536 FINDLAY AND CREIGHTON : THE INFLUENCE OF COLLOIDS ANDLVIL-The InJuence of Colloids and Fine Suspensionson the Solubility of Gases in Water. Part I.Sotubildy of Carbon Dioxide and Nitrous Oxide.By ALEXANDER FINDLAY and HENRY JERMAIN MAUDE CREIGHTON,M.A., M.Sc. (1851 Exhibition Science Scholar of DalhousieUniversity, Halifax, Nova, Scotia).FOR many years the problem of the absorption of gases, moreespecially of oxygen and carbon dioxide, by blood has claimed theattention of physiologists. In the case of oxygen the absorptionhas been regarded as being due, in greatest measure, to theformation of a, compound with the hemoglobin of the blood;whereas, in the case of carbon dioxide, the increased absorptionas compared with a corresponding salt solution has been attributedto the reputed alkalinity of the blood, and the consequent formationof carbonate and bicarbonate.In recent years, however, different investigators (compare Hoeber,Pfliiger’s Archiu, 1903, 99, 572; Parkas, ibid., 1903, 98, 551;Friedenthal, Verworns Archiv f .ullgem. Physiologie, 1904, 4, 44 ;van Westenryk, Arch. exp. Path. Pharm. SuppZ., 1908, p. 517) haveshown by different methods that blood-serum is practically “ water-neutral.” In view of these results, it seemed that possibly theabsorption of carbon dioxide by the blood had been ascribed tooexclusively to the alkalinity of blood, and it seemed not improbablethat the colloids present in blood play an important r81e (Findlayand Harby, Zeitsch. Chem. Ind. Rolloide, 1908, 3, 169; Wo.Ostwald, ibid., 1908, 2, 264).Before this view could be tested, itwas necessary to study the influence of colloids of various kinds onthe absorption of gases, since our knowledge of this depended almostentirely on the few experiments carried out by Geffchen (Zeitsch.physikal. Chem., 1904, 49, 298).Preliminary experiments had shown that the increased absorptionof carbon dioxide which occurred under atmospheric pressure inpresence of certain colloids was due, probably, to chemical inter-action. To obtain a deeper insight into the influence of colloids ongas solubility, it was deemed of importance ts study the solubilityunder a series of presures, and also to exclude effects due to chemicalcombination by studying the absorption of a neutral gas, nitrousoxide (with regard to the neutrality of nitrous oxide, see Geffchen,Zoc. cit., p.301). This gas was chosen because its solubility in wateris nearly the same as that of carbon dioxide. Experiments on thelines indicated were carried out during the year 1908-9, theinfluence of ferric hydroxide, gelatin, arsenious sulphide, silicic acidFINE SUSPENSIONS ON THE SOLUBILITY OF GASES IN WATER. 537dextrin, starch, glycogen, egg-albumen, and serum-albumen, as wellas suspensions of charcoal and silica, on the absorption of carbondioxide and nitrous oxide having been investigated at pressuresvarying from about 750 mm. to 1400 mm. of mercury.Apparatus.The apparatus employed was, in its essential points, the sameas that used by Gefichen (Zoc. cit.), the manometer tube, however,being graduated and considerably lengthened to perinit of absorp-tions being carried out at pressures higher than atmospheric. Theburette was connected with the absorption pipette by means ofcapillary copper tubing, in order to impart the necessary flexibilityto the apparatus.The burette was contained in a glass mantlethrough which water was caused to circulate, the temperature beingmaintained constant within 0.lo throughout a determination.So long as the absorption of gas was allowed to take place underatmospheric pressure only, the dead space (that is, the ungraduatedportion at the top of the burette and the volume of the tubesconnecting it with the absorption vessel) does not require to betaken account of, as the initial and final conditions under whichthe gas is measured are the same. When, however, the absorptionis allowed to take place at higher pressures, the volume of the deadspace must be known.This was ascertained by measuring the totalcontraction of a known volume of gas and the volume in the deadspace, produced by a known increase of pressure.Since the gas in the measuring burette was always kept dry, thepoint of saturation of the solution with gas was approached fromthe side of least pressure only, but precautions were taken to makesure that the process of absorption at any given pressure wascomplete.The liquid used for the absorption of the gas was previously wellboiled to free it from air; or in those cases where boiling was notpermissible, the liquid was freed from air by being placed underdiminished pressure.I n all the following experiments the temperature of absorptionwas 25*0°, and the experimental error did not exceed +O-25 percent., and was in most cases less than this.The solubilitySolubility =C’alcutation of Resutts.was calculated by means of the formula,538 FINDLAY AND CREIGH'I'ON : THE INFLUENCE OF COLLOIDS ANDwhere01 = concentration of the gas in the liquid phase.vl = initial volume of the gas in the burette measured at thev2 = 6nal volume of the gas in the burette measured under theT = absolute temperature of experiment (thermostat temperature).Yl = absolut4e temperature of the gas in the burette a t thebeginning of the oxperiment when v1 wits measured.Tz = absolute temperature of the gas in the burette at the end ofthe experiment when o2 was measured.P = barometric pressure.p = increaseof prossuro as shown by the manometer.p' = vnpour pressure of the liquid in the absorption pipette at theVz = volume of the gas space in the pipette at the temperature T.V2 = volume of absorbing liquid in the pipette.The volume v2 was corrected, when necessary, for the dead spaceOP the apparatus.Considering the experimental errors of deter-mination, no correction was applied to the burette readings forcubical expansion of glass, nor were the barometric readingscorrected for temperature.cg = > ) Y ) ,, ,, gaseous phase.pressure P.pressure P + p .temperature 9'.I.