This example furnishes us with a real link between primitive intracellular digestion and the perfected and derivative extracellular digestion. In the same group of Gasteropods may be followed out several stages of this evolution so that in the higher representatives of the group, such as the slugs and the snails, we meet with digestion carried on only by secreted juices in the gastro-intestinal contents. In these Mollusca a voluminous glandular organ, the liver, which is certainly derived from coecal appendices similar to those of Phyllirhoë, is now met with. Regarded from this point of view the liver is, as Claude Bernard has stated, an organ of second digestion. I think that a detailed study of the liver of the Mollusca, guided by this idea, will give results of considerable importance.

In the Vertebrata intracellular digestion in the gastro-intestinal canal almost disappears and is replaced by digestion carried on by means of ferments contained in secreted juices. We cannot, of course, offer to the reader anything like a complete account of this extracellular digestion in the higher animals. It is necessary, however, to draw attention to several aspects of this function which have been established, thanks to the progress made during recent years, in obtaining digestive juices and in the study of their action.

For the study of intracellular digestion the sea-anemone is the most suitable animal for our purpose; for that of extracellular digestion the dog. In this latter animal, an omnivorous flesh-eater, the food substances are treated by digestive juices of great activity which contain a whole series of soluble ferments. The stomach secretes two of these: rennet and pepsin. The pancreas elaborates three: trypsin, amylase and saponase, which act on the three main groups of food substances. To these the small intestine adds a special ferment, described by Pawloff^[83] under the name of enterokynase. Every one recognises the proteolytic function of pepsin and trypsin and the analogies and differences between these two diastases. Nor need I dwell on amylase or on the ferment which saponifies fats. But enterokynase merits special attention in connection with the study of immunity. Pawloff entrusted to his pupil Chépowalnikoff the study of the digestive rôle of the intestinal juice concerning which, up to this, very little was known. It was known indeed that this juice contained weak saccharifying and inverting ferments, but it was generally regarded as a secretion of little importance. Chépowalnikoff^[84] has demonstrated that this view is absolutely erroneous. The intestinal juice fulfils the very important function of accelerating the action of the three pancreatic ferments. The duodenal juice of the dog, especially, contains enterokynase. When this juice is mixed with a pancreatic juice that by itself actively digests fibrin and albumen, digestion takes place still more rapidly, the action being from three to four times as great. The part played by the intestinal juice becomes even more evident when it is mixed with a pancreatic juice that has little or almost no activity, as is the case of that from dogs that have recently been operated upon. Thus pancreatic juice, which has no action upon albumen, digests it promptly when a certain quantity of duodenal juice is added. When Chépowalnikoff took 500 c.c. of inactive pancreatic juice diluted with 500 c.c. of water or soda solution and added to it but a single drop of intestinal juice, the mixture exerted a manifest digestive action on coagulated albumen.

If, in place of pancreatic juice, we take the aqueous or glycerinated extract of the pancreas, which by itself exerts a very insignificant digestive action on albumen, and add to it intestinal juice, digestion takes place immediately. If it be admitted, as several physiologists maintain, that the inactivity of the pancreas is due to the fact that we have zymogen present in place of trypsin, one might conclude with Chépowalnikoff that “the intestinal juice possesses the power of transforming the zymogen into trypsin, and that this transformation takes place in a much more marked degree than in the presence of acids or the oxygen of the air” (p. 137).

The intestinal juice, from whatever region of the small intestine it be derived, exercises an undoubtedly favourable influence on the digestion of starch by the pancreatic juice, but this action is much more feeble than that on trypsin digestion. The action of the intestinal juice on the saponification of fats is even less marked. But here it is to the bile that the more important rôle is transferred. This fluid also augments the activity of the pancreatic juice, but in a manner different from the intestinal juice, for it acts especially by accelerating the digestion of fatty substances.

The action on the pancreatic digestion is not in any way interfered with when the bile is heated to boiling point. On the other hand the intestinal juice, under these conditions, completely loses its accelerating rôle. It follows from this, as has been formulated by Pawloff, that, in the intestinal juice, the existence of a soluble ferment which is destroyed by heat must be admitted; to this ferment he proposes to give the name of enterokynase. Without exercising a digestive power on any of the alimentary substances, it may act as a ferment of the pancreatic ferments.

