substance may be mixed with another substance of much lower specific gravity in such proportion that the specific gravity of the mixed substances may be as close to that of either of them as may be desired. For this purpose, the author uses the paraffin sold in the form of candles. The transparency of the paraffin enables the appearance of the embedded mineral to be minutely examined. Results are given showing the accuracy of the method.

B. H. B.

Dissociation of Copper Sulphate. By W. MÜLLER-ERZBACH (Ann. Phys. Chem. [2], 32, 313).-The author has studied the dissociation of copper sulphate at higher temperatures than those which he previously employed, and finds that his results agree with those obtained by Lescoeur (Abstr., 1887, 208). The paper also contains a discussion of the dependence of chemical affinity on temperature (Abstr., 1887, 628). With sodium phosphate containing 5 mols. H2O, and sulphuric acid of 1.294 sp. gr., water passes from the acid to the salt at 32°, but the change is reversed, and water passes from the salt to the acid at 47°. The equilibrium between the affinity of copper sulphate and of dilute sulphuric acid for water occurs, as might be expected, at higher temperatures the more dilute the acid.

H. C.

Rate of Dissociation as a Measure of the Vapour-tension of Hydrated Salts. By R. SCHULZE (Ann. Phys. Chem. [2], 32, 329). -A reply to Müller-Erzbach. The author seeks to justify his former conclusions with regard to Müller-Erzbach's method of determining the vapour-tension of hydrated salts (Abstr., 1887, 766). MüllerErzbach having objected to the use of zinc sulphate as being a salt which admittedly exhibits irregularities in its behaviour, copper sulphate is here shown to act in an irregular manner also when investigated by the above method. In two out of three tubes containing copper sulphate, evaporation set in at 20°, but the third did not exhibit any change even at the end of 10 weeks.

H. C.

Interaction of Metals and Sulphuric Acid. By V. H. VELEY (Chem. News, 56, 221-222). In this communication, the author points out that the results obtained by Spring and Aubin in their investigation on the action of acids on zinc containing lead (Abstr., 1887, 1074) do not adequately represent the rate of chemical change as comparable, for example, with the rate of evolution of a gas from a homogeneous liquid. Thus the initial retardation or induction" observed may be due to the adherence of bubbles of gas to the surface of the metal, and, secondly, when the change has set in, the metal is surrounded by a concentrated solution of the metallic salt, which is only in part removed by the gas bubbles. The hydrogen evolved is a resultant of a series of changes, each one of which is variable at any moment, such as the rate of diffusion of the salt of the metal in the acid liquid, the amount of surface exposed (which Spring and Aubin in some experiments kept approximately constant), and the local rise of temperature. The amount of gases other than hydrogen, such as sulphurous anhydride and hydrogen sulphide, is doubtless also dependent on the more or less perfect removal of the products of the

change from the sphere of the dissolving metal as well as on the concentration of the acid solution. On the other hand, it does not seem that variations in the relative masses of zinc could make any difference either in the rate of solution or in the products of the change, provided that the surfaces exposed were equal. The dissolution of a solid in a liquid must be regarded as a superficial action only. The author is at present studying the rate of solution of metals in acid liquids under such conditions that not only fresh surfaces of a regular geometrical figure are continuously being exposed, but also the products of the change, whether gas or metallic salt, are at once and continuously removed from the vicinity of the dissolving metal. V. H. V.

Velocity of the Formation of Ethereal Salts. By N. MENSCHUTKIN (Compt. rend., 105, 1016-1019).-The particular reaction investigated was the action of acetic anhydride on alcohols, Ac2O + RHO ACOR + AcOH, at 100°. With most alcohols, the reaction is complete. The formation of the ethereal salt is accompanied by a change of volume, which is least with methyl alcohol and increasingly greater with ethyl, propyl, and isobutyl alcohols. In order to eliminate this variable, the mixture of acetic anhydride and alcohol was diluted with 15 volumes of benzene. The constants of velocity are calculated dx from the equation = C(A — x)(B — x), in which A and B are the quantities of the substances originally present, and x the quantity of the new substance formed in time t. A and B being equal, and x and t being 0, = CAt, which gives the constant C. The results A -x given in the following tables are the mean of several concordant experiments, the constants of velocity being referred to that observed with methyl alcohol, which is taken as 100:

dt

The ethereal salts of the tertiary alcohols, phenols, and propargyl alcohol are decomposed by acetic acid, and hence in these cases the reactions are not comparable with those of primary alcohols.

