Protoplasmic Action and Nervous Action

It is evident, however, that the demarcation potential depends primarily on the special physical and chemical properties of the cell surface, since whatever modifies the structure or chemical character of the protoplasmic surface layer also alters the potential. This is shown by the conditions under which the so-called currents of injury arise. Mechanical or other injury and the application of cytolytic substances, which demonstrably increase permeability, always lower the potential; i.e., the altered region becomes negative relatively to unaltered regions. This effect is often reversible if the tissue is not exposed too long; thus potassium salts render a voluntary muscle locally negative, in accordance with their permeabiHty-increasing action; and if the tissue is soon afterward bathed in Ringer's solution, the original isoelectric condition returns. Any local alteration which impairs semi-permeability thus induces local negativity; i.e., decreases the potential-difference between the protoplasm and the surroundings. The fact that the variation of potential is always in a negative direction is consistent with the theory that the normal negative variation accompanying stimulation is also the effect of an alteration of the cell surface, involving a temporary and rapidly reversed increase of permeabiHty. We may thus understand why the bioelectric variation of stimulation is similar in its direction and range to that accompanying loss of semi-permeability, while differing in being reversible or evanescent.

It has long been recognized that variations in the permeability of a semi-permeable partition separating two electrolyte solutions must involve variations in the potential difference across the partition, and the chief modern attempts to explain the bioelectric potentials have been based on this ground ("membrane theory" of Ostwald, 1890,^ followed by Cybulsky, Bernstein,'' and others). Ostwald's original suggestion was that the plasma membrane may act as an "ion sieve," allowing the cations of some intracellular electrolyte to pass but not the anions. This hypothesis was adopted by Bernstein as affording a point of view from which the sudden fall of potential during stimulation might be explained; at this time Bernstein supposed the membrane to become permeable to both classes of ions. The actual conditions, however, are undoubtedly more complex, and include other factors than the simple diffusion potentials considered by Ostwald and Bernstein. Nevertheless, it must be recognized that the breakdown of a semi-permeable partition between the protoplasm and its medium (the two adjoining electrolyte solutions concerned) must decrease the potential between the two, whatever the detailed conditions of this potential may be. The "membrane theory" of the bioelectric potentials need not necessarily have the form of a. modified diffusion theory, as some of its opponents seem to have supposed. More recent developments of this theory have aimed at correlating the bioelectric phenomena with the chemical as well as the physical processes occurring in the protoplasmic boundary layers. It may now be taken as well established that the cell surface possesses electrode-like properties, and that in the determination of its electromotor behavior other conditions enter than merely a selective or differential hindrance to the diffusion of ions.

The work of Macdonald^ is of special interest since it first showed that the demarcation-potential of nerve varies with the concentration of the salts in the adjoining solution in the same manner as the potential difference between a metallic electrode and its adjoining solution (e.g., Zn in ZnS04 solution). When the demarcation potential between the cut surface of the sciatic nerve and an uninjured area one centimeter distant was measured (by the usual compensation arrangement), after leaving the nerve for five minutes in differently concentrated solutions of a given salt, results of the following kind were obtained. (The potential in the original physiological salt solution is represented in the bracketed expression by E.)

It will be noted that the potential difference increases with increasing dilution, and very nearly in direct proportion to the logarithm of the dilution. This is the characteristic relation found in electrode potentials, and Ci represents the concentration corresponding to zero potential (equivalent to that required to compensate the ionic solution-pressure of the metal) , and C2 the concentration of the ions in solution; thus with zinc in contact with ZnS04 solution, the potential decreases with increase in the concentration of the zinc ions in a logarithmic curve. The fact that a similar relation is found with the living tissue, and that the results obtained are reversible, as Macdonald found, shows that the living tissue behaves as if its surface were an electrode reversible

