the Recalescence of Iron. Ill 



so that, according to the above theory, there should be at least 

 two points at which a sudden reheating takes place ; but not 

 necessarily two sensible reglows*, or even, as we have seen, 

 two sensible reheatings. This is all the more to be expected ; 

 because when, on heating, a wire under moderate stress has 

 attained a temperature of about 550° C, there is evidently a 

 sudden permanent yielding showing sudden softening of the 

 metal, though this is not nearly so marked as the permanent 

 yielding which takes place at the higher critical temperature. 

 The next experiment appeared to show, not one recalescence, 

 but several. 



Experiment II. — The end of a steel poker was heated to a 

 very bright red in a fire and was then taken to a corner of 

 the room on one side of the fire. The room was quite dark 

 save where it was faintly lighted up by the fire, and the eyes 

 of the observer had been previously rested by shutting them 

 for five minutes before the experiment. The heated part of 

 the poker was intently watched whilst it was cooling and 

 nothing was observed of any note for some little time. Pre- 

 sently, however, the external surface appeared to lose in 

 temperature rapidly, and shortly afterwards to brighten most 

 perceptibly. The apparent rapid darkening and reglow 

 occurred no less than seven times during this same cooling. 

 The experiment was repeated again and again, not only on 

 the same night but on several nights, and as often as it was 

 repeated seemed to show most conclusively that there were 

 several reglows. It was noticed, however, that when the 

 poker was loosely held in the hand the apparent reglows 

 occurred more frequently than when it was fixed ; and further 

 that, when the outside of the poker was seen to darken, a 

 slight motion of the poker in any direction caused it to 

 brighten. Evidently convection-currents of air could not 

 account for the phenomenon ; for though a current of air 

 might cause a sensible darkening of the external surface by 

 its cooling effect, it could not cause the brightening which 

 seemed to result from the motion. As it was thought that 

 the brightening might be produced by the motion of the iron 

 through the earth's magnetic field, the poker was reheated 

 and fixed in a horizontal position above the poles of a rather 

 powerful electromagnet. As soon as one of the darkenings 

 preliminary to a reglow occurred, the circuit of the battery 

 which actuated the electromagnet was closed and then opened 

 again ; but there was no sensible brightening as a result of 

 these operations, either at any point of this particular trial or 



* The reheating may take place at too low a temperature to produce a 

 reglow. 



