QD Anomalous Osmosis with Gold Beaters Skin Membranes, and the Relation of Osmosis to Cell Potential A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE UNIVERSITY OF MICHIGAN BY ORIN EDWARD MADISON 1918 EASTON, PA.: ESCHSNBACH PRINTING Co. 1921 Anomalous Osmosis with Gold Beaters Skin Membranes, and the Relation of Osmosis to Cell Potential A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE UNIVERSITY OF MICHIGAN BY ORIN EDWARD MADISON 1918 EASTON, PA.; ESCHBNBACH PRINTING Co. 1921 The author wishes gratefully to express his indebted- ness and gratitude to Professor Floyd E. Bart ell, under whose direction this research was carried out, in most sincere apprecia- tion of invaluable advice, kindly encouragement, helpful criticism, and many personal favors throughout the course of the work. CONTENTS I. HISTORY Discovery of the Phenomenon of Osmosis 5 First Quantitative Experiments on Osmosis 5 Later Work of Graham and Others, and their Views as to the Cause of Osmosis 6 Similarity Between Anomalous Osmose and Electric Osmose 7 Electric Osmose and Adsorption 8 II. THEORETICAL CONSIDERATIONS Introduction 9 Capillary System and Membrane System 9 Adsorption Potentials 9 Possible Electrical States Associated with the Membrane and Capillary Systems 10 Examples of Each 12 III. EXPERIMENTAL PART Apparatus and Methods 13 a. Two-Compartment Cell 15 b. One-Compartment Cell 16 Osmose of Chlorides in One-Compartment Cell 16 Osmose of Chlorides in Two-Compartment Cell 18 Osmose of Nitric Acid 20 Osmose of Sodium Hydroxide 20 Measurement of Cell Potential 21 The Electromotive Force of . 05 M Chlorides against Water 22 The Electromotive Force of Nitric Acid and of Sodium Hydroxide against Water 23 Summary and Conclusions 23 Osmose of . 05 M Chlorides in Two-Compartment Cell 23 a. Acid throughout the Cell 28 b. Acid on the Solution Side 29 c. Acid on One Side; Solution on the Other Side 32 d. Alkali throughout the Cell 33 The Electromotive Force of . 05 M Chlorides with Nitric Acid throughout the Cell 35 The Electromotive Force of . 05 M Chlorides with Sodium Hydroxide throughout the Cell 35 Summary ., 35 458723 ANOMALOUS OSMOSIS WITH GOLD BEATERS SKIN MEMBRANES, AND THE RELATION OF OSMOSIS TO CELL POTENTIAL The phenomenon of osmosis appears to have been dis- covered in 1748 by Abbe Nolle!. 1 He filled a vessel with alco- hol, closed it with bladder, and submerged the whole in pure water. The volume of the alcohol was increased and the bladder distended, thus showing that the water had passed through the membrane more rapidly than the alcohol. This discovery, however, was accorded little attention in scientific circles outside that of medicine and was apparently forgotten until 1819 when Sommering 2 made a similar discovery. He found that when a hog's bladder, filled with an alcohol- water solution, was suspended in air, the alcohol became more concentrated. When the experiment was repeated using an India rubber bag, the alcohol became more dilute. These two opposite effects with different membrane materials early estab- lished the importance of the nature of the membrane itself. The first quantitative experiments on osmosis were car- ried out by Dutrochet 3 and Vierordt 4 between the years 1826 and 1848. They found that when a salt solution was separated from water by means of pig's bladder, the water diffused through the membrane more readily than the salt solution, thus producing a hydrostatic pressure. Dutrochet observed that there was always a current inward to the more concen- trated solution side; this he called the endosmotic current. Simultaneously there was an outward current which he called the exosmotic current. In 1827 Dutrochet 5 brought forth an electrical theory to explain osmosis. He concluded that the two sides of the membrane developed different "degrees of electricity," but that the difference could not be detected with a galvanometer. The researches of Dutrochet and Vierordt established the fact that the difference between the rates of diffusion of pure water and of salt solutions de- pended not only on the concentration of the solution, but also on the nature of the salt solution and, as they later found, on the nature of the permeable septum used. Dutrochet also found that osmotic pressures were developed by porous in- organic membranes as well as by organic membranes. About twenty-five years later, extensive investigations were carried out by Thomas Graham, 6 who used a variety of membranes, both organic and inorganic, with many different types of solutions. He obtained osmotic effects covering a wide range of magnitude. Certain anomalous effects he attributed mainly to the chemical disintegration of the mem- branes; in fact, he advanced the theory that an alteration of the membrane was an indispensable condition to the mainte- nance of the " osmotic force." He considered that one side of the membrane was always acid and the other side alkaline. The direction of the endosmotic current, he believed, was al- ways from the acidic to the basic side. The effects he ob- tained with organic membranes were generally opposite to those he obtained with unglazed porcelain; however, he offered no explanation for this difference in behavior. Later, influenced by his own work on dialysis, and by the work of L/Hermite 7 on selective or preferential solubility of two liquids in a separating membrane, Graham came to the conclusion held by Liebig 8 that osmosis is due to the ability of the mem- brane to absorb the separated liquids. From this time on for nearly half a century the work on osmosis was directed mainly to the study of unidirectional currents. "Semi-permeable" mem- branes were used which were capable of producing maximum osmotic pressures. It had been pointed out by van't Hoff 9 that such pressures were expressible by the gas law formulations. As work progressed and quantitative data increased, many of the investigators in this field appear to have almost entirely neglected to take into account the fact that anomal- ous osmotic effects of considerable magnitude are obtained when solutions of electrolytes are used with osmotic membranes. Abnormal effects were in nearly all cases