been precipitated by means of sodium carbonate. The difference is due to the retention of a small quantity of alkali by the strontium carbonate. C. H. B.

Fluorescence of Manganese Compounds in a Vacuum under the Influence of the Silent Discharge. By L. DE BOISBAUDRAN (Compt. rend., 103, 468–471).-All the sulphates used in these experiments were heated to dull redness, and the carbonates and oxides were strongly heated.

Manganese sulphate alone, and trimanganese tetroxide prepared from the carbonate, show no fluorescence.

Calcium sulphate alone gives a feeble fluorescence with a continuous spectrum, but if it contains a trace of manganese sulphate it gives a brilliant green fluorescence, the spectrum of which consists of a broad band which begins at about 6600, attains its maximum intensity at about 5400, and fades away between the blue and the violet.

Calcium carbonate alone shows a feeble fluorescence, but if mixed with a small quantity of manganese, it shows a brilliant orange-yellow fluorescence characterised by a band which begins about 6700, attains its maximum intensity at about 5890, and fades away at 5260. This fluorescence is much more strongly marked than that observed with the sulphates.

Magnesium sulphate alone gives a faint greenish fluorescence, but if mixed with manganese sulphate it gives a brilliant red fluorescence characterised by a band which begins nebulously at about 6720, attains its maximum at 6200, and fades away at about 5540.

Magnesium carbonate mixed with manganese shows the same fluorescence as the sulphates.

Zinc sulphate alone shows a feeble rose-coloured fluorescence, but if mixed with manganese sulphate it shows an orange-red fluorescence, which is characterised by a band which begins nebulously at 6720, attains a maximum at about 6280, and fades away near 5380.

Cadmium oxide alone or mixed with manganese oxide gives no fluorescence. Cadmium sulphate alone shows only a feeble pale greenish-yellow fluorescence, but if it is mixed with manganese sulphate it gives a very brilliant yellowish-green fluorescence, the spectrum of which consists of a broad band which begins at 6620, attains its maximum at 5590, and fades away at about 4560. This is the most brilliant of all the manganese fluorescences.

Strontium sulphate alone shows a pale lilac fluorescence with a continuous spectrum, and the carbonate or oxalate after being strongly heated shows a blue fluorescence. These fluorescences are not materially affected by the presence of manganese.

Lead sulphate alone gives a feeble violet fluorescence, but if mixed with manganese sulphate it gives a beautiful yellow fluorescence characterised by a band which begins at about 6560, attains its maximum at 520-5760, and ends at 5310-5260. Lead oxide, whether pure or mixed with manganese, gives no notable fluorescence. Beryllium sulphate alone gives a somewhat marked green fluorescence, but if it is mixed with a trace of manganese sulphate the fluorescence is stronger and is yellowish-green. It gives a spectrum

consisting of a fairly brilliant band which begins at 6690, attains a maximum at 5640, and fades away at 4840-4800. An excess of beryllium sulphate prevents the development of the manganese fluorescence.

The character of the fluorescence and the position and character of the bands in its spectrum vary with the nature of the substance with which the manganese is mixed. These fluorescences constitute an extremely delicate test for manganese. C. H. B.

Fluorescence of Bismuth Compounds. By L. DE BOISBAUDRAN (Compt. rend., 103, 629-631).-Bismuth sulphate alone, previously heated to dull redness, shows no fluorescence when subjected to the action of the silent electrical discharge in a vacuum, but the admixture of small quantities of this salt with various other sulphates confers upon them the power of fluorescing under these conditions. In all cases, the sulphates were previously heated to dull redness. If the proportion of bismuth is gradually increased, the fluorescence at first increases in brilliancy, attains a maximum, then becomes weaker, and finally disappears even whilst the proportion of bismuth is still very small.

With calcium sulphate, the fluorescence is orange-red, and its spectrum consists of a band which begins very nebulously at 6730, attains its maximum intensity at 6140, and fades away gradually at about 5780. With strontium sulphate, the fluorescence is orange, and is more brilliant than with the calcium salt. Its spectrum consists of a band which begins very nebulously at 6640, attains its maximum intensity at 5980, and fades away at about 5637. With barium sulphate, the fluorescence is redder than with calcium sulphate, and is characterised by a band which begins nebulously at 6540, attains its maximum intensity at about 6220, and fades away gradually at 5840-5850. With magnesium sulphate, the fluorescence is still redder but is not so brilliant as in the preceding cases. It is characterised by a band which begins nebulously at 6750-6760, attains its maximum intensity at 6320-6330, and fades gradually away at about 5860. The fluorescence of the corresponding oxides is not appreciably affected by the presence of bismuth oxide. No fluorescence was observed with zinc, cadmium, and lead sulphates, or zinc and cadmium oxides, in the presence of bismuth.

The spectrum of this fluorescence is a more delicate test for bismuth than the spark spectrum. C. H. B.

