CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. A STUDY OF THE CONDUCTIVITY AND VISCOSITY OF SOLUTIONS OF CERTAIN ELECTROLYTES IN WATER, METHYL ALCOHOL, ETHYL ALCOHOL, AND ACETONE; AND IN BINARY MIXTURES OF THESE SOLVENTS. BY HARRY C. JONES, PROFESSOR OF PHYSICAL CHEMISTRY IN THE JOHNS HOPKINS UNIVERSITY, AND C. F. LINDSAY, C. G. CARROLL, H. P. BASSETT, E. C. BINGHAM, C. A. ROUILLER, L. McMASTER, W. R. VEAZEY. WASHINGTON, D. C. PUBLISHED BY THE CARNEGIE INSTITUTION OF WASHINGTON. 1907 CARNEGIE INSTITUTION OF WASHINGTON. PUBLICATION No. 80. / / / NorfoooD J. 8. Gushing Co. — Berwick & Smith Co. Norwood, Mass., U.S.A. CONTENTS. PAGE INTRODUCTION 1 HISTORICAL SKETCH 3 Inorganic Solvents . 3 Hydrocyanic Acid 3 Water 3 Ammonia 4 Nitric Acid 4 Sulphur Dioxide 4 Organic Solvents 7 Hydrocarbons 7 Alcohols 7 Ethyl Alcohol 8 Higher Alcohols 9 Ether 10 Ketones 10 Acids 11 The Nitriles and Cyanogen 11 Pyridine 12 Other Organic Solvents . . 12 Mixed Solvents 13 Hydrogen Dioxide and Water 13 Mixtures of Water and Alcohols 13 Conductivity and Viscosity 16 Viscosity 19 WORK OF LINDSAY 24 Experimental 24 Apparatus 24 Solvents 25 Water 25 Methyl Alcohol 25 Ethyl Alcohol 25 Propyl Alcohol 25 Solutions .... 26 Conductivity Measurements 26 Potassium Iodide 26 Ammonium Bromide 30 Strontium Iodide 32 Cadmium Iodide 34 Lithium Nitrate .35 Ferric Chloride . .38 Summary 40 iii IV CONTENTS. PAGE WORK OF CARROLL . . 43 Experimental 43 Apparatus 43 Solvents .............. 43 Method of Preparing the Solutions 44 Conductivity Measurements .......... 44 Cadmium Iodide 44 Sodium Iodide 47 Calcium Nitrate 49 Hydrochloric Acid 50 Sodium Acetate in Mixtures of Acetic Acid and Water .... 52 Dissociation in Fifty Per Cent Methyl Alcohol 53 Potassium Iodide 53 Sodium Iodide 54 Cause of the Minimum 58 Discussion of Results 66 Viscosity and Conductivity 68 Summary and Conclusions 73 WORK OF BASSETT 75 Experimental Work 75 Conductivity Apparatus Employed ........ 75 Conductivity Measurements 76 WORK OF BINGHAM 81 Experimental 81 Apparatus .............. 81 Conductivity ............. 81 Viscosity 81 Preparation of Solutions 82 Solvents 83 AVater 83 Methyl Alcohol 83 Ethyl Alcohol 83 Acetone 83 Conductivity Measurements 83 Lithium Nitrate ............ 84 Potassium Iodide 89 Calcium Nitrate 95 Viscosity Measurements 103 Discussion of Results 107 Summary 113 WORK OF ROUILLER 115 Object of this Investigation ........... 115 Solvents 115 Water 115 Methyl Alcohol 115 Ethyl Alcohol 116 Acetone 116 Mixed Solvents 116 Conductivity 116 Apparatus 116 CONTENTS. v PAGE WORK OF ROUILLER — Continued. Preparation of Solutions . . . . . . . . . .117 Conductivity Measurements . . . . . . . . . .117 WORK OF MCMASTER . . . ' 126 Experimental ............. 126 Apparatus ............. 126 Conductivity 126 Viscosity ............ 126 Solvents 127 Water 127 Methyl Alcohol 127 Ethyl Alcohol .127 Acetone 127 Solutions 128 Conductivity Measurements 128 Lithium Bromide 129 Conductivity and Viscosity of Certain Salts 139 Cobalt Chloride 139 Viscosity Measurements 151 Discussion of Results 159 Fluidity and Conductivity 159 Temperature Coefficients 164 Summary 168 WORK OF VEAZEY 170 Experimental ............. 170 Apparatus ............. 170 Conductivity 170 Viscosity ............ 171 Solvents ............. 172 Water ............. 172 Methyl Alcohol 173 Ethyl Alcohol 173 Acetone ............. 173 Solutions ............. 173 Conductivity Measurements 173 Cobalt Nitrate 174 Copper Chloride 174 Potassium Sulphocyanate 181 Viscosity Measurements 193 A Summary of the Facts Established 202 DISCUSSION OF RESULTS 207 Negative Viscosity Coefficients .......... 213 Conclusions ... 216 GENERAL SUMMARY AND CONCLUSIONS 219 INDEX 229 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. INTRODUCTION. The earlier work in physical chemistry had to deal almost exclusively with aqueous solutions. This was due to the fact that water is by far the most universal solvent. It dissolves a much larger number of substances than any other known liquid. Further, water has greater power to break molecules down into ions than any other common solvent. For these reasons water is the most important solvent chemically; indeed, chemistry is largely a science of aqueous solutions. Again, water, of all the common solvents, is the most easily obtained, and in a fair degree of purity. These are some of the reasons why solutions in water as the solvent were studied first. During the last few years, the measurement of dissociation has been extended, to a greater or less extent, to solutions in many solvents, both inorganic and or- ganic, and in several cases interesting and important results have been obtained. The study of non-aqueous solutions has led to a comparison of the dis- sociating power of the various solvents, and this in turn has given rise to several generalizations which attempt to connect dissociating power with other physical and chemical properties of solvents. J. J. Thomson l and Nernst 2 have sought to connect the dissociating power of a solvent with its dielectric constant. Nernst says: The greater the dielectric constant of a medium, the greater becomes its electrolytic dissociation of dissolved substances under exactly similar conditions. J. J. Thomson, after showing that molecules condensed on the surface of a conducting sphere will be completely dissociated, goes on to say : The same effect would be produced by a substance possessing a very large specific inductive capacity. Since water is such a substance it follows, if we accept the view that the forces between the atoms are electrical in their origin, that when the molecules of a substance are in solution, the forces between them are very much less than they are when the molecule is free and in the gaseous state. Briihl 3 showed that while certain organic bodies, as the oximes and the alcohols, exist in a polymerized state when dissolved in hydrocarbons, chloroform, or carbon disulphide, the molecular complexes are more or less broken down in aqueous solution, and to a less extent in alcohols, ethers, esters, ketones, and phenols. All of these also exert more or less dissociating power. According to his theory of the tetravalence of oxygen they are 'Phil. Mag., 36, 320 (1893). 2 Ztschr. phys. Chem., 13, 531 (1894). 3 Ibid., 18, 514 (1895); 27, 319 (1898); 30, 1 (1899). Ber. d. chem. GeselL, 28, 2847, 2866 (1895); 30, 163 (1897). 2 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. unsaturated compounds, and it is to this unsaturation that he ascribes dis- sociating power. He also attributes the dissociating power of solvents con- taining nitrogen to the fact that in these compounds nitrogen exists as a triad, while it has the power of becoming pentavalent. He predicted disso- ciating power for compounds containing trivalent nitrogen, including the hydrazines, the amines, the diazo compounds, and liquid hydrocyanic acid; and even for other classes of unsaturated compounds, as the trichlorides of arsenic and phosphorus, the mercaptans, and alkyl sulphides. This prophecy is borne out, to some extent at least, by the later work of Walden. In his last paper Briihl sought rather to connect dissociating power with a high dielectric constant, and with the tautomerizing power of the solvent. Ciamician 1 concluded that dissociating power is a function of the chemical properties of the substance, and that those substances which resemble water chemically, as methyl and ethyl alcohols, should have the greatest dissoci- ating power. Konowalow,2 from a study of the conductivity of the compounds which amines form with acids, holds that only those solutions conduct in which there is chemical action between the solvent and the dissolved substance. Dutoit and Aston3 advanced the idea that dissociating power is related to the amount of the polymerization of the solvent. Water and the alcohols, which are good dissociants, also exist as polymerized molecules, as is shown by the surface-tension method of Ramsay and Shields.4 For further views on this subject consult the paper by Crompton.5 Donnan says : In a solution in which the solute is more or less ionized, one might suppose the ions to be surrounded by clusters of solvent molecules which had, so to speak, condensed around them, and opposed an obstacle to their recombination. Now one might suppose this state of things as being caused by some sort of specific attraction between the solvent molecules and electricity, i. e., the electrons or electrical charges which are associated with the ions. Were this the case, one might expect this specific attraction to mani- fest itself in other ways. For example, if electrical nuclei were present in, or were pro- duced by any means in air which was saturated with the vapor of an ionizing liquid, then it would be just possible that the specific attraction referred to above might help to produce condensation of the vapor around these nuclei under suitable conditions, i.e., if the vapor were supersaturated by a sudden adiabatic expansion. If the liquid in question did not act as an ionizing solvent, it would be natural to expect that the condensa- tion just alluded to would only occur when the vapor entered the really unstable (labile) region, or at any rate would only be produced by a much higher degree of supersaturation. From his experiments, however, he does not feel warranted in drawing any final conclusion. lZtschr. phys. Chem., 6, 403 (1830). 2 Wied. Ann., 49, 733 (1893). 3 Compt. rend., 125, 240 (1897). See also Dutoit and Friderich: Bull. Soc. Chim., [3] 19, 325 (1898). 'Ztschr nhys. Chem., 12, 433 (1893). 6 Journ. Chem. Soc., 71, 925 (1897). HISTORICAL SKETCH. 3 HISTORICAL SKETCH OF WORK IN NON-AQUEOUS SOLVENTS. In considering more in detail the results which have led to the above generalizations, we shall consider the work done, first in inorganic solvents, second in organic solvents, and third in mixed solvents. INORGANIC SOLVENTS. Water has always been regarded as the best dissociant, but recently another solvent, liquid hydrocyanic acid, has been found to produce greater disso- ciation. HYDROCYANIC ACID. Schlundt1 has measured the dielectric constant of liquid hydrocyanic acid, and found the very large value of 95 at 21°, a value which exceeds that of water, which is 80, at the same temperature. It was, therefore, important, as bearing on the Nernst-Thomson theory of dissociation, that measurements of the conductivity of solutions in this solvent be made. This has been done by Centnerszwer,2 with the result that not only do solu- tions in hydrocyanic acid show greater conductivity, but the dissociation is also greater than in water. The substances worked with were potassium iodide and trimethylsulphonium iodide. Their conductivity at 0° was nearly the same as the conductivity of aqueous hydrochloric acid at 25°. WATER. The dissociation of a great number of substances in aqueous solution has been determined by a variety of methods, including the conductivity method of Kohlrausch,3 the freezing-point method of Jones,4 Loomis,5 and others, and the solubility method of Nernst6 and Noyes.7 The result of this work has been to show that for a strong acid or strong base, or a salt of a strong acid and a strong base, at a dilution of about 1000 liters the dissociation is practically complete. In most solvents, however, it is impossible to deter- mine directly, by the conductivity method, the value of the molecular con- ductivity for complete ionization, since the dilution at which this is reached is so great as to preclude the application of the conductivity method. The best that we can do in these cases is to compare the values of /ic, at varying dilutions, with the corresponding values of /*„ in aqueous solution. In this way an approximation to the dissociating power of various solvents can be obtained. 1 Journ. Phys. Chem., 5, 157 (1901). 6 Wied. Ann., 61, 500 (1894); 67, 495, 591 'Ztschr. phys. Chem., 39, 217 (1902). (1896); 60,523(1897). 3 Wied. Ann., 26, 160 (1885). 6 Ztschr. phys. Chem., 4, 372 (1889). 4Ztschr. phys. Chem., 11, 110,529; 12,623 7 Ibid., 6,241 (1890); 9,603(1892); 12,162 (1893). (1893); 16, 125 (1895). 4 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. AMMONIA. Several years ago Cady l noticed that solutions of salts in liquid ammonia conduct the current. Goodwin and Thomson2 made some measurements of the conductivity of such solutions, while at work on the dielectric constant of liquid ammonia. The most elaborate work, however, on this subject is that of Franklin and Kraus.3 They measured the conductivity of potassium bromide and nitrate, sodium bromide and bromate, ammonium chloride and nitrate, silver iodide and cyanide, besides other inorganic salts, and organic compounds over very great changes in dilution. A direct comparison of the values of the conductivities in liquid ammonia with similar values in water, shows that the former are much larger than the latter. This, however, does not necessarily mean a larger dissociation, since the conductivity is dependent on two factors, namely, the dissociation and the velocity of the ions. The percentage dissociation («=— ) is larger V /*»/ in water than in ammonia, and hence the large conductivity of solutions in liquid ammonia is due rather to the high velocity of the ions than to the large number present. NITRIC ACID. The only work on solutions in nitric acid is that of Bouty,4 who has meas- ured the conductivity of certain alkaline nitrates when dissolved in nitric acid. The conductivities are nearly as large as in water, but the work is too fragmentary to permit making comparisons between the dissociating power of nitric acid and of water. The former is, however, in all probability, a good dissociant. SULPHUR DIOXIDE. Walden and Centnerszwer 5 published the results of an extensive investiga- tion on sulphur dioxide as a solvent. This is an extension of the older work of Walden.8 They investigated the conductivity of nineteen salts, consisting of iodides, bromides, chlorides, and sulphocyanates of inorganic and organic bases. They show, first, that while in aqueous solution the molecular con- ductivities at 25° of monobasic halogen salts generally lie between 100 and 140, the corresponding values in sulphur dioxide vary between 3 and 157; second, that Kohlrausch's law of the independent migration velocities of the ions does not hold for solutions in sulphur dioxide. They have also shown by a series of conductivity measurements at different temperatures between -78° and 157° (the freezing-point and critical temperature of 1 Journ. Phys. Chem., 1, 707 (1896). " Compt. rend., 106, 595 (1888). 2Phys. Rev., 8, 38 (1899). 6 Ztschr. phys. Chem., 39, 513 (1902). 3Amer. Chem. Journ., 23, 277; 24, • Ber. d. chem. Gesell., 32, 2862 (1899). 83 (1900). INORGANIC SOLVENTS. 5 sulphur dioxide) , that the molecular conductivity at first increases with the temperature, passes through a maximum, and then diminishes as the critical temperature is approached. This is seen at once to be just what would be expected from the polymerized solvent theory of Dutoit and Aston. As the temperature rises the association of the solvent decreases, and this would be expected to diminish the ionizing power. They have also determined the molecular weights of a number of electrolytes (salts) in liquid sulphur dioxide, by the boiling-point method; reaching the remarkable result that many of these salts show a molecular weight greater than normal, or what is the same thing, the value of the van't Hoff coefficient "i" is less than unity. This they endeavor to show is due to the fact that, in addition to the electrolytic dissociation, an association takes place in the solutions in sulphur dioxide. Facts similar to the above were noticed by Franklin and Kraus * in their work on liquid ammonia. The remaining work in inorganic solvents we owe chiefly to Walden.2 He has investigated the solvent and ionizing power of the following com- pounds : Phosphorus trichloride, phosphorus tribromide, phosphorus oxy- chloride, arsenic trichloride, antimony trichloride, antimony pentachloride, boron trichloride, silicon tetrachloride, tin tetrachloride, sulphur monochlo- ride, sulphuryl chloride, thionyl chloride, sulphur trioxide, and liquid bro- mine. Of these, sulphur monochloride, sulphuryl chloride, thionyl chloride, phosphorus oxychloride, arsenic trichloride, and antimony trichloride show considerable ionizing power, while solutions in the remaining solvents ex- hibit only the very slightest conductivity. In his next paper Walden adds a study of arsenic tribromide, chlorsulphuric acid, sulphuric acid, and the dimethyl ester of sulphuric acid. All of these show a strong tendency to ionize dissolved electrolytes. It is important to notice from this work of Walden that there appears to be no connection between dissociating power and chemical constitution. Antimony pentachloride does not dissociate electrolytes, while the trichloride dissociates to a very considerable extent. On the other hand, phosphorus trichloride does not dissociate, while phos- phorus oxychloride does. It is thus evident that, among the inorganic sol- vents at least, a knowledge of the dissociating power of one solvent tells us nothing as to the dissociating power of substances closely related chemically. Oddo 3 has also shown that phosphorus oxychloride strongly ionizes elec- trolytes. Tolloczko,4 as well as Garelli and Bassani,5 have worked with the halides of arsenic and antimony, showing them to have ionizing power. 1 Amer. Chem. Journ., 20, 836 (1898); 24, 3 Atti R. Accad. dei Lincei Roma, [5] 10 83 (1900). 452. 1 Ztschr. anorg. Chem., 25, 209 (1900); 29, 4 Ztschr. phys. Chem., 30, 705 (1899). 371 (1903). B Atti R. Accad. dei Lincei Roma, [5] 10 255. 6 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. Kahlenberg and Lincoln l studied solutions of ferric chloride, antimony tri- chloride, bismuth trichloride, and mercuric chloride in phosphorus trichlo- ride and arsenic trichloride, with results which confirm those mentioned above. Centnerszwer 2 is authority for placing cyanogen among the solvents that do not dissociate. Frankland and Farmer3 have also shown that nitrogen peroxide does not dissociate, while Skilling 4 demonstrates that solu- tions in hydrogen sulphide show no conductivity. The following table of inorganic solvents is given to show what relations exist between dissociating power, dielectric constants, and the association factor : TABLE 1. — Inorganic solvents which effect dissociation. Solvent. Dielectric constant. Association factor. Considered as — Hydrocyanic acid 95 81.12 16.2 13.75 (?) 12.35 (?) 13.9 33.2 9.05 9.15 (?) (?) (?) 4.8 (?) 3.7 1.0 1.0 1.7-1.9 (?) (?) 1.00 (?) 1.08 0.97 (?) (?) 32.0 0.95-1.05 Saturated. Do. Unsaturated. Do. Saturated. Unsaturated. Do. Saturated. Unsaturated. Do. Saturated. Do. Do. Do. Unsaturated. Water Ammonia Sulphur dioxide Nitric acid Arsenic trichloride Arsenic tribromide Phosphorus oxychloride . . . Antimony trichloride .... Thionyl chloride ..... Sulphuryl chloride Dimethyl sulphate Chlorsulphuric acid Sulphuric acid Sulphur monochloride .... Inorganic solvents which do not dissociate electrolytes. Bromine 3.18 2.52 3.56 (?) 3.36 (?) 3.78 (?) 3.2 (?) (?) 1.2-1.3 (?) Saturated. Do. Do. Do. Unsaturated. Do. Saturated. Do. Do. Do. Do. Cyanogen Sulphur trioxide Boron trichloride (?) 1.02 (?) (?) 1.06 (?) (?) (?) Phosphorus trichloride Phosphorus tribromide Antimony pentachloride . . . Silicon tetrachloride .... Tin tetrachloride ..... Hydrogen sulphide Nitrogen peroxide The values for the association factors are taken from the researches of Ramsay and Shields,5 and Ramsay and Aston;6 while those for the dielectric constants are almost wholly taken from the work of Turner.7 'Journ. Phys. Chem., 3, 12 (1899). 2Ztschr. phys. Chem., 39, 217 (1902). 3 Journ. Chem. Soc., 79, 1356 (1901). 4Amer. Chem. Journ., 26, 383 (1901). 5 Ztschr. phys. Chem., 12, 433 (1893). 8 Journ. Chem. Soc., 65, 167 (1894). 7 Journ. Phys. Chem., 5, 503 (1901). ORGANIC SOLVENTS. 7 ORGANIC SOLVENTS. HYDROCARBONS. Kahlenberg and Lincoln l have shown that solutions of ferric chloride in a large number of hydrocarbons do not conduct the current, while Kablukoff2 showed that the conductivity of hydrochloric acid in benzene, xylene, and hexane is very small. This is in perfect accord with what would be expected from the fact that non-electrolytes dissolved in hydrocarbons tend, in a number of cases, to give a complex molecular weight when this is determined by the boiling-point or freezing-point method. ALCOHOLS. When we come to study the work on solutions in the alcohols, we find that a considerable amount has been done, especially in the case of the two lowest members of the aliphatic series. Fitzpatrick 3 studied the conduc- tivities of calcium nitrate, lithium nitrate, lithium chloride, and calcium chloride in methyl alcohol, and found values which, though less than in water, were very considerable. Hartwig4 measured the conductivity of formic, acetic, and butyric acids in methyl alcohol. Paschow5 studied the conductivities in methyl alcohol of potassium iodide, cadmium iodide, calcium nitrate, and potassium and sodium acetates. Vollmer6 worked out the conductivities of potassium and sodium iodides, potassium and sodium acetates, and lithium chloride in methyl alcohol, over a considerable range of dilution. Holland7 studied the effect of non-electrolytes on the conductivity in methyl alcohol of potassium, sodium, calcium, lithium, and ammonium nitrates, and sodium chloride. Carrara8 carried out by far the most extensive investigation which has yet been made of salts in methyl alcohol. He measured the conductivities of the following substances at various dilutions : Potassium chloride, bromide, iodide, methylate; sodium chloride, iodide, methylate, acetate; lithium chloride; ammonium chloride, bromide, iodide, fluoride; tetraethylam- monium chloride, bromide, iodide; tetramethylammonium iodide ; triethyl- amine, diisopropylamine, and a number of sulphur derivatives. Kerler,9 working in Beckmann's laboratory, determined the conductivities of lithium and calcium chlorides; lithium, sodium, and barium bromides; potassium iodide, ammonium nitrate, and potassium acetate. The conductivity of 1 Journ. Phys. Chem., 3, 12 (1899). 6 Wied. Ann., 62, 328 (1894). 2 Ztschr. phys. Chem., 4, 429 (1889). 7 Ibid., 50, 263. 3 Phil. Mag-, 24, 378 (1887). 8 Gazz. Chim. Ital., 26, [1] 119 (1896). 4 Wied. Ann., 33, 58 (1888) ; 43, 838 (1891). 8 Dissertation Erlangen (1894). 8 Charkow, 1892. 8 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. mercuric iodide in methyl alcohol was measured by Cattaneo;1 Schall 2 determined the conductivity of hydrochloric, picric, oxalic, and dichloracetic acids in methyl alcohol; Kablukoff 3 also studied the conductivity of hydro- chloric acid in methyl alcohol, and Kahlenberg and Lincoln4 measured the conductivity of ferric chloride and antimony trichloride in this solvent. The most satisfactory work on the whole that has ever been done on the conductivity of solutions in methyl alcohol is that of Zelinsky and Krapiwin.5 Their work included a number of salts in pure methyl alcohol, as well as in a mixture of this solvent and water, as we shall see later. They used in their work potassium bromide and iodide, ammonium bromide and iodide, cadmium iodide, tetramethylammonium bromide and iodide, tetraethylammonium iodide, a number of the substituted amines and sulphines, diethyl- and tri- ethyl-stannic iodides, "fumaroid" dimethylsuccinic acid, oxalic acid, iodic acid, and trichloracetic acid. Jones 6 has applied his boiling-point apparatus to the determination of the dissociation of salts in methyl alcohol. The salts used were: Potassium, sodium, and ammonium bromides; potassium, sodium, and ammonium iodides ; potassium and sodium acetates, and calcium nitrate. The dissocia- tion in methyl alcohol, as found by the boiling-point method, is about two- thirds of that in water under the same conditions. ETHYL ALCOHOL. A considerable amount of work has also been done in ethyl alcohol. The conductivity of the following substances has been determined by Fitz- patrick : 7 Calcium chloride, calcium nitrate, lithium chloride, lithium nitrate, mercuric, magnesium, and ferric chlorides. Hartwig 8 has determined the conductivity of formic, acetic, and butyric acids in alcohol. Vicentini 8 worked on the chlorides of ammonium, lithium, magnesium, calcium, cad- mium, zinc, and copper. Cattaneo10 has studied the conductivities of ferrous, ferric, and mercuric chlorides, and cadmium bromide and iodide. He found that these substances have a negative temperature coefficient of con- ductivity. Vollmer used a larger number of salts in ethyl alcohol than he did in water." These were potassium and sodium iodides, potassium and sodium acetates, sodium, lithium, and calcium chlorides, and calcium and silver nitrates. Kawalki,12 by a comparison of the rates of diffusion of a series of salts in water and in ethyl alcohol, showed that the rates of diffusion 1 Rend. R. Ace. Line. Roma (1895). 8 Wicd. Ann., 33, 58 (1888); 43, 838 2Ztschr. phys. Chcm., 14, 701 (1894). (1891). 3 Ibid., 4, 429 (1889). • Biebl. Wied. Ann., 9, 131 (1885). 4 Journ. Phys. Chem., 3, 26 (1899). 10 Ibid., 18, 219, 365 (1894). 6Ztschr. phys. Chem., 21, 35 (1896). ll Wied. Ann., 62, 328 (1894). 8 Ibid., 31, 114 (1899). 12 Ibid., 62, 324 (1894). 7 Phil. Mag., 24, 378 (1887). ORGANIC SOLVENTS. 9 in the two solvents bear the same relation to one another as the maximum molecular conductivities in the two solvents. Reference should be made to the work of Paschkow,1 who measured the conductivities in ethyl alcohol, of potassium and cadmium iodides, potassium and sodium acetates, and calcium nitrate; of Schall,2 who used picric, oxalic, and dichloracetic acids; of Wildermann,3 who studied the conductivities of di- and tri-chloracetic acids; of Kahlenberg and Lincoln,4 who worked with ferric chloride and antimony trichloride in ethyl alcohol; and of Kablukoff,5 who has measured the conductivity of hydrochloric acid in ethyl alcohol. Jones6 measured the dissociation of a number of salts in ethyl alcohol, using the boiling-point method. These include potassium and sodium iodides, sodium and ammonium bromides, potassium and sodium acetates, and calcium nitrate. These salts were found to be dissociated by ethyl alcohol, to from one-third to one-fourth the extent that they are dissociated in water at the same dilution. It should be observed, however, that dissociation as measured by the boiling-point method would not seem to be directly comparable with dissociation as measured by conductivity, since the two sets of measurements are made at different temperatures. It has, however, been established by Jones and Douglas,7 and later confirmed by Noyes and Coolidge 8 and Jones and West,9 that the temperature coefficient of dissociation is, in aqueous solu- tions, small ; in which case, if this holds for alcoholic solutions, there should be only a very small difference between the results obtained by the two methods. HIGHER ALCOHOLS. Comparatively little work has been done on the dissociating power of the higher alcohols. Schlamp10 has shown, from the results of his measurements on solutions of lithium and calcium chlorides, sodium iodide, and lithium salicylate, that their conductivity in propyl alcohol is somewhat less than one-half that in ethyl alcohol. In propyl and amyl alcohols Carrara " has made a few measurements, while Hartwig12 determined the conductivity of formic, acetic, and butyric acids in amyl alcohols. Among the isoalcohols Carrara 13 worked with isopropyl, and Kablukoff 14 with isobutyl and isoamyl alcohols, obtaining the remarkable result that in isoamyl alcohol solution the molecular conductivity of hydrochloric acid decreases with increase in dilution. Schall 15 has deter- mined the conductivity of picric acid in isobutyl alcohol. 1 Dissertation, Charkow, 1892. • Amer. Chem. Journ., 34, 357 (1905). 2 Ztschr. phys. Chem., 14, 701 (1894). 10 Ztschr. phys. Chem., 14, 272 (1894). 3 Ibid., 14, 267 (1894). "Gazz. Chim. Ital., 27, I, 221 (1897). 4 Journ. Phys. Chem., 3, 26 (1899). l2 Wied. Ann., 33, 48 (1888); 43, 838 6 Ztschr. phys. Chem., 4, 429 (1889). (1891). 6 Ibid., 31, 133 (1899). 13 Gazz. Chim. Ital., 27, I, 221 (1897). 7 Amer. Chem. Journ., 26, 428 (1901). " Ztschr. phys. Chem., 4, 432 (1889). 8 Ztschr. phys. Chem., 46, 323 (1903). "Ibid., 14, 707 (1894). 10 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. ETHER. Practically the only work on ethereal solutions is that of Cattaneo 1 and Kablukoff.2 Cattaneo measured the conductivity of ethereal solutions of cadmium iodide, cadmium bromide, ferrous and ferric chlorides, aluminium mercurous and stannous chlorides, salicylic and hydrochloric acids. He found that ethereal solutions have a negative temperature coefficient of conductivity, and that the molecular conductivity of hydrochloric acid in ether decreased with increase in dilution. This is analogous to the results of Kablukoff in the case of isoamyl alcohol. Indeed, Kablukoff found also that the conductivity of hydrochloric acid decreased with the dilution. KETONES. The conductivity of a number of salts of the alkalies in acetone were published by Cattaneo 3 several years ago. About the same time a paper appeared from St. v. Lasczynski 4 on the conductivity of some salts in acetone. Among these were included lithium and mercuric chlorides, potassium iodide, silver nitrate, and potassium, sodium, and ammonium sulphocyanates. The conductivity of solutions in acetone has also been measured by Carrara.5 Kahlenberg and Lincoln 6 measured the conductivity of solutions of ferric, cupric, and stannous chlorides, and antimony trichloride in acetone; and Dutoit and Aston,7 as well as Dutoit and Friderich,8 studied a number of solutions in acetone and other ketones. Dutoit and Aston pointed out, as has been mentioned, that those solvents which dissociate to the greatest extent are polymerized, as shown by the surface-tension method of Ramsay and Shields.9 In addition to acetone they worked with methylethyl ketone and methyl- propyl ketone. In the former solvents they used mercuric chloride, cadmium iodide, ammonium sulphocyanate, and sodium salicylate; in the latter, cadmium iodide, ammonium sulphocyanate, and sodium salicylate. They found that in the methylethyl ketone the conductivities were larger than in the methylpropyl ketone, but that the conductivity in acetone was the greatest of the three. Dutoit and Aston conclude from their work, to- gether with that of Kablukoff,10 that there is a general relation between the polymerization of the molecules of a solvent and its dissociating power. 1 Atti R. Ace. del le Scienze, Torino, 28, 5 Gazz. Chim. Ital., 27, I, 207 (1897). 329. Rend. R. Accad. del Lincei, [5] 8 Journ. Phys. Chem., 3, 27 (1899). 2, 295. 7Compt. rend., 125, 240 (1897). 2Ztschr. phys. Chem., 4, 431 (1889). 8 Bull. Soc. Chim., [3] 19, 321 (1897). 8 Rend. R. Accad. dei Lincei, [51 4°, 2 8 Ztschr. phys. Chem., 12, 423 (1893) sem., 63-75. I0 Ibid., 19, 251 (1896). 4 Ztschr. Elektrochem., 2, 55 (1895). ORGANIC SOLVENTS. 11 In connection with their work in acetone they make the following remark- able statement: We have found by the boiling-point method that the following salts in acetone have normal molecular weights : Cadmium iodide, lithium chloride, sodium iodide, mercuric chloride, and ammonium sulphocyanate. And that these substances in acetone conduct the current. Results l obtained in the physical-chemical laboratory of the Johns Hopkins University indicate that this statement is erroneous. ACIDS. The dissociating power of formic acid has been quite elaborately investi- gated by Zanninovich-Tessarin.2 In his work he used mainly the freezing- point method, but also studied the conductivity of a few salts in this solvent. He measured the freezing-point lowering of formic acid produced by the following substances, at dilutions varying from 0.34 to 3.4 normal (and in some cases at even greater concentration) : Potassium, sodium, ammonium, and lithium chlorides; potassium, sodium, and ammonium bromides; hydro- chloric, acetic, and trichloracetic acids. Formic acid is one of the strongest dissociating solvents next to hydrocyanic acid and water. The behavior of hydrochloric acid in this solvent is very remarkable. Not only does it show no dissociation, but the molecules are actually polymerized. Although, as just mentioned, the freezing-point lowering showed no dissociation, the conductivity in this solvent was very considerable. This may be due to the fact that while a majority of the molecules were polymerized, some were disso- ciated into ions which conducted the current. The conductivities of potassium and sodium chlorides in this solvent were also found to be very large. Zanninovich-Tessarin 3 has also determined the freezing-point lowering pro- duced by sodium bromide and lithium chloride in acetic acid. The former gave normal values, indicating no dissociation; while the latter showed marked polymerization. The conductivity of sulphuric acid in acetic acid has been measured by Jones,4 who found that the molecular conductivity, which was small at all dilutions, increased with the dilution to a certain point, and then decreased with further increase in the dilution of the solution. This is somewhat analogous to the result obtained by Kablukoff 5 for hydrochloric acid in ether and in isoamyl alcohol. THE NITRILES AND CYANOGEN. Dutoit and Aston 8 determined the conductivities of mercuric chloride, sodium bromide, cadmium bromide and iodide, ammonium sulphocyanate, and silver nitrate, in propionitrile. The investigation was extended by 1 Amer. Chem. Journ., 27, 16 (1902). * Amer. Chem. Journ., 16, 13 (1894). 2 Ztschr. phys. Chem., 19, 251 (1896). 6 Loc. cit. 3 Ibid., 19, 255 (1896). • Compt. rend., 125, 240 (1897). 12 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. Dutoit and Friderich l to solutions in acetonitrile and butyronitrile. It was shown that the dissociating power is greater in the first members of the nitriles, but in no case do they at all approach the dissociating power of liquid hydro- cyanic acid as determined by the recent work of Centnerszwer.2 Centnerszwer 3 has also shown that liquid cyanogen is a non-dissociant. PYRIDINE. Werner * found that certain inorganic salts, when dissolved in pyridine, conduct the current very well, but show very little or no dissociation by the boiling-point method. It is, however, to the work of St. v. Lasczynski and St. v. Gorski 5 that we owe what knowledge we have of the dissociating power of pyridine. They measured the conductivity of lithium chloride, potassium, sodium, and ammonium iodides, and potassium, sodium, and ammonium sulphocyanates in pyridine, over a wide range of dilutions. OTHER ORGANIC SOLVENTS. Such a few measurements have been made in other organic solvents that they can be passed over with brief reference. Thus, Werner6 found that cuprous chloride in ethyl sulphide conducts very poorly. Cattaneo 7 studied a few solutions in glycerol, and found that they had a larger conductivity than the corresponding solutions in ether. They also had larger temperature coefficients of conductivity. Dutoit and Aston 8 measured the conductivities of electrolytes in benzene chloride, ethyl bromide, and amyl acetate, and found that these solutions conduct very poorly. They found, on the other hand, that solu- tions in nitroethane conduct very well. Dutoit and Friderich 9 worked with acetophenone as a solvent, and with cadmium iodide, mercuric chloride, and ammonium sulphocyanate as electrolytes. The conductivity in this solvent was very small. Four other solvents have thus far been employed; namely, ethyl acetate, benzaldehyde, ethyl acetoacetate, and nitrobenzene. This work was done by Kahlenberg and Lincoln.10 As electrolytes they used ferric and stannous chlorides, bismuth trichloride, and antimony trichloride. The conductivi- ties in these solvents are in general small, but vary considerably with the nature of the electrolyte used. The most recent work in organic solvents is that of Walden.11 Five large pieces of work, yielding important and interesting results, have recently been published. A large number of types of organic compounds have been 1 Bull. Soc. Chim., [3] 19, 321 (1898). 8 Compt. rend., 125, 240 (1897). 2Ztschr. phys. Chera., 39, 217 (1902). "Bull. Soc. Chim., [3] 19, 325 (1898). 3 Loc. cit. 10 Journ. Phys. Chem., 3, 12 (1899). 'Ztschr. anorg. Chem. ,15, 1, 39 (1897). ll Ztschr. phys. Chem., 46, 103 (1903); 54, 'Ztschr. Elektrochem.,4,290(1897). 129 (1906); 55, 207 (1906); 55, 281 « Ztschr. anorg. Chem. ,15, 1, 139(1897). (1906); 55, 682 (1906); 58, 479 (1907); 7 Beibl. Wied. Ann., 17, 365 (1893). 59, 192 (1907). MIXED SOLVENTS. 13 brought within the scope of these investigations. These include alcohols, aldehydes, acids, acid anhydrides, esters, acid amides and amines, nitriles, sulphocyanates, mustard oils, nitro-compounds, nitrosodimethylene, ethald- oxime, epichlorhydrine, and ketones. The work has had to do with the conductivity of electrolytes in these solvents, with the relation between conductivity and internal friction, with boiling-point determinations, and with the solvent power of these dif- ferent substances. For details the original papers must be consulted. MIXED SOLVENTS. HYDROGEN DIOXIDE AND WATER. The dielectric constant of a mixture of hydrogen dioxide and water is greater than that of pure water. This has been shown by Calvert,1 and would lead one to suspect that electrolytes dissolved in such mixtures would have a greater conductivity than in pure water, in accordance with the Thomson- Nernst rule. The dissociating power of such mixtures has, however, not yet been determined. Reference should also be made to the later work of Cal- vert,2 showing that hydrogen dioxide has acid properties, and to the work of Jones, Barnes, and Hyde 3 along the same line. MIXTURES OF WATER AND THE ALCOHOLS. Only brief mention need be made of the work of Lenz,4 Kerler,5 Stephan,6 Kablukoff,7 Carrara,8 Schall,9 and Arrhenius.10 Wakeman,11 in quite an elab- orate investigation, measured the conductivity of organic acids in mixtures of ethyl alcohol and water in varying proportions. The results show that the conductivity becomes gradually smaller as the amount of alcohol be- comes larger and larger. This is just what would be expected from the relative conductivities in these two solvents. Zelinsky and Krapiwin 12 have, however, obtained results of a very different character. They found that the salts with which they worked, when dissolved in a mixture of methyl alcohol and water containing 50 per cent methyl alcohol, gave a conductivity considerably less than the conductivity in either alcohol or water. Similar results were obtained by Cohen 13 with ethyl alcohol and water, but only when the mixture contained very little water, and at dilutions which were quite large, as is shown by table 2. 1 Ann. der Phys., 1, 483 (1900). 7 Ztschr. phys. Chem., 4, 432 (1889). 2 Ztschr. phys. Chem., 38, 513 (1901). 8 Gazz. Chim. Ital., 16, 1 (1886). 3 Amer. Chem. Journ., 27, 22 (1902). 9 Ztschr. phys. Chem., 14, 701 (1894). 4 Mem. de 1'Acad. de St. Petersbourg, 10 Ibid., 9, 487 (1892). [7] 30, 1881. u Ibid., 11, 49 (1893). 5 Dissertation Erlangen, 1884. 12 Ibid., 21, 35 (1896). 8 Wied. Ann., 17, 673 (1882). 13 Ibid., 26, 31 (1898). 14 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 2. — Potassium iodide. V Pure alcohol. M,18° 80 p. ct. alcohol. M*18° V Pure alcohol. M«18° 80 p. ct. alcohol. M8° 64 128 256 26.1 29.2 31.8 30.9 32.2 33.2 512 1024 2048 34.4 36.0 36.3 34.1 34.5 35.0 From an examination of the above results it is seen that the conductivities in the mixtures of water and alcohol are the greater until a dilution of 512 liters is reached. At higher dilutions the conductivity in the pure alcohol becomes greater than that of the alcohol containing 20 per cent of water. In general, however, Cohen found that addition of water increased the con- ductivity, as we should expect. Lenz l measured the conductivities of various salts (potassium iodide, bromide, and chloride, sodium chloride, etc.) in mixtures of methyl and ethyl alcohols and water. He found that in certain cases the relative re- sistances can be obtained from the equation r = 100 where 100 is taken as the resistance of an aqueous solution of the same per cent, v is the volume per cent of alcohol, and b a constant. The formula holds best for the mixtures of methyl alcohol and water. Stephan 2 studied the conductivities of dilute solutions of sodium, potas- sium and lithium chlorides, and sodium and potassium iodides in mixtures of ethyl alcohol and water. Kablukoff3 determined the conductivity of hydrochloric acid in ethyl alcohol containing varying amounts of water. Arrhenius 4 investigated the changes in the conductivity of aqueous solu- tions, resulting from the addition to them of small quantities (less than 10 per cent by volume) of non-electrolytes, such as methyl or ethyl alcohol, cane-sugar, acetone, etc. He found that the changes could be expressed by the empirical formula — where I is the conductivity in water, 10 that in the mixture, x the volume per cent of added non-electrolyte, and a a constant peculiar to each non- electrolyte. Where two non-electrolytes were added a similar empirical formula was found to hold. The coefficient « differs not only for different 1 Mem. de 1'Acad. de St. P6tersbourg, 7, 30 (1881). 2 Wied. Ann., 17, 673 (1882). 3 Ztschr. phys. Chem., 4, 432 (1889). 4 Ibid., 9, 487 (1892). MIXED SOLVENTS. 15 non-electrolytes and different electrolytes, but varies also with concentration, and is greatest when dissociation is least. Arrhenius concludes that the amount of dissociation is not appreciably changed by addition of small quantities of non-electrolytes. This follows from the fact that the alteration in conductivity is independent of the con- centration. Further, he found that the velocity of inversion of cane-sugar is not appreciably influenced by addition of small amounts of non-electrolytes. Holland l worked in the same field as did Arrhenius. His results will be referred to again. Strindberg 2 repeated and confirmed some of Arrhenius's work. Wakeman 3 measured the conductivities of various electrolytes, sodium and potassium chlorides, hydrochloric acid, and numerous organic acids, in mixtures of ethyl alcohol and water (containing 10, 20, 30, 40, and 50 per cent of alcohol). For the cases studied, the equation — A = constant p(100 - p) was found to hold, where A is the difference between the conductivity of the electrolyte in water and in the mixture, respectively, and p is the per cent of alcohol by volume. Schall 4 determined the conductivity of picric acid in aqueous alcohol. Zelinsky and Krapiwin 5 studied the conductivities of sodium and am- monium iodides and bromides in water, methyl alcohol, and a mixture of the two containing 50 per cent of water by weight; for dilutions from v = 16 to v = 1024. Here a striking phenomenon was observed. The conductivities in the 50 per cent mixture were found to be decidedly less than the corre- sponding conductivities in the pure solvents. This minimum is best seen when the results are plotted as curves, with the conductivities as ordinates and the composition of the mixture as abscissae. Cohen 6 observed the minimum in the case of potassium iodide. He made a study of the conductivity of potassium iodide in mixtures of ethyl alcohol and water (containing 20, 40, 60, 80, and 99 per cent alcohol). The dilutions ranged from v = 64 to v = 2048. The minimum manifested itself in the 80 per cent mixture beyond the concentration v = 512. From his own observations, and from those of Wakeman (loc. cit.), he con- cludes that the relation w"n2?i = Constant fj.vH2O . Ale. 1 Wied. Ann., 50, 261 (1893). 4 Ibid., 14, 701 (1894). 2Ztschr. phys. Chem., 14, 161 (1894). * Ibid., 21, 35 (1896). 3 Ibid., 11, 49 (1893). • Ibid., 25, 31 (1898). 16 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. holds, being independent of both temperature and concentration; that is, the conductivities compared are approaching a limiting value at the same rate, and either the dissociation is the same in the cases compared, or for mixtures of alcohol and water conductivity is not a direct measure of disso- ciation. Cohen is inclined to the latter view. Walker and Hambly 1 studied the conductivity of diethylammonium hydrochloride in mixtures of water and ethyl alcohol. Hantzsch 2 made some interesting applications of results obtained by studying conductivities in various mixtures. Tijmstra 3 investigated the conductivities of solutions obtained by the action of mixtures of methyl or ethyl alcohol and water on sodium. In the case of the mixtures of methyl alcohol and water the minimum was observed. Roth 4 made a careful study of the conductivity of potassium chloride in mixtures of ethyl alcohol and water containing 8 and 20 per cent alcohol by weight. He found that the relation given by Wakeman (loc. cit.) holds, while that given by Cohen (loc. cit.) does not. The quotient — /vH2O.Alc. was found to decrease with increasing dilution, and with increase in the amount of alcohol in the mixture. This, Roth thinks, may indicate a decrease in dissociation. The relation given by Arrhenius (loc. cit.) was also found to be valid. The work of Wolf 5 and of Rudorf 6 needs no special consideration. CONDUCTIVITY AND VISCOSITY. That viscosity and conductivity are related is by no means a new idea. As early as 1856, G. Wiedemann 7 studied aqueous solutions of copper sulphate, and concluded that the conductivity of a solution is directly proportional to the concentration, and inversely proportional to the viscosity. When for- mulated this would be •„• Kr] — = constant P where K is the conductivity of the solution of concentration p, and t] the viscosity. Grotrian,8 in 1876, measured the conductivity and viscosity of solutions at different temperatures, but obtained indecisive results. 1 Journ. Chem. Soc., 71, 61 (1899). 4 Ztschr. phys. Chem., 42, 209 (1903). 2 Ztschr. anorg. Chem., 25, 332 (1900). B Ibid., 40, 222 (1902). Ber. d. chem. Gesell., 35, 1001 (1902). • Ibid., 43, 257 (1903). 'Proc. Kon. Akad. te Amsterdam, 7 Pogg. Ann., 99, 229 (1856). 1903, p. 104. 8 Ibid., 157, 130 (1876). CONDUCTIVITY AND VISCOSITY. 17 Stephan,1 in 1883, tried a third possibility, by using mixtures of alcohol and water as a solvent. He found that the temperature coefficients of con- ductivity and of fluidity (the reciprocal of viscosity) closely resembled each other. He observed a minimum in his curves and proposed the formula : KH - — = constant, to hold up to the minimum point, Krj and —. — = constant, to hold from that point on ; WKrj where K is the conductivity of the equivalent aqueous solution, k the con- ductivity of the mixture, and H and 17 the corresponding viscosities ; w and w' are the per cents of water in the given aqueous mixture and in the aqueous alcoholic mixture of minimal fluidity, respectively. He believed that each ion carries with it neighboring molecules of the solvent, and that ionic fric- tion results from the friction between these and the rest of the solvent. Dutoit and Friderich2 introduced the association factor and concluded that — The values of MD for a given electrolyte dissolved in different solvents, are a direct function of the degree of polymerization of the solvent and an indirect function of the coefficient of viscosity of these solvents. A fourth method of changing the fluidity was resorted to by Rontgen,3 and later by Warburg and Sach,4 and more exhaustively by Cohen.5 They subjected the aqueous solution to high pressure. Cohen found that, at low temperatures, the viscosity is decreased by the pressure, but that above 40° the viscosity increases with the pressure. In concentrated solutions of sodium and ammonium chlorides the viscosity increases nearly proportional to the pressure, and nearly independent of the temperature. Hauser 6 showed that, at 32°, the pressure ceased to affect the fluidity of water. Grossman,7 in 1883, recalculated Grotrian's results, and found that the conductivity multiplied by the viscosity gave a constant independent of the temperature, and that the temperature coefficients were the same to within 1 per cent. Arrhenius 8 worked with aqueous solutions to which small amounts of non- electrolytes, such as acetone and methyl and ethyl alcohols, had been added. He pointed out an empirical relation between the conductivity and the fluidity ; but he saw that these quantities are not simply dependent on each other, since the conductivities of dilute solutions of different salts are not the same. This empirical relation was further developed by Euler.9 'Wied. Ann., 17, 673 (1883). 'Ann. d. Phys., 6, 597 (1901). 2 Bull. Soc. Chim., [3], 19, 321 (1898). 7 Wied. Ann., 18, 119 (1883). 8 Wied. Ann., 22, 510 (1884). 8 Ztschr. phys. Chem., 9, 487 (1892); 4 Ibid., 22, 514 (1884). 1, 285 (1887). 'Ibid., 46, 666 (1892). • Ibid., 36, 536 (1898). 18 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. We need only mention in this connection the work of Strindberg * and of Holland.2 Voilmer 3 investigated solutions of various salts in methyl and ethyl alcohols. He found the temperature coefficients of conductivity and fluidity to be very nearly identical. Kohlrausch and Deguisne 4 used the formula Kt = Ku[l + a(t - 18°) + 0(t- IS0)2] to represent the influence of temperature on conductivity, starting from 18° as a mean temperature. Kohlrausch 5 noted that on extrapolating this curve, aqueous solutions would reach a zero value of conductivity at about — 39°, which is about the temperature where the fluidity would become zero. Bousfield and Lowry 6 showed that the viscosity of water may be represented accurately by a formula similar to the above, 1713 = 17, [! + «(«- 18) +£(«- IS)2] They found that the constants a and @ are the same in the two formulas, to within the limits of experimental error. They believe, however, that these formulas will not hold at low temperatures, and that the zero values can not be experimentally realized. This belief is borne out by the work of Kunz.7 In an exceedingly interesting paper, Kohlrausch 8 proposes the view that — About every ion there moves an atmosphere of the solvent, whose dimensions are determined by the individual characteristics of the ion. . . . The electrolytic resist- ance is a frictional resistance that increases with the dimensions of the atmosphere. The direct action between the ion and the outer portion of the solvent diminishes as the atmosphere becomes of greater thickness. . . . For a very sluggish ion there will be only the friction of water against water, and the electrolytic resistance will have the same temperature coefficient as the viscosity of water, provided the atmosphere itself does not change its dimensions with the temperature. If, however, the atmosphere becomes, for example, smaller with increasing temperature, the temperature gradient of the conductivity might be greater than that of the fluidity. According to observations now at hand, this would seem to be the case for the slowest moving univalent ion, Li. Bousfield and Lowry 9 have gone farther and have shown that we should also expect to find an upper limit of conductivity, on account of the decrease in dissociation with rise in temperature. A maximum conductivity of this sort has been observed by Franklin and Kraus10 in liquid ammonia. Potas- sium iodide gives a maximum in conductivity, in methyl alcohol, at 160°.u 1 Ztschr. phys. Chem., 14, 221 (1894). 7 Compt. rend., 135, 788 (1902). 2 Wied. Ann., 50, 261 (1892). 8 Proc. Roy. Soc., 71, 338 (1903). 3 Ibid., 52, 328 (1894). 9 Loc. cit. 4 Dissertation Strassburg, 1893. 10 Amer. Chem. Journ., 24, 83 (1900). 6Sitz. Berlin. (1901), 1028. » Phys. Rev., 18, 40 (1904). 6 Proc. Roy. Soc., 71, 42 (1902). VISCOSITY. 19 Noyes * observed a maximum conductivity with N/10 potassium and sodium chlorides, in water, at 280°. The formula of Slotte 2 for variation of fluidity holds at low temperatures, so that combining this formula with that of Abegg and Seitz for decrease in dielectric constant, - «-« ~ Bousfield and Lowry 3 give, as the complete formula representing the effect of temperature on conductivity, Reference should also be made to the work of Hechler.4 VISCOSITY. The majority of workers have confined themselves either to viscosity determinations alone, or to conductivity determinations alone. We must, therefore, consider some of these if we wish to see clearly the relations be- tween the phenomena. We need simply mention here the work of Poiseuille,5 Noack,6 Pagliani and Battelli,7 Slotte,8 Gartenmeister,9 and Traube.10 The monumental work of Thorpe and Rodger u merits more careful attention. They worked with very great accuracy both with pure liquids and with mixtures,12 and over a con- siderable range of temperature. They have shown that the formula of Slotte gives the best results. They proved, conclusively, what had been hinted at before, that — Viscosity may be taken as the sum of the attractive forces in play between the mo'.e- cules ; . . . that an increment of CH2 in chemical composition, or the substitution of an atom of Cl, Br, or I for an atom of hydrogen, brings about a definite change in the magnitude of the viscosity. It is, therefore, made evident that viscosity or intermolec- ular attraction is, in reality, a property of the atoms of which the molecules are com- posed. Isomers have nearly but not the same viscosity, yet the effect of CH-2 is the same as in the normal compounds. The effects due to ring grouping, iso- and double-linkages, and changes in the condition of the oxygen may be quantitatively allowed for. . . . But water and the alcohols show no agreement with the calculated values. 'Ztschr. phys. Chem., 46, 323 (1903). 7 Atti di R. Ac. delle Sc. d. Torino, 20, Journ.Amer.Chem.Soc., 26, 134(1903). 607 (1885). 2Beibl.. 16, 182 (1892). « Loc. cit. 3 Proc. Roy. Soc., 74, 280 (1904). • Ztschr. phys. Chem., 6, 524 (1890). * Dissertation Minister (1904). 