THE STRUCTURE OF THE COLLODION MEMBRANE AND ITS ELECTRICAL BEHAVIOR : XI. THE PREPARATION AND PROPERTIES OF "MEGAPERMSELECTIVE" COLLODION MEMBRANES COMBINING EXTREME IONIC SELECTIVITY WITH HIGH PERMEABILITY

1. The electronegative membranes described in the literature which show a high degree of ionic selectivity (permitting cations to pass and restricting the anions) have serious shortcomings: their absolute permeability is extremely low, much too small for convenient experimentation; their ionic selectivity in most cases is not as perfect as would be desirable, and is moreover adversely affected by prolonged contact with electrolyte solutions. 2. A method has been worked out to prepare membranes substantially free from these defects. Porous collodion membranes were cast on the outside of rotating tubes and then oxidized with 1 M NaOH. By allowing the oxidized porous membranes to dry in air on the tubes membranes of desirable properties are obtained. These membranes are smooth, have a well defined shape, and allow considerable handling without breaking. 3. This new type membrane when tested for ionic selectivity by the measurement of the "characteristic concentration potential," consistently gives potentials of 54 to 55 mv., the maximum thermodynamically possible value (at 25°C.) being 55.1 mv. This high degree of ionic selectivity is not lost on prolonged contact with water, and is only very slowly affected by electrolyte solutions. 4. The absolute permeability of the new type membranes can be varied over a very wide range by changing the time of oxidation. Under optimum conditions membranes can be obtained with a resistance in 0.1 N KCl solution of only 0.5 ohms per 50 cm.² membrane area. The absolute rate of cation exchange through these membranes between solutions of different uni-univalent electrolytes is very high, in one case, e.g. 0.9 m.eq. cations per 4 hours, the anion leak being 0.02 m.eq. Thus, the absolute permeability of the new type membranes is two to four orders of magnitude greater than the permeability of the dried collodion membranes and the oxidized ("activated") dried collodion membranes used heretofore. Because of the characteristic properties of the new type membranes the term "megapermselective" (or "permselective") collodion membranes is proposed for them.     

THE STRUCTURE OF THE COLLODION MEMBRANE AND ITS ELECTRICAL BEHAVIOR : I. THE BEHAVIOR AND PROPERTIES OF COMMERCIAL COLLODION

1. The electrochemical behavior of membranes prepared from commercial collodion preparations varies widely, some preparations showing very high, other ones very low electrochemical efficiency ("activity"). 2. The electrochemical activity of a collodion membrane depends entirely upon impurities of an acidic nature contained in the collodion used for casting the membrane. 3. The active acidic impurities are substantially due to partial oxidation which occurs in the manufacturing process. Sulfuric acid compounds; e.g., acid sulfuric acid esters play only a minor rôle, if any. 4. The electrochemical behavior of collodion membranes in solutions of strong electrolytes is decisively dependent upon the acidic groups built permanently into the collodion surfaces. Preferential ion adsorption plays only a minor, if any, rôle.     

THE STRUCTURE OF THE COLLODION MEMBRANE AND ITS ELECTRICAL BEHAVIOR : III. THE BASE EXCHANGE PROPERTIES OF COLLODION

1. Theoretical considerations lead to the conclusion that dissociable acidic groups present to a varying extent in different collodion preparations determine the electrochemical behavior of membranes cast from these preparations. It is further reasoned that the base exchange capacity of the collodion surfaces is the true quantitative measure of the abundance of the dissociable groups. 2. The concept of base exchange capacity and the base exchange method are discussed. The conditions which allow a purposeful application of the latter are stated. 3. The base exchange properties of a number of fibrous collodion preparations of different origins and after various types of treatment, having widely varying electrochemical activities, are determined. 4. With the chemical (titration) and physical (electrometric) methods employed, no regular correlation can be found between electrochemical activity and base exchange. The base exchange capacity which is necessary to cause even great electrochemical activity of collodion is extremely small. 5. Measurable to high base exchange capacity always seems to be associated with good or high electrochemical activity; but base exchange capacity too low to be definitely measurable with the available methods may be found with collodion preparations of high as well as with preparations of low electrochemical activity. 6. The bearing of these results upon the problem of the spatial and electrical structure of the collodion membrane is indicated briefly.     

