On the Magnetic Permeability of an Iron Bar. 275 



fi magnetic resistance " of the yoke may be approximately 

 allowed for by making a small addition to the actual length 



Fig. 1. 



W 



of the bar between the faces of the yoke : in what follows 

 this has been done by treating the bar as having a virtual 

 length of 13 centims. The bar was wound over the whole clear 

 length with a magnetizing solenoid through which currents 

 were sent from a battery of accumulators. A small induction- 

 coil less than 1 centim. long was wound at the middle of the 

 bar, and was connected to a ballistic galvanometer, which was 

 calibrated in the usual way by means of an earth induction- 

 coil. To measure the magnetization the magnetizing-current 

 was reversed several times, and readings w r ere taken of the 

 transient currents induced in the small coil by reversal of the 

 magnetizing force acting on the bar. In every case a series 

 of readings was taken, beginning with small magnetizing 

 currents and going on to stronger ones, and from these a 

 curve of 33 (the magnetic induction) and «£) (the magnetizing 

 force) was drawn. 



The first group of experiments show how transverse cuts 

 reduce the permeability of a wrought-iron bar, and how 

 longitudinal compression partially neutralizes the effect of 

 the cuts. 



The bar was turned to be of the same diameter throughout, 

 and to fit without jamming in the holes of the yoke. It was 

 first tested in the solid state, then when cut in the lathe into 

 two parts, then in four parts, lastly in eight parts ; in each 

 case two curves of 93 and Q were determined, one when the 

 bar was not loaded, the other when it was compressed with a 

 stress of 22$ kilogs. per square centim. The results are 

 shown by the curves of fig. 2, where the curves in full lines 

 refer to the experiments made without load, and the dotted 

 lines refer to those made when the bar was loaded. Com- 

 paring, in the first place, the no-load curves of fig. 2, we see 



T2 



