During the second period, the presence of a flame like that of a candle proves that the combustion of carbon is followed by the production of carbonic oxide, and the oxygen is now entirely absorbed by the combustible matter in the fluid mass, since the jet of gas at the mouth of the converter burns only on the surface when it comes in contact with the air.
The flames remain; strong detonations take place in the apparatus; they are accom panied by the projection of viscid portions of the interior mass, mixed with metallic grains, against the wall of the chimney. The flames appear more clear during several seconds after each explosion. These explosions do not seem to be essential to the pro cess; and a skilful operator will at length succeed with most irons in working without them. The noise produced at this time is similar to that of a fluid mass in a state of violent ebullition. Much smoke now issues from the chimney. At this period the iron burns directly and its combustion is very lively. The oxide of iron is carried off as smoke, and the loss in weight is considerable.
The iron has now reached a period when it is somewhat viscid, having been refined till it has reached a condition of malleability. Bubbles are formed, and when they receive oxygen and carbonic oxide in proper proportions they become explosive.
A calm now succeeds. A long white flame extends from the mouth of the converter. Soon this flame experiences an entire change. It becomes torn on the edges, trembles, and suddenly falls, as though the blowing-engine had in a moment lost nine-tenths of its power. After this fall the flames are not bright, and much smoke issues from the converter. The operation is now stopped. If it be continued, there will be a rapid formation of silicate by the fused lining of the converter combining with the oxidized metal, and ultimately an oily fluid silicate of iron will fill the converter.
Practice teaches how to seize the exact moment to stop, that the required degree of hardness may be produced. It is always a delicate point to determine the exact moment when the operation shall be discontinued.
When the flame falls, the converter has within it a very fluid metal holding much gas in solution, which, if allowed to cool, is not malleable, and resembles in its fracture what is known as "burned iron." The action of the blast is terminated by reversing
the apparatus and shutting the air-valve.
During the fourth period, the continuance of the flame proves that the combustible gas is produced in abundance; but the cessation of detonations shows that the iron has acquired a temperature sufficiently high to absorb all the oxygen thrown into it. The combustion of the last portions of carbon (probably the graphite) appears to take place indirectly by the decomposition of the oxide of iron, free or combined. The iron has become almost exclusively the combustible substance. Its affinity for oxygen increases with its temperature, and the action of the mass is in favor of its combustion, in prefer ence to the sulphur, the phosphorus, and other bodies that have escaped burning at former periods. The refining, therefore, can be carried no further. The flame produced at this period is white and brilliant; it falls generally when the last particles of carbon have disappeared.
The final product is always very liquid. The slags formed towards the last of the operation run easily, are glassy-black without, and yellowish-green within.
Seven and one-third per cent. of fused spiegeleisen, from Siegen, or other manganif erous iron, melted in a special compartment of the reverberatory furnace, is then intro duced by means of a small crane and ladle. The addition of this metal always pro duces a tumultuous action, which is more or less violent in proportion to the amount of air that has been driven in, and varying with the different irons treated.
After allowing the metal to mix thoroughly, it may be poured into the ladle on the crane without turning up the converter; but if it is desired to make the mixture still more complete, it may be done by bringing the converter again to a vertical position and blowing air into it for a few seconds. The metal is then poured into the distributing ladle. During this last blast the flame passes from red to yellow, and the same phe nomena are seen to occur in rapid succession as have been described above.
From the distributing ladle the diffused metal must be run, before it cools, into the ingot-mould.