culties encountered and space required in stowing the two fuels. With respect to the comparative heating power of anthracite coal and petroleum, let us first mention their absolute theoretical steam-generating capacities; that is, if the combustible elements in both cases could be completely consumed, and all the heat thus generated be utilised in making steam. According to the accurate and elaborate experiments of the eminent chemists, MM. Favre and Silberman, a pound of carbon completely burned, so as to make carbonic acid (that is, by combining with 2 lbs. of oxygen), will evaporate 15 lbs. of water from a temperature of 212° Fahr.; and similarly, 1 lb. of hydrogen, by combining with 8 lbs. of oxygen, will evaporate 642 lbs. of water from the same temperature. The chemical composition of petroleum is said to be C2H12; or, in other words, it is composed of 12 atoms of carbon and 12 atoms of hydrogen. As an atom of carbon exce‹ ds an atom of hydrogen six times in weight, it is seen that in a pound of petroleum, 6 parts are carbon and 1 part hydrogen. The absolute theoretical evaporative power of a pound of petroleum, in pounds of water, from 212°, is therefore × 15+ 642=2202 lbs. Allowing that there is 20 per cent. of noncombustible matter in anthracite coal, there remains in a pound of a pound of carbon; hence × 15-12 lbs., the weight of water that theoretically may be evaporated from 212 by a pound of anthracite, under these conditions. Consequently, the absolute heating power of petroleum is 1835 times greater than that of a pound of anthracite. experiments on the Palos, conducted under the immediate supervision of the petroleum retort inventor, it will be seen that the result is still further against the economical use of the oil as a steam fuel. = We have now to consider the cost and durability of the apparatus for burning the oil. As to the cost, let us look at the expense of applying the retort system to vessels of the Wampanoag class, which have 58 furnaces in their boilers. It is estimated that such an apparatus would cost at least 250 dollars for each furnace, or 250 X 58 : 14,500 dollars, and when to this is added the cost of the air-pumps for pumping air into the retort, the engines for driving them, together with the pumps for forcing the petroleum into the retorts, and the paraphernalia of cocks, valves, and pipes connected with this machinery, we think it is safe to assume that the application of the retort system to such a ship would cost at least 60,000 dollars. When we come to. look at the durability of this apparatus, the case is still more strongly against the petroleum; and when the high temperature to which the apparatus is exposed is considered, it becomes evident that portions of it will require constant renewal. But the most fatal practical objection to the retort system remains to be mentioned: it is the deposition of carbon, coke, and incombustible matter within the retorts and pipes. In experiments with a retort apparatus, the pipes and passages became so choked in less than 48 hours that the fire went out, and could only be renewed by taking the fixture apart and cleaning it. With regard to storing petroleum on board ship, another serious difficulty is encountered. A great portion of the crude oil evolves, at ordinary temperatures, an inflammable gas; and this gas, when mixed with atmospheric air, forms an explosive compound, which instantly takes fire when brought in contact with flame. Other samples do not evolve this gas until the temperature is raised from 80° to 100° Fahr. Now let us look at the comparative cost: the average price of crude petroleum is about 21 cents per gallon, and coal about 5 dollars per ton: in other words, the petroleum costs 3 cents per lb., while the cost of the coal is 23 of 1 cent. Hence, a pound of petroleum, while it has but 1835 times more absolute heating power than a pound of anthracite, costs 13 times more than the coal; that is, the heat produced by petroleum costs over seven times more than the heat produced by coal. The greater cost of the petroleum is therefore decisive against its economical use as a steam fuel, and this, too, without considering the complication of the apparatus necessary for burning it. With regard to the compara-taining the petroleum should be immersed in water; tive cost of the attendance on the fire, economy in this particular is decidedly in favour of the petroleum. It is probable that a steam vessel which requires, say 24 men in the fire room, could dispense with at least 14 of them by the use of petroleum. So far, we have assumed that not only was all the heat generated that is attainable from the perfect combustion of the two fuels, but that the whole of this heat was utilised in the evaporation of water. It need scarcely be remarked that such results are quite out of the question in practice. The experiments carried on with the machinery of the U. S. gunboat, Palos, in