10 Ber. d. chem. Gesell., 19, 871 (1886). 6 Mem. Inst. Paris, 9, 433 (1896). "Phil. Trans., 185A, 307 (1894). eWied. Ann., 27, 289 (1886). n Journ. Chem. Soc., 51, 360 (1897). 20 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. Hydrogen is calculated to have an effect on molecular viscosity of 44.5, carbon of 31, and iso-linkage, for example, of —21. The effect is shown in table 3. TABLE 3. — Molecular viscosity at slope 0.0000323. Normal compound. Iso-compound . Found. Calculated. Found. Calculated. Pentane Hexane Heptane Octane 687 818 931 1035 689 809 929 1049 663 799 908 668 788 908 These investigators found both maxima and minima when working with mixtures, and in the particular case of chloroform and ether they found a point of inflection. They believe that the maximum value is caused by a "feeble chemical combination or molecular aggregation, which is destroyed by heat or dilution." It is to be expected that water and the alcohols would give abnormal results, since Ramsay and Shields l have shown that these liquids are asso- ciated. They are also abnormal in possessing a high dielectric constant.2 Some of the characteristic properties of water, methyl alcohol, ethyl alco- hol, and acetone are grouped together in table 4. TABLE 4. Mol. vol. Viscosity. Association. Dielectric constant. Water .... f 18.0 at 0° \ 18.1 25 0.01778 at 0° .00891 25 1.707 at 0° 1.644 20 79.46 at 0° 73.92 20 Methyl alcohol / 39.5 0 1 40.3 20 .0080846 0 .005530 25 2.65 - 90 2.32 20 34.05 Ethyl alcohol . . /57.1 0 I 58.3 20 .017761 0 .0108545 25 2.03 - 90 1.65 20 25.02 Acetone .... f 70.£ 0 I 73.2 20 .0039496 0 .0030726 25 1.26 17 to 78 21.85 No such quantitative relation has been worked out for changes in viscosity caused by salts brought into solution, and still less is known about the con- ductivity which a given salt may be expected to give. Yet Bredig,3 Wagner,4 and Euler 5 have worked on this problem with considerable success. Wagner found that the viscosity of a salt solution is an additive function of the metallic and non-metallic radicals of the dissolved salt. For allied metals the viscosity decreases as the atomic weight increases. The dissociated ions appear in 'Ztschr. phys. Chem., 12, 433 (1893); 16, 111 (1894). Ibid., 14, 286 (1894). 3 Ibid., 13, 243 (1894). 4 Ibid., 6, 31 (1890). • Ibid., 25, 536 (1898). VISCOSITY. 21 some cases to have greater, and in other cases less viscosity than the original solution. To explain this "negative viscosity" shown by certain salts in their power to lower the viscosity of pure water, Euler employed the " electrostriction theory" proposed by Drude and Nernst.1 According to this theory the ion is surrounded by a strong electrical field, in virtue of its charge, which causes a strong compression of the liquid in this field. Euler holds that the effect of a salt on viscosity is the result of two tendencies : First, that of the atoms tending to increase the viscosity in inverse proportion to their migration velocities; and second, the electrostriction tending to lessen the viscosity. Euler has calculated viscosity constants for a large number of ions, and finds the relation between them and the migration velocities to be expressed by the formula — (A - 0.68) . C7(or (K - 0.68) . F) = a constant, where A and K are the viscosity constants and U and V are the migration velocities of the anion and cation, respectively. Hydrogen and hydroxyl ions are exceptions. Some of the values are given in table 5. TABLE 5. Ion. Migration velocity. Viscosity constant. Li ... 398 1 15 Ca . . 62 0 1 023 K 706 0962 Br .... 730 0946 NO3 0919 Wagner 2 has shown that Mullenbein's 3 measurements have discounted Euler's explanation of negative viscosity, since the viscosity of the solvent may be lowered by the addition of certain non-electrolytes, even when the viscosity of the dissolved substance is higher than that of the solvent. With ethyl alcohol, o-nitrotoluene gives an inversion-point, m-nitrotoluene a mini- mum and p-nitrotoluene a maximum. He proposes, as an explanation, that the solute diminishes the quantity of the solvent in a given space, and this leads to a diminution of the viscosity, which diminution is partly compensated, however, by the solute itself. Ac- cording to the relative magnitude of the various factors, the viscosity may be increased or diminished. Dunstan 4 has investigated a large number of mixtures. He believes that the increase in viscosity is due to the formation of loosely held complexes. 1 Ztschr. phys. Chem., 16, 79 (1894). 4 Journ. Chem. Soc., 85, 817 (1904). Ztschr. 2 Ibid., 46, 867 (1903). phys. Chem., 49, 590 (1904). 3 Dissertation, Leipzig, 1901. 22 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. He thinks that the cause of the minimum in viscosity is more deep-seated than Wagner supposed, and attributes it to "some change in molecular aggre- gation or dissociation." Blanchard l found that the addition of 1 equivalent of ammonia for 1 equivalent of silver, and the addition of 4 equivalents of ammonia for 1 equivalent of copper and zinc in aqueous solutions of their salts, very greatly decreases the viscosity. He reasons that this can not be due to increased ionization, and therefore rejects the electrostriction theory and proposes a hydrate theory. He says that if the positive ion consists solely of a metallic atom bearing an electric charge, combination with ammonia molecules can not decrease its mass or readily increase its sym- metry, so as to reduce the viscosity. The only explanation seems to be that the ions in solution are hydrated. The hydrate water is replaced by ammonia, which forms with the ion a more stable complex and one of smaller mass, or greater symmetry, or both. He believes that this theory also accounts for negative viscosity and for the effect of pressure on viscos- ity. Evidently, this is the same conclusion as that reached by Kohlrausch in his hypothesis of ionic spheres, but by a somewhat different method of approach. Blanchard added small amounts of water to alcoholic solutions of sodium hydroxide, and found the viscosity smaller than would be expected from a study of the pure solvents. This is due, as he thinks, to the formation of a complex between the alkali, water, and alcohol, which is, however, smaller or more symmetrical than the alcohol-water complex originally present. Mixtures of alcohol and water give a maximum in viscosity. Blanchard finds that cupric chloride increases this effect. He further applies this theory to the work of Jones and Lindsay,2 on conductivity. The existence of hydrates, or solvates (in the case of non-aqueous solvents) in one form or another is an old conception. Poisseuille 3 first suggested it in working with alcohol and water. Graham * confirmed and extended Pois- seuille's work. Wijkander5 supposed that acetic acid forms a hydrate with water, C2H402 . H20, which would account for abnormal viscosity. The changes due to temperature he attributed to dissociation changes in the liquid. Thorpe and Rodger 6 and Traube a also assume the presence of hydrates. Recently, Varenne and Godefroy 7 have found evidence from viscosity curves for the existence of various hydrates in mixtures of water with methyl and ethyl alcohols and acetone. These are shown in table 6. 'Journ. Amer. Chem. Soc., 26, 1315 6 Wied. Beibl., 8, 3 (1879). (1904). "Loc. cit. 'Arner. Chem. Journ., 28, 329 (1902). 7Compt. rend., 137, 992 (1903); 138, 3 Loc. cit. 990 (1904). 4 Phil. Trans., 151, 373 (1861). VISCOSITY. 23 TABLE 6. Methyl alcohol and water. Ethyl alcohol and water. Acetone and water. CH3OH . H2O CH3OH . 2 H2O CH3OH . 3 H2O CHsOH . 5 H2O CHsOH . 8 H2O CH3OH . 20 H2O C2H6OH . 2 H2O C2H6OH . 3 H2O C2H5OH . 6 H2O 3 (C«H6OH) . 2 H2O C2H5OH . 22 H2O CH3COCH3 . 3 H2O CH3COCH3 . 4 H2O CH3COCH3.8H20 CHsCOCHs . 34 H2O This conception of the existence of hydrates differs from the view put forward by Jones,1 according to which the composition of the hydrates formed by any substance is a function of the concentration, the temperature remaining constant. The composition may vary all the way from one molecule of water to a large number, every intermediate step being rep- resented. Various lines of evidence have been furnished for this view by Jones,2 Jones and Getman,3 Jones and Bassett,4 and Jones and Uhler.5 The more important of these have to do with the relation between water of crystallization and lowering of the freezing-point, dissociation as measured by freezing-point and by conductivity, certain color changes in solution, and the relation between water of crystallization and temperature. 1 Amer. Chem. Journ., 23, 89 (1900). 2 Loc. cit. 3Amer. Chem. Journ., 31, 303 (1904). Ztschr. phys. Chem., 46, 244 (1903); 49, 385 (1904). 4 Amer. Chem. Journ., 33, 584 (1905). 6 Carnegie Institution of Washington Publication No. 60. Amer. Chem. Journ , 37, 126 (1907). WORK OF LINDSAY. EXPERIMENTAL. This work was undertaken as an extension of the older work of Zelinsky and Krapiwin,1 and Cohen,1 on the conductivity of electrolytes in mixtures of methyl and ethyl alcohols with water. Zelinsky and Krapiwin, in their work, have shown that solutions in a 50 per cent mixture of methyl alcohol and water have a much less conductivity than in the pure alcohol itself. They have also shown that the slightest addition of water to a solution of an electrolyte in absolute methyl alcohol produced a lowering of its conducting power. We have extended this work, by making conductivity measurements of solutions in which the solvents were mixtures of methyl alcohol and water of varying composition. By this means we have been able to plot curves showing, for each salt worked with, the mixture of methyl alcohol and water having the least dissociating power. We have also extended the investi- gation to ethyl alcohol, propyl alcohol, and to mixtures of ethyl alcohol and water, propyl alcohol and water, and methyl and ethyl alcohols. The work has, for the most part, been done both at 0° and 25°. In this way we have been able to calculate the temperature coefficients of conduc- tivity of the various salts in the different solvents and, what is of more im- portance, to show the influence of temperature on the minimum values mentioned above. The salts used are potassium iodide, strontium iodide, ammonium bromide, cadmium bromide, ferric chloride, and lithium nitrate. APPARATUS. In all this work the Kohlrausch method of measuring conductivity was employed. The bridge wire used was a meter in length and made of " man- ganin." The resistance coils were manufactured by Leeds & Co., of Phila- delphia, and were found to be accurate to 0.04 per cent. The cells are of the form shown in fig. 1, the difference between them and the ordinary Arrhenius cell being that they are provided with a ground-glass top to prevent evaporation of the more volatile solvents, and to protect the anhydrous alcohols from the moisture of the baths and air. In some cases the ground-glass joint was also covered with paraffin as an extra precaution. 1 Loc. cit. 24 SOLVENTS. 25 The glass tubes carrying the electrodes were shoved through thin rubber tubes, and then inserted into the glass tubes in the cap. Sealing wax was then run over the outside of the joint. The zero-bath was prepared as follows: A large glass battery-jar was filled with pure, finely crushed ice, and moistened with distilled water. This was placed in a water-bath, the space between the two being filled with finely crushed ice. By this means it was possible to keep a cell within 0.02° of zero for hours. The 25° bath was of the ordinary form and was stirred by means of a small hot-air engine. The thermometers used were graduated to 0.04° and were carefully calibrated. The burettes and flasks were also carefully calibrated. SOLVENTS. WATER. FIG. 1. All the water used in this work was purified as follows: Ordinary distilled water was first distilled from acidified potassium dichromate. This water was redistilled from potassium dichromate acidified with sulphuric acid, and then from barium hydroxide. The water purified in this way had a conductivity of never more than 2 x 10~6, and sometimes as low as 0.8 x 10 ~6. METHYL ALCOHOL. The methyl alcohol used was the best commercial article that could be obtained. It was first boiled with calcium oxide, then distilled and allowed to stand over anhydrous copper sulphate for weeks. Before use it was dis- tilled from the copper sulphate and then from sodium. None of the alcohol used in making up the solutions in absolute alcohol had been distilled from sodium more than twenty-four hours before use. It had a conductivity of about 2.3 x 10~6. ETHYL ALCOHOL. The ethyl alcohol was the best obtainable article, and was purified in the same manner as the methyl alcohol. Its conductivity had a mean value of 2xlO-7. PROPYL ALCOHOL. The propyl alcohol was Kahlbaum's best, and before use was distilled from anhydrous copper sulphate and sodium. It had a conductivity of 0.8 X 10~7" 26 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. SOLUTIONS. The method of making up the original mother-solutions will be given when the various electrolytes are considered. From this mother-solution the remaining solutions were made by successive dilutions by means of burettes and measuring flasks. In the cases where this would necessitate the use of small quantities of the solution, a second mother-solution was made, and from this successive dilutions were prepared. In preparing the solutions in the mixed solvents, a sufficient quantity of the solvent was made by mixing the constituents in the required proportions. This was then used in the same manner as a simple solvent. In preparing these mixed solvents the following method was employed : x c.c. of an alcohol were diluted to 100 c.c. In the following discussion such a solution would be designated as — alcohol x per cent. In making mixtures of methyl alcohol and ethyl alcohol, the methyl alcohol was measured and diluted with the ethyl alcohol, and the concentration expressed in terms of the methyl alcohol. CONDUCTIVITY MEASUREMENTS. In all determinations of conductivity from three to five different resistances were used, and the values given in the tables are the mean of these values. POTASSIUM IODIDE. The salt used in this work was recrystallized a number of times. It was then carefully dried and kept in a desiccator. All the mother-solutions were made by direct weighing. TABLE 7. — Molecular conductivity of potassium iodide. V In water. In methyl alcohol. In ethyl alcohol. JM«0° M»25° M«0° Mv25° /O>° /u25° 64 128 256 512 1024 74.09 76.4 D 77.01 78.0 D 77.96 132.1 135.4 138.0 139.6 140.7 59.32 63.88 67.73 69.85 71.23 82.87 88.49 93.73 98.36 102.0 19.12 21.36 22.66 25.00 27.43 29.40 33.02 36.02 38.63 41.35 V In methyl alcohol. In methyl alcohol (20 p. ct.) and water. In methyl alcohol (40 p. ct.) and water. In methyl alcohol (50 p. ct.) and water. V«25° 3/ut,25° M«0° Mv25° ° M,25° M,0° MB25° 64 128 256 512 1024 78.7 84.7 88.2 90.8 93.0 82.52 88.69 93.85 98.19 102.2 45.69 47.26 47.79 48.45 49.07 91.91 93.78 95.64 97.12 98.10 35.48 35.92 36.52 37.02 37.85 72.14 73.69 75.14 76.25 77.68 33.73 34.44 35.13 36.05 36.76 67.46 68.79 70.37 71.72 72.57 1 The values of the conductivity at 0° marked "D" were obtained by Jones and Douglas (Amer. Chem. Journ., 26, 428), while those at 25° are taken from the work of Ostwald. 3 Carrara. » Zelinsky and Krapiwin. POTASSIUM IODIDE. 27 TABLE 7. — Molecular conductivity of potassium iodide. — Continued. V In methyl alcohol (65 p. ct.) and water. In methyl alcohol (80 p. ct.) and water. In ethyl alcohol (50 p. ct.) and water. In methyl alcohol (50 p. ct.)and ethyl alcohol. ^0° M«25° fj*0° ^25° M,0° M,25° /M)° M»25° 64 128 256 512 1024 35.12 35.71 36.49 37.23 37.75 65.04 67.25 68.78 70.00 70.94 39.03 40.51 41.83 43.23 44.45 67.78 70.33 71.83 73.16 74.81 19.26 19.82 20.35 20.92 21.43 48.30 50.07 50.80 51.97 52.52 36.74 39.46 41.93 44.46 46.89 54.18 58.52 62.13 65.93 69.61 The conductivity of solutions of potassium iodide in methyl alcohol had already been determined by both Zelinsky and Krapiwin,1 and Carrara,2 but with such different results that the above measurements seemed neces- sary. Our measurements agree very well with those of Zelinsky and Krapiwin, as is seen from table 7. TABLE 8. — Temperature coefficients o/ conductivity of potassium iodide. V In water (0° to 25°) . In methyl alcohol (0°to25°). In ethyl alcohol (0°to25°). 64 2.26 0.942 0.411 128 2.30 0.984 .466 256 2.35 1.04 .534 512 2.40 1.14 .545 1024 2.52 1.23 .557 V In mixtures of methyl alcohol and water of various compositions. In a 50 p. ct. mixture of ethyl alcohol and water (0° to 25°). In a 50 p. ct. mix- ture of methyl and ethyl alco- hols (0°to25°). 20 p. ct. 40 p. ct. 50 p. ct. 65 p. ct. 80 p. ct. 64 1.83 1.47 1.35 1.17 1.15 1.16 0.698 128 1.86 1.51 1.37 1.26 1.17 1.25 .762 256 1.91 1.54 1.41 1.29 1.20 1.22 .808 512 1.95 1.57 1.43 1.31 1.20 1.24 .859 1024 1.96 1.59 1.43 1.32 1.21 1.24 .909 Some of the results given in table 7 are plotted in fig. 3. The curves are of the same general form as the preceding. The chief points of difference are : The minimum point has shifted to the right, corresponding now to an alcohol- water mixture of about 65 per cent. The increase to the right of the mini- mum is much less rapid than that to the left, the difference being due to the fact that the temperature coefficient of conductivity is much greater in water than in methyl alcohol. The alcohol-water mixture, having the same con- ducting power as the solution in pure methyl alcohol, has also changed. In this case it changes from an alcohol of about 19 per cent to an alcohol of about 30 per cent, depending upon the concentration of the solution. 1 Ztschr. phys. Chem., 21, 35 (1896). 2 Gazz. Chim. Ital., 26, (1) 119 (1896). 28 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 9. — Comparison of the molecular conductivity of potassium iodide. V In ethyl alcohol, water, and a 50 p. ct. mixture of these solvents. AtO°. At 25°. Water. Me Mixture. Me Ethyl alcohol. Me Water. Me Mixture. Me Ethyl alcohol. Me 64 128 256 512 1024 74.09 76.4 77.01 78.0 77.96 19.26 19.82 20.35 20.92 21.43 19.12 21.36 22.66 25.00 27.43 132.1 135.4 138.0 139.6 140.7 48.30 50.07 50.80 51.97 52.52 29.40 33.02 36.02 38.63 41.35 V In water, methyl alcohol, and mixtures of these solvents at 0°. 0 p. ct. 20 p. ct. 40 p. ct. 50 p. ct. 65 p. ct. 80 p. ct. 100 p. ct. 64 128 256 512 1024 74.09 76.4 D 77.01 78.0 D 77.96 47.26 47.79 48.45 49.07 35.48 35.92 36.52 37.02 37.85 33.73 35.12 34.44 35.71 35.13 36.49 36.05 37.23 36.76 37.75 39.03 40.51 41.83 43.23 44.45 59.32 63.88 67.73 69.85 71.23 V In water, methyl alcohol, and mixtures of these solvents at 25°. 0 p. ct. 20 p. ct. 40 p. ct. 50 p. ct. 65 p. ct. 80 p. ct. 100 p. ct. 64 128 256 512 1024 132.1 135.4 138.0 139.6 140.7 91.91 93.78 95.64 97.12 98.10 72.14 73.69 75.14 76.25 77.68 67.46 65.04 68.79 67.25 70.37 68.78 71.72 70.00 72.57 70.94 67.78 70.33 71.83 73.16 74.81 82.87 88.49 93.73 98.00 102.0 V In methyl alcohol, ethyl alcohol, and a 50 p. ct. mixture of these solvents. AtO°. At 25°. Methyl alcohol. V* Mixture. Me Ethyl alcohol. Mi> Methyl alcohol. Me Mixture. Me Ethyl alcohol. MB 64 128 256 512 1024 59.32 63.88 67.73 69.85 71.23 36.74 39.46 41.93 44.46 46.89 19.12 21.36 22.66 25.00 27.43 82.87 88.49 93.73 98.36 102.0 54.18 58.52 62.13 65.93 69.61 29.40 33.02 36.02 38.63 41.35 Cohen1 has shown that solutions of potassium iodide in a mixture of alcohol and water show a minimum in the conductivity, but only at great dilutions (512) and when the amount of water present is small. His work, however, was all done at 18° C. From the results in table 9 we see that at 0° we have a minimum in the conductivity values for an alcohol as dilute as 50 per cent, and in solutions which are comparatively strong, namely, fromv = 128. In all probability in alcohol of 75 to 80 per cent a much greater depression •Ztschr. phys. Chem., 25, 31 (1898). POTASSIUM IODIDE. 29 would be found. This is, however, a subject for future investigation. At 25° all trace of a minimum has disappeared. In order to see the connection existing between the conducting power of the solutions in the various solvents, the preceding table is given for the sake of comparison. ivlv Curve J, V =* 64 .II, v=r 128 " III, V => 256 " IV, V = Y,v«= Strength of Methyl Alcohol FIG. 2. — POTASSIUM IODIDE AT 0°. Some of the values in table 9 are plotted in fig. 2. It is seen that the values of the molecular conductivity reach a minimum in a mixture of methyl alcohol and water containing 50 per cent methyl alcohol. It is also seen that an addition of approximately 10 per cent of alcohol lowers the conductivity of the aqueous solutions to that of an alcoholic solution. 30 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. Table 9 makes it clear that in a mixture of methyl and ethyl alcohols, the conductivity of potassium iodide shows no minimum value when compared with the conductivity in the pure solvents. In fact, the conductivity values Curve I, v =• 64 . " II, v => 128 " III, V = 256 Concentration of Methyl Alcohol FIG. 3. — POTASSIUM IODIDE AT 25°. for the solutions in the mixed solvents approach the mean value of the con- ductivities in the pure solvents. Thus, at 0° the observed value of the conductivity in the mixture, for v = 64, is 36.74, while the mean of the conductivities at the same dilution in the pure solvents is 39.22. In all cases, however, the conductivity in the mixture lies below this mean value. AMMONIUM BROMIDE. The salt used in this work was carefully recrystallized, and on sublimation left no residue. It was thoroughly dried and kept in a desiccator. All the mother-solutions were made by direct weighing. AMMONIUM BROMIDE. 31 TABLE 10. — Molecular conductivity of ammonium bromide. V In water. In methyl alcohol. In ethyl alcohol. 1*0° M.25° M.0° M»253 /*.o° 64 128 256 512 1024 74.22 75.23 76.62 77.49 77.78 135.3 138.6 141.2 143.5 145.6 58.71 63.16 66.45 68.51 70.40 79.56 85.80 90.88 94.99 98.24 16.71 18.83 19.66 22.66 22.88 V In methyl alco- hol (50 p. ct.) and water. In ethyl alcohol (50 p.ct.) and water. In methyl alco- hol (50 p. ct.) and ethyl al- cohol. In methyl alcohol, water, and mixtures of these solvents atO°(i;=64). /*»o° M«0° MvO° Alcohol, p. ct. MvO° 64 128 256 512 1024 34.85 35.78 36.36 37.11 37.49 19.42 19.89 20.09 20.70 21.50 34.15 38.40 39.75 41.06 42.00 0 20 50 65 80 100 74.22 47.96 34.85 34.68 40.55 58.71 We see that in these results we have practically the same phenomenon as in the case of potassium iodide. In the case of ammonium bromide the minimum point in the conductivity values appears to be reached with an alcohol of 50 per cent. TABLE 11. — Temperature coefficients of conductivity of ammonium bromide. V In water (0° to 25°). In methyl alcohol (0° to 25°). 64 2.44 0.834 128 2.54 0.906 256 2.58 0.977 512 2.64 1.059 1024 2.71 1.114 TABLE 12. — Comparison of the molecular conductivity of ammonium bromide. V In water, methyl alcohol, and a 50 p. ct. mixture of these solvents at 0°. 1 In water, ethyl alcohol, and a 50 p. ct. mixture of these solvents at 0°. 2 In methyl alcohol, ethyl alcohol, and a 50 p. ct. mix- ture of these solvents at 0°. Water. MvO° Mixture. M*0° Methyl alcohol. MvO° Water. M»0° Mixture. A0>° Ethyl alcohol. A0>° Methyl alcohol. MvO° Mixture. M>° Ethyl alcohol. MvO° 64 128 256 512 1024 74.32 75.23 76.62 77.49 77.78 34.85 35.78 36.36 37.11 37.49 58.71 63.16 66.45 68.51 70.40 74.22 75.23 76.62 77.49 19.42 19.89 20.09 20.70 21.50 16.71 18.83 19.66 22.66 22.88 58.71 63.16 66.45 68.51 70.40 34.15 38.40 39.75 41.06 42.00 16.71 18.83 19.66 22.66 22.88 1 Here also a 50 per cent ethyl alcohol gives a minimum, but only in the case of the more dilute solutions, namely, for dilutions t> = 512 and 1024. 2 In this comparison there is no trace of a minimum, nor does there appear to be a probability of a minimum for any mixture of methyl and ethyl alcohols. 32 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. STRONTIUM IODIDE. The strontium iodide used in this work was a sample of Bender and Ho- bein's best material. It was freed from all impurities. The material was dried as follows : It was carefully heated with a little ammonium iodide in a current of pure, dry hydrogen, until all the water and ammonium iodide had been driven off. After cooling in the stream of hydrogen it was at once dissolved. The solutions were perfectly neutral and showed not the slight- est coloration with a starch solution. No trace of ammonium salts could be detected. The solutions were preserved in the dark in tightly stoppered bottles. TABLE 13. — Molecular conductivity of strontium iodide. V In water. In methyl alcohol. In ethyl alcohol. In propyl alcohol. M«0° M»25° M*0° M.,250 |U«00 1^25° /M)° ^,25° 32 64 128 256 512 1024 113.1 117.7 122.1 126.0 129.8 132.6 205.3 214.5 223.1 231.8 240.2 245.9 75.82 85.01 94.76 104.4 114.0 123.4 101.4 115.3 128.6 141.4 153.9 166.3 17.44 20.28 23.66 27.00 32.07 36.01 24.00 28.88 33.53 38.88 46.13 51.25 4.70 5.62 6.52 7.41 7.58 8.84 10.20 11.32 V In methyl alcohol (25 p. ct.) and water. In methyl alcohol (50 p. ct.) and water. In methyl alcohol (75 p. ct.) and water. In ethyl alcohol (50 p. ct.) and water. In propyl alcohol (50 p. ct.) and water. /0>° M,25° /M>° ^25° M»0° At.,250 MvO° ^25° ^0° Ah25° 32 64 128 256 512 1024 63.06 66.05 68.62 70.98 73.10 75.51 131.3 138.5 145.3 152.3 157.4 161.9 50.19 52.61 55.05 57.18 61.08 103.8 109.9 115.3 120.1 124.3 128.5 55.53 59.24 62.85 66.68 69.98 73.22 98.09 104.8 111.4 118.0 124.8 131.4 28.32 29.72 31.25 32.23 33.22 34.16 72.51 76.89 80.21 83.21 86.44 89.32 27.40 28.63 29.83 30.98 67.67 71.63 75.44 79.32 TABLE 14. — Temperature coefficients of conductivity of strontium iodide. In ethyl In propyl In various mixtures of V In water In methyl alcohol In ethyl alcohol In propyl alcohol alcohol (50 p. ct.) and alcohol (50 p. ct.) methyl alcohol and water. (.0 to ^o ; . (0° to 25°) (0° to 25°) (0° to 25°) water (0°to25°). (0° to 25°). 25 p. ct. 50 p. ct. 75 p. ct. 32 3.29 1.02 0.262 0.115 1.77 1.61 2.73 2.14 1.70 64 3.87 1.21 .344 .129 1.89 1.72 2.90 2.29 1.82 128 4.04 1.35 .396 .147 1.96 1.82 3.07 2.41 1.94 256 4.23 1.48 .475 .156 2.04 1.93 3.25 2.52 2.05 512 4.42 1.60 .562 • • • 2.13 3.37 2.59 2.19 1024 4.53 1.72 .610 2.21 .... 3.45 2.70 2.33 Some of the results in table 13 are plotted in fig. 4. It is seen that the curve is of the same form as that for potassium iodide at the same tempera- ture. The minimum point is reached with an alcohol of about 50 per cent. STRONTIUM IODIDE. 33 Curve I , v = 33 II, v = 64 III, v = 128 IV, V = 256 V , v = 512 VI, v = 25 $ 50 # 75/o Concentration of Methyl Alcohol FIG. 4. — STRONTIUM IODIDE AT 0°. TABLE 15. — Comparison of the molecular conductivity of strontium iodide. In water, methyl alcohol, and mixtures of these solvents. V AtO°. At 25°. 0 p. ct. 25 p.ct. 50 p. ct. 75 p. ct. 100 p. ct. 0 p. ct. 25 p.ct. 50 p.ct. 75 p. ct. 100 p. ct. 32 113.1 63.06 50.19 55.53 75.82 205.3 131.3 103.8 98.09 101.4 64 117.7 66.05 52.61 59.24 85.01 214.5 138.5 109.9 104.8 115.3 128 122.1 68.62 55.05 62.85 94.76 223.1 145.3 115.3 111.4 128.6 256 126.0 70.98 57.18 66.68 104.4 231.8 152.3 120.1 118.0 141.4 512 129.8 73.10 59.51 69.98 114.0 240.2 157.4 124.3 124.8 153.9 1024 132.6 75.51 61.03 73.22 123.4 245.9 161.9 128.5 131.4 166.3 34 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 15. — Comparison of the molecular conductivity of strontium iodide. — Continued. In water, ethyl alcohol, and a 50 p. ct. mixture of these solvents. AtO°. At 25°. V Water. Mixture. Ethyl alcohol. Water. Mixture. Ethyl alcohol. Mt> MB MB MB M» M» 32 113.1 28.32 17.44 205.3 72.51 24.00 64 117.7 29.72 20.28 214.5 76.89 28.88 128 122.1 31.25 23.66 223.1 80.21 33.53 256 126.0 32.23 27.00 231.8 83.21 38.88 512 129.8 33.22 32.07 240.2 86.44 46.13 1024 132.6 34.16 36.01 245.9 89.32 51.25 Some of the results in table 15 are plotted as fig. 5. It is seen that in this curve the effect of temperature has been such as almost to blot out the minimum value in the curve for v = 32 ; and in the other dilutions the minimum is much less pronounced than in the curves thus far studied. The effect of temperature is also to shift the minimum point to the right, the minimum point existing for an alcohol of about 65 to 70 per cent. In table 15 for 0° we see that the values for pure ethyl alcohol are, in the stronger solutions, much smaller than those for the mixture. They, however, increase more rapidly, and in the most dilute solutions pass the values for the mixture, giving us again the minimum point. At 25° there is not the slightest trace of a minimum point, although the values are well below the mean of the values for the pure solvents. In comparing the values for a mixture of propyl alcohol and water with those for the pure solvents, we find that there is not the slightest trace of a minimum either at 0° or 25°. CADMIUM IODIDE. The cadmium iodide which was used was a sample which had been em- ployed in some previous work in this laboratory, and had then been very carefully purified. The solutions were made by direct weighing. TABLE 16. — Comparison of the molecular conductivity of cadmium iodide. V In methyl alcohol, water, and a 50 p. ct. mix- ture of these solvents at 25°. In varying mixtures of methyl alcohol and water at 25°. Water. M» Mixture. MB Methyl alcohol. M» Concentration of alcohol, p. ct. v= 16. Mr V = 64. /J-v 16 62.98 20.31 13.07 0 62.98 104.7 32 81.96 24.22 13.59 50 20.31 31.17 64 104.7 31.17 14.16 60 25.66 128 129.3 42.03 14.78 80 14.70 18.41 256 153.6 50.43 15.44 100 13.07 14.16 CADMIUM IODIDE. 35 When the results in the second part of table 16 are plotted as curves, no trace of a minimum appears (fig. 6). A considerable difference, however, is noticed between the values obtained and those required from the law of mixtures, the conductivity values obtained being always lower. 10M Concentration of Methyl Alcohol FIG. 5. — STRONTIUM IODIDE AT 25°. LITHIUM NITRATE. The lithium nitrate used in this work was a sample obtained from Kahlbaum. It was dried in an air-bath at 150° and kept in a desiccator. The solutions were made by direct weighing. From table 17 we see that at 0° the conductivity in pure methyl alcohol, although starting lower than the conductivity in water, increases more rapidly, so that we have solutions in methyl alcohol with greater conductivity than 36 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. solutions of the same strength in water. That our measurements are fairly accurate is made more probable by the close agreement with the values obtained by Ostwald. The solutions measured at 0° were the same as at 25°. The conductivity of a number of the solutions in methyl alcohol was redetermined, using a different sample of the alcohol and salt. In all cases the agreement was all that could be desired. 20 202 40$ 602 80^ Concentration of Methyl Alcohol FIG. 6. — CADMIUM IODIDE. From table 17 it is seen that with lithium nitrate in ethyl alcohol a minimum point is found in the conductivity values at 0°, and through all the dilutions employed. At 25° no trace of minimum values is apparent. LITHIUM NITRATE. 37 TABLE 17. — Molecular conductivity of lithium nitrate. V In water. In methyl alcohol. In ethyl alcohol. MP Mr (Ostwald) w ^ M> 1^25° 32 64 128 256 512 1024 50.00 51.49 52.51 53.40 54.70 55.30 91.83 94.62 98.00 99.68 101.3 102.3 91.8 94.5 97.7 100.0 101.5 102.0 45.97 50.12 53.95 56.67 60.06 63.40 63.51 69.32 74.51 80.57 83.31 86.46 14.29 15.60 17.52 19.39 21.36 23.29 21.99 24.85 27.72 30.84 33.25 35.52 V In ethyl alcohol (50 p. ct.) and water. In methyl alcohol (25 p. ct.) and water. In methyl alcohol (50 p. ct.) and water. In methyl alcohol (75 p. ct.) and water. /^o° ^25° ^° /425° M*0° M,25° ^ *,25° 32 64 128 256 512 1024 13.10 13.56 14.27 14.63 15.45 16.25 33.73 35.57 37.08 38.85 40.14 41.35 29.15 29.68 30.15 30.70 31.35 32.56 60.56 62.16 63.77 64.96 66.78 69.02 23.59 24.49 25.03 25.71 26.35 27.35 47.87 49.92 51.50 53.57 54.62 55.60 26.67 27.95 28.66 29.51 30.64 31.91 47.06 49.52 51.64 54.36 56.68 58.56 TABLE 18. — Temperature coefficients of conductivity of lithium nitrate. V In water (0°to25°). In methyl alcohol (0°to25°). In ethyl alcohol (0° to 25°). In ethyl alcohol (50 p. ct.) and water In methyl alcohol (25, 50, and 75 p. ct.) and water (0° to 25°). (0° to 25°). 25 p. ct. 50 p. ct. 75 p. ct. 32 1.67 0.702 0.308 0.83 1.25 0.97 0.82 64 1.72 .768 .370 .88 1.30 1.01 .86 128 1.82 .822 .408 .91 1.35 1.06 .92 256 1.85 .956 .458 .97 1.37 1.11 .99 512 1.86 .930 .476 .99 1.42 1.13 1.04 1024 1.88 .922 .489 1.04 1.46 1.13 1.07 TABLE 19. — Comparison of the molecular conductivity of lithium nitrate in methyl alcohol, water, and mixtures of these solvents. AtO°. At 25°. V 0 p. ct. 25 p. ct. 50 p. ct. 57p.ct. 100 p. ct. 0 p. ct. 25p.ct. SOp.ct. 75p.ct. 100 p. ct. 32 50.00 29.15 23.59 26.67 45.97 91.83 60.56 47.87 47.06 63.51 64 51.49 29.68 24.49 27.95 50.12 94.62 62.16 49.92 49.52 69.32 128 52.51 30.15 25.03 28.66 53.95 98.00 63.77 51.50 51.64 74.51 256 53.40 30.70 25.71 29.51 56.67 99.68 64.96 53.57 54.36 80.57 512 54.70 31.35 26.35 30.64 60.06 101.3 66.78 54.62 56.68 83.31 1024 55.30 32.56 27.35 31.91 63.40 102.3 69.02 55.60 58.56 86.46 These values, table 19, when plotted (figs. 7 and 8), give curves very similar to those already described in connection with potassium iodide and strontium iodide. The only difference worthy of special mention is the fact that at 0° the molecular conductivity in methyl alcohol rises above that in water. 38 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. FERRIC CHLORIDE. It was desired to make a complete investigation of the changes in the conductivity of solutions of ferric chloride in the various solvents. This was desirable on account of the great solubility of the substance in the different alcohols, and because of the large number of ions into which it can dissociate. This part of the investigation had to be postponed. Some few observations were, however, made, and these are recorded. The ferric chloride used was I , v = 32 II, v= 64 III, v= 128 IV, v — 356 V, v= VI, t; = 100* Concentration of Methyl Alcohol FIG. 7. — LITHIUM NITRATE AT 0°. prepared as follows: Iron filings, which had been washed with alcohol and ether to remove any adhering grease, were heated in a current of pure, dry chlorine, in a large combustion tube of hard glass. The ferric chloride formed was allowed to distill into a cooled portion of the tube, and then it was re- distilled into a wide-mouthed salt bottle. The excess of chlorine was removed by heating the chloride in a current of dry nitrogen. The chloride thus formed dissolved completely in both alcohol and water. Water solutions were standardized as follows: They were reduced with zinc and sulphuric acid, and the ferrous iron determined with standard FERRIC CHLORIDE. 39 potassium permanganate. Alcoholic solutions were first precipitated with aqueous ammonia; filtered, washed, dissolved in a little hydrochloric acid, reduced, and titrated as above. 110- V IV III II 90- o 'O o O o 0> „, •370- 50- VI V IV III ir i 25$ 50$ 75# Concentration of Methyl Alcohol FIG. 8. — LITHIUM NITRATE AT 25°. 100!$ The values in table 20 agree fairly well with those found by Goodwin,1 but are in all cases slightly lower, possibly due to the fact that the conduc- tivity of our solutions was measured immediately after standardization, while those used by Goodwin had been made up for several months. Accurate measurements at a dilution greater than v = 256 could not be made, since hydrolysis took place to a very marked extent, as was noticed by Goodwin in the work just referred to. TABLE 20. — Conductivity of ferric chloride in water. V Molecular conductivity. Temperature coefficient (0°to25°). M»0° M«25° 32 64 128 256 163.7 187.0 213.0 240.0 319.2 370.2 422.6 476.7 6.22 7.32 8.38 9.47 'Phys. Rev., 11, 193 (1900); Ztschr. phys. Chem., 21, 1 (1896). 40 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. In mixtures of the alcohols and water the same hydrolysis was found to take place, while in the solutions in absolute alcohol a different change occurred. The solution, originally of a pale straw color, gradually became lighter and lighter when in contact with the platinum black of the electrodes. This was accompanied by a steady rise in the molecular conductivity, which, at the end of 24 hours, was still appreciable. The colorless solution showed only the slightest trace of ferric iron, having apparently been reduced to the ferrous condition. These changes are shown in table 21. TABLE 21. — Changes in the molecular conductivity of ferric chloride (v = 512), with time. In mixture of methyl alcohol (50 p.ct.) and water. In methyl alcohol. Date. M»25° Date. M.253 AprilS. 12 h 25m p.m. April 15. 2h 15m p.m. 12 27 130.1 2 18 62.90 12 29 135.6 2 21 65.47 12 31 141.8 2 24 67.18 12 33 145.4 2 28 68.16 12 36 150.8 2 35 69.64 12 38 153.9 2 45 71.91 12 41 156.3 3 00 75.00 12 45 159.5 3 15 78.06 12 56 162.4 3 30 81.75 1 00 166.7 3 46 84.11 1 10 169.1 4 15 88.59 1 30 172.9 4 45 92.93 2 00 175.0 6 15 103.6 3 00 180.6 April 16. 10 35 a.m. 132.6 4 00 182.2 11 15 133.7 5 00 183.0 2 35 p.m. 134.3 SUMMARY. The preceding investigation leads to the conclusion that the minimum point, discovered by Zelinsky and Krapiwin, is not an isolated phenomenon restricted to the mixtures of methyl alcohol and water, but is much more general. This minimum point in the conductivity has been found for all the salts studied in mixtures of methyl alcohol and water, with the exception of cadmium iodide. Ethyl alcohol and water yield a minimum in the conduc- tivity of all the salts investigated at 0°. At 25° this minimum had disappeared. Mixtures of methyl alcohol and ethyl alcohol do not exhibit this phenomenon, but the conductivity of a salt dissolved in a 50 per cent mixture of methyl and ethyl alcohols is less than the mean of the conductivities of the substance in the pure solvents at the same dilution. To explain these facts we advance tentatively the following suggestion : According to the theory of Dutoit and Aston it is only those substances whose molecules are polymerized that can dissociate dissolved electrolytes. If this SUMMARY. 41 be true, it is probable, since those substances which dissociate dissolved elec- trolytes also show in general a normal molecular weight for dissolved non- electrolytes, that this breaking down of the polymerized molecule can be accomplished best by another associated molecule. From this it follows that the effect of mixing two associated solvents would be to lower the state of association of one or both until a state of equilibrium is reached. Such a mixture would be that of water and either methyl or ethyl alcohol, or a mix- ture of methyl alcohol and ethyl alcohol. In these cases, since the molecules are less associated than the constituents, we should expect dissolved electro- lytes to show a conductivity lower than that required by the law of mixtures. In every solvent with which we have worked this is exactly what has been observed. In the mixtures of methyl alcohol and water, where the association of the constituents is the greatest, the lowering of conductivity is also the greatest, as would be expected. In support of the above view that one associated solvent can diminish the association of another associated solvent, we have experimental evidence in the results of freezing-point measurements. The molecular weights of the alcohols in water, as determined by the freezing-point method, are normal; while the surface-tension method of Ramsay and Shields shows, beyond question, that the alcohols are associated compounds. The effect of temperature on the lowering of the conductivity is in accord with the above suggestion. Since the effect of rise in temperature is to lower the state of aggregation of an associated liquid, it would be expected that at the higher temperature the influence of the solvents on each other would be less than at the lower temperature. That such is the case can be seen by comparing the results at 0° with those at 25°. The conclusion reached from this investigation, that one associated solvent can diminish the association of another associated solvent, was subsequently confirmed by the work of Jones and Murray.1 They worked with water, and formic and acetic acids, and determined the molecular weight of each in the other by the freezing-point method. These, as is well known, are all strongly associated substances when in the pure, homogeneous condition. Jones and Murray found that the molecular weights of these substances, in the most concentrated solutions which could be studied, were always less than the molecular weights of the pure substances as determined by the method of Ramsay and Shields;2 and, further, the molecular weights de- creased in every case with increase in the dilution of the solution, as would be expected from the law of mass action. Both of these facts point to the same conclusion, viz, that one associated solvent breaks down the complex molecules of another associated solvent into simpler molecules. 'Amer. Chem. Journ., 30, 193 (1903). 2 Ztschr. phys. Chem., 12, 433 (1893). 42 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. The action of one associated solvent on another associated solvent will be seen to be analogous to the action of an associated solvent on an electrolyte, if we take into account that an associated solvent is a non-electrolyte. An associated solvent breaks the molecules of an electrolyte down into ions. An associated solvent breaks the complex molecules of a non-electro- lyte down into simpler molecules, which is the nearest approach to ions that can be obtained from a non-electrolyte. WORK OF CARROLL. The first part of this work is a continuation of the investigation of Jones and Lindsay. EXPERIMENTAL. APPARATUS. The Kohlrausch method of measuring conductivity was used throughout this investigation. The bridge-wire was of "manganin." The resistance coils were carefully calibrated. The conductivity cells were of the form used by Jones and Lindsay. The constants of these were determined by means of N/50 and N/500 solutions of potassium chloride. Cells used to determine conductivities of the solvent and of highly diluted solutions were treated in the manner recommended by Whetham.2 The electrodes were first coated with platinum black in the usual manner, and were afterwards heated to a high temperature in the flame of a blast-lamp. It was found, as Whetham states, that the usual coating of platinum black, in spite of careful and long- continued washing, retains traces of salt that subsequently pass slowly into solution. The oxidizing action of the platinum black is also avoided by this treatment. For the purposes mentioned, electrodes of this kind can not be too highly recommended. The tone-minimum in the telephone is fully as good as with the ordinary type of electrode. The 25° thermostat was of the usual (Ostwald) form, and the stirrer was driven by a small hot-air motor. The zero-bath was of the type used by Jones and Lindsay, consisting of an outer and an inner vessel. The inner vessel and the annular space between the two were filled with finely crushed ice. The outer portion of ice was moistened with a small quantity of distilled water, and to the ice in the inner vessel about an equal weight of water was added. By the foregoing means the temperature of a cell immersed in the ice and water of the inner vessel could be kept for any desired period at 0.02° to 0.05° C. The measuring flasks, pipettes, and burettes were carefully calibrated. SOLVENTS. The water used was purified in the following manner: Ordinary distilled water, after addition of potassium dichromate and sulphuric acid, was re- distilled. The distillate was again distilled from chromic acid into, and then out of, a solution of barium hydroxide. When the conductivity of the water thus obtained was greater than 2 x 10~6, the above process was repeated. In many cases the conductivity was much less than this value. 1Amer. Chem. Journ., 28, 329 (1902). 2 Phil. Trans., 94 (A), 321 (1900). Ztschr. phys. Chem., 33, 346 (1900). 43 44 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. The methyl and ethyl alcohols were prepared from the purest commercial preparations obtainable. Each was subjected to the same treatment. The commercial alcohol was dehydrated by standing in contact with freshly burned lime for several weeks. From this it was distilled, and then allowed to stand over dehydrated copper sulphate for a week or more. When required for use, it was distilled from the copper sulphate, small quantities of sodium being added, and precautions were taken to protect the distillate from access of moisture. The conductivity of the methyl alcohol thus obtained was usually from 1 to 2 x 10~6. That of the ethyl alcohol was less. The acetic acid used was obtained from Bender and Hobein, and was designed for cryoscopic work. The amount of water contained in it was determined, as suggested by Rudorf, by observation of its freezing-point. Its conductivity was less than 2 x 10 ~6. METHOD OF PREPARING THE SOLUTIONS. The mixtures of solvents were prepared as follows : n c. c. of alcohol, for example, were diluted to, say, 100 c. c. This is designated as a mixture of n per cent alcohol. Calibrated flasks were used for the dilutions, and the temperature was kept within a few tenths of a degree of the temperature of calibration. In making up the solutions, the exact amount of the salt in question was put into a measuring-flask, and after adding a portion of the solvent, the substance was dissolved and the flask filled to the mark. Here also the temperature was kept under control. Usually, the original solutions were N/16 or N/32. From these, others were made by adding the solvent to a measured portion of the solution. Where the quantity to be used would be too small to be measured with reasonable accuracy, one of the intermediate solutions was taken as a starting-point for further dilution. CONDUCTIVITY MEASUREMENTS The constants of the cells used were determined or checked at intervals of a few days. For each conductivity determination, from three to seven or eight different resistances were used. The values given in the tables are, therefore, the mean of several determinations. Conductivities throughout are expressed as molecular conductivities. CADMIUM IODIDE. The cadmium iodide used was a preparation of which a part had been used by Jones and Lindsay in their work. Jones and Lindsay measured the conductivity of cadmium iodide in water, methyl alcohol, and mixtures at 25° only. The minimum was not observed. It seemed desirable, therefore, to complete the study of the compound. The cadmium iodide was dried by being allowed to stand in a desiccator over calcium chloride for a week or more. At first the attempt was made to CADMIUM IODIDE. 45 dry it by prolonged heating in an air-bath at 70° to 80°. It was found that when thus treated, the salt assumed a pinkish hue, which immediately gave place to the pinkish white of the salt in the ordinary condition when a small quantity of water was added. No mention of this color change can be found in the literature. Though no traces of decomposition could be detected, the other method of drying was chosen. The original solutions were made by direct weighing. 80- V 70- IV I v=* 16 II V=> 32 III V = 64 IV V=128 § O o "o 30' 20- 10- 25# 501 75$ 100^ Per cent Alcohol by Volume FIG. 9. — CONDUCTIVITY OF CADMIUM IODIDE IN WATERV METHYL ALCOHOL, AND MIXTURES AT 0°. TABLE 22. — Conductivity of cadmium iodide. In water at 0° and 25°. In 25 p. ct. methyl alcohol at 0° and 25°. In 50 p.ct. methyl alcohol at 0° and 25°. V Tempera- Tempera- Tempera- /M>° ^25" ture coeffi- M«0° M»0° ture coeffi- /M>° At«25° ture coeffi- cient. cient. cient. 16 31.16 62.86 1.268 14.57 33.83 0.770 9.96 20.82 0.434 32 40.07 81.82 1.670 17.68 42.10 0.977 11.23 24.21 0.519 64 51.93 107.77 2.114 22.68 55.02 1.294 14.21 31.25 0.682 128 63.85 130.24 2.666 28.59 70.22 1.665 18.92 41.61 0.908 256 76.54 155.55 3.160 35.36 87.31 2.078 21.58 51.4 1.193 In 75 p.ct. methyl alcohol at 0° and 25°. In 100 p.ct. methyl alcohol at 0° and 25°. In ethyl alcohol, ^25°. V Tempera- Tempera- MvO° M.258 ture coef- ^0° /^25° ture coef- 25 p.ct. 50 p.ct. 75 p. ct. 100 p.ct. ficient. ficient. 16 8.94 15.78 0.274 10.96 13.39 0.097 26.97 14.03 9.43 2.29 32 9.71 17.08 .295 11.55 14.51 .118 34.90 16.19 9.52 2.30 64 11.24 20.08 .354 12.66 14.83 .087 44.21 18.93 10.94 2.32 128 14.37 25.68 .452 13.69 16.82 .124 54.54 25.66 12.27 2.39 256 18.58 33.52 .598 17.62 20.01 .096 69.71 34.89 15.16 2.66 46 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. From the data given, especially in table 23, it is seen that cadmium iodide does not show the minimum in mixtures of methyl alcohol and water at 25°. At 0°, however, in a 75 per cent mixture, at volumes 16, 32, and 64, the mini- mum appears. Beyond these concentrations it disappears. 25$ 50$ 75$ Per cent Alcohol "by Volume FIG. 10. — CONDUCTIVITY OF CADMIUM IODIDE IN WATER, METHYL ALCOHOL, AND MIXTURES AT 25°. In various mixtures of ethyl alcohol and water at 25° (table 23) no minimum appears, although the values observed are in all cases less than would be expected from the rule of averages. The results are plotted as curves in figs. 9 and 10, the ordinates being conductivities and the abscissae representing the per cent by volume of alcohol. SODIUM IODIDE. 47 TABLE 23. — Comparison of conductivities. Cadmium iodide in methyl alcohol. AtO°. At 25°. V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. Op.ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. CH3OH 16 31.16 14.57 9.96 8.94 10.96 62.86 33.83 20.82 15.78 13.39 32 40.07 17.68 11.23 9.71 11.55 81.82 42.10 24.21 17.08 14.51 64 51.93 22.68 14.21 11.28 12.66 104.77 55.02 31.25 20.08 14.83 128 68.53 31.09 18.92 14.37 13.69 130.24 70.22 41.61 25.68 16.82 256 76.54 35.36 21.58 18.58 17.62 155.55 87.31 51.40 33.52 20.01 Cadmium iodide in ethyl alcohol at 25°. V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. C,H5OH 16 62.86 26.97 14.03 9.43 2.29 32 81.82 34.90 16.19 9.52 2.30 64 104.77 44.21 18.93 10.94 2.32 128 130.24 54.54 25.66 12.27 2.39 256 155.55 69.71 34.89 15.16 2.66 SODIUM IODIDE. The sodium iodide used was a preparation that had been carefully purified by Jones and Lindsay. The salt was dried in an air-bath for three days at a temperature of 1 10° to 130°. This prolonged treatment was found necessary to bring it to constant weight. The original solutions were made up by direct weighing. TABLE 24. — Conductivity of sodium iodide at 0° and 25°. In water. In 25 p. ct. methyl alcohol. V Tempera- Tempera- jU»0° £iv25° /x.j25o(0) ture UTJQ° A<*250 ture coefficient.1 coefficient. 32 57.46 106.0 105.7 1.942 33.63 70.62 1.48 64 59.37 109.35 109.3 2.000 34.68 72.77 1.52 128 60.71 112.44 112.3 2.069 35.63 73.78 1.53 256 62.35 115.5 115.2 2.126 36.73 74.32 1.504 512 64.28 118.08 117.9 2.152 37.83 74.98 1.49 In 50 p. ct. methyl alcohol. In 75 p. ct. methyl alcohol. In 100 p.ct. methyl alcohol. Temper- Temper- Temper- ature ature ature V ^.(,0 fJ-i-25 coeffi- M»0 ^25° coeffi- /AjjO ^25° coeffi- cient. cient. cient. 32 27.91 57.18 1.171 31.70 56.50 0.992 51.09 72.03 0.838 64 28.73 58.30 1.183 33.08 59.47 1.056 55.95 77.63 0.867 128 29.04 59.16 1.205 34.16 61.49 1.093 58.89 82.70 0.955 256 30.32 60.87 1.232 34.72 62.74 1.121 61.02 86.19 1.009 512 32.08 61.62 1.181 35.00 63.77 1.151 62.56 88.27 1.026 1 These values are those given by Ostwald [Ztschr. phys. Chem., 7, 74 (1887)]. The agreement s seen to be quite satisfactory. 2 The values at 0° are uncorrected for conductivity of solvent. 48 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 25. — Comparison of conductivities. V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. CH3OH r 32 57.46 33.63 27.91 31.70 51.09 64 59.37 34.68 28.73 33.08 55.95 AtO° 128 60.71 35.63 29.04 34.16 58.89 256 62.35 36.73 30.32 34.72 61.02 . 512 64.28 37.83 32.08 35.00 62.56 r 32 106.0 70.62 57.18 56.50 72.03 64 109.35 72.77 58.30 59.47 77.63 At25°. 128 112.44 73.78 59.16 61.49 82.76 256 115.49 74.33 60.87 62.74 86.19 512 118.08 74.98 61.62 63.77 88.27 From the data given in table 25, and from the curves plotted in the man- ner already indicated (fig. 11), it is evident that sodium iodide exhibits the minimum in mixtures of methyl alcohol and water. Only two dilutions at the two temperatures of obser- vation have been plotted, since the curves would be too close together if all were shown. The minimum, as Jones and Lindsay have observed in other cases, is more pro- nounced at 0° than at 25°. Further, the shifting effect of change of temperature and of concentration, also observed by Jones and Lindsay, appears at 25°, v = 32. The minimum oc- curs in a 75 per cent mix- ture ; beyond this dilution the minima occur solely in the 50 per cent mixture. At 0° the minimum appears to be exhibited in the 50 per cent mixture alone. 13 = 512 (C5C) V = 3-2 (£5°) V = 512 (0 V •= 32 (0 25 tf 00 -i 7j;'t Perec nt Methyl. Alcohol "by Volume ice;; FIG. 11. — CONDUCTIVITY OF SODIUM IODIDE IN MIXTURES OF METHYL ALCOHOL AND WATER. CALCIUM NITRATE. 49 CALCIUM NITRATE. It was thought that the study of a ternary salt might prove interesting. Jones and Lindsay had already made a study of strontium iodide, and had found it to exhibit the same phenomena as did binary salts. The calcium nitrate used was a preparation obtained from Kahlbaum. This substance was found to be particularly difficult to dehydrate. Heating for several days to a temperature of 103° to 140° was necessary to bring it to constant weight. This treatment caused no perceptible decomposition. All the original solutions were made by direct weighing. TABLE 26. — Conductivity of calcium nitrate. V In water at 0° and 25°. In 25 p. ct. methyl alcohol at 0° and 25°. *o° ^25° Temperature coefficient. ^ *25° Temperature coefficient. 