dencies [A] and [D], the iron-oxidising [B] and [C], the carbonreducing [C], and the carbon-oxidising [D] and [E] tendencies. Knowledge of the relative magnitude of these forces has been gained from the careful experiments of Sir Lowthian Bell and Dr. Alder-Wright. In tracing out the general chemical changes undergone by the oxide of iron, the furnace may be conveniently divided into three regions. In the uppermost region, tendencies [A] and [D] jointly are stronger than [B] and [C] jointly, and consequently rapid reduction of ferric oxide takes place. In this region, too, tendency [C] is more powerful than tendencies [D] and [E], and consequently carbon deposition takes place to a large extent. Here, too, the limestone flux is for the most part calcined into quicklime, raw coal, if used, is coked, and if iron carbonate is used, instead of oxide, it becomes converted into oxide, the reducing and carbon-depositing reactions going on simultaneously with the formation of the oxide.

In the middle region, the iron-reducing tendencies are almost balanced by the iron-oxidising ones, whilst the carbon-depositing tendencies are equalled by the carbon-oxidising ones. Here reduction takes place, but only languidly, the chief effect produced in passing through this region being an increase of temperature.

In the lowest region, the reduction of the residual iron oxide is completed chiefly through the agency of cyanides formed in the vicinity of the tuyeres; the reduced iron melts dissolving a certain amount of the finely divided carbon in contact with it, together with small quantities of silicon, phosphorus, and sulphur reduced by subsidiary reactions. The non-metallic constituents of the ore and the lime of the limestone also fuse, forming slag.

The nitrogen brought into the furnace acts solely as a diluent, except in the lower portions of the furnace. It has long been known that when nascent potassium or sodium vapour finds itself simultaneously in contact with carbon and nitrogen, the three elements combine, forming a metallic cyanide. The cyanide thus formed acts on the last portions of unreduced oxide of iron, converting it into metal and becoming itself changed to cyanate. This is probably decomposed with the formation of an alkaline carbonate and the elimination of nitrogen. The alkaline salts condensed in the upper portions of the furnace are again brought down to the level of the tuyeres as the materials sink. Consequently each particle of alkali metal does duty over and over again, the alkalies introduced in small quantities in the fuel accumulating in the furnace to a very large extent. As much as 4 cwt. of alkali metals and 2 cwt. of cyanogen per ton of iron made have been repeatedly found in the gases near the level of

the tuyeres. This concentration of alkali explains the fact that furnaces reduce more readily after they have been some time in blast.

In order to effect the smelting of iron, a definite amount of heat is required to perform the general work of the furnace, the various items in the appropriation of heat in an 80-foot blastfurnace, during the production of 20 cwt. of pig-iron from Cleveland ore being as follows:

In order to produce this heat with a minimum expenditure of fuel, it is necessary that the whole of the carbon used as fuel should be oxidised to carbonic anhydride, an ideal that can only be realised in the reduction of lead oxide. The relative strengths of the forces involved in the nine reactions described are, in the case of iron ores, such that it is not possible to convert more than 35 or 40 per cent. of the carbon burnt into carbonic anhydride, the rest necessarily escaping as carbonic oxide. Consequently more fuel must be used to do the work of the furnace. The most economical results are obtained when the ratio

CO,

CO

is greatest. This ratio, however, itself is modified by the temperature, the velocity, and the distribution of the gaseous current. Care must be taken to lessen, as far as possible, the amount of

the carbon burnt by the carbonic anhydride. The solid carbon is divided into two portions-one burnt in the zone of fusion, and the other in the zone of reduction. In the second case the reduction is effected with consumption of solid carbon, whilst in the former case, it is effected solely by the gases. The ideal case would be that in which carbon descends to the level of the tuyeres, and is there converted into carbonic oxide.

A convenient classification of blast-furnaces is that based on the ratio of the maximum diameter to the height. In this manner the following three classes may be distinguished:

1. Squat furnaces, in which the height is less than, or equal to, three times the diameter,

H
D

<3.

H

2. Ordinary furnaces, in which the ratio

varies between 3

D

and 4, but is usually about 3.5.

3. Elongated furnaces, in which the ratio is greater than 4.

The following table gives comparative data of the dimensions and workings of ten typical blast-furnaces used for the smelting of iron :

H

D

Cubic

Total

at

Diameter Diameter Diameter Daily
at
Out-
Capacity. Height. Hearth. Boshes. Throat. turn.

at

In the first six furnaces coke is the fuel employed, whilst charcoal is used in the last four. The names and situations of the furnaces are: 1. Newport, England; 2. Edgar Thomson "F,” Pittsburgh, United States; 3. North Chicago, No. 6, Chicago, United States; 4. North Lonsdale, No. 3, Lancashire; 5, Union, No. 3 and No. 4; 6. Union, No. 2, Chicago; 7. Midland, Crawford County, Missouri, United States; 8. Treibach, Austria; 9.

Ferdinand, Hieflau, Austria; 10. Wrbna, Eisenerz, Austria.
The first four of these furnaces belong to the second class, in
H
which is less than 4, whilst the last six are elongated furnaces

D

In furnaces of this type, the working is irregular on account of the contraction at the throat, which renders it difficult to apply any mechanical method of distributing the charge uniformly.

Mr. E. Walsh* shows that by placing the boshes, or widest diameter of the furnace, entirely within the zone of complete fusion a constant supply of fuel at the level of the tuyeres would be secured. He also concludes that, within practical limits, the narrower the furnace-shaft is constructed the more energetic the actions of reduction and carbon-impregnation will be. This view is supported by the outlines of the ten furnaces shown in Fig. 59. It may be noted that the Union No. 2 (6 in table) and the Treibach No. 3 (8 in table) do not owe their low fuel consumption to great capacity, to the employment of high temperature of blast, nor to the great length of time the materials are allowed to remain in the furnace.

It was formerly the practice to allow the waste gases to burn at the throat of the furnace. Since the introduction of the hot blast, however, it has become usual to apply some arrangement for closing the throat of the furnace and for collecting the waste gas. The and cone, cup invented by Parry, is the arrangement generally used. The throat is closed by an iron cup-shaped casting, the diameter of which at the lower end is about one-half of that of the throat of the furnace. Beneath this cup, a cast-iron cone is suspended from its apex, and when the charge has descended

*Engineering, vol. xlii. (1886), p. 513; Trans. Amer. Inst. Min. E. vol. xv., 1887, p. 419.