The greatest velocity is observed with methyl alcohol. The velocity is affected by the isomerism of the radicles in the alcohols, but is highest with primary alcohols, and much lower with secondary alcohols, whilst in the case of tertiary alcohols it is very small indeed. In homologous alcohols of analogous constitution, the constant of velocity diminishes as the molecular weight increases, the difference being greatest in the normal primary alcohols. Non-saturated alcohols have a lower constant of velocity than the corresponding saturated alcohols. C. H. B.

Inorganic Chemistry.

Action of Carbon Bisulphide on Metals. By A. CAVAZZI (Chem. Centr., 1887, 888, from Mem. R. Acc. Sc. Inst. Bologna [4], 7, 27-33).-Carbon bisulphide vapour when passed over the heavy metals in a fine state of division and heated to a high temperature yields metallic sulphides with separation of carbon in a graphitoïdal form. Other compounds of carbon with sulphur seem not to be formed in any case. V. H. V.

Formation of Hydrates of Lithium Hydroxide from Alcoholic Solutions: Quantitative Determination of Lithium. By C. GÖTTIG (Ber., 20, 2912-2915).-Lithium hydroxide generally separates from hot saturated solutions in 96.8 per cent. alcohol with mol. H2O; when in contact with water, it shows the movements previously observed with crystals of potassium and sodium hydroxide (Abstr., 1887, 636).

When lithium hydroxide is crystallised from 62.8 per cent. alcohol, a hydrate with 1 mol. H2O is obtained; the crystals do not move when placed in contact with water.

In determining lithium as sulphate, the sulphate must be ignited for a long time until of constant weight. N. H. M.

Transformation of Ammonium Nitrate. By M. BELLATI and R. ROMANESE (Nuovo Cimento [3], 21, 5-24).-Frankenheim, and, more recently, Lehmann have shown that ammonium nitrate crystallises in various forms, according to the temperature at which the

crystallisation is effected. Thus at 36° it crystallises in the trimetric system, at 87° in the rhombohedral, and at 120° in the monometric system. In the present paper, it is shown that at these various points the salt undergoes other physical modifications. Thus, on warming, the temperature of the salt increases in direct proportion to the time up to a temperature of 35.67°; so also the rate of cooling is regular up to 30-3°, reaching a minimum at 30-07°; it then increases to 31·05°, at which point it remains constant for some time. Similar phenomena are also observable at a temperature of 85.5° to 86.5° with ascending temperature, and 82.2° to 82.6° with descending temperature, as also at 124.8° and 124.05°.

The variations of volume corresponding with these crystalline changes is also determined in an accurately calibrated dilatometer containing oil of turpentine, a liquid whose expansion is regular, and which, when properly dried, does not dissolve the salt. Results show that the curve of coefficient of expansion has two points of inflection: one at a temperature between 33 29° and 41 29°, and another at about 85°; the formula expressing the rate of expansion from 0° to the former of these points is vivo (1 + 0·000339t + 0.000000346ť2), whilst between 40° and 85° this becomes v1 = vo (1.04957 +0.00038756t + 0·000008976ť + 0·00000004322). There is also an alteration in the value of the mean specific heat at the temperature of these crystalline transformations; applying the method of mixtures and using oil of turpentine as the liquid, it is shown that the mean specific heat from 0-31' is 0.407, from 31° to 82.5° is 0.355, and from 82.5° to 124° is 0426. Hence the following values are deduced for the heats of transformation at these points :

The values obtained for the specific heats of ammonium nitrate are compared with those of Kopp and Tillinger, and the methods of correction applied by the latter are criticised. V. H. V.