THE STRUCTURE OF THE COLLODION MEMBRANE AND ITS ELECTRICAL BEHAVIOR : VIII. QUANTITATIVE STUDIES CONCERNING THE ACIDIC PROPERTIES OF COLLODION AND THEIR CORRELATION WITH MEMBRANE STRUCTURE AND ACTIVITY

1. The electrochemical behavior ("activity") of collodion membranes depends upon acidic, dissociable groups located in the interstices of the membranes. The active groups can be determined by base exchange measurements. High base exchange capacity is always found with preparations of great "electrochemical activity;" medium and low base exchange capacities occur with electrochemically active as well as with inactive preparations. The observed base exchange capacity is determined by two factors: the inherent acidity of the collodion (its mean equivalent weight) and the submicroscopic micellar structure of the collodion. A comparison of the base exchange capacity of various collodion preparations and their inherent acidities therefore allows certain conclusions to be drawn concerning the relative availability of the micellar surfaces in the different preparations. 2. The inherent acidity of various collodion preparations, their "acid number," was determined by electrometric titration. Collodion in the acidic state, i.e. after exchange of all other cations for H⁺ ions, was titrated in an organic solvent mixture with alcoholic KOH using a quinhydrone electrode. Details of the experimental procedure are given in the paper. The acid numbers, expressed in milliliters of 0.01 N KOH per gram dry collodion, vary from 1.0 for a highly purified collodion preparation of very low electrochemical activity to 3.3 for a highly oxidized sample of very high activity. Acid numbers of about 1.5 (corresponding to an equivalent weight of about 67,000) are found both with inactive commercial and with fairly active oxidized preparations. The base exchange capacity of the same preparations in the fibrous state as measured after 48 hours of exchange time varies from 0.0013 ml. 0.01 N NaOH per gm. dry collodion for the most inactive preparation up to 0.26 ml. 0.01 N NaOH per gm. for the most active preparation. Thus the acid numbers over the whole range investigated differ only in the ratio of 1:3.3, whereas the base exchange values differ in the range of 1:200. 3. In the inactive preparation only one in 770 acid groups is available for base exchange, in the most active collodion one group in 13; values between these extremes are found with commercial and alcohol purified oxidized preparations. 4. The high base exchange capacity of the electrochemically active preparations is not so much due to their higher acid number as to their more open structure. This difference in structure is ascribed to the presence of a small fraction of low molecular weight material which inhibits normal formation and arrangement of the micelles. 5. Short time base exchange experiments with fibrous collodion indicate that the number of acid groups available for the typical electrochemical membrane functions may be estimated to be about 50 to 1000 times less numerous than those found in the 48 hour base exchange experiments. It is estimated that in membranes prepared even from the most active collodion not more than one in 500 acid groups may be available for the typical membrane functions; with the less active preparations this ratio is estimated to be as high as one in 1,000,000 or more.     

THE STRUCTURE OF THE COLLODION MEMBRANE AND ITS ELECTRICAL BEHAVIOR : V. THE INFLUENCE OF THE THICKNESS OF DRIED COLLODION MEMBRANES UPON THEIR ELECTROMOTIVE BEHAVIOR

