Principles of General Physiology

The refractory state which follows a second effective stimulus, applied during the relative refractory state following a previous stimulus, is shorter than the normal one. The duration of the refractory state is, therefore, dependent on the magnitude of the disturbance. The refractory state is not merely a local effect at the point of application of the stimulus, but is the same at any point in the nerve after the passage of the propagate 1 disturbance (Bramwell and Lucas, 1911).

Keith Lucas (1911) shows further that the refractory state is associated with the propagated disturbance by the fact that a stimulus falling within the absolute refractory period of a previous one does not prolong that refractory state, and :i third stimulus is effective at the same interval of time after the first, whether the second has been interpolated or not. Fatigiie. — If we regard fatigue as that result of activity by which a cell is less readily put into action again until a certain time for recovery has been allowed, it is clear that the refractory state itself is one of fatigue. Under ordinary conditions, however, the recovery is so rapid and complete that it is impossible to demonstrate that a nerve is less excitable at the end of a long period of activity than at the beginning. From certain experiments by von Baeyer (1902), it appears that, in the absence of oxygen, signs of fatigue are to be detected ; while the refractory period was found by Frohlich (1904) to be prolonged to O'l second in the absence of oxygen. In view of the definite proof by A. V. Hill, that the heat evolved is so minute as to make it very doubtful whether there is any metabolism in the nerve fibre, it becomes necessary to consider for a moment what are the experimental facts with regard to the effect of oxygen. The experiments of von Baeyer showed that a region of nerve, exposed to currents of nitrogen or hydrogen, failed to respond to induction shocks after some five hours' action. The excitability returned in oxygen. This behaviour is quite similar to that when a typical anaesthetic is used, so that it seems that the process may well be the same. Von Baeyer (1902, 2) did not obtain any evidence that the time required to produce the state of inexcitability was made shorter by continued stimulation of the nerve. Thorner (1909), however, found that continuous tetanic stimulation, in the absence of oxygen, caused an earlier appearance of the inexcitable condition ; recovery took place, to a considerable extent, when the excitation ceased, ivilhont the necessity of the presence of oxygen.

It seems possible that the local effect of the current at the electrodes was not sufficiently excluded in these experiments. Polarisation is not easily prevented. Anyone who has excited the vagus nerve of the cat is familiar with the fact that the inhibitory effect on the heart rapidly disappears during stimulation and that reappearance occurs when the electrodes are moved to another spot on the nerve. The results of von Baeyer may possibly have been din- to traces of impurity in the gases used, although they were purified by the usual chemical methods. Exposure for five hours might enable an effect to be produced by traces which would be incapable of detection.

r On the other hand, since we must suppose that nerve fibre is living, it is difficult to believe that it is absolutely devoid of respiratory activity. We know that it requires a certain minimal amount of energy to start a disturbance ; but the fact that this disturbance in normal nerve is propagated without diminution, suggests a physical process, although it might be argued that energy is supplied to it as it travels. In the latter case, it is conceivable that the energy-giving material might require replacement by an oxidation process ; but we are again met with the difficulty of the absence of heat production. A. V. Hill suggests that oxygen acts by keeping the machine in order, as it were, somewhat as oil in a motor does.

It must be confessed that this seems a rather unusual function for oxygen to perform, and it does not appear to me that the dependence of excitability on oxygen has been satisfactorily demonstrated. It would be desirable to test the effect of a simple vacuum, although the experimental difficulties of exposing the nerve to the vacuum, while allowing access of oxygen to the muscle, seem insuperable ; the use of the electrical change as indicator would be possible, though less satisfactory.

In any case, it will have been abundantly clear, from the various facts given in previous pages of this book, that the food requirements of cell machinery in general are extremely small ; food is required to afford energy for the numerous physiological processes, and this is done by oxidation under the action of the cell mechanisms. An infinitesimal amount of oxygen may actually be necessary in such a process as that of conduction in nerve, where there is practically no energy change involved. We shall meet with some further facts bearing on the question presently.

