508 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



in the circuit was made to grow from zero by steps so far apart in 

 time that, in every interval, the current in the secondary circuit had 

 time to die sensibly out. After a maximum current of the desired inten- 

 sity had been attained, the current was then reduced by steps to zero, 

 and was then built up in the opposite direction to the same maximum 



Figure 3. 



value by steps. The areas under the oscillograph curves of the secondary 

 circuit evidently furnish means of obtaining a hysteresis curve for the 

 core of the magnet, but this well-known method of procedure is in my 

 hands and for this particular magnet not quite so satisfactory as the 

 one indicated above. 



The leakage in a magnet of the form of this one is of course considerable, 



