Principles of General Physiology

The experiments of von Frey (1876), previously referred to, were undertaken to test the hypothesis of an interference process, in which the excitation and inhibition were supposed to act on the same cell mechanism. It was found that, if the vaso constrictor and vaso-dilator nerves to the submaxillary gland were stimulated simultaneously, the former obtained the victory during the period of stimulation, but the latter showed their effect afterwards. Now, although this result is inconsistent with a purely physical interference process, which should

result in a total abolition of the dilator effect, it does not disprove the hypothesis that the excitatory and inhibitory nerves do actually play on the same muscle cell or nerve cell, although not in the way of interference of wave motion. It will be found, in fact, that experimental results necessitate this view. We must remember that, at the time of von Frey's experiments, it was not known that the activity of the gland cells gives rise to certain products, " metabolites," possibly of an acid nature, which diffuse to the arterioles and, acting there as chemical agents, cause dilatation. This, naturally, continues its action after the actual stimulation of the nerve has ceased. Attention has been directed to it chiefly by the work of Barcroft(1907).

In cases where a process is a spontaneous one, it will be clear that it will resume its activity on cessation of the inhibitory influence. There is more evidence in favour of a second group of theories, founded on the nutrition of cells. The foundation of this view was laid by Hering in his papers on sensations of light (1878) and worked out in more detail in a celebrated paper in "Lotos " (1889). It rests on the idea of the opposition between assimilation and dissimilation in Bering's words, or anabolism and catabolism in those of Gaskell (1886, p. 46). Verworn has adopted it and elaborated it in connection with his biogen theory.

There are, however, many objections to be brought against the theory. In the first place, we may consider the main principle. When a cell is actually increasing in the substance of its protoplasmic machinery, there is no doubt of the fact that it is engaged in building up complex systems out of simpler food materials ; this is assimilation or anabolism ; it may be considered as analogous to the manufacture of a petrol motor. Again, there are certain cell processes about which there can be no doubt that they consist in disintegration or breaking down, with the giving off of energy ; the oxidation of glucose to carbon dioxide and water is such a case. This is dissimilation or catabolism. Now the theory of Hering is based on the idea that the two phenomena are mutually exclusive, and, no doubt, this might be the case if the molecules in question were of the same kind, as is assumed in the theory of biogens. So that if, as is pretty clear, the phenomena known especially as vital, or obvious manifestations of activity, are associated with the breaking down of molecules and giving off of energy, catabolic action will be synonymous with excitation. Further, if we accept the view that building up of new material is inconsistent with catabolic activity, we are justified in regarding the opposite process, or anabolism, as being associated with inhibition.

The theory also supposes that both these processes can be accelerated or started by nervous influences. As to the catabolic processes, there is no dispute, but we have already seen (page 288) that there is no satisfactory evidence of the direct influence of nerves on growth of protoplasm ("trophic nerves"). We have, moreover, also found evidence in various directions that material used for energy purposes does not become an integral part of the protoplasmic molecules, but is used by the protoplasmic machinery in a way analogous to that in which fuel is burnt in an internal combustion engine. The building up of a material of high potential energy, for the purpose of giving off this energy in an available form on its breakdown, appears to be effected by the concurrence of another reaction in which energy is set free by oxidation, as in the case of secretion, and we shall find a further striking instance in muscular contraction.

Again, it is very difficult to form an idea of how the increase of anabolism can result in a decrease of catabolism. Take the illustration used by Forbes (1912, 1, p. 152). The cell is compared to a water tank, provided with an inlet pipe and an outlet, both supplied with adjustable stopcocks. The outlet is supposed to be partly opened, and the stream of water represents the outgoing energy, catabolism. The inlet is connected with a supply at a somewhat higher level, and is opened to such an extent that the level in the tank is kept constant. This inflow, anabolism, is thus equal to the outflow. Now the theory under discussion implies that, if we increase anabolism by opening the inlet wider, we shall diminish the rate of outflow. In point of fact, of course, by increasing the inflow, we raise the level of

