lo Lange, Some Remarkable Steel Crystals. 



carbon ; this is the temperature of the well-known phenom- 

 enon of recalescence. Brinell does not state the actual 

 temperatures for V and IV in this case, but Howe states 

 in reference to the same that Brinell's temperature V 

 appears to be Arj (690^0.), and his W appears to be 

 Ac2, 3, and therefore to vary with the percentage of 

 carbon as shown in Roberts-Austen's diagram {vide Fifth 

 Report of the Alloys Research Committee, 1899). 



If steel is heated to a temperature above W, and 

 allowed to cool slowly without work, the crystal grains 

 continue to grow until the temperature V is reached, 

 beyond which there is no further growth. Now the 

 action of forging or rolling strongly opposes crystalli- 

 zation in the case of both iron and steel. Like agitation 

 of salts crystallising from an aqueous solution, it appears 

 to arrest the development of crystalline structure. It 

 follows from this that all forged work should be finished 

 as near to the temperature V as possible, so as to prevent 

 the formation of a coarse structure during subsequent 

 undisturbed cooling, and it also follows from this that 

 all forgings which are not finished at one heat should have 

 their structure refined by re-heating the whole forging to 

 a suitable temperature as a last operation, in order to 

 destroy any coarsely crystalline structure produced in the 

 parts first finished by reason of their proximity to the 

 portion last heated up for finishing. 



The experiments of Stead and Heyn show that, for 

 breaking up coarsely crystalline structures in very mild 

 steels, the refining temperature should be about 900 'C, or 

 beyond the highest critical point for iron, and an 

 examination of Brinell's later experiments on the heat- 

 treatment of steels of varying carbon contents would 

 appear to make this hold good for carbon steels also, if 



