An Introduction to the Study of Experimental Medicine

there are numberless others which are still totally unknown. I believe that the most useful path for physiology and medicine to follow now is to seek to discover new facts instead of trying to reduce to equations the facts which science already possesses. This does not mean that I condemn the application of mathematics to biological phenomena, because the science will later be established by this alone ; only I am convinced that, since a complete equation is impossible for the moment, qualitative must necessarily precede quantitative study of phenomena.

Physicists and chemists have already often tried to reduce the physico-chemical phenomena of living beings to figures. Among the ancients, as well as among the modems, the most eminent physicists and chemists wished to establish principles of animal mechanics and laws for chemical statistics of animals. Though the progress of physico-chemical science has made these problems more accessible to-day than in the past, it seems to me impossible to reach accurate conclusions at present, because foundations are lacking on which to base our calculations. We may, of course, strike a balance between what a living organism takes in as nourishment and what it gives out in excretions; but the results would be mere statistics incapable of throwing light on the inmost phenomena of nutrition in living beings. According to a Dutch chemist's phrase, this would be like trying to tell what happens inside a house by watching what goes in by the door and what comes out by the chimney. We can accurately fix the extreme terms of nutrition ; but if we afterward try to interpret the intermediary between them, we find ourselves in an unknown region the greater part of which is created by the imagination, and this the more easily because figures often lend themselves marvellously to demonstrating the most diverse hypotheses. Twenty-five years ago, at the outset of my career as a physiologist, I was one of the first, I think, to carry experimentation into the inner environment of the organism, so as to follow experimentally, step by step, all the transformations of substances that chemists explained theoretically. I therefore devised experiments to investigate how sugar, one of the best defined of alimentary substances, is broken down in living beings.

But instead of informing myself about the breaking down of sugar, my experiments led me to discover that sugar is continually produced in animals, no matter what they eat. Moreover, these in- " See the third part of this Introduction. vestigations convinced me that numberless very complex physicochemical phenomena take place, in the organic environment, which give rise to many other products, still unknown to us, with which the chemists do not at all reckon in their static equations. In the chemical statics of life, as well as in the various quantitative estimates of physiological phenomena, certainly neither chemical thinking nor rigor in calculation is lacking; but physiological foundations, which most of the time are false, simply because they are incomplete. We are afterwards led astray all the more easily because we start from an incomplete experimental result and reason without verifying our deductions at every step. Let me cite examples of calculations which I condemn, taking them from works which I nevertheless hold in the highest esteem. In 1852 Bidder and Schmidt of Dorpat published highly important works on digestion and nutrition. Their investigations include excellent and very numerous raw data, but in my opinion the deductions from their calculations are often risky or erroneous. Thus, for example, they took a dog weighing 16 kilograms; in the duct of the submaxillary gland they placed a tube through which the secretion flowed ; and in one hour they obtained 5.640 grams of saliva, from which they concluded that for both glands this should make 11.280 grams. They afterward placed another tube in the duct of the same animal's parotid gland; and in an hour they obtained 8.790 grams of saliva which for both parotid glands would make 17.580 grams. Now, they went on, if we wish to apply these numbers to man, we must take a man weighing 64 kilograms or about four times as much as the dog in question; a calculation based on this ratio consequently gives us, for the man's submaxillary glands, 46 grams of saliva per hour, or 1.082 kilograms per day.

For the parotid glands, we have 70 grams per hour, or 1.687 kilograms per day which reduced one half gives about 1.40 kilograms of saliva secreted in twenty-four hours by the salivary glands of an adult man.^^ As the authors themselves feel, only one thing is true in the above : the crude result found in the dog ; all the calculations deduced from this rest on false or doubtful foundations; first of all, doubling the product of one gland to get the product of both is incorrect, because physiology teaches us that in most cases double glands secrete alter-

" Bidder and Schmidt, Die Verdauungssdfte und der 8toffwechsel. Mittau and Leipzig, 1852. p. 12. nately, and that, when one secretes a great deal, the other secretes less ; then, besides the two submaxillary and parotid salivary glands, there are others which are not mentioned. Next, it is a mistake to believe that multiplying one hour's output of saliva by 24 gives the saliva poured into an animal's mouth in 24 hours. In fact, salivary secretion is highly intermittent and takes place only at meal time or when stimulated ; during the rest of the time, the secretion is nil or insignificant. Finally, the quantity of saliva got from the salivary glands of the dog in this experiment was not absolute; it would have been nil if the mucous membrane of the mouth had not been stimulated ; it might have been greater or less if another stimulant, stronger or weaker than vinegar, had been used.

