Fluorescent Mixtures. By L. DE BOIS BAUDRAN (Compt. rend., 452-455).—The author withdraws the statements made in Compt. rend., 105, p. 1231, lines 19-33, and p. 1232, lines 1-4. The quantities of chromium oxide given on p. 1229 are four times as great as they should be.

Purified calcium oxide shows distinct although not very strong greenish or bluish-white fluorescences, which are usually of short duration, and show no spectral bands. These fluorescences are much feebler after the calcium oxide has been strongly heated in hydrogen. Some specimens show a rose fluorescence which is not affected by heating in hydrogen.

Comparatively pure calcium carbonate was dissolved in hydrochloric acid, and fractionally reprecipitated. The first fraction showed a rose fluorescence, becoming blue-violet when heated, the manganese band being readily recognisable; the second fraction gave a blueviolet fluorescence, whilst in succeeding fractions the fluorescence became weaker and weaker. It would seem from these results that the fluorescence of calcium oxide is due to the presence of minute quantities of impurities. C. H. B.

Galvanic Polarisation. By F. STREINTZ (Ann. Phys. Chem. [2], 33, 465-474).-In continuation of his experiments on the above (this vol., p. 99), the author has examined the polarisation of mercury, gold, palladium, and platinum electrodes. With mercury, the polarisation of the hydrogen plate is at a maximum when the E.M.F. of the cell used is that of two Daniells, and for further increase remains constant. This maximum is higher than that observed for any other metal. The polarisation of the oxygen plate varies with the time of immersion, and continually increases with the E.M.F. of the cell used. With the three other metals, the results agree generally with those of Fromme (Abstr., 1887, 541). When the intensity of the current used is greater than that required to produce maximum polarisation, in each case a decrease in that of the oxygen plate sets in, that of the hydrogen plate remains constant in the case of platinum, but shows a variation with the time of immersion for gold and palladium.

H. C.

Conductivity of Illumined Air. By S. ARRHENIUS (Ann. Phys. Chem. [2], 33, 638-644).—It is shown that air under a pressure of 1 to 20 mm., when illumined by the sparks of an electrical machine, acts as a conductor, the maximum conductivity being for a pressure of 4 to 5 mm. This conductivity is shown to be electrolytic, as a current may be obtained flowing from a zinc to a platinum plate in illumined air, just as it would were these plates immersed in water. No manifestation of the above could be obtained for pressures other than those stated, although the author considers that it is not confined within those limits. In support of this he quotes the experiments of Hertz (Ann. Phys. Chem. [2], 31, 983), in which it was shown that sparks are more easily formed in air at the ordinary pressure when the path is illumined than when it is kept in the dark.

H. C.

Molecular Conductivity of Fuming Nitric Acid. By E. BOUTY (Compt. rend., 106, 595-597).-Fuming nitric acid, which always contains a slight excess of water, dissolves large quantities of potassium nitrate, forming compounds which can be crystallised, and the solutions behave with respect to conductivity like aqueous solutions of very soluble salts. The conductivity of the acid at first increases more rapidly than the proportion of dissolved salt, attains a maximum, and then decreases as the solution becomes more viscous. With small quantities of salt, the increase in conductivity is proportional to the weight of salt dissolved.

Sodium nitrate is only slightly soluble in nitric acid, and does not form crystallisable compounds. The nitric acid solution, like the aqueous solution, behaves abnormally, and has a higher resistance than similar solutions of other nitrates.

The unit of conductivity in these experiments was the resistance of normal nitric acid. The nitric acid used was of sp. gr. 1.552. A molecular resistance of 15 72 ohms is the limit of conductivity of the solutions of potassium and ammonium nitrates in nitric acid. This value approximates very closely to the molecular resistance of potassium chloride in aqueous solution. The resistance of solutions of nitrates in nitric acid increases rapidly on addition of water.

C. H. B. Electrical Resistance of Bismuth and its Alloys. By E. V. AUBEL (Phil. Mag. [5], 25, 191-201).-The electrical resistance of bismuth and its alloys in a powerful magnetic field is determined. The bismuth was prepared in three forms-(1) melted and cooled slowly; (2) melted and cooled quickly; (3) compressed. The first was obtained by melting the metal in capillary tubes immersed in sand; for the second the metal was melted and poured into a V-shaped iron mould; the third was prepared by Spring's method. The alloys were prepared by mixing weighed quantities of the metals. The resistances were measured by Thomson's method, and the results show that in some samples of bismuth the resistance rises, in others falls, with rise of temperature. The experiments with alloys show that this is not due to the presence of tin or arsenic as an impurity, nor was there any relation shown with their fusing points or specific gravities. Magnetism produces an increase in resistance, but the effect is feeble. Compressed bismuth shows hardly any alteration in resistance with change of temperature, but if it is fused and cooled slowly the usual effect is observed. In the compressed bismuth which had been drawn through a draw-plate, the particles form parallel fibres. H. K. T.

