The attempts quoted having proved unsuccessful in lessening the the loss of heat, owing to the temperature of the escaping gases failing to fall below that which the enlargement to 12,000 cubic feet had effected, it occurred to me that the chemical action which took place in the reducing zone might be of a heat-producing character; and that to raise the height of the furnace beyond a certain point merely elevated this zone of heat evolution. If so, its distance from the final exit of the gases would remain unaltered, and the power of the incoming minerals to absorb heat from such gases would not be increased.

From the large amount of heat, evolved by the generation of carbonic acid in the reducing zone, has, however, to be deducted that required for dissociating, or tearing away, the oxygen from the oxide of iron. Now the experiments for determining the heat produced by the combinations of different substances are of a very delicate nature, and it was difficult to say, with a narrow margin to work on, how far any calculation respecting the estimated balance between heat generated by burning carbonic oxide to the state of carbonic acid, and heat absorbed by reducing peroxide of iron, might be affected by experimental error. The extent of the margin referred to was known to be small, because the investigations of the best authorities went to show that the heat absorbed, in splitting up, or reducing, peroxide of iron, was a little less, but only a little less, than that generated by the combination of oxygen with the carbonic oxide used to effect the dissociation.

Under such circumstances nothing short of experiment upon the furnace itself could be regarded as a satisfactory solution of the problem; and this I conducted in the following way. It was ascertained that a mixture of flints and blast furnace slag had as nearly as possible the same specific heat as calcined Cleveland ironstone, and hence that there would be little or no disappearance of sensible heat from any difference between the materials in this respect. The rapidity with which the mixture absorbed heat also corresponded pretty nearly with that of the ironstone. The temperature of the upper portion of the furnace was of course far below that required to fuse such materials as

'That is, to heat up the same weights to the same temperature, the same quantity of heat was absorbed in each case. This consideration is of importance, for the difference in some kinds of material is very great.

those just mentioned; so that there would be an absence, so far as they were concerned, of all chemical action likely to affect the temperature of the gases, and these would therefore be simply cooled by the heat-absorbing properties of the flints and slag.

room.

The ironstone was then withdrawn from the charges of a furnace, and an equal weight of the mixed slag and flints was put on in its The mean temperature of the gases gradually fell, as the unreduced oxide remaining in the furnace lost its oxygen; and as soon as the ore was all deoxidized, the mean temperature remained stationary. After continuing it at this point for a short time, the use of the slag and flints was discontinued, and that of ironstone resumed; when the temperature of the gases gradually rose until it reached its original level.

This experiment seems conclusive as to the reducing zone being one of a heat producing character, and that in consequence there was, as I supposed, a point beyond which, for the reasons already given, it was useless to add to the height of the furnaces. This opinion received confirmation from actual experience in the case of the furnaces of 90 and 103 feet in height; for it did not appear that the gases left them at a perceptibly lower mean temperature than those from furnaces of 80 feet in height.

The other alternative alluded to, as a means of obtaining from the fuel consumed in the blast furnace an increase in the quantity of useful heat, consisted in a greater degree of saturation of the gases by oxygen. Any addition from this source, however, must be obtained by the action of the gases on the ore; because it has already been shewn that all which the atmospheric oxygen could effect was the conversion of the carbon at the tuyeres into carbonic oxide.

With a view to supply this increased quantity of oxygen to the burthen of a furnace in good working order, let us suppose an additional quantity of ore to be added. One of two things would happen: either the carbonic oxide, by its combination with the oxygen in this ore to form carbonic acid, would at first raise the temperature of the contents of the furnace in the reducing zone, or else the carbonic oxide, having already taken up as much oxygen as it was capable of holding, would be unable to act further on a fresh supply of ore.

Practically the effect in each case might be the same. In the first the rise in temperature would intensify the action of the carbonic acid

in the gases, as explained above in describing the four reactions, and would thus interfere with the process of reduction. The result of this change would be that a portion of the ore would pass down through the reducing zone with a part of its oxygen unexpelled: this part would then have to be driven off either by solid carbon, or if by carbonic oxide, yet under such conditions that the resulting carbonic acid by dissolving carbon, would, immediately revert to the state of carbonic oxide. If, on the other hand, no action took place in the uppermost zone between the carbonic oxide and the unreduced ore, owing to the inability of the former to separate a further quantity of oxygen from the oxide of iron, this separation would have to be performed by solid carbon under precisely the same unfavourable conditions as those just indicated.