Now the application of the above calculations to man is still more questionable. If the quantity of saliva had been multiplied by the weight of the salivary glands, a closer relation would have been found ; but I cannot concede the validity of calculating the quantity of saliva from the weight of the body taken as a whole. Estimating a phenomenon in kilograms of the animal's body seems to me wholly incorrect, when all sorts of tissues foreign to the phenomenon in question are included.

In the part of their investigation devoted to nutrition. Bidder and Schmidt described a very notable experiment, perhaps one of the most laborious ever performed. From the point of view of elementary analysis, they kept a balance sheet of everything taken in and given out by a cat during eight days' nourishment and nineteen days' fasting. But this cat was in a physiological condition of which, they were unaware; she was pregnant, and she had her kittens on the seventeenth day of the experiment. In these circumstances, our authors considered the kittens as excretions, and calculated them with other eliminated materials as a simple loss of weight. I believe that these interpretations should be rectified when trying to define such complex phenomena.

In a word, I think that, if figures correspond with reality in these works of chemical statics applied to vital phenomena, it is only by chance or because the experimenters' feeling guides and rearranges the calculation. I repeat, nevertheless, that the criticism which I have just made is not directed against the principle of using calculations in physiology, but against its application under present conditions. I am fortunate, moreover, in being able here to rely on the opinion of the physicists and chemists most competent in such matters. Regnault and Reiset, in their fine work on respiration, express themselves as follows about the calculations used to establish the theory of animal heat: "We have no doubt that animal heat is produced wholly by chemical reactions occurring in the bodily economy; but we think the phenomenon much too complex for possible calculation of the heat from the quantity of oxygen consumed. The substances burned in respiration are generally composed of carbon, hydrogen, nitrogen or oxygen, often in considerable proportions ; when they are completely destroyed in respiration, the oxygen which they contain contributes to the formation of water and carbonic acid ; and the heat liberated is therefore necessarily quite different from what would be produced in burning the supposedly free carbon and hydrogen. These substances, moreover, are not wholly destroyed ; a portion is transformed into other substances which play special parts in the animal economy or escape, in excretions, in the form of highly oxidized materials (urea, uric acid). Now, in all these transformations and in the assimilation of substances taking place in the organs, heat is liberated or absorbed ; but the phenomena are obviously so complex that there is little chance that we shall ever succeed in reducing them to calculation.

It was therefore by a fortuitous circumstance in the experiments of Lavoisier, Dulong and Despretz, that the quantity of heat liberated by an animal was found to be about equal to what the carbon (contained in the carbonic acid produced) and the hydrogen would give off in burning, — the quantity of hydrogen being determined by a quite gratuitous assumption that the quantity of oxygen consumed, but not found in the carbonic acid, had been used in turning the hydrogen into water."

Chemico-physical phenomena of living organisms are therefore still too complex to-day to be embraced as a whole, except by means of hypotheses. To find correct solutions of such vast problems, we must begin by analyzing the results of complicated reactions, and by separating them experimentally into distinct and simple questions. In several attempts which I have made on this analytic path, I have shown that we should not handle the problem of nutrition en bloc,

^ Cf. Regnault and Reiset, Recherches chimiques sur la respiration des animaux des diverses classes. {Annates de chimie et de physique. 3d Series, Vol. XXVI, p. 217.) but rather should first define the nature of the physico-chemical phenomena taking place in an organ made of some definite tissue, such as a muscle, gland or nerve ; that we must at the same time take account of the organ's state of activity or rest. I have also shov^n that v^e can regulate an organ's state of rest or activity at will, by means of its nerves, and that we can even act on it locally withaut reverberation through the organism, if we first separate the peripheral nerves from the nervous centres. When we have analyzed the physicochemical phenomena peculiar to each tissue and each organ, then only can we try to understand nutrition as a whole and to found biochemistry on a solid base, that is to say, on the study of definite, complete and comparable physiological facts.

