only fusion has to be effected, but reduction of oxides must take place, this reduction is hastened by complete contact of the fuel with the ores; and hence it has been recommended that the fuel and ore should be mixed before being charged into the furnace. But it is well known, that it is only necessary to begin reduction on the surface of a body, and that the process penetrates to the interior without any immediate contact with the reducing agent being necessary. M. Le Play has, it is true, recently and repeatedly called attention to the reducing power of carbonic oxide. He has shown that oxides of iron, in circumstances where they could not come in contact with solid carbon, were reduced at a certain temperature by carbonic oxide made in a tube in which carbon was at one end, oxide of iron at the other, and the air was sealed up with the two-that the carbonic oxide first formed was transformed to carbonic acid by the oxygen of the ore, and reconverted to carbonic oxide by the carbon, and so on in succession, till either the carbon was burned or the ore was completely reduced. "But," says Karsten, "the oxides of iron are unquestionably not reduced in this way in the blast furnace, because the CO formed by the incandescent carbon in contact with CO2 rushes too rapidly past to the top of the furnace, and because the oxides of iron are everywhere in contact with incandescent carbon, by which the reduction can be begun without its being necessary to decompose a gas, which would, besides, have to take place under circumstances which would rather favor the generation than the decomposition of the gas (CO2).” แ Assuming that the process of reduction above described was correctly conceived, Karsten proceeds to take into consideration the products of combustion, and concludes, “that the formation of carbonic oxide takes place in the upper part of the furnace by the reduction of the ores by solid carbon." "But," says he, "the reduction when only CO is formed takes twice as much fuel as when CO2 is formed; and if there were any means of preventing the formation of CO here, this would be at the same time the means of reducing the ores with the least expenditure of fuel.” We now know that the air blown into the furnace at the twyres is almost instantaneously transformed into carbonic oxide, and this gas, in its passage up the body of the furnace, acts more or less directly in reducing ores; that is, with or without the aid of the solid carbon (see ante, p. 25). We also know that, in the presence of CO2, iron is practically reoxidized, and that in the presence of equal volumes of CO and CO2, peroxide and metallic iron are both brought to the state of oxide. The doctrine of Lampadius was much nearer the truth than that of Karsten, and is, in fact, the explanation still given of what goes on in the blast furnace (see Percy, vol. ii. p. 444), but great strides had yet to be made, and that very year, 1839, before the printing of the number of Karsten's Archiv in which his notions above given was finished, Bunsen's first experiments on the composition of the gases at different depths of the blast furnace, at Veckerhagen, were published as an Appendix to Karsten's paper. In his report of the experiments Bunsen writes: "They prove that too much weight has hitherto been laid on the desoxidation of CO2 to CO at the cost of incandescent carbon. Some of my experiments demonstrate that, in the combustion of charcoal, the first immediate product is CO, provided there be no surplus supply of oxygen to burn the CO to CO2, and thus one-half the calorific effect of the fuel is lost." Thus began to be formed a theory of the blast furnace. But such as it was, the theory was still only a qualitative analysis, so to speak. Bunsen first began the quantitative analysis, and first gave a calculation of the quantity of caloric used in the furnace, and of the quantity carried off in the gases, and pointed out for what purposes this lost heat might be made available. For a particular case of analysis of the gases collected 5 feet under the tunnel-head, 49.55 per cent., or onehalf of the fuel was proved to go off as CO, and was lost, and that besides the sensible heat of the gases themselves being 993° C., 25 % more of the fuel was lost in this way; so that 75% altogether was lost.* The researches of Ebelmen on the composition of the blast furnace gases were begun at the same time as Bunsen's researches, but were published a year or two later.† Ebelmen made his calculations of the quantities of caloric used, and of the temperature in the furnace, by adopting Dulong's calorimetric results, given in the Comptes Rendus, 1838, p. 874, and arrived at the conclusion that 64.8 % of the fuel went off in the gases and was lost. Ebelmen pointed out that the gases escaping would suffice for heating the blast, supplying the steam for blast engine, and for the desiccation of the wood used in the furnace itself. * Karsten's Archiv, vol. xii. p. 547; 1839. † Annales des Mines, 1841-43, and collected Memoirs, published in Paris, 1861. For the combustion of C to CO2 Dulong found 7167 cals., but it is 8080 CO to CO2 66 66 2634 cals., 66 "2473. Such was the progress made towards a rational theory of the blast furnace-chemical and physical-and of its application to determining the calorific efficiency of the fuel consumed, in 1842; and very little progress was made for several years afterwards either in diffusing or extending the notions then attained to. And now it is interesting to show how, in this department as in many others, art is the mother of science. Long before an analysis of the gases of blast furnaces was made, or could have been made, practical men perceived the uses to which the gases could be turned. For this object a quotation from Dr. Percy's great work, vol. ii. p. 663, will suffice: "In June, 1814, Berthier published an interesting and important paper* on the successful application, in France, of the waste gas to various purposes, such as the conversion of iron into steel by cementation, and the burning of lime and bricks. The credit of this application is due to M. Aubertot, who was a proprietor of ironworks in the department of Cher. He obtained a patent for it in France in 1811. Berthier visited four works, either belonging to, or managed by, M. Aubertot, so that his description was founded on personal observation; and it is very just to his memory to state, that he seems clearly to have foreseen the value of the application in question. The calorific effect of the waste gases was rightly attributed by Berthier, partly to the sensible heat, and partly to the heat developed by combustion in contact with atmospheric air." Mr. Moses Teague took a patent on this matter in 1832. Mr. James Palmer Budd obtained a patent for the applica * Journal des Mines, 35, p. 375; June, 1814. tion of the "heat, flames, and gases of the blast furnace to the heating of hot-blast stoves" in 1840. This application was carried into practice at Ystelyfera, and so satisfied was Mr. Budd with the success that he even ventured to draw the following conclusion: "It would appear to be more profitable to employ a blast furnace, if as a gas generator only, even if you smelted nothing in it, and carried off its heated vapors (sic) by flues to your boilers and stoves, than to employ a separate fire to each boiler and each stove. These considerations irresistibly suggest to me a great revolution in metallurgical practice, a new arrangement, in fact, of furnaces and works by which considerably above £1,000,000 a year might be saved in the iron trade alone (1848)." This was excusable in the first enthusiasm, but it is well known that it is only when a low class of coal has to be used that generator or "producer" gases are not wasteful, except where intensity of heat is required. Let us revert now to the progress made since 1842 in the chemical and physical theory of the blast furnace. Bunsen's calculations were founded on the truth of "Welter's law," by which the caloric evolved in the furnace would be to the potential caloric of the gases escaping as the weight of the oxygen burned in the furnace is to that of the oxygen required for complete combustion of the gases escaping. Welter's law is not universally true, and it is certainly not true for this application. Ebelmen could only refer to the calorimetric determinations of Dulong for his calculation of the development of caloric in the combustion of carbon and carbonic oxide.* Since then * It is interesting to mention here that in the Handbuch der Chemie of L. Gmelin, published in 1843, he remarks-"Dulong's determination of the |