--Solubility of Carbon Dioxide.The carbon dioxide employed for the following experiments wasthe commercial product, which analysis showed to contain 0.58 percent.of impurity. The following values were found for itssolubility in pure water at 2 5 O (table I) :TABLE l.--Solubilz'ty of Cm4on Dioxide in Water.Pressure (mm. Hg) ... 752 800 955 1059 1153 1351Solubility .............. 0.817 0'815 0.816 0.817 0'818 0'820Pressure ............... 743 841 955 1064 1243 1351Solubility ............... 0'816 0.817 0.817 0.819 0.819 0-820As the mean of these and a number of other determinations weobtained the value 0.817 for the solubility of the carbon dioxideemployed, the solubility being independent of the pressure withinthe limits of experimental error.The value found by Geffchenfor pure carbon dioxide was 0.826PINE SUSPENSIONS ON THE SOLUBILITY OF GASES IN WATER. 539( a ) Ferric Hydroxide Solution.In preparing the solution of ferric hydroxide, the method recom-mended by A. A. Noyes ( J . Amer. Chem. SOC., 1905, 37, 94) wasemployed. To a molar solution of ferric chloride, molar ammoniumcarbonate solution was added until the precipitate which formed oneach addition barely dissolved. This mixture was then thoroughlydialysed, first against tap water, and finally against distilled water,until soluble salts were removed. The concentration of the solutionwas determined by precipitation of the hydroxide with ammoniumsulphate.FIG. 1.Pressure mm, Hg.Carbon dioxide and ferric hydrox:c?c.TABLE 2.-Solubility of Carbon Dioxide in Ferric HydroxideSolutions (see Fig.1).Concentration: 0.569 gram of Fe(OH), in 100 C.C. of solution.Density = 1*000.Pressure ........... 750 848 928 1015 1146 1356Solubility ......... 0.848 0.843 0'841 0'842 0.845 0.846Concentration: 0.854 gram of Fe(OH), in 100 C.C. of solution.Density = 1.003.Pressure ............ ?50 847 923 1040 1234 1322Solubility ......... 0.862 0.858 0'856 0.857 0.860 0.861Concentration: 1.277 gram of Fe(OH), in 100 ,c.c. of solution.Density= 1'005.Pressure ............ 746 841 985 1071 1133 1266Solubility ......... 0.886 0.881 0'880 0.878 0.878 0'88540 FINDLAY AND CREIGHTON : THE INFLUENCE OF COLLOIDS ANDTABLE 2 (continued).Concentration: 1.661 grams of Fe(OH), in 100 C.C.of solution.Density = 1.009.Pressnre ............ 747 831 918 1002 1150 1267Solubility ......... 0.904 0.901 0.896 0 900 0.900 0'902Geffchen (Zoc. c i t . ) has stated that the comparatively rapid initialabsorption of carbon dioxide is succeeded by a slow further absorp-tion. I n our experiments this slow absorption was barelyappreciable except at higher pressures, and was, even then, notgreat.The above numbers, when plotted, show that the increase ofsolubility under atmospheric pressure is proportional to the con-centration of the ferric hydroxide.( b ) Dextrin.Itcontained a slight quantity of impurity insoluble in water, and thiswas separated from the solutions before they were used forabsorbing carbon dioxide. The solubility values are given intable 3.The dextrin employed was the purest supplied by Kahlbaum.TABLE S.-SolubiZity of Carbon Dioxide in Deztrin S o l ~ i i ~ n s .(See also Fig.2).Concentration: 3-50 grams of dextrin in 100 C.C. of solution.Density = 1.008.Pressure ........... 753 819 888 1060Solubility ......... 0T99 0,800 0.800 0.800Concentration : 5-60 grams in 100 C.C. of solution.Pressure ............ 754 806 856 971Solubility ......... 0.785 0.785 0.784 0 787Concentration : 9.50 grams in 100 C.C. of solution.Pressnre ............ 753 817 864 960Solubility ......... 0.761 0.756 0.758 0.759Concentration : 13.00 grams in 100 C.C. of solution.Pressure ........... 741 798 882 962Solubility ......... 0.746 0.741 0'742 0'745Concentration : 18.90 grams in 100 C.C.of solution.Pressnre ............ 748 846 934 1031Solubility ......... 0.715 0.710 0.713 0'716Concentration : 20.60 grams in 100 C.C. of solution.Pressure ............ 728 826 920 1008Solubility ......... 0.703 0'697 0'698 0.7001171 12620'802 0.803Density = 1.015.1078 12470.787 O*i91Density = 1.034.1115 12860.764 0.768Density = 1.040.1131 12560.749 0.751Density= 1.064.1180 13440.720 0.725Density = 1.069.1161 13560'704 0-71FINE SUSPENSIONS ON THE SOLUBILITY OF GASES IN WATER. 541As shown in Fig. 8, the solubility diminishes in almost exactproportionality with the increase in concentration of the dextrin.( c ) Arsenious Sulphide.The arsenious sulphide was prepared by passing hydrogen sulphideinto a solution of pure arsenious oxide until the latter was saturated.The greater part of the excess of hydrogen sulphide was thenexpelled by bubbling hydrogen through the liquid, which was thenfilt,ered before being used.As colloidal solutions of arseniousFIG. 2.O*S3000‘81000.79000*7700SJ u .*ro 2 0.75000.73000.71000-6900750 850 950 1050 1150 1250 1350 1 50Pressure in r m a . Hg.Carbon dioxide and dexhin.sulphide decompose on boiling, the last traces of dissolved air wereremoved by placing the liquid under diminished pressure, althoughthis led to the formation of a very thin film on the surface of theliquid. This behaviour is similar to the formation of films on thesurface of peptone solutions observed by Metcalf (Zeitsch.physikal.Chem., 1905, 62, l), and is no doubt to be regarded similarly asan illustration of Gibbs’s principle of increased surfaceconcen trathn.The amount of arsenious sulphide in the solutions was determine542 FINDLAY AND CREIGHTON : THE INFLUENCE OF COLLOIDS ANDby precipitating with hydrochloric acid and drying the precipitatea t SOo.TABLE 4.