Electrolysis of Copper. By T. GRAY (Phil. Mag. [5], 25, 179 -184; compare Abstr., 1887, 315).-The electrochemical equivalent of copper is determined for varying current-densities. The copper sulphate solutions had a density of 115 to 118, and were always acid. The plates were cleaned with glass-paper, and after deposition were washed with acidified water and dried before a fire. At temperatures of 10-15°, the results are constant, but the loss in weight of the plate increases rapidly as the temperature rises to 35°. The

error can be ascertained within 0.1 per cent. by keeping a similar plate in a cell with no current passing through it. At 2°, there is no change in the amount of metal deposited until the area of the plate exceeds 200 sq. cm. per ampère. A table of apparent electrochemical equivalents of copper for different current-densities and temperatures is given. H. K. T.

Thermal and Electrical Behaviour of some Bismuth-tin Alloys in the Magnetic Field. By A. v. ETTINGSHAUSEN and W. NERNST (Ann. Phys. Chem. [2], 33, 474-492).-The authors were led by former experiments to suspect some connection between the thermoelectrical properties of metals, and the "rotatory power in a magnetic field observed by Hall (Phil. Mag., 1885, 19, 419). Since the thermoelectrical properties of bismuth are greatly modified by alloying with tin, the present investigations were undertaken with the object of ascertaining whether this modification would be attended with similar change in the rotatory power. A series of four alloys containing varying amounts of tin were examined, but although the rotatory power was very largely influenced by the presence of the tin no definite connection with the thermoelectrical properties could be traced.

H. C.

The Recalescence of Iron. By H. TOMLINSON (Phil. Mag. [5], 25, 103–116).—The internal friction of iron at different temperatures is determined by suspending a wire vertically and noting its period of horizontal vibration, the wire being heated by an electric current. A table of results is given. At 550° the internal friction rises rapidly, and still more rapidly at 1000°, so that at this temperature the wire comes to rest after two or three vibrations. From 1100° to 1200° it seems to decrease. At 550° the magnetic and thermoelectric properties also change. At 1000° heat also becomes latent, as shown by alterations at this temperature due to stress and strain. The author considers that recalescence is similar to regelation, being a sudden evolution of heat at temperatures somewhat below these points. When iron has been strained either by bending or hammering, the strained portion as the iron cools appears clouded, owing to the more rapid cooling of the strained portion, its specific heat being less than that of the unstrained parts, probably because the consumption of heat in separating the particles is prevented by the strain. Shortly before recalescence, the clond disappears. Recalescence does not seem to be prevented by shaking or hammering. With well annealed iron, recalescence cannot be detected. The author considers that it does take place, but at a point near the critical temperature (1000), and hence is not visible. H. K. T.

Evaporation of Liquids. By W. HEMPEL (Ber., 21, 900-902).Liquids can be evaporated about six times more rapidly (with, however, the combustion of about three times the amount of gas per hour) than on a steam-bath by employing a Siemens' inverted regenerative burner placed just above the surface. The liquids do not enter into ebullition, so all spirting is avoided. Experiments show

that no appreciable amount of sulphuric acid is absorbed by the liquids during evaporation, and that whilst hot the iron parts of the burner are not attacked by acid vapours. W. P. W.

Bumping during Distillation. By A. REISSMANN (Arch. Pharm. [3], 25, 970; from Pharm. Centralb., 28, 501).—A closely wound platinum spiral is charged with several longish bits of pumice, and its ends are then closed. One or more of these spirals placed in a liquid undergoing distillation effectually prevents bumping, and the operation goes on with perfect regularity. The platinum must be heavy enough to sink the pumice.

J. T.

Relation of Gases to Mariotte's Law at High Temperatures. By C. PUSCHL (Monatsh., 9, 93-98).-If at ordinary temperatures the volume of a liquid be v and the pressure p, pv increases with the pressure, so that d(pv)/dp = h is positive. If t be the temperature, and the coefficient of expansion of the liquid, then dh/dt = d(apv)/dp, so that by following the change in h with temperature, the change in pv or in Mariotte's law may be ascertained.