Summation and Facilitation. — In the experiments of Adrian and Lucas (1912), from which the curve of Fig. 108 (p. 389) was constructed, we see evidence that the refractory period is succeeded by one of slightly increased excitability, in which a less strength of stimulus is required to excite a propagated disturbance. This phenomenon is met with in any part of the nerve after the passage of a disturbance. There is, however, another form of increased excitability, shown at the point of excitation only (Adrian and Lucas, 1912, pp. 69-72). The first effect of a stimulus at its place of application, is a process which, on Nernst's theory of excitation (see later, page 393), we should interpret as a concentration of ions against a semipermeable membrane. Now this happens even when the stimulus is too weak to set up a propagated disturbance. It is shown by the fact that a second stimulus, also inadequate by itself, following the first after about O0008 second, sets up a propagated disturbance. It is clear that the first stimulus has left behind it a change of some kind which persists for a measurable time, and is added on to that produced by the second stimulus when this is put in. The propagated disturbance, on the other hand, as we have seen, leaves behind it a stage of diminished excitability at this interval after a previous stimulus ; so that there are evidently two factors involved in the excitatory process, one t)f which is confined to the point of application of the stimulus. The importance of this fact for the theory of excitation will be seen presently.

A narcotic, such as alcohol, does not prolong the time required for recovery, the refractory state, even at the stage in which the disturbance is conducted with considerable decrement and slowing of rate of conduction (Keith Lucas, 1913). This fact suggests that the recovery process is not of the nature of a chemical, oxidation process under the control of living protoplasm. In fact, it seems to exclude the view of the necessity of oxygen for recovery, as an oxidative process.

The Electrical Response. — We have seen that the disturbance in nerve is associated with a temporary state of negativity. The meaning of this will be discussed presently. It is held by some observers that the two processes are not necessarily connected, but Keith Lucas (1912, pp. 502-508) shows that none of their experimental results are free from objection and that there is no reason for doubting the identity of the two. On the other hand, he points out that more strict proof is desirable before definitely accepting the electrical change as a basis for a physico-chemical explanation of the excitatory process. In the case of muscle we shall find evidence that, although the excitatory process and the electrical response may be the same phenomenon, yet both these may be present without a contractile response, which is, so to speak, an additional process, whose conditions of appearance may be absent.

Macdonald (1902) gives good reasons for regarding the potential difference between cut end and longitudinal surface of nerve as due to the high concentration of inorganic salts in the axis cylinder, in connection with the presence of membranes impermeable to these. This potential difference was found to be abolished by a certain concentration of ions outside the nerve, 7 to 10 per cent, of potassium chloride being necessary. The salts of the nerve cannot be regarded as combined chemically and split off on excitation, but must be adsorbed on surfaces of colloids in the axis cylinder. In this way, as is pointed put, these salts are prevented from manifesting their great osmotic pressure. The difficulty, however, still exists, since ions, in order to give the necessary Helmholtz double layer, must be free and not adsorbed. The electrical state of nerve and muscle is often spoken

of as that of a concentration battery. As we shall see later (page 393 and Chapter XXII.) it is only in a modified sense that this statement can be made. Under certain conditions it is possible to observe an electrical change in the opposite direction, after cessation of the stimulus, both after tetanising and after single stimuli (Garten, 1903, p. 59). This phenomenon, which was first noticed by Ewald Hering, was correlated by him with the restitution or assimilation process, by which the excited nerve returns to its original state. This view is in agreement with Bering's well-known theory of assimilation, which will be discussed under the head of "inhibition." But it cannot be said that we have, as yet, a satisfactory explanation of this positive electrical response. It may, perhaps, have some connection with the stage of increased excitability of Adrian and Lucas. According to Ve'szi (1912), however, the magnitude of an electrical response is decreased in the stage of positivity after a prolonged tetanic excitation, but it would \*e more to the point if the observations had been made on the actual propagated disturbance itself. Cremer (1909) suggests that nerve in the resting state may be in a condition of partial excitation or negativity, which disappears, of course, immediately after a disturbance, and would give rise to the appearance of a stage of less negativity, that is, of positivity, until the normal tonic stattt is re established. If the resting state is a balance of two opposite processes, as is likely, there is some justification for Cremer's view, although no other evidence has been brought forward in favour of the existence of such a tonic condition of partial excitation.

