118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure.This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point.These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility.The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure.This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point.These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility.The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No.13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure.This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13. ON THE MEASUREMENT OF MECHANICAL PROd PERTIES OF BINARY INORGANIC SALT MIXTURES. BY K. LAYBOURN AND W. M. MADGIN. Received 4th October, 1932. In the following study of the mechanical properties of the three binary inorganic salt systems Pb(N03).,/NaN0,, Pb(N03),/KN03, and NaNOJKNO,! it has been found that, in addition to standardised con- ditions of casting and cooling of specimens, annealing and solvent treat- mcnt are of the utmost importance in obtaining consistent results. While Kurnakow and co-workers have measured Brinell hardness and pressure necessary to produce flow in certain binary mixtures of inorganic salts, they do not appear to have considered the possible effects of annealing such specimens.The present work was undertaken both as a study of the conditions necessary for consistent results and as a comparison of the properties of the three systems under review. Both of the binary systems involving Pb(NO,), have been shown to be of the simple eutectic type,, while NaNO, with KNO, forms a continuous series of solid solu- t i o n ~ . ~ A convenient means of comparing the systems and of examining the effects of solvent treatment is afforded by measurements of transverse breaking strength (the horizontal pull necessary to break a rod casting over two knife edges). Compression strengths also have been deter- mined for the system NaNO,/KNO,, and values have been obtained which show the effects of varying periods of annealing.In addition, Brinell hardness figures have been measured for all three systems, but it should be noted that both transverse breaking tests and compression tests possess certain advantages (discussed later) over determinations of Brinell hardness in the case of inorganic salts. Joff6 and Ewald * and Polanyi 5 have observed that the tensile strength of rock salt is greater under water than in air, and it seemed reasonable to suppose that atmospheric moisture might have a marked effect on the strengths of the castings used in the present work, since all the nitrates possess a considerable solubility in water. In the experiments now described it has been found necessary to treat all specimens with solvents to remove the effects of previous exposure, and to protect against sub- sequent attack by atmospheric moisture by coating the specimens with oil.Experimental. I. Transverse Breakiqg Strength. The specimen S (&-inch. diam. x 3&-inch long) is held loosely in a double collar C Apparatus.-This is shown diagrammatically in Fig. I. 2. anorg. Chem., 74, 89, 1912 : and earlier papers. Glass, Laybourn and Madgin, J. Chem. SOL, 874, 1932. Laybourn and Madgin, J. Chem. SOC., 2582, 1932. 2. PhySik, 22, 286, 1924. Trans. Faraday SOC., 24, 72, 1928. 857 56858 BINARY INORGANIC SALT MIXTURES (12 inches between collars), the inner surfaces of which are bevelled to ensure point contact only.A brass wire, attached to C, passes C # I ,1 FIG. I. as in tensile tests, and, in view of the brittle nature of salt castings, this is an obvious advantage. casting of Specimens.-The salts used in these and all other tests described were purified as stated by Glass, Laybourn and Madgin.2 The various mixtures used were synthesised by fusing together the appro- priate quantities of pure components in hard glass test tubes heated electrically. These melts were poured into gun metal split moulds which contained three separated circular casting holes (&inch x 3) inches), each provided with a feed trough to repair " pipes '' developed during setting. Cold casting was useless on account of premature setting of the melt with consequent irregular castings and coring.The moulds were therefore heated in an electric furnace packed with sand. It was also necessary to overheat the melts above the freezing-points so that (a) they did not set in the tube during pouring, and (b) extrusion of molten material within the moulds (resulting in winged castings) was avoided. The conditions finally adopted were : (I) KNO,!NaNO, mixtures : 1 Mould 100' below, melt 25" above F.P. of (2) { Pb(NOa)a/KNO, mixtures f melt. Pure sodium and potassium nitrates and the mixture of lowest freezing-point in each system were found to develop large " pipes " in the centre when setting, and feeding with melt was very necessary in these cases. Annealing and Solvent Treatment.