580 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



of the past experiences of the iron, the form of the hysteresis diagram 

 is precisely the same, whether the half cycle be carried out by one 

 reversal of the main switch or in a very large number of steps. In 

 general agreement with these results are some less accurate ones which 

 I obtained three years ago in experimenting upon a transformer which 

 has an exciting coil of 1394 turns and a core of about 120 square 

 centimeters in cross-section, built up of thin strips of varnished sheet 



Figure 36. Growth from an originally neutral core of a current in a trans- 

 former wdth a laminated core. The effects of eddy currents are here noticeable. 



iron about ten centimeters wide. This transformer was connected in 

 simple circuit with a storage battery and a rheostat besides a suitable 

 oscillograph. When the circuit was suddenly closed, with such a 

 resistance (jr) in the rheostat that the final strength of the current was 

 about 1.10 amperes, the current curve was of the form R as shown in 

 Figure 35, and when after a few seconds a' was suddenly removed, so 

 as to bring the final strength of the current up to about 2.30 amperes, 

 the current curve was Q. When the whole journey was made without 

 a: the current curve was T. The sum of the flux changes represented 

 by the shaded areas as measured by a Coradi " Grand Planimtstre Rou- 

 lant et h Sphere" was 1126, while the flux change corresponding to 

 the area above the curve T was 1130. The core w'as not sufticiently 

 well divided to avoid all evidence of eddy currents, for the curve Q 

 does not exactly conform throughout with the upper part of T. This 

 is shown more clearly in Figure 36, taken with the same transformer. 

 Here the area of the shaded portion above K multiplied by the resis- 

 tance then in the circuit should be equal to so much of the area above 

 C, multiplied by the resistance belonging to it, as lies to the left of the 