Boston harbour, throw much light on the evaporative power practically attainable with the two fuels. The petroleum in these experiments was burned by the contrivance known as Foote's retort apparatus, and the result of these trials was, that if the evaporative power of the anthracite is represented by I, that of the petroleum is 138-a result much inferior to that which we have shown is due to the combustible elements of the two substances. This, perhaps, is due to the fact that it is more difficult to completely burn a hydrocarbon fuel than it is to burn one composed almost wholly of carbon. If we modify the calculations we have before entered into, according to the results attained with the These facts point out at once the extraordinary care that must be taken in storing this substance on board ship, in order to guard against accidents of the most frightful character. It therefore seems clear that a proper regard for safety demands that the tanks conwhen the weight of these tanks and cisterns is borne in mind, together with the bulk of petroleum and coal composition (a cubic foot of the former weighs 50 lbs., and a cubic foot of anthracite, as stored in bunkers, weighs 525 lbs.), and also their relative heating values, which may be set down as 14 for the former and I for the latter, it may be with safety assumed that there would be a saving in weight and space of not over 30 per cent. in stowage capacity by substituting the oil for the coal. It is therefore evident, from what we have said, that even assuming that the oil and the coal can be burned with equal facility, and with an equal degree of liability to derangement, in steam-boiler furnaces, the excess of the heating power of the petroleum over that of the oil is so very much less than its excess of cost, that there is not the slightest probability, as long as these ratios exist, of petroleum ever taking the place of coal as a steam fuel. When we look at the great complication and danger that must be added to steam machinery in order to burn the oil, and the liability to derangement and the want of durability, it is not likely that any prudent steam navigation company would allow it to be employed in their vessels, even if they could find an engineer to recommend it.-American Artisan, ON MONOCHROMATIC LIGHT. BY HENRY MORTON, PH.D., PROF. CHEM. AND PHYS. UNIVERSITY PENNA. THE difficulty of producing monochromatic light in great quantity prevents us from demonstrating satisfactorily many points of interest in connection with the composition of light, and the theory of vision. Coloured glasses placed in the path of powerful beams of white light are doubly unsatisfactory; they reduce the amount of light enormously, and, with few exceptions (e. g., red glass coloured with gold), yield a beam of mixed colour. The old experiments of snap-dragon, or the spirit lamp with salted wick, are admirable as far as they go, but yield a very faint light at best. Something far superior to this is furnished by the arrangement of a ring of cotton wick wound on a wire, and supported immediately over and around a large Bunsen burner, the wick being soaked with an aqueous solution of some flame-colouring salt. This plan in effect is suggested in Sir David Brewster's natural magic; but as the Bunsen burner was not then known, a less simple arrangement is described to perform the same office. The drawbacks to this arrangement are: the trouble of adjusting the cotton wicks, the delay in changing to produce a new colour, and the brief duration and, to a certain extent, irregular amount of the effect. After lighting, the burners increase their effect to a certain point, and then soon rapidly diminish in the intensity of their light. When a large number of burners are to be used these difficulties become serious, and I have therefore devised the following arrangement, which has proved, on trial, thoroughly efficient:-The burners to be used, varying in number from 5 to 30, in different experiments, are enclosed below in a box with a single large entrance, opposite to which is placed an atomiser, operated either by steam or compressed air. A spray of the colouring solution is thus mixed with the air supplying the burners, and their flames are thus tinged with the greatest ease, certainty, and intensity of effect, the whole action being entirely under control, and capable of being maintained indefinitely, while the change from one colour to another is effected by simply transferring the tube of the atomiser from one solution to another. For experiments demanding diffused light, this apparatus is most satisfactory. Five large Bunsen burners, thus arranged, light up the lecture room of the Franklin Institute, which seats about 350 persons; and with 60 burners, arranged in two groups of 30, I expect to illuminate sufficiently, at my next lecture, the Academy of Music, which has 3,500 stalls. Of course in neither case is the