GIFT OF Rheology Seminar EHQiNEERING LJBRAEY STUDIES ON SOLUTION IN ITS RELATION TO LIGHT ABSORPTION, CONDUCTIVITY, VISCOSITY, AND HYDROLYSIS A REPORT UPON A NUMBER OF EXPERIMENTAL INVESTIGATIONS CARRIED OUT IN THE LABORATORY OF THE LATE PROFESSOR HARRY C. JONES COMPILED BY PAUL B. DAVIS CARNEGIE INSTITUTION OF WASHINGTON 1918 / r/ STUDIES ON SOLUTION IN ITS RELATION TO LIGHT ABSORPTION, CONDUCTIVITY, VISCOSITY, AND HYDROLYSIS , ^, t-.v A REPORT UPON A NUMBER OF EXPERIMENTAL INVESTIGATIONS CARRIED OUT IN THE LABORATORY OF THE LATE PROFESSOR HARRY C. JONES COMPILED BY PAUL B. DAVIS CARNEGIE INSTITUTION OP WASHINGTON 1918 CARNEGIE INSTITUTION OF WASHINGTON PUBLICATION No. 260 !.!3RAf PRESS OF GIBSON BROTHERS WASHINGTON, D. C. PREFACE. The several chapters comprising this report represent the various lines of investigation pursued under the direction of the late Professor Harry C. Jones during the year 1915-16 and hi the case of the work of Davis and Johnson continued in 1916-17. Although somewhat diverse in nature, they all bear directly or indirectly upon the concep- tions of solution in general and of solvation in particular which have been developed in this laboratory during the past fifteen years. Dr. Hulburt and Dr. Hutchinson have measured the absorption coefficient of solutions of a number of salts in differents solvents for monochromatic radiation. They have calculated from this the molec- ular absorption coefficient for such solutions and have made a careful comparative study of the molecular absorption-concentration curves. The investigation of formamid as a solvent, begun by Davis and Putnam, has been continued by Dr. Davis and Dr. Johnson. In addi- tion to observing the behavior of a series of nitrates and formates in this solvent, they have determined the conductivity and viscosity of solutions of a number of salts of the organic acids and have also studied several representative salts in mixtures of formamid with ethyl alcohol. Dr. Davis has also made some observations on the viscosity of csesium salts hi binary mixtures of glycerol and of formamid with water. Dr. Lloyd and Dr. Pardee have extended the work in absolute ethyl alcohol to include a study of the conductivities of the sodium salts of a number of organic acids and have succeeded in applying the formula of Noyes and Johnston for aqueous solutions to the calculation of disso- ciation in this solvent. Dr. Ordeman has completed his study of the relative dissociating power of free and combined water reported on in part in Publication No. 230 of the Carnegie Institution of Washington. Dr. Connolly has investigated the different chemical activity of free and semi-combined water, using as an illustration the effect of neutral salts in the hydrolysis of acetic anhydride. A preliminary paper on this work is also to be found in Publication No. 230. The results of all these investigations, which have been carried out with aid of generous grants from the Carnegie Institution of Washing- ton, are recorded in this volume. The writer also wishes to thank that Institution for making possible the completion of certain investi- gations left unfinished by the untimely death of Professor Jones, and the Chemical Staff of this University for their courtesy in extending the facilities of the laboratory. PAUL B. DAVIS. THE JOHNS HOPKINS UNIVERSITY, June 1917. i 912121 // . * lAtoiftlafti al* )- CONTENTS. CHAPTER I. THE ABSORPTION COEFFICIENT OF SOLUTION FOR MONO-CHROMATIC RADIATION: PAGE. Introduction 6 Statement of the problem 9 Historical Apparatus 12 Procedure . . 14 Sut Errors and corrections 17 THE ABSORPTION COEFFICIENT OF THE SOLVENTS: Water 18 Methyl alcohol 20 Ethyl alcohol 21 Propyl alcohol 21 Iso-butyl alcohol 21 Iso-amyl alcohol 21 Discussion of results with the solvents 21 THE ABSORPTION COEFFICIENT OF THE SOLUTIONS: Cobalt chloride in water 23 Cobalt chloride in methyl alcohol 26 Cobalt chloride in ethyl alcohol 29 Cobalt chloride in propyl alcohol 31 Cobalt chloride in iso-butyl alcohol 32 Cobalt chloride in iso-amyl alcohol 33 Discussion of results for cobalt chloride 35 Cobalt chloride in methyl alcohol with water 36 Cobalt chloride in ethyl alcohol with water 37 Cobalt chloride in propyl alcohol with water 40 Cobalt nitrate in water 40 Cobalt sulphate in water . . 43 Nickel chloride in water. 47 Hydrated nickel chloride in the alcohols 51 Nickel nitrate in water 54 Nickel sulphate in water 59 Ferric ammonium sulphate in water 62 Chromium chloride in water 64 Chromium nitrate in water 66 Chromium sulphate in water 67 Potassium permanganate in water 67 Conclusion , 69 CHAPTER II. THE CONDUCTIVITY AND VISCOSITY OF CERTAIN ORGANIC AND INORGANIC SALTS IN FORMAMID AND IN MIXTURES OF FORMAMID WITH ETHYL ALCOHOL: Introduction 71 Experimental 72 Preparation of the solvents 72 Preparation of the salts 73 Preparation of the solutions 73 Apparatus 73 Procedure 75 3 4 Contents. Tables: PAGE. Ammonium nitrate in f ormamid 76 Potassium nitrate in f ormamid Sodium nitrate in formamid 77 Calcium nitrate, barium nitrate ......; Strontium nitrate 79 Rubidium formate 79 Ammonium formate 80 Sodium formate 80 Lithium formate 81 Barium formate 81 Strontium formate 82 Sodium benzoate 82 Sodium metabrombenzoate 83 Sodium metamido benzoate 83 Sodium dinitro benzoate /!** ~ 84 Sodium salicylate l v i* : 7 84 Sodium benzene sulpbonate 85 Sodium succinate L *. J?t 85 Tetraethylammonium iodide in formamid with alcohol 86 Rubidium iodide in formamid with alcohol 87 Lithium nitrate in formamid with alcohol 89 Calcium nitrate in formamid with alcohol 90 Discussion of results 91 CHAPTER III. A NOTE ON THE VISCOSITY OF CESIUM SALTS IN GLYCEROL- WATER MIXTURES: Caesium chloride 97 Caesium nitrate 98 CHAPTER IV. A STUDY OP THE ELECTRICAL CONDUCTANCE OP THE SODIUM SALTS OF CERTAIN ORGANIC ACIDS IN ABSOLUTE ETHYL ALCOHOL AT 15, 25, AND 35: Introduction 99 Historical 99 Experimental 103 Reagents 103 Apparatus 105 Procedure 106 Measurements: Explanation of tables 109 Sodium formate and acetate 109 Sodium monochloro, dichloro, trichloro, and phenyl acetates: propionate and ido-propionate; butyrate and oxy isobutyrate 110 Sodium benzoate; ortho amido and p-amido benzoates, m- and p-bromben- zoates; o-, m-, and p-chlorobenzoates Ill Soduim salicylate; m-andp-hydroxybenzoates; acetyl, iodo and sulphosalicyl- ates; o- and meta nitrobenzoates 112 Sodium p-nitro and 2.4 dinitro ben/oates, ortho, meta, and paratoluates and picrate 113 Discussion of results Summary 118 Contents. 5 CHAPTER V. A STUDY OF THE DISSOCIATING POWER OP FREE AND OP COMBINED WATER : PAGE. Introduction 119 Experimental 119 Apparatus 119 Solvents 121 Salts 122 Solutions 123 Procedure 124 Measurements 125 Tables 125-127 Discussion of results 127 CHAPTER VI. THE DIFFERENCE IN CHEMICAL ACTIVITY OF FREE AND SEMI-COMBINED WATER AS ILLUSTRATED BY THE EFFECT OF NEUTRAL SALTS ON THE HYDROLYSIS OF ACETIC ANHYDRIDE: Hydrolysis 131 Hydrotysis of acetic anhydride 131 Hydrolysis of salts 134 Neutral salt action 134 Effect of neutral salts on catalytic activity of acids 135 Effect of neutral salts on hydrolysis by water alone 135 Statement of the problem 137 Experimental: Purification of acetic anhydride 137 Purification of salts 138 Apparatus 138 Solutions 138 Method of procedure 139 Calculations 140 Data 140-143 Discussion of results 143 STUDIES ON SOLUTION IN ITS RELATION TO LIGHT ABSORPTION, CONDUCTIVITY, VISCOSITY, AND HYDROLYSIS A REPORT UPON A NUMBER OF EXPERIMENTAL INVESTIGA- TIONS CARRIED OUT IN THE LABORATORY OF THE LATE PROFESSOR HARRY C. JONES Compiled by PAUL B. DAVIS CHAPTER I. THE ABSORPTION COEFFICIENT OF SOLUTION FOR MONOCHROMATIC RADIATION. BY E. O. HULBURT AND J. F. HUTCHINSON. INTRODUCTION. STATEMENT OF THE PROBLEM. Experiments have shown that in the case of certain solutions the absorption of monochromatic radiation may be represented by the formula 7 = / XlO ' (1) where 7 is the original intensity of the radiation, I is the intensity of the radiation after passing through a layer of solution of thickness t millimeters, and a is a quantity, called the absorption coefficient of the solution for the specified frequency of radiation. Experiments have also shown that different values of a are obtained if there is any change in: (a) the nature of the solvent or of the dissolved substance. (6) the concentration of the solution. (c) the temperature. (d) the wave-length of the radiation, etc. To solve the problem of light-absorption in solutions it is necessary to determine the explicit form of the relation between the absorption coefficient a and the quantities of which it is a function. At present our knowledge is far too meager to indicate more than a qualitative idea of the nature of this relation. In the present investigation a has been measured in those regions of the spectrum where the pure solvents possess appreciable absorption. It is assumed that the total absorption of the solution is the sum of two parts, the first being the absorption due to the presence of the salt, the second being the absorption due to the pure solvent. In calcu- lating this second part, it is assumed that the absorption due to the solvent is the same as it would be if there were no dissolved salt present. We therefore write 10 Studies on Solution. where a is the absorption coefficient for the pure solvent, c is the concen- tration in gram-molecules of salt per liter of solution, and A is called the molecular absorption coefficient of the salt in the solution. From this relation it follows that The present investigation has consisted of a systematic and thorough study of the absorption coefficient a. This quantity has been measured at intervals of 20juju to 40/iM throughout the region of the spectrum from 600/z/* to l,300ju/z for many solutions. The work has been restricted to a study of inorganic salts in aqueous and alcoholic solution. All the measurements have been carried out with solutions at room tempera- ture. The values of a, when plotted as ordinates against the corre- sponding wave-lengths as abscissas, form the absorption curve. For each salt a series of solutions varying in concentration from satura- tion to moderate dilution was prepared and the absorption curve has been drawn for each solution. From the measured values of a and OQ and from the known value of c, A has been calculated for each wave- length by means of formula (2). The values of A for a given wave- length have been plotted as ordinates against the corresponding values of c as abscissas. The curves thus formed will be referred to as the A-c curves. It was the purpose of the present investigation to deter- mine the form of the A-c curves. HISTORICAL. The general problem of the absorption of radiation by solutions has been the subject of many investigations. Only those papers are of primary interest here which concern determinations of the numerical values of the absorption coefficient as a function of the concentration. Beer 1 measured the absorption coefficient for red light of a number of aqueous solutions of inorganic salts. The results of his experiments showed that within the error of experiment A was a constant with respect to c. The statement that "A is a constant" has been men- tioned by subsequent workers in this field as ''Beer's law." This "law" has since been shown to be the exception rather than the rule, and therefore in this paper but few references have been made to "Beer's law." A paper by Rudorf 2 entitled " Lichtabsorption in Losungen vom Standpunkt der Dissociationstheorie" reviews the literature up to the year 1904 and gives a very good statement of the conclusions reached at that time. Rudorf concluded the section of his paper concerning Beer's law with the following observation: l Pogg. Ann., 86, 78 (1882). 'Sammlung Chemischer und Chemish-Technischer Vortrage, 9, 1 (1904). The Absorption Coefficient of Solution for Monochromatic Radiation. 11 "We have seen that in general Beer's law can be true only within certain limits, though many observers believe that it holds accurately within wide limits. The experimental data is in many cases unsatisfactory, and in still others contradictory." A survey of the literature since 1904 bearing on the relation between A and c yields few definite conclusions. The reason for the unsettled state of the problem is not difficult to find. None of the researches has been carried out with the necessary completeness. The investi- gators have been content with a determination of the molecular absorp- tion coefficient A for a few concentrations at a very limited number of points of the spectrum. In 1906 Miiller 1 measured A for three solutions of copper chloride in water. The values of A were determined for 5 wave-lengths in that part of the visible region of the spectrum where this solution was fairly trans- parent, Muller's results snowed that A was not only variable with c, but also that the rate of variation was different for each wave-length. Hantzsch ? and his co-workers (the reference is to the final one of a series of papers) have recorded the value of A for a number of solutions of inorganic colored salts. A was measured for a single wave-length for a few concentrations and was found in general to decrease with cin the ease of the monochromates, the ferrocyanides, and the permangan- ates of the alkali metals, and to be fairly constant for dilute solutions of certain organic colored salts. . , Sheppard, 3 in his researches, has included determinations of A for alcoholic solutions of a few dyes. The values of A were constant within the; error of experiment, except for the most dilute solutions, where they experienced a perceptible increase, which was ascribed to chemical change taking place in the solution. Garrett 4 has recorded the values, of A for aqueous solutions of a num- ber of salts of copper. A was determined for 3 wave-lengths on the violet side of the red absorption band for 3 concentrations and was found in all cases to decrease with dilution. In the work thus far cited the values of A have been determined for wave-lengths lying in the visible region of the spectrum by means of a visual spectro-photometer. The photographic method of testing Beer's law, as used by previous workers in this laboratory, 5 is applicable to both the ultra-violet and visible regions of the spectrum. This method, however, yields informa- tion concerning the variations of A with c only for those wave-lengths on the edge of an absorption band. In studying a large number of solutions in this way, many bands were found whose edges obeyed Beer's law and many more whose edges did not. 'Ann. d. Phys., 21, 515 (1906). 4 Zeit. Elektrochein., 19, 1 (1913). Zeit. phys. chem., 84, 321 (1913). 6 Carnegie Inst. Wash. Pub. Nos. 110, 130, 160, 190. Journ. Chem. Soc., 95, 15 (1909); Proc. Roy. Soc., 82-A, 256 (1909). 12 Studies on Solution. A very important quantitative study of the light-absorption of solu- tions has been carried out by Houstoun and his co-workers 1 (the refer- ence is to the last of a series of eleven papers) . Many phases of the gen- eral problem were considered and frequent reference will be made here to the separate papers. His work is unique in that it is the only record we have of the determination of A for solutions for wave-lengths in the infra-red. Even this work, although of a more complete character than any of the researches hitherto attempted, did little more than touch upon the relation between A and c. The absorption curves were determined for the region of the spectrum from 645/zju to 1,270/A/x for the chloride, bromide, iodide, nitrate, and sulphate of cobalt. 2 This was done for a strong and for a dilute aqueous solution of each salt, in all cases the values of A for the more concentrated solution were found to be greater than the corresponding values for the dilute solution. Houstoun also made a further study of the chloride and bromide of cobalt, nickel, iron, and copper. 3 Solutions of each salt were prepared varying in concentration from saturation to moderate dilution. A was determined for a single wave-length lying on the edge of an absorption band. The results for nickel chlo- TABLE i. Nickel ride as an example are given in table 1. The values Chloride in Water. of A are seen to decrease with dilution reaching a Wave-length 434/1/1. minimum value, and then to remain fairly constant. Table 1 and other similar tables show that A increased again for the more dilute solutions. This increase was considered either as within the error of experiment or due to the chemical change taking place in the solution. In all of Houstoun's work A was determined by comparing a cell containing the solution with a cell of the same thickness containing the pure solvent. This method is open to criticism, but the difference be- tween the A thus determined and the true value was probably less than the errors in the values of A due to other experimental causes. APPARATUS. The apparatus used for determining the coefficient of light-absorp- tion has been developed by previous workers in the Johns Hopkins laboratory. The quantitative work was begun by Guy, 4 who built a sensitive radiomicrometer and used this in connection with a glass- prism spectrograph. The apparatus was greatly improved by Shaeffer 5 during the following year, and the apparatus used in the present inves- tigation and described in this paper is the same in all respects, except for minor details, as that used by Shaeffer and his co-workers. ^oc. Roy. Soc. Edinburgh, 33, 156(1912-13). /Kd.,31,521 (1910-11). 3 /bw*.,33, 147(1912-13). 'Carnegie Inst. Wash. Pub. No. 190, 29 (1913). */Kd., 230, 44 (1915). The Absorption Coefficient of Solution for Monochromatic Radiation. 13 The arrangement of the apparatus is shown in figure 1. The light from a Nernst glower g, operated at 110 volts on 0.8-ampere direct current from a constant potential storage battery, was rendered parallel by a lens Zi, 3.8 cm. in diameter and with a focal length of 20 cm. The light after passing through cell K' was focussed on the slit A of the spectrograph by a second lens k, 3.8 cm. in diameter and with a focal length of 20 cm. A shutter s was placed between the glower g and lens /i, by means of which the light could be turned on and off. The optical system thus far described, con- sisting of the glower, the two lenses Zi and Z 2 , and the cells, was held by a solid metal frame- work and was perpendicular to the plane of the drawing in figure 1. The light after passing through lens Z 2 was reflected onto slit A by a right-angle glass prism (not shown in figure 1) close to slit A. The temperature of the solution was recorded by a thermometer not placed in the solution but fastened on the metal frame supporting the cells. The spectrograph con- sisted of the Littrow mount- ing of a plane grating. The grating had a ruled area 6 cm. by 7.5 cm. and was ruled 15,000 lines to the inch. The cone of light from slit A was reflected by a right- angle glass prism through the large achromatic lens Z 3 , 10 cm. in diameter and with a focal length of 75 cm. The spectrum was brought to a focus at slit B. The grating possessed a bright first-order, and this first-order spectrum FIG. i. Schematic diagram was used throughout the present work. The dispersion was such that with slit B 1 mm. wide a beam of light containing a wave-length range of 20 A. or 2ju/i passed through. In this work both slit A and slit B were always 1 mm. in width. The grating was mounted on a turntable, which was rotated from the outside by a worm-screw, thus causing various wave- lengths to pass through slit B. The approximately monochromatic beam of light from slit B was focussed on the junction of the radiomicrometer r by a lens Z 4 , 3.5 cm. in diameter and with a focal length of 6 cm. A complete description of the construction of the radiomicrometer is given in Shaeffer's paper. 1 To eliminate the drift of the zero-point Carnegie Inst. Wash. Pub. No. 230, p. 44. 14 Studies on Solution. of the instrument, due to temperature changes in the air of the room, it was encased hi a large box surrounded with an excelsior packing. When the room temperature was kept fairly constant, the drift was negligible. Occasionally readings were taken in the presence of a slight drift, and in this case the zero was redetermined after each deflection and one-half the drift added to the observed deflection. The deflections of the radiomicrometer were observed on a ground- glass scale at a distance of 5 meters. This scale was placed on the table on which was mounted the Nernst glower and cells. This arrangement enabled a single observer to carry out all the measure- ments, i. e., to manipulate the cells, to watch the glower current, and to read the deflections. The cells, which were made by Shaeffer 1 and described in his paper, were used in the present work on a few salts only. These cells, which were of brass, gold-plated, and of adjustable depth, although perfectly workable, were found to be somewhat clumsy for this investiga- tion. A cell was required which could be easily and quickly opened, cleaned, and filled. The form of cell finally chosen was very satis- factory. This cell (fig. 2) consisted simply of a glass ring, 4.2 cm. in diameter, closed on each end by a plane-par- allel plate of glass 2 mm. thick. The glass ring was ground to a uniform thickness within 4.acms. *-* 0.001 inch. 1 : It was found unnecessary to FlG . 2 . Cross-section of cell. cement the glass plates on the glass ring. To fill the cell the glass ring was placed on the bottom plate, the solutions poured in, and the upper plate slid on. In the case of water solutions, the cell thus filled Was quite tight and remained free from bubbles for several hours; in the case of solutions of methyl alcohol small bubbles appeared in about half an hour. It was some- times convenient to seal the bottom plate on to the glass ring with rubber cement. Six cells were made varying in thickness from 1.844 to 21.996 mm. A thick cell K' and a thin cell K (fig. 1), were held in a frame (not shown hi fig. 1) and either in turn could be quickly interposed in the path of the light. PROCEDURE. The solution for which a was to be determined was placed in two cells exactly alike, except that one was thin and the other thick. The energy / of the monochromatic beam of light after passing through the thin cell containing a thickness h of solution, and the energy /' after passing through the thick cell containing a thickness h' of solution, were measured in arbitrary units i. e., deflections of the radiomicrom- *Camegie Inst. Wash. Pub. No. 230, p. 50. The Absorption Coefficient of Solution for Monochromatic Radiation. 15 eter. If the initial intensity, 7 , of the light falling on the cell was the same in each case: h h 5 I a= 7 logl 3~' where d and d f are the deflections produced by / and I', respectively, and t is the difference in thickness in millimeters of the two cells. This method eliminated all corrections for reflections from the glass surfaces and thus gave a directly. For the study of each salt, solutions of the salt in the solvent were prepared varying in concentration from saturation to moderate dilution. The absorption curve for each solution was then drawn. This involved the determination of a at intervals of 20^ to 40ju/* throughout the available region of the spectrum i. e., from 600/x/z to 1,300/iju. The experimental procedure was as follows: The two cells, filled with the solution whose absorption was to be measured, were mounted in place in their frame and were adjusted until the image of the Nernst glower on slit A suffered no displacement when either cell was inter- posed in the path of the light. The zero-reading of the radiomicrom- eter was taken, and then the deflections were noted for each cell in turn in the path of the light. This was done for each wave-length, the shutter (s, fig. 1) being closed, usually after every four readings, to see if the zero remained unchanged. Readings were taken fo? wkve-length intervals of 20ju/x to 40juM throughout the entire available spectrum, and the whole set was repeated in reverse order. Thus; each point on an absorption curve, i. e., each measurement of #, was the mean of two, and often more, separate determinations. As an illustration of the method of procedure, the complete read- ings for a solution of NiSO 4 in water are given in table 2. The data from which the curves have been plotted are arranged iri tables. For each solution the following data are recorded in these tables: the temperature of the solution in degrees centigrade; t, the. difference in thickness of the two cells; c, the concentration in granv molecules of salt per liter of solution; the values of a calculated from equation (3); and the values of A calculated from equation (2). The short-wave limit of the absorption curves is at about 600/xju because the deflections of the radiomicrometer for light of wave-length shorter than 600/z/i are too small to give accurate values of a. The long-wave limit is at about 1,200/i/v although the limit set by the transparency of glass is at about 2,OOOjuM- The reason for this was that in order to study regions beyond 1,200/iju a color screen had to be used. Wave-length l,200juM in the first-order is overlapped by wave-length of the second-order. A thin layer of a strong solution of cnro- 16 Studies on Solution. mium chloride in water served as a color screen, and such a screen was used whenever a was determined for wave-lengths greater than 1,200/z/*. This absorbs the light up to 700juju (see fig. 24) and is transparent for wave-lengths above this. Water itself is quite opaque above 1,300/i/z (see fig. 3) and hence this screen cut down the deflections to such an extent that the values for a were liable to great inaccuracy. In most cases, therefore, the long-wave limit of the absorption curves is at about 1,200MM- TABLE 2. Nickel Sulphate in Water. Temperature 18.6. / = 10mm. c=0.4. Deflections of radiomicrometer , in millimeters. Wave-length. Cell of thickness Cell of thickness d/d' djd' a = 11 mm. = 1 mm. d' d Mean. 645wi 30 30 37 37 1.25 1.25 1.25 0.0097 866 38 38 49 50 1.29 1.31 1.30 0.0114 585 38 38 38 59 63 59 1.55 1.66 1.55 1.60 0.0204 605 34 35 71 74 2.08 2.10 2.09 0.0320 625 25 25 82 77 3.28 3.08 3.18 0.0502 644 17 17 89 89 5.22 4.94 5.09 0.0707 664 19 17 110 95 5.80 5.60 5.70 0.0756 684 21 19 131 110 6.22 5.80 6.01 0.0779 704 20 19 142 127 7.10 6.68 6.89 0.0838 724 20 21 156 140 7.80 6.67 7.23 0.0859 744 26 27 171 168 6.58 6.22 6.40 0.0806 764 41 43 188 186 4.58 4.33 4.45 0.0648 783 55 63 178 198 3.24 3.15 3.20 0.0505 803 81 91 192 212 2.37 2.33 2.35 0.0371 823 110 117 207 221 1.88 .89 1.89 0.0277 842 130 141 215 230 1.66 .63 1.65 0.0218 861 151 158 223 234 .48 .48 1.48 0.0170 881 158 165 227 236 .44 .43 1.44 0.0158 901 159 158 232 229 .46 .45 1.46 0.0164 920 149 149 233 228 .57 .53 1.55 0.0190 940 129 123 232 222 .80 .81 1.81 0.0258 960 85 86 221 214 2.60 2.49 2.55 0.0407 979 74 78 222 233 3.00 2.98 2.99 0.0476 998 63 70 216 229 3.43 3.26 3.34 0.0524 1018 61 58 220 212 3.61 3.64 3.63 0.0560 1037 52 53 211 207 4.06 3.90 3.98 0.0600 1056 41 43 185 196 4.51 4.56 4.53 0.0656 1076 33 37 177 190 5.36 5.13 5.24 0.0719 1095 28 27 173 177 6.18 6.56 6.37 0.0804 1115 23 24 168 167 7.30 6.96 7.13 0.0853 Kahlbaum materials were used, and when possible the salts were purified by recrystallization. In preparing the solutions a uniform method was adopted. A solution saturated at room temperature was prepared, and the concentration was determined by a standard method. The solutions of lower concentration were then prepared by diluting this mother solution. The Absorption Coefficient of Solution for Monochromatic Radiation. 