Nor is this all; for it must not

over that of pig and iron sponge. be forgotten that the metal made by the direct process constitutes only 47 per cent. of the whole charge, and that to this the whole of the extra cost is to be debited; so that, in reality, the steel obtained from the sponge has been produced at a cost of above 30s. per ton over that which the other ingredients are capable of affording, by the alternative mode of treatment.

A further loss falls on the steel made from the iron sponge, due to the smaller output of the open hearth in which it is melted, viz. 760 tons, as against probably 1,000 tons in the ore process. It is, therefore, not improbable that the difference of 30s. given above, may be increased to 33s.; and this would, it is submitted, amply justify the opinion already expressed respecting the future of Mr. Blair's process.

More recently Mr. Dupuy has communicated to the Franklin Institute of Philadelphia a mode of protecting the sponge, when made, by enclosing the ore and carbonaceous matter in cases of sheet iron. When reduction is completed the whole is brought to a welding heat, and in that state is drawn out into a bar. According to his own estimate the cases will cost 23s. 5d. per ton of iron obtained; which cannot fail to prove a serious obstacle to the introduction of the system recommended by this gentleman.

In all the trials hitherto spoken of in connection with the direct process, the purer ores of the hematite class alone constituted the subjects of experiment. These were purposely selected in order not to encumber the apparatus unnecessarily with inert matter, and thereby add proportionately to the expense of obtaining the actual product. But such ores were also at the same time free from phosphorus; whereas the elimination of this substance, had phosphoric ores being used, would have been at all events a partial set off against the costliness of the process itself. Dr. Siemens, Past President of the Iron and Steel Institute, impressed with the advantage of using the regenerative principle in combination with a rotating furnace, has not confined himself to the use of the purer oxides of iron, but has operated for a considerable time on the earthy ores of the oolitic measures in the Midland counties, containing a large quantity of phosphorus. A mixture of the crushed mineral and coal is introduced into the furnace, which is gradually heated up to a full red heat, until the ore is supposed to enter into a

state of fusion. By this mode of treatment it was believed that reduction would be effected by the carbon suspended in the liquid mass, which would neutralize in a great measure, if not entirely, the oxidizing character of the carbonic acid and vapour of water, produced by the complete combustion of the coal.

A very instructive and highly interesting series of papers is to be found in the Transactions of the Iron and Steel Institute and of the Chemical Society, giving in ample detail the results obtained by Dr. Siemens in the course of his experiments. These results are such as to have inspired their author with great confidence in the future reserved for the process, on which he has bestowed so much thought and perseverance. In the course of the papers referred to, stress is laid on the fact that, whereas in the blast furnace the carbon chiefly escapes as carbonic oxide, in his regenerative apparatus the whole of it is practically converted into carbonic acid. Undoubtedly the heat obtained from the combustion of every unit of fuel is, on this account, much greater in the latter than in the former case. Owing to a variety of causes, however, even in this improved form of furnace not much more, if any more, than 25 per cent. of the heat evolved in the fire place is made available. On the other hand, it will be hereafter seen that in the blast furnace, with a less perfect character of combustion, the fuel actually performs a much larger comparative duty than that just mentioned.

I am quite prepared, from personal observation, to admit that a very considerable portion of the ore is ultimately fused by Dr. Siemens; and that possibly reduction under the favourable conditions claimed may, to some extent, take place. I am nevertheless assured, from my own experiments, that it will be extremely difficult to heat up a mixture of carbon and oxide of iron towards fusion, without the operation being accompanied by the reduction, previous to fusion, of a very large quantity of the metal contained in the ore. The moment this reduced iron meets the highly oxidizing atmosphere of the furnace-which must, in a rotating furnace, happen more or less-waste will begin.

Undoubtedly the useful product of this direct process contains only a fractional part of the phosphorus found in the ore; whereas the blast furnace, as has been mentioned, gives us metal containing practically the whole of this deleterious ingredient, as originally associated with the

minerals employed. Its absence, however, from the iron obtained in the rotating furnace is entirely due to the oxidizing nature of the vast quantity of iron which either escapes reduction, or, being reduced, passes again as oxide into the cinder. Were it possible, therefore, to carry on the operation without this loss, it is highly probable that the malleable iron obtained would be found contaminated with phosphorus.

The ew just expressed, as to the nature of the action, appears to coincide with the personal observations of my friend, P. Ritter von Tunner, who calculated, from the particulars given him of 100 heats with a mixture of calcined Northampton ore and calcined Black band operated on, that 27.6 per cent. of the metallic iron was lost in the slag.2

The product obtained possessed an advantage over that of the Chenot process, inasmuch as it was in the form of a solid hammered bloom, containing 99-71 per cent. of iron, and free from the great liability to waste in the subsequent stages of the manufacture. It only contained 074 per cent. of phosphorus, although made from materials which would probably give a pig iron having 1 or 1.25 per cent. of this metalloid, out of which 2 or 3 per cent. would remain in the bar iron when puddled in the ordinary way. We have here a material which can be used either for steel making or for the production of bar iron; but it is chiefly in connection with the former that Dr. Siemens attaches a value to the process.