light brilliant, but simply sufficient to show in the most distant portions of the room the peculiar effects of this light on the faces and dresses of the audience themselves. Any object near the flame is, however, brightly illuminated. The stage being set with the most brilliant scenery and coloured decorations, suddenly illuminated by yellow light thus developed, the ordinary gas-light being turned down, presents an appearance of ghastly change, which cannot be appreciated until seen, while sufficient light reaches all parts of the house to allow the audience to repeat the strange observation of metamorphosis upon their neighbours. In certain experiments, however, it is necessary to have monochromatic light of great intensity and concentration. This I have found could be furnished in the case of yellow light, to a certain extent, by substituting in the gas microscope or polariscope a soda-glass rod for the ordinary lime cylinder, and so adjusting its position that the rays from the heated glass should be cut off from the lenses, and only the light of the yellow flame should reach them. We can thus show, upon a screen five feet in diameter, the greatly-increased number of rings developed by a section of Iceland spar in monochromatic light. With the spectroscope we can also project the sodium line on the screen, so as to be well seen by an audience of 500 persons. By afterwards producing the absorption-band due to the vapour of the substance, with the aid of a small Bunsen burner and iron spoon with sodium, and again showing the absorption-bands of nitric peroxide and of cochineal, the characteristic phenomena of spectrum analysis may be sufficiently illustrated, without the trouble and expense of the electric light. We have, in fact, with a large and efficient battery at command, generally employed the above arrangement by preference. The details of adjustment we will furnish, with pleasure, should any one think it of sufficient interest to ask for them. ON THE PREPARATION OF MAGNESIA FOR RESISTING HIGH TEMPERATURES. BY M. H. CARON. ABOUT two years ago I briefly pointed out the alvantages which would accrue to metallurgy from the employment of magnesia as a refractory material. I regretted, at the same time, the high price of this earth, the employment of which appeared likely to be confined to the laboratory. Now, circumstances have happily changed; recent modifications introduced into the manufacture of cast steel, and especially the employment of Siemens's furnace and Martin's process, absolutely demand more refractory bricks than those at present in use, irrespective of price. On the other hand, native carbonate of magnesia, which formerly cost 250 fr. the 1,000 kilos., may row be obtained at the price of 70 fr. delivered at Marseilles, or 100 fr. delivered at Dunkirk. Calcination at the place where the carbonate is obtained, may still further reduce its price.* I desire now to describe my processes for its agglomeration, which may be employed by the chemist for the ready preparation of refractory vessels of all forms; by the physicist to obtain pencils for oxyhydrogen lighting purposes; and also by the manufacturer, to replace, in some cases, fire-bricks which have become insufficient in carrying out different processes of heating. The magnesia which I employ comes from the island of Elba, where it is found in considerable quantities as a native carbonate, white, very compact, and of great hardness. This carbonate contains traces of lime, silica, and iron; it is, besides, interspersed sometimes with serpentine and large plates of silica, which would diminish the infusibility of the substance, and render This preliminary calelnation requires less heat than burning lime, and diminishes the weight of the carbonate one half. CHEMICAL NEWS, it especially unfit for oxyhydrogen illumination if their removal is neglected (I will presently give the reason). These plates are, however, easily recognised, and may be readily separated, even in working on the large scale. In the case of refractory bricks, the presence of a small quantity of these foreign bodies would, at the most, give rise, under the influence of the highest temperatures, to a slight vitrification, offering no serious inconvenience. taining siliceous matter, that the vessels before or after firing do not possess quite the desirable solidity; they should then, to acquire this, be simply moistened in a cold saturated solution of boracic acid, dried, and then fired as before. This operation does not render the magnesia more fusible: it only causes the grains of the substance to cohere more strongly together. Magnesia, very pure, strongly calcined and finely pulverised, may be employed in the form of paste (barbotine) and yield the most delicate and translucent crucibles, as well as the sharpest and most complicated impressions. I am convinced that some time hence this earth will be advantageously employed in the ceramic art in spite of the difficulties of its moulding compared with those of porcelain paste.