As the amount of sulphur is not sensibly reduced, 0.3 per cent. is the limit which may be present in the pig-iron.

One important feature of the basic process is the value of the slag incidentally produced. It consists chiefly of lime and phosphoric anhydride. Of the latter there is as much as 13 to 20 per cent. Both these materials are used as fertilisers by the agriculturist, and basic slag, as might have been expected, has been found to be a valuable manure. According to Gilchrist,* in 1889 as much as 600,000 tons of this slag were made, containing 17 per cent. of phosphoric anhydride and 60 per cent. of lime. It is, of course, not a suitable fertiliser for all soils, calcareous ones for instance, but for sour, peaty, and clay soils it is of great value, as is shown by the fact that all the 600,000 tons were sold at from 208. to 308. per ton at the works. It is interesting to note that the phosphoric anhydride is combined with the lime in an unusual manner. Instead of being an insoluble tribasic phosphate, it is a readily soluble tetra-basic phosphate, and, if it be finely ground, the phosphorus it contains is readily assimilated by plants. At first attempts were made to treat it by various chemical methods, but it has been found best to simply grind it and use it in fine powder.

Of late years an attempt has been made to revert to a type of converter employed by Sir Henry Bessemer in his earlier experiments. It is of small size and has horizontal tuyeres. In the case of the Clapp-Griffiths converter † these are symmetrically arranged in the walls of the converter near the base. There is a very low pressure of blast, 5 to 6 lbs. per square inch, and the valves are adjusted automatically. The converters are adapted to charges varying from 1 to 3 tons. The Robert converter ‡ is small, and is adapted to charges of 1 ton or more, the charges in no case exceeding 3 tons. By the action of the blast through a range of tuyeres inclined at different angles, a rotary motion is gradually imparted to the fluid metal. These tuyeres are placed high up in the converter, and by tilting it and by manipulating the blast-valve, the operator can change the volume and pressure of the blast. The converters are mounted on trunnions in the usual way, and are revolved by means of hand-gearing. The main point about these little appliances is that they bring the production of soft steel within the reach of small manufacturers. In Clapp-Griffiths' practice it has been found that by * Journ. Iron and Steel Inst., 1887, No. i. p. 212, where there is also a bibliography of the subject.

+Ibid., 1883, No. ii. p. 705.

F. L. Garrison, Ibid., 1889, No. ii. p. 266.

keeping the carbon low, a certain amount of phosphorus may be allowed to remain in the steel without injury to the properties of the metal.

The Bessemer converter has been modified and employed for smelting copper regulus. The early experiments were not successful. In 1880, however, Pierre Manhès, the proprietor of the Vedènes copper works, Department of Vaucluse, France, obtained patents for its use in copper smelting, and smelting works were erected at Eguilles, near Avignon.*

The saving in fuel effected by the use of the converter is very considerable, and it is claimed that the Manhès process renders copper smelting possible in countries where, owing to the high price of fuel, the Welsh process is out of the question. As will subsequently be shown, the Welsh process consists of six to eight successive roastings and fusions in order to obtain coarse copper from the ore, all the operations being effected in large reverberatory furnaces. In the Manhès process from one to three of these operations are effected by blowing air through copper regulus in a converter which differs considerably from that employed for ironsmelting

Its later form is a horizontal cylinder, holding about 1 ton of regulus, and turning on axial trunnions, through one of which the blast passes to a row of tuyeres, parallel to the axis. The level of the tuyeres can be altered as the "blow" progresses, and the copper as it is reduced sinks to the bottom of the converter, which is gradually turned so as to permit the blast to pass into the regulus above the level of the copper, which thus escapes chilling. The blow lasts 15 to 20 minutes, and at one operation "blister copper" may be obtained from the "white metal" of the Welsh smelter. For poorer regulus of 50 to 60 per cent. of copper (blue metal), it is well to blow twice in separate converters. regulus may be run into the converter direct from the smelting furnace.

The

The electrical furnace, originally devised by Sir W. Siemens, will probably become a powerful instrument of research in the near future.

It consists, in one of its latest forms, shown in Fig. 73a, of a metallic framing K, lined with a non-conducting refractory material R, such as magnesia, and containing a crucible C R, that may be of carbon or other material. The framing has two metal castings M M mounted upon it, each of which carries a terminal A, and a carbon-holder V, placed at an angle of 45°

* Annales des Mines, 8th series vol. iii., 1883, p. 429.

in either direction; M M are insulated from K, and from each other by sheets of mica. A plug B closes an aperture in the roof of the furnace, through which material may be introduced from time to time into the crucible. The front and back of the furnace may be closed by sheets of mica held in position by the plates I I, so that the reactions taking place in the crucible may be watched by the experimenter, who must, however, protect himself from the light from the interior by

suitable dark glasses.