16 32 64 128 256 94.33 102.47 108.35 113.59 118.02 177.56 189.45 199.24 209.93 215.93 3.329 3.440 3.636 3.854 3.916 53.17 57.30 59.81 63.39 66.19 111.45 120.63 128.95 136.81 141.45 2.333 2.533 2.776 2.937 3.010 V In 50 p. ct. methyl alcohol In 75 p. ct. methyl alcohol at 0° and 25°. at 0° and 25°. In 100 p. ct. methyl alcohol at 0° and 25°. ^00 *»' TcCoT M,OO 25o Temp. f***0 coef. ju,0° ,,25o Temp, coef. 16 32 64 128 256 41.07 44.70 49.15 53.94 54.82 79.04 1.518 90.11 1.816 98.35 1.968 103.68 1.990 109.19 2.175 39.59 43.60 48.58 51.90 70.06 1.219 80.16 1.462 89.98 1.645 97.72 1.883 3V.30 37.27 46.66 55.17 32.79 41.88 50.79 60.52 73.98 0.423 .541 .555 .752 V In 25 p. ct. ethyl alcohol at 0° and 25°. In 50 p. ct. ethyl alcohol at 0° and 25°. M,0o ,,25o Temperature coefficient. *00 ,,25o Temperature coefficient. 16 32 64 128 256 38.80 41.84 48'.02 50.26 92.93 100.56 112'.93 119.04 2.065 2.359 2.596 2.750 23.70 2589 27.37 28.83 30.30 60.43 65.30 69.83 74.13 78.31 1.469 1.576 1.698 1.812 1.920 In 75 p. ct. ethyl alcohol at 0° and 25°. In 100 p. ct. ethyl alcohol at 0° and 25°. V M- M»25° Temperature coefficient. a 0° LL.2*i° Temperature ^* ^" coefficient. 16 32 64 128 256 19.13 21.41 26*71 28.61 39.08 44.57 56.08 60.79 0.798 0.926 1.175 1.287 4.89 7.11 8.59 10.74 12.48 7.08 9.57 12.43 15.47 18.57 0 .088 .098 .153 .189 .244 50 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 27. — Comparison of conductivities. V Calcium nitrate in methyl alcohol. AtO°. At 25°. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. CH3OH 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. CH3OH 16 32 64 128 256 94.33 102.47 108.35 113.59 118.02 53.17 57.30 59.81 63.39 66.19 41.07 44.70 49.15 53.94 54.82 39.59 43.60 48.85 51.90 31.30 37.27 46.66 55.17 177.56 189.45 199.24 209.93 215.93 111.45 120.63 128.95 136.81 141.45 79.04 90.11 98.35 103.68 109.19 70.06 80.16 89.98 97.72 32.79 41.88 50.79 60.52 73.98 V Calcium nitrate in ethyl alcohol. AtO°. At 25°. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. C2H5OH 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. C2HBOH 16 32 64 128 256 94.33 102.47 108.35 113.59 118.02 38.80 41.84 48.02 50.26 23.70 25.89 27.37 28.83 30.30 19.13 21.41 26.71 28.61 4.89 7.11 8.57 10.74 12.48 177.56 189.45 199.24 207.93 215.93 92.93 100.56 li2.93 119.04 60.43 65.30 69.83 74.13 78.31 39.08 44.57 56.08 60.79 7.08 9.57 12.43 15.47 18.57 From the data given in table 27, it is evident that calcium nitrate in no case exhibits the minimum. The conductivities are always less than the proper average. The relation found by Wakeman, = const, does not hold in the p(100 - p) cases thus far studied; nor does that found by Cohen. HYDROCHLORIC ACID. In the study of hydrochloric acid the solutions were prepared as follows : Into a portion of the solvent, kept cool by ice, dry hydrochloric-acid gas was conducted. This was obtained by allowing concentrated sulphuric acid to drop slowly from a dropping-funnel into pure, aqueous, hydrochloric acid. The gas was dried by passing through gas-washing bottles containing con- centrated sulphuric acid. The vessel containing the solvent into which the gas was passed, was protected from extraneous moisture by a drying-tube containing phosphorus pentoxide. The strength of the original solution was determined volumetrically by means of a standard solution of ammonium hydroxide, methyl orange being used as the indicator. From this solution the dilutions were made. Control determinations were carried out in a number of cases, since only the fairly dilute solutions in mixtures and in pure methyl alcohol were found to be stable. The composition of the 69.75 per cent mixture was determined by means of its specific gravity. HYDROCHLORIC ACID. 51 TABLE 28. — Conductivity of hydrochloric acid. In 50 p ct. methyl alcohol at 25°. In 69.75 p. ct. methyl alcohol at 0° and at 25°. V ^25° V /M>° M«25° 33.05 172.45 44.67 63.83 116.0 132.21 166.09 92.42 67.06 123.77 264.42 165.30 178.75 66.^49 117.8 528.83 155.29 357.5 64.66 118.1 714.72 66.19 116.87 In 90 p. ct. methyl alcohol at 0° and 25°. In 100 p. ct. methyl alcohol at 0° and 25°. V M»0° M»25° V ttfl" M»25° 31.53 46.59 73.79 8.42 67.36 95.83 63.07 47.60 76.19 32.8 77.06 110.50 157.67 51.84 84.27 130.26 79.84 118.79 252.28 51.84 84.27 481.44 89.79 129.87 504.56 .... 84.22 2104.0 95.46 133.44 A consideration of the re- sults, table 28, shows that in certain cases they are irregular and unexpected. In the 50 per cent mixture the conductivity falls from the first dilution, which is analogous to what has been observed for hydrochloric acid in ether and isoamyl alcohol by Cattaneo and Kablukoff, and for sul- phuric acid in acetic acid by Jones — the molecular conductivities decreased for decreasing dilution. In the 69.75 per cent mixture a maximum is reached at v = 92.42 at both 0° and 25°. From this point the con- ductivities decrease slightly. However, at 0°the mean of the last four conductivities, including the maximum, differs from the maximum value by only 1.5 per cent. •d a 6 50 3 o o 3 30 30 10 Per cent Alcohol by Volume FIG. 12. — CONDUCTIVITY OF HYDROCHLORIC ACID IN MIXTURES OF METHYL ALCOHOL AND WATER. 52 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. In the 90 per cent mixture the results are perfectly regular, and, what is surprising, a limiting value is reached at v = 157.67. In methyl alcohol the results are also regular, but there is no indication of a maximum value for conductivity. The values found agree in most cases very well with those of Carrara, whose results are also irregular (see fig. 12). SODIUM ACETATE IN MIXTURES OF ACETIC ACID AND WATER. The acetic acid froze at 15.47°. From the tables of Riidorff it contained in 100 parts by weight, 0.6 part by weight of water. Knowing the composi- tion of the acid, the proper amounts to be used in making the mixtures could be calculated. The acid was partially frozen and the liquid portion rejected. After melting the solid acid the process was repeated. The specimen thus obtained was used for the conductivities in the pure solvent. Owing to inevitable exposure to the moisture of the air, it was not thought profitable to try to remove the last traces of moisture. In this part of the problem new complications arose, since acetic acid in aqueous solution conducts the current. The specific conductivity of the solvent (containing 25, 50, 75, and 100 per cent by volume of acetic acid) was determined; then that of the solution in question. The difference between the latter and the former was multiplied by the volume to give the (apparent) molecular conductivity. The sodium acetate used was a specimen of the fused salt obtained from Kahlbaum. It was dried for two days in an air-bath at 120° to 130°. The original solutions were made up by direct weighing. TABLE 29. — Conductivity of sodium acetate. V Specific conductivity of solution. M«.25°. In 25 p. ct. acetic acid at 2.5° (specific / 32 2.040 17.89 conductivity of solvent = 1.631) . . \ 64 1.469 -10.37 In 50 p. ct. acetic acid at 25° (specific conductivity of solvent = 0.8584). . ( 32 ] 64 (128 1.388 0.9263 0.8986 16.84 4.36 5.15 In 75 p. ct. acetic acid at 25° (specific conductivity of solvent = 0.0558). . f 32 4 64 (128 0.7050 0.4446 0.2999 20.77 24.87 31.24 In acetic acid at 25° (conductivity of solvent — 1 x 10 ~6) ( 32 \ 64 0.00418 0.00236 0.134 0.161 (128 0.00129 0.165 The results are seen to be irregular, and no final conclusions can be drawn from them. In the 25 per cent mixture for v = 64, the molecular conduc- tivity is apparently negative. This, of course, means nothing more than that the specific conductivity of the solvent is greater than that of the solution. SODIUM ACETATE. 53 Only in the 75 per cent mixture is there any regularity observed. Here the (apparent) molecular conductivities increase with decreasing concentration. In the pure solvent the conductivities are so small as to be almost neg- ligible. This is not surprising. Wakeman * has determined the conductivity of hydrochloric acid in acetic acid, and has found it to be exceedingly small. For example, for v = 98.56 /><.„ was found to be 1.78. The irregular results in mixtures of acetic acid and water are to be ex- plained as being due to mutual isohydric influence of dissolved substance and solvent. The dissociation of the sodium acetate is driven back by the acetic acid, and vice versa. This influence is most marked where the disso- ciation of each separately would be greatest, i.e., in mixtures of lower per cent of acetic acid, and of minimal concentration of sodium acetate. Since these phenomena have no direct connection with the problem in hand, no further discussion is necessary, especially as the question has been treated by Wolf 1 and by Riidorff .2 DISSOCIATION IN FIFTY PER CENT METHYL ALCOHOL. It has been seen that in the case of hydrochloric acid in mixtures of methyl alcohol and water, limiting values for conductivity are reached at a smaller dilution than in either water or methyl alcohol. It is important to see whether this relation is general. The limiting values in the case of sodium and potassium iodides and potas- sium bromide were determined. Throughout this part of the work the utmost care was taken in the preparation of solvents and solutions, and in making the dilutions. The cells used were standardized before and after each series of measurements. The conductivity of the solvent was carefully deter- mined, and the necessary corrections were made. The water and the methyl alcohol used had a conductivity of not over 1 x 10~6. Every result given is the mean of from five to ten different values. POTASSIUM IODIDE. The potassium iodide was prepared by Kahlbaum. The flame test showed that no appreciable impurity was present. The salt was dried to constant weight at 100° to 110°. The values obtained are given: V A^25° 400 2000 4000 69.29 70.58 70.60 The value for the conductivity at v = 400 agrees with the (interpolated) value obtained by Zelirisky and Krapiwin, and our results may therefore be 'Ztschr. phys. Chem., 15, 181 (1894). 2 Loc. cit. CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. incorporated with theirs. This has been done in table 30. The degree of dissociation has been calculated for each dilution. Included in the table, for comparison, are similar values calculated from data obtained by Ostwald for potassium iodide in water, and from data given by Carrara for potassium iodide in methyl alcohol. TABLE 30. — Conductivity and dissociation of potassium iodide in water, methyl alcohol, and 50 per cent methyl alcohol. Conductivity of potassium iodide in water, Dissociation of potassium iodide in water, methyl alcohol, and 50 p. ct. methyl alcohol, and 50 p. ct. methyl alcohol, at 25°. methyl alcohol. V H2O 50 p.ct. CHSOH CH3OH H2O 50 p. ct. CH3OH CH3OH ^25° /i,,25° M,.25° a a a 16 124.5 62.13 68.14 0.873 0.880 0.697 32 128.5 64.37 74.42 .900 0.912 .762 64 130.5 66.01 79.85 .915 0.936 .818 128 133.0 67.45 84.70 .931 0.956 .868 256 135.8 68.28 88.25 .953 0.968 .904 400 .... 69.20 .... • • • • • ... 512 137.9 69.65 90.82 .967 0.987 .931 1024 140.9 70.55 93.07 .989 1.000 .954 2000 .... 70.58 • • • • ... * .... .... 4000 .... 70.60 . . • * • * • * . • .... 00 142.6 70.58 97.63 * * • « .... • • • * The values of the dissociation factors are shown in table 30. From an inspection of the data, it is seen that a limiting value for conductivity is reached at a lower dilution in the mixture than in the other solvents. An inspection of this table shows that, at corresponding dilutions, dissocia- tion as calculated from conductivity is greater in the mixture than in methyl alcohol or in water. The only other solvent thus far shown to have a greater dissociating power than water is liquid hydrocyanic acid, as appears from the work of Centnerszwer. Potassium iodide was one of the substances used. SODIUM IODIDE. The sodium iodide was the preparation previously employed, panying measurements were made: The accom- V *,25° 500 60.92 2000 61.72 4000 61.65 Combining these with the values previously obtained, and using Carrara's values for conductivities in water and methyl alcohol, respectively, we have the values given in table 31. SODIUM IODIDE. 55 TABLE 31.— Conductivity and dissociation of sodium iodide in water, methyl alcohol, and 50 per cent methyl alcohol. Conductivity of sodium iodide in water, methyl alcohol, and 50 p. ct. methyl Dissociation of sodium iodide in water, methyl alcohol, and 50 p ct. methyl alcohol, at 25°. ' alcohol. 2 V H2O 50 p. ct. CH3OH CH2OH H,O 50 p.ct. CH3OH CHsOH M,25° AH-250 Mv25° a a a 32 106.0 57.18 68.75 0.865 0.927 0.766 64 109.3 58.30 73.11 .901 .946 .816 128 112.4 59.16 77.31 .917 .959 .861 256 115.5 60.87 79.90 .949 .987 .890 512 118.1 61.27 82.15 .971 .993 .915 2000 • • • • 61.72 1.000 4000 «... 61.65 .... 1.000 1.000 00 121.4 61.68 89.77 .... .... .... 1 It is evident, in this case, that the results are of the same general character as those found for potassium iodide. 2 Here also, as was found for potassium iodide, the dissociation is greater in the mixture. The accompanying measurements were made for potassium bromide in 50 per cent methyl alcohol : V Mt25° 250 1250 2500 65.78 69.26 69.35 At v = 250 the dissociation is 0.949 for the mixture. At v = 256 the dissociation is 0.927 for water, and 0.806 for methyl alcohol, as calculated from Ostwald's l and Carrara's 1 observations. The values for the dissociation constants are seen to be about the same as those for potassium iodide. Data for conductivity in aqueous solution could not be found. The values for methyl alcohol are calculated from Carrara's data ; those for the mixture, from the results of Zelinsky and Krapiwin. TABLE 32. — Dissociation of ammonium bromide, ammonium iodide, and lithium nitrate in 50 per cent methyl alcohol. Dissociation of ammonium Dissociation of ammonium Dissociation of lithium nitrate in bromide in 50 p. ct. methyl alcohol. iodide in 50 p. ct. methyl alcohol. 50 p. ct. methyl alcohol at 0° and at 25°. V 50 p. ct. CH3OH 50 p. ct. CH3OH oO° a25° a a 16 0.871 0.875 32 0.910 0.909 0.863 6.862 64 0.942 0.943 0.899 0.899 128 0.962 0.967 0.915 0.926 256 0.974 0.973 0.944 0.964 512 0.986 0.994 0.963 0.982 1024 1.000 1.000 1.000 1.000 1 Loc. cit. 56 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. In table 32 the values are about the same as for ammonium bromide. The data for the mixture and for methyl alcohol were furnished by Zelinsky and Krapiwin and by Carrara, respectively. From results given by Jones and Lindsay it is possible to calculate the dis- sociation of lithium nitrate in the 50 per cent mixture at two temperatures, 0° and 25°, assuming that here, as in the other cases, complete dissociation is reached at v = 1024. It is known that in aqueous solution complete disso- ciation is not reached as soon in the case of lithium salts as with potassium or sodium salts. Less value must therefore be attached to the results given in table 32. Values of the dissociation for hydrochloric acid in the 90 per cent and in the 69.75 per cent mixtures at 0° and at 25° are also given in table 33. It will be remembered that the results in the latter mixture were irregular. No stress can, therefore, be laid upon these figures. We have taken the deter- minations for v = 92.4 as the limiting values in this case. TABLE 33. — Dissociation of hydrochloric acid in 69.75 per cent and 90 per cent methyl alcohol, in water, and in methyl alcohol. 69.75 p. ct. CH3OH 90 p. ct. CH3OH H20 CH3OH V aO° a25° oO° a25° a25° a25° v 31.5 0.900 0.875 0.909 0.890 37.74 44.67 0.98i 0.937 • • . • • * * • • • * 0.916 75.47 63.0 • • . * 0.919 0.924 .... f .... 92.42 1.000 1.000 • • • • .... 0.972 150.9 158 • • • • • • • • 1.000 1.000 • • • • 252 .... .... 1.000 1.000 0.934 .... .... The values for conductivities in water are taken from Ostwald; 1 those for methyl alcohol from Carrara. It is observed in table 33 that the dissociation is greater in the 69.75 per cent mixture than in water at the corresponding dilution. This, however, is not the case for the 90 per cent mixture. It is also interesting to note that the dissociation is apparently greater at 0° than it is at 25°, and this is true for the 90 per cent mixture, where the results are more reliable than those for the 69.75 per cent mixture. We have suggested that the dissociation in the 50 per cent mixture may be due in part to the presence of the hydrate CH3OH . 3 H2O. This compound would be more stable at a lower than at a higher temperature, and would be present to a greater extent at the lower temperature, and therefore the dis- sociation might be greater. This result should be shown by salts as well as 1 Journ. prakt. Chem., 140, 300 (1885). POTASSIUM AND SODIUM IODIDES. 57 by hydrochloric acid, but data are not at hand for comparison. Jones and Lindsay's values for lithium nitrate are available and have been used (table 29). From this table, apparently, the dissociation is greater at the higher temperature. But, as we have said, no final conclusions can be drawn from these data, since we can not be certain that limiting conductivity values were reached by Jones and Lindsay. In the case of hydrochloric acid, however, they were reached in at least one instance. Further investigation will be needed to decide this matter. It is interesting at this point to see whether the hypothesis of Dutoit and Aston * is quantitatively true for the cases that have been considered. This hypothesis states that the dissociating power of a solvent is dependent upon its association, as determined by the surface-tension method of Ramsay and Shields.2 If the hypothesis holds quantitatively, it may be formulated thus : a 3/ — = - , where a and a' are the dissociations of the solutions compared, and x Ct & and x' the association factors of the solvents. The relation may be put into the form - = constant. x In comparing solutions in different solvents, there should be the same number of gram-molecules of electrolyte dissolved in the same number of gram-molecules of each solvent. Where solutions in water, methyl alcohol, and ethyl alcohol are to be compared, the volumes will have the ratio 18, 40, and 58 approximately. A comparison is made on this basis for potassium and sodium iodides in water and methyl and ethyl alcohols (table 34). For the ethyl alcohol solu- tion the dissociation was calculated from the data of Vollmer.3 The others were taken from the preceding tables. TABLE 34. for potassium and sodium iodides in water and methyl and ethyl alcohols at 25°. H,O a;=3.68 constant. CH3OH z=3.43 constant. C,H5OH af= 2.74 constant. KI . . 249 25 3 249 Nal (v = 32, 64, 100) . . (v = 64, 128) . . . KBr (v = 128, 256) . . . 23.5 24.5 25.8 23.7 25.1 26.2 23.6 • • • Similarly, assuming that Dutoit and Aston's hypothesis holds for 50 per cent methyl alcohol, we may calculate the degree of association. Taking the 1 Compt. rend., 125, 240 (1897). 2 Ztschr. phys. Chem., 12, 433 (1893). 3Wied. Ann., 62, 328 (1894). 58 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. value of the constant as 23.6 and the comparable volume for the mixture as 48. from the relation - = 23.6 we can find x. We have — — = 23.6, whence x x a; = 3.96. A mixture of methyl alcohol and water, containing 50 per cent methyl alcohol by volume, has very approximately the composition corresponding to the hydrate CH3OH . 3 H2O. The existence of alcoholic hydrates has been made probable on other grounds. It is possible that such hydrates, in virtue of their complexity, have high dis- sociating power. The greater dissociation found in the 50 per cent mixture may be due to this hydrate, in which four simple molecules combine to form a complex molecule. CAUSE OF THE MINIMUM. The first observers of the conductivity minimum, Zelinsky and Krapiwin, offered no satisfactory explanation of it. They suggested that it might be connected with the formation of hydrates of methyl alcohol. Further than this they did not go. Jones and Lindsay would explain the existence of the minimum as due to the effect of one associated solvent on the association of another associated solvent (see p. 41). The explanation offered by Jones and Lindsay was later strengthened by an investigation by Jones and Murray. They showed, from a study of the freezing-points of solutions of acetic and formic acids, and water — acetic acid in formic acid, formic acid in acetic acid, acetic acid in water, etc. — that the association of one solvent is apparently diminished by the presence of another associated solvent. According to the explanation offered by Jones and Lindsay, the chief cause producing the minimum is a diminution of the dissociation of the dissolved substance, due to a diminution of the association of the solvents, and, conse- quently, a decrease in conductivity. We have shown that, in the 50 per cent mixture of methyl alcohol and water, the dissociation, instead of being di- minished by the presence of the alcohol (or by bringing together water and the alcohol), is actually increased. This fact alone makes it evident that the explanation offered by Jones and Lindsay does not account wholly for the phenomenon. Two factors determine conductivity — amount of dissociation and ionic mobility. Decrease in one or both of these produces decrease in conductivity. It has been shown that the decrease in conductivity in question can not be due alone to decrease in dissociation. The inevitable conclusion is, then, that it is due to a decrease in ionic mobility. A complete explanation of the minimum in conductivity will have to ac- count for the following facts : (1) The effect itself. CAUSE OF THE MINIMUM. 59 120 100 -MS 80 CO 40 -I 20- IV at II 0 at III 0 at IV at 100 (2) The fact that the effect is more pronounced at a lower temperature than at a higher. (3) The fact that rise in temperature (and in some cases increase in con- centration) shifts the min- imum towards a mixture containing a larger per cent of alcohol. There is a close connec- tion between the viscosity or fluidity of a solvent and the conductivity of electro- lytes when dissolved in that solvent. The greater the fluidity, or the less the viscosity, other things being equal, the greater is the conductivity. This close relationship is shown by the fact that for cer- tain aqueous solutions the temperature coefficients of conductivity and of fluid- ity are identical. The connection between conductivity and fluidity, or viscosity, will be considered in detail in the concluding part of this section. The investigations of Jones and Lindsay, Zelin- at o ° sky and Krapiwin, and this II at 10° III

— A//,, A0 — A/it, A<£ — AM,, A A0 A0 Atf. 128 0.287 0.258 0.263 0.339 256 .282 .247 .258 .333 512 .239 .231 .232 .283 1024 .239 .231 .232 .262 2048 .239 .231 .232 .262 64 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. It is evident that the effect of variation of fluidity on conductivity is great- est in the 40 per cent mixture, for here the values of the quotients are least. It is seen that the effect increases with increasing dilution, and finally becomes constant. Similar comparisons are given for potassium iodide in mixtures of methyl alcohol and water at 0° and at 25°, using the results of Jones and Lindsay. TABLE 41. — Variation (percentage fall) in conductivity. V Potassium iodide in mixtures of methyl alcohol and water at 0° and at 25°. 20 p AA .ct. 40p.ct. A/Xv SOp.ct. 65 p. ct. A^u 80 p. ct. CH3OH Aju,, 0° 25° 0° 25° 0° 25° 0° 25° 0° 25° 64 128 256 512 1024 6.361 .365 .366 .361 .543 0.237 .257 .262 .262 .262 0 479 507 509 510 503 695 0.356 .368 .377 .379 .379 0.494 .510 .502 .513 .508 .695 0.372 .386 .394 .399 .402 0.456 .478 .486 .489 .490 .682 0.350 .360 .371 .381 .386 .... 0.267 .283 .300 .313 .320 Lithium nitrate in mixtures of methyl alcohol and water at 0° and 25° (from Jones and Lindsay's data). Sodium iodide in mixtures of methyl alcohol and water at 0° and 25°. V 25p.ct. 50 p. ct. A/t, 75 p. ct.CH3OH A/t* 25 p. ct. 50p.ct. 75 p. ct.CH3OH 0° 25° 0° 25° 0° 25° 0° 25° 0° 25° 0° 25° 32 64 128 256 512 1024 0.406 .417 .432 .436 .441 .434 .600 0.285 .296 .307 .314 .310 .300 .456 0.508 .518 .532 .535 .541 .541 .695 0.384 .392 .403 .406 .408 .411 .575 0.432 .443 .466 .472 .477 .477 .661 0.333 .345 .358 .362 .354 .354 .481 0.400 .406 .409 .408 .409 0.277 .283 .297 .313 .321 0.485 .505 .515 .508 .500 0.357 .378 .395 .395 .402 0.411 .419 .423 .433 .446 0.298 .298 .318 .330 .333 TABLE 42. — Comparison of variation in fluidity and in conductivity. V Potassium • A' A A0 — A/*u A<£ 20 p. ct. A0 — A/*,, 40 p. ct. A0 — Afj.v 50 p. ct. A0 — AjOly 65 p. ct. A0 — A/jLv 80 p. ct. CHsOH A0— A/ty A<£ A A<£ A<£ A<£ 0° 25° 0° 25° 0° 25° 0° 25° 0° 25° 64 128 256 512 1024 6.335 .325 .326 .335 0.354 .300 '.289 .286 0.311 .275 .268 .266 .276 0.309 .285 .264 .264 0.289 .266 .261 .262 .270 0.304 .276 • • * • .251 .246 0.331 .299 .300 .282 .282 0.329 .321 '.28 i .272 0.309 0.385 .346 '.279 .262 VARIATION IN FLUIDITY AND CONDUCTIVITY. 65 TABLE 42. — Comparison of variation in -fluidity and in conductivity. — Continued. V Lithium nitra Arf> — A/UU to at 0° and at 25°. Sodium iodide. te A > a.i A<£ 25 p. ct. A0 — A/Uv 50 p. ct. A — Aju,, 75 p. ct. CH3OH A — A/J.V 25 p. ct. A0 — A/*,, 50 p. ct. A<£— Afj.v 75 p. ct. CH3OH. A0 — A/JT, A0 A0 A0 A0 A A0 0° 25° 0° 25° 0° 25° 0° 25° 0° 25° 0° 25° 32 64 128 256 512 1024 0.323 .305 .280 .273 .265 .277 0.375 .351 .327 .311 .318 (.344) 0.269 .255 .235 .230 .222 .222 0.332 .318 .300 .294 .290 .285 0.345 .330 .294 .284 .277 .277 0.308 .283 (.255) (.247) .264 .264 0.333 .323 .320 .320 .320 0.392 .379 .348 .313 .296 0.302 .273 .260 .269 .281 0.379 .343 .318 .313 .301 0.378 .366 .360 .345 .325 0.384 .355 .339 .314 .308 Finally, in table 43 we give a comparison of the temperature coefficients of conductivity for various electrolytes and the temperature coefficients of fluidity for the various mixtures. The volumes were 1024 for KI, LiNO3, SrI2; 256 for'Nal; 128 for CdI2; 256 for Ca(N03)2. The temperature coefficients for the ethyl alcohol mixtures are seen to be most nearly equal in the 65 per cent mixture for potassium iodide; for the other salts in the 50 per cent mixture. The temperature coefficients of the mixtures were calculated from Noack's data. It is evident from table 43 that for the four salts KI, LiN03, SrI2, and Ca(N03)2, in most mixtures, the quotient — is constant to within 10 per cent

- 0° . . . . -25° . . . 1 A0 55.39 112.00 0.0249 22.8 60.7 0.0250 14.15 41.77 0.0264 25.6 55.1 0.0216 55.5 85.88 0.0142 0-25° ' At 1 AM ... M25° A t KIO = 1024) . LiNO8(t> = 1024) Sri (v = 1024) . Ca(NO3)2(v = 256) 0.0172 0.0184 0.0272 0.0182 0.0237 0.0251 0.0247 0.0245 0.0135 0.0138 0.0199 0.0131 0.0233 0.0210 66 CONDUCTIVITY AND VISCOSITY IX MIXED SOLVENTS. TABLE 43. — Comparison of temperature coefficients of conductivity and fluidity for various electrolytes. — Continued. In methyl alcohol-water mixtures. 25 p. ct. 40 p. ct. 50 p. ct. 65 p. ct. 75 p. ct. 100 p. ct. CH3OH 0-0° . . . 4> - 25° . . . 1 A0 025° A£ 1 A/*, KI 30.6 69.04 0.0223 27.0 63.4 0.0230 0.0205 27.8 60.65 0.0216 0.0197 0.0202 0.0203 0.0210 0.0220 34.5 68.9 0.0200 0.0187 39.75 81.83 0.0205 131.3 — 3.8° 182.1 — 25.4° 0.0123 0.0121 0.0117 0.0118 0.0104 A A •""•• M25° At Nal .... LiNO3 . . . SrL(v=256). . CdI2 .... 0.0202 0.0211 0.0213 0.0238 0.0179 0.0182 0.0177 0.0178 DISCUSSION OF RESULTS. The value of the quotient Ac/. - is a measure of the parallelism between the two phenomena — decrease in conductivity and decrease in fluidity. If the decrease were the same in both cases, other conditions being the same, the value of the quotient would be zero. The fact that it is not zero indicates that the decrease in ionic mobility resulting from the decrease in fluidity is not proportional to the latter. When we come to compare the effect in the case of potassium iodide in mixtures of the two alcohols and water (tables 37 and 38), it is seen that the effect of decrease of fluidity on ionic mobility is greatest in the ethyl alcohol mixtures, or the two effects are here most nearly parallel. The effect is less for potassium iodide in methyl alcohol mixtures at both temperatures of ob- servation, 0° and 25°. It is to be remembered that we are leaving out of account possible differences of dissociation. Increase in dissociation (in the mixture) over that of the corresponding aqueous solution would diminish the effect of decrease in fluidity. Apparently, this possible change in disso- ciation can not be very great, as some of our measurements show. A far greater and entirely impossible change in dissociation would be necessary to account for the difference in the two effects. The minimum is much more pronounced in the methyl alcohol mixtures. It occurs, however, in the ethyl alcohol mixtures generally at 0°, but is not very marked. We have just seen that the real effect is greater in the latter case, when we make a proper comparison. The reason why it is not so evi- dent in the case of the ethyl alcohol mixtures is to be found in the small con- ductivities in ethyl alcohol; these, in turn, as will be shown in the last part of this section, being small, on account of the relatively great viscosity of ethyl alcohol and its rather small dissociating power. On the other hand, DISCUSSION OF RESULTS. 67 the phenomenon does exhibit itself in the other mixtures, and this is because of the high conductivities in methyl alcohol, these being high on account of the small viscosity of methyl alcohol and its relatively great dissociating power. Considering all of the salts in the various mixtures, it is seen that, in general, the effect of increased viscosity on conductivity is greatest in that mixture in which the minimum in conductivity occurs. In some cases the maximum effect occurs elsewhere. For example, for lithium nitrate (table 40) at 25° the maximum effect is in the 75 per cent mixture, while the minimum in conductivity occurs for the most part in the 50 per cent mixture. The explanation of this is found in the fact that, although in the one mixture the effect is greater, it is offset by the effect of the smaller fluidity of the other mixture — that of the 75 per cent mixture being 81.8, and that of the 50 per cent mixture being only 68.9. In cases where the minimum shifts with increase in dilution, the probable explanation is to be found in variation in increase of dissociation accompanying further dilution. The result of variation in the temperature is also shown in the tables — particularly in the case of the lithium salt (table 40). At 0° the maximum effect is in the 50 per cent mixture; at 25°, in the 75 per cent mixture. In table 43 is given a comparison of the temperature coefficients of fluidity and of conductivity for various electrolytes in the various mixtures of methyl alcohol and water, and also for ethyl alcohol mixtures, for which the data are more meager. From table 43 it is evident that the temperature coefficients of conductivity and fluidity do not differ markedly, particularly for some salts. For potas- sium iodide they are most nearly equal in the 65 per cent mixture, differing by only 7 per cent. For the other salts in the 50 per cent mixture the agree- ment is closest, and the differences are in no cases greater than for potassium iodide in the 65 per cent mixture. In some instances the agreement is seen to be much closer (SrI2, CdI2). For the ethyl alcohol mixture of 50 per cent, the temperature coefficients differ to about the same degree as in the methyl alcohol mixtures. In other words, — = constant. These facts are significant, as will appear from the latter part of this section. The conclusion which can be drawn is, then, that the decrease in conductivity of electrolytes in binary mixtures of various alcohols and water, which is in some cases accompanied by the minimum conductivity observed by Zelinsky and Krapiwin, Jones and Lindsay, and ourselves, is caused primarily by a diminution in the fluidity of the solvent, and a consequent decrease in ionic mobility, and secondarily by the effect of one associated solvent on the association of another such solvent. 68 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. VISCOSITY AND CONDUCTIVITY. The viscosity of a liquid or solution is denned as being the force (in dynes) necessary to move a layer of the liquid or solution one molecule in thickness, and of unit area (1 sq. cm.) over another layer of the liquid, with unit velocity (1 cm. per sec.). The symbol 77 is used for the coefficient of viscosity. The fluidity of a liquid is the reciprocal of its viscosity = -• ~n It is plain that the hypothesis of Dutoit and Aston does not go very deeply into the phenomena; all that it asserts is that there is a parallelism between the amount of dissociation effected by the solvent and the amount of its own association. The Thomson-Nernst hypothesis, however, offers an explana- tion of more profound significance. Dutoit and Friderich l make an attempt to find a connection between con- ductivity, viscosity, and association, and thus take a step in the direction of a more general hypothesis. From a study of the conductivities of solutions of different electrolytes in different solvents they came to the following conclusion : The values of M°° for a given electrolyte dissolved in different solvents are a direct function of the degree of polymerization of the solvents, and an indirect function of the coefficient of viscosity of these solvents. The relation was found to hold only in a general way — indeed, hardly more than qualitatively. When we come to consider the proposed relation, it is difficult to see why it should exist ; it is wholly empirical, and it is not in accord with the hypothe- sis of Dutoit and Aston ; for when complete dissociation is reached the asso- ciation of the solvent is no longer an influencing factor. It is superfluous to discuss the matter further because, as stated, the relation does not hold quantitatively. This will suffice to indicate what has thus far been done to show a connec- tion between conductivity and viscosity. The relations hitherto brought to light are qualitative. Of far more importance is the establishment of a quantitative relationship. We shall show that such does exist, and propose the following hypothesis : The conductivities of comparable, equivalent solutions of binary electrolytes in certain solvents (methyl and ethyl alcohols, other alcohols of the same series, acetone, etc.} are inversely proportional to the coefficient of viscosity of the solvent in question, and directly proportional to the association factor of the solvent. In case the hypothesis of Dutoit and Aston does not hold for the solvent in question, for " association factor of the solvent" must be substituted " amount of dissociation of the solution." 1 Bull. Soc. China., [3] 19, 321 (1898). VISCOSITY AND CONDUCTIVITY. 69 II KIinC2H5OH III Li NO 3 IV LiCl V Fluidity Curve IV Formulated, the hypothesis is expressed by the relation ^ = constant, x or ^ = constant, where the symbols have the usual significance. This be- et comes, when fiv = p.oo, /*„ 17 = constant. The meaning of the term "comparable equivalent solutions" needs to be defined. In comparing aqueous solutions those of the same normality (con- taining equal gram-molecules of electrolyte in equal volumes) are strictly com- parable. It is evident that this is not the case when, for example, we come to compare solutions of the same electrolyte in different solvents. In order to be strictly comparable, the so- lutions must contain the same number of gram-molecules of electrolyte dissolved in the same number of gram-mole- cules of the different solvents, or equal weights of the (same) electrolyte dissolved in volumes of the solvents which are proportional to the molecular volumes of the solvents in question. It is obvious that this is the only proper basis of com- parison. To illustrate, comparable solutions in water and in methyl alcohol would be those containing the same weight of electrolyte dis- solved in 18 volumes of water and 40 volumes of methyl alcohol, because the molecular volume of water is 18, and that of methyl alcohol approximately 40. The volumes compared should always be in the ratio 18 (of water) to 40 (of methyl alcohol). Similarly, in the case of methyl and ethyl alcohols, the volumes would be as 40 to 57.5. In the first place we have plotted (fig. 15) the variation in the fluidity of methyl and ethyl alcohols with temperature, making the different fluidities ordinates and the different temperatures abscissae. Plotted with these, for the sake of comparison, are the conductivities of various binary electrolytes 33 o 30- 20- 10 I Fluidity Curve II LiN03 V = 256 III Nal v=256 IV NH4Br J>=256 V KI v=25G VI LiCl 10 20 30 Temperature FIG. 15. — COMPARISON OF CONDUCTIVITY AND FLUIDITY. 70 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. in solution in the two solvents. Since only two observations were made, one at 0° and the other at 25°, the variation in conductivity is shown by a straight line. If observations at intermediate points could be obtained, the line would probably be curved. The first ordinate — that at 0° — is made the same in both cases. The second in the case of the conductivities is then found from the observed value by multiplication by a factor. From an inspection of fig. 15 it is evident that the lines representing va- riation of conductivity and variation of fluidity with temperature are almost coincident. That is, the temperature coefficients are almost the same in both, the difference never being over 8 per cent, and in most cases much less than this, particularly for solutions in ethyl alcohol. Vollmer has already called attention to this fact. This amounts, then, to a proof of the Kohlrausch hypothesis for solutions in methyl and ethyl alcohols, and such proof is necessary to establish the validity of the one proposed. It has already been shown that this is also true for certain electrolytes in certain mixtures of solvents. Though the temperature coefficients of fluidity and conductivity are the same, the value of the quotient — is much greater Of for an electrolyte in the mixture than for the same electrolyte in the pure alcohol. It is difficult to see why this is the case, unless it be that in the mixture we have a complex solvent not to be compared with the simple one. Certainly the presence of compounds in the mixture would complicate mat- ters. This interesting point needs further investigation. Knowing the values of the constant for the pure solvent, and assuming the validity of the relation for the mixture, we might calculate the ionic friction. This would not be the same as the coefficient of viscosity of the mixture, but its variation with temperature would be identical with that of con- ductivity. The next step is the discussion of certain data as to conductivities in the two solvents — proof that the expression — holds for solutions in these solvents. The values for the coefficients of viscosity have for the most part been taken directly, or interpolated from the results of Thorpe and Rodger.1 The values for the conductivities in the various solvents are taken from observations of Vollmer,2 Carrara,2 Jones and Lindsay,2 and others, and from our own. The values for the association factors are those given by Ramsay and Shields in their first paper,3 at ordinary temperatures for methyl alcohol 3.43, for ethyl alcohol 2.74. 1 Phil. Trans., 185 A, 397 (1894). 2 Loc. cit. 3 Ztschr. phys. Chem., 12, 433 (1893). VISCOSITY AND CONDUCTIVITY. 71 TABLE 44. — — in methyl and ethyl alcohols. For lithium nitrate. For ammonium bromide.1 CH3OH C2H5OH CH3OH C2Hr,OH Volume. 0° 25° Volume. 0° 25° 25° 0° 128 256 512 0.1288 .1353 .1484 0.1238 .1339 .1385 191 381 763 0.1239 .1368 .1500 0.1211 .1323 .1402 0.1369 .1441 .1513 0.1293 .1416 .1531 Volume. For potassium iodide. CH3OH 25° (Jones & Lindsay). CH3OH 25° (Carrara) . CHsOH 0° (Jones & Lindsay). C2H5OH 25° (Jones & Lindsay). C2H5OH 0° (Jones & Lindsay). f-v'n a CH3OH M»»» C2H5OH I II III IV 0.1377 .1471 .1558 .1636 0.1308 .1408 .1466 .1513 0.1417 .1526 .1617 .1665 0.1298 .1428 .1544 .1698 0.1359 .1479 .1592 .1760 0.516 .525 .539 .539 0.571 '.57 i .571 1 The volumes of comparison are the same as those in the case of lithium nitrate. The conductivities compared were, for the volumes 128, 256, 512 for methyl alcohol, 190.7, 381.4, 762.8 for ethyl alcohol, these being comparable dilu- tions. In the case of solutions in ethyl alcohol values for the conductivities were found by interpolation. As is seen from inspection of table 44, the constants for the comparable volumes are equal to within a few per cent. Further, the constant is the same at the lower and at the higher temperature. Of course the constants are not the same for the different volumes, for here we have used the associa- tion factor as a constant in the equation ; it represents dissociation. If we were to substitute per cent of dissociation for the association factor, the values for the constant would be the same in all cases. All together, the agreement of theory and fact is all that could be expected, when we consider the errors involved in the determination of the quantities used. The figures for association are certainly only approximate. Conduc- tivities are liable to an appreciable error, and different observers give values for the viscosity coefficients differing by as much as 4 or 5 per cent. Values for the conductivities in ethyl alcohol are taken from the work of Jones and Lindsay; those given by them for methyl alcohol solutions are not used. Those of Carrara are taken, as they agree with values given by Zelinsky and Krapiwin, while those of Jones and Lindsay do not. In all the cases the agreement is as close as could be expected. The volumes compared were 64, 128, 256, and 512 for methyl alcohol; for ethyl alcohol, 95.7, etc. The values found by both Jones and Lindsay and 72 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. Carrara, for conductivities in methyl alcohol, were used in order to compare the results. In some cases those of the one give better agreement, in other cases those of the other, indicating that the differences are due to experi- mental errors. It should be stated that in this case Gartenmeister's values for the coeffi- cients of viscosity were used. The calculations were made before it was decided to use those given by Thorpe and Rodger. It would hardly be profit- able to recalculate, since the result is the same whichever set of values be employed. We have satisfied ourselves of this by numerous trials. In many cases three or four calculations were made, using the different constants given in the tables of Landolt and Bornstein. TABLE 45. — — for lithium chloride in methyl, ethyl, and propyl alcohols. x CH,OH, 25° 0.1213 C2H5OH 18° .1166 .1256 Vollmer— -18° C3H,OH 15° 1164 The volumes compared in table 45 are 256, 403, 604. The values for the conductivities are taken from the work of Carrara and Vollmer, and for n-propyl alcohol from Schlamp.1 The agreement is satisfactory. It is interesting to note that the relation holds for picric acid in solution in methyl and ethyl alcohols. Where the volumes compared are 205 and 270, respectively, — in the one case equals 0.0665, and in the other, 0.0641. The •I/ data for the comparison are furnished by Schall.2 Other results may be summarized (table 46) without further detail. the instances given the agreement continues to be satisfactory. In all TABLE 46. -- --for various electrolytes in different solvents. CH3OH C2H5OH n- CsH7OH Nal 0 1438 0 1447 f Jones & Carroll (25°). j v = 200, 283 .... \Vollmer (18°). v = 750, 575 .... .157 0.159 /Jones & Carroll (25°). CHsCOONa .... .1156 I Schlamp (15°). Vollmer (18°) v - 832, 1200 .... .1123 1093 Carrara (25°) • Vollmer (18°) CH,COOK NaCl . .1279 1247 Vollmer (18°) B v = 228. 340 .... .1457 .1366 Carrara (25°) ; Vollmer (18°) 1 Landolt und Bernstein's Tabellen. 2 Ztschr. phys. Chem., 14, 707 (1894). SUMMARY AND CONCLUSIONS. 73 In the foregoing pages we have discussed all the material available in the literature. In all of the cases — some nine in number — fact and theory are in accord. It can, therefore, be fairly claimed that the proposed hypoth- esis becomes highly probable. Further investigation is, however, desirable, to see how widely it applies. For solutions in which dissociation is complete the formulated hypothesis becomes /«; = constant. The proof of the validity of this relation is, it seems to us, a crucial test. Table 47 gives the necessary comparisons. TABLE 47. — M^ /or various electrolytes in various solvents. Electrolyte. C,Hr>OH n = 0.012385 18° CH,OH n = 0.00666 (18°) 0.00552 (25°) CH,COCHS n = 0.00353 25° n-C,H7OH n = 0.002554 15° KI ... 0 567 0 530 — 18° 0.541 Nal 466 .528 — 25° .505—18° .494 0.481 NH4I .496 — 25° .581 .538 .431 LiCl .377 .416 — 18° NaCl .434 .427—25° .479 The values for acetone were taken from the work of Carrara,1 those for ethyl alcohol from that of Vollmer,2 those for methyl alcohol from that of Vollmer2 and Carrara,2 and those for propyl alcohol from the investigation of Schlamp. When we consider the necessarily large experimental error involved in the determination of the limiting values for conductivity — an error which must certainly be greater than that involved in the determination of conductivi- ties at ordinary dilutions — the agreement is as good as could be expected. Further, for lithium and sodium chlorides in ethyl alcohol, limiting values were probably not reached, for with them Vollmer did not go to a dilution as high as in other cases. The true values for these salts would probably be greater than the values given in the table, thus making a still better agreement. SUMMARY AND CONCLUSIONS. (1) The investigations of Zelinsky and Krapiwin and of Jones and Lindsay have been extended, and the occurrence of the minimum in conductivity has been shown for three substances, cadmium iodide, sodium iodide, and hydro- chloric acid, in mixtures of methyl alcohol and water. (2) The dissociation (as determined from conductivity) of sodium and potassium iodides, and potassium bromide in 50 per cent methyl alcohol, has been determined and has been found to be greater than that in water at the corresponding dilution. 1 Gazz. Chim. Ital., 27 (1), 207 (1897). 2 Loc. cit. 74 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. (3) It has been shown that the explanation of the minimum in conductivity offered by Jones and Lindsay is not sufficient. The phenomenon has been shown to be dependent primarily upon the decrease in fluidity which results when the components of the solvent are mixed. (4) The hypothesis of Dutoit and Aston has been proved quantitatively for certain salts in three solvents — water, methyl alcohol, and ethyl alcohol. (5) The hypothesis of Kohlrausch (formation of an atmosphere of the solvent around the ions in solution) has been shown to hold for binary electrolytes in methyl and ethyl alcohols. (6) A hypothesis correlating conductivity, association, and viscosity (or fluidity) has been proposed, and has been shown to hold for all the cases available for discussion. WORK OF BASSETT. EXPERIMENTAL WORK. This investigation was undertaken for the purpose of determining what effect mixtures of methyl alcohol and water would have on the relative velocities of the ions of such a salt as silver nitrate. The work of Jones and Lindsay 1 on the conductivity of certain salts in water, methyl, ethyl, and propyl alcohols, and mixtures of these solvents, sug- gested this work. In their work, Jones and Lindsay found that the conductivities of such salts as potassium iodide, ammonium bromide, strontium iodide, etc., were less in mixtures of the solvents than in either of the solvents alone. Especially was this the case in mixtures of methyl alcohol and water. Considering these facts, the first thing to determine was whether silver nitrate would give similar conductivity results, and if so, whether there was any relation between this phenomenon and the relative velocities. The conductivities of silver nitrate in these solvents and varying mixtures of them were determined. The water and methyl and ethyl alcohols were purified by the methods de- scribed by Jones and Lindsay. In each case a mother-solution was made in the solvent in question, and the remaining solutions were obtained by successive dilutions with some of the solvent of the same composition. In this way an error was avoided which would result from the contraction when alcohol and water were mixed, and also prevent the accompanying heat effect. In some cases, as in very dilute solutions, where such small quantities of the mother- solution were required, a second mother-solution was made from the first, and the more dilute solutions made from it in the way described. The strength of these mother-solutions was determined by titrating with a standard solution of ammonium sulphocyanate. CONDUCTIVITY APPARATUS EMPLOYED. The apparatus described and used was similar to that employed by Jones and Lindsay. The cells differed from the ordinary Arrhenius cell, being pro- vided with a ground-glass top to prevent evaporation of the more volatile solvents, and also to protect the anhydrous alcoholic solutions from the moisture of the baths and air. The glass tubes carrying the electrodes were passed through thin rubber tubes in the cap. Sealing-wax was then run over the outside of the joint. 'Amer. Chem. Journ., 28, 329 (1902). 75 76 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. The zero-bath was prepared as follows: A large glass battery-jar was filled with finely crushed ice and distilled water. It was then placed in a water- bath and the space between filled with finely crushed ice and water. This proved very efficient, as it was possible to keep within 0.05° of zero for hours. The bath at 25° was of the ordinary form, and was kept in constant motion by a stirrer driven by means of a hot-air engine. The thermometers used could be accurately read to 0.02°. The burettes and flasks were carefully calibrated. CONDUCTIVITY MEASUREMENTS. All the conductivity measurements were made at the two temperatures, 0° and 25°. In tables 48 to 50, v = number of liters of solution containing a gram-molecular weight of the salt; /xyO° = molecular conductivity at 0°; /u.,,250 = molecular conductivity at 25°. TABLE 48. — Molecular conductivity of silver nitrate. In water. In ethyl alcohol. In methyl alcohol. V V ^°° ^25° Mi-O0 ^25° 1^0° M»25° 10 55.72 99.46 9.71 7.11 10 25.96 35.77 20 58.63 105.72 19.43 8.90 14.26 20 32.63 44.67 40 63.10 110.22 38.86 11.35 16.96 40 39.71 53.42 80 65.38 115.81 77.73 13.10 20.11 80 45.28 62.95 160 65.38 119.86 155.47 15.13 23.87 160 51.09 70.36 320 69.91 12508 310.95 17.04 26.46 320 56.71 80.17 640 71.05 125.86 621.89 19.43 30.62 640 61.42 88.22 1280 70.59 125.35 In 25 p. ct. ethyl alcohol In 50 p. ct. ethyl alcohol In 75 p. ct. ethyl alcohol and water. and water. and water. V M,0° Mt'25° MrO° M,25° ^0° ^25° 19.43 15.25 37.87 13.12 27.01 38.86 25.95 59.70 15.81 39.50 14.30 30.43 77.73 26.65 62.75 17.04 42.42 16.79 33.65 155.47 26.45 63.82 17.92 45.15 19.48 35.94 310.95 28.72 64.87 19.10 47.13 18.00 39.01 621.89 28.84 68.28 19.90 49.39 19.41 40.14 1243.78 .... .... 20.79 52.80 .... .... In 25 p. ct. methyl alcohol In 50 p. ct. methyl alcohol In 75 p. ct. methyl alcohol and water. and water. and water. V M.QO M,25° ju,.0° M,.25° ^0° *,25° 20 27.27 53.33 27.98 48.20 40 35.63 72.68 28.63 56.80 30.03 52.33 80 36.95 75.56 29.93 59.75 32.81 57.17 160 39.03 79.34 31.47 63.22 35.22 61.31 320 41.03 82.93 32.29 65.85 35.71 63.23 640 41.23 83.91 34.67 68.67 40.27 69.42 SILVER NITRATE. 77 TABLE 49. — Temperature coefficients of conductivity of silver nitrate. In mixtures of methyl alcohol and water V In water In methyl alcohol of various compositions. (0°to25°). (0°to25°). 25 p. ct. 50 p. ct. 75 p. ct. 10 1.75 0.392 20 1.88 0.482 • • • • 1.04 0.810 40 1.88 0.548 1.48 1.13 0.892 80 2.10 0.707 1.54 1.19 0.974 160 2.14 0.771 1.61 1.27 1.024 320 2.21 0.938 1.68 1.34 1.100 G40 2.20 1.702 1.71 1.36 1.166 1280 2.19 .... • • • • • • • • • • • • In mixtures of ethyl alcohol and water of V In ethyl alcohol various compositions. 25 p. ct. 50 p. ct. 75 p. ct. 19.43 0.214 0.905 0.556 38.86 .224 1.35 0.962 .645 77.73 .280 1.44 1.015 .674 155.47 .350 1.49 1.089 .658 310.95 .377 1.45 1.121 .840 621.89 .488 1.58 1.180 .829 1243.78 .... .... 1.280 .... TABLE 50. — Comparison of the molecular conductivities of silver nitrate. In ethyl alcohol and mixtures of it with water at 25° C. and 0° C. V At 25° C. At 0° C. 25 p.ct. 50 p.ct. 75 p.ct. 100 p.ct. 25 p. ct. 50 p.ct. 75 p.ct. 100 p.ct. alcohol. alcohol. alcohol. alcohol. alcohol. alcohol. alcohol. alcohol. 9.71 7.11 19.43 • • • • 37.87 27.01 14.26 15.25 13.12 890 3886 59.70 39.50 30.43 16.96 25.95 15.81 14.30 11 35 77.73 62.75 42.42 33.65 20.11 26.65 17.04 16.79 13.10 155.47 63.82 45.15 35.94 23.87 26.45 17.92 19.48 15.13 310.95 64.87 47.13 39.01 26.46 28.72 19.10 18.00 17.04 621.89 68.28 49.39 40.14 30.62 28.84 19.90 19.41 19.43 1243.78 .... .... .... .... .... 20.79 .... .... In water, methyl alcohol, and mixtures of these solvents at 25° C. and 0° C. At 25° C. At 0° C. V Water. 25 p.ct. alcohol. 50 p.ct. alcohol. 75 p.ct. alcohol. 100 p.ct. alcohol. Water. 25 p. ct. alcohol. 50 p.ct. alcohol 75 p.ct. 100 p.ct. alcohol, alcohol. 10 99 46 35 77 55 72 25 96 20 105.72 53.33 48.20 44.67 58.63 27.27 27.98 32.63 40 110.22 72.68 56.80 52.33 53.42 63.10 35.63 28.63 30.03 39.71 80 115.81 75.56 59.75 57.17 62.95 63.16 36.95 29.93 32.81 45.28 160 119.86 79.34 63.22 61.31 70.36 65.38 39.03 31.47 35.22 5109 320 125.08 82.93 65.85 63.23 80.17 69.91 41.03 32.29 35.71 56.71 640 125.86 83.91 68.67 69.42 88.22 71.05 41.23 34.67 40.27 61.42 1280 125.35 70.59 78 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. 50 # 7554 "Percentage of Ethyl Alcohol 100^ FIG. 16. — CONDUCTIVITY OF SILVER NITRATE IN MIXTURES OF ETHYL ALCOHOL AKD WATER AT 25°. In order to see the connection existing between the conductivities in each solvent, table 50 is given for comparison. It is seen from the values in table 50 that the molecular conductivity in ethyl alcohol and mixtures with water does not show a minimum over the ordinary range of dilution, but still does not obey the law of mixtures. This is in accordance with the work of Jones and Lindsay on potassium iodide. In that case they did not find a trace of a minimum at 25°. The values in the first part of table 50 are plotted as curves in fig. 16, the abscissae representing the different per cents of alcohol and the ordinates the molec- ular conductivities. The curves in fig. 17 are of the same general form as those in fig. 16. No distinct minimum is shown, but the form of the curve indicates that a minimum value is approached. In nearly every case the minimum, as shown by the curves in fig. 18, SILVER NITRATE. 