solution, the three systems merge into one. It follows, therefore, that the differentiation takes place only at the moment of a change in the physical conditions of the systems.

The vapour-tension of a solution of sodium acetate diminishes regularly in proportion to the amount of anhydrous sodium acetate dissolved, and since, according to Wüllner, the diminution in the maximum vapour-tension is proportional to the weight of the compound actually existing in the system, it follows that the solution of sodium acetate contains the anhydrous salt almost exclusively.

The second system, which corresponds with the saturation of the solution, has a constant maximum vapour-tension of 124 mm., but when the composition of the system approaches that of the hydrate, NaC2H3O2 + 3H2O, the tension drops suddenly to 44 mm., and retains this value until dehydration is complete.

The maximum vapour-tension of the hydrated salt, NaC2H,O2 + 3H2O, increases with the temperature up to 58°, at which point the salt melts. The water is given off as a whole, and no intermediate hydrates are formed. The tension of dissociation of the hydrate is lower, and the maximum vapour-tension of its saturated solution is higher than the average tension of the aqueous vapour in the atmosphere, hence the hydrate is neither efflorescent nor deliquescent. As the vapour-tension of the saturated solution of anhydrous sodium acetate is lower than the average tension of atmospheric aqueous vapour, anhydrous sodium acetate is deliquescent.

These results explain Reischauer's observations concerning the hydration of sodium acetate (Ann. Chim. Pharm., 95, 116).

C. H. B.

Relation between the Theories of the Theories of Capillarity and of Evaporation. By J. STEFAN (Ann. Phys. Chem. [2], 29, 655-665). -Laplace has based his theory of capillarity on the assumption that there is a force of attraction between the particles of a liquid which decreases very rapidly with the distance, becoming insensible at a certain small distance known as the radius of molecular attraction. From this it follows that the molecular forces acting on a particle are in equilibrium so long as the distance of the particle from the liquid surface exceeds the above, but that if it is less, they have a resultant inwards, and consequently an expenditure of work is required to transfer liquid from the interior to the surface. Adopting Clausius' theory of evaporation, and assuming that the particies of a liquid do not differ from those of its vapour, it is shown that in the case of a plane surface the work done in moving a particle from the interior to the surface, is equal to that required to transfer it from the surface into the vapour above it. This work is equivalent to the heat required for the process. If the density of the liquid were uniform up to the boundary between it and its vapour, the equation P2 - Pi = ΡΑ would give the relation between p2, the pressure in the interior, p, the pressure at the surface, p the density, and A the mechanical equivalent of the latent heat of evaporation of the liquid, a relation which in the case of ether gives for p2, 2574 atmospheres! Assuming that the density is a linear function of the pressure, a similar though

somewhat smaller result is obtained, as well as a value of p2 = 2728 atmospheres for carbon bisulphide. It is important, however, to mark that the above results are based on the assumption that no work is expended on the molecule itself in transferring it from the liquid to the vapour.

The author also explains on this theory the fact, first noticed by Sir W. Thomson, that the pressure of saturated vapour is less in the presence of a concave liquid surface, and greater in the presence of a convex, that when the surface is plane. A. H. F.

Coefficients of Affinity of Bases. By W. OSTWALD (J. pr. Chem. [2], 35, 112-121).-In this paper, the author determines the coefficients of affinity of various bases by means of the rate of hydrolysis of ethyl acetate, and compares the results with those obtained from the corresponding coefficients of electrical conductivity. The following results were obtained :—

The discrepancies in the case of the feeble bases, ammonia and allylamine, are probably due to the difficulty of obtaining the coefficient of initial change by interpolation. H. K. T.

Influence of Heat on the Decomposition of Oxalic Acid by Ferric Chloride. By G. LEMOINE (Bull. Soc. Chim., 46, 289–294). -The author formerly investigated (Abstr., 1884, 381) the influence of sunlight on the progress of the reaction FeCl + H2C2O1 = 2 FeCl2 + 2HCl + 2CO2. He now describes the action of heat on the same reaction. The aqueous solutions used contained 63 grams of crystallised oxalic acid and 162.5 grams ferric chloride respectively per litre. When equal volumes of these were solutions heated together in the dark, scarcely any action took place until the temperature reached 50°, the action then increased rapidly as the temperature was increased. At 100°, the evolution of gas

was at first rapid, gradually decreasing in the same ratio as the oxalic acid remaining undecomposed in the mixture decreased. tion with water very much increased the evolution of gas. The author considers this to be due to the action of the water in decomposing the ferric chloride, and thus assisting the formation of ferric oxalate, which is then further decomposed. Dilution with normal solutions of ferric chloride, ferrous chloride, or hydrochloric acid had scarcely any influence on the rate of reaction, the additional water in these cases not acting as if in the free state. Dilution with normal oxalic acid greatly increased the rate of reaction until enough oxalic acid had been added to form an acid ferric oxalate (the colour changing to green), but further addition of oxalic acid gradually diminished the rate of reaction. The author has not succeeded in isolating the acid oxalate which appears to be formed.

The influence of heat on this reaction seems, therefore, to be similar to that of light. L. T. T.

