Every living system possesses, as we know, a peculiar and characteristic manner of reacting to stimulation. The muscle responds with a contraction, the salivary cell with production of saliva, the luminous cell with the emission of light. This is the specific energy in the sense of Johannes Müller. Every living system is likewise characterized by a certain degree of irritability, which can be expressed by the threshold value of the stimulus at which the specific reaction is just perceptible. This degree of irritability, by which the system concerned is distinguished, may be termed its specific irritability.

From the standpoint of the conditional method of investigation it is at once apparent that specific energy, as well as specific irritability, must be solely determined by the specific conditions existing in the particular system. It follows from this that every alteration in the conditions of the system, that is, every change of its state, likewise entails a corresponding alteration of its specific energy and its specific irritability. It is, therefore, self-evident that the alteration of the state, which is undergone by the living system in the process of excitation, brings about an alteration of its specific irritability. Likewise as the original state of the system is restored by the metabolic self-regulation after the course of an excitation, the specific irritability of the system must be reestablished. The specific irritability is, therefore, a property of the living system, which, like the metabolic equilibrium, undergoes restitution by the process of self-regulation after variation produced by a stimulus of any kind. It is scarcely necessary to repeat each time that this is only applicable within the physiological variations and for a limited period, during which the alterations in development need not be considered.

These alterations of the specific irritability following an excitation and their compensation through the metabolic self-regulation will now claim our attention.

That the specific irritability of a living system undergoes a diminution as the result of a stimulus of long duration has been long known through the study of fatigue. This is especially so with frequently recurring excitating stimuli. It is only within the last decade, however, that the observation has been made in a few instances that a single momentary excitation is likewise followed by such a reduction of the specific irritability. But that this is a fact of general physiological fundamental importance for the whole field of response to stimulation in the living substance has only been recognized within the last few years.

In 1876 Marey119 found that the irritability of the heart in response to artificial stimulation was greatly reduced during the systole, and that recovery took place during the following diastole. (Figure 29.) This fact was already apparent from the observations made by Bowditch120 and Kronecker,121 that by stimulation of the isolated frog’s heart with single induction shocks, an artificial systole can only be produced with certainty when the stimuli succeed each other at certain intervals, which must be the longer as the strength of the stimulation is weaker. Marey calls this period of reduced irritability “phase réfractaire” of the heart. The refractory period of the heart has been made the subject of a great number of investigations, especially by Engelmann and his pupils. It was Engelmann122 especially who determined more exactly the duration of the course of the refractory period. He found, namely, that irritability disappears immediately before each systole and reappears shortly before the beginning of the diastole, and again reaches its original height at the end of the diastole. For a long time, however, this refractory period was looked upon as a special peculiarity of the heart. It was not until Broca and Richet,123 twenty years after Marey’s