At pressures above the critical, heating the liquid gradually to above the critical temperature, the quantity h will be found to pass through two maxima with one minimum in between, and the latter if the pressure be continually increased at high temperatures will change from negative to positive values, and may in the end be brought into coincidence with the higher maximum, which at the same time falls to lower temperatures. If the pressure then were high enough, h would, from the lower maximum, continually decrease with rise of temperature.

For pressures below the critical, h is negative and increases with the temperature. This is the case with all gases and vapours, with the exception of hydrogen. If heated, then, a point h=0 will be reached at which the gas will obey Mariotte's law. Above this point, h takes a positive value, reaches a maximum, and then falls again, it crosses the zero-line a second time, and becomes permanently negative.

There is also for every gas an interval of temperature for which h is positive at every pressure, and pv has neither maximum nor minimum. On each side of this interval, there are two pressures for every temperature at which h = 0, for the smaller of which pv is a maximum and for the greater a minimum. At the higher and lower limiting values for these temperatures the two pressures will coincide. H. C.

Easy Method of Finding the Specific Gravity of Liquids. By A. B. TAYLOR (Chem. News, 51, 138-139). For the sake of simplicity, the author suggests that the weight of a convenient solid should be so adjusted that, expressed in grains, it corresponds with its sp. gr., then, on weighing it in the liquid to be tested, the loss of weight in grains will give the sp. gr. of that liquid without calculation.

D. A. L.

The

Compressibility of Water. By W. C. RÖNTGEN and J. SCHNEIDER (Ann. Phys. Chem. [2], 33, 644-660).--The authors repeat their experiments on the above (Ann. Phys. Chem. [2], 29, 197), using greater pressures, and thus obtaining a greater contraction. apparatus, with the exception of a new manometer, is that previously used. Experiments were conducted at the temperatures 0°, 90°, and 17.95°, and gave as the apparent compressibility per atmosphere at these temperatures 0-00004910, 0-00004602, and 0.00004413. The experiments being made with great care, and account taken of all known sources of error, the last figure only in the above is looked on as doubtful. The constant of deformation of the piezometer was determined by comparing the observed apparent compressibility of rock salt with the true compressibility obtained by Voigt (Ber., 1884, 990); by introducing this correction, the true compressibility of water at the above temperature is obtained. This is 0.0000462 at 17·95°, a value which exactly agrees with that obtained by Grassi, 0·0000481 at 9°, and 0.0000512 at 0°. H. C.

Decrease in the Solubility of Sulphates. By A. ETARD (Compt. rend., 106, 206–208).-At temperatures below 100°, the solubility of cupric sulphate increases with the temperature, but between 103° and 190° the solubility diminishes as the temperature rises. This deflection in the curve of solubility is observed with almost all sulphates. With cadmium, zinc, manganese, and iron sulphates, the point of deflection is below 100°. A similar decrease in solubility is observed in the case of salts of carbonic, sulphurous, and succinic acids, but not with salts of monobasic acids, except those of feeble organic acids.

In the case of zinc sulphate, y = 276 +0.2604t between -5° and 81°, and y = 5000-2244t between 81° and 175°. At 180°, the solubility is the same as at -5°. As the temperature of a saturated solution rises, the hydrate ZnSO, + 2H,O is deposited on the sides. of the vessel in hard, insoluble, porcelain-like concretionary masses. At all temperatures there is, in all probability, a condition of equilibrium between this and the other hydrates existing in the solution.

In the case of manganese sulphate, y = 300+ 0.2828t between -8° and 57°, and y = 48′0 — 04585t between 57° and 150°. The solution deposits anhydrous manganese sulphate as a hard, rosecoloured, porcelain-like mass, almost insoluble in water. The equation indicates that the solubility vanishes at 161°, and experiment shows that at 180° the liquid retains mere traces of the salt.

For potassium sulphate, y = = 7-501070t between 0° and 163°, and between 163° and 220° the solubility remains constant.

C. H. B. Laws of Chemical Equilibrium. By H. LE CHATELIER (Compt. rend., 106, 355-357).-A mathematical paper, not admitting of useful abstraction.

Chemical Equilibrium. By P. DUHEM (Compt. rend., 106, 485-487). The results obtained by Le Chatelier are identical with those obtained by the author (Compt. rend., 99, 1113).