with respect to the cations of the solution. Results similar to the foregoing were obtained also with NaCl and HCl. More recently Loeb and Beutner/ in an extended and important series of researches, have shown that the characteristic logarithmic relation between the concentration of the ions in the solution and the potential difference holds for organic membranes of a variety of kinds and also for solutions of lipoids in organic solvents. The organic membranes act as if they were reversible to cations as a class. This result is highly significant, for it seems to imply that reversible combinations between these ions and components of the membrane (e.g., proteins or lipoids) occur, and that the formation of these combinations is the essential factor determining the potential equilibria observed in a given solution. Just as a metallic electrode, like Zn in contact with a solution of ZnS04, may be regarded as '^dissociating off" zinc ions until an equilibrium (to which corresponds a definite potential difference) exists between the ions tending to pass into solution from the metal and those already in solution, so in the case of a salt solution in contact with an organic membrane a certain potential difference corresponds to the equilibrium existing between the ions in solution and the ion-membrane compounds formed by the combination of these ions and the membrane components (proteins, etc.). Any increase of the ions in solution decreases the potential difference in logarithmic ratio.

The demarcation potentials of living tissues appear to exemplify this general condition/ In the usual method of studying these potentials in muscle or nerve the normal cell surface in part of the tissue is first altered by mechanical or other means. The electrodes leading to the galvanometer then touch two surfaces, altered and normal, which differ in their physico-chemical condition; a corresponding potential difference is shown, the injured region being negative. Hence the term ''alteration current" for the current between the two regions. Loeb and Beutner found the same to be true for a simple organic membrane like an apple skin; when one region is crushed, this region shows itself negative to an unaltered region; potential differences of 20 to 100 millivolts were observed in different experiments, the values varying with the concentration of the salt solution in contact with the tissue.^

Loeb and Beutner found also that the potential changed with changes in the concentration of the surrounding solution in a manner similar to that observed by Macdonald for nerve. To produce a constant arithmetic change in the potential-difference, the concentration of the salt had to be changed in a constant ratio; i.e., to a geometric series of concentrations ^ The importance of the Donnan membrane potential (potential across a membrane permeable to only a part of the ions present), in the case of protein solutions separated from electrolyte solutions by collodion or similar membranes, has recently been demonstrated by Loeb in an important series of researches (summarized in his book. Proteins and the Theory of Colloidal Behavior). In this case a colloidal ion (protein) is the one to which the membrane is impermeable. In living tissues, however, with protein ions in about equal concentration on both sides of the membrane (in protoplasm and in lymph) this source of potential can scarcely play a part.

corresponds a linear series of potentials, the relation characteristic of electrode-potentials in general. That the effect observed in any single case depends on the special character of the surface was shown in a series of experiments in which Loeb and Beutner compared the effects of varying the concentration of the solution in contact with (A) the uninjured surface of an apple, and (B) a surface from which the skin had been removed/ In all such experiments one electrode in contact with the apple remained unchanged; the other was connected with the solution which was varied; the latter was in contact with another portion of the surface at some distance from the first electrode. A quadrant electrometer was used.

Both surfaces show the same kind of variation with varying concentration of electrolyte, but the altered area shows a smaller change of potential for a given change of concentration; thus on the average a tenfold dilution increases the positivity of the unaltered surface by about 0.06 volt, and of the altered by about^o.03 volt. With every solution used the altered surface exhibits the lower potential; i.e., is negative relatively to the unaltered; the conditions also suggest that it represents an area which is reversible to anions as well as to cations,^

2 This would correspond to freer penetration by anions and an entrance of diffusion potentials in the total effect. in which also the side containing the dilute solution is positive; AgCl may be substituted for metalhc silver with the same result. Similarly the arrangement: gave potential differences of a similar order to those found with the organic structure, and varying similarly with the concentration of the dilute salt solution. In other words, the surface of the non-aqueous phase acts

in the same manner as an electrode reversible to cations. The remarkable feature is that the reversibility relates to salts of cations in general, and not only to those of a single cation, as in the case of silver or other metallic electrode. All of the alkali and alkali-earth cations (those of the chief physiological interest) gave typical results; thus with KCl the following observations were made;^ the non-aqueous phase was salicylic aldehyde saturated with salicylic acid:

It was further shown that this effect is dependent on the acid character of the non-aqueous phase; i.e., on its ability to take up cations (reversibly) by salt formation; other solutions of weak acids, e.g., of benzoic acid in phenol, behaved similarly. But when the nonaqueous phase is basic in chemical character, the potential changes in the opposite direction, the dilute solution becoming more negative, instead of more positive, with increasing dilution, and the behavior is such as to indicate reversibility to anions. Beutner found this to be the case when the weakly basic compounds, anihne and toluidine, were used as the water-insoluble phase in an arrangement similar to the foregoing.^

^ For a complete account of Beutner's investigations cf. his recent book, Entstehung eleklrischer Strome in lebenden Geweben (Stuttgart, Such facts suggest that substances having the properties of weak acids determine the type of electromotor behavior shown by the cell surfaces in the demarcation potentials. Both proteins and lipoids (e.g., lecithin) belong in this class. Loeb and Beutner^ therefore carried out further experiments of the kind described, using solutions of lecithin in organic solvents, and found again the same relation between the concentration of the salt and the observed potential difference. When lo per cent solutions of lecithin in guaiacol were used, the behavior, both quahtative and quantitative, was found closely similar to that of plant tissues. Oleic and palmitic acids gave similar results, but not cholesterol. Extracts of various plant and animal tissues (muscle, brain, frogskin) in organic solvents also exhibited this behavior; a further interesting fact was that potassium salts had a greater influence than sodium salts in altering the potentials, a pecuHarity which the authors ascribe to a greater solubiHty of potassium salts in the lipoid phase.

It would seem, therefore, as if the ordinary potential differences observed between altered and intact portions of the cell surface were phenomena of the same general type as those just described; i.e., referable to the presence of a water-insoluble phase containing weakly acid substances and forming a thin film or partition between two dissimilar electrolyte solutions, represented respectively by the living protoplasm and its surrounding medium.^

' As to the specific nature of the electrolytes concerned, little definite can be said at present. In general the two cations whose concentration is higher inside than outside the cell are K and H. A recent calculation The sudden fluctuations of potential accompanying normal vital processes like stimulation evidently require a different type of explanation. In such cases rapid and reversible alterations of the electromotor properties of the protoplasmic surfaces are apparently involved ; these effects can only be referred to the chemical or metabolic processes characteristic of living matter. The external surface layer of the living cell consists of a thin film of chemically alterable material in immediate contact with the surrounding medium on the one side, and with the internal protoplasm on the other; it is, therefore, subject not only to purely physical changes, such as local thinning or interruption, but also to changes of chemical composition, resulting from variations in oxidative or other metabolism. Along with such alterations must go alterations in physical properties, thickness, permeability to electrolytes, acid or basic character, etc.; and these must alter correspondingly the electromotor properties of the cell surface. It has already been pointed out that the reversible variations of potential seen in the action-currents of tissues like muscle and nerve have a range closely similar to that of the demarcation-currents {ca. 0.05 volt); and this fact receives a consistent

by Adams favors the idea that the difference between the internal and external H-ion concentrations is an important factor in the bioelectric potentials {Journal of Physical Chemistry^ XXVI [1922], 639). Recently Rohonyi has opposed Beutner's conception of the importance of an oil-like phase in the determination of the bioelectric potentials; cf. his critique: Biochem. Zeitschrift, CXXX (1922), 68. He regards semi-permeability (permeability to water, but not to electrolytes) as the essential property.

