536 



PROCEEDINGS OP THE AMERICAN ACADEMY. 



the first case it was necessary of course to add non-inductive resistance 

 to the circuit to reduce the current to this value. The time required for 

 the current to attain practically its final value is far less in the first case 

 than in the second. The dotted curves show the records of the oscillo- 

 graph in the secondary circuit on an arbitrary scale. The two curves 

 are quite unlike in shape, but the areas under them, as measured by a 

 planimeter, are almost exactly the same. The general forms of the dotted 

 curves here shown are like all the scores of others which I have obtained 

 under all sorts of conditions of current strength and resistance. 



Figure 20. 



The pole pieces of the magnet shown in Figure 1 are fastened to the 

 cylindrical arms by long bolts which extend outside the frame and carry 

 nuts which press upon the yokes and serve to keep the jaws apart. When 

 one of these bolts was removed it left a long axial hole of about an inch 

 in diameter through the arm, and into this I put a long rod of soft iron 

 upon which a layer of fiue insulated wire had been wound pretty uni- 

 formly, and this coil was connected with the secondary oscillograph 

 already mentioned. This coil and its long core filled the cavity fairly 

 completely except at the outside ends, but doubtless the joint at the inner 



