Protoplasmic Action and Nervous Action

Whenever a sufficient uncompensated potential difference is established between an electrode and a solution, as in any battery with closed circuit, the conditions for chemical change are present; there is a transfer of electricity associated with a chemical decomposition or other reaction (oxidation, synthesis, etc.) at the interface. It is well known that a certain critical decomposition-voltage must be exceeded in order to carry out any definite electrolysis, e.g., of a metallic salt; and if the cell surface possesses the general properties of an electrode, the chemical reactions there occurring must be subject to similar conditions. The need for a certain minimal or ''threshold" current-intensity in stimulation is thus explained ; it is evident that if the foregoing theory of transmission is well founded the potential

* Bredig and Kerb, loc. cit.; Wilke and Meyerhof, Arch. ges. Physiol., difference of the local bioelectric circuit in a conducting tissue like a nerve must exceed the critical value required for the electrochemical process which initiates the chemical reaction of stimulation. The general nature of the conditions will be considered more fully later when the phenomena of transmission are discussed in detail. For the present we may conclude that the significance of the polarization change involved in electrical stimulation is simply to furnish the condition required for some critical chemical decomposition at the cell surface. Presumably this chemical change alters locally the physical properties of the surface-film in such a way as to involve local breakdown or increase of permeability; and then, just as in the passive iron model, an automatically self-propagating wave of chemical decomposition is initiated. In this process the altered and the unaltered portions of the cell surface act as two electrode areas, in a manner analogous to that observed in the passive wire and similar systems during transmission.

In living tissues the conditions are more complex than in the simple model considered by Nernst, which takes account of only one of the conditions of electrical stimulation. Two chief conditions which this simple theory disregards are: (i) the existence of a critical threshold current-intensity, independent of duration; and (2) the character of the response to currents of changing intensity. Nernst's theory, however, explains the essential fact of polar stimulation, in addition to assigning a definite condition, viz., change of polarization, for the initiation of the stimulation-process. On the basis of the law of polar stimulation we may now say further that a change of polarization in a definite direction^ such as

to render the external layer of the solution in contact with the cell surface less positive than before, i.e., a depolarization, is the critical or initiatory event in stimulation/ But a current of too weak intensity, or one rising to its maximum too slowly, will not stimulate, whatever its duration. These discrepancies from the simple polarization theory must be referred to the special properties which the irritable tissue possesses by virtue of being a living structure. Apparently the irritable element is able to compensate slight or gradual changes of polarization as a part of its general regulatory capacity. Thus, if a current be led gradually into a nerve or muscle, a considerable intensity may be reached without stimulation. But if then the current be suddenly broken, stimulation results. This behavior seems to imply that while the current is gradually increasing, the cell by some regulatory process maintains its normal or resting physiological polarization essentially unaltered. During the flow of the external current, part of the polarization at the cell surface must depend on the presence of this current, which steadily conveys ions to (or from) the surface. When this influence is suddenly withdrawn, by breaking the current, the effect is to alter the polarization more rapidly than can be compensated by the activity of the cell, and stimulation results. The fact that the rate of change to which the irritable element can thus adjust itself without undergoing stimulation is rapid for rapidly reacting tissues (i.e., those with brief chronaxie) and gradual for ''slow" tissues indicates that

* Cf. Briinings, Arch. ges. Physiol., C (1903), 367. Hermann also recognized that the polarization change of stimulation is in the direction of rendering the external surface of the irritable element less positive than before {loc. cit.). some specific chemical process, whose rate is determined by the characteristic metabolic properties of the tissue, is what preserves the normal resting state of the irritable elements. Thus we may imagine the material of the surface-film as being constructed and replaced as rapidly as it is removed by the chemical (reducing) action of the current; under these conditions the film (with its resting polarization) remains unaltered and no stimulation results. But if the rate of removal exceeds the rate of replacement, the consequence is alteration of the film and stimulation. Conditions of essentially this kind exist in the passive iron model, which shows a similar type of behavior.

A typical case of stimulation by the constant current will illustrate how the stimulating effect varies with the duration of the stimulus. The following table from Keith Lucas^ gives results obtained with the sartorius muscle of the frog. Constant currents of known intensity were passed through the tissue by non-polarizable electrodes, and the minimal duration required for stimulation was determined by varying the distance between two contacts (one making, the other breaking the current) in a swinging pendulum.