17 ERRORS AND CORRECTIONS. The values of a and of A have been plotted against wave-length and concentration respectively. It was thought better to connect the plotted points by straight lines rather than to draw smooth curves through them. A glance at the figures shows that the absorption curves i. e., the curves of a against wave-length are fairly smooth, but that the curves of A against c are quite irregular. The inaccuracy hi the values of A, as shown by the irregularities hi the curves, is quite large. In many cases the deviations of the broken line indicate as much as 10 per cent variations in A. The causes of such errors are many. Without going into tedious and obvious details, it is only necessary to state that the accurate determination of a depended upon the proper choice of cell depth, which was regulated by the actual value of the absorption coefficient, as well as upon the care used hi preparing the solutions and hi cleaning and adjusting the cells. Errors also resulted from the poor keeping qualities of certain solutions. The deflections of the radiomicrometer could be duplicated to within a milli- meter. Hence the ratio of the deflections for each cell was usually accurate to within 2 per cent. In cases where the absorption coefficient was large, the deflection for the thick cell was small and the error pro- portionately greater. The values of a hi the tables are considered to be accurate to within 3 per cent, the error being greater for very high and very low values of a. A was calculated from formula (2) and devi- ations of 5 to 10 per cent were within the error of experiment. The chance for error in A was much greater for the dilute solutions than for the more concentrated ones, so that it was the practice to make up the solutions below a concentration c = l in smaller steps than in the case of solutions for which c was greater than 1. The calculations of a and A have been carried out to three figures hi most cases, although quite often the third figure is not significant. The concentration c is defined to be the number of gram-molecules of salt per liter of solution, and the solutions were prepared hi con- formity with this. The calculation for A, however, has been made on the basis that c is the concentration in gram-molecules of salt per liter of solvent. The procedure of calculating A by formula (2) presupposes that hi 1 mm. layer of solution there is a 1 mm. layer of solvent plus the dissolved salt. This, however, is not strictly true, because the addition of the salt to the solvent produces sometimes expansion and some- times contraction. The error in the value of A due to this is, how- ever, negligible in comparison with the errors arising hi other ways. For example, consider the case of an aqueous solution of CoCl 2 , when c = 1.90. At wave-length 979/^u, a = 0.0742. Assuming no expansion upon dissolving, 18 Studies on Solution. Correcting for expansion upon dissolving, using data from Landolt and Bornstein, we have In this case the correction amounts to 1 per cent. Furthermore, the example just cited is one in which this correction is at its maximum. In the cases for solutions which are more dilute and for wave-lengths where the water absorption is smaller, this correction is much less. In measuring absorption bands which are narrow in comparison with the range of wave-lengths passing through the second slit of the spectrograph, a correction for the finite width of the slit must be made. All of the bands studied in this investigation were so broad as to make such a correction negligible. It should be noted that the spectrograph and radiomicrometer of this investigation are useful for a detailed quantitative study of band structure. At no time in the present work has the full resolving power of the instrument been called upon. Readings could be taken at wave-length intervals of 4/iju without fear of measuring overlapping regions of the spectrum. THE ABSORPTION COEFFICIENT OF THE SOLVENTS. WATER. The water used throughout this investigation was the same as that used in the work on conductivity carried on in this laboratory. The water was dust-free and had a mean specific conductivity of 1.8 X 10" 6 reciprocal ohms. In view of the fact that the values of a for water are used in the calculations of A for all the water solutions, the absorption curve of water was repeated 6 times, and the recorded values are thus each a mean of 12 separate measurements. The absorption curve for water in this region of the spectrum has been drawn by one other observer, Aschkinass. 1 His curve is also plotted on figure 3 for the sake of comparison. The lack of agreement in the location of the position of the bands at 979/i/x and at l,190juM is probably due to the fact that Aschkinass used a quartz-prism spectro- graph. The determination of wave-lengths in this region of the spectrum is more uncertain in the case of the prism than the grating spectrograph. The values of a for the maximum of the sharp band at 979ju/z given by Aschkinass are lower and those for the minimum at l,070ju/i higher than the corresponding values recorded in the present 'Wiedem. Ann., 55, 401 (1895). The Absorption Coefficient of Solution for Monochromatic Radiation. 19 work. These discrepancies between the values of a are such as arise from the use of an instrument of low dispersion with relatively wide slit- widths. Although Aschkinass does not record the width of the slits, it is believed that the employment of different spectrographs is the cause of the differences in the values of OQ~. For wave-lengths shorter than 900ju/i the values of a recorded here are much greater than those found by Aschkinass. The absorption of water is quite small in this region and probably Aschkinass is more nearly correct, for he was able to use longer cells for the measurements. However, the values of a found in this work are the ones used for the calculations of A. TABLE 3. The Absorption Coefficient of the Solvents (Fig. 3). Wave- length. Water. Temp. =20.0 = 20 mm. Methyl alcohol. Temp. = 19.2 4 = 20.2 mm. Ethyl alcohol. Temp. = 18.9 * = 20.2 mm. Propyl alcohol. Temp. = 19.5 4 = 20.2 mm. Iso-butyl alcohol. Temp. = 18.9 <=20.2 mm. Iso-amyl alcohol. Temp. = 18.0 4=20.2 mm. 605uu 0.0002 625 .0002 644 .0004 664 0006 684 0006 704 0010 0004 0008 724 0015 0004 0010 744 0020 0010 .0012 764 0020 0007 .0012 783 0018 0007 .0010 0.0002 803 0017 0010 .0010 .0004 823 0018 0012 .0010 .0004 842 .0026 .0010 0.0005 0.0004 .0010 .0004 861 .0028 .0010 .0015 .0009 .0012 .0004 881 .0032 .0016 .0016 .0012 .0018 .0010 901 .0036 .0043 .0045 .0034 .0038 .0028 920 .0046 .0052 .0038 .0055 .0046 .0050 940 .0082 .0027 .0036 .0030 .0038 .0028 959 .0191 .0034 .0028 .0026 .0028 .0022 978 .0206 .0055 .0036 .0026 .0024 .0016 998 .0181 .0071 .0051 .0043 .0041 .0032 1018 .0139 .0084 .0061 .0055 .0056 .0048 1037 .0099 .0081 .0058 .0053 .0048 .0043 1056 .0075 .0071 .0056 .0048 .0046 .0038 1075 .0071 .0063 .0050 .0045 .0045 .0032 1095 .0084 .0051 .0045 .0038 .0034 .0024 1114 .0106 .0045 .0038 .0036 .0030 .0022 1133 .0161 .0083 .0056 .0050 .0045 .0038 1151 .0430 .0192 .0138 .0157 .0160 .0151 1170 .0525 .0248 .0272 .0256 .0269 .0261 1190 .0532 .0495 .0495 .0536 .0583 .0599 1210 .0530 .0403 .0394 .0422 .0420 .0450 1229 .0521 .0243 .0306 .0314 .0318 .0311 1248 .0489 .0200 .0196 .0204 .0175 .0176 1267 .0467 .0192 .0154 .0151 .0130 .0122 1287 .0494 .0151 .0123 .0130 .0094 .0089 1306 .0564 .0135 .0111 .0103 .0081 .0074 1325 .0680 .0164 .0106 .0107 .0075 .0075 1344 .0685 .0262 .0149 .0132 .0111 .0102 20 Studies on Solution. .0500 0400 600 700 800 900 1.000 MOO 1.200 FIG. 3. The Absorption Curves for the Solvents. METHYL ALCOHOL. The methyl alcohol was refluxed and distilled twice over lime and once over metallic calcium. Its specific gravity at 15 referred to water at 15 was 0.7956. The figure for anhydrous methyl alcohol given by the Bureau of Standards, Bulletin 19, page 22 (1916), is 0.79647. This indicates that the methyl alcohol used in this work was free from water. The absorption curve for this alcohol also indicates absence The Absorption Coefficient of Solution for Monochromatic Radiation. 21 of water, for the curve shows that the alcohol becomes transparent again at 1,320/zju, which would perhaps not be the case if water were present even in small quantities, as pure water is quite opaque at this point. Therefore it is believed that the maxima shown by this curve are characteristic of the alcohol and not of any impurity. ETHYL ALCOHOL. The ethyl alcohol was refluxed and distilled repeatedly over lime. Its density at 25 referred to water at 4 was 0.7851, which compares favorably with the figure 0.78506 given by the Bureau of Standards, Bulletin 19, page 7 (1916). PROPYL ALCOHOL. The propyl alcohol was refluxed and distilled once over lime. Its density at 20 referred to water at 4 was 0.8037. The figure for the anhydrous propyl alcohol given in Van Nostrand's Chemical Annual, 1913, page 312, is 0.80358. ISO-BUTYL ALCOHOL. The iso-butyl alcohol was refluxed and distilled twice over lime. Its specific gravity at 20 referred to water at 20 was 0.8033. The figure given by Biedermann, Chemiker Kalender, 1915, page 96, is 0.8031. This alcohol showed signs of slight cloudiness in the cell. The ab- sorption curve also shows general slight absorption in the visible region of wave-lengths. ISO-AMYL ALCOHOL. The iso-amyl alcohol was refluxed and distilled once over lime. Its density at 20 referred to water at 4 was 0.8111. The figure given hi Van Nostrand's Chemical Annual, 1913, page 278, is 0.8104. DISCUSSION OF RESULTS WITH THE SOLVENTS. The absorption curves for water and the five alcohols have been plotted together for the sake of comparison as shown in figure 3. All the curves have a common axis of ordinates; the zero of the ordinate axis is different for each curve, so that as a result each curve is trans- posed a convenient distance above the neighboring curve. The similarity in the positions of the maxima and minima of the curves and the concordance hi the values of a at these points are interesting. Although the infra-red transmission of the alcohols has been studied by a number of observers, 1 no determinations of the absorption coeffi- cients in the region from 600^M to 1,300/4/x have been recorded. The absorption spectra of the above five alcohols and many other sub- stances have been photographed by Abney and Testing. 2 In then- work the light was passed through a thickness of 3 inches or more of , Handbuch, vol. 3, p. 304. 2 Phil. Trans. 172, 887 (1881). 22 Studies on Solution. liquid, and the spectrum was photographed throughout the region from 600MM to l,280jzju on special plates with a glass-prism spectroscope. Then- spectrograms of the five alcohols used in this investigation show the existence of a very complicated set of absorption bands and sharp lines in this region of the spectrum. It was not possible, however, to identify any of these bands and lines with the maxima of the curves hi figure 3, for these absorption curves have not been drawn with the necessary detail. .0900 .0800 .0700 .0200 . .0100 _ 600 700 800 900 1,000 \,\00ju/i FIG. 4. The Absorption Curves for Cobalt Chloride in Water. The Absorption Coefficient of Solution for Monochromatic Radiation. 23 THE ABSORPTION COEFFICIENT OF THE SOLUTIONS. COBALT CHLORIDE IN WATER. Twenty-three solutions were prepared varying in concentration from c = 3.23 to c = 0.1. The more concentrated solutions were quite stable and showed no signs of decomposition, even after standing in the bottles for several days. In the more dilute solutions, however, there appeared a flocculent precipitate which increased their absorption mate- rially. On this account a second set of solutions, whose concentrations varied from c = 1.0 to c = 0.1, was prepared and the measurements of these appear in table 4. The absorption curves include the long-wave side of the yellow- green cobalt absorption band and the short-wave side of the infra-red band, and show the region of transmission between the two bands. The minimum of absorption is at 764ju/z. 704^/4 .0300 U .0100 _ .5 i.O 1.5 2.0 2.5 C 3.0 FIG. 5. The A-c Curves for Cobalt Chloride in Water. 24 Studies on Solution. TABLE 4. Cobalt Chloride in Water (Figs. 4 and 6). Temp. -16.5 Temp. =15.7 Temp. = 15.7 Temp. = 15.9 Temp. = 17.8 Temp. = 18.3 e <=5 mm. t=5 mm. *=5 mm. *=5 mm. t = 10 mm. ( = 10 mm. Wave- Cone. =3.227 Cone. =3.0 Cone. =2.8 Cone. =2.6 Cone. =2.4 Cone. =2.2 length. a A a A a A a A a A a A 664/K/t 0.0735 0.0334 684 0.0795 0.0331 .0640 .028: 704 0.107 0.0600 0.146 0.0483 0.0104 0.0367 0.0652 0.0297 .0610 .0250 .0480 .0214 724 .0876 .0268 .0708 .0231 .0576 .0201 .0466 .0174 .0403 .0162 .0362 .015? 744 .0571 .0171 .0501 .0161 .0444 .0151 .0370 .0135 .0328 .0128 .0296 .012C 764 .0461 .0138 .0418 .0133 .0387 .0131 .0335 .0122 .0298 .0116 .0265 .0111 783 .0445 .0133 .0413 .0132 .0376 .0128 .0324 .0118 .0296 .0116 .0265 .0112 803 .0466 .0130 .0435 .0140 .0380 .0120 .0334 .0122 .0311 .0122 .0286 .0122 823 .0406 .0148 .0475 .0152 .0401 .0137 .0374 .0137 .0340 .0134 .0316 .013C 842 .0545 .0161 .0512 .0162 .0461 .0155 .0415 .0140 .0383 .0149 .0352 .0148 861 .0576 .0170 .0558 .0177 .0510 .0172 .0450 .0162 .0416 .0162 .0376 .0158 881 .0610 .0170 ,0500 .0186 .0536 .0180 .0474 .0170 .0437 .0169 .0398 .0167 001 .0647 .0100 .0615 .0103 .0568 .0100 .0505 .0180 .0460 .0177 .0426 .0177 020 .0701 .0200 .0680 .0212 .0610 .0201 .0542 .0101 .0500 .0175 .0460 .0188 040 .0783 .0218 .0760 .0225 .0680 .0213 .0626 .0209 .0564 .0201 .0526 .0202 060 .0050 .0235 .0040 .0250 .0841 .0232 .0755 .0217 .0720 .0220 .0689 .0226 070 .0834 .0261 .0807 .0274 008 .100 .0342 .0921 .0336 Temp. =21.3 Temp. =22.0 Temp. 22.8 Temp. =23.4 Temp. =23.8 Temp. =22.3 t = 10 mm. t = 10 mm. t = 10 mm. < = 10mm. t = 10 mm. < = 10 mm. Wave- Cone. = 1.08 Cone. = 1.0 Cone. = 1.7 Cone. = 1.6 Cone. = 1.5 Conc. = l. length. a A a A a A a A a A a A 605/zM 0.0048 0.0478 0.0865 0.0455 0.0755 0.0444 0.0695 0.0434 0.0652 0.0435 0.0589 0.0420 625 .0834 .0418 .0783 .0413 .0670 .030* .0614 .0384 .0574 .0383 .0502 .0359 644 .0726 .0366 .0602 .0364 .0587 .0345 .0540 .0338 .0506 .0337 .0446 .0319 664 .0620 .0317 .0500 .0315 .0 08 .0203 .0471 .0294 .0434 .0289 .0378 .0270 684 .0500 .0255 .0458 .0241 .0407 .0230 .0387 .0241 .0357 .0237 .0315 .0225 704 .0370 .0182 .0340 .0170 .0315 .0170 .0290 .0175 .0267 .0171 .0229 .0156 724 .0276 .0132 .0260 .0120 .0220 .0126 .0208 .0121 .0198 .0122 .0172 .0112 744 .0233 .0108 .0213 .0101 .0101 .0100 .0173 .0096 .0164 .0096 .0144 .0087 764 .0215 .0008 .0107 .0003 .0173 .0090 .0164 .0090 .0153 .0089 .0136 .0083 783 .0235 .0110 .0201 .0006 .0181 .0096 .0168 .0094 .0159 .0094 .0143 .0089 803 .0246 .0116 .0225 .0100 .0200 .0108 .0187 .0106 .0178 .0107 .0161 .0103 823 .0275 .0130 .0250 .0127 .0228 .0124 .0215 .0125 .0202 .0123 .0188 .0121 842 .0300 .0143 .0202 .0140 .0260 .0138 .0247 .0138 .0232 .0137 .0217 .0136 861 .0336 .0156 .0318 .0153 .0285 .0151 .0271 .0152 .0256 .0152 .0234 .0147 881 .0358 .0165 .0330 .0162 .0307 .0162 .0290 .0162 .0270 .0159 ,0250 .0156 001 .0384 .0175 .0362 .0171 .0327 .0171 .0310 .0171 .0290 .0170 .0269 .0167 020 .0414 .0186 .0302 .0182 .0353 .0181 .0336 .0181 .0320 .0182 .0301 .0182 040 .0472 .0107 .0456 .0107 .0414 .0195 .0398 .0198 .0381 .0199 .0358 .0198 060 .0638 .0224 .0615 .0223 .0558 .0222 .0541 .0218 .0522 .0221 .0492 .0215 070 .0754 .0276 .0742 .0282 .0661 .0268 .0637 .0270 .0618 .0274 .0596 .0278 008 .0858 .0342 .0826 .0338 .0751 .0335 .0716 .0335 .0688 .0338 .0653 .0336 1018 .0848 .0415 .0798 .0408 .0776 .0425 .0725 .0418 1037 .0916 .0510 .0870 .0513 .0820 .0516 The Absorption Coefficient of Solution for Monochromatic Radiation. 25 TABLE 4. Cobalt Chloride in Water Continued. Temp. = 18.9 Temp. =20.7 Temp. =20.5 Temp. =20.3 Temp. =20.2 Temp. = 19.8 * = 10 mm. < = 10 mm. < = 10 mm. =20mm. * = 20mm. <=20mm. Wave- Cone. = 1.3 Cone. = 1.2 Cone. = 1.1 Cone. = 1.0 Cone. =0.8 Cone. =0.6 length. a A a A a A a A a A a A 605MM 0.0528 0.0406 0.0500 0.0417 3.0447 0.0406 0.0384 0.0384 0.0317 0.0396 0.0235 0.0392 625 .0463 .0356 .0427 .0355 .0384 .0349 .0334 .0338 .0290 .0362 .0222 .0370 644 .0398 .0306 .0371 .0308 .0340 .0309 .0292 .0292 .0239 .0299 .0204 .0340 664 .0338 .0260 .0317 .0264 .0293 .0267 .0246 .0246 .0214 .0269 .0175 .0292 684 .0272 .0209 .0254 .0212 .0228 .0207 .0195 .0195 .0163 .0204 .0139 .0232 704 .0200 .0146 .0187 .0145 .0165 .0141 .0147 .0137 .0127 .0148 .0109 .0165 724 .0152 .0105 .0140 .0104 .0126 .0101 .0114 .0099 .0110 .0120 .0087 .0120 744 .0130 .0085 .0123 .0088 .0107 .0079 .0097 .0077 .0085 .0081 .0081 .0102 764 .0123 .0078 .0110 .0075 .0104 .0076 .0095 .0075 .0081 .0076 .0068 .0080 783 .0131 .0087 .0122 .0088 .0109 .0083 .0100 .0082 .0082 .0080 .0065 .0078 803 .0147 .0100 .0139 .0102 .0125 .0098 .0114 .0097 .0095 .0098 .0082 .0108 823 .0173 .0119 .0162 .0120 .0143 .0113 .0133 .0115 .0106 .0110 .0087 .0115 842 .0202 .0135 .0185 .0133 .0169 .0130 .0159 .0133 .0131 .0131 .0111 .0142 861 .0220 .0148 .0205 .0146 .0189 .0145 .0178 .0150 .0149 .0151 .0124 .0160 881 .0234 .0155 .0221 .0157 .0205 .0157 .0187 .0155 .0155 .0154 .0132 .0167 901 .0252 .0174 .0237 .0168 .0218 .0165 .0200 .0164 .0172 .0170 .0140 .0173 920 .0281 .0181 .0265 .0182 .0241 .0177 .0221 .0175 .0181 .0174 .0154 .0180 940 .0335 .0195 .0319 .0194 .0302 .0200 .0276 .0194 .0239 .0196 .0197 .0192 960 .0472 .0216 .0453 .0218 .0422 .0219 .0402 .0211 .0363 .0215 .0314 .0205 979 .0555 .0268 .0541 .0269 .0512 .0278 .0476 .0270 .0417 .0276 .0371 .0275 998 .0615 .0334 .0598 .0348 .0546 .0333 .0518 .0337 .0455 .0342 .0387 .0343 1018 .0676 .0413 .0648 .0424 .0596 .0414 .0560 .0421 .0480 .042^ .0400 .0417 1037 .0767 .0511 .0726 .0522 .0658 .0509 .0610 .0511 .0525 .0536 .0423 .0540 1056 .0466 .0652 1076 .0528 .0762 1095 .0607 .0872 Temp. =20.0 Temp. =20.0 Temp. =20.4 Temp. =20.0 Temp. =20.4 * = 20 mm. < = 20mm. t 20 mm. t = 20 mm. J = 20 mm. Wave- Cone. =0.5 Cone. =0.4 Cone. =0.3 Cone. =0.2 Cone. =0.1 length. a A a A a A a A a A 605/z/z 0.0210 0.0420 0.0163 0.0408 0.0124 0.0413 0.0082 0.0410 0.0043 0.0430 625 .0186 .0371 .0144 .0360 .0115 .0383 .0071 .0355 .0042 .0420 644 .0173 .0346 .0137 .0342 .0102 .0340 .0065 .0325 .0040 .0400 664 .0149 .0298 .0116 .0290 .0088 .0293 .0055 .0275 .0033 .0330 684 .0123 .0246 .0095 .0242 .0068 .0227 .0047 .0235 .0028 .0280 704 .0088 .0156 .0076 .0165 .0056 .0153 .0039 .0145 .0025 .0150 724 .0076 .0122 .0062 .0117 .0047 .0107 .0035 .0100 .0025 .0100 744 .0072 .0104 .0058 .0095 .0045 .0083 .0035 .0075 .0024 .0090 764 .0058 .0076 .0055 .0087 .0043 .0077 .0034 .0070 .0027 .0070 783 .0063 .0090 .0052 .0085 .0043 .0083 .0034 .0080 .0027 .0090 803 .0074 .0113 .0065 .0120 .0050 .0110 .0036 .0095 .0027 .0100 823 .0075 .0114 .0070 .0130 .0065 .0156 .0034 .0080 .0032 .0140 842 .0097 .0142 .0080 .0135 .0068 .0140 .0045 .0135 .0038 .0120 861 .0104 .0152 .0091 .0157 .0073 .0150 .0049 .0105 .0041 .0120 881 .0113 .0162 .0103 .0177 .0078 .0153 .0064 .0160 .0043 .0110 901 .0121 .0190 .0114 .0195 .0088 .0173 .0070 .0170 .0050 .0140 920 .0134 .0176 .0126 .0200 .0102 .0186 .0081 .0175 .0064 .0180 940 .0180 .0196 .0160 .0195 .0144 .0207 .0124 .0210 .0106 .0240 960 .0303 .0224 .0278 .0219 .0261 .0233 .0233 .0210 .0212 .0210 979 .03 6 .0280 .0318 .0280 .0290 .0313 .0269 .0315 .0237 .0310 998 .0354 .0346 .0326 .0362 .0284 .0343 .0250 .0345 .0216 .0350 1018 .0354 .0430 .0321 .0455 .0279 .0467 .0232 .0465 .0183 .0440 1037 .0366 .0534 .0320 .0552 .0255 .0523 .0207 .0540 .0153 .0540 1056 .0401 .0652 .0348 .0682 .0267 .0640 .0204 .0645 .0147 .0720 1076 .0459 .0776 .0402 .0827 .0297 .0753 .0228 .0785 .0144 .0730 1095 .0548 .0928 .0460 .0890 .0359 .0917 .0266 .0910 .0178 .0940 1115 .0626 .104 .0550 .111 .0427 .140 .0313 .104 .0212 .106 1134 .0742 .116 .0615 .109 .0483 .107 .0405 .122 .0280 .119 26 Studies on Solution. The A c curves for wave-lengths 605/i/i to 764/z/z, inclusive, lying on the edge of the yellow-green band, show that A decreases in a marked manner with dilution and reaches a minimum value at about c = 1 .0. Below c = 1 .0, A shows a slight increase. The A c curves for those wave-lengths in the region of transparency, from 842/x/x to 979jujLi, are straight lines parallel to the abscissae, show- ing that A in this region is constant for all concentrations. For wave- lengths greater than 979/z/x, which lie on the edge of the infra-red band, A is a constant within the error of experiment. The two band-edges in question are thus seen to behave quite differently as dilution proceeds. Houstoun 1 has drawn the absorption curves for two solutions of cobalt chloride in water, and table 5 shows the comparison between his values and the values interpolated from table 4. TABLE 5. 4 for Cobalt Chloride in Water. Wave-length. c=0.65 c=3.10 Houstoun. From table 4. Houstoun. From table 4. 645MM 684 720 750 794 850 910 980 1070 0.041 .024 .031 .028 .028 .028 .028 .040 .070 0.0340 .0232 .0123 .0090 .0109 .0147 .0175 .0275 .0762 6.200 .041 .037 .016 .018 .029 .038 .074 0.0330 .0150 .0138 .0165 .0198 The agreement between Houstoun's values and the values of A found in the present investigation is far from satisfactory. However, both sets indicate similar changes in A with c. COBALT CHLORIDE IN METHYL ALCOHOL. Seven solutions were prepared varying in concentration from c = 0.7 to c = 0.1. The solutions appeared to keep very well, and no such precipitate was formed as was noticed hi the aqueous solutions. The absorption curves show that the character of the absorption of the alcohol solutions was quite different from that of the aqueous solutions, the absorption curve for the alcohol solution being shifted towards the red, so that the minimum of absorption was now at 842/i/i, the shift thus amounting to about SOjuM- The shift towards the red of the edge of the band in the green was sufficient to make this band absorb nearly all of the visible red light. (Instead of speaking of the "shift of a band/' some have preferred to speak of the bands in the 1 Proc. Roy. Soc. Edinburgh, 31, 521 (1910-11). The Absorption Coefficient of Solution for Monochromatic Radiation. 27 different solvents as entirely different bands.) As a consequence, the more concentrated solutions appeared a deep purple, becoming more and more pink as the dilution increased. The A -c curve for 744/iM shows that A decreases by a large amount with dilution, dropping from 0.128 for c = 0.7 to 0.080 for c = 0.1. This is the only A -c curve which has been plotted for a wave-length .1 . .3 4 .5 .6 C .7 700 800 900 1,000 '1,100 . 1,20 !.$00/ 0.0240 0.0568 .176 [).0227 .0703 0.0555 .160 0.0278 .0798 0.0454 .114 .260 .149 .0816 .0523 .0578 .0971 .122 .170 .221 .257 0.0303 .0763 .172 .0978 .0527 .0327 .0355 .0520 .0672 .104 .142 .165 0.0258 .0681 .153 .089S .0512 .0341 .0371 .06& .0822 .118 .137 .149 0.0258 .0681 .151 .0882 .0487 .0309 .0325 .0494 .0616 .104 .129 .141 .268 .151 .0963 .0956 .155 .182 .250 .107 .0593 .0375 .0362 .0544 .0645 .0940 .217 .114 .0666 .0769 .125 .153 .225 .108 .0556 .0317 .0362 .0527 .0662 .105 .353 .108 .164 .221 .265 .0887 .0404 .0404 .0517 .0620 .274 .156 .138 .200 .248 .0774 .0437 .0380 .0516 .0648 .206 .121 .115 .172 .213 .200 .0678 .0301 .0368 .0510 .0641 .0010 Wave- length. Temp. -22.0 t -6.36 mm. Cone. -0.0 Temp. =20.4 t 6.36 mm. Cone. -0.8 Temp. =20.2 t =6.36 mm. Cone. -0.7 Temp. =21.0 <=6.36 mm. Cone. -0.6 Temp. =21.5 <- 6.36 mm. Cone. =0.5 Temp. -22.7 t -6.36 mm. Cone. -0.4 a A a A a A a A a A a A 665w* 605 645 684 724 764 803 842 881 020 050 078 1018 1056 1005 1183 0.0263 .0652 0.0202 .0723 0.0225 .0563 0.0281 .0704 3.0225 .0404 .114 .125 .145 .115 .0638 .0343 .0230 .0268 .0522 .0642 .0803 .0073 .117 0.0321 .0706 .163 .178 .204 .162 .0887 .0453 .0283 .0317 .0473 .0623 .0949 .128 .154 0.0191 .0453 .0931 .106 .125 .0933 .0557 .0313 .0216 .0249 .0494 .0598 .0723 .0873 .107 0.0318 .0755 .155 .177 .205 .152 .0900 .0478 .0307 .0338 .0505 .0653 .0973 .133 .164 0.0125 .0332 .0787 .0890 .105 .0793 .0456 .0268 .0180 .0211 .0445 .0527 .0630 .0757 .0906 0.0250 .0764 .158 .178 .206 .155 .0878 .0484 .0296 .0330 .0508 .0642 .0982 .0364 .164 0.0125 .0304 .0685 .0750 .0850 .0635 .03SO .0230 .0146 .0169 .0390 .0459 .0538 .0621 .0763 .0894 0.0313 .0760 .171 .188 .207 .153 .0908 .0510 .0285 .0383 .0498 .0633 .0998 .137 .170 .183 .138 .0816 .0470 .0300 .0346 .0608 .0775 .100 .121 .145 .151 .0888 .0403 .0308 .0333 .0463 .0633 .0060 .127 .152 .120 .0750 .0406 .0272 .0308 .0580 .0715 .0027 .114 .135 .158 .0016 .0475 .0300 .0328 .0408 .0636 .0085 .134 .158 The*Absorption Coefficient of Solution for Monochromatic Radiation. 