Transverse Fields and Allied Phenomena. 



413 



stray field outside of the gap, the galvanometer for a loop 

 within the gap will respond to B, — (B 2 —jB 1 ). Two measure- 

 ments were therefore made to compute B r ■ -•"'.; i &\ { H 

 To prevent excessive heating currents between 5 and 6 

 amperes were usually employed, though these were not suffi- 

 cient to saturate the core of the helix as the above data and 

 the results below show. I may add that the ordinary effects 

 of viscosity and slip are eliminated by the method of experi- 

 ment, since the field was alternately made and broken several 

 times. 



m4M 



The change 8n of rigidity n, is at once given in terms of 

 the deflections 80 of the mirror; for 8n/n = 80/0, very nearly, 

 if for the same section and length, is the twist simultane- 

 ously imparted to each wire. In fact -the above method was 

 devised to secure this convenience. If 80/0 is positive, i. e. if 

 in the above apparatus the twist is imparted clockwise looking 

 down the wire and the deflection due to magnetization is also 

 clockwise from the same point of view, then the magnetic 

 change of rigidity is an increment for upper or magnetized 

 wire ; and vice versa. 



4. /Results. — The chief difficulty encountered has already 

 been suggested. If the wire is carefully straightened and 

 adjusted in the middle line of the vertical air gap, the wire 

 nevertheless becomes slightly sinuous or wavy when the mag- 

 netic field is excited. Different parts of the length of the 

 wire are unequally attracted. If these parts lie across the field 

 such attraction is accompanied by rotation and the observed 