Delezenne, at the Pasteur Institute, has repeated Chépowalnikoff’s experiments. He has confirmed the accuracy of his results and has added new data of great importance, not only as regards the physiology of digestion but also in relation to the study of immunity. Enterokynase appears from Delezenne’s experiments to be a true ferment; carried down by the same precipitants (collodion, phosphate of lime, alcohol) which enable us to obtain the greater number of the known ferments; it is sensitive to high temperatures, and even that of 65° C. is sufficient to do away with the greater part of its activity. Yet another property of enterokynase, which it possesses in common with the soluble ferments and which has for us a very special interest, is the facility with which it attaches itself to fibrin. By means of flakes of this substance we can at any time remove from a fluid the whole of the enterokynase contained therein. This fixative property is very important in connection with the part which enterokynase plays in digestion. The fibrin to which it has become attached absorbs trypsin with great avidity. If we introduce flakes of fibrin impregnated with enterokynase along with other flakes which have not been in contact with this ferment into a solution of trypsin, the former are digested with great rapidity, whilst the latter do not undergo any change. The fibrin that has fixed enterokynase is capable of clearing a fluid of its trypsin. On the other hand, that which has not been acted upon by the intestinal juice leaves it there almost unaltered.

It is of the utmost importance that we should inform ourselves as to the origin of the enterokynase of the intestinal fluid. This fluid, when obtained from a fistulous opening, for example, contains mucus and a considerable amount of débris of various kinds of cells. What are the elements which furnish such a remarkable ferment? Delezenne has obtained a very precise answer to this question. The enterokynase is not contained in the mucus and is not secreted by the intestinal glands; it comes from the lymphoid organs.

If the small intestine of a fasting dog be washed carefully with water all the pre-existing enterokynase is removed from it. The Peyer’s patches are then removed and treated with chloroform water. The other parts of the small intestine are similarly treated. This fluid dissolves the enterokynase, as it does the other soluble ferments. We find that the Peyer’s patches furnish enterokynase, but that the rest of the intestine, including Lieberkühn’s glands, give none.

We know that the Peyer’s patches are lymphoid organs in which are a large number of amoeboid mononucleated cells, and that these elements are even capable of ingesting foreign bodies and of submitting them to intracellular digestion. It is therefore not at all astonishing that Delezenne should have succeeded in finding enterokynase in the mesenteric glands of several Mammals (dog, pig, rabbit). These glands, when treated by the method just mentioned, yield a substance which assists the action of trypsin just as does the intestinal juice. Having reached this point, Delezenne asked himself whether the mononucleated white corpuscles, so closely allied to the mononucleated cells of the lymphoid organs, may not also contain enterokynase. With the object of settling this point he collected exudates that were rich in mononucleated leucocytes; in these also he found this same soluble ferment. Moreover, the leucocytic layer of the blood showed itself equally capable of increasing, very energetically, the action of trypsin.

The results of the old experiments carried out by Schiff and by Herzen on the adjuvant rôle of the extract of the spleen in pancreatic digestion, must without doubt be ranged alongside those we have just indicated. In fact the mononucleated cells of the spleen, like those of Peyer’s patches and of the mesenteric glands, contain a substance which acts like enterokynase. Delezenne has given us a definite demonstration of its presence and action.

In intracellular digestion it is the chemical side which has been most difficult of demonstration. The purely physiological functioning, the sensitiveness of the digestive cells and the amoeboid movements of their protoplasmic processes are, on the other hand, so manifest that it has even been suggested that intracellular digestion should be looked upon as a protoplasmic phenomenon purely vital in character.

In extracellular digestion through the agency of secreted juices we have a very different condition. Here the chemical side is the striking feature, the physiological factor being veiled more or less completely. Nevertheless, thanks to recent advances and above all to the labours of Pawloff’s disciples in St Petersburg, this problem has been elucidated in a very remarkable fashion.