49 TABLE 18. Nickel Chloride in Water (Figs. 16 and 17) Continued. Temp. =23.3 Temp. =23.3 Temp. =23.0 Temp. =21.0 Temp. =20.2 Temp. -20.2 t =6.36 mm. t = 10.5 mm. t = 10.5 mm. t =20.2 mm. t =20.2 mm. t =20.2 mm. Wave- length. Cone. =0.3 Cone. =0.25 Cone. =0.20 Cone. =0.15 Cone. =0.10 Cone. -0.06 a A a A a A a A a A a A 665/i/z 0.0102 0.0307 0.0079 0.0316 0.0062 0.0310 0.0038 0.0246 0.0036 0.0360 0.0022 0.0440 605 .0239 .0797 .01S7 .0748 .0153 .0765 .0113 .0752 .0083 .0830 .0043 .0860 645 .0513 .171 .0417 .167 .0345 .173 .0265 .177 .0168 .168 .0089 .178 684 .0573 .191 .0457 .183 .0377 .189 .0294 .196 .0193 .193 .0099 .198 724 .0645 .210 .0533 .207 .0443 .214 .0321 .204 .0225 .210 .0113 .196 764 .0485 .155 .0405 .154 .0335 .158 .0243 .149 .0174 .154 .0093 .0460 803 .0291 .0913 .0236 .0876 .0190 .0865 .0143 .0840 .0105 .0890 .0055 .0760 842 .0169 .0477 .0139 .0452 .0115 .0445 .0086 .0400 .0074 .0480 .0043 .0340 881 .0125 .0310 .0099 .0268 .0086 .0270 .0066 .0226 .0060 .0280 .0041 .0180 920 .0146 .0333 .0128 .0328 .0106 .0300 .0089 .0286 .0078 .0320 .0058 .0240 059 .0346 .0517 .0312 .0484 .0289 .0490 .0257 .0440 .0238 .0470 .0209 .0360 978 .0406 .0667 .0368 .0648 .0335 .0645 .0302 .0640 .0266 .0600 .0232 .0520 1018 .0439 .100 .0381 .0968 .0329 .0950 .0284 .0967 .0237 .0930 .0178 .0780 1056 .0477 .134 .0410 .134 .0359 .142 .0277 .135 .0215 .140 .0137 .124 1095 .0595 .170 .0504 .168 .0427 .172 .0337 .169 .0256 .172 .0160 .152 1133 .0723 .187 .0622 .184 .0546 .193 .0436 .184 .0347 .186 .0255 .188 Houstoun has measured two solutions of nickel chloride in water. His values are shown in table 19 for the sake of comparison, and it is seen that the two sets of values are not greatly at variance. TABLE 19. A for Nickel Chloride in Water. c =0.757 c4.09 Wave-length. Houstoun. From table 16. Houstoun. Remarks. 684 0.19 0.178 0.229 720 .19 .200 .258 750 .146 .160 .168 794 .071 .090 .101 No data for 850 .033 .042 .045 comparison. 910 .030 .030 .039 980 .062 .047 .089 1070 .146 .140 .170 Houstoun's values indicate that A decreases considerably with dilution, which is contradictory to the results of the present work. Plate 25 of the paper by Jones and Anderson 1 would seem to indicate that A is very nearly constant, possibly decreasing slightly with dilution, for wave-lengths on the short wave-length side of the red absorption band. However, they state that the photographic method, such as they used, is not the best method for studying a band whose edge is hazy and not sharply defined as is the case for this nickel band. Carnegie Inst. Wash. Pub. No. 110. 50 Studies on Solution. .1500 600 700 800 900 1.000 Fio. 16. The Absorption Curves for Nickel Chloride in Water. The Absorption Coefficient of Solution for Monochromatic Radiation. 51 .2000 .. -... .1500 A .1000 .0500 7Z4 842 1.0 2.0 3.0. . C 4.0 FIG. 17. The A-c Curves for Nickel Nitrate in Water, NiCI 2 -6H,0 IN ALCOHOLS. In view of the fact that the cobalt bands exhibit such interesting changes with the character of the solvent, an attempt was made to prepare alcoholic solutions of nickel chloride. The dehydrated salt, however, did not dissolve perceptibly in any of the three lower alcohols. 52 Studies on Solution. In the case of the methyl alcohol, a pale greenish-yellow solution resulted after the salt had been allowed to remain in the alcohol for several days. It is believed that this was due to traces of water, for the addition of the slightest amount of water produced a similar greenish- yellow solution. The ethyl and propyl alcohols remained colorless even after standing above the salt for days. Three solutions of unknown concentration were prepared by dropping a few crystals of the hydrate NiCk 6H 2 into methyl, ethyl, and propyl alcohols. The resulting solu- tions all showed the green color characteristic of the aqueous solution. TABLE 20.NiClf6H& in Alcohols (Fig. 18). NiCl inHjO. NiCV6H0 inCHjOH. NiCl s -6H 2 O in C*HOH NiCU-6H 2 O in CsHjOH. Wave- length. Temp. 21.5 t =6.36 mm. Cone. -0.5 Temp. -21.6 t 6.36 mm. Cone, unknown. Temp. =21.4 t =6.36 mm. Cone, unknown. Temp. -21.3 t =6.36 mm. Cone, unknown. a a a a 605/iM 615 625 635 645 655 664 674 684 694 704 714 724 734 744 754 764 774 783 794 803 813 823 833 842 0.0382 .0593 .0787 0.0428 .0489 .0555 .0577 .0626 .0624 .0640 .0647 .0710 .0720 .0745 .0729 .0682 .0668 .0597 .0543 .0505 .0442 .0390 .0350 .0308 0.0550 .0607 .0700 .0740 .0781 .0740 .0661 .0621 .0593 .0560 .0587 .0607 .0615 .0591 .0569 .0533 .0518 .0435 .0415 .0413 .0332 0.0215 .0257 .0268 .0295 .0308 .0282 .0264 .0239 .0230 .0230 .0258 .0263 .0258 .0258 .0243 .0243 .0210 .0195 .0174 .0157 .0146 .0851 .0890 .0962 .105 .0979 .0793 .0638 .0456 The absorption curves for these three solutions show that the band in the red possesses two maxima. The absorption curve of nickel chloride hi water, c = 0.5, is also plotted hi figure 18 for the sake of comparison. The absorption band for the aqueous solution has a single maximum at 724/*ju. In the curve for the methyl-alcohol solution this maximum has been shifted to 744/i/i and there appears a second small maximum at 684/x/u. The curve for the ethyl-alcohol solution shows that the first peak has experienced a still further shift towards the red to 764/iju, and that the second peak at 684ju/z has The Absorption Coefficient of Solution for Monochromatic Radiation. 53 greatly increased in height. In the curve for the propyl-alcohol solution the positions and relative height of the two peaks are much the same as in the case of the ethyl-alcohol solution. In this connection a paper by T. R. Merton 1 deserves mention. The absorption curves for solutions of uranous chloride in various solvents were drawn, and the bands were shown to undergo interesting modifications, depending on the solvent used. Of course the cases of a .1000 .0500 I A / Curve I -NiCI 2 mH 2 0; C-.5 n-NiC! 2 -6H 2 OinCH,OH " JH- " "C 2 H 5 OH V-NiCI 2 -6NH 3 inH 2 700 800 900 1,000 1,100/t/l FIG. 18. Comparison of the Absorption Curves of Nickel Salts in Water and the Alcohols. the uranous chloride and the nickel chloride hydrate solutions are not exactly comparable, for the uranous chloride actually does dissolve in the solvents, and in the light of other work we are not surprised at the difference hi the character of the bands, but the nickel chloride goes into solution in the alcohols only in the presence of water. Under such 1 Proo. Roy. Soc. A, 87, 138 (1912). 54 Studies on Solution. conditions we would perhaps not expect to find such changes as have been observed above in the case of the nickel band. Possibly similar examples of this same phenomenon may be found. A solution of NiCU-GNHg hi water of unknown concentration was measured, and its absorption curve also appears hi figure 18. NICKEL NITRATE IN WATER. Twenty-four solutions were prepared, varying in concentration from c = 4.2 to c = 0.05. As was the case hi the aqueous nickel-chloride solutions, the Ac curves show that the values of A experience changes, with c depending upon the wave-length. The changes in the case of the nitrate, however, although similar in their general character, are nowhere so marked as in the case of the chloride. Throughout the region from 565/i/x to 724juju, which includes the short wave-length side TABLE 21. Nickel Nitrate in Water {Figs. 19 and 20} . Wave- length. Temp.' =20.8 * = 2.73 mm. Cone. =4.13 Temp. = 10.5 t =2.73 mm. Cone. =3.8 Temp. = 10.8 t =2.73 mm. Cone. =3.5 Temp. = 10.8 t =2.73 mm. Cone. =3.2 Temp. = 20.2 t =2.73 mm. Cone. =2.0 Temp. - 20.5 t =2.73 mm. Cone. =2.6 a A a A a A a A a A a A 565 M M 605 i 803 842 881 020 050 078 0,11$ fcjV2 0.0267 0.103 .-0271 0.0060 0.0274 0.0820 0.0256 0.0758 0.0262 0.0678 .215 .248 .128 .0865 .0970 .164 .207 0.0260 .0827 .0949 .0480 .0320 .0355 .0558 .0716 .340 .182 .116 .141 .220 .275 .101 .0511 .0382 .0388 .0600 .0726 ,306 .162 .102 .132 .204 .257 .0050 .0406 .0311 .0306 .0578 .0737 .277 .142 .0006 .110 .175 .226 .0948 .0513 .0309 .0361 .0538 .0706 .210 .138 .168 .270 .0501 .0328 .0305 .0631 .102 .132 .147 .24$ .0407 .0340 .0373 .0500 Wave- length. Temp. =24.8 t" 6.36 mm. Cone. -2.3 Temp. =23.0 <= 6.36 mm. Cone. =2.0 Temp. =24.5 t =6.36 mm. Cone. =1.7 Temp. =24.7 4=6.36 mm. Cone. =1.4 Temp. =22.0 * = 6.36 mm. Cone. =1.1 Temp. =22.1 4=6.36 mm. Cone. -1.0 a A a A . a A a A a A a A 565MM 605 764 803 842 881 020 050! 078 1018 1056 0.0506 0.0250 0.0508 0.0254 0.0370 +w 0.0233 .0728 0.0338 .0017 0.0278 .0654 0.0303 .0887 0.0275 .0806 0.0262 .0749 .158 .0900 .0501 .0316 .0362 .0703 .0870 .111 .139 0.0262 .0749 .158 .0883 .0475 .0284 .0316 .0511 .0604 .0971 .131 .200 . .110 .0706 .0810 .130 .165 .0000 .0478 .0337 .0336 .0521 .0627 ,176 .0026 .0601 .0710 .118 .152 .0870 .0450 .0285 .0332 .0404 .0655 .145 .0705. .0544 .0503 .105 .127 .0868 .0452 .0302 .0315 .0505 .0624 .127 .0667 .0436 .0510 .0005 .100 .152 .0803 .0450 .0288 .0332 .0510 .0632 .0086 .0081 .0516 .0346 .0406 .0738 .0031 .110 .144 .0878 .0446 .0283 .0328 .0488 .0659 .105 .133 The Absorption Coefficient of Solution for Monochromatic Radiation. 55 TABLE 21. Nickel Nitrate in Water (Figs. 19 and 20} Continued. Temp. =22.5 Temp. =22.5 Temp. =18.3 Temp. =18.5 Temp. =19.8 Temp. =20.0 t =6.36 mm. t =6.36 mm. t =6.36 mm. t =6.36 mm. t =6.36 mm. t =6.36 mm. Wave- Cone. =0.9 Cone. =0.8 Cone. =0.7 Cone. =0.6 Cone. =0.5 Cone. =0.4 length. a A a A a A a A a A a A 565/z/t 0.0286 0.0318 0.0204 0.0255 0.0153 0.0218 0.0174 0.0257 0.0102 0.0204 0.0107 0.0267 605 .0697 .0775 .0596 .0745 .0518 .0740 .0484 .0866 .0371 .0742 .0338 .0845 645 .106 .177 .0889 .178 .0689 .172 684 .0978 .196 .0786 .195 724 .107 .210 .0868 .213 764 .142 .148 .121 .151 .108 .152 .0950 .155 .0817- .159 .0651 .158 803 .0810 .0881 .0703 .0857 .0595 .0826 .0502 .0808 .0456 .0878 .0382 .0913 842 .0437 .0457 .0386 .0450 .0325 .0427 .0281 .0425 .0239 .0426 .0211 .0463 881 .0281 .0277 .0258 .0283 .0220 .0267 .0190 .0263 .0184 .0304 .0157 .0313 920 .0338 .0325 .0272 .0283 .0258 .0 02 .0228 .0310 .0204 .0316 .0184 .0345 959 .0641 .0501 .0589 .0473 .0512 .0457 .0489 .0497 .0446 .0510 .0397 .0515 978 .0791 .0628 .0717 .0637 .0657 .0644 .0608 .0670 .0534 .0656 .0458 .0630 1018 .102 .0979 .0956 .102 .0856 .102 .0743 .101 .0616 .0954 .0550 .103 1056 .125 .130 .115 .134 .101 .134 .0906 .139 .0788 .143 .0636 .140 1095 .152 .161 .132 .156 .117 .156 .111 .171 .0978 .179 .0759 .169 Temp. =20.5 Temp. =20.2 Temp. =20.0 Temp. =20.8 Temp. =21.5 Temp. =21.4 <=10.5 mm. t = 10.5 mm. *=10.5 mm. t = 10.5 mnii *=20.2 mm. t =20.2 mm. Wave- Cone. =0.3 Cone. =0.25 Cone. =0.2 Cone. =0.15 Cone. =0.10 Cone. =0.05 length. a A a A a A a A a A a A 565/jju 0.0071 0.0237 0.0072 0.0284 0.0065 0.0325 0.0047 0.0319 0.0025 0.0250 0.0015 0.0300 605 .0237 .0790 .0205 .0820 .0150 .0750 .0124 .0326 .0075 .0750 .0053 .106 645 .0524 .175 .0429 .171 .0352 .176 .0265 .176 .0173 .173 .0101 .202 684 .0585 .195 .0478 .191 .0386 .193 .0275 .184 .0186 .186 .0114 .228 724 .0672 .219 .0556 .216 .0432 .209 .0335 .213 .0217 .202 .0134 .240 764 .0496 .159 .0412 .157 .0336 .158 .0262 .160 .0167 .147 .0104 .168 803 .0294 .0923 .0241 .100 .0195 .0890 .0154 .109 .0100 .0830 .0058 .0820 842 .0158 .0440 .0139 .0452 .0121 .0475 .0089 .0420 .0064 .0380 .0046 .0400 881 .0128 .0320 .0095 .0242 .0089 .0285 .0069 .0253 .0052 .0200 .0041 .0040 920 .0148 .0340 .0124 .0328 .0109 .0315 .0089 .0277 .0071 .0250 .0056 .0200 959 .0354 .0543 .0308 .0468 .0287 .0480 .0266 .0500 .0231 .0400 .0212 .0420 978 .0412 .0673 .0369 .0668 .0328 .0610 .0302 .0640 .0270 .0640 .0238 .0640 1018 .0448 .103 .0379 .0960 .0333 .0970 .0298 .106 .0229 .0900 .0187 .0960 1056 .0502 .142 .0424 .140 .0349 .137 .0285 .13 .0203 .128 .0148 .146 1095 .0586 .167 .0501 .167 .0423 .160 .0347 .176 .0246 .162 .0175 .182 of the red absorption band, the A c curves show that A is a constant. From 803/4/z to 978/4/4, a region including the long wave-length side of the red band and the region of low absorption beyond this, the A c curves show that A decreases with dilution. Beyond 1,018/4/4, on the edge of the infra-red, again A is a constant for all values of c. 56 Studies on Solution. '500 .0500 600 700 800 900 , IjOOO UOO/ 1 . ^^J. 724 .0100 - /*\^*~\ r^ Y / ~^._, 744 1 I A .4000 .3000 .ZDOO .1000 Cr 2 (S0 4 ) 3 inH 2 704 m* '724 .744 J L J L .5 C -7 1.0 1.5 C 2.0 700 800 900 1.000 1,100 1.200 FIG. 25. The A-c Curves for Chromium Nitrate, Chromium Sulphate, and Potassium Per- manganate in Water; the Absorption Curves for Potassium Permanganate in Water. was to fill the cells and then make the few necessary measurements as quickly as possible. The A c curves for wave-lengths 744/iju, 764/*ju, The Absorption Coefficient of Solution for Monochromatic Radiation. 69 and 783/>tM> the region on the long wave-length edge of the green absorp- tion band, show that A is a constant with respect to c for concentra- tions greater than c = 0.05. The increase observed in A for solutions of lower concentration than this is very probably due to the effect of decomposition. TABLE 29. Potassium Permanganate in Water (Fig. 25). Wave- length. Temp. <=5 Cone. = 19.5 mm. -0.278 Temp. =21.1 t =5 mm. Cone. =0.250 Temp. =18.8 t =5 mm. Cone. =0.200 Temp. = 18.8 t =5 mm. Cone. =0.150 Temp. =18.9 t =5 mm. Cone. =0.100 Temp. -18.5 t =5 mm. Cone. =0.050 a A a A a A a A a A a A 724MM 744 764 783 803 823 0.115 0.481 .0235 .0129 .0091 .0067 2.28 0.92 .43 .22 0.2228 .0753 .0305 .0136 .0091 1.10 0.352 .143 .059 0.160 .0571 .0248 .0121 .0067 1.05 0.368 .167 0.112 .0445 .0214 .0114 .0098 1.10 0.425 .196 0.102 .0397 .0166 .0083 0.358 .136 .0535 0.09 .03 .01 .01 54 0.374 86 .117 50 .053 06 Wave- length. Temp. =18.2 t =5 mm. Cone. =0.025 Temp. =18.0 t =5 mm. Cone. =0.020 Temp. - 18.2 t =5 mm. Cone. =0.015 Temp. =17.5 <=5 mm. Cone. =0.010 Temp. -17.5 t =5 mm. Cone. =0.005 a A a A a A a A a A Mfyy* 664 684 704 724 744 764 783 803 823 0.0816 .0657 .0435 .0248 .0178 .0121 .0098 0.120 .0813 .0481 .0270 .0158 .0083 0.1204 .0639 .0381 .0187 .0129 .0091 .0083 i!77 2.49 1.32 0.67 0.167 .0872 .0508 .0273 .0121 .0091 0.122 .0704 .0381 .0221 .0106 .0083 4.31 2.46 1.26 4.63 2.44 1.34 4.71 2.55 1.38 4.70 3.26 2.00 CONCLUSION. The relation between A, the molecular light-absorption coefficient of the solution, denned by formula 2, and c, the concentration of the solution in gram-molecules of salt per liter of solution, has been deter- mined. It has been found that in general A is not a constant. In certain cases A decreases with dilution, in other cases A increases with dilution, and still other cases as dilution proceeds A decreases to a minimum, and then increases again. Another possible combination, namely, that A should increase to a maximum and then decrease, was not met with. The deviations from a constant value observed in A were usually comparatively small, except at certain points in the 70 Studies on Solution. spectrum for the cases of certain solutions. These points have been spoken of by Houstoun 1 as "sensitive points/ 7 These sensitive points have been found in general to be situated at the edges of absorption bands. At present there is no adequate theory to explain the facts which have been recorded here. The fact that A varies with the concen- tration has been probably correctly attributed by Jones and Anderson and others to the formation of complexes, which were considered to be loose chemical compounds of molecules of the salt with molecules of the solvent. Undoubtedly the changes in A with c observed in this investigation may be explained in a qualitative manner by the hypothe- sis of complexes, or "solvates" as they have been called; but before it can be useful for the interpretation of quantitative data, the solvate hypothesis must be couched in more mathematical terms. Roy. Soc. Edinburgh, 33, 151 (1912-13). CHAPTER II. THE CONDUCTIVITY AND VISCOSITY OF CERTAIN ORGANIC AND INORGANIC SALTS IN FORMAMID AND IN MIXTURES OF FORMA- MID WITH ETHYL ALCOHOL. BY P. B. DAVIS AND H. I. JOHNSON. INTRODUCTION. The study of the conductivity and viscosity of salts in formamid as a solvent was begun in the Johns Hopkins Laboratory in 1915 by Davis, Putnam, and Jones. 1 In the report on their investigations a compre- hensive survey is made of the work of previous experimenters on forma- mid as well as a detailed comparison of the physical and chemical properties of this solvent with those of water. Their work comprised at first a study of the methods available for obtaining formamid of sufficiently low specific conductivity. Repeated fractional distillation in vacuo was finally adopted as the most suitable process and an efficient vacuum distillation apparatus was devised and constructed. This appa- ratus and the scheme of fractionation are described in detail in their paper. Having obtained pure formamid in sufficient quantity for conduc- tivity purposes, a preliminary study was made of the conductivity, dis- sociation and viscosity of electrolytes in this solvent. They found that in general conductivity values are much lower in formamid than in water, but that complete dissociation is reached at a much lower dilution. The first fact is attributed to the greater viscosity of forma- mid as compared with water, the second to its higher dielectric constant and greater association factor. From a study of the temperature co- efficients some evidence was also obtained for the formation of solvates. The viscosities of solutions of all the salts studied were greater than that of formamid itself. Even caesium salts, which produce the greatest lowering in the viscosity of water and glycerol, increase the viscosity of formamid, although to a lesser extent than the other salts of the alkalis. The present investigation, which is a continuation of the earlier work, has comprised a study of the conductivity and viscosity of (1) salts with a common anion i. e., a series of nitrates of the inorganic salts and of formates of the organic salts; (2) salts with a common cation i. e., the sodium salts of the organic acids; (3) a study of the behavior of certain representative salts in mixtures of formamid with ethyl alcohol. 'Carnegie Inst. Wash. Pub. No. 230, 16. 72 Studies on Solution. EXPERIMENTAL. PREPARATION OF THE SOLVENTS. Formamid. The formamid used in this work was prepared in the same manner as that used by Davis and Putnam i. e.j the so-called c. p. material was subjected to repeated fractionation in the vacuum distilling apparatus described by them. By this method it was pos- sible to obtain formamid of a specific conductivity comparable to that of water with a minimum loss of material. The conductivity values for the solvent used in this work was some- what lower than that used earlier, ranging from 0.7 to 1.5X10" 3 as compared with 2.7X10" 5 . The average density was 1.130 at 25, the viscosity 0.0332 at the same temperature. A very small fraction, representing only about one-tenth of the original volume, was obtained after about three fractionations more than required for preparing the solvent in large quantities which had a specific conductivity of about 2X10" 6 , a viscosity of 0.03358, and a density of 1.1331. Merry and Turner 1 mention having obtained a similar fraction by repeated crys- tallization with a viscosity of 0.03359 and density of 1.1312. After formamid had been recovered from salts used in making about 15 "sets" 2 of solutions, it was found by continued fractionation that a product could be obtained which showed a specific conductivity of 0.83 X 10~ 5 at 25. This value is quite comparable with those obtained when formamid is purified from the commercial product. The infer- ence drawn from this observation is that the salts do not alter the pur- ity of the solvent. It was also observed that formamid upon standing in sealed glass-stoppered Erlenmeyer flasks for a period of four months, June-October 1916, in a dark closet, increased in specific conductivity only about ten-fold. The values observed were 0.7 X 10~ 5 and 0.97 X 10- 4 at 25. Formamid with a specific conductivity of 0.70 X 10~ 5 at 25 offers no great advantage over that with an average specific conductivity of about 1.5X 10~ 5 at 25, with the important exception of a lower solvent correction. When the formamid was recovered from mixtures with ethyl alcohol its specific conductivity would reach a value of the order of 1.5 X 10~ 5 at 25 in about the same number of fractionations as when recovered from pure formamid solutions, but on standing the specific conductivity soon increased and in the course of 3 or 4 days became too large for conductivity measurements. This suggests a possible reaction between formamid and alcohol, the products of which are more difficult to remove by fractionation than the traces of ammonium formate result- ing from hydrolysis of pure formamid by moisture from the air. Fur- l Journ. Chem. Soc., 106, 748 (1914). 2 By "set" is meant all the solvent required for the solutions of various dilutions. Conductivities and Viscosities in Formamid and in Mixed Solvents. 73 ther fractionation, however, yielded a product which maintained a high specific conductivity during the same interval of time as the pure solvent. Ethyl Alcohol. The ethyl alcohol used in preparing the mixed sol- vents was obtained by refluxing a good grade of commercial alcohol over lime for about 24 hours, then distilling. The middle fraction of about seven-tenths of the total distillate was collected and kept in receiver similar to that described by Lloyd and Pardee. (See Chapter III.) It had a mean specific conductivity of 4.1X10" 7 at 25 and a density of 0.78506 to 0.78507 at the same temperature. Mixed Solvents. The mixed solvents containing formamid and alcohol were prepared by weighing directly into glass-stoppered flasks the quantities of each component to make a mixture of the desired weight per cent of each, all weighings being reduced to a vacuum. SALTS. As in the earlier work, all salts used were carefully recrystallized and dried to constant weight at a suitable temperature depending upon the nature of the salt. In the case of calcium nitrate, the salt was prepared from the purified carbonate, the solution evaporated to dryness, and the salt heated to constant weight at 150, since it was practically impossible to recrystallize it. The aqueous solution showed only traces of alkalinity. The formates and the sodium salts of the other organic acids were purified by recrystallization or were prepared from the purified acids. Just before using they were dried to constant weight in the vacuum drying-oven described under the head of apparatus. In the case of all hygroscopic salts the drying process was repeated after weighing out the required amount of salt for the solutions. SOLUTIONS. All solutions were made up at 20, the more concentrated by direct weighing, those below one-tenth molar by dilution. Special precau- tions were used to protect both solvent and solutions from access of moisture, the procedure followed being essentially that outlined by Davis and Putnam; 25 to 50 cubic centimeters only of each solution were prepared, as this amount was sufficient both for conductivity and viscosity measurements. APPARATUS. The conductivity apparatus used was identical with that employed in the earlier work. The plate type of cell, previously described, served for measuring the conductivities of solutions both in pure formamid and in the mixed solvents. The cells were carefully standardized at regular intervals. 74 Studies on Solution. The viscosity measurements were obtained in a modified form of the Ostwald viscosimeter, the diameter and length of the capillary being adjusted so as to render the instrument suitable for measuring liquids more viscous than water. The viscosimeters were calibrated according to the more accurate method proposed by Thole, 1 using as calibrating liquids ethyl, propyl, and isobutyl alcohols, 30, 40, and 50 per cent by weight mixtures of ethyl alcohol and water and a 40 per cent solution of pure sucrose. The values for the density and viscosity of these calibrating liquids were obtained from the data compiled by Thole, Bingham, 2 and others from the most reliable measurements of various investigators. The average constant obtained for each instru- ment with this method gave somewhat larger values for the viscosity of formamid solutions than when calibrated with water alone, the time of flow of water being too short for accurate measurements i. e., less than 100 seconds. The following will serve as an example of the constants obtained: VISCOSIMETER I A . r/25 D25/4 tfxio- 4 Ethyl alcohol 001096 78506 1.253 30 p ct ethyl alcohol 00218 95967 1.243 40 p. ct. ethyl alcohol 40 p. ct. sucrose .00235 005187 .93148 1 . 10188 1.248 1.244 Av. 1.247 All measurements, both of conductivity and viscosity, were carried out in the thermostats described in a previous paper, in which a con- stant temperature to within 0.01 was maintained. In order to obtain completely anhydrous samples of the salts studied a vacuum drying-oven was designed and constructed with the aid of Dr. Pardee. This apparatus consisted of a tubulated bell-jar 18 cm. X 24 cm. mounted on a heavy iron vacuum-plate. Two pairs of electrical connections lead into the center of the plate through a rubber stopper, one pair to a stove consisting of a 50-watt carbon-filament lamp incased in a metal chimney open at the top and having a circular window near the bottom, the other pair leading to a miniature fan motor in series with an 8 candle-power carbon lamp placed on the outside base. The fan maintained circulation within the oven by driving the air through the open side of the chimney, up around the lamp, and then out over two dishes containing either sulphuric acid or phosphorus pentoxide. The material to be desiccated was placed in watch crystals on per- forated trays set above the motor and chimney. The tubular in the l Journ. Chem. Soc., 105, 2009 (1914). Zeit. Phys. Chem. 83, 644 (1913) ; Bureau Standards Scientific Paper No. 298. Conductivities and Viscosities in Formamid and in Mixed Solvents. 75 bell-jar was closed with a rubber stopper carrying the thermometer and pump connection. At about 90 mm. of mercury water boils at 49.6. The heater maintained a temperature of 65, the suction pump a vacuum of 70 to 80 mm., so that with the rapid circulation of the warm residual air over the material and the drying agent all traces of moisture could be removed from a sample with much greater ease than in an ordinary vacuum desiccator. PROCEDURE. Each "set" (i. e., M/2, M/4, M/10, M/50, etc.) of solutions in pure formamid were made up the day before the conductivity measurements were taken, since experiments showed that measuring the solutions on the same day they were prepared did not increase the accuracy of the work. In the case of mixed solvents, however, it was necessary to make up the solutions and measure them the same day. Cells were read consecutively in the 15, 25, and 35 baths. This order was always followed. The bridge readings, however, could be duplicated for the more concentrated solutions when allowed to come to temperature again in the 15 or 25 baths. The molecular conductivity values were repeated for a number of salts, representing each series measured, to within 0.5 mm. reading on the bridge for all more concentrated solutions. Therefore, consider- ing the errors in making up "check" solutions, the values below should be approximately correct. In the tables all conductivity values are expressed in reciprocal ohms and are the molecular conductivities of gram-molecular weights of the various salts. These molecular conductivities (/*) were calculated VOL from the formula ^^K-^r, where K represents the cell constant, v the volume of concentration, R the resistance in ohms as measured by the rheostat, (a) and (6) the readings on the two sides of the bridge. The percentage dissociation, a, was calculated from the equation a = X 1 00, M oo where M oo is the highest value of n, obtained. The temperature coefficients in conductivity units (T) were derived by means of the formula -~ jr- = T, in which jj,jt represents the t t molecular conductivity at the higher temperature t, and /*/ at the lower temperature t'. The coefficients expressed as percentages were T calculated from the formula A = - M The values representing the molecular conductivity in these tables are mean of three bridge-readings involving different values for R. The 76 Studies on Solution. term V in the tables represents the number of liters containing a gram- molecular weight of the solute. K expresses the specific conductivity of the solvent. The viscosity data were calculated from the formula rj = K* d- t-, where t\ presents the viscosity coefficient, k the constant of the instru- ment determined by calibration with a number of liquids of known vis- cosity, d the density of the solution at the temperature in question, and t the tune of flow of the liquid or solution under investigation at that temperature. The fluidity

