A. From Four Pot Crucible Furnace: Each Heat 400 Pounds in 4 Hours or 100 Pounds an Hour, Each Heat Pours 4 Ingots 3 × 3 × 36″. B. From Fifteen Ton Bessemer Converter: 30,000 Pounds in 20 Minutes or 90,000 Pounds an Hour, Each Heat Pours 6 Ingots 19 × 20 × 62″. C. From Fifty Ton Open-Hearth Furnace, 100,000 Pounds in 8 Hours or 12,500 Pounds an Hour, Each Heat Pours 6 Ingots 24 × 32 × 72″. ] Reddish-brown smoke and a shower of sparks come from the converter. These gradually develop into a flame. The blast shows considerable partiality in selecting for its first attention the metalloids silicon and manganese, in preference to the iron itself or any other of the metalloids present. After from three to five minutes half of the silicon and manganese have been burned out. If the temperature of the metal and other conditions have become right the carbon then begins to burn. This gives a change in the nature of the flame which becomes large and of a dazzling whiteness. The metal is hot—very hot—so much so that pieces of cold steel often must be dropped in to cool it somewhat. This is known as “scrapping” the charge. An experienced blower can judge through every period of the operation of the condition of his metal and just how things are progressing. After some minutes the flame begins to waver and later “drops”; i.e., there is scarcely a flame at all. This signal, which is very definite to an experienced man, cannot be lightly disregarded. Oxygen has affinity for iron as well as for the metalloids and it is only because of its greater love for silicon, manganese and carbon that it has thus far largely neglected the iron. With the metalloids mentioned out of the way, as they are when the drop occurs, the iron will begin to burn. Were the “blowing” continued we would shortly have no iron left, but in its place a mass of iron oxide and slag. Thus we see that during the first minutes of the blow, more than one-half of the silicon and manganese are burned. The remainder of these and all of the carbon are removed in the subsequent five or six minutes. At the end of this short blowing period we have practically pure iron. [Illustration: TWO CONVERTERS IN OPERATION AND A THIRD POURING ] The metal is not yet in condition to pour well, however, largely because of the dissolved air and gases which it holds. Something akin to the “killing” of the steel which we observed in the crucible process must be accomplished or ingots from it will be spongy. And, having practically no carbon, it is not yet “steel.” Bessemer, knowing that the finished steel should contain carbon, tried to stop the blow long enough before the drop of the flame to leave exactly the desired amount of this element. He found this difficult to do and therefore uncertain. It was found to be far better to blow until the drop of the flame and then put back sufficient carbon to give the proper composition.