and 27,000 cubic feet capacity, all charged with the same charges, and of the uniform height of 95 feet (figs. 7 and 9), asserts, on the other hand, that he cannot find any advantage in economy or otherwise in the large furnaces. His long experience has convinced him that beyond the limit of 11,000 to 12,000 cubic feet capacity, the large furnaces of Cleveland really offer no advantage in economy of fuel consumed.

It is also to be observed that the blast furnaces are 15 feet higher than the Clarence furnaces, and yet they burn no less coke, and the temperature of the escaping gases is not lower. And now, a last remark on the large furnaces of Clarence Works and of Eston.

In § 2 the three types of Clarence and the two huge furnaces of Eston (figs. 7 and 9) of 15,000 and 27,000 cubic feet capacity have been alluded to. These furnaces consume the same weight of coke per ton of pig yielded, 1·125 ton, but this equality is in reality obtained by the much slower working of the large furnaces, which require 420 to 490 cubic feet of capacity per ton of yield in place of 280 to 320 cubic feet. If therefore the large furnaces were made to yield proportionally as much as the smaller furnaces, the consumption of fuel would inevitably be greater, and therefore the working of the monster furnaces is in truth less advantageous. We cannot escape from this alternative, less production per cubic foot of capacity, or greater consumption of fuel. This conclusion. comes out more strongly from comparison of the two furnaces at Ferryhill already mentioned in § 2.

30,500 66

The one measures 82 feet high, and 16,000 capacity.
The other 66 104 66
The first takes 310 c. ft. capacity, the second 420 c. ft.

But, notwithstanding the great difference in height, the

escaping gases have within 6° the same temperature in the 104 feet as in the 82 (191° and 197°), and the consumption of fuel is as nearly as possible the same in the two-1000 tons in the high furnace, and 1025 in the other.*

What is the cause to which we are to attribute this apparent anomaly in blast furnaces in which the gases do not vary in temperature after a certain height, which appears to be from 75 to 80 feet? It is an important question.

§ 24. Beyond a certain height the temperature of the escaping gases does not diminish by reason of the dissociation of the oxide of carbon.-We know that the gaseous current of blast furnaces, which at first are composed only of CO and N, become mixed in increasing quantities as they ascend with CO2, and also with watery vapor when the ores and combustibles are moist, or chemically hydrated. The reducing action of the CO is thus partially neutralized in the upper regions of blast furnaces, and we know that certain mixtures of CO2, CO, and HO have no action on the oxide of iron. Thus M. Debray has proved that the peroxide is simply transformed at intense red heat to protoxide by a mixture of equal volumes of CO and CO2, and that, on the other hand, this mixture converts metallic iron to this same degree of oxidation. Mr. Bell has verified this double reaction; but this mixture of equal volumes corresponds in weight to

* In one of the last numbers of the Journal of the Iron and Steel Institute, that of November 1871, Mr. Bell, recurring to the same subject, thus sums up his conclusions as to the working of the Cleveland furnaces: “After a certain capacity, 15,000 c. ft., for example, the escaping gases are as cool and as rich in CO2 as those of furnaces of double this capacity” (p. 391). From discussions of Mr. Samuelson's account of two monster furnaces, read before the Institute of Civil Engineers.

Whenever, therefore, the CO2 reaches this proportion, reduction cannot go further than the protoxide, and if the current of gas contains steam as well, reduction would reach its limits even before this proportion, m = 1.581, was reached.

Besides, there is a certain time required—a prolonged action of the two gases-to reduce ores in small pieces so long as their temperature is not very high.

But the ores remain only a few hours in the upper regions of the furnaces; consequently they will be little modified by the gases as soon as CO2 abounds. Thus Mr. Bell has proved that in the Cleveland furnaces, for which the value of m varies between 0-50 and 0.70, reduction almost ceases towards the top of the furnace. At the temperature of melting zinc (about 417°) when the gases contain 31 volumes of CO2 to 100 CO, which corresponds to m = 0.49, it requires

5 hours to carry off 1.9 % of the oxygen contained, and

And when the gases are composed of 50 volumes of CO2— 100 of CO-which gives m = 0.79, it takes 5 hours to carry 0.9% of the oxygen on the ores.

We thus see that even when the gases are dry and hot, as in the Cleveland district, reduction at the level of the tunnel is insignificant, and it must therefore be still less when the ores are hydrates or carbonates. And indeed this has been recognized by MM. Ebelmen and Tunner long since. In the experiments at Eisenerz, M. Tunner ascertained that reduc

tion commences when the ratio of

CO2
CO

=

2, but that metallic

* Section 33 of Mr. Bell's work.

iron does not appear until the ratio descends to 0.70-about 23 feet below the tunnel-head.*

It is however certain that if there were no other chemical reaction in the upper regions of the furnace than the partial reduction of the ores by CO-reduction effected as we know (§ 11) almost without change of temperature—there ought to be some real advantage in increased height of furnace above the point at which the gases still retain a temperature in escaping of 300° to 400°. By such increased height we should inevitably cool the gases more perfectly, and thus utilize more perfectly the caloric produced. But there is another reaction which always takes place in the upper regions of the furnace: this is the dissociation of the carbonic oxide, and it is this transformation of 2 CO into CO2 + C which, after a certain limit, prevents the further lowering of the temperature of the gases by any increase of height of the furnace.

Mr. Bell was the first to point out this singular reaction of the ores on the gases of blast furnaces, and since then I have with much care investigated the principal circumstances attending it. The results of this investigation were communicated to the Academy of Sciences in July, 1871, and the memoir has been published in the Recueil des savants étrangers, and in the May number of the Annales de physique et de chimie for 1872. I shall here transcribe the conclusions:

1. If we pass CO over iron ore, heated to 300° to 400°, the oxide of iron is gradually reduced from the surface inwards of each small piece. But from the moment any small portion of the extreme crust of these morsels is brought to the state of metallic iron, the ore cracks in all directions and gets

* Memoir published at Leoben in 1859.

covered with a dust of carbon. This reaction takes place by whatever method the CO is prepared.

2. As the reduction approaches completion the carbon deposit becomes less abundant, and would altogether cease from the moment the oxide of iron (Fe2O3) is completely reduced, if this absolute reduction could be realized in the conditions under which our experiments were made. But, at all events, this would require a very long time.

3. If we pass CO over metallic iron at the temperature of 300° to 400° C., the iron, in like manner, becomes covered with a dust of carbon from the moment that the reducing action of the CO is partially tempered down, whether by the presence of a small proportion of CO2, or by any source of oxygen which could transform any part of CO into CO2.

4. On the other hand, CO, pure and dry, gives off so much the less carbon to the metallic iron, as the iron is itself free of all admixture of oxygen, so that there would be scarcely any reduction at 300° or 400°, if the experiment could be made with iron absolutely free of any admixture of oxide.

5. The carbon dust which deposits, either on the ores at the time of their reduction, or on the metallic iron when the CO acts together with a small proportion of CO2, is a sort of ferruginous carbon-a composition of iron and carbon with 5 % to 7 % of metallic iron as a maximum. This carbon differs from that of steel, or the hexagonal graphite of gray pig-iron.

The ferruginous carbon contains also a small proportion of oxidized iron, chiefly magnetic, the part played by which seems essential in the reaction which determines the deposit of carbon.

6. Carbonic acid acts as an oxidizing agent on iron; but