water in the tank, and this in itself increases the outflow owing to the rise of driving pressure. That is, increase of anabolisni, or rather the anabolic state of the cell, increases catabolism, instead of decreasing it. We arrive at the same result if we regard the process from the point of view of a reversible chemical reaction. Increase of the mass of a material undergoing decomposition will increase the amount of decomposition taking place in a given time. Indeed, one cannot imagine a process in which increase of activity in one direction necessarily involves decrease in the opposite one. As Forbes puts it, to assume that increase of anabolisni necessarily implies decrease of catabolism, is to suppose that increasing a man's salary ensures decrease of his expenditure. To return to the tank, suppose that we reduce catabolism by narrowing the outlet, the level will rise, and consequently the inflow will diminish. That is, so far as the tank itself is concerned, the effect is the same as increasing anabolisni. If we increase the outflow, we increase also the inflow. Indeed it appears as if Hering had fixed his attention too exclusively on the static condition of the protoplasm of the cell, which is certainly increased in amount by increasing anabolisni. But if we look at the really important dynamic condition, there seems no doubt that increase of anabolisni must also increase catabolism.

There are, on the contrary, certain facts which must not be overlooked, which appear to support this nutrition theory. If we turn to Fig. 113 (page 405) we notice that after the inhibitory pause, the first few beats are larger than those preceding the pause. It looks as if inhibition had, by increasing the contractile material, raised the functional capacity of the tissue. The question is whether this result is any greater than it would be after an equal rest produced in any other way. A further consequence of the anabolic theory would appear to be that the longer the rest, the greater the subsequent improvement. In Fig. 113 there is no relation between the two, and in Fig. 109 of Gaskell's article (1900, p. 205) the first beats are smaller than normal. We may have, in fact, after an inhibitory pause, the same condition as that shown by the " staircase " phenomenon of a ventricle which has been at rest for some time, owing to separation from the sinus.

In the case of inhibitory reflexes to skeletal muscles, we frequently find a subsequent augmentation of contraction, called by Sherrington " successive induction " or " rebound contraction." The effect of inhibition of various durations on this phenomenon has been studied by Forbes (1912, 1) and several important facts relating to inhibition in nerve centres have been brought out. There are two different phenomena concerned : the " rebound " after a brief inhibition, a contraction which is too great to be explained by mere " damming up " of " energy " ; and, secondly, the effect of a prolonged period of inhibition on a subsequent excitation. It is shown that this latter effect depends on the strength of the stimulus of the inhibitory nerve. If moderate, it has a favouring effect, if strong, a depressing one ; so that there is a " critical value " between the two, where no effect* results. An important fact is that this critical value is lowered if the inhibitory stimulus is accompanied by an excitatory one. This result indicates that the two kinds of synapse have a more or less close relation to one another and will be found to have a bearing on the theory of inhibition. In the depressor reflex on the blood pressure in the rabbit, I found (1893, p. 320) that the state of the centre was the same before and after sixteen minutes' continuous stimulation of the inhibitory nerve, during which time the centre was in a state of inhibition, as shown by the unchanged fall of blood pressure.

The part played by fatigue is of interest. If a nerve which causes an excitatory reflex be stimulated, fatigue is produced after a certain time. Now it might be thought that, if an inhibitory nerve be stimulated at the same time, fatigue would be diminished. Forbes shows that the contrary is the case. Fatigue comes on earlier (p. 170). Further, an inhibitory reflex itself is capable of fatigue (p. 179); or rather, a prolonged inhibition, with fairly strong stimulation, diminishes the inhibitory effect of a test stimulus made immediately afterwards. To interpret this result, we must bear in mind some facts as to the seat of fatigue. As Sherrington has pointed out (1906, pp. 103-105), the seat both of excitation and of

inhibition is not in the actual motor neurone itself, but in the synapse of the afferent or intermediate neurone with it. This fact, in itself, is difficult to bring into agreement with any recognisable amount of metabolism, a conception foreign to that of a boundary surface. The motor neurones of the flexor muscles of the hind leg can be used for the scratch reflex when inhibited from being used for the ordinary flexion reflex. Of course, strictly speaking, the synaptic membrane is common to both neurones of which it forms the connecting link, but it is convenient to speak of either one without including the membrane. When fatigue of a particular reflex is brought about by stimulation of a certain afferent nerve, it is found that its motor neurones are not fatigued for a reflex brought about by stimulation of another afferent nerve. Similarly with inhibitory reflexes, we must conclude, then, that the synapses of various afferent (or intermediate) neurones with the same motor neurone are practically independent of one another. But before drawing conclusions as to the irreconcilability of fatigue with increase of anabolism, we !