The essential feature of these results is that below a certain definite intensity of current (0.18 units), increasing the duration has no effect in lessening the intensity required to stimulate; i.e., weaker currents will not stimulate, whatever their duration. This intensity represents the critical or threshold value for any current. But with stronger currents, the duration required for stimulation becomes less as the intensity increases, and a close approximation to Nernst's square root law is found. This signifies that a certain minimal change of polarization is required to initiate the stimulation process; with currents above a certain critical intensity this polarization is attained with briefer and briefer durations as the intensity is progressively increased. Lapicque and other observers have obtained similar results. The current of threshold value, i.e., of the least intensity that will stimulate with any duration, must traverse the tissue for a certain minimal time in order to stimulate. This time is characteristic for the tissue in question, and apparently is a direct function of the rate of certain specific metabolic processes; probably those concerned in the alteration of the surface-film, as indicated by the duration and other features of the refractory period (see below). According to Keith Lucas and Mines, the length of this minimal time varies with the temperature of the tissue in accordance with a somewhat low temperature-coefhcient, similar to that of diffusion

(Qio = i.3).^ For the irritable tissues of the frog, Lucas^ gives the following determinations (at ca. 13°): substance/3 (nerve end-plate) of the sartorius, .001 second; motor nervetrunk, .003 second; muscle fiber (sartorius), .02 second; ventricle, 2 seconds; for smooth muscle the period is much longer (several seconds) . Lapicque finds the least effective duration of the minimal stimulating current to vary widely for the muscles of different animals, and gives the following data:^

These determinations illustrate the specificity of this time-factor for different animals. It is interesting to note that the velocities of the motor nerve impulses in different animals vary in a closely parallel manner. To designate this characteristic time-factor in the electrical stimulation of different irritable systems, Lapicque has introduced the term ''chronaxie." As now defined, the term has reference to the least duration required by a current of exactly twice the threshold intensity (or so-called "rheobase").

The characteristic time-factor or chronaxie of a tissue also expresses itself in the rate of variation of intensity required by the stimulating current; this rate is greater the briefer the chronaxie; it is also greater the more rapidly the stimulation-process develops in the tissue, as indicated 'by the rate at which the accompanying bioelectric variation rises to its maximum. The chronaxie also varies directly with the duration of the summation-interval for subminimal stimuli.^

The stimulating effect of a current of continuously changing intensity, or of a change in the intensity of a current already traversing the irritable tissue, varies, in a manner which is characteristic for the tissue, with the rate of change, and is largely independent of the actual intensity. It is significant that this rule, relating stimulating effect to rate of change, applies also to mechanical, chemical, and other forms of stimulation, in aU of which a sudden change is more effective than a gradual one. A general property of living matter is apparently here involved. In the activation of the foregoing metalHc model (passive iron wire in nitric acid) the same rule holds; e.g., in order to activate the metal mechanically by scraping with glass, the movement must be rapid; a slow movement is ineffective. Similarly in electric activation a current which is gradually increased up to a sufficient intensity has no effect, while one of the same intensity, attained suddenly, causes instant activation.

The rate of change which a current requires in order to stimulate a tissue varies with the nature of the tissue, and is a function of the characteristic chronaxie. When the chronaxie is brief, the rate of change must be rapid. When the rate of change of an increasing current is gradual, a greater final intensity of current is needed for stimulation than when this rate is rapid. The following observations of Lucas, on the stimulation of the frog's sartorius, illustrate the conditions for a single typical tissue. The rate of change of the exciting current was controlled by varying the rate of movement of a shutter which opened and closed a slot in a partition set across a zinc sulphate solution, forming part of the stimulating circuit.^ Comparison was made between the current-strength required: (i) when the circuit was closed instantaneously; and (2) when the intensity was increased from subminimal to a stimulating value at varying rates. The muscle was also stimulated by currents of the same linear gradient or rate of change under two conditions, (A) while immersed in pure 0.7 per cent NaCl solution and (B) in a mixture of 0.65 per cent NaCl plus 0.05 per cent CaCla. The following results are typical:^

Strength of Current Required Time Required to Reach Full for Stimulation The more slowly the current changes its intensity the less effective it is as a stimulus. The sensitivity to rate of change varies with temperature and the composition of the medium; the necessary rate of change is greater at higher temperatures; it is also greater when calcium is present (B) than in the pure NaCl solution (A). According to Lucas, ''an increase in the concentration of the calcium would appear to necessitate a more rapid concentration of the ions concerned in excitation."' Observations by Mines, on the minimal duration of the threshold current of constant intensity,^ have shown that in this case also the duration is briefer when Ca is present. Such facts indicate that the chronaxie of a tissue is determined not only by its specific constitution but also by the external conditions to which it is exposed. This is well shown in certain studies, by Adrian, on the effects of peripheral nerve injury.^