-Solubility of Curbon Dioxide in Solutions of ArseniousSulp7uide.Concentration: 0.392 gram of As,S, in 100 C.C. of solution.Density = 0.997.Pressure ............ 756 891 951 1047 1172 1259Solubility ......... 0.816 0.817 0.814 0 816 0.818 0'820Concentration : 1.410 grams in 100 C.C. of solution. Density= 1.003.Pressure ............ 756 851 972 1082 1137 1281Solubility .........0'810 0310 0'812 0.810 0 812 0.811Concentration : 2.289 grams in 100 C.C. of solution. Density = 1.007.Pressure ............ 754 853 938 1003 1068 1211Solubility ......... 0.806 0.806 0'806 0.806 0.806 0'806(d) Starch.For these experiments Kahlbaum's pure soluble starch wasemployed.TABLE 5.-SolubiEity of Carbon Dioxide in Solutions of Starch.(See also Fig. 3.)Density = 1.009.Pressure ............ 752 849 951 1050 1182 1334Solubility ...... .. 0.796 0.797 0999 0.801 0.804 0-806Concentration: 2-50 grams of starch in 100 C.C. of solution.Concentration : 5-00 grams in 100 C.C. of solution. Density= 1.016.Pressure ............ 753 840 912 1021 1198 1298Solubility ......... 0.778 0.780 0.781 0.784 0.789 0.790Concentration : 7.50 grams in 100 C.C.of solution. Density= 1.023.Pressure ............ 752 860 1016 1078 1201 1351Solubility ......... 0,762 0'764 0.767 0.769 0.772 0,774Concentration : 10.00 grams in 100 C.C. of solution. Density = 1.030.Pressure ............ 758 893 9E2 1087 1163 1337.Solubility ......... 0.750 0.753 0'754 0.756 0.759 0.760It was observed in the case of the above solutions that the timerequired to saturate the solution with gas was much greater than inmost of the other cases studied. The relation between starchconcentration and solubility is shown in Fig. 8FINE SUSPENSIONS ON THE SOLUBILITY OF GASES IN WATER. 543( e ) Gelatin.French sheet gelatin, which was found to be free from salts, wasSolutions containing as much as 6 per cent.of gelatin were used.quite mobile at 2 5 O .FIU. 3.0'82000-8000sl u 0-78004,.,'u 60.76000.7400750 850 950 1050 1150 1250 1350 1450Pressure in mm. R g .Carbon dioxide and starch.TABLE 6.--SolumZ)ility of Carbon Dioxide in Solutions of Gelatin.See also Fig. 4.)Density = 0.999.Pressure .......... 746 825 901 1011 1184 1369Solubility ......... 0.815 0'814 0.814 0.815 0.815 0.815Concentration: 1.06 grams of gelatin in 100 C.C. of solution.Concentration : 1.68 grams in 100 C.C. of solution. Density= 1-000.Pressure ............ 740 837 938 1072 1219 1324Solubility ......... 0.819 0'816 0.816 0.816 0.817 0.817Concentration : 3.36 grams in 100 C.C. of solution. Density= 1.003.Pressure ........... 741 826 943 1068 1230 1387Solubility .........0.826 0'819 0'818 0.818 0.819 0-820Concentration : 6.09 grams in 100 C.C. of solution. Density= 1.008.Pressure ............ 746 836 936 1015 1191 1371Solubility ........ 0.835 0.827 0'824 0.824 0.825 0.826The influence of concentration of gelatin on the solubility a tatmospheric pressure is shown in Pig. 7.Although dilute solutions of gelatin quickly become saturatedwith gas, the absorption takes place more slowly in the case of themore concentrated solutions. On reducing the pressure, the gasescaped rapidly from the solution, so as to cause considerablefrothing. The question of rate of evolution of gas is, however, 544 FINDLAY AND CREIGHTON : THE INFLUENCE OF COLLOIDS ANDspecial one, and, on account of its importance in various directions,will require to be investigated specially.Further, absorption of carbon dioxide appreciably lowered thegelatinising temperature of the solution, thus producing an effectsimilar to peptonisation.Whether the effect is st temporary or apermanent one, we have not yet investigated.( f ) GZycogelt.I n order to free it.from the small quantities of the salts which it contained, it wassubjected to dialysis, toluene being added to prevent putrefaction.Kahlbaum's pure glycogen was employed.FIG. 4.0 *a6000*8400 sr;;9u -I.2ru ;3 of32000*8000750 850 950 1050 1150 1150 1350 1450Pressure i n mm. H g .Carbon dioxide and gelatiit (-).Carbon dioxide and glycogen (- - -1.TABLE 7.--rS)~~t~biI?'it~ of Carbon Dwxide in h'olutions of GLycogen(see also Fig.4).Concentration: 0.34 gram of glycogen in 100 C.C. of solution.Density = 0.998.Pressure ............ 759 859 959 1132 1247 1369Solubility ........ 0.819 0.805 0*810 0.812 0.810 0'810Concentration : 0.68 gram in 100 C.C. of solution. Density = 1-000.Pressure ............ 759 842 954 1114 1277 1371Solubility ......... 0.817 0.805 0.807 0.807 0.807 0'807As it was impossible to remove all the toluene from the glycogensolutions, the experimental values of t.he solubility had to becorrected for the slight lowering of solubility produced by thetoluene. The numbers in the above table are such corrected values,but they are probably not quite so accurate as in the previouscases. We may assume, however, that the relative values at differentpressures and concentrations are unaffected by the correctionFINE SUSPENSIONS ON THE SOLUBILITY OF GASES 1N WATER.545(9) Egg-Albumen.This was prepared from fresh eggs by the improved method ofHofmeister *(J. Phgsiol., 1898,23, 130). Pure crystals were obtainedfrom the first crystalline precipitate as follows. The precipitatewas washed with three changes of half saturated ammoniumsulphate solution, which contained one part of glacial acetic acidper thousand. The crystals were then dissolved in the minimalquantity of water, and, while constantly stirring, a saturatedsolution of ammonium sulphate was added slowly until a distinctprecipitate was formed; then, in addition, further 2 C.C. of thesulphate solution were added for each 1000 C.C.of albumen solution.A t the end of several days crystals were obtained.As these crystals are not pure albumen, but contain ammoniumsulphate either in combination or in solution, they were dissolvedFIG. 