investigations, discovered an analogous refractory period for the motor centers of the cerebral cortex of the dog. They first made this observation on a dog affected with chorea, in which the choreic movements rhythmically occurred in intervals of one second. They found that after each movement electrical stimulation of the cortex remained without result for about .5 seconds. During the next .25 seconds stimulation was followed by a weak response and it was not until the last .25 seconds before the next movement that a strong effect was produced. They also found in the normal dog a refractory period after every artificial stimulation equal to .1 second, so that the number of contractions brought about by rhythmical electrical stimulation were only ten per second. Following this, numerous other investigations of the refractory period have been made on the central nervous system. Zwaardemaker124 and Lans have observed a refractory period in the eyelid reflex of the human being which, on stimulation of the optic nerve, amounts to about .5–1 second; on the stimulation of the trigeminus produced by blowing on the cornea on the other hand, it is somewhat shorter, less than .25 seconds. Zwaardemaker125 also was able to demonstrate an analogous refractory period for the swallowing reflex of the cat. Further a refractory period was found and closely analyzed by Verworn126 for the reflexes in the spinal cord of the strychninized frog. Dodge127 found a refractory period in the knee jerk reflex of man. Gotch and Burch128 showed, by two induction shocks following each other in quick succession, a refractory period of the nerve, which is characterized by its extremely brief duration. They found, depending upon the temperature, a period of nonirritability of .001–.008 seconds after every stimulus. The investigations of Miss Buchanan129 lead us to conclude that there is a refractory period for the cross striated skeletal muscle. Miss Buchanan stimulated the muscle at times through the nerve, at other times directly after elimination of the nervous element, with very frequent electrical stimuli (about 1000 in the second) and found by means of the capillary electrometer a rhythmical reaction of the muscle of about 50–100 excitation shocks per second. Likewise the Ritter tetanus produced by the breaking of an increasing current proved to be a rhythmical reaction of an analogous nature. In a more direct manner Keith Lucas130 has determined the refractory stage for the musculus sartorius of the frog. He allowed two induction shocks to act successively on the muscle at intervals of varied duration and then registered the action currents by means of the capillary electrometer. He then found that the second stimulus was ineffective for about .005 seconds after the application of the first stimulus. If the second stimulus follows somewhat later, it produces a contraction which is weaker and has a longer latent period the nearer the second stimulus approaches the first in point of time. (Figure 30.) Massart131 and Jennings132 likewise observed the existence of a refractory period for the myoids of unicellular organisms brought about by mechanical stimuli. Massart attributes this cessation of reaction to stimuli following each other at certain intervals, to fatigue, an explanation which has been disputed by Jennings as the result of his investigations made on Stentor and Vorticella. Jennings looks upon the behavior of the infusoria rather as an “adaptation” to the stimulus. Pütter was the first to see in this the existence of a refractory period. His experiments on Spirostomun ambiguum in 1900 showed a refractory period in the reaction to rhythmical mechanical stimuli. I wish to state, however, that these observations of Pütter have not as yet been published. Thus the existence of a refractory period has even today been proved for a whole series of very different kinds of substances.