-Comptes Rendus, Ixvi. 839. Before powdering the carbonate, it is advisable to bake it at the temperature necessary for the expulsion of the carbonic acid; the material then becomes very friable, and may be pulverised more easily. It is then possible to separate the serpentine and silica, which do not become friable under the influence of heat. This preliminary treatment does not permit of the agglomeration of the magnesia, and even were this difficulty to be overcome, a temperature higher than that of the original calcination would cause an enormous contrac- ON A NEW PROCESS IN MINERAL ANALYSIS. tion, producing fissures and alterations in shape, which would interfere with the use of this substance. It is, therefore, indispensable, before moulding the magnesia, to submit it to a very intense heat, at least equal to that which it is intended to support subsequently. Thus calcined it is not plastic, its appearance is sandy, and compression does not cause it to acquire any cohesion; a mixture of magnesia, less calcined, imparts to it this quality. The quantity of the latter to be added necessarily varies with the degree of calcination of the two magnesias; it is scarcely one-sixth of the weight of that which has been exposed to the temperature of melting steel. It only now remains to moisten it with 10 or 15 per cent. of its weight of common water, and strongly compress it in iron moulds, as adopted in making artificial fuel. The brick produced in this operation hardens on drying in the air, and becomes still more resisting when it is subsequently calcined at a red heat. The same process would appear practicable, varying the form of the moulds, for obtaining crucibles of great capacity; but compression is difficult in large masses, as well as when the moulds have a large surface, as the magnesia adheres strongly to the sides. Although I have been able to obtain small crucibles for the laboratory, I do not consider this process adapted to make the large crucibles employed in steel melting. In this and other cases it is preferable to agglomerate the magnesia in the humid way. To endow magnes'a with a sort of plasticity, I have profited by a property of this earth pointed out in "Berzelius's Chemistry." When strongly calcined, and then moistened, it hardens in drying. This fact is doubtless due to a hydration which takes place without sensible increase of temperature. I have also remarked that when solidified in this manner the magnesia only loses the assimilated water at a high temperature. Then calcination not only does not disintegrate it, but on the contrary confers upon it a hardness and resistance comparable to those of ordinary crucibles after their baking. This being understood, the application of this fact is obvious. Thus, the magnesia to be employed in the manufacture of crucibles, should only be moistened, moulded into shape, dried, and then ignited. For the construction of steel melting furnaces, a paste of moistened magnesia should be plastered over the walls; it will become ignited in due course without any particular precautions being taken. It sometimes happens, however, either owing to the magnesia being too much or too little hydrated, or owing to its con BY FRANK WIGGLESWORTH CLARKE, S.B. In the course of some experiments upon the alkaline fluorides, I found that the most refractory minerals, such as emery, chromite, tinstone, rutile, &c., were completely and easily resolved by fusion with a mixture of fluoride of sodium and bisulphate of potassa. The operation is performed as follows:-One part of the finely pulverised mineral is mixed in a platinum crucible with three parts of fluoride of sodium, and upon the top of this mixture are placed twelve parts of bisulphate of potash, which may be either in powder or in small lumps. Upon heating, the mixture boils up strongly, and after a while settles into a clear, tranquil fusion. The boiling is chiefly owing to the action of the reagents upon the mineral, and not, as might be supposed, merely to the influence of the bisulphate upon the fluoride. This is shown by the fact that, whenever the reagents are heated together without minerals, although some boiling takes place, the addition of a little powdered chromite or iron ore fully doubles the violence of the action. In quantitative analysis it is necessary to keep the crucible closely covered in order to avoid loss from spattering; and to heat carefully, so that the mass may not boil over. The bisulphate should never be mixed with the fluoride and mineral, because a portion of the assay is then apt to escape action, being left on the sides of the crucible by the boiling of the mass; but should be placed upon the top of the mixture as above directed, as then the decomposition is complete. The mass obtained by this fusion is, in the case of