). Z>25/4 1,15 i?25 1735 ,W ,26' ** 15-25 25-35 0.5 1 . 1436 0.04679 0.03515 0.02826 21.37 28.45 35.39 0.0331 0.0244 .25 1.1376 .04546 .03409 .02746 22.00 29.33 36.42 .0332 .0242 .10 1.1330 .04474 .03384 .02746 22.35 29.55 36.42 .0322 .0232 Solv. 1.1302 .04369 .03298 .02632 22.89 30.32 37.99 .0325 .0252 Conductivities and Viscosities in Formamid and in Mixed Solvents. 77 TABLE 31. Potassium Nitrate in Formamid. Temperature Coefficients of Conductivity. V Molecular Conductivity. Per cent. Conductivity Units. 15 25 35 15-25 25-35 15-25 25-35 2 14.08 18 .17 22.72 0.0290 0.0250 0.409 0.455 4 16.67 21 .62 26.82 .0297 .0240 .495 .520 10 19.04 24 .39 30.30 .0280 .0242 .535 .591 50 21.27 27 .29 33.90 .0283 .0242 .602 .661 100 21.90 28 .14 34.85 .0284 .0238 .624 .671 200 22.48 29 .05 36.18 .0292 .0245 .657 .713 400 23.70 30.53 37.84 .0288 .0239 .683 .731 K= 0.945 X10~ 6 1.24 X 10-* 1.52 X10~ 5 Mol. Viscosity and Fluidity. Temperature Coefficients (15 ^25 *>35 15-25 25-35 0.5 1 . 1570 0.05166 0.03858 0.03040 19.36 25.92 32.89 0.0339 0.0269 0.25 1.1444 .04836 .03611 .02819 20.68 27.69 35.47 .0339 .0281 0.10 1 .1359 .04591 .03450 .02751 21.78 28.99 36.35 .0331 .0254 Solv. .04369 .03298 .02632 22.89 30.32 37.99 .0325 .0253 TABLE 32 Sodium Nitrate in Formamid. V Molecular Conductivity. Dissociation. Temperature Coefficients of Conductivity. Per cent. Conductivity Units. 15 25 35 15 25 35 15-25 25-35 15-25 25-35 2 4 10 50 100 200 400 800 1600 12.96 15.44 17.72 19.38 20.05 20.35 20.77 21.30 21.26 16.74 20.23 23.32 25.03 26.18 26.62 27.14 27.73 27.56 . 21.13 24.69 29.72 31.62 33.01 33.20 33.98 34.47 34.38 K=0.67; 60.8 72.4 83.1 90.9 94.1 95.5 97.5 100.0 60.4 72.9 84.0 90.2 94.4 95.9 97.8 100.0 61.3 71.6 86.2 91.7 95.7 96.3 98.5 100.0 0.0291 .0310 .0317 .0291 .0305 .0306 .0305 .0302 .0295 :io-* 0.0260 .0220 .0273 .0263 .0260 .0247 .0251 .0243 .0247 0.378 .479 .560 .565 .613 .627 .637 .643 .630 0.439 .446 .640 .659 .683 .658 .684 .674 .682 K10~ 5 0.87 X 10-* 1.07 X Mol. Cone. Viscosity and Fluidity. Temperature Coefficients () D25/4 Tjl5 7/25 7735 ^15 ?25 V>35 15-25 25-35 0.5 0.25 0.10 Solv. 1.1542 1.1429 1.1361 1.1314 0.05585 .05012 .04651 .04403 0.04112 .03726 .03509 .03338 0.03223 .02972 .02785 .02665 17.91 19.95 21.50 22.71 24.32 26.84 28.50 29.96 31.03 33.65 35.91 37.52 0.0358 .0345 .0326 .0319 0.0276 .0254 .0260 .0252 78 Studies on Solution. TABLE 33. Calcium Nitrate in Formamid. Temperature Coefficients of Conductivity. Molecular Conductivity. V Per cent. Conductivity Units. 15 25 35 15-25 25-35 15-25 25-35 10 30.44 39.50 49.37 0.0297 0.0249 0.906 0.987 50 37.80 48.56 60.71 .0284 .0250 1.076 .215 100 41.18 52.98 66.56 .0286 .0256 1.180 .358 200 42.55 54.89 70.13 .0290 .0274 1.234 .524 400 43.46 55.90 72.15 .0286 .0271 1.244 .625 1600 46.03 58.54 75.44 .0272 .0306 1.251 .690 # = 1.41X10-* 1.91 X10~ 5 2.34 XlO- 5 TABLE 34. Barium Nitrate in Formamid. Molecular Conductivity. Temperature Coefficients of Conductivity. V Per cent. Conductivity Units. 15 25 35 15-25 25-35 15-25 25-35 4 20.90 27.19 34.14 0.0300 0.0255 0.629 0.695 10 28.20 36.93 47.08 .0309 .0274 .873 1.015 50 ... 36 79 47 73 60 77 .0298 .0273 .094 .304 100 39.73 51 91 65 05 . 0306 .0257 .218 .314 200 40.86 53.16 66.52 .0300 .0251 .230 .336 400 41.53 53.94 67.39 .0299 .0249 .241 .345 800 44.20 57.51 71.90 .0300 .0250 .331 .439 1600 45.24 58.78 74.05 .0296 .0259 .354 .527 A" =0.79 XlO- 6 1.99 XlO- 5 1.27 XlO- 5 Viscosity and Fluidity. Temperature Coefficients (). Mol. Cone. Z>25/4 ij!5 ij25 7735 >15 ^25 ?35 15-25 25-35 0.25 1.1785 0.05815 0.04286 0.03393 17.20 22.33 29.47 0.0298 0.0306 0.10 1 . 1504 . 04903 .03688 .02933 20.40 27.11 34.09 .0329 .0257 Solv. 1.1313 .04440 .03328 .02651 22.52 30.05 37.72 .0334 .0255 Conductivities and Viscosities in Formamid and in Mixed Solvents. 79 TABLE 35. Strontium Nitrate in Formamid. V Molecular conductivity. Dissociation. Temperature Coefficients of Conductivity. Per cent. Conductivity Units. 15 25 35 15 25 35 15-25 25-35 15-25 25-35 4 10 50 100 200 400 800 1600 22.62 29.46 37.73 39.18 41.09 42.87 42.94 42.47 30.18 39.24 49.73 51.29 53.75 56.75 56.91 55.74 37.62 49.77 62.99 64.82 67.51 70.84 71.38 69.30 K = l.'< 52.6 68.6 87.0 91.2 95.6 99.8 100.0 53.0 68.5 87.0 90.1 94.4 99.7 100.0 52.7 69.0 87.4 90.7 94.5 99.2 100.0 0.0334 .0326 .0319 .0309 .0308 .0320 .0325 0.0246 .0269 .0265 .0263 .0256 .0248 .0254 0.756 .958 1.205 1.211 1.266 1.388 1.391 0.744 1.053 1.326 1.353 1.370 1.409 1.447 J5X10- 6 ' 1.32X10- 5 1.96X10- 5 Mol. Cone. Viscosity and Fluidity. Temperature Coefficients (15 25 25/4 "25 *25 0.25 1 . 1462 0.03561 28.08 0.10 1 . 1370 .03432 29.14 Solv. 1.1213 .03286 30.43 80 Studies on Solution. TABLE 37. Ammonium Formate in Formamid. Temperature Coefficients of Conductivity. V Molecular Conductivity. Dissociation. Per cent. Conductivity Units. 15 25 35 15 25 35 15-25 25-35 15-25 25-35 10 18.82 24.22 29.99 79.9 80.1 80.9 0.0287 0.0238 0.540 0.577 50 21.86 28.09 34.68 91.2 92.9 93.0 .0285 .0234 .623 .659 200 23.24 29.82 36.82 97.0 98.7 99.4 .0281 .0234 .653 .700 400 23.97 30.21 37.04 100.0 100.0 100.0 .0260 .0226 .624 .683 A'=2.01X10~ 5 2.51 X10- 5 3.34 X10~ 5 Viscosity and Fluidity. Temperature Coefficients (25 *>35 15-25 25-35 0.25 1 . 1324 0.04734 0.03544 0.02829 21 .12 28.22 35.35 0.0336 0.0253 0.10 1.1310 . 04497 . 03403 . 02720 22 .24 29.39 36.76 .0322 .0251 Solv. 1.1303 .04389 .03332 .02640 22 .78 30.01 37.88 .0317 .0262 TABLE 38. Sodium Formate in Formamid. Y Molecular Conductivity. Dissociation. Temperature Coefficients of Conductivity. Per cent. Conductivity Units. 15 25 35 15 25 35 15-25 25-35 15-25 25-35 2 4 10 50 100 200 400 10.32 12.76 15.22 17.35 18.07 18.50 18.45 13 16 19 22 23 24 24 .67 .48 .91 .64 .61 .15 .01 17.24 21.46 25.00 28.44 29.62 30.29 30.21 tf=0. 55.8 68.9 82.1 93.8 97.6 100.0 56.6 68.2 82.4 93.8 97.7 100.0 56.9 70.8 82.5 93.8 97.7 100.0 0.0324 .0302 .0309 .0305 .0305 .0300 0.0261 .0290 .0256 .0255 .0254 .0254 0.335 .372 .469 .529 .553 .565 0.357 .498 .509 .580 .601 .614 BX10- 6 1. 03X10-* 1.27X10- 5 Mol. Cone. Viscosity and Fluidity. Temperature Coefficients (25/4 7715 7725 1735 ,15. ,28- ^ 15-25 25-35 0.50 0.25 0.10 Solv. 1 . 1469 1 . 1393 1 . 1345 1.1314 0.05869 .05166 .04672 .04403 0.04299 .03812 .03510 .03338 .03348 .03037 .02798 .02665 17.04 19.36 21.40 22.71 23.26 26.23 28.49 29.96 29.87 32.93 35.74 37.52 0.0365 .0355 .0331 .0319 0.0284 .0255 .0254 .0252 Conductivities and Viscosities in Formamid and in Mixed Solvents. 81 TABLE 39. Lithium Formate in Formamid. Molecular Conductivity. Temperature Coefficients of Conductivity. Dissociation. V Per cent. Conductivity Units. 15 25 35 15 25 35 15-25 25-35 15-25 25-35 4 10.42 13.47 16.83 56 5 56.6 57.0 0.0293 0.0249 0.306 0.335 10 13.00 16.82 20.97 70. 9 70.9 71.0 .0293 .0246 .381 .416 50 16.53 21.22 26 49 89 6 89.4 89.8 .0283 .0248 .438 .449 100 17.24 22.31 27 56 93 4 94.4 93.4 .0293 .0235 .507 .525 200 17.79 22.90 28.35 96. 4 96.3 96.2 .0287 .0238 .546 .511 400 18.03 23.25 28.88 97 7 98.1 97.8 .0289 .0242 .454 .564 800 18.26 23.51 29.18 99 98.9 98.9 .0287 .0241 .548 .567 1600 18.44 23.55 29 50 100 100.0 100.0 .0289 .0232 .533 .572 K =0.54 X10~ 5 0.71 X10~ 5 0.87 X10~ 6 Mol. Viscosity and Fluidity. Temperature Coefficients (25 ^35 15-25 25-35 0.5 1 . 1399 0.05680 0.04224 0.03358 17.61 23.67 29.78 0.0344 0.0258 0.25 .. 1 . 1358 .05091 .03787 .03043 19.64 26.41 32.86 .0345 .0256 0.15 1 . 1328 .04637 .03495 .02791 21.57 28.61 35.83 .0326 .0252 Solv. 1.1314 .04403 .03338 .02665 22.71 29.96 37.52 .0319 .0252 TABLE 40. Barium Formate in Formamid. Temperature Coefficients of Conductivity. V Molecular Conductivity. Per cent. Conductivity Units. 15 25 35 15-25 25-35 15-25 25-35 50 33.35 43.48 53.72 0.0296 0.0235 0.993 .024 200 37.26 48.87 60.52 .0311 .0238 1.161 .165 400 37.94 49.93 61.69 .0316 ,0236 1.199 .176 800 38.59 50.63 62.64 .0311 .0236 1.204 .201 1600 39.27 51.67 63.71 .0316 .0233 1.240 .204 #=0.77XKr 5 .99X10~ 5 1.24X10~ 5 82 Studies on Solution. TABLE 41. Strontium Formate in Formamid. Temperature Coefficients of Conductivity. V Molecular Conductivity. Dissociation. Per cent. Conductivity Unite. 15 25 35 15 25 35 15-25 25-35 15-25 25-35 Satui ated soli ition. 50 32.42 41.54 51.54 77.8 76.2 76.8 0.0281 0.0241 0.912 1.000 200 37.52 48.71 60.78 90.2 89.4 90.6 .0298 .0248 1.119 1.207 400 39.14 51.60 63.72 94.1 94.7 95.2 .0318 .0234 1.246 1.212 800 40.37 52.97 65.25 97.1 97.2 97.3 .0312 .0232 1.260 1.228 1600 41.59 54.48 67.05 100.0 100.0 100.0 .0310 .0231 1.289 1.257 K =0.54 X10- 5 0.71 X10- 6 0.87 XlO" 6 TABLE 42. Sodium Benzoate in Formamid. Molecular Conductivity. Dissociation. Temperature Coefficients of Conductivity. V Per cent. Conductivity Unite. 15 25 35 15 25 35 15-25 25-35 15-25 25-35 4 9.24 12.41 15. 73 57. 6 60.1 62.1 0.0353 0.0267 0.317 0.332 8 10.94 14.51 18. 37 68. 2 70.3 72.5 .0326 .0266 .357 .386 10 11.40 15.03 19. 15 71. 1 72.8 75.6 .0319 .0273 .363 .412 50 13.22 17.54 22. 25 82. 5 85.0 87.8 .0326 .0268 .432 .471 200 14.22 18.58 23. 46 88. 7 90.1 92.6 .0306 .0263 .436 .488 400 14.45 18.81 23. 67 90. 2 91.2 93.3 .0294 .0259 .436 .486 1600 16.02 20.62 25. 33 100. 100.0 100.0 .0287 .0223 .460 .471 #=0.6X10- 5 0.8 X10- 6 1.06 X10-* Mol. /~Vr/> Viscosity and Fluidity. Temperature Coefficients (??). uonc. Z>25/4 1/15 1/25 n35 *>15 35 15-25 25-35 0.25 1.1392 0.05492 0.04047 0.03164 18.21 24.71 31.61 0.0357 0.0279 0.10 1.1342 .04808 .03604 .02853 20.80 27.75 35.05 .0334 .0263 Solv. 1.1295 . 04402 .03319 .02678 22.72 30.13 37.34 .0326 .0239 Conductivities and Viscosities in Formamid and in Mixed Solvents. TABLE 43. Sodium-Meta-Brom-Benzoate in Formamid. 83 F Molecular Conductivity. Temperature Coefficients of Conductivity. Per cent. Conductivity Units. 15 25 35 15-25 25-35 15-25 25-35 10 50 10.47 14.95 13.92 19.50 K=Q. 17.69 24.73 8XHT 5 0.0330 .0313 1.06XK 0.0271 .0264 r 5 1.32) 0.345 .477 ). 15-25 25-35 c 0.10 Solv. 1.1395 1 . 1307 0.04943 .04409 0.03682 .03325 0.02947 .02655 20.23 22.68 27.16 30.08 33.93 37.66 0.0343 .0326 0.0249 .0252 TABLE 46. Sodium Salicylate in Formamid. V Molecular Conductivity. Temperature Coefficients of Conductivity. Per cent. Conductivity Units. 15 25 35 15-25 25-35 15-25 25-35 4 9.65 11.74 13.70 14.49 14.80 17.10 / 13.06 15.51 18.04 19.01 19.27 22.13 C=0.68X 16.56 19.79 22.95 24.12 24.40 27.26 10~ 6 0. 0.0353 .0321 .0317 .0312 .0300 .0295 83X10- 5 0.0267 .0276 .0272 .0269 .0266 .0232 1.06X10 0.341 .377 .434 .452 .447 .503 -6 0.350 .428 .491 .511 .513 .513 10 ... 50 200 400 1600 Mol. Cone. Viscosity and Fluidity. Temperature Coefficients (?). Z>25/4 7,15 i?25 i;35 *>15 y>25 *>35 15-25 25-35 0.25 0.10 Solv. 1.1417 1.1345 1.1306 0.05374 .04787 .04379 0.03988 .03571 .03317 .03136 .02859 .02648 18.61 20.89 22.84 25.08 28.00 30.15 31.89 34.98 37.76 0.0348 .0340 .0320 0.0272 .0249 .0252 Conductivities and Viscosities in Formamid and in Mixed Solvents. 85 TABLE 47. Sodium Benzene Sulphonate in Formamid. V Molecular Conductivity. Temperature Coefficients of Conductivity. Per cent. Conductivity Units. 15 25 35 15-25 25-35 15-25 25-35 10 12.05 13.90 14.52 14.76 16.40 15.87 18.23 19.01 19.04 20.90 K>0.8X 20.21 23.06 24.00 24.39 26.82 10~ 5 1.0 0.0317 .0301 .0309 .0290 .0274 5X10- 5 0.0273 .0265 .0262 .0280 .0283 1.32X10- 0.382 .433 .449 .428 .450 6 0.434 .483 .499 .547 .592 50 200 400 1600 Mol. Cone. Viscosity and Fluidity. Temperaturc Coefficients (v>). >25/4 7715 i;25 1735 V>15 *>25 *>35 15-25 25-35 0.10 Solv. 1 . 1360 1.1306 0.04727 .04379 0.03554 .03317 .02836 .02648 21.17 22.84 28.14 30.15 35.26 37.76 0.0329 0.0253 .0320 .0252 TABLE 48. Sodium Succinate in Formamid. Temperature Coefficients of V Molecular Conductivity. Dissociation. Conductivity. Per cent. Conductivity Units. 15 25 35 15 25 35 15-25 25-35 15-25 25-35 10 21.72 28.71 36. 59 62.2 64.3 66.6 0.0321 0.0274 0.69S 0.788 50 29.54 38.82 49. 04 84.4 87.0 89.2 .0314 .0263 .928 1.012 200 32.56 42.67 53. 84 90.3 95.6 97.9 .0308 .0252 1.011 1.117 400 33.20 43.32 54. 24 90.5 97.1 98.6 .0304 .0252 1.012 1.092 1600 34.88 44.59 54. 95 100.0 100.0 100.0 .0298 .0233 .979 1.036 K =0.6X10- 0.8 X10~* 1.06X10" 5 Mol. Viscosity and Fluidity. Temperature Coefficients (). Cone. D25/4 7,15 T725 ,35 V>15 *>25 *35 15-25 25-35 0.10 1.1381 0.05254 0.03907 .03110 19.03 25.60 32.15 0.0345 0.0256 Solv. 1 . 1295 .04402 .03319 .02678 22.72 30.13 37.34 .0326 .0239 Studies on Solution. TABLE 49. Tetraethylammonium Iodide. In 7 per cent formamid and 25 per cent ethyl alcohol. Specific conductivity 25. Formamid, 14 XH)- 6 . Ethyl alcohol, 4.1 X10~ 7 . V Molecular Conductivity. Temperature Coefficients of Conductivity. Per cent. Conductivity Units. 15 25 35 15-25 25-35 15-25 25-35 4 13.31 19.96 23.38 24.64 24.75 24.96 24.82 tf = 1.0 17.13 25.70 29.76 31.44 31.69 31.95 31.95 XIO- 5 !.! 21.14 31.72 36.86 38.91 39.15 39.53 39.24 J5X10-* 0.0287 .0287 .0274 .0276 .0280 .0280 0.0234 .0234 .0328 .0237 .0236 .0237 0.382 .374 .638 .680 .694 .699 0.401 .602 .710 .747 .746 .758 10 50 100 200 400 1600 1. 57X10-* Mol. Cone. Viscosity and Fluidity. Temperature Coefficients fa). Z>25/4 Tjl5 1725 i;35 ^15 *>25 V>35 15-25 25-35 0.25 0.10 Solv. 1.0436 0.03639 0.02763 1.0330 .03469 .02661 1.0260 .03389 .02577 02228 02132 02066 27.48 28.83 29.51 36.19 37.72 38.80 44.88 46.90 48.40 0.0317 .0308 .0315 0.0240 .0243 .0247 In 50 per cent formamid and 50 per cent ethyl alcohol. > Molecular Conductivity. Temperature Coefficients of Conductivity. Per cent. Conductivity Units. 15 25 35 15-25 25-35 15-25 25-35 4 18.80 22.23 27.24 29.49 29.96 31.00 X-l 23.29 27.92 34.01 36.71 37.24 38.64 .01x10- 28.59 33.94 41.30 44.67 45.26 46.95 ' 1.24 X 0.0239 .0250 .0248 .0245 .0245 .0246 10- 6 1.4( 0.0227 .0219 .0214 .0215 .0215 .0215 >xio-^ 0.449 .559 .677 .722 .734 .765 0.530 .613 .729 .796 .802 .831 10 50 ........... 100 200 ....... . 400 . Mol. Cone. Viscosity and Fluidity. Temperature Coefficients (). D25/4 1715 1725 1735 >15 *>25 ^35 15-25 25-35 0.25 .10 .02 Solv. 0.9570 0.02666 0.02086 C .9437 .02571 .02002 .9369 .02494 .01953 .9346 .02488 .01939 .01707 .01631 .01597 .01580 37.51 38.90 40.10 47.94 49.95 51.20 58.58 61.31 62.62 0.0278 .0284 .0277 .0283 0.0222 .0227 .0223 .0227 Conductivities and Viscosities in Formamid and in Mixed Solvents. 87 TABLE 49. Tetraethylammonium Iodide Continued. In 25 per cent formamid and 75 per cent ethyl alcohol. Molecular Conductivity. 15 25 35 Temperature Coefficients of Conductivity. Per cent. 15-25 25-35 Conductivity Units. 15-25 25-35 4 10 50 100 200 400 800 1600 Saturated solution. 22.28 29.21 32.43 34.08 36.04 36.72 38.13 27.86 33.92 35.56 39.41 .41.34 43.79 44.46 46.08 42.61 47.17 49.36 52.29 53.15 54.90 0.0219 .0219 .0215 .0213 .0215 .0210 .0209 0.0208 .0198 .0194 .0194 .0194 .0195 .0191 0.482 .635 .698 .726 .775 .774 .795 0.560 .705 .776 .802 .850 .869 .882 1.56 X10~ 6 1.80 X10- 5 1.99 XIO" 6 Mol. Cone. Z)25/4 Viscosity and Fluidity. 7J25 7735' Temperature Coefficients (25/4 rj!5 r;25 1.0417 1.0300 1.0257 0.03468 .03394 .03366 0.02657 .02598 .02573 7725 0.02147 .02086 .02070 28.84 29.46 29.71 37.64 38.49 38.87 46.58 47.94 48.31 Temperature Coefficients ( 15-25 C 25-35 c 0.25 .10 Solv. 0.9763 .9515 .9345 0.02790 0.02190 .02572 .02031 .02471 ! .01934 0.01769 .01647 .01577 35.84 38.88 40.47 45 . 66 49.24 51.71 56.53 60.72 63.41 0.0274 .0266 .0278 0.0238 .0233 .0226 In 25 per cent formamid and 75 per cent ethyl alcohol. Molecular Conduct ivity. 15 25 35 Temperature Coefficients of Conductivity. Per cent. 15-25 25-35 Conductivity Units. 15-25 25-35 4 10 50 100 200 400 1600 20.92 24.75 30.49 32.80 34.06 35.13 36.69 25.27 29.98 37.06 39.85 41.44 42.75 44.54 30.03 35.61 44.05 47.40 49.38 51.02 53.28 0.0208 .0211 .0215 .0215 .0216 .0216 .0214 0.0188 .0187 .0188 .0189 .0191 .0193 .0194 0.436 .523 .627 .705 .738 .762 .785 0.476 .563 .699 .755 .793 .826 .878 K =0.425 X10~ 5 0.515 X10~ 5 0.608 XIO" 6 Mol. Cone. Viscosity and Fluidity. 7715 j25 1735 V?35 C Temperature Coefficients () 15-25 C 25-35 c 0.25 .10 Solv. 1.0353 1 . 0292 1.0249 0.03800 .03522 .03373 0.02879 .02711 .02576 0.02312 .02169 .02091 26.32 28.39 29.65 34.73 36.89 38.82 43.25 46.10 47.87 0.0320 .0299 .0309 0.0245 .0252 .0232 In 50 per cent formamid and 50 per cent ethyl alcohol. Molecular Conductivity. lo c 25 C 35 C Temperature Coefficients of Conductivity. Per cent. 15-25 25-35 Conductivity Units. 15-25 25-35 10. 50. 200. 400. 1600. 20.95 23.72 25.87 26.41 27.74 24.74 29.57 32.09 33.04 35.11 29.92 35.87 39.03 40.26 43.35 0.0180 .0204 .0240 .0251 .0265 0.0209 .0214 .0216 .0218 .0234 0.379 .485 .622 .663 .737 0.518 .630 .694 .722 .824 #=0.75X10- 5 0.93 X10- 5 1.21 XHT 6 Mol. Cone. Viscosity and Fluidity. D25/4< 7,15' r?25 c Temperature Coefficients (). 15-25 25-35 0.25 .10 Solv. 0.9457 .9392 .9344 0.02861 .02623 . 02472 0.02219 .02083 .01932 0.01809 .01674 .01575 34.95 38.12 40.45 45.07 48.01 51.76 55.28 59.74 63.49 0.0289 .0259 .0280 0.0227 .0244 .0227 90 Studies on Solution. TABLE 51. Lithium Nitrate Continued. In 25 per cent formamid and 75 per cent ethyl alcohol. Molecular Conductivity. lo c 25 35 C Temperature Coefficients of Conductivity. Per cent. 15-25 25-35 Conductivity Units. 15-25 25-35 10 50 200 400 1600 19.61 25.49 28.57 29.64 30.63 23.81 30.85 34.84 36.23 38.24 28.12 36.82 41.53 43.27 46.53 0.0214 .0210 .0219 .0222 .0248 0.0181 .0193 .0192 .0194 .0217 0.420 .536 .627 .659 .761 0.431 .597 .669 .704 .829 0.74 X10~ 5 0.94 X10~ 5 Mol. Cone. Viscosity and Fluidity. D25/4 r;35 c Temperature Coefficients (25/4 C i;25 e i?35 Temperature Coefficients (25 *>35 Temp. Coeff. (*>). 25-35 In 75 per cent glycerol with water f 0.5 .25 0.3092 .3184 0.1945 .1983 3.234 3.141 5.141 5.043 0.590 .606 1 .10 .3242 .2016 3.085 4.960 .608 ( Solv. .3303 .2207 3.028 4.531 .496 In 50 per cent glycerol with water {0.50 .25 0.05974 .06127 0.04336 .04477 16.74 16.32 23.06 22.34 0.378 .369 .10 .06189 . 04478 16.16 22.33 .388 Solv. .06255 .04536 15.99 22.05 .379 In 75 per cent glycerol with water. . 0.50 1 .25 0.02019 .02052 0.01597 49.53 48.73 62.62 0.264 .10 .02063 .01619 48.47 61.77 .274 Solv. .02070 .01260 48.31 61.73 .278 Carnegie Inst. Wash. Pub. No. 180. 97 98 Studies on Solution. tivity of these salts being materially increased in concentrated solutions on account of the greater fluidity of the solution. In order to complete the series of salts lowering the viscosity of these solvents, caesium compounds remained to be measured. A supply of caesium carbonate was finally obtained and converted into the nitrate and chloride. The viscosities of these salts have already been measured in water and in mixtures of water with methyl alcohol, ethyl alcohol, and acetone. Tables 57 and 58 give similar results in mixtures of glycerol and water. The viscosities were measured in the apparatus described in the preceding chapter on formamid. TABLE 58. Viscosity and Fluidity of Ccesium Nitrate. < Mol. Cone. r?25 7735 *>25 >35 Temp. Coeff. (if). 25-35 In 75 per cent glycerol f 0.25 0.3089 0.1933 3.237 5.173 0.598 with water .10 .3207 .1990 3.118 5.025 .612 1 Solv. .3303 .2207 3.028 4.531 .496 In 50 per cent glycerol with water f 0.50 j .25 0.05774 .06043 0.04223 .04407 17.32 16.55 23.68 22.69 0.367 .371 .10 . 06149 . 04443 16.26 22.51 .384 [ Solv. .06255 .04536 15.99 22.05 .379 In 25 per cent glycerol with water f 0.50 .25 0.01981 .02031 0.01566 .01611 50.48 49 . 24 63 . 86 62.07 0.265 .261 ) .10 .02056 .01615 48.64 61.92 .273 1 Solv. .02070 .01615 48.31 61.92 .282 It will be seen from tables 57 and 58 that caesium salts decrease the viscosities of glycerol-water mixtures, the decrement being greater, however, than in the case of rubidium salts. It should also be noted that when salts of both metals increase the viscosity of a solvent, as in the case of certain mixtures of water with acetone and the alcohols, the caesium salts produce a smaller increment than rubidium salts. CHAPTER IV. A STUDY OF THE ELECTRICAL CONDUCTANCE OF THE SODIUM SALTS OF CERTAIN ORGANIC ACIDS IN ABSOLUTE ETHYL ALCOHOL AT 15, 25, AND 35. BY H. H. LLOYD AND A. M. PARDEE. INTRODUCTION. In the Johns Hopkins laboratory, for some years past, a compre- hensive study has been made of the electrical conductance and dissoci- ation of various organic acids in aqueous solution. 1 This work was extended to absolute-alcohol solutions by Wightman, Wiesel, and Jones, 2 and by Lloyd, Wiesel, and Jones. 3 These investigators were unable to obtain, or even to approach, experimentally A , the equiva- lent conductance at zero concentration. The authors have therefore investigated the behavior of the sodium salts of the organic acids in absolute alcohol in order to obtain first the A values for these salts and then, by substitution in the Kohlrausch equation, 4 the A values for the acids themselves. The writers are interested also in the accumulation of accurate conductance data, as well as in such questions as tempera- ture coefficients of conductance, conductance in relation to chemical constitution, limits of experimental accuracy in working with dilute solutions in absolute alcohol, and the general phenomenon of alco- holysis. HISTORICAL. The measurement of the electrical conductance of the sodium salts of organic acids in absolute alcohol up to the present time has received but scant attention. With few exceptions, all investigations were incidental in nature and the compounds studied were chosen simply as types of organic salts. Dutoit and Rappeport, 5 in a study of the limiting conductances of some electrolytes in absolute alcohol, measured sodium acetate, among other salts, evidently taking the same as an example of the salts of organic acids. They subjected their results to some rather inter- esting deductions, but their conductances were measured at 18, mak- ^arnegie Inst. Wash. Pub. No. 170, part n; No. 210, chap. n. 2 Carnegie Inst. Wash. Pub. No. 210, chap, in; Journ. Amer. Chem. Soc. 36, 2243 (1914). 'Carnegie Inst. Wash. Pub. No. 230, chap. VH; Journ. Amer. Chem. Soc. 38, 121 (1916). 4 W, Ostv aid : Zeit. physik. Chem., 2, 561 (1888) ; 3, 170 (1889) ; Amer. Chem. Journ. 46, 66 (1914) . 6 Jour. chem. Phys. 6, 545 (1908). 99 100 Studies on Solution. ing exact comparison with those at 25 an impossibility. They inter- preted their results in a manner similar to that of Goldschmidt, and so their deductions are really illustrated in the latter's communication. Dhar and Bhattacharyya 1 carried on some work in alcohol with various salts and studied among others the following organic deriva- tives: sodium propionate, sodium benzoate, and sodium salicylate. Then* measurements at odd concentrations and temperatures render comparison impossible. Heinrich Goldschmidt, 2 incidental to his study of the esterification of organic acids in absolute alcohol, found it necessary to measure the conductances at 25 of a number of sodium salts of these acids. The salts were made by neutralizing the alcoholic solutions of the acids with an alcoholic solution of sodium ethylate. Goldschmidt measured the conductances from N/10 to N/5120 concentrations, and the values determined for five different salts are shown in tables 59 to 63. These results are given to enable us to discuss them and the deductions leading from them, as well as to point out later wherein we differ from him as to certain conclusions. These salts are sodium trichloroacet- tae, dichloroacetate, picrate, salicylate, and sulphosalicylate. There is appended to each table his calculation of A for the salt at specified dilutions. TABLE 59. Sodium Trichloroacetate. TABLE 60. Sodium Dichloroacetate. V Ai Ail 10 11.07 20 13.95 40 17.27 17.33 80 20.99 20.96 160 24.94 25.12 320 28.83 29.04 640 32.39 32.50 1280 35.28 35.29 2560 37.61 37.48 5120 39.23 38.92 A* 1(320-1280) = 46.10 An ,, A _M= 46.20 (1280-5120) 4 Mean A V Ai An |Ain 10 Q RS 20 12^64 40 ie!ii 15. 