We will now examine the alterations of irritability which are perceptible during the refractory period to complete restitution of the specific irritability of the particular system, and endeavor by the analysis of their special conditions to render them comprehensible from a physical standpoint of view.

The first fact to take into consideration is, that, as is shown in the heart, the refractory period begins at the moment of the appearance of the systolic excitation. The irritability of the heart is absent and remains so until the excitation has reached its highest point, that is, shortly before the beginning of the diastole. From this point the restitution of irritability begins, which does not reach the maximum until the end of the diastole. In other words: irritability undergoes the greatest reduction by disintegration produced by the stimulus and is restored by the metabolic self-regulation following the decomposition.

This point of view enables us to interpret this state from a physical standpoint. In this discussion on the relations between irritability and the extension of excitation, I have taken the amount of energy which is produced during the time unit and space unit in a living system as the general standard for the degree of irritability, at the same time duly regarding the individual components involved. This amount of energy is determined in a given system by the quantity of substance broken down by a stimulus of a given intensity. It is, therefore, clear that during the time in which an increased disintegration produced by a stimulus takes place, the irritability in response to a second stimulus must be reduced, as during this period the second stimulus has less of necessary decomposable substances at its disposal, and at the same time there are more products of disintegration in a given space. If a living organism is the subject of consideration, to which the “all or none law” is applicable, as, for instance, the heart at the moment of the beginning of excitation, irritability is completely obliterated, as shown by the fact that the second stimulus of any strength remains without response, for during the excitation there is a complete breaking down of all the substances capable of decomposition. If, on the contrary, a system is the subject of observation, for which the “all or none law” is not valid, then irritability is merely reduced but not wholly obliterated during an excitation, and whether or not a response is obtained to the stimulus depends upon its strength. To impress the relations between the degree of irritability and the intensity of the stimulus, I have, therefore, employed the term “relative refractory period” in contrast to the “absolute refractory period,” in which irritability is obliterated even for the strongest stimuli. It is self-evident that irritability must again increase in the same degree as the restitution of the living system by metabolic self-regulation takes place, for the more molecules capable of disintegrating are restored and the more products of disintegration removed, the more molecules necessary for decomposition in the unit of space are attacked and broken down by the stimulus. All these are self-evident facts which are in accordance with the conception we have here developed of the course of the process of excitation and its physical nature. But another important point is evolved from the observations we have made of the nature of the process of self-regulation. The process of self-regulation is founded on the same principle as that which governs the taking place of all chemical equilibrium, for metabolic equilibrium is merely a special kind of a chemical equilibrium. The development of a chemical equilibrium between reacting substances and reaction products has, as known, a characteristic course in regard to its duration. If the rapidity with which the equilibrium is reached is expressed by a curve in which the abscissa represents the time, while the ordinates signify the number of contacts of the interacting molecules, the rapidity of reaction is altered with the approach to the equilibrium in the form of a logarithmic curve; that is, the approach to the state of equilibrium, which is represented by ordinate value zero, takes place at first very rapidly, then with more and more decreasing speed, for with the decrease of the number of reacting molecules and the increase of the amount of products of reaction, the contact of the interacting molecules and with this the opportunity for the reaction occurs less and less frequently. Although the self-regulation of metabolic equilibrium is by no means such a simple process as, for instance, that of the well-known example of the forming of ethylester from acetic acid and æthyl alcohol, we have still in every case to deal with the taking place of a chemical mass equilibrium. Hence the progress to the metabolic equilibrium must likewise correspond with a logarithmic curve, i.e., restitution after a disturbance of the equilibrium must take place at first rapidly, then at a constantly decreasing rate. For reasons readily to be understood the special form of this restitution curve has so far not been accurately ascertained for any kind of living substance. Even in those cases where the restitution occurs very slowly we meet with the difficulty that, when the tests are applied which are necessary to determine the restitution at different intervals, with each testing stimulus irritability is each time reduced. Hence the construction of the restitution curve can only be achieved by indirect means, and we must content ourselves with the ascertainment of a smaller number of its points from which by interpolation its form can be constructed. Indeed in this connection a certain number of results have already been gained quite sufficient to experimentally confirm the correctness of these types of curves, primarily obtained by purely theoretical deductions. That irritability very gradually reaches its maximal height has been already shown, as previously mentioned by Bowditch133 in his investigations on the influence of rhythmical induction shocks on the apex of the heart of the frog. He found that in order to produce response, the weaker the stimuli the longer must be the intervals between them. It follows from this, that after a discharge the irritability in response to strong stimuli reappears more rapidly than for weak, i.e., that they only gradually regain their maximum. The exact periods of time for the course of the return of irritability for the heart have unfortunately not been so far ascertained. On the other hand, the investigations of Ishikawa134 furnish the material for the construction of the restitution curve for the centers of the spinal cord of the frog. Ishikawa did not employ the threshold of stimulation as an indicator for the course of restitution, but used instead the duration of the reflex time following on a stimulus of a certain strength. The reflex time is greatly prolonged after an excitation of extended duration and only regains its normal value in the same degree as restitution takes place. By a great number of painstaking experiments Ishikawa ascertained the duration of the reflex time at intervals of thirty seconds to one minute, and obtained figures which show that restitution does actually take place, at first rapidly and then with constantly decreasing speed. The detailed study of the course of self-regulation of the individual forms of living substance will doubtless be more exactly determined in the near future. But even at the present we are fully justified in describing the form of restitution curve as a logarithmic in type. Therefore, a relative refractory period must be present in every metabolic self-regulation after an excitation, during which stronger stimuli produce response, while weaker are still without result. This is a fact which, as we shall see later, is of fundamental importance for the comprehension of the various kinds of interference responses to stimuli.

