ION SERIES AND THE PHYSICAL PROPERTIES OF PROTEINS. I By Jacques Loeb Published in: The Journal of General Physiology DOI: 10.1085/jgp.3.1.85 ABSTRACT 1. This paper contains experiments on the influence of acids and alkalies on the osmotic pressure of solutions of crystalline egg albumin and of gelatin, and on the viscosity of solutions of gelatin. 2. It was found in all cases that there is no difference in the effects of HCl, HBr, HNO3, acetic, mono-, di-, and trichloracetic, succinic, tartaric, citric, and phosphoric acids upon these physical properties when the solutions of the protein with these different acids have the same pH and the same concentration of originally isoelectric protein. 3. It was possible to show that in all the protein-acid salts named the anion in combination with the protein is monovalent. 4. The strong dibasic acid H2SO4 forms protein-acid salts with a divalent anion SO4 and the solutions of protein sulfate have an osmotic pressure and a viscosity of only half or less than that of a protein chloride solution of the same pH and the same concentration of originally isoelectric protein. Oxalic acid behaves essentially like a weak dibasic acid though it seems that a small part of the acid combines with the protein in the form of divalent anions. 5. It was found that the osmotic pressure and viscosity of solutions of Li, Na, K, and NH4 salts of a protein are the same at the same pH and the same concentration of originally isoelectric protein. 6. Ca(OH)2 and Ba(OH)2 form salts with proteins in which the cation is divalent and the osmotic pressure and viscosity of solutions of these two metal proteinates are only one-half or less than half of that of Na proteinate of the same pH and the same concentration of originally isoelectric gelatin. 7. These results exclude the possibility of expressing the effect of different acids and alkalies on the osmotic pressure of solutions of gelatin and egg albumin and on the viscosity of solutions of gelatin in the form of ion series. The different results of former workers were probably chiefly due to the fact that the effects of acids and alkalies on these proteins were compared for the same quantity of acid and alkali instead of for the same pH. ION SERIES AND THE PHYSICAL PROPERTIES OF PROTEINS. I. BYJACQUESLOEB. (From the Laboratories of The Rockefeller Institute for Medical Researek.) (Received for publication, July 17, 1920.) I. INTRODUCTION. It had been shown in a preceding paper ~that the relative influence of ions on the rate of diffusion of water through collodion membranes can be expressed by arranging the ions in two definite series, as follows: (1) Rb < K < Na < Li < divalent < trivalent cations. (2) C1 < Br < I < divalent < trivalent anions. These results are intelligible on the assumption that we are dealing in this case with electrostatic effects of ions, the electrostatic effects of the cations being due to the excess charges of the positive nucleus and the electrostatic effects of the anions being due to the charges of the captured valency electrons. In this case the relative efficiency of monatomic and monovalent cations should increase inversely with the radius of the ion; and the relative efficiency of the monovalent, monatolnic anions should increase directly with the radius of the ion. In the cation series the Li ion should be next to the divalent cations, while in the anion series the iodion should be next to the divalent anions. This order differs from that usually given for the action of ions on the physical properties of proteins. Hofmeister~ and afterwards Pauli3 determined the relative order of efficiency of ions for the pre- cipitation of proteins. This order is according to Pauli: i Loeb, J., J. Gen. Physiol., 1919-20,ii, 673. Hofmeister, F., Arch. exp. Path. u. Pharm., 1888,xxiv, 247. 3Pauli, W., Beitr. chem. Physiol. u. Path., 1903, iii, 225; Fortschr. naturwies. Forschung, 1912,iv, 237. 