must remember that there are, in all probability, one or more nerve impulses are FIG. 127. TYPICAL, FORM OF FALL OF BLOOD PRESSURE PRODUCED of the same excita- BY THE DEPRESSOR NERVE IN THE RABBIT. tory nature, that in f^e tracing shows the beginning and end of a period of stimulation lasting the axis Cylinder seventeen minutes. During the whole of this time the blood pressure remained which forms the inhibitory synapse with the motor neurone, is an excitation, and therefore that at the intermediate synapses is also excitatory, and it may be in these synapses that the apparent fatigue of inhibition is situated.

In some other cases, there is no indication of fatigue in inhibition. Gaskell (1900, p. 205) has kept the heart of the toad in complete rest for twenty-eight minutes by continuous weak stimulation of the intracranial vagus. In vasomotor reflexes, I found (1893, p. 314) that, by stimulation of the central end of the depressor nerve, the blood pressure was reduced to about half its height, and remained at this level without change for seventeen minutes ; when the stimulation was stopped, the blood pressure returned to its previous level (see Fig. 127).

In the interpretation of experiments such as those of Forbes, there are two further points to be remembered. The "subsequent augmentation" might have been due to auto-stimulation by the contracting muscle of afferent fibres in its own substance. To exclude this, the experiments were repeated on preparations in which the afferent fibres from the muscle had been cut. No difference in behaviour could be detected. The other point is that if we accept the results of various observers on " all-or-nothing " in the excitatory process, increase or decrease in the height of the reflex contraction must mean a greater or smaller number of cells in the centre in a state of activity. It would appear, therefore, that an increased response, following a period of inhibition, must be due to the inhibitory stimulus having made some synapses accessible to excitation which were previously inaccessible.

Drainage or Diversion Theories. — Another set of theories is based on tinidea of a stream of excitation, flowing in a particular direction, being diverted to a different course by the presentation to it of an easier way. Apart from the "animal spirits" of Descartes, about which something will be said later, the first form in which the theory was expressed seems to be that of William James (1890, 1, p. 585, footnote). It was given more definite expression to by von Uexkull (see 1909, p. 185), who uses the names "tonus" and "excitation," which flows from one part to another of the neuro-muscular mechanisms, and by McDougall (1903), who speaks of a stream of "energy," to which he, originally, gave the name "neurin."

The simplest illustration we can take is that of a water tank feeding a fountain at a lower level than itself ; there is a continual stream of water, possessing energy, passing along the pipe to the fountain. If the gardener opens a large tap on the course of this pipe, in order to fill his watering can, the fountain stops for the time, since the pressure in it falls to zero ; we may say that we have " inhibited " the fountain. Now, to begin with, I find some difficulty in discussing this view of the mechanism of the nerve centres, because the conception does not readily fit in with what we know of the nature of the excitatory process in nerve. There arc. however, certain assumptions made which cannot be accepted in the form stated. It must be admitted that von Uexkiill (1909, p. 58) defends himself from the imputation of using a misleading image, in speaking of a liquid flowing about in the nervous system, on the ground that the object of science is not " truth " but " order," so that we must accept his assurance that he uses such expressions as " quantity " and " pressure " of excitation and " varying capacities of reservoirs " merely to facilitate the description of experimental facts. At the same time, there appears to me to be a fundamental misconception at the base of all theories of this nature, namely, that the nerve " energy " in a given organism is a definite, limited quantity. When a part of it is diverted into some channel, it must, therefore, be drained away from some other place. But there is no reason to suppose that, when a nerve fibre divides into two, the magnitude of the propagated disturbance is diminished to half in each of the branches. Indeed, Adrian's results show that 'the excitation is "all-or-nothing" in both branches; since it cannot be nothing in both, it must be " all " in both. Similar conclusions are forced upon us by consideration of the electrical organ of Malapterurus (see Gotch and Burch, 1896, p. 387), in which the single efferent nerve fibre divides into about 1,800 branches, one to each plate of the organ. If any diminution occurred, the whole excitatory process must be frittered away to practical non-existence.