From the information here gained on the nature and origin of the refractory period the conclusion must inevitably be drawn that in all living substance there must exist, directly following an excitation, a period of time in which its irritability is reduced, that is, under proper conditions a refractory period can be demonstrated for every living organism. Every living system possessing irritability undergoes a period of reduced irritability at the time of and subsequent to every excitation, for every excitation momentarily decreases the amount of products capable of disintegration and increases the disintegration products in the unit of space. As restitution involves time, a stimulus occurring in the phase preceding complete restitution cannot break down the same quantity of molecules as would be the case after the establishment of complete restitution, that is, the response is weaker, the irritability is decreased. The refractory period during and subsequent to excitation is as much a general property of the living substance as irritability and metabolic self-regulation.

This conclusion appears so self-evident that it would seem hardly to call for emphasis were it not that even at the present time the view is still widely held that the refractory period is a special characteristic of certain forms of living substance. This assumption is explained on the one hand by the fact that our information concerning the refractory period is still of comparatively recent date and that few physiologists are in the habit of connecting special observations with general physiological conceptions, but also for the reason that some investigators have vainly tried to find a refractory period in certain forms of living substance. Langendorff and Winterstein,135 for instance, have not succeeded in proving a refractory period for the spinal cord of the frog. Langendorff stimulated the central sciatic stump with two stimuli in quick succession and used the contractions of the triceps as indicator of the response. He found that when the stimuli, if consisting in either single induction shocks or faradic shocks, followed each other even at intervals of .004 seconds the second stimulus was still operative, this being perceptible in an increase of the contraction or with greater intervals of time in a summation of two contractions. Winterstein concludes from this that the development of a refractory period after a stimulation is not a general property of all nerve centers. If the experiments of Langendorff failed to show the presence of a refractory period it is not for the reason that this does not take place in the centers of the spinal cord but rather results from the fact that the conditions for the investigation were not suited for its demonstration. In fact, Fröhlich136 and especially Vészi137 have incontestably proved the existence of relative refractory periods in the normal spinal cord.

If the existence of the refractory period is based on the fact that during the time of and subsequent to an excitation the quantity of substances necessary for disintegration is decreased and that of the breaking down products increased, and if it is limited by the restitution of the substances required for decomposition and the elimination of the disintegration products, its duration must be dependent upon the length of these processes. All factors which lessen the decomposition and hasten the metabolic self-regulation must, therefore, shorten its duration. This is completely confirmed by experimental investigations. As can be understood, the factors of special interest for us are those which influence the duration of the refractory period in the physiological occurrences of the organism.

One of these factors is temperature. As we know, the rapidity of chemical reactions increases with ascending and decreases with falling temperature. As in the disintegration as well as in the restitution, processes are chemical in nature, it is to be expected that the duration of the refractory period is influenced in like manner by temperature. Indeed, Kronecker138 found some time ago that in the isolated frog’s heart a much more frequent rhythm of stimulation is effective at a higher than at a lower temperature. When the heart is stimulated at a temperature of 11–12° C. with twelve rhythmical induction shocks in the second, every stimulus is operative and produces a systole. If a stimulus of the same frequency is used at a temperature of 5° C., the heart responds merely to every second stimulus. This shows that the refractory period is of longer duration at a lower than at a higher temperature.

