drawing off the gas until a certain interval has passed, until the usual régime is re-established. As to the analysis of the gases thus collected nothing is more simple, as it is only required to determine the proporCO2 tions of It is not necessary to measure or weigh the CO gas which has to be examined. The way to operate is as follows: The gas comes out slowly by the cock g from the Mariotte's vase, by letting in mercury by the straight tube. The gases are dried in U tubes with chloride of calcium, or pumice stone with sulphuric acid. The carbonic acid is taken up by potash tubes. Burn the carbonic acid (CO) by the oxide of copper, retain the water formed by the small quantity of hydrogen present, and then determine the carbonic acid produced from the carbonic oxide by means of a second system of potash tubes. This analysis will only be inaccurate in cases where the gases contain appreciable quantities of carburetted hydrogen, which never takes place when the furnace is working with coke. A last precaution is perhaps necessary in taking the specimens of gas. There are furnaces which smoke a good deal, or in which the gases carry off quantities of fine dust. In such cases the slit in the copper tube might get obstructed more or less. This happened to M. Scheurer-Kestner in his experiments when there was much smoke. This difficulty may be overcome as this able chemist did it. A little rake, composed of a short slip of copper put into the copper tube, could be drawn backwards and forwards by a rod passing on the outside, and thus the slit kept clear. Having determined the ratio CO2 the mean temperature of the gases should be determined. When the mine is not hydrated ores the temperature of the gases may rise to 400° or even 600° C., which renders the mercurial thermometer inapplicable, and there the thermo-electric pyrometer, or the resistance thermometer of Siemens, or the pyrometer of Lamy, based on the variable tensions of CO2 derived from the decomposition of the carbonates of lime and magnesia.* § 10. Determination of the caloric consumed in Blast Furnaces. -Suppose we have determined the two elements we have and the temperature of the gases been considering CO2 = m, as they leave the furnace. Let us now see how we can employ these elements to determine the working of the furnace. The question is to compare exhaustively the caloric received and the caloric consumed. It has been already shown how the caloric generated in the furnace by combus. tion may be calculated, when we know the total weight of CO2 and CO in the escaping gases. Further on it will be shown how the caloric produced near the twyres-in the zone of fusion, and that generated in the zone of reduction, may be separately estimated. In both cases the caloric carried in by the hot blast must be added to the caloric produced in the furnace, in which there is no difficulty if we know the temperature of the blast. It is the sum of these two quantities, caloric produced and caloric thrown in, which makes the total quantity received. For the present, let us show how the caloric consumed may be determined. It is composed of four parts: 1. The caloric absorbed by the reduction of the ores and * Comptes Rendus, tome lxix. p. 347. by the fusion of the pig-iron. This is a constant element for a given quality of pig, and one which varies very little for different qualities of pig. 2. The caloric absorbed by the fusion of the slags, the decomposition of the limestone, the evaporation of the water in the coke and in the charges of mineral, and lastly, the decomposition of the water in the air. This second part is essentially variable, not only on account of the different quantities of lime used, and of water in the ore and fluxes, but also from the very varying composition of the slags, which on this account require very different quantities of caloric for their fusion. 3. The sensible heat carried off by the gases. This is also a variable element, but always easily calculated when we know the composition and temperature of the gases. 4. The caloric lost by radiation from the walls of the furnaces by contact, or by artificial means of cooling. Some experiments have been made to determine the losses in this way, but in general they can only be appreciated by deducting from the caloric received the sum of the quantities due to the first three causes. Let us now endeavor to estimate the value of these different elements. § 11. Caloric absorbed by the reduction of the ores and the fusion of the pig-iron.-We may here admit as in § 7 that the pig-iron contains 0·94 of iron, and 0·03 of carbon, and 0·03 of silicium, phosphorus, sulphur, and earthy metals, etc. The caloric absorbed by the reduction is equal to that given off by 0.94 iron burned to the state of peroxide, plus the caloric given off by the oxidation of 0.03 silicium, phosphorus, etc. The caloric produced by the oxidation of iron has been determined by several experimenters. By burning iron in oxygen Dulong found that the caloric produced for each lb. of oxygen is 4327 calories.* Supposing there was formed magnetic iron or Fe3O1, this would amount per lb. of iron to or, applying the law of Welther, for the passage of Fe3O1 to Fe203 per lb. of iron giving peroxide, 1854 and this gives for the peroxide Again, Favre and Silbermann* found for the transformation of a lb. of iron into protoxide by the wet way was which gives for the peroxide 66 1582 66 1780 If we adopt the mean of these three determinations we have 1887 calories. This is the figure we shall adopt for the caloric absorbed in the reduction of the peroxide of iron for each lb. of metallic iron yielded. But it must be borne in mind that this number is only a more or less accurate approximation. Not only the experiments quoted do not agree with each other, but we do not know that the law of Welther is exact for the transformation of Fe3O4 and FeO into Fe203; and besides, the quantities of caloric given off and absorbed vary with the density and molecular condition * Annales de physique et de chimie, 3e série, tom. viii. † Ibid., tom. xxxvii. p. 435. of the products. Lastly, if we accept the experiments of Despretz, we get a much higher number. According to this physicist the combustion of iron produces 5325 per lb. of oxide, which corresponds to 2019 per lb. of iron passing to the state of Fe3O4, and 2271 when the metal is transformed into the peroxide.* As to the caloric arising from the oxidation of the 0·03 of silicium, phosphorus, earthy metals, etc., it is still more difficult to get accurate data. We know from the experiments of MM. Troost and Hautefeuille, that the caloric produced by the oxidation of silicium is 7830 calories, and according to Mr. Andrews the oxidation of phosphorus yields 5767 calories. We have no knowledge of the caloric corresponding to the earthy metals. But as silicium is generally the predominant element, we shall not probably be far wrong in adopting 7000 calories for the caloric of each lb. of these elements. It is also probable that the combination of silicium with iron produces caloric, and therefore we may with more confidence reduce the number 7830 calories formed for pure silicium. We may therefore admit that the 77 3 * Let us here call to mind that the 1887 calories per lb. of iron corresponds to 1887 X 4403 cals. per lb. of oxygen. This is the value of C in § 4. It follows hence that the reduction of peroxide of iron by CO is accompanied by slight absorption of caloric, 4403 — 4205 4205 198 calories. Whilst if with Mr. Bell we adopt 1780 calories found by Andrews, we have per lb. of oxygen 1780 × 4153 calories, from which would follow that 77 3 there is a slight disengagement of caloric at the moment of reduction by CO, amounting to 52 calories. |