342 PROCEEDINGS OF THE AMERICAN ACADEMY 



the workman, stirring the mass as it begins to lose its fluidity, gathers 

 into rough masses the aggregated particles, which are always spheri- 

 cal in general form. From the masses, which are veiy porous and 

 unequal, a bar of regular form is obtained by the usual means of 

 pressing, or hammering and rolling. 



" There is in this process strictly a segregation of particles of pure 

 iron from the crude mass, which, under the agitation of stirring, unite 

 to form rounded aggregations, and the heat of the furnace being in- 

 creased, the separation of pure iron continues, until the melted im- 

 purities alone remain. The change of crude to pure iron is accom- 

 panied by the production in part of the impurities which I'emain ; 

 they are not educts. Aside from the loss of carbon in the form of 

 carbonic oxide, the phosphorus and sulphur, — which my experiments 

 have proved are always united to the metallic bases of the earths or 

 alkalis, — with these bases, burn into oxidized products ; while the 

 silicium and a portion of the iron, also oxidized, form the melted 

 slag, or cinder, as an additional foreign matter. To the loss of im- 

 purities we must also add the weight of iron burned in forming sec- 

 ondary products, so that, if the operations were performed on crude 

 iron containing ninety -two per cent of pure iron, no more than eighty- 

 two per cent of malleable iron will be obtained. By the method of 

 Bessemer, crude iron in a melted state is exposed in a nearly 

 closed receptacle to jets of air forced into and under the fluid, and 

 it is alleged that such an excess of heat is produced in the process, 

 ' that the metal continues to boil even after the blast has ceased.' 

 The direct statement is made, that ' the air, dividing into globules, 

 and diffusing itself among the particles of fluid iron, and thus coming 

 in contact at numerous points with the carbon contained in the crude 

 iron, and producing thereby a vivid combustion,' and the same action 

 is implied in other parts of Mr. Bessemer's patent-specification. 



" Now it is well known to chemists, that the combustion of the 

 carbon of crude iron cannot take place under the conditions. This 

 carbon exists in gray iron in the allotropic state of graphite, and is 

 not combustible even alone, when exposed highly heated to a current 

 of atmospheric air. We burn it in the laboratory by the application of 

 oxygen in some condensed state only. The proper chemical expla- 

 nation of this point is a very simple one. Iron, which is a highly 

 combustible body at ordinary temperatures, has its attraction for 

 oxygen enormously increased by the heat of fluidity, and in com- 