79 v= 38.80 v= T7.73 =* 155.47 V= 310.95 V= 021.89 5051 75; Percentage of Ethyl Alcohol FIG. 17. — CONDUCTIVITY OF SILVER NITRATE IN MIXTURES OF ETHYL ALCOHOL AND WATER AT 0°. I V = 40 II v =80 III v =100 50 2554 50# 75J Percentage of Methyl Alcohol FIG. 18. — CONDUCTIVITY OF SILVER NITRATE IN MIXTURES OF ML.THYL ALCOHOL AND WATER AT 25°. 80 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. lies between mixtures of the solvents containing respectively 50 and 75 per cent of alcohol. The curves in fig. 19 are of the same general form as those in fig. 18. They differ, however, in some respects. The chief difference is that the minimum point has shifted to the left, corresponding now to an alcohol- water mixture not far from 50 per cent. 100 # Percentage of Methyl Alcohol FIG. 19. — CONDUCTIVITY OF SILVER NITRATE IN MIXTURES OF METHYL ALCOHOL AND WATER AT 0°. From the above results it is clear that silver nitrate is not an exception to the general relation found by Jones and Lindsay. WORK OF BINGHAM. EXPERIMENTAL. APPARATUS. CONDUCTIVITY. The Kohlrausch method of measuring conductivity, with Wheatstone bridge, telephone receiver, and induction coil, was employed. It was not difficult to read to less than 0.1 of 1 per cent. The bridge-wire was made of "manganin" and was calibrated before beginning the work. The resistance coils were carefully standardized. In order to work successfully with acetone, it was necessary to provide cells of special construc- tion, so as to avoid the presence of rubber or wax, which would be dissolved by the solutions. The cells were made of hard glass with ground-glass stoppers, and had the form shown in fig. 20. The glass tubes carrying the electrodes were sealed into both the upper and the lower walls of the glass stopper. The distance between the electrodes thus remained permanently fixed. The zero-bath was prepared by filling a large battery-jar with clean, finely crushed ice, moistened with water. This was placed in a pail made of compressed paper pulp, and the annular space filled with finely crushed ice. Thus protected, the bath could be used for a much longer time than with the methods usually employed. The 25° bath was of the usual form, the stirrer being driven by means of a hot-air engine. An Ostwald regulator was employed. The thermometers were regulated to tenths of a degree. They were tested at the beginning of the work. Burettes and flasks were care- fully calibrated. VISCOSITY. The viscometer was of the form recommended by Ostwald.1 A fixed volume of the liquid to be measured was introduced into the apparatus. The FIG. 20. Physiko-Chemische Messungen, 2d ed., p. 260. 81 82 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. liquid was raised exactly to the mark above the bulb, by air-pressure. The air was dried over sulphuric acid. The pressure was released by means of a Mohr pinchcock, on a thick-walled rubber tube. By this arrangement the readings agreed to within 0.2 of a second. It was important that the liquids should be given time to drain out of the tube above the upper mark, since, otherwise, a drop of liquid might collect in the upper capillary and impede the entering air, thus introducing large error. For the zero-bath a battery-jar was filled with very finely crushed ice moistened with water. The ice was renewed as often as was necessary to keep the temperature constant. For the •25° bath a 5-liter beaker was employed. The bath was stirred by means of a hot-air engine, the temperature being kept to within 0.1° of 25°, as in the conductivity method. The room was kept as near this temperature as possible. To measure the specific gravities at zero, which was necessary in order to calculate the viscosities, we constructed a pycnometer (fig. 21) which would allow the large ex- pansion of the alcohols and acetone and avoid loss by evaporation. PREPARATION OP SOLUTIONS. In making up mixtures of solvents, n c. c. of acetone diluted to 100 c. c. was designated FIG. 21. as a mixture of " n per cent acetone." Since acetone, especially, has a high coefficient of expansion, it was important to have the temperature always the same, 20°. The acetone or alcohol was brought to this temperature before making up the mixture. On mixing acetone and water, contraction took place and heat was generated, so that the mixture was brought to the temperature before diluting to the mark. The mother-solution was made by weighing into a measuring-flask the exact amount of salt required, and adding the mixed solvent. Since, however, heat was again generated, especially with the calcium nitrate, the solution was again brought to the designated temperature before diluting to the mark. From the mother-solution the other solutions were made by successive dilutions. Where this would necessitate the use of small quantities of solu- tion, a new mother-solution was made, and from this successive dilutions prepared. SOLVENTS. 83 SOLVENTS. WATER. The water was purified by the method of Jones and Mackay,1 and had a conductivity of 1 X 10~6 at 0°. METHYL ALCOHOL. The methyl alcohol was the best commercial article obtainable. It was boiled with calcium oxide for a day, distilled, and allowed to stand over anhydrous copper sulphate for a long time. Before use it was distilled, using a Linnemann fractionating head. Precautions were taken against ab- sorption of moisture. The first and last portions of the distillate were discarded, giving a liquid which boiled constantly at 66°. The mean value of the conductivity was 2 x 10~6 at 25°. ETHYL ALCOHOL. The ethyl alcohol was the best commercial alcohol obtainable. It was purified in the same manner as the methyl alcohol. Its conductivity had a mean value of 2 x lO"6 at 25°. ACETONE. The acetone was dried over fused calcium chloride for weeks and distilled with a fractionating head as above. Its conductivity was 0.6 X 10~6. CONDUCTIVITY MEASUREMENTS. In all determinations of conductivity at least three different resistances were used, and the values given are the mean. However, if the readings did not agree to 0.1 of 1 per cent, they were usually repeated. The constants of the cells were checked at frequent intervals. The cells were not allowed to remain in contact with the solution when not in use, nor to remain empty after being dried out with alcohol and ether. In the former case, small quanti- ties of salt were found to be slowly absorbed, and in the latter acetic acid was formed by the action of the platinum on the alcohol or ether in the presence of air. When not in use the cells were filled with pure distilled water. A N/50 and a N/500 solution of potassium chloride were used in determin- ing the cell constants. The conductivity of the former was taken as 129.7 at 25°. The value of the latter was determined several times in different cells. The mean value obtained agrees well with the interpolated values of other observers. The temperature coefficients are obtained by dividing the increase in the conductivity per degree by the conductivity at the lower temperature. 1 Ztschr. phys. Chem , 22, 237 (1897). Amer. Chem. Journ., 19, 91 (1897). CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 51. — Conductivity of N/ 500 potassium chloride at 25°. Observed value corrected for the conductivity of water 136.5 Corresponding value obtained by Jones and West l 135.5 Deduced value, from Kohlrausch 2 136.94 Interpolated value, from Ostwald 3 138.7 LITHIUM NITRATE. The lithium nitrate used in this work was a sample obtained from Kahlbaum. No appreciable impurity could be detected. It was dried in an air-bath at 150° until the weight remained constant. The salt was kept in a desiccator. The operation of drying was repeated whenever the salt was exposed to the air. TABLE 52. — Conductivity of lithium nitrate at 0° and 25°. In methyl alcohol. In a mixture of 25 p. ct. acetone and methyl alcohol. In a mixture of 50 p. ct. acetone and methyl alcohol. V o 0 Temp. a o Temp. Temp. M"° Mv coef. Mr Mr coef. Mi' M« coef. 5 31.17 43.05 0.0154 33.50 44.27 0.0128 32.79 41.26 0.0103 10 37.62 51.31 .0145 40.78 53.85 .0128 41.11 51.68 .0100 25 43.40 60.49 .0158 48.36 6463 .0134 51.38 65.39 .0109 50 48.31 67.2 .0157 54.29 72.8 .0136 58.82 75.42 .0113 100 52.0 72.6 .0155 58.9 79.8 .0148 65.2 84.8 .0120 200 55.1 76.8 .0147 63.6 85.9 .0140 69.6 93.2 .0135 400 56.8 80.0 .0172 66.2 90.6 .0147 75.2 99.2 .0128 800 59.6 83.7 .0162 69.8 95.7 .0148 80.0 105.3 .0127 1200 60.8 85.3 .0160 71.0 97.7 .0150 83.3 109.6 .0126 1600 61.9 86.7 .0160 73.1 99.8 .0146 86.8 112.2 .0117 V In a mixture of 75 p. ct. acetone and methyl alcohol. In acetone. In ethyl alcohol. 5 25.53 29.98 0.00703 7.78 9.25 0.00755 9.14 14.65 0.0241 10 34.06 39.45 .0158 9.67 10.87 .00493 10.75 17.17 .0238 25 45.86 53.69 .00683 11.35 12.86 .00532 13.55 21.71 .0240 50 55.50 66.30 .0078 14.07 15.64 .00447 15.54 24.9 .0241 100 65.2 78.4 .0081 18.1 19.5 .00310 16.9 27.6 .0251 200 74.8 92.2 .0093 23.8 25.3 .00252 18.4 30.3 .0260 400 82.6 103.0 .0099 30.6 32.4 .00232 19.1 32.1 .0271 800 90.0 113.3 .0103 43.4 45.5 .00193 20.3 34.1 .0251 1200 94.4 119.0 .0104 48.7 52.5 .00311 20.0 34.4 .0288 1600 96.6 123.6 .0112 55.3 59.8 .00325 21.7 35.4 .0252 In a mixture of 25 p. ct. ace- In a mixture of 50 p. ct. ace- In a mixture of 75 p. ct. ace- V tone and ethyl alcohol. tone and ethyl alcohol. tone and ethyl alcohol. 5 12.5 18.6 0.0195 15.2 19.7 0.0142 14.1 16.9 0.00795 10 15.9 22.9 .0176 19.6 25.4 .0119 19.3 22.5 .0066 25 20.2 29.2 .0178 26.1 33.8 .0118 27.2 31.3 .0060 50 23.1 33.9 .0187 30.9 40.4 .0123 34.3 39.7 .0063 100 26.0 38.4 .0191 35.7 47.6 .0133 41.7 49.6 .0074 200 29.0 43.2 .0196 40.4 54.8 .0142 50.2 60.1 .0079 400 31.2 46.6 .0198 44.4 61.3 .0152 54.2 70.8 .0122 800 33.0 50.3 .0210 47.4 67.0 .0165 64.9 80.9 .0098 1200 34.6 51.8 .0222 49.4 69.6 .0164 69.5 86.5 .0098 1600 35.1 54.5 .0220 50.7 71.9 .0167 70.6 91.7 .0119 1 Amer. Chem. Journ., 34, 357 (1905). 3 Lcitvermogen der Elektrolyten. 3 Lehrbuch der ed.f p. 732. allgcmeinen Chem., 2d LITHIUM NITRATE. 85 TABLE 52. — Conductivity of lilhium nitrate at 0° and 25°. — Continued. •71 In water. In a mixture of 25 p. ct. acetone and water. V 0 oro Temp. oro Temp. '*" Mt. coef. Mu Me coef. 5 44.45 80.72 0.0323 27.31 55.69 0.0415 10 46.39 83.87 .0323 27.37 56.35 .0421 25 49.57 91.4 .0334 30.32 62.76 .0427 50 51.4 94.4 .0334 31.50 65.3 .0430 100 52.5 97.0 .0340 32.8 70.5 .0460 200 53.1 98.8 .0342 34.1 72.5 .0450 400 54.3 100.8 .0341 34.6 76.0 .048 800 55.0 102.0 .0340 37.6 77.8 .043 1200 55.9 102.6 .0332 39.0 80.7 .0425 1600 56.3 102.8 .0330 40.0 83.1 .0430 (58.3) (107.0) In a mixture of 50 p. ct. acetone and In a mixture of 75 p. ct. acetone and V water. water. 5 21.81 43.37 0.0394 21.25 36.78 0.0292 10 23.48 47.84 .0415 24.41 42.65 .0298 25 24.90 51.33 .0425 27.64 48.92 .0308 50 26.26 54.4 .0427 30.25 54.14 .0316 100 27.2 57.3 .0442 31.8 57.4 .0322 200 29.1 60.0 .0425 33.5 61.4 .0333 400 28.8 60.0 .0433 35.1 63.6 .0325 800 29.9 62.7 .0440 36.7 66.3 .0322 1200 31.6 65.9 .0425 38.1 68.6 .0320 1600 32.3 67.5 .0435 37.8 69.1 .0331 TABLE 53. — Comparison of the conductivities of lithium nitrate. In mixtures of acetone and methyl alcohol. V AtO°. At 25°. 0 p. ct. 25 p. ct. 50p.ct. 75p.ct. 100 p. ct. Op.ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 5 31.17 33.50 32.79 25.53 7.78 43.05 44.27 41.26 29.98 9.25 10 37.62 40.78 41.11 34.06 9.67 51.31 53.85 51.68 39.45 10.87 25 43.40 48.36 51.38 45.86 11.35 60.49 64.63 65.39 53.69 12.86 50 48.31 54.29 58.82 55.50 14.07 67.2 72.8 75.42 66.30 15.64 100 52.0 58.9 65.2 65.2 18.1 72.6 79.8 84.8 78.4 19.5 200 55.1 63.6 69.6 74.8 23.8 76.8 85.9 93.2 92.2 25.3 400 56.8 66.2 75.2 82.6 30.6 80.0 90.6 99.2 103.0 32.4 800 59.6 69.8 80.0 90.0 43.4 83.7 95.7 105.3 113.3 45.5 1200 60.8 71.0 83.3 94.4 48.7 85.3 97.7 109.0 119.0 52.5 1600 61.9 73.1 868 96.6 55.3 86.7 99.8 112.2 123.6 59.8 V In mixtures of acetone and ethyl alcohol. 5 9.14 12.5 15.2 14.1 7.78 14.65 18.6 19.7 16.9 9.25 10 10.75 15.9 19.6 19.3 9.67 17.17 22.9 25.4 22.5 10.87 25 13.55 20.2 26.1 27.2 11.35 21.71 29.2 33.8 31.3 12.86 50 15.54 23.1 30.9 34.3 14.07 24.90 33.9 40.4 39.7 15.64 100 16.9 26.0 35.7 41.7 18.1 27.60 38.4 47.6 49.6 19.5 200 18.4 29.0 40.4 50.2 23.8 30.3 43.2 54.8 60.1 25.3 400 19.1 31.2 44.4 54.2 30.6 32.1 46.6 61.3 70.8 32.4 800 20.3 33.0 47.4 64.9 43.4 34.1 50.3 67.0 80.9 45.4 1200 20.0 34.6 49.4 69.5 48.7 34.4 51.8 69.6 86.5 52.5 1600 21.7 35.1 50.7 70.6 55.3 35.4 54.5 71.9 91.7 59.8 86 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 53. — Comparison of the conductivities of lithium nitrate. — Continued. V In mixtures of acetone and water at 0°. In mixtures of acetone and water at 25°. Op. ct. 25 p. ct. 50 p. ct. 75p.ct. 100 p.ct. 0 p. ct. 25 p.ct. 50 p. ct. 75 p. ct. 100 p.ct. 5 44.45 27.31 21.81 21.25 7.78 80.72 55.69 43.37 36.78 9.25 10 46.39 27.37 23.48 24.41 9.67 83.87 56.35 47.84 42.65 10.87 25 49.57 30.32 24.90 27.64 11.35 91.4 62.74 51.33 48.92 12.86 50 51.4 31.50 26.26 30.25 14.07 94.4 65.3 54.4 54.14 15.64 100 52.5 32.8 27.2 31.8 18.1 97.0 70.5 57.3 57.4 19.5 200 54.8 34.1 29.1 33.5 23.8 98.8 72.5 60.0 61.4 25.3 400 54.3 34.6 28.8 35.1 30.6 100.8 76.0 60.0 63.6 32.4 800 55.0 37.6 29.9 36.7 43.4 102.0 77.8 62.7 66.3 45.5 1200 55.9 39.0 31.6 38.1 48.7 102.6 80.7 65.9 68.6 52.5 1600 56.3 40.0 32.3 37.8 55.3 102.8 83.1 67.5 69.1 59.8 Table 53 (figs. 22 and 23) shows that lithium nitrate, in mixtures of methyl alcohol and acetone, gives a pronounced maximum in conductivity. ^ 50$ 75j* ^Percentage of Acetone 1000 FIG. 22. — CONDUCTIVITY OF LITHIUM NITRATE IN MIXTURES OF ACETONE AND METHYL ALCOHOL AT 0°. At high concentrations the maximum is rather small. As the dilution is in- creased the maximum appears in the 75 per cent mixture and even beyond. It should also be noticed that, in the dilute solutions, the rise in conductivity is directly proportional to the amount of acetone, up to the 75 per cent mixture. The maximum is increased by rise in temperature. The points will be made clear by a study of the figures. In all cases the curves represent the molecular conductivities at the successive dilutions. LITHIUM NITRATE. 87 120- .110 100- . 90- ? 80- I i I ; 60 I : 50- t I ! 40- 30- 20 10- 25$ 50# -75# "Percentage of Acetone FIG. 23. — CONDUCTIVITY OF LITHIUM NITRATE IN MIXTURES OF ACETONE AND METHYL ALCOHOL AT 25°. TABLE 54. — Comparison of the temperature coefficients of conductivity of lithium nitrate. In mixtures of acetone and methyl alcohol. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 5 0.0152 0.0129 0.0103 0.00703 0.00755 10 .0145 .0128 .0100 .00158 .00493 25 .0158 .0134 .0109 .00683 .00532 50 .0157 .0136 .0113 .0078 .00447 100 .0155 .0148 .0120 .0081 .00310 200 .0147 .0140 .0135 .0093 .00252 400 .0172 .0147 .0128 .0099 .00232 800 .0162 .0148 .0127 .0103 .00192 1200 .0160 .0150 .0126 .0104 .00311 1600 .0160 .0146 .0117 .0118 .00325 V In mixtures of acetone and ethyl alcohol. 5 0.0241 0.0195 0.0142 0.00795 0.00755 10 .0238 .0176 .0119 .0066 .00493 25 .0240 .0178 .0118 .0060 .00532 50 .0241 .0187 .0123 .0063 .00447 100 .0251 .0191 .0133 .0074 .00310 200 .0260 .0196 .0142 .0079 .00252 400 .0271 .0198 .0152 .0122 .00232 800 .0251 .0210 .0165 .00985 .00192 1200 .0288 .0222 .0164 .0098 .00311 1600 .0252 .0220 .0167 .0119 .00325 88 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 54. — Comparison of the temperature coefficients of conductivity of lithium nitrate. — Continued. V In mixtures of acetone and water. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 5 0.0323 0.0415 0.0394 0.0292 0.00755 10 .0323 .0423 .0415 .0298 .00493 25 .0334 .0427 .0425 .0308 .00532 50 .0334 .0430 .0427 .0316 .00447 100 .0340 .0460 .0442 .0322 .00310 200 .0342 .0450 .0425 .0333 .00252 400 .0341 .048 .0433 .0325 .00232 800 .0340 .043 .0440 .0322 .00192 1200 .0332 .0425 .0425 .0320 .00311 1600 .0330 .0430 .0435 .0331 .00325 Percentage of Acetone FIG. 24. — CONDUCTIVITY OF LITHIUM NITRATE IN MIXTURES OF ACETONE AND ETHYL ALCOHOL AT 0°. Table 54 (figs. 24 and 25) shows the same characteristics for lithium nitrate in mixtures of acetone and ethyl alcohol as those observed for the same salt in mixtures of acetone and methyl alcohol, but there is not such a well- defined maximum in these curves. Table 54 (figs. 26 and 27) for lithium nitrate, in mixtures of acetone and water, shows, at low temperatures and at high dilution, the minimum which is familiar under similar circumstances in mixtures of the alcohols and water. There is, however, even in this case, a tendency towards a maximum, which POTASSIUM IODIDE. 89 results in an inflection-point in most of the curves. It should be especially noticed that the curves diverge from each other rapidly between the 75 per cent mixture and pure acetone. This seems to indicate that the dissociation is greatly increased by the addition of small amounts of water. POTASSIUM IODIDE. The salt gave no test for the presence of an iodate, and the flame test showed no appreciable impurity. The salt was dried at 100° to 110° and kept in a des- iccator. The salt dissolved in acetone, giving only a very slight coloration. 100 90 80 70 60 50 ft a 3 40 o 30 10- 35^ 50$ .75# Percentage of Acetone FIG. 25. — CONDUCTIVITY OF LITHIUM NITRATE IN MIXTURES OF ACETONE AND ETHYL ALCOHOL AT 25°. .100/5 Percentage of Acetone FIG. 26. — CONDUCTIVITY OF LITHIUM NITRATE IN MIXTURES OF ACETONE AND WATER AT 0°. 90 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. 120- 100^ Percentage of Acetone FIG. 27. — CONDUCTIVITY OF LITHIUM NITRATE IN MIXTURES OF ACETONE AND WATER AT 25°. TABLE 55. — Conductivity of potassium iodide at 0° and 25°. In methyl alcohol. In a mixture of 25 p. ct. acetone and methyl alcohol. Temperature 0 o Temperature Mt-0 Mt.25 coefficient. Mi> Mu coefficient. 200 65.7 91.4 0.0156 74.1 101.5 0.0148 300 68.1 94.6 .0156 75.6 103.7 .0151 400 68.8 96.3 .0159 77.6 106.1 .0147 600 70.1 99.0 .0165 78.1 108.8 .0157 800 70.6 100.5 .0169 82.3 113.4 .0151 1000 71.1 101.9 .0174 80.8 113.8 .0163 1200 71.4 102.4 .0174 82.8 113.8 .0149 1600 71.7 103.3 .0176 83.9 116.5 .0155 In a mixture of 50 p. ct. ace- tone and methyl alcohol. In a mixture of 75 p. ct. ace- tone and methyl alcohol. In acetone. V /J^Q" Mr Temperature coefficient. *r Mr Tempera- ture coef- ficient. - ^25° Temperature coefficient. 200 82.7 110.3 0.0133 93.1 117.0 0.0103 100.4 118.0 0.00701 300 85.2 114.7 .0139 95.6 123.0 .0115 105.8 126.2 .00772 400 87.2 114.8 .0127 97.8 124.3 .0108 108.9 128.7 .00730 600 89.6 118.6 .0129 100.4 129.0 .0110 112.3 134.1 .00775 800 91.9 121.8 .0130 104.1 131.8 .0106 116.2 138.6 .00772 1000 93.6 123.1 .0126 103.7 132.7 .0120 118.4 141.6 .00785 1200 93.7 126.2 .0139 104.3 135.7 .0118 118.2 140.3 .00750 1600 94.1 129.2 .0149 106.5 137.7 .0117 120.0 141.1 .00704 POTASSIUM IODIDE. 91 TABLE 55. — Conductivity of potassium iodide at 0° and 25°. — Continued. V In ethyl alcohol. In a mixture of 25 per cent acetone and ethyl alcohol. (J-vO° M*25° Temperature coefficient. M.OO M,25° Temperature coefficient. 200 300 400 600 800 1000 1200 1600 22.0 23.2 23.8 25.5 26.2 27.3 27.9 28.6 34.6 36.5 37.3 39.1 39.9 41.2 41.8 42.8 0.0230 .0241 .0238 .0212 .0209 .0202 .0200 .0194 35.5 36.4 37.8 38.8 40.3 39.8 39.7 40.1 52.5 54.2 56.3 58.0 60.5 59.7 60.8 63.5 0.0192 .0196 .0196 .0197 .0200 .0200 .0212 .0232 V In a mixture of 50 p. ct. acetone and ethyl alcohol. In a mixture of 75 p. ct. acetone and ethyl alcohol. M,0° M,25° Temperature coefficient. M»0° M.25- Temperature coefficient. 200 300 400 600 800 1000 1200 1600 52.2 53.4 56.1 57.5 59.5 59.7 59.8 61.3 71.4 74.0 76.8 79.1 82.6 82.9 82.9 85.4 0.0146 .0155 .0148 .0150 .0155 .0155 .0154 .0157 72.0 75.3 78.0 79.5 82.4 81.8 82.7 84.8 92.7 97.5 100.1 102.4 106.0 105.4 107.1 109.0 0.0115 .0118 .0113 .0115 .0114 .0115 .0118 .0114 V In water. In a mixture of 25 p. ct. acetone and water. ^0° *25° Temperature coefficient. AMP ^250 Temperature coefficient. 200 300 400 600 800 1000 1200 1600 76.7 77.2 77.5 78.0 78.0 78.0 78.0 78.0 136.3 138.3 138.8 139.8 140.1 140.6 140.7 140.7 0.0310 .0316 .0318 .0312 .0318 .0320 .0322 .0322 44.6 44.7 45.3 46.1 47.0 46.3 47.5 47.8 91.1 92.7 92.4 95.6 96.9 96.2 102.9 100.1 0.0392 .0430 .0416 .0430 .0425 .0432 .0485 .0438 V In a mixture of 50 p. ct. acetone and water. In a mixture of 75 p. ct. acetone and water. M- ^25" Temperature coefficient. M° ^25° Temperature coefficient. 200 300 400 600 800 1000 1200 1600 36.3 36.8 37.2 37.6 38.6 39.0 38.4 37.5 73.8 75.9 76.3 77.7 79.4 80.9 80.6 78.8 0.0413 .0425 .0420 .0427 .0423 .0430 .0439 .0440 41.6 41.9 42.2 42.8 43.7 42.2 42.8 44.1 74.1 74.1 74.6 75.4 79.7 76.4 76.1 79.7 0.0311 .0308 .0304 .0302 .0330 .0320 .0310 .0321 92 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 56. — Comparison of the conductivities of potassium iodide. In mixtures of acetone and methyl alcohol. V AtO°. At 25°. Op.ct. 25 p. ct 50 p. ct 75 p. ct 100 p. ct. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. et. 100 p. ct. 200 65.7 74.1 82.7 93.1 100.4 91.4 101.5 110.3 117.0 118.0 300 68.1 75.6 85.2 95.6 105.8 94.6 103.7 114.7 123.0 126.2 400 68.8 77.6 87.2 97.8 108.9 96.3 106.1 114.8 124.3 128.7 600 70.1 78.1 89.6 100.4 112.3 99.0 108.8 118.6 129.0 134.1 800 70.6 82.3 91.9 104.1 116.2 100.5 113.4 121.8 131.8 138.6 1000 71.1 80.8 93.6 103.7 118.4 101.9 113.8 123.1 132.7 141.6 1200 71.4 82.8 93.7 104.3 118.2 102.4 113.8 126.2 135.7 140.3 1600 71.7 83.9 94.1 106.5 120.0 103.3 116.5 129.2 137.7 141.1 In mixtures of acetone and ethyl alcohol. AtO°. At 25°. 200 22.0 35.5 52.2 72.0 100.4 34.6 52.5 71.4 92.7 118.0 300 23.2 36.4 53.4 75.3 105.8 36.5 54.2 74.0 97.5 126.2 400 23.8 37.8 56.1 78.0 108.9 37.3 56.3 76.8 100.1 128.7 600 25.5 38.8 57.5 79.5 112.3 39.1 58.0 79.1 102.4 134.1 800 26.2 40.3 59.5 82.4 116.2 39.9 60.5 82.6 106.0 138.6 1000 27.3 39.8 59.7 81.8 118.4 41.2 59.7 82.9 105.4 141.6 1200 27.9 39.7 59.8 82.7 118.2 41.8 60.8 82.9 107.1 140.3 1600 28.6 40.1 61.3 84.8 120.0 42.8 63.5 85.4 109.0 141.1 In mixtures of acetone and water. AtO°. At 25°. 200 76.7 44.6 36.3 41.6 100.4 136.3 91 1 73.8 74.1 118.0 300 77.2 44.7 36.8 41.9 105.8 138.3 92.7 75.9 74.1 126.2 400 77.5 45.3 37.2 42.2 108.9 138.8 92.4 76.3 74.6 128.7 600 78.0 46.1 37.6 42.8 112.3 139.8 95.6 77.7 75.4 134.1 800 780 47.0 38.6 43.7 116.2 140.1 96.9 79.4 79.7 138.6 1000 78.0 46.3 39.0 42.4 118.4 1406 96.2 80.9 76.4 141.6 1200 78.0 47.5 38.4 42.8 118.2 140.7 102.9 80.6 76.1 140.3 1600 78.0 47.8 37.5 44.1 120.0 140.7 100.1 78.8 79.7 141.1 TABLE 57. — Comparison of the temperature coefficients of conductivity of potassium iodide Jrom 0° to 25°. V In mixtures of acetone and methyl alcohol. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 200 0.0156 0.0148 0.0133 0.0103 0.00701 300 .0156 .0151 .0139 .0115 .00772 400 .0159 .0147 .0127 .0108 .00730 600 .0165 .0157 .0129 .0110 .00775 800 .0169 .0151 .0130 .0106 .00772 1000 .0173 .0163 .0126 .0120 .00785 1200 .0174 .0149 .0159 .0118 .00750 1600 .0176 .0155 .0149 .0117 .00704 POTASSIUM IODIDE. 93 TABLE 57. — Comparison of the temperature coefficients of conductivity of potassium iodide from 0° to 25°. — Continued. In mixtures of acetone and ethyl alcohol. V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 200 0.0230 0.0192 0.0146 0.0115 0.00701 300 .0241 .0196 .0155 .0118 .00772 400 .0238 .0196 .0148 .0113 .00730 600 .0212 .0197 .0150 .0115 .00775 800 .0209 .0200 .0155 .0114 .00772 1000 .0202 .0200 .0155 .0115 .00785 1200 .0200 .0212 .0154 .0118 .00750 1600 .0194 .0232 .0157 .0114 .00704 V In mixtures of acetone and water. 200 0.0310 0.0392 0.0413 0.0311 0.00701 300 .0316 .0430 .0425 .0308 .00772 400 .0318 .0416 .0420 .0304 .00730 600 .0312 .0430 .0427 .0302 .00775 800 .0318 .0425 .0423 .0330 .00772 1000 .0320 .0432 .0430 .0320 .00785 1200 .0322 .0485 .0439 .0310 .00750 1600 .0322 .0438 .0440 .0321 .00704 25 }J 505« 7556 Percentage of Acetone 100 $ FIG. 28. — CONDUCTIVITY OF POTASSIUM IODIDE IN MIXTURES OF ACETONE AND METHYL ALCOHOL AT 0°. 94 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. Tables 55 and 56 (figs. 28 and 29), for potassium iodide in mixtures of acetone and methyl alcohol, show that the conductivity is almost exactly what we should expect from the law of averages. There is, however, a slight tendency towards a maximum as we raise the temperature. In this respect the results are similar to those obtained with lithium nitrate. The values for the conductivity of potassium iodide in pure water, and ethyl and methyl alcohols, were taken from the work of Jones and Lindsay. 140 50$ 75$ Percentage of Acetone 100$ FIG. 29. — CONDUCTIVITY OF POTASSIUM IODIDE IN MIXTURES OF ACETONE AND METHYL ALCOHOL AT 25°. Tables 56 and 57 (figs. 30 and 31) show the same characteristics for potassium iodide, in mixtures of acetone and ethyl alcohol, as those observed for the same salt in mixtures of acetone and methyl alcohol, but there is less of a tendency towards a maximum. In fact, there is a slight sagging in the curves. It is observed that this statement is almost identical with the one in regard to lithium nitrate in mixtures of acetone and ethyl alcohol. Tables 55 and 56 (figs. 32 and 33), for potassium iodide in mixtures of acetone and water, display a minimum in molecular conductivity. There is no tendency towards a maximum. It should, however, be noticed that the divergence of the curves between the 75 per cent mixture and pure acetone is small. The salt is, therefore, quite largely dissociated at all dilutions in all of the mixtures. It was thought advisable, at this stage, to use solutions of sodium iodide in the various mixtures. Solutions of sodium iodide in acetone had been investigated by Carrara,1 by Dutoit and Friderich,2 and by Jones.3 It was 1 Gazz. Chim. Ital., [1] 27, 207 (1897). 2 Bull. Soc. Chim., [3] 19, 334 (1898). sAmer. Chem. Journ., 27, 16 (1902). POTASSIUM IODIDE. 95 noticed, however, that the solution had a high colora- tion, which deepened on standing. Moreover, there was a slight deposit formed at the same time. On evapo- ration the residue was still colored. The strongest solu- tions of the sodium iodide, in acetone, gave a test for iodine with starch paste, while the more dilute solu- tions, though still somewhat colored, gave none. Free iodine was then added to pure acetone until the same color was reproduced. This solution, tested with starch 140- 130- 120- 110- 100- '> 90- '•3 o -§ 80- o | GOJ "o ^ 50- 40 30- 20- 10- 120- .110 100 FIG. 30. — CONDUCTIVITY OF POTASSIUM IODIDE IN MIXTURES OF ACETONE AND ETHYL ALCOHOL AT 0°. FIG. 31. — CONDUCTIVITY OF POTASSIUM IODIDE IN MIXTURES OF ACETONE AND ETHYL ALCOHOL AT 25°. ;£ 50 $ Percentage of Acetone paste, gave no test for io- dine. Time did not permit the further investigation of this interesting point. CALCIUM NITRATE. The calcium nitrate used was an anhydrous preparation obtained from Kahlbaum. It was heated for several days at 140°, until it had a constant weight. Subse- quently, it was dried for some time at 140° after each exposure to the air. The salt contained no calcium oxide after heating, and showed no appreciable im- purity by the flame test. 25$ 50^ Percentage of Acetone 100^ 9G CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. 120 110 100 90- 70- T3 I 60- 80 J BO ao Percentage of Acetone FIG. 32. — CONDUCTIVITY OF POTASSIUM IODIDE IN MIXTURES OF ACETONE AND WATER AT 0°. 40-| 50# 75£ Percentage of Ace"ton~e FIG. 33. — CONDUCTIVITY OF POTASSIUM IODIDE IN MIXTURES OF ACETONE AND WATER AT 25°. CALCIUM NITRATE. 97 TABLE 58. — Conductivity of calcium nitrate at 0° and 25°. In methyl alcohol. v At,-0° Att,25° Temperature MvO ..-. /^i,-2o (Carroll). (Carroll). coefficient. 5 14.33 19.67 0.0149 10 18.98 25.38 • • » • .0135 16 .... 32.79 25 27.66 36.77 .... .0131 32 31.30 41.88 .0135 50 34.6 .... 45.7 * • • • .0128 64 .... 37.27 50.79 .0145 100 42.2 .... 55.9 .... .0130 128 46.66 60.52 .0119 200 49.9 .... 65.4 • • • • .0124 256 55.17 73.98 .0136 400 58.2 75.6 .... .0120 800 65.7 .... 86.6 .... .0127 1200 74.4 95.0 • * .0111 1600 77.2 98.2 .... .0109 In a mixture of 25 p. ct. acetone and In a mixture of 50 p. ct. acetone and At methyl alcohol. methyl alcohol. u QCO Temperature MrO° ro Temperature M" MD-O coefficient. //.„ 0 coefficient. 5 13.13 17.36 0.0129 10.16 12.97 0.0111 10 17.76 23.08 .0120 13.82 17.29 .0100 25 26.33 33.74 .0111 21 .30 26.11 .00903 50 33.9 43.2 .0110 28.11 34.44 .00895 100 42.1 53.9 .0112 35.7 42.8 .00795 200 50.7 66.8 .0127 45.2 54.8 .0085 400 60.9 76.8 .0105 55.3 68.3 .0088 800 71.0 89.4 .0104 66.8 83.0 .0097 1200 79.7 98.0 .0092 73.5 91.3 .0097 1600 82.6 102.7 .0098 79.2 98.8 .0099 V In a mixture of 75 p. ct. acetone and methyl alcohol. In acetone. 5 6.07 7.65 0.0104 3.93 4.96 0.0105 10 8.10 9.78 .0083 4.44 5.67 .0111 25 12.40 14.56 .0070 5.06 6.55 .0118 50 16.29 19.47 .0078 5 .34 6.90 .0117 100 21.65 25.1 .00635 5 .48 7.06 .0115 200 28.5 33-0 .00630 5 .93 7.54 .0109 400 38.1 44.3 •0065 6 .69 8.22 .00916 800 49.2 57.8 .00698 7 .93 9.69 .00884 1200 57.3 66.9 .0067 9 .26 11.22 .00847 1600 64.2 75.1 .0068 10 .36 12.62 .00873 V In ethyl alcohol. In a mixture of 25 p. ct. acetone and ethyl alcohol. 5 3.80 5.94 0.022 4.45 6.18 0.0156 10 5.13 7.86 .021 6 .01 8-29 .0151 25 7.69 11.67 .020 9 .29 12.5 .0140 50 9.80 14.9 .024 12.20 16.59 .0144 100 11.9 18.4 .024 15.4 21.1 .0148 200 14.3 22.4 .024 , f f r • • • • • • • 400 15.2 23.7 .022 22 .3 32.2 .0177 800 17.2 27.5 .024 27.2 40.1 .0190 1200 18.1 29.5 .024 28 .9 42.7 .0191 IfiOO 18.8 33.3 .031 31 .6 46.3 .0186 98 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 58. — Conductivity of calcium nitrate at 0° and 25°. — Continued. In a mixture of 50 p. ct. acetone and In a mixture of 75 p. ct. acetone and ethyl alcohol. ethyl alcohol. V OKO Temperature o n O^O Temperature **" Mr coefficient. *** AH'^O coefficient. 5 4.42 5.86 0.0130 3.89 5.08 0.0123 10 6.00 7.64 .0109 4.80 5.99 .00992 25 9.25 11.48 .00963 6.92 8.22 .00753 50 12.33 15.24 .00945 9.20 11.34 .00930 100 16.1 19.9 .00942 12.1 13.8 .0056 200 20.7 26.0 .0102 16.2 18.3 .0052 400 26.1 32.8 .0102 21.3 24.3 .00562 800 31.8 41.4 .0121 28.1 32.7 .0065 1200 35.8 46.8 .0123 33.2 39.1 .0073 1600 38.0 50.7 .0134 36.2 42.4 .0068 V In water. In a mixture of 25 p. ct. acetone and water. 5 80.3 146.8 0.0331 49.9 101.0 0.0410 10 89.8 165.5 .0333 55.0 112.3 .0417 25 98.2 184.2 .0350 60.2 123.9 .0423 50 107.0 192.2 .0313 64.8 134.0 .0427 100 110.4 204.7 .0341 70.3 144.7 .0423 200 114.9 214.2 .0343 72.3 154.1 .0452 400 119.9 222.2 .0341 74.1 154.9 .0441 800 123.4 232.2 .0351 76.8 161.0 .0438 1200 .... 238.5 .... 79.1 165.4 .0436 1600 128.3 249.8 .0378 80.0 167.7 .0439 V In a mixture of 50 p. ct. acetone and water. In a mixture of 75 p. ct. acetone and water. 5 36.9 73.4 0.0392 25.0 43.1 0.0290 10 42.2 84.6 .0402 31.3 52.8 .0272 25 46.7 94.8 .0413 40.1 68.3 .0281 50 50.9 104.7 .0423 45.8 79.2 .0291 100 54.8 115.4 .0449 52.6 92.3 .0300 200 .59.6 120.3 .0458 58.4 102.5 .0301 400 61.4 128.0 .0435 62.6 114.9 .0331 800 64.0 133.1 .0433 71.6 126.9 .031 1200 66.1 137.9 .0434 75.1 133.5 .031 1600 66.2 139.8 .0444 76.7 137.7 .032 TABLE 59. — Comparison of the conductivities of calcium nitrate. In mixtures of acetone and methyl alcohol. V AtO°. At 25°. 0 p. ct. 25 p. ct. 50 p. ct. 75 p.ct. 100 p.ct. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p.ct. 5 14.33 13.13 10.16 6.07 3.93 19.67 17.36 12.97 7.65 4.96 10 18.98 17.76 13.82 8.10 4.44 25.38 23.08 17.29 9.78 5.67 25 27.66 26.33 21.30 12.40 5.06 36.77 33.74 26.11 14.56 6.55 50 34.6 33.9 28.11 16.29 5.34 45.7 43.2 34.44 19.47 6.90 100 42.2 42.1 35.7 21.65 5.48 55.9 53.9 42.8 25.1 7.06 200 49.9 50.7 45.2 28.5 5.93 65.4 66.8 54.8 33.0 7.54 400 58.2 60.9 55.3 38.1 6.69 75.6 76.8 68.3 44.3 8.22 800 65.7 71.0 66.8 49.2 7.93 86.6 89.4 83.0 57.8 9.69 1200 74.4 79.7 73.5 57.3 9.26 95.0 98.0 91.3 66.9 11.22 1600 77.2 82.6 79.2 64.2 10.36 98.2 102.7 98.8 75.1 12.62 CALCIUM NITRATE. 99 TABLE 59. — Comparison of the conductivities of calcium nitrate. — Continued. In mixtures of acetone and ethyl alcohol. V AtO°. At 25°. 0 p. ct. 25 p. ct. 50p.ct. 75 p.ct. 100 p.ct. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p.ct. 5 3.80 4.45 4.42 3.89 3.93 5.94 6.18 5.86 5.08 4.96 10 5.13 6.01 6.00 4.80 4.44 7.86 8.29 7.64 5.99 5.67 25 7.69 9.29 9.25 6.92 5.06 11.67 12.52 11.48 8.22 6.55 50 9.80 12.20 12.33 9.20 5.34 14.91 16.59 15.24 11.34 6.90 100 11.9 15.4 16.1 12.1 5.48 18.4 21.1 19.9 13.8 7.06 200 14.3 19.1 20.7 16.2 5.93 22.4 26.5 26.0 18.3 7.54 400 15.2 22.3 26.1 21.3 6.69 23.7 32.2 32.8 24.3 8.22 800 17.2 27.2 31.8 28.1 7.93 27.5 40.1 41.4 32.7 9.69 1200 18.1 28.9 35.8 33.2 9.26 29.5 42.7 46.8 39.1 11.22 1600 18.81 31.6 38.0 36.2 10.36 33.3 46.3 50.7 42.4 12.62 In mixtures of acetone and water. AtO°. At 25°. 5 80.3 49.9 36.9 25.0 3.93 148.6 101.0 73.4 43.1 4.96 10 89.8 55.0 42.2 31.3 4.44 165.5 112.3 84.6 52.8 5.67 25 98.2 60.2 46.7 40.1 5.06 184.2 123.9 94.8 68.3 6.55 50 107.0 64.8 50.9 45.8 5.34 192.2 134.0 104.7 79.2 6.90 100 110.4 70.3 54.8 52.6 5.48 204.7 144.7 115.4 92.3 7.06 200 114.9 72.3 59.6 58.4 5.93 214.2 154.1 120.3 102.5 7.54 400 119.9 74.1 61.4 62.6 6.69 222.2 154.9 128.0 114.9 8.22 800 123.4 76.8 64.0 71.6 7.93 232.2 161.0 133.1 126.9 9.69 1200 79.1 66.1 75.1 9.26 238.5 165.4 137.9 133.5 11.2 1600 128.3 80.0 66.2 76.7 10.4 249.8 167.7 139.8 137.7 12.6 TABLE 60. — Comparison of the temperature coefficients of conductivity of calcium nitrate from 0° to 25°. In mixtures of acetone and methyl alcohol. Op. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 5 0.0149 0.0129 0.0111 0.0104 0.0105 10 .0135 .0120 .0100 .0083 .0111 25 .0131 .0111 .00903 .0070 .0118 50 .0128 .0110 .00895 .0078 .0117 100 .0130 .0112 .00795 .0063 .0115 200 .0124 .0127 .0085 .0065 .0109 400 .0120 .0105 .0088 .0065 .00916 800 .0127 .0104 .0097 .00698 .00884 1200 .0111 .0092 .0097 .0067 .00847 1600 .0109 .0093 .0099 .0068 .00873 V In mixtures of acetone and ethyl alcohol. 5 0.022 0.0156 0.0130 0.0123 0.0105 10 .021 .0151 .0109 .0083 .0111 25 .020 .0140 .00963 .0070 .0118 50 .024 .0144 .00945 .0078 .0117 100 .024 .0148 .00942 .0093 .0115 200 .024 .0155 .0102 .0056 .0109 400 .022 .0177 .0102 .00562 .00916 800 .024 .0190 .0121 .0065 .00884 1200 .024 .0191 .0123 .0073 .00847 1600 .031 .0186 .0134 .0068 .00873 100 TABLE 60. CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. • Comparison of the temperature coefficients of conductivity of calcium nitrate from 0° to 25°. — Continued. V In mixtures of acetone and water. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 5 0.0331 0.0410 0.0392 0.0290 0.0105 10 .0333 .0417 .0402 .0272 .0111 25 .0350 .0423 .0413 .0281 .0118 50 .0313 .0427 .0423 .0291 .0117 100 .0341 .0423 .0449 .0300 .0115 200 .0343 .0452 .0458 .0301 .0109 400 .0341 .0441 .0435 .0331 .00916 800 .0351 .0438 .0433 .031 .00884 1200 .0436 .0434 .031 .00847 1600 .0378 .0439 .0444 .032 .00873 100- Percentage of Acetone FIG. 34. — CONDUCTIVITY OF CALCIUM NITRATE IN MIXTURES OF ACETONE AND METHYL ALCOHOL AT 0°. Tables 58 and 59 (figs. 34 and 35), for calcium nitrate in mixtures of acetone and methyl alcohol, give a pronounced maximum in conductivity at high dilutions. It will be recalled that Jones and Carroll obtained a mini- mum conductivity with this salt, in mixtures of ethyl alcohol and water, only at low temperatures and at high dilution. It should be noted that calcium nitrate is very slightly dissociated in acetone, and that the maximum occurs in the 25 per cent mixture. The temperature coefficients of conductivity decrease with the dilution, and they are also less in the mixtures than in the pure solvents, the minimum appearing in the 75 per cent mixture. CALCIUM NITRATE. 101 Tables 58 and 59 (figs. 36 and 37), for calcium nitrate in mixtures of ace- tone and ethyl alcohol, show the same characteristics as were observed in the tables for this salt in mixtures of acetone and methyl alcohol; but here the Percentage of Acetone Fio. 35. — CONDUCTIVITY OF CALCIUM NITRATE IN MIXTURES OF ACETONE AND METHYL ALCOHOL AT 25°. maximum is more prominent, being present at all dilutions and at both tem- peratures. The dissociation of calcium nitrate in ethyl alcohol, however, is small. With increasing dilution the maximum shifts from the 25 per cent mixture to the 75 per cent mixture. 50^ 75;£ 100# Percentage of Acetone FIG. 36. — CONDUCTIVITY OF CALCIUM NITRATE IN MIXTURES OF ACETONE AND ETHYL ALCOHOL AT 0°. 102 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. The temperature coefficients in ethyl alcohol are almost constant, perhaps increasing slightly. The temperature coefficients of the mixtures show a Percentage of Acetone FIG. 37. — CONDUCTIVITY OF CALCIUM NITRATE IN MIXTURES OF ACETONE AND ETHYL ALCOHOL AT 25°. minimum in the 75 per cent mixture. A similar phenomenon was noticed in the mixtures of acetone and methyl alcohol. 50 g T5# 1003 Percentage of Acetone FIG. 38.— CONDUCTIVITY OF CALCIUM NITRATE IN MIXTURES OF ACETONE AND WATER AT 0°. CALCIUM NITRATE. 103 Tables 58 to 60 (figs. 38 and 39) , for calcium nitrate in mixtures of acetone and water, show a point of inflection at low temperatures and high dilution. In concentrated solutions, at high temperatures, the conductivity is what we should expect from the law of averages. These results are similar to those obtained with lithium nitrate in mixtures of acetone and water. 25 # 50$ Percentage of Acetone 1005S FIG. 39. — CONDUCTIVITY OF CALCIUM NITRATE IN MIXTURES OF ACETONE AND WATBR AT 25°. The temperature coefficients of conductivity of water increase with the dilution, while the temperature coefficients of acetone decrease with the dilution. In the 75 per cent mixture the temperature coefficients are nearly independent of the dilution. VISCOSITY MEASUREMENTS. In the table of viscosity data (table 61), the values of Thorpe and Rodger * for pure water, at 0° and 25°, are taken as the standard, and the other values are referred to them; 17 represents viscosity, <£ fluidity, and D the density of the liquid in question, at 0s0 and 25i°, compared with the density of water at 0° and 25°, respectively. IPhil. Trans., 185A, 307 (1894). 104 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 61. -- Viscosity measurements. • A r\o J ocro -./\O *4*r\o DO0 ^.r>-O j OrO D25° Temp. At 0 and 25 . V 7?0 ZO 25° coef. Water i p g_ 0 01778 56.24 1.000 0.00891 112.3 1.000 0.0426 Mixture of 25 p. ct. acetone VJ.VJ J- 1 ( O and water P S .0293 34.12 0.9802 001276 78 37 0 9710 .0518 Calcium nitrate in mixture 10 !01397 7l!59 !948 of 50 p.ct. acetone and 1600 .0133 75.13 .936 water P.S. .03027 33.03 .9522 0133 74.96 937 .0508 Calcium nitrate in mixture 10 100S617 104!o '.892 of 75 p. ct. acetone and 1600 .009019 110.9 .880 water P S .0170 58.80 .9023 008904 112 3 879 .0364 10 !003544 282'. 1 !sos Calcium nitrate in acetone 1600 .003255 307.3 .790 P.S. .004097 244.1 .8132 .003237 308.9 .788 .0106 Calcium nitrate in mixture 10 .005319 188.0 .811 of 25 p. ct. acetone and 1600 .004604 217.2 .790 methyl alcohol .... P.S. .006498 153.9 .816 .004615 216.7 .792 .0163 Calcium nitrate in mixture 10 .004532 220.6 .816 of 50 p. ct. acetone and 1600 .003926 254.7 .794 methyl alcohol .... P.S. .005336 187.4 .818 .003891 257.0 .794 .0148 Calcium nitrate in mixture 10 .003909 255.8 .812 of 75 p. ct. acetone and 1600 .003480 287.3 .793 methyl alcohol .... P.S. .004501 222.2 .817 .003446 290.1 .792 .0122 Calcium nitrate in ethyl 5 .01373 72.84 .81612 alcohol P.S. .01856 53.88 .80820 .01106 90.35 .7895 .0271 Calcium nitrate in mixture of 25 p. ct. acetone and ethyl alcohol 5 P.S. .01041 96.08 £1244 .008444 .006714 118.4 148.9 .81822 .791 .0220 Calcium nitrate in mixture of 50 p. ct. acetone and ethyl alcohol 5 P.S. .006801 147.0 .81394 .005861 .004874 170.6 205.2 .81818 .7904 .0148 Calcium nitrate in mixture of 75 p. ct. acetone and ethyl alcohol 1 5 1 P.S. .004990 200.4 .81380 .004484 .003776 223.0 264.8 .81709 .7896 .01296 1 P.S.=pure solvent. TABLE 62. — Comparison of fluidities. Mixtures. V 0 p.ct. 25 p.ct. 50 p.ct. 75 p. ct. 100 p. ct. Mixtures of acetone and water at 0° 'P.S. 56.2 34.12 33.03 58.80 244.1 Calcium nitrate in mixtures of acetone and water at 25° . . . r 10 1 1600 [ P.S. 112.3 78.37 71.59 75.13 74.96 104.0 110.9 112.3 282.1 307.3 308.9 Mixtures of acetone and methyl alcohol at 0° }.... 122.2 153.9 187.4 222.2 244-1 Calcium nitrate in mixtures of ( 10 161.8 188.0 220.6 255.8 282.1 acetone and methyl alcohol ] 1600 180.4 217.2 254.7 287.3 307.3 at 25° [ P.S. 176.7 216.7 257.0 290.1 308.9 Mixtures of acetone and ethyl alcohol at 0° }.... 53.88 96.08 147.0 200.4 244.1 Calcium nitrate in mixtures of ace- / 5 72.84 118.4 170.6 223.0 263.5 tone and ethyl alcohol at 25°. . I P.S. 90.35 148.9 205.2 264.8 308.9 1 P.S. = pure solvent. TABLE 63. — Comparison of the temperature coefficients of fluidity. Mixtures. From 0° to 25°. Op.ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. Mixtures of acetone and water . . . Mixturesof acetone and methyl alcohol Mixtures of acetone and ethyl alcohol 0.0398 .0178 .0271 0.0518 .0163 .0220 0.0508 .0148 .0148 0.0364 .0122 .01296 0.0106 .0106 .0106 FLUIDITY. 105 Tables 61 to 63 (figs. 40 and 41) show that there is a minimum of fluidity only in the case of acetone and water. In the mixtures of acetone with methyl alcohol we get somewhat larger values than would be expected from the fluidities of the pure solvents. This effect is not so apparent, however, in the case of acetone and ethyl alcohol. These last values were compared with those derived from Dunstan's results, and were found to be almost iden- tical with them. 100"5 Percentage of Acetone FIG. 40. — FLUIDITY OF SOLVENT MIXTURES AT 0°. If we compare the viscosity curves of acetone and water, with the fluidity curves, we find that the maximum is more pronounced than the fluidity minimum. The viscosity curves for mixtures of acetone and the alcohols show a marked sagging, as Dunstan has pointed out. Table 64 shows that although the temperature coefficients of conductivity and fluidity vary in the same manner, the former are uniformly smaller than the latter. 106 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 64. — Comparison of the temperature coefficients of conductivity and fluidity. V Solute. Op.ct. 25 p. ct. 50p.ct. 75 p. ct. 100 p. ct. Mixture. Fluidity . . Pure solv't 0.0398 0.0518 0.0508 0.0364 0.0106 •. {5 Ca(N03)2 .0331 .0410 .0392 .0290 .01050 1600 .0378 .0439 .0444 .0320 .00873 Acetone Conductivity 5 1600 LiNO3 .0323 .0330 .0415 .0430 .0394 .0435 .0292 .0331 .00755 .00325 and water. 200 KI .0310 .0392 .0413 .0311 .00701 1600 II .0322 .0438 .0440 .0321 .00704 - Fluidity . . Pure solv't 0.0271 0 0220 0.0148 0.01296 0 0106 - {5 Ca(N03)2 .022 .0156 .0130 .0123 .0105 1600 '* .031 (?) .0186 .0134 .0068 .00873 Acetone Conductivity 5 1600 LiNO3 .0241 .0252 .0195 .0220 .0142 .0167 .00795 .0119 .00755 .00325 and ethyl alcohol. 200 KI .0230 .0192 .0146 .0115 .00701 1600 II .0194 .0232 .0157 .0114 .00704 - Fluidity . . Pure solv't 0.0178 0.0163 0.0148 0 0122 0.0106 {5 Ca(NO3)2 .0149 .0129 .0111 .104 .0105 1600 4* .0109 .0098 .0099 .0068 .00873 Acetone Conductivity 5 1600 LiNO3 .0152 .0160 .0129 .0146 .0103 .0117 .0104 .0112 .00755 .00325 > and methyl alcohol. 200 KI .0156 .0148 .0133 .0103 .00701 1600 II .0176 .0155 .0149 .0117 .00704 50 $ 75 # Percentage of Acetone FIG. 41.— FLUIDITY OF SOLVENT MIXTURES AT 25°. 100$ DISCUSSION OF RESULTS. 107 DISCUSSION OF RESULTS. The curves for the conductivity of potassium iodide in all of the different mixtures are very similar to the curves of fluidity in the corresponding mix- tures. Lithium nitrate and calcium nitrate in all mixtures, at low tem- peratures, show a deviation from the fluidity curves, particularly in the 75 per cent mixtures, tending to produce a maximum in conductivity. The work of Jones and Lindsay and of Jones and Carroll showed that solu- tions of lithium nitrate, in mixtures of methyl alcohol and water, gave curves with a simple minimum, like the fluidity curves for the corresponding mixtures. Calcium nitrate, dissolved in the same mixtures, and also in mixtures of ethyl alcohol and water, in no case gave a minimum according to Jones and Carroll ; consequently their curves are not similar to the corresponding fluidity curves. Our results differ fundamentally from those heretofore observed, in that the mixtures of acetone with the alcohols and water show a tendency towards a maximum in the conductivity of solutions of certain salts, such as lithium nitrate and calcium nitrate. Since conductivity is dependent upon fluidity, and not vice versa, we shall discuss first the fluidity curves, and then the conductivity curves in connec- tion with them. When methyl alcohol or ethyl alcohol is mixed with acetone, the fluidity curve of the mixtures is a straight line. This is what we should expect, if the fluidity of each of the components has its proportionate effect. Hence we may conclude that the molecular aggregations of these pure solvents are not essentially changed in regard to size by mixing the two solvents. We have already shown that the work of Thorpe and Rodger, Traube, Varenne and Godefroy, and others has made it evident that viscosity is dependent upon the character of the molecular aggregations present. It may be objected that a straight line is not the "normal" fluidity curve, since, heretofore, a straight line has been considered to be the normal viscosity curve, and the two conceptions are, in general, incompatible. To make this clear, let us suppose that we mix two liquids which are made up of particles which have no unusual action on each other, i. e., do not form new aggrega- tions of any kind. Two monomolecular liquids which do not form complexes on mixing would fulfill this condition. Further, let us suppose that the liquid is allowed to flow through a tube. The resulting fluidity would be the sum of the partial fluidities of the components. That is, the more rapidly moving particles would be held back by the slower ones, and the motion would be a mean value, proportional to the relative amounts of the components. For- mulated, this would be (1) 108 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. where m^ and ra2 are the amounts of the components, and ^ and <£2 are their respective fluidities. This is similar to the conception which we have in electricity, where the conductance of two or more conductors is represented by the sum of their separate conductances. The conductance of a single conductor is expressed by the value -^, where c is the specific conductivity, a- the cross-section of L the conductor, and I the length. The conductance of a pair of conductors of different material, in parallel, is, per unit length, fa + ?2 — *7i which is the equation of the equilateral hyperbola, the Y-axis of which is the distance ; - to the left of the origin, to which equation (1) is referred. Thus, we seem justified in concluding that the hyperbola is the normal curve for viscosities. From the above considerations we are led to the belief that inferences drawn from viscosity curves alone may lead to erroneous conclusions. For example, Wijkander 1 reached the conclusion that in no case is the viscosity identical with that calculated by the admixture rule. In the case of mixtures of ether with chloroform, and of ether with carbon disulphide, there were 1 Beibl. Wied. Ann., 8, 3 (1879). FLUIDITY. 109 inflection-points in the curves, but no simple relation between the viscosity coefficients of a mixture and those of its constituents could be deduced. Linebarger 1 found that the observed viscosities, in general, were less than those that were calculated by the mixture rule, except, perhaps, in the case of mixtures of benzene and chloroform, and mixtures of carbon disulphide and benzene, toluene, ether, and acetic ether, where, according to Dunstan, the temperature of observation, 25°, was possibly too near the boiling-point of the carbon disulphide to make any specific influence which that liquid might exert at lower temperatures perceptible. Dunstan 2 makes the significant statement that " the law of mixtures is never accurately obeyed, and divergences seem to be more clearly marked in the case of viscosity than with other properties, such as refractive index." These discrepancies are explained if our view be accepted, since the diver- gence in every abnormal case thus far investigated is smaller for the fluidity curves than for the corresponding viscosity curves, and the mixtures with carbon disulphide, which give "normal" viscosity curves, also give fluidity curves that are equally satisfactory. In the particular case of acetone and methyl alcohol or ethyl alcohol, the fluidity is a straight line, nearly to within the limits of experimental error, so that these two pairs of liquids may be considered as perfectly normal. It must be stated explicitly that many of the conclusions arrived at by the above-mentioned workers are not changed by this new method of comparing results, especially since, in many cases, they obtained curves with actual maxima and minima. These effects are reproduced in the fluidities as minima and maxima, respectively, which are generally less prominent than before. When most of the organic solvents worked with up to this time are mixed with water, there is a very large increase in viscosity. In general, there is also a contraction on mixing these solvents and water. Some of those that give a pronounced maximum of viscosity are methyl alcohol, ethyl alcohol, propyl alcohol, isopropyl alcohol, acetic acid, propionic acid, butyric acid, isobutyric acid, and acetone. The workers in this field have attributed the increase in viscosity to increase in the size of the molecular aggregations. This decrease in fluidity retards the movement of the ions, hence there is a fall in molecular conductivity, which explains the minimum in conductivity heretofore observed by Zelinsky and Krapiwin, Cohen, Jones and Lindsay, Jones and Carroll, and ourselves. This relation between viscosity and con- ductivity, as has been shown, has long been recognized. Wiedemann, Stephan, Dutoit and Friderich, and Jones and Carroll have been connected with the development of the exact relation between them. 1 Amer. Journ. Sci. [4] 2, 331 (1896). 2 Journ. Chem. Soc., 85, 817 (1904.) Ztschr. phys. Chem., 49, 590 (1904). 110 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. If the salt happens to be highly dissociated in water and very little disso- ciated in the other solvent, the curves may be at such an angle with the axis of X that there will be only a sagging of the curve and not an actual minimum. Therefore, the amount of deviation from the normal curve appears to us to be a matter of prime importance, while the finding of a minimum is not. If we could make a correction at the different parts of the curve for the different degrees of dissociation in the pure solvents and in the mixtures, we should then have a curve exactly parallel to the fluidity curve, if it is true that fluidity and dissociation are the only factors concerned, as Dutoit and Friderich, and Jones and Carroll supposed. In our case it is almost impos- sible to make the correction with the data at hand, except, possibly, in the case of potassium iodide. The conductivity values for potassium iodide show that it is nearly dissociated in all mixtures, but with the other salts com- plete dissociation is not even approximately reached. Thinking that it might be possible to calculate /too, we have tested Kohlrausch's formula, fM 00 — fj.v __ T^ r l U 3 where C is the concentration and K is a constant. We found that it did not apply except in aqueous solutions. Vollmer 1 has already shown that the Ostwald dilution law does not apply to solutions in ethyl alcohol and methyl alcohol. Solutions of potassium iodide, in mixtures of methyl alcohol and water, were investigated by Zelinsky and Krapiwin and Jones and Lindsay. The conductivity curves resemble the fluidity curves for methyl alcohol and water, as shown by Jones and Carroll. We have found the same similarity in the case of potassium iodide and water. If we accept the Kohlrausch and Jones's hypotheses of ionic spheres, it is evident that the atmosphere about the ions remains of the same size through- out all the mixtures ; otherwise the ions would tend to show a maximum in conductivity in those mixtures where the atmosphere is smallest, causing a divergence from the fluidity curves. Difference in dissociation would also cause a divergence between the conductivity and fluidity curves. In the above case, however, the dissociation is large and all the curves are parallel. If we pass now to potassium iodide in mixtures of acetone with methyl alcohol and ethyl alcohol, we find, again, that the conductivity curves and fluidity curves are very similar, i. e., nearly straight lines, with a tendency towards a maximum, which is greater in the case of methyl alcohol than in that of ethyl alcohol. Evidently the changes in the size of the ionic spheres and the changes in the dissociation have either counteracted each other or remained zero. 1 Ann. der Phys., 62, 328 (1894). FLUIDITY. Ill Now let us consider lithium nitrate. In mixtures of acetone with the alcohols, we get a pronounced maximum in conductivity in the 75 per cent mixtures, at high dilutions. Since these solvents gave no such maximum in the case of potassium iodide, the maximum must be connected with the lithium nitrate itself. There are two possible explanations of the phenome- non: (1) Increase in dissociation in the 75 per cent mixture; (2) increase in the mobility of the ions, due to the diminution in the size of the ionic spheres. We shall attempt to decide between these two possibilities. We have shown by consideration of the fluidities that the liquids are not more associated in the mixtures than in the pure solvents; hence, if we accept the hypothesis of Dutoit and Aston, that dissociating power increases with the association of the solvent, the maximum can not be due to increase in dissociation in the mixture. For example, calcium nitrate shows a pro- nounced maximum in conductivity in mixtures of alcohol and acetone, even though the dissociation of calcium nitrate in pure acetone is very small. It hardly seems probable that the acetone increases the dissociation of pure alcohol if it does not form complexes with it. We also have the fact, found by Jones and Carroll, that even in the case of alcohol and water, where molecular aggregations are known to be formed, there is not an increase in dissociation larger than the possible experimental error. Furthermore, if the 75 per cent mixture has the highest dissociating power, we should not expect to find the maximum moving from the 25 per cent mixture to beyond the 75 per cent mixture, as the concentration of the dissolved substance decreases. This is the case with lithium nitrate, in mixtures of the alcohols and acetone. Finally, the maximum in conductivity should manifest itself in the most concentrated solutions of potassium iodide, in mixtures of ace- tone with the alcohols. This is contrary to the facts. We, therefore, accept, tentatively, the view that the maximum in conductivity is due, primarily, to a change in the dimensions of the ionic spheres. The determination, however, of the dissociation of lithium nitrate in pure alcohol, in pure acetone, and in a 75 per cent mixture of these solvents, would be a very important check. Through the kindness of Mr. L. McMaster this point has been tested for acetone and ethyl alcohol, as shown in table 65. TABLE 65. — Conductivity of lithium nitrate corrected), V In pure ace- tone at 25°. In mixture of 75 p. ct. ace- tone and ethyl alcohol at 25°. In pure alcohol at 25°. 