Such were the conditions under which the first charge of pig iron was converted in a vessel neither internally nor externally heated by fire.

I, however, desired to convert a second charge of pig Iron which had been put into the cupola; and in order to prevent this dangerous projection upwards of sparks and molten slags, a temporary expedient was resorted to, which, however, failed in its object.

I procured one of those circular, chequered cast-iron plates so much used in the London pavements to allow coals to be put into the cellars below the pavement. This plate, which was about a foot in diameter, was suspended by a chain at a distance of about 18 in. above the central opening in the top of the converter, as shown in Fig. 39, Plate XIII.

This, as a mere temporary device, was deemed sufficient to allow the conversion of another 7 cwt. charge to be effected, without any danger of setting fire to the premises. The converting operation went on quietly as before, but when the eruption commenced, I saw the suspended plate get rapidly red-hot, and in a few minutes more it melted and fell away, leaving the chain dangling over the opening, and allowing the slags and splashes of metal to shoot upwards as before. Thus it happened that the first converter that I constructed was at once condemned as commercially impracticable, owing to this vertical eruption of cinder, and for this reason only.

All attempts to lessen the violence of the process by the reduction of the number of tuyères,or by lessening their diameter, or by diminishing the pressure of the blast, only resulted in a reduction of the necessary temperature, and in preventing the conversion of the molten pig into malleable iron. In one case the trial of a diminished area of tuyére openings resulted in nearly the whole charge of metal, after more than an hour's blowing, being converted into a solid mass of brittle white iron, similar to ordinary refiner's plate metal. Indeed, I may say the result of all my early investigations proved to me, beyond the possibility of a doubt, a fact which has since been confirmed in every Bessemer steel works throughout Europe and America, viz.: that rapidity of action, ending in a violent eruption, is an absolutely necessary condition of success. Not only must the converted metal acquire an enormously high temperature, so that it may not be chilled when pouring it out of the converter, or when a relatively large quantity of much cooler metal be added to deoxidise it, but it must not chill and form a shell in the ladle during the comparatively long time required for casting the ingots. Hence, to carry out the Bessemer process successfully, a temperature must be obtained very considerably above the mere melting temperature of malleable iron; and in order to secure this it is necessary to drive powerful streams of air into the metal, so as to divide it into innumerable tiny globules diffused through out the whole body of iron under treatment which, for the time being, may be likened to a fluid sponge with the active combustion of carbon with oxygen going on in every one of its myriads of ever-changing cavities.

It has been found that the union of carbon and oxygen takes place so rapidly at this high temperature as to produce a series of mild explosions. In the large converters in common use, a space some 8 ft. or 10 ft. in height above the normal level of the metal is provided, in which this violent action expends itself unseen, and is only partially recognised by a small quantity of slags leaping out of the mouth of the converter.

With these facts before us, it must be self evident that all attempts to produce malleable iron in a plain cylindrical vessel that has no top to it, and in which the metal normally rises to within 6 in. of the open mouth, must utterly fail from two causes: first, because heat would fly off so freely that the temperature of molten malleable iron could never be reached; and secondly, because nearly all the metal contained in such a shallow, open-topped vessel would have leaped out of it, and have been scattered in all directions on the occurrence of the explosive eruption, without which no charge of molten pig iron has, or can be, converted into fluid malleable iron by a blast of air.

I had no sooner condemned my first cylindrical converter than I commenced to remedy its defects. The most obvious and ready way of doing this would have been simply to make an opening on one side of