Nobel Lecture, December 11, 1913

Anaphylaxis

It is not without emotion that I address this assembly on the experiments that have brought me, through the most gracious favour of the Caroline Institute, the highest reward that a scientist has the right to hope for. I ask your indulgence in speaking of my own research, as I must do, and in setting out the findings that have given anaphylaxis a leading place in general pathology over the last decade.

First I feel I must explain and indeed justify the use of the word itself, for it may seem somewhat barbarous at first glance. This neologism I invented twelve years ago on the assumption, which I think is still valid, that a new idea calls for a new word in the name of scientific precision of language.

Phylaxis, a word seldom used, stands in the Greek for protection. Anaphylaxis will thus stand for the opposite. Anaphylaxis, from its Greek etymological source, therefore means that state of an organism in which it is rendered hypersensitive, instead of being protected.

To make this plain, we will consider the example of a subject that has received a poison.

Let us suppose the dosage to be moderate and that after a few days the subject is, or at least appears to be, normal. If, at this point, a further injection is given of the same dosage of the same poison, what will happen?

There are three possibilities.

The first and simplest is that there has been no change in the organism and that in receiving the same dosage as one month previously, exactly the same phenomena will result, in exactly the same conditions. Naturally this is what happens most of the time. Specialists and doctors work on this assumption when they repeat the intoxication at one month intervals.

The second possibility is that the subject has become less sensitive. In other words, the preceding intoxication has produced a certain condition of tolerance or non-sensitivity. This will mean that a stronger dose is necessary at the second injection to give the same results. This is the case of (relative) immunization or, as it is sometimes called, of mithridatism. The most remarkable case of this tolerance is to be seen when opium or morphine are used. People who take morphine injections need stronger and stronger doses for the morphine to take effect. Some unhappy morphine addicts get to the point of standing a dose of 20 grams, whereas one decigram is dangerous in a normal subject. It has been known for persons to drink one litre of laudanum per day, while one drop of laudanum produces already some effect.

These two cases, of unchanged sensitivity or stability, and of diminished sensitivity or habituation, have been known since long. Now I have shown that there is a third possibility, frequently to be observed in certain conditions which I have specified: this is of heightened sensitivity. The first injection, instead of protecting the organism, renders it more fragile and more susceptible. This is anaphylaxis.

These are the circumstances under which I first observed this phenomenon. You will allow me to go into some details on the origins. You will find that it is by no means the result of profound thought but a simple observation, almost a fortuitous one; so that my merit has only been in letting myself see the facts which were plain before me.

In tropical waters, Coelenterata are to be found floating on the surface, also known as Physalia (Portuguese galleys). The basic structure of these creatures is a pocket filled with air so that they can float like a bladder. A bucco-anal cavity is subjoined to this pocket, with very long tentacles which hang in the water. These feelers sometimes run to two or three meters long and are equipped with small devices which adhere like sucking cups to objects encountered. Within each of these innumerable suction-cups is a pin-point which drives into the foreign body that is being touched. At the same time, this pin-point causes penetration of a subtle but strong poison, which is contained in the tentacles, so that contact with a feeler of the Physalia is tantamount to a multiple injection of poison. On touching a Physalia an acute sensation of pain is felt immediately, due to the penetration of this liquid venom. This is similar in relative intensity to a swimmer’s mishap when he bumps into a jelly-fish in the water.

During a cruise on the yacht of Prince Albert of Monaco, the Prince advised me to study Physalia poison, together with our friends Georges Richard and Paul Portier. We found that it is easily dissolved in glycerol and that by injecting this glycerol solution, the symptoms of Physalia poisoning are reproduced.

When I came back to France and had no more Physalia to study, I hit upon the idea of making a comparative study of the tentacles of the Actinia (Actinia eqnina, Anemone sulcata) which can be obtained in large quantities, for Actinia abound on all the rocky shores of Europe.

Now Actinia tentacles, treated with glycerol, give off their poison into the glycerol and the extract is toxic. I therefore set about finding how toxic it was, with Portier. This was quite difficult to do, as it is a slowly acting poison and three or four days must elapse before it can be known if the dose be fatal or not. I was using a solution of one kilo of glycerol to one kilo of tentacles. The lethal dose was of the order of 0.1 liquid per kilo live weight of subject.

But certain of the dogs survived, either because the dose was not strong enough or for some other reason. At the end of two, three or four weeks, as they seemed normal, I made use of them for a new experiment.