A factor of particular interest is the supply of oxygen, for we know its fundamental importance in all aërobic organisms in the breaking down of the living substance. The life of these organisms is primarily dependent upon the supply of oxygen from without. Organic reserve substances for restitution after disintegration are contained in ample quantity in the reserve stores in the living cell substance, whereas oxygen is present in very small quantities in relation to the former. It is, therefore, self-evident that the rapidity of the breaking down processes is very closely dependent upon the amount of available oxygen at hand. Nevertheless it is not the absolute quantity but the relative amount of oxygen in relation to the momentary requirement which is of importance. For instance, the quantity of oxygen present may completely suffice for the oxydative disintegration in the metabolism of rest or at lower temperature, whereas the same amount would be much too small to meet the demand increased by excitation or at higher temperature. In the latter case “a relative deficiency of oxygen” occurs. I have introduced the term “relative deficiency of oxygen”139 for I have found that a number of authors by neglecting the relations of the available oxygen to that which is required at the moment have been led to false conclusions. There is no living object so preëminently fitted to demonstrate in such a striking manner the dependence of the duration of the refractory period upon the supply of oxygen as the spinal cord centers of the frog, when their irritability has been increased to the maximum by strychnine.140 Various observers, such as Loven, Buchanan, H. von Baeyer and others, investigated the action current by the capillary electrometer. As a means of studying the number of impulses in the strychnine tetanus, we can upon the basis of their figures roughly assume the number of impulses to equal ten per second at room temperature. In short, in the freshly strychninized frog the duration of the refractory period is about .1 second. By means of the method of artificial circulation already mentioned a deficiency of oxygen can readily be brought about. It has been demonstrated that the rhythmic in contrast to the continuous method of introduction of circulatory fluid is superior in that the former reproduces more closely the natural conditions of the circulation of the blood and renders the smallest capillaries more permeable. In consequence I have recently constructed a small appliance for artificial circulation, which accomplishes this in a manner as simple as it is complete. (Figure 31.)

The fluid flows from a vessel, E, provided with an outlet tube through a thin rubber tube into a glass canula, which is introduced into the general aorta of the frog, F. The tube is automatically occluded by the rhythmical movement of the armative of an electromagnet, D, produced by a metronome, B. The pressure of the circulating fluid can be readily changed at will by varying the level of the vessel and the frequency of the pulse by the rhythm of the metronome, which makes and breaks the current to the electromagnet.141 In this way it is possible to artificially replace the normal circulation with satisfactory exactitude and substitute for the blood, circulating in the vessels of the frog, any desired fluid. If the entire quantity of blood of a frog is displaced by a continuous stream of oxygen-free saline solution and a weak strychnine solution is injected with a Pravaz syringe, a violent strychnine tetanus appears after the lapse of a few seconds. (Figure 32, A.) If the artificial circulation with oxygen-free saline solution is now contained in the rhythm of the natural heart beat, the further reactions can then be readily observed. The first long-continued tetanic attack, which can be produced by a slight touch of the skin, is followed by a whole series of tetanic convulsions of prolonged duration, which are repeatedly followed by periods of exhaustion. I wish to emphasize this fact once more, as it appears to me as not without interest for the understanding of the question of reserve substances.

