estimated in a preliminary experiment from the amount of mercury forced into a capillary tube containing water, and opening into mercury. The iron block was placed in a bucket of ice, or of water surrounded by ice. All experiments were made with solutions carefully saturated at 0°. Pressure was applied in each case for varying times, and the gain or loss of salt estimated from the specific gravity of the solution. Full authorities are given for the constants used in the calculations. The more important are as follows: Sal ammoniac. L = -344, Vo= 0.356; and if E= quantity of salt which goes into the saturated solution when the pressure is raised 1 atmosphere, ē = -0.000125. The saturated solution must therefore be partly precipitated by pressure. Sodium chloride. (determined by the author) = −2, and › = −0·177. Here -0.177 varies with the pressure, at 1530 atmospheres becoming 0 (see later). For 100 atmospheres = 0.000062. Alum. λ= -12, and v = 0.1033. = 0.0001648. and 0.0000351. λ=-187, 0·1496 v = The three last, therefore, must show increased solubility at high pressures. The experimental results were as follows: Saturated alum solution dissolved, after 23 hours, 11; after 20 hours, 1.8; after 18 hours, 13; after 18 hours, 2.9; after 3 hours, 3.27 grams per 100 grams of solution. The greatest change thus corresponded with a pressure of 200 atmospheres, the pressure having evidently fallen off as a result of contraction. Sodium sulphate solution dissolved, after 20 hours, 1.57, and, after 19 hours, 0.92 gram per 100 of solution. The greatest change here corresponded with 500 atmospheres. Sodium chloride solution dissolved, after 36 hours, 0.32, after 28 hours, 0.32 gram per 100 of solution. On the other hand, sal ammoniac solution deposited 218 grams per 100 of solution after 38 hours. A fall of 10° in temperature would be required to produce the same change. The author has also measured the compressibilities of the salts given below, and of the saturated solutions, by means of Oersted's piezometer. The coefficients were first measured for the solutions, and subsequently from the compressibilities of mixtures of salt and solution; the constants were calculated for the solids. The unit of pressure is one atmosphere, temperature 1°. As a rule, solution is attended by contraction; ammonium chloride, bromide, and iodide, magnesium chloride and tartaric acid, furnish the few known exceptions at ordinary temperatures and pressures. Under other conditions, the final volume of a solution may be assumed to be the same whether the water and salt are separately compressed before mixture, or compressed after mixture; and it may be shown that the rate at which the coefficient of contraction changes with pressure, or ov/cp, will be negative, nothing, or positive, accordingly as the change of volume of a solution by pressure is >=< the change of volume of the components. Now for the four solutions investigated, the sum of the compressibilities of the components is always considerably greater than that of the mixture. Three cases are then possible:-1st. The water in the solutions may regain its compressibility in the free state. Evidently then the dissolved salts must expand on compression, which is hardly probable. 2nd. The salts in solution may retain the above determined compressibilities. 3rd. The salts may behave as though incompressible. In either of the latter cases, it can be seen from the table that the compressibility of the water in the solution must be much less than that of pure water (= 51 × 10−6). In all cases, the effect of solution has therefore been to make the water more rigid; and since dv/op is always positive, the attraction attending solution must diminish, and the dilatation increase, with increased pressure. Again, since for all easily soluble salts, at least up to 50°, the coefficient of expansion by heat of a saturated solution is greater than that of a less saturated solution, or of the solid salt; it can easily be shown that dv/dt is likewise positive. Thus the changes of volume accompanying solution, whether produced by rise of temperature or by increased pressure, always tend toward expansion. The author shows also that must always change in the same sense both with increasing pressure and increasing temperature. The experiments of Winkelmann prove that A diminishes at high temperatures. With increasing temperature and pressure, it must, therefore, diminish rapidly; and the author suggests that the solution of a solid might mix with the solvent heated above its critical temperature, as soon as λ becomes = 