95 15.86 80 19.78 19.59 19.53 160 23.78 23.65 23.54 320 28.00 27.70 27.52 640 31.96 31.51 31.49 1280 35.66 34.87 34.96 ^2560 38.42 37.74 38.02 .5120 40.71 40.86 A (320- 1280) = 48 -54 * ' AQ AO ^0 (640-2560) ~^'^ An /oon loom Tr .UO I 46 (640-2560) A (1280-5120) A (320-1280) AO (640-2560) A (1280-5120) = 48.36 = 50.64 = 47.36 49.14 50.90 ii in Value An = , *Zeit. anorg. Chem. 82, 357 (1913). 2 Zeit. physik. Chem. 89, 129 (1914) ; 91, 46 (1916). Electrical Conductance in Absolute Ethyl Alcohol. 101 TABLE 61. Sodium Salicylate. TABLE 62. Sodium Sulphosalicylate. V A 10 9.57 20 12.21 40 15.27 80 18.78 160 22.67 320 26.58 640 30.14 1280 33.20 2560 35.48 5120 36.29 V Ai An Mean. 40 13.50 13.54 13.5 80 16.72 16.74 16.7 160 20.21 20.18 20.2 320 23.76 23.69 23.7 640 27.06 27.02 27.0 1280 30.0 30.03 30.0 2560 32.22 32.23 32.2 5120 33.84 34.12 34.0 A O (320- 1280) "fi*- 8 AO (640-2560) = jfg- A (1280-5120) =41.55 Most probable value = 44.5 A (320-1280) ~ 2? 'I AO (640-2560) ~41. 1 A ( 1280-5120) =40.8 An =40.9 TABLE 63. Sodium Picrate. V Ai An 40 18.04 18.14 80 22.06 22.11 160 26.34 26.34 320 30.61 30.64 640 34.59 34.59 1280 37.94 38.07 2560 40.43 40.65 5120 42.03 42.75 L (320-1280) l (640-2560) L ( 1280-5120) = 50.421 = 50.10 h 48.99 J ^0(320-1280) = ^*^ 1 A (640-2560) =50.97 II A ( 1280-5120) ~ OU '' Z J Selected value = 51 Goldschmidt thought that it was evident, after carrying his dilutions to 5,120 liters, that A could not be reached by ordinary experimental methods. He attempted to calculate A for these organic salts and expected to obtain the relative velocity of the organic anion from the salt and introduce the same into the equation To determine A for the organic salt he made use of the Kohlrausch formula 1 in which A is the unknown conductance at infinite dilution, A the conductance at a known dilution v, and a an unknown constant. Two equations involving the use of different A values are equated, the A being the same in both cases, and the expression solved for the value a. Once having this, it is a simple matter to solve for A in one of the two Ann. 26, 161 (1885). 102 Studies on Solution. TABLE 64. KI in Abso- lute AlcoholConductances in mhos at 25. original equations. By reference to the tables quoted above we can observe how such values are derived. It is to be noticed that alter- nate A values are equated. This is done so that the difference may be of sufficient degree of magnitude and that any inaccuracy in an individual measurement may not affect two successive derivations. A glance at the tables and calculations will show that the calculated values of A are by no means concordant. The higher the value of A used in the equation, the lower becomes the calculated A . His final conclusions are vague and inconclusive. The value chosen for A must be regarded as only approximate; it was usually the highest possible. Goldschmidt seems to have overlooked the very exact and admirable piece of work done on the subject of the limiting conductance and degree of ionization of alcoholic solutions by B. B. Turner 1 in the Johns Hopkins laboratory. Turner carried his dilutions to far greater limits, as table 64 illus- trates. We have repeated this work and have every reason to believe that it is unquestioned and is remarkably accurate, especially when one considers that it was done without the more re- cent conductivity apparatus now at our disposal. Turner showed that up to 5,000 liters dilution it is easy to obtain concordant results; but the values for A as calculated according to the Kohlrausch method are not constant for these higher dilutions. Like those of Goldschmidt, they decrease the higher the values of A used in the equation. Turner also showed that plot- ting A against the reciprocal of the cube root of the volume does not give a straight line as in aqueous solutions of equal dilutions, but rather a smooth curve slightly convex towards the dilution axis. He there- fore assumed that the Kohlrausch method fails to answer the require- ments of absolute alcoholic solutions. Extrapolation of his results with the formula would give us a value of 56 for A instead of the experi- mental value of 48.5 obtained. He thought that accidental introduc- tion of water into his solutions might affect the readings, and to test this he added as much as 0.2 to 0.3 per cent of water by weight to his alcoholic solutions, with a variation in conductivity of only 0.01 X 10~ 6 units, showing that no accidental experimental error of this nature had crept hi. Furthermore, Dutoit and Rappeport 2 showed identically the same phenomenon with a number of inorganic salts in work to which reference has already been made (page 99). This work, like that of Turner, V A 10 22.2 12 23.0 16 24.1 32 27.5 64 31.1 128 35.0 250 38.2 500 41.4 1000 44.0 5000 47.8 10000 48.4 20000 48.5 00 48.5=*=0.5 lAmer. Chem. Journ. 40, 558 (1908). 2 Journ. chem. Phys. 6, 545 (1908). Electrical Conductance in Absolute Ethyl Alcohol. 103 seems to have escaped the notice of Goldschmidt, as he does not mention either piece of work in any of his papers. In other words, the problem as undertaken by Goldschmidt is very incomplete from this standpoint. No reason can be given why he should use arbitrarily chosen limits for v in applying the Kohlrausch formula, nor is it shown how accurately measured conductances up to 20,000 liters dilution can be reconciled with such a falling-off in the calculated A for the salt. Whether such a method could be applied or not, or whether another can be substituted in its place, is a question of very great importance. Furthermore, Goldschmidt based his conclusions on the results of only six or seven salts. It was therefore deemed advisable by the present writers, in the first place, to obtain more conductance data on a larger number of salts, and, in the second place, to make these measure- ments at several temperatures in order to look at this subject in a broad way. EXPERIMENTAL. REAGENTS. The alcohol used in this investigation was prepared in the following manner: Ordinary 95 per cent ethyl alcohol was heated for several days with lime in a copper tank with a glass condenser attached. A minimum of refluxing in the condenser was obtained by inserting into the tank through the stopper a coil of 3/16-inch lead-pipe containing running water and serving to cause condensation immediately below the reflux tube. The alcohol was distilled off, using a glass still-head with a bulb blown in it and containing glass wool soaked in alcohol in order to prevent any dusting over of the dry calcium hydroxide. The middle fraction was treated in the same manner as above and again fractionated. This process was continued until a specific gravity of 0.78507 was obtained, the extreme limits of variation being 0.78505 to 0.78510, which, according to Circular No. 19 of the Bureau of Stand- ards, corresponds to a purity of from 100 to 99.987 per cent. The specific conductance of the alcohol varied with the different samples from 0.46 to 1.6X10" 7 mhos. Upon the final distillation the alcohol was collected in a 6-liter alcohol-extracted Jena bottle with a sealed stopper carrying a siphon for drawing off the liquid, a calcium chloride- soda lime tube, and an adapter with a ground-glass stopcock. Alcohol prepared and stored in this manner, after several days following the distillation, remained practically unchanged as to its conductance for a period of several weeks. It was found that our discarded alcoholic solutions and washings, when distilled once in a glass vessel with a few drops of concentrated sulphuric acid before the final lime treatment, produced a very superior grade of "absolute" alcohol, being generally better than that obtained from fresh supplies of the 95 per cent material. 104 Studies on Solution. The organic salts used in this investigation were prepared by adding the necessary amount of sodium ethylate in absolute alcohol to the organic acid hi alcoholic solution, as advised by Goldschmidt and pre- viously mentioned hi the historical section (page 100). The acids employed were taken from the various samples purified in the work of Lloyd, Wiesel, and Jones. When such were lacking new material was obtained from well-known firms and purified in the following man- ner: Whenever possible the acid was recrystallized from hot absolute- alcoholic solution, but when necessary a small amount of water was added. In every case the fractionation was carried out several times. The halogen-substituted aliphatic acids were fractionally crystallized from hot benzol, placed in a sulphuric-acid desiccator, and the final traces of benzol were removed by introducing into the container pieces of paraffine, which acted as an absorbent for the solvent. To purify the liquid aliphatic acids we resorted to both fractional crystallization by means of a refrigerant and repeated distillations under reduced pres- sure, hi the latter case collecting the various fractions in a specially constructed receiver for small quantities. The ethylate was prepared as needed in the following manner, as suggested by J. H. Shrader: 1 A special grade of metallic sodium, free from other metals, was wiped carefully with filter paper, the approx- imate amount was pared to fresh surfaces, and in small pieces was put first in a good grade of alcohol, then transferred into some conduc- tivity alcohol for final washing, and finally dropped into a measuring flask of the best alcohol, so that upon solution it could be made up to the mark. With practice it was possible to estimate successfully the amount of sodium to produce a nearly N/10 solution. This solution was standardized and used within an hour or two for the salt preparation. It was found necessary to use the ethylate immediately, as evidences of decomposition giving a straw color to the solution appeared within 24 hours of its preparation, and even sooner in the case of more concentrated solutions. This ethylate solution was immediately standardized by means of an N/10 aqueous solution of hydrochloric acid. This latter reagent was prepared by the method of Hulett and Bonner, 2 lately extended by Hendrixson. 3 As a check on this solution four series of silver chloride gravimetric analyses were made at various times throughout the year, none of which varied more than 0.1 per cent. Phenolphthalein served as the indicator for the various titrations, special precautions noted in a later paragraph being used to prevent the interference of carbon dioxide from the atmosphere. As a final proof of the correctness of our choice of indicators, the ethylate was standardized with hydrochloric acid, using in this case methyl red as *J. H. Shrader: Dissertation, Johns Hopkins University 14-16 (1913). 2 Journ. Amer. Chem. Soc. 31, 390 (1909). Mourn. Amer. Chem. Soc. 37, 2352 (1915). Electrical Conductance in Absolute Ethyl Alcohol. 105 an indicator, and it showed results concordant with the phenolphtha- lein values previously obtained. The methyl red naturally was use- less in the titration of most of the organic acids, so its use was aban- doned after proving the value of the phenolphthalein procedure. In order to dry completely our various pieces of apparatus, acetone was used, as suggested by Barnebey. 1 The acetone was dehydrated over calcium chloride and then redistilled. APPARATUS. The cylindrical type of conductivity cells was used in all save the more concentrated solutions, where the ordinary plate type was adopted. The reason for using the cylindrical cell lies in the fact that the organic salts in absolute alcohol, although having greater conductance than the organic acids, are nevertheless of sufficient resistance to warrant such a procedure. White 2 and Wightman 3 have described the method for obtaining the constants of these cells. Both the temperature coefficients of expansion of alcohol and the temperature coefficients of conductance of substances in it as a solvent are so large that it was especially necessary to maintain the solutions at a constant temperature to within 0.01. The thermometers were of the differential Beckmann type and were carefully compared with a standard Reichsanstalt instrument which had in turn been calibrated at the Bureau of Standards. The combined gas-regulator and thermo- regulator was devised by Davis and Hughes. 4 The improved form of constant-temperature bath, as devised by Davis, 5 was used in our investigation. These baths are capable of even finer temperature adjustment than that stated above as employed in our work. The resistance-box used throughout this work was calibrated at the Bureau of Standards. The improved Kohlrausch slide-wire bridge was employed, by means of which it was possible to read distances on the slide wire corresponding to tenths of a millimeter (the total length of the wire being 5 meters). Special precautions were taken to remove all external resistance in the circuit. No. 10 B. & S. insulated copper wire was used, and all leads coming to the bridge were dipped into a mercury-contact rocking commutator. In the volumetric work Jena flasks were employed (50, 100, 200, 250, 500, 1,000 c.c.) which had been previously calibrated in this labora- tory and recalibrated by ourselves, using weight methods. Reichsan- stalt double-mark pipettes were recalibrated before use. In filling and draining the pipette the following device was suggested by Dr. Davis. It consisted of a right-angled T-tube with a glass stopcock on the base of the T, the pipette being attached by rubber to one end of 1 Journ. Amer. Chem. Soc. 37, 1835 (1915). 4 Zeit. physik. Chem. 85, 519 (1913). 'Amer. Chem. Journ. 42, 527 (1909). 'Carnegie Inst. Wash. Pub. No. 210, 21 (1914). 3 Amer. Chem. Journ. 44, 64 (1911). 106 Studies on Solution. the cross-piece, held vertically with the regulating finger on the opposite end of the cross-piece. The control finger is maintained throughout the operation at this opening and the danger of contamination by suction is removed. A tube filled with a mixture of calcium chloride and soda lime is inserted in the rubber tube leading from the glass stop- cock on the base of the T to the mouth, for obvious reasons. The 50 c.c. burettes adopted were calibrated at 2 c.c. intervals by weight. In order to titrate with phenolphthalein in an atmosphere free from carbon dioxide the following apparatus was constructed, partially as suggested by Hendrixson i 1 A carboy was connected to an ordinary tire pump and served as a gas reservoir. The air was led through three wash bottles, the first containing concentrated potassium hydrox- ide solution, the second a more dilute solution, and the third pure water. The titration was effected in an Erlenmeyer flask closed with a rubber stopper, which in turn was fitted loosely around the burette tip, serving in this way as a vent for the stream of air passed slowly through the solution. The difficulty in desiccating our acids when once purified was solved by means of a vacuum drying-oven designed by Dr. Davis and constructed with the help of the authors (see Chapter II). In this apparatus the lamp heating-unit maintained a temperature of 65 and an ordinary suction-pump kept a reduced pressure of 70 to 80 mm., so it is easily seen that with the added help of a strong dehydrating agent, such as sulphuric acid or phosphorus pentoxide, all traces of the crystallizing solvent could be removed, since water boils at about 47 at this pressure. In proof of this practically all the organic acids titrated theoretically. PROCEDURE. The sodium e thy late was standardized by titration with N/10 HC1 in a carbon-dioxide-free atmosphere, as described previously. When the ethylate was standardized, the organic acid from which the salt was to be made was weighed out in quantity sufficient to give 100 c.c. N/10 salt solution and this weight was confirmed by titration, which showed a very general concordance, giving added proof of the purity of the acids. In dealing with very deliquescent substances, as trichloracetic acid for example, we weighed by difference, making ap- proximate standard solutions rather than exactly N/10 strengths; but even in this case we obtained confirmation of our work. The non- deliquescent, crystalline acids were weighed on a watch crystal, the deliquescent ones in glass-stoppered weighing bottles; but in both cases the acids were washed through a funnel into the 100 c.c. measuring flasks with conductivity alcohol and made up to mark at 25. Several Uourn. Amer. Chem. Soc. 37, 2352 (1915). Electrical Conductance in Absolute Ethyl Alcohol. 107 salts of N/50 dilution were made up in this same manner at the begin- ning of our work, but this dilution was omitted later as unnecessary. Let us notice a few of the necessary steps in the titrations. All such were made in 70 c.c. solution (50 c.c. water, 10 c.c. acid, and approximately 10 c.c. ethylate). The carbon-dioxide-free air was allowed to bubble through the solution for 2 minutes before titration. It was found that the presence of some alcohol retarded the end-point and a number of titrations were made throughout the year to enable us to correct for this. We found as a result of our work : 70 c.c. water and c.c. alcohol required 0.03 c.c. to produce color. 60 c.c. water and 10 c.c. alcohol required 0.04 c.c. 50 c.c. water and 20 c.c. alcohol required 0.05 c.c. Therefore it was necessary to apply this correction, as our accuracy in titration was made to check to 0.02 c.c. After calculating the amounts necessary, 100 c.c. N/100 salt solution in absolute alcohol at 25 was prepared, placed in a 150 c.c. glass- stoppered Erlenmeyer flask, and sealed with rubber cement until the conductances were to be determined. It was possible to make up three or four different mother solutions of various organic salts in one day, another day being devoted to the dilution down to weaker con- centrations, measurement of the conductances, and calculation of results for each salt. These last three operations on a single salt at various dilutions we have designated as a "run." It is deemed advisable at this point to introduce an example of the calculations upon which a single salt was prepared as described above : Acid orthonitrobenzoic, C?H6O4N. Strength of standard HC1, 0.10027. I. Standardization of the Ethylate. 10.005 c.c. HC1 used in each titration. Ethylate burette. II. Standardization of the OrganicAcid. 10.005 c.c. acid used in each titration. Ethylate burette. Readings. Corrected. Difference. c.c. c.c. c.c. 2.77 2.76 10.84 10.83 8.07 10.85 10.84 .... 18.91 18.92 8.08 18.92 18.93 26.99 27.02 8.09 Readings. Corrected. Difference. c.c. c.c. c.c. 1.98 1.98 10.03 10.02 8.04 10.03 10.02 18.05 18.07 8.05 18.05 18.07 26.08 26.11 8.04 Mean 8.08 less 0.04 correction = 8. 04 c.c. ethylate. 10.005 :8.04 : : x : 0.10027 a: = 0.1248 normality of the ethylate. To make 100 c.c. N/100 salt solution requires 8.015 c.c. Mean 8.04 less 0.04 correction = 8.00 c.c. ethylate. 8.04 : 8.00 : : 0.10027 : x x = 0.09977 normality of organic acid. To make 100 c.c. N/100 salt solution requires 10.02 c.c. plus 0.01 c.c.; excess equals 10.03 c.c. 108 Studies on Solution. It should be mentioned that this work was carried on in a rather small room with one window and one door at opposite ends of the room, so that with care it was possible to keep the room temperature at 25 with less than 0.3 variation. Thus it was possible to measure out the solutions in burettes and pipettes, provided that such were not handled unnecessarily to cause heating and were always kept dry to prevent cooling in evaporation. All burettes and pipettes were con- nected with a tube filled with a mixture of calcium chloride and soda lime to prevent contamination from moisture and carbon dioxide. In handling the "run" the N/100 solution of one of the salts served as a basis for the preparation of all the more dilute solutions. The following scheme represents the method by which these solutions were prepared : N /,oo ZO^cto lOOCejsoln "Aoo iQcclto iOCcc|soM %ooo AQC.C tO lOOcc soln. 20c.c to iOOcc|so! r.oooo lOcc IOOc.c .lto Jsol After a number of experiments it was deemed inadvisable to wash the measuring flasks with water; they were therefore rinsed with a good grade of alcohol and then three tunes with conductivity alcohol. The cells were filled with conductivity water until several hours before use. They were then rinsed three times with good alcohol. Each cell was finally washed three tunes with the solution of the particular dilution to be "run" in that cell before filling. These cells, together with one containing the conductivity alcohol, were then introduced into the 15 bath, gently agitated twice within an hour's time to insure absence of bubbles as well as to hasten diffusion, and then read. They were placed successively in the 25 and 35 baths, allowing for the same time and procedure as in the 15 bath. It will be remembered that the solutions were made up at 25 and that the molecular conductances were measured at 15, 25, and 35. Alcohol has such an appreciable temperature coefficient of expansion that it was necessary to correct for the contraction and expansion at the other temperatures. One liter of alcohol at 25 expands to 1.01114 liters at 35 and contracts to 0.98923 liter at 15. Therefore, to obtain the molecular conductance at 35, one must multiply the specific con- ductance at that temperature by the product of the molecular volume and the factor 1 .01 1 14. Likewise, to obtain the molecular conductance at 15, the specific conductance at that temperature must be multiplied by the product of the molecular volume and the factor 0.98923. Electrical Conductance in Absolute Ethyl Alcohol. 109 MEASUREMENTS. EXPLANATION OF TABLES. In the following tables V signifies the volume at which a solution was made up, A the molecular conductance of that solution at the vari- ous temperatures. The method of calculating A is thoroughly familiar. Corrections were applied as described, allowing for the contraction and expansion of the solutions. (The solutions were so dilute that their volume changes with variation in temperature were assumed to be the same as that of pure alcohol.) The values of A 25, therefore, rep- resent the molecular conductance of a solution of volume V at 25. The values of A 15 and A 35, however, represent the molecular con- ductance of a solution of volume 0.98923 V at 15 and 1.01114 V at 35. Only the one value V is given in the tables to save space. All conduct- ances are expressed in reciprocal ohms. Concerning the calculation of the temperature coefficients of con- ductance, we have adopted this expression: where At' and At represent the molecular conductivities of the same solution at t' and f (t f >t), and T the temperature coefficient of con- ductance. To find the percentage coefficient of conductance we have used the formula T ' where A is the percentage coefficient and At the conductivity at the lower temperature. At first the values of At and At' at 15 and 35 were corrected for the difference in volume between 0.98923 V and F, and 1.01114 V and F, respectively. This was done in order that com- parison might be made between solutions of the same volume. Later this correction was omitted because of its small value. TABLE 65. Sodium Formate. V A25 A35 A25-35 100 20.09 22.70 1.30 250 25.03 28.53 1.40 500 28.48 33.25 1.67 1,000 32.62 38.13 .69 2,000 35.34 41.88 .85 5,000 37.75 44.67 .83 10,000 39.03 46.35 .88 20,000 39.76 47.24 .88 TABLE 66. Sodium Acetate. V A25 A35 A25-35 100 17.20 19.10 1.10 250 22.20 25.08 1.30 500 26.07 29.76 1.42 1,000 29.99 34.67 .56 2,000 32.80 38.54 .75 5,000 35.42 41.62 .75 10,000 36.36 42.84 .78 20,000 36.79 42.95 .67 110 Studies on Solution. TABLE 67. Sodium Chloroacetaie. TABLE 68. Sodium Dichloroacetate. V A15 A25 A35 415-25 A25-35 50 12.92 14.66 16.40 .35 .19 100 16.15 18.45 20.74 .42 .24 250 20.52 23.76 26.92 .58 .33 500 23.76 27.79 32.04 .70 .53 1,000 26.51 31.34 36.56 .82 .67 2,500 29.19 34.79 41.00 1.92 .81 5,000 30.73 36.81 43.83 1.98 .91 10,000 31.53 37.84 45.02 2.00 1.90 TABLE 69. Sodium Tnchloroacetate. V A15 A25 A35 A15-25 A25-35 100 19.03 22.05 25.13 1.59 .40 250 23.26 27.24 31.36 1.71 .51 500 26.18 30.94 36.10 1.82 .67 1,000 28.80 34.31 40.31 1.91 .75 2,000 30.36 36.40 43.04 1.99 .82 5,000 32.24 38.79 46.12 2.03 .89 10,000 33.02 39.84 47.39 2.07 .90 20,000 33.71 40.54 48.39 2.03 .94 TABLE 71. Sodium Propionate. V A15 A25 A35 A 15-25 A25-35 100 14.68 16.50 18.18 1.24 .02 250 18.80 21.41 23.95 1.39 .19 500 22.16 25.71 28.95 1.60 .29 1,000 25.13 29.38 33.85 1.68 .52 2,000 27.35 32.28 37.66 1.80 .67 5,000 29.26 34.73 40.64 1.87 .70 10,000 30.27 36.11 42.52 1.93 .77 20,000 30.52 36 . 23 42.41 1.87 .71 TABLE 73. Sodium Butyrate. V A15 A25 A'iS A 15 -25 A25-35 100 14 . 39 16.16 17.82 1.23 .03 250 18.50 21.03 23 . 49 1.37 .17 500 21.79 24.97 28.36 1.46 . 36 1,000 24.74 28.89 33.30 .68 .53 2,000 26.93 31.80 37.07 .81 .66 5,000 28.95 34.29 40.16 .84 .82 10,000 2Q.85 35 . 67 42.06 .95 .79 20,000 30.27 35 . 98 42.41 .89 .79 V A15 A25 A35 A15-25 A25-35 100 18.12 20.84 23.67 .50 1.36 250 22.38 26.06 29.91 .64 1.48 500 25.55 30.00 34.92 .74 1.64 1,000 28.36 33.61 39.40 .85 1.72 2,000 30.30 36.12 42.79 .92 1.85 5,000 32.16 38.62 46.01 2.01 1.91 10,000 33.08 39.77 2.02 20,000 33.97 40.67 48.71 1.97 1.98 TABLE 70. Sodium PhenylacetaU . F A15 A25 A35 A15-25 A25-35 50 11.14 12.44 13.60 .17 0.93 100 13.86 15.58 17.16 .24 1.01 250 17.96 20.51 22.91 .42 1.09 500 21.03 24.31 27.58 .56 1.35 1,000 24.09 28.23 32.36 .72 1.46 2,500 26.68 31.60 36.82 .84 1.65 5,000 28.06 33.46 39.31 .92 1.75 10,000 28.40 34.06 40.25 1.99 1.82 TABLE 72. Sodium B-iodopropion"t< . V A15 A25 A35 A15-25 A25-35 50 15.25 17.67 21.25 1.59 2.03 100 18.54 21.67 26.48 1.69 2 22 250 22.59 26.70 33.27 1.82 2.46 500 25.51 30.47 38.13 1.94 2.51 1,000 27.91 33 . 