attributed either to electrolytic dissociation, or to molecular association, or to hydration. The attention of these investigators has for the most part been directed to a study of the more perfect semi- permeable membranes such as copper ferrocyanide with solu- tions of non-electrolytes such as sugar. The tendency of electrolytes to produce osmotic pressures at variance with the values calculated from van't Hoff's generalization, even with the best of "semi-permeable" mem- branes, is easily detected when sufficiently refined measure- ments are made. This has been clearly shown by Lord Berkeley and E. G. J. Hartley 10 Morse and his collaborators, 11 and by other investigators who have observed the lack of conformity between the experimental and the calculated values of osmotic pressures of salt solutions. No generally accepted theory has been given to account for this osmotic behavior. The anomalous effects of salt solutions with natural cells and tissues in the presence of an acid or alkali medium has been a perplexing problem and has been studied by Girard, 12 Lillie, 13 Osterhout, 14 Loeb 15 and others. Girard studied the osmotic pressures of electrolytes with various animal membranes. He found that the os- motic pressures of electrolytes vary greatly with their nature, as well as with their concentration; in fact, he noted that in some cases the exosmotic current was greater than the endos- motic current, i. e., negative osmosis was obtained. In seeking an explanation, Girard announced his electrostatic theory. He considered osmosis of electrolytes to be due primarily to an electrical effect, and the process of osmosis to be dependent mainly upon two electrical factors: (i) the sign of the charged, movable, liquid layer adjacent to the walls of the capillaries in the membrane, and (2) the difference of potential existing between the two faces of the membrane. He regarded the membrane as being electrically charged. He considered the charge on the walls of the capillaries to be due to the effect of a small excess of hydrogen or hydroxide ions. The movable layer of liquid within the capillary was assumed to possess a charge opposite to that of the capillary wall. Girard found that the difference of potential between two solu- tions with a membrane interposed, may be greater or less than the potential between the two solutions when in direct con- tact, and further, that the orientation of the cell system may even be reversed by the interposition of the membrane. A reversal of this kind means that the sign of the interface potential has been changed. It appears to be the rule that permeable membranes of almost any material whatever, interposed between a solution and water, or between two solutions, give differences of poten- tial between the two faces of the membrane which are differ- ent from the contact potential of the two liquids. Examples of such potential differences exhibited by membranes are to be found in the work of Brtinings, 16 Lillie, 17 Loeb, 18 Beutner, 19 Bartell and Hocker, 20 and others. The precise nature of the membrane seems to be an im- portant factor in determining the nature of the osmotic effect and the electrical condition of a cell system. The propor- tionality which has been shown to exist in osmotic cell systems between the osmosis measured in terms of hydrostatic pressure and the E. M. F. of the same cell system seems to conform fairly closely to Wiedemann's third law 21 for electrical osmose, which states that for a given diaphragm material, the differ- ence in hydrostatic pressure maintained between the two sides of the porous diaphragm is proportional to the applied potential. Further analogies may be shown to exist between the phenomena of anomalous osmose and that of electrical osmose; 22 for example, in both cases a reversal of flow of liquid can be brought about by the introduction of acid, base, or a salt of polyvalent ions, into the cell system. Both phe- nomena are dependent upon the existence of an electrical double layer along the walls of the capillary pores. In the process of electrical osmose, a difference of poten- tial is enforced upon the cell system and is caused to be opera- tive within the two solutions which bathe the two faces of the membrane; whereas in the process of anomalous osmose the difference of potential is self-induced, and it too may be assumed to function between the two faces of the membrane. The effects, resulting in a flow of solution through the mem- brane, are the same in either case. Freundlich, 23 influenced by his own work on adsorption and by the theories of Perrin 24 regarding the analogies be- tween the behavior of suspensions and the peculiarities of electrical osmose, was probably the first to point out clearly the intimate relations existing between adsorption and elec- trical osmose. Bancroft 25 has further contributed to our understanding of the relation between the sign of the charge on a membrane and the selective adsorption of anion or cation. It is only a short step forward to apply to osmotic phe- nomena, which, as above stated, have been shown to be very similar fundamentally to electrical osmose, a definite theory based upon selective or preferential adsorption of ions. Theoretical In attempting to explain the osmose of electrolytes by an electrical theory similar to that used to account for elec- tric osmose, two determining factors must always be con- sidered: (i) the electric charge of the capillary pore wall in respect to the charge on the liquid layer bathing this wall (i. e., the Helmholtz electrical double layer), represented in Fig. I, which we shall refer to as the capil- s y stem >' an( ^ ( 2 )> ^ e orientation of the electrical potential existing, between the two faces of the membrane, represented in Fig. II, which we shall refer to as the membrane system. The magnitude of these two electrical factors is dependent upon the extent of diffusion of electrolyte through the mem- brane, upon the relative migration velocities of the ions and upon the extent of selective ion adsorption. These three fac- tors are operative simultaneously and, since each factor affects to some degree the effect of the others, the