If there is no diminution on branching of the nerve fibre, it is clear that adding or removing a branch will have no effect on the disturbance in the main fibre. The process is more analog >us to the propagation of an explosion along a train of gunpowder, which can be made to branch as many times as desired without affecting the intensity of the explosion along the main track. In the nerve, of course, the process is reversible and perhaps unaccompanied by evolution of energy, thus differing from that of an explosion (see page 396 above).

The use of the words "stream of energy" is also inappropriate. The actual energy involved in the propagation of a nerve impulse is quite infinitesimal, as we have seen. If it be said that the words are used metaphorically, it is misleading to take a word which has a definite quantitative, mechanical meaning. The application of the drainage theory to explain " reciprocal innervation," in which excitation of a particular muscle is associated with inhibition of its antagonist, is shown by McDougall (1903, p. 175) in an ingenious diagram. It shows how the two are always associated and is curiously similar to that of Descartes, which will be found described on page 495. It implies, however, that all cases of inhibition must be associated with excitation somewhere else. In our illustration, " inhibition " of the fountain is associated with " excitation " of the watering can, or of something into which the water runs. In fact, in McDougall's scheme, inhibition of a centre controlling a certain muscle can only take place by stronger excitation of the centre of its antagonist. This does not agree with experimental facts. Sherrington (1906, p. 203), moreover, objects to the theory on the ground that it makes inhibition in nerve centres a different process from peripheral inhibition of smooth muscle, heart, etc. In these latter cases, it does not seem possible to apply the drainage theory. Von Uexkiill, nevertheless, does apply it, in his form, to the case of the claw of the crayfish and in the following way (1909, p. 213, and Uexkiill and Gross, 1913, p. 354). There are two motor nerve tracts, ending at each muscle in a network. These networks are connected by bridges, so that any fresh excitation of either network sucks off the remaining excitation of the antagonist. In front of the network of the closing muscle there is a block of high resistance, so that weak stimuli do not reach the network. The chief evidence for the existence of such bridges seems to lie in the fact that direct stimulation of the opening muscle causes contraction in the extensor of the carpopodite, and in the presence of dividing fibres in the trunk of the nerve. It will be seen that the anatomical facts of Biedermann and of Mangold do not support this view.

There is no sign _of anastomosis of nerve fibres to form a network, nor of fibres which could be pointed out as connecting one muscle with the other. It is much simpler to suppose that the two different kinds of fibres which enter the same muscle fibre have opposite functions on account of the difference of the way in which they end in the muscle fibre. If each fibre of the nerve divides, as seems most probable, a branch going to each muscle, it seems that " axonereflexes " in Langley's sense (1899, p. 388) should cause excitation or inhibition of the antagonist muscle when either muscle is stimulated directly. Which of the two would occur would depend on the strength of the stimulus, as in Richet's experiments.

Hofmann's work (1914), indeed, gives us definite information on this question. The two axis cylinders, which we have seen to run together, were found, in the case of the opening muscle, to come from two separate nerve trunks, so that they could be excited each apart from the other. One is inhibitory, the other excitatory. There is no nerve network in the neighbourhood of the muscle. It was found that the axis cylinders of the excitatory nerve send branches to the muscle of the preceding joint of the appendage, so that the result of von fJexkiill, described above, turns out to be an axone-reflex, as suggested. Since the two opposing muscles are innervated by the same axis cylinder, which divides, it is clear that both muscles would be stimulated to contraction at the same time, unless the one were provided with means of inhibition from the centre. The balance is peripheral here, instead of central, as in the vertebrate. The opening muscle is innervated merely by two branching axones from the centre, and there are no anastomoses at all.

Block. — It might be thought that a very simple way of putting an end to the excitatory impulses playing on a nerve cell would be to make some synapse in their course impervious to excitation. Sherrington (1906, pp. 100-103) holds that inhibition involves more than this. In the decerebrate cat, the extensors of the knee can be put into contraction by a slight pinch of the opposite foot. The discharge continues for some time, gradually passing off. But if the central end of a branch of the hamstring nerve of the same leg be stimulated for a quarter of a second, the after-discharge is suddenly and completely inhibited. It seems that the efferent neurone is put into a state of excitation which continues after the exciting impulse has ceased, and that this state of excitation can be quelled by inhibitory nerves, although there is no exciting impulse to be blocked off.