The secretion of digestive fluids follows definite laws, the most potent factor being the reflex action of the nervous system. To use the expression of Pawloff, the study of the process of salivary secretion has revealed a real psychology of these organs. You may fill the mouth of a dog with small polished pebbles or with snow; you may pour into it very cold water—the saliva will not flow. But merely allow the animal to see sand in the distance—the glands at once begin to secrete fluid saliva. Tempt the dog with flesh—and immediately a thick saliva appears; show him dry bread—saliva is secreted in abundance, even if the dog has no great desire to eat.

The same phenomena may be observed in the stomach. Mechanical stimulation by inert bodies, such as stones, provokes no secretion; but the suggestion of a meal or the sight of food is sufficient to call forth a large quantity of gastric juice. The quantity and quality of the gastric juice are regulated by the quantity and quality of the food. Bread given to a dog provokes the secretion of a gastric juice endowed with the greatest digestive power. That which flows after the ingestion of milk contains only one-fourth as much pepsin.

In spite of these differences in the gastric secretion in relation to food, Pawloff and his pupils have never been able to assure themselves that there was any prolonged and chronic adaptation of the gastric function. They were struck by the uniformity of the digestive power of a great number of their dogs. Samoïloff^[85] had under observation three dogs placed on different diets. In spite of the very long periods during which these diets were given, the gastric juice, in all the dogs, presented the same properties and manifested no appreciable difference. This result harmonises with that indicated above as obtained in the Actinians fed with blood by Mesnil. In spite of repeated feedings on blood from the same species of animal, the extract from the mesenterial filaments was in no way different from that of the fasting Actinians used for control.

The pancreatic secretion is, in many respects, a more perfect type. We have here to do with the principal agent in the digestive function, without which the organism could not continue to exist. The advances made in surgery have enabled us to remove the stomach, first in the dog and then in man, and there are already several persons^[86] from whom the stomach has been removed and who, in spite of this operation, have continued to live. A portion of the small intestine may also be removed, but, in order that life may not be endangered, a considerable portion of it must be left intact. It is evident then that the pancreatic digestion is an admirably organised function both in animals and in man. One of the main regulators of this process of digestion consists in the great sensitiveness of the intestinal mucous membrane. Just as the organs of the buccal cavity possess in the specific sense of taste an excellent means of discrimination in the choice of foods, so the mucous membrane of the small intestine is endowed with a special sensitiveness, comparable to the chemiotaxis of unicellular organisms and of the cells of more highly developed organisms. Hirsch and Mehring have satisfied themselves that the passage of the contents of the stomach through the pyloric orifice depends on a reflex mechanism which proceeds from the upper reaches of the small intestine. To the researches of the school of Pawloff, however, we owe what light has been thrown on this question. The duodenal mucous membrane is endowed with a well-developed chemiotaxis for acid substances. The passage of the acid content of the stomach into the duodenum determines this chemiotaxis and brings about a secretion of alkaline juice which neutralises the acid. This contest between acid and alkali forcibly calls to our mind the analogous phenomena in those plants that defend themselves against the alkaline secretions of parasites by the production of an acid (see Chapter II). As in these lower organisms, this battle of the chemical secretions is regulated by the action of living and sensitive parts.

When the acidity of the mass which passes through the pylorus is too marked, the reflex contraction starting from the duodenal mucosa arrests its passage. Then takes place a neutralisation of the acid, thanks to the alkaline secretion, and the pylorus is again allowed to open. This mechanism thus regulates the passage of the contents of the stomach into the duodenum, the passage taking place in instalments.

The sensitive intestinal mucous membrane can estimate not only the degree of acidity, but also the other chemical characters of the aliments which pass into the duodenum. This chemiotaxis is, as it were, the starting-point of the reflex action which excites the pancreatic secretion with its contained three ferments. The passage of bread through the pylorus excites the secretion of a juice very rich in amylase and very poor in saponase. The passage of milk into the duodenum brings forth, on the other hand, a juice very much richer in saponase but poorer in amylase and in trypsin. Flesh-meat provokes the secretion of a pancreatic juice which is less rich in amylase than the juice poured on bread, but richer in saponase. Fat causes the secretion of a juice still richer in saponase than is the juice poured out in the presence of bread or milk. These facts now carefully established—especially by Walter^[87]—demonstrate that the pancreatic function is carefully regulated as regards its adaptation to the characters of the food substances on which it is to act. Such adaptation may even become permanent.