The chronaxie of a tissue appears to be closely related both to the rate of response and to the rate of recovery of the irritable elements. Thus it shows a close correlation with the characteristic duration of both the bioelectric variation of the tissue and the refractory period. The more slowly a tissue responds to a constant current, i.e., the longer the minimal duration of the current of threshold intensity, the more gradual is the rate of change required for excitation by a current of changing intensity. The length of the summation-

3 The chronaxie of a muscle with nerve supply interrupted increases progressively until innervation is re-established {Archives oj Radiology and Electrotherapy, May, 191 7). interval also appears to be determined by the same conditions. The phenomenon of summation is of great importance in the analysis of the stimulation process. It shows clearly that a single subminimal stimulus produces an effect on the tissue, but that this effect is transient; within a certain brief time the tissue resumes the same condition as before the stimulus. But if before this time has elapsed a second similar stimulus is applied, its effect is added to that of the first, and the critical level of disturbance required to initiate an excitationwave may be reached. The second stimulus, in order to be effective, must be sent in before the effect of the first has subsided; and the more rapid the rate of this subsidence the shorter is the summation-interval. The summation-interval is therefore defined as the longest interval separating the successive subminimal stimuli of an effective series of two or more such stimuli.^

When the shocks were lo per cent below the threshold, the interval was much shorter. The summation-interval is thus longer the more gradual the excitation-process (the longer the chronaxie) of the tissue. It varies with temperature and with the state of the tissue. Lucas finds the temperaturecoefficient to be low {Qzo = ca. 1.3), a fact suggesting that purely physical changes, e.g., dift'usion-processes (which have a similar temperature-coefficient), are chiefly concerned in the return of the tissue to the normal after a slight disturbance.' The influence of the inorganic salts is again highly interesting. The presence of Ca shortens the summation-interv^al, just as it shortens the minimal duration of the threshold constant current and increases the rate of change required for stimulation by a changing current.^

According to Lucas and Mines, the effect of temperature on the minimal duration of the threshold current is the same as on the summation-interval. Such facts again emphasize the distinction between the local change produced by the stimulating agent and the propagated ^ Cf. Lucas' discussion, op. cit., p. 473. It is noteworthy that the time required for the return to the normal properties after complete stimulation, as measured by the length of the refractory period, is much longer than the summation-interval, and that the temperature-coefficient of this return or recovery process is high (Qio = ca. 3); these facts indicate that chemical processes play the chief part in the recovery from a complete stimulation. The "local change" may thus be of a purely physical kind (e.g., polarization change), while in the complete or propagated excitation the chemical factor is essential. This conclusion agrees with the fact that the refractory period is much longer than the summation interval. Apparently the former represents a period of metabolic and structural restitution.

effect or stimulation-process proper. The former is, or may be, a purely physical change; in electrical stimulation its essential feature is apparently a change of polarization resulting from changes of ionic concentration at the cell surface. This change does not initiate a propagated effect unless it exceeds a certain critical limit, and unless the state of the tissue is favorable; thus in an anaesthetized tissue the local change of polarization is produced by a current, but no propagated excitation follows. The differences between the physical conditions and manifestations of the two processes, local change and propagated disturbance, and the differences in their temperature-coefffcients show that the propagated process is more complex than the initiatory local process and includes chemical or metabolic factors among its chief components.

The factors determining the characteristic chronaxie of a tissue would thus appear to be largely factors determining the rate at which the critical polarization change occurs in the irritable elements. This rate depends on the rate of movement of ions and also on the ' special structural conditions within the tissue. An important advance in the theory of the local change has been made by Hill,^ who has modified Nernst's simple theory and brought it into closer conformity both with the facts of organic structure and with the actual behavior of the tissue in electrical stimulation. Hill points out that in any case of electrical stimulation the concentration-changes at two semi-permeable surfaces

situated a short distance apart must be considered; these are the surfaces where the current-lines intersect the two opposite faces of the irritable element. If these surfaces are close together, the diffusion-gradient set up by a given current is steep, and hence the backdiffusion opposing the polarization is relatively rapid. The production of the critical polarization change will then require a stronger current than with the membranes far apart. This conception explains also why a current passing crosswise through a nerve or parallel-fibered muscle is so much less effective than one passing lengthwise. Hill's calculation leads him to the formula,

