(2.) Transferring the oil into metallic tanks. (3.) Making a large tank in masonry, filling it with water, and plunging into it, mouth downward, a vessel, like a gasholder, containing the petroleum, which is to float on the water within the inverted vessel. (4.) Attaching weights to the ordinary barrels and sinking them in water. The Author points out that if a vessel laden with petroleum takes fire in a crowded port, it is worse than useless, so far as the other vessels are concerned, to scuttle her, because the water rushing in displaces the petroleum, and thus causes it to float about over the surface of the water instead of being confined to the burning ship. The Paper concludes with two tables, one of which shows the relative proportion of the various products obtained by fractional distillation from different petroleums, and is given below in extenso. The other table, based on the experiments of M. Henri SaintClair Deville, gives the specific gravities, the co-efficient of dilatation and the weight of water that can be evaporated by each of fortyone different mineral oils. The specific gravity varies from 0.786 for petroleum of Parma to 1.044 for the heavy oil of the Parisian Gas Company. The co-efficient of dilatation ranges from 0.000641 in the case of petroleum of Hanover (Wilze) to 0.001 in the case of the Parma petroleum (before mentioned) and of Canada West; while the power of evaporating water lies between 12.240 times the weight of the oil for the crude petroleum of the schists of Vagnas (Ardèche) to 15 364 times the weight of the combustible for the oil of Schwabwiller (Bas-Rhin). D. K. C. Respective Merits of Blast-furnace or Cupola Castings. (Berg-und Hüttenmännische Zeitung, No. 2, 1874.) It being often specified in orders for gas and water tubes that the casting be made from the cupola, and not direct from the blast furnace, the Author has been led to investigate the inherent changes, produced by remelting pig iron in a cupola, that make it denser and stronger than when flowing directly from the blast furnace. No doubt there is an oxidising effect' great in proportion to the height above the tuyeres at which the iron melts; greater, therefore, in the cold-blast furnace, and in the old form of cupola, worked with small blow-pipes and a high pressure of blast, than in the more recent forms. The best of all is Krigar's cupola, in which the blast, passing through wide openings partly filled with red-hot coke, has its free oxygen taken up and neutralised before coming in contact with the iron. The value of this oxidising effect of melting in cupolas depends upon the purpose for which the casting is intended. Iron, carbon, silicon, in a less degree sulphur, and in a still less degree phosphorus, are burnt out, and a less carbonised, purer, denser iron is the result. For most purposes an iron is wanted soft enough to work, but strong and uniform in texture, and free from hard lumps, formed of an agglomeration of crystalline grains, with pure carbon, much of it as graphite, very free from silicon, sulphur, and phosphorus. Too much graphite gives the iron a looser texture and renders it weak, as does also silicon; too little, makes it liable to break up under the tool. Sulphur renders it sluggish when molten, and honeycombed when cold, from the fact of sulphur gases being given off. Phosphorus, though making it liquid when molten, causes it to be brittle when cold, and tends to make white iron. Pig such as No. 1, always rich in silicon, and often containing much sulphur, is improved by one, two, and even three remeltings. But less graphitic iron, such as No. 3, coke pig and all charcoal pigs which contain the right proportion of graphite, are not improved by it; charcoal pig, in fact, cannot be remelted alone, though it improves No. 1 when remelted with it. The absorption of carbon is directly proportional to the time the iron remains unmolten after having been reduced from the ores. The proportion of combined to uncombined carbon is a function of the temperature of the smelting zone, or rather of the degree to which the iron is superheated. A high temperature works powerfully in producing graphitic iron, and this in two ways; directly, by changing the already formed carburet (Fe, C) into one less 1 This oxidising effect of the blast is relatively much less in the blast furnace, because there the iron is protected not only by the slags, but also by the greater proportion of fuel to iron; and the blast being hot, the active oxygen is more rapidly taken up, and its effect neutralised by the carbon. soluble; and indirectly, by reducing and dissolving foreign substances, such as silicon, which prevent the intimate chemical combination between iron and carbon. The uncombined carbon of the first sort is easily changed into combined by remelting at a low heat and sudden cooling. But the uncombined carbon of the second sort, such as Scotch foundry pig, remains grey even when rapidly cooled. In general it may be assumed that ores easy to reduce and smelt with a liquid slag yield, when the temperature of the smelting zone is low, a white highly-carbonised pig, and when the temperature of this zone is higher, a grey or mottled pig. With ores easy to reduce, but hard to smelt, the remelting pig is grey with little combined carbon, such as No. 2, made up chiefly of crystalline grains of pure iron. Ores hard to reduce, but easy to smelt, always yield a white pig iron poor in carbon. The first group includes spathic ores and spherosiderites, which are the best for getting highly-carbonised white irons. Between the first and second would come brown hæmatites of recent geological formation, such as bog ores, minettes, and bean ores (the last two from Luxembourg), which are suitable for either grey or white iron. To obtain grey irons from these, a high heat in the hearth, got by a highly heated blast, is required, and the slag should be a basic refractory one. The second group is represented by nearly all the red hæmatites, as well as the brown hæmatites from the older formation. The charcoal irons of the Harz and Nassau, and the Cumberland pigs, are good examples of the produce of these last. Magnetic ores belong principally to the third group, from which most Swedish pig is produced. The last group is made up almost unexceptionally of forge and finery cinders. When grey iron is to be produced only a small percentage of these can be used. As resistance to breaking may generally be taken as the chief requisite in cast iron, the Author gives some tables of breaking strains for comparison. The experiments were made on bars, cast in a half-upright mould, 0.99 inch (25 millimètres) square, laid on supports 31.5 inches (0·8 mètre) apart, the load being applied at the centre. They represent the mean of several hundred trials. 