some minerals, completely soluble in water. In other cases, basic salts are formed, which although insoluble in water, dissolve readily in hydrochloric acid. Almost all of the latter class may be rendered soluble in water by the following process:-The fused mass, after cooling without removal from the crucible, is treated with a small quantity of strong sulphuric acid, and again fused. The mass thus obtained is entirely soluble in water. There are exceptions to this rule, however. I will now describe the results I have obtained with various minerals, and for the sake of brevity, will speak of the fusion with bisulphate and fluoride as fusion No. I, and the subsequent treatment with sulphuric acid, as fusion No. 2. Tinstone is completely resolved, giving a white mass almost entirely soluble in cold water. A very small quantity of stannic acid remains undissolved, however, but by first treating the mass with hydrochlo acid and then adding water, complete solution is ensured. By rendering the solution nearly neutral, taking care, however, to leave a slight excess of sulphuric acid, then reducing the iron by sulphuretted hydrogen or hyposulphite of soda, and then boiling for some time, all the tin is precipitated in the form of stannic acid. If the ore contains tungstic, niobic, or tantalic acids, these also will be completely precipitated. A second fusion is of no advantage with tinstone. This is the only mineral which I have yet tested which was not completely resolved by this process, over an ordinary Bunsen's gas burner. This substance required the heat of a blast lamp. Wolfram is entirely decomposed, affording a pale yellowish mass, partly soluble in water. Hydrochloric acid dissolves a part of the residue, but a white powder, probably the hydrate of tungstic acid, remains unattacked. Fusion No. 2 possesses no advantages with wolfram. I should not recommend the use of this process for the analysis of wolfram, as it is so readily decomposed by nitric acid. I made the experiment, however, because as wolfram is so often mixed with tinstone, it se med to me necessary to know what the reaction would be. Rutile. This mineral is rapidly and easily resolved, giving a mass when cooled, of a yellowish white colour, and, as I obtained it, presenting a beautiful mottled appearance, being crystalline in structure, and in some parts nearly transparent, but opaque in others. This mass dissolved entirely in cold water, rendering a second fusion unnecessary. From this solution all the titanic acid may be thrown down by boiling. Niobite is decomposed with the greatest ease. The mass is pale yellow, and partially soluble in cold water, but a small quantity of niobic acid remains undissolved, even in hydrochloric acid. But it is always best to treat with hydrochloric acid, in order to dissolve any basic sulphates of iron that may be formed. Upon boiling the filtered solution, niobic acid is precipitated. Fusion No. 2.-A larger proportion of the mass is soluble in water than with the first fusion, as no basic salts of iron are present. Ilmenite is completely decomposed, giving a mass closely resembling that obtained from rutile. Cold water dissolves a large part of this, but leaves some basic salts which dissolve readily in hydrochloric acid. Fusion No. 2.-The resulting mass dissolves completely in cold water, and by boiling the solution, the titanic acid can be precipitated. Chromic iron ore is decomposed very easily. In one case in which I timed the operation, the fusion was complete in less than three minutes from the time I began to heat, and that over an ordinary Bunsen's gas burner. The cooled mass is light green, partly soluble in water alone, and entirely soluble in hydrochloric acid. Fusion No. 2.-The mass possesses a deeper green colour than that obtained by the first fusion, and a larger proportion of it dissolves in water. In every fusion that I have yet made of chromite, however, a small quantity of basic salts was formed, requiring treatment with hydrochloric acid. Emery is rapidly and easily resolved. The mass contains basic compounds that require hydrochloric acid for complete solution; although water dissolves a large proportion of it. Fusion No. 2.-The mass thus obtained is entirely soluble in water. Hæmatite.-(An exceedingly hard specular ore from the Tilden mine, Lake Superior). It was completely resolved, giving a mass partially soluble in water, but dissolving entirely in hydrochloric acid. Fusion No. 2.