Theory of Fractional Precipitation. By J. J. HOOD (Phil. Mag. [5], 21, 119-127). When the separation of the rarer closely allied earths is effected either by decomposition of the nitrates by heat, or by precipitation of highly dilute solutions by ammonia, the relative amount of each salt decomposed depends on its "basic power," or coefficient of resistance to decomposition; a quantity which may or may not vary with the temperature, and with the nature of the precipitant. In any separation, it is of importance to determine the conditions for the greatest difference of basic powers. In practice, the basic power, E, might be determined by measuring the intensity of the action when the same amount of precipitant is added to solutions each containing a definite mass of some sulphate (Ni, Co, Mn); or, by the action of the precipitant on a solution containing one or more of these sulphates, the relative values of E might be arrived at. these cases, E might be regarded as a measure of the attraction of a particular base for a particular acid.

In

In the following theoretical investigation, no account is taken of secondary or inverse actions. Let the masses of two salts in solution be A and B, and let E and E' be the respective basic powers. Assuming that the rate of precipitation is proportional to the products of the active materials in solution, let x of A and y of B be precipitated at time t, and let p and q be quantities of precipitant consumed in each case. Then

E

— — 1, (B − y) (C — p − 2),

E

E'

log (1-2)

Thus the ratio should be calculable from a single experiment,

E'

and should be independent of the time taken by the precipitate to form, and of the mode of addition of the precipitant. It can also be shown that the ratio of the less to the more basic material in the precipitate increases, the smaller the fraction precipitated; and if the ratio of the salts in solution be V, the maximum value of the preceding ratio will E be for an infinitesimal precipitate. If E = E', no separation is VE possible; the experiments of Marignac (Abstr., 1884, 813) are not therefore decisive as regards the homogeneous character of the commoner elements.

Taking some observations by Mills and Bicket (Abstr., 1882, 689) on the precipitability of manganese and nickel sulphates by dilute sodium carbonate, the author finds the value of = 2·97; that is,

E

E'

the former resists the decomposing action of sodium carbonate with a force 2.97 times greater than does the latter. Similarly for nickel and cobalt sulphates, precipitated by dilute aqueous soda (Mills and Smith, Abstr., 1879, 877) =0.97. The separation of these two metals by fractional precipitation in this way would therefore be extremely tedious.

E
E'

The author shows, however, that basic power is not related to atomic weight, as might be suspected. Debus (Annalen, 85, 86, and 87) found that if a solution containing baryta and lime in the ratio a is precipitated by carbonic anhydride, the ratio of the oxides in the precipitate is given by the equation a = Kß. For a small precipitate approximately.

E K =

E'

Сн. В.

Halogen Carriers. By L. MEYER (J. pr. Chem. [2], 34, 502— 504). The author does not agree with Wildgerodt that the property of transferring halogens is a function of the atomic weights of the elements, provided they form the necessary combinations (compare this vol., p. 130). Thus antimony acts in this way, bismuth does not. The author maintains his view that a combination first takes place between the haloid compound and the substance acted on, that a combination then takes place between the halogen of the double compound and the hydrogen of the same, their places being taken by the free halogen. He quotes several reactions in support of his view. H. K. T.

Halogen Carriers. By C. WILLGERODT (J. pr. Chem. [2], 34, 547-550).-A reply to L, Meyer (preceding Abstract). The author points out the difference in the methods employed by himself and those used by Meyer and his pupils. G. H. M.

Indium and Gallium as Halogen Carriers. By C. WILLGErodt (J. pr. Chem. [2] 35, 142–144).—The author continues his researches on halogen carriers. With metallic indium in benzene, scarcely any chlorination took place until the indium had been converted into the chloride by the small quantity of hydrogen chloride formed. The

University of

chlorination then proceeded rapidly. The product consisted almost entirely of paradichlorobenzene. With gallium, the chlorination commenced at once; mono- and di-chlorobenzene and a small quantity of benzene hexachloride were produced. The author regards these experiments as giving further evidence of his law that the halogen carrying power of an element is a function of its atomic weight provided it form the necessary combination. H. K. T.

Expansion produced by Amalgamation. By W. E. AYRTON and J. PERRY (Phil. Mag., 22, 327).-A brass bar, one foot long and three-quarters of an inch thick, rapidly became curved when it was amalgamated along one edge, this edge becoming convex; and on hammering the bar to straighten it the curvature increased. Very great force must be produced by amalgamation.

The authors think that the mercury amalgam used in polishing Japanese "Magic Mirrors" may assist in making the thin portions more convex than the thicker.

Сн. В.

Inorganic Chemistry.

Oxy-acids of Iodine. By C. W. BLOMSTRAND (J. pr. Chem. [2], 34, 433—462).-A review of the evidence pointing to the pentavalent nature of iodine in periodic acid.

Production of Ozone. By J. J. THOMSON and R. THRELFALL (Proc. Roy. Soc., 40, 340-342).-Experiments were made with a view of ascertaining whether ozone is produced in an electric field, just not strong enough to permit of the passage of the spark through oxygen gas. The gas, before passing into the ozonising apparatus, was carefully dried and purified, and the issuing gas tested with paper soaked with a most sensitive solution of potassium iodide and starch, as also by the solution itself. Practically no ozone was produced.

It is further shown that no ozone is produced when a spark is passed through oxygen dried with the utmost care. V. H. V.

Action of Sulphur on Ammonia and Metallic Bases in Presence of Water. By J. B. SENDERENS (Compt. rend., 104, 58-60).-Contrary to Brunner's statement, an aqueous solution of ammonia of ordinary concentration acts gradually on sulphur at the ordinary temperature, with formation of a dark red liquid which contains a polysulphide and a thiosulphate. If this solution is exposed to the air, it deposits sulphur. The same products are formed according to Fluckiger, when sulphur is heated at 100° with ammonia in sealed tubes for several days.

Barium and calcium oxides likewise act on sulphur in the cold in