1. Charcoal iron direct from the blast furnace, from a mixture of magnetic and hydrated brown hæmatite peroxide, broke at. 2. Charcoal iron direct, from red and brown hæmatites and a little magnetic ore. 3. The same iron melted with charcoal in an old form of cupola 6. Charcoal iron from bog ores. 8. The same mixed with equal parts of English pig (Clarence) and melted like the last. 9. Scotch iron, Langloane No. 1, melted once with coke 1,401 [1874-75. N.S.] 2 E The Author deduces for these that it is sufficient, in giving orders, to specify the resistance to rupture, and that in many cases castings direct from the furnace may be actually better than those from a cupola. On the Size of Blast-furnace Charges. (Berg-und Hüttenmännische Zeitung, No. 7, 1874.) F. W. G. Ringel has shown by statistics, that small charges are better than large ones. Other conditions, including that of the production of iron, being equal, a small charge will be at least three hours longer in the furnace than a large one; it remains so much longer exposed to the furnace gases, and the charges follow each other past the tuyeres more than twice as quickly; by which means the temperature in the carbonising and reduction zones is reduced, while the charges reach the zone of smelting in a better state of preparation. Small charges have, moreover, the following advantages :— (a) The layers of ore being thinner allow a more intense action of the gases. (b) The more rapid alternation of the charges prevents too high a temperature in the zone of preparation, and lessens the danger of scaffolding. (c) The fuel is more equally distributed, and by presenting a larger surface to the incoming blast is more thoroughly utilised. (d) The better intermixture of the fuel with the batch allows the injurious ingredients of the former to be more quickly absorbed by such ingredients as are added to neutralise them. (e) The temperatures in the zones above the smelting zone being lower, and the iron therefore more completely reduced and more thoroughly carbonised, the injurious influence of the sulphur given off by the fuel is lessened. (f) The temperature above the boshes being lower, opposes the reduction of silicon. A great error is made in preferring large to small charges, especially for the fuel, as the temperature of the furnace is too much raised by so doing. Practice has shown that 17 to 20 cwt. is a good fuel charge, and far preferable to one of twice that size. As a basis for calculating the charges of fuel, it is often laid down that it should be sufficient to form a layer over the broadest section of the furnace, which should be thick enough (according to some metallurgists, 4 inches and more) to prevent the next ore charge from getting through; but this is an unsafe rule to act upon, as one component of the calculation, the density of the coke, is very variable. The Author, after restating the advantages of small charges, adds that lime also has the effect of more thoroughly carbonising the iron, and that by using small charges and lime, the sulphur from the fuel can be kept out of the pig. F. W. Inquiries into the Texture of Iron. By M. JANOYer. (Annales des Mines, No. 1, 1874, pp. 80–109.) The Author starts with the assumption, that the granular texture is the only arrangement of particles inherent in the metal, and adduces arguments and experiments to confirm the statement. The different classes of iron, such as granulated and fibrous, or iron into the texture of which both these elements enter, are the results of an imperfect or defective process of manufacture, not admitting of that perfect welding or amalgamation of particles, which alone constitutes the true condition of the metal. All descriptions of wrought iron may, therefore, be classed under the two general heads of granulated, or iron perfectly welded throughout the entire mass, and fibrous, in which these conditions do not obtain. Iron, manufactured with wood fuel, when it is very pure and homogeneous, has always a granulated texture. Without denying the influence of the hammer, to which some attribute the texture in question, it is due essentially to the high temperature, which promotes the repulsion of the scoria and the perfect welding of the entire mass. In order to prove that high temperature is indispensable to the production of granular iron, it is sufficient for the puddler, engaged in making blooms for iron of that description, to lower the temperature of the furnace, to roll the bloom about in the scoriæ, and to place it in that condition under the hammer. The iron produced will be altogether of a fibrous character. Nevertheless, in this operation the puddling process has suffered no alteration, and yet the granular iron has become fibrous. The lowering of the temperature, which favoured an imperfect welding, and the interposition of scoriæ are the only agents accountable for the change in the texture of the metal. As the transformation of granular into fibrous iron is produced by lowering the temperature, at a certain stage in the manufacture, the reverse of the operation changes fibrous into granular iron. It is merely necessary to raise the temperature sufficiently to effect the welding. Thus iron can be manufactured of either character, as may be required : and a bar may present at one extremity a texture of the one nature, and of the other at the opposite end. What has been already stated relates to the production of raw iron; it remains to be seen how the different operations of forging, rolling, and converting it into wrought iron, affect the texture. If several bars of granular iron be made into a pile, reheated to a white heat, and then passed through the rolls, the rolled bar will be also of a granular texture, provided always, when it leaves the last groove, the temperature is sufficiently high to maintain the welding complete. If, on the other hand, the bar leaves the rolls at a simple red heat, the texture will be fibrous, because the amalgamation of the particles is incomplete. The bar being drawn out in too cold a condition for the particles |