upon which it floats. Four-fifths of this water is then run off, containing about one per cent. of sulphate of potassa, which is utilized either by evaporating it in drying-houses, or by running it off upon dry earth exposed to the air, which, when sufficiently charged with the salt, is washed. Directly after this operation, the basin is filled again with a fresh portion of soap-suds, which float the fatty matter and permit it to be run off into a side tub. The product obtained is a mixture of unaltered oil, the acids, animal matters, and a large quantity of water which forms with them a species of hydrate. This water is disengaged by injecting several times into the mass a current of steam, which heats it and facilitates its evaporation. The fatty matter is then run off into a boiler, where it is submitted to a rapid ebullition, aided by continual agitation, which drives off the last portions of water. The product contains twenty or twenty-five per cent. of impure matters which colour it and render it turbid. To purify it, it is poured into basins of copper and mixed with two per cent. of concentrated sulphuric acid. After two days the limpid oil comes to the surface, while the impurities are precipitated to the bottom.

The oil is carefully separated, and the deposit, when filtered through cloths in a press, gives still a large quantity of oily products, which are added to the preceding and made into soap by treating them with common soda.

The residuum is black and very thick; from it M. Houzeau produces the gas for lighting, but before introducing it into the retort, he liquifies it by means of the empyreumatic oil obtained in the preceding operation.

The gas thus prepared is purified by lime, and the water from the washing contains sufficient cyanide of calcium for the preparation of Prussian blue from it by treating it with sulphate of iron and washing the precipitate with muriatic acid.

This gas possesses a considerable lighting power, and in order to apply it to the lighting of the establishments scattered throughout the city of Rheims, M. Houzeau has contrived a manner of transporting it, at the same time simple, economical and free from danger.

On the Want of Identity between Microlite and Pyrochlore. By CHARLES UPHAM SHEPARD, M.D., Prof. Chem. in the Medical College of the State of South Carolina.*

PROF. SILLIMAN having apprised me on my recent arrival in town, that a paper was about to appear in this journal by Messrs. Teschemacher and Hayes of Boston, the object of which would be to establish the identity of Microlite with Pyrochlore, and being permitted by the editors, I hasten to reply in the same number.

From the "American Journal of Science and Arts."

The opinion of the identity originated with Mr. Teschemacher; and is expressed without the slightest reserve, resulting as he remarks, "from a close examination of several crystals of the mineral named microlite;" but Mr. T. singularly enough omits altogether the results of any researches bearing on the point to be established, if we except the mention of a mineralogical property which is simply descriptive, never characteristic, viz., the circumstance that in microlite, the colour is "transparent (!) straw-yellow to brick-red and dark brown!"* For the full confirmation of his opinion he then refers to the analysis of Mr. Hayes, which was undertaken at the request of the Chemical Society of Boston.

I shall first exhibit the mineralogical differences between the two minerals, in order to show the fallacy of any attempt to unite the two species on natural-history grounds; and shall then inquire what support the new view acquires from chemistry.

Pyrochlore.

Microlite.

Crystals, unmodified regular Crystals, regular octahedron, octahedrons.

Cleavage, none (fracture conchoidal).

Lustre, resinous to vitreous.
Colour, deep red-brown to black.

Streak, brown.

with edges truncated and angles surmounted by four-sided pyramids, whose faces correspond to the octahedral planes.

Cleavage, octahedral, distinct.

Lustre, resinous.

Colour, pale honey-yellow. Surfaces of crystals often reddish or blackish brown from implanted minerals.

Streak, pale yellowish white.

Translucent on the edges, to Semi-transparent to translucent.

I consider it apparent therefore, that all attempts to unite two minerals, whose individuals afford no transition-links to constitute

• In this account of colour however, my specimens do not agree.