In view of Sherrington's later work on " Plastic Tonus," to be described in Chapter XVIII., it might be suggested that this after-discharge, being itself reflex from receptors in the muscle itself, is subject to the same process as excitation from other sources. But the experiments of Forbes (1912, 1, p. 182) show that a certain amount of after-discharge is still present when the afferent fibres from the muscle have been divided. It appears, then, that there is direct evidence that the process of inhibition implies more than the mere cutting off of impulse*.

In a somewhat different sense, however, inhibition may perhaps be regarded as a kind of block. We have found reason to look upon an increase of permeability of a membrane as an intimate part of the excitatory process, and inhibition, as the opposite process, would thus be associated with decrease of permeability. But this is obviously quite a different matter from blocking the propagated disturbance itself. Other Contributions to the Theory of Inhibition. — Although the protoplasm of the neurone is probably a liquid, it contains various substances in the colloidal state. Further, we have seen the necessity of the presence of electrolytes to account for the electrical changes in nerve Now, if these electrolytes lower surface energy, they will be adsorbed on the surface of the colloidal particles (see page 55). This is the foundation of the theory of Macdonald (1905) as to the nature of excitation and inhibition. This investigator points out (p. 335) how the concentration of electrolytes in the external phase would be increased if anything in the nature of aggregation or coagulation in the nerve colloids occurred. The adsorbing surface would be diminished. Assuming that these electrolytes are essential to excitation, it will be seen how a coagulation process would be associated with excitation, while a greater dispersion than normal would be associated with greater adsorption of electrolytes and inhibition (Macdonald, p. 348). This is a brief account of this important theory. Details of its application require more knowledge than we possess as yet of the phenomena taking place at the membranes. If we place the seat of excitation and of inhibition of nerve cells at the synaptic membrane, we may suppose that the system of colloids and electrolytes in question either forms the membrane itself or is intimately associated with it.

In discussing Hill's modified form of Nernst's equation for excitation, we saw (page 394) that it contains a constant, C, which is connected in some way with adsorption (or disappearance) of ions. Keith Lucas (1910, p. 243) shows that it is altered by removal of calcium. It seems not unlikely that changes in the value of this factor, C, might be made use of to investigate the hypothesis of Macdonald. With regard to the properties of the synaptic membrane, Keith Lucas (1911) points out that it must present a* greater resistance to conduction than the axis cylinder does ; it is thus similar to the junction between nerve and muscle or to a narcotised region in nerve. There is thus a possibility that it might obliterate very rapid, and therefore small (see Wedensky inhibition, page 420), nerve impulses, so that they would not get through the synapse. Adrian (1912, p. 411) calls attention to the fact that an impulse, if it has been able to pass a region of decrement at all, recovers its full size on arriving in normal nerve again ; so that, in order that the block mechanism above referred to may be effective, the impulses must be reduced to zero in one of the synapses ; unless this were the case, it would not matter how many separate regions of decrement the impulse had to traverse.

It cannot be said that any one of the theories suggested is a satisfactory one. Perhaps each, when modified in accordance with certain important parts of the others, has a part of the truth, and there may be particular fields in which aspects of one theory have more part to play than in other fields. There are certain facts which admit of no doubt, and any theory must reckon with these. The function of any particular nerve fibre depends on its termination ; the nature of this termination determines whether it excites or inhibits. In the case of smooth muscle or heart and certain nerve centres, which have an inherent state of excitation, each element requires two nerve fibres to modify its state, increase it or decrease it. In Langley's view, each of these nerve fibres ends in a distinct receptive substance, whose qualities determine whether the effect is excitatory or inhibitory. It appears also that, of the various nerve fibres forming synapses with a particular nerve cell, each has its definite character of inhibiton or of excitation. In such cases as those of reciprocal innervation, Sherrington points out (1906, p. 105) that the inference must be made that each afferent nerve fibre concerned in the reflex divides into two parts in the spinal cord, one set of

these subdivisions being excitatory, the other inhibitory, in respect of the motor neurones. This is similar to the most probable arrangement in the crayfish claw, namely, that each nerve fibre divides into two, one division going to the closing muscle, the other to the opening muscle ; the two parts into which each fibre divides are always one excitatory, the other inhibitory. Michailov (1911) describes two different kinds of nerve endings in the muscular tissue of the heart. He regards one of them as sensory, the other as the inhibitory endings of the vagus. It is, of course, also possible that the former might be the terminations of the sympathetic supply.