Whilst, as already stated, the stomach, under the influence of a fixed diet, is incapable of effecting any lasting modification in the composition of its secreted juice, the pancreas may reach this degree of perfection. When a dog is fed for several weeks on bread or on milk and is then placed on flesh diet its pancreatic juice is found to become progressively richer in trypsin. Whilst this augmentation of the proteolytic power is being brought about, the juice becomes poorer and poorer in amylase. Wassilieff^[88] has carried out a large number of experiments on this point and has demonstrated a very remarkable adaptation of the pancreatic juice to the wants of nutrition, an adaptation that may become permanent. A dog which has been accustomed to digest bread and milk adapts itself to this nourishment: its pancreatic juice contains less and less trypsin, but, on the other hand, becomes richer in amylase. Pawloff observed that in dogs great variations in the composition of the pancreatic juice are often present; this he attributes to the diet to which these animals had been previously subjected.

Not only does the quality of the digestive juices accommodate itself to the wants of digestion; their quantity also undergoes variations according to the part that these juices have to play. Thus, Pawloff has observed that his dogs secreted a saliva which was very fluid and very abundant when he gave them acids, bitter substances or other substances they did not like. On the other hand, the presence of food in the mouth, or even the sight of it, excited the secretion of a thick saliva containing a large quantity of mucin. In the first case the part played by the saliva was that of diluting the injurious substances as much as possible, in the second that of facilitating the deglutition of the food.

In general the organism manifests a tendency to produce more digestive ferments than it actually needs for digestion. It is for this reason probably that they are often found outside the digestive canal. Among these ferments pepsin and amylase, especially, have been definitely proved to be present in the urine of man and of some mammals, notably the dog. The data as to rennet and trypsin are not so well established. But, as several of these ferments, such as amylase and trypsin, may be derived from several sources in the organism, their elimination by the urine is less important for the thesis I have just formulated than is that of pepsin.

Pepsin was found in the urine by Brücke exactly forty years ago. It is more frequently found in the morning urine, but is absent from that passed immediately after the principal meal. Leo and Senator^[89] found only traces of pepsin during the prolonged fast of the Italian Cetti; but the day he broke his fast they were able to demonstrate the presence of a considerable quantity of this ferment in his urine.

Delezenne and Froin, with the object of seeking the source of the urinary pepsin, extirpated the stomach of a dog. After the animal had recovered, they fed it well and examined its urine at different periods of the day. By the methods which had shown the presence of pepsin in all the normal dogs taken as controls they could never discover the faintest trace of this diastase in the urine of the dog that had been operated upon. On the other hand, the urine of a dog whose stomach had simply been isolated, contained very much the same quantity of pepsin as that of normal dogs. This experiment proved among other things that the pepsin, before it could be eliminated by the kidneys, must have been re-absorbed by the wall of the stomach. From these data, combined, it must therefore be admitted that the pepsin found in the blood and which passes thence into the urine can only be of gastric origin. As it serves no useful purpose in the organism we must conclude that a portion of the pepsin, secreted by the stomach and not used for digestion, has been rejected as superfluous.

The study of the digestive function of animals gives us information on a large number of points of the highest importance for the comprehension of immunity. Intracellular digestion, a function so widely distributed in the lower animals, is very intimately connected with the phenomena which are observed when micro-organisms are destroyed in the animal organism. Extracellular digestion furnishes us with information concerning many of the features of progressive adaptation, similar to those which are observed in connection with acquired immunity.

When we examine the phenomena of intracellular digestion and those of secretory digestion as a whole, we see that, in both, the chemical processes are subjected to the influence of the living parts of the organism. In the lower animals, it is the protoplasm of the amoeboid cells which regulates the chemical processes in digestion; in the higher animals, this rôle is taken by a very complicated apparatus, in which the nervous system plays a predominant part.