Such cinders as run from the furnaces, when producing white pig at Nancy, contain from 2 to 3 per cent. of protoxide of iron. The metal is very suitable for puddling purposes, and has the following composition:

In instances of extraordinary derangement, the cinder may become so overloaded with iron oxide as to reduce the iron to the condition of a pasty mass, due partly to the reduced temperature of the hearth, and partly to the infusible nature of the metal itself. Under such circumstances it becomes impossible to work a furnace satisfactorily. The slag resembles mill cinder in appearance, and the iron is so thick that it sets on the hearth, giving rise ultimately, if continued, to what is technically known as "gobbing." There is besides so imperfect a separation of the two, that a good deal of the iron is carried away mechanically along the slag runner. In such cases it is more than probable that during the mere passage of the iron through the cinder, it may lose a portion of the carbon with which it had become associated in an upper region of the furnace.

Cooling attends the "overburdening" of the furnace. The carbonic oxide has then imposed upon it more duty than it can perform, unreduced ore descends into the hearth, and the result is a black cinder. The same effect may be brought about by anything which lowers the temperature of the zone of fusion, even when, at the time this occurs, reduction is being suitably carried on in its proper place. A very common cause of this temporary disturbance is the admission of water by a faulty tuyere.

The decomposition of water is attended with a great absorption of heat-no less than 34,000 calories per unit of its combined hydrogen, besides that which is required for its conversion into vapour. If we assume each unit of coke, as it is burnt with hot air, to give 3,300 calories, then, after giving the water credit for the heat it generates by

oxidation of the fuel, it may be taken that the admission of 15 lbs., ie. 1 gallons, of water per minute, into a furnace making about 400 tons per week, requires an addition of 3.2 cwts. of coke per ton of iron.

So great indeed is the absorption of heat from the injection of water at the tuyeres, that differences in the extent of moisture in the atmosphere may occasionally be enough to account for variations in the work performed. It has frequently been remarked in winter, at a time of extreme cold, that blast furnaces work more steadily and with less fuel than in summer. Indeed this fact was constantly quoted as a reason against the use of the hot blast in its early days. No doubt the change was due to the comparative absence of moisture in the air; and thus a virtue was attributed to the mere coldness of the air which did not belong to it.

It is I believe capable of demonstration that the reduction of a certain amount of oxide of iron within a very short distance of the tuyeres, and the escape of a trace of the same into the slag, is probably unavoidable; but for many reasons the smaller this amount is the better. If this undecomposed oxide be unduly increased, it is beyond the power of the fuel at the hearth to effect a proper amount of deoxidation; and the unreduced metal is carried off in undue quantity by the slag, giving to it the well known dark colour.

The cinder from blast furnaces, certainly from those in the Cleveland district, possesses a property which, were it possible to reduce all the oxide of iron it contains, ought, one would imagine, to secure that result. This property consists in the occlusion, or power of absorption of a considerable volume of the gases found in the furnace itself, nearly 50 per cent. of these being gas of a highly deoxidizing nature.

For the first time, so far as I know, these have lately been made the subject of analysis at the Clarence works. Two of these were as follows:

There is no doubt that the volume of gas so absorbed is very considerable, because, being emitted as the cinder cools, it often gives rise to the bursting of the square masses of slag into which it is run. The outer portion of the mass being first solidified, forms a shell which is frequently fractured by the pent up gas. This is often attended with considerable explosive force, not altogether free from danger to those who may be in the immediate neighbourhood. Besides the true gases, which escape from the flowing stream of cinder, a portion of its solid constituents are occasionally emitted in the form of a white smoke; the actual weight of which however is very small. This always happens in smelting iron at a high temperature from Cleveland stone. By condensing and collecting this vapour, probably oxidized after leaving the slag, it was found to consist of:

Furnace Fume.-A certain amount of fine dust is propelled mechanically from the blast furnace, by the mere force of the current of gas escaping at the tunnel head. This consists of powder of ironstone, and minute particles of coke, the limestone being rarely in such a condition as to size to permit of its being so ejected. Besides this, there issues from most furnaces a vast volume of a white vapourlike matter or fume, of a varying composition, but differing always widely from any of the materials charged in at the top. A considerable portion of this condenses very speedily; but a part probably from its very minute state of division, passes through the boiler and stove fire places, and floats away for a very great distance from the chimneys of our iron works. Any questions connected with the generation and emission of this sublimed matter will be more con

veniently considered in the following Section, which it is intended to devote more exclusively to the chemical changes, as they occur in the furnace.

As a matter of practical interest this sublimate, if such it be, is of little importance: except that its presence is the source of considerable inconvenience and expense to the furnace manager. It gradually forms. a coating on the boilers and hot air apparatus which interferes greatly with the transmission of the heat to their contents. The removal of this, and the emptying of the flues of a large quantity of hot light dust, occupies a considerable time, and is a very disagreeable occupation for the men engaged in the operation.

SECTION IX.

ON THE CHEMICAL CHANGES AS THEY TAKE PLACE IN THE BLAST FURNACE.

IN the four preceding Sections the results obtained by the use of the blast furnace have been described. In doing this some general explanation of the chemistry of the process was unavoidable; but in order to render this more complete, I have deemed it desirable to devote a separate Section to a more minute examination of the changes experienced by the materials, during their descent from the top to the lower part of the structure. This mode of dealing with the subject involves some repetition, but, even with this inconvenience, the separation of the chief functions of the furnace from others of a more scientific character, but occupying perhaps a subordinate position as a question of technical interest, has its advantages.

The sequence of chemical actions which takes place in the blast furnace is complicated, and as regards any particular locality is liable to considerable changes in its nature. The immediate action on the materials, viz. ore, coke and limestone, when they enter a close topped furnace, is a mere elevation of temperature. The first to manifest any chemical change is the ore, which quickly commences to lose a portion of its oxygen; then follows the coke, which is attacked to a limited extent by carbonic acid; and lastly the limestone, which does not part with its carbonic acid until it reaches a zone where the temperature is more intense than that required for the changes experienced by the oxide of iron and the coke.

The order in which chemical action is established in the materials, as just set forth, must be understood as relating only to the commencement of the process. A large block of ironstone may pass through the hotter zones, of carbon solution and of limestone decomposition, before