the Recalescence of Iron. 109 



the metal and also in the electrical resistance. These sudden 

 changes would appear to indicate an equally sudden change 

 in the molecular architecture of the iron ; and it seems reason- 

 able to suppose that when, on heating, the critical temperature 

 is reached, a certain amount of the whole thermal energy 

 which may be imparted by an electric current or a burner to 

 the wire is used up in producing this change, whilst an equal 

 amount of thermal energy is given out when the wire, on 

 cooling, reassumes its original molecular structure. It by no 

 means follows, however, that the so-called latent heat con- 

 nected with the sudden change can be detected unless by very 

 refined apparatus ; for it is not at all unlikely that the sudden 

 loss of magnetic properties may be due simply to a rotation 

 of the molecules about their axes ; and the researches of Pro- 

 fessor Ewing* prove that there may be considerable rotation 

 of the molecules about their axes, and yet at the same time 

 very minute change of temperature resulting from this rotation. 



Near the higher temperature of 1000° C. there is again a 

 most remarkable alteration in the behaviour of an annealed 

 iron wire which is being heated when under stress or strain. 

 If it is stretched by a slight weight it suddenly unstretches f ; 

 if it is under a slight bending-stress it suddenly unbends; and 

 if it is under slight twisting-stress it suddenly untwists ; 

 whilst, on the contrary, if it has been previously bent or 

 twisted permanently and then entirely released from stress, it 

 suddenly bends more or twists more as the case may be J. 

 Further, there are sudden changes, evidently at the same 

 temperature, in the thermoelectrical relations which exist be- 

 tween strained and unstrained iron, and also between stressed 

 and unstressed iron ; these changes, like those before men- 

 tioned, being opposite in direction for stress and strain §. 

 All these phenomena indicate, according to the author's belief, 

 a second notable change in the molecular architecture of 

 the iron ; they certainly seem to indicate a change which 

 must involve the expenditure of thermal energy to produce 

 it w T hen the iron is being heated, and the giving out of thermal 

 energy when the iron is being cooled : in other words, here 

 must be, one would think, a second point at which heat becomes 

 latent. 



Now suppose that the iron has been raised in temperature 

 above 1000° C. and is cooling : when the critical temperature 

 is reached the molecules would begin to come back to their 



* PHI. Trans, part ii. p. 553 (1885). 



t Phil. Mag. [4] vol. xlvi. p. 472. 



\ Ibid. [5] vol. xxiv. No. 148, p. 256 (1887). 



§ Ibid. [5] vol. xxv. No. 152, p. 45 (1888). 