5.750 850 950 1020 1150 1250 1350 1450Pressure in mna. Hg.Carbon dioxide and egg-albzrmen (-1.L’ai-bon diozidc and sert6nt-dburnen (- - -).in water and dialysed until free from ammonium salts. A smallquantity of toluene was added to prevent putrefaction.The amount of albumen in solution was determined by heatingthe solution until the albumen was completely coagulated, thecoagulum being then dried at 100° and weighed.TABLE 8.--SolubiEity of Curbon Dioxide in Solutions of E g gAlbumen (see also Fig. 5).Concentration: 0.50 gram of albumen in 100 C.C.of solution.Density= 0.999.Pressure ...... . .,.., 729 849 1004 1125Solubility ......... 0.806 0.795 0.802 0.810Concentration: 1.00 gram in 100 C.C. of solution.I’ressiire .. ...... ... 734 836 984 1089Solubility .. ,... . .. 0*800 0-784 0.794 0-8011236 13500.812 0.816Density = 1.002.07310 0’8121257 135546 FlNDLAY AND CREIGHTON : THE INFLUENCE OF COLLOlDS ANDTABLE 8 (continued).Concentration : 1-61 grams in 100 C.C. of solution. Density = 1.005.Pressure ............ 735 841 966 1123 1239 1359Solubility __....... 0.791 0.773 0.783 0.797 Os801 0.804The influence of concentration of albumen on the solubility atatmospheric pressure is shown in Fig. 7.(h) SerumAlbumen.Neutral serum-albumen was prepared from fresh ox-blood by amethod due to Pauli.The blood-serum, to which was added a0*96000-94000.92000*9000d u.cl0.88003 h;, 60*86000-8400OT32000*8000Carbon dioxide am? charcoal and silica.small quantity of toluene t o prevent putrefaction, was placed insmall parchment cells suspended in closed glass vessels filled withdistilled water. The serum was dialysed for six weeks againstdistilled water saturated with toluene. During the first three weeksthe water was changed daily, thereafter every second day. ThFlNE SUSPENSIONS ON THE SOLUl3ILITY OF GASES IN WATER. 547concentration of the solutions was determined as in the case ofsolutions of egg-albumen.TABLE 9.-Solubility of Cwbon Dioxide in Solutions of Semtni-Albumen (see also Fig.5).Concentration: 0.44 gram of albumen in 100 C.C. of solution.Density = 0.998.Pressure ............ 748 844 945 1089 1246 1415Solubility ........ 0.804 0.800 0.802 0’804 0.806 0.806Concentration : 1-29 grams in 100 C.C. of solution. Density= 1.000.Pressure ............ 744 838 966 1066 1261 1431Solubility ......... 0.779 0.7T4 0.778 0.785 OT89 0.792The influence of concentration on the solubility at atmosphericpressure is shown in Fig. 7.(i) Silicic Acid.Solutions of silicic acid were prepared by dissolving pure silicain potassium hydroxide and adding excess of hydrochloric acid.The liquid was then dialysed, first against tap water, and thenagainst distilled water, until free from chloride.TABLE lO.--Solubility of Carbon Dioxide in Solwtions of SilicirA cid.Concentration: 1-40 grams of SiO, in 100 C.C.of solution.Density = 1.000.Pressure ............ 731 829 936 1064Solubility ......... 0.822 0.819 0.816 0’816Concentration : 2-20 grams in 100 C.C. of solution.Pressure ............ 732 836 938 1038Solubility ......... 0 828 0 822 0.820 0’820Concentration: 2.80 grams in 100 C.C. of solution.Pressure ........... 731 873 960 1050Solubility ......... 0.831 0’825 0.824 0.8231193 13540 3 1 6 0.816Density= 1.002.1178 13350.820 0.820Density = 1.003.1203 13300.824 0.825I n this case the solubility-pressure curves are similar in form toThe influence of concentration on the solubility at atmosphericthose for carbon dioxide and ferric hydroxide (Fig. 1).pressure is shown in Fig.7.( j ) Suspensions of Charcoal and of Silica.Suspensions of Kahlbaum’s well-powdered bone charcoal and ofThe solubility of carbon dioxide in pure silica were employed.presence of such suspensions is given in the following table:VOL. XCVII. 0 548 FINDLAY AND CREIGHTON THE INFLUENCE OF COLLOIDS ANDTABLE 11 (see also Fig. 6).0,236 gram of charcoal in 100 C.C. Density= 1.000.Pressure ............ 743 812 909 1069 1160 1250 1372Solubility ........... 0.815 0'823 0'845 0.892 0.919 0.940 0.9500.253 gram of silica in 100 C.C. Density=1*000.Pressure ........... 748 849 962 1048 1182 1274 1359Solubility ............ 0.814 0'815 0'818 0.819 0-821 0.822 0'824FIG. 7.N0V%0 1 2 3 4 5 6 7Conmlztrcction in grams per 100 C.C.solution.In the case of charcoal suspensions, the initial comparativelyrapid absorption of gas was followed by a comparatively slowFIG. 8.0-82000*80000.78000.76000'7400 *0 2-5 5.0 7.5 10.0 12.5 15.0 17'5Concentration in grams per 100 C.C. solution.absorption lasting from six to ten hours. The solubility valuesgiven in the above table are calculated from the maximum volumFINE SUSPENSIONS ON THE SOLUBILITY OF GASES IN WATER. 549of gas absorbed by the liquid. In the case of silica suspensions,t.he liquid quickly became saturated with the carbon dioxide, noslow absorption being observed.II.--Solubility of Nitrous Oxide.The nitrous oxide was prepared by heating pure ammoniuninitrate in a flask a t about 210-225O.Before the heat was applied,the flask was thoroughly exhausted. When the pressure of nitrousoxide in the apparatus had become equal to atmospheric pressure(a manometer was attached to the apparatus), a certain amount ofthe gas was allowed to escape into the air. The outlet to the airwi19 then closed, and the nitrous oxide caused to bubble throughsolutions of potassium hydroxide and ferrous sulphate before beingstored in a gasholder filled with brine. Before being used, it wasdried by means of calcium chloride and phosphoric oxide. Thesolubility of the nitrous oxide in water, and in water containingcolloids and suspensions, was then determined in exactly the samemanner as with carbon dioxide. The following tables contain theresults obtained.TABLE 12.