We have found evidence that the membranes which are the seat of the excitatory process are of such a nature as to be impermeable to one only of the two oppositely charged ions into which certain electrolytes in the nerve are dissociated. It may possibly happen that, when the impermeable ion is allowed passage by the arrival of the propagated disturbance, it produces chemical or physical changes in the substance of the inhibitory synapse of such a kind as to render excitatory synapses on the same cell incapable of undergoing the normal change of permeability associated with the passage of an excitation. In the present state of knowledge, it is unprofitable to follow such speculations far, and the suggestion is made merely as an indication of a possible mode of explanation.

Although under ordinary conditions, such as change of intensity of stimulus, altered time course, and so on, an inhibitory termination of an afferent arc cannot be made to give any other effect than inhibition, nor an excitatory termination anything but excitation, there is evidence, as Sherrington points out (1906, p. 105), that under special conditions an inhibitory termination can be made into an excitatory one, and vice versa. Certain drugs have effects of this kind. Moreover, from the cerebral cortex, antagonistic muscles can be put into action at the same time. We saw above that stimulation of the central ends of various afferent nerves of the hind leg produces inhibition of the contraction of the vasto-crureus in the decerebrate cat. Now, Sherrington found (1905, p 287) that a small dose of strychnine converts this effect into a reflex contraction. I found the same fact in the case of the inhibition of the vaso-constrictor centre by the depressor nerve (1908, 2), and Seeman (1910) in respiratory reflexes. Sherrington also showed that tetanus toxin does the same. It is to be noted that the vastocrureus is a purely extensor muscle and can be completely isolated. By careful gradation of dose, a stage can be obtained in which the inhibitory effect is diminished, but not replaced by excitation. This fact seems to show that the result is not to be explained by decreased resistance allowing the stimulation to spread to other neurones in the centre, since the resistance at this stage is increased. Another possibility is that the afferent nerve stimulated might contain, along with the fibres causing inhibition, others causing contraction, and that the effect of strychnine might be to paralyse the inhibitory reflex fibres or synapses before the excitatory ones.

Now these inhibitory fibres would be associated with those causing contraction of the flexors and no trace of depression in the force of the flexor contractions is to be detected. The antagonist muscles are simultaneously thrown into contraction. I have shown, moreover (1908, 2), that the strychnine reversal of the inhibition of the vasomotor centre is not due to unmasking of excitatory fibres mixed with the inhibitory ones, by paralysis of the latter. The dose of the drug required in the rabbit is a comparatively large one and, at the time that the reversal is most perfect, it is found that the excitatory vasoconstrictor fibres in other afferent nerves are completely paralysed, so that the usual rise of pressure from ordinary sensory nerves is absent. It seems on the whole that the view taken by Sherrington at the first (1906, p. Ill) is the most probable one, namely, that the action of strychnine " is to convert in the spinal cord the process of inhibition — whatever that may essentially be — into the process of excitation — whatever that may essentially be. The reflex nexus was preexistent, but the effect across it was signalised by a different sign, namely, minus prior to the strychnine or tetanus toxin, instead of plus, as afterwards."

In a later paper bv Sherrington and Owen (1911) the question is discussed further, but the impossibility of a definite decision is pointed out until we have a nerve which can be proved to contain no afferent excitatory fibres whatever. Chloroform is well known to be, pharmacologically, a general antagonist to strychnine ; it is not unexpected, therefore, to find that its action on nerve processes is the reverse of that of strychnine. I showed (1908, 2) that it converts the excitatory component of a vasomotor reflex into an inhibitory one. That is, the rise of pressure produced by a nerve which excites the vaso-constrictor centre is converted into a fall. Examples of this fact are given in Fig. 128, from a paper by Mathison (1912), and Fig. 129 from one of mine. Similarly, Sherrington and Sowton (1911, 2) showed that the reflex contraction of the vasto-crureus, evoked by stimulation of the popliteal, internal saphenous or genito-crural nerve, was converted into inhibition by the administration of chloroform.