85 The Journal of General Physiology 86 ION SERIES AND PROTEINS. I For cations, Mg < NH4 < K < Na < Li. For anions, CNS < I < Br < C1 < acetate < tartrate < citrate < phosphate < sulfate < F1. The reader will notice that in the cation series the divalent cation Mg is next to NH4 and K, while we should expect it to follow Li in order of efficiency. On the other hand, if the efficiency of the monovalent, monatomic cations increases in the order of K < Na < Li as Hof- meister and Pauli observed we should expect the efficiency of the anions to increase in the order of C1 < Br < I instead of in the re- verse order. We must ~onclude that the phenomena of precipitation of proteins by ions are either not determined by the electrostatic forces of the ion or that other variables enter which are not yet known. This latter possibility exists since the concentration of electrolytes required for precipitation is very high and the writer has shown that the influence of ions causing the anomalous diffusion of water through membranes already ceases to be noticeable in comparatively low con- centrations of electrolytes.4 It therefore seems advisable to restrict our attention to such influ- ences of ions where they act in low concentrations. This is possible when we compare the effects of low concentrations of different acids and alkalies on the physical properties of proteins. The first one who undertook such a study on a larger scale was Pauli. He recognized that in order to investigate the effect of acids and alkalies on proteins it was necessary to free them first from salts. When Pauli and Handovski 5added to blood albumin dialyzed for 6 weeks various quan- tities of different acids they found that all the acids increased the vis- cosity of the blood albumin but in a different degree. Their results are expressed in curves from which we gather that the relative order of efficiency of various acids tried by them was HC1 > monochloracetic > oxalic > dichloracetic > citric > acetic > sulfuric > trichloracetic acid where HC1 raises the viscosity most, and trichloracetic least. The viscosity was measured by the time of outflow through a viscometer. Pauli assumes, as Laqueur and Sackur~had done previously in their Loeb, J., J. Gen. Physiol., 1919-20,ii, 173. 5Pauli, W., and Handovski, H., Biochem. Z., 1909,xviii, 340. 6Laqueur, E., and Sackur, 0., Beitr. chem. Physiol. u. Path., 1903,iii, 193. JACQUES LOEB 87 experiments on casein solutions, that the increase in the viscosity Of the protein caused by the addition of acid (or of alkali) is due to an increase in the ionization of proteins, as a consequence of the salt for- mation betweenprotein and acid or alkali, the protein salt beingcapable of a greater degree of ionization than the protein not treated with acid or alkali. In order to explain the relative differences in the effect of various acids on the viscosity, Pauli assumes that the protein salts formed with different acids differ in their degree of electrolytic dis- sociation. He states; e.g., that "the strong trichloracetic acid and the slightly weaker sulfuric acid furnish few protein ions.''7 He also states that the relative efficiency of different anions on the other properties of proteins, as osmotic pressure, precipitation by alcohol, is the same as that found for viscosity. In his experiments on the influence of acids and alkalies on the osmotic pressure of gelatin solutions the writer had arrived at the con- clusion that the results he obtained cannot be expressed in terms of ion series.8 These experiments showed only an influence of the val- ency of the anion or cation in combination with the gelatin, but no other influence. Solutions of Li, Na, K, and NH4 gelatinate of the same pH and the same concentration of originally isoelectric gelatin had the same osmotic pressure. The same was true for Ca and Ba gelatinate; but the osmotic pressure of Ba and Ca gelatinate was only one-half or less than that of Li, Na, K, or NH4 gelatinate of the same pH and the same concentration of originally isoelectric gelatin. The influence of acids on the physical properties of proteins was still more interesting. Solutions of gelatin chloride, bromide, nitrate, acetate, phosphate, citrate, tartrate, succinate, of