-Solubility of Nitrous Oxide in Water.Pressure ............758 842 967 1041 1185 13628olubility ......... 0.592 0.593 0.592 0.593 0592 0.592Pressure ............ 758 831 997 1082 1214 1351Solubility ........ 0,592 0.593 0.592 0.593 0.594 0.592Pressure ........... 758 888 971 1091 1190 1281Solubility ......... 0.591 0'592 0'591 0.592 0.593 0.593From these and other similar determinations, the mean value ofthe solubility of nitrous oxide in water was found to be 0.592, thesolubility being independent of the pressure within the limitsinvestigated.TABLE 13.-h'ohbdity of Nitrous Oxide in Fer& HydroxideSolutions (see also Fig. 9).Concentration : 0.625 gram of Fe(OH), in 100 C.C. of solution.Density = 1.001.Pressure ............758 846 934 1010 1121 1383Solubility ......... 0.590 0.586 0.584 0.588 0.588 0.588Concentration : 1.49 grams in 100 C.C. of solution. Density= 1*008.Pressure ............ 734 828 935 1078 1215 1432Solubility ........ 0.586 0-579 0.577 0.581 0'585 0.586Concentration : 4.061 grams in 100 C.C. of solution. Density = 1-029.Pressure ............ 754 835 883 1093 1208 1358Solubility ......... 0.578 0.573 0.571 0574 0.579 0.5800 0 550 FINDLAY AND CREIGHTON : THE INFLUENCE OF COLLOIDS ANDContrary to the behaviour of carbon dioxide, the solubility ofnitrous oxide is lowered by ferric hydroxide, the diminution of solu-bility being practically proportional to the concentration, as shownin Fig. 14. Further, the behaviour of nitrous oxide is unlike thatof carbon dioxide, in that there is no long period of slow absorptionobservable, neither at high nor at low pressures.FIG.9.750 850 950 1050 1150 1250 1350 1450Pressure in <mm. Eg.Nitrous oxide and dextrin (-).Nitrous oxide and ferric hydroxide (----).TABLE 14.-8olub&ty of Nitrous Oxide in Sohtions of Dextrin(see also Fig. 9).Concentration: 6.98 grams of dextrin in 100 C.C. of solution.Density = 1.018.Pressure ............ 739 822 949 1092 i239 1368Solubility ......... 0.549 0-550 0.555 0.560 0-562 0'569Concentration : 13.01 grams in 100 C.C. of solution. Density= 1.039.Pressure ............ 729 836 914 1023 I237 1358Solubility ......... 0.529 0.523 -- 0.526 0.533 0.540 0.544Concentration : 20.30 grams in 100 C.C.of solution. Density = 1.062.Pressure ............ 740 836 911 1149 1290 1360Solubility ......... 0.503 0'499 0.503 0.509 0'513 0-516Compare also Fig. 15FINE SUSPENSIONS ON THE SOLUBILITY OF GASES IN WATER. 5510.54000.5200TABLE 15.-h'olubility of Nitrous Oxide in Solutions of ArseniousSulphide.Concentration: 1-85 grams of As,S3 in 100 C.C. of solution.Pressure ............ 746 820 924 1055 1196 1346Solubility ......... 0.591 0.590 0'590 0'592 0.593 0'593Density = 1.004.Concentration : 2-29 grams in 100 C.C. of solution. Density = 1.007.Pressure ............ 746 850 1006 1110 1209 1300Solubility ......... 0.590 0586 0'588 0.589 0.589 0.590From these figures it is seen that arsenious sulphide is withoutinfluence on the solubility of nitrous oxide.-"732FIG.10.750 850 950 1050 1150 1250 1350 1450Pressure in mm. Hg.Nitrous oxide and starch (-).Nitrous oxide and glycogen (- - -1,TABLE 16,--SoZub%ty of Nitrous Oxide in Sohtions of Starch(see also Fig. 10).Concentration: 2.50 grams of starch in 100 C.C. of solution.Density = 1.009.Pressure ............ 742 871 1020 1166 1284 1441Solubility ......... 0.580 0.576 0.575 0-578 0.581 0.582Pressure ........... 742 848 929 1046 1261 1381Solubility ........., 0.561 0.554 0.553 0554 0.562 0.567Concentration : 6.89 grams in 100 C.C. of solution. Density = 1.021.Concentration : 10.00 grams in 100 C.C. of solution. Density = 1.030.Pressure ..... . ...... 742 860 948 1071 1235 1350Solubility ......... 0.550 0.544 0.545 0-545 0.553 0.555Concentration : 13.73 grams in 100 C.C.of solution. Density = 1.040.Pressure ............ 739 836 982 1136 1252 1387Solubility ...... .. 0.557 0.532 0.530 0'535 0-536 0.53552 FINDLAY AND CREIGHTON : THE INFLUENCE OF COLLOIDS ANDThe influence of concentration on the solubility is shown inFig. 15.TABLE 17.--SolubiEity of Nitrous Oxide in Solutions of Gelatin(see also Fig. 11).Concentration: 1.31 grams of gelatin in 100 C.C. of solution.Density = 0.999.Pressure ............ 731 849 937 1069 1176 1328Solubility ......... 0-589 0.590 0.590 0.592 0592 0'592Concentration : 3.09 grams in 100 C.C. of solution. Density = 1.003.Pressure ........... 730 858 950 1089 1230 1373Solubility ......... 0.581 0.582 0.584 0.586 0.588 0.588Concentration : 6.06 grams in 100 C.C.of solution. Density = 1.008.Pressure ............ 730 850 961 1097 1247 1379Solubility ......... 05$0 0-563 0'566 0-568 0'570 0.571The influence of concentration of gelatin on the solubility isshown in Fig. 14.FIG. 11.0*60005 5 0*5800rnb ru60.5600750 850 950 1050 1150 1250 1350 1450Pressure in mna. ZT9.Nitroits oxide and gelatin.TABLE 18.--Xolubility of Nitrous Oxide in Solutions of Glycogen(see Fig. 10).Concentration: 0.49 gram of glycogen in 100 C.C. of solution.Density = 0.999.Pressure ............ 738 889 977 110.2 1239 1386Solubility ......... 0.590 0.588 0.591 0'594 0.594 0-594Concentration: 1.00 gram of glycogen in 100 C.C. of solution.Density= 1.002.Pressure ............737 871 991 1050 1201 1360Soliihility ........ 0.585 0.584 0'589 0.591 0.594 0-596The influence of concentration of glycogen on the solubility ofnitrous oxide is shown in Fig. 14FINE SUSPENSIONS ON THE SOLUBlLITY OF GASES IN WATER. 