Allusion has been made to the power of carbonic oxide to take up oxygen from ore. A few words of additional explanation on this important subject appears necessary for a perfect understanding of the nature of this reaction, for virtually it is the key to the discovery of the extent to which economy in fuel can be carried in smelting iron.

It must be conceded in the first place, that it is quite possible to convert carbonic oxide wholly into carbonic acid by means of iron ore; but this can only be done by using a very large excess of the latter. This is equivalent to our supposing that in reducing peroxide of iron, the oxygen which is separated at first is less firmly held by the iron than those portions which are withdrawn at a later period of the process-a supposition which experiment and observation entirely confirm. Those who may wish to study in greater detail the chemical reactions of the furnace are referred to the separate Section devoted to this particular branch of the subject: here it may be stated that so tenaciously is the last portion of oxygen retained by the iron, that it is believed that the trace of that metal always found in the cinder is due to the impossibility of effecting an absolutely perfect separation of all the oxygen from its combination with it.

In the meantime it must be granted that, having regard to the temperature which is set up in the reducing zone of the furnace, and is inseparable from it, a position of equilibrium is necessarily established, when the carbon as carbonic acid bears a certain proportion to the carbon as carbonic oxide. The maintenance of the conditions most

favourable for saturating the carbonic oxide with oxygen, as far as the nature of the action permits, in other words for obtaining the highest possible percentage of carbonic acid in the gases, is perhaps a difficult matter. It would indeed be equally difficult to determine by actual observation what is the maximum ratio which the carbon as carbonic acid can bear to the carbon as carbonic oxide; because the determination of the mean amount of each in the gases is not without a certain degree of uncertainty, owing to the circumstance that these amounts are constantly varying, according to the time, after the period of charging fresh ore, at which the samples are taken. Under such difficulties as those referred to, the smelter finds it best to allow himself a little margin in the fuel he supplies to his furnace; or, what amounts to the same thing, if his "burthen" is adjusted with the view of producing No. 3 iron, he must be prepared for having sometimes a higher and sometimes a lower quality than the one he aims at obtaining.

Were it required to name a limit beyond which it would be impracticable, if not impossible, to carry the mean saturation with oxygen of the gases of a blast furnace, the answer probably would be-When the two compounds of carbon and oxygen contain an equal quantity of oxygen. This happens when, out of three equivalents of carbonic oxide, one has been converted into carbonic acid, leaving the remaining two unchanged (CO, + 2CO). We shall however see later that this is a degree of oxidation of the gases rarely, it may be doubted indeed if ever, attained in actual practice-at all events through any considerable length of time.

SECTION VI.

ON THE USE AND THEORY OF THE HOT BLAST.

IN the Section immediately preceding the present, it has been shewn that the heat produced in a blast furnace is that caused, in the first place by burning all the carbon to carbonic oxide, and in the second place by converting something like one-third of such carbonic oxide into carbonic acid. Unavoidable losses of the heat so evolved, are occasioned by some minor sources of escape and by the temperature at which the gases are necessarily permitted to leave the throat. The balance between the heat evolved and the heat so lost is what is actually expended in the operation itself.

If the heat absorbed by the general requirements of a furnace were such as could be supplied by the combustion of a quantity of fuel, which would permit the gases to escape fully oxidized and cooled, so far as is shewn in the last Section to be possible, then all the conditions necessary for a complete utilization of the heating power of the fuel, consistent with the nature of the process, would have been secured. A different state of things however might obtain. A proportion of say 20 units of carbon per 20 units of iron made, might suffice in the furnace, so far as reduction is concerned, supposing a little above 6 units should escape as carbonic acid, and a little under 14 units as carbonic oxide; but the heat evolved by the combustion of the 20 units, might be insufficient to melt the iron and slag, and provide for other minor causes of absorption. If we assume that 24 units of carbon, burnt to carbonie oxide, would suffice for the general requirements of the operation, it is clear we might add the heat represented by the 4 units of carbon by a direct introduction of heat at the hearth, without in any way interfering with the proper relation which the carbonic acid and carbonic oxide must bear to each other in the zone of reduction.