-The mass dissolves completely in water. Limonite and magnetite behave exactly like hæmatite. Zircon is entirely decomposed. All its silica is converted into the gaseous fluoride of silicon, and driven off. The mass obtained resembled that given by rutile. It dissolved almost entirely in cold water, but the solution speedily became turbid and deposited a white precipitate, which was either zirconia or some basic salt of that oxide. By first digesting the mass with a little strong hydrochloric acid and afterward adding water, the whole went into solution. Fusion No. 2, afforded a mass of a beautiful waxy lustre, which was completely soluble in water. Kyanite is entirely resolved, and, like zircon, freed from silica. The white mass contained basic compounds, and consequently required hydrochloric acid for complete solution. A second fusion gave a mass entirely soluble in water. Orthite is completely decomposed, and deprived of silica. The mass was white, and dissolved partly in water. The insoluble residue contained basic salts and some sulphate of lime, and hydrochloric acid dissolved all but the latter. Fusion No. 2.-With the exception of a little sulphate of lime, the mass dissolved in water. The Quartz sand.-I subjected some of this substance to fusion with the mixture of bisulphate of potash and fluoride of sodium, in order to ascertain to what extent silica would be converted into fluoride of silicon. fusion took place very easily, giving a white mass which dissolved almost entirely in water. Only a very small trace of insoluble residue remained, probably not more than one-tenth of one per cent. of the quantity of sand taken. It is very probable that more careful treatment would get rid of even that small amount. After this it seems impossible to doubt that any and all silicates would be decomposed and freed from silica by this process. A convenient method is thus afforded for the estimation of bases in silicates, bearing in mind, however, that a second fusion with sulphuric acid will in most cases be necessary. As far as concerns the complete resolution of any mineral, pure, finely pulverised cryolite may be substituted for fluoride of sodium. It labours under the disadvantages, however, of introducing alumina into the mass obtained, but nevertheless it can be advantageously employed in cases where the introduction of alumina is a matter of indifference, as in the technical analysis of tinstone and chromite, and the e-timation of iron in ores, slags, and cinders. The perfectly white translucent specimens of cryolite should be chosen for this purpose. Either the bisulphate of potash or of soda may be used, and although neither seems to possess any advantage in point of thoroughness, the potassa salt appears to be the most readily fusible, and is therefore to be preferred. The following are the advantages that I claim for this process: First. Speed. Among the different minerals upon which I have tested the action of this mixture, I have found no case where I was obliged to heat longer than five minutes, and many fusions are complete in three, and even less. When bisulphate of potash alone is used for a similar purpose, it is usually necessary to heat for from half to three-quarters of an hour; and even then in the cases of emery, chromite, and some other minerals, it is almost impossible to obtain absolutely complete resolution. By my process, even when the second fusion with sulphuric acid is necessary, not more than twenty minutes should be consumed in both fusions and the time for cooling between them. Second. In every case except tinstone that has come under my observation, in my process, an ordinary Bunsen's burner may be used as the source of heat; whereas, by most other methods, a blast lamp is necessary. Third. When bisulphate of potash alone is used for the decomposition of chromite, &c., it is necessary that the mineral should be reduced to extremely fine powder; but, when the mixture of bisulphate and fluoride is employed, although the mineral should be in fine powder, such an extreme state of subdivision is by no means required, and thus much labour is saved. Fourth. In all cases when this mixture is used, the resolution of the mineral is absolutely perfect. Furthermore, all the silica is got rid of at once, and all the bases present are converted into sulphates. This last is a great advantage whenever iron is to be determined volumetrically by means of hypermanganate of potassa solution. As far as thoroughness of action is concerned, or generality of application, I can claim no advantage for my process over the use of acid fluoride of potassium; but the last-named reagent is difficult to prepare, and when prepared, cannot be preserved in glass vessels. Fluoride of sodium is subject to neither of these disadvantages, and in the mixture with bisulphate of potassa, has slight advantages as regards rapidity of action. It may not be out of place for me to mention here a fact connected with the use of fluoride of ammonium for decomposing silicates, as described by Rose. After fusing the mineral with the fluoride, he directs treatment with sulphuric acid, for the purpose of converting the bases into sulphates. I find that if, in the first place, sulphate of ammonia is mixed in excess with the fluoride, all the bases are directly converted into sulphates, thereby obviating the necessity of treatment