2000 2500 3000 55.28 62.87 66.42 92.66 100.55 101.88 37.01 41.04 41.48 112 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. The conductivity of the pure acetone used in these experiments was 1.516 x 10-6; that of the pure alcohol, 0.857 X KT6. These results show that complete dissociation is nearly reached only in the pure alcohol, and that the mixture is dissociated about as we might expect from the law of averages. We note that the maximum is very pro- nounced. The conclusion of Dutoit and Friderich and of Jones and Carroll, that conductivity is inversely proportional to viscosity, and directly propor- tional to the association factor of the solvent (or to the amount of dissocia- tion), is incomplete. It fails to take into consideration changes in the dimensions of the ionic spheres. In the conductivities of lithium nitrate, in mixtures of acetone and water, the decreased fluidity manifests itself again. We notice, however, that the power of the acetone to produce smaller (or more symmetrical) ionic spheres is not destroyed by substituting water for methyl alcohol or ethyl alcohol. Practically all of the dilutions in the 75 per cent mixture of acetone and water, show a decided elevation of the conductivity curves above those we should expect from similar measurements with potassium iodide. That increase in dissociation would make itself manifest in this way is doubtful, and our theory is thus strengthened. Jones and Lindsay's results with lithium nitrate, in mixtures of water and methyl alcohol, do not show this effect; hence the acetone acts peculiarly in this respect. However, acetone behaves exceptionally in other ways. At this point we should call attention to the fact that lithium forms a very slowly moving ion, i. e., one with a large ionic sphere, while potassium forms a comparatively rapidly moving ion, i. e., one with a small ionic sphere. The anions used do not differ greatly. Calcium forms an ion with a migration velocity between that of lithium and potassium. It was thought best to measure the conductivities of calcium nitrate in all of the mixtures, exactly as with the other salts. The results show that the tendency towards a maximum in conductivity in the mixtures is very marked indeed. Moreover, they show that calcium nitrate is dissociated in the solvents very differently from lithium nitrate. Calcium nitrate is dissociated to a large extent in water and methyl alcohol, very much less in ethyl alcohol, and still less in acetone. In spite of all differ- ences, we are struck by the fact that the maximum divergence still tends to manifest itself in the 75 per cent mixture. Especially is this the case with acetone and water. Let us now turn our attention to the concentrated solutions. In these we find the tendency towards a maximum very small, or entirely absent. If our explanation is correct, then, as the concentration of the salt increases there will be less of the solvent for the formation of ionic spheres, or, in the SUMMARY. 113 case of mixtures which tend to reduce the large atmospheres of the dilute solutions, the mass action of the solvent is much diminished. Calcium nitrate is intermediate in its behavior between potassium iodide and lithium nitrate. It seems reasonable to connect this with the migration velocity. It would be interesting to experiment with sodium, the ion of which has a slow migration velocity; but, as we have shown, sodium iodide, although soluble, is unsuited for this purpose. SUMMARY. We have measured the fluidities of mixtures of acetone with methyl alcohol, ethyl alcohol, and water, and of a few solutions of calcium nitrate in these mixtures. We have measured the conductivity of various concentrations of lithium nitrate, potassium iodide, and calcium nitrate, dissolved in the above mixtures. These conductivities, in the case of mixtures of acetone and water, exhibit the minimum in conductivity previously observed by several other workers. Moreover, this minimum in conductivity has been shown to be intimately connected with the minimum in fluidity observed in these mixtures, but the conductivity curves of different salts show marked differences. In the mixtures of acetone and the alcohols, the fluidities are what we should expect from the law of averages, i. e., the fluidity curve is nearly a straight line. From this fact we have concluded that acetone and the alco- hols thus far studied do not form more complex molecular aggregations when mixed than were originally present before mixing. The conductivities of potassium iodide, in mixtures of acetone with methyl alcohol or ethyl alcohol, are also what we should expect from the law of averages — the conductivity curves are nearly straight lines at all dilutions. Again, the conductivity has been shown to be intimately connected with fluidity. Lithium nitrate and calcium nitrate, however, give a very pronounced maximum in conductivity, in mixtures of acetone with methyl alcohol or ethyl alcohol. Evidently this was an unexpected phenomenon; to explain it, all of the factors that could reasonably influence conductivity were collected. After the elimination of several of them, the possible explanations were shown to be, either an increase in dissociation giving rise to more ions, or a diminution in the size of the ionic spheres already in the solution. It was then shown to be possible to eliminate one of these factors by the following considerations : (1) The fluidity of the mixtures of acetone and alcohol shows that there is no increase in molecular aggregation, hence we should not expect increased dissociation, if we accept the hypothesis of Dutoit and Aston. 114 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. (2) Jones and Carroll have proved that, even in the case of alcohol and water, there is practically no increase in dissociation in the mixtures. (3) Furthermore, the maximum migrates from the 25 per cent mixture, at high concentration, to the 75 per cent mixture in the more dilute solutions. This would hardly be expected if the dissociating power is greatest in a certain mixture. (4) Potassium iodide shows no tendency towards a maximum of conduc- tivity in the most concentrated solution with which we worked. (5) Very recent measurements, at extreme dilution, have failed to show any great difference in the dissociating power of the mixtures from that of the pure solvents. We, therefore, seemed justified in the conclusion that the maximum in conductivity is due to a change in the dimensions of the atmospheres about the ions. The conclusion of Dutoit and Friderich and of Jones and Carroll, that con- ductivity is proportional to the dissociation, and inversely proportional to the viscosity, has been shown to be incomplete in not taking into consideration possible changes in the size of the ionic spheres. The conductivities of lithium nitrate and calcium nitrate, in mixtures of acetone and water, again show a tendency towards a maximum, in spite of the great diminution in fluidity. Finally, it has been pointed out that the tendency to form a maximum in conductivity increases from potassium iodide, through calcium nitrate, to lithium nitrate, which seems to show these effects most strongly. This may be connected with the migration velocities of these ions. WORK OF ROUILLER. OBJECT OF THIS INVESTIGATION. This work is a direct continuation of the investigation carried out in Johns Hopkins University two years earlier by Jones and Bassett.1 They wished, if possible, to trace a connection between the phenomena of minimum conduc- tivity, first observed by Zelinsky and Krapiwin2 and later extensively studied by Jones and Lindsay,3 which solutions in certain mixtures of alcohol and water exhibit. Jones 4 and his students have extended the investigation of this problem to mixtures of other solvents, including acetone, and have obtained interesting results. It \vas, therefore, thought desirable to extend also the work of Jones and Bassett. The conductivity of silver nitrate and the transport number of its anion in binary mixtures of water, methyl alco- hol, ethyl alcohol, and acetone have been determined at two temperatures, 0° and 25°. SOLVENTS. WATER. The water used in preparing the solutions was purified essentially by the method of Jones and Mackay.5 Ordinary distilled water wras twice redistilled from an acidified solution of potassium dichromate, and the stream from the second distillation passed through a boiling solution of barium hydroxide. It had, at 0°, a conductivity of about 1.0 X 10~6. 4 METHYL ALCOHOL. The purest obtainable product was boiled 1 to 2 days with lime, distilled, and allowed to stand over anhydrous copper sulphate till needed. All dis- tillations were made through a Singer fractionating head and a block-tin condenser, and the liquid was protected during distillation from the moisture in the air by means of a soda lime U-tube. To prevent any possibility of soda-lime dust being drawn back into the liquid, the end of the tube nearest to the bottle was covered with filter paper. The first and last portions of the distillate were always discarded. The mean conductivity at 0° = 0.8 X 10" 1-6 1 Amer. Chem. Journ., 32, 409 (1904). 4 Ibid., 32, 521 (1904); 34, 481 (1905). 2 Ztschr. phys. Chem., 21, 35 (1896). 5 Ibid., 19, 83. 9 Amer. Chem. Journ., 28, 329 (1902). 115 116 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. ETHYL ALCOHOL. The best commercial article was treated in the same manner as the methyl alcohol. Its conductivity at 0° was about 0.0 X 10~7. ACETONE. The acetone was allowed to stand over fused calcium chloride before using. Conductivity at 0°= 1.0 x 1Q-6. MIXED SOLVENTS. The mixture obtained by adding n c. c. of solvent A to 100 — n c. c. of solvent B, was designated as an "n per cent A — (100 — n) per cent B mixture." This method of preparing mixed solvents was deemed preferable to the more common method of diluting n c. c. of solvent A to 100 c. c. with solvent B. In the latter case it is always necessary to state which solvent is used as diluent, and the mixture must always be allowed to cool down to the tempera- ture of the unmixed solvents before the final dilution to the mark on the measuring-flask can be effected. CONDUCTIVITY. APPARATUS. In making conductivity measurements, the usual Kohlrausch method, with Wheatstone bridge, induction coil, and telephone receiver, was used. The bridge- wire was of "manganin." The resistance coils were calibrated to within 0.04 per cent. The conductivity cells were of the type devised by Jones and Bingham1 for use with volatile solvents, which, like acetone, attack rubber and wax. They differ from the simple Arrhenius cells in that the cup is closed by a ground-glass stopper, into which are sealed with glass the tubes that carry the electrodes. The electrodes were carefully cleansed with chromic acid and covered with platinum black, by electrolyzing a dilute solution of platinic chloride. All absorbed chlorine was removed with sodium hydroxide, and the plates were washed with dilute hydrochloric acid and distilled water, dried, and heated to redness in the flame of a blast-lamp. Electrodes thus treated gave a good tone minimum, did not appreciably absorb salts from their solutions, and showed no tendency to cause the rapid oxidation of alcohol to acetic acid. When not in use the cells were filled with distilled water. Solutions were never allowed to stand in them longer than was necessary. The zero-bath was of the form commonly used in this laboratory. A tin pail was filled with finely crushed ice and water, and the cells were then packed into the ice as tightly as possible. The pail was placed in a larger, indurated 'Amer. Chem. Journ., 34, 481 (1905). CONDUCTIVITY MEASUREMENTS. 117 filter bucket, and the intervening space filled with finely crushed ice and water. Such a bath will maintain a temperature of 0° to within 0.1° for hours. The 25° bath was a galvanized-iron tub, lined on the outside with asbestos. The water in it was kept at a uniform temperature by means of a stirrer driven by a small hot-air motor, and could easily be kept to within 0.1° of any desired temperature. The thermometers were graduated to 0.1° and standardized. The burettes and measuring-flasks were all carefully calibrated by the method of Morse and Blalock.1 PREPARATION OF SOLUTIONS. The silver nitrate used in this work was obtained from Kahlbaum, and was perfectly neutral. It was finely pulverized, dried for several hours at 100° to 105°, and kept in a desiccator in the dark. Somewhat more of the salt than was necessary to prepare a N/50 solution was weighed into the solvent, and the exact concentration was determined by titration with a N/25 solution of ammonium sulphocyanate, with ferric ammonium sulphate as indicator. The sulphocyanate solution was standardized against a N/25 solution of silver nitrate and weighed quantities of thoroughly dried potassium chloride, which had been especially purified in the physical chemical laboratory of the Johns Hopkins University for use in conductivity work. The solution whose conductivity was to be measured was then made exactly N/50 by the addition of the proper amount of solvent. From this mother-solution the other solutions were prepared by successive dilution. The N/800 and N/1200 solutions were prepared from the N/400. CONDUCTIVITY MEASUREMENTS. In measuring the conductivity of a solution, readings were always made with three different resistances, and the values given are the mean. Before using, the cells and electrodes were carefully dried and rinsed out with the solution whose conductivity was to be measured. The cell constants were determined with 0.02 N and 0.004 N solutions of pure potassium chloride. The molecular conductivity of the former was taken as 129.7 at 25°. In tables 66 and 67, under v is given the concentration, expressed in number of liters of the solution containing a gram-molecular weight of the salt; under /* 0°, the molecular conductivity at 0°; and under ^25°, the molecular conductivity at 25°. The temperature coefficients are obtained by dividing the increase in con- ductivity per degree, by the conductivity at 0°. The values for the molecular conductivities in water and methyl and ethyl alcohols are obtained, by interpolation, from the measurements of Jones and Bassett, and the temperature coefficients are calculated from these inter- polated values. 1 Amer. Chem. Journ., 16, 479 (1894). 118 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 66. — Molecular conductivity of silver nitrate. 50 100 125 200 400 800 1200 In a 75 p. ct. water, 25 p. ct. acetone mixture. 38.97 40.91 42.36 41.65 43.24 45.01 78.79 81.82 85.70 85.00 88.24 91.71 Conductivity of solvent at 0°= 1.15 x 10-°. In a 50 p. ct. water, 50 p. ct. acetone mixture. AM)0 31.31 31.72 33.00 33.74 34.32 34.57 62.97 64.34 67.69 68.50 71.12 70.60 Conductivity of solvent at 0°= 1.66 x 10-6. In a 25 p. ct. water, 75 p. ct. acetone mixture. 30.79 34.59 36.49 38.75 40.35 40.26 53.04 60.05 63.81 68.34 70.63 71.28 Conductivity of solvent at 0° = 2.06 x 10-6. In acetone. 7.93 8.18 8.58 9.64 10.54 10.08 10.36 11.57 12.08 13.11 Conductivity of solvent at 0° = 1.0xlO-6. In water and acetone, and mixtures of these solvents. V Percentage of acetone at 0°. Percentage of acetone at 25°. 0 p. ct. (Bassett). 25 p. ct. . 50 p. ct. 75 p.ct. 100 p.ct. Op. ct. (Bassett). 25 p.ct. 50 p.ct. 75 p. ct. 100 p. ct. 50 63.11 38.97 31.31 30.79 111.62 78.79 62.97 53 .04 100 63.71 40.91 31.72 34.59 .... 116.82 81.82 64.34 60 .05 • • • 200 66.51 42.36 33.00 36.49 8.18 121.16 85.70 67.69 63 .81 10.36 400 70.19 41.65 33.74 38.75 8.58 125.27 85.00 68.50 68 .34 11.57 800 70.94 43.24 34.32 40.35 9.64 125.73 88.24 71.12 70 .63 12.08 1200 70.65 45.01 34.57 40.26 10.54 125.43 91.71 70.60 71 .28 13.11 In a 75 p. ct. methyl alcohol, In a 50 p. ct. methyl alcohol, In a 25 p. ct. methyl alcohol, 25 p.ct. ethyl alcohol 50 p. ct. ethyl alcohol 75 p. ct. ethyl alcohol V mixture. mixture. mixture. AM)0 M.25' AM)0 M*25° AM)0 M,25° 50 32.39 45.69 24.06 35.24 17.45 26.54 100 38.01 54.09 29.04 42.39 20.78 31.89 200 44.94 64.40 32.17 47.29 23.40 36.41 400 47.09 67.72 35.93 53.04 26.45 41.32 800 50.45 72.85 39.22 58.10 28.38 44.71 1200 51.66 74.88 40.06 59.68 28.91 46.03 Conductivity of solvent at Conductivity of solvent at Conductivity of solvent at 0°= 7.4 x 10-7. 0°=3.1 x 10-7. 0° = 3.8 x 10-7. In methyl and ethyl alcohols, and mixtures of these solvents. Percentage of ethyl alcohol at 0°. Op. ct. (Bassett). 25 p. ct. 50 p. ct. 75 p.ct. 100 p.ct. (Bassett). Percentage of ethyl alcohol at 25°. Op.ct (Bassett). 25 p.ct. 50 p.ct. 75 p. ct. 100 p.ct. (Bassett), 50 100 200 400 800 1200 41.10 46.73 52.49 57.89 32.39 38.01 44.94 47.09 50.45 51.66 24.06 29.04 32.17 35.93 39.22 40.06 17.45 20.78 23.40 26.45 28.38 28.91 11.79 13.61 15.61 17.64 55.80 64.80 72.81 82.18 45.69 54.09 64.40 67.72 72.85 74.88 35.24 42.39 47.29 53.04 58.10 59.68 26.54 31.89 36.41 41.32 44.71 46.03 17.75 21.05 24.52 27.50 CONDUCTIVITY OF SILVER NITRATE. 119 TABLE 66. — Molecular conductivity of silver nitrate. — Continued. 50 100 200 400 800 1200 In a 75 p. ct. methyl alcohol, 25 p. ct. acetone mixture. 40.45 46.36 55.60 63.98 65.62 71.85 53.81 64.24 73.65 83.08 92.69 96.70 Conductivity of solvent at 0°= 1.3xlO-«. In a 50 p. ct. methyl alcohol, 50 p. ct. acetone mixture. 38.46 48.31 58.22 67.87 77.83 81.57 45.94 61.69 68.61 83.88 100.90 100.40 Conductivity of solvent at 0°=1.36x 10-8. In a 25 p. ct. methyl alcohol, 75 p. ct. acetone mixture. AM>° /^25 25.49 28.40 33.67 37.78 42.42 47.00 51.50 59.87 63.19 75.12 66.71 78.19 In methyl alcohol, acetone, and mixtures of these solvents. Percentage of acetone at 0°. Percentage of acetone at 25°. 0 p. ct. (Bassett). 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct, Op. ct. (Bassett). 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct 50 100 200 400 800 1200 41.10 46.73 52.49 57.89 40.45 46.36 55.60 63.98 65.62 71.85 38.46 48.31 58.22 67.87 77.83 81.57 25.49 33.67 42.42 51.50 63.19 66.71 8.18 8.58 9.64 10.54 55.80 64.80 72.81 72.18 53.81 64.24 73.65 83.08 92.69 96.70 45.94 61.69 68.61 83.88 100.90 100.40 28.40 37.78 47.00 59.87 75.12 78.19 10.36 11.57 12.08 11.13 v In a 75 p. ct. ethyl alcohol, 25 p. ct. acetone mixture. In a 50 p. ct. ethyl alcohol, 50 p. ct. acetone mixture. ^25° In a 25 p. ct. ethyl alcohol, 75 p. ct. acetone mixture. 50 100 200 400 800 1200 15.86 19.90 25.06 27.40 31.51 32.85 22.39 28.38 33.64 39.82 46.30 49.04 16.91 22.13 25.72 33.55 40.51 43.90 21.42 28.12 33.64 43.49 53.51 58.69 13.38 17.42 22.20 29.80 37.84 43.76 16.36 23.06 27.60 45.80 55.33 Conductivity of solvent at 0° = 2.4 x 10~7. Conductivity of solvent at 0° = 4.6 x 10~7. Conductivity of solvent at 0° = 4.4x 10-'. In ethyl alcohol, acetone, and mixtures of these solvents. Percentage of acetone at 0°. 0 p. ct. (Bassett). 25 p. ct 50 p. ct. 75 p. ct. 100 p. ct Percentage of acetone at 25° 0 p. ct. (Bassett). 25p.ct 50 p. ct 75p.ct 100 p. ct. 50 100 200 400 800 1200 11.79 13.61 15.61 17.64 15.86 19.90 25.06 27.40 31.51 32.85 16.91 22.13 25.72 33.55 40.51 43.90 13.38 17.42 22.20 29.80 37.84 43.76 8.18 8.58 9.64 10.54 17.75 21.05 24.52 27.50 22.39 28.38 33.64 39.82 46.30 49.04 21.42 28.12 33.64 43.49 53.51 58.69 16.36 23.06 27.60 45.80 55.33 10.36 11.57 12.08 13.11 120 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 67. — Temperature coefficients of conductivity of silver nitrate. In water, acetone, and mixtures of these In ethvl alcohol, acetone, and mixtures solvents, 0° to 25°. of these solvents, 0° to 25°. V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. acetone. acetone. acetone. acetone. acetone. acetone. acetone. acetone. acetone. acetone. 50 0.0307 0.0409 0.0404 0.0289 0.0202 0.0165 0.0107 0.0089 • . • • 100 .0334 .0400 .0411 .0294 .0189 .0170 .0108 .0130 200 .0329 .0409 .0430 .0299 0.0107 .0203 .0137 .0123 .0097 0.0107 400 .0314 .0416 .0424 .0305 .0139 .0224 .0181 .0118 .0139 800 .0309 .0416 .0429 .0300 .0101 .0188 .0128 .0084 .0101 1200 .0310 .0415 .0417 .0308 .0097 .... .0197 .0134 .0106 .0097 In methyl and ethyl alcohols, and mixtures In methyl alcohol, acetone, and mixtures of these solvents, 0° to 25°. of these solvents, 0° to 25°. V Op.ct. ethyl alcohol. 25 p. ct. ethyl alcohol. 50 p. ct. ethyl alcohol. 75 p. ct. ethyl alcohol. 100 p. ct. ethyl alcohol. 0 p. ct. acetone. 25 p. ct. acetone. 50 p. ct. acetone. 75 p. ct. acetone. 100 p. ct. acetone. 50 0.0143 0.0164 0.0186 0.0208 0.0202 0.0143 0.0132 0.0078 0.0046 100 .0155 .0169 .0184 .0214 .0189 .0143 .0154 .0119 .0049 200 .0155 .0173 .0188 .0222 .0203 .0155 .0130 .0071 .0068 0.0107 400 .0168 .0175 .0190 .0225 .0224 .0168 .0119 .0119 .0065 .0139 800 .0178 .0192 .0230 .0165 .0119 .0076 .0101 1200 .0180 .0196 .0237 .0138 .0092 .0069 .0097 Some of the values in table 66 are plotted as curves in figs. 42 and 43, the abscissae representing the different percentages of acetone, and the ordinates the molecular conductivities. It will be seen that at 0°, for all dilutions but the lowest investigated, N / 50, there is a pronounced point of inflection that appears in the 75 per cent acetone mixture. At 25° it has almost dis- appeared, but still manifests itself at the higher dilutions. The curves are almost identical in form with those obtained by Jones and Bingham 1 for calcium nitrate, in mixtures of the same solvents. Table 67 shows that the temperature coefficients increase with the pro- portion of acetone up to the 50 psr cent mixture, but a further increase in the proportion of acetone produces a rapid fall in the values for the tempera- ture coefficients. With increase in concentration, the maximum value tends to shift to the 25 per cent acetone mixture. These results are also nearly identical with those obtained by Jones and Bingham with calcium nitrate. The similarity of the curves (figs. 48 and 49), plotted from the values given in table 66, to the corresponding curves obtained by Jones and Bingham for calcium nitrate, is even more striking than in the case of the methyl alcohol and acetone mixtures. A pronounced maximum manifests itself at both 0° and 25°, appearing in the 25 per cent acetone mixture in the more concentrated solutions, and shifting, with increase in dilution, through the 50 per cent to the 75 per cent mixture. 1 Amer. Chem. Journ., 34, 481 (1905). CONDUCTIVITY OF SILVER NITRATE. 121 130- 130- 110- 100- 90- ft 80- c 3 §70- 8 60- o a) -3 50- 40- 30- ?0- 10- 70- 65 Percentage of Acetone FIG. 42. — CONDUCTIVITY OP SILVER NITRATE IN MIXTURES OF WATER AND ACETONE AT 0°. 25^ 50$ 75^ Percentage of Acetone FIG. 43. — CONDUCTIVITY OF SILVER NITRATE IN MIXTURES OF WATER AND ACETONE AT 25°. 100 Jt 122 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. 60- 55- 50- 45- .-S .40- |35- o, O 30- t-i g.25- 3) •3 S.20- 15- 10- 6- Percentage of Ethyl Alcohol FlQ. 44. — CONDtTCTIVITT OF SILVER NlTBATE IN MIXTURES OF MKTHYL AND ETHYL ALCOHOLS AT 0°. 90- 80 ,70 § 60- •d c°3 50- s 40- u | 30- 20- JO- .25$ 505; Percentage of 7554 100 } Fia. 45. — CONDUCTIVITY OF SILVER NITRATE IN MIXTURES OF METHYL AND ETHYL ALCOHOLS AT 25°. CONDUCTIVITY OF SILVER NITRATE. 123 As with calcium nitrate, the temperature coefficients (table 67) show a minimum in the 75 per cent mixture. In view of the great similarity in conductivity phenomena exhibited by silver and calcium nitrates, in mixtures of water and of methyl and ethyl joo- 'Percentage of Acetone FIG. 46. — CONDUCTIVITY OF SILVER NITRATE IN MIXTURES OF METHYL ALCOHOL AND ACETONE AT 0°. 25 f> 50$ Percentage of Acetone FIG. 47. — CONDUCTIVITY OF SILVER NITRATE IN MIXTURES OF METHYL ALCOHOL AND ACETONE AT 25°. 124 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. alcohols with acetone, it is interesting to know whether this is also true of solutions of the two salts in mixtures of water with the alcohols. By com- paring the values for calcium nitrate in these mixtures, obtained by Jones 45 H 40- >> -b> !H o 3 £ o O 35- 225° ture coef- cient. cient. ficient. 2 42.09 76.17 0.0324 24.46 49.86 0.0415 18.42 37.01 0.0404 5 45.33 83.33 .0335 26.57 54.90 .0426 20.57 41.58 .0408 10 47.25 86.09 .0329 27.52 57.46 .0435 21.71 43.99 .0410 50 51.43 95.62 .0344 30.04 63.69 .0448 24.24 49.46 .0416 100 51.92 96.41 .0342 30.50 64.78 .0450 24.40 50.27 .0424 200 53.27 99.51 .0347 31.24 66.25 .0448 25.16 51.74 .0423 400 53.71 100.34 .0347 31.65 66.65 .0442 25.96 53.24 .0420 800 55.45 104.81 .0356 32.87 68.83 .0438 26.42 54.48 .0425 1600 56.12 106.23 .0357 33.07 69.24 .0437 26.91 55.15 .0420 In a mixture of 75 p. ct. methyl alcohol and water. In methyl alcohol. In a mixture of 25 p. ct. ethyl alcohol and water. V ^25° Tempera- Tempera- f^0° Tempera- MtO° ture coeffi- M»0° ^,25° ture coeffi- A^250 ture coef- cient. cient. ficient. 2 19.51 33.66 0.0290 22.02 31.35 0.0169 17.82 41.21 0.0525 5 21.67 38.31 .0307 30.55 42.57 .0157 18.92 45.03 .0552 10 23.50 41.99 .0314 35.92 50.21 .0159 19.66 47.05 .0557 50 27.05 48.68 .0320 46.48 64.92 .0159 21.60 52.52 .0572 100 27.73 49.98 .0321 49.36 69.19 .0161 21.67 52.58 .0570 200 28.92 52.29 .0323 52.51 73.62 .0161 22.09 54.37 .0584 400 29.55 53.56 .0325 55.18 78.25 .0167 22.50 54.41 .0567 800 30.22 55.29 .0332 57.45 82.56 .0175 24.51 56.46 .0521 1600 31.35 57.59 .0335 57.63 83.64 .0181 25.72 57.10 .0448 In a mixture of 50 p. ct. ethyl In a mixture of 75 p. ct. ethyl alcohol and water. alcohol and water. In ethyl alcohol. V /aO° ^25° Temperature ,, no coefficient. ^"u /^.25° Temperature coefficient. M,,0° ^25° Temperature coefficient. 2 11.70 27.50 0.0540 9.20 18.96 0.0424 6.11 10.18 0.0266 5 12.48 29.51 .0546 10.55 22.11 .0438 8.67 14.00 .0246 10 12.62 31.41 .0595 11.59 24.26 .0437 10.55 17.22 .0253 50 13.71 35.41 .0633 11.87 25.62 .0463 14.40 23.28 .0247 100 13.98 36.12 .0633 12.14 26.29 .0466 15.69 25.57 .0252 200 14.34 37.26 .0639 12.77 27.66 .0466 17.29 28.26 .0254 400 14.96 38.39 .0626 13.13 28.62 .0472 18.55 30.32 .0254 800 15.77 39.94 .0613 13.62 29.78 .0474 19.71 31.73 .0244 1600 16.06 40.01 .0596 13.83 29.91 .0465 20.79 33.36 .0242 130 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 68. — Conductivity of lithium bromide at 0° and 25°. — Continued. In a mixture of 25 p. ct. methyl alcohol and ethyl alcohol. In a mixture of 50 p. ct. methyl alcohol and ethyl alcohol. In a mixture of 75 p. ct. methyl alcohol and ethyl alcohol. M'O- ^25° Temperature coefficient. *Q- Ai.,250 Temperature coefficient. ^ M,25° Temperature coefficient. 2 8.59 13.51 0.0229 12.22 18.37 0.0201 16.46 24.22 0.0188 5 12.55 19.08 .0208 17.38 25.54 .0188 23.34 33.38 .0172 10 15.17 22.97 .0206 21.03 30.72 .0184 27.71 39.60 .0171 50 21.07 31.56 .0199 28.69 41.81 .0183 37.02 52.80 .0170 100 22.70 34.55 .0209 30 .42 44.48 .0185 38.96 55.92 .0174 200 24.50 37.61 .0214 32 .95 48.86 .0193 41.64 60.03 .0176 400 26.08 40.03 .0214 34 .76 51.31 .0190 44.21 63.76 .0177 800 26.94 41.82 .0221 35 .83 53.29 .0195 46.51 66.73 .0174 1600 29.08 45.07 .0220 37 .53 55.74 .0194 49.57 71.37 .0176 In a mixture of 25 p. ct. In a mixture of 50 p. ct. In a mixture of 75 p. ct. acetone and water. acetone and water. acetone and water. V M° M»25° Temperature coefficient. *r ^,25° Temperature coefficient. M» ,,25' Temperature coefficient. 5 27.58 56.23 0.0415 20.37 41.82 0.0421 17.40 31.47 0.0323 10 28.82 59.71 .0429 21 .70 45.16 .0432 20.40 36.49 .0315 50 31.49 65.64 .0434 24 .81 51.95 .0437 25.76 46.77 .0326 100 31.60 65.82 .0433 25.03 52.46 .0438 27.29 49.87 .0331 200 32.29 67.88 .0441 26 .10 54.64 .0437 28.87 52.87 .0332 400 32.98 68.69 .0433 26 .81 56.44 .0442 30.10 55.53 .0338 800 35.18 70.25 .0399 27 .07 57.54 .0430 30.20 56.46 .0348 1600 35.70 71.47 .0401 28 .34 59.28 .0437 31.65 61.10 .0372 In acetone. In a mixture of 25 n. ct. In a mixture of 50 p. ct. acetone and methyl alcohol. acetone and methyl alcohol. V *«P ^25° Temperature coefficient. ^.00 /*,.25° Temperature coefficient. *flP M.25° Temperature coefficient. 5 9.35 10.82 0.00629 29 .65 40.04 0.0140 27.99 35.37 0.0105 10 11.91 14.08 .00729 35 .03 47.57 .0143 34.27 42.32 .00939 50 22.36 26.77 .00789 46 .71 63.44 .0143 48.65 61.77 .0108 100 28.86 34.52 .00784 49 .68 67.40 .0142 52.67 68.29 .0118 200 37.33 47.45 .01080 53 .28 73.81 .0154 57.69 74.13 .0114 400 48.28 57.96 .00802 56 .45 78.11 .0153 61.74 79.77 .0117 800 59.42 72.16 .00858 58 .91 81.55 .0154 64.70 85.57 .0129 1600 70.89 85.90 .00847 60 .38 83.53 .0153 67.02 89.45 .0134 In a mixture of 75 p. ct. acetone and In a mixture of 25 p. ct. acetone and methyl alcohol ethyl alcohol. V Temperature Temperature M«0 Mv25° coefficient. M> ,,,25 coefficient. 5 23.36 30.06 0.0115 11.80 16.94 0.0174 10 29.77 38.15 .0112 14.72 20.91 .0168 50 48.45 61.89 .0111 21.70 31.00 .0171 100 55.00 70.87 .0115 23.70 34.38 .0180 200 62.74 80.63 .0114 26.38 38.27 .0180 400 72.25 91.17 .0105 28.28 41.74 .0190 800 82.20 100.28 .0088 28.88 42.97 .0195 1600 84.15 106.14 .0104 29.21 44.24 .0206 LITHIUM BROMIDE. 131 TABLE 68. — Conductivity of lithium bromide at 0° and 25°. — Continued. In a mixture of 50 p. ct. acetone and ethyl alcohol. In a mixture of 75 p. ct. acetone and ethyl alcohol. V o Temperature Temperature M" coefficient. fJ-vO Ho 5 coefficient. 5 14.70 19.38 0.0122 14.48 18.07 0.00992 10 19.23 25.75 .0135 19.16 23.23 .00850 50 30.25 39.36 .0120 33.20 40.08 .00829 100 34.50 45.12 .0123 39.32 48.94 .00978 200 39.20 52.76 .0138 47.31 58.01 .00905 400 43.83 59.26 .0141 55.55 68.28 .00917 800 46.92 64.33 .0148 61.20 77.57 .01070 1600 50.98 71.22 .0159 66.28 83.36 .01030 TABLE 69. — Comparison of the conductivities of lithium bromide. In mixtures of methyl alcohol and water at 0° and at 25°. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p.ct. 0 p. ct. 25 p. ct. 50 p. ct. 75 p.ct. 100 p.ct. 2 42.09 24.46 18.42 19.51 22.02 76.17 49.86 37.01 33.66 31.35 5 45.33 26.57 20.57 21.67 30.55 83.33 54.90 41.58 38.31 42.57 10 47.25 27.52 21.71 23.50 35.92 86.09 57.46 43.99 41.99 50.21 50 51.43 30.04 24.24 27.05 46.48 95.62 63.69 49.46 48.68 64.92 100 51.92 30.50 24.40 27.73 49.36 96.41 64.78 50.27 49.98 69.19 200 53.27 31.24 25.16 28.92 52.51 99.51 66.25 51.74 52.29 73.62 400 53.71 31.65 25.96 29.55 55.18 100.34 66.65 53.24 53.56 78.25 800 55.45 32.87 26.42 30.22 57.45 104.81 68.83 54.48 55.29 82.56 1600 56.12 33.07 26.91 31.35 57.63 106.23 69.24 55.15 57.59 83.64 In mixtures of ethyl alcohol and water at 0° and at 25°. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p.ct. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. lOOp.ct. 2 42.09 17.82 11.70 9.20 6.11 76.17 41.21 27.50 18.96 10.18 5 45.33 18.92 12.48 10.55 8.67 83.83 45.03 29.51 22.11 14.00 10 47.25 19.66 12.62 11.59 10.55 86.09 47.05 31.41 24.26 17.22 50 51.43 21.60 13.71 11.87 14.40 95.62 52.52 35.41 25.62 23.28 100 51.92 21.67 13.98 12.14 15.69 96.41 52.58 36.12 26.29 25.57 200 53.27 22.09 14.34 12.77 17.29 99.51 54.37 37.26 27.66 28.26 400 53.71 22.50 14.96 13.13 18.55 100.34 54.41 38.39 28.62 30.32 800 55.45 24.51 15.77 13.62 19.71 104.81 56.46 39.94 29.78 31.73 1600 56.12 25.72 16.06 13.83 20.79 106.23 57.10 40.01 29.91 33.36 In mixtures of methyl alcohol and ethyl alcohol at 0° and at 25°. V C2H6OH 25 p.ct. CH3OH 50 p.ct. CH3OH 75 p. ct. CH3OH CH3OH C«H6OH 25 p.ct. CH3OH 50 p.ct. CH3OH 75 p.ct. CH3OH CH3OH 2 6.11 8.59 12.22 16.46 22.02 10.18 13.51 18.37 24.22 31.35 5 8.67 12.55 17.38 23.34 30.55 14.00 19.08 25.54 33.38 42.57 10 10.55 15.17 21.03 27.71 35.92 17.22 22.97 30.72 39.60 50.21 50 14.40 21.07 28.69 37.02 46.48 23.28 31.56 41.81 52.80 64.92 100 15.69 22.70 30.42 38.96 49.36 25.57 34.55 44.48 55.92 69.19 200 17.27 24.50 32.95 41.64 52.51 28.26 37.61 48.86 60.03 73.62 400 18.55 26.08 34.76 44.21 55.18 30.32 40.03 51.31 63.76 78.25 800 19.71 26.94 35.83 46.51 57.45 31.73 41.82 53.29 66.73 82.56 1600 20.79 29.08 37.53 49.57 57.63 33.36 45.07 55.74 71.37 83.64 132 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 69. — Comparison of the conductivities of lithium bromide. — Continued. In mixtures of acetone and water at 0". In mixtures of acetone and water at 25°. V Op. ct. 25 p. ct. 50 p. ct. 75 p. ct. lOOp.ct 0 p. ct. 25p.ct. 50 p. ct. 75p.ct. lOOp.ct. 5 45.33 27.58 20.37 17.40 9.35 83.33 56.23 41.82 31.47 10.82 10 47.25 28.82 21.70 24.00 11.91 86.09 59.71 45.16 36.49 14.80 50 51.43 31.49 24.81 25.76 22.36 95.61 65.64 51.95 46.77 26.77 100 51.92 31.60 25.03 27.29 28.86 96.41 65.82 52.46 49.87 34.52 200 53.27 32.29 26.10 28.87 37.33 99.51 67.88 54.64 52.87 47.45 400 53.71 32.98 26.81 30.10 48.28 100.34 68.69 56.44 55.53 57.96 800 55.45 35.18 27.07 30.20 59.42 104.81 70.25 57.54 56.46 72.16 1600 56.12 35.71 28.34 31.65 70.89 106.23 71.47 59.28 61.10 85.90 In mixtures of acetone and methyl alcohol at 0° and at 25°. V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 0 p. ct. 25p.ct. 50p.ct. 75 p. ct. 100 p. ct. 5 30.55 29.65 27.99 23.36 9.35 42.57 40.04 35.37 30.06 10.82 10 35.92 35.03 34.27 29.77 11.91 50.21 47.57 42.32 38.15 14.08 50 46.48 46.71 48.65 48.45 22.36 64.92 63.44 61.77 61.89 26.77 100 49.36 49.68 52.67 55.00 28.86 69.19 67.40 68.29 70.87 34.52 200 52.51 53.28 57.69 62.74 37.33 73.62 73.81 74.13 80.63 47.45 400 55.18 56.45 61.74 72.25 48.28 78-25 78.11 79.77 91.17 57.96 800 57.45 58.91 64.70 82.20 59.42 82.56 81.55 85.57 100.28 72.16 1600 57.63 60.38 67.02 84.15 70.89 83.64 83.53 89.45 106.14 85.90 In mixtures of acetone and ethyl alcohol at 0° and at 25°. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. et. 100 p. ct. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 5 8.67 11.80 14.70 14.48 9.35 14.00 16.94 19.38 18.07 10.82 10 10.55 14.72 19.23 19.16 11.91 17.22 20.91 25.75 23.23 14.08 50 14.40 21.70 30.25 33.20 22,36 23.28 31.00 39.36 40.08 26.77 100 15.69 23.70 34.50 39.32 28.86 25.57 34.38 45.12 48.94 34.52 200 17.29 26.38 39.20 47.31 37.33 28.26 38.27 52.76 58.01 47.45 400 18.55 28.28 43.83 55.55 48.28 30.32 41.74 59.26 68.28 57.96 800 19.71 28.88 46.92 61.20 59.42 31.73 42.97 64.33 77.57 72.16 1600 20.79 29.21 50.98 66.28 70.89 33.36 44.24 71.22 83.36 85.90 TABLE 70. — Comparison of the temperature coefficients of conductivity of lithium bromide fvom* 0° to 25°. V In mixtures of methyl alcohol and water. In mixtures of ethyl alcohol and water. Op. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 2 0.0324 0.0415 0.0404 0.0290 0.0169 0.0324 0.0525 0.0540 0.0424 0.0266 5 .0335 .0426 .0408 .0307 .0157 .0335 .0552 .0546 .0438 .0246 10 .0329 .0435 .0410 .0314 .0159 .0329 .0557 .0595 .0437 .0253 50 .0344 .0448 .0416 .0320 .0159 .0344 .0572 .0633 .0463 .0247 100 .0342 .0450 .0424 .0321 .0161 .0342 .0570 .0633 .0466 .0252 200 .0347 .0448 .0423 .0323 .0161 .0347 .0584 .0639 .0466 .0254 400 .0347 .0442 .0420 .0325 .0167 .0347 .0567 .0626 .0472 .0254 800 .0356 .0438 .0425 .0332 .0175 .0356 .0521 .0613 .0474 .0244 1600 .0357 .0437 .0420 .0335 .0181 .0357 .0448 .0596 .0465 .0242 LITHIUM BROMIDE. 133 TABLE 70. — Comparison of the temperature coefficients of conductivity of lithium bromide from 0° to 25°. — Continued. In mixtures of methyl alcohol and ethyl alcohol. In mixtures of acetone and water. V C2H5OH 25 p. ct. CH3OH 50 p. ct. CH3OH 75 p. ct. CH3OH CH3OH Op. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 2 00266 0 0229 0 0201 00188 0 0169 5 .0246 .0208 .0188 .0172 .0157 0.0335 0.0415 0.0421 0.0323 0.00629 10 .0253 .0206 .0184 .0171 .0159 .0329 .0429 .0432 .0315 .00729 50 .0247 .0199 .0183 .0170 .0159 .0344 .0434 .0437 .0326 .00789 100 .0252 .0209 .0185 .0174 .0161 .0342 .0433 .0438 .0331 .00784 200 .0254 .0214 .0193 .0176 .0161 .0347 .0441 .0437 .0332 .01080 400 .0254 .0214 .0190 .0177 .0167 .0347 .0433 .0442 .0338 .00802 800 .0244 .0221 .0195 .0174 .0175 .0356 .0399 .0430 .0348 .00858 1600 .0242 .0220 .0194 .0176 .0181 .0357 .0401 .0437 .0372 .00847 In mixtures of acetone and methyl alcohol. In mixtures of acetone and ethyl alcohol. V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 5 0.0157 0.0140 0.0105 0.0115 0.00629 0.0246 0.0174 0.0122 0.00992 0.00629 10 .0159 .0143 .00939 .0112 .00729 .0253 .0168 .0135 .00850 .00729 50 .0159 .0143 .0108 .0111 .00789 .0247 .0171 .0120 .00829 .00789 100 .0161 .0142 .0118 .0115 .00784 .0252 .0180 .0123 .00978 .00784 200 .0161 .0154 .0114 .0114 .01080 .0254 .0180 .0138 .00905 .01080 400 .0167 .0153 .0117 .0105 .00802 .0254 .0190 .0141 .00917 .00802 800 .0175 .0154 .0129 .0088 .00858 .0244 .0195 .0148 .01070 .00858 1600 .0181 .0153 .0134 .0104 .00847 .0242 .0206 .0159 .01030 .00847 60 1 "Percentage oOfethyl Alcohol FIG. 60. — CONDUCTIVITY OF LITHIUM BROMIDE IN MIXTURES OF METHYL ALCOHOL AND WATER AT 0°. 134 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. Table 68 (figs. 50 and 51) shows that lithium bromide, in mixtures of methyl alcohol and water, gives a pronounced minimum in conductivity. The minimum is more marked at 0° than at 25°. At 25° the minimum occurs in the 75 per cent mixture up to v = 100. Beyond this dilution the minimum occurs solely in the 50 per cent mixture. At 0° the minimum appears in the 50 per cent mixture alone. We also notice that at 0° the values of \iv for the pure methyl alcohol at the higher dilutions exceed the values of /*„ for the corresponding aqueous solutions. These points will be made clear by a study of the figures. In all cases the curves represent the molecular conductivities at the successive dilutions. 50 f, 755t 10054 Percentage of Methyl Alcohol FIG. 51. — CONDUCTIVITY OF LITHIUM BROMIDE IN MIXTURES OF METHYL ALCOHOL AND WATER AT 25°. The temperature coefficients of conductivity increase with the dilution, and they are also greater in the mixtures than in the pure solvents, the max- imum appearing in the 25 per cent mixture. The temperature coefficients of conductivity of salts in water generally increase with the dilution, as Jones l has pointed out in a recent article. Table 68 (figs. 52 and 53) shows that lithium bromide, in mixtures of ethyl alcohol and water, also gives a minimum in conductivity. At high concentrations the minimum does not appear, either at 0° or at 25°. At 'Amer. Chem. Journ., 35, 445 (1906). LITHIUM BROMIDE. 135 100$ Percentage of Ethyl Alcohol FIG. 52. — CONDUCTIVITY OF LITHIUM BROMIDE IN MIXTURES OF ETHYL ALCOHOL AND WATER AT 0°. 10 25$ 50$ 75$ 100$ Percentage of Ethyl Alcohol FIG. 53. — CONDUCTIVITY OF LITHIUM BROMIDE IN MIXTURES OF ETHYL ALCOHOL AND WATER AT 25°. 136 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. both temperatures the minimum appears only in the 75 per cent mixture. At 0° the minimum begins with v — 5Q; at 25° it begins at v = 2QQ°. It is to be noticed that the minimum is more marked at the lower temperature, being very slight at 25°. The temperature coefficients in ethyl alcohol are almost constant. This same fact was found by Jones and Bingham l in the case of calcium nitrate. The temperature coefficients are greater in the mixtures than in the pure solvents, the maximum appearing in the 50 per cent mixture. 75 Jf 10055 Percentage of Methyl Alcohol FIG. 54. — CONDUCTIVITY OF LITHIUM BKOMIDE IN MIXTURES OF METHYL ALCOHOL AND ETHYL ALCOHOL AT 0°. Tables 68 and 69 (figs. 54 and 55) show that lithium bromide, in mixtures of methyl and ethyl alcohols, gives no minimum in conductivity. In fact, the conductivity values for the solutions in the mixed solvent approach the mean value of the conductivities in the pure solvents. The conductivity curves at the high concentrations show a slight sagging at both temperatures, but beyond v = 50 the curves are nearly straight lines. This same fact has been observed by others working with mixtures of methyl and ethyl alcohols. The temperature coefficients obey, approximately, the law of averages. 1Amer. Chem. Journ., 34, 529 (1905). LITHIUM BROMIDE. 137 Percentage of Methyl Alcohol FIG. 55. — CONDUCTIVITY OF LITHIUM BROMIDE IN MIXTURES OF METHYL ALCOHOL AND ETHYL ALCOHOL AT 25°. 70 50 56 75# Percentage of Acetone FIG. 56. — CONDUCTIVITY OF LITHIUM BROMIDE IN MIXTURES OF ACETONE AND WATER AT 0°. 138 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. Table 68 (figs. 56 and 57), for lithium bromide in mixtures of acetone and water, gives the minimum which is exhibited in mixtures of the alcohols and water under similar circumstances. The minimum is more marked at the lower temperature. The curves diverge from each other rapidly between the 75 per cent mixture and pure acetone at both temperatures. This seems to indicate that the addition of small amounts of water greatly increases the dissociation. It is also to be noticed that, at the lower temperature, the values of /u.B for the pure acetone at the higher dilutions exceed those of p.v for the corresponding aqueous solutions. A similar phenomenon was noticed in the solution in methyl alcohol and water. 110- 25$ 5054 75j« 100 # Percentage of Acetone FIG. 57. — CONDUCTIVITY OF LITHIUM BROMIDE IN MIXTURES OF ACETONE AND WATER AT 25°. In pure acetone the temperature coefficients increase with the dilution. In the mixtures the increase is very small. Tables 69 and 70 (figs. 58 and 59), for lithium bromide in mixtures of acetone and methyl alcohol, give a maximum in conductivity in the 75 per cent mixture at both temperatures. The maximum is increased by rise in temperature. This same phenomenon was found by Jones and Bingham, working with lithium nitrate in mixtures of acetone and methyl alcohol. It will be recalled that we obtained a minimum conductivity with this salt in mixtures of the alcohols and water. COBALT CHLORIDE. 139 Tables 69 and 70 (figs. 60 and 61), for lithium bromide in mixtures of acetone and ethyl alcohol, show the same characteristics as were observed in the tables for this salt in mixtures of acetone and methyl alcohol. The values for //,„ in acetone are greater than the corresponding values in the pure ethyl alcohol at practically all dilutions. CONDUCTIVITY AND VISCOSITY OF CERTAIN SALTS. The temperature coefficients increase slightly with the increase in dilution. The values are highest in ethyl alcohol, from which there is a regular grada- tion to the values of pure acetone. 100* Percentage of Acetone FIG. 58. — CONDUCTIVITY OF LITHIUM: BROMIDE IN MIXTURES OF ACETONE AND METHYL ALCOHOL AT 0°. COBALT CHLORIDE. The cobalt chloride used in this work was obtained from Kahlbaum. No appreciable impurity could be detected. This salt can not be dehydrated in contact with the air, and special precautions must be taken to prevent the formation of the oxychloride. The salt containing 6 molecules of water was first placed over concentrated sulphuric acid, in a vacuum desiccator, for several days. It was thus deprived of part of its water of crystallization. It was then placed in an air-bath and dried at 140° to 150°, in a stream of dry hydrochloric acid gas. The salt was subsequently kept in a vacuum desic- cator over sulphuric acid and potassium hydroxide. It gave no test for free hydrochloric acid, and possessed a pale, sky-blue color. 140 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. 75$ 1005$ Percentage of Acetone FIG. 59. — CONDUCTIVITY OF LITHIUM BROMIDE IN MIX- TURES OF ACETONE AND METHYL ALCOHOL AT 25°. .25$ 50^ 75jt 100;* Percentage of Acetone FIG. 60. — CONDUCTIVITY OF LITHIUM BROMIDE IN MIX- TURES OF ACETONE AND ETHYL ALCOHOL AT 0°. COBALT CHLORIDE. 141 Cobalt chloride, dissolved in methyl and ethyl alcohols and acetone, gives rise to a number of color phenomena. Spectroscopic observations concerning the color changes of cobalt chloride in water, methyl and ethyl alcohols, acetone, and binary mixtures of these solvents have been carried out in the physical chemical laboratory of the Johns Hopkins University by Jones and Uhler. The results of this work have been published by the Carnegie Institution of Washington.1 25^ 507. Percentage of Acetone FIG. 61. — CONDUCTIVITY OF LITHIUM BROMIDE IN MIX- TURES OF ACETONE AND ETHYL ALCOHOL AT 25°. Hydrolysis probably comes into play, to some extent, in the more dilute lueous solutions. TABLE 71. — Conductivity of cobalt chloride. V In water at 0° and 25°. In a mixture of 25 p. ct. methyl alcohol and water at 0° and 25°. In a mixture of 50 p. ct. methyl alcohol and water at 0° and 25°. ^0° ^25° Temperature coefficient. M,0° ^25° Temperature coefficient. *o- ,,25° Temperature coefficient. 10 50 100 200 400 800 1600 86.42 104.31 109.27 113.73 115.87 116.71 117.40 156.36 189.27 197.41 213.40 219.63 220.04 221.50 0.0324 .0326 .0322 .0350 .0358 .0354 .0355 43.97 52.02 52.39 57.20 57.99 60.14 61.98 92.35 110.24 112.95 126.60 125.17 127.46 133.98 0.0440 .0447 .0462 .0485 .0463 .0447 .0472 32.94 39.96 41.39 44.95 46.15 48.51 51.40 65.85 82.44 85.58 92.90 96.35 100.54 104.19 0.0400 .0425 .0427 .0427 .0435 .0429 .0411 Publication No. 60, "Hydrates in Aqueous Solution." 142 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 71. — Conductivity of cobalt chloride. — Continued. V In a mixture of 75 p. ct. methyl alcohol and water at 0° and 25°. In methyl alcohol at 0° and 25°. fn a mixture of 25 p. ct. ethyl alcohol and water at 0° and 25°. M*0° /^25° Temperature coefficient. M*0° 1^25° Temperature coefficient. /O>° Mt<25° Temperature coefficient. 10 50 100 200 400 800 1600 27.40 37.45 38.07 43.16 45.54 48.55 51.39 49.58 61.58 66.17 76.01 77.82 80.51 84.39 0.0324 .0258 ,0296 .0304 .0283 .0263 .0257 33.46 51.76 60.89 70.45 75.64 86.57 95.54 41.78 65.22 76.08 87.37 99.96 117.18 133.33 0.00995 .01040 .00998 .00961 .01280 .01410 .01580 30.53 33.74 34.64 35.64 37.63 41.89 42.42 73.68 85.65 87.60 94.42 97.64 101.33 103.86 0.0565 .0615 .0611 .0660 .0638 .0568 .0579 V In a mixture of 50 p. ct. ethy; alcohol and water at 0° and 25°. In a mixture of 75 p. ct. ethyl alcohol and water at 0° and 25°. In ethyl alcohol at 0° and 25°. /M)° Ait,25° Temperature coefficient. M*0° M»25° Temperature coefficient. M*0° ^25° Temperature coefficient. 10 50 100 200 400 800 1600 18.17 21.55 22.46 23.83 24.76 26.46 28.68 44.75 55.21 57.41 63.42 64.00 69.34 70.51 0.0585 .0625 .0622 .0664 .0634 .0648 .0583 15.21 19.85 21.34 24.02 25.67 27.74 29.61 30.39 41.12 44.72 51.70 54.45 59.56 63.87 0.0399 .0428 .0438 .0461 .0448 .0459 .0463 '6.06 10.59 12.79 15.43 17.66 20.70 23.99 7.64 13.75 17.33 20.93 24.18 28.45 33.59 0.0104 .0119 .0142 .0142 .0148 .0150 .0160 V In a mixture of 25 p. ct. methyl alcohol and ethyl alcohol at 0° and 25°. In a mixture of 50 p. ct. methvl alcohol and ethyl alcohol at 0° and 25°. In a mixture of 75 p. ct. methyl alcohol and ethyl alcohol at 0° and 25°. ^0° ^25° Temperature coefficient. Mtfl" Mr25° Temperature coefficient. M*0° /^25° Temperature coefficient. 10 50 100 200 400 800 1600 8.82 14.66 17.32 20.23 23.75 28.34 32.23 10.76 17.61 21.75 26.58 30.08 35.23 41.14 0.00880 .00805 .01022 .01255 .01070 .00972 .01106 12.89 21.46 24.31 28.63 32.28 39.21 46.28 16.89 27.38 30.39 35.89 41.80 48.22 58.73 0.0124 .0110 .0100 .0101 .0118 .0092 .0108 22.49 35.33 40.79 48.55 54.78 63.06 69.29 25.53 44.31 49.90 60.62 69.42 81.13 95.07 0.00541 .01020 .00893 .00994 .01070 .01150 .01490 V In a mixture of 25 p. ct. acetone and water at 0° and 25°. In a mixture of 50 p. ct. acetone and water at 0° and 25°. /^o° oso Temperature coefficient. /u*0° M.250 Temperature coefficient. 100 200 400 800 1600 64.40 67.88 69.13 72.93 79.34 115.38 0.0316 118.42 .0298 122.39 .0308 126.88 .0296 134.69 .0279 54.42 57.79 60.29 64.54 67.80 96.89 106.88 112.70 117.97 125.54 0.0312 .0340 .0348 .0331 .0341 V In a mixture of 75 p. ct. acetone and water at 0° and 25°. In acetone at 0° and 25°. AO>° ,i OKO Temperature ^^ coefficient. M.0° ^,25° Temperature coefficient. 100 200 400 800 1600 33.14 37.98 53.41 61.16 68.51 70.57 0.0452 84.48 .0490 94.84 .0310 109.76 .0318 124.95 .0329 10.18 10.96 11.63 12.66 12.79 9.47 9.70 9.94 10.11 10.45 - 0.00279 - .00460 .00581 .00806 - .00732 COBALT CHLORIDE. 143 TABLE 71. — Conductivity of cobalt chloride. — Continued. V In a mixture of 25 p. ct. acetone and methyl alcohol at 0° and 25°. In a mixture of 50 p. ct. acetone and methyl al- cohol at 0° and 25°. In a mixture of 75 p. ct. acetone and methyl alcohol at 0° and 25°. ^0° ^25° Tempera- ture coeffi- cient. ^0° Mv25« Tempera- ture coeffi- cient. M«0» M«25° Temperature coefficient. 100 200 400 800 1600 48.10 57.33 64.43 74.60 91.57 56.28 64.87 73.04 84.46 100.09 0.00680 .00526 .00534 .00529 .00372 35.31 42.12 49.39 56.97 67.55 39.20 46.94 55.47 66.85 74.11 0.00441 .00458 .00492 .00694 .00388 23.85 28.72 32.74 42.05 50.94 24.83 28.67 31.85 38.69 47.20 4- 0.00164 - .00007 - .00108 - .00319 - .00294 V In a mixture of 25 p. ct. ace- tone and ethyl alcohol at 0° and 25°. In a mixture of 50 p. ct. ace- tone and ethyl alcohol at 0° and 25°. In a mixture of 75 p. ct. acetone and ethyl alcohol at 0° and 25°. M,00 M,25° Tempera- ture coeffi- cient. MvO° ^25° Temperature coefficient. MvO° M,25° Temperature coefficient. 100 200 400 800 1600 14.40 18.05 21.19 26.10 31.80 15.43 19.47 23.67 30.20 36.41 0.00286 .00314 .00562 .00628 .00580 12.78 15.20 17.95 22.47 28.67 12.41 13.91 15.78 18.76 23.47 -0.00116 - .00339 - .00483 - .00660 - .00726 13.46 14.41 14.69 15.74 17.28 11.95 12.49 12.69 13.48 14.19 -0.00449 - .00533 - .00544 - .00574 - .00715 TABLE 72. — Comparison of the conductivities of cobalt chloride. V In mixtures of methyl alcohol and water at 0°. In mixtures of methyl alcohol and water at 25°. Op.ct. 25 p.ct. 50 p.ct. 75 p. ct. 100 p.ct. Op.ct. 25 p. ct. 50 p.ct. 75 p.ct. 100 p.ct. 10 50 100 200 400 800 1600 86.42 104.31 109.27 113.73 115.87 116.71 117.40 43.97 52.02 52.39 57.20 57.99 60.14 61.98 32.94 39.96 41.39 44.95 46.15 48.51 51.40 27.40 37.45 38.07 43.16 45.54 48.55 51.39 33.46 51.76 60.89 70.45 75.64 86.57 95.54 156.36 189.27 197.41 213.40 219.63 220.04 221.50 92.35 110.24 112.95 126.60 125.17 127.46 133.98 65.85 82.44 85.58 92.90 96.35 100.54 104.19 49.58 61.58 66.17 76.01 77.82 80.51 84.39 41.78 65.22 76.08 87.37 99.96 117.18 133.33 V In mixtures of ethyl alcohol and water at 0°. In mixtures of ethyl alcohol and water at 25°. 10 50 100 200 400 800 1600 86.42 104.31 109.27 113.73 115.87 116.71 117.40 30.53 33.74 34.64 35.64 37.63 41.89 42.42 18.17 21.55 22.46 23.83 24.76 26.46 28.68 15.21 19.85 21.34 24.02 25.67 27.74 29.61 6.06 10.59 12.79 15.43 17.66 20.70 23.99 156.36 189.27 197.41 213.40 219.63 220.40 221.50 73.68 85.65 87.60 94.42 97.64 101.33 103.86 44.75 55.21 57.41 63.42 64.00 69.34 70.51 30.39 41.12 44.72 51.70 54.45 59.56 63.87 7.64 13.75 17.33 20.93 24.18 28.45 33.59 V In mixtures of methyl alcohol and ethyl alcohol at 0°. In mixtures of methyl alcohol and ethyl alcohol at 25°. 10 50 100 200 400 800 1600 6.06 10.59 12.79 15.43 17.66 20.70 23.99 8.82 14.66 17.32 20.23 23.75 28.34 32.23 12.89 21.46 24.31 28.63 32.28 39.21 46.28 22.49 35.33 40.79 48.55 54.78 63.06 69.29 33.46 51.76 60.89 70.45 75.64 86.57 95.54 7.64 13.75 17.33 20.93 24.18 28.45 33.59 10.76 17.61 21.75 26.58 30.08 35.23 41.14 16.89 27.38 30.39 35.89 41.80 48.22 58.73 25.53 44.31 49.90 60.62 69.42 81.13 95.07 41.78 133.33 76.08 87.37 99.96 117.18 133.33 144 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 72. — Comparison of the conductivities of cobalt chloride. — Continued. In mixtures of acetone and water at 0°. In mixtures of acetone and water at 25°. V 0 p. ct. 25 p. ct. 50 p. ct. 75p.ct. lOOp.ct. 0 p. ct. 25p.ct. 50 p. ct. 75 p. ct. lOOp.ct. 100 109.27 64.40 54.42 33.14 10.18 197.41 115.38 96.89 70.57 9.47 200 113.73 67.88 57.79 37.98 10.96 213.40 118.42 106.88 84.48 9.70 400 115.87 69.13 60.29 53.41 11.63 219.63 122.39 112.70 94.84 9.94 800 116.71 72.93 64.54 61.16 12.66 220.04 126.88 117.97 109.76 10.11 1600 117.40 79.34 67.80 68.51 12.79 221.50 134.69 125.54 124.95 10.45 In mixtures of acetone and methyl In mixtures of acetone and methyl V alcohol at 0°. alcohol at 25°. 100 60.89 48.10 35.31 23.85 10.18 76.08 56.28 39.20 24.83 9.47 200 70.45 57.33 42.12 28.72 10.96 87.37 64.87 46.94 28.67 9.70 400 75.64 64.43 49.39 32.74 11.63 99.96 73.04 55.47 31.85 9.94 800 86.57 74.60 56.97 42.05 12.66 117.18 84.46 66.85 38.69 10.11 1600 95.54 91.57 67.55 50.94 12.79 133.33 100.09 74.11 47.20 10.45 In mixtures of acetone and ethyl In mixtures of acetone and ethyl V alcohol at 0°. alcohol at 25°. 100 12.79 14.40 12.78 13.46 10.18 17.33 15.43 12.41 11.95 9.47 200 15.43 18.05 15.20 14.41 10.96 20.93 19.47 13.91 12.49 9.70 400 17.66 21.19 17.95 14.69 11.63 24.18 23.67 15.78 12.69 9.94 800 20.70 26.10 22.47 15.74 12.66 28.45 30.20 18.76 13.48 10.11 1600 23.99 31.80 28.67 17.28 12.79 33.59 36.41 23.47 14.19 10.45 TABLE 73 — Comparison of the temperature coefficients of conductivity of cobalt chloride from 0° to 25°. In mixtures of — V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. f 10 0.0324 0.0440 0.0400 0.0324 0.00995 50 .0326 .0447 .0425 .0258 .01040 Methyl alcohol and 100 - 200 .0322 .0350 .0462 .0485 .0427 .0427 .0296 .0304 .00998 .00961 water 400 .0358 .0463 .0435 .0283 .01280 800 .0354 .0447 .0429 .0263 .01410 1600 .0355 .0472 .0411 .0257 .01580 f 10 0.0324 0.0565 0.0585 0.0399 0.0104 50 .0326 .0615 .0625 .0428 .0119 Ethyl alcohol and 100 < 200 .0322 .0350 .0611 .0660 .0622 .0664 .0438 .0461 .0142 .0142 water 400 .0358 .0638 .0634 .0448 .0148 800 .0354 .0568 .0648 .0459 .0150 1600 .0355 .0579 .0583 .0463 .0160 10 0.0104 0.00880 0.0124 0.00541 0.00995 50 .0119 .00805 .0110 .01020 .01040 Methyl alcohol and 100 • 200 .0142 .0142 .01023 .01255 .0100 .0101 .00893 .00994 .00998 .00961 ethyl alcohol . . 