An unexpected phenomenon arose, which we thought extraordinary. A dog when injected previously even with the smallest dose, say of 0.005 liquid per kilo, immediately showed serious symptoms : vomiting, blood diarrhoea, syncope, unconsciousness, asphyxia and death. This basic experiment was repeated at various times and by 1902 we were able to state three main factors which are the corner-stone of the history of anaphylaxis: (1) a subject that had a previous injection is far more sensitive than a new subject; (2) that the symptoms characteristic of the second injection, namely swift and total depression of the nervous system, do not in any way resemble the symptoms characterizing the first injection; (3) a three or four week period must elapse before the anaphylactic state results. This is the period of incubation.

Once these first factors in anaphylaxis were well grounded, the field opened right up, thanks to the skilled and fruitful research of many investigators.

In 1903 Arthus, in Lausanne, showed that a first intravenous injection of serum on a rabbit causes anaphylaxis, i.e. three weeks after the first injection the rabbit is hypersensitive to the second injection. The phenomenon of anaphylaxis was becoming of general application. Instead of applying only to toxins and toxalbumins, it held good for all proteins, whether toxic at the first injection or not.

Two years later Rosenau and Anderson, two American physiologists, demonstrated in a noteworthy piece of work that the phenomenon of anaphylaxis occurs after every injection of serum, even when the injection is minute, for example of 0.00001 ml which is an infinitely small amount but nevertheless sufficient to anaphylactize an animal. They quoted examples of anaphylaxis from all organic liquids: milk, serum, egg, muscle extract. They specified the reaction and clearly showed that of all the subjects, the guinea-pig appeared the most sensitive in anaphylactic terms.

In 1907 I conducted an experiment which shed much light on the pathogeny of anaphylaxis. An anaphylactic state is produced by taking the blood of an anaphylactized animal and injecting it into a normal animal subject. The anaphylactogen poison is therefore a chemical substance contained in the blood.

Such are, I think, the main stages through which our knowledge has passed. I pass now to particular points I wish to stress.

The incubation period varies according to the poison used rather than according to the type of animal subject. There is however a minimum period of one week (in the guineapig, following the injection of milk). With mytilin extracted from the common mussel (Mytilus edulis), the incubation period is a fortnight. With the dog, using crepitin extract from Hura crepitans, the period is longer, of some four weeks. With the guinea-pig, following the injection of serum – on an exhaustive series of experiments – the incubation period is of some eleven days and the reaction symptoms reach their peak at the fourteenth day, always allowing for considerable variation according to subject.

But it is a much harder task to state when the anaphylactic period has actually passed. Most writers incline to the view (and I myself would think them correct in their view) that the anaphylactic state never passes. In other words, once a subject has been anaphylactized and consequently modified in his chemical constitution, then the subject can never go back to his former state. Return to normal is not possible. Subjects have been known who even after four years from the date of the first serum injection, were still sensitive to the unleashing reaction.

Let me add in passing that it is an extraordinary phenomenon that so insignificant a quantity of poison can modify the organism to the extent that the succeeding days down long years can not eradicate this indelible modification. Unfortunately minute researches on just this point are still lacking. But it certainly looks as though considerable differences will be found in the duration of anaphylactization.

Anaphylactic symptoms also vary to a great extent, although the differences are marked rather according to the nature of the experimental animal than according to the nature of the poison used. It is indeed worthy of note to find that the phenomena are constant, whatever the poison used.

I have made especial study of anaphylaxis in dogs, which permits of greater accuracy in specifying symptoms than in experiments with the guinea-pig. In the dog, four degrees of anaphylaxis may be distinguished, according to intensity.

In the lightest form, the main symptom is prurience or itching. The animal, let loose, sneezes and gives various shakes of the head as if there was something inconvenient in his ears. The dog scratches his head and sides with his paws, sometimes frantically. Sometimes he rubs his muzzle against the ground and rolls over.

The next stage in anaphylactic intensity is characterized by itching again, but this time more violent. This is followed almost immediately by various symptoms; more rapid breathing, lowered arterial pressure, faster heart-beat, vomiting, blood diarrhoea and rectal tenesmus.