If we assume that at the moment when the entire amount of blood is removed from the vascular system, no oxygen remains in the cells of the spinal cord and muscle, then disintegration of the living substance could from this instant take place exclusively anoxydatively, and there would be no further oxydative breaking down into carbon dioxide and water. The energy production compared in equal number of molecules, taking the figures of Lesser for the fermentation of sugar, would approximately amount to about 3.8 per cent. of that of the energy production in the oxydative disintegration of dextrose into carbon dioxide and water. In reality, however, the tetanic convulsions are at first exactly as violent as in the frog with a normal circulation. There simply remains the assumption, therefore, that either the disintegration as soon as it becomes anoxydative involves relatively greater number of molecules than would be the case if it were oxydative in nature, or to suppose that even after the complete displacement of the blood a certain, though relatively small, amount of oxygen is present in the cells which for a short time suffices for the taking place of oxydative disintegration and with this an almost maximal production of energy which naturally decreases as the oxygen is consumed. It seems to me that the latter supposition contains more probability than the first. To return, however, from this observation to a further consideration of the animal we are studying, we see how the complete tetanic convulsions in the refractory period which we assumed to be .1 second are gradually transformed into incomplete tetanus. After a time the tetanic convulsions become shorter after each stimulus (Figure 32, B) and permit us to distinguish their individual movements, even though the latter at first succeed each other still very rapidly. Gradually this incomplete tetanic convulsion assumes the form of a short series of individual contractions, distinctly separated from each other and soon a stage is reached in which each reaction to a peripheral stimulus consists merely in a single contraction. (Figure 32, C.) The refractory period is, however, even now less than a second. Nevertheless, with a further continuation of the experiment, the refractory period becomes more and more prolonged, so that stimuli succeeding each other at intervals of less than a second are without effect. It is possible at this stage, as Tiedemann142 did, to graphically record the reactions. He severed the sciatic nerve on one side and stimulated its central stump, at the same time connecting the triceps with a writing lever. It is then found that when the single induction shocks follow each other at intervals of a second or more every stimulus produces a contraction, but that on the contrary only the first stimulus of a rhythmical series is operative and all those succeeding ineffectual, if the stimuli follow each other at shorter intervals. The refractory period becomes, however, more and more prolonged. The rhythm of the stimulus must become continually slower if each individual stimulus is to remain effective. If the rhythm is even slightly too rapid only the first few stimuli of a rhythmical series are effective and this with decreasing response and later no contraction at all is observed. With a further continuance of the experiment, the stimuli are only effective when following each other at long intervals. It is necessary that a period of recovery lasting several seconds must take place before the following stimulus can meet with response. (Figure 33.) The refractory period can gradually be prolonged for the space of a minute or longer, until finally irritability does not reappear at all, and even the strongest stimuli fail to produce the least contraction. The continuous manner in which the refractory period is, in the absence of oxygen, more and more prolonged until eventually a prolonged state of nonirritability is developed, can be better followed by observing the experiment than when described in words. If at this stage instead of the oxygen-free saline solution, defibered blood of the ox shaken in air or a saline solution saturated with oxygen is circulated in the frog, restitution is often within a few minutes so complete that tetanic attacks are once more produced by a single stimulus, that is, the refractory period has from being practically nil returned to the normal. This experiment can be repeated several times on the same animal. It is invariably found that the refractory period is prolonged by the withdrawal of oxygen and shortened with a renewed supply.

I have described this experiment somewhat in detail as it contains facts which are the key for the comprehension of a general physiological process of paramount importance. I refer to fatigue. The refractory period and fatigue are inseparably connected, for fatigue is founded on the existence of the refractory period and is an expression of prolongation of the former, brought about by want of oxygen. This is shown at once by closer analysis. It is here necessary to differentiate somewhat more in detail the factors which bring about the prolongation of the refractory period in deficiency of oxygen.

If we first turn our attention to the normal refractory period which occurs in a system in metabolic equilibrium of rest in direct connection with dissimilatory excitation, following a momentary stimulus, we find that reduction of irritability or, more exactly expressed, the lessening of the response is, as we have seen, determined by the time involved in the metabolic decomposition and recovery. Both these processes require time and until their completion the quantity of substance demanded for the oxydative disintegration is decreased in a given space, and every stimulus must consequently be followed by a weaker response. Our conceptions of the physical details of these processes depend essentially upon the question, if the oxydative disintegration itself in the given living system occurs in one single phase, in that the oxygen is the activator for the oxydative splitting up of the carbon chain, or if this takes place in two periods, in which the carbon chain is first anoxydatively split up into larger fragments by the stimulus, which are then seized upon by the oxygen to be split up into carbon dioxide and water. As we have seen, this question must remain for the present undecided as far as the metabolism of rest as well as the excitation produced by a single momentary stimulus is concerned. It is highly probable that a uniformity of the process for all living systems does not exist. We are, therefore, not justified in assuming that these special chemical processes resulting from single stimuli are uniform throughout the refractory period.

On the contrary it is different in the case of oxygen deficiency. Here we see with increasing want of oxygen a constantly increasing duration of the refractory period, a prolongation which may be attributed to the retardation of the oxydative disintegration. It is necessary, however, that we now study more clearly these alterations brought about by the deficiency of oxygen.