--Widely varying results ob- tained with earlier castings suggested that annealing would be desir- able, and accordingly all rods were finally heated in close fitting glass tubes (to prevent distortion) for six days at temperatures roo below the Mould 75" below, melt 50' above F.P.Pb(NO,),/NaNO, and Castings were kept in desiccators pending testing. centrally through the knife - edges piece aa over the pulley p , to a spring balance B, which is in turn attached by a second brass wire, over pulleys p , and p,, to a screw wind- ing handle H . With the handle H a steady pull can be transmitted to C, and the specimen S then breaks in a uniform manner on the knife - edges. These knife - edges are 2 inches apart. In transverse breaking tests of this kind the speci- men does not re- quire to be grippedK. LAYBOURN AND W. M. MADGIN 859 minimum freezing-point in the appropriate systems.The freezing-point data are given by Briscoe and Madgin and Glass, Laybourn and Madgin.2 A specially constructed electric furnace was used to give uniform heating during annealing, and specimens were finally cooled off in the furnace. This annealing, resulted in greater consistency, but the values were stil1 conditioned by the time of exposure in the ordinary atmosphere between removal from the desiccator and testing. The effect of varying exposure to the atmosphere is shown by the following figures, obtained for seven annealed specimens of the same composition (45 per cent. KNO,, 5 5 per 20 2 bn .- I5 rd si 3 22 I0 3 E z C .I - M 0 : 5 0 Component A per cent. by weight. FIG. 2.-Transverse breaking strength/composition diagrams for the systems NaNO,/KNO,, Pb(NO,),/NaNO, and Pb(N08),/KN0,.cent. NaNO,). simultaneously, and tested consecutively in the order shown : The specimens were all removed from the dessicator Timeof exposure (mi-.) . o 5 10 15 20 25 30 Breakingstrength (lbs.) . 6 8 9-5 11.5 12.5 14.5 15 Believing these wide variations to be due to the effects of moisture, the following method of solvent treatment was adopted. Each specimen was dipped in distilled water to remove the effects of atmospheric mois- ture on the surface. The specimen was then dipped in alcohol and ether successively to remove the water, and, when dry, i t was dipped in medicinal paraffin oil ; this gave a protective coating from moisture, and J . Chem. SOC., 123, 1608, 1923.860 BINARY INORGANIC SALT MIXTURES adhered to the specimen during testing.Tests on specimens thus treated gave consistent and reproducible results, although the values were con- siderably lower than those for untreated specimens. The same degree of reproducibility was not attained, however, in the case of castings which had been treated with solvents, but had not been annealed, and i t is therefore evident that the conditions necessary for really satisfactory results are annealing, solvent treatment and surface protection. TABLE I .-EXPERIMENTAL DATA FOR TRANSVERSE BREAKING STRENGTHS (T.B.S.). (Compositions expressed as weight per cent. of the component stated.) Series I.-NaNO, with KNO,. 1bs. 8.5 11 12.7 13.3 13 10-3 7 5-8 4.5 3-7 4-2 5-2 7 6-9 6.5 6 4.8 Series 11.-Pb(NO,), with NaNO,.% NaNO, 40 45 47 50 55 57'7 60 65 70 75 80 85 90 95 T.B.S. in lbs. 7.5 8.5 7.7 6.5 1'75 1-25 1.5 3.7 3-9 3.7 3-5 3-2 2.8 2-5 Series 111.-Pb(NO,), with KNO,. The results of breaking strength tests on annealed and solvent Each treated specimens in the three binary systems are shown in Fig. 2 . value plotted was the mean of a t least six closely agreeing results. 2. Compression Strength. Apparatus.-Compression strength was measured with a 5-ton Buckton tensile machine, adapted for compression measurements, by determining the load necessary to crush each specimen. In the first series of experiments (referred t o as " continuous loading " experiments) the load was steadily increased until fracture occurred; in the second series (" intermittent loading " experiments), specimens were compressed at a series of increasing loads, each load being applied for thirty seconds, when it was released, and the length of the specimen measured with a screw gauge reading to OQOOI inch.Preparation of Specimens.-The specimens used were cast in cold Q-inch test-tubes, the casting melts being 25' above the respective freezing-points, and the test-tubes were completely filled. The middle portions of the thick rods thus obtained were sawn into suitable lengths, and these were trued up to I inch & 0.001 inch by polishing the ends on emery paper. Both annealed and unannealed specimens have been examined, and the annealing was done in closely fitting test-tubes under the same temperature conditions as with transverse breaking strengths.The usual solvent treatment with water, alcohol and ether was employed and the specimens were coated with paraffin oil. Only the system NaNO,/KNO, was examined by this method,K. LAYBOURN -4ND W. M. MADGIN 861 (a) Continuous Loading Experiments.