Let us examine the subject under both these heads, regarding Dr. Siemens' product as a substitute, wholly or partially, for iron in the form of pig which, with the addition of ore, is capable of affording steel. The economy of the blast furnace, in the matter of fuel, has been already alluded to incidentally, where one ton of iron was stated to require only two tons of coal. A word now as to the labour. The very nature of the smelting process secures the exercise of an economy, in this respect, for which it would be difficult to find a parallel in the entire range of manufacturing operations. As an illustration, let us take a pair of furnaces making 1,000 tons of iron per week, which is by no means an uncommon rate of production. The three classes of raw

"Das Eisenhüttenwesen der Vereinigten Staaten.”

Although Professor Tunner, in his Preface, mentions that the centner (cwt.) made use of is one of 112 lbs., his calculation of the loss is on one of 100 lbs. The waste of iron, therefore, in reality amounts to 35.3 per cent. instead of 27·6.

material consumed in the Cleveland works for this quantity will weigh together about 4,750 tons, the whole of which can be so dealt with as to be brought in train loads to two or three points in close proximity to the furnaces. From the depôts they slide by gravitation into the wagons or barrows, from which they are shot into the tunnel heads. In like manner the pig iron and the refuse are so concentrated, in point of space, that effective mechanical means can be largely employed as a substitute for manual labour. All this is entirely changed, if anything approaching the weight referred to above is to be treated in a range of furnaces, each of which occupies four and a half hours in the reduction of 20 cwts. of ore, according to the weights given by Professor Tunner. A large area of ground would necessarily be occupied; and the materials having to be distributed in small quantities at various points over its entire surface, such labour-saving contrivances as are usefully employed at the blast furnace become impracticable.

In my opinion, the advocates of these direct processes underrate the importance of the duty performed by the blast furnace, in connection with such ores as those operated on by Dr. Siemens at Towcester, while they magnify its cost. In the case of an ore containing 40 to 42 per cent. of iron, there will be found almost exactly the same weight of earthy matter as there is of metal. These earths have a very powerful affinity for oxide of iron, and therefore seriously interfere with the process of reduction at high temperatures. Besides this serious inconvenience, we are thus, in the direct process, crowding up the space of an expensively worked piece of apparatus with inert matter, equal in weight and more than equal in bulk to the iron itself. Under such circumstances as those just related, it is needless to say how important it is to rid the process of this foreign matter as speedily as possible.

It will be more convenient to wait till we are dealing with the action of the blast furnace itself, before we show how very small a quantity of fuel is there required to effect the fusion of these earths. Suffice it now to say that the entire cost-for fuel, wages and flux-of freeing the metal from its earthy admixture can be shewn not to exceed 5s., on the ton of metal obtained from an ore containing 40 to 42 per cent. of iron. Professor Tunner gives no less than 36s. 8d. (20 gulden) per ton as the cost for labour at the Towcester works, carried on by Dr. Sie

mens.

This seems excessive, and may be partly due to the fact that the process at that establishment had scarcely passed-if, indeed, it had passed-the mere experimental stage. This item cannot however fail to be high, for in its nature the process closely resembles mechanical puddling, in which the wages per ton of product are at least four times as great as those incurred at the blast furnace. This difference will unquestionably be further increased by the circumstance that, while pig iron loses but little in the process of puddling, the ore in the rotary furnaces, with this direct process, only yielded about 33 per cent. of its weight in blooms.

In the matter of fuel, according to the authority already quoted, no less than 4 tons were being consumed per ton of hammered iron obtained; but for the purpose of comparing its cost with that of pig iron we will accept the weight named by Mr. Rose, the Manager of the works, to Professor Tunner, viz., 3 tons. From the data just described, we may, for our present object, accept the following as factors:

The question the steel maker has to consider is the extent, if any, of the superiority of malleable iron so obtained over pig iron, for the future stages of his operation. Speaking from such information as I possess, it would not appear that a furnace melting a mixture of pig and wrought iron gives a much greater weight of produce than one working pig and ore. If this be so, the actual expense of melting either mixture will be about the same; and we are left simply to compare the cost of the one set of ingredients with that of the other.

Let it be assumed, where the charge consists of blooms and pig iron, that it is made up of 80 per cent. of the former and 20 of the latter, while in the Siemens Landore mode of treatment 80 per cent. of pig iron has added to it 20 to 25 per cent. of ore.

Under such circumstances, and having regard to the great waste of iron and extra cost incurred for labour and fuel, in producing the blooms by the direct process, I cannot see how the materials employed for each ton of steel can cost less by this process, than by that where hematite pig iron (at 60s.) and ore are used, as at Landore.