The specific gravity as heretofore quoted, was 4.75 to 5.00; but very opportunely for this examination, I had brought with me from my Charleston cabinet an excellent crystal, weighing 3.805 grains; and possessing one of Robinson's best eight inch-beam balances (which turns readily with onethousandth of a grain), I was able to correct the specific gravity, as above, The original determination of this property was effected by means of a balance much less sensitive, and on a crystal whose weight was only four-tenths of a grain. Indeed, this was the largest crystal I had seen, when I described the species. Observing however, that my crystal of 3.805 grains had two of its octahedral sides pitted from the implantation of calcareous spar and green tourmaline, I at first took its specific gravity with these impurities attached. The result was 5.485 I then cleaved off these faces, thereby reducing the crystal to 2.410 grains, when its gravity rose to 5.562.

A a

the passage of insensible gradation, in properties so essential to specific gravity, internal structure and hardness, must prove unavailing; and according to the rules of forming the species in mineralogy, microlite must still stand distinct from pyrochlore.

Let us now turn to a review of the analysis of Mr. Hayes. This gentleman has not in any form, alluded to a paper of mine, entitled "Chemical Examination of Microlite," and published in vol. xxxii, p. 338 of this journal. To save the reader the trouble of referring to that article, I will simply quote from it so far as to say, that I found the microlite to be a columbate of lime and yttria, with moisture and traces of uranium and tungstic acid. The details of the examination were fully given in that communication; and in particular, the proofs that the mineral was a columbate. My surprise was therefore not inconsiderable, to find the subject of my former labours held up in a light so novel, as that of being a salt of a totally new and non-isomorphic genus.

The paper of Mr. Hayes in support of his views is out of proportion full on points where the evidence is unsatisfactory; and silent where the reader might appreciate its value as affecting the point at issue; for I hold the blowpipe characters of complex bodies like the microlite to be simply sufficient for furnishing the chemist to tolerable guesses, in aid of subsequent and more certain experiments with chemical reagents upon the decomposed and separated ingredients of the mineral. After nearly a page of blowpipe results, many, if not all of which would be likely in other hands (with different lamps, blowpipes, and lungs, as well as different sized specimens and proportions of fluxes), to be afforded from specimens of half a dozen other species as well as from microlite, we have the steps of the analysis given in the following words. "One grain of the mineral which had been dried was decomposed; the titanic acid carefully separated and dried, it weighed 80; this had the chemical characters of pure titanic acid.

"The solution of the oxide from the titanic acid gave a precipitate of sulphate of tin, when treated with sulphohydric acid, and the sulphate oxidized before the blowpipe gave with soda a globule of tin. Sulphohydrate of ammonia gave a black precipitate, which when roasted contained oxides of iron and uranium, with traces of oxide of cerium. The fluid remaining gave with oxalate of ammonia, a precipitate which was converted into sulphate of lime, equivalent to 0.08 lime. Thus,

Titanic acid....

Oxide of tin

80.0

• The sulphuret of tin is here referred to by Mr. Hayes, in this as well as in the previous line.

"This mineral is therefore identical with that analyzed by Wöhler. The absence of protoxide of manganese and water, and the smaller proportion of oxide of iron in this specimen, as indicated by the blowpipe experiments, will account for the larger proportion of titanic acid given in the analysis."

Then follows a quotation of the analysis of pyrochlore by Wöhler, with which Mr. Hayes identifies his own results upon the microlite. But when we find, on comparing the two analyses, a difference of more than seventeen per cent. in the acid, and of one-third in the principal base, as well as a wide discrepancy in volatile matter, and call to mind that these results are deduced from a single trial, and that upon one grain's weight of the mineral, we must believe, that the conclusion of Mr. Hayes cannot be sustained upon such slender grounds; and, inasmuch as Mr. Teschemacher refers to the analysis for the confirmation of his views, is it not possible that Mr. H. in turn, has unconsciously been indebted to the mineralogical evidence for a portion of his confidence?