If we follow the course of the changes from that of the normal state of equilibrium of metabolism, wherein oxygen is sufficient to bring about complete disintegration of the molecules to the formation of carbon dioxide and water, we must assume in spite of the great explosive rapidity of this process on the basis of our chemical knowledge, that first a series of intermediate products are produced before finally the end products are formed. In this way the oxydative disintegration produced by a stimulus becomes more and more prolonged by an increasing want of oxygen. If, as I have previously suggested, the amount of energy which is liberated in a given space and time by an excitating stimulus is taken as a standard of irritability, it is apparent that the more the oxydative disintegration following a stimulus is retarded, the greater must be the decrease in irritability. The less oxygen there is at disposal and the more incomplete the oxydative breaking down, the smaller is the degree of irritability, the weaker the response and the slower the return of irritability after every stimulus. In other words, with the increasing deficiency of oxygen, the response is not merely reduced for every stimulus, but the duration of the refractory period is likewise progressively prolonged until finally with an absolute want of oxygen, constant and complete depression takes place. In the genesis of this process another factor, however, has the same effect.

While with a sufficient supply of oxygen disintegration leads to the formation of carbon dioxide and water, therefore to end products, which can quickly and easily be removed by diffusion, the want of oxygen produces complex products of incomplete combustion and finally of anoxydative decomposition, such as lactic acid, fatty acids and even more complex substances in constantly increasing quantities. These products permeate the protoplasmic surfaces with great difficulty, if at all, and as they cannot subsequently be oxydatively split up, constantly accumulate. These asphyxiation substances, as they may be briefly termed, produce a depressing effect on further disintegration. This can be experimentally demonstrated.

For this purpose I have modified the experiment previously described in the way that after every introduction into the blood of oxygen-free saline solution and after the injection of strychnine, the artificial circulation was stopped so that stagnation of the oxygen-free saline solution took place in the vascular system. The processes then occurred in exactly the same manner with the exception that the state of non-irritability appeared somewhat earlier. If after the beginning of complete depression artificial circulation with oxygen-free saline solution was again started, a certain degree of recovery took place within one or more minutes. The stimuli were once more effective and produced a number of contractions. At times, several single contractions, following each other in more or less quick succession, could be brought about. But complete recovery or the appearance of even incomplete tetanic convulsions was never again obtained, whereas by the introduction of oxygen complete recovery could at once be brought about. If, however, the circulation with oxygen-free saline solution was continued, irritability gradually decreased. The refractory periods after the individual stimuli became longer, and in spite of continuous artificial circulation irritability again disappeared. The experiment shows that by the circulation of oxygen-free solution irritability can simply be reduced up to a certain degree. This partial restitution is produced by washing out the depressing metabolic products. Being desirous to verify the results of this investigation with greater exactitude I have requested Dr. Lipschütz143 to repeat the experiments, taking the utmost possible precaution in respect to the absolute exclusion of oxygen. Lipschütz has tested the normal saline solution made oxygen free with the sensitive Winkler method, in which the slightest trace of oxygen is shown by the oxydation of manganous chloride to manganic chloride in which the latter in a saline solution sets free an amount of iodide from iodide of potassium corresponding to that of the consumed oxygen. These experiments of Lipschütz have shown that even with the absolute exclusion of the slightest trace of oxygen a partial recovery can be brought about by artificial circulation. There can be, therefore, no doubt that recovery is actually founded on the removal of the depressing asphyxiation substances by artificial circulation. Moreover Fillié144 has previously succeeded in the laboratory at Göttingen in obtaining by the same methods a corresponding result for the nerve. In both cases the experiments are extremely complicated and must be carried out with the most painstaking care. The depressing influence of the asphyxiation products need not be regarded as a specific effect of poisoning. It can be solely an expression of mass relations, if we assume that the anoxydative decomposition is controlled by a chemical equilibrium between masses capable of disintegrating and products of the disintegration. It is not possible to give any detailed account as to the part taken by accumulating asphyxiation substances in the prolongation of the refractory period. Indeed, we must for the present relinquish the attempt to delimitate quantitatively the part taken by the individual constituent processes in the symptoms of depression resulting from the deficiency of oxygen. We can merely say, the individual alterations produced by the want of oxygen, that is, the restriction and retardation of the oxydative disintegration, the corresponding increase of the anoxydative decomposition and the accumulation of the products of incomplete oxydation and anoxydative breaking down have the same influence in that they decrease the strength of the response and retard the rapidity of the decomposition process. These are the general effects perceptible in the refractory period by the deficiency of oxygen.

