120 THE DISSIPATION OF ENERGY 



was also got on treating in the same manner a portion of wire 

 already heated past 375 C. 



EFFECT OF EXTENSIONAL STRAIN 



A length of the original wire was hung with a 14-lb. weight 

 attached in a long vertical shaft, and portions cut from it 

 were tested at intervals of days or weeks, but no change was 

 observed. The weight was then replaced by one of 28 Ibs., 

 and latterly of 56 Ibs., but still the same values of n were 

 observed. The wire after heating to 400 C., however, could 

 be easily elongated by a pull ; and on subsequent testing it 

 was found to give results tending to approach those obtained 

 with an unheated length of wire. A wire 12 inches in length 

 was raised to 400 C. and then stretched till rupture occurred. 

 The value of a was found to have changed to 100, whilst n 

 was found to be 1*34, the points again all falling into one 

 straight line. The diameter was now 0*90 mm., and the 

 change of length 3 inches. Thus stretching is followed by a 

 reversion to the original conditions of the unheated wire ; 

 and it is quite probable that, with more careful stretching, 

 the value of n could be reduced to its original amount. The 

 effects of repeated extensional and torsional strains on brass 

 are treated subsequently. 



It was found that a wire, of diameter 1*2 mms., gave results 

 similar to those got for that of diameter I/O mm. The exact 

 values got were 



a=95, n= -70, 6=439. 

 A thicker wire, of diameter 1'65 mms., gave the results 



a=85, rc=-66, 6=360. 



In this case the readings had to be taken very quickly, 

 since the oscillations died down with extreme rapidity, and 