55 42.09 2.02 2.55 2,500 30.18 36 . 45 45 . 50 2.08 2.48 5,000 31.39 37.91 47.39 2.07 2.50 10,000 31.93 38 . 69 48.06 2.12 2.43 TABLE 74. Sodium Oxyixobulymlt F A15 A25 A35 A 15-25 A25-35 50 12.28 14.21 16.13 .57 1.35 100 15.39 17.87 20.32 .61 1.37 250 19.73 23.07 26.51 .69 1.49 500 22.74 26.81 31.11 .79 1.60 1,000 25.71 30 . 52 35 . 76 .87 1.72 2,500 28.07 33.46 39.52 .92 1.81 5,000 29.34 35.07 41.74 .95 1.90 10,000 30.23 36.05 42.97 .93 1.92 Electrical Conductance in Absolute Ethyl Alcohol. Ill TABLE 75. Sodium Benzoate. TABLE 76. Sodium Orthoamidobenzoate. V A15 A25 A35 A 15-25 A25-35 100 250 18.94 21.66 24.29 .45 1.21 500 22.11 25 . 66 29.21 .61 1.38 1,000 25 . 03 29 . 40 33.93 .75 1.54 2,000 27.00 31.99 37.39 .85 1.69 5,000 29.18 34.78 40.83 .92 1.74 10,000 30.01 35.80 42.20 .92 1.82 20,000 30.41 36.18 42.64 .90 1.79 TABLE 77. Sodium p-amidobenzoate. V A15 A25 A35 A15-25 A25-35 100 12.26 13.64 14.92 .13 0.94 250 16.30 18.50 20.34 .35 0.99 500 1,000 22.54 26.13 29.81 .59 1.41 2,000 24.87 29.20 33.80 .74 1.58 5,000 27.37 32.37 37.52 .79 1.63 1 10,000 28.33 (33.21) (39.58) (1.72) (1.92) 20,000 28.92 34.16 (40.01) (1-81) (1.71) TABLE 79. Sodium p-brombenzoate. F A15 A25 A35 A15-25 A25-35 100 15.71 17.94 20.11 .42 .21 250 19.78 22.83 26.00 .54 .39 500 22.84 26.76 30.89 .72 .51 1,000 25.37 30.02 34.94 .83 1.64 2,000 27.19 32.46 38.15 .94 .75 5,000 28.62 34.31 40.59 .99 1.84 10,000 29.41 (35.09) 42.05 20,000 29.84 35.75 42.73 2.01 1.89 TABLE 81. Sodium Melachlorobenzoate. V A15 A25 A35 A15-25 A25-35 100 15.53 17.69 19.82 .39 .20 250 19.74 22.76 25.80 .53 .34 500 22.83 26.62 30.59 .66 .49 1,000 25.60 30.18 35.15 .79 .65 2,000 27.38 32.68 38.36 .94 .74 5,000 29.55 35.40 41.81 .98 .81 10,000 30.51 36.49 43.40 .96 .89 20,000 31.01 37.03 43.87 .94 1.85 V A15 A25 A35 A 15-25 A25-35 100 13.19 14.84 16.37 .25 .07 250 17.24 19.60 21.86 .37 .15 500 20.76 23.94 26.99 .53 .27 1,000 23 . 90 27.82 31.92 .64 .47 2,000 26.17 30.81 35.75 .77 .60 5,000 28.44 33.54 38.91 .79 .60 10,000 29.29 34.71 40.68 .85 .72 20,000 29.52 34.82 40.50 .80 .63 TABLE 78. Sodium m-brombenzoatc >.. V A15 A25 A35 A15-25 A25-35 100 14.64 16.66 18.67 .38 1.21 250 18.52 21.35 24.20 .53 1.33 500 1,000 23.94 28.20 32.86 .78 1.65 2,000 25.55 30.38 35.80 .89 1.78 5,000 27.56 32.85 38 . 83 .92 1.82 10,000 28.30 33 . 93 40.20 .99 1.85 20,000 29.06 34.71 41.15 1.94 1.86 TABLE 80. Sodium Orthochlorobenzoate . V A15 A25 A35 A15-25 A25-35 100 14.54 16.49 18.34 1.34 .12 250 18.71 21.52 24.25 1.50 .27 500 21.90 25.44 29.09 1.62 .43 1,000 24.83 29.23 33.86 1.77 .58 2,000 26.85 31.96 37.42 1.90 .71 5,000 28.82 34.65 40.85 2.02 1.79 10,000 29.97 36.04 2.03 .... 20,000 30.26 36.76 43.59 2.15 1.86 TABLE 82. Sodium p-chlorobenzoaie. V A15 A25 A35 A15-25 A25-35 100 15.80 18.04 20.22 1.42 1.21 250 19.92 23.04 26.18 1.57 1.36 500 22.91 26.78 30.90 1.69 1.54 1,000 25.65 30.37 35.39 1.84 1.65 9 000 5,000 29.25 35.27 41.64 2.06 1.81 10,000 30.17 (36.24) 43.13 (2.02) (1.90) 20,000 30.33 36.80 43.42 2.13 1.80 112 Studies on Solution. TABLE 83. Sodium Salicylate. TABLE 84. Sodium m-hydroxybenzoate. V A15 A25 A35 A15-25 A25-35 50 14.19 16.32 18.42 .51 .29 100 17.19 19.87 22.58 .56 .36 250 21.52 25.13 28.84 .67 .48 500 24.56 28.92 33.60 .78 .62 1-,000 27.48 32.62 38.31 .87 .74 2,000 29.18 34.87 41.18 .95 .81 2,500 30.00 35.92 42.45 1.97 .82 5,000 31.20 37.47 44.57 2.01 .87 10,000 31.90 38.38 45.61 2.03 1.88 20,000 32.38 38.99 46.36 2.04 1.89 TABLE 85. Sodium p-hydroxybenzoatc. V A15 A25 A35 A15-25 A25-35 50 100 12.54 14.04 15.21 1.20 0.83 250 16.63 18.84 20.93 1.33 1.11 500 19.39 22.31 25.21 1.51 1.30 1,000 22.31 26.05 29.85 1.68 1.46 2,000 24.17 28.82 33.50 1.92 1.62 5,000 26.27 31.58 36.73 2.02 1.63 10,000 27.28 32.95 38.74 2.08 1.76 20,000 27.34 33.27 (38.69) 2.17 TABLE 87. Sodium lodosalicylate. V A15 A25 A35 A15-25 A25-35 50 100 250 500 1,000 2,000 2,500 17.66 21.93 25.07 27.48 29.24 20.47 25.67 29.55 32.67 34.95 23.44 29.59 34.49 36.28 41.23 .59 .71 .79 .89 .95 1.45 1.53 1.67 1.73 1.80 5,000 10,000 20,000 30.89 31.38 31.57 37.17 37.69 38.35 44.03 45.08 2.03 2.01 2.15 1.85 1.96 TABLE 89. Sodium Orthonitrobenzoate. V A15 A25 A35 A15-25 A25-35 50 11.81 13.28 14.62 .24 .01 100 14.79 16.73 18.59 .31 .11 250 18.94 21.76 24.45 .49 .24 500 22.13 25.69 29.34 .61 .42 1,000 24.87 29.26 33.82 .77 .56 2,500 27.60 32.83 38.51 .89 .73 5,000 29.10 34.89 41.29 .99 .83 10,000 30.04 36.07 42.89 2.00 .89 V A15 A25 A35 A 15-25 A25-35 50 100 250 500 1,000 2,000 2 500 13.61 17.64 23! 61 25.62 15.31 20.13 27161 30.27 16.97 22.51 3l!85 35.42 1.25 1.41 i!e>9 1.81 1.08 1.18 1^54 1.70 5,000 10,000 20,000 27.67 28.33 29.02 32.88 33.97 34.65 38.51 39.98 40.78 1.S8 1.99 1.94 1.71 1.77 1.77 TABLE 86. Sodium Acetylsalicylale. V A15 A25 A35 A15-25 A25-35 50 14.16 16.29 18.38 .50 .28 100 17.16 19.81 22.50 .54 .36 250 21.50 25.08 28.71 .67 .45 500 24.44 28.73 33.41 .76 .63 1,000 27.62 32.76 38.36 .86 .71 2,500 30.06 35.89 42.38 .94 .81 5,000 31.28 37.44 44.56 .97 .90 10,000 32.56 38.97 46.45 .97 .92 25,000 32.73 39.17 46.70 .97 .92 TABLE 88. Sodium Sulphosalicylaie. V A15 A25 A35 A15-25 A25-35 50 12.38 14.28 16.21 1.53 .35 100 15.30 17.73 20.29 1.59 .44 250 19.19 22.48 25.92 1.71 .53 500 21.92 25.90 30.22 1.82 .67 1,000 24.34 29.00 34.11 1.95 .76 2,000 26.12 31.27 37.11 1.97 .87 2,500 26.58 31.93 37.86 2.01 .86 5,000 28.18 33.88 40.42 2.02 .93 10,000 29.37 35.47 42.41 2.08 .97 20,000 30.65 37.13 44.16 2.11 .89 TABLE 90. Sodium m-nitrobenzoate. V A15 A25 A35 A15-25 A25-35 50 13.61 15.55 17.49 .43 .25 100 16.43 16.94 21.44 .53 .32 250 (20.13) (23.09) 26.52 ( .47) .49 500 23.32 27.54 31.95 .81 .60 1,000 26.16 31.08 36.44 .88 .72 2,500 28.41 33.89 40.05 .93 .82 5,000 29.52 35.21 42.12 .93 .96 10,000 29.80 35.66 42.95 .97 2.04 Electrical Conductance in Absolute Ethyl Alcohol. 113 TABLE 91. Sodium Paranitrobenzoate. TABLE 92. Sodium 2, 4, Dinitrobenzoate. V A15 A25 A35 A15-25 25-35 50 14.31 16.47 18.61 .51 1.30 100 17.22 19.96 22.72 .59 1.38 250 21 04 28 19 500 23.98 28.36 33.09 .83 1.67 1,000 26.59 31.69 37.30 .84 1.77 2,500 28.63 34.25 40.51 .96 1.83 5,000 29.75 35.66 42.53 1.99 1.93 10,000 30.70 36.98 44.10 2.05 1.93 TABLE 93. Sodium Orthotoluate. V A15 A25 A35 A15-25 A25-35 50 11.42 12.80 14.11 1.22 1.01 100 14.19 16.03 17.81 1.30 .11 250 18.32 21.07 23.67 .50 .23 500 21.28 24.69 28.16 .60 .41 1,000 24.53 28.80 33.24 .74 .54 2,500 27.17 32.19 37.52 .85 .66 5,000 28.63 34.23 40.23 .96 .75 10,000 29.56 35.36 41.75 .96 .81 25,000 TABLE 95. Sodium Paratoluate. F A15 A25 A35 A15-25 A25-35 50 11.25 12.53 13.74 1.14 0.96 100 14.05 15.79 17.43 1.24 1.04 250 18.11 20.68 23.00 .42 1.12 500 21.14 24.45 27.75 .57 1.35 1,000 24.15 28.29 32.55 .71 1.51 2,500 26.64 31.46 36.62 .82 1.64 5,000 28.00 33.29 39.21 .89 1.78 10,000 28.49 33.80 39.91 .86 1.81 40,000 V A15 A25 A35 A15-25 A25-35 100 18.20 21.09 24.09 1.59 .42 250 22.18 25.98 29.93 1.71 .52 500 24.95 29.45 34.39 1.80 .69 1,000 27.41 32.52 38.33 1.86 .79 2,000 28.74 34.42 40.88 1.98 .88 5,000 30.45 36.63 43.71 2.03 .93 10,000 31.21 37.66 44.94 2.07 .93 20,000 31.80 38.27 45.86 2.03 .98 TABLE 94. Sodium m-toluate. V A15 A25 A35 A15-25 A25-35 50 11.16 12.46 13.64 .16 0.97 100 14.16 15.92 17.57 .24 1.04 250 18.26 20.85 23.32 .42 1.18 500 21.42 24.75 28.17 .55 1.38 1,000 24.10 28.24 32.55 .72 1.53 2,500 26.60 31.52 36.74 .85 1.66 5,000 27.97 33.38 39.36 .93 1.79 10,000 28.68 34.24 40.33 .94 1.78 25,000 29.36 35.05 41.22 .94 1.76 TABLE 96. Sodium Picrate. V Alo A25 A35 A 15-25 A25-35 100 19.77 23.28 27.09 1.78 1.64 250 24.49 28.93 33.80 1.81 1.68 500 27.81 33.04 38.79 1.88 1.74 1,000 30.65 36.56 43.24 1.93 1.83 2,000 32.78 39.24 46.60 1.97 1.88 5,000 34.73 41.67 49.85 2.00 1.96 10,000 35.73 42.91 51.46 2.01 1.99 20,000 36.32 43.63 52.43 2.01 2.02 40,000 43.86 .... .... DISCUSSION OF RESULTS. The most apparent observation from tables 65 to 96 is the great similarity in amount of conductance of these organic salts in alcohol. At 25 in a 1,000-liter dilution the extreme limits for the conduct- ance are from 26 to 36 mhos, with an average value from 28 to 33. The obvious reason for this is the uniform effect of the sodium ion in the solution and the similarity in the velocities of the organic anions. As naturally expected, the conductances of these salts are much greater than those of the corresponding acids. Very little can be said as to the relation between chemical composi- tion and conductance. The aliphatic and aromatic derivatives show no 114 Studies on Solution. difference, and the conductance of the aromatic compounds seems to be independent of the position of the various substituent groups. Sodium picrate has a much larger conductance than any other salt, and the monosodiumsulphosalicylate at high dilutions gives abnormally large and increasing conductance values, due probably to the secondary ionization of the carboxyl group at these high dilutions. In discussing the temperature coefficient of conductance it is to be noticed that this value becomes gradually larger with increase in dilution, and at the highest dilutions approximates the value 0.0200. Just as in the conductance results, there is here no definite relation between the values for the temperature coefficient and chemical composition. It is of importance to note that this work on the sodium salts of the organic acids in absolute alcohol has been greatly restricted, owing to the almost complete insolubility of a great many of these salts in this solvent. If the work were carried out in alcohol which was not absolute, practically all the salts could be studied, for it is necessary to add only a very small amount of water to obtain a sufficient degree of solubility. We have approximately covered the field of available compounds. It is of interest to note that the polybasic acids of both the aliphatic and aromatic series are excluded from study for this reason, as well as all unsaturated acids of both series. A number of salts of aromatic acids with di- and tri-substitutions in the ring were likewise impossible to study. Reference has already been made (see pages 100-103) to the work of Heinrich Goldschmidt on the conductance of alcoholic solutions of sodium salts. We have purposely investigated most of the salts which he studied. A comparison of these results is conveniently made by reference to the following tables : Salt. Goldschmidt. Authors. Sodium dichloroacetate. . . Sodium trichloroacetate . . Sodium salicylate Sodium sulphosalicylatc . . Sodium picrate Table 60, p. 100 59, 100 61, 101 62, 101 63, 101 Table 68, p. 110 69, 110 83, 112 88, 112 96, 113 It can be seen from these tables that the two series of conductance values are in accordance, but an exact comparison can not be made because of the fact that the values of A in the two series refer to some- what different concentrations. In order to make an effective com- parison we have plotted the values of A against the logarithms of the volume V in the case of sodium trichloroacetate (see fig. 26). The points circled refer to the data of Goldschmidt and the crosses to Electrical Conductance in Absolute Ethyl Alcohol. 115 data obtained in the present work. With few exceptions all the points lie on one curve, and the slight deviations which occur are within the limits of error of the conductance method. The four other salts give similar results; therefore their graphs are omitted. 44- 4-0 36 32 28 A 24 20 16 LogV FIG. 26. Comparison of Conductance Values in Absolute Alcohol. o = Goldschmidt; x = Lloyd and Pardee. It has been found impossible to obtain, experimentally, a value for the limiting conductance, although measurements have been carried out to 10,000 and 20,000 liter dilutions. It is therefore necessary to determine A by some method of extrapolation. It will be recalled that Goldschmidt used the Kohlrausch formula for this purpose (see page 101), although its applicability to alcoholic solutions and even to aqueous solutions 1 had been previously questioned. We applied this formula to our experimental data with a similarly unsatisfactory result. The calculated values of A vary to such an extent that it is impossible to make a selection. A function of another form, suggested by A. A. Noyes, 2 which has been successfully used in connection with researches upon the electrical *A. A. Noyes: Journ. Amer. Chem. Soc. 30, 344 (1908). 2 Journ. Amer. Chem. Soc. 30, 335 (1908). 116 Studies on Solution. conductance of aqueous solutions, presented a possible means of deter- mining AO in alcoholic solutions. This function has the form where A is the equivalent conductance at the concentration c ( 1 / V) . K is a constant, and n is a number which, for aqueous solutions, lies between 1.3 and 1.7. The value of n is so chosen that the graph obtained by plotting the reciprocal of the equivalent conductance (I/A) at the various concentrations (c) against (cA) n-1 is nearly a straight line. Two other graphs corresponding to neighboring values of n, on opposite sides of the first line, are also drawn so as to aid in determining the most probable point at which the graphs cut the I/A axis. 1 This point is 1/A , the reciprocal of the limiting conduc- tivity. This procedure was followed, using the data at 25 of sodium tri- chloroacetate, salicylate, orthonitrobenzoate, 2, 4 dinitrobenzoate, and picrate. The graphs obtained are in every respect similar to those for aqueous solutions, except that the value of n lies between 1.7 and 1.8. The values of Ao obtained for the above salts at 25 are: Sodium trichloroacetate ......................... 41 . 6 Sodium salicylate ................................ 39.9 Sodium orthonitrobenzoate ....................... 38 . Sodium 2, 4 dinitrobenzoate ...................... 39.2 Sodium picrate .................................. 44. 7 From these figures the percentage dissociation of these salts is obtained by means of the familiar formula a = T A While the procedure outlined above is thus proved to give satisfactory results hi alcoholic solutions, the calculations are quite laborious, and advantage is taken of a much shorter method of approximating A , suggested by Randall. 2 It is a fact that as the zero of concentration (infinite dilution) is approached, the difference in the percentage ionization of all salts approaches zero. Randall makes the provisional assumption that the ionization of salts of the same type (such as thallous chloride and potassium chloride) is the same. Knowing the percentage dissociation of potassium chloride at various dilutions very accurately, he calculates the value of A for thallous chloride by means of the equation A =A/a' in which a' is the percentage dissociation of KC1 at any given dilution and A is the molecular conductance of T1C1 at the same dilution. Such a calculation gives values for A which approach a constant figure with increasing dilution. . Johnston: Journ. Amer. Chem. Soc. 31, 1010 (1909). 2 Journ. Amer. Chem. Soc. 38, 788 (1916). Electrical Conductance in Absolute Ethyl Alcohol. 117 In applying this method to our results we have made use of the values of percentage dissociation obtained by means of the equation of Noyes. It has been found that the three salts, sodium trichloroacetate, salicylate, and orthonitrobenzoate, include examples of all the various types of salts encountered in the present investigation. The calculation of A is illustrated by table 97 TABLE 97. y 100 a 25 A 25 A Sodium Acetate. Sodium Salicylate. Sodium Acetate. a Sodium Salicylate. 100 49.8 17.20 34.5 250 63.0 22.20 35.2 500 72.5 26.97 35.9 1,000 81.8 29.99 36.6 2,000 87.4 32.80 37.5 5,000 94.0 35.42 37.7 10,000 96.3 36.36 37.8 20,000 97.8 36.79 37.7 Probable Ao =37.8. Table 98 contains the most probable values of A at 25 for all the salts studied by the authors and calculated in the manner just indi- cated. TABLE 98. Sodium. Ao Sodium. Ao Sodium. Ao Formate 40.7 Orthoamidobenzoate . . 36.6 lodosalicylate 39 2 Acetate 37 8 Para-amidobenzoate 35 Sulphosalicylate Chloroacetate 39 5 Metabrombenzoate 35 6 Orthonitrobenzoate 38 Dichloroacetate Trichloroacetate Phenylacetate Propionate /3-iodopropionate . 41.6 41.6 35.6 37.3 40 2 Parabrombenzoate .... Orthochlorobenzoate . . Metachlorobcnzoate . . Parachlorobenzoate . . . Salicvlate 36.9 37.9 38.0 37.9 39 9 Metanitrobenzoate . . Paranitrobenzoate . . . 2, 4 dinitrobenzoate . . Orthotoluate Metatoluate 37.3 38.7 39.2 37.0 35 7 Butyrate 37 Metahydroxybenzoate. 35 7 Paratoluate 35 6 Oxyisobutyrate 37 6 Parahydroxybcnzoate 35 Picrate 44 7 Benzoate 37 4 Acetylsalicvlate 39 9 The values of A for the organic acids are calculated from those of the sodium salts by means of the following equation: AO acid=A Na salt+A HCl-A NaCl The values of A HC1 and A NaCl have been obtained from the conductance data of Goldschmidt 1 by means of the equations of Noyes and of Randall. A HC1 can be very precisely fixed at 82.0 mhos. A NcCl is most probably 42.0 mhos, with a possible variation of ==0.5 mho. Substituting these values in the equation above, we have A acid=A Na salt +40 physik Chem. 89, 131, 142 (1914). 118 Studies on Solution. Table 99 contains the probable values of A at 25 for the organic acids, calculated in the manner just indicated. TABLE 99. Acid. Ao Acid. Ao Acid. Ao Formic Acetic 80.5 78 Orthoamidobenzoic . . . Meta~amidobenzoic . . . 76.5 75.0 lode-salicylic Sulphosalicylic 79.0 Chloroacetic 79.5 Para-amidobenzoic . . . 75.5 Orthonitrobenzoic . . . 78.0 Dichloroacetic 81.5 Orthochlorobenzoic . . . 77.5 Metanitrobenzoic .... 77.5 Trichloroacetic Phenylacetic .... 81.5 75.5 Metachlorobenzoic Parachlorobenzoic .... 78.0 78.0 Paranitrobenzoic .... 2, 4 dinitrobenzoic . . . 78.5 79.0 Propionic 77 5 Salicylic 80.0 Orthotoluic 77.0 /3-:odopropionic Butyric 80.0 77 Metahydroxybenzoic. . Parahydroxybenzoic. . 75.5 75.0 Metatoluic Paratoluic 75.5 75.5 77 5 Acetylsalicvlic 80 Picric 84 5 Benzoic 77.5 SUMMARY. The authors have prepared absolute alcohol solutions of 32 sodium salts of organic acids, and have measured the electrical conductance of these solutions at 15, 25, and 35, over a concentration range extend- ing from N/50 to N/20000. Five of the salts had been previously studied by Goldschmidt and his pupils, and our results present a striking confirmation of their data. The A values for the salts can not be obtained experimentally, although they may be closely approached in many instances; they must therefore be obtained by some method of extrapolation. Goldschmidt used the Kohlrausch formula In common with Turner and with Dutoit and Rappeport we have been unable to get satisfactory results with this formula. We have been entirely successful, however, in the use of a function developed for aqueous solutions by A. A. Noyes and J. Johnston: By means of this function we have obtained the A values at 25 for all of the salts which have been studied. By combining these values with A HC1 and A NaCl we have been able to calculate the limiting conductance at 25 of 31 organic acids in absolute alcohol solution. These A values, in the case of the 5 acids studied also by Goldschmidt, are uniformly lower than those obtained by the latter. With a knowledge of the A values of the organic acids, it will be possible to estimate the dissociation and affinity constants of these acids in absolute alcohol solution. CHAPTER V. A STUDY OF THE DISSOCIATING POWERS OF FREE AND OF COMBINED WATER. BY G. FRED. ORDEMAN. INTRODUCTION. The work of Uhler, 1 Anderson, 2 Strong, 3 Guy and Shaeffer, 4 and Paulus 5 on the absorption spectra of solutions in their relation to the phenomenon of solvation has been reviewed in a preliminary paper on this subject. 6 These investigators having found a marked physical difference between free water and combined or water of hydration in their behavior towards light, it was believed that a determination of the dissociation power of this combined water might lead to the establish- ment of further differences between it and free water. A few prelim- inary measurements showed the probability of such a difference in dissociating power. The present investigation is a continuation of this work along somewhat broader lines. For the sake of completeness, certain details of the method, although described in the preliminary paper, are repeated here. The object has been to ascertain the differ- ence, if any, between the dissociating power of combined water or water of hydration and the dissociating power of uncombined or free water. EXPERIMENTAL. APPARATUS. Conductivity Apparatus. The improved slide- wire bridge used for the conductivity measurements was manufactured by The Leeds and Northrup Company, of Philadelphia. In this instrument the resist- ance wire, 5 meters in length, is wrapped around a porcelain drum. Readings were made corresponding in most cases to at least 0.25 mm. The resistance box had been standardized by the Bureau of Standards, Washington. An alternating current was supplied by an induction coil specially constructed for such work. The coil was actuated by a single lead accumulator and the strength of the current was regulated by adjusting the length of a thin manganin wire inserted between battery and coil. A telephone receiver was employed to determine the point of equilibrium. A double system of wiring was used between the 'Carnegie Inst. Wash. Pub. No. 60, 160 (1907). 3 Ibid., 130 (1910). 5 Ibid, 210, 9 (1915). ., 110 (1909). *Ibid., 190 (1913). *IUd, 230, 161 (1915). 119 120 Studies on Solution. rheostat, bridge, and cell. Thus, by means of a rocking commu- tator with mercury contacts, the positions of rheostat and cell relative to the bridge could be interchanged so that both a and b could be read directly. All copper wire in the external circuit was of such a gage that the resistance was negligible. All connections were soldered. Cells. Because of the high resistance of the water used a special " water cell" 1 having large electrodes was necessary. The electrodes consist of two concentric platinum cylinders held in position by small drops of fusion glass in such a manner that they are about 1 mm. apart. Because of the large conductivity of the solutions measured a differ- ent type of cell was demanded. The cell finally adopted for this work was that which has been previously used in this laboratory 2 for measure- ments of the conductivity of concentrated solutions. It consisted of a U-shaped tube made of difficultly soluble glass and fit ted with ground- glass stoppers. A glass tube carrying a small platinum electrode is sealed by means of sealing-wax into the hole bored in the center of each stopper. The tubes were first held in position by tamping wet asbestos between them and the walls of the stoppers. The distance between the electrodes can be changed by removing the wax, adjusting the tubes, and resealing. The platinum plates are coated with platinum black. Numbers are etched upon the stoppers and the corresponding arms of the U-tubes, so that the electrodes will always be placed in th9 same U-tube and in the same position. Constant Temperature Bath. A constant temperature was main- tained by the application of a principle most clearly stated by Morse : 3 "If all the water or air in a bath is made to pass rapidly (1) over a continu- ously cooled surface which is capable of reducing the temperature slightly below that which it is desired to maintain, then (2) over a heated surface which is more efficient than the cooled one but which is under the control of a thermo- stat, and (3) again over the cooled surface, etc., it should be practicable to maintain in the bath any temperature for which the thermostat is set, and the constancy of the temperature should depend only on the sensitiveness of the thermostat and the rate of flow of the water or air." The bath used is fully described by Davis and Putnam. 4 By means of an improved toluene-mercury thermo-regulator 5 and an electrically controlled gas valve 6 the temperature was maintained constant to within 0.01. A Beckmann thermometer graduated to 0.05 was used in the bath. Comparison was made with a thermometer recently standardized at the Bureau of Standards, Washington. Glassware. Measuring flasks, burettes, and pipettes were recali- brated by direct weighing. Jena glass bottles were used for keeping the solutions. l Carnegie lust. Wash. Pub. No. 180, 89 (1913); Amer. Chem. Journ., 45, 282 (1911). 2 Zeit. physik. Chem., 49, 389 (1904). Carnegie Inst. Wash. Pub. No. 210, 119 (1915). 'Carnegie Inst. Wash. Pub. No. 198, 56 (1914). 6 /6iW.,230, 13(1915). Ibid.,2W, 121 (1915). The Dissociating Powers of Free and of Combined Water. 121 SOLVENTS. Water. The water was purified by the method of Jones and Mackay 1 as modified by Schmidt/* and had a mean specific conductivity of 1.8 X 10~ 6 at 25 C. Isohydric Solutions. If two solutions of electrolytes are mixed the conductivity of the mixture is, in general, less than the mean of the conductivities of the constituents. If the two solutions contain a common ion, however, there are concentrations at which they can be mixed without affecting each other's conductivity. This fact was first explained by Arrhenius. 