-Under these conditions of loading a complete set of compression strengths has been determined for annealed specimens. Annealing for four days gave very consistent results, and in all cases the values were less than for unannealed speci- mens. Still lower compression strengths were obtained after fourteen days’ annealing, but the values were relatively the same as those for four days (Fig. 3), and i t is therefore concluded that the shorter period is sufficient. Figures obtained for unannealed specimens showed such specimens to be considerably stronger than those which had been annealed, but consistent crushing strength figures were not given.TABLE II.-COMPRESSION STRENGTHS (C.S.) FOR THE SYSTEM NaNO,/KNOs (Compositions expressed as weight per cent. of KNO,.) (a) Continuous Loading. I. Specimens annealed for 4 days. %moil * 0 10 25 40 55 70 85 IOO C.S. (lbs.) . . 320 1900 3700 2200 450 2100 1800 760 2 . Specimens annealed fov 14 days. yo KNO, . . o 1 0 25 40 55 70 85 IOO C.S. (lbs.) . - 1650 2600 1650 160 goo 750 - 3. Unannealed specimens. yo KNO, . 0 10 40 55 70 85 100 C.S. (lbs.) . . 350 2150 4800 1650 3700 3500 1100 (b) Intermittent Loading. Specimens annealed for 4 days. yo KNOs . 0 10 25 40 55 70 85 IOO C.S. (lbs.) . . 240 1700 3100 2000 290 1800 1300 580 (b) Intermittent Loading Experiments.-The length of each specimen, as measured after release of load, remained constant up to a small load (elastic limit), but thereafter a small but increasing permanent deformation occurred up to a certain load which varied with the com- position of the mixture.Beyond this point considerable plastic Aow set in and a further slight increase in load resulted in complete collapse of the specimen, the crushed material showing the typical cone-shaped fracture pieces similar to those obtained with crushed concrete. Fig. 4 summarises the data thus obtained. In all cases the final crushing strengths were markedly lower than those obtained in the parallel experiments under continuous loading (see Fig. 3). 3. Brine11 Hardness. Apparatus and Method.-The hardness figures for each of the three systems have been determined, using an “ Avery ” Hardness Testing862 BINARY INORGANIC SALT MIXTURES 0 25 50 75 I00 KNO, per cent.by weight. I. 4 days' annealing {continuous loading). 11. 4 days' annealing (intermittent loading). 111. 14 days' annealing (continuous loading). FIG. 3.-Compression strength/composition diagram for the system NaNO,/KNO,. TABLE III.-BRINELL HARDNESS FIGURES (B.H.). (Compositions expressed as weight per cent. of the component stated.) - ~ ~- Series I.-NaNO, with KNO,. yo KNO, . . o 10 25 40 55 70 85 IOO B.H. . . 3-0 13-1 28.2 25.8 1.6 28.2 14-3 8.8 Series 11.-Pb (NO,) with NaNO,. %NaNO, 90 85 80 75 70 65 60 57'7 55 50 45 40 B.H. . 10-0 11.8 13-7 16.7 19.4 16.7 11.8 8.6 17-4 18.8 19.4 17-4 Series III.-Pb(NO,), with KNO,.% KNO, * 95 90 85 80 75 70 65 60 55 50 45 40 35 B.H. . . 1.9 2-7 2-7 3-5 4.3 7-4 10.9 11.8 10.9 8-6 10.9 11.8 10.9K. LAYBOURN AND W. M. MADGIN 863 Machine by measuring the diameter of the impression made by a 10 mm. steel ball. Small loads, varying between 10 and 50 Kgms., were used, and the period of loading was 300 seconds. Specimens ( I inch diam. x 9 inch thick) were cast in cold nickel crucibles, and the melts used were poured a t 25' above the respective freezing-points. Both of the flat surfaces of the annealed specimens were polished parallel before testing. Solvent treatment was applied as 0 500 IOOO 1500 2000 2500 3000 Compression Ioad in lbs. weight. I. Pure NaNO,. V. 55 per cent. KNO,. VI. 70 per cent. KNO,.VII. 85 per cent. KNO,. VIII. Pure KNO,. FIG. 4.-Compression/load diagram for the system NaNO,/KNO, intermittent loading 11. 10 per cent. KNO,. 111. 25 per cent. KNO,. IV. 40 per cent. KNO,. experiments. before, but a paraffin oil coating was not applied, as it rendered the impression difficult to measure. Owing to the dull white colour and the softness of the material, the impression marks were not always well defined. In view of this and the fact that no protective oil coating was used, the same degree of precision and accuracy cannot be claimed for these tests as for transverse breaking strength and compression strength measurements. Nevertheless, Brine11864 BINARY INORGANIC SALT MIXTURES hardness evidently varies widely with composition, and the general nature of this variation is the same in all the three systems.Brine11 hardness was calculated from the formula : where : L = Load in Kilograms. D = Ball diameter (10 mm.). d = Diameter of impression in mm. The results are shown in Fig. 5 . o 10 20 30 40 50 60 70 8 2 Pb(NO,), per cent. by weight. M I. NaN03/KN0, 111. PblNO,),/KNO, 11. Pb NO,),/NaNO, I5 I0 5 0 100 go 80 70 60 50 40 30 20 10 o KNO, per cent. by weight. FIG. 5.