At the third degree, depression of the nervous system is such that the itching has gone or almost gone. The animal has no strength to vomit, diarrhoea is marked while the fluid passed from the rectum is often almost wholly blood. The nervous symptoms often develop so suddenly and violently that there is no time for colic and diarrhoea. Ataxia follows at once. The animal reels as if drunk, the pupils are dilated, the eyes haggard and after heart-rending cries, the animal falls to the ground, urinating and defecating underneath himself, unconscious, no longer reacting to the excitations and in complete mind-blindness. Breathing is laboured and agonized. The heart beats are so faint as to be barely perceptible: blood pressure hardly reaches the one or two centimetre mercury level. To sum up, all the symptoms point to the central nervous system being the seat of severe and sudden intoxication. This brutal assault of the poison on the nervous system has been called anaphylactic shock.

There is a fourth degree of anaphylaxis, it may be said, which is more serious still: when all the symptoms, instead of passing off, worsen so that within a quarter or a half hour the subject is dead.

In the dog such death at the onset is rare. In most instances, following the anaphylactic shock, the dog revives. After fifteen or thirty minutes, he gets to his feet, staggering a bit, regains feeling and consciousness and is left with only blood diarrhoea still persisting from the anaphylaxis. Often death takes place during the night following the injection; but constantly after a period of apparent recovery.

In the rabbit, according to Arthus, respiration becomes polypneuic. The animal falls on its side, throws its head back, makes running movements with the legs and then suddenly breathing stops. Heart failure is systolic and death ensues within a couple of minutes.

Arthus also observed some interesting local effects of anaphylaxis in the rabbit. The second injection being given in the same ear as the initial injection, ulcers and gangrene appear, although there are almost no general symptoms. This local effect of anaphylaxis is often called the “Arthus phenomenon”.

The guinea-pig is extremely sensitive to anaphylaxis. If the anaphylaxis is slight, only symptoms of itching, excitation and heightened breathing appear. Often the animal falls on its side, sometimes in violent convulsions, sometimes on the contrary paralysed and powerless. In both of these cases, death takes place fast and it is almost a matter of seconds between the injection and the final failure of the heart.

Anaphylaxis has been observed in all animals: the horse, the goat, the ox, the rat, the pigeon, the duck and even recently in frogs.

Anaphylaxis takes place also in human subjects and has caused death in certain instances. It is indeed probable that sudden death following the bursting of a hydatid cyst is an anaphylactic phenomenon. Some years back I was in Brazil and I heard the story of a doctor who had given himself a preventive injection of anti-plague serum. The next year a new outbreak of plague was feared so he persuaded his students to have a preventive injection of the same serum. He set the example by giving himself another one. This was however an unleashing injection and his body had been affected by the first. The second injection was fatal and within two hours he was dead.

Now however the effects of anaphylaxis in mankind are very well known. Two doctors from Vienna, Pirquet and Schick, have studied the matter with the greatest care. They have described serum-sickness (“Serum-Krankheit”) in children subjected to injections of diphtheria serum and they saw that it was in most cases an anaphylactic phenomenon. It is only in the rarest cases that the first injection is productive of immediate reaction. When it comes to the second injection, an immediate reaction follows for 90% of the cases, that is to say when the period between the first and second injection is from ten to thirty days.

The symptoms to be observed are very close to symptoms observed in animal subjects: urticaria, erythema, pangs of pain, itching and in the worst cases demi-syncope, with nausea, vomiting, hyperthermia, edema over the whole skin area and general urticaria.

Thus by comparison of anaphylactic effects in man and the animals, it will be seen that they are akin. It is as if poison had been produced, which reacts upon the nervous system, especially on the vaso-motor nerves or the trophic nerves of the skin.

It is now opportune to examine the substances apt to develop the anaphylactic state. They can be defined very simply, by using a fairly arbitrary system of classification, which groups substances in colloids on the one hand and crystalloids on the other.

Crystalloids are on the whole non-active. I am not aware of any successful attempt to induce anaphylaxis by one crystallizable salt or by any alkaloid. On the other hand all the proteins without exception produce anaphylaxis: one has seen this with all sera, milks, organic extracts whatsoever, all vegetable extracts, microbial proteinotoxins, yeast cells, dead microbial bodies. It would be of more interest now to find a protein which does not produce anaphylaxis than to find one that does.

But what is above all important is to know the degree of specificity of these injections.

At first sight it looks as if the specificity is pushed very far. For example if the preparatory injection is of goat’s milk, then the unleashing injection will be much stronger and will have more intensive effects if made from goat’s milk than if made from cow’s or sheep’s. Again, for the unleashing injection of horse serum to take maximum effect, the first injection should also be of horse serum. It is obvious that the animal in this case is still somewhat sensitive to a second injection with serum from a dog or rabbit, but the effect is far less. It is thus permitted to conclude that there is specificity, that is to say necessary identity between the preparatory and the unleashing injection.

