “autopurification” of the female genital organs. He himself, as well as his predecessors, Doederlein and Stroganoff, tried to make out the mechanism of this purification. In the new-born female child the phenomenon is less complicated than in the adult. According to Menge the acidity of the vaginal secretion in these infants at first prevents the development of a large number of bacteria. Associated with this factor is a marked emigration of leucocytes, which destroy the bacteria by an act of phagocytosis, or perhaps by their products that have escaped into the vaginal mucus. As a third element to which much importance is attributed, we must accept the intervention of acidophile bacteria which grow well in acid secretions but which hinder the development of other micro-organisms. Doederlein concludes that it is more especially to the bacillus which bears his name that the vagina owes its protection against infective germs. Menge, however, attributes this action to a whole series of bacteria.

After introducing a quantity of the Staphylococcus pyogenes into the vagina of new-born females, Menge found that they grew for a certain length of time. Their presence excited a great accumulation of leucocytes in the vaginal mucus, this being followed by a very marked ingestion of the micro-organisms, but it was only from the moment when the vagina became peopled with the bacteria which constitute its normal flora that the staphylococci began to disappear. This process of autopurification only ceased three days after the introduction of these bacteria. Menge asked himself whether some purely mechanical element did not contribute to rid the vagina of the micro-organisms which had entered it. To settle this point he introduced into this cavity grains of vermilion, and as these latter remained there for a longer period than did the micro-organisms, he concluded that the vagina was incapable of purifying itself by mechanical means. We must, however, in these experiments take into account the fact that the micro-organisms which Menge introduced into the vagina excited considerable reaction, accompanied by a marked leucocytosis. Under these conditions there should be produced a greater quantity of the mucous secretions which could much more readily carry off with them the micro-organisms that had come into the vagina than the smaller quantity could deal with the vermilion. It is very probable, therefore, that, just as in the case of the other mucous membranes, that of the female genital organs is capable of mechanically expelling fine particles, and especially micro-organisms.

With the object of throwing further light on the problem of the autopurification of the vagina, Cahanescu^[721], working in my laboratory, undertook experiments on the females of several species of mammals. The mare, as producing the greatest amount of vaginal mucus, was selected by this observer as suitable for the settling of this question of the bactericidal power of this secretion. The result was absolutely negative, even when such an inoffensive saprophyte as the Coccobacillus prodigiosus was used. The autopurification of the vagina of the female dog, rabbit and guinea-pig, was found to be neither very marked nor very active. The micro-organisms introduced into the vagina usually remained there for some time. Of all the factors in the microbial destruction which Cahanescu was able to make out that of the accumulation of leucocytes was the most active. Sometimes he observed an extraordinary amount of phagocytosis, whilst in other experiments this was slight or even absent. Many of the leucocytes being killed in the vaginal mucus, it is possible that in some cases a certain bactericidal action of the cytases which have escaped from these dead leucocytes is set up. It is true that the vaginal secretion of the mare did not exhibit this antimicrobial property in vitro, but in the other animals experimented upon it was found impossible to make similar experiments, the quantity of mucus being too small. In woman the acidity of the surface of the mucous membrane of the vulva and of the vagina, so frequently present, may play a certain part in the protective action against those bacteria which cannot tolerate the acid medium, but the animals studied by Cahanescu, even female dogs, do not possess this advantage, their mucous membranes usually having an alkaline reaction.

In the urinary channels this acid reaction also plays a part, as one of the defensive agencies against the penetration of bacteria. This may also be effective in man and other animals that have an acid urine. In many other animals, however, where the urine is alkaline micro-organisms do not pass into the deeper parts of the urinary organ under normal conditions. Here it is to the outflow of the urine that the bladder owes its immunity against pathogenic micro-organisms and saprophytes. When we connect two flasks containing sterilised broth in such a way that the fluid flows slowly from one of them into the other, the former never becomes contaminated by the micro-organisms which are present in the latter, in which latter the broth is soon transformed into a purée of bacteria, whilst in the former the broth remains unaffected and aseptic. This purely mechanical factor has been well brought out by Preobrajensky^[722] as the result of work carried out in Duclaux’s laboratory. The sterility of the normal urinary bladder must be attributed to a similar cause. When the urine begins to stagnate in the bladder it very readily becomes contaminated.

Since the acceptance of the view that the suprarenal capsules serve to neutralise the effect of certain toxic substances elaborated in the body, there has been an inclination to assume that these organs might also fulfil an antitoxic rôle against microbial poisons. The hypothesis was advanced that this function might be shared by the suprarenal capsules with the thyroid gland and with certain other problematical organs. We have already stated (Chapter V) that the suprarenal capsules, in some experiments where spermotoxin was injected into rabbits, exhibited a certain antispermotoxic power. But, up to the present, no exact fact has been observed that would favour the idea of an antitoxic action of the above-mentioned organs against bacterial toxins. Roux and Vaillard^[723], in their great work on tetanus, have made experiments in this direction, but their results did not justify them in giving a positive answer to the question.

Nature does not make use of antiseptics to protect the skin and the mucous membrane. The fluids which moisten the surface of the mouth and of other mucous membranes are not microbicidal, or are so to a very slight degree, and then rather of an exceptional nature. Nature rids the mucous membranes and the skin of a large number of micro-organisms, eliminating them by epithelial desquamation, and expelling them along with fluid secretions and excretions. Nature, like the doctors of the present day who replace antisepsis of the mouth, intestine, and other organs by washing with pure physiological saline solution, has chosen this mechanical method. She avails herself of the help of inoffensive micro-organisms to prevent pathogenic micro-organisms from taking up their abode in these positions, and she is constantly sending to all the mucous membranes and the skin an army of mobile phagocytes which explore the ground and rid it of micro-organisms. When these begin to get more numerous the phagocytic reaction becomes more intense. A struggle takes place between the two living elements—phagocytes and micro-organisms. In those cases where the animal remains unaffected the former gain the upper hand.