Electrical Conductivity of Gases and Vapours. By J. LUVINI (Compt. rend., 103, 495-497). The author's experiments, combined with those of previous investigators, show that gases and vapours at any temperature or pressure are perfect insulators, and cannot be electrified by internal friction or friction with solids and liquids. Measurements were made by observing the divergence between a pith ball suspended by cocoon silk and a brass sphere suspended in the same way in an atmosphere of the gas to be examined, both sphere and pith ball being electrified. The gases used were air saturated with moisture at various temperatures between

16° and 100°, moist hydrogen, moist carbonic anhydride, mercury vapour at 100°, vapour of ammonium chloride, air heated by burning charcoal and by a candle flame, smoke from a smouldering candlewick, vapours from burnt sugar, incense, &c.

The belief that gases at very low pressures or high temperatures are conductors arises from a confusion between resistance to a disruptive discharge and resistance to a conductive discharge.

All theories in which gases and vapours are regarded as conductors or capable of being electrified by friction must be abandoned..

C. H. B.

Relation of the Conductive Capacity of Gases to their Temperature. By A. WINKELMANN (Ann. Phys. Chem. [2], 29, 68-113). In these experiments, the conductive capacity is determined by means of three horizontal copper plates immersed in the gas and separated from one another by glass balls, the upper and lower plates being maintained at fixed and different temperatures. The temperature of each plate was indicated by a delicate thermometer. The results obtained are

For air, mean of four experiments..

carbonic anhydride, four experiments..

,, hydrogen, two experiments...

0.00206

0.00366

0.00206

The heat given off by a plate immersed in a gas is different according to its orientation, hence that given off by a given area cannot be determined from the rate of cooling of the whole.

H. K. T.

Electrical Conductivity of Solid Substances at a High Pressure. By L. GRAETZ (Ann. Phys. Chem. [2], 29, 314-330).Hitherto our knowledge of the conductivity of electrolytes is derived for the most part from results obtained with the substances in the dissolved and not in the fused condition. Although it is à priori to be expected that the law governing these phenomena in the case of a homogeneous media would be of a simpler character than those in the case of heterogeneous media, namely, a solution of salt in water, yet the result of experience is the reverse.

The only generalisation at present deduced is that solid substances are not conductors at low temperatures, but their conductivity commences at temperatures far below the melting point, and increases with the temperature.

If then this increase of conductivity is due to an increase of molecular mobility and of the number of molecular impacts, increase of pressure without alteration of temperature should be effective to a like degree. In the paper, this point is examined by means of a compression apparatus, capable of giving a pressure of upwards of 4000 atmospheres. The results are given for the halogen-compounds of lead and silver, and for sodium nitrate. The substances examined are divisible into two classes, in the one of which, with application of maximum pressure, the resistance rapidly decreases to a constant point, and in the other this minimum resistance is, under the same condition, only attained after several hours. To the former class

belong the halogen-compounds of silver, to the latter those of lead and sodium nitrate. In the paper full details are given of the apparatus used, and of the methods of experiments; the results are tabulated. V. H. V.

Specific Heats of Homologous Series of Liquid Organic Compounds. By R. SCHIFF (Annalen, 234, 300-337).-The author has determined the specific heats of a large number of hydrocarbons, acids, alcohols, and ethereal salts, at different temperatures, and has arrived at the following results. All the ethereal salts of fatty acids of the formula CnH2Ö2, have the same specific heat at the same temperature.

The specific heat of these substances at t is represented by the equation Kt = 0·4416+ 0·00088t, and the mean specific heat between t and t' by Ctt' 04416 +0.00044(t + t').

xylene

0.3834+0.001043t.

Ethyl and propyl phenoxide

creso-oxide..

Ethylbenzene, pseudocumene, mesityl-0-3929 +0-001043t.

ene...

Pawlewsky (Abstr., 1883, 276) has pointed out that the critical temperature of the ethereal salts of the fatty acids is 182° higher than their boiling point.

The author divides the distance between the absolute zero and the critical temperature of an ethereal salt into 100 equal parts, which he ternis critical degrees.

In the table (p. 7) LX cr represents 60 of these critical degrees; KLX the true specific heat at this temperature; Dix the sp. gr. at this temperature, and KDLx the product of these two values.

Equal volumes of these substances have the same specific heat at the same "critical degree."

The author disputes the accuracy of the conclusions which De Heen (Physique comparée, 39) draws from his researches on this subject, and he also criticises the results obtained by v. Reis.

Saturation of Normal Arsenic Acid with Barium Hydroxide. By C. BLAREZ (Compt. rend., 103, 746).-The developments of heat accompanying the addition of successive equivalents of barium hydroxide to a dilute solution of arsenic acid are given in the following table:

The phenomena are not similar to those observed when arsenic acid is neutralised by alkalis, lime, or strontia. Cochineal and helianthin change when the first function of the acid is somewhat more than neutralised, but phenolphthaleïn indicates accurately the neutralisation of the second acid function. Even in presence of four or five equivalents of baryta, the precipitate is tribarium arsenate, a result different from that obtained when arsenic acid is neutralised with lime or strontia, or when orthophosphoric acid is neutralised with baryta. C. H. B.