result ob- tained is necessarily a differential one, it being the combined effect of all three factors ; any one factor may play a predominat- ing part in any particular case. The value of the electrical charges may be materially altered by even traces of acids or alkalis. It has been pointed out by Bancroft 26 that adsorption is a specific process, the neutralization of the charge on a given colloid depending on the nature of the colloid, and upon the nature of both cation and anion. This harmonizes with the view of Freundlich, 27 Michaelis 28 and others that adsorption IO potentials are dependent upon the nature of the adsorbing material and upon the extent of selective ion adsorption. It seems probable, then, that the sign of the charge upon the capillary pore wall of an osmotic membrane is dependent mainly upon the relative adsorption of the cation and anion from the solution present in the capillary pore. An indica- tion of the magnitude of the charge on the capillary wall may be obtained by reducing some of the membrane ma- terial to a fine suspension which can be placed in the solution in question and then subjected to the influence of a differ- ence of potential (i. e., the process of cataphoresis) . The direction and velocity of migration of the particle indicates the sign and magnitude of the charge upon it. An application of the above concept brings out the fact that a complete cell system must exist in some one of nine different conditions of electrification. Each of the following diagrams (Fig. Ill) represents a single capillary pore extending i n ui o JL A Fig. Ill through a membrane; in connection with this pore there is indicated also the sign of the electrostatic charge on the II membrane, the corresponding opposite charge of the liquid layer bathing the pore wall, and the electrical orientation of the membrane system. In each case the arrow on the left, pointing upward, represents the direction of the tendency to produce normal osmose, such for example, as would be ob- tained with sugar solution. The arrow on the right indicates the direction of the superimposed effect. The direction of this superimposed effect may be the same as, or opposite to, the normal osmotic effect. In the latter case negative osmose may result. The solution is understood to be on the upper side of the membrane, and water (or the more dilute solu- tion), on the lower side. The osmose due to this superimposed effect, is assumed to be caused by the passage of a charged liquid layer along the walls of the capillary pores of the mem- brane under a driving force of potential which acts as though it were set up between the two faces of the membrane. 1 If we consider all the cases in which it is possible for the cell systems to exist, we find, referring to diagrams in Fig. 1 The term normal osmose has been used throughout to designate that process which tends to produce an equilibrium difference of pressure, of magni- tude expressible by the gas law formulations, when solution and solvent are separated by a membrane permeable to the solvent alone. In the present paper absolutely no attempt has been made to "explain" the phenomena of normal osmose. It has been assumed that a tendency to produce normal osmose does exist within a system whenever two aqueous solutions of unequal concentra- tion or a solution and water are separated by a truly "semi-permeable" membrane. It is also assumed that, in case the membrane is not strictly semi- permeable but is, instead, permeable to solute as well as solvent, the tendency to produce positive normal osmose still exists and will continue to exist until the solutions on the two sides of the membrane become of the same concentra- tion. Further, it has been tacitly assumed that the rate of flow of liquid through the membrane in normal osmose should be very nearly the same with different kinds of solutions which are isotonically equal. In those cases in which the rate of flow of liquid through the membrane is different than the rate obtained as the result of normal osmose alone, it is assumed that some superimposed effect is operative within the system. The superimposed effect may act in conjunc- tion with the force tending to produce normal positive osmose, resulting thereby in abnormally great positive osmose, or the superimposed effect may act in op- position to the normal osmotic tendency and may in some cases even become so great as to completely overcome the normal osmotic effects and produce as a result a flow of liquid from concentrated to dilute solution. This we have desig- nated as negative osmose. 12 III, that in cases I, II, III, IV and VII, normal osmotic ef- fects would be obtained; in cases V and IX abnormally high positive osmose would be produced; while in cases VI and VIII an abnormally low osmose would be produced, which osmose might even become negative. Cases I, II and III represent a membrane which is iso- electric with the solution. This condition, even though a difference in potential might exist between the faces of the membrane, would produce normal osmose. Case I would be obtained with a membrane electrically neutral, with a sugar solution. Cases II and III may be con- sidered to exist when a membrane such as porcelain is in con- tact with a solution of an electrolyte at such concentration that the membrane material is at the iso-electric point. Case IV represents a membrane such as porcelain (elec- tro-negative) in a sugar solution. The membrane is negative to the sugar solution; however, owing to the fact that no polarization exists between the two faces of the membrane, only normal osmose would result. Case V represents a membrane such as porcelain with a solution such as KNO 3 ; the membrane is electro-negative to the solution and the electrical orientation of the cell sys- tem is such that the solution side is electro-negative to the water side. This condition would result in an abnormally great positive osmose. Case VI exists when a porcelain membrane is in contact with a dilute solution of a base within the cell. The mem- brane is negative to the solution, but owing to selective ad- sorption of ions and also to the more rapidly moving anion, the dilute solution side is electro-negative to the other side. An abnormally small, or even negative osmose would result. Case VII represents a membrane such as aluminium oxide (electro-positive), with a sugar solution. The aluminium oxide is positive to the sugar solution, but since no polariza- tion exists between the two faces of the membrane, only normal osmose would result. Case VIII exists with a concentrated solution of HC1. 