explanation on the hypothesis that dissolution^ and re-formation of the semi-permeable surface layers of the cells are the main factors in the production of the normal electromotor variations. The specifically vital functions of metabolic construction and destruction which determine the physical properties of the cell structures would thus determine also the normal variations of the bioelectric potentials. The general physico-chemical conditions of these potentials are of a kind present at all phase-boundaries; but the special peculiarities of the protoplasma boundary layers, and hence of the dependent electromotor phenomena, are determined by the specific metabolic activities of the protoplasm and vary with these activities. It is evident that metabolic destruction and re-formation of surface-films would involve electromotor variations; and in those cases where the filmmaterial is susceptible to chemical alteration (e.g., oxidation or reduction) under the influence of the local electric currents thus arising, the conditions would also be furnished for the processes of spreading and transmission, which are essential to stimulation. Conditions closely resembling in their general features those just defined are in fact reaUzed in the passive iron model and related inorganic systems described above.

The present view, therefore, refers the normal bioelectric phenomena to variations in the phase-boundary ^ The precise nature and degree of this alteration are unknown; the term "dissolution" may be regarded as indicating an alteration sufficient to deprive the surface layer temporarily of its properties as a membrane, i.e., as a semi-permeable partition. This effect is equivalent to increase of permeability. potentials of the polyphasic living system, protoplasm, which in its nature is subject (especially if highly ''irritable") to rapid variations of chemical or metabolic activity. Such variations imply corresponding alterations (breakdown, construction, etc.) of those protoplasmic structures, including the interfacial films, whose formation and maintenance depend on this metabolic activity. Apparently during the normal stimulation of any irritable cell, e.g., a muscle cell, the electromotor properties of the cell surface change in such a manner that the surface adopts temporarily properties like those of an injured or altered surface; i.e., one which has lost its semi-permeability. This effect, the result of a metabolic change of some kind, in which oxidations probably play a chief part, is a reversible one; consequently a reversible electromotor variation accompanies it. It is as if the injury-current were temporary, and lasted for only a brief period, whose duration depends on the rate at which a complete surface layer with the original properties can be re-formed. It has already been pointed out that certain forms of injury-current, those caused by KCl solution, are reversible if the injury is not too extensive.^ On such a hypothesis we may understand why the whole reversible bioelectric variation occupies a definite time, characteristic for each irritable tissue; this time is determined by the tissue's own specific rate of metaboHc construction and destruction. The special reaction-velocities characteristic of the

^ It is well kno\\Ti that with excessive stimulation of any kind the return of the normal positivity is delayed or incomplete (Hermann and others). All transitions from complete reversibility to irreversibility can be obtained. (Cf. Ebbecke, " Membrananderung u. Nervenerregung," Arch. ges. Physiol, CXCV [1922], 555; cf. pp. 581 £f.) particular living system under consideration thus determine the duration and other special features of its action-currents, and hence also the time-relations of the dependent or correlated phenomena; e.g., velocity of transmission, refractory period, summation, and chronaxie.

The passive iron model again affords a clear and simple illustration of the manner in which rapid electromotor fluctuations can result from changes in the character of the boundary layer between the two chemically interacting phases. When such a wire, immersed in a solution of nitric acid and connected through a voltmeter with an indifferent electrode (platinum wire) also immersed in the acid, is activated, a sudden change of potential (of about 0.7 volt) is observed. With strong acid (60 vols, per cent 1.42 acid or stronger) this variation is automatically and rapidly reversed and the metal resumes its former potential within a second or two; this reversal is a result of the re-deposition of the passivating surface-film; the chemical reaction then ceases. Under certain conditions the return of complete passivity may be delayed, or rhythmical fluctuations of potential and chemical activity may occur; the latter phenomenon is frequent in a somewhat weaker acid (between 50 and 55 volumes per cent), and depends upon the alternating formation and dissolution of the passivating film.^ Many striking electrical phenomena, having features which have been regarded as especially characteristic of bioelectric processes, are in fact exhibited by this model. These phenomena show that rapid variations of

potential, with associated chemical effects (shown in the rapid breakdown and replacement of the surfacefilm of oxidation-product), may occur in a three-phase system of relatively simple constitution, when the third phase has the form of a thin chemically alterable film between the other two. In a chemically reactive and film-partitioned system such as living protoplasm processes of a similar kind are to be expected; such processes would here also necessarily be associated with variations of potential. And conversely, since in all processes of this type the factors controlling the formation and breakdown of the films are mainly electrical, such systems would be influenced in their chemical activity by electric currents passing through them from external sources. This is in fact true both for metallic systems of this type (which include the mercury-peroxide system) and for living protoplasm.