current, where X is a direct function both of the proximity of the two membrane-surfaces concerned (proximity being the reciprocal of the distance apart) and of the rate of movement of the ions; i represents the intensity of the current, t its duration, and \x and B are constants having reference to the conditions of movement of the ions in the tissue. This formula gives a remarkably close agreement with observation through a wide range of intensities. A further essential feature of Hill's theory is its recognition that the polarization change is in reality merely the determining condition of a chemical change which must proceed at a certain minimal rate in order to cause excitation. From this point of view it is possible to understand why not only the degree of polarization attained, but also the rate at which this critical degree is reached determines whether stimulation shall be initiated or not. The stimulation process is a process sui generis^ distinct from the initiatory physical change;

yet it is of such a nature as to be initiated only by physical changes proceeding at more than a certain rate. We are thus led again to consider those special properties of the living system which it possesses by virtue of being living; i.e., metabolically and synthetically active. The features above of stimulation cannot be understood except by reference to what is distinctive in vital processes as such. Yet it is to be noted that the case of the living system is by no means unexampled in the respect just considered. There are many natural processes in which, if a certain effect is to be produced, the effecting agent must act at more than a certain minimal rate; if it acts slowly, the effect fails entirely. For example, a swift current of air will extinguish a candle flame, while a slow one will not; a rapid stroke will ignite a match, a rapid projectile penetrates a plate, a rapid attack succeeds in war or on the football field. The lighting of a match shows many analogies with stimulation. Ignition occurs at a certain critical temperature, which is attained when the match head is drawn uniformly over a rough surface for a certain time at more than a certain rate. If the movement is too slow, ignition will never occur; in this case the stationary condition at which the^ gain of heat from friction is equal to that lost to the surroundings by conduction and radiation is reached at a temperature below that of ignition. A movement at a certain rate must last for a certain time, which is shorter the more rapid the rate. Summation phenomena and summation intervals may also readily be demonstrated in this system; i.e., a succession of brief strokes but not a

single stroke will ignite if the interval between the strokes is not too long. It will be evident that the general condition common to all cases of this kind is that two active processes or sets of processes, with resultants acting in opposite directions, are concerned; the equilibria are ''dynamic" rather than static. In an irritable tissue through which a uniform electric current is flowing, the effect which the current produces in the direction of stimulation is presumably counterbalanced by contrary processes depending largely on the metabolic or synthetic activity of the tissue. Interruption of an already existing uniform current, as well as its sudden increase or decrease, disturbs this equilibrium and may result in stimulation.

Further analysis of the process of stimulation requires, therefore, a consideration of the special nature of the processes occurring in the irritable protoplasmic system. According to the foregoing conception of the conditions of stimulation in living cells, the primary or initiatory process is a local alteration of the protoplasmic surface-film, or ''plasma membrane," of the irritable element. This alteration has a self-propagating character, like that shown by other chemically alterable surface-films at the boundary between two electrically conducting phases, and leads secondarily to the characteristic manifestation of cell-activity, or response. If this conception of the stimulation process is a true one, all forms of stimulation should exhibit definite evidence of accompanying surface-processes of the kind indicated.

Certain effects which apparently accompany all forms of stimulation and activation, whatever the special nature of the response may be, constitute evidence of this kind. These are: (i) The bioelectric variations; (2) the presence of a refractory or temporarily inexcitable period immediately following stimulation; and (3) a temporary loss of semi-permeability or increase in the permeability of the cell surface to water-soluble substances. We have seen above that the electric current is a universal stimulating agent; and that, conversely, when irritable tissues respond, they give rise to electric currents which traverse the surroundings and may be there detected by appropriate means. Similarly, mechanical or chemical alteration of the cell surface causes excitation and also gives rise to bioelectric currents. Further, during many forms of normal excitation there is direct evidence that the surface layer of protoplasm undergoes a sudden and pronounced change in its properties, one effect of which is to increase the permeability to water-soluble substances; in many cases this change is distinct and easily demonstrated; e.g., the turgor mechanisms of plants, gland cells, and egg cells during activation; in others (nerve, muscle) the indications of changing permeability are indirect. The loss of irritability during the refractory period is also in harmony with the hypothesis that the surface layer breaks down or is otherwise altered during stimulation, since semi-permeability, implying electrical polarizability, is apparently essential for stimulation; any temporary loss of semi-permeability must therefore involve loss of irritability.