the same pH and the same concentration of originally isoelectric gelatin, had approximately the same osmotic pressure, while the osmotic pressure of solutions of gelatin sulfate was only half or less than half of that of gelatin chloride, etc. The osmotic pressure of solutions of gelatin oxalate was almost but as a rule not quite as high as that of gelatin chloride,s This peculiarity found its explanation in an investigation of the combining ratios of these acids with gelatin. The writer was able to Pauli, W., Fortschr. naturwiss. Forschung, 1912,iv, 243. s Loeb, J., J. Gen. Physiol., 1918-19,i, 483, 559. 88 ION" SERIES AND PROTEINS. I show that the strong dibasic acid H~SO4 combines in equivalent pro- portions with gelatin while the weaker acids, e.g. phosphoric or oxalic, combine in molecular proportions. At a given pH, e.g. pH = 3.5, 1 gin. of originally isoelectric gelatin is in combination with three times as much 0.1 N. acid when the acid is H~PO4 than when it is HC1; and with almost twice as much 0.1 N acid when the acid is oxalic than when it is HC1; while the ratio of HC1 and H~SO, is 1. It follows from this that the strong dibasic acid H2SO, forms a salt with gelatin in which the anion is divalent, namely SO4, while the weak tribasic acid HsPO4 forms a salt with gelatin in which the anion is monovalent; namely, H~PO4 (instead of the trivalent anion PO4). Likewise, citric, succinic, and tartaric acids form gelatin salts in which the anion is always monovalent; namely, H-succinate, H-tartrate, etc. In the case of oxalic acid this is also the case though a slight amount of gelatin salt with divalent anion is probably formed and this might account for the fact that the osmotic pressure of gelatin oxalate is generally slightly less than that of gelatin chloride? It was also found that Ca(OH)2 and Ba(OH)~ combine with gelatin in equivalent and not in molecular proportion thus showing that the cation of Ca gelatinate and Ba gelatinate is divalent. 8 All these facts together show that solutions of salts of gelatin with a bivalent ion have an osmotic pressure of one-half or less of that of solutions of salts of gelatin with monovalent ions at the same pH and the same concentration of originally isoelectric gelatin. This then proves that while the valency of the ion in combination with the gelatin has a strong influence on the osmotic pressure of the gelatin solution differences in the nature of ions" of the same valency have either no effect on the osmotic pressure or if they have such an effect it is too slight to be noticeable by our method of experimen- tation. We can certainly say that it is impossible to express our observations in terms of the Hofmeister series. In the Hofmeister series phosphates and chlorides stand at almost opposite ends of the anion series while in our experiments the effects of H~PO, and HC1 are identical if we compare the effects at the same pH of the gelatin solu- tions. The reason that Pauli arrived at a different conclusion lies 9Loeb,J., J. Gen. Physiol., 1918-19,i, 483. JACQUES LOEB 89 probably in the fact that he compared the effects of the addition of equal quantities of acid to a protein while we compared the effects of different acids at equal hydrogen ion concentrations of the solutions. II. Action of Weak and Strong Monobasic Acids. We will first indicate why it is necessary to choose equal hydrogen ion concentrations as a standard of comparison instead of using equal quantities of acid. We always use as standard material isoelectric protein, to which we add enough acid or alkali to bring it to the de- sired hydrogen ion concentration. In the case of gelatin we proceed usually in this way; we add to 1 gin. of powdered gelatin brought to the isoelectric point (pH = 4.7) small quantities of acid to bring the samples to the desired hydrogen ion concentration, then melt the gelatin, and bring the volume of the solution to 100 cc. by adding H20. The pH is then determined. It is hardly necessary to state that it requires greater quantities of weak than of strong