In living organisms variations of electrical potential are associated with physiological activities of all kinds; and for detailed descriptions of the bioelectric phenomena and for a review of the extensive literature reference must be made to the special treatises on electrophysiology. Bioelectric currents have been shown to accompany the following vital processes:^ automatic^ or reflex activity of the central nervous system, rhythmical processes like the heartbeat or the activity of automatic nerve cells, muscular contraction, nervous and other forms of protoplasmic transmission, glsCndular secretion, stimulation of special sense receptors (retinal currents) , the movements

^ For a more detailed account cf. Garten's article, "Die Produktion von Elektrizitat," loc. cit., 1910. of plants (Mimosa, Venus' fly-trap), and growth processes; they are probably also associated with celldivision and ciliary movement. They appear in fact to be as essential a feature of protoplasmic action as the consumption of oxygen or the evolution of CO2. The fact that their rate of development and their rhythm are influenced by changes of temperature in the manner characteristic of chemical reactions (Qio = 2-3)^ indicates their dependence upon the fundamental metabolic processes of protoplasm. Direct proof that the bioelectric rhythms are accompanied by rhythmical chemical reactions is at present lacking, but there are many indications that this is the case. The evidence is clearest in those instances where the rhythm is slow. Thus the production of CO2 by dividing sea-urchin eggs follows a rhythm which runs parallel with the rhythm of cleavage;^ the latter rhythm is accompanied by a parallel rhythm of variation in the physical properties of the egg surface;^ and this rhythm is almost certainly associated with a variation of potential."* In a certain sense it is self-evident that the energy of the bioelectric currents, as of other organic activities, represents the transformed energy of chemical reactions;

* Cf. Piper's results on tortoise muscle {Elektrophysiologie menschlicher Muskeln, Berlin [1911], chap, ix, 130). Cf. also the data in Garten's article, loc. ciL; also Lucas, Journal of Physiology, XXXIX (1909), 3 Cf . my paper on the physiology of cell-division in the Journal of Experimental Zoology, XXI (191 6), 369. < This is indicated by Miss Hyde's observations in the Fundulus e?g, American Journal of Physiology, XII (1904), 241. but the problem of the precise nature of the transformation still presents many difficulties.

The special development of bioelectric currents as a means of attack and defense in the electric fishes is a fortunate circumstance for general physiology, since the structural conditions found in the electric organs are full of suggestion for the general theory of the bioelectric processes. These conditions indicate clearly, first, that the powerful effects produced by these organs depend on the summation of the potentials of numerous cellular elements or ''disks" (apparently modified muscle cells) arranged in series; and, second, that the action of each single element depends on the alteration of a special and definitely oriented portion of its protoplasmic surface layer. This area is structurally characterized by the rich branching of the nerveterminals in contact with it; it forms one face of each element, posteriorly directed in the case of the electric eel, while the opposite or anterior face is free from nerve fibers.^ One face of each element is thus innervated and apparently undergoes alteration or activation during activity, in a manner which may be compared with that of an activated gland cell, while the opposite face presumably remains unchanged. The innervated and non-innervated surface layers of adjacent elements thus alternate in position in a manner comparable with the alternation of positive and negative metallic plates in a battery of galvanic elements in series. A closer comparison would be with the original galvanic ''pile," where each pair of plates, copper