13 The capillary wall is positive to the solution as a whole, while the water or dilute solution side is electro-positive to the concentrated solution side. This condition would give an abnormally low or negative osmose. Case IX is obtained with an A1C1 3 solution. The capil- lary wall is positive in respect to the solution, while the dilute solution side of the system is electro-negative, and would therefore result in an abnormally great positive osmose. The anomalous osmosis due to the effect of electrolytes in general, used singly or in combination, and its relation to the equilibrium of emulsions, sols, jellies, blood coagulation, living plant and animal cells, etc., may be accounted for on the basis above outlined. This explanation is further con- firmed by the various data obtained in connection with the action of electrolytes in many different physiological and bio- logical systems. The same fundamental principles underlie all these inter-related phenomena. Apparatus and Methods The object of this investigation has been to study the osmotic effects produced by solutions of electrolytes with an animal membrane such as gold beaters skin, and to ascertain whether any parallelism exists between the observed osmotic effects produced, and the difference of potential associated with the same cell system. The gold beaters skin used was of very fine grade and was of uniform texture. That we are justified in considering this membrane capillary in nature is evident from the work of Bigelow, 29 who found that Poiseuille's law for the passage of liquids through capillary tubes applies to the passage of water through collodion, parchment paper, and gold beater's skin. It may be well to point out the fact that gold beaters skin membranes are far from being semi-permeable. From the beginning of an experiment to the end, there is a continual diffusion of solute from the more concentrated to the more dilute solution. This diffusion of solute, which results in a change in concentration, will continue until the solutions 14 on the two sides of the membrane are of the same concen- tration. With a membrane of this type, we are unable to even approach the theoretical maximum osmotic pressure values. 1 What we actually have obtained in this work, is data showing the rate of flow of solution through the mem- brane. In some cases we have measured also the equilibrium pressure, expressed in terms of hydrostatic pressure, of the different solutions when the rate of flow of liquid through the membrane in one direction was just balanced by the rate of flow of liquid in the other direction. If the rate of flow of liquid was practically the same as that of a sugar solution of the same concentration, we have considered the rate of flow normal and have designated the process as nor- mal osmose. If the rate of flow of liquid is far different from that of sugar solution, we have characterized the osmose as abnormal and the process as one of anomalous osmose. If the rate of flow of liquid was greater in the direction of the more concentrated solution, we have designated that as a positive osmotic flow or positive osmose, while if the rate of flow was greater in the direction of the more dilute solution, we have designated that as a negative osmotic flow, or nega- tive osmose. It will readily be appreciated that a comparison of the rates of flow of different solutions is in no way an exact means of comparing the absolute osmotic activity of the differ- ent solutions. It does, however, give us a fairly accurate in- dication of the order of the maximum equilibrium pressures which may be obtained with these solutions. Furthermore, in those instances in which the direction of flow of solution is opposite to that obtained in normal osmose, there seems to be no logical argument against the view that some force must be operating in the system in addition to that tending to produce positive osmose. It is for the purpose of throw- ing some light on the nature and source of this additional force, or superimposed effect, that the work of this paper is directed. 1 It may be mentioned that this is quite the type of osmotic membranes we encounter in practically all living organisms, both animal and vegetable. Osmotic experiments were carried out in a cell of two compartments, each half of which consisted of a glass L-tube of approximately 20 cc capacity Fig. IV. The ends of the L-tubes in contact with the membrane were ground to make water-tight joints. The ends were covered with a thin coating of low-melting paraffin, which served as a pro- tective cushion for the membrane when the cell was set up. The membrane was held in place by a piece of tightly fitting rubber tubing, which, in turn, was held firmly to the paraf- fined glass cell by means of tightly wound copper wires, the extensions of which served as legs to support the cell in an upright position. All the stoppers in the cell were coated with paraffin each time a cell was set up. As a re- sult no difficulty was experi- enced from leakage. When concentrations of alkali greater than o.oi M were used, it was necessary to pro- tect the face of the rubber stoppers with paraffin. The outlet tubes, used to measure the osmotic effects, were of about 3 mm internal diameter. Fig. IV At the beginning of each experiment the cell was filled and the liquids were brought to the same height in both tubes. The temperature was kept at approximately 20 C. Read- ings were taken every two hours for twelve hours, at which time the cells had reached very nearly their maximum or minimum osmotic values. The main advantages of these cells are as follows: (i) Any leak in the cell is easily detected, (2) evaporation is practically eliminated, (3) tem- perature changes