874 Traxsactions of the A^ieeican Institute. 



If the slow heating is continued it next becomes purple, after which 

 blue and last black. 



Professor Henry, of the Smithsonian institute, informed the writer 

 a few years since, that these colors indicated the thickness of the 

 scale or Him of oxyd upon the surface ; and that the blue surface 

 might be removed by careful polishing, so as to expose the purple, 

 and the operation continued would expose the straw color. If, after 

 the black scale forms, which is black oxyd of iron, the same as scales 

 which are found about the blacksmith's anvil, the piece be cooled 

 slowl}^ it will be annealed. The same temperature will produce the same 

 degree of hardness in tempering, whether the scale forms or not. If 

 air should be absolutel}^ excluded no scale would form. If steel be 

 heated too hot, part of the carbon will be vaporized, which, bursting 

 out, will disintegrate the metal ; some borax applied will prevent 

 further escape of the carbon, when, by welding and condensation 

 under the hammer it may be partially restored ; but it will be loioer 

 steel and possibly could not be tempered. 



High steel, such as is used for tools, to be tempered, has about one 

 and a half per cent of carbon. Low steel, about a half of one per cent 

 of carbon, and is of the quality used for tires for locomotive drivers, 

 <fec. ; it cannot be tempered by heating and cooling, but may be con- 

 siderably hardened by condensation under the hammer when nearly 

 cold. 



Steel is, or should be, composed only of iron and carbon. 



Case-hardening is simply carbonizing or making steel of the surface. 

 Decarbonizing is expending the carbon, and the product, if from cast- 

 steel, is homogeneous iron. 



There is no chemical change of quality in steel which is hardened 

 or annealed ; hardening is simply the inauguration of tensions by 

 unequal cooling, when conducted in the manner described, and anneal- 

 ing is relieving those tensions, by heating the steel until the tension 

 is relaxed, and then cooling slowly and uniformly. 



If the thin plate above referred to could be caught jfym^, as it fell 

 hot from between the cast-iron plates, between two sponges having 

 flat surfaces, well saturated with water, in such a manner as to cool 

 the whole of both surfaces simultaneously, it would not be hardened 

 at all. If the experiment should be conducted with a steel plate half 

 an inch thick, then it would be hardened ; because the cooling would 

 not be uniform, the surfaces would be cooled in advance of the middle 

 metal of the mass ; steel being comparatively a non-conductor of heat, 