0. Is the variation of solubility with temperature due to a change of molecular forces, or to a specific influence of heat? It is conceivable that thermal expansion may alter the molecular forces by changing the distance between the molecules. But in this case all substances which go into solution with diminished volume should be partly precipitated by warming their saturated solutions, which is contrary to experience. The influence of heat on solubility is therefore specific, although its exact nature is as yet undecided. Regarding the volume changes due to pressure and to temperature as having equal effect on solubility, the author calculates that the increase due to a specific heat influence alone is equal to the actual increase multiplied by 1-107 for alum; by 1016 for sodium sulphate; by 48 for sodium chloride; and by 0·596 for sal ammoniac. For sodium chloride, the diminution of solubility due to molecular forces is insignificant when compared with the increase directly due to heat. For ammonium chloride both influences work in the same direction. CH. B. Liquid Diffusion. By J. J. COLEMAN (Phil. Mag. [5], 23, 1–10). Determinations of velocity of diffusion were made by Graham's jar method, and by a burette method in which the solutions were introduced, and the different strata drawn off through the tap at the bottom. The rates of diffusion of mercurous nitrate, mercuric chloride, lithium sulphate, cadmium sulphate, silver sulphate, manganese sulphate, nickel sulphate, and lead nitrate were determined. The author finds that the velocities of liquid diffusion of the elements vary in a manner similar to that of their atomic volumes, being less for elements at the centre than for those at the ends of the horizontal series in the periodic arrangement. Thus calcium chloride is less diffusible than potassium chloride, and strontium chloride less diffusible than potassium and rubidium chlorides, which have the same diffusive velocity. In this last case, the author supposes that the effect of the large molecular volume of the rubidium chloride is neutralised by the greater molecular weight. In order to examine this point, the author diffused solutions of magnesium and zinc sulphates, which have equal molecular, volumes, for 50 days. The magnesium sulphate proved the most diffusive. In the same way, chromic acid was found to be more diffusive than tungstic acid. The author further considers the equal diffusibility of the chlorides, bromides, and iodides of potassium and sodium, to be due to the neutralisation of the effect of increase of molecular volume by that of increase in molecular weight. The diffusion rate of the sulphates of the 8th periodic group were the same as that of the 2nd or dyads, whilst the monads contrast strikingly with the latter. H. K. T. Catalytic Actions. By O. LOEW (Ber., 20, 144-145).—When a 15 per cent. solution of formaldehyde is mixed with an equal volume of concentrated aqueous soda, a slight evolution of gas takes place only on warming. The addition of cuprous oxide to the mixture causes an extremely violent evolution of hydrogen, accompanied by a inoderate rise of temperature; sodium formate is produced. This evolution of gas could be produced by no other metallic oxide. Hydroxylamine is very rapidly decomposed by caustic soda in presence of platinum-black. N. H. M. Conditions of Equilibrium in Aqueous Solutions: Action of Aqueous Soda on some Normal Sodium Salts. By T. THOMSEN (J. pr. Chem. [2], 35, 145-161).-When sodium hydroxide is added to a solution of sodium hydrogen tartarate, the molecular rotary power at first increases, and reaches a maximum when the acid is completely converted into neutral salt (J. pr. Chem. [2], 34, 81). Further addition of soda causes a diminution in the rotation, which varies both with the concentration of the solution and with the excess of alkali. Thus, if p = percentage of acid, and n = total number of = molecules of NaOH present in the solution, the initial molecular rotation, (m)D 60.6°, of a weak solution of neutral salt (p = 2, n = 2) becomes 59.7° for n = 5. On the other hand, when p = 18, n = 2, the initial rotation, (m)D = 57.58°, becomes 18.57° when n is increased to 5. For solutions of intermediate strength, or for intermediate amounts of alkali, the loss of rotatory power lies between the above limits. Other alkalis (KOH,NH,) produce a slight increase of rotation with increased concentration. The effects of the several alkalis are graphically represented by means of curves for the different values of n, of which the values of p are abscissæ, those of (m)D ordinates. These curves appear to converge to about the same point for low values of p. The author explains these results by assuming that in weak solutions soda, even in excess, simply neutralises tartaric acid, but that when the concentration is great a new lavorotatory compound is formed, probably by displacement of the hydroxylic hydrogen of the acid. And, in fact, by very largely increasing the excess of alkali, the rotation is rendered left-handed. Thus, for a 6 per cent. solution of tartaric acid, to which 39 mol. NaOH had been added, (m)» = = -38°. Very similar results are obtained with malic and quinic acids. The author has redetermined the molecular rotation of sodium malate. This salt crystallises with extreme difficulty from its syrupy solution, unless brought into contact with previously formed crystals, when it slowly separates as a tough mass of the formula Na,C,H,O + 4н2O. The table refers to the anhydrous sodium salt. These values differ somewhat, especially for strong solutions, from those found by Schneider (Abstr., 1881, 892). The rotation is thus different in weak and in strong solutions. Excess of soda invariably causes the solution to become more dextrorotatory. For example, excess of 1 mol. NaOH added to a solution containing 14:48 per cent. of malic acid as neutral sodium salt, reduces the molecular rotation from -11.24° to -7.21°; while excess of 3 mol. of alkali changes the rotation of a 20.5 per cent. solution from -861° to +19·12°. In this case also, the concentrated alkali probably acts on the hydroxylgroup. Results of the same kind were obtained with quinic acid, which is monobasic but contains four hydroxyl-groups. On the other hand, the right-handed rotation of camphoric acid, which contains no hydroxyl-groups, was not sensibly affected by excess of soda. The rotation of strong solutions of tartrates varies considerably VOL. LII. 2 g with the nature of the base. But by investigating the effect of dilution in each case, it is possible to calculate the molecular rotation for an infinitely dilute solution. Applying this method to the nentral potassium, sodium, and ammonium tartrates, the author shows that the limiting values thus obtained are approximately the same in the three cases, although the molecular rotations for strong solutions are very different. Similar limiting and nearly constant values may be calculated from Schneider's observations on the alkaline malates, and Landolt's on the alkaline quinates. The author has already shown (Abstr., 1881, 147 and 257) that the molecular rotations of the carbohydrates in aqueous solution tend towards simple multiples of a constant as the dilution is increased; and the same holds for alcoholic solutions of the cinchona alkaloïds. In all cases which have been carefully examined, heat acts in the same sense as dilution, and doubtless the action in both cases is of a chemical nature; a conclusion strengthened by the fact that when the rotation of a tartrate is changed from right to left by heat, the transition is perfectly regular (Abstr., 1881, 911). The polarimetric method can probably be applied to determine the distribution of a base between several acids in solution. Cн. B. Lecture Experiments. By A. VALENTINI (Gazzetta, 16, 399-401). -In this paper forms of apparatus are described, to show the synthesis of ammonia by the passage of nitrogen and hydrogen over a platinum spiral rendered incandescent by an electric current, and for attaining continuously a flame of nitric oxide and carbon bisulphide. V. H. V. Inorganic Chemistry. Method for Obtaining Chlorine from Chloride of Lime, using Kipp's Apparatus. By C. WINKLER (Ber., 20, 184—185).— Dry chloride of lime is intimately mixed with burnt gypsum, and moistened to such a degree that it can only with difficulty be rolled into balls between the fingers. It is made homogeneous by powdering with an iron mortar, and beaten into an iron frame 10 to 12 mm. high by means of an iron mallet. It is then covered with a piece of oilcloth, and submitted to great pressure. The plate of chloride of lime is then cut into cubes whilst still in the frame, taken out whole, and dried as quickly as possible at 20°. The cubes are then preserved in well-closed vessels. It is used in a Kipp's apparatus with hydrochloric acid (sp. gr. 1.124) diluted with an equal volume of water. The acid must be free from sulphuric acid. N. H. M. Action of Hydrochloric Acid on Sphalerite. By F. STOLBA (Chem. Centr., 1887, 169).-The author has previously recommended the preparation of hydrogen sulphide by the action of hydrochloric |