1. Experiments were carried out to decide whether or not the electromotive properties of dried collodion membranes depend upon their thickness. 2. A number of dried collodion membranes of varying thickness, 3–160 µ, were prepared from collodion preparations of different electrochemical activity. The characteristic concentration potentials across them were measured and the means of these values determined for each thickness. 3. The characteristic concentration potentials across dried collodion membranes are a function of their thickness. The thinnest membranes yield in all cases the lowest concentration potentials; increasingly thicker membranes give increasingly higher potential values, until a constant value is reached which is characteristic of the particular collodion preparation used. With electrochemically active collodion, characteristic concentration potentials approaching the thermodynamically possible maximum are obtained with membranes of only 10 µ thickness, thinner membranes giving appreciably lower values. With two rather inactive commercial collodion preparations the characteristic concentration potential increases from about 30 mv. for membranes 3 µ thick to about 42 mv. for 20 µ membranes; still thicker membranes do not show a significant increase in the potential values. With a highly purified collodion preparation the constant maximum value was found to be about 32 mv., 4 µ thick membranes giving only about 22 mv. 4. These results do not support the homogeneous phase theory as applied to the dried collodion membrane. They are readily compatible with the micellar-structural theory. Several special possible cases of the latter as applied to the dried collodion membrane are discussed.     

THE STRUCTURE OF THE COLLODION MEMBRANE AND ITS ELECTRICAL BEHAVIOR : IX. WATER UPTAKE AND SWELLING OF COLLODION MEMBRANES IN AQUEOUS SOLUTIONS OF ORGANIC ELECTROLYTES AND NON-ELECTROLYTES

1. Dried collodion membranes are known to swell in water and to the same limited extent also in solutions of strong inorganic electrolytes (Carr and Sollner). The present investigation shows that in solutions of organic electrolytes and non-electrolytes, the swelling of dried collodion membranes is not as uniform, but depends on the nature of the solute. 2. The solutions of typically "hydrophilic" substances, e.g., glycerine, glucose, and citric acid, swell collodion membranes only to the same extent as water and solutions of strong electrolytes. In solutions of typically carbophilic substances (e.g., butyric acid, valeric acid, isobutyl alcohol, valeramide, phenol, and m-nitrophenol) the swelling of the membranes is much stronger than in water, according to the concentration used. For the brand of collodion used the swelling in 0.5 M solution was in some cases as high as 26 per cent of the original volume, as compared to 6 to 7 per cent in water. Therefore, in these solutions the "water-wetted dried" collodion membrane is not rigid, inert, and non-swelling, but behaves as a swelling membrane. 3. The solutes which cause an increased swelling of the membranes are accumulated in the latter, the degree of accumulation being markedly parallel with the degree of their specific swelling action. 4. The anomalously high permeabilities of certain carbophilic organic solutes reported by Michaelis, Collander, and Höber find an explanation in the specific interaction of these substances with collodion. 5. The use of the collodion membrane as a model of the ideal porous membrane is restricted to those instances in which no specific interaction occurs between the solute and the collodion.     

THE STRUCTURE OF THE COLLODION MEMBRANE AND ITS ELECTRICAL BEHAVIOR : VI. THE PROTAMINE-COLLODION MEMBRANE,A NEW ELECTROPOSITIVE MEMBRANE

1. Strongly electropositive porous membranes were prepared by the adsorption of protamine (salmine) on porous collodion membranes. These membranes retain their electrochemical chracteristics for at least 12 months without change. 2. They are distinctly electropositive between pH 1 and 10, the range of most pronounced electropositive behavior occurring in solutions between pH 3 and pH 8. The filtration rates and ohmic resistance of these membranes do not differ significantly from similar uncoated membranes. 3. The porous protamine-collodion membranes show very pronounced positive anomalous osmosis, the observed effects with proper electrolytes being similar to those obtained with oxidized collodion membranes. They also show very conspicuous negative osmosis with strong acids. 4. Protamine-collodion membranes which correspond in their properties to the activated dried collodion membranes were prepared by the adsorption of protamine on porous collodion membranes followed by drying in air. The concentration potentials across such dried protamine-collodion membranes closely approach the thermodynamically possible maximum.     