3 He showed that if equal volumes of two solutions of acids of certain concentrations be mixed, the conductivity of the mixture is the mean of the conductivities of the solutions, pro- vided there be no appreciable change in volume. Such solutions are said to be isohydric. Arrhenius 4 defines them as follows: "Two solutions of acids are isohydric whose conductivity, or in other words, whose electrolytic dissociation, is not changed if they are mixed." Arrhenius worked out the condition for two solutions containing a common ion to be isohydric and has expressed it thus: In this equation, a = percentage dissociation of the salt in solution, Vi = number of liters of solution containing a gram-molecular weight of the salt, m = number of common ions in each molecule of the salt. |3, # 2 , and n are the respective symbols for the second solution. It was further shown by Arrhenius that two acids are isohydric if in a unit volume they contain the same number of hydrogen ions. With this principle in mind the investigator, in the course of his work, determined the concentrations of five different pairs of salt solutions when they fulfill the condition of being isohydric. This method follows. In calculating the percentage dissociation by the conductivity method " tt= ~ '- 2 Here, ju' and M" = molecular conductivities of the two solutions; and //OQ = the conductivities at infinite dilution. From the method of Kohlrausch for calculating conductivity (3) 1 Amer. Chem. Journ., 19, 90 (1897). 3 Wied. Ann., 30, 51 (1887). 2 Carnegie Inst. Wash. Pub. No. 180, 135 (1913). 4 Zeit. physik. Chem., 2, 284 (1888). 122 Studies on Solution. ki and k% are cell constants, -oi, 61, and 02 and 6 2 are the respective readings on the bridge for the resistances w\ and w 2 . Substituting the values of (3) in (2), we obtain (4) Substituting now the values of (4) in (1) m k\a\v\ _ n (5) If the same cell be used in determining the conductivities of the two solutions, then k\ equals k 2 , and simplifying (5) we obtain mttl Ua * (6) This is the general equation, which becomes further simplified for a particular case. A single illustration will make its meaning clear. What concentration is necessary for a calcium nitrate solution that it be isohydric with regard to a molar solution of potassium nitrate? At 25 C., and this temperature was used throughout the work, for calcium nitrate, M'OO equals 257.99, 1 and for potassium nitrate, //'> equals 148.39. Considering the nitrate ion, m = 2 and n = 1 . Therefore 2 ai a 2 257.99 biWi 148.39 b 2 w 2 Whence b\w\ b 2 w 2 Now by measuring a and b of the molar solution of potassium nitrate for the resistance w 2 , the right-hand side of the equation becomes a constant. A concentrated solution of calcium nitrate was taken in different portions and diluted until, by trial, that concentration was found such that -^- became equal to the value for the right-hand side Mi of the equation. The concentration of the calcium nitrate was found to be 0.698 molar. And so a calcium nitrate solution of this concentration contains the same number of nitrate ions per unit volume as a molar solution of potassium nitrate. SALTS. The majority of the salts used were the purest obtainable from Kahlbaum. All the non-hydrated salts were carefully recrystallized from conductivity water and thoroughly dried at the temperature best 1 M\ MOO values were taken from Carnegie Inst. Wash. Pub. No. 170 (1912), but are expressed here in reciprocal ohms instead of in Siemens units. The Dissociating Powers of Free and of Combined Waters. 123 suited to each salt. The hydra ted salts, being so soluble, were in most cases not recrystallized, but dissolved in conductivity water, filtered, and used as concentrated solutions. SOLUTIONS. Solutions for the first three salts in table 100 were made as follows. Quantities of the isohydric solutions of the two chlorides were made. The amount of added salt necessary for each concentration was weighed upon a watch glass and introduced into a calibrated flask through a short-stemmed funnel. The salt was dissolved by one of the isohydric solutions and the flask placed in the bath regulated for 25 C. When the solution had come to temperature it was diluted to the mark with more of the isohydric solution. The process was now repeated, using the same flask and the other isohydric solution. But it was found that the volume change caused by the added salts was considerable. This means that the solutions when made would be of the proper strength for the added salt but weaker for the isohydric solutions. The results are, however, still comparable, as the volume change in the two iso- hydric solutions was found to be about the same. Solutions for the other three salts in table 100 were made in a dif- ferent manner. Instead of using the stock isohydric solutions, they were made up as needed. The amount of potassium chloride necessary to make the isohydric solution molar was weighed into the flask. The added salt, being a hydrated one, could not be weighed directly. It was added in the form of a concentrated solution of known strength from a small burette. The solutions were finally brought to the mark with conductivity water. For a comparable solution the necessary number of cubic centimeters of a concentrated calcium chloride solu- tion of known strength to make the solution 0.6951 molar was used in place of potassium chloride. In this way solutions were obtained accurate with respect to the isohydric solutions but not with regard to the added salt, because of the errors due to improper drainage of con- centrated solutions in a burette of such small bore. It was finally decided to take the densities of the various concen- trated solutions and to add these solutions by weight rather than by volume. The densities were taken with a pycnometer. The solutions were added to the weighed flasks (capacity 25 c.c.) from a burette with a finely drawn tip. With care the solutions could be weighed to within 1 or 2 mg., and this proved to be the most accurate method of handling these salts. As before the same flask was used for comparable solutions and the possibility of errors was thus, in part, avoided. The strengths of the different concentrated solutions of the chlorides were determined by an estimation of the chlorine as silver chloride. The other concentrated solutions were analyzed for the cations. All solutions were made up at 25 C. 124 Studies on Solution. PROCEDURE. Six U-shaped cells were used in this work. Five of them had con- stants in the neighborhood of 14,000, while the sixth cell had a constant of 29,941. The constants were determined by means of a half-molar solution of potassium chloride. The molecular conductivity of this solution was found to be 115.71 reciprocal ohms at 25 C. Two solutions were first made isohydric by the means described above. The specific conductivities of these solutions were measured by the usual method and calculated from the formula s = k r The same cell was employed for both solutions, so that any change in the cell constant or any error in its determination would be eliminated for comparison. Pairs of solutions were made which were isohydric with regard to each other and of a known molarity for an added salt. Three concentrations of each added salt were used. The specific conduc- tivities of these solutions were now determined. When the conduc- tivity of a solution, say molar with regard to potassium nitrate and half-molar with regard to sodium nitrate, had been measured, the cell was thoroughly cleaned and dried. The same cell was now used for the determination of the conductivity of a solution 0.6984 molar with respect to calcium nitrate that is, isohydric with potassium nitrate and half-molar in regards to sodium nitrate. Thus possible errors were avoided. The increase in conductivity of each isohydric solution was calculated for each added salt at every concentration and results are given in the third and fifth columns of each of the following tables. The difference between the increases for comparable solutions is found in the last column of each table. At the top of each table the concentrations of the two solutions which were isohydric are given. In all cases the numbers given for conductivities are in reciprocal ohms and represent the mean of at least three readings on the bridge for different resistances. The Dissociating Powers of Free and of Combined Water. 125 MEASUREMENTS. The headings in tables 100 to 104 require some explanation. The two main headings in each table are the concentrations of the two solu- tions which were made isohydric; s and s' are the specific conductivi- ties of solutions, say for 111.78 in table 100, molar for KC1 and eighth- molar for NaCl. As and As' are the results obtained by subtracting from s and s f the specific conductivities of the corresponding isohydric solutions. As As' is the difference in the increases As and As'. TABLE 100. Molar for KC1; 0.695 molar for CaCl 2 . Added salt. V s As s' As' s-As' NaCl 8 118.82 8.70 106.91 7.92 0.78 NaCl 2 144.33 34.21 127.91 28.92 5.28 NaCl 1 172.30 62.17 151.66 52.67 9.50 KC1 8 122.65 12.52 109.86 10.87 1.65 KC1 2 158.12 47.99 140.91 41.92 6.07 KC1 1 203.76 93.63 180.69 81.70 11.93 NH 4 C1 8 122.56 12.44 109.75 10.77 1.67 NH 4 C1 2 157.59 47.46 140.38 41.29 6.18 NH 4 C1 1 202.12 91.99 180.69 81.70 10.29 MgCl 2 8 124.39 14.27 110.58 11.60 2.67 MgCl 2 2 157.44 47.31 139.66 40.67 6.64 MgCl 2 1 185.54 75.41 158.21 59.22 16.19 CaCl 2 8 126.47 16.34 112.39 13.40 2.93 CaCl 2 2 166.88 56.75 147.27 48.28 8.47 CaCl 2 1 204.78 94.65 177.69 78.70 15.95 SrCl 2 8 125 . 82 15.69 112.43 13.45 2.24 SrCl 2 2 166.13 56.00 145.86 46.88 9.12 SrCl 2 1 204.47 94.34 177.83 78.84 15.50 I TABLE 101. Molar for NaCl; 0.597 molar for CaCl. Added salt. V s As 9 As' As -As' NaCl 8 93.21 8.87 97.38 8.59 0.28 NaCl 8 117.77 33.42 119.70 30.91 2.51 NaCl 1 146.76 62.41 147.31 58.52 3.89 NH 4 C1 8 97.17 12.82 99.67 10.87 1.95 NH 4 C1 2 132.59 48.24 133.06 44.26 3.98 NH 4 C1 1 177.05 92.70 175.16 86.37 6.33 MgCl 2 8 98.05 13.70 101.16 12.36 1.34 MgCl 2 2 129.95 45.60 129 . 80 41.01 4.59 MgCl 2 1 157.52 73.17 164.44 65.65 7.52 CaCl 2 8 99.37 15.02 102.72 13.93 1.09 CaCl 2 2 138.24 53.89 138.73 49.94 3.95 CaCl 2 2 175.42 91.07 172.73 83.93 7.13 SrCl 2 8 99.79 15.45 103.16 14.37 1.07 SrCl 2 2 138.74 54.39 139.10 50.31 4.08 SrCl 2 1 175.88 91.53 110.22 85.21 6.32 KNO 3 8 94.69 10.34 98.17 99.38 0.97 KNO 3 2 122 . 59 38.24 122.52 33.73 4.51 KNO 3 1 156.38 72.03 151.31 62.52 9.51 126 Studies on Solution. TABLE 102. Molar for NaNO 8 ; 0.681 molar for Ca(NO 3 ),. Added salt. V 1 A* s f As' As -As' NaNO 3 8 81.77 7.25 83.53 6.06 1.19 NaNO 3 2 101.13 26.62 98.83 21.36 5.26 NaN0 3 1 123.40 48.89 116.71 39.24 9.65 KNO 3 8 83.49 8.97 84.92 7.45 1.52 KNO 3 2 108.83 34.31 105.44 27.97 6.35 KNO 3 1 138.18 63.66 129.79 52.32 11.34 NH 4 NO 3 8 84.83 10.31 86.52 9.05 1.27 NH 4 NO 8 2 114.10 39.59 111.81 34 . 33 5.25 Mg(N0 3 ) 2 8 87.01 12.49 87.52 10.05 2.44 Mg(N0 3 ) 2 2 116.62 41.11 109.53 32.06 9.05 Mg(N0 3 ) 2 1 140.45 65.94 128.26 50.79 15.15 Ca(N0 3 ) 2 8 85.28 10.77 86.00 8.53 2.24 Ca(N0 3 ) 2 2 109.01 34.49 104 . 34 26.87 7.62 Ca(N0 3 ) 2 1 125.69 51.17 115.68 38.20 12.97 Sr(N0 3 ) 2 8 84.00 9.48 85.01 7.54 1.95 Sr(N0 3 ) 2 o 104.46 29.94 99.74 22.27 7.67 Sr(N0 3 ) 2 1 115.42 41.23 106.92 29.44 11.78 KC1 8 84.81 10.29 87.01 99.54 0.75 KC1 2 115.09 40.57 113.72 36.25 4.33 KC1 1 155.72 81.20 149.74 72.27 8.93 TABLE 103. 0.5 molar for NaNO 3 ; 0.310 molar for Ca(NO 3 ) 2 . Added salt. V s A.s s' A*' As -A-?' NaNO 3 8 50.78 8.43 51.78 7.67 0.75 NaNOj 2 74.56 32.21 73.73 29.63 2.58 NaNO 3 1 101.50 59.15 98.06 53.96 5.19 KNO 3 8 52.68 10.33 53.67 9.57 0.77 KNO 3 2 81.80 39.45 80.72 36.62 2.83 KNO 3 1 115.64 73.29 111.90 67.80 5.50 NH 4 N0 3 8 53.58 11.23 54.44 10.33 0.89 NH 4 N0 3 2 86.60 44.25 85.36 41.26 3.00 NH 4 NO 3 1 126.48 84.13 122.66 78.56 5.57 Mg(NO 3 ) 2 8 57.33 14.98 57.83 13.72 1.25 Mg(N0 3 ) 2 2 94.39 52.04 91.71 47.60 4.44 Mg(NO 3 ) 2 1 126.18 83.83 120.63 76.53 7.30 CaNO 3 ) 2 8 56.12 13.77 56.72 12.62 1.15 Ca(N0 3 ) 2 2 88.13 45 . 78 86.46 42.36 3.42 Ca(N0 3 ) 2 1 112.60 70.25 108.49 64.39 5.87 Sr(N0 3 ) 2 8 55.13 12.78 56.01 11.91 0.87 Sr(N0 3 ) 2 2 83.23 40.88 81.96 37 . 85 3.03 Sr(N0 3 ) 2 1 103.27 60.92 99 . 05 54.95 5.97 KC1 8 53.94 11.59 55.14 1 1 . 03 0.55 KC1 2 89.30 46.95 88.67 44.56 2.39 KC1 1 132.99 90.64 130.94 86.84 3.81 The Dissociatin Powers of Free and of Combined Water. TABLE 104. Molar for KNO 3 ; 0.698 molar for Ca(NO 3 ) 2 . 127 Added salt. V s As s' As' s-As' NaNO 3 8 97.84 7.09 84.69 5.80 1.29 NaNO 3 2 116.64 25.89 98.83 20.94 4.95 NaNO 3 1 138.90 48.15 117.73 38.84 9.31 KN0 3 8 100.08 9.33 86.26 7.34 1.95 KN0 3 2 125.02 34.26 106.61 27.83 6.43 KNO 3 1 155.41 64.66 131.04 52.15 12.51 Sr(N0 3 ) 2 8 99.75 9.00 86.00 7.11 1.89 Sr(N0 3 ) 2 2 117.89 27.14 100.25 21.37 5.77 Sr(N0 3 ) 2 1 128 . 86 38.11 108.09 19.20 8.91 KC1 8 102.38 11.63 88.13 9.25 2.38 KC1. 2 136.72 45.97 115.05 36.16 9.81 KC1 1 180.38 89.63 150.27 71.38 18.25 NaCl 8 100.08 9.33 86.08 7.20 2.14 NaCl 2 125.94 35.20 107.22 28.34 6.86 Nad 1 156.15 65.41 131.75 52.86 12.54 DISCUSSION OF RESULTS. The conductivity values to be found in the second and fourth columns of tables 100 to 104 are not the sums of the specific conduc- tivities of the two salts present in each case, but are less than this sum because of the common ion effect. Furthermore, since the two solutions in any given case contain the same number of anions, the added salt not being considered, the driving back of the dissociation of the added salt by these anions, other things being equal, would be the same. An inspection of the tables will show that for every pair of solutions studied this suppression is more pronounced in the hydrated solutions. Or, stating it in another way, the increase in conductivity caused by the addition of the same amount of added salt is always greater in the non-hydrated solutions. This means that the added salts dissociate more in the last-named solutions than in the comparable isohydric solutions of hydrated salts. A closer inspection of the tables reveals the fact that the driving back of the ionization of the hydrated salts added is much greater than the driving back of comparable quantities of non-hydrated salts in both isohydric solutions of every pair studied. A comparison of tables 102 and 103 will show that for any one added salt the difference in the increases of conductivity in table 102 is approximately double the corresponding difference in table 103. Finally, a few salts were added which do not have ions in common, and these behaved in somewhat the same manner as the other added salts, though the results are somewhat irregular. How can all these facts be explained? 128 Studies on Solution. A tentative explanation based upon these somewhat limited observa- tions is offered which is by no means final. When a salt is added to water or to the solution of another added salt, the added salt is dis- sociated by the water present. It is believed that combined water i. e., water of hydra tion in the solution of hydrated salts possesses less ionizing power than the uncombined water, in which case the salts added would be less dissociated. And further, this effect would be greater the greater the concentration, since more combined water would then be present. The hydrated salts used as added salts are less dissociated than the other added salts because water of hydration now exists in both of any pair of solutions. However, the dissociation is always less in the case of the hydrated salt of any pair because of the less dissociating power of the water of hydration already present in that solution. These results and conclusions which follow are to be regarded as preliminary. The nitrates and chlorides have been used and not the sulphates, principally because they are less liable to form double salts. But in the concentrated solutions it can not be said with certainty that no complexes were present. Valu e s for the degree of dissociation based upon the equation u, a are somewhat open to doubt. 1 The conductivity of a solution Moo (apart from experimental errors) is dependent to a greater or less extent upon the viscosity of the medium and the migration velocity of the ions. The latest relation between viscosity and conductivity has been deduced by Washburn and Clark. 2 Unfortunately this is of little value in applying a viscosity correction, since one of the factors is dependent upon the nature of the medium and there is at present no means of evaluating it for solutions of strong electrolytes. Considering the speed of the ions, no quantitative correction can be made. It will be noticed that normal solutions of sodium and potas- sium nitrates have been paired with calcium nitrate solutions. The migration velocity of potassium is greater, while that of sodium is less than the migration velocity of \ calcium, yet this fact hardly affects the results. Lewis 3 has recently held that the speed of ions actually increases rather than decreases with increasing concentration, and so the degree of dissociation based upon the conductance ratio is always too high. The evidence either way, however, is not conclusive. From his extensive work upon dielectric properties of solutions, Walden 4 concludes that the presence of salts in solutions increases the ionizing power of the solvent. With this granted, the hydrated salts may be said to alter the dielectric constant differently from the non- KJourn. Amer. Chem. Soc., 38, 788 (1916). *Ibid., 37, 1043 (1915). ., 38 (1916). "Zeit. physik. Chem., 55, 683 (1906). The Dissociating Powers of Free and of Combined Water. 129 hydrated salts, since we believe their ionizing powers to be different. This does not mean that the combined water is more or less associated, for while the dissociating solvents with highest dielectric constants are usually most highly associated, the principle is not without exception. The results presented here are relative. It would be interesting to make a further study along the same lines, but eliminating any influ- ence of viscosity by the use of some indifferent substance such as sucrose. Then, too, a determination of the dielectric constants of the various pairs of solutions by one of the methods suggested by Drude 1 or Smale 2 should lead one to a more definite statement. l Wied. Ann., 59, 17; 60, 600. 2 /6w*. t 57, 215; 60, 625. CHAPTER VI. THE DIFFERENCE IN CHEMICAL ACTIVITY OF FREE AND SEMI-COM- BINED WATER AS ILLUSTRATED BY THE EFFECT OF NEUTRAL SALTS ON THE HYDROLYSIS OF ACETIC ANHYDRIDE. 1 BY GERALD C. CONNOLLY. INTRODUCTION HYDROLYSIS. The term "hydrolysis" is applied to a number of chemical reactions in which there is first the addition of water to a complex, and then the decomposition of the product into simpler substances. From this definition it is evident that the reactions included under hydrolysis are numerous and varied. There are, in general, four main divisions of hydrolysis: (1) Hydrolysis of metallic salts. (2) Hydrolysis of esters and closely associated substances, such as amides, nitriles, acid chlorides, acid anhydrides, etc. (3) Hydrolysis of complex carbohydrates and glucosides. (4) Hydrolysis of polypeptides and proteins. In this discussion we will confine ourselves almost entirely to the first two divisions, for these are the only forms of hydrolysis which come within the scope of this investigation. HYDROLYSIS OF ACETIC ANHYDRIDE. The hydrolysis of acetic anhydride has been studied by several inves- tigators with varying degrees of success. The term "hydrolysis of acetic anhydride" is used here in preference to the term "hydration of acetic anhydride" used by other investigators, since it is more in accordance with the definition of hydrolysis previously stated. The work of previous investigators has been carefully reviewed in a prelim- inary paper on this subject. Therefore it need only be referred to here when bearing directly on the present work. Menschutkin and Vasilieff, 2 in studying the decomposition of acetic anhydride by water, found that with a mixture of equal portions of acetic anhydride and water at 19 only about one-half the anhydride was hydrolyzed at the end of 6 hours, and 1 1 days were necessary for complete hydrolysis. In table 105, taken from their work, a com- parison is made between the velocities of decomposition of acetic *See preliminary paper on this subject in Carnegie Inst. Wash. Pub, No. 230. 2 Jour. Russ. Phys. Chem. Soc., 21, 188 (1889). 131 132 Studies on Solution. anhydride, acetamide, and ethyl acetate by 1 gram-equivalent of water at 100 under the same conditions. The experiments were carried out in the presence of acetic acid. TABLE 105. 1 Substance Acet. Anhyd. + 1H 2 Acetamide. + 1H 2 Ethyl Acetate. +1H 2 Acetic acid added Per cent. 11.86 Per cent. 15.85 Per cent. 11.45 Decomposition: 1 inin. 25. 6S 4.51 0.2 11 83.9 4.64 .5 61 98.5 4.94 .87 121 99.5 5.82 .99 181 99.7 6.41 ... TABLE 106. Hydrolysis of Acetic Anhydride by Water. The acetic anhydride was almost entirely decomposed at the end of 1 hour, while the decomposition of the acetamide was slight and that of the ethyl acetate had hardly begun. A. and L. Lumiere and Barbier showed that when acetic anhydride is dissolved in water the solution possesses practically all the properties of acetic anhydride itself, but that if more than 12 parts of the anhydride are used solution is incomplete. Table 106 shows their results with 5 and 10 per cent solutions of the anhydride in cold water. Aliquot parts of each solution were withdrawn at 10-minute intervals and added to a known slight excess of aniline, which reacted quantitatively with the nonhydrolyzed portion of the acetic anhydride, forming acetanilid and an equivalent of acetic acid. Subse- quent titration with a normal solution of sodium by dioxide gave the total acid present, from which the degree of hydro- lysis of the acetic anhydride was calcu- lated. From their results it can be seen that the rate of hydrolysis is fairly rapid at first and then gradually decreases. It is the more rapid the greater the initial dilution of the anhydride and the higher the temperature. Alcoholic solutions of the anhydride were also prepared, and it was found that when molecular proportions were used, esterification was incomplete, even after a month. l Bull. Soc. Chim. (Ill) 33, 783 (1905); 35, 625 (1906). 5 per cent 10 per cent Solution. Solution. Time 15 15 9.2 4.6 11.5 9.8 10 52.5 35.0 58.2 34.6 20 74.2 48.4 71.0 51.1 30 89.7 60.8 78.9 60.0 40 95.7 69.0 86.6 67.0 50 100.0 76.2 91.7 73.3 60 80.4 93.3 77.9 70 85.5 94.6 81.5 80 89.6 96.4 85.1 90 93.8 97.9 88.9 100 96.9 100.0 92.8 110 .... 100.0 94.8 120 95.8 140 . . . . 98.5 160 100.0 Chemical Activity of Free and Semi-Combined Water. 133 Orton and M. Jones, in addition to studying the velocity of hydrolysis of acetic anhydride in acetic acid and water, investigated the effect of catalysts. It was found that acids are powerful catalysts of the hydrolysis. The effect is most noticeable in media containing but little water, and diminishes as the proportion of the water increases, being least obvious in pure water. The value of the velocity factor is a linear function of the concentration of the acid. Alkalis and hydro- lyzed salts were also found to act as strong catalysts of the hydrolysis in aqueous solutions. The following equations were given to represent the mechanism of the hydrolysis: (I) AC 2 O+H 2 O = 2AcOH (II) AC 2 0+H 2 O+H+ = 2AcOH+H+ (III) AC 2 O+H 2 O+HX=2AcOH+HX (IV) AC 2 O+H 2 + OH=2AcOH+OH Any one of the four forms could predominate, according to the con- ditions, medium, etc. In aqueous solutions the choice lies between (I), (II), and (IV). HYDROLYSIS OF SALTS. It is a well-known fact that certain salts, even though they contain the strictly equivalent quantities of acid and base required for "neu- trality," when dissolved in water are not neutral to indicators, but react either acid or alkaline. This was first noticed by H. Rose, in working with certain basic salts, but was not explained satisfactorily until Arrhenius proposed his theory of electrolytic dissociation. In the light of this theory acidity is due to the presence of an excess of hydro- gen ions, while alkalinity is due to the presence of an excess of hydroxyl ions. These ions can not be accounted for by the salts themselves; therefore they must be accounted for by the water. Water must contain both hydrogen and hydroxyl ions. The ioniza- tion constant of water can be calculated by the equation H+XOH- H 2 Since the active mass of the nonionized water is so great in comparison with the active mass of the ions, it may be considered constant. We then have H+ XOH~ = A; H 2 O, the value of & being 1.2X10-14 at 25. This ionization is the same in all aqueous solutions. The value &H 2 o> however, increases with rise in temperature. This increase is most probably due to the breaking down of the associated molecules into the simpler ones, which are more easily dissociated. Pure water contains an equal number of hydrogen and hydroxyl ions, and there- fore must react neutral. Furthermore, this relation holds for any neutral solution. To be acidic, a solution must contain an excess of 134 Studies an Solution. hydrogen ions; to be basic, an excess of hydroxyl ions. To determine whether a solution is neutral or not, we therefore make use of indicators, such as litmus, methyl orange, phenolphthalein, which give evidence by their color changes. When a normal salt is dissolved in water, partial hydrolysis takes place, yielding free acid and free base. Whether the solution will react acid or alkaline will depend on the degree of dissociation of these products of hydrolysis. It follows, therefore, that there are four types of salts which may undergo hydrolysis: (1) salts derived from strong acids and strong bases; (2) salts of weak acids and strong bases; (3) salts of strong acids and weak bases; (4) salts of weak acids and weak bases. All salts except those of the first type are hydrolyzed to a considerable extent, due to the small degree of dissociation of one or of both of the products of hydrolysis. Salts of strong acids and strong bases under ordinary conditions do not undergo hydrolysis. The determination of the degree of hydrolysis is not accomplished without difficulty. The free acid or base can not be directly titrated with a standard solution, for equilibrium would be destroyed at once and neutrality would be reached only when the salt was completely decomposed. A method must then be employed which will not destroy the hydrolytic equilibrium. The methods most generally used are: 1 (1) the determination of the velocity constant for the hydrolysis of an ester, for this is proportional to the amount of free acid or alkali present; (2) the determination of the rate of inversion of cane sugar; (3) the determination of the electrical conductivity of the solution; (4) the determination of the coefficient of distribution between two solvents. There are also many other methods of more or less limited applicability. Only those salts were used in this investigation which were proved by the above methods to be nonhydrolyzed. For these salts the values of k calculated according to the equation (Salt) _ k (Acid) X (Base) ~A: H2 o are so small that they need not be taken into account. The salts were further tested according to an observation made by Salm, 2 that salts which give no reaction with litmus have a concentration of H+ and OH" ions less than 1 X 10" 6 , a value so small that it is negligible. NEUTRAL SALT ACTION. In a discussion of neutral salt action one must distinguish clearly between the effect produced by a neutral salt on the catalytic activity of an acid (or alkali), and the effect of the neutral salt on hydrolysis by water alone. It is the latter effect in which we are most interested 1 R. C. Farmer: B. A. Reports, 240 (1901). 