acid to bring the gelatin to the same pH on the acid side of the isoelectric point. Fig. 1 gives the quantities of 0.1 N acetic, monochloracetic, dichloracetic, and trichloracetic acids required to bring 1 gm. of isoelectric gelatin in a 1 per cent solution to the same pH. The abscissae in Fig. 1 are the pH of the gelatin solution resulting from the addition of acid and the ordinates are the numbers of cc. of 0.1 N acid which must be contained in 100 cc. of a 1 per cent solu- tion of originally isoelectric gelatin to produce the pH. The curve indicates that the quantity of acid required is the less the stronger the acid. If we now measure the osmotic pressure of the solu- tions of the four gelatin acetates (all 1 per cent in regard to the originally isoelectric gelatin) and if we plot the curves with the values for the osmotic pressure as ordinates and the pH as abscissa~, the curves for all four gelatin-acetate salts are practically identical (Fig. 2). All four curves have a minimum at the isoelectric point pH ---4.7; theyall rise identically with a diminution in pH (i.e. a rise in hydrogen ion concentration) ; all reach their maximum at a pH .of about 3.5, and all drop almost as steeply with a further diminution of pH as they rise on the ascending side of the curve. The drop will not be discussed in this paper. 90 ION SERIES AND PROTEINS, I The slight differences in the maximal height of the four curves are within the limits of the accuracy of these experiments, and the same or even greater variations may be observed when the same experiment is repeated several times with the same acid. The maximal values of osmotic pressure reached in these experiments at pH 3.5 are the same as those reached with gelatin salts of any other monovalent anion, C1, Br, NO3, phosphate, citrate, succinate, etc. 22 / ~-- 14 % k\ * i "0 IS ..-.4 z I X. ~; pHzo zz N ~..~zs ao 3z at 3.6 a8.4o 4z ,~ ~e 48~ FIG. 1. The ordinates represent the number of cc. of 0.1 N acetic, mono-,di-, and trichloracetic acids required to bring 1 gin. of isoelectric gelatin to the pH indicated by the abscissm. Enough H20 was added to bring the gelatin-acid solution to a volume of 100 cc. It follbws from this that the conclusions which are based upon a comparison of the quantifies of the acid added instead of upon the pH of the protein solutions cannot be correct. The reason why we get identical curves when we plot the osmotic pressures as ordinates over the pH as abscissm seems to be as follows. When we add a definite small quantity of acid to 1 gm. of isoelectric gelatin, melt, and bring the volume of the solution to 100 cc. by adding H20, part of the acid will combine with gelatin and part will remain JACQVES LOEB 91 free in the solution. The more acid we add, the greater the amount of isoelectric gelatin transformed into gelatin-acid salt. We assume that there is a definite equilibrium between the hydrogen ion concen- tration of the solution, the amount of gelatin-acid salt formed, and isoelectric gelatin. If this view is correct, it follows that at the same 340 320 300 280 Z60 240 (~ 2.20 ~o 200 o~ ~o 180 160 I40 ~ i20 0 100 80 60 40 20 0 pI-Iz J noc hlo c, tr .qlol,ac~ tare ~.i~'; ~1~ ~ ~,t~ i t. i 3 4. 5 6 4.7 FIG. 2. The ordinates indicate the osmotic pressure (in mm. of the height of a column of the gelatin solution) of 1 per cent solutions of originally isoelectric gelatin which have been brought to different pH by the addition of the acids used in Fig. 1. The abscissm are the pH. The curves are practically identical, the slight differences at the summit being inside the range of the variation found for the same acids. 