Certain support is given to the idea of the electro-positive response as representative of a restitution process by the expei'iments of Sochor (1911), who found that, in a current of nitrogen, this positive after-action is abolished much more rapidly than the negative excitatory change is. The result might be interpreted as showing the necessit}7 of oxygen for restitution, but the fact that carbon dioxide was found to abolish the effect much more quickly than nitrogen does suggests rather narcotic action. As Garten remarks, granting the necessity of oxygen for the restoration process does not prove that it is an assimilation in the sense of Hering.

Rate of Conduction. — The fact that the nerve impulse takes time to traverse a nerve was first definitely shown by Helmholtz (1850) and had an important effect on views taken with regard to mental phenomena, since here was a nervous process capable of numerical expression. The value obtained by Helmholtz for the frog was 29 m. per second. In man, the latest value, obtained by Piper (1912, p. 52), is 123 m. per second. This was obtained by the use of the string galvanometer and may be taken as a very accurate one.

All investigators agree that the rate is independent of the strength of the stimulus. Narcotics, such as alcohol, slow the rate of conduction (Keith Lucas, The temperature coefficient as determined by the most accurate method, that of Keith Lucas (1908), is T79 for 10°. I have already pointed out (page 42) that it is not permissible to draw conclusions as to whether a process is chemical or physical from this value alone ; one may say this much, that a simple chemical reaction with a temperature coefficient lower than 2, at ordinary temperatures, is extremely rare, if not unknown.

Changes in Permeability. — When a nerve is cut across and electrodes placed on the cut end and on the longitudinal surface, as in Macdonald's experiments referred to above, there is found to be a difference of electrical potential between these points, such that the cut end is negative to the normal surface. As we shall see in Chapter XXII., the only satisfactory way of explaining such electrical states is by the assumption of a membrane which is permeable to one of the ions into which an electrolyte inside the axis cylinder is dissociated, but not permeable to the oppositely charged fellow ion. We have, indeed, described such a case in that of Congo-red, separated from water by a parchment-paper membrane. A Helmholtz double layer is formed at the membrane or, as it is sometimes expressed, the membrane is "polarised," having anions on one side. cations on the other side. Suppose the membrane to become suddenly permeable to both ions, what will happen ? Since the constraint preventing the two layers of ions from freely mixing is removed, the ordered arrangement of ions ceases to exist and, with it, the potential difference between the two sides of the membrane and the possibility of polarisation. Now this is precisely what happens when a nerve is put into a state of excitation. If the cut end is negative at rest and the other electrode on the longitudinal surface becomes negative when excited, as experiment shows, the potential difference is either greatly reduced or abolished,

according to the extent of the loss of impermeability at the excited spot. This is why the electrical response of nerve or mtiscle may be called, as by its discoverer, Du Bois Reymond, the " negative variation " ; negative does not refer to the sign of the electrical response, but means diminution. This manner of origin of the electrical response is sometimes described as a "concentration battery," but, if the description on pages 190-191 above be referred to, it will be seen that a concentration battery in the original sense requires electrodes of one of the elements of the dissociated salt. The electromotive force of the kind of battery with which we are here concerned is also a function of the relative concentration of the two solutions in the ion to which the membrane is permeable, and is expressed by the formula which Nernst worked out for the concentration battery proper, as will be shown in Chapter XXII.

If a potential difference is applied to such an arrangement, so that, for example, the anode is on that side of the membrane where are the cations, to which we will suppose the membrane to be impermeable, and the cathode on the opposite side, it will be clear that no current will flow, since no cations can travel to the cathode to be discharged there. The membrane is said to be polarised. If, however, the membrane becomes completely permeable, the current can pass freely, and the polarisation ceases. This change was shown by Hermann (1879, pp. 165-167) to occur in the excitation of nerve; it becomes less polarisable, as it might be expressed.