THE STRUCTURE OF THE COLLODION MEMBRANE AND ITS ELECTRICAL BEHAVIOR : X. AN EXPERIMENTAL TEST OF SOME ASPECTS OF THE TEORELL AND MEYER-SIEVERS THEORIES OF ELECTRICAL MEMBRANE BEHAVIOR

1. The Teorell, Meyer-Sievers theory characterizes the electrochemical behavior of membranes by their selectivity constant "A_p" which is derived conventionally from concentration potential measurements at various concentration levels. The selectivity constant may, however, be derived also from entirely independent, different experimental data, namely base exchange studies. The constants arrived at in this second way are designated as "A_b." The selectivity constants derived by these two methods must be in reasonable, at least semiquantitative agreement if the basic assumptions of the theory are correct. 2. The selectivity constants A_p and A_b were determined for eleven different sets of membranes of different electrochemical activity and of different (8.2 to 80 volume per cent) water content. 3. The potentiometric selectivity constants A_p are in most cases several orders of magnitude greater than the corresponding A_b values. With membranes of great porosity and high electrochemical activity the A_b values approach at least in order of magnitude the A_p values. 4. It is concluded that the unexpectedly large discrepancy between the A_p and A_b values is due to some inherent weakness of the Teorell, Meyer-Sievers theory, most likely to its neglect of any structural factors.     

THE STRUCTURE OF THE COLLODION MEMBRANE AND ITS ELECTRICAL BEHAVIOR : IV. THE RELATIVE MERITS OF THE HOMOGENEOUS PHASE THEORY AND THE MICELLAR-STRUCTURAL THEORY AS APPLIED TO THE DRIED COLLODION MEMBRANE

1. Experiments were carried out to decide whether a homogeneous phase (solubility) theory or a micellar-structural theory more adequately describes the behavior of dried collodion membranes with solutions of strong electrolytes. 2. A number of dried collodion membranes were prepared from an electrochemically inactive collodion preparation (state I); the characteristic concentration potentials across them were low, about 30 mv. The membranes were activated by oxidation (state II) to give maximum or nearly maximum concentration potentials (about 50 mv.). The oxidized membranes are dried, dissolved in alcohol-ether, and a new set of dry collodion membranes prepared from this solution (state III). The concentration potentials across these membranes are low. 3. Since the properties of a homogeneous phase should not be influenced by a rearrangement of its constituent particles, the experimental results do not support a homogeneous phase (solubility) theory, but they agree with the predictions of the micellar-structural theory. The characteristic behavior of dried collodion membranes in solutions of strong inorganic electrolytes is therefore due to the micellar character of its interstices.     

STUDIES ON PERMEABILITY OF MEMBRANES : VIII. THE BEHAVIOR OF THE DRIED COLLODION MEMBRANE TOWARD BIVALENT CATIONS.

A study of the behavior of the dried collodion membrane toward the bivalent calcium ion showed that: 1. There is almost no potential difference established across a membrane separating two calcium chloride solutions of 0.1 and 0.01 N concentrations. 2. The transfer numbers of chlorine and calcium, as measured in electrical transfer experiments, are both close to 0.5. 3. A sample of membrane in equilibrium with a solution of calcium chloride has an extremely high electrical resistance, greater than is observed with solutions of the chlorides of any of the monovalent cations. 4. The total electrolyte content of a membrane in equilibrium with a solution of calcium chloride was only 20 per cent of that observed when the solution was lithium chloride and 10 per cent of that found when the solution was potassium chloride. In explaining these various results it is supposed that (1), (2) and (3) are all the result of (4), that is, of the inability of the calcium ion to penetrate any but the largest of the membrane pores. As the total quantity of electrolyte able to penetrate the membrane is very small the electrical conductivity must also be very small. Moreover, the few larger pores that are large enough to transport the hydrated calcium ion are too large to exert any appreciable effect in decreasing the mobility of the anion. Thus the membrane has no effect in modifying the potentials established across concentration chains with CaCl₂ and the transfer numbers determined experimentally are what one would expect if no membrane were present.     

STUDIES ON THE PERMEABILITY OF MEMBRANES : III. ELECTRIC TRANSFER EXPERIMENTS WITH THE DRIED COLLODION MEMBRANE.