Another very frequent application of mathematics to biology is the use of averages which, in medicine and physiology, leads, so to speak, necessarily to error. There are doubtless several reasons for this ; but the greatest obstacle to applying calculation to physiological phenomena is still, at bottom, the excessive complexity which pre^ vents their being definite and comparable one with another. By destroying the biological character of phenomena, the use of averages in physiology and medicine usually gives only apparent accuracy to the results. From our point of view, we may distinguish between several kinds of averages: physical averages, chemical averages and physiological and pathological averages. If, for instance, we observe the number of pulsations and the degree of blood pressure by means of the oscillations of a manometer throughout one day, and if we take the average of all our figures to get the true or average blood pressure and to learn the true or average number of pulsations, we shall simply have wrong numbers. In fact, the pulse decreases in number and intensity when we are fasting and increases during digestion or under different influences of movement and rest; all the biological characteristics of the phenomenon disappear in the average. Chemical averages are also often used. If we collect a man's urine during twenty-four hours and mix all this urine to analyze the average, we get an analysis of a urine which simply doe^ not exist; for urine, when fasting, is different from urine during digestion. A startling instance of this kind was invented by a physiologist

" Claude Bernard, Sur le chart gement de couleur du sang dans I'etat de fonction et de repos des glandes. — Analyse du sang des muscles au repos et en contraction. Legons sur les liquides de I'organisme. Paris, 1859. who took urine from a railroad station urinal where people of all nations passed, and who believed he could thus present an analysis of average European urine ! Aside from physical and chemical, there are physiological averages, or what we might call average descriptions of phenomena, which are even more false. Let me assume that a physician collects a great many individual observations of a disease and that he makes an average description of symptoms observed in the individual cases ; he will thus have a description that will never be matched in nature. So in physiology, we must never make average descriptions of experiments, because the true relations of phe^ nomena disappear in the average ; when dealing with complex and variable experiments, we must study their various circumstances, and then present our most perfect experiment as a type, which, however, still stands for true facts. In the cases just considered, averages must therefore be rejected, because they confuse, while aiming to unify, and distort while aiming to simplify. Averages are applicable only to reducing very slightly varying numerical data about clearly defined and absolutely simple cases.

Let me further point out that the reduction of physiological phenomena to an expression in kilograms of body weight is vitiated by many sources of errors. Eor a certain number of years this method has been used by physiologists studying the phenomena of digestion (see p. 131). We observe, for instance, how much oxygen or how much food an animal consumes in a day; we then divide by the animal's weight and get the intake of food or of oxygen per kilogram. This method may also be applied to measure the action of toxic or medicinal materials. We poison an animal with a maximum dose of strychnine or curare, and divide the amount by the weight of the body, to get the amount of poison per kilogram. For greater accuracy in the experiments just cited, we should have to calculate, not per kilogram of the aninaal's body taken as a whole, but per kilogram of blood and of the unit on which the poison acts ; otherwise we could not deduce any direct law from the reductions. But other conditions would still remain to be established similarly by experiment, conditions varying with age, height, state of digestion, etc. ; in these measures, physiological conditions should always hold first rank.

To sum up, every possible application of calculation would be excellent if the physiological conditions were quite accurately defined. Physiologists and physicians should therefore concentrate their effort, for the moment, on defining these conditions. We naust first accurately define the conditions of each phenomenon; this is true biological accuracy, and, without this preliminary study, all numerical data are inaccurate, and the more inaccurate because they include figures which mislead and impose on us by a false appearance of accuracy.