with sulphuric acid. tion, when I came to heat the resulting mass with sulphuric acid, red fumes were given off, which were probably the so-called terfluoride of chromium. Technical Estimation of Iron in Ores, Slags, and Cinders. After fusion with cryolite and bisulphate of potash, the mode of treatment varies according to the method it is desired to use for determination of the iron. If Penny's process of estimating iron volumetrically with bichromate of potassa, or the ordinary method of precipitation with ammonia, is to be employed, the mass may be treated with hydrochloric acid, and thus brought immediately into solution. If, on the contrary, the iron is to be determined volumetrically by means of hypermanganate of potassa, the use of hydrochloric acid is interdicted, and it becomes necessary to fuse again with sulphuric acid and dissolve in water. When volumetric processes are used, the whole determination of the iron, from the time the assay is placed in the crucible, should take less than an hour to perform. This process has a great advantage over all others, in the examination of ores, slags, and cinders containing iron, both as regards speed and convenience. A perfectly clear solution is immediately obtained without filtering, all the silica is got rid of, and it is only necessary to reduce the iron with hydrogen, and then to titrate. If in an iron ore it is desired to determine also titanic acid and manganese, it is best to make the subsequent fusion with sulphuric acid. The clear solution obtained is diluted to a known quantity, and by means of a graduated pipette divided into several portions. In one part the iron may be reduced and determined volumetrically; in another, the titanic acid thrown down by boiling. In still another portion, the iron, alumina, and titanic acid may be thrown down by boiling with acetate of soda, and in the filtrate, the manganese may be precipitated by a current of chlorine gas, or by boiling with hypochlorite of soda. If fluoride of sodium be used instead of cryolite in the fusion, lime and magnesia also may be estimated. For determining silica, phosphorus, and sulphur, other methods must be employed. Preparation of Fluoride of Sodium.-When pure finely pulverised cryolite is boiled with caustic soda solution, it is, as is well known, decomposed. Fluoride of sodium is deposited as a jelly-like mass at the bottom of the vessel, and the supernatant liquid contains alumi of soda. The deposit of fluoride is to be washed with cold water until the washings have no longer an alkaline reaction; and is then purified by repeated solution and evaporation. It is so difficultly soluble, and crystallises so badly, that this last is a matter of some difficulty; but the first part of the process is so very easy that the crude fluoride of sodium can be prepared by the hundred-weight without much trouble. Iron vessels are suitable for the operation, but must be very clean and free from rust. If caustic potassa be substituted for soda, the deposit of fluoride of sodium is smaller, and the supernatant solution contains aluminate of potassa, fluoride of potassium, and a little fluoride of sodium. Technical Determination of Chromium in Chromite.-nate of soda, a little fluoride of sodium, and the excess After fusion with cryolite and bisulphate of potash, as previously directed, the mass is to be treated with a little strong hydrochloric acid, and allowed to digest for about ten minutes. Then, upon boiling with water, the whole dissolves. The solution should then be neutralised, acetate of soda added, and the chromium oxidised to chromic acid by a current of chlorine gas, or by boiling with hypochlorite of soda solution. The chromium may then be separated from other substances, as directed in Professor Gibbs's paper, (Am. Journ. Sci., January, 1865). | When chromite is fused with bisulphate of potash and cryolite, and saltpetre is added to the mass, as soon as clear fusion is obtained, the chromium is nearly all oxidised to chromic acid. If the mass be boiled with a solution of carbonate of soda, and the liquid filtered, a filtrate is obtained which contains nearly all, but not quite all the chromium as alkaline chromates, free from iron or alumina; but, invariably, the residue upon the filter contains traces of chromium. When chromite is fused with the acid fluoride of potassium, a part of the chromium is usually oxidised to chromic acid by the oxygen of the air; and in one case that came under my observa Possibly the fluoride of potassium might be prepared in a state of purity from this solution, but it is extremely problematical whether this could be done economically. When pure cryolite is fused with carbonate of soda, and the used mass powdered and treated with water, fluoride of sodium is dissolved out. This method, however, cannot compare with the first for convenience and economy. It may not be altogether out of place to remark in |