The strongest evidence that the surface-change is the essential and primary change in stimulation is the constancy with which the foregoing three manifestations of stimulation are associated. Especially significant also is the fact that closely analogous phenomena are observed in the inorganic models considered above (mercuryhydrogen peroxide catalysis, passive iron), in which the process of activation is known to depend upon the electrolytic disintegration of thin interfacial films. In the temporary activation of passive iron, the dissolution of the oxide-film involves (i) a change of potential, (2) a marked increase in permeability, allowing ready access of the acid to the metal, and (3) a delay in the recovery of the former state of susceptibility after the return of passivity. It seems highly improbable that those parallels are accidental; the indications are that they are expressions of an underlying identity in the essential structural constitution and conditions of activity of the living system and of the inorganic model. The essential features of structure and composition common to both systems are, briefly; (i) the presence in both cases of a thin film separating two electrically conducting phases, one or both of which is an electrolyte solution, and (2) the susceptibiHty of the film to alteration under the influence of electric currents (breakdown or construction by local electrolysis). Before dealing in greater detail with these parallels and their bearing on the problem of the essential constitution of living matter, it will be necessary to review briefly the essential facts which have been established with regard to the foregoing three general accompaniments of stimulation.

A complete review of this large field of research is not possible in the space at our disposal.' It is necessary, however, to consider in some detail the chief facts bearing on the present problem; these may be conveniently grouped under the two headings: (i) bioelectric potentials in resting cells and tissues; and (2) variations of bioelectric potentials in relation to physiological activity. The existence of potential-differences between the resting cell or other protoplasmic element and its surroundings has in itself nothing unexampled or surprising. Typically such potentials are found at all phaseboundaries unless special compensating conditions are present. The conditions on either side of the interface are asymmetric with respect to chemical composition and physical condition, and corresponding to this assymmetry there is an electrical asymmetry or potentialdifference. The facts of electrical convection and electrical endosmose show the presence of potentialdifferences between all kinds of insoluble materials and the adjacent layer of solution. These potentials

^ For an exhaustive account of the earlier work, cf. Biedermann's Electro physiologie, English translation. For a more recent account, cf. Bemstem's Eledrobiologie (191 2), and the article of Garten, "Produktion von Elektrizitat," in Winterstein's Handhuch der vergl. Physiol. j III (1910), 105; also the interesting but more special work of Bose, Comparative Electro physiology (1907). are known to vary with the composition of the nonaqueous phase and with the electrolyte content of the aqueous phase; i.e., ions have a special influence, although surface-active substances other than electrolytes may also have an effect, as already shown. The case of the living cell falls partly in this general category. Suspended cells travel in the electric field, usually toward the anode; and the rate and even the direction of this travel may be changed, just as in non-living suspended particles, by changing the electrolyte content of the medium; especially active in this regard are H ions and the ions of polyvalent metals. Constant currents passed through living tissues (muscle) effect transport of fluid. This is a special case of electrical endosmose; evidently any passage of electricity through a cell must involve some displacement or transport of fluid, a fact which must be considered in relation to physiological processes hke secretion, absorption, and cell-division. The potentials between suspended particles and suspension-media are apparently in large part adsorption potentials; their range is comparatively narrow, usually between 0.02 and 0.05 volt; according to Freundlich they represent the potentials between an adhering immobile layer of solution and the mobile layer adjoining.^

These facts, while relevant to the general theory of ^ the bioelectric potentials, do not in themselves explain sufficiently the special peculiarities of the latter. Apparently the closest resemblances are with electrode potentials; e.g., those between a metal and an adjoining solution. From the physiological point of view the most ^ Freundlich, Kapillarchemie, p. 243 ; Report on the Physics and Chemistry of Colloids, Faraday Society and Physical Society of London (London, 1921), p. 146.

significant fact is that the demarcation potentials (shown in the injury-currents of muscle and similar phenomena) vary with the condition of the protoplasmic surface layer. The loss of semi-permeability accompanying death is always associated with a decline or disappearance of the demarcation potential, a fact indicating that the latter depends on the presence of a semipermeable partition between the internal protoplasm and the surrounding medium. Such a partition allows the existence of permanent differences of electrolytecontent between the solutions adjoining the outer and inner faces of the plasma membrane, and so provides the asymmetric conditions necessary for a potentialdifference. Just why this potential-difference should have the observed orientation (positive externally) and range (of the order of 0.05 to o.i volt) is not entirely clear; possibly these conditions are referable to a higher total electrolyte content of the cell interior as. compared with the surroundings, or to a preponderance of certain ions (e.g., H ions) in the cell interior. The chemical relations between the ions present in solution and the materials composing the surface-film are undoubtedly an important factor, and it seems probable that oxidationreduction potentials and adsorption potentials (in the Freundlich sense) are also concerned. The total potential as observed thus represents an additive effect.