the Recalescence of Iron. 113 



had it not been for the contrast caused by the unclouded por- 

 tions AB, CB. Very shortly after its formation the cloud rolled 

 away, and immediately afterwards a reglow occurred through- 

 out the whole of the heated portion of the bar. The eyes had 

 been kept upon the convex side of the bar in this trial ; and 

 as it was thought that the concave side might possibly show 

 up bright instead of dark, by contrast with the unstrained 

 metal, the bar was heated a second time and again watched on 

 cooliug. Rather contrary to expectation, the portion D, when 

 the cooling had reached a certain point, showed up as being 

 slightly clouded ; as before, the cloud rolled away and the re- 

 glow immediately succeeded, which also, as before, seemed to 

 raise the whole of the heated portion of the bar to the same 

 temperature. The explanation of the formation of the cloud 

 appears to be this : above but near the critical temperature 

 the specific heat of the strained portion is lower than that of 

 the unstrained portion, and, as a consequence, very shortly 

 after the bar has been removed from the fire the temperature 

 of the former is more rapidly lowered than that of the latter. 

 Again, the reason why the strained portion has a less specific 

 heat than the unstrained, and why, after the reglow, the strained 

 and unstrained portions become apparently once more uniform 

 in temperature, may well be that, in consequence of the greater 

 internal friction of the strained portion, or from some other 

 cause, the contraction of its molecules does not go on at so great 

 a rate. When the temperature of a body is raised, part of the 

 thermal energy expended on it is expended in producing vis 

 viva of the molecules, and part in pushing them asunder. If 

 we can prevent the molecules from being pushed asunder, a 

 given amount of thermal energy imparted to the metal will 

 produce a greater rise of temperature. Similarly, when the 

 metal is cooling, a given loss of thermal energy will be at- 

 tended by a greater lowering of temperature when the con- 

 traction of the molecules is lessened. As a consequence, when 

 the strained and unstrained portions are cooling near the cri- 

 tical temperature, there is a greater amount of molecular 

 potential energy, and therefore a greater rise of temperature 

 when the explosive change occurs at reglow, in the former 

 than in the latter. The next experiment shows that effects 

 similar to the above can be produced by hammering. 



Experiment IV. — The straight portion of the bar used in 

 Experiment III. was heated to a white heat, and whilst hot 

 was placed upon a small anvil and struck twice very hard with 

 the sharp end of a hammer at two separate parts, so that two 

 rather deep dents were made in the bar about 10 centim. apart 

 from each other. The bar was now heated a second time to a 



Phil. Mag. S. 5. Vol. 25. No. 153. Feb. 1888. I 