I will be coming back to the meaning of this term, specificity. First, I will mention a curious use to which anaphylaxis has been put in forensic medicine, on this principle that there is specificity.

Suppose for the sake of example some blood drops of unknown provenance, which however must be discovered in the name of medical jurisprudence. Let us say, it has to be established whether the blood is human or of a dog or a pig or an ox. Guinea-pigs are used; one is injected with human serum, another with dog serum, another with ox serum, and another with pig serum. Then one month later, the blood of unknown provenance is made into a water solution. The same small quantity of the unknown blood is then injected into each of the guinea-pigs in turn. If one of them shows morbid symptoms and dies, for example the guinea-pig that had the human-blood serum injection, then we will conclude that the blood in question was in fact human blood.

I will recount at this point another experiment which was out of the ordinary. Flesh was taken from the mummified form of a man, three or four thousand years old. Muscle extract was made from this. The injection of this fluid into guinea-pigs made them sensitive to muscle serum and to human muscle serum only. This would show, were it necessary, that the chemical components of the human body have undergone no great variation in the course of the last four thousand years.

This series of evidence gives good reason for recognizing the specificity of anaphylaxis. However there must be no overstatement. Let us note that guinea-pigs sensitive to cow-milk serum are not altogether non-sensitive to goat- or sheep-milk serum, although their preparatory injection was only of cow-milk serum.

Two further series of observations I have made quite recently do lead me to question the hard and fast rules for specificity in anaphylaxis one is tempted to lay down. First, when I gave a preparatory injection of crepitin and I determined one month later the emetic dose (that means the dose causing vomiting) of apomorphine, I saw that with normal dogs a dose of apomorphine hydrochloride equal to 0.00275 of the salt per kilogram caused vomiting in 21% of the dogs, whereas with dogs initially injected with crepitin, for the same dose of apomorphine hydrochloride, vomiting ensued in 63% of them.

Anaphylactic dogs are thus more sensitive to apomorphine than normal dogs, and it follows that there exists general anaphylaxis, as apomorphine in no way resembles crepitin.

Further, the second experiment to be adduced against the specificity of anaphylaxis I conducted with two kinds of toxalbumin, extracted from the Actinia, a substance which I named congestin, as its property is to bring on grave congestion of the circulatory system in the intestines and stomach. Two congestins may be prepared at some pains: yellow congestin, soluble in a fluid containing 50% alcohol, and black congestin, completely insoluble in a fluid containing 25% alcohol. Now I was able to show that black congestin is not unleashing, but is better as the preparatory injection than the yellow congestin. This gives us authority for thinking that the sensitizing (or preparatory) property and the unleashing property belong to allied protein groups, but not identical ones. Biological chemistry will no doubt unravel these two substances. In practice the two substances, preparatory and unleashing, are almost always lined up together, so that we have a near right to pronounce on strict specificity.

Another experiment of prime importance is this, for it shows the very nature of the anaphylactic process. In April 1907 I showed that the injection of serum from an anaphylactized dog induced an anaphylactic state in untreated dogs, as if this serum contained the toxic substance which activates the unleashing injection.

With actino-congestin, the experiment is clear-cut. Almost harmless doses cause death within a matter of hours in dogs that had not been anaphylactized, but had had injections of serum from anaphylactized animals. This is what is known as passive anaphylaxis.

At about the same time, in May and June 1907, Gay and Southard in America, and Otto in Germany, also showed quite clearly that passive anaphylaxis exists. It has become one of the classic tenets of anaphylaxis.

Another finding, that I call anaphylaxis in vitro, allowed me as it were to synthesize the poison that is released during the unleashing injection.

The experiment worked best with crepitin. The immediate toxic effect of a certain dosage of crepitin was first determined, say of 0.004 g. Then serum is taken from an animal anaphylactized by crepitin, and in this serum is dissolved 0.004 g of crepitin. This injection is harmless, providing the crepitin had been diluted with water. It is however very offensive when the crepitin is dissolved in serum from a dog that has been anaphylactized. It must thus be admitted that by some chemical combination the crepitin in conjunction with the unknown substance in the anaphylactic serum has given rise to a veritable poison.