The fact may also be stated in the form that the excitatory change is increased at the anode, diminished at the cathode. Verzar (1912) has obtained results which show that this diminished polarisability lasts considerably longer than the electrical excitatory change proper, although in a considerably diminished degree. Further direct evidence of increased permeability of the membrane in excitation will be found in the case of muscle below. Confirmation of this view of the source of the electromotive force of nerve is to be found in the experiments of Macdonald (1900) on the magnitude of the "demarcation current," when immersed in solutions of electrolytes. This " demarcation current " or potential difference between cut end and normal surface follows the Nernst formula for concentration batteries, as we have seen that the membrane process of the hypothesis in question does.

It must be confessed that it is difficult to make out at present which of the two changes referred to is the cause of the other, or whether they are different expressions of the same phenomenon. The loss of impermeability may be the cause of the disappearance of polarisation, or the disappearance of polarisation, as in excitation by an electrical current, may affect the membrane, which must be colloidal in nature, in such a way as to make it permeable. But it is not easy to see how mechanical stimuli can directly affect polarisation.

When the excitability of nerve is spoken of as a colloidal phenomenon, what is to be understood is that the membranes of which we have spoken are of complex colloidal structure and, as such, sensitive to electrolytes, etc. Hoeber (1910) has shown that electrolytes, in their action on nerve, follow the Hofmeister series, a characteristic of their action on lyophile colloids, as w^ have seen. Loewe (1913), also, shows how the action of narcotics is to be explained as a (decrease of the possibility of the membrane becoming permeable on excitation. This decrease is due to adsorption of the narcotic by the preponderant lipoid constituents of the membrane, which are thus changed from lyophile to lyophobe colloids.

The Nernst Theory of Excitation. — Nernst (1899), considering the reasons why ery rapidly alternating currents do not excite nerves, was led to the view that the rocess of excitation by an alternating electrical current is essentially connected h the production at some membrane of a certain minimal concentration of ions to which the membrane is impermeable. If the time during which the current passes in any one direction is too short, the opposite current will carry back these ions before they have had time to reach the effective concentration. This view leads to the simple law that single currents of variable duration will be of the same just effective strength if the product of their strength and the square root of their duration is constant. This follows from the mathematical expression for diffusion. Now, experimentally, this simple relation is found to hold only in a limited region of very short durations of current flow. In fact, Nernst himself regards it only as a first approximation and suggests factors that have to be taken into consideration in a law of wider application. Some of these factors have been considered by A. V. Hill (1910) and modified formulae put forward.

The factors in question may be discussed briefly here. It may be pointed out that, from the standpoint of general physiology, the vakie of formulaof the kind in question is not so much that of being able to express the relation between the exciting power of an electrical stimulus and its physical properties, but the light that they throw on the nature of the excitatory process itself. The first point is that, in Nernst's treatment of the problem, only one membrane is taken account of. But it is clear that there may be another membrane at no great distance from the one under consideration, which will make a considerable difference in the diffusion of the ions, since the ions of opposite sign will be concentrated there. By the introduction of this conception, Hill deduces a formula which was found by Keith Lucas (1910) to satisfy experimental data when currents _ of long duration are used. The effect of the proximity of the membranes in its tendency to cause the equalisation of concentration by diffusion, owing to the rapid fall of concentration in a short distance, would naturally not come into play in very short periods of closure of the current.

A second point, which was suggested by Nernst in order to account for the fact that, if a current is allowed to rise in strength at a rate less than a certain critical value, it does not excite at all, is that there is reason to suppose that the separation of ions brought about by the current is accompanied by a slow, independent, automatic process, by which the ions are taken out of the sphere of action in some way before they have attained sufficient concentration to excite. The precise manner in which this happens is not clear, but it is probably a reversible process of the nature of adsorption.