cause practically no alteration of the hydro- i6 static pressure on the membrane, since a change in tempera- ture causes approximately the same rise in each of the outlet tubes, (4) no difficulty is experienced in working with solu- tions which must be protected from atmospheric contamina- tions, since both solutions are well enclosed, (5) any change in the concentration of the two solutions due to the dissolving of the membrane material is practically the same, (6) correc- tions due to the capillarity are negligible, since the effects are practically the same in the two outlet tubes, (7) the whole cell can easily be immersed in a constant temperature bath when quantitative measurements are desired. Osmose of Chloride Solutions of Different Concentrations in a One-Compartment Cell In the first series of experiments carried out, but one compartment of the above described cell was used, Fig. V. Fig. V 17 A membrane was fastened over one end of the glass L-tube and the whole was suspended in a vessel of water, about 1000 cc. With this set-up we determined the osmotic effect of the cell with a very large volume of water present. It was de- sired to compare osmotic effects obtained with large volumes of water present with those obtained when smaller volumes (about 20 cc) were present. The following tables contain the results of the experi- ments on the osmose of chloride solutions of different concen- trations, and of sugar solutions of the same concentrations as these, against a large volume of water. The data are given as rise in millimeters (from the original level of the meniscus). The height of the liquid in the outlet tubes was measured by means of a millimeter scale and estimated to 0.5 of a milli- meter. TABLE i Concentration o.oi M Solutions of chlorides in cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 A1C1 3 ThCl 4 Sugar o O 2 2 i-5 3 6 6 21 79 2 4 4 3 5 ii "5 49 153 4 6 5 5 6-5 16 17 81 230 6 8 6 6-5 7 19 21 H3 297 8 10 7 8 8 21 25 150 365 ii 12 7 8 9 22.5 27 . i87 403 12 TABLE 2 Concentration 0.02 M Solutions of chlorides in cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 A1C1 3 ThCl 4 Sugar 2 2-5 3 2-5 3 2 61 9i 7 4 4 5 3-5 8 8 133 202 ii 6 5 7 5 10.5 ii 213 315 15 8 6 8-5 6-5 16 17 270 403 18 10 6-5 9 8-5 25 27 345 509 21 12 7-5 9 10 30 35 420 564 23 i8 TABLE 3 Concentration 0.05 M Solutions of chlorides in cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 A1C1 3 ThCl 4 Sugar O O o o 2 2 2 2.5 8 9 93 131 12 4 4 4 4 15 18 198 325 20 6 5 5-5 5-5 22 23-5 285 456 30 8 6 7 7 28 29 367 5H 38 10 7 7-5 9 32 33 468 579 4 6 12 8-5 10 ii 36 38 550 660 51 Concentration o. i M Solutions of chlorides in cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 A1C1 3 ThCl 4 Sugar 2 3 3-5 3-5 20 23 137 201 15 4 5-5 6 6-5 35 32 281 461 35 6 8 8-5 9 45 45 390 657 54 8 10 10 10 50 55 494 777 69 10 ii n-5 12 55 65 615 891 84 12 12 12.5 H 58 75 738 977 99 Osmose of 0.05 M Chlorides in Two-Compartment Cells This set of experiments was made with the volumes of solu- tion and solvent on opposite sides of the membrane, as nearly equal as possible. This was done, in contrast to the condi- tions in the previous experiments in which the volume of solu- tion and solvent were made very unequal, in order to study the influence of the relative volumes of solution and water on the osmotic effect. For these experiments, likewise for those in which the effect of acid and alkali was studied, as also for those in which measurements were made of the E. M. F., the two-compartment type of osmotic cell previously described was employed. Using this type of cell, experi- ments were made to determine the osmose of o . 05 M chlorides. The results thus obtained are given in the following table: TABLE 5 Solutions of chlorides in two- Concentration 0.05 M compartment cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 Aids ThCl 4 o o o O 2 9 3-5 5-5 27 42.5 112.5 9i 4 20 8-5 10.5 42.5 66.5 129 112.5 6 25 15 19 55 84 152-5 150 8 28 21 27 62.5 97-5 H7-5 1 66 10 29 27-5 3i 61 107 142 170 12 29 31 36 61 112 137 169 The data contained in the preceding tables show that: 1. The order of osmose of the several salt solutions was the same with the double cell as with the single type of cell. 2. The osmose of the salt solutions of univalent and di- valent cations was decidedly greater in the two-compartment cell than in the single cell. 3. On the other hand, the salt solutions containing polyvalent cations, as aluminium and thorium, gave decidedly smaller effects in the two-compartment cell than in the single cell. The Osmose of Acid and Alkali Against Water In these experiments nitric acid and sodium hydroxide were employed. Carbon dioxide-free water was used to make the solutions of alkali. The outlet tubes were closed with very small soda lime tubes to prevent the adsorption of carbon dioxide from the atmosphere. A positive effect signifies a flow of liquid toward the side of the membrane in contact with the electrolyte, and a nega- tive effect, indicated as ( ), signifies a passage of liquid in the opposite direction. The osmose is expressed in terms of rise in mms of solution, which is half the actual hydrostatic pressure, or half the difference in level of the liquids in the two outlet tubes. Concentrations of both acid and alkali varying from 2O o.oooi M to 0.5 M were employed, and the results of the tests are shown in Tables 6 and 7. TABLE 6 The Osmose of Nitric Acid Time (hrs.) o.oooi M o.ooi M o.oi M o.i M 0.2 M 0.5 M o o 2 o o 3-5 -i-5 7-5 -14-5 4 0.3 0-5 6-5 2 12 22 6 0-5 i 9-5 2.5 -H-5 26 8 0.7 i-5 "5 2.7 15-5 2/.5 10 i 2 13 2-5 16 28 12 i 2-5 H-5 2 -16.5 -28.5 From Table 6 it may be pointed out that: 1. Nitric acid gives both positive and negative osmose, depending on the concentration of the acid employed. 2. The osmose is positive at concentrations of o.oi M or less, and negative at higher concentrations. 3. The osmose increases from o.oooi M to o.oi M as the concentration increases, but at o . i M concentration the osmose becomes slightly negative, and continues to become increasingly negative as the concentration of the acid is in- creased. TABLE 7 The Osmose of Sodium Hydroxide Time (hrs.) o.oooi M o.ooi M o.oi M o.i M 0.2 M 0.5 M O 2 i-5 I 4 5 4 0-5 2 2 o 3-5 7 6 0-5 2 3 2-5 7-5 8 i 2 4 2 6 10 i-5 3 5 0.2 2 5-5 12 i-5 3 5 I 2 5-5 Table 7 shows that: i. Sodium hydroxide also gives both positive and nega- tive osmose. 