553TABLE 19.--Solubility of Nitrous Oxide in, Solutions of Egg-Albumlen (see also Fig. 12).Concentration: 0.35 gram of albumen in 100 C.C. of solution.Density = 0.998.Pressure ............ 735 830 954 1139 1249 1363Solubility ......... 0.580 0.578 0.580 0.581 0.580 0.580Concentration : 0.75 gram in 100 C.C. of solution. Density = 1.000.Pressure ............ 735 820 872 951 1104 1344Solubility ......... 0.569 0.562 0-564 0'567 0573 0.577Pressure ............ 729 811 886 946 1199 1399Solubility ......... 0.548 0.535 0.540 0.544 0.553 0.558Concentration : 1.60 grams in 100 C.C. of solution.Density = 1.005.The influence of concentration on solubility is shown in Fig. 14.FIG. 12.750 850 950 1050 1150 1250 1360 1450Pressure in mm. Hg.Nitrous oxide and egg-albumen (-).NitTom oxide and serum-albumen (-- -).TABLE 2 0 . 4 o l u b i l i t y of Nitrous Oxide in Solutions of Sepum-A l b u m e n (see also Fig. 12).Concentration: 0.32 gram of serum-albumen in 100 C.C.Density = 0.998.Pressure ............ 746 873 978 1126 3.259 1395Soliibility ......... 0.583 0.581 0.579 0.586 0.588 0,591Concentration : 1-40 grams in 100 C.C. of solution. Density = 1.001Pressure ............ 743 842 913 1048 1228 1358Solubility .........0.537 0'538 0.545 0.550 0'558 0-562The influence of concentration on solubility is shown in Fig. 14554 FINDLAY AND CREIGH'I'ON : THE lNFLUENCE OF COLLOlDS ANDTABLE 21.--SoEubility of Nitrous Oxide in Solutions of Silicic Acid(see also Fig. 13).Concentration: 1.87 grams of SiO, in 100 C.C. of solution,Density = 1-001.Pressure ............ 748 825 921 1046 1217 1349Solubility ......... 0.596 0.598 0.598 O%OO 0.602 0'604Concentration : 3.63 gra.ms in 100 C.C. of solution.Pressure ............ 741 848 994 1122 1217 1394Solubility ......... 0'601 0'602 0.605 0.607 0-608 0'609Density= 1.005.The influence of concentration on the solubility is shown inFig. 14.O.iOO00'68000.6600A.4 .-0'6400 - m3t/l0'620Co '60000'580C7TABLE 22.--Solubility of Nitrous Oxide in Water containingCharcoal and Silica in Suspension (see also Fig.13).100 C.C. of liquid contained 0.227 gram of charcoal. Density= 1.000.Pressure ..._........ 729 824 936 1034 1150 1254 1356So!ubility ............ 0.596 0.600 0.618 0.635 0.648 0.661 0.674100 C.C. of liquid contained 0.30 gram of SiO,. Density=1'000.Pressure ............ 730 846 960 1081 1224 2365 1481Solubility ............ 0.592 0-593 0.595 0.597 0.597 0.600 0-60FlNE SUSPENSIONS ON THE SOLUBILITY OF GASES IN WATER. 555111.-Solubility of Carbon Dioxide in Solutions of Aniline.In order that the solubility curves previously obtained mightbe compared directly with a case where chemical combination isknown to occur, the solubility of carbon dioxide in solutions ofaniline was determined.The results are contained in table 23.FIG. 14.Concentration in grams per 100 C.C. solution.TABLE 23.Concentration: 0.206 gram of aniline in 100 C.C. of solution.Pressure ............ 748 808 920 1053 1159 1243Solubility ......... 0.865 0.855 0-857 0.855 0'862 0'860Concentration: 0.425 gram in 100 C.C. of solution.Pressure ............ 760 816 921 1150 1236 1380Solubility ......... 0'909 0.897 0.897 0.897 0'902 0.908Concentration: 0.566 gram in 100 C.C. of solution.Pressure ............ 760 823 941 1082 1223 1341Solubility ......... 0.935 0.929 0.925 0.923 0-924 0.930Concentration: 0.743 gram in 100 C.C. of solution.Pressure ............ 760 895 983 1063 1223 1302Solubility .........0-953 0.941 0'940 0-940 0.940 0-942The solubility-pressure curves are similar in form to those forcarbon dioxide and ferric hydroxide (Fig. 1).IV.-Solubility of Carbon Dioxide in Solu,tions of PotassiumChloride.Although many investigators (see Steiner, AnnaZem, 1894, 52,275; Gordon, Zeitsch. phgsikal. Chem., 1895, 18, 1; Braun, dbid.556 FINDLAY AND CREIGHTON : THE INFLUENCE OF COLLOIDS AND1900, 33, 721; Knopp, ibid., 1904, 48, 97; Hufner, ibid., 1907,57, 611) have studied the influence of dissolved substances, bothelectrolytes and non-electrolytes, on the solubility of gases, suchFrG. 15.-5 10-0 12-5 15.0 17.Concentration in gram per 100 C.C. solution.investigations have always been made at only one pressure. I norder that a comparison might be made between the influence ofcolloids and suspensions (emulsoids and suspensoids) and of truePressure in warn. Hg.Carbon dioxide and potassium chloride.solutions, the solubility of carbon dioxide in solutions of potassiumchloride at different pressures was determined. The results arecontained in table 24, and represented in Fig.16FINE SUSPENSIONS ON THE SOLUBILITY OF GASES IN WATER. 557TABLE 24 (see also Fig. 16).Concentration: 7.45 grams of KCl in 100 C.C. of solution.Density = 1.043.Pressure ............ 756 850 953 1116 1249 1362Solubility ........ 0.694 0-693 0.688 0-700 0.709 0.710Concentration: 5.00 grams of KCl in 100 C.C. Density=1*031.Pressure ............ 756 832 901 1050 1150 1223Solubility .........0.731 0-727 0'724 0.726 0.735 0.736Concentration: 2-56 grams of KCI in 100 C.C. Density=1*016.Pressure ........... 756 852 981 1079 1190 1362Solubility ......... 