FIG. 73a.

The chamber may be rendered partly vacuous, or may be charged with any particular gas by means of the two tubes 0 0.

By the aid of such an appliance, and using a current varying from 25 to 450 amperes, with a difference of potentials at the terminals of the furnace of from 45 to 70 volts, Moisson has succeeded in fusing such refractory substances as lime, and in readily preparing samples of the difficultly reducible metals, such as zirconium, chromium, calcium. He has since † made the furnace reverberatory, by placing the material to be treated in a hollow tube of pure carbon placed at right angles to the electrodes. The heating and electrolytic effects of the arc are thus kept distinct.

Comptes Rendus, vol. cxv. (1892) p. 1031. † Ibid. vol. cxvii (1893) p. 679.

CHAPTER VIII.

MEANS OF SUPPLYING AIR TO FURNACES.

Methods of Producing Draught. In every furnace it is necessary to conduct away the gaseous products of combustion to enable fresh air to enter and to give up its oxygen to the fuel. This passage of the fire gases from the furnace and of the air to the furnace may be effected in two ways:-First, by exhausting the products of combustion; and, second, by forcing in air for combustion. In the former method, a space containing rarefied air is formed in the furnace, and atmospheric air flows in from outside so as to preserve the equilibrium, whilst in the latter method the pressure in the furnace is greater than that of the air outside, and consequently the air and the fire gases are forced out. Although the current is usually the same in both cases, the influence on the combustion may be different when the movement is effected by the compression or rarefaction of the air.

The exhaustion of the air is usually effected by means of chimneys. The chimney or stack may be regarded as a vertical pipe containing heated and expanded gaseous products of combustion. The column of gas within the chimney is, in consequence of the expansion due to heat, considerably lighter than a column of air of the same height at the ordinary temperature. The consequence is, that owing to the difference of weight, there is an excess of pressure of air under the grate, and movement ensues. This difference in the weight of the hot and cold columns is equal to the weight of the increase in volume that would be produced by heating the cold column of air to the temperature of the chimney. If the elongation thus produced be represented by h, the velocity of the movement v = √2gh; but h is dependent both on the height of the cold column of air and on the difference of temperature within and outside the chimney, whence it follows that theoretically the velocity of a current of gas within a chimney increases proportionately with the square root of the height and the square root of the difference of the external and internal temperatures, and that consequently the action of equal increments of height

and temperature becomes continually smaller. But the action of a chimney does not depend so much on the velocity produced as on the weight of air supplied in a given period of time. The higher the products of combustion are heated, when they pass into the chimney, the greater is their volume, and with equal velocity the less weight of gas would actually pass through the same chimney. The velocities, it has been shown, increase with the difference of the external and internal temperatures, but only in proportion to the square roots, whilst the relative weight of the gases decrease in direct proportion to the temperature by of the original volume for every degree Centigrade. There must therefore be a limit where the action of the chimney reaches its maximum, and it has been calculated that this maximum is attained when the difference of temperature amounts to 273°, or, in other words, when the external air is at the mean temperature, and the chimney gases have a temperature of 300°. Similarly it has been calculated that the quantities of air supplied to the chimney between 200° and 400° differ but slightly, so that for the chimney-draught these temperatures appear to be the most suitable, whilst for the utilisation of heat in the furnace the lower chimney temperature is obviously to be preferred, and even when the temperature of the chimney gases is only 100°, the quantity of air drawn in is to that drawn in by a chimney having a temperature of 300° as 7: 8.

The amount of gas passing out of the chimney could be determined from the height and temperature of the shaft and the specific gravity of the gases, if only the resistance given by friction to the gases in motion and the so-called "free section" were known. The term "free section" is applied to the sum of the areas of the interstices between the lumps of fuel on the grate. The smaller this free section is, the quicker is the motion of the air in it, and the more perfect the combustion, provided that the reduction in section is due to a diminution of the area of the grate and not to undue clogging of the layers of fuel which, by increasing the friction, at once diminishes the action of the shaft as the volume of air supplied is necessarily lessened. For this reason small grates give rapid combustion and large grates slow combustion, with the same chimney. It has, however, previously been shown that a rapid motion of the air, in this case equivalent to rapid combustion, facilitates the production of carbonic anhydride, whilst a slow motion facilitates the production of carbonic oxide. As a rule, therefore, for the production of heat and for high temperature, rapid combustion on a small grate is to be preferred to slow combustion on a large grate.