meteoric minerals were soluble in acid, the augite and enstatite were subImitted to this solvent action. Digested for several hours at 100° C. in hydrogen chloride diluted with half its volume of water, and subsequently in potash for some hours to remove the free silica, the augite and each of the three forms of enstatite proved to be acted upon, the results in all cases showing that the acid simply exercises a solvent action on the mineral, without separating it into two or more distinct silicates.

The subjoined Table gives the results of the experiments. The degree in which the acid dissolved the mineral was due to the more or less complete trituration of the material before treatment. In one case, for which the transparent variety was selected, a repetition of the process three times gave results that left no doubt as to the nature of the action of the acid.

Of the greyish-white variety of enstatite, after treatment for 20 hours with acid and 12 hours with potash, 9.414 per cent. dissolved, an analysis of which is given in column I.

Of the grey tubular variety of enstatite, after treatment with acid for 16 hours and with potash for a similar time, 7.779 per cent. dissolved, that gave on analysis numbers the approximate value of which is found in column II.

Of the white variety, after the first treatment for 20 hours with acid and subsequently with potash, 12.68 per cent. dissolved, the composition of which is given in column III. By a second treatment of the residual enstatite from this experiment, after 2 hours' trituration with acid for 30 hours and potash for 12 hours, 67.84 per cent. dissolved; and on subjecting the mineral to a third treatment in a similar way, 51·18 per cent. were dissolved in acid and potash. In the last of these experiments the ratio of the silica to the bases, neglecting the small amount of the former dissolved in the acid, is as 58.4 to 42.0, that of an analysis of an enstatite being as 58.4 to 41.6.

The solubility of the augite was determined by subjecting it to similar treatment with acid during 18 hours, and with potash for a like time, these reagents removing 7.384 per cent. of the mineral.

X. The Iron of the Busti Meteorite.

A small pepita of the iron contained in the meteorite was analyzed.

Omitting the silicate attached to the iron, the results of the analysis were as

The quantity was far too small to encourage a search for cobalt and other metals.

Besides the nickeliferous iron, which is disseminated very sparsely, and in particles singularly unequal in size and distribution, and with which troilite is associated in very small quantity, chromite is present as a constituent of small but appreciable amount. The crystals of this mineral are distinct and brilliant, and sometimes present good angles for measurement. One gave the solid angle of a regular octahedron.

The Manegaum Meteorite of 1843.

This meteorite fell at Manegaum in Khandeish, India, on the 26th July, 1843. Only a small fragment was preserved, and of this a portion was given by the Asiatic Society of Bengal to the British Museum in 1862. In 1863 I described its appearance as seen in section in the microscope, and gave the particulars of its fall (Phil. Mag. August 1863).

From the minuteness of the specimen I had very little material to work upon. One mineral is conspicuous in the stone, namely, a primrosecoloured transparent crystalline silicate in small grains, loosely cemented by a white flocculent mineral. This greenish-yellow mineral (I.) and a fragment of the entire meteorite (II.) were analyzed, and crystalline grains of the former were measured on the goniometer. The prism angles (1) for the prism {110} were about, and (2) for the prism {101} were 98° 8' for {100, 110} about 46°; for {100, 101}, 49° 4'; and for {110, 101} 58° 39'.

88°

The analyses gave the following numbers :

81°52'

The specific gravity of the granular mineral is 3.198, and its hardness

5.5.

The result of the above analyses is to show that, except for a little chromite and a little augite, with possibly in the crystallized mineral a little free silica, both that mineral and the collective silicate of the stone consist of a ferriferous enstatite.

The formula most in accordance with the analysis would be

(MgFe)O, SiO, ;

that of the enstatite in the Breitenbach meteorite is ( Mg Fe)O, SiO,. The bulk of the Busti meteorite consists of a purely magnesian enstatite; this of Manegaum is almost entirely an enstatite richer in iron than any yet examined. Both bear evidence to the white flocculent mineral which characterizes the microscopic sections of many meteorites, being composed of this now important mineral enstatite.

In publishing the results I have obtained in the attempt, so far as this memoir goes, to treat exhaustively of the mineralogy of two important meteorites, I wish to record the obligations I am under to Dr. Flight, Assistant in my Department at the British Museum, for his valuable aid in the chemical portion of the inquiry.

II. "On Fluoride of Silver.-Part I. By GEORGE GORE, F.R.S. Received October 5, 1869. (Abstract.)

This communication treats of the formation, preparation, analysis, composition, common physical properties, and chemical behaviour of fluoride of silver.

The salt was prepared by treating pure silver carbonate with an excess of pure aqueous hydrofluoric acid in a platinum dish, and evaporating to dryness, with certain precautions. The salt thus obtained invariably contains a small amount of free metallic silver, and generally also traces of water and of hydrofluoric acid, unless special precautions mentioned are observed. It was analyzed by various methods: the best method of determining the amount of fluorine in it consisted in evaporating to dryness a mixture of a known weight of the salt dissolved in water, with a slight excess of pure and perfectly caustic lime in a platinum bottle, and gently igniting the residue at an incipient red heat until it ceased to lose weight. By taking proper care, the results obtained are accurate. The reaction in this method of analysis takes place according to the following equation, 2AgF+CaO CaF,+2Ag+0. Sixteen parts of oxygen expelled equal thirty-eight parts of fluorine present. One of the methods employed for determining the amount of silver consisted in passing dry ammonia over the salt in a platinum boat and tube at a low red heat. The results ob

=

tained in the various analyses establish the fact that pure fluoride of silver consists of 19 parts of fluorine and 108 of silver.

