than one whose gases of combustion are wholly within its shell, because the first will have a greater external surface than the second, and because the temperature of the gases of combustion are on one side of the setting, while in the case of the second the temperature in contact with the felt or other covering is only that of the contained steam, or about half as much. In either case, however, the loss by external radiation of heat is comparatively trifling. During the experiments repeated tests were made of the steam, in order to ascertain if it carried over any water in suspension, and, if so, how much. The various methods gave reasonably accordant results, and showed that no more than 2 of the water in the boiler was so entrained. This quantity is not deducted from the vaporization given in the preceding table, because it was desired to have that vaporization comparable with the results of other similar experiments, and because there is great doubt of the accuracy of such determinations, which are really among the most delicate, complex and difficult in physical science, while the methods practically adopted are so very crude as not to be even approximately reliable. The proportion of water foamed over, or entrained in boilers, has been, in consequence, absurdly exaggerated. Finally, there may be remarked that the hygrometric moisture in coal will, in general, about offset the water that may be carried over from the boiler by foaming, so that the apparent vaporization, including the entrained or foamed over water, will be about the true vaporization, exclusive of the entrained water given by the coal, exclusive of its hygrometric water. Magnetic and Diamagnetic Elements. The magnetic elements are N, O, Fe, Ni, Co, Mn, Pt, Os, Pd, Ir, Rh, Cr, Ti, Ce, C, K and U. The diamagnetic are H, Na, Cu, Ag, Au, Hg, Zn, Cd, Pb, Sn, P, As, Sb, Bi, S., Se, Cl, Br, I, Ti, Si. Carnelly has observed that all the elements which are found in the even series, in Mendelejeff's classification, are magnetic, while those of the odd series are diamagnetic. This result furnishes new and interesting evidence of the shrewdness of the investigators who are endeavoring to trace all chemnd physical phenomena to the action of primitive laws of motion. Chem. Gesell. V -(THIRD SERIES, Vol. lxxx.) C. 3 PROGRESS OF THE DEPHOSPHORIZATION OF IRON. By F. GAUTIER. Read before the Society of Civil Engineers of Paris, January 9th, 1880. In my communication of July 4th, 1879,* I endeavored to show the state in which this very important question was at that time. The result of that study was that two processes were shown to have arrived at results of a certain value: the Thomas-Gilchrist process, by which Bessemer steel can be obtained from a pig iron as highly charged with phosphorus as the iron of Cleveland and the Moselle, and the Krupp or Narjes-Bender process, producing purified pig iron (fontes mazée), partially dephosphorized, and capable of being used in puddling or in the Martins-Siemens furnace. The Thomas-Gilchrist process, as it has been explained in numerous communications read before the Iron and Steel Institute, at its meeting in London, and by public experiments at Eston near Middleboro', in the works of Messrs. Bolckow and Vaughan, presented the following characteristics: Employment of a dolomitic casing, replacing the old-styled silicious lining of the converters. Addition, at the commencement of the operation, of a mixture of lime and oxide of iron, intended to absorb the silica and phosphoric acid produced. Continuation of the blowing beyond the ordinary stopping point in Bessemer operations. Determination of the practical end of the dephosphorisation by a series of samples and repeated trials. Under these conditions steel very low in carbon and containing no more than 0.15 of phosphorus has been produced from pig iron carrying 0.15; but in operating in this manner there are numerous causes of cooling. To avoid projections only a small charge was treated in a large converter, which augmented the loss by radiation. The addition of the oxide of iron was also a notable element in lowering the tem *See the JOURNAL for February, 1889, p. 83. perature of the operation; finally, the repeated tests at the close necessitated the loss of precious time. In order to obviate these inconveniences, they have treated iron carrying as much as 3 per cent. of silica, the calorific power of which is very high. But this only places the problem in a deceptive circle. It is necessary to increase the basic additions in order to absorb the great quantity of silica thus produced, for there cannot be left more than 20 per cent. of silica in the slag without endangering the stability of the phosphates formed. At the moment of the recarburation this very fluid slag, is moreover, traversed by the carbonic oxide resulting from the reaction of the carbon of the spiegel on the oxide of iron of the bath. There is then produced a partial rephosphorization of the purified metal; about a thousandth of phosphorus is added to that which had been left in the steel, and probably can only proceed from the action of the manganese of the spiegel, or of the carbonic oxide, on the phosphates of the slag. These reactions, moreover, whatever their nature, are produced with violence; the final addition should be made in several portions to avoid a sudden projection of the slag out of the converter. The metal obtained remained cold; the loss was as much as 18 to 20 per cent., that is to say, it required 1200 kilogrammes of metal, from the furnace, to obtain 1000 kilogrammes of steel ingots, whereas 1130 or 1140 suffice when the metal is pure and is submitted to ordinary treatment by the Bessemer process. The refuse from rolling attained 10 per cent. in the month of May last, and seldom was lower than 5 per cent. PROGRESS IN ENGLAND.