92 ION SERIES AND PROTEINS. I pH and with the same concentration of originally isoelectric gelatin all mixtures of acid and gelatin solution must have the same concen- tration of gelatin-acid salt; and if this is the case they must all have the same osmotic pressure if the valency of the anion is the same, since the osmotic pressure is almost entirely determined by the gela- tin-acid salt, that of the isoelectric gelatin being very low. This conclusion is in harmony with the view expressed by Laqueur and Sackur, Pauli, and others. The writer differs from Pauli only in regard to the statement that the different acids, e.g. HC1, H3PO4, tartaric, acetic, and trichloracetic, have different effects on the physi- cal properties of proteins. III. Combining Ratios of Acids and Alkalies with Crystalline Egg Albumin and the Osmotic Pressure of the Albumin Solutions. Crystalline egg albumin was prepared under Dr. Northrop's super- vision according to S6rensen's method, 1 and crystallized three times. The only difference in procedure was in the dialysis. Instead of put- ring the water under negative pressure as was done by S6rensen, pres- sure was put on the egg albumin by attaching a long glass tube full of water to the dialyzing bag so that the solution was under about 150 cm. water pressure during dialysis. This was necessary to avoid too great an increase in volume. The same stock solution of albumin served for all the experiments and was diluted before the experiment to a 1 per cent solution. The concentration of ammonium sulfate left in the solution was between M/l,000 and M/2,000. The pH of the stock solution was about 5.20. By adding about 1 cc. of 0.1 N HC1 to 100 cc. of a 1 per cent solution of this albumin the solution was brought to the isoelectric point of the egg albumin, which is according to S6rensen at pH = 4.8. The 1 per cent solutions were made up with different quantities of acid (or alkali) and the pH of the albumin solution was determined electrometrically, n In Fig. 3 are plotted the curves in which the pH are the abseissa~ and the cc. of 0.1 N acid required to obtain the various 10Stirensen, S. P. L., Compt.rend.tray.Lab. Carlsberg,1915-17,xii, 1. n The colorimetric determination which gives fairly good results in the case of gelatin is unreliable in tile case of egg albumin. IACQUES LOEB 93 pH as ordinates. The curves represent these values for four acids, HC1, H~S04, H3PO4, and oxalic acid. Beginning with the lowest curve we notice that the curve is the same for 0.1 N HC1 and 0.1 N H~SO4, since both are strong acids; or, in other words, H~SO4combines "~ i7 i6 o .%: \5. 'b- I1 ~ o *.-- 10 N ~ \g 8 ~ ,,~ , . 5 7 ~ o 6 o 4 z a I ~ 2 i 0 pPI 2.0 22 2.4 2.6 2.8 3.0 8.2 VL\ \\ \\\ 3.4 3.6 38 4.0 4.2 4.4 4.8 4.5 FIG. 3. The ordinates represent the number of cc. of 0.1 N HC1,H2SO4,oxalic, and phosphoric acids required to bring 1 gm. of isoelectric crystalline egg albumin to the pH indicated on the axis of abscissae. Enough H20 was added to bring the solutions of albumin and acid to a volume of 100 cc. For the same pH the ordi- nates for HC1, H2SO4, and phosphoric acid are approximately as 1 : 1 : 3. The ratio of HC1 : oxalic acid is a little less than 1 : 2. 94 ION SERIES AND PROTEINS. I in equivalent proportions with egg albumin. The curve for HsPO4 is the highest curve and if we compare the values for HsPO4 with those for HC1 (or HsSO4) we notice that for each pH the ordinate for HsPO, is as nearly three times as high as .that for HC1 as the accuracy of our experiments permits. This means that under the conditions of our experiments phosphoric acid combines with albumin in molecular proportions and that the anion of albumin phosphate is the mono- valent anion H2P04. 20O 180 160 ~ 140 u'/ e~ lZ0 g~ 100 80 ~ 60 O 40 20 0 /~' ~ X~lbt~ ninal~lte / ' pH 1 2 3 5 FIG. 4. Osmoticpressure of different albumin-acidsalts. 6 The ordinates indi- cate the osmotic pressure (in ram. of 1 per cent albumin solution); the abscissa~ are the pH. All solutionsare 1 per cent in regard to isoelectricalbumin. The curves for albumin chlorideand albumin phosphate are identical. The values for oxalic acid are for pH above 3.2 not quite twice as high as those for HC1, indicating that for these values of pH oxalic acid combines to a greater extent in molecular and only to a smaller extent in equivalent proportions with albumin. If egg albumin behaves like gelatin we should expect that the curves of osmotic pressure for albumin phosphate and albumin chloride when plotted as a function