21 2. The osmose is distinctly positive at concentrations of o.o i M or less, but seems to be practically zero at o. i M con- centration, and becomes increasingly negative as the concen- tration^ increases . It is a peculiar and interesting fact that the turning point for both the acid and alkali is at about the same order of con- centration, namely o.oi M to 0.02 M. (See Fig. VI.) Fig. vi _In this same connection it may be noted that Bartell and Hocker, 30 in their work with porous porcelain mem- branes, make mention of a similar turning point in the case of hydrochloric acid and sodium hydroxide, but they did not ob- serve negative osmose with the concentrations of acid used. Measurement of Cell Potential The relation of osmose to cell potential was studied by measuring the potential of the cell system when the cells were set up precisely as when measurements of osmose were to be made. These potential measurements were made by the compensation method, using calomel electrodes, a Wolff potentiometer, and a sensitive galvanometer. Two modified Hulett batteries connected in series served as a source of poten- tial for the external balancing current. These batteries 22 maintained a very good constancy. One electrode was brought in direct contact with the solution and the other electrode in similar contact with the water, giving the chain: Hg- Hg 2 Cl 2 -o.i M KCl-solution-membrane-water-o.i M KC1- Hg 2 Clo-Hg. In the case of electrolytes used in combination, the chain was: Hg-Hg 2 Cl 2 -o.i M KCl-solution A-membrane- solution B-o.i M KCl-Hg 2 Cl 2 -Hg. Only the initial constant values were utilized since, as pointed out by Bayliss in his work with parchment paper, 31 they seem to be the more re- liable for comparative data. However, a sufficient number of time measurements were taken to ascertain that the E. M. F. steadily falls after the cell has been set up for a time. This is due probably to diffusion of the electrolyte through the membrane, causing a change in concentration of the solutions bathing the faces of the membrane. The Electromotive Force of 0.05 M Chlorides vs. H 2 These and subsequent electromotive force measurements were made in an endeavor to ascertain whether there was any relation between the osmotic effect produced and the electro- motive force of the same cell system. The measurements were carried out as previously described and are contained in the following table. The results given are the averages of two or more measurements, none of which varied more than two millivolts from the average value given. All the E. M. F. measurements were made within 5 minutes after the cell was set up. TABLE; 8 E. M. F. of 0.05 M Chlorides against H 2 O Solution Potential Solution side Solution Potential Solution side KC1 NaCl LiCl BaCl 2 +0.002 +0.015 +0.046 +0.050 MgCl 2 A1C1 3 ThCl 4 +0.060 +0.067 +0.070 TABLE 9 The E. M. F. of HNO 3 and NaOH against H 2 O Concentration HN0 3 Potential Solution side Concentration NaOH Potential Solution side o.ooi M o.oi M o.i M 0.050 0.092 o. no O.OOI M o.oi M o.i M +O.OI8 +0.040 +0.059 Summary of Results and Conclusions i. The principal relationships found have been brought together in the following table: Solution Osmose Sign of Potential Solution side Sign of liquid layer Osmotic Tendency Single cell (12 hrs.) Double cell (12 hrs.) 0.05 M Sugar 51 0.000 Normal (Positive) 0.05 M KC1 29 8-5 + O.OO2 + Negative 0.05 M NaCl 31 IO + 0.015 + Negative 0.05 M LiCl 36 I I +0.046 + Negative 0.05 M BaCl 2 61 36 + 0.050 + Negative 0.05 M MgCl 2 I 12 38 + 0.060 + Negative o. 05 M A1C1 3 137 550 + 0.067 (Abnormally posi- tive) 0.05 M ThCl 4 169 660 +0.070 (Abnormally posi- tive) o.ooi MHNO 3 2-5 0.050 + (Abnormally posi- tive) o.oi M HNO 3 H-5 o . 092 (Probably near turning point) o.i MHN0 3 2.0 o. no Negative o.ooi M NaOH 3 +0.018 + Negative o.oi M NaOH 5 +0.040 + Negative o.i MNaOH i +0.059 + Negative 2. The osmose of sugar solutions indicate that the rate of osmose is very nearly proportional to the concentration of the solution. 3. It is noted that the direction and magnitude of flow of solution is, in practically every case, that which we would predict from the postulates above stated. If the solution 24 side of the membrane system is of the same electrical sign as the capillary liquid layer the resulting osmose will be ab- normally low, or negative; whereas if these parts of the sys- tem are of opposite sign the resulting osmose will be abnormally high. 4. With salts of univalent and divalent cations the super- imposed effect is found to work in opposition to normal osmose, with the result that the observed rate of osmose is less than normal. 5. With salts of Al and Th the superimposed effect works in conjunction with the normal osmose and the result is an abnormally great osmose. 6. Increase in concentration causes but slight increase in osmose of solutions of univalent cations, a marked increase in osmose of solutions of divalent cations and a decidedly greater increase in osmose of solutions of trivalent and quadri- valent cations. A logical explanation, for the facts just men- tioned, seems to be that with dilute solutions of univalent and divalent cations, the charge of the membrane against the solu- tion is at all times electro-negative which tends to produce an abnormally low osmose. In the case of the solutions of divalent cations there is a marked tendency to neutralize the negative charge of the membrane, with the result that with the more concentrated solutions the membrane ap- proaches the iso-electric point and osmose now approaches the normal rate. In the case of solutions of trivalent and quadrivalent cations, the sign of the membrane is electro- positive, even with the very dilute solutions ; this results in an abnormally great positive osmose in every case. 7. With the two-compartment cells, the concentrations of the solutions on the two sides of the membrane are much more nearly equal than in the one-compartment cell. This is due to the small initial water volume, with the result that the K. M. F. of the membrane system is, in this case, much less than in the case of the one-compartment cell. Owing to the smaller potential difference between the two faces of the membrane, the resulting osmose is in all cases more nearly 25 normal. In the case of the solutions of imivalent cations, there exists a lesser tendency toward negative osmose, whereas in the case of solutions of polyvalent cations, as Al and Th, there exists a lesser tendency for an abnormally great posi- tive osmose. 