400 .0148 .01070 .0118 .01070 .01280 800 .0150 .00972 .0092 .01150 .01410 1600 .0160 .01106 .0108 .01490 .01580 COBALT CHLORIDE. 145 TABLE 73. — Comparison of the temperature coefficients of conductivity of cobalt chloride from 0° to 25°. — Continued. In mixtures of V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. r 10 50 100 0.0322 0.0316 o.6si2 0.0452 -0.00279 Acetone and water . . 200 .0350 .0298 .0340 .0490 - .00460 400 .0358 .0308 .0348 .0310 - .00581 800 .0354 .0296 .0331 .0318 - .00806 ^1600 .0355 .0279 .0341 .0329 - .00732 100 0.00998 0.00680 0.00441 + 0.00164 -0.00279 Acetone and methyl alcohol 200 400 800 .00961 .01280 .01410 .00526 .00534 .00529 .00458 .00492 .00694 - .00007 - .00108 .00319 - .00460 - .00581 - .00806 L1600 .01580 .00372 .00388 - .00294 .00732 100 0.0142 0.00286 -0.00116 -0.00449 -0.00279 Acetone and ethyl alcohol 200 400 800 .0142 .0148 .0150 .00314 .00562 .00628 - .00339 - .00483 - .00660 - .00533 - .00544 - .00574 - .00460 - .00581 - .00806 1600 .0160 .00580 - .00726 - .00715 - .00732 Tables 71 and 72 (figs. 62 and 63), for cobalt chloride in mixtures of methyl alcohol and water, show a minimum in conductivity at both tempera- 10054 Percentage of Methyl Alcohol FIG. 62. — CONDUCTIVITY OF COBALT CHLORIDE IN MIX- TURES OF METHYL ALCOHOL AND WATER AT 0°. 146 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. tures. It should be noticed that the minimum, which occurs in the 75 per cent mixture at both temperatures, is more marked at 25° than at 0°. In the case of lithium bromide, in mixtures of methyl alcohol and water, the minimum appeared more pronounced at the lower temperature. Between the 75 per cent mixture and pure ethyl alcohol the curves diverge rapidly from each other. This seems to indicate that the dissociation is greatly in- creased by the addition of small amounts of water. The temperature coefficients increase with the dilution, especially in the aqueous and pure methyl alcohol solutions. The temperature coefficients are greater in the mixtures than in the pure solvents, reaching a maximum in the 25 per cent mixture. 240^ 220 200 180- x 43 "i> 1CO- o -§ 140 o ^120 jg loo "o 80 CO 40 20 250 500 750 Percentage of Methyl Alcohol FIG. 03. — CONDUCTIVITY OF COBALT CHLORIDE IN MIXTURES OF METHYL ALCOHOL AND WATER AT 25°. Tables 71 and 72 (figs. 64 and 65), for cobalt chloride in mixtures of ethyl alcohol and water, show a point of inflection, At 0° the inflection of the curves exists at all dilutions, while at 25° the inflection is marked only at the high dilutions. Jones and Bingham obtained curves showing points of inflection while working with calcium nitrate, in mixtures of acetone and water. COBALT CHLORIDE. 147 o a) 240 220 200 180 160 140 120 100 80 CO 40 20- 130 110 100 $ Percentage of Ethyl Alcohol FIG. 64. — CONDUCTIVITY OFCOBALT CHLORIDE IN MIXTURES OF ETHYL ALCOHOL AND WATER AT 0°. 25# 50^ 75$$ Percentage of Ethyl Alcohol FIG. 65. — CONDUCTIVITY OFCOBALT CHLORIDE IN MIXTURES OF ETHYL ALCOHOL AND WATER AT 25.° 148 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. In the pure ethyl alcohol the temperature coefficients increase with the dilution. They are largest, for the most part, in the 50 per cent mixture. Tables 71 and 72 (figs. 66 and 67) make it clear that in a mixture of methyl and ethyl alcohols the conductivity of cobalt chloride exhibits no minimum value. There is a sagging of the curves, which demonstrates that the values obtained are less than what we should expect from the law of averages. Tables 71 and 72 (figs. 68 and 69), for cobalt chloride in mixtures of acetone and water, show a point of inflection at low temperatures and high dilution. Attention should be called to the fact that the conductivity values in pure acetone at 25° are less than the corresponding values at 0°, thus giving negative temperature coefficients. The values in pure acetone for /u,v are small at both temperatures and at all dilutions. Percentage of Methyl Alcohol FIG. 66. — CONDUCTIVITY OF COBALT CHLORIDE IN MIXTURES OF METHYL ALCOHOL AND ETHYL ALCOHOL AT 0°. From a study of tables 71 and 72 (figs. 70 and 71) we see that cobalt chloride, in mixtures of acetone and methyl alcohol, gives neither a minimum nor a maximum in conductivity. The values at most of the dilutions are what we should expect from the law of averages. It should be recalled that lithium bromide, in mixtures of acetone and methyl alcohol, gives a maximum in conductivity at both temperatures. VISCOSITY MEASUREMENTS. 149 By studying the temperature coefficients we see that we have to deal with a peculiar phenomenon. In the pure acetone, as already stated, we have negative temperature coefficients. In the 75 per cent mixture of acetone and methyl alcohol, beginning with v = 400, we again have negative temperature coefficients. For v = 100 we have the temperature coefficients positive, while at v = 200 we have practically no temperature coefficient of conductivity. Tables 71 and 72 (figs. 72 and 73), for cobalt chloride in mixtures of acetone and ethyl alcohol, give a maximum in conductivity in the 25 per cent mix- ture, especially at high dilutions. In the mixtures of acetone and ethyl alcohol we have negative temperature coefficients, not only in the pure acetone and 75 per cent mixture, but also in the 50 per cent mixture. In the case of acetone and methyl alcohol, the tem- perature coefficients were negative only in the pure acetone and 75 per cent mixture. With increase of dilution we have an increase in value of the temper- ature coefficients, not only in the pure solvents, but also in the mixtures. 1002 Percentage of Methyl Alcohol FIG. 67. — CONDUCTIVITY OF COBALT CHLORIDE IN MIXTURES OF METHYL ALCOHOL AND ETHYL ALCOHOL AT 25°. 150 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. 120 110 Percentage of Acetone FIG. 68. — CONDUCTIVITY OF COBALT CHLORIDE IN MIXTURES OF ACETONE AND WATER AT 0°. 230. 220^ 210 200; 190-^ 180- 170- 25* 50* 75* 100* Percentage of Acetone FIG. 69. — CONDUCTIVITY OF COBALT CHLORIDE IN MIXTURES OF ACETONE AND WATER AT 25°. COBALT CHLORIDE. 151 VISCOSITY MEASUREMENTS. dx It can be shown that the rate — , at which the angular distortion of a portion U(/ of fluid changes, is proportional to the shearing stress upon the portion of the ii i* fluid. This ratio, shearing stress (£) divided by --, for a given fluid, is called at its coefficient of viscosity. It may be written thus, dt in which rj is the coefficient of viscosity. Percentage of Acetone FIG. 70. — CONDUCTIVITY OF COBALT CHLORIDE IN MIXTURES OF ACETONE AND METHYL ALCOHOL AT 0°. The viscosity of fluids is often determined by the methods of Poisseuille and of Hagenbach. The viscosity is calculated by the formula * /* = • 8vl where P is the actual pressure, diminished by that pressure which would be necessary to give to the fluid, while flowing through capillary tubes, the 1 Hagenbach : Pogg. Ann., 19, 385 (1860). 152 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. kinetic energy possessed by it; t is the time of flow through the capillary tube of radius r and length I; v is the volume of fluid flowing in time t. The viscosity of fluids is, however, most commonly found by the method recommended by Ostwald.1 We find by this method the time of flow of a fixed volume through a capillary tube, under the pressure due to the differ- ence in level of the free surfaces of the fluid. The viscosity is found from the ratio of the time of flow for the fluid in question to the time of flow for an equal volume of water, multiplied by the specific gravity of the solution. This is expressed by the formula „ iot in which r;0 is the coefficient of viscosity for water, S0 its specific gravity, and 10- 75 ;J 100-6 Percentage of Acetone FIG. 71. — CONDUCTIVITY OF COBALT CHLORIDE IN MIX- TURES OF ACETONE AND METHYL ALCOHOL AT 25°. t0 its time of flow, at a given temperature. The specific gravity and time of flow of the liquid in question are given by S and t, respectively. 1 Ostwald-Luther : Physiko-Chemische Messungen, Zweite Aufl., p. 259. COBALT CHLORIDE. 153 In tables 74 to 76, containing viscosity data, the values for pure water, at 0° and 25°, taken from the work of Thorpe and Rodger, are used, and all the other values are referred to them ; t\ is the viscosity, while $, which rep- resents fluidity, is obtained from the expression The density of the liquid in question at 0° and 25° was compared with the density of water at 0° and 25°, respectively. The density of water at 0° and 25° was taken from data given in Landolt and Bernstein's tables. The values of 77 and for the solvents methyl alcohol, ethyl alcohol, acetone, and mixtures of these, are taken from the work of Jones and Bingham. 40- 25?£ 50f 7 Percentage of Acetone FIG. 72. — CONDUCTIVITY OF COBALT CHLORIDE IN MIXTURES OF ACETONE AND ETHYL ALCOHOL AT 0°. .Percentage of Acetone FIG. 73. — CONDUCTIVITY OF COBALT CHLORIDE IN MIXTURES OF ACETONE AND ETHYL ALCOHOL AT 25°. 154 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 74. — Fluidities. At 0 and 25°. V juO° 00° M25° 025° Tempera- ture coef- ficient. Water Solvent 0 01778 56 24 0 OOSQ1 112 3 0 0398 Lithium bromide in water . LITHIUM BROMIDE IN Mixture of 25 p. ct. methyl f 10 1600 (. Solvent f 10 1600 .01874 .01827 .01778 .034216 53.36 54.73 56.24 29.22 .009077 .008963 .008910 .01440 .01420 110.17 111.57 112.30 69.43 70 42 .0425 .0415 .0398 .0550 alcohol and water . . . Mixture of 50 p. ct. methyl I. Solvent f 10 1600 .03335 .03703 29.98 27.00 .01409 .01680 .01639 70.94 59.53 61.01 .0546 .0482 alcohol and water . TVTi"5rtiiT*p r\f 7^ T\ pf TYiptVi vl I Solvent f 10 .03642 27.46 .01611 .01345 62.04 74.32 .0503 1600 .01298 77.00 Methyl alcohol TVTi vturp c\i *}^ T\ pf pfVi vl ( Solvent f 10 1600 ( Solvent f 10 .02576 .008994 .008346 .008185 38.32 111.18 119.82 122.20 .01283 .006124 .005635 .005659 .01832 77.92 163.28 177.46 176.70 54.57 .0402 .0187 .0192 .0178 1600 .01818 55.00 Mixture of 50 p. ct. ethyl (. Solvent f 10 1600 .05264 .06922 18.99 14.45 .01810 .02453 .02407 55.22 40.77 41.53 .0763 .0728 IVTivtnrp of 7^ T\ r*i~ ptVi\rl I Solvent f 10 .06720 14.88 .02405 .02204 41.56 45.36 .0717 1600 .02139 46.75 Ethyl alcohol I Solvent f 10 .05167 .02441 f>O TOO 19.35 40.96 A n: A Q .02118 .01366 ni oo/i 47.21 73.18 Q1 RO .0575 .0314 f»Q1 C Solvent f 10 .u^iyy .01856 4o.4o 53.88 .U1ZZ4 .01106 .01366 oi-oy 90.35 73.19 .UO-lo .0271 Mixture of 25 p. ct. acetone HTiH wflfpr 1600 • • • . .01356 73.72 .... Mixture of 50 p. ct. acetone and water ( Solvent f 10 1600 .0293 .03102 34.12 32.23 .01276 .01428 .01369 78.37 70.00 73.00 .0518 .0468 ( Solvent f 10 .03027 33.03 .01330 .009562 74.96 104.58 .0508 Mixture of 75 p. ct. acetone 1600 .009263 107.95 Acetone ( Solvent ( 10 1600 .017 .004302 004117 58.8 232.45 242 85 .008904 .003484 003339 112.30 286.96 299 42 .0364 .0093 0093 ( Solvent f 10 .004097 244.10 .003237 .005455 308.90 183.30 .0106 Mixture of 25 p. ct. acetone 1600 .005068 197.29 and methyl alcohol . . . Mixture of 50 p. ct. acetone and methyl alcohol . . . ( Solvent f 10 1600 ( Solvent f 10 .006498 .006265 .005774 .005336 153.90 159.6 (?) 173.17 187.40 .004615 .004527 .004234 .003891 .003775 216.70 220.87 236.18 257.10 264.88 .0163 .0153 .0145 .0148 Mixture of 75 p. ct. acetone \ 1600 .003581 279.25 and methyl alcohol . . . ( Solvent f 10 .004501 222.20 .003446 .008218 290.10 121.69 .0122 Mixture of 25 p. ct. acetone 1600 .007731 129.34 and ethyl alcohol . . . Mixture of 50 p. ct. acetone ( Solvent f 10 1600 .01041 .007353 96.08 136.0 .006714 .005554 .005165 148.90 180.05 193.59 .022 .0129 and ethyl alcohol . . . ( Solvent f 10 .006801 147.0 .004874 .004237 205.20 235.99 .0148 Mixture of 75 p. ct. acetone 1600 .003874 258.10 and ethyl alcohol . . . ( Solvent .00499 200.4 .003776 264.80 .01296 LITHIUM BROMIDE FLUIDITIES. 155 TABLE 75. — Comparison of fluidities of lithium bromide at 0° and 25°. AtO \ Mixtures of — V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. Methyl alcohol and water . . ( 10 1600 53.36 54.73 29.22 27.00 .... 111.18 119.82 ( Solvent 56.24 29.98 27.46 38.82 122.20 EthyL alcohol and water . . ( 10 1600 53.36 54 73 .... 14.45 .... 40.96 45 48 ( Solvent 56.24 18.99 14.88 19.35 53.88 Acetone and water .... f 10 1600 53.36 54.73 .... 32.23 .... 232.45 242.85 ( Solvent 56.24 34.12 33.03 58.80 244.10 Acetone and methyl alcohol . f 10 1600 111.18 119.82 .... 159.6(?) 173.17 .... 232.45 242.85 I Solvent 122.20 153.90 187.40 222.20 244.10 Acetone and ethyl alcohol . . f 10 1600 40.96 45.48 .... 136.00 .... 232.45 242.85 ( Solvent 53.88 96.08 147.00 200.40 244.10 Mixtures of — At 2 5°. Methyl alcohol and water . . f 10 1600 110.17 111.57 69.43 70.42 59.53 61.01 74.32 77.00 163.28 177.46 (. Solvent 112.30 70.94 62.04 77.92 176.70 Ethyl alcohol and water . . ( 10 1600 110.17 111.57 54.57 55.00 40.77 41.53 45.36 46.75 73.18 81.69 I Solvent 112.30 55.22 41.56 47.21 90.35 Acetone and water .... f 10 1600 110.17 111.57 73.19 73.72 70.00 73.00 104.58 107.95 286.96 299.42 I Solvent 112.30 78.37 74.96 112.30 308.90 Acetone and methyl alcohol . f 10 1600 163.28 177.46 183.30 197.29 220.87 236.18 264.88 279.25 286.96 299.42 I Solvent 176.70 216.70 257.00 290.10 308.90 Acetone and ethyl alcohol . . f 10 1600 73.18 81.69 121.69 129.34 180.05 193.59 235.99 258.10 286.96 299.42 I Solvent 90.35 148.90 205.20 264.80 308.90 TABLE 76. — Comparison of the temperature coefficients of lithium bromide from 0° to 25' Mixtures of — V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. ( 10 0.0425 0.0728 0.0314 Ethyl alcohol and water . . 1600 .0415 .... .... .0318 ( Solvent .0398 0.0763 .0717 .00575 .0271 f 10 .0425 .0550 .0482 .... .0187 Methyl alcohol and water . 1600 .0415 .0192 I Solvent .0398 .0546 .0503 .0402 .0178 f 10 .0425 .... .0468 * • • • .0093 Acetone and water .... \ 1600 .0415 .... .... .0093 I Solvent .0398 .0518 .0508 .0364 .0106 f 10 .0187 .0153 .0093 Acetone and methyl alcohol . 1600 .0192 • • • • .0145 .0093 I Solvent .0178 .0163 .0148 .0122 .0106 f 10 .0314 .0129 .0093 Acetone and ethyl alcohol . . 1600 .0318 • • • • • • • • .0093 v Solvent .0271 .022 .0148 .01296 .0106 156 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. 180 160 140 120 I. Methyl Alcohol and Water at 0°. II. Ethyl Alcohol and Water at 0°. III. Methyl Alcohol and Water at 25°. IV. Ethyl Alcohol and Water at 25°. V. 0.1 N LiBr in CH3OH and H2O at 25°. VI. 0.1 N LiBr in C2HBOH and H2O at 25°. 100 £ Percentage Composition FIG. 74. 25* Percentage of Acetone FIG. 75. — FLUIDITY OF SOLVENT MIXTURES AT 0°. FLUIDITY OF SOLVENT MIXTURES. 157 Tables 74 and 75 (fig. 74) show that there is a minimum of fluidity not only in the case of the mixtures of the pure solvents — methyl alcohol and water and ethyl alcohol and water — at both temperatures, but also in the case of lithium bromide dissolved in the above solvents. The fluidity of a liquid being the reciprocal of its viscosity, the viscosity curve would pass through a maximum in the above cases. A large number of investigations have been carried out on the viscosities of the alcohols and water, notably 350- 300 250 200- 150 100- 50 75$ 100j£ Percentage of Acetone FIG. 76. — FLUIDITY OF SOLVENT MIXTURES AT 25°. by Poisseuille,1 Stephan,2 Pagliani and Battelli,3 Noack,4 and Traube.5 Quite recently some work has been done on the viscosity of mixtures of the alcohols and water by Dunstan,6 Blanchard,7 and Varenne and Godefroy.8 All of these 1 Mem. Inst. Paris, 9, 433 (1896). 2Wied. Ann., 17, 673 (1883). 3 Atti. di R. Ac. delle Sc. d. Torino, 20, 607 (1885). 4 Wied. Ann., 27, 289 (1886). 5Ber. d. chem. Gesell., 19, 871 (1886). • Journ. Chem. Soc., 85, 817 (1904). 7Journ. Amer. Chem. Soc., 26, 1315 (1904). "Compt. rend., 137, 992 (1903); 138, 990 (1904). 158 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. workers have found that, in the case of mixtures of the alcohols and water, the viscosity of the mixture is much greater than would be expected from the law of averages. Jones and Carroll 1 have calculated the various fluidities of mixtures of methyl and ethyl alcohols and water, for the temperatures 0°, 10°, 20°, and 30°, from the results of Pagliani and Battelli 1 and of Traube,1 and have plotted the fluidity curves. Their curves were similar to those which we obtained. 25$ 50$ 75# 100f« Percentage of Acetone FIG. 77. — N/10 LITHIUM BROMIDE IN ACETONE MIXTURES AT 25°. Tables 74 and 75 (figs. 75, 76, and 77) show that there is a minimum of fluidity for the solvents only in the case of acetone and water. This fact was pointed out by Jones and Bingham. Lithium bromide, in a mixture of ace- tone and water, shows a similar minimum. Jones and Bingham also found that in the mixtures of acetone with methyl alcohol somewhat larger values were obtained than would be expected from the fluidities of the pure solvents. This effect is not quite so apparent in the case of acetone and ethyl alcohol. We obtained similar results in the case of lithium bromide in these solvents. It is to be especially noticed that the viscosity curves in all cases, for mixtures of acetone with the alcohols, show a marked sagging. 1 Loc. cit. FLUIDITY AND CONDUCTIVITY. 159 TABLE 77. — Comparison of the temperature coefficients of conductivity and fluidity. Mixtures of — V Dis- solved sub- stance. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. Methyl al- f*r^l~)nl *} r\ rl Fluidity . . Solvent f 10 LiBr 0.0398 .0329 0.0456 .0435 0.0503 .0410 0.0402 .0314 0.0178 .01590 tyUJLUJJ. . 1 i I < ( water. Conductivity 1600 10 LiBr CoCh .0357 .0324 .0437 .0440 .0420 .0400 .0335 .0324 .01810 .00995 1600 CoCh .0355 .0472 .0411 .0257 .01580 Ethyl alco-- li r\ 1 in H : Fluidity Solvent f 10 LiBr .0398 .0329 .0763 .0557 .0717 .0595 .0575 .0437 .0271 .0253 i n M (i i 1 1 i water. Conductivity 1600 10 LiBr CoCh .0357 .0324 .0448 .0565 .0596 .0585 .0465 .0399 .0242 .0104 [ 1600 CoCl3 .0355 .0579 .0583 .0463 .0100 A Ppt ATI P Fluidity . . Solvent .0398 .0518 .0508 .0364 .0106 j».l_,C ULHJ.C f\ n f\ ( 10 LiBr .0329 .0429 .0432 .0315 .00729 dil\JL water. Conductivity 1600 100 LiBr CoCl2 .0357 .0322 .0401 .0316 .0437 .0312 .0372 .0452 .00847 .00279 I 1600 CoC]2 .0355 .0279 .0341 .0329 .00732 A fptnnp Fluidity . . Solvent * • • * .0178 .0163 .0148 .0122 .0106 iHjC Lw JJ.C JITlH \ 10 LiBr .0159 .0143 .00939 .0112 .00729 cuj.u methyl o l/>r»Tir^1 Conductivity 1600 100 LiBr CoCh .0181 .00998 .0153 .0068 .01340 .00441 .0104 .00164 .00847 .00279 Ou\t\JlL\jlm 1600 CoCh .0159 .00372 .00388 .00294 .00732 r Fluidity . . Solvent .0271 .022 .0148 .0129 .0106 Acetone f 10 LiBr .0253 .0168 .0135 .0085 .00729 and ethyl alcohol. Conductivity 1600 100 LiBr CoCl2 .0242 .0142 .0206 .00286 .0159 .00116 .0103 .00449 .00847 .00279 1600 CoCh .0160 .00580 .00726 .00715 .00732 Table 77 shows that the temperature coefficients of fluidity and conductiv- ity, in the case of lithium bromide, vary in the same manner, although the latter are, for the most part, smaller than the former. Abnormal results were obtained with cobalt chloride in mixtures of acetone with the alcohols, in so far as the temperature coefficients of conductivity are concerned. DISCUSSION OF RESULTS. FLUIDITY AND CONDUCTIVITY. That there is a parallelism between conductivity and fluidity has been previously pointed out in the introduction to this section; and since the con- ductivity of a solution is dependent, in part, upon its fluidity, we shall first discuss the fluidity curves and then, in connection with them, the conductivity curves. When methyl alcohol, ethyl alcohol, or acetone is mixed with water, there is a contraction in volume and an evolution of heat, and we have a large devia- tion of the fluidity curve from a straight line. In other words, we have a marked viscosity maximum — a fact which frequently manifests itself when water and organic solvents are mixed. In addition to methyl and ethyl alcohols and acetone, propyl and isopropyl alcohols, propionic acid, butyric 160 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. and isobutyric acids, when mixed with water, show viscosity maxima. In every case there is a greater decrease in fluidity than would be expected from the law of averages. When methyl and ethyl alcohols are mixed, they do not exhibit the same phenomena as is shown in the case of mixtures of organic solvents with water. Arrhenius 1 states that there is no observable change when these alcohols are brought together. When acetone is mixed with methyl or ethyl alcohol, the fluidity curve of the mixture is approximately a straight line. The same is true when we have an electrolyte dissolved in mixtures of these solvents. From a consideration of the fluidity curves and conductivity curves for lithium bromide in mixtures of methyl alcohol and water, it is quite evident that the minimum of fluidity corresponds to the minimum of conductivity, both generally occurring in the 50 per cent mixtures of the solvents. The drop in fluidity is more pronounced at the lower temperature, and, similarly, the drop in conductivity. In fact, at the higher temperature the minimum of conductivity occurs in the 75 per cent mixture until v = 100, where there is a shifting of the minimum to the 50 per cent mixture. In the case of mixtures of ethyl alcohol and water, the minimum of fluidity is in the 50 per cent mixture, and more marked at the lower temperature. The conductivity minimum in this case occurs in the 75 per cent mixture, and is very slight indeed at 25°, although the values in each case are much less than we should expect from the law of averages. Thus, we have in both the case of methyl alcohol and water and of ethyl alcohol and water, a tendency for the shifting of the minimum towards the mixture containing the greater per cent of alcohol, whenever we have a rise in temperature. Stephan,2 working with mixtures of ethyl alcohol and water, found that the temperature coefficients of conductivity and of fluidity were very similar. A minimum in his curves was observed, and he proposed the relations — KH wKH k = and k = — -. — ?) Wr) the first holding for the mixtures up to the minimum point, and the second from that point on. K is the same in both formulae, and is the conductivity of the equivalent aqueous solution of the electrolyte; k is the conductivity in the mixture; H and rj are the viscosity coefficients for water and for the mixture, respectively, w and w' are the per cents of water in the mixture and in the aqueous alcoholic mixtures of minimal fluidity, respectively. Stephan concluded that each ion carries with it molecules of the solvent, and that the ionic friction consists in friction between these molecules and the rest of the solvent. Thus we have very early the idea of ionic hydration introduced 1 Ztschr. phys. Chem., 1, 285 (1887). 2 Wied. Ann., 17, 673 (1882). FLUIDITY AND CONDUCTIVITY. 161 This idea of ionic hydration, or ionic spheres, was further extended by Kohlrausch,1 who proposed the hypothesis that — About every ion there moves an atmosphere of the solvent, the dimensions of which are determined by the individual characteristics of the ion. . . . The electrolytic resistance is a fractional one that increases with the dimensions of the atmosphere. The direct action between the ion and the outer portions of the solvent diminishes as the at- mosphere becomes of greater dimensions. For a slow-moving ion there would be only the friction of water against water, and the electrolytic resistance will have the same temperature coefficient as the viscosity of water, providing the atmosphere does not change its dimensions with temperature. However, if the atmosphere becomes smaller with rise in temperature, the temperature gradient of the conductivity might be greater than that of the fluidity. This seems to be true for the slowest-moving, univalent ion, Li. It will be recognized that this is essentially the theory of hydrates proposed by Jones. In the case of the alcohols and water, this minimum of conductivity is entirely accounted for by some investigators on the basis of the formation of hydrates. This view was suggested by Zelinsky and Krapiwin.2 The mini- mum in fluidity is attributed to the formation of molecular aggregations, which are formed by mixing the solvents. In the case of methyl alcohol and acetone, or ethyl alcohol and acetone, since the fluidity curve is a straight line, we conclude that the molecular aggregations of these solvents are not changed in size when the solvents are mixed. This is what we should expect if the fluidities are additive, a fact which has been shown recently by Bing- ham 3 to be true. From the above, we see that the conductivity minimum is generally accompanied by a fluidity minimum, and that both minima are more marked at the lower temperatures. Also, that an increase in temperature tends to shift the minimum towards the mixture containing a greater per cent of alcohol or acetone, as the case may be. Thus, we believe that a diminution in the fluidity of the solvent, which would bring about a corresponding decrease in ionic mobility, is an important factor in causing the minimum of conductivity. In the case when the conductivity minimum is in the 75 per cent mixture, while the minimum of fluidity is in the 50 per cent mixture, as it is with ethyl alcohol and water, we believe that the explanation is to be found in the fact that the ethyl alcohol and water mixtures have a much greater viscosity than those of methyl alcohol and water. When the minimum shifts with an in- crease in dilution, there may be an increase in dissociation. However, we do not believe that the above explanation accounts entirely for the conductivity minimum. Since conductivity is dependent upon the number and velocity of the ions, there is no doubt that an increase in JProc. Roy. Soc., 71, 338 (1903). 3 Amer. Chem. Journ., 35, 195 (1906). Ztschr. phys. Chem., 21, 35 (1896). 162 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. viscosity retards the velocity of the ions, but we think that the change in the size of the ionic sphere, or the atmosphere which surrounds the ion, should also be taken into account. The movement of the ion depends not only upon its composition, but also upon its attraction for the surrounding solvent, which causes an atmosphere of the solvent to be formed about the ion. Lithium is an extremely slow-moving ion, or, in other words, one with a very large ionic sphere. This atmosphere becomes larger with decreasing temperature. Thus, we have evidence that the change in dimensions of the ionic spheres must be taken into account in dealing with fluidity and, consequently, with con- ductivity from the fluidity data of lithium bromide. The fluidity values are, on the whole, small, due to the large atmosphere surrounding the lithium ion. This atmosphere increases with decrease in temperature, and thus produces the smaller fluidity values at the lower temperature. We should not, however, lose sight of the fact that if viscosity is in any way dependent upon the attractions between the molecules, whenever there is a contraction in mixing two solvents, as there is in the case of the alcohols and water, then the molecules will be brought closer together and the attractions will be increased between the molecules. It is obvious from this that we should get a fluidity value different from that calculated from the law of averages. The statement was made above that fluidities are sometimes additive. That is, the resulting fluidity of a mixture of two solvents is equivalent to the sum of the fluidities of each solvent. Formulated, this would be (1) where <£ is the resultant fluidity, <£, and <£2 the respective fluidities of the components of amounts ki and k2. Jones and Bingham * have pointed out that this expression is similar to the conception which we have in electricity, where the conductance of several conductors, in parallel, is represented by the sum of their separate conduct- ances. The conductance of a pair of conductors, of different material, is, for a unit length, (o-j + 0° M,25° Temp, coef. M»0° M,25° Temp, coef. 10 50 100 200 400 800 1600 30. 42. 47. 52. 57. 59. 63. 28 17 95 48 06 87 67 48.20 69.75 80.59 89.97 99.83 106.13 114.25 0.0237 .0262 .0272 .0286 .0300 .0309 .0318 15.04 27.28 33.58 40.46 46.36 53.51 60.25 17.98 32.98 42.37 50.06 57.78 65.78 72.64 0.00984 .00936 .01047 .00949 .00985 .00917 .00823 34.45 40.50 43.27 45.12 46.31 47.45 49.06 81.17 99.17 107.01 112.44 116.73 119.85 126.90 0.0543 .0578 .0589 .0597 .0608 .0610 .0635 V In 50 p. ct. ethyl alcohol and water. In 75 p. ct. ethyl alcohol and water. In ethyl alcohol. 10 50 100 200 400 800 1600 19. 24. 26. 28. 29. 30. 32. 66 41 60 28 79 87 57 46.98 61.24 67.56 73.05 77.84 81.43 88.16 0.0556 .0604 .0616 .0633 .0645 .0655 .0683 13.24 18.64 21.20 24.17 26.33 28.25 32.36 25.17 36.83 42.46 48.81 53.92 59.05 67.31 0.0360 .0390 .0401 .0408 .0419 .0436 .0432 3.53 6.42 8.48 10.20 12.01 13.95 15.91 4 8 12 13 17 19 25 .88 .30 .07 .59 .18 .92 .24 0.0153 .0117 .0169 .0133 .0172 .0171 .0235 V In 25 p. ct. methyl alcohol ethyl alcohol. and In 50 p. ct. methyl alcohol and ethyl alcohol. M,,0° Ak>25° Temperature coefficient. /M)° ^,25° Temperature coefficient. 10 50 100 200 400 800 1600 5.12 9.63 12.35 15.16 18.17 20.76 23.97 6.39 11.83 15.48 17.47 24.18 28.58 34.90 0.00992 .00914 .01014 .00610 .01320 .01510 .01830 7.81 14.68 18.43 22.39 26.54 29.77 34.83 0.57 18.09 23.21 28.94 35.54 40.70 48.90 0.00901 .00929 .01037 .01170 .01356 .01469 .01616 V In 75 p. ct. methyl alcohol and ethyl alcohol. In ethyl alcohol and water. M»0° Ak,25° Temp, coef. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 10 50 100 200 400 800 1600 11.44 21.10 26.08 31.28 37.33 40.95 48.76 13.88 25.93 32.61 39.98 48.80 53.75 64.12 0.00853 .00916 .01002 .01113 .01229 .01250 .01260 0.0329 .0346 .0372 .0347 .0350 .0347 .0371 0.0543 .0578 .0589 .0597 .0608 .0610 .0635 0.0556 .0604 .0616 .0633 .0645 .0655 .0683 0.0360 .0390 .0401 .0408 .0419 .0436 .0432 0.0153 .0117 .0169 .0133 .0172 .0171 .0235 COPPER CHLORIDE. 177 In studying the temperature coefficients of conductivity, it is to be noted that in every case, with the exception of pure methyl alcohol, there is an increase in the temperature coefficient with increase in dilution. This in- crease, although not perfectly regular, is, however, decidedly marked when the difference in the value for the most concentrated and for the most dilute solution is considered. It is also to be observed that the temperature coeffi- cients in the 25 per cent mixture are decidedly larger than the corresponding values in either of the other mixtures used. TABLE 79. — Comparison of the conductivities of copper chloride. In mixtures of methyl alcohol and water In mixtures of methyl alcohol and water V atO°. at 25°. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. lOOp.ct. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. lOOp.ct. 10 89.28 48.12 35.35 30.28 15.04 162.60 98.73 68.27 48.20 17.98 50 103.94 56.57 43.92 42.07 27.28 193.87 119.03 88.14 69.75 32.98 100 106.14 59.96 46.99 47.95 33.58 204.86 128.00 95.63 80.59 42.37 200 115.03 62.14 50.14 52.48 40.46 214.84 133.80 102.24 89.97 50.06 400 117.47 65.36 53.22 57.06 46.36 220.34 141.60 110.48 99.82 57.78 800 123.73 67.45 54.91 59.87 53.51 231.03 146.90 114.56 106.13 65.78 1600 129.52 67.37 59.01 63.61 60.25 249.70 147.30 123.50 114.25 72.64 In mixtures of ethyl alcohol and water In mixtures of ethvl alcohol and water at 0°. at 25°. V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. lOOp.ct. 0 p. ct. 25 p.ct. 50 p. ct. 75 p. ct. lOOp.ct. 10 89.28 34.45 19.66 13.24 3.53 162.60 81.17 46.98 25.17 4.88 50 103.98 40.50 24.41 18.64 6.42 193.87 99.17 61.24 36.83 8.30 100 106.14 43.27 26.60 21.20 8.48 204.86 107.01 67.56 42.46 12.07 200 115.03 45.12 28.28 24.17 10.20 214.84 112.44 73.05 48.81 13.59 400 117.47 46.31 29.79 26.33 12.01 220.34 116.73 77.84 53.92 17.18 800 123.73 47.45 30.87 28.25 13.95 231.03 119.85 81.43 59.05 19.92 1600 129.52 49.06 32.57 32.36 15.91 249.70 126.90 88.16 67.31 25.24 In mixtures of methyl alcohol and ethyl In mixtures of methyl alcohol and ethyl alcohol at 0°. alcohol at 25°. V Ethyl alcohol. 25 p. ct. 50 p. ct. 75 p. et. Methyl alcohol. Ethyl alcohol. 25 p. ct. 50 p. ct. 75 p. ct. Methyl alcohol. 10 3.53 51.2 7.81 11.44 15.04 4.88 6.39 9.57 13.88 17.98 50 6.42 96.3 14.68 21.10 27.28 8.30 11.83 18.09 25.93 32.98 100 8.48 123.5 18.43 26.08 33.58 12.07 15.48 23.21 32.61 42.37 200 10.20 1.51.6 22.39 31.28 40.46 13.59 17.47 28.94 39.98 50.06 400 12.01 181.7 26.54 37.33 46.36 17.18 24.18 35.54 48.80 57.78 800 13.95 207.6 29.77 40.95 53.51 19.92 28.58 40.70 53.75 56.78 1600 15.91 239.7 34.83 48.76 60.25 25.24 34.94 48.90 64.12 72.64 Tables 78 and 79 (figs. 82 and 83) show that copper chloride, in mixtures of ethyl alcohol and water, also gives a dropping below the rule of averages for the curves in the 25 per cent and 50 per cent mixtures. Here also, the values between the 75 per cent and 100 per cent mixtures show a general tendency towards the rule of averages. It is to be noted also that there is a bunching of the curves in the 25 per cent and the 51 per cent mixtures, or, 178 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. in other words, that there is a comparatively small increase in conductivity with increase in dilution in these mixtures. It is further to be observed that the conductivity values in ethyl alcohol are very much smaller than those in water, although the general increase in conductivity with increase in dilution is about the same in both cases. The temperature coefficients of conductivity in every case increase with 130 120 110- 100- • 90- 80- ; TO- 60- o 0> 50- 40- 30- 20- 10- Percentage of Methyl Alcohol FIG. 80. — CONDUCTIVITY OF COPPER CHLORIDE IN MIXTURES OF METHYL ALCOHOL AND WATER AT 0°. TABLE 80. — Comparison of the temperature coefficients of conductivity of copper chloride V In mixtures of methyl alcohol and water from 0° to 25°. In mixtures of methyl alcohol and ethyl alcohol from 0° to 25°. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. Ethyl alcohol. 25 p. ct. 50 p. ct. 75 p. ct. Methyl alcohol. 10 50 100 200 400 800 1600 0.0329 .0346 .0372 .0347 .0350 .0347 .0371 0.0421 .0442 .0454 .0461 .0467 .0471 .0475 0.0372 .0403 .0414 .0416 .0431 .0435 .0437 0.0237 .0262 .0272 .0286 .0300 .0309 .0318 0.00984 .00836 .01047 .00949 .00985 .00917 .00823 0.0153 .0117 .0169 .0133 .0172 .0171 .0235 0.00992 .00914 .01014 .00610 .01320 .01510 .01830 0.00901 .00929 .01037 .01170 .01356 .01469 .01616 0.00853 .00916 .10002 .01113 .01229 .01250 .01260 0.00984 .00836 .01047 .00949 .00985 .00917 .00823 COPPER CHLORIDE. 179 increase in dilution, not with any great regularity, but the values for the most dilute solutions are decidedly greater than those for the most concen- trated. Here again, we note that temperature coefficients are larger in the mixtures than they are in either of the pure solvents, and that these values are on the whole largest in the 50 per cent mixture instead of in the 25 per cent, as was the case with methyl alcohol and water. 25$ 50 $ 75$ Percentage of Methyl Alcohol FIG. 81. — CONDUCTIVITY OF COPPER CHLORIDE IN MIXTURES OF METHYL ALCOHOL AND WATER AT 25°. Tables 78 and 79 (figs. 84 and 85) show that copper chloride, in mixtures of methyl alcohol and ethyl alcohol, gives no minimum in conductivity, but it is to be noted that there is a decided dropping of the values in the 25 per cent and 50 per cent mixtures, below the values calculated from the rule of averages, and that this is less pronounced in the more concentrated solutions than in the dilute. A slight exception is found in the case of the N/1600 solution at 25°, where a slight increase in the value above the average value is to be observed in the 75 per cent mixture. The remainder of the curve shows the same decrease in values as is shown by the other curves. Here again it is to be noted that the conductivity values in ethyl alcohol are much smaller than those in methyl alcohol, and increase much less rapidly with the dilution. 180 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. Percentage of Ethyl Alcohol FIG. 82. — CONDUCTIVITY OF COPPER CHLORIDE IN MIXTURES OF ETHYL ALCOHOL AND WATER AT 0°. Percentage of Ethyl Alcohol FIG. 83. — CONDUCTIVITY OF COPPER CHLORIDE IN MIXTURES OF ETHYL ALCOHOL AND WATER AT 25°. POTASSIUM SULPHOCYANATE. 181 80 70-1 £60 M-4 I 50^ a 5 40 e "3 30 o i— i g 20- 10 25^ 50 <£ 75# 1005; Percentage of Methyl Alcohol FIG. 84. — CONDUCTIVITY OF COPPER CHLORIDE IN MIXTURES OF METHYL ALCOHOL AND ETHYL ALCOHOL AT 0°. 2554 50-4 75 # 100 $ Percentage of Methyl Alcohol FIG. 85. — CONDUCTIVITY OF COPPER CHLORIDE IN MIXTURES OF METHYL ALCOHOL AND ETHYL ALCOHOL AT 25°. POTASSIUM STJLPHOCYANATE. The potassium sulphocyanate used in this work was purified in the follow- ing manner : The purest salt obtainable was twice recrystallized from redis- tilled, 95 per cent ethyl alcohol, and then once recrystallized from absolute alcohol. The crystals were drained by suction, and dried at 100°. In every case the salt was crystallized in as finely divided condition as possible. The salt was preserved in glass-stoppered bottles, in a sulphuric acid desiccator. 182 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 81. — Conductivity of potassium sulphocyanate at 0° and 25°. In a mixture of 25 p. ct. In a mixture of 50 p. ct. In water. methyl alcohol and water. methyl alcohol and water. V Tempera- Pempera- Tempera- /U)° ^25° ture coef- JM>° Ah,25° ;ure coef- M,0° M,25° ture coef- ficient. ficient. ficient. 10 62.98 112.70 0.0316 37.04 75.00 0.0410 29.59 59.19 0.0400 50 67.42 121.44 .0321 39.09 80.35 .0422 31.35 63.59 .0411 100 68.78 125.06 .0327 39.45 81.14 .0423 31.78 64.62 .0413 200 69.73 126.47 .0326 40.31 82.86 .0422 32.53 66.47 .0417 400 71.62 129.99 .0326 41.19 84.94 .0425 32.63 66.93 .0421 800 71.42 128.96 .0322 41.97 85.29 .0413 32.85 68.17 .0430 1600 72.64 121.61 .0325 41.94 85.84 .0419 33.72 69.87 .0429 V In a mixture of 75 p. ct. methyl alcohol and water. In methyl alcohol. In a mixture of 25 p. ct. ethyl alcohol and water. 10 32.06 56.46 0.0304 46.83 64.81 0.0154 27.28 62.60 0.0518 50 35.21 62.77 .0313 56.94 79.40 .0158 28.14 65.90 .0537 100 36.22 64.86 .0316 60.63 84.97 .0161 28.73 67.88 .0545 200 27.34 67.37 .0322 63.78 89.91 .0164 29.51 69.71 .0545 400 37.96 68.75 .0324 66.01 92.91 .0163 29.57 70.73 .0557 800 37.74 70.10 .0343 68.80 97.19 .0165 29.49 71.70 .0573 1600 37.16 69.88 .0352 71.94 102.41 .0169 28.56 71.43 .0600 V In a mixture of 50 p. ct. ethyl alcohol and water. In a mixture of 75 p. ct. ethyl alcohol and water. In ethyl alcohol. 10 16.93 41.92 0.0590 15.43 32.73 0.0449 14.72 22.97 0.0224 50 17.39 44.37 .0621 16.66 36.31 .0472 18.97 30.13 .0235 100 17.84 45.79 .0627 17.32 37.89 .0475 20.74 33.15 .0239 200 18.02 46.36 .0629 17.76 39.06 .0480 22.45 36.20 .0245 400 18.23 46.94 .0630 17.95 39.70 .0485 23.66 38.82 .0256 800 18.39 47.13 .0625 18.21 40.34 .0486 24.88 41.02 .0260 1600 18.36 47.56 .0636 18.64 41.38 .0490 25.58 42.73 .0268 In a mixture of 25 p. ct. methyl In a mixture of 50 p. ct. methyl In a mixture of 75 p. ct. methyl V alcohol and ethyl alcohol. alcohol and ethyl alcohol. alcohol and ethyl alcohol. 10 21.09 31.85 002.04 28.60 41.69 0.0183 37.06 52.40 0.0166 50 26.47 40.36 02.10 35.56 52.08 .0186 45.39 64.85 .0172 100 28.89 44.27 02.13 38.55 56.98 .0191 48.95 70.42 .0175 200 30.89 47.66 02.17 40.77 60.30 .0192 51.29 74.17 .0178 400 32.32 50.39 02.22 42.74 64.12 .0200 54.09 78.17 .0178 800 33.37 52.20 02.26 43.94 66.10 .0202 54.95 79.75 .0181 1600 34.98 55.74 02.37 45.40 69.09 .0209 57.55 83.61 .0181 In a mixture of 25 p. ct. In a mixture of 50 p. ct. In a mixture of 75 p. ct. V acetone and water. acetone and water. acetone and water. 10 39.47 78.74 0.0398 31.81 63.74 0.0402 34.59 59.07 0.0283 50 40.45 82.36 .0414 33.41 67.59 .0409 38.98 68.02 .0298 100 41.56 84.00 .0409 33.69 68.62 .0415 40.21 70.21 .0298 200 42.82 87.12 .0414 34.70 70.56 .0413 41.93 74.01 .0307 400 43.95 89.18 .0412 35.25 71.68 .0413 43.05 76.00 .0306 800 43.46 87.58 .0406 35.55 72.81 .0419 43.76 77.40 .0308 1600 43.59 87.56 .0404 36.86 76.62 .0432 44.00 78.48 .0312 POTASSIUM SULPHOCYANATE. 183 TABLE 81. — Conductivity of potassium sulphocyanate at 0° and 25°. — Continued. In acetone. In a mixture of 25 p. ct. acetone and methyl alcohol. In a mixture of 50 p. ct. acetone and methyl alcohol. V Tempera- Tempera- Tempera- l^ft M(,25° ture coeffi- Mw^" M^25 ture coeffi- H0Q° /^250 ture coeffi- cient. cient. cient. 10 43.44 48.66 0.00481 52.06 69.97 0.0138 53.92 70.51 0.0123 50 69.63 78.84 .00529 63.84 86.48 .0142 67.56 88.93 .0127 100 73.83 87.65 .00749 68.44 93.20 .0145 73.33 97.31 .0131 200 94.45 109.00 .00616 71.95 97.90 .0144 77.99 103.77 .0132 400 105.95 126.81 .00788 75.85 103.58 .0146 82.09 109.93 .0136 800 118.80 141.81 .00775 77.29 106.35 .0150 85.42 114.54 .0136 1600 126.20 151.92 .00815 79.75 101.11 .0157 88.41 119.41 .0140 In a mixture of 75 p. ct. acetone and methyl In a mixture of 25 p. ct. acetone and ethyl alcohol. alcohol. V 0 0 Temperature ,»25° Temperature M"° /*» 5 coeffi cient. coefficient. 10 57.40 71.33 0.00971 23.97 34.28 0 .0172 50 74.50 93.21 .01005 30.86 44.55 .0177 100 81.96 104.08 .01080 34.13 49.59 .0181 200 86.53 111.05 .01133 36.43 53.55 .0188 400 94.19 120.98 .01138 38.74 57.54 .0194 800 98.00 126.14 .01148 40.55 60.58 .0198 1600 102.90 133.16 .01176 42.14 63.26 .0201 V In a mixture of 50 p. ct. acetone and ethyl alcohol. In a mixture of 75 p. ct. acetone alcohol. and ethyl 10 34.62 45.51 0.0126 44.87 546 .1 0.00868 50 45.06 60.17 .0134 59.92 744 .1 .00967 100 49.92 67.40 .0140 67.52 850 .6 .01039 200 53.52 73.13 .0147 72.90 925 .1 .01076 400 57.35 79.11 .0152 78.97 1015 .6 .01144 800 59.06 82.98 .0162 83.23 1078 .6 .01184 1600 60.19 86.54 .0175 87.62 1147 .3 .01237 TABLE 82. — Comparison of the conductivities of potassium sulphocyanate. In mixtures of methyl alcohol and water at 0°. In mixtures of methyl alcohol and water at 25°. V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. Op. ct. 25 p. ct. 50 p. ct. 75p.ct. 100 p. ct. 10 62.98 37.04 29.59 32.06 46.83 112.70 75.00 59.19 56.46 64.81 50 67.42 39.09 31.35 35.21 56.94 121.44 80.35 63.59 62.77 79.40 100 68.78 39.45 31.78 36.22 60.63 125.06 8L14 64.62 64.86 84.97 200 69.73 40.31 32.53 37.34 63.78 126.47 82.86 66.47 67.37 89.91 400 71.62 41.19 32.63 37.96 66.01 129.99 84.94 66.93 68.75 92.91 800 71.42 41.97 32.85 37.74 68.80 128.96 85.29 68.17 70.10 97.19 1600 72.64 41.94 33.72 37.16 71.94 131.61 85.84 69.87 69.88 102.41 V In mixtures of ethyl alcohol and water at 0°. In mixtures of ethyl alcohol and water at 25°. 10 62.98 27.28 16.93 15.43 14.72 112.70 62.60 41.92 32.73 22.97 50 67.42 28.14 17.39 16.66 18.97 121.44 65.90 44.37 36.31 30.13 100 68.78 28.73 17.84 17.32 20.74 125.06 67.88 45.79 37.89 33.15 200 69.73 29.51 18.02 17.76 22.45 126.47 69.71 46.36 39.06 36.20 400 71.62 29.57 18.23 17.95 23.66 129.99 70.73 46.94 39.70 38.82 800 71.42 29.49 18.39 18.21 24.88 128.96 71.70 47.13 40.34 41.02 1600 72.64 28.56 18.36 18.64 25.58 131.61 71.43 47.56 41.48 42.73 184 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 82. — Comparison of the conductivities of potassium sulphocyanate. — Continued. V In mixtures of methyl alcohol and ethyl alcohol at 0°. In mixtures of methyl alcohol and ethyl alcohol at 25°. Ethyl alcohol. 25 p. ct. 50 p.ct. 75 p. ct. Methyl alcohol. Ethyl alcohol. 25 p. ct. 50 p. ct. 75 p. ct. Methyl alcohol. 10 50 100 200 400 800 1600 14.72 18.97 20.74 22.45 23.66 24.88 25.58 21.09 26.47 28.89 30.89 32.31 33.37 34.98 28.60 35.56 38.55 40.77 42.74 43.94 45.40 37.06 45.39 48.95 51.29 54.09 54.95 57.55 46.83 56.94 60.63 63.78 66.01 68.80 71.94 22.97 30.13 33.15 36.20 38.82 41.02 42.73 31.85 40.36 44.27 47.66 50.39 52.20 55.74 41.69 52.08 56.98 60.30 64.12 66.10 69.09 52.40 64.85 70.42 74.17 78.17 79.75 83.61 64.81 79.40 84.97 80.91 92.91 97.19 102.41 V In mixtures of acetone and water at 0°. In mixtures of acetone and water at 25°. Op.ct. 25 p. ct. 50 p.ct. 75 p.ct. 100 p.ct. 0 p.ct. 25 p.ct. 50 p. ct. 75 p.ct. 100 p.ct. 10 50 100 200 400 800 1600 62.98 67.42 68.78 69.73 71.62 71.42 72.64 39.47 40.45 41.56 42.82 43.95 43.46 43.59 31.81 33.41 33.69 34.70 35.25 35.55 36.86 34.59 38.98 40.21 41.93 43.05 43.76 44.10 43.44 69.63 73.83 94.45 105.95 118.80 126.20 112.70 121.44 125.06 126.47 129.99 128.96 131.61 78.74 82.36 84.00 87.12 89.18 87.58 87.56 63.74 67.59 68.62 70.56 71.68 72.81 76.62 59.07 68.02 70.21 74.10 76.00 77.40 78.48 48.66 78.84 87.65 109.00 126.81 141.81 151.92 V In mixtures of acetone and methyl alcohol atO°. In mixtures of acetone and methyl alcohol at 25°. 10 50 100 200 400 800 1600 48.83 56.94 60.63 63.78 66.01 68.80 71.94 52.06 63.84 68.44 71.95 75.85 77.29 79.75 53.92 67.56 73.33 77.99 82.09 85.42 88.41 57.40 74.50 81.96 86.53 94.19 98.00 102.90 43.44 69.63 73.83 94.45 105.95 118.80 126.20 64.81 79.40 84-97 89.91 92.99 97.19 102.41 69.97 86.48 93.20 97.90 103.58 106.35 111.11 70.51 88.93 97.31 103.77 109.93 114.54 119.41 71.33 93.21 104.08 111.05 120.98 126.14 133.16 48.66 78.84 87.65 109.00 126.81 141.81 151.92 V In mixtures of acetone and ethyl alcohol atO°. In mixtures of acetone and ethyl alcohol at 25°. 10 50 100 200 400 800 1600 14.72 18.97 20.74 22.45 23.66 24.88 25.58 23.97 30.86 34.13 36.43 38.74 40.55 42.14 34.62 45.06 49.92 53.52 57.35 59.06 60.19 44.87 59.92 67.52 72.90 78.97 83.23 87.62 43.44 69.63 73.83 94.45 105,95 118.80 126.20 22.97 30.13 33.15 36.20 38.82 41.02 42.73 34.28 44.55 49.59 53.55 57.54 60.58 63.26 45.51 60.17 67.40 73.13 79.11 82.98 86.54 54.61 74.41 85.06 92.51 101.56 107.86 114.73 48.66 78.84 87.65 109.00 126.81 141.81 151.92 TABLE 83. — Comparison of the temperature coefficients of conductivity of potassium sulphocyanate from 0° to 25°. In mixtures of methyl alcohol and water. In mixtures of ethyl alcohol and water. V Op. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 10 0.0316 0,0410 0.0400 0.0304 0.0154 0.0316 0.0518 0.0590 0.0449 0.0224 50 .0321 ..0422 .0411 .0313 .0158 .0321 .0537 .0621 .0472 .0235 100 .0327 .0423 .0413 .0316 .0161 .0327 .0545 .0627 .0475 .0239 200 .0326 .0422 .0417 .0322 .0164 .0326 .0545 .0629 .0480 .0245 400 .0326 .0425 .0421 .0324 .0163 .0326 .0557 .0630 .0485 .0256 800 .0322 .0413 .0430 .0343 .0165 .0322 .0573 .0625 .0486 .0260 1600 .0325 .0419 .0429 .0352 .0169 .0325 .0600 .0636 .0490 .0268 POTASSIUM SULPHOCYANATE. 185 TABLE 83. — Comparison of the temperature coefficients of conductivity of potassium sulphocyanate from 0° to 25°. — Continued. In mixtures of methyl alcohol and ethyl alcohol. In mixtures of acetone and water. V Ethyl alcohol. 25 p. ct. 50 p. ct. 75 p. ct. Methyl alcohol. Op. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. 10 0.0224 0.0204 0.0183 0.0166 0.0145 0.0316 0.0398 0.0402 0.0283 0.00481 50 .0235 .0210 .0186 .0172 .0158 .0321 .0414 .0409 .0298 .00529 100 .0239 .0213 .0191 .0175 .0161 .0327 .0409 .0415 .0298 .00749 200 .0245 .0217 .0192 .0178 .0164 .0326 .0414 .0413 .0307 .00616 400 .0256 .0222 .0200 .0178 .0163 .0326 .0412 .0413 .0306 .00788 800 .0260 .0226 .0202 .0181 .0165 .0322 .0606 .0419 .0308 .00775 1600 .0268 .0237 .0209 .0181 .0169 .0325 .0404 .0432 .0312 .00815 In mixtures of acetone and methyl alcohol. In mixtures of acetone and ethyl alcohol. V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. Op. ct. 25p.ct. 50 p. ct. 75 p. ct. 100 p. ct. 10 0.0154 0.0138 0.0123 0.00971 0.00481 0.0224 0.0172 0.0126 0.00868 0.00481 50 .0158 .0142 .0127 .01005 .00529 .0235 .0177 .0134 .00967 .00529 100 .0161 .0145 .0131 .01080 .00749 .0239 .0181 .0140 .01039 .00749 200 .0164 .0144 .0132 .01133 .00616 .0245 .0188 .0146 .01076 .00616 400 .0163 .0146 .0136 .01138 .00788 .0256 .0194 .0152 .01144 .00788 800 .0165 .0150 .0136 .01148 .00775 .0260 .0198 .0162 .01184 .00775 1600 .0169 .0157 .0140 .01176 .00815 .0268 .0201 .0175 .01237 .00815 Tables 81 and 82 (figs. 86 and 87) show that potassium sulphocyanate gives a decided minimum in conductivity, for all the dilutions studied at 0°, in the 50 per cent mixtures of methyl alcohol and water; and that at 25° the so X — • 1> ~J o 3 O O o 0> "o a co- 50 40 30- 20- 25# 5Q?» 75$ lOOji Percentage of Methyl Alcohol FIG. 86. — CONDUCTIVITY OF POTASSIUM SULPHOCYANATE IN MIXTURES OF METHYL ALCOHOL AND WATER AT 0°. minimum shifts its position in the first two dilutions to the 75 per cent mixture. It is also to be observed that in the mixtures there is but a very slight increase in molecular conductivity in the more dilute solutions, with increase in dilu- 186 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. 160- Percentage of Methyl Alcohol FIG. 87. — CONDUCTIVITY OF POTASSIUM SULPHOCYANATE IN MIXTURES OF METHYL ALCOHOL AND WATER AT 25°. tion. It is to be noted also that the increase in the conductivity with increase in dilution is greater in methyl alcohol than it is in water, although the values are considerably smaller. The temperature coefficients of conductivity show in all cases a general increase with increase in dilution, and they are also larger in the 25 per cent and 50 per cent mixtures than in any of the other mixtures of the solvents. so Percentage of Ethyl Alcohol TIG. 88. — CONDUCTIVITY OF POTASSIUM SULPHOCYANATE IN MIXTURES OF ETHYL ALCOHOL AND WATER AT 0°. POTASSIUM SULPHOCYANATE. 187 25 # 5'Otf TSji 100 fo Percentage of Ethyl Alcohol FIG. 89. — CONDUCTIVITY OF POTASSIUM SULPHOCYANATE IN MIXTURES OF ETHYL ALCOHOL AND WATER AT 25°. Percentage of Methyl Alcohol FIG. 90. — CONDUCTIVITY OF POTASSIUM SULPHOCYANATE IN MIXTURES OF METHYL ALCOHOL AND ETHYL ALCOHOL AT 0°. 188 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. Tables 81 and 82 (figs. 88 and 89) show that potassium sulphocyanate, in mixtures of ethyl alcohol and water, exhibits a minimum in conductivity at 0°, in the 50 and 75 per cent mixtures, in all cases except the N/10 solu- tion, and at 25° the minimum is shown only in the N/800 and N/1600 solutions, in the 75 per cent mixture. The curves which do not show actual minima, exhibit a decided drop below the average values for the two sol- vents. They also show that there is only a very slight increase in conduc- tivity with increase in dilution in the mixed solvents, and particularly in the 25 per cent and 50 per cent mixtures. It is to be noted, however, that the increase in conductivity with increase in dilution is greater in ethyl alcohol no-J 100- oo- 80- •3 § •d 6 d 3 53 70- 50H 40- 30- 20- 2554 5054 7556 Percentage of Methyl AlcohoL FIG. 91. — CONDUCTIVITY OF POTASSIUM SULPHOCYANATE IN MIXTURES OF METHYL ALCOHOL AND ETHYL ALCOHOL AT 25°. than it is in water for this particular salt. It is also to be noted that the actual value for conductivity in water is much greater than it is in ethyl alcohol. The temperature coefficients of conductivity show an increase with increase in the dilution of the solution, in every case, and these values are greatest in the 50 per cent mixture. Tables 81 and 82 (figs. 90 and 91) show that potassium sulphocyanate, in mixtures of methyl alcohol and ethyl alcohol, does not exhibit a minimum in conductivity, but nevertheless there is an appreciable dropping of the curves, plotted from these values, below the average value. It is to be noted also that the increase in conductivity with increase in dilution is practically the average increase as calculated from the increase in the pure solvents. It is POTASSIUM SULPHOCYANATE. 189 FIG. 92. — CONDUCTIVITY OF POTASSIUM SULPHOCYANATE IN MIXTURES OF ACETONE AND WATER AT 0°. 160- 150 •04JD- 130 120 & £ 110 •*» o •gioo o 0 9 o ,2 80 'o 70 60 50 40 60 25 ;« 50 % 75 f. Percentage of Acetone FIG. 93. — CONDUCTIVITY OF POTASSIUM SULPHOCYANATE IN MIXTURES OF ACETONE AND WATER AT 25°. 100 £ 75* . Percentage of Acetone 190 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. also worthy of note that the conductivity values are greater in methyl alcohol than they are in ethyl alcohol. Here, again, there is a general increase in the value of the temperature coefficients with increase in the dilution of the solu- tion, but in this case the temperature coefficients are greatest in ethyl alcohol and not in the mixtures. Tables 81 and 82 (figs. 92 and 93) show that potassium sulphocyanate, in mixtures of acetone and water at 0°, exhibits, in all the dilutions studied, a decided mimimum in conductivity. It is to be noted, however, that the minimum is somewhat less decided in the most concentrated solution (N/10). At 25° the minimum occurs in all dilutions except N/10, where it entirely disappears. Here also in the more dilute solutions there is only a slight in- crease in conductivity, in the mixed solvents, with increase in dilution. It is also seen that even in the case where the minimum in conductivity has dis- appeared, there is a decided dropping of the curve below the average values. Also the increase hi the conductivity in acetone, with increase in dilution, is very much greater than the corresponding increase in water. So great is this difference that although the values are less in acetone for the more con- centrated solutions than the corresponding values in water, yet they become much greater in acetone than they do in water for the more dilute solutions. The temperature coefficients show a general increase with increase in the dilution of the solutions, and the values of the temperature coefficients them- selves are greatest in the 50 per cent mixture. Tables 81 and 82 (figs. 94 and 95) show that potassium sulphocyanate, in mixtures of acetone and methyl alcohol, exhibits a maximum in conductivity in the 75 per cent mixture, for the first three dilutions (N/10, N/50, N/100) at 0°, and for the first four dilutions (N/10, N/50, N/100, N/200) at 25°. Also, that for the more dilute solutions the curves show a decided drop below the average values in the mixtures. It is also to be noted that the in- crease in conductivity with increase in dilution is very much greater in acetone than it is in methyl alcohol, and, further, that in all the mixtures the increase in conductivity with increase in dilution is practically what would be calcu- lated from the law of averages. Here, again, we note a general increase in the temperature coefficients with increase in dilution, but in this case as in that of potassium sulphocyanate in methyl alcohol and ethyl alcohol, the greatest values do not occur in any of the mixtures, but in one of the pure solvents. In the case of acetone and methyl alcohol the greatest values are in pure methyl alcohol. Tables 81 and 82 (figs. 96 and 97) show that potassium sulphocyanate, in mixtures of acetone and ethyl alcohol, exhibits a maximum in conductivity in the N/10 solution, in the 75 per cent mixture. However, there is a marked tendency towards a maximum and a very marked increase above the average values in N/50, N/100,and N/200 solutions. The remaining dilutions, while POTASSIUM SULPHOCYANATE. 