The transfer numbers of the ions of electrolytes in the dried collodion membrane, as determined in a previous paper indirectly from the E.M.F. of concentration chains, can also be determined directly by electrical transfer experiments. It is shown that the difficulties involved in such experiments can be overcome. The transfer numbers obtained by the two methods are in satisfactory agreement. The experimental results obtained in the transfer experiments furnish an additional argument in favor of maintaining the theory that the electromotive effects observed in varying concentrations of different electrolytes with the dried collodion membrane may be referred to differences in the mobilities of the anions and cations within the membrane. As was shown by the method of the previous paper, the transfer number depends largely on concentration. There are some minor discrepancies between the values of the transfer numbers obtained by the two methods which, as yet, cannot be completely explained.     

STUDIES ON PERMEABILITY OF MEMBRANES : I. INTRODUCTION AND THE DIFFUSION OF IONS ACROSS THE DRIED COLLODION MEMBRANE.

The theoretical aspects of the problem of sieve-like membranes are developed. The method of preparing the dried collodion membrane is described, and the method of defining the property of a particular membrane is given. It consists of the measurement of the Co P, that is the P.D. between an 0.1 and an 0.01 M KCl solution separated by the membrane. Co P is in the best dried membranes 50 to 53 millvolts, the theoretically possible maximum value being 55 millivolts. Diffusion experiments have been carried out with several arrangements, one of which is, for example, the diffusion of 0.1 M KNO₃ against 0.1 M NaCl across the membrane. The amount of K⁺ diffusing after a certain period was in membranes with a sufficiently high Co P (about 50 millivolts or more) on the average ten times as much as the amount of diffused Cl^-. In membranes with a lower Co P the ratio was much smaller, down almost to the proportion of 1:1 which holds for the mobility of these two ions in a free aqueous solution. When higher concentrations were used, e.g. 0.5 M solution, the difference of the rate of diffusion for K⁺ and Cl^- was much smaller even in the best membranes, corresponding to the fact that the P.D. of two KCl solutions whose concentrations are 10:1 is much smaller in higher ranges of concentration than in lower ones. These observations are confirmed by experiments arranged in other ways. It has been shown that, in general, the diffusion of an anion is much slower than the one of a cation across the dried collodion membrane. The ratio of the two diffusion coefficients would be expected to be calculable in connection with the potential difference of such a membrane when interposed between these solutions. The next problem is to show in how far this can be confirmed quantitatively.     

STUDIES ON PERMEABILITY OF MEMBRANES : V. THE DIFFUSION OF NON-ELECTROLYTES THROUGH THE DRIED COLLODION MEMBRANE.

A study has been made of the relative rates at which various organic non-electrolytes diffuse through the dried collodion membrane. It was found that acetone and urea pass through the membrane many times more rapidly than glycerine and that glycerine in its turn diffuses much faster than glucose. It was also demonstrated that the rate of diffusion varies directly with the difference in concentration between the solutions on the two sides of the membrane. It was shown that the presence of glycerine on the two sides of the membrane did not appreciably affect the rate of diffusion of acetone. In a study of the changes going on during the establishment of the stationary diffusion gradient with glucose experiments were described which strongly suggested that many of the membrane channels may gradually become clogged up with glucose molecules so that the diffusion rate decreases from day to day until the stationary gradient is finally reached. In explaining the various experimental data the conception of the collodion membrane as a sieve with pores approximating in smallness the size of individual molecules was utilized. The large differences in the diffusion rates between different substances were then referred to differences in molecular size, the relatively large molecules of glycerine and glucose being unable to pass through many of the smaller pores available for urea and acetone. From the data available it was possible to estimate that 98 per cent of the pore area distributed among holes large enough for the diffusion of acetone was unavailable for the passage of glycerine and that only 0.3 per cent of the pore area available for acetone could be utilized by glucose. In trying to correlate the ratio between the diffusion rates of two different substances with the characteristic concentration potential (Co P) given by the same membrane it was found (1) that with the acetone-glycerine ratio there is no correspondence (2) that with the acetone-glucose ratio a suggestive relation exists and (3) that with the glycerine-glucose ratio a definite correspondence can be shown, the higher ratios being obtained only with membranes giving high CO P values. A rational explanation for these facts was proposed.     