2 Zcit, physik. Chem. 57, 471 (1907). Chemical Activity of Free and Semi-Combined Water. 135 in this investigation, although the former is what is generally under- stood by the term "neutral salt action." EFFECT OF NEUTRAL SALTS ON THE CATALYTIC ACTIVITY OF ACIDS. It was early found 1 that the addition of a substance which is largely ionized in aqueous solution alters the rate of hydrolysis of esters or of carbohydrates by strong acids. This has been proved by the addi- tion of metallic chlorides to mixtures in which hydrochloric acid is the catalyst, the addition of bromides to hydrobromic acid, and of nitrates to nitric acid. Those chlorides which are highly dissociated have much the same effect, while a salt like mercuric chloride, which is only partially ionized, has a much feebler action. Non-electrolytes, such as the alcohols of sugars, have but little effect on the hydrolytic activity of the hydrogen ions. The action of the neutral salt is not always to accelerate the hydroly- sis; often there is a retardation. There are also well-defined differ- ences between the influence of neutral salts on the rate of inversion of cane sugar in the presence of acids and their influence in the catalytic hydrolysis of esters. The velocity of the inversion of cane sugar is increased to a much greater extent by the addition of certain concen- trations of salts than is the velocity of the hydrolysis of esters. Neutral salts have in general a retarding effect upon the hydrolysis of esters and amides by alkalis. Senter, 2 however, found that the hydrolysis of sodium chloroacetate by sodium hydroxide was greatly accelerated by the presence of neutral salts. It has been shown that neutral salt action is independent of the concentration of the compound hydrolyzed, is proportionally greater the more dilute the acid solution, is not greatly influenced by temperature or pressure, and is independent of the nature of the acid employed as catalyst. In addition, Poma 3 has determined that the intensity of the action developed by neutral salts bears a strict relation to the chemical nature of the ions of the salts and diminishes in passing from chlorides to bromides to nitrates to iodides, in succession; that it is independent of the chemical nature of the cations; and, finally, that it seems to be pro- portional, not to the concentration of the salt in the solution, but to the concentration of the ions. EFFECT OF NEUTRAL SALTS ON HYDROLYSIS BY WATER ALONE. Probably the first work done on neutral salt action in the absence of an acid was by Smith, 4 who investigated the effect of neutral salts on the rate of inversion of cane sugar. He found that salts of weak acids had almost no effect, while potassium chloride and sodium sulphate, the more nearly neutral salts, had considerable effect. Mourn, prakt. Chem. 85, 321, 401 (1862). 3 Medd. K.Vetenskapsakad. Nobelinst., 2, No. 11, 1-28. 2 Journ. Chem. Soc. 91, 473 (1907). 4 Zeit. physik. Chem. 25, 144 (1898). 136 Studies on Solution. Senter showed that neutral salts have practically no effect on the decomposition of sodium chloroacetate by water Kellogg 1 studied the effect of the neutral salts, potassium chloride, potassium bromide, and potassium iodide on the velocity of the hydro- lysis of ethyl acetate. The reactions were carried out in sealed tubes at 100, using a fixed quantity of ester and varying concentrations of the salt solution. The results obtained show that the specific influence of salts is greater in somewhat dilute solutions. As the concentration is increased, the effect gradually becomes less until it reaches zero, and then becomes negative in character; for example, a 4-normal solution of potassium chloride hydrolyzes the ester more slowly than pure water itself. Kellogg found a decrease in the accelerating power from chloride to bromide to iodide, which is in reverse order to their stability. Henderson and Kellogg 2 continued the investigation, using the chlorides of sodium, lithium, calcium, strontium, and barium, and the chloride and iodide of cadmium. They carried out the work under the same conditions as before and also measured the conductivities and viscosities of the solutions at the concentrations and temperatures employed in the experiments; and from these calculated the degree of ionization. They found that the salts which produce the greatest effect are those which are the least ionized. The accelerating effect of lithium chloride is greater than that of sodium chloride, although the degree of ionization of the former is less, while the chlorides of calcium, barium, and strontium have a greater effect than either sodium chloride or potassium chloride, although they too are less ionized. Cadmium chloride, the least ionized of all the chlorides studied, produced the greatest effect, due probably to the hydrolysis of the salt. Henderson and Kellogg concluded that the effect produced by a neutral salt on the hydrolysis of ethyl acetate is due to a specific influence on the non- ionized portion of the salt, rather than to any function of the ions. There have been several suggestions put forward to explain neutral salt action. Arrhenius 3 proposed that the salts may affect the sub- stance which is being hydrolyzed; that there may be present in the solution an equilibrium between an active and an inactive form of the substrate, and that this equilibrium may be altered through changes of temperature or ionic concentration. Armstrong and Caldwell concluded that the salts act by removing part of the water in the form of definite hydrated compounds, and in this manner increase the concentration of the reacting substance. Stieglitz explained salt effect in general by the theory that the presence of salts in the solution increases the dielectric constant, or at any rate the ionizing power of the solvent. All of these theories are plausible, but it is highly improb- able that neutral salt action is due to any one cause exclusively. l Journ. Amer. Chem. Soc. 31, 403, 886 (1909). 3 Zeit. phys. Chem. 4, 226 (1889). 35, 396 (1913). Chemical Activity of Free and Semi-Combined Water. 137 STATEMENT OF THE PROBLEM. The object of this investigation and the methods used were fully outlined in the preliminary paper. However, in order that the com- pleted work may be more readily understood, they are repeated here in some detail. The studies on the absorption spectra of solutions carried out in this laboratory by Anderson, Strong, Guy, Shaeffer, and others led to the conclusion that a marked physical difference exists between free and combined water. It seemed desirable, therefore, to determine whether a similar chemical difference was to be found. With this in view, Holmes and Jones 1 took up a study of the action of strongly hy- drated salts and slightly hydrated salts on the hydrolysis of methyl acetate and methyl formate. The method used consisted in measuring the velocity of hydrolysis of the ester by pure water and by solutions of slightly and strongly hydrated salts. The solutions were prepared in such a way that the amount of water in each was the same and was equal to the amount of pure water employed. Taking into account the hydrolysis of the strongly hydrated salts, they found that these salts hydrolyzed the ester much more rapidly than pure water itself. The reaction studied by Holmes and Jones was a very slow one and indicated that combined water has greater activity than free water. We wished to investigate the same problem, using a reaction that pro- ceeded much more rapidly; therefore we chose the reaction involving the conversion of acetic anhydride into acetic acid. EXPERIMENTAL. PURIFICATION OF ACETIC ANHYDRIDE. Pure acetic anhydride was necessary for the work. The physical properties as described in the literature vary greatly. The boiling- points given range anywhere from 135 to 140 at 760 mm. pressure. The densities given vary between 1.07 and 1.09. From this it can be seen that it was impossible to test its purity by the ordinary simple means. Acetic acid is the impurity most likely to be present in the anhydride, and is very difficult to detect if only small amounts are present. 0.51 gram of pure acetic anhydride, when completely hydro- lized, is equivalent to 100 c.c. N/10 solution of sodium hydroxide, while the same weight of a mixture containing 1 per cent of acetic acid is equivalent to 99.85 c.c. This is within the experimental error. Methods of finding the actual percentage of acetic acid and anhydride in a mixture have been given by Pickering, 2 Menschutkin and Vasilieff , 3 Treadwell, 4 Edwards and Orton, 5 and Orton and Jones. 6 Pickering Carnegie Inat. Wash. Pub. No. 230 (1915), Analytical Chemistry, 1914, vol. u. 'Journ. Chem. Soc. 63, 1000 (1893). Mourn. Chem. Soc. 99, 1181 (1911). Uourn. Russ. Phys. Chem. Soc. 21, 190 (1889). 'Ibid., 101, 1720 (1912). 138 Studies on Solution. determined the freezing-points of the solutions of anhydride and water, and compared them with the freezing-points of known concentrations of acetic acid. Menschutkin and Vasilieff treat with aniline and water, and determine the acidity after the reaction C 6 H 5 NH 2 + (CH 3 CO) 2 O = CeHsNHCOCHs+CHaCOOH has taken place. Treadwell recommends treatment with barium- hydroxide solution and titration of the excess of the latter, while Edwards and Orton convert the anhydride into acetanilid, the latter into phenylacetylchloramine, and then determine the chloramine volu- metrically. The method finally adopted to purify the acetic anhydride was that of repeated distillation, using a 5-bulb distilling head and discarding the first and last fractions. This gave an anhydride which distilled prac- tically constant at 138 to 139. Specific gravity determinations, using a 10 c.c. pycnometer, gave a mean value of 1.0852 at 15/4. The acetic anhydride was further tested by titrating weighed samples both directly and by the method advocated by Menschutkin and Vasilieff. PURIFICATION OF SALTS. Only the purest salts obtainable were used. They were usually Kahlbaum preparations, although some of other well-known firms were used. These salts were dissolved in conductivity water, filtered from any foreign matter present, and then recrystallized one or more times. APPARATUS. Thermostats. The constant-temperature baths were of the improved form designed by Davis 1 of this laboratory. The thermometers were of the differential Beckmann type. They were compared with a standard thermometer, which had been calibrated at the Bureau of Standards. Flasks, pipettes, and burettes for measuring purposes were all carefully calibrated by weight. All bottles used (varying in content from 50 to 6.1 c.c.) and all measuring flasks were of Jena glass. A special apparatus was used for the alkali solution, to protect it from carbon dioxide and water vapor in the air. SOLUTIONS. The water used in the preparation of the solutions was purified by the method of Jones and Mackay 2 as modified by Schmidt. 3 It had a con- ductivity at no tune greater than 2X10" 6 . The aniline used to combine with the excess of acetic anhydride was the purest obtainable. It was further distilled as many times as necessary to remove all decomposition products. The slightly colored product was then kept in a cupboard protected from light. Carnegie Inat. Wash. Pub. No. 210 (1914). 3 /Wd., 19, 90 (1897). J Amer. Chem. Journ., 17, 83 (1895). Chemical Activity of Free and Semi-Combined Water. 139 The solutions of the non-hydrated salts were made up directly by weight, while those of the hydrated salts were analyzed gravimetrically and diluted to the required strengths. The chlorides of barium, strontium, calcium, and magnesium were determined as silver chloride and the sulphates of sodium and magnesium were determined as barium sulphate. The solution of sodium hydroxide used in titrating the acetic acid formed by the hydrolysis of the acetic anhydride was made up approx- imately half-normal, using "sodium hyolroxide from alcohol." It was preserved in an apparatus protected from the impurities in the air. It was standardized by titration against a solution of sulphuric acid of about the same strength (0.4115 N). The sulphuric acid had been standardized as barium sulphate. The indicator used was phenolphthalein, as it gives the best results in titrating a weak acid with a strong alkali, the only objection being that it is also sensitive to carbonic acid. Corallin had been tried, but was not so satisfactory. METHOD OF PROCEDURE. The method in principle is a modification of that of Menschutkin and Vasilieff, 1 and later employed by A. and L. Lumiere and Barbier. 2 In order that the results should be comparable, the amount of water present must be kept constant; therefore the specific gravity of the salt solution was first taken, giving the weight of 1 c.c. From analysis, that part of the weight due to the anhydrous salt alone was known for each cubic centimenter. This known weight of salt, subtracted from the weight of 1 c.c. of solution, gave the weight due to the pure water alone. This, divided into the weight of 1 c.c. of pure water at that temperature, gave the amount of solution in cubic centimenters equivalent to 1 c.c. of pure water. The amount of solution thus calculated was pipetted into a 250 c.c. Jena bottle. An equivalent of 100 c.c. of pure water was taken in all determinations. The bottle was suspended in the constant- temperature bath. There was also placed in the bath a bottle containing the anhydride and a number of small empty bottles of 50 c.c. capacity. When all had come to the temperature of the bath, the bottle was removed and 5 c.c. of the anhydride introduced. Time was reckoned from when the anhydride was first added. Solution took place immediately on shaking, except in the case of the very concentrated solutions. Aliquot portions were removed and placed in the small 50 c.c. bottles, the whole being kept in the bath. These small bottles were removed, first every 5, then every 10 minutes, and a slight known Carnegie Inst. Wash. Pub. No. 60, 160 (1907). -Ibid., 130 (1910) ; 190 (1910). 140 Studies on Solution. excess of aniline added. On shaking, this combines with the residual acetic anhydride, precipitating acetanilid and liberating an equivalent of acetic acid. In one bottle of each series the reaction was allowed to go to completion without the addition of aniline, so as to control the results obtained. The total amount of acetic acid was then determined in the bottle by titration with the half-normal solution of sodium hydroxide in the presence of phenolphthalein as indicator. Never less than 10 c.c. nor more than 25 c.c. of alkali, as measured in a 50 c.c. burette, was required to neutralize the acetic acid. Two temperatures, 15 and 25, were employed. Only one concen- tration of acetic anhydride was used (approximately 5 per cent), because if two were employed the results would not be comparable on account of volume changes. For the salts molar, half-molar, and quarter-molar solutions were taken in all cases, and whenever possible solutions of greater concentration. Measurements of the velocity were not taken for longer than 60 minutes at 15 and 40 minutes at 25, for it was found that the hydroly- sis of the acetic anhydride by water was then practically complete. CALCULATIONS. From the total amount of acetic acid, as determined by titration with the alkali, that due to the water alone must be calculated. The simple formula y = 2z x is used, where y is the amount of acetic acid due to the water alone, z is the total amount of acetic acid measured by titration, and x is the total amount of acid that can be formed if all the acetic anhydride has been hydrolyzed. The results obtained for the "control" bottles, when substituted in the formula, should give the same values for x and y, which would be equivalent to 100 per cent hydrolysis. DATA. In tables 107 to 116 the concentrations of salt solutions are M, molar; M/2, half-molar, etc. Time is expressed in minutes. All results are expressed in percentages, 100 per cent meaning complete hydrolysis of the acetic anhydride. In each table there is placed for comparison a column showing the percentage decomposition of acetic anhydride by water alone. Chemical Activity of Free and Semi-Combined Water. TABLE 107. 141 Time. Concentration Potassium Chloride at 15. Concentration Potassium Chloride at 25. Water. 3M 2M M M/2 M/4 Water. 3M 2M M M/2 M/4 5 10 20 30 40 50 60 30.99 54.15 78.43 90.22 96.87 98.18 99.01 14.68 30.16 49.67 63 67 73.70 81.65 86.54 21.71 38.96 59.93 73.57 83.09 89.24 92.36 26.79 48.85 69.72 82.70 90.76 94.61 97.08 29.2 50.07 73.51 86.53 93.21 96.61 98.13 31.28 53.55 77.33 89.16 94.32 97.48 98.63 44.54 72.76 93.71 98.31 99.53 24.82 45.68 70.82 84.63 91.86 27.58 55.12 84.63 91.50 94.81 36.56 64.98 87.87 95.99 98.13 42.18 68.88 91.08 96.99 98.64 44.14 71.98 92.16 97.59 99.00 TABLE 108. Time. Concentration Sodium Chloride at 15. Concentration Sodium Chloride at 25. Water. 4M 3M 2M M M/2 M/4 Water. 4M 3 M 2M M M/2 M/4 5 10 20 30 40 50 60 30.99 54.15 78.43 90.22 96.87 98.18 99.01 19.24 24.49 40.87 56.04 64.75 73.23 79.02 21.37 33.64 54.07 67.90 77.06 84.53 88.78 24.84 42.71 65.04 79.43 86.47 90.95 94.08 30.44 52.79 75.14 87.42 92.12 97.13 98.21 32.21 54.87 77.88 89.09 94.20 97.70 98.53 33.71 55.98 79.38 89.42 94.65 97.99 98.73 44.54 72.76 93.71 98.31 99.53 21.80 36.85 60.09 75.73 85.90 28.93 50.48 75.62 87.85 93.91 35.44 60.97 84.93 93.45 97.28 42.23 60.07 90.30 97.56 98.75 44.65 72.53 91.90 97.95 99.26 46.91 73.87 93.51 98.57 99.69 TABLE 109. Time. Concentration Calcium Chloride at 15. Concentration Calcium Chloride at 25. Water. 4M M M/2 M/4 Water. 4M M M/2 M/4 5 10 20 30 40 50 60 30.99 54.15 78.43 90.22 96.87 98.18 99.01 2.80 15.60 41.37 57.97 69.51 77.85 83.95 33.98 55.67 80.86 90.45 95.72 97.49 98.90 34.34 56.48 81.08 91.87 96.20 99.13 99.84 34.01 56.58 81.19 92.04 96.96 98.25 99.19 44.54 72.76 93.71 98.31 99.53 20.18 49.93 78.20 92.23 96.16 48.36 75.24 93.76 99.81 100.00 48.61 76.23 94.52 97.94 99.37 48.79 76.81 94.56 97.51 99.65 TABLE 110. Time. Concentration Magnesium Chloride at 15. Concentration Magnesium Chloride at 25. Water. 4M 2M M M/2 M/4. Water. 4M 2-M M M/2 M/4 5 10 20 30 40 50 60 30.99 54.15 78.43 90.22 96.87 98.18 99.01 1.30 14.15 24.80 52.32 70.00 86.61 89.73 20.37 37.30 58.84 73.62 82.22 88.35 92.09 30.01 49.83 73.29 85.01 92.15 94.62 97.76 30.49 51.71 75.81 88.17 93.74 96.96 98.73 32.95 54.64 78.32 89.93 95.09 97.43 99.31 44.54 72.76 93.71 98.31 99.53 1.80 25.36 73.30 92.54 96.73 29.52 56.29 79.97 91.74 97.15 40.01 69.41 89.79 96.16 98.45 41.60 70.89 92.39 97.07 98.84 43.09 72.08 93.31 97.89 99.08 142 Studies on Solution. TABLE 111. Time. Concentration Barium Chloride at 15. Concentration Barium Chloride at 25. Water. M M/2 M/4 Water. M M/2 M/4 5 10 20 30 40 50 60 30.99 54.15 78.43 90.22 96.87 98.18 99.01 30.51 50.56 74.68 86.83 92.96 97.09 98.10 31.49 54.98 78.60 90.30 95.76 97.66 99.10 34.48 56.88 81.05 91.09 96.20 98.10 99.88 44.54 72.76 93.71 98.31 99.53 39.57 68.43 91.25 97.85 98.44 42.04 73.23 92.00 97.99 98.72 46.89 75.90 93.94 98.19 98.85 TABLE 112. Time. Concentration Strontium Chloride at 15. Concentration Strontium Chloride at 25. Water. 2M M M/2 M/4 Water. 2M M M/2 M/4 5 10 20 30 40 50 60 30.99 54.15 78.43 90.22 96.87 98.18 99.01 20.80 39.43 63.31 76.46 84.93 89.84 92.18 27.44 48.17 74.88 87.19 94.40 97.28 98.42 32.33 55.15 78.86 90.56 95.89 98.00 99.08 35.22 57.00 80.89 92.07 96.30 98.43 99.32 44.54 72.76 93.71 98.31 99.53 31.40 57.58 84 43 94.38 98.19 41.11 72.15 90.86 96.47 98.71 47.17 74.35 92.87 98.79 99.58 47.89 76.01 94.37 98.85 99.64 TABLE 113. Time. Concentration. Sodium Sul- phate at 15. Concentration. Sodium Sul- phate at 25. Water. M M/2 M/4 Water. M M/2 M/4 5 10 20 30 40 50 60 30.99 54.15 78.43 90.22 96.87 98.18 99.01 38.98 65.30 86.15 94.07 96.64 98.32 99.65 37.16 60.57 83.74 93.27 96.43 98.13 98.97 44.54 72.76 93.71 98.31 99.53 61.51 87.56 98.16 99.38 99.83 54.62 82.84 96.95 99.22 99.71 50.52 79.32 96.02 99.16 99.63 TABLE 114. Time. Concentration M agnesium Sulphate at 15. Concentration Magnesium Sulphate at 25. Water. M M/2 M/4 Water. M M/2 M/4 5 10 20 30 40 50 60 30.99 54.15 78.43 90.22 96.87 98.18 99.01 41.62 66.49 86.65 92.71 94.85 97.44 98.37 41.88 67.34 88.45 96.43 98.30 98.83 99.13 37.99 61.09 83.71 93.33 96.96 98.96 99.43 44.54 72.76 93.71 98.31 99.53 56.61 84.26 95.49 96.48 98.13 50.98 78.72 95.82 98.17 99.71 55.23 80.46 95.99 98.29 99.82 Chemical Activity of Free and Semi-Combined Water. TABLE 115. 143 Time. Concentration Potassium Nitrate at 15. Concentration Potassium Nitrate at 25. Water. 2,M M M/2 M/4 Water. 2M M M/2 M/4 5 10 20 30 40 50 60 30.99 54.15 78.43 90.22 96.87 98.18 99.01 21.93 36.35 58.29 71.53 81.37 89.93 92.27 26.50 44.79 64.15 80.91 87.11 93.80 96.14 30.59 51.24 74.33 86.18 92.74 95.91 97.43 31.78 53.35 77.15 88.76 94.85 97.20 98.49 44.54 72.76 93.71 98.31 99.53 34.44 51.90 77.03 88.69 94.94 35.56 54.37 82.07 94.58 97.30 37.03 65.93 88.45 96.48 98.60 41.64 69.71 90.92 98.01 98.84 TABLE 116. Time. Concentration. Sodium Nitrate at 15. Concentration. Sodium Nitrate at 25. Water. 2M M M/2 M/4 Water. 2M M M/2 M/4 5 10 20 30 40 50 60 30.99 54.15 78.43 90.22 96.87 98.18 99.01 21.69 36.23 59.80 73.51 81.49 89.11 92.47 26.38 45.73 70.35 81.95 91.10 94.38 97.01 32.36 53.12 77.27 87.93 93.61 97.81 98.07 33.66 54.52 79.14 89.11 94.85 96.85 98.37 44.54 72.76 93.71 98.31 99.53 33.14 54.72 80.20 91.87 96.59 36.80 63.93 86.68 95.77 97.89 41.07 69.01 90.58 97.06 98.24 41.64 69.94 90.92 97.54 98.83 DISCUSSION OF RESULTS. There is one difficulty in the study of this problem that must first be pointed out, i. e., the use of a strong alkali solution (half-normal NaOH) with which to titrate the acetic acid formed. This necessarily intro- duces some error, since a difference of 0.1 c.c. in reading the burette would make a difference of over 1 per cent. A more dilute solution of alkali could not be used, since too large a quantity of such a solution would be required. As noted in the preliminary paper on this subject, the rate of decom- position of the acetic anhydride is at first very rapid, being almost com- plete at 25 in 5 minutes and nearly three-quarters complete at the end of 10 minutes, then gradually decreasing as the reaction approaches completion. In this respect the reaction differs from similar ones studied, such as the hydrolysis of esters, since in these cases the reac- tions are reversible. Temperature has a marked accelerating influence on the hydrolysis, the velocity of the reaction as a whole and the increase for succeeding intervals of time being much greater at 25 than at 15. 144 Chemical Activity of Free and Semi-Combined Water. All the salts studied, with the exception of sodium sulphate and perhaps also magnesium sulphate, have in the case of the greater con- centrations a retarding influence on the hydrolysis. This retardation diminishes as the salt solution becomes more and more dilute. With sodium sulphate solutions the reverse is true the more concentrated the solution the greater is the accelerating effect. This is also true to a certain extent with magnesium sulphate, although the effect is not so pronounced. In the case of both magnesium salts studied, magnesium chloride and magnesium sulphate, it was difficult to get clear, clean-cut results. In titrating the acetic acid with the alkali in the presence of these salts a good end-point could not be reached. The color of the indicator, phenolphthalein, appeared to be masked, especially in the more con- centrated solutions. All the non-hydrated salts studied have a hindering effect on the hydrolysis. The amount of this hindrance under the same conditions is practically the same for the four salts studied, there being at no time a variance of more than a few per cent. With the most dilute solutions studied, quarter-molar, the results for the decomposition are practically the same as for pure water. The hydrated salts, with the exception of magnesium chloride, all give results for the decomposition greater than those of the non- hydrated ones, while with the more dilute solutions there is an appre- ciable acceleration of the hydrolysis of the acetic anhydride over that due to pure water alone. Sodium sulphate and magnesium sulphate at all concentrations studied have a very marked accelerating effect on the hydrolysis. Greater concentrations of these salts were not used for the reason that they do not mix with the anhydride at once on simple shaking. Calcium chloride, strontium chloride, and barium chloride also have an accelerating influence on the hydrolysis in the more dilute solution. Magnesium chloride acts as do the non-hydrated salts, having a retarding influence at all dilutions. Engineering Library THE UNIVERSITY OF CALIFORNIA LIBRARY