Stress or Permanent Strain to a Bright-red Heat. 261 



On heating, the untwisting always showed itself 

 at the temperature of bright red. 



In consequence of the sudden change on cooling not occur- 

 ring till a dull red heat had been reached in his experiments, 

 Professor Barrett connected the phenomenon in question with 

 another of equal interest. It is well known that, at a tempe- 

 rature of dull red, iron begins very rapidly to lose its magnetic 

 properties ; and, according to Professor Barrett, the two phe- 

 nomena occur simultaneously. It is no doubt the case that, 

 on cooling, the sudden jerk occurs simultaneously with the 

 sudden regaining of magnetic properties in some specimens! of 

 iron or steel ; but it by no means follows that we have not 

 two very distinct critical temperatures — one at or about a 

 dull red, at which iron loses or regains its magnetic pro- 

 perties according as the wire is being heated or cooled; 

 and another at a much higher temperature, namely near a 

 bright red, at which sudden changes, certainly not less pro- 

 found, occur. This is shown in the next experiment, which at 

 the same time illustrates the fact already mentioned, that the 

 effect of permanent strain is opposite to that of temporary 

 stress. 



Experiment IV. — The glass tube (fig. 1) was placed inside 

 a magnetizing solenoid J, consisting of a single layer of cotton- 

 covered copper wire ^ nicn diameter, and connected through 

 a key with a battery of two Grove's cells. Surrounding the 

 solenoid and concentric with it is a secondary coil, consisting 

 of 840 turns of cotton-covered copper wire partly ^ inch 

 diameter and partly ^ inch diameter. This coil has a re- 

 sistance of about 1 ohm, and is connected through a key 

 with a very delicate Thomson's reflecting-galvanometer of 

 about 7 ohms' resistance. A piece of the same soft iron wire 

 already mentioned was subjected to thirty complete revolu- 

 tions of permanent torsion, and was finally released from all 

 torsional stress. When the wire, on heating, reached the 

 temperature of dull red, a momentary deflection of the needle 

 of the Thomson's galvanometer took place, indicating a sudden 

 loss of magnetic permeability ; and as soon as a bright red 

 heat had been attained it, as in the -preliminary observations, 

 twisted sharply and suddenly. The battery which was em- 

 ployed to heat the wire had its circuit now broken, and the 



* This fact, which was also noticed by Mr. Gore and Prof. Barrett, is 

 significant. 



t The reason of this will be found in the remarks on the recalescence 

 of iron. 



X Only the ends of this solenoid can be seen in the figure, the rest 

 being enveloped by the secondary coil. 



