360 Dr. L. T. More on the Changes in Length 



Possible Errors. 



It has been shown that the change in elasticity occurring 

 when an iron wire is magnetized will modify the relation 

 between the intensity of magnetization and the elongation, 

 and it has been assumed that, except for heavily loaded wires, 

 this effect will be very small. Bock* found as a result of his 

 experiments that magnetizing soft iron made it more incom- 

 pressible, and that the change in elasticity was less than one 

 half per cent. He could not find any relation between the 

 intensity of magnetization and change in elasticity, so that it 

 is impossible to give more than a probable maximum to the 

 correction it is necessary to make in the curves given in this 

 investigation. 



If we take B = 20,000, 



B 2 1 



•5— = — x 10 8 dynes per square centimetre, 



and 



a -r- x l ° 8 



^ X 10'=-^! TT _ = _73. 



A change of \ per cent, in y would give a correction of 



•005x73 = *4 of a unit. This correction may therefore be 

 neglected, as errors greater than this would occur in measur- 

 ing B. 



It was also assumed that when the wire was not stretched 

 except by the jacket, the correction might be neglected. 



The weight of this was 350 grams or 42,000 dynes per square 

 centimetre, then 



«xl0'=**™xl0'-*X>. 



A change of \ per cent, in 200 is 1*0. Consequently the 

 curves marked no load are not significantly changed by 

 making this correction. 



But when we consider the case of the wire loaded with 

 750 kilograms per square centimetre, we get, by similar 

 calculation, a correction of 15 units ; and for the wire under 

 the greatest tension, 1750 kilograms per square centimetre, a 

 correction of 40 units. Since the magnetism makes the metal 

 more incompressible, this correction enters as a contraction 



* Bock, he. tit. 