Ammonium Phosphites. By L. AMAT (Compt. rend., 105, 809— 811). A solution of phosphorous acid mixed with ammonia until neutral to methyl-orange, and then concentrated until the weight of the liquid is one-fourth or one-fifth more than the calculated weight of the salt, yields large, deliquescent crystals which can be dried over sulphuric acid or at 100°. Similar crystals are obtained if the liquid is concentrated in a vacuum at the ordinary temperature. The crystals have the composition NH, H,PO, and seem to be monoclinic prisms; they melt at 123° and are very soluble in water. At 145° they lose half their ammonia without evolution of hydrogen phosphide, and yield a gummy mass which seems to contain crystals. At a higher temperature, ammonia and hydrogen phosphide are given off and phosphoric acid is formed.

. Hydrated diammonium phosphite, (NH),HPO, + 2H2O, when kept in a dry vacuum at the ordinary temperature or heated at 100°, loses water and ammonia, and yields the monammonium salt just

described. Both at the ordinary temperature and at 100°, the water is given off before the ammonia.

Monammonium phosphite has no appreciable action on ammonia gas at the ordinary temperature, but at 80° to 100° absorption is rapid, and the anhydrous diammonium salt is obtained as a white powder. The corresponding compounds of sodium and potassium have not yet been obtained. C. H. B.

Effects produced by Small Quantities of Bismuth on the Ductility of Silver. By J. SCULLY (Chem. News, 56, 224-226; 232-244; 247-248).-It is observed that the Indian method of wet assay is incidentally a delicate test for bismuth in presence of a large excess of silver. The bullion is dissolved in nitric acid, the solution diluted, excess of hydrochloric acid added, and the whole vigorously agitated to facilitate the aggregation of the silver chloride, which settles down and generally leaves a clear supernatant liquid; if, however, the liquid is turbid and the silver chloride on exposure to light, while still under the liquid, remains white, the turbidity is due to mercury, if on the other hand, the silver chloride becomes discoloured, the turbidity is due to bismuth; tin and antimony having been proved to be absent when dissolving in nitric acid. In such cases, to prevent the vitiation of the silver assay, the following modified method has proved successful:-The assay pound of bullion is dissolved in 5.5 c.c. of nitric acid, sp. gr. 1.200, the solution is mixed with 5 ozs. of water and 10 c.c. of nitric acid, sp. gr. 1-320, then 2·5 c.c. of hydrochloric acid are added, and the method proceeds as usual. For the estima tion of the bismuth, having obtained a rough idea of the amount of bismuth present from the amount of turbidity in the trial assay, sufficient bullion to yield a weighable amount of bismuth is dissolved in a small quantity of nitric acid, the solution diluted and heated with excess of ammonium carbonate, which dissolves the silver and copper carbonates, but leaves the bismuth carbonate insoluble; the latter is then washed, dried, ignited, and weighed. If lead or cadmium are present they would remain with the bismuth carbonate; the latter, however, is not likely to be present, and the former may be separated by dissolving the bismuth carbonate in nitric acid, and evaporating down with sulphuric acid; the lead sulphate is treated in the usual manner and weighed, whilst the bismuth is reconverted into carbonate and estimated as described above.

Fine silver, or silver containing 10 per 1000 of copper, alloyed with 1 to 5 per 1000 of bismuth and cooled rapidly, had its ductility, as tested by rolling, sensibly but slightly impaired, the straps having jagged edges; with 6 per 1000 of bismuth, the decrease in ductility was more evident, whilst fine silver with 9 to 11 per 1000 of bismuth was so brittle as to break with a mere tap. When, however, the cooling was gradual, 4 per 1000 of bismuth was sufficient to make the silver or silver-copper alloy mentioned above highly brittle, the fracture being crystalline in the case of fine silver and granular in the silvercopper alloy. With Indian standard silver containing 83-4 per 1000 of copper, 2 per 1000 of bismuth produced red shortness and jagged straps: as the quantity of bismuth increased the evidence of diminished