Magnetic Hysteresis with Frequency. 105 



Ring I. Ring II. 



No. of laminae 18 31 



Internal diameter (r x ) 7*6 cm. 15-223 



External diameter (>\,) 11-58 cm. 16-477 



Mean thickness (*) . . . . '0475 cm. *0443 



Section of magnetic circuit (a) 1*701 cm.- -861 



Length of magnetic circuit (1) 30*12 cm. 49'79 



Specific resistance of iron at 12° 0. (p) .... 12590 13600 



No. of primary turns (n x ) 164 406 



No. of secondary turns (?i 2 ) o or 10 10 



The laminae were well annealed, the oxide removed, and 

 they were insulated from each other by oiled paper. The 

 mean thickness was determined from weight, area, and 

 specific gravity. The values obtained for the specific re- 

 sistance seemed high, so the