8. In the case of dilute acid the tendency is toward an abnormally great positive osmose. As the concentration of acid is increased, the sign of the capillary system is changed (reversed), and the osmotic tendency passes from abnormally great positive to normal, then to abnormally small, and finally to negative osmose. 9. In the case of sodium hydroxide, a negative tendency persists throughout. At the higher concentrations the elec- trical factors of the system are such that negative osmose results. 10. Work with porcelain membranes showed somewhat similar results for the osmotic behavior of acids and alkalis. In some investigations it has been found that with certain concentrations of acid or alkali, (approx. o.oi M cone.) positive osmose may be of a very considerable magnitude, whereas at still higher concentration of acid or alkali, nega- tive osmose may result. At higher concentrations, about 10 M cone., positive osmose again results. 32 That is, the curve in Fig. VI for acid against water comes above the concentra- tion axis again and thus forms practically a sine curve. All these facts coincide with many physiological observa- tions which up to the present time have received no satis- factory explanation. In our previous studies of the relation of osmose of solu- tions of electrolytes to the electrical states of the membrane system, we concluded that the nature and magnitude of the resulting osmose was dependent largely upon two factors: (1) the electrical orientation of the membrane system, and (2) the electrical orientation of the capillary wall system. The four conditions responsible for abnormal osmose may be represented by the following diagrams, Fig. VII. 26 B D Fig. VII. With conditions represented in A and D, an abnormally great positive osmose would be obtained; while with condi- tions represented in B and C, an abnormally low, or even negative, osmose would result. Gold beaters skin in pure water is electro-negative to the water. With dilute salt solutions of univalent cations, the solution side of the membrane system is electro-positive to the other side (case B), which should give as a result a tendency to produce an abnormally low osmose. In our ex- perimental work we have found that this prediction correctly represents the facts. With salt solutions of polyvalent cations as aluminium and thorium, the membrane becomes electro-positive to the solution. The solution side of the membrane is electro-posi- tive (case D). The resulting osmose should be abnormally positive. The experimental results were entirely in accord with this prediction. It is well known that small amounts of acids or bases play an important role in adsorption, and that comparatively small amounts of these substances tend to alter greatly the sign of the charge of any adsorbing materials placed in such solutions. It was our aim in the present investigation to study the effect of the presence of different concentrations of acids and bases upon the osmose of different salt solutions. If our fundamental hypothesis is correct, we should be able, by altering the sign of the charge of the membrane by having present acids or bases, to greatly alter the osmotic effects of the different salt solutions. For example, those salt solutions which show an abnormally great osmose in neutral solution 2 7 should be caused to show an abnormally low or even negative osmose when the electrical sign of the system is properly altered by the presence of acid or alkali. Solutions of chlorides of K, Na, Li, Ba, Mg, Al and Th (the same salts that were used in our earlier investigation), of 0.05 concentration, were used. Three series of experiments were run in which were used both HNOs and NaOH solutions of different concentrations. The acid or alkali was used (1) throughout the cell system, (2) on the solution side of the membrane with distilled water on the opposite side, and (3) on the side of the membrane opposite to that of the solution. The apparatus and methods used were the same as those described in the previous paper. The results obtained are shown in the following tables. TABLE 1 Concentration of . 05 M. Solutions of Chlorides in Two-Com- partment Cells Time (hrs.) KC1 NaCl LiCl Bad* MgCl 2 A1C1 3 ThCl 4 2 9 3.5 5.5 27 42.5 112.5 91 4 20 8.5 10.5 42.5 66.5 129 112.5 6 25 15 19 55 84 152.5 150 8 28 21 27 62.5 97.5 147.5 166 10 29 27.5 31 61 107 142 170 12 29 31 36 61 112 137 169 TABLE 2 Acid throughout the Cell System Concentration of Acid 0.0001 M. 0.05M Chloride in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 Aid. ThCU 2 9.5 12.5 16 22.5 64 72 42 4 16 21 29.5 36 112 90 49 8 20 27.5 39 44.5 142 81 40 8 22.5 33 42 49 162 72 33 10 25 34.5 44 55.5 174 56 27 12 27 37 46 60 182 46 24 28 TABLE 3 Acid throughout the Cell System Concentration of Acid 0.001 M. Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 A1C1 S ThCU 2 17 29.5 35 82 116 105 14 4 32 33 71 150 204 112 12 6 43.5 44 91 202 306 100 10.5 8 50 54 108 251 380 92 8 10 53 57 118 271 414 81 4.5 12 55 68 121 312 466 71 3 TABLE 4 Acid throughout the Cell System Concentration of Acid 0.01 M. 0.05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 A1CU ThCU 2 14 26.5 32 49 93.5 61 11 4 23.5 35 51 99 17S 50 9 6 29 38 60.5 144 270 40 7 8 34.5 40 62 173 336 35 5 10 38 46 52 208 392 29 3 12 42 48- 50 227 429 23 3 TABLE 5 Acid throughout the Cell System Concentration of Acid . 1 M. . 