0.767 0'761 0.761 0'762 0.768 0.766Discussion of Results.A glance a t the curves given on the preceding pages will showthat many peculiarities of behaviour are found in the solubilityof gases in liquids when that solubility is investigated, not, aspreviously, at only one pressure, but a t different pressures. Sovaried, indeed, is the influence, not only in degree, but in kind,of the different solutes or pseudo-solutes and suspensions on thesolubility of carbon dioxide and nitrous oxide, that conclusionsdrawn from the behaviour under one pressure might be veryerroneous when considered for another pressure.The substances the influence of which on the solubility of thetwo gases, carbon dioxide and nitrous oxide, has been studied,may be divided into emulsoids and suspensoids.To the formerclass belong ferric hydroxide, gelatin, starch, glycogen, egg-albumen, serum-albumen, and silicic acid; to the latter clam,arsenious sulphide, charcoal, and silica. Dextrin may, perhaps, beregarded as intermediate between a true solute and an emulsoid.Aniline and potassium chloride have been included merely for thepurposes of comparison.In the case of the emulsoids, we see that under atmosphericpressure, silicic acid increases the solubility both of carbon dioxideand of nitrous oxide; ferric hydroxide and gelatin increase thesolubility of carbon dioxide but diminish the solubility of nitrousoxide; and the other emulsoids decrease the solubility of both thegasea Of the suspensoids, arsenious sulphide is practically withoutinfluence on the solubility of either gas, while charcoal and silicaincrease the solubility of both gases.Lastly, dextrin decreasesthe solubility of both carbon dioxide and of nitrous oxide.What the nature of the interaction may be in the case ofgelatin and carbon dioxide is not, perhaps, quite easily decided558 FLNDLAY AND CREIGHTON : THE INFLUENCE OF COLLOIDS ANDGelatin, as is known, is an amphoteric substance, and may thereforefunction as a weak base. It seems to us, however, tu be doubtfulif this basic property is sufficient in itself to explain t,he wholeincrease in the solubility.Possibly some more complicated actionoccurs, an indication of which appears to be given by the effectof carbon dioxide in lowering the gelatinisation point of gelatinsolutions already referred to (p. 544).I n the case of ferric hydroxide, as has already been pointed outby Luther and Krsnjavi (Zeitsch. physikaZ. Chem., 1905, 46, 170),there is probably complex ion formation. The formation of aferric carbonate appears from the work of Raikow (Chem. Zeit.,1907, 31, 87) and of Cameron and Robinson (J. Physical Chem.,1908, 12, 561) to be excluded.Although in these cases we may regard chemical combination asbeing one of the causes, perhaps the main cause, of the increasedsolubility of carbon dioxide, it is difficult to adopt a similarexplanation in the case of silicic acid, which increases the solubilityboth of carbon dioxide and of nit-rous oxide.In the latter casean explanation is more probably to be sought in the phenomena ofadsorption (see also p. 560).Solubility Referred to the Water in the- Solutions.-With regardto the lowering effect of electrolytes and non-electrolytes on thesolubility of gases in water, the view has been expressed, moreespecially by J. C. Philip (Trans., 1907, 91, 711), that the observeddepression wn be explained on the assumptions: (1) that onlythe water in the solution acts as solvent for the gas, and that thissolvent power is not affected by the presence of the solute; (2) thatthe solute molecules are more or less hydrated, and thereforediminish the amount of active solvent for the gas; (3) that thegas does not dissolve in the solute, whether anhydrous or hydrated.It must be borne in mind that the cases investigated by us arenot generally comparable with those to which Philip applied histheory, for with the exceptions of the solutions of potassiumchloride, and, possibly, dextrin, the solvent systems were not homo-geneous, but must be regarded, most probably, as heterogeneous.And that alters the case entirely.With regard to the solutions of potassium chloride, it may bementioned that the values calculated for the degree of hydrationvary from 6.42 to 8.68 molecules of water to one molecule of salt.These numbers are rather lower than those calculated by Philip,but not greatly so.In the cas0 of dextrin, however, it is evident that the solutionsof dextrin show considerably different behaviour, according asthe solubility of carbon dioxide or of nitrous oxide is investigatedFINE SUSPENSIONS ON THE SOLTJBILITY OF GASES IN WATER.559In the former case practically no hydration is evidenced; in thelatter case a slight amount of hydration would be calculated. Butin the case of carbon dioxide it will be seen that the numbersrepresenting the solubility referred to water in the solutiondiminish with increase of concentration, whereas in the case ofnitrous oxide, the numbers increase.It does not appear to us that sufficient evidence has yet beenadduced in support of the theory put forward by Philip.More-over, we believe that the solubility curves which we have obtaineda t higher pressures show the necessity of extending the range ofinvestigation in this direction. On this we are at presentengaged. *Change of Solubility with Pressure.-Whatever conclusions maybe drawn as regards the influence of the suspensoids and emulsoidson the solubility of carbon dioxide and nitrous oxide from deter-minations at one pressure, they must to a greater or lesser degree befound inaccurate when applied over a range of pressures; for asthe figures previously given show, the solubility is not independentof the pressure (as it is in the case of pure water), nor are thesolubility curves for solutions of different concentration in all casesparallel.Assuming that the influence of ferric hydroxide and of gelatinis mainly due to chemical combination with formation of alargely hydrolysed compound, we should expect that thesolubility-pressure curve would first fall, owing to hydrolysis,and then remain nearly horizontal, owing to the diminution ofhydrolysis by addition of carbonic acid.This is the type of curveobtained with aniline and ferric hydroxide, but is better seen inthe case of the more weakly basic substance gelatin. Looked a tin this way, the curve for carbon dioxide and silicic acid wouldalso indicate chemical combination, and we should therefore haveto assume that silicic acid is amphoteric (for which we do notknow of any other evidence), or that between silicic acid andcarbonic acid a reaction takes place comparable with that betweensilicic acid and hydrofluoric acid, the compound formed beinghighly hydrolysed.As regards the influence of suspensions of charcoal and silica onthe solubility of carbon dioxide and nitrous oxide, and of silicicacid on the solubility of nitrous oxide, it will be noticed that weare here dealing with curves similar to those obtained by otherinvestigators for the absorption of gases or of dissolved substancesby charcoal.