The effects of this new poison are extremely strong, as the following experiment will show. It was carried out on a bitch that had been given an active dose of crepitin mixed with the anaphylactic serum. “Severe vomiting, diarrhoea, rectal tenesmus: unable to keep standing, she urinates under herself; the pupils are dilated, the eyes haggard; complete mind-blindness, near-total failure of reflexes, deep unconsciousness, breathing dyspneic, heart-beat faint and very fast, pulse barely perceptible; dead in thirty-six hours.”

Thus, the mixture of the antigen with the blood of an animal anaphylactized by this same antigen, produces a strong violent poison which is different from the antigen itself.

To evaluate this reaction, we must mention a valuable experiment of Claude Bernard carried out long ago. Bitter almonds contain two substances: amygdalin which is harmless and emulsin which is harmless too. Animal subjects survive an injection of either amygdalin or emulsin. But emulsin is a diastase and has the property of breaking up amygdalin, liberating hydrocyanic acid, which is one of the most virulent toxic gases known. Thus if an animal that has been given amygdalin is then injected with emulsin, hydrocyanic acid will be formed in the blood stream and death will take place at once. Yet injected separately, neither the amygdalin nor the emulsin has any effect.

It is just the same with anaphylactic serum and the antigen. Separate, they are harmless. Together, they are fatal.

A simple hypothesis suggests itself, even though Wolf-Eissner has not yet been able to accept it. Let us assume the existence of a substance in the anaphylactized blood, which we will call toxogenin. It is in itself harmless as animals have it in the blood and seem to enjoy good health. It may moreover be injected into other animal subjects without harm. But if toxogenin is mixed with antigen, then a new poison is produced, which has immediate and serious consequences. This poison, derived as it is from the antigen, I propose to call apotoxin. The chemical reaction is straightforward: toxogenin + antigen = apotoxin.

This appears to be a general law of biological chemistry: that bodies that are non-active and harmless in themselves become harmful and activated when in reaction one to the other. Trypsin is non-active when it has not been in contact with enterokinase. The sperm must perforce contact the ovum in order for fertilization to take place, Hydrochloric acid must contact pepsin, for digestion, etc. All workers on anaphylaxis have had to assume the existence of this sensitizing substance that I called toxogenin. Besredka later called it sensibilisin, while Friedberger called it anaphylatoxin. The name matters little. The fact is that there exists in anaphylactized blood a substance harmless in itself but which releases a strong poison when mixed with the antigen.

I omit the details of the successful experiments undertaken by Besredka on antianaphylaxis, together with the painstaking work of Friedberger and his pupils on deviation of the complement. I will only mention the course of my own original research, for I have no hope in this lecture of covering the whole field of anaphylaxis research.

It is relevant here to indicate the relationship I have been able to establish between leucocytosis and anaphylaxis, a relationship that is hard to grasp without elaborate techniques and prolonged observations. All my experiments have been conducted on dogs, with the help of my friend P. Lassablière who did the calculations.

The number of white corpuscles or leucocytes in the normal dog is 100 per hundredth of a millimeter cubic on average, varying from 70 to 130. In animals, now, that have been anaphylactized, even after a considerable time-lag of say six months, when they appear to be completely normal and in perfect health, the number of leucocytes reaches and often exceeds 200.

An initial injection which makes the body anaphylactic, therefore, induces a marked leucocytosis and this is the only symptom that can be observed.

With weaker doses of antigen and with antigens that are harmless or practically so, such as peptone, the anaphylactic leucocytosis does not last as long but is nevertheless pronounced. A quantity of peptone equal to 0.005 per kilo live weight will still give leucocytosis and bring about either immunity or anaphylaxis. There is no reaction more sensitive than that of leucocytosis. By systematic analysis of this subtle phenomenon it seems clear to me that certain conclusions may be made which would have been utterly out of the question otherwise.

I will cite as illustration some experiments which I am still making on the action of chloroform on dogs. On a dog chloroformed for the first time, the number of leucocytes in the blood undergoes no modification either under the anasthaetic or after, whether on the second or the tenth or the twentieth day. If however a second chloroformization is carried out a month or so after the first, in conditions as nearly identical with the first as may be permissible, then on the third or the fourth or the fifth day in particular, severe leucocytosis will appear, reaching 220 or 250 leucocytes.

What is the explanation of this curious phenomenon? There can be no question of real anaphylaxis, for anaphylaxis is always severe, immediate and terrible, whereas in this instance, the leucocytosis only appeared on the third or the fourth day.