05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 A1C1 3 ThCU 2 3.5 7.5 9.5 25 26.5 21 6.5 4 5.5 10 12.5 44 43 36 5 6 7 16 19 62 63 31 4 8 9 19 26.5 72 78 25 1 10 10 20.5 30 SI 93 21.5 12 10 22.5 31.5 91 105 16 2 9 TABLE 6 Acid on Solution Side ; Distilled Water on Other Side Concentration of Acid 0.0001 M. 0.05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 A1C1 3 ThCl 4 o 2 7 8 10.5 23 41 71 80 4 11 17 29 38 102 1 55 170 6 13 ' 20 44 37 155 213 250 8 12 17.5 54 26 195 247 275 10 11 14 66 19 227 256 282 12 9 9.5 72 12 242 252 270 TABLE 7 Acid on Solution Side; Distilled Water on Other Side Concentration of Acid 0.001 M. 0.05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 . A1C1.3 ThCl 4 2 22 31 36 52 96 162 175 4 1C) 53 72 49 196 291 300 6 58 69 103 37 285 343 350 8 68 76 121 21 359 353 376 10 71 si 135 13 429 :M7 355 12 71 80 144 8 483 334 330 TABLE 8 Acid on Solution Side; Distilled Water on Other Side Concentration of Acid 0.01 M. 0.05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 A1C1 8 ThCU 2 20 25 29 75 75 120 1 12 4 37 46 61 61 163 240 265 8 54 59 85 57 246 :7 374 8 61 67 110 43 316 392 422 10 67 69 1 !>:> 30 380 432 461 12 70 71 132 18 436 455 500 TABLE 9 Acid on Solution Side ; Distilled Water on Other Side Concentration of Acid 0.02 M. 0.05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaClo MgCl 2 A1CU ThCl 4 2 18 25 26 61 63 77 98 4 31 38 51 111 133 156 240 6 41 44 70 156 190 232 280 8 50 54 85 189 215 289 340 10 56 55 99 235 263 333 400 12 51 62 109 252 287 274 445 TABLE 10 Acid on Solution Side; Distilled Water on Other Side Concentration of Acid . 05 M. . 05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 A1C1 3 ThCl 4 2 8 13 14 35 43 101 105 4 13.5 24 27 81 92 216 228 6 16.5 36 37 116 130 244 262 8 15.5 39 41 138 151 275 294 10 15.5 44 47.5 157 167 306 324 12 15.5 48 51 172 - 184 340 360 TABLE 11 Acid on Solution Side ; Distilled Water on Other Side Concentration of Acid . 1 M. . 05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 AlCla ThCl 4 2 4 9 9.5 25 24 41 54 4 7 15 17 44 42 78 86 6 8.5 21 23 62 62 124 142 8 10 25 28 74 76 158 186 10 10.5 28 32 86 90 193 245 12 11.5 30 35.5 95 100 221 ,282 TABLE 12 Acid on Solution Side; Distilled Water on Other Side Concentration of Acid 0.2 M. . 05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 Aldi ThCU 2 1.5 4 4 10 10 19.5 24 4 2.5 7 8 20 20 39 48 6 3 11 12.5 30 32 70 88 8 3.5 12.5 13.5 35 38 84 116 10 4 13 17 42 46.5 107 132 12 3.5 15 19 46.5 52.5 122 146 TABLE 13 Acid on One Side ; Solution on the Other Side Concentration of Acid 0.0001 M. 0.05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 AlCla ThCU 2 9.5 7.5 19.5 39 49 69 25 4 13 12 33.5 56.5 98 140 17 6 15.5 14.5 38 63.5 124 70 14 8 17 16 41 64 136 15 10 10 18 19 41 62 >142 15 7 12 18 22 41 59 141 15 4 TABLE 14 Acid on One Side; Solution on the Other Side Concentration of Acid 0.001 M. 0.05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 Aids ThCl 4 2 24 29 31 76 22 80 67 4 42 59 53 80 30 144 87 6 54 78 68 84 29 152 137 8 61 94 87 64 27 162 125 10 64 104 102 54 26 167 110 12 66 109 114 49 25 167 90 TABLE 15 Acid on One Side; Solution on the Other Side Concentration of Acid 0.01 M. 0.05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 A1C1, ThCl 4 2 28 33.5 35 81 11.5 147 275 4 48 66 71 92 12.5 200 350 6 61 89 90 104 11 222 425 8 69 106 113 80 11 222 500 10 73 121 123 72 11 222 525 12 75 130 133 63 11 222 540 16 Acid on One Side ; Solution on the Other Side Concentration of Acid 0.1M. 0.05M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl-2 A1C1 8 ThCl 4 2 2 20 24 45 47.5 51 220 4 2 27 38.5 54 60 65 310 6 2 30 44 60 68 71 340 8 2 28 47 48 68 74 355 10 2 26 47 45 66 76 350 12 2 26 47 45 66 75 350 TABLE 17 Alkali throughout the Cell Concentration of Alkali 0.0001 M. 0.05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 A1C1 3 ThCl 4 "o 2 2 4 5 13 15 97 101 4 3 5 7 24 23 176 186 6 4 7 11 29 27 225 240 8 6 9 14 33 32 256 262 10 9 12 17 33 35 280 290 12 11 14 18 33 35 285 310 33 TABLE 18 Alkali throughout the Cell Concentration of Alkali . 001 M. . 05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl LiCl BaCl 2 MgCl 2 ALC1 3 ThCU 2 2.5 10 11 7 8 35 85 4 4.5 17 20 9.5 11 40 90 6 7 22 25 10.5 13 39 65 8 7.5 25 28 9.5 17 39 44 10 9 27 30 9.5 16 32 40 12 11 29 33 9.5 16 32 38 TABLE 19 Alkali throughout the Cell Cone, of Alkali 0.01 M. 0.05 M Solutions of Chlorides in Cells Time (hrs.) KC1 NaCl UC1 2 2 3 3.5 4 3 4 5 C) 4 6 7 8 6 8 9 10 7 12 14 12 7 15 IS TABLE 20 Alkali throughout the Cell Cone, of Alkali 0. 1 M. 0.05 M Solutions of Chlorides in Cells Time (hrs.) KCl NaCl LiCl 2 2 3 4 i 4 4.5 8 2 5 7 10 3 5 9 12 4 5 10 34 ^ d co 1 CO 000 o o o o o d + + 1 H + tO O CO ^ CO rH o o o g o odd (j ! ^ d + + 1 1 + rH O CO C^ tS* *-O o GO 88 bJO O CO O o be O O O a O ^ 'fl i> to o 2 00 00 CO tO f o CO CM o odd hM c3 PQ d + + 1 6 O rfi 3?6 (1914); Jour. Phys. et Pathologic gen., 12, 471 (1910). 13 Lillie: Am. Jour. Physiol., 38, 194 (1911). 14 Osterhout: Biol. Chem., 19, 493, 561 (1914). 16 Loeb: Jour. gen. Physiol. (1914). 18 Briinings: Pfluger's Archiv., 98, 241 (1903); 117, 409 (1907). 17 Lillie: Loc. cit. 18 Loeb: Science, 34, 884. 19 Beutner: Jour. Phys. Chem., 17, 344 (1913). 20 Bartell and Hocker: Jour. Am. Chem. Soc., 37, 1036 (1916); 38, 1029 (1916). 21 Wiedemann: Pogg. Ann., 87, 321 (1852); 99, 177 (1856). 22 Quincke: Pogg. Ann., 107 (1859); no, 38 (1860); Engelmann: Arch. Neerl., 9, 332 (1874); Perrin: Comptes rendus, 136, 1388 (1903); Baudouin: Ibid., 138, 898 (1904); Holmes: Dissertation, Johns Hopkins (1907); Bose and Guillaume: Comptes rendus, 147, 55 (1908); Larguier des Bancels: Ibid., 149, 316 (1909); Barratt and Harris: Zeit. Elektrochemie, 18, 221 (1912); Bancroft: Jour. Phys. Chem., 16, 312 (1912); Byers: Jour. Am. Chem. Soc., 36, 2284 (1914); Briggs: Jour. Phys. Chem., 21, 198 (1917); 22, 256 (1918). 23 Freundlich: Kapillarchemie, 245 (1909); Zeit. phys. Chem., 79, 407 (1912). 24 Perrin: Jour. Chim. phys., 3, 85 (1905). 26 Bancroft: Jour. Phys. Chem., 19, 349, 363 (1915). 26 Bancroft: Trans. Am. Electrochem. Soc., 21, 233 (1912) 27 Freundlich: Loc. cit. 28 Michaelis: "Dynamics of Surfaces." 29 Bigelow: Jour. Am. Chem. Soc., 29, 1675 (1907). 30 Bartell and Hocker: Loc. cit. 31 Bayliss: Proc. Roy. Soc., 848, 246 (1911). 32 From unpublished data obtained n this laboratory. p m. . IB m 4587X3 UNIVERSITY OF CAUFORNIA LIBRARY , ,v-