* Since this was written, a paper has appeared (this vol., p.66) by F. L. Usher,who fails t o find confirmation of the theory put forward by Philip560 FINDLAY AND CREIGHTON : THE INFLUENCE OF COLLOIDS ANDSimilarity is also shown by the fact that when one examines therelation between the concentration of gas in the water and in thesolid, the general relationship cg/cl = const., found by previousworkers for “ adsorption ” phenomena, also holds in the presentcases.Here c2 is the weight of gas taken up by the water in100 C.C. of the suspension, and c1 the weight taken up by thesuspended solid. x we have found to be equal to 4.TABLE 25.Charcoal and Carbon Dioxide,0.236 gram of Charcoal in 100 c.c.Charcoal and Nitrous Oxide,0.227 gram of Charcoal in 100 C.C.Pressure. c,. C,. C y C I .950 0’0086 0-1837 0.1321000 0.0127 0’1934 0.1101050 0.0178 0’2030 0.0991150 0.0266 0-2224 0.0921250 0’0365 0‘2417 0.0981350 0.0430 0.2610 0.108Pressure. cl. %* c&950 0-0059 0.1332 0.0531000 0.0085 0.1402 0.0481050 0-0109 0.1472 0.0431150 0.0151 0-1613 0.0451250 0.0200 0.1753 0-0471350 0.0259 0’1893 0.050In the above cases, therefore, increased solubility would beascribed to “ adsorption,” accompanied or unaccompanied byabsorption.The solubility curves so far discussed are comparatively simplein form, and the influence of the colloid or suspension may plausiblybe explained on the basis of partial chemical combination or of‘ I adsorption,” accompanied or unaccompanied by absorption.Inthese cases the solubility of the gases is increased a t all pressures.In most of the cases examined, however, where dealing withemulsoids, the solubility of the gas is diminished even when onetakes into account the volume of water in the solution. This maybe explained, formally, by the assumption of hydrate formation.But even if this be accepted, the remarkable behaviour observedat higher pressures remains to be accounted for.An examinationof the solubility-pressure curves shows that, with the exception ofthe cases already discussed, there exists for a number of thecolloids a very well-defined minimum of solubility, this minimumbeing more marked in concentrated than in dilute solutions. I nall such cases t h influence of the colloid must be a very complexone, and two effects a t least must enter into play, one causing adiminution of solubility with pressure, the other an increase. Indilute solutions the former is sometimes absent or negligible.So far as the rising portion of thecurves is concerned, we assume that the rise is due t o “ adsorption,”whatever the true nature of this process may be. For this portionof the different curves we have also found that the ratioci/cl = const., as is shown by the following table:What are these two factorsFINE SUSPENSIONS ON THE SOLUBILITY OF GASES IN WATER. 561TABLE 26.Ses.umAZbwnen and Nitrous Oxide.1-40 grams of albumen per 100 C.C.915 0.0017 0'1164 0.1081000 0*0026 OC1272 0~1001050 0*0031 0.1336 0-1031150 0.0047 0.1463 0.0981250 0*0061 0.1590 0.1051350 0.0079 0.1717 0.1110.537 is the valnc of the solubility used for calculating c2.Pressure. el - c,. c;/cl.A similar degree of constancy is obtained in the case of theother curves.As regards the factor producing the lowering of the solubility,we believe that the simplest assumption to make is that ofsolubility of the gas in the colloid phase. As has already beenpointed out, we are dealing here with heterogeneous systems,comparable with a mixture of two partly miscible liquids. So faras we are aware, the solubility of a gas in such a system has notyet been investigated; but we may very properly assume that thegas dissolves (unequally) in the two phases, namely, in the casesunder discussion, in the aqueous phase and the colloid phase. Wemust also further assume that the solubility in the colloid phaseno longer follows Henry's law, but that the solubility increasesless rapidly than the pressure. Under such conditions the solubilitycurve would no longer be a straight line, but would fall withincrease of pressure. Such a deviation from Henry's law signifiesthat the molecular weight of the gas in the gaseous phase and inthe colloid phase is no lodger the same; and we must thereforeassume that the gases have a higher molecular weight in the colloidphase than in the water phase. That is, we must assume poly-merisation of the gas in the colloid phase. By these assumptionswe are enabled to explain, formally a t least, the behaviourobserved, and it must be left to future investigation to showwhether the explanation is only formal or may be regarded asessential. It is clear, however, from the foregoing investigationthat colloids in solution will not necessarily increase the solubilityof a gas. The action is a specific one, and depends both on thecolloid and on the gas.CHEMICAL DEPARTMENT.UNIVERSITY OF BIRMINGHAM