Hill gives (1910, p. 208) as an illustration a tube of a mixture of oxygen and hydrogen gases. Suppose that this is heated at one end to a temperature at which explosion occurs. This corresponds to an effective stimulus setting up a propagated disturbance. But, if we heat very gradually, not allowing the temperature to rise to the explosion point, the gases combine slowly without explosion, and, if the heating is continued for a sutlicicntly long time, there will be a very small tension of th« gases left uncombined, and no explosion will result even when the temperature arrives at the degree usually sufficient.

According to Hill, the experimental results available at present are not of such a form as to enable his formula to be applied to cases of exciting currents slowly rising in strength. Although the complete derivation of the formula is beyond the space that can be given here, it may be of interest to enumerate the factors of which it consists. In its simplest form it is : — where i and t are the variables, i being the smallest current that will excite when of the duration t. X, p and 6 are constants, whose precise form and significance will be found in the original paper and in that by Keith Lucas (1910, p. 234). It must suffice to say that each of these constants is compounded of other constants to which a definite meaning can be attached. They are : —

b, the distance from the membrane at which the concentration changes are v, the number of ions, each carrying a given quantity of electricity. C, a constant expressing the rate of " recombination " of the ions in the manner referred to above; or, as Lucas prefers to put it, the ease with which the propagated disturbance is set up in a particular condition. Lucas shows further (1910) how the various constants are affected by certain changes of condition, such as temperature and presence of calcium, and the part played by each in the process of excitation. We note especially the changes in

C and in k. Now — is the diffusion time of the ions concerned in the pi-ore^. and a- the constant 0 of the simplified equation is defined by Hill as k so that a convenient measure of — is log 6. Lucas calculates the value of log (9 for various excitable tissues from his own experimental data, to which reference will be made presently. It may be noted that, in all probability, this quantity is essentially the same thing as Waller's " characteristic." In fact, any considerable alteration in the shape of the curve correlating excitation with stimulus is due to changes in 6, since A. and //, are not so readily affected, as may be seen by the consideration of what they mean.

A is the smallest current that will excite at all, however long it be continued. If t becomes very large, 1 — \iQl becomes unity, and i is equal to A.. fi refers only to the distance from the membrane at which the change of concentration is being considered, and will not be liable to important changes. I fear that this necessarily brief account gives but an imperfect view of this important work ; the original papers of Hill and Lucas should be consulted.

There is another point to which a little attention must be given. The fact, that on closing a current through a nerve, the excitation wave starts from the cathode shows that the cations are the important agents. How then is the fact of excitation at the anode, which occurs on breaking the circuit, to be explained? It is pointed out by Keith Lucas- (1912, p. 519) that "the one feature which is common to the cathode when the current is made, and the anode when the current has just ceased to flow, is an increase of the concentration of cations above the value which occurred at each of these points immediately before." At the anode, however, the concentration of cations only rises to its normal level by diffusion, after having been decreased: Nernst and A. V. Hill give what is essentially the same explanation on the ground of the " combination" of ions with some substance in the nerve. During the passage of the current, the diminished concentration of cations at the anode results in a different equilibrium in the reversible "compound," or adsorption, between the ions and the assumed substance. When the current ceases to flow, there is a sudden concentration of cations in the system in excess of that with which it was previously in equilibrium ; a condition which is the same as that at the cathode when the current is first established. Thus the excitation at the anode and the failure of slowly rising currents to excite appear to depend on the same conditions. It will be clear that more experimental work is required before the question can be decided.

A word is perhaps necessary as to the position of the membranes about which we have been speaking. There is no evidence of the existence of transverse membranes and, in fact, their assumption would raise considerable difficulties. It seems most likely that it is the cell membrane covering the axis cylinder that is concerned. This axis cylinder, as we shall see, is a part of a long cell, the "neurone," which includes the cell body with its nucleus, etc. Bernstein (1902), indeed, put forward the view that this surface membrane is the structure responsible for the electrical phenomena of nerve and muscle.