191 they do not exhibit a true minimum, do, however, show a dropping below the average values in the mixtures. It is to be observed that the values for conductivity in acetone are much larger than the corresponding values for ethyl alcohol, and also, that the increase in conductivity with increase in dilution is very much larger in acetone than it is in ethyl alcohol. The increase in conductivity with dilution, in the mixtures, is, however, practically what would be calculated from the values in the pure solvents. 25 # 100 $ Percentage of Acetone FIG. 94. — CONDUCTIVITY OF POTASSIUM SULPHOCYANATE IN MIXTURES OF ACETONE AND METHYL ALCOHOL AT 0°. Here, again, as in all the preceding cases, with a very few exceptions, we find a general increase in the temperature coefficients with increase in dilution. In this particular case the coefficients are largest in the pure solvent, ethyl alcohol, and not in the mixtures. A general examination of the temperature coefficients and conductivity in all of the pure solvents and mixtures thus far studied shows that although the increase in conductivity in acetone is rela- tively very large as compared with the increase in the other solvents, and that the temperature coefficients are relatively very small, yet the percentage increase in the temperature coefficients with increase in dilution is relatively large. 192 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. FIG. 95. — CONDUCTIVITY OF POTASSIUM SULPHOCYANATE IN MIXTURES OF ACETONE AND METHYL ALCO- HOL AT 25°. 13 130 130- 110 100 ? 90 4 4 > 80 70H g 60 1 50 40 25$ 50 /« 75$S Percentage of Acetone FIG. 96. — CONDUCTIVITY OF POTASSIUM SULPHOCYANATE IN MIXTURES OF ACETONE AND ETHYL ALCO- HOL AT 0°. 100 tf 100? Percentage of Acetone VISCOSITY MEASUREMENTS. 193 20- 100 # Percentage of Acetone FIG. 97. — CONDUCTIVITY OF POTASSIUM SULPHOCYANATE IN MIXTURES OF ACETONE AND ETHYL ALCOHOL AT 25°. VISCOSITY MEASUREMENTS. The method used for measuring the viscosity of solutions throughout this work is that employed by Ostwald.1 The viscosities have been calculated from the following formula : in which n0 is the coefficient of viscosity for water, £0 is the specific gravity of water, and to the time of flow of water through any given capillary at a given temperature ; n is the viscosity coefficient of the solution investigated, S is its specific gravity as compared with water as unity at any given tempera- ture, and t is the time of flow of the given solution at that temperature. In the following tables the values for pure water at 0° and 25° were taken from the 1 Ostwald-Luther : Physiko-chemische Messungen, Aufl. 2, p. 259. 194 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. work of Thorpe and Rodger,1 and these values were used as the basis for the calculation of all the values given in tables 84 to 86. The fluidity values were calculated from the following formula : 1 n in which is the fluidity, and n is the corresponding viscosity value. In other words, the fluidity values are simply the reciprocals of the viscosities. All the values given in tables 84 to 86 have been obtained both for pure solvents and solutions, using the modified form of viscometer which has been already described. Many of the values given have been carefully checked by using entirely different solutions in the several determinations. TABLE 84. — Fluidity of potassium sulphocyanate at 0° and 25°. Mixture. V ?iO° 00° «25° 25° Tempera- ture coef- ficient. In water . . . ( 10 1600 0.01737 57.58 0.008823 .008836 113.34 113.17 0.0387 25 p. ct. methjd alcohol [ Solvent 10 1600 .01778 .03203 56.24 31.22 .008910 .01297 .01310 112.23 77.08 76.34 .0398 .0588 and water 50 p. ct. methyl alcohol I Solvent ( 10 1600 .03304 .03526 30.27 28.36 .01312 .01462 .01474 76.18 68.40 67.82 .0607 .0565 and water • • • 75 p. ct. methyl alcohol and water [ Solvent f 10 1600 .03586 .02479 27.89 40.35 .01477 .01201 .01197 67.72 83.24 83.56 .0571 .0425 In methyl alcohol . . . I Solvent j 10 1600 .02451 .009599 40.81 104.18 .01196 .006355 .006085 83.60 157.37 154 34 .64i9 .0204 25 p. ct. ethyl alcohol and water I Solvent | 10 1600 .009032 .04904 110.72 20.39 .006084 .01630 .01653 165.36 61.35 60.50 .0194 .0804 50 p. ct. ethyl alcohol and water I Solvent 1 10 1600 .05135 .06742 19.47 14.83 .01661 .02148 .02168 60.19 46.55 46.13 .0837 .0856 75 p. ct. ethyl alcohol and I Solvent f 10 1600 .07005 .04960 14.27 20.16 .02170 .01968 .01939 46.08 50.81 51 57 .0892 .0608 water In ethyl alcohol .... [ Solvent 1 10 1600 .04996 .02237 20.01 44.71 .01935 .01261 .01197 51.68 79.28 83 54 .0633 .0309 I Solvent .02108 47.44 .01145 87.36 .0337 1 Phil. Trans., 185A, 307 (1894). FLUIDITY. 195 TABLE 84. — Fluidity of potassium sulphocyanate at 0° and 25°. — Continued. Mixture. V nO° 00° 1125° $25° Tempera- ture coef- ficient. 25 p. ct. acetone and water I 10 \ 1600 0.02849 35.10 0.01202 .01206 83.20 82.92 0.0548 50 p. ct. acetone and water | Solvent ( 10 1600 .02868 .03006 34.87 33.27 .01205 .01268 .01260 83.00 78.88 79.40 .0552 .0548 75 p. ct. acetone and water [ Solvent J 10 1600 .02992 .01737 33.42 57.58 .01258 .009000 .008763 79.52 111.11 114.12 .0552 .0372 In acetone I Solvent / 10 .01695 .005294 59.00 188.91 .008727 .004126 114.58 242.38 .0377 .01132 1 1600 .004010 249.36 25 p. ct. acetone and I Solvent f 10 1600 .005045 .007429 198.24 134.61 .003977 .005380 .005115 251.46 185.86 195.49 .01074 .01522 methyl alcohol .... 50 p. ct. acetone and I Solvent 1 10 1600 .006497 .005632 153.92 177.56 .005087 .004932 004675 196.56 202.74 213.92 .0111 .00567 methyl alcohol .... 75 p. ct. acetone and I Solvent / 10 1600 .005177 .004726 193.16 211.62 .004498 .004567 .004253 222.33 218.99 235.12 .00604 .00139 methyl alcohol .... 25 p. ct. methyl alcohol I Solvent 1 W 1600 .004338 .01707 230.54 58.58 .004155 .009892 .009500 240.60 101.10 105.27 .00175 .0290 and ethyl alcohol . . . 50 p. ct. methyl alcohol I Solvent 1 10 1600 .01617 .01341 61.84 74.56 .009481 .008270 .007905 105.48 120.91 126.51 .0282 .0249 and ethyl alcohol . . . 75 p. ct. methyl alcohol I Solvent 1 10 1600 .01259 .01059 79.45 94.46 .008009 .007115 .006791 124.86 140.56 147.24 .0229 .0195 and ethyl alcohol . . . 25 p. ct. acetone and ethyl I Solvent ( 10 1600 .01003 .01302 99.68 76.80 .006790 .007808 .007617 147.29 128.07 131.29 .0191 .0267 alcohol _ 50 p. ct. acetone and ethyl I Solvent / 10 1600 .01156 .007934 86.53 126.04 .007332 .006027 .005377 lob, 39 165.92 185.98 .0261 ,0127 alcohol 75 p. ct. acetone and ethyl I solvent / 10 1600 .007080 .005353 141.24 186.81 .005333 .004663 .004453 187.51 214.45 224.57 .0131 .00592 alcohol I Solvent .004900 204.07 .004393 227.62 .00462 196 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. TABLE 85. — Comparison of fluidities of potassium sulphocyanate. Mixture. At- V 0 p. ct. 25 p. ct. 50 p. ct. 75 p. ct. lOOp.ct. I 10 57.58 31.22 28.36 40.35 104.18 f 0° { Solvent 56.24 30.27 27.89 40.81 110.72 Methyl alcohol and f 10 113.34 77.08 68.40 83.24 157.37 I 25° 1600 m!7 7fi 34 67 8^ 83 56 164 34 \. ^ »_* [ Solvent • X f 112.23 f \J , O^ 76.18 \J 1 * — 67.72 <_7O« *J\J 83.60 i \J X.t O^ 164.36 I 10 57.58 20.39 14.83 20.16 44.71 f 0° | Solvent 56.24 19.47 14.27 20.01 47.44 Ethyl alcohol and water 10 113.04 61.35 46.55 50.81 79.28 I 25° 1600 113.17 60.50 46.13 51.57 83.54 ( Solvent 112.23 60.19 46.08 51.68 87.36 10 57.58 35.10 33.27 57.58 188.91 f 0° Solvent 56.24 34.87 33.42 59.00 198.24 Acetone and water . . 10 113.34 83.20 78.88 111.11 242.38 I 25° 1600 113.17 82.92 79.40 114.12 249.36 Solvent 112.23 83.00 79.52 114.58 251.46 10 104.18 134.61 177.56 211.62 188.91 Acetone and methyl alcohol ( 0° Solvent 10 110.72 157.37 153.92 185.86 193.16 202.74 230.54 218.99 198.24 242.38 1 25° 1600 164.34 195.49 213.92 235.12 249.36 Solvent 164.36 196.56 222.33 240.60 251.46 f 10 44.71 76.80 126.04 186.81 188.91 Acetone and ethyl alcohol .... f 0° \ Solvent 10 47.44 79.28 86.53 128.07 141.24 165.92 204.07 214.45 198.24 242.38 I 25° 1600 83.54 131.29 185.98 224.57 249.36 [ Solvent 87.3G 136.39 187.51 227.62 251.46 f 10 44.71 58.58 74.56 94.46 104.18 Methyl alcohol and ethyl alcohol . ... 1 0° I 25° \ Solvent 1 10 1600 47.44 79.28 83.54 61.84 101.10 105.27 77.45 120.91 126.51 99.68 149.56 147.24 110.72 157.37 164.34 • [ Solvent 87.36 105.48 124.88 147.29 164.36 TABLE 86. — • Comparison of temperature coefficients of fluidity of potassium sulphocyanate from 0° to 25°. Mixture. V 0 p. ct. 25 p. ct. 50 p. ct. 3 75 p. ct. 100 p. ct. Methyl alcohol and water . . \ 10 [ Solvent 0.0387 .0398 0.0588 .0607 0.0565 .0571 0.0425 .0419 0.0204 .0194 Ethyl alcohol and water . . ! 10 [ Solvent .0387 .0398 .0804 .0837 .0856 .0892 .0608 .0633 .0309 .0337 Acetone and. water I 10 1 Solvent .0387 .0398 .0548 .0552 .0548 .0552 .0372 .0377 .0113 .0107 Acetone and methyl alcohol I ' 10 [ Solvent .0204 .0194 .0152 .0111 .00567 .00604 .00139 .00175 .0113 .0107 Acetone and ethyl alcohol . / 10 I Solvent .0309 .0337 .0267 .0231 .0127 .0131 .00592 .00462 .0113 .0107 Methyl alcohol and ethyl al- cohol ! 10 \ Solvent .0309 .0337 .0290 .0282 .0249 .0229 .0195 .0191 .0204 .0194 FLUIDITY. 197 Tables 84 and 85 (fig. 98) show that potassium sulphocyanate, in mixtures of methyl alcohol and water, exhibits a marked minimum in fluidity in the 50 per cent mixture at 25° and at 0°, although in the latter case there is only a slight difference between the values in the 25 per cent and the 50 per cent mixtures. It is of especial importance to note here that potassium sulphocya- nate shows a marked negative viscosity (or positive fluidity) in water, and that the viscosity values do not become positive until a mixture about midway between 25 Percentage of Methyl Alcohol FIG. 98. 100 i Curve I, fluidities of mixture of methyl alcohol and water at 0°. Curve II, fluidities of N/10 potassium sulphocyanate in mix- tures of methyl alcohol and water atO°. Curve III, fluidities of the above solvent mixtures at 25°. Curve IV, fluidities of N/10 potassium sulphocyanate in the above solvent mixtures at 25°. the 50 per cent and 75 per cent mixtures is reached. In other words, N/10 solution of potassium sulphocyanate in water is much less viscous (or has a much greater fluidity) than pure water itself. On the other hand, a N/20 solution of potassium sulphocyanate in methyl alcohol has a greater viscosity (or smaller fluidity) than the pure solvent itself. These two effects become equal, and we have the viscosity (or fluidity) of the N/10 solution and the pure solvent equal in a mixture intermediate between the 50 per cent and 75 per cent mixtures. The normal or general action of a salt, when dissolved in water, is to increase the viscosity. Cases of negative viscosity have been noted by other workers, and this subject will be discussed later in this memoir. It should be noted that the difference in viscosity between the solution and the pure solvent is greater in methyl alcohol than it is in water, or any of the mixtures. The temperature coefficients of fluidity are greater in the pure solvents than in the solutions, in the 0 per cent, 25 per cent, and 50 per cent mixtures; and vice versa in the 75 per cent and 100 per cent mixtures, and 198 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. the temperature coefficients of fluidity themselves are greatest in the 25 per cent mixture. Tables 84, 85, and 86 (fig. 99) show that potassium sulphocyanate, in mixtures of ethyl alcohol and water, exhibits a decided minimum in fluidity in the 50 per cent mixture. Here, also, there is the negative viscosity coeffi- cient in the aqueous mixtures, up to a mixture intermediate between the 50 per cent and 75 per cent mixtures; from this point on it is positive. The difference in viscosity between the solutions and pure solvent is much greater in the pure ethyl alcohol than it is in pure water; and it will also be noted that the difference between the viscosity of the solution and the pure solvent, in the case of ethyl alcohol, is greater at 25° than at 0°. A study of the tem- perature coefficients of fluidity shows that in all the mixtures they are greater Curve I, fluidities of mixtures of ethyl alcohol and water at 0°. Curve II, fluidities of N/10 potassium sulphocyajnate in mix- tures of ethyl alcohol and water at 0°. Curve III, fluidities of the above solvent mixtures at 25°. Curve IV, fluidities of N/10 potassium sulphocyanate in the above solvent mixtures at 25°. Percentage of Ethyl Alcohol FIG. 99. 100 £ for the pure solvent than they are for the solution. The temperature coefficients of fluidity are greatest in the 50 per cent mixture. The temperature coefficients of fluidity decrease with increase in dilution in the 0 per cent, 25 per cent, and 100 per cent mixtures, and increase with increase in dilution in the 50 per cent and 75 per cent mixtures, which are the mixtures in which the maximum fluidity is shown. Tables 84, 85,. and 86 (fig. 100) show that potassium sulphocyanate, in mixtures of acetone and water, exhibits a minimum in the fluidity curves in the 50 per cent mixtures at both 0° and 25°. Here, again, we have the negative viscosity coefficient in the aqueous solutions and in the 25 per cent mixtures. This negative coefficient becomes zero at some point intermediate between the 25 per cent and the 50 per cent mixtures ; and is positive from that point on throughout the remaining mixtures. We must also call at- tention to the fact that the increase in viscosity with increase in dilution is FLUIDITY. 199 greater in acetone than it is in water, and that the viscosity of acetone is very much less than that of water. The temperature coefficients of fluidity increase in all the mixtures, with the exception of the 100 per cent mixtures, with increase in dilution; and the largest coefficients are in the 25 per cent and 50 per cent mixtures; the values in these two cases being identical to the fourth decimal place. Tables 84, 85, and 86 (fig. 101) show that potassium sulphocyanate, in 260 340 220- 200- 180- "3 140- 120- 100- 80- 60- 40- IV 50^ Percentage of Acetone FIG. 100. 100$ Curve I, fluidities of mixtures of acetone and water at 0°. Curve II, fluidities of N/10 potassium sulphocyanate in mix- tures of acetone and water at 0°. Curve III, fluidities of the above solvent mixtures at 25°. Curve IV, fluidities of potas- sium sulphocyanate in the above solvent mixtures at 25°. mixtures of acetone and methyl alcohol, exhibits a marked maximum which disappeared at 25°. The curves at 25°, however, show an increase in the fluidity at 0°, in the 75 per cent mixtures, but that this maximum has values above the average values for the two solvents. It is to be noted also, that although the increase in fluidity with increase in dilution is nearly the same in acetone and methyl alcohol, yet this increase is very much greater in the mixtures, and especially in the 50 per cent and 75 per cent mixtures. Tables 84, 85, and 86 (fig. 102) show that tenth-normal potassium sulpho- cyanate, in mixtures of acetone and ethyl alcohol, exhibits a maximum in fluidity in the 75 per cent mixture at 0°. The pure solvent at 0°, and the solutions and solvent at 25° do not exhibit this maximum, although they do 200 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. show an increase above the values calculated on the basis of averages. It is to be noted also that the increase in fluidity with increase in dilution is greatest in the 50 per cent and 75 per cent mixtures. The temperature coefficients of fluidity increase with increase in dilution in the 0 per cent and 50 per cent mixtures; but decrease with increasing dilu- tion in the 25 per cent, 75 per cent, and 100 per cent mixtures. The largest values are found in the 0 per cent solutions. Tables 84, 85, and 86 (fig. 103) show that potassium sulphocyanate, in mixtures of methyl alcohol and ethyl alcohol, exhibits an increase in fluidity above the average values in the 50 per cent and 75 per cent mixtures. The temperature coefficients of fluidity decrease with increase in dilution 260- 250 220 200 » 180 1'60 140 120 100 Curve I, fluidities of N/10 potassium sulphocyanate in mix- tures of acetone and methyl alcohol at 0°. Curve II, fluidities of mixtures of acetone and methyl alcohol at 0°. Curve III, fluidities of N/10 potassium sulphocyanate in the above mixtures at 25°. Curve IV, fluidities of the above solvent mixtures at 25°. 25jS 5056 75$ Percentage of Acetone FIG. 101. in every case, except the 0 per cent mixture, the maximum value being in the 0 per cent mixture. Table 85 shows that the temperature coefficients of fluidity of potassium sulphocyanate, in mixtures of methyl alcohol and water, increase with increase in dilution of the solutions in the 0 per cent, 25 per cent, and 50 per cent mixtures; and decrease with increase in dilution in the 75 per cent and 100 per cent mixtures. The temperature coefficients of conductivity of copper chloride in these mixtures increase with increasing dilution in every case, except the 100 per cent mixture, where the opposite condition holds. The temperature coefficients of conductivity of potassium sulphocyanate increase with increase in dilution in all of the solvents. The temperature coefficients of fluidity are largest in the 25 per cent mixture, and such is the case also for FLUIDITY. 201 the temperature coefficients of conductivity of both copper chloride and potassium sulphocyanate. Table 87 shows that the temperature coefficients of fluidity of potassium sulphoc}^anate, in mixtures of ethyl alcohol and water, increase in all cases with increase in dilution. The same is true of the temperature coefficients of conductivity of both copper chloride and potassium sulphocyanate. The temperature coefficients of fluidity are largest in the 50 per cent mixtures, Curve I, fluidities of N/10 potassium sulphoeyanate in mix- tures of acetone and ethyl alcohol at 0°. Curve II, fluidities of mixtures of acetone and ethyl alcohol at 0°. Curve III, fluidities of N/10 potassium sulphocyanate in the above mixtures at 25°. Curve IV, fluidities of the above solvent mixtures at 25°. Percentage of Acetone FIG. 102. 1000 and the same is true of the temperature coefficients of conductivity of both the above-mentioned salts. Table 87 shows that the temperature coefficients of fluidity of potassium sulphocyanate, in mixtures of acetone and water, increase with increasing dilution, with the exception of the solutions in the pure acetone. The maxi- mum values for these temperature coefficients are found in the 25 per cent and 50 per cent mixtures. The temperature coefficients of conductivity of potassium sulphocyanate, in these mixtures, increase with increase in dilu- tion, the maximum values being in the 50 per cent mixture. Table 87 shows that the temperature coefficients of fluidity of potassium sulphocyanate, in mixtures of acetone and methyl alcohol, decrease with 202 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. increase in dilution in the 0 per cent, 25 per cent, and 100 per cent mixtures, and vice versa in the 50 per cent and 75 per cent mixtures. The maximum values are in methyl alcohol. The temperature coefficients of conductivity of potassium sulphocyanate, on the other hand, increase with increase in dilution in every case, and the maximum values are in methyl alcohol. Table 87 shows that the temperature coefficients of fluidity of potassium sulphocyanate, in mixtures of acetone and ethyl alcohol, increase with increase Curve I, fluidities of N/10 potas- sium sulphocyanate in mixtures of methyl alcohol and ethyl alcohol at 0°. Curve II , fluidities of mixtures of methyl alcohol and ethyl alcohol at 0°. Curve III, fluidities of N/10 potas- sium sulphocyanate in the above mix- tures at 25°. Curve IV, fluidities of the above mixed solvents at 25°. 25;* 50ft 75ji Percentage of Methyl Alcohol FIG. 103. in dilution in the 0 per cent and 50 per cent mixtures, but decrease in the 25 per cent, 75 per cent, and 100 per cent mixtures. The maximum values are in the 0 per cent mixture. The temperature coefficients of conductivity of potassium sulphocyanate in these mixtures increase in every case with in- crease in dilution. The maximum values are in the 0 per cent mixture. Table 87 shows that the temperature coefficients of conductivity of copper chloride increase with increase in dilution in all cases, except the 100 per cent mixture, the maximum values being in the 0 per cent mixture. The corresponding temperature coefficients for potassium sulphocyanate increase with increasing dilution in all cases, and the maximum values are in the 0 per cent mixture. A SUMMARY OF THE FACTS ESTABLISHED. (1) A minimum in conductivity has been noted in the following cases : Potassium sulphocyanate in 50 per cent methyl alcohol and water at 0°. This shifts to the 75 per cent mixture at 25°. Also in 50 per cent acetone TEMPERATURE COEFFICIENTS. 203 and water at 0° and 25°, with the single exception of the N/ 10 solution at 25°. Likewise, in the 50 per cent and 75 per cent mixtures of ethyl alcohol and water, with the exception of the N/ 10 solution at 0°. At 25° the N/800 and N/1600 solutions show minimum values, these occurring in the 75 per cent mixture. TABLE 87. — Comparison of the temperature coefficients of conductivity and fluidity. V Dissolved substance. 0 p. ct 25 p. ct. 50 p. ct. 75 p. ct. 100 p. ct. In mixtures of r Fluidity . . ( 10 1 Solvent KCNS 0.0387 0398 0.0588 .0607 0.0565 0571 0.0425 0419 0.0204 D1Q4 methyl alcohol " i 10 CuCl2 .0329 .0421 !0372 !0237 t\j i • ' i .00984 and water. L Conductivity 1600 10 CuCl2 KCNS .0371 .0316 .0475 .0410 .0437 .0400 .0318 .0304 .00823 .0154 I 1600 KCNS .0325 .0419 .0429 .0352 .0169 In mixtures of Fluidity . . j 10 1 Solvent KCNS .0387 .0398 .0804 .0837 .0856 0892 .0608 .0633 .0309 0^7 ethyl alcohol 10 CuCl2 .0329 .0543 !0556 !0360 • v/OO i .0153 and water. Conductivity j 1600 10 CuC!2 KCNS .0371 .0316 .0635 .0518 .0683 .0590 .0432 .0449 .0235 .0224 i 1600 KCNS .0325 .0600 .0636 .0490 .0268 In mixtures of 'Fluidity . . ( 10 1 Solvent KCNS .0387 .0398 .0548 .0552 .0548 .0552 .0372 .0337 .0113 .0107 acetone and 10 CuCl2 .0329 water. 1600 CuCl2 .0371 Conductivity 10 KCNS .0316 .0398 .0402 .0283 .00481 I 1600 KCNS .0325 .0404 .0432 .0312 .00815 In mixtures of 'Fluidity . . j 10 I Solvent KCNS .0204 .0194 .0152 .0111 .00567 .00604 .00139 .00175 .0113 .0107 acetone and 10 CuCl2 .00984 methyl alcohol. . . 1600 CuCl2 .00823 Conductivity 10 KCNS .0154 .0138 .0123 .00971 .00481 L 1600 KCNS .0169 .0157 .0140 .01176 .00815 In mixtures of Fluidity . . ( 10 1 Solvent KCNS .0309 .0337 .0267 .0231 .0127 .0131 .00592 .00462 .0113 .0107 acetone and • f 10 CuCl2 .0153 ethyl alcohol. 1600 CuCl2 .0235 Conductivity 10 KCNS .0224 .0172 .0126 .00868 .00481 [ 1600 KCNS .0268 .0201 .0175 .01237 .00815 In mixtures of Fluidity . . j 10 1 Solvent KCNS .0309 .0337 .0290 .0282 .0249 0229 .0195 0191 .0204 .0194 methyl alcohol and ethyl alco- hol. Conductivity ( 10 1600 1 10 CuCl2 CuCl2 KCNS .0153 .0235 .0224 .00992 .01831 .0204 .00901 .01620 .0183 .00853 .01260 .0166 .00984 .00823 .0154 L 1600 KCNS .0268 .0237 .0209 .0181 .0169 (2) A decided fall below the average values for the two pure solvents has been noted in the following mixtures : (a) Copper chloride in methyl alcohol and water, the fall being most pronounced in the 50 per cent mixture ; in ethyl alcohol and water in the 25 per cent and 50 per cent mixtures; in methyl alcohol and ethyl alcohol in the 25 and 50 per cent mixtures. (6) Potassium sulphocyanate in mixtures of methyl alcohol and ethyl alco- hol, the concentrations being N/400, N/800, and N/1600; in mixtures of acetone and methyl alcohol, and in the mixtures of acetone and ethyl alcohol. (3) It has been noted that in every case where there is not an actual mini- mum shown by the curves, but simply the falling below the average values, there is a wide difference between the conductivity values in the two unmixed or pure solvents. It is quite evident that this difference must play a large 204 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. part in determining whether there will be an actual minimum in the curves or not; and we wish especially to point out the fact that although an actual minimum is not shown in every case, yet we have virtually a minimum in the conductivity values whenever the pure solvents are mixed ; and that this minimum occurs in the 50 per cent and 75 per cent mixtures. There are only a very few exceptions to the above statement, and these exceptions will be pointed out in paragraph 4. (4) A maximum in conductivity has been noted in the following cases: Potassium sulphocyanate in 75 per cent acetone and methyl alcohol in the N/10, N/50, and N/100 solutions at 0°, and in the N/10, N/50, N/100,and N/200 solutions at 25°. In acetone and ethyl alcohol the N/10 solutions, in the 75 per cent mixture, show a pronounced maximum; and a decided ten- dency towards a maximum is shown by the N/50, N/100, and N/200 solutions. (5) It has been noted that there is a marked difference not only between the actual numerical values for molecular conductivity, but also between the corresponding increase in the values with increase in dilution, for copper chlo- ride (a ternary electrolyte) and potassium sulphocyanate (a binary electrolyte) in a given pure solvent. In pure water, for example, the conductivity of copper chloride at 25° increases from 162.6, in the N/10 solution, to 249.7, in the N/1600 solution, while the conductivity of potassium sulphocyanate under the same conditions increases from 112.7 to 131.6. In other words, the conductivity values for copper chloride in water are much greater than the corresponding values for potassium sulphocj^anate. In pure methyl alcohol the conductivity of copper chloride at 25° increases from 17.98 to 72.64, while the conductivity of potassium sulphocyanate under the same conditions increases from 64.81 to 102.41. Thus we see that although the conductivity of copper chloride, in water, is much greater than the con- ductivity of potassium sulphocyanate in aqueous solution, yet in methyl alcohol exactly the reverse is true, i. e., the conductivity of copper chloride in methyl alcohol is much less than that of potassium sulphocyanate under the same conditions. In pure ethyl alcohol the conductivity of copper chloride at 25° increases from 4.88 to 25.24. Potassium sulphocyanate, on the other hand, under the same conditions, increases from 22.97 to 42.73. Data similar to those just referred to have been obtained by other workers in this field. From the work of Jones and McMaster l we see that lithium bromide (a binary electrolyte) shows an increase in conductivity from the N/10 solution to the N/1600 solution in water, of from 86.09 to 106.23. Cobalt chloride (a ternary electrolyte) shows an increase of from 156.36 to 221.50, over the same range in dilution. 1 Amer. Chem. Journ., 36, 335-381 (1906). SUMMARY OF FACTS. 205 In methyl alcohol they found that lithium bromide increases from 50.21 to 83.64, and cobalt chloride from 41.78 to 133.33 for the same increase in dilution. In ethyl alcohol lithium bromide increases from 17.22 to 33.36; and cobalt chloride from 7.64 to 33.59, between the N/10 and the N/1600 solutions. In acetone lithium bromide increases from 110.82 to 85.90 over the above range in dilution, and cobalt chloride increases from 9.47 in the N/100 solution to 10.45 in the N/1600 solution. Jones and Bingham * obtained the following data with lithium nitrate, potassium iodide, and calcium nitrate: Lithium nitrate in water, between the N/10 and N/1600 solutions increases from 83.9 to 102.8 ; potassium iodide in water between the N/200 and N/1600 solutions increases from 136.3 to 140.7; calcium nitrate in water between the N/10 and N/1600 solutions increases from 165.5 to 249.8. Here, again, we observe that the ternary electrolytes show a much greater conductivity in aqueous solutions than do the binary electrolytes. Lithium nitrate in methyl alcohol, between the N/10 and N/1600 solutions increases from 51.31 to 86.7; potassium iodide in methyl alcohol between the N/200 and N/1600 solutions increases from 91.4 to 103.3; calcium nitrate in methyl alcohol between the N/10 and the N/1600 solutions increases from 17.17 to 35.4; potassium iodide in ethyl alcohol between the N/200 and N/1600 solutions increases from 34.6 to 42.8. Calcium nitrate in ethyl alcohol between the N/10 and the N/1600 solu- tions increases from 7.86 to 33.3; lithium nitrate in acetone under the same conditions increases from 10.87 to 59.8; potassium iodide in acetone between the N/200 and the N/1600 solutions increases from 118.0 to 141.1; and calcium nitrate in acetone between the N/210 and the N/1600 solutions increases from 5.67 to 12.62. (6) The temperature coefficients of conductivity increase with increase in dilution, with one exception. This exception is copper chloride in methyl alcohol. The increase is not regular, but it is quite decided when the differ- ence in the values for the N/10 and the N/1600 solutions is considered. (7) The temperature coefficients of conductivity are always a maximum in the mixtures of water with the alcohols or acetone, and are never a maximum in the mixtures of the alcohols with each other or with acetone. The individual facts upon which the above statement is based are as fol- lows: Copper chloride in 25 per cent methyl alcohol and water; in 50 per cent ethyl alcohol and water; potassium sulphocyanate in 25 per cent and 50 per cent mixtures of methyl alcohol and water, and in 50 per cent acetone and water, and in 50 per cent ethyl alcohol and water. The data obtained by Jones and McMaster 2 show that this is also true for 1 Amer. Chem. Journ., 34, 497-534 (1905). 2 Loc. cit. 206 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. lithium bromide in 25 per cent methyl alcohol and water, in 50 per cent ethyl alcohol and water, and in 50 per cent acetone and water. It is also true for cobalt chloride, in 25 per cent methyl alcohol and water, in 50 per cent ethyl alcohol and water, and in 75 per cent acetone and water. By reference to the work of Jones and Bingham 1 we see that this is likewise true for lithium nitrate in 25 per cent acetone and water, for potassium iodide in 50 per cent acetone and wrater, and for calcium nitrate in 25 per cent and 50 per cent acetone and water. (8) The molecular conductivities of potassium sulphocyanate in acetone, as compared with the corresponding conductivities in water, bring out some important facts. The conductivity values in the N/10 solutions are usually much smaller in acetone than they are in water, but with increase in dilution they become very much larger in acetone than they do in water. The same thing is seen to be true when the conductivity of the acetone solutions is com- pared with the solutions in the alcohols, but to a smaller degree. The values become greater in acetone than they do in either of the alcohols. (9) A marked minimum in fluidity has been noted in the following cases : 50 per cent methyl alcohol and water at 0° and 25°, 50 per cent ethyl alcohol and water, 50 per cent acetone and water. (10) A maximum in fluidity has been noted in the following cases: In 75 per cent acetone and methyl alcohol at 0°. Although this maximum has disappeared at 25°, yet there is a decided increase of the values in the mix- ture above the average values; in N/10 potassium sulphocyanate in 75 per cent acetone and ethyl alcohol at 0°. The pure solvent does not show the maximum at 0°, and the maximum has entirely disappeared at 25° in both the solution and the solvent, but nevertheless there is an increase above the average values in each case in the 75 per cent mixture. This increase above the average values is also shown by the mixtures of methyl alcohol and ethyl alcohol. (11) A marked negative viscosity coefficient is shown by potassium sul- phocyanate in aqueous solution. In the other pure solvents, methyl alcohol, ethyl alcohol, and acetone, potassium sulphocyanate gives a positive viscosity coefficient, and in mixtures of the other solvents with water the viscosity coefficient becomes zero in the following mixtures : At a point about midway between the 50 per cent and 75 per cent mixtures of methyl alcohol and water; and in the same mixtures of ethyl alcohol and water, and between the 25 per cent and 50 per cent mixtures of acetone and water. (12) The temperature coefficients of fluidity are a maximum in the mixtures of the other solvents with water in the 25 and 50 per cent mixtures, and in no case are they a maximum in the mixtures of the alcohols and acetone with one another. ' Loc. cit. DISCUSSION OF RESULTS. IT would be desirable for some reasons to discuss the facts brought out by our study of conductivity in one section, and the results of the work on fluidity in a separate section ; but on account of the close relations between conduc- tivity and fluidity it seems best to take up the discussion in such a manner as to bring out more clearly the bearing of these relations upon one another. An explanation of the minimum in molecular conductivity has been offered by Jones and Lindsay,1 and, as has already been pointed out in the introduc- tory section of this paper, this has been experimentally substantiated by the work of Jones and Murray.1 This explanation was, however, applied only to the cases where an actual minimum occurs, but, as we have already pointed out, there are fully as many cases in which the curves show simply a falling below the values as calculated from the rule of averages. We have shown that in these cases we have what we may call a virtual minimum, and we extend the hypothesis of Jones and Lindsay to these cases also, since it bears as we believe on both the actual minima and the virtual minima. Since these two conditions cover by far the greater majority of the conductivity results that have been obtained up to the present in mixed solvents, it appears that the hypothesis of Jones and Lindsay when applied to the problem of conductivity in mixed solvents is perfectly general. A further proof of this explanation has been brought out by our study of fluidity. A marked minimum in fluidity occurs in the 50 per cent mixtures of water and methyl alcohol, water and ethyl alcohol, and water and acetone. It is in these same mixtures that the minimum in conductivity occurs. F. H. Getman,2 in discussing the maximum viscosity (or minimum fluidity, which occurs when the alcohols and water are mixed, drew the conclusion that if the association of one liquid is diminished by the presence of another associated liquid, then the viscosity of a 50 per cent mixture of these liquids should be less than the viscosity as calculated from the rule of averages. This conclusion, as we shall see, is erroneous. The work of Thorpe and Rodger 3 has clearly shown that viscosity may be taken as the sum of the forces in play between the molecules. Therefore, when two associated liquids are mixed, if they mutually decrease the association of one another, the total number of molecules in a given volume of the mixtures is increased, and consequently 1 Loc. cit. 2 Journ. chim. Phys., 4, 403 (1906). 3 Phil. Trans., 185A, 307 (1894). 207 208 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. the total surface of the molecules of the solvents is increased, i. e., the total frictional surface is increased, and an increase in viscosity results. The point at which the viscosity of the mixed solvents becomes a maximum, is, then, the point at which the effect above mentioned of the pure solvents on each other is the greatest. Thus we see that at the point where we find the maxi- mum in viscosity we have the greatest number of simple molecules of the two solvents present. According to this view of the cause of the fluidity minimum (or viscosity maximum), the reason for the minimum in the conductivity curve follows at once. Since the association of the solvents is a minimum at the point of minimum fluidity, the dissociating power of the solvents is also at a minimum, and the minimum in conductivity is a natural consequence. As has already been stated, it is at the minimum points of conductivity that we have the phenomenon of a very small increase in molecular conductivity with increase in dilution, as is shown by the conductivity curves for the different dilutions approaching one another as they approach the minimum. This also is a natural consequence of our theory. The association of the solvents being a minimum at these points, as has already been shown, their dissociating power is also a minimum, and, consequently, an increase in the amount of the solvent present has little influence on the dissociation of the dissolved salt. To elabo- rate this a little more fully, let us suppose that the salt were dissolved in a solvent which had no dissociating power; then no matter how much of the solvent were present, the dissociation of the salt and, consequently, the molecular conductivity of the solution would remain zero. If, now, a solvent is used which has very small dissociating power, it is evident that compara- tively large amounts of it would be required to produce any very appreciable effect upon the dissociation of the dissolved salt. In a few cases a maximum in conductivity was observed. These cases occur in certain of the mixtures of acetone with the alcohols. Similar examples, and in these same solvents, were first observed by Jones and Bingham,1 who suggested the explanation that these maxima are due either to an increase in the number of ions present, or to a diminution in the size of the ionic spheres, making possible a more rapid movement of the ions. As has been pointed out, their final conclusion is that the latter explanation, i. e., a diminution in the size of the ionic spheres, is the more probable. The question, however, arises: Why does this change in the size of the ionic spheres take place ? The experimental data, obtained up to the present, are not sufficient to justify a final answer to this question. We offer the following suggestion, however, as probable : The increase in fluidity, in view of what has already been said with regard to the cause of maxima and minima in the fluidity 1 Loc. cit. DISCUSSION OF RESULTS. 209 curves, is probably due to an increase in the molecular aggregation of the solvent. This increase is not necessarily caused by an increase in the association of either of the pure solvents. Indeed, the association of the pure solvents is in all probability less in the mixtures. The increase in molecular aggregation above referred to may simply be due to a molecular combination of the un- associated parts of the pure solvents. The view that is now generally held with regard to chemical combination is that such combination usually takes place between the ions and not between the molecules. We may, in a sense, regard a molecule that is broken down into its ions as in a state of diminished atomic aggregation, just as we regard the diminution in association, when two pure solvents are mixed, as a state of diminished molecular aggregation. The ions, to be sure, are in a different physical condition from the undis- sociated molecule, in that they carry equal and opposite electrical charges ; and we have no experimental reason whatsoever for supposing that the un- associated portions of the solvent molecules, in the mixed solvents, are in the same condition. It is, however, the difference in energy relations between the ions that is at the basis of our present views of chemical combination. Rea- soning from analogy, it is not unthinkable, to say the least, that a similar, although not identical difference in energy relations is the cause of molecular aggregation or association. If such an energy relation as has been referred to does exist, then we have an explanation of the formation of complex mole- cules, such as hydrates, double salts, etc. If the above reasoning is correct, the explanation of the fluidity maximum is simple. According to Ramsay and Shields l the molecular complexity of the four solvents is as follows : Substance. Molecular complexity between 0° and 25°. Water (H"O)4. Methyl alcohol .... Ethyl alcohol .... Acetone (CH3OH)3.4 (C2H5OH)2.7 (CHsCOCHaW From these data we see that the molecules of methyl alcohol and the mole- cules of ethyl alcohol are much more complex than the molecules of acetone. For the sake of simplicity we shall limit our discussion to the case of mixtures of acetone and methyl alcohol. The maximum in fluidity occurs in the 75 per cent mixture of the two solvents. We should expect the effect of the acetone upon the methyl alcohol to be greatest in about this mixture, because of the relatively large mass of the acetone present ; and, consequently, in such a mixture we should have the largest number of unassociated molecules of methyl alcohol. This is shown to be the case by reference to the temperature 'Ztschr. phys. Chem., 12, 433 (1893). 210 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. coefficients of fluidity of the pure solvents, which are a minimum in the 75 per cent mixture. The methyl alcohol, being much more associated than the pure acetone, a small mass of it would cause a proportionally greater decrease in the association of the acetone, and in about the 75 per cent mixture we should expect to find the condition for molecular combination of these two solvents most favorable. The complex molecule thus formed explains the fluidity maximum, in terms of the view of viscosity and fluidity maxima and minima that we have already suggested. Since the fluidity of the solvent has become much greater, the ionic velocity is increased. Thus we conclude that the maximum in conductivity is dependent upon two factors : The cliange in the size of ionic spheres and the change in the fluidity of the solvent. Reference to the work of J. J. Thompson,1 Briihl,2 Nernst,3 Ciamician,4 Dutoit and Aston,5 Ramsay and Shields,8 Crompton,7 and Donnan,8 shows that the prevailing idea concerning the action of dissociating solvents upon dis- solved electrolytes is that the dissociating action of the solvent is mainly a function of the dielectric constant and the degree of association of the solvent. We wish, however, to call attention again to the fact that the dissociating action of the solvent is also dependent largely upon the nature of the dis- solved electrolyte. We have already seen that the conductivity of copper chloride in water is much greater than the corresponding conductivity of potassium sulphocyanate in water. Now if the dissociating action of all solvents were solely a function of the properties of the solvent, we should expect the conductivity of copper chloride in any other solvent to be greater than the corresponding conductivity of potassium sulphocyanate in that solvent. But such is not the case, since we find that in methyl alcohol, for example, the conditions are exactly the reverse of what they are in water, and, further, that potassium sulphocyanate has a much greater conductivity in methyl alcohol than copper chloride has under the same conditions. A number of other similar cases have been mentioned in the "General summary of facts established" in this section. These facts show conclusively that the nature of the dissolved salt also plays a large part in determining what the dissociating action of any given solvent will be. The prob- able explanation of the marked difference in the action of dissociating sol- vents towards binary and ternary electrolytes is, in the case of water and methyl alcohol for example, as follows: Water, being a highly associated solvent, has the power of breaking down the ternary electrolytes into the simplest ions, each molecule of the ternary electrolyte yielding three ions. 'Phil. Mag., 36, 320 (1893). 6 Compt. rend., 125, 240 (1897). 2 Ztschr. phys. Chem., 13, 531 (1894). 6 Ztschr. phys. Chem., 12, 433 (1893). 3 Ibid., 18, 514 (1895); 27, 319 (1898); 30, 7 Journ. Chem. Soc., 71, 925. 1 (1899). 8 Phil. Mag., (6) 15, 305. 4 Ibid., 6, 403 (1890). DISCUSSION OF RESULTS. 211 Methyl alcohol, on the other hand, being less associated than water, probably has the power to break down the ternary electrolytes into only two ions, whereas the binary electrolytes are broken down into ions in both the water and the alcohol. The increase in the temperature coefficients with increase in dilution in aqueous solutions has been explained by Jones,1 and the same explanation doubtless holds for the increase in the temperature coefficients with increase in dilution in the mixed solvents. In the mixed solvents we probably have a sphere around the ions, which is composed of molecules of both the pure solvents. The fact that such solvates are formed in other solvents than water has been established by the work of Jones and McMaster.2 The temperature coefficients of conductivity are a maximum in the 25 per cent and 50 per cent mixtures of water and the other solvents, as has al- ready been pointed out. At first glance this seems to be a remarkable fact, in view of the large amount of experimental evidence that has been furnished by Jones and his co-workers upon the hydrate theory as proposed by Jones. There seems to be but one explanation, viz, that in these particular mixtures the most complex solvates are formed. These complex solvates change in complexity more rapidly with change in temperature than the less complex solvates, and, consequently, the temperature coefficients are a maximum. The question, however, naturally arises, Why are the solvates more complex in one mixture than in another? It will be observed that the maximum temperature coefficients of conductivity in mixtures of methyl alcohol and water are found, as a rule, in the 25 per cent mixtures; yet they also occur in some cases in the 59 per cent mixtures. The maximum values of these ternf .-ature coefficients of conductivity in mixtures of ethyl alcohol and water occur most frequently in the 50 per cent mixture, although in some cases they are found in the 25 per cent mixture. The maximum values of the temperature coefficients of conductivity in mixtures of acetone and water are mainly in the 50 per cent mixture, a few values occurring in the 75 per cent mixture. These facts are significant when considered in the light of the degree of association of the solvents involved, as shown by the table previously given. Acetone being the least associated of this group of solvents, we should ex- pect its greatest action in diminishing the association of water to occur in about the 75 per cent mixture, where the mass of the acetone is very great. We should expect this effect to manifest itself in the smaller percentage mix- tures, when mixtures with water, of more highly associated solvents than acetone, are considered. A glance at the fluidity values shows that although mixtures of methyl alcohol and water show a minimum in the 50 per cent 'Amer. Chem. Journ., 35, 445 (1906). 2 Ibid., 35, 316 (1906). 212 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. mixture, yet there is a much more pronounced tendency towards a minimum in the 25 per cent mixture than there is in the 75 per cent mixture. We see also that although the mixtures of ethyl alcohol and water show a minimum in the 50 per cent mixture, yet there is a more marked tendency towards a minimum in the 75 per cent mixture than there is in the 25 per cent mixture. The case of mixtures of acetone and water is not quite so clean cut, but here again the minimum in fluidity is in the 50 per cent mixture; and although the actual value of the fluidity in the 75 per cent mixture is greater than it is in the 25 per cent mixture, yet, when the very great difference between the fluidity of acetone and the fluidity of water is considered, it is readily seen that in this case also the tendency towards a minimum in fluidity is more marked in the 75 per cent mixture than it is in the 25 per cent mixture. These facts show clearly that in about the 25 and 50 per cent mixtures of the other solvents with water we have the most favorable conditions for the formation of molecular aggregations, such as solvates, between the ions of the dissolved salt and the molecules of the solvents. In other words, we should have the greatest number of simple solvent molecules present in the mixtures above mentioned ; and, as we have already pointed out in another connection, these are the conditions under which we can most reasonably expect the greatest combination between the solvent and the dissolved salt. The fact that the molecular conductivities of potassium sulphocyanate in acetone are smaller than they are in water for the more concentrated solu- tions, but are much greater in acetone than they are in water for the more dilute solutions, might be due to several causes. First, a much greater degree of dissociation, and a more rapid increase in dissociation with increase in dilution, in acetone than in water. This view, however, is untenable in view of the- fact that acetone is very much less asso- ciated than water, and has a much smaller dielectric constant. Indeed, these facts lead us to the conclusion that potassium sulphocyanate is much less dissociated in acetone than it is in water. Second, a much greater velocity of the ions in acetone than in water. That this is the probable explanation is shown by the following considerations: The fluidity of acetone is very much greater than the fluidity of water, and this, to be sure, is one of the factors that governs an increase in the ionic velocity; but a far more important factor becomes manifest from a study of the temperature coefficients of conductivity. It is a very significant fact that the temperature coefficients of conductivity in water are nearly ten times as great as the corresponding coefficients in acetone. This shows quite clearly that the solvates formed in the acetone solutions are very much less complex than those formed in the aqueous solutions. Since the ions in acetone have, as we have just seen, a much smaller atmosphere of solvent to DISCUSSION OF RESULTS. 213 carry with them as they move through the solution than they have in the aqueous solutions, their velocity becomes much greater. These two factors both act in the same direction, i. e., they both tend to increase the ionic velocity, and are quite sufficient to justify the conclusion that the molecular conductivities of potassium sulphocyanate are greater in acetone than they are in water, because of a very great increase in the velocity of the ions in the acetone solutions. NEGATIVE VISCOSITY COEFFICIENTS. It has been known for a number of years that when certain salts are dis- solved in water the resulting solutions have a smaller viscosity than the pure water. Several theories have been proposed to explain this phenomenon. A brief but comprehensive outline of these theories has been given by Jones and Bingham,1 and a mere reference to the literature will suffice in this connection. Euler 2 employed the " electrostriction theory " of Drude and Nernst 3 as a probable explanation of the negative viscosities ; but it was later shown by Wagner 4 that this theory is incorrect, because the vis- cosity of a solvent may be lowered by the addition of certain non-electro- lytes. Dunstan,5 Blanchard,6 Varenne and Godefroy,7 Thorpe and Rodger, and Traube 8 all seem to attribute the abnormalities in viscosity to the pres- ence of hydrates. The hydrate work of Jones and his co-workers has clearly shown that potassium chloride and similar salts are but little hydrated, even in dilute solutions; and since potassium chloride is one of the salts which produces a marked negative viscosity when dissolved in water, it is very diffi- cult to see how hydrates could enter into the question of negative viscosity to any appreciable extent. Wagner 9 has made a study of the effect on the viscosity of water of about forty-five inorganic salts. From his data we find that the only salts which produce negative viscosity are the salts of caesium, rubidium, potassium, and thallium (in the thallous conditions), viz, caesium chloride, rubidium chloride, potassium chloride, potassium nitrate, and thallous nitrate. All potassium salts do not have this property of diminishing the vis- cosity of water. Potassium sulphate, potassium ferrocyanide, potassium fer- ri cyanide, and potassium chromate, all give positive viscosity coefficients. But this is not at all surprising, since, as has already been mentioned, it was clearly shown that the viscosity of a salt solution is an additive function of the metallic and the non-metallic ions of the dissolved salt. In other words, the cations and anions seem to work counter to each other in some cases, such as potassium sul- •Loc. cit. 7Compt. rend., 137, 992 (1903); 138, 'Ztschr. phys. Chem., 25, 536 (1898). 990 (1904). 3 Ibid., 15, 79 (1894). 8 Phil. Trans., 185A, 307 (1894). 4 Ibid., 46, 867 (1903). 9 Ztschr. phys. Chem., 5, 31 (1890). 8 Ibid., 49, 590 (1904). 6 Journ. Amer. Chem. Soc., 26, 1315 (1904). 