-The first question which occupied the English engineers was to reduce the basic addition, the absorbing capacity of which was complicated with a very marked cooling action. They suppressed the oxide of iron, and forced in the limestone by an artifice analogous to that used at Neuberg in the treatment of white and mottled iron by the Bessemer process, where charcoal dust was introduced with the blast. They succeeded thus in eliminating half the phosphorus, at the same time retaining 2.5 per cent. of carbon, whereas in the old way of operating, when the carbon was reduced from 3.5 to 2.5 the diminution in the phosphorus was seldom more than 5 or 6 per The suppression of the oxide of iron diminished the cooling, and protected the lining, the slag being less corrosive; but at the same time it produced obstructions at the mouth of the converter, and, to cent. remedy this they modified the form of the latter. They supplanted the lower demi-sphere by a right cone, decidedly truncated; the gases. were then freely disengaged, and projections tumbled back into the bath without encumbering the throat. To prevent the final explosions which resulted, as we have said, from the too quick reaction of the oxide of iron of the bath on the carbon of the spiegel or of the ferro-manganese, they called to their aid silicon. In acting on the dissolved oxide the silicon gives origin to silica, which remains in the bath and has no tendency to be pro-jected outside. 3 Fe3O1+Si=9FeO+SiO3 If the silicon finds itself in presence of carbonic oxide, formed by the inevitable action which the carbon accompanying it exercises on the oxide of iron of the bath, . Fe3O1+C=3FeO+CO the silicon decomposes this gas, giving origin to silica and to a deposit of carbon, which is redissolved in the metal Si+3C0=3C+SiO3 It will be remembered that it is this reaction which explains the formation of steel without blowholes by means of silicon. This addition of silicon was made at Eston under the form of grey pig carrying 3 per cent. of silicon. In conformity with the preceding theory, the disengagement of carbonic oxide, which gave to the operation of overblowing a dangerous character, was entirely removed; the subsequent addition of the spiegel, destined to remove by the manganese the last traces of "red shortness," and to communicate to the metal the proper degree of carburation, was made with quiet and precision. In order to simplify, it has been proposed in France to employ silicide of iron and manganese, which ought to have several important advantages. First, it reduces the additions to one only; and, further, in these alloys, high in silicon, the carbon is in very small proportion,. not often passing 1.5 per cent., the chance of the production of carbonic oxide is reduced to a minimum; finally, the simultaneous presence of manganese and silicon ought, under an oxidizing action, to lead to the production of a very fusible silicate of iron and of manganese, and in consequence to give great homogeneity to the metal. The trial of silicate of iron and of manganese was not made in Eng-land, except under imperfect conditions; the alloy carried but 3 per cent. of silicon, and did not permit obtaining all the effect which one might expect. It will be remembered that the experiments at Eston had led to the noting of a singular phenomenon: the metal dephosphorized to 0·140 was suddenly changed to 0.235 of phosphorus after the addition of the spiegel. As the spiegel contained both carbon and manganese, it is to the presence of one or the other of these substances, and to their action. on the phosphate of the slag, that we must attribute this rephosphorization, which sometimes exceeded one-thousandth. The silicon, in combatting with the carbonic oxide, prevented the rephosphorization. This would seem to give reason to the opinion attributing this augmentation of phosphorus to the action of the carbonic acid on the slag. Another means to avoid the rephosphorization was also tried at Eston. The mass of metal and slag was cooled in a ladle furnished with a stopper and an orifice at the bottom; then the metal alone was • led into a second ladle, where, protected from the contact of the slag, the introduction of the spiegel was made. This arrangement succeeded well enough, but it was renounced, owing to the additional cooling which resulted from it, and which it was easy to foresee. Finally, in order to comply with certain criticisms expressed in regard to the accuracy of the diagrams representing the relative oxidation of the different elements which accompany the iron in the metal treated, Mr. Richards submitted to the meeting of the Iron and Steel Institute, at Liverpool, the results of samples taken every three minautes after the commencement of the operation. |