-Brine11 hardness/composition diagrams for the systems NaNO,/KNO, Pb(NO,),/NaNO, and Pb(NO,),/KNO,. Discussion and Conclusions. It appears from the work now recorded that any examination of (a) Carefully regulated conditions of temperature for casting speci- (b) Adequate annealing of specimens. (c) Solvent treatment, followed by application of a protective oil The authors are of opinion that there are surface effects removed largely by solvent treatment, and volume effects, which can only be mechanical properties of inorganic salt mixtures demands : mens.coating.K. LAYBOURN AND W. M. MADGIN removed by annealing. I t may be expected that annealing would not only relieve the internal strains set up during cooling, but would also promote the well-known phenomena of grain-growth, with consequent weakening of the structure, and this would explain the fact that values for annealed specimens are always markedly less than those for un- annealed specimens. The surface effects have already been ascribed to the influence of atmospheric moisture. It is thought that such moisture may produce a recrystallisation film of fine crystals which has a cementing effect on the surface; solvent treatment would remove such a film.Minute surface cracks have been quoted as a source of weakness, and the increased strength of rock salt under water has,been explained by some as a result of the removal of such cracks.’ In the results now reported a pro- nounced weakening has been caused by solvent treatment, and it there- fore seems unlikely that the repair of surface cracks can be other than an insignificant factor by comparison with the cementation effects of atmospheric moisture. In the three systems examined, the results are evidently of the same general type, and do not afford a means to distinguish a eutectic from a solid solution system. The curves (Figs. 2, 3 and 5 ) are of the form which Kurnakow associates with partially miscible solid solutions, and in the case of the system NaNO,/KNO, this agrees with Tammann’s conclusion 8 that mixtures of these two salts segregate on cooling. How- ever, the two systems containing lead nitrate show behaviour somewhat different from that anticipated by Kurnakow, who has concluded that the property/composition curve should be a straight line for eutectic systems.In the present work, for certain intermediate mixtures values have been found which are much greater than would correspond with such a straight line relationship, although the actual eutectic mixtures have values very similar to those of the pure components. The authors find that the form of the curves is not altered by subjecting the specimens to as much as fourteen days’ annealing.The present results may possibly be explained in terms of internal strain. When mixtures of not exactly eutectic composition solidify, crystals of one pure component first separate and, a t a lower temperature, eutectic solid forms as a magma in which the primary crystals are em- bedded. The setting of the eutectic magma between the primary crystals would be expected to set up a condition of strain, with a resulting hardening effect in the solid mass.9 Such a hardening effect should reach a maximum for compositions lying between the eutectic composi- tion and the pure components, a conclusion which is amply supported by the results obtained in the present investigation. In the case of a solid solution system there are similar intermediate compositions of high internal strain where complete diffusion has probably not occurred on setting. In such a system, however, segregation in the solid state can take place on cooling, and in the resultant miscibility gap a condition of mere mechanical mixture is approached, with a corresponding weaken- ing of the structure.10 Unless segregation were quite complete, the two-peaked type of curve would thus persist. Cf. Desch, Trans. Faraday SOC., 24, 53, 1928. Cf. Sachs, 2. Metallk., 17, 85, 1925. * 2. anorg. Chem., IW, 65, 1931. lo Cf. Vrshesnevsky, J. Russ. Phys. Chem. SOL, 43, 1364, 1911.866 SURFACE TENSION OF MERCURY Summary. Transverse breaking strength, compression strength, and Brinell hard- ness figures have been measured for the systems NaNO,/KNO,, Pb(NO,),/ NaNO,, and Pb(NO,),/KNO,. Property/composition diagrams are of the same general form in all three systems, and an explanation of this, in terms of internal strains developed during crystallisation, is put forward. It is claimed that atmospheric moisture produces surface hardening which is removable by solvents. The special conditions of casting, annealing and surface treatment necessary to obtain consistent results are stated. We wish t o thank Professor C. J. Hawkes for facilities which enabled us to carry out Brinell and compression tests, and Mr. C. E. Pearson for his helpful advice. Armstrong College (University of Durham), Newcastle- upon - Tyne.