of the pH should be identical, since the anion is in both cases monovalent; that the curve for albumin sulfate should JACQUES LOEB 95 be considerably lower since the anion SO4 in combination with albu- min is bivalent; while the curve for the osmotic pressure of albumin oxalate should be between the curves for albumin chloride and albu- min sulfate, but much nearer the chloride than the sulfate curve. Fig. t9 "~ I8 x:l iT .M t6 t5 o 14 0 o ,,.,.i -~ 1.1 0 10 tO -,,-4 d u 8 o o 7 -c3 *~ 5 Z 4 d ~ 2 t 0 pH 2.0 2.2 Z4 26 26 3.0 3.Z 3.4 3.6 3.8 4.0 4.Z 44 4.6 46 FIG. 5. Curves expressing the cc. of 0.1 Nacetic and dichloracetic acids required to bring 1 gin. of isoelectric albumin to different pH. 4 giving the curves of the osmotic pressures of the four albumin salts shows that this expectation is fulfilled. Fig. 5 gives the curves for the combining ratios of acetic acid and dichloracetic acid with isoelectric albumin, showing the same differ- 96 ION SERIES AND PROTEINS. I ence as the corresponding curves for gelatin in Fig. 1. Fig. 6 gives the curves for the influence of the two acids upon the osmotic pressure of 1 per cent solutions of originally isoelectric albumin. The two curves are identical and are also identical with those of albumin chlor- ide and albumin phosphate in Fig. 4, thus confirming our theory. In Fig. 7 are given the curves for combining ratios of NaOH, Ca(OH)~, and NH4OH with isoelectric albumin. The curve for Ca(OH).., is identical with that for the strong base NaOH, indicating that Ca(OH)~ combines with egg albumin in equivalent proportions. ZOO 180 160 O 140 ~I20 ~" 100 80 6o 4O 20 0 pi-i -- I / - - i ! , I ~ _ _ _ _ ' 2 3 4 5 6 FIG.6. Curves of osmoticpressure of solutions of albumin acetate and albumin dichloracetate. The curves for both acids are identical. Fig. 8 shows that the curve for the osmotic pressure of Ca albuminate is only one-half as high as that of Na albuminate as was to be ex- pected. The curve for NH4 albuminate is identical with that for Na albuminate, which was to be expected since the NH4 is monovalent. The results with albumin are therefore identical with those obtained in the case of the corresponding gelatin salts. The result that gelatin phosphate and albumin phosphate behave like gelatin chloride and albumin chloride may be considered as a crucial test against the col- loidal conception of ion effects on proteins and in favor of a purely chemical theory. JACQUES LOEB 97 c . ~ 17 .~ i6 :g t+ e0 13 o to "6-" 123 .~ lO o O 8 d 7 o o ~z a ,,---i d 2 d 0 pI-I 4 5 c tN ! ..... ,, c :N ;a(C ~z // f j~ / / ! 6 7 8 9 10 ii FIG. 7. Curves representing the number of cc. of 0.1 N NH4OH, NaOH, and Ca(OH)~ required to bring 1 gin. of isoelectric, crystalline egg albumin to differ- ent pH. The curves for NaOH and Ca(OH)2 are identfcal. 98 ION SERIES AND PROTEINS. I 340 320 300 280 280 240 qS) 2?.0 ~o 200 ~ 180 t60 t40 o 120 Io0 8o 6o 4O 2O 0 p~4 'NH4alt"'u'mina:e N~ ,e /. o\ // Ca ~bt mirNe j !,/- ~' ~^ 5 6 ~ 8 g 10 1~ FIo. 8. Curves of the osmotic pressure of NH4, Na, and Ca albuminate at differ- ent pH. The curves for NH4 and Na albuminate are practically identical. JACQUES LOEB 99 IV. Combining Ratios of A rids and Bases with Gelatin and the Viscosity of Gelatin Salts. Since Pauli's ion series was based primarily on the influence of acids or their anions on the viscosity of protein solutions it seemed necessary to find out whether or not viscosity measurements confirm the conclusions at which we arrived on the basis of osmotic pressure experiments on gelatin and egg albumin. A few remarks concerning our method are required. The gelatin is first rendered isoelectric in the following way. 25 gin. of powdered gelatin of pH about 7.0 are put into 1 liter of M/128 acetic acid for 30 minutes at 10C., after which time the acetic acid is renewed and left in contact with gelatin again for30 minutes at 10C. The acid is then decanted and replaced with distilled water of about 5C. The mixture is filtered in a Buchner funnel through muslin, employing slight suction. The gelatin is then washed about