STUDIES ON THE PERMEABILITY OF MEMBRANES : IV. VARIATIONS OF TRANSFER NUMBERS WITH THE DRIED COLLODION MEMBRANE PRODUCED BY THE ELECTRIC CURRENT.

The transfer number of Cl in a KCl solution within the pores of a dried collodion membrane is always lower than 0.5. It depends on the concentration of the solution and decreases in general with decreasing concentration. However, the transfer number for any given KCl concentration has the significance of a definite and constant figure only when an infinitely small amount of coulombs is allowed to pass through the system. For finite durations of electric transfer experiments a polarization effect will always change the original transfer number. This polarization consists in an accumulation of the salt at the one boundary and a diminution at the other boundary of the membrane. Again, as the transfer number strongly depends on concentration, this change in concentration will bring about in its turn a gradual change in the transfer number too. It is shown under what conditions the transfer numbers for the anion as obtained by electic transfer experiments are higher or lower than the ones expected without polarization effect. Thus, by changing the character and magnitude of the force driving the ions across the membranes, and according to the history of previous treatment of the membrane, the whole character of what we may call the specific permeablity for ions of the membrane may be varied without any substantial change of the membrane itself concerning its structure, its chemical composition, or its pore size. Contemplation of the results obtained in this series of experiments in the light of the theoretical considerations just outlined has impressed us with the fallacy of speaking of the definite permeability of any type of membrane for electrolytes. The behavior of the membrane toward the passage of electrolytes depends on a variety of conditions. It may be recalled that different investigators have reported widely varying results concerning the permeability of certain physiological membranes for electrolytes. Such experiments as have been described in this paper may lead to an understanding of some of the factors responsible for such variations. We are aware that the collodion membrane in its simplicity is scarcely comparable to the extremely complicated biological membranes. Nevertheless any attempts to understand better the behavior of biological membranes may wisely begin with a study of the simplest prototypes.     

STUDIES ON PERMEABILITY OF MEMBRANES : VI. MENSURATION OF THE DRIED COLLODION MEMBRANE (CALCULATION OF DIMENSIONS AND OF RELATIONS TO CERTAIN BIOLOGICAL MEMBRANES).

The flat type of dried collodion membrane used by Michaelis and his associates in numerous investigations has been subjected to mensuration in order that the dimensions of these membranes may be placed on record. The membranes had a functioning area of about 30 cm., were approximately 0.1 mm. in thickness and were composed on the average of 87 per cent by volume of collodion and 13 per cent by volume of pores. In reviewing some of the previously reported results of diffusion experiments with non-electrolytes in the light of the calculated values for the total pore area for the same membranes additional evidence was presented to show that a smaller molecule (acetone) probably utilizes a much larger percentage of the total pore area for its diffusion than is available for a larger molecule (glycerol). By using the figures of Fricke and McClendon for the thickness of the membrane of the red blood cell some comparisons were drawn between the dried collodion membrane as a model for certain biological membranes and the red blood cell membrane. In these comparisons emphasis was placed on the exaggerated importance of small electromotive forces and very slight permeabilities when these were associated with membranes of such extreme thinness as the red blood cell membrane.     

THE POTENTIOMETRIC ANALYSIS OF MEMBRANE STRUCTURE AND ITS APPLICATION TO LIVING ANIMAL MEMBRANES

1. The electromotive forces which arise, if two electrolyte solutions are separated from each other by a layer of any kind, are discussed. A general equation is derived comprising the known equations for diffusion, partition, and membrane (Donnan) potentials as special cases. 2. A method is proposed to analyse membranes potentiometrically with respect to their cation or anion selectivity, their dissolving power for ions, and their influence on ion mobility (migration velocity). 3. The possibility of analysing a membrane composed of several layers of different permeability is discussed. 4. The investigation of the skin of the belly of Rana temporaria leads to the following results. It is composed of at least four layers of different permeability, one of which is specifically permeable to H ions and is very likely identical with the "basal membrane" situated between the stratum germinativum and the corium. The major part of the resting potential of the skin is located across this membrane and is due to the difference of H⁺ concentrations on both sides of the membrane. 5. Experiments on muscle show that the sarcolemma is specifically permeable to H ions. The injury potential of the muscle is attributed to the difference of H⁺ concentration inside and outside the fibre.     