214 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. phate, where the S04 ions tend to produce a positive viscosity coefficient to such a great extent as to overcome entirely the negative effect of the potassium ion ; and the resultant action is the production of a positive viscosity coefficient by potassium sulphate when dissolved in water. The above facts show very clearly that there is some close connection between the physical properties of caesium, rubidium, and potassium and the negative viscosity phenomena. In dis- cussing viscosity in general, Thorpe and Rodger l state that " Viscosity is, no doubt, the net result of at least two distinct causes. When a liquid flows, during the actual collision or contact of the molecules, a true, friction-like force will be called into play, opposing the movement. But in addition to this cause, even after the actual collision, molecular attractions will exercise a resistance to forces which tend to move one molecule past another." Our study of viscosity has led us to the same conclusion as that arrived at by Thorpe and Rodger, viz, that viscosity phenomena are largely a function of the frictional surfaces of the ions, molecules, and molecular aggregates in any given solution. If now by any possible means the total frictional surface should be decreased, then the viscosity of the solution would also be decreased. Suppose that into a solvent containing molecules with comparatively small molecular volume we bring a salt which yields particles having relatively large atomic or ionic volumes. The larger particles of the dissolved substance would be distributed among the smaller molecules of the solvent, and these smaller molecules, instead of coming in contact with each other so frequently, would come in contact with the larger salt particles, and thus the total fric- tional surface involved would be decreased and, consequently, the viscosity of the solution would also be decreased. In other words, the effect of bringing those salts whose ions have large atomic volumes into pure water, would be the same as if some of the molecules of the water had combined into larger spheres, and thus caused a diminution in the total frictional surface. Since most salts do not produce such an effect on water, we must assume that their atomic volumes are too small, and that only the salts of those metals having very large atomic volumes could produce a diminution in the viscosity of water. The explanation proposed above for the negative viscosity phenomena can then be easily tested by a glance at the atomic volume curves of the ele- ments. When this test is applied, we find that csesium, rubidium, and potassium stand at the very maxima of the atomic volume curve, and that none of the other elements have anything like as large atomic volumes as the three elements just named. Furthermore, if our hypothesis is correct, then that ion which has the largest atomic volume should produce the greatest decrease in the viscosity of water, and the ion having the next smaller atomic volume should produce smaller decrease in the viscosity of water. Here again we find that our theory 1 Loc. cit. DISCUSSION OF RESULTS. 215 is in perfect accord with the facts, as far as they are recorded. According to the work of Wagner,1 caesium chloride in normal solutions lowers the value of t}2 from 1.0000 to 0.9775; rubidium chloride under the same conditions lowers the viscosity of water from 1.0000 to 0.9846; and potassium chloride in normal solution lowers the viscosity of water from 1.0000 to 0.9872. The atomic volume curve shows that caesium has the largest atomic volume (about 74), rubidium is next in order (about 57), and potassium has the smallest atomic volume (about 47) of this group of metals which produce the negative viscosity in water. It is worth noting that the difference between the values of t\ for caesium chloride and rubidium chloride is much greater than the difference between rubidium chloride and potassium chloride. This is in keeping with the rel- ative atomic volumes of the three elements. The difference between the atomic volumes of rubidium and caesium is much greater than the difference between the atomic volumes of rubidium and potassium. If we extend the above conception to other cations with large atomic volumes, but with much smaller atomic volumes than the alkalies, we shall find that it holds satisfactorily. Take calcium, strontium, and barium, which, of all cations, have the next larger atomic volumes, and compare their chlorides with respect to the values of rj; we have, for normal solutions: •n Calcium chloride 1.1563 Strontium chloride 1.1411 Barium chloride 1.1228 These values are all positive, as we should expect them to be, but their order is exactly the reverse of the atomic volumes, just as we should expect. Chlorides with cations of smaller atomic volumes have values of 17 much larger than the above. This will be seen from the following table, where all the values of »/ refer to normal solutions. In the same table are given the approximate atomic volumes of the cations of the salt. At. vol. rj Magnesium chloride Cupric chloride . . Manganous chloride Nickel chloride . . Cobalt chloride . . 14 8 7 7 7 1.2015 1.2050 1.2089 1.2055 1.2041 It is obvious that for the atomic volumes of the same order of magnitude, the values of 77 are of the same order of magnitude; and, in general, the larger 1 Ztschr. phys. Chem., 6, 35 (1890). 2 1?, the time of flow of water through the viscometer, is taken as unity. 216 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. the atomic volume the smaller the value of y, just as we should anticipate in terms of our hypothesis. The small values of t\ for cadmium and mercury are due to the fact that the salts of these metals are only slightly dissociated. Thallium seems at first glance to present an exception, but it is to be re- membered that it is the thallous salt which produces the negative viscosity. The atomic volume curve deals with thallium in the thallic condition, and, consequently, thallium can not at present be considered as a test of our hypothesis. CONCLUSIONS. (1) We have measured the conductivities of various concentrations of cop- per chloride in water, methyl alcohol, ethyl alcohol, and in binary mixtures of these solvents. We have also measured the conductivities of various con- centrations of potassium sulphocyanate in water, methyl alcohol, ethyl alcohol, acetone, and in binary mixtures of these solvents. (2) Further, we have measured the fluidities of the above-named solvents and mixtures of these with one another; also the fluidities of solutions of potassium sulphocyanate in these solvents. (3) A minimum in conductivity was observed in certain of the mixtures of the solvents. The hypothesis of Jones and Lindsay as substantiated by the work of Jones and Murray has been discussed. (4) It has been shown that in nearly all cases where actual minima do not occur, there is nevertheless a decided dropping of the conductivity curves below the values as calculated from the rule of averages; and that this drop is most pronounced in the 25 per cent and 50 per cent mixtures, which are the points at which the actual minima usually occur. It has also been shown that these cases of dropping of the values below the values calculated from the rule of averages constitute cases of virtual minima, and we have extended the hypothesis of Jones and Lindsay to such cases, and have shown that the hypothesis when applied to the problem of conductivity in mixed solvents is perfectly general. (5) A minimum in the fluidity curves of the above solvents has been ob- served, and the cause of this minimum has been explained as follows : It has been shown that viscosity and fluidity are largely frictional phenomena; and that since when two associated liquids are mixed they mutually decrease the association of one another, they thus increase the number of molecules pres- ent, and, consequently, increase the total frictional surface, thus causing an increase in viscosity (or decrease in fluidity). Further, it has been shown that the point of maximum viscosity is the point where the effect of the sol- vents upon each other is greatest; and, consequently, is the point at which we have the greatest number of simple molecules present. This has been DISCUSSION OF RESULTS. 217 shown to be an additional proof of the hypothesis of Jones and Lindsay to account for the minima in conductivity. (6) A maximum in conductivity has been observed similar to, and in the same solvents as the maxima in conductivity found by Jones and Bingham. The explanation offered by them to account for these maxima has been dis- cussed, and it has been shown that their explanation only partly accounts for the facts. We have offered an additional explanation which is as follows : We have seen that at these maximum points of conductivity, the fluidity of the mixed solvent is also a maximum. We have shown that this maximum fluidity is due primarily to an increase in the size of the molecules of the sol- vents. We have shown further that this enlarging of the molecular spheres can not be due to increased association of either of the pure solvents, and must be due to a molecular aggregation of the solvents with one another. The conditions for such a molecular aggregation are probably most favorable in the particular mixtures in which the maximum fluidities occur. Since the • fluidity of the solvent is increased, the velocity of the ions is increased; and, consequently, we conclude that the maxima in conductivity are dependent upon two factors — the change in the size of the ionic spheres, and the change in the fluidity of the solvent. (7) We have shown that the dissociating action of any given solvent towards electrolytes is not dependent solely upon the physical properties of the solvent, but is also dependent upon the nature of the dissolved salt. That this is true is seen from the fact that although ternary electrolytes, in general, have a higher molecular conductivity when dissolved in water than binary electrolytes, yet when dissolved in methyl alcohol, or ethyl alcohol, or acetone, the conditions are exactly reversed; and the binary electrolytes show the greater molecular conductivities in these solvents. This is probably due to the breaking down of ternary electrolytes into their simplest ions in some solvents, while in other solvents they yield only two ions; whereas the binary electrolytes are dissociated in the same manner in all solvents. (8) We have observed that the temperature coefficients of conductivity in the 25 per cent mixtures of the organic solvents with water are a maximum. This is probably due to the formation of more complex solvates between the dissolved substance and solvents in these particular mixtures than in any other mixtures of these solvents. These are the mixtures in which we have the simplest solvent molecules present, and, consequently, the most favorable conditions for the formation of solvates. (9) We have observed a much greater molecular conductivity of potassium sulphocyanate in solutions in acetone than in aqueous solutions. We have shown that this might be due to either of two causes: (1) A greater degree of dissociation in acetone than in water. This view is untenable on account of the small association and dielectric constant of acetone as compared with that 218 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. of water. (2) A much greater velocity of the ions in acetone than in water. This has been demonstrated to be correct, since the fluidity of acetone is much greater than that of water, and especially because the solvates formed in water are much more complex than those in acetone, as is shown by the great difference in the temperature coefficients of conductivity. These are about ten times as great in aqueous solutions as they are in acetone solutions. (10) A negative coefficient of viscosity has been found for potassium sulpho- cyanate in aqueous solution, and we have offered the following explanation to account for it, after having called attention to the fact that all previously offered explanations are inadequate. From the experimental data recorded by Wagner we find that caesium, rubidium, and potassium are practically the only ions that produce the nega- tive viscosity in aqueous solution. We have further pointed out that all potassium salts do not give a negative viscosity, and that this is due to the fact that viscosity is an additive function of both the ions involved ; and that in some cases the action of the anion, tending to produce a positive viscosity, is sufficient to overcome the action of the potassium ion which tends to produce a negative viscosity in aqueous solution. We have pointed out the fact that the total frictional surface in a liquid would be diminished by the introduction into that liquid of a substance which gives ions with atomic volumes much larger than the molecular volumes of the molecules of the liquid itself; and we find that such is probably the case when caesium, rubidium, and potassium salts are dissolved in water. We have tested our suggestion by reference to the atomic volume curve of the elements as drawn by Lothar Meyer, which, as is well known, shows that potassium, rubidium, and caesium have by far the largest atomic volumes of any of the elements. We have still further tested our hypothesis by showing that the amount of negative viscosity produced by the chlorides of these three elements is in the same order as their relative atomic volumes. GENERAL SUMMARY AND CONCLUSIONS. WE have now recorded the details in connection with seven investigations dealing with the condition of electrolytes in certain non-aqueous solvents, and in mixtures of these solvents with one another. At the conclusion of each investigation a brief summary of the facts and relations established by that investigation has been made. It now seems desirable at the close of this monograph to give a general summary of the work done in this field, and the conclusions which can be drawn from it. The work of Lindsay included the following solvents : Water, methyl alco- hol, ethyl alcohol, and propyl alcohol, and mixtures of these with one another. The electrolytes which he dissolved in these solvents are potassium iodide, ammonium bromide, strontium iodide, cadmium iodide, lithium nitrate, and ferric chloride. A minimum in the molecular conductivity was found for all the salts studied in mixtures of methyl alcohol and water, with the exception of cadmium iodide. A minimum was also found in the mixtures of ethyl alcohol and water especially at 0°, the minimum disappearing at 25°. Mixtures of methyl alcohol with ethyl alcohol do not show the minimum in conductivity, but in the 50 per cent mixture of these solvents the molecular conductivity of dis- solved electrolytes is less than the mean of the conductivities in the separate solvents. An explanation that was offered to account for the conductivity minimum in the mixed solvents, is that in these associated solvents each solvent diminishes the association of the other. Since dissociating power is a function of the association of the solvent, anything that will diminish the association will diminish its dissociating power. The effect of mixing two dissociating solvents would thus be to diminish the association of both, and, consequently, the dissociating power of each of the solvents. A mixture of two such solvents would, then, dissociate less than either alone, and the con- ductivity of an electrolyte in such a mixture would be less than in the individ- ual solvents — the conductivity curve would pass through a minimum. This explanation would account for the conductivity minimum in the mixed solvents. The fundamental question, however, is this: Is this explanation correct ? We have now considered experimental evidence bearing upon this question. The molecular weights of the alcohol when dissolved in water are, in general, normal, i. e., the molecules of the alcohol are the simplest possible. In pure 219 220 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. alcohols, however, the molecules are quite complex, as is shown by the sur- face-tension method of Ramsay and Shields. The water has thus broken down the complex molecules of alcohol into the simpler molecules. A more direct line of evidence bearing upon this problem has been furnished by the work of Jones and Murray. They worked with mixtures of water, formic acid, and acetic acid each with the other. It will be recognized that these are all associated liquids. The molecular weight of each of these liquids dissolved in each of the other two was determined by the freezing-point method. It was found that the molecular weight was smaller the more dilute the solution, and even in the most concentrated solutions that could be studied, the molecular weight was always much less than the molecular weight of the substance when in the pure homogeneous condition. This showed beyond question that the action of an associated liquid is to diminish the association of another associated liquid. The above explanation to account for the conductivity minimum in the mixed solvents is, then, undoubtedly an important factor. We shall, however, see that this is only one factor in conditioning the existence of this minimum. The investigation by Carroll included the same solvents that had been used by Lindsay, i. e., water, methyl alcohol, ethyl alcohol, and binary mixtures of these solvents; and in addition acetic acid was also used. The electro- lytes employed by Carroll are cadmium iodide, sodium iodide, calcium nitrate, hydrochloric acid, and sodium acetate, in mixtures of acetic acid and water. The minimum in the molecular conductivity was found for cadmium iodide, sodium iodide, and hydrochloric acid, in mixtures of methyl alcohol and water. The dissociation of sodium iodide, potassium iodide, and potassium bromide in a 50 per cent mixture of methyl alcohol and water was determined directly, and was found to be apparently slightly greater than in water alone at the same dilution. The question as to the cause of the minimum in conductivity was then taken up. It was shown that there was a parallelism between the conductivity minimum and the minimum in the fluidity of the solvent. It was further shown that both minima are more pronounced at lower temperatures, and that both occur at approximately the same points. Again, the effect of rise in temperature is the same upon both minima, i. e., a shift towards the mixture containing a greater per cent of alcohol. From a quantitative study of these two classes of phenomena, the con- clusion was drawn that the decrease in conductivity in electrolytes when dissolved in binary mixtures of various alcohols and water, which is frequently accompanied by a well-defined minimum in conductivity, is due largely to the diminution in the fluidity of the solvent which takes place when the two solvents are mixed. This diminishes the velocity with which the ions move, and, consequently, diminishes the conductivity. GENERAL SUMMARY AND CONCLUSIONS. 221 A quantitative comparison is then made of the conductivity of the different solvents with their viscosities. In order that this comparison can be made for different solvents, we must deal with " comparable equivalent solutions," i. e., solutions containing the same number of gram-molecules of electrolyte in the same number of gram-molecules of the different solvents. The result is to show that conductivities of "comparable equivalent solu- tions" of binary electrolytes in such solvents as members of the methyl alcohol series, acetone, etc., are inversely proportional to the coefficient of viscosity of the solvent in question, and directly proportional to the association factor of the solvent, or to the amount of the dissociation of the substance dissolved in that solvent. This is essentially the same as to say that the hypothesis of Dutoit and Aston holds quantitatively for certain electrolytes in such solvents as water, methyl alcohol, and ethyl alcohol. While the larger part of the work of Bassett had to do with the question of the relative velocities of the ions of silver nitrate in non-aqueous solvents, and in mixtures of these solvents with water and with one another, yet he took up the problem of the conductivity of silver nitrate in water, in methyl alcohol, in ethyl alcohol, and in binary mixtures of these solvents with one another. It was shown that silver nitrate in mixtures of methyl alcohol and water gives a minimum in conductivity at both 0° and 25°. Silver nitrate, however, in mixtures of ethyl alcohol and water does not show the minimum at either 0° or 25°. When the conductivities at 0° are compared with those of 25°, we see that the influence of one solvent on the other is greater the lower the temperature. This is exactly what would be expected in terms of the suggestion put for- ward in the work of Lindsay. As the temperature is raised the association of associated solvents becomes less and less ; consequently, each solvent will diminish the association of the other less and less as the temperature becomes higher and higher. The investigation carried out by Bingham included the solvents used in the earlier work — water, methyl and ethyl alcohols, and in addition acetone was employed. The conductivities of lithium nitrate, potassium iodide, and cal- cium nitrate in these solvents, and in binary mixtures of one with another, were measured. In this investigation a fairly large number of viscosity measurements were also made. These included not simply the measurement of the viscosities of the pure solvents and of the mixtures of these with one another, but also the viscosities of solutions of calcium nitrate in the different solvents and in the mixtures of these with one another. The temperature coefficients of conductivity and of fluidity in mixtures of acetone with the other solvents were also compared. 222 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. The conductivities in mixtures of acetone with water show the minimum previously observed in a number of cases. This minimum is closely connected with the minimum in fluidity observed in these mixtures. The conductivities of potassium iodide in mixtures of methyl or ethyl alcohol with acetone, obey the rule of averages, and the conductivity curves are nearly straight lines at all the dilutions. In mixtures of acetone with the alcohols the fluidity curve is nearly a straight line. One of the most important facts brought out in this investigation, and one that had not been observed in any of our previous work, is that lithium nitrate and calcium nitrate give a very pronounced maximum in conductivity in mix- tures of acetone with methyl alcohol or ethyl alcohol. It was pointed out that this must be due either to an increase in the dissociation, increasing the number of ions present, or to a diminution in the size of the ionic spheres, causing the ions present to move more rapidly. It was shown to be possible to eliminate one of these. The fluidities of mixtures of acetone with the alcohols were shown to obey the rule of averages, which would indicate that there is no increase in the mo- lecular aggregation when the alcohols and acetone are mixed. In terms of the well-established hypothesis of Dutoit and Aston, such a mixture would not dis- sociate to a greater extent than the constituent solvents. Further, direct measurements of dissociation at extreme dilutions have failed to show any great difference between the dissociating power of the mixtures and that of the pure solvents. Further, the maximum moves from the 25 per cent mixture at large concentration to the 75 per cent mixture in the more dilute solutions. This would not be expected if the dissociating power is greatest in a certain mixture. We have thus eliminated increase in dissociation as being the cause of the maximum and are compelled to accept the other alternative, that the maxi- mum in conductivity is due to a change in the dimensions of the atmospheres about the ions. In dealing with conductivity in single solvents and in mixtures of these with one another, we must take into account not only the effect of each con- stituent of the mixture on the association of all the other constituents, and the viscosity of the individual solvent or solvents, but also the sizes of the spheres of the solvents around the ions, and any changes in the sizes of these spheres in different mixtures of the solvents. These ionic spheres, or solvates, have been shown by Jones to exist generally in solutions, and the ions must drag these spheres with them in their move- ments under the action of the current. Any change in the size of these spheres would produce a change in the effect- ive mass of the ion, and, consequently, a change in the velocity with which it would move. GENERAL SUMMARY AND CONCLUSIONS. 223 It was also shown in this investigation that the hyperbola, and not the straight line, is the normal curve of viscosity. While the work of Rouiller had to do more especially with the velocities of ions in mixed solvents, yet he studied the conductivities of silver nitrate in acetone, in mixtures of acetone with water, in mixtures of methyl alcohol with ethyl alcohol, in methyl alcohol and acetone, and in ethyl alcohol and ace- tone. Silver nitrate, in mixtures of methyl alcohol and acetone, and ethyl alcohol and acetone, shows a pronounced maximum in conductivity strictly analogous to calcium nitrate, which had been studied in these solvents by Bingham. This maximum is pronounced at both 0° and 25°, appearing in the 25 per cent acetone mixture in the more concentrated solutions, and shifting with increase in dilution through the 50 per cent to the 75 per cent mixture. The work of Rouiller on the migration velocity of ions in mixtures of these solvents would indicate that the explanation of the conductivity maximum offered by Jones and Bingham is correct — there is a change in the atmos- phere of the solvent about the ions. The investigation of McMaster included the same solvents that had been used by Bingham — water, methyl alcohol, ethyl alcohol, and acetone, and binary mixtures of these solvents. The electrolytes used were lithium bromide and cobalt chloride. The con- ductivities of a large number of solutions of these substances in the above solvents and in binary mixtures of these solvents were measured. The fluidities of water, methyl alcohol, ethyl alcohol, and acetone, and binary mixtures of these solvents, were also measured. The conductivities in mixtures of the alcohols with water, and in the mixtures of acetone with water, show a well-marked minimum. This minimum in conductivity was more pronounced at the lower temperature, and was intimately connected with the minimum in fluidity observed in the above mixtures. The conductivity curves in mixtures of methyl and ethyl alcohols are nearly straight lines, obeying the law of averages, as we should expect. Lithium bromide dissolved in mixtures of methyl or ethyl alcohol with acetone, shows a pronounced maximum in conductivity. The conductivity maximum was also given by cobalt chloride in mixtures of acetone with ethyl alcohol. The fluidities of lithium bromide in mixtures of acetone with the alcohols were found to be what would be expected from the rule of averages — the fluidity curves being nearly straight lines. The same was found for the pure solvents. This would indicate that the alcohols and acetone do not form more complex aggregates when mixed than when alone. We are therefore forced to the same conclusion as that reached by Jones and Bingham, i. e., that the size of the ionic spheres is an important factor in determining conductivity, and that changes in the sizes of these spheres in 224 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. different mixtures of the solvents are the prime factor in conditioning the maximum in the conductivity curve. Some interesting results were obtained in this investigation bearing on the temperature coefficients of conductivity. Jones l has recently pointed out the bearing of hydrates on the temperature coefficients of conductivity. Jones and West showed that with rise in temper- ature there is a decrease in the dissociation, and that increase in conductivity with rise in temperature was due primarily to an increase in the velocities with which the ions move. As the temperature is raised the hydrates in combination with the ion become simpler and simpler, and, therefore, the effective mass of the ion decreases with rise in temperature. The ion being smaller and the solvent less viscous, it will move faster the higher the tempera- ture. Jones has also pointed out in terms of his hydrate theory that the greater the dilution the greater should be the temperature coefficient of conductivity. The more dilute the solution, the more complex the hydrate; the more complex the hydrate in combination with any given ion, the greater the change in the complexity of the hydrate with rise in temperature. The temperature coefficient of conductivity should, therefore, be greater in the more dilute solution, and such is the fact. The work of McMaster showed that the tem- perature coefficients of conductivity and of fluidity for lithium bromide are of the same order of magnitude, the temperature coefficients for this substance in the mixed solvents all being positive. In certain of the mixtures of acetone with the alcohols, cobalt chloride showed negative temperature coefficients of conductivity. This was true in the 75 per cent mixture of acetone with methyl alcohol, and also in the 50 per cent and 75 per cent mixtures of acetone with ethyl alcohol. Negative tem- perature coefficients had previously been found in a few cases at low tem- peratures, but in this work negative temperature coefficients were found at ordinary temperatures. What is the meaning of negative temperature coefficients of conductivity ? With rise in temperature the solvent becomes less viscous, and this would increase the velocity of the ions. With rise in temperature, however, the association of the solvent becomes less and, consequently, its dissociating power is diminished ; which means that there would be a smaller number of ions present the higher the temperature. These two influences act counter to one another — the former increasing the conductivity and the latter diminish- ing it. When we have negative temperature coefficients of conductivity, it means that the latter influence more than overcomes the former — the decrease in the number of the ions with rise in temperature more than counterbalancing 1 Carnegie Institution of Washington, Publication No. 60. GENERAL SUMMARY AND CONCLUSIONS. 225 the effect on conductivity of the increased velocity with which the ions move. The result is a decrease in conductivity with rise in temperature. This explanation accounts entirely satisfactorily for the facts in the above case. The alcohols and acetone are at ordinary temperatures highly asso- ciated liquids. The effect of rise in temperature is to diminish their asso- ciation and, consequently, their dissociating power. It is interesting to note in this connection that a solution was found that had a zero temperature coefficient of conductivity. It was a solution of cobalt chloride in a 75 per cent mixture of acetone with methyl alcohol, the solution having a value of v = 200. The last investigation upon this problem — that of Veazey — has brought out a number of points of interest. It consisted experimentally in measuring the conductivities and viscosities of solutions of copper chloride and potassium sulphocyanate in water, methyl alcohol, ethyl alcohol, and acetone, and in binary mixtures of these solvents. The minimum in conductivity observed in the earlier investigations was found for the above substances, and was shown to be a much more general phenomenon than was supposed from any of the earlier work. It has been pointed out that the minima in fluidity, or maxima in viscosity, correspond to the minima in conductivity of the solutions of electrolytes in these solvents. An explanation of why solutions of methyl alcohol or ethyl alcohol and water are more viscous than either of the pure solvents alone was offered. These liquids are all highly associated. When one associated liquid is mixed with another associated liquid, each diminishes the association of the other. This means that the large molecules of each solvent are torn down into a larger number of small molecules. This would increase the frictional surfaces of the molecules that would be exposed to one another, just as small shot would exert greater friction, in moving over one another in a manner analogous to that followed by the molecules in producing viscosity, than larger shot. The result of each associated liquid diminishing the association of the other, would thus be to increase the viscosity of the mixture over that of either pure solvent. This explains also why the conductivity curves for different dilutions of the same substance generally approach one another as they approach the minimum. Those mixtures of the solvents in which the conductivity minima occur are the least associated, and, therefore, have the least dissociating power. It is obvious that such mixtures would produce the least increase in dissociation with increase in dilution, and, consequently, the conductivity curves for the different dilutions would approach one another as they approach the minima. Fact and theory are here in perfect accord. The minima in conductivity observed in the earlier work were also found in a number of cases studied in this investigation. It was shown that this is 226 CONDUCTIVITY AND VISCOSITY IN MIXED SOLVENTS. due in part to a change in the size of the ionic spheres, but that another factor also comes into play. It was found that the maxima in conductivity correspond to the maxima in fluidity of the mixed solvents. These maxima in fluidity are probably due, as has been shown, to an increase in the size of the molecules of the solvent. From the work of Jones and Murray this can not be caused by an increased association of the molecules of the several solvents, since the action of each solvent on the other is to diminish its association. This increase in size of the molecular aggregates of the solvents must be due to the combination of one solvent with the other, forming a molecular complex. This would diminish the viscosity, or increase the fluidity, and, consequently, increase the velocity with which the ions would move through the solvent. This factor must also be taken into account in explaining the conductivity maxima observed in a number of these investigations. A reason is suggested to account for the fact that different solvents show different relative dissociating powers in the case of binary and ternary electro- lytes. Weakly dissociated solvents are probably capable of breaking down ternary electrolytes into only two ions; while strongly dissociating solvents can break these molecules down into three ions. This would also explain why ternary electrolytes, showing greater conductivity in water than binary electrolytes, often show much smaller conductivity in the alcohols and in acetone than binary electrolytes. The temperature coefficients of conductivity are a maximum in the 25 and 50 per cent mixtures of the organic solvents with water. These are about the mixtures in which the solvents show least association — the molecules would be in the simplest condition and therefore most favorable for chemical action. The solvents probably combine with the dissolved substance to the greatest extent in such mixtures,, and the effect of rise in temperature, breaking down these complexes, would therefore be a maximum. The conductivity of solutions of potassium sulphocyanate in substances like acetone is much greater than in water. This was proved to be due to the greater fluidity of the acetone. Potassium sulphocyanate dissolved in water lowers the viscosity of water, i. e., the solution has a smaller viscosity than water itself. On examining the literature it was found that salts of potas- sium, rubidium, and caesium are practically the only known electrolytes which lower the viscosity of water when dissolved in it. Certain salts of potassium, however, do not lower the viscosity of water, just as might be expected, since viscosity is an additive property of both the ions present in the solution. The anions tend to increase the viscosity of the solution, while certain cations, viz. potassium, rubidium, and caesium, have a tendency to diminish the viscosity of a solution. If the effect of the negative ion more than overcomes that of the positive ion, potassium, rubidium, and caesium, then the solution is GENERAL SUMMARY AND CONCLUSIONS. 227 more viscous than water. If it does not, then the solution is less viscous than pure water. The explanation of the diminution in viscosity produced by the above- named cations is comparatively simple in the light of the conception of vis- cosity proposed earlier in this investigation. If the atomic volume of the ions introduced was much larger than the molecular volumes of the solvent molecules, the effect would be to diminish the frictional surfaces that would come in contact with one another in the solution, and, consequently, the friction of the movement of the molecules over one another would be di- minished. The question, then, is: Are the atomic volumes of potassium, rubidium, and caesium very large? And are they much larger than the atomic volumes of other elementary cations? If we turn to the well-known atomic volume curve, we see that potassium, rubidium, and csesium occupy the maximum of the curve, and have much larger atomic volumes than any other known elements. Even the atomic volume of potassium, which is smaller than that of rubidium and caesium, is much larger than that of any other known element except rubidium and csesium. If we test this relation quantitatively, the result is very satisfactory. By comparing the viscosities of solutions of the same concentration of potas- sium chloride, rubidium chloride, and csesium chloride, we find that, while all these viscosities are less than the viscosity of pure water, the viscosity of the solution of rubidium chloride is less than that of potassium chloride, and the viscosity of the solution of caesium chloride is less than that of rubidium chloride. More work will be done in the physical chemical laboratory of the Johns Hopkins University to test whether these relations are perfectly general. In conclusion it gives me great pleasure to express my hearty thanks to my seven cooperators, who, the one after another, have taken up and studied uninterruptedly during the past six years the various problems that have arisen in connection with this line of investigation. INDEX. PAGE Acids 11 Alcohols 7 Ammonia 4 Ammonium bromide 30 comparison of the molecular con- ductivities of 31 molecular con- ductivity of 31 temperature co- efficients of conductivity of 31 Apparatus used by Bassett 75 used by Bingham 81 used by Carroll 43 used by Lindsay 24 used by Rouiller 116 used by Veazey 170 Arrhenius, on conductivity as affected by the presence of a non-electrolyte 14 on the effect of non-elec- trolytes on the viscos- ity of water 17 work in mixtures of the alcohols and water. . . 13 Aston and Dutoit, on the relation be- tween dissociating power and polymerization 2 Bassett, summary of his work 221 work of 75 Bingham, summary of his work 221 work of 81 Blanchard, on viscosity 22 Bousfield and Lowry, effect of tem- perature on viscosity of water. . . 18 Bouty, on the conductivity of nitrates in nitric acid 4 Bredig, on viscosity 20 PAGE Briihl, on the relation between disso- ciating power and unsaturation . . 1 Cadmium iodide 34, 44 comparison of con- ductivities of 47 comparison of the molecular con- ductivities of 34 conductivity of 45 Cady, on the conductivity of solutions in liquid ammonia 4 Calcium nitrate 49, 97, 98 comparison of conduc- tivities 50 comparison of the con- ductivities of ... 98, 99 comparison of the temperature co- efficients of con- ductivity of .. .99, 100 conductivity of 49, 97 Calvert, on the dielectric constant of a mixture of hydrogen dioxide and water 13 Carrara, conductivity in acetone 10 conductivity in isopropyl al- cohol 9 conductivity in methyl alco- hol 7 on conductivity in propyl and amyl alcohols 9 work in mixtures of the alco- hols and water 13 Carroll, summary of his work 220 work of 43 Cattaneo, conductivity in acetone 10 conductivity in ethereal so- lutions 10 conductivity in ethyl alco- hol 8 conductivity in glycerol. . 12 229 230 INDEX. PAGE Cattaneo — continued. conductivity in methyl al- cohol 8 Centnerszwer, on the dissociating power of liquid hydrocyanic acid 3 on the non-dissociating power of liquid cyanogen 12 work with liquid cyano- gen as a solvent . . 6 Ciamician, on the relation between dissociating power and chemical properties 2 Cobalt chloride 139 comparison of the con- ductivities of, 143, 144 comparison of the tem- perature coeffi- cients of conduc- tivity of 144, 145 conductivity of 141, 142, 143 nitrate 174 Coefficients, negative viscosity 213 Cohen, observed a minimum in con- ductivity 28 on change in fluidity 17 on conductivity in mixtures of alcohol and water 15 work in mixtures of ethyl alco- hol and water 13 Comparison of molecular conductivi- ties in water and mixtures of the alcohols. 62 Conclusions and summary 219 Conductivity and fluidity, comparison of temperature coefficients of, 66, 203 and viscosity 16 of certain salts 139 measurements 26 measurements made by Bassett 76 measurements made by Bingham 83 measurements made by Carroll 44 measurements made by McMaster 126 measurements made by Rouiller 116,117 measurements made by Veazey 170 PAGE Copper chloride 174 comparison of the con- ductivities of. ... 177 comparison of the temperature co- efficients of con- ductivity of 178 conductivity of 175 Discussion of results 207 Dissociation in fifty per cent methyl alcohol 53 in mixture of water and methyl alcohol greater than in either pure solvent 54 of salts in fifty per cent methyl alcohol 55 Donnan, on the dissociating power of liquids 2 Dunstan, on viscosity 21 Dutoit and Aston, conductivity in ke- tones 10 on conductivity in benzene chlo- ride, ethyl bro- mide and amyl acetate 12 on conductivity in propionitrile . . 11 on the relation be- tween disso- ciating power and polymeri- zation 2 and Friderich, conductivity in keton 10 and Friderich, conductivity in nitriles 12 on conductivity in acetophenone . 12 and Friderich, on the relation between conductivity and viscosity 17 Electrostriction theory of Euler to account for "negative viscosity" 21 Ether 10 Ethyl alcohol 8, 25 Euler, on negative viscosity 21 on the relation between fluidity and viscosity 17 on viscosity 20 Ferric chloride 38 changes in the molec- ular conductivity of. 40 INDEX. 231 PAGE Ferric chloride — continued. conductivity of, in water 39 Fitzpatrick, conductivity in methyl alcohol 7 Fluidities, comparison of 104 of potassium sulphocya- nate, comparison of. . 196 Fluidity and conductivity 59, 159 and conductivity, compari- son of the temperature coefficients of 106, 203 and conductivity, compari- son of variation in. ... 64, 65 comparison of the tempera- ture coefficients of 104 of potassium sulphocyanate, 194, 195 of potassium sulphocyanate, comparison of the tem- perature coefficients of . . 196 Franklin and Farmer, on the dissociat- ing power of nitro- gen peroxide 6 and Kraus, a maximum in the conductivity in liquid ammonia. ... 18 and Kraus, on the conduc- tivity of solutions in liquid ammonia. 4 Garelli and Bassani, work with the halides of arsenic and antimony . . 5 Goodwin and Thomson, on the con- ductivity of solutions in liquid ammonia 4 Graham on viscosity 22 Grossman, conductivity multiplied by viscosity gave a constant 17 Grotrian, on conductivity and viscos- ity 16 Hantzsch, work in mixed solvents. . . 16 Hartwig, conductivity in methyl alco- hol 7 on conductivity in amyl alco- hols 9 on the conductivity in ethyl alcohol 8 Hauser, on the effect of pressure on the fluidity of water 17 Higher alcohols 9 Historical sketch of work in non-aque- ous solvents 3 Holland, conductivity in methyl alco- hol 7 on conductivity in mixed sol- vents. . 15 PAGE Hydrocarbons 7 Hydrochloric acid 50 conductivity of. ... 51 dissociation of, in mixtures of methyl alcohol and water 56 Hydrocyanic acid 3 Hydrogen dioxide and water 13 Hyperbola the normal curve for vis- cosities 108 Increase in conductivity with rise in temperature, due mainly to an increase in the velocity of the ions 165 Inorganic solvents 3 Jones and Bassett, evidence for hy- drate theory 23 and Douglas, on the temperature coefficients of dissociation . . 9 and Getman, evidence for hy- drate theory 23 and Murray, the effect of one associated liquid on the as- sociation of another asso- ciated liquid 41 and Murray, summary of their work 220 and Uhler, evidence for hydrate theory 23 and West, on the temperature coefficients of dissociation 9 applies the boiling-point method to measure dissociation in methyl alcohol 8 Barnes and Hyde, on the disso- ciating power of hydrogen dioxide 13 hydrate theory 23 on dissociating as measured by the boiling-point method. . 9 on the conductivity of sulphuric acid in acetic acid 11 the dissociating power of water measured by freezing-point lowering 3 Kablukoff, conductivity in ethereal solutions 10 conductivity in methyl alcohol conductivity in mixtures of ethyl alcohol and water 14 on conductivity in ethyl alcohol . . 9 232 INDEX. PAGE Kablukoff — continued. on conductivity in iso- butyl and isoamyl alcohols 9 on the conductivity in hydrocarbons 7 work in mixtures of the alcohols and water. . 13 Kahlenberg and Lincoln, conductivity in acetone 10 and Lincoln, conductivity in methyl alcohol .... 8 and Lincoln, on conduc- tivity in ethyl alcohol 9 and Lincoln, on conduc- tivity in organic sol- vents 12 and Lincoln, on the con- ductivity in hydro- carbons 7 and Lincoln, work with the halides of arsenic and antimony 6 Kanowalow, on the relation between dissociating power and chemical action 2 Kawalki, on diffusion in ethyl alcohol 8 Kerler, conductivity in methyl alcohol 7 work in mixtures of the alco- hols and water 13 Ketones 10 Kohlrausch and Deguisne, formula for the influence of tem- perature on conduc- - tivity 18 on the dissociating power of water determined by its conductivity . . 3 sphere of solvent about the ions 18 Lenz, on the conductivities in mixtures of the alcohols and water. 14 work in mixtures of the alcohols and water 13 Lindsay, summary of his work 219 work of 24 Lithium bromide 129 conductivity of, 129, 130, 131 comparison of the con- ductivities of, 131. 132, 155 comparison of the tem- perature coefficients of conductivities of, 132, 133 PAGE Lithium bromide — continued. comparison of the temperature co- efficients of flu- idities of 155 fluidities 154 Lithium nitrate 35 comparison of the con- ductivities of, 85, 86 comparison of the molecular conduc- tivities in mix- tures of water and the alcohols, 62 comparison of the molecular conduc- tivities of 37 comparison of the temperature co- efficients of con- ductivity of 87,88 conductivity of . . . .84, 111 molecular conductiv- ity of 37 temperature coeffi- cients of conduc- tivity of 37 Loomis, the dissociating power of water as measured by freezing- point lowering 3 Maximum in conductivity due pri- marily to a change in the dimen- sions of the ionic spheres Ill McMaster, summary of his work 223 work of 126 Methyl alcohol 25 Minimum in conductivity a general phenomenon 40 in conductivity, cause of, 58, 67 in conductivity, explana- tion of 41 in the molecular conduc- tivity of potassium iodide 29 Mixed solvents 13 used by Rouiller 116 Mixtures of water and the alcohols ... 13 Negative temperature coefficients of conductivity, 148, 149, 167 viscosity coefficients 213 Nernst, on the relation between disso- ciating power and dielec- tric constant 1 the dissociating power of water as measured by solubility. . 3 INDEX. 233 PAGE Nitric acid 4 Nitriles and cyanogen 11 Non aqueous solvents, historical sketch of work in 3 Noyes and Coolidge, on the tempera- ture coefficients of disso- ciation 9 the dissociating power of water as measured by solubility 3 Oddo, work in inorganic solvents. ... 5 Organic solvents 7, 12 Paschkow, on conductivity in ethyl alcohol 9 Paschow, conductivity in methyl alco- hol 7 Poisseuille, on viscosity 22 Potassium and sodium iodides in mixed soh'ents, ratio of dissociation to as- sociation of the sol- vent 57 iodide 26,53,89 iodide, comparison of the conductivities of. ... 92 iodide, comparison of the molecular conduc- tivities in water and mixtures of the alco- hols 61 iodide, comparison of the molecular conduc- tivities of 28 iodide, comparison of the temperature coeffi- cients of conductivity of 92,93 iodide, conductivity and dissociation of, in va- rious solvents 54 iodide, conductivity of. .90, 91 iodide, molecular conduc- tivity of 26 iodide, temperature coeffi- cients of conductivity of 27 sulphocyanate 181 sulphocyanate, compari- son of fluidities of. . 196 sulphocyanate, compari- son of the conduc- tivities of 183,184 sulphocyanate, compari- son of the tempera- ture coefficients of conductivity of. .184, 185 PAGE Potassium — continued. sulphocyanate, compari- son of the tempera- ture coefficients of fluidity of 196 sulphocyanate, conduc- tivity of 182, 183 sulphocyanate, fluidity of, 194, 195 Propyl alcohol 25 Pyridine 12 Ramsay and Aston, on the association factors of various solvents 6 and Shields, on the associa- tion factors of vari- ous solvents 6 Results, discussion of 207 obtained by Bingham, dis- cussion of 107 obtained by Carroll, discus- sion of 66 obtained by McMaster, dis- cussion of 159 Rontgen, on change in fluidity 17 Roth, conductivity in mixtures of ethyl alcohol and water 16 Rouiller, summary of his work 223 work of 115 Rudorf, conductivity in mixed sol- vents 16 Schall, on conductivity in ethyl alco- hol. 9 on conductivity in methyl al- cohol 8 on the conductivity of pic- ric acid in aqueous alco- hol 15 work in mixtures of the alco- hols and water 13 Schlamp, on conductivity in propyl alcohol 9 Schlundt, on the dielectric constant of liquid hydrocyanic acid 3 Silver nitrate, comparison of the mo- lecular conductivi- ties of 77 molecular conductivity of 76, 118,119 temperature coefficients of conductivity of, 77, 120 Skilling, on the non-conductivity of solutions in liquid hydrogen sul- phide 6 234 INDEX. PAGE Sodium acetate, conductivity of 52 in mixtures of acetic acid and water, 52 Sodium iodide 47, 54 comparison of conduc- tivities 48 conductivity and disso- ciation of, in mixed solvents 55 conductivity of 47 Solutions 26 as prepared by McMaster. . . 128 as prepared by Rouiller. ... 117 method of preparing the .... 44 preparation of, by Bingham, 82 Solvents 25 used by Bingham 83 used by Carroll 43 used by McMaster 127 used by Rouiller 115 used by Veazey 172, 173 Stephan, on the relation between con- ductivity and viscosity, 17 work in mixtures of ethyl alcohol and water 14 work in mixtures of the alco- hols and water 13 Strindberg, on conductivity in mixed solvents 15 Strontium iodide 32 comparison of the molecular con- ductivities in water and mix- tures of the alco- hols 61 comparison of the molecular con- ductivity of. ... 33 molecular conductiv- ity of 32 temperature coeffi- cients of conduc- tivity of 32 St. v. Lasczynski and St. v. Gorski, on the dissociating power of pyri- dine 12 conductivity in ace- tone 10 Sulphur dioxide 4 Summary and conclusions 219 and conclusions from the work of Carroll 73 of Lindsay's work 40 PAGE. Summary — continued. of the facts established .... 202 of the work of Bingham. . . 113 of the work of McMaster. . 168 Temperature coefficients of conductiv- ity due mainly to an increase in the velocity of the ions 165 coefficient of conductiv- ity 164 coefficient of conductiv- ity, zero 149 coefficients of conductiv- ity and fluidity, comparison of 65 coefficients of conductiv- ity, negative 148 Thomson, on the relation between dis- sociating power and dielectric constant 1 Thorpe and Rodger, on viscosity 19 viscosity and hy- drates 22 Tijmstra, conductivity in mixtures of the alcohols and water 16 Tolloczko, work with the halides of arsenic and antimony 5 Trane, hydrates and viscosity 22 Turner, on the dielectric constants of liquids 6 Varenne and Godefroy, on viscosity and hydrates 22 Variation in conductivity 64 in conductivity and fluidity, comparison of 63 in conductivity of potassium iodide in mixtures of ethyl alcohol and water 63 Veazey, summary of his work 225 work of 170 Vicentini, on conductivity in ethyl alcohol 8 Viscosities, hyperbola the normal curve for 108 Viscosity 19 and conductivity 16, 68 and conductivity of certain salts 139 coefficients, negative 213 measurements 103, 104, 193 by Bingham . . 81 by McMaster, 126, 151 by Veazey. ... 171 INDEX. 235 PAGE Vollmer, conductivity in ethyl alco- hol 8 conductivity in methyl alco- hol 7 on conductivity and viscosity in the alcohols 18 Wagnor, on viscosity 20 Wakeman, on conductivity in mix- tures of alcohol and water 15 work in mixtures of ethyl alcohol and water. ... 13 Walden and Centnerszwer, on conduc- tivity of solutions in liquid sulphur dioxide, 4 on conductivity in organic solvents 12 work on inorganic solvents 5 Walker and Hambly, on conductivity in mixtures of ethyl alcohol and water 16 Warburg and Sach, on change in fluid- ity 17 PACK Water 25 dissociating power of 3 Werner, conductivity in ethyl sulphide, 12 on the conductivity of inor- ganic salts in pyridine. . 12 Wiedemann, G., on the relation be- tween conductivity and viscosity, 1 6 Wikander, on viscosity 22 Wildermann, on conductivity in ethyl alcohol 9 Wolf, conductivity in mixed solvents, 16 Zanninovich-Tessarin, on the disso- ciating power of formic acid .... 11 Zelinsky and Krapiwin, conductivity in methyl alcohol 8 and Krapiwin, on conductiv- ity in mixtures of alco- hol and water 15 and Krapiwin, work in mix- tures of methyl alcohol and water 13 Zero temperature coefficient of con- ductivity 149 J II 4 7 MBI. WHOI I.IItltAKV UH Ifllb