six times with 100 cc. of distilled water of 5 each, and is made into a 5 per cent solution which serves as a stock solution. The pH of this solution is about 4.7, or, in other words, the gelatin is isoelectric. Some of the stock solution is heated to 45 and made up to a 2 per cent solution in quantity sufficient for a day's experiments. This 2 per cent solution is kept during the day at 24C. To 50 cc. of this solution is added the desired acid or alkali in sufficient quantity and then the volume is raised to 100 cc. by the addition of enough distilled water. This 1 per cent solution is then rapidly brought to a tempera- ture of 45, kept there for 1 minute, and is then rapidly cooled to 24. The solution is stirred constantly during the heating and cooling. The viscosity is measured immediately after the solution is cooled to 24C. The measurements were all made at 24C. by using the time of outflow through a viscometer. The time of outflow of distilled water through an Ostwald viscometer at 24C. was exactly 1 minute. Each measurement of viscosity was repeated with the same gelatin solution and the beginning and the end of a series consisted in the measurement of viscosity of isoelectric gelatin. These latter meas- urements agreed in all experiments within 1 second varying only between 80 and 81 seconds, thus guaranteeing the reproducible char- acter of the experiments. I00 ION SERIES AND PROTEINS. I The results can be given briefly. Fig. 9 gives the curves for the cc. of 0.1 N HNOs, H~SO4, oxalic, and phosphoric acids required to bring 1 gin. of isoelectric gelatin to different pH. The curve shows again that for each pH the number of cc.of 0.1 1~acid required are forHNO3, H2SO4, oxalic, and phosphoric acids approximately as 1:1:2:3. 40 , 38 .... ~ 34 ,,. .~ 30 .~ 26 W C: 24 i~ o za "~ 2o f ~ 10 4 . \ \ ,% o -"~*~ -'-- p)~ 20 ~.2 24 ~ 2.8 3.o 3.2 3.4 3.6 a~ 4.o ~ 4.4 4~ FIG.9. Curves for the number of cc. of 0.1 I~IHNO3, H2SO4,oxalic, and phos- phoric acids required to bring 1 gin. of isoelectric gelatin to different pH (in 100 cc. of solution). Curves similar to those for egg albumin (Fig. 3). For the same pH the ratio of HNO.%H2SO4,oxalic, and phosphoric acids required is approxi- mately as 1: 1: 2: 3. IACQVES LOEB 101 Fig. 10 gives the curves for the viscosity of 1 per cent solutions of gelatin chloride, sulfate, oxalate, and phosphate. The absdssa~ are the pH, the ordinates the ratio of the time of outflow of the gelatin solutions divided by the time of outflow of pure water. For the sake .,,/ 6( lati~~ o:cal(~e 2.7 ;~ e0 2.5 o 2.3 / t o /% u Z.1 _- p \ ::,-, / o 1.8 *~ 1.7 o 1.6 O.~u~ 1.5 u~ 1.4 1.3 I 12 1. Lo t phi 2 3 4 5 6 FIO. 10. The curves of specificviscosity of 1 per cent solution of originally iso- electric gelatin brought to different pH by the acids mentioned in legend of Fig. 9 except that HC1 is used for HNOa. The curves of viscosity for gelatin chloride, phosphate, and oxalate are practically identical. Specificviscosity given as time of outflow of gelatin solution divided by time of outflow of water through vis- cometer at 24C. 102 I0N SERIES AND PROTEINS. I of brevity we will call this quotient the specific viscosity of the gelatin solution. The curves for the four acids all rise steeply from the iso- electric point with increasing hydrogen ion concentration until they reach a maximum at pH about 3.0. The curves then drop again. 3.0 29 2.8 2.7 2.6 2.5 2.4 2.3 2.2 2.1 2.O 1.9 1.8 1.7 1.5 1.4 i.3 1.2 i.! 1.0 g-, ~o O co O D .5 > tO o o 2 .) ~0 / v L.... FIG. 11. Curves of specific viscosity of gelatin succinate, tartrate, and citrate. The curves are practically identical with those for the viscosity of gelatin chloride and phosphate. The curves for the three acids, gelatin chloride, oxalate, and phos- phate, are practically identical while the curve for gelatin sulfate is considerably lower. L ! r I ,, 5eI,~tin 3u, cin ~te o " tat ~&t i 2 3 4 5 6 JACQUES LOV.B 103 Fig. 11 gives the curves for the viscosity of gelatin citrate, tartrate, and succinate. The three curves are practically identical and also identical with the curves for gelatin chloride and gelatin phosphate in Fig. 10. i / 2.9 2.8 2.7 ..