I have necessarily arrived at the following hypothesis. Namely, that the chloroform works on the hepatic cells and causes the break-up of certain protein substances in them, which pass thence to the blood stream. If it is the first time that these proteins have been released to reach the blood, then there is no leucocytic reaction. If, after an interval of three weeks, a new break-up takes place in the liver as a result of the second chloroformization, then this behaves like a second protein injection, the unleashing injection on an anaphylactized animal.

There does exist then, besides direct anaphylaxis, an indirect anaphylaxis. about which little is so far known. But it seems that indirect anaphylaxis greatly widens the scope of anaphylactic action. Anaphylactic phenomena have been the subject of much medical research. It would take too long even to list. As I have not myself undertaken work in this respect, I forbear to dwell on it

I cannot however pass over the possible relationship between anaphylaxis and tuberculin reactions. This topic is highly controversial and undoubtedly worthy of further studies.

From the start of our research on anaphylaxis, we noticed the analogy existing between anaphylaxis and sensitivity of tuberculous animals to tuberculin. The admirable contributions of R. Koch, which has since been borne out by numberless experiments undertaken by others, showed that a normal animal does not react to tuberculin, whereas tuberculous animals do react to doses a thousand times weaker. What is this heightened sensitivity, if not anaphylaxis?

When it came to questions of detail, considerable differences were found to appear. In fact a first injection of tuberculin does not make normal animals sensitive to a second injection. The blood of tuberculous animals does not induce passive anaphylaxis. Lastly, the anaphylactic reaction is on the whole one of hypothermy, while the tuberculin injection on tuberculous subjects always causes hyperthermy.

However, I do not believe that these are fundamental objections. At most they prove that the growth of the Koch bacillus produces preparatory substances which are not to be found in tuberculin. Tuberculin contains unleashing substances, but the preparatory substances are lacking, probably because the numerous chemical changes that must take place before the tuberculin can be extracted from tuberculous cultures have themselves caused change in the preparatory substance. I am of the firm belief that in the animal organism infected by the tubercle bacillus, the infection creates substances that act as preparatory, but which are not found in tuberculin as we use it. This is not paradoxical, at all.

It may be thought that general application can be made of this anaphylactic method of diagnosis. Two methods lie open. One, the patient may be given a subcutaneous injection of specific serum to see if he is sensitive to the reaction. The other is to take the patient’s serum and inject it into guinea-pigs, seeing after the passage of two or three days if the guinea-pigs are sensitive to such and such bacterial toxin.

I considered whether this method of diagnosis by anaphylaxis might not be made use of in cancer. Taking cancer tumours and precipitating by alcohol the aqueous extract of such tumours, a precipitate results which admits of purification by being dissolved and precipitated in successive steps. This dry product can then be dissolved in water and injected into patients suffering from cancer. If anaphylacto-diagnosis of cancer did really exist, this injection would produce a certain reaction. This was not the case. Some of my colleagues made the injection of this product into patients with cancer. The effects of the injection were absolutely nil.

While on the subject of negative experiments, I wish to say a word on what I call homogenic anaphylaxis. The aim was to discover if the injection into an animal of blood from another subject of the same species, provokes a stronger reaction at the second injection than at the first, always given the same source for the transfusion in both cases.

Here again the results were absolutely nil. A dog A was injected with 70 gram per kilo of the blood of another dog B. Not much happened. A month later, the same dog A that had been treated was given a further injection of 70 gram per kilo of blood from the same transfusion source dog B. No symptom was observed. It seems thus there is no such thing as homogenic anaphylaxis, and the blood of one species of animal injected into an animal of the same kind is harmless both at the first and at the second injection.

To date, all experiments mentioned above have been carried out by parenteral injections, that is to say that the substance introduced into the blood was introduced by other means than the digestion, and namely by means of subcutaneous, intravenous, intraspinal and peritoneal injections. But there is also anaphylaxis which comes after ingestion by way of the digestive system. This is alimentary anaphylaxis and it follows ingestion by the digestive duct.

It was for the first time demonstrated by Rosenau and Anderson in 1906 that guinea-pigs were sensitive to horse serum after first ingesting horse serum by way of the digestive tract.

It should be understood that the term alimentary anaphylaxis does not signify anaphylaxis by alimentary substances but anaphylaxis by the introduction of the anaphylactizing substance by way of the digestive channels. Alimentary anaphylaxis is characterized by the antigen, whether alimentary or not, being introduced into the organism by means of the digestive tube. Introduction by the rectal duct is not included, as the essential feature of alimentary ingestion is absent, which is the modification of the antigen by the digestive juices.