STUDIES ON THE PERMEABILITY OF MEMBRANES : II. DETERMINATION OF IONIC TRANSFER NUMBERS IN MEMBRANES FROM CONCENTRATION CHAINS.

The ionic transfer number in an electrolyte solution in the pores of a narrow pored collodion membrane depends much more on the concentration than it does in a free aqueous solution. The potential difference of two solutions of the same electrolyte in different concentration depends largely on the concentration range. The ratio of the concentrations on the two sides was always 1:2 in the experiments; the concentration range was varied. It is shown that the transfer number of Cl, calculated from the P.D. measured, is very small in dilute solution (down to .02 and less in some cases), whereas it approaches the value .5 holding for free aqueous solutions when the concentration range is raised. The differences for the transfer number of Cl, according to the cation (H, K, Na, Li), can be recognized and show the same order as in free aqueous solution. But even in LiCl, where in an ordinary aqueous solution the transfer number of Cl is always > .5, this number is very low in the case of the membrane (e.g. < .05 in .01 M solution).     

STUDIES ON PERMEABILITY OF MEMBRANES : VII. CONDUCTIVITY OF ELECTROLYTES WITHIN THE MEMBRANE.

Two methods of measuring the electrical conductivity of the dried collodion membrane in contact with an electrolyte solution are described and the results of such measurements with different electrolytes in different ranges of concentration recorded. Some of the difficulties encountered in making these measurements are outlined. Of special interest was the fact that each membrane with each electrolyte showed a maximum level of resistance at a certain point in the dilution scale, a level which was not surpassed by further dilution. It is believed that this level was fixed by the collodion itself rather than by the contiguous electrolyte solution. Its existence limited the results available for reasonable interpretation. In relatively concentrated solutions the conductivity was shown to be approximately proportional to the concentration. With different electrolytes in the same concentration it was shown that the conductivities varied much more than in simple solutions without a membrane and that they fell in the order HCl > KCl > NaCl > LiCl. A method was described whereby the electrolyte content of a membrane in contact with different chloride solutions could be determined. It was shown that a membrane saturated with either 0.5 N HCl or 0.5 N KCl had practically the same total electrolyte content whereas the same membrane in contact with 0.5 N LiCl contained only half the quantity. These results were used in interpreting the conductivity data, the evidence presented strongly suggesting that two factors are operative in causing the widely divergent conductivities recorded with different electrolytes. The first factor depended on the quantity of electrolyte which can enter the membrane pores, a quantity dependent on the size of the pores and the volume of the larger of the two hydrated ions of the electrolyte. This factor was the chief one in determining the difference in conductivity between KCl and LiCl. The second factor was concerned with differences in the mobility of the various cations within the membrane brought about by friction between the moving ions and the pore walls. With KCl and HCl the quantity of electrolytes entering the membrane was in each case the same, being determined by the size of the larger Cl^- ion. The widely different conductivity values were explained as due to the changes in the mobility of the two cations within the membrane pores.     

THE VARIATION OF ELECTRICAL RESISTANCE WITH APPLIED POTENTIAL : II. THIN COLLODION FILMS

The resistance of very thin collodion membranes to direct current bears some resemblance to that of living cells since it varies with the applied potential. With membranes separating two different solutions the resistance varies with the direction (and the voltage) of the applied potential, rising when less mobile ions are carried across the membrane and falling when more mobile ones are so carried. (With some membranes the resistance varies with potential when the same solution is on both sides.) These changes are very prompt and regular. There is a hysteretic effect of previous current flow. But the membranes differ from Valonia cells in that the rise of resistance is largely ohmic, there being little or no polarization potential.