Argentic fluoride is usually in the form of yellowish brown earthy fragments; but when rendered perfectly anhydrous by fusion, it is a black horny mass, with a superficial satin lustre, due to particles of free silver. It is extremely deliquescent and soluble in water; one part of the salt dissolves in 55 part by weight of water at 15°5 C.; it evolves heat in dissolving, and forms a strongly alkaline solution. It is nearly insoluble in absolute alcohol. The specific gravity of the earthy-brown salt is 5.852 at 15°5 C.; the specific gravity of its aqueous solution, at 15°5 C., saturated at that temperature, is 2·61. By chilling the saturated solution, it exhibited the phenomenon of supersaturation and suddenly solidified, with evolution of heat, on immersing a platinum plate in it. The solution is capable of being crystallized, and yields crystals of a hydrated salt; the act of crystallization is attended by the singular phenomenon of the remainder of the salt separating in the anhydrous and apparently non-crystalline state, the hydrated salt taking to itself the whole of the water. The fused salt, after slow and undisturbed cooling, exhibits crystalline markings upon its surface.

The dry salt is not decomposed by sunlight; it melts below a visible red heat, and forms a highly lustrous, mobile, and jet-black liquid. It is not decomposed by a red heat alone; but in the state of semifusion, or of complete fusion, it is rapidly decomposed by the moisture of the air with separation of metallic silver; dry air does not decompose it. In the fused state it slightly corrodes vessels of platinum, and much more freely those of silver.

The salt in a state of fusion with platinum electrodes conducts electricity very freely, apparently with the facility of a metal, and without visible evolution of gas or corrosion of the anode; a silver anode was rapidly dissolved by it, and one of lignum-vitæ charcoal was gradually corroded. A saturated aqueous solution of the salt conducted freely with electrolysis, crystals of silver being deposited upon the cathode, and a black crust of peroxide of silver upon the anode; no gas was evolved; with dilute solutions gas was evolved from the anode. By electrolysis of anhydrous hydrofluoric acid with silver electrodes, the anode was rapidly corroded.

The electrical order of substances in the fused salt was as follows, the first-named being the most positive: silver, platinum, charcoal of lignum-vitæ, palladium, gold. In a dilute aqueous solution of the salt, the order found was aluminium, magnesium, silicon, iridium, rhodium, and carbon of lignum-vitæ, platinum, silver, palladium, tellurium, gold.

The chemical behaviour of the salt was also investigated. In many cases considerable destruction of the platinum vessels occurred, either in the experiments themselves, or in the processes of cleaning the vessels from the products of the reactions.

Hydrogen does not decompose the dry salt, even with the aid of sunlight, nor does a stream of that gas decompose an aqueous solution of the salt, but the dry salt is rapidly and perfectly decomposed by that gas at an incipient red heat, its metal being liberated.

Nitrogen has no chemical effect upon the salt, even at a red heat, nor upon its aqueous solution. Dry ammonia gas is copiously absorbed by the dry salt. In one experiment the salt absorbed about 844 times its volume of the gas. The salt in a fused state is rapidly and perfectly decomposed by dry ammonia gas, and its silver set free. A saturated solution of the salt is also instantly and violently decomposed by strong aqueous ammonia.

Oxygen has no effect either upon the dry salt at 15° C., or at a red heat, nor upon its aqueous solution. Steam perfectly and rapidly decomposes the salt at an incipient red heat, setting free all its silver. No chemical change took place on passing either of the oxides of nitrogen over the salt in a state of fusion.

By passing anhydrous hydrofluoric acid vapour over perfectly anhydrous and previously fused fluoride of silver, at about 60° Fahr., distinct evidence of the existence of an acid salt was obtained. This acid salt is decomposed by a slight elevation of temperature.

Numerous experiments were made to ascertain the behaviour of argentic fluoride in a state of fusion with chlorine, and great difficulties were encountered in consequence of the extremely corrosive action of the substances when brought together in a heated state. Vessels of glass, platinum, gold, charcoal, gas carbon, and purified graphite were employed*. By heating the salt in chlorine, contained in closed vessels, formed partly of glass and partly of platinum, more or less corrosion of the glass took place, the chlorine united with the platinum and fluoride of silver to form a double salt, and a vacuum was produced. By similarly heating it in vessels composed wholly of platinum, the same disappearance of chlorine, the same double salt, and a similar vacuum resulted. Also, by heating it in vessels composed partly of gold, an analogous double salt, the same absorption of chlorine and production of rarefaction were produced. And by employing vessels partly composed of purified graphite, a new compound of fluorine and carbon was obtained.

III. "Approximate determinations of the Heating-Powers of Arcturus and a Lyræ. By E. J. STONE, F.R.S., First Assistant at the Royal Observatory, Greenwich. Received October 13, 1869.

About twelve months ago I began to make observations upon the heatingpower of the stars. My first arrangements were simply these: I made

In the next communication will be described the results obtained with vessels formed of other materials.