Alimentary anaphylaxis has been studied on various hands since Rosenau and Anderson, but the results are not so far constant nor uniform. I have tried to tackle the problem from another angle, that is to see under what conditions substances introduced into the stomach can pass into the blood. I used a reagent that is extremely sensitive, namely leucocytosis.

A dog is given cooked meat: no leucocytosis results. A dog is given raw meat, even one fifth in quantity compared with the cooked meat, then in three or four hours time, leucocytosis results. The most likely and simplest explanation is that when cooked meat is ingested, all the proteins have become non-soluble and can not be made soluble except by the action of digestive juices: pepsin, trypsin and erepsin. The products of the break-up of the protein that are formed are non-toxic and do not induce the leucocytic reaction. It is therefore not surprising that cooked meat should be ingested without affecting the leucocytes, for no soluble protein has been introduced into the stomach, and the only proteins which can pass it are those that have been modified, transformed and homogenized by the digestive juices.

Now if muscle serum or raw meat is ingested, then soluble proteins are introduced into the stomach. The digestive juices have powerful action, but it is probable that part of the protein escapes and certain particles pass into the circulation, thus effecting a true antigen injection, which can thus set off the leucocyte reaction.

It follows that each time soluble protein is introduced by the digestive channels, anaphylactic reaction may result, as it is equivalent to an antigen injection.

This may explain away the divergences of opinion among physiologists in respect of alimentary anaphylaxis, for following the introduction of a protein, depending on whether it is soluble or not, whether it is absorbed or not, whether it is resistant to the action of the ferments or not, it will or will not penetrate into the blood system.

In fact I have been able with crepitin to cause a clear instance of alimentary anaphylaxis.

I have indicated that there are three methods of alimentary anaphylaxis. Let us call the alimentary ingestion A, and the parenteral injection P. The following combinations are possible: (I) A preparatory, A releasing; (2) A preparatory, P releasing ; (3) P preparatory, A releasing. Even in the first of these three cases (A + A) where the anaphylaxis is strictly alimentary, for the initial ingestion as well as the subsequent ingestion, there is no doubt about anaphylaxis having taken place. When a dog ingests crepitin for the first time, he never vomits. When he ingests it for the second time, some three weeks later, he always vomits. This is the anaphylactic protective vomit. In the second case (A + P), the preparatory ingestion being alimentary and the releasing injection parenteral, the results are clearer still. In effect the anaphylactic shock is violent and plainly proves that a small quantity of crepitin must have escaped the digestive juices at the first ingestion and passed to the blood, as the lasting leucocytosis to be found in animals that have ingested crepitin also shows.

I have observed in this connection a remarkable fact: a period of one year between the initial ingestion and the subsequent parenteral injection. A dog ingested in June 1911 a strong dose of crepitin and survived. (Whatever the ose, it is not possible to poison dogs by ingesting crepitin.) After one year had passed, in June 1912, this dog had a harmless crepitin injection and died within an hour and a half as if struck by lightning. The death of a dog at this speed from anaphylactic shock is very rare indeed.

To these experiments, I must add the work of Gideon Wells and Thomas Osborne. In January 1911, they made a close study of the anaphylactizing and immunizing action of vegetable proteins.

The general conclusion is as expected but nevertheless necessary to be shown: (1) through the digestive mucous membranes never passes more than tiny amounts of colloids, but sometimes it does pass them; (2) these minute amounts are enough on occasion to cause the anaphylactic state either preparatory or unleashing; (3) the amounts of colloids that pass into the digestive juices are weak enough to give immunity rather than anaphylaxis, especially if it be remembered that most are cases of ingestion repeated and increased at various intervals: all which conditions favour antianaphylaxis immunity rather than true anaphylaxis.

These findings in the field of alimentary anaphylaxis are perhaps not without importance to clinical medicine. It may be that many cases of dyspepsia are nothing more than light attacks of anaphylaxis. Doctors have long found that regular diet on strictly uniform lines was to be preferred to all other regimens. It is as if by the repeated ingestion of one some protein substance the organism had accustomed itself to it and had immunized itself against this usual antigen.

No need to go over the more extraordinary aspects of alimentary anaphylaxis that had hitherto remained unexplained. It has long been known that some people are sensitive to cheese or to strawberries or to fish or to shellfish or to eggs or even to milk. Now the symptoms to be seen in such individuals on ingesting such and such foods are analogous to the effects of anaphylaxis: acute stomach pains, vomiting, diarrhoea, colic, erythema, urticaria, severe itching and sometimes cardiac troubles and fever. We know now that these are anaphylactic phenomena; this has become a pathological commonplace.

We shall conclude by reiterating the various phenomena and attempting to establish their import in general terms.

In the first place anaphylaxis, like immunization, creates humoral differentiations between different individuals.

A guinea-pig that is anaphylactized by horse serum will not be identical to untreated guinea-pigs nor to guinea-pigs anaphylactized by ox or dog serum. This means that over and above the individual differences due to diverse means of immunization, there are individual differences due to diverse anaphylactizations. One has only to think of the innumerable quantity of substances that are anaphylactizing and the substances that can immunize, and one will conclude that the chemical or humoral diversity is so to say unlimited with the different individuals.

To be different from other members of the same species, an animal has only to receive into his blood a small quantity of alien protein which anaphylactizes him in a special way, or for a microbe to evolve in his blood which gives him immunity in a special way. In the course of some years’ life span, the same organism that is unique will accumulate immunities or anaphylaxia that appertain to it, diversely grouped in diverse subjects until each one of these persons will differ from all others.

Each one of us, by our chemical make-up, above all by our blood and probably also by the protoplasm of each cell, is himself and no one else. In other words, he has a humoral personality. We all know very well what the personality of the psyche is. The multiplicity and the variety of our memories make each one of us different from all other human beings. We all have a body of stored impressions which preclude our being confused with any other specimen of our kind. Nothing could be clearer than this idea of the personality in terms of psyche which stands to reason and is valid in all human conscience.

Now, in the light of notions of immunity and of anaphylaxis, we can conceive of another personality in juxtaposition to the moral personality and that is the humoral personality, which makes us different from other men by the chemical make-up of our humours.

This is an entirely new idea. It was thought up to now, perhaps from lack of after-thought, that with individuals of the same age, race and sex the humors would no doubt be chemically identical. Well, it is not like that at all. Every living being, though presenting the strongest resemblances to others of his species, has his own characteristics so that he is himself and not somebody else. This means that henceforth study of the physiology of the species is no longer enough. Another physiology must be taken up, which is very difficult and barely broached, namely that of the individual.

It may be asked how anaphylaxis fits in to that general law, which admits of no exceptions, that living organisms exist in an optimum state of protection.

It does indeed seem absurd that an organic disposition should make beings more fragile, more susceptible to poisons, for in most cases everything in living beings seems disposed to assure them a greater power of resistance.

But some reflection on the final aim of anaphylaxis will give the answer.

It is in fact important that animal species are of determined chemical entity. If, following the hazard of ingestion or injection, alien proteins were found in the cellular juices as part of our humours, then the chemical make-up of beings would be modified and consequently perverted. Crystalloids dialyse through membranes and are speedily eliminated. In a few days, even in a few hours, they are completely gone. Colloids however, that no dialysis can eliminate, do not disappear once they have penetrated to the blood. They fix on cellules and end up by being integral to them.

Grave danger would thus face the animal species, were they not nicely balanced in their hereditary chemical make-up. If heterogenous substances got fixed into our cellules and definitely intermingled with our humours, that would be the end of the chemical constitution of each animal species, which is the fruit of slow evolution down the generations, and all the progress that has been achieved through selection and heredity would be lost.

It does not matter much that the individual becomes more vulnerable in this regard. There is something more important than the salvation of the person and that is integral preservation of the race.

In other words, to formulate the hypothesis in somewhat abstract terms but clear ones all the same: the life of the individual is less important than the stability of the species.

Anaphylaxis, perhaps a sorry matter for the individual, is necessary to the species, often to the detriment of the individual. The individual may perish, it does not matter. The species must at any time keep its organic integrity intact. Anaphylaxis defends the species against the peril of adulteration.

We are so constituted that we can never receive other proteins into the blood than those that have been modified by digestive juices. Every time alien protein penetrates by effraction, the organism suffers and becomes resistant. This resistance lies in increased sensitivity, a sort of revolt against the second parenteral injection which would be fatal. At the first injection, the organism was taken by surprise and did not resist. At the second injection, the organism mans its defences and answers by the anaphylactic shock.

Seen in these terms, anaphylaxis is an universal defence mechanism against the penetration of heterogenous substances in the blood, whence they can not be eliminated.