As for statistics, they are given a great role in medicine, and they therefore raise a medical question which we should examine here. The first requirement in using statistics is that the facts treated shall be reduced to comparable units. Now this is very often not the case in medicine. Everyone familiar with hospitals knows what errors may mark the definitions on which statistics are based. The names of diseases are very often given haphazard, either because the diagnosis is obscure, or because the cause of death is carelessly recorded by a student who has not seen the patient, or by an employee unfamiliar with medicine. For this reason pathological statistics can be valid only when compiled from data collected by the statistician himself. But even then, no two patients are ever exactly alike; their age, sex, temperament and any number of other circumstances involve diiferences, with the result that the average, or the relation deduced from our comparison of facts, may always be contested. But I cannot accept even the hypothesis that facts can ever be absolutely alike and comparable in statistics ; they must necessarily differ at some point, for statistics would otherwise lead to absolute scientific results, while they can actually show only probability, never certainty. I acknowledge my inability to understand why results taken from statistics are called laws; for in my opinion scientific law can be based only on certainty, on absolute determinism, not on probability. I should stray from my subject, if I went into all possible explanation of the value of statistical methods based on the calculus of probabilities ; yet I cannot but say here what I think about the application of statistics to physiological science in general and to medicine in particular.

In every science, we must recognize two classes of phenomena, first, those whose cause is already defined; next, those whose cause is still undefined. With phenomena whose cause is defined, statistics have nothing to do ; they would even be absurd. As soon as the circumstances of an experiment are well known, we stop gathering statistics : we should not gather cases to learn how often water is made of oxygen and hydrogen ; or when cutting the sciatic nerve, to learn how often the muscles to which it leads will be paralyzed. The effect will occur always without exception, because the cause of the phenomena is accurately defined. Only when a phenomenon includes conditions as yet undefined, can we compile statistics ; wemust learn, therefore, that we compile statistics only when we cannot possibly help it ; for in my opinion statistics can never yield scientific truth, and therefore cannot establish any final scientific method. A single example will illustrate my meaning. Certain experimenters, as we shall later see, published experiments by which they found that the anterior spinal roots are insensitive; other experimenters published experiments by which they found that the same roots were sensitive. These cases seemed as comparable as possible ; here was the same operation done by the same method on the same spinal roots. Should we therefore have counted the positive and negative cases and said : the law is that anterior roots are sensitive, for instance, 25 times out of a 100 ? Or should we have admitted, according to the theory called the law of large numbers, that in an immense number of experiments we should find the roots equally often sensitive and insensitive ? Such statistics would be ridiculous, for there is a reason for the roots being insensitive and another reason for their being sensitive ; this reason had to be defined ; I looked for it, and I found it; so that we can now say: the spinal roots are always sensitive in given conditions, and always insensitive in other equally definite conditions.

I will cite still another example borrowed from surgery. A great surgeon performs operations for stone by a single method; later he makes a statistical summary of deaths and recoveries, and he concludes from these statistics that the mortality law for this operation is two out of five. Well, I say that this ratio means literally nothing scientifically and gives us no certainty in performing the next operation; for we do not know whether the next case will be among the recoveries or the deaths. What really should be done, instead of gathering facts empirically, is to study them more accurately, each in its special determinism. We must study cases of death with great care and try to discover in them the cause of mortal accidents, so as to master the cause and avoid the accidents. Thus, if we accurately know the cause of recovery and the cause of death, we shall always have a recovery in a definite case. We cannot, indeed, admit

that cases with different endings were identical at every point. In the patient who succumbed, the cause of death was evidently something which was not found in the patient who recovered ; this something we must determine, and then we can act on the phenomena or recognize and foresee them accurately. But not by statistics shall we succeed in this ; never have statistics taught anything, and never can they teach anything about the nature of phenomena. I shall further apply what I have just said to all the statistics compiled with the object of learning tbe efficacy of certain remedies in curing diseases. Aside from our inability to enumerate the sick who recover of themselves in spite of a remedy, statistics teach absolutely nothing about the mode of action of medicine nor the mechanics of cure in those in whom the remedy may have taken effect.