In order to speak from present knowledge on the subject, although I have no reason to doubt the correctness of former results, I subjected 2142 grains of the above-mentioned crystal to the following examination; premising, however, that my inquiry was chiefly directed to the proofs of columbic acid in the mineral, and not to a rigid determination of proportions in the different constituents.

A. The powdered mineral was white, with only a faint tinge of yellow. Heated for half an hour to redness, it scarcely changed in colour; and lost 0.060 in weight, which equals 2.801 per cent. It was then intimately mixed with six times its weight of bisulphate of potash, and the mixture maintained in fusion for half an hour. Successive portions of water were boiled upon the fused mass for upwards of half an hour, until every thing soluble was taken into solution. A dense white powder remained undissolved.

B. Hydrosulphate of ammonia produced in a portion of the sulphate A, a precipitate, without sensitive discolouration, and which, from former proofs in my paper of 1835, was taken for yttria. The remainder of the fluid was then precipitated by oxalate of ammonia, the precipitate ignited, redissolved in hydrochloric acid, the yttria thrown down by ammonia, and the lime by oxalate of ammonia, which on ignition, weighed 0.032, or 1.49 per cent.*

C. The white insoluble matter A, was digested for some time in a saturated solution of hydrosulphate of ammonia, the operation being conducted on a filter in the way recommended by Berzelius. The insoluble matter was blackened by the affusion of the hydrosulphate.

• The calcareous content here obtained is quoted only as an approximation; but falling so much below what I previously found, and considered in conjunction with my observation of implanted calcareous spar upon the crystal, it is possible that lime may yet prove to be an accidental ingredient in the mineral.

D. The sulphohydric fluid was decomposed by nitric acid and heat; hydrochloric acid was added; no precipitation ensued. But the fluid, on the addition of ammonia, afforded a precipitate of peroxide of tin, which after ignition, weighed 001, which equals 0.047 per cent. The oxide was reduced with carbonate of soda on charcoal to the metallic state.

E. The blackened powder C was treated with dilute hydrochloric acid, and gently warmed. The filtered solution was precipitated by ammonia, and ignited. The peroxide of iron was apparently blended with traces of yttria. It weighed 0.022, or about 1 per cent.

F. The insoluble matter, washed by hydrochloric acid (E), was drenched upon the filter with an abundance of hot water: it showed no tendency to pass through the filter in a milky state, as titanic acid is well known to do, under such circumstances. It was ignited in a platina crucible, and exhibited a white colour while hot, as well as after cooling. Its weight after some accidental losses was 1.052 gr., 49.11 per cent. But I still regard my former determination of the proportion of acid, and which was 75.70 per cent., to be very near the truth.

A portion of the acid was fused with carbonate of potash in a platina crucible; water was boiled upon the mass, whereby a complete solution was effected. To a portion of the solution hydrochloric acid was added. It first occasioned a cloudiness, but subsequently the fluid became clear. The same effect was produced by oxalic acid. Ferrocyanide of potassium afforded with the oxalic solution a yellow precipitate, and the tincture of nut-galls with the same, a rich orange precipitate.

A portion of the acid was fused with biphosphate of soda on a platina hook, in the inner flame of the blowpipe: a transparent glass was obtained, which was colourless while hot, as well as after cooling. Fused with soda before the blowpipe, an opaque white bead was obtained.

My conclusion concerning the composition of the microlite, derived from the foregoing examination, coupled with that in 1837, is this, that it is a columbate of yttria and lime, with a little moisture, adventitious traces of iron, tin, tungstic acid, and uranium.

I do not possess enough of the mineral to enable me to undertake its analysis but it is a question which I regard with interest, and shall welcome its final settlement from whatever quarter it comes; though I must be excused, in the face of such evidence as I have already cited, from acquiescing in the mere declaration of any authority however high, that its prime ingredient is titanic, in place of columbic acid.

To show that a zeal to defend a species because it was put forth by myself, is not the motive for the present reclamation, I am free to say, that I have long perceived several striking analogies which point towards an identity between microlite and the yellow yttrotan