The establishment of these facts of the dependence of the refractory period upon oxygen are of the utmost importance for the genesis of fatigue, for the state of fatigue in all aërobic organisms is invariably brought about by deficiency of oxygen. In other words: fatigue is invariably asphyxiation. A deficiency of organic reserve substances never occurs in fatigue before the effect of oxygen deficiency leads to complete depression, for the quantity of organic reserve substances at the disposal of the cells is greater comparatively than that of oxygen. This is shown by transfusion experiments in which the time involved before complete paralysis was brought about in the frog by the introduction of an oxygen-free saline solution was ascertained and compared with the period which elapsed before complete paralysis took place, when the same solution saturated with oxygen was used.

Although the previously described experiments on the strychninized frog show clearly the relations of fatigue to the refractory period, I should, nevertheless, like to illustrate them somewhat further.

The state of fatigue as it is developed in a living system by a continuous functional activity is characterized by a series of symptoms which can be best studied in the fatigue of the muscle, the nervous centers, and the peripheral nerves.

If the muscle of the frog is isolated and rhythmically stimulated with single induction shocks and the muscle contractions graphically recorded, it will be found that the first perceptible alteration during the course of stimulation is the increasing height in the curve, which appears directly after the first contraction and becomes more and more noticeable after every succeeding one. With the isolated apex preparation of the frog’s heart an effect is produced which Bowditch145 has termed the “Treppe” and Tiegel,146 Minot147 and others have obtained the same result for the skeletal muscle. The Treppe has been often regarded as an expression of increasing of capability of the muscle following each succeeding stimulus in spite of the fact that it is physiologically incomprehensible that an isolated muscle can become more capable by increased demands. Fröhlich148 first threw light on this seeming contradiction by showing that the increase in height of the muscle contraction in the Treppe is in reality the first indication of the beginning of fatigue, and Fr. Lee149 arrived at the same result. The increase in height of the contraction curve depends upon the retardation of the course of contraction. As the contraction extends over the muscle substance in the form of a wave, a longer stretch of the muscle will be in a state of contraction when the wave is more extended than when it is shorter, that is, the shortening of the muscle will be greater, the contraction curve higher, when the wave is more extended. With increasing fatigue the retardation in the course of contraction, as Rollet150 already has shown, becomes continuously greater. (Figure 34.) The consequence of this retardation in the course of contraction is, therefore, perceptible in the rhythmically activated muscle in the form of contracture. As fatigue increases, the muscle requires an increasing length of time to relax to its full extent and in consequence the period between the two stimuli is very soon insufficient for this to occur. There remains a certain amount of shortening, when the next contraction begins. This characteristic extension of the individual contraction curve of the fatigued muscle is an expression of the retardation of the oxydative disintegrating processes and of the Treppe. It shows us that fatigue is perceptible to a slight degree even after the first excitation. After every succeeding stimulus the oxydative decomposition in the fatigued muscle is increasingly prolonged. It is, therefore, self-evident that the capability of action of the muscle likewise becomes less with increasing fatigue. Every state of fatigue is, in fact, distinguished by the decrease of response. This is perceptible in the later stages by the decline of the height of contraction. Hence all symptoms of fatigue which we observe form the expression of one single process; it is the constantly increasing slowness of oxydative disintegration with increasing fatigue.