assumed a priori. But in reference to calorific effects, there is very nearly compensation; the excess of blast, 0.375, would take in 25 calories more into the furnace, whilst the excess of gases would carry off 14 to 15 calories above the numbers deduced by calculation.

There would therefore be in reality an excess of caloric carried in of 10 to 11 calories, which should be compensated by a less caloric of combustion; that is, the caloric of combustion would be reduced from 3642 to 3632. If we would go very accurately to work, we must begin the calculation again with this last number in order to determine the values of m and y. The number Q, in the preceding formulas, would be reduced by 10 units. We should have 4553 calories, instead of 4563. But it is quite evident that this correction would be insignificant. It would give the value of CO 01b .004 more (1·649 instead of 1·645), and the ratio m would scarcely be affected. In short, the figures above given represent sufficiently near, for practical purposes, the real state of things; and the result is, that the working of the Pouzin furnace is very satisfactory.

The ratio m has almost attained the limit 0-80, at which the dissociation of CO does not take place. The gases are rich in CO2 and yet cool. In these circumstances an increase of the height of the furnace could produce no sensible useful effect (§ 24).

The consumption could not be reduced below 0.970 of pure carbon, unless the blast were superheated, as at Consett. This advantageous working of the Pouzin furnace must, in my opinion, be attributed, in great measure, to the high tapering form of action. The gases are very uniformly distributed, and the reduction goes on very regularly. At one

time, the director of the works tried to enlarge the tunnelhead without modifying the system of charging. There resulted immediately a less regular working, and a greater consumption. Since that time the yield has been increased by forcing the weight of blast, but the consumption has not sunk below 0.970 of pure carbon.

We now see, that indirectly, even without determining by

CO2

experimenting the value of m = we can come at an ap

со

proximate estimate of the manner in which the carbon is consumed in blast furnaces, and thus appreciate the relative economy of their working. But this indirect determination

of the value of depends on an estimate more or less

CO2
CO

hypothetic of the caloric consumed.

It will always be preferable to determine the ratio by an analysis of the gases, which allows of our calculating with perfect exactness the caloric received.

This is a system of controlling the working of blast furnaces easily realized. It may be confidently recommended to iron-masters. They can thus ascertain in what respect their apparatus is faulty, and remedy the fault with confidence.

§ 28. Another example of French blast furnaces.—I have now shown by the example of Pouzin, that it is not always necessary to have great height of blast furnace to get satisfactory results. I might multiply examples. I shall take the blast furnaces of Denain of 1865, in which, with a volume of 2750 cubic feet, and a height of 42 feet only, they have arrived at a yield of 30 tons per day of forge iron, consuming only 1.150 to 1.200 tons of coke of 15 % ashes per ton of pig,

and this with argillaceous ores yielding only 32 % to 35 %. These favorable results are realized chiefly by a more rational system of charging, and also, as at Pouzin, by their having a more contracted region of fusion than most of the English blast furnaces.

Take now, as example, No. 1 furnace of Montluçon, St. Jacques' Works, of which in 1867 the principal elements

Temperature of escaping gases, 100°, arising from the hydrated state of the ores.

Yield in twenty-four hours, 24 tons of gray forge iron.

This is again an example of economical working, and if we apply to this the methods of calculation above given in detail for Pouzin, we should again find a high value for the ratio CO2

CO

An increase in height would certainly not diminish

the consumption, already very low.

The only possible economy is that which would result from the use of higher temperature of blast.

If now I am asked whence arises this great difference between the working of the French furnaces here named and those of the older types of Cleveland, I can only suggest the following reasons:

1. The section of the French furnaces is generally more tapering than that of English furnaces. Hence there is a better distribution of the gases; the reduction goes on in the upper regions with a less combustion of solid carbon. It has been already remarked that it is to this difference in section, in shape, that part of the economy of the furnace of Clarence 1866, compared to Ormesby 1867, and Clarence 1853, is to be attributed.

2. The manner of charging is more studied in France. Its influence on the distribution of the blast is well known. It is seen that thus they approach most nearly to the ideal working, that is, to the reduction of the ores without burning solid carbon.

3. Lastly, the region of fusion near the twyres, the hearth, appears to me to be generally too wide in England. In the smallest furnaces of Cleveland the diameter of the hearth is never less than 6 feet, and in most of those of 10,000 cubic feet and 17,000 cubic capacity, it reaches 7 feet 3 inches to 8 feet 3 inches, and even 9 feet, and yet the height is often less than this diameter. But the capacity of the region of fusion has a direct influence on the consumption. For fusing a refractory ore, it is not necessary alone to have a certain amount of caloric, the caloric must be well distributed, so that the local temperature shall be intense, which cannot be realized save in a contracted space.

We know by Tunner's experiments on the temperatures of the region of fusion, that near each twyre the zone of combustion is always very limited, and that if we desire to have a high mean temperature the zone of fusion must be contracted. Whatever be the dimensions of the body of a blast furnace this high temperature cannot be realized with a relatively low consumption, except by the use of narrow hearths high relatively to their diameter. This is a precaution too often neglected, not only in blast furnaces but also in lead and copper furnaces and in cupola furnaces.

ADDITIONAL NOTE.

During the impression of these pages, I have engaged in certain experiments on the caloric of the iron and slags running from the blast furnace. I shall publish these results soon. In the mean time I must say that the numbers adopted by Mr. Bell now appear to me somewhat high, and that' in the estimates above given there may be some correction to make.

All the general conclusions to which my study of the blast furnace has led me are in no way affected by this-only, the caloric carried off by radiation from the walls of the furnace should be about 100 calories more than above given.