( 532 ) 



But this does not solve the problem yet, for it is a fact that 

 cementite can l)e formed in the sohd iron-carbon mixtures by heating 

 them to temperatures lying even some hundi-eds of degrees below 

 the eutectic temperature. So we can only accept the new 7'A'-tigure, 

 when the phenomenon mentioned just now has been explained. 



Before proceeding to this explanation, I will first mention Royston's ^) 

 very interesting experiments, from wiiich the said formation of 

 cementite appeared with the greatest distinctness. Royston worked 

 with malleable cast iron, which had been obtained by heating cast 

 iron up to 850°, and then allow it to cool exceedingly slowly in 

 an atmosphere free from oxygen. In (his process, which is called 

 the Blackheartproci-ss, a conglomerate of ferrite and graphite is obtained, 

 and so the cementite has entirely disappeared as phase. In this it is 

 to be expected that ferrite will contain besides a little carbon a 

 trace of iron carbide in solulion. We may now assume with great 

 probability that this conglomerate is a fixed condition, which was a 

 state of equilibrium at a temperature under that of the mixed crystal 

 eutectic point (so under about 700). 



Now RoYSTON found that when this malleable cast iron was 

 heated at 720° for three hours, and then comparatively quickly 

 cooled, cementite had formed, which appears clearly fi-om the following 

 analysis results. 



graphite. Ci\^sFe^C 

 composition before the heating 3.5 "/o trace 



after „ „ 2.65 0.85 



In a following experiment a new (juantity of malleable cast iron 

 ^vas heated to 1030°, and appeared, as is shown below, to possess 

 a greater quantity of bound carbon. 



In perfect agreement with the above analysis results perlite, ce- 

 mentite, and graphite was found on micrographical investigation. So 

 these experiments show with perfect clearness that when a conglo- 

 merate of iron and graphite is heated, a formation of cementite may 

 take place at temperatures under the initial melting-point of the 

 mixture. This fact, which at first sight is in direct contradiction 

 with the assumption that cementite is always metastable, has not 

 been accounted for up to now, and yet in my opinion it is not 



Ï) Journ. h"on and steel Institute 1, IGG (1897). 



