ON THE INFLUENCE OF SILICON ON THE PROPERTIES OF STEEL. 271 



grey it also becomes weaker. By a still further addition a closer grain 

 results, but it is even more brittle than in its white condition. Too much 

 silicon also produces, as in the silicon steel now being described, lack of 

 fluidity and greater shrinkage. Mr. Keep thinks that silicon in cast-iron 

 is evidently, to some extent, combined with the iron and carbon ; but 

 whether it exists also in a form corresponding with graphitic carbon, 

 mechanically mixed with the remaining mass, is a question still in dispute 

 and unanswered. As now well known, the principal effect of silicon is to 

 change the carbon from the combined to the graphitic state. One point 

 particularly mentioned by Mr. Keep, and an important one, is that silicon 

 irons have always had the reputation of imparting fluidity to other brands ; 

 and naturally this was at first supposed to be owing to the silicon added. 

 It has now been found that this is not directly so, and that it is merely 

 from the fact that the silicon present causes an increase in the quantity of 



Diagram 1. 



I 



Diagram by Mr. King, showing the influence of Silicon during dccarbiirisation 

 upon the graphitic and combined carbon in cast-iron, 



graphite, and consequently a more fluid cast-iron. It is not, therefore 

 directly the cause except by its indirect action on the carbon. 



In conclusion on this point, Mr. Snelus said more than seventeen 

 years ago, it is generally supposed that the absorption of much silicon 

 tends to set free the carbon in the graphitic state. No statement more 

 concisely expresses the influence of silicon on what is termed cast-iron 

 than that given some eight years ago by Mr. C. F. King of Newport, 

 U.S.A., in an able paper read before the American Institute of Mining 

 Engineers on ' The Chemical Action of the Bes.semer Process.' He said 

 it is due to the presence of silicon in pig-iron that carbon is set free 

 on cooling, and it is in proportion to the elimination of the silicon that 

 the carbon remains chemically combined. Mr. King gives a diagram 

 showing the rate of elimination of the metalloids in the process named, 

 and it is a somewhat remarkable coincidence that the percentage point 

 (18 per cent, silicon), where the diminishing silicon curve cuts the 

 combined carbon and graphitic curves, is exactly that which gave the 