--- 26 7 2.5 C9 z4 o ~ 2.2 o Z.1 t..) ::r, t9 o 1.8 N t.7 ,.,~ > o 1.6 1.5 ~" 1.4 1.3 12 1.1 1.0 pI-i ,~,~ ,,6e[at~ac~tale ~ ~ ', m0n~ch orac ~ ,, ~ric~or ~cel~te ! 1 2 3 4 4.7 5 8 FIG. 12. Curves of specific viscosity of gelatin acetate, mono-, di-, and trichlor- acetate. Curves identical with those for gelatin chloride and phosphate. Fig. 12 gives the curves for the viscosity of 1 per cent solutions of originally isoelectric gelatin to which acetic and mono-, di-, and tri- chloracetic acids have been added. The curves are again identical i04 ION SERIES AND PROTEINS. I o 0.1~ N~)PI -, 0.1~ KOI I 7 ~ 4 8 o 3 Z 0 ~ pn4.7 I / d / t 5.2 .'17 fiZ ft.7 72 7.7 &?. a7 92 97 10.210.7I12 I12 FIG. 13. Curves for the number of cc. of 0.1 N NaOH, KOH, Ba(OH)~, and Ca(OH)2 required to bring 1 gm. of isoelectric gelatin to different pH (in 100 cc. of solution). All four curves are identical. Z3 ~o 2.2 2.1 ~- 2.0 ~ L9 ~ 1.8 ~ 1.7 ~ :,1.6 ,~ 1.5 "'~ . 1.4 5 eo 12 ~" i.1 1.0 ,N / I 4 ~"~--'7~J J i / E * l 4 5 6 7 8 9 tO il lZ Fro. 14. Curves for specific viscosity of Na and Ca gelatinate for different pH. .]'ACQI.TES LOEB 105 with those for gelatin chloride, phosphate, etc. These curves are interesting in connection with Pauli's statement that trichloracetic acid causes the same low values of viscosity as sulfuric acid. This was the case neither in our experiments on the osmotic pressure nor in those on the viscosity of gelatin solutions. Fig. 13 gives the cc. of 0.1 N NaOH, KOH, Ca(OH),, and Ba(OH), required to bring 1 gin. of originally isoelectric gelatin in 100 cc. solu, tion to a given pH. One curve suffices for the four alkalies thus prov- ing that Ca and Ba combine with gelatin as bivalent ions. We should expect the curve for the viscosity of Ca gelatinate to be considerably lower than that of Na gelatinate. Fig. 14 shows that this is true. SUMMARY. 1. This paper contains experiments on the influence of acids and alkalies on the osmotic pressure of solutions of crystalline egg albu- min and of gelatin, and on the viscosity of solutions of gelatin. 2. It was found in all cases that there is no differencein the effects of HC1, I-IBr, HNOs, acetic, mono-, di-, and tHchioracefic, succinic, tartaric, citric, and phosphoric acids upon these physical properties when the solutions of the protein with these different acids have the same pH and the same concentration of originally isoelectric protein. 3. It was possible to show that in all the protein-acid salts named the anion in combination with the protein is monovalent. 4. The strong dibasic acid H, SO4 forms protein-acid salts with a divalent anion SO4 and the solutions of protein sulfate have an osmotic pressure and a viscosity of only half or less than that of a pro- tein chloride solution of the same pH and the same concentration of originally isoelectric protein. Oxalic acid behaves essentially like a weak dibasic acid though it seems that a small part of the acid com- bines with the protein in the form of divalent anions. 5. It was found that the osmotic pressure and viscosity of solutions of Li, Na, K, and NH4 salts of a protein are the same at the same pH and the same concentration of originally isoelectric protein. 6. Ca(OH), and Ba(OH), form salts with proteins in which the cation is divalent and the osmotic pressure and viscosity of solutions of these two metal proteinates are only one-half or less than half of 106 ION SERIES AND PROTEINS. I that of Na proteinate of the same pH and the same concentration of originally isoelectric gelatin. 7. These results exclude the possibility of expressing the effect of different acids and alkalies on the osmotic pressure of solutions of gelatin and egg albumin and on the viscosity of solutions of gelatin in the form of ion series. The different results of former workers were probably chiefly due to the fact that the effects of acids and alkalies on these proteins were compared for the same quantity of acid and alkali instead of for the same pH.