540 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



current be sent through the coil again, the retardation of the crest of the 

 secondary will be perhaps twice or thrice as great if the direction of 



the new current be opposed to 

 that of the next preceding one, 

 as it would be if the new current 

 had the same direction as its 

 predecessor. 



In Figure 23 the abscissas 

 represent currents in amperes 

 and the ordinates the lag, in tens 

 of seconds, of the crest of the 

 secondary current behind the 

 closing of the primary circuit 

 when the voltage was about 84 

 and when the current considered 

 had the opposite direction to the 

 one before it. The gap was in 

 this case closed. For a current 

 of 1 ampere, as the diagram 

 shows, the lag is about 40 sec- 

 onds, while for a current of 5 

 amperes it is only 10 seconds. 

 In Figure 24 the ordinates rep- 

 resent the relative heights of the 

 crests of the secondary currents 

 corresponding to primary cur- 

 rents the intensities of which, in 

 amperes, are represented by the 

 abscissas. 



Figure 24. 



In the next portion of this paper I hope to discuss some of the results 

 given here, with a large number of others which throw some light upon 

 the march of the eddy currents in a massive iron core of the kind here 

 used. In this connection the records of the secondary coils wound upon 

 the separate members of a set of loosely-fitting, coaxial, cylindrical shells 

 of soft steel, made part of the core of the magnet represented in Figure 1, 

 will be interesting. 



Jefferson Physical Laboratory, 

 Harvard University. 