^ Cf. Gotch's article in Schafer's textbook, II, 561, for an account of the essential structure of the electric organ; also Biedermann's Electro physiology and Bernstein's Elektrohiologie. and zinc, in direct contact with each other, is separated from the next pair in the series by disks of cloth or paper soaked with electrolyte solution. Similarly the innervated and non-innervated surfaces of two adjacent elements are in contact, while the two surfaces of the same element are separated by the mass of internal protoplasm which is modified in a characteristic manner. This arrangement constitutes strong evidence in favor of the theory that the surface-films or plasma membranes of these cellular elements play essentially the same part as the electrodes or metallic plates in batteries.^ In the active electric organ the current of the discharge (positive stream) runs within each cell or element from the innervated to the non-innervated surface, just as in the usual type of bioelectric circuit (e.g., of a single muscle cell) the intracellular direction of the current is from active to inactive (i.e., in the external medium from inactive or ''positive" to active or '^negative"). Although the E.M.F. of each cellular element is small, apparently the same as that of a single muscle cell (viz., 0.04-0.05 volt), a high potential between the terminals of the series is attained by means of the summation of many elements. Briinings has shown that by arranging several frogs' muscles in series, with the cut surface of one apposed to the uninjured surface of the next, a summation of potentials may be obtained.^ The structure of the electric organ is thus in striking conformity with the theory that the bioelectric currents originate in a manner essentially similar to that

^ Compare Bernstein's account in his Elektrobiologie, chap, vi, p. 121. * Brunings, Arch. ges. Physiol., XCVIII (1903), 241. According to Briinings, potentials of a volt or more can be obtained by arranging muscles in series. of the currents produced by the usual combinations of metallic electrodes and electrolyte solutions (or so-called ^'batteries"), with the difference that in the living system the electromotor surfaces consist of thin protoplasmic films having a composition and structure which are subject to rapid variation.

The potential changes are relatively small in single cellular elements and their approximate range may be readily determined in parallel-fibered muscles like the frog's sartorius. Here, with a symmetrical side-by-side arrangement of the elements, there can be no summation of potentials; and the conditions are like those of a battery arranged ''in parallel." The maximum range of variation during contraction does not usually appear to exceed 0.05 volt, a potential-difference similar to that of the demarcation-current. According to some observers, however, the potential of the action-current in muscle during strong contraction may be greater than that of the demarcation current, and may even attain 0.08 volt. Comparative observations of the demarcation-current potentials throughout a wide range of invertebrate and vertebrate forms give magnitudes of the order of 0.03 to 0.05 volt.^ The exact physico-chemical significance of these values cannot be stated at present. Bernstein has investigated the influence of temperature on the demarcation potential of muscle, and finds that within the physiological range (from 5° to 30°), its magnitude is closely proportional to the absolute temperature, as in the case of electrode or diffusion potentials.*

Considering the fact that the potentials as measured must be less than those actually existing in the living tissue — because of the partial cross-circuiting of the current through the fluids of the tissue — it seems probable that in typical active tissues the normal fluctuations of potential in single cells have a range of 50 to 100 millivolts. The bioelectric currents are thus weak; their intensity, however, is amply sufficient to excite sensitive tissues, as is demonstrated in the laboratory experiments in which nerves and muscles are excited by demarcation and action currents ('^rheoscopic frog" experiments) .

In each special instance the duration, the rate of variation, and the rhythm — in the cases where the variations are rhythmical — exhibit special features characteristic of the tissue and of the species. Comparison of the conditions in different tissues and animals reveals the existence of highly significant correlations between the timerelations of the electrical response and of the normal functional response or mode of activity of the tissue. When the bioelectric variation develops rapidly and is of brief duration, the response of the tissue to stimulation is also rapid; e.g., in a muscle the duration of the latent period and of the single twitch is brief, the chronaxie is also brief, and the propagation of the excitation-wave is rapid; the refractory period and the summationinterval are also brief. On the other hand, tissues with slowly developing bioelectric variations exhibit a slower rate of response and a slower subsidence of their activity; the muscular twitch, the chronaxie, the refractory period, and the summation-interval are relatively prolonged and the transmission is slow.