It is said that coincidence may play so large a part in causes of statistical errors, that we should base conclusions only on large numbers. But physicians have nothing to do with what is called the law of large numhers, a law which, according to a great mathematician's expression, is always true in general and false in particular. This amounts to saying that the law of large numbers never teaches us anything about any particular case. What a physician needs to know is whether his patient will recover, and only the search for scientific determinism can lead to this knowledge. I do not understand how we can teach practical and exact science on the basis of statistics. The results of statistics, even statistics of large numbers, seem indeed to show that some compensation in the variations of phenomena leads to a law ; but as this compensation is indefinite, even the mathematicians confess that it can never teach us anything about any particular case ; for they admit that if the red ball comes out fifty times in succession, that is no reason why a white ball would be more likely to come out the fifty-first time.

Statistics can therefore bring to birth only conjectural sciences; they can never produce active experimental sciences, i.e., sciences which regulate phenomena according to definite laws. By statistics, we get a conjecture of greater or less probability about a given case, but never any certainty, never any absolute determinism. Of course, statistics may guide a physician's prognosis ; to that extent they are useful. I do not therefore reject the use of statistics in medicine, but I condemn not trying to get beyond them and believing in statistics as the foundation of medical science. This false idea leads cer-

tain physicians to believe that medicine cannot but be conjectural; and from this, they infer that physicians are artists who must make up for the indeterminism of particular cases by medical tact. Against these anti-scientific ideas we must protest with all our power, because they help to hold medicine back in the lowly state in which it has been so long. All sciences necessarily began by being conjectural ; even to-day science has its conjectural parts. Medicine is still almost wholly conjectural. I do not deny it; I only mean to say that modern medical science must exert itself to get out of the temporary condition which is no more a final scientific state for medicine than for any other science. The scientific state will be harder to reach and will take longer to establish in medicine, because of the complexity of the phenomena; but the goal of scientific physicians in their own science, as in the rest, is to reduce the indeterminate to the determinate. Statistics therefore apply only to cases in which the cause of the facts observed is still indeterminate. In these circumstances, statistics in my opinion can serve only to guide the observer toward investigation of the indeterminate cause, but they can never lead to any real law. I emphasize this point, because many physicians have great confidence in statistics when based on wellobserved facts which they consider mutually comparable, and they believe that such statistics may lead to knowledge of phenomenal law. I have already said that facts are never identical ; therefore statistics are simply an empirical enumeration of observations.

In a word, if based on statistics, medicine can never be anything but a conjectural science; only by basing itself on experimental determinism can it become a true science, i.e., a sure science. I think of this idea as the pivot of experimental medicine, and in this respect experimental physicians take a wholly different point of view from so-called observing physicians. Indeed, if a phenomenon appears just once in a certain aspect, we are justified in holding that, in the same conditions, it must always appear in the same way. If, then, it differs in behavior, the conditions must be different. But indeterminism knows no laws; laws exist only in experimental determinism, and without laws there can be no science. Most physicians seem to believe that, in medicine, laws are elastic and indefinite. These are false ideas which must disappear if we mean to found a scientific medicine. As a science, medicine necessarily has definite and precise laws which, like those of all the sciences, are derived from

the criterion of experiment. To the explanation of these ideas I shall especially devote the work which I have named Principles of Experimental Medicine, in order to show that the principles of experimental determinism must be applied to medicine, if it is to become an exact science founded on experimental determinism, instead of remaining a conjectural science based on statistics. A conjectural science may indeed rest on the indeterminate; but an experimental science accepts only determinate or determinable phenomena.

Only determinism in an experiment yields absolute law ; and he who knows the true law is no longer free to see a phenomenon otherwise. The indeterminism of statistics leaves to thought a certain liberty limited by the numbers themselves; and in this sense philosophers were able to say that liberty begins where determinism ends. But when determinism increases, statistics can no longer grasp and confine it within a limit of variations. There we leave science, for we are forced to invoke chance or an occult cause to regulate phenomena. We shall certainly never reach absolute determinism in everything; man could no longer exist. There will always be some indeterminism then, in all the sciences, and more in medicine than in any other. But man's intellectual conquest consists in lessening and driving back indeterminism in proportion as he gains ground for determinism by the help of the experimental method. This alone should satisfy his ambition, for by this alone is he extending, and can he further extend, his power over nature.

X. The Physiologist's Laboratory and Various Methods Necessary to the Study of Experimental Medicine Every experimental science requires a laboratory. There the man of science withdraws, and by means of experimental analysis tries to understand phenomena that he has observed in nature. A physician's subject of study is necessarily the patient, and his first field for observation is the hospital. But if clinical observation teaches him to know the form and course of diseases, it cannot suffice to make him understand their nature; to this end he must penetrate into the body to find which of the internal parts are injured in their functions. That is why dissection of cadavers and microscopic study of diseases were soon added to clinical observa-

tion. But to-day these various methods no longer suffice; we must push investigation further and, in analyzing the elementary phenomena of organic bodies, must compare normal with abnormal states. We showed elsewhere how incapable is anatomy alone to take account of vital phenomena, and we saw that we must add study of all physico-chemical conditions which contribute necessary elements to normal or pathological manifestations of life. This simple suggestion already makes us feel that the laboratory of a physiologist-physician must be the most complicated of all laboratories, because he has to experiment with phenomena of life which are the most complex of all natural phenomena.

Libraries may also be considered as part of the laboratory of a man of science or experimenting physician. But this is on condition that he shall read the observations, experiments and theories of his predecessors in order to know them and verify them in nature, and not to find opinions ready-made in books, thus saving himself the trouble of working and of trying to further the investigation of natural phenomena. Misconceived erudition has been, and still is, one of the greatest obstacles to the advancement of experimental science. Thus erudition, setting man's authority in the place of facts., halted science through several centuries at Galen's ideas, without any one's daring to touch them; and this scientific superstition was such that Mundini and Vesalius, who first contradicted Galen by confronting his opinions with animal dissections, were considered innovators and revolutionaries. Yet such should always be the practice of scientific erudition. It should always be accompanied by critical investigations of nature, planned to verify the facts about which we speak, and to decide the opinions which we discuss. In this way science in advancing would be simplified and cleansed by sound experimental criticism, instead of being encumbered by exhuming an accumulation of numberless facts and opinions among which it is soon impossible to distinguish falsehood from truth. It would be out of place for me here to say more of the mistakes and misdirection of most of the studies of medical literature, characterized as historical or philosophical. I may perhaps have occasion to explain myself elsewhere on this subject ; for the moment, I shall limit myself to saying that, in my opinion, all these mistakes have their origin in a perpetual confusion between literary or artistic production and scientific production, between criticism of art and

scientific criticism, between the history of science and the history of men. Literary and artistic productions never grow old, in this sense, that they are expressions of feeling, changeless as human nature. We may add that philosophical ideas stand for aspirations of the human spirit which are also of all time. But science, which stands for what man has learned, is essentially mobile in expression; it varies and perfects itself in proportion to the increase of acquired knowledge. Present day science is therefore necessarily higher than the science of the past; and there is no sort of reason for going in search of any addition to modern science through knowledge of the ancients. Their theories, necessarily false because they do not include facts discovered since then, can be of no real advantage to contemporary science. No experimental science, then, can make progress except by advancing and pursuing its work in the future. It would be absurd to believe that we should go in search of it in the study of books bequeathed to us by the past. We can find there only the history of the human mind, which is quite another matter.

We must of course be familiar with what we call scientific literature, and know what our predecessors have done. But scientific criticism in the literary manner can be of no possible use to science. Indeed, we need not ourselves be poets or artists to judge literary or artistic work, but this is not true of experimental science. We cannot judge of a memoir on chemistry without being chemists nor of a memoir on physiology if we are not physiologists. In deciding between two different scientific opinions, it is not enough to be a good philologist or a good translator, we must above all be deeply versed in technical science; we must even be masters of the special science and ourselves be able to experiment and do better than the men whose opinions we discuss. Some time ago I discussed an anatomical question concerning the anastomoses of the pneumogastric and spinal nerves.^* Willis, Scarpa and Bischoif had expressed different and even opposite opinions on this subject. A mere scholar could only have quoted these various opinions and more or less correctly compared the texts; that would not have answered the scientific question. It was therefore necessary to dissect and