280 Prof. J. A. Ewing and Mr. W. Low on the 



lies decidedly below curve 1 ; and curve 3, which is shown 

 only by the points marked thus O, also below curve 1 ; curve 4 



Fig. 3. 



(1) Solid bar, no load. 



(2) Solid bar, loaded. 



(4) Cut bar ; ends rough ; loaded. 



(5) Cut bar ; end3 faced true ; no load. 

 X Cut bar ; ends faced true (6). 



O Cut bar ; ends rough ; no load (3). 



15000 































14000 

















13000 



12000 



-f 11000 



^ 10000 



o 



£Q 9000 



c 

 ■S 8000 



V 













































IHHHMPSfl 











































Magnetic In 



© © « 



© © c 













































4000 

















3000 

















2000 

 1000 































10 12 14 .6 18 20 22 24 26 



Magnetizing Force, § (C.G.S.). 



28 30 3* 34 36 



lies below curve 3 and below curve 2. But it is remarkable 

 that curve 5 is nearly coincident with curve 3 ; and curve 6, 

 which is shown only by the points marked thus x , is almost 

 absolutely coincident with curve 2. In other words, when 

 the cut surfaces are faced to true planes, the cut bar does not 

 behave very differently as regards magnetic permeability from 

 the same bar with its cut surface only roughly shaped so long 

 as it is not compressed by an external load. It is only when 

 33 is raised to high values that there is a marked difference ; 

 then the faced ends seem brought into better contact, and the 

 thickness of the apparent air-space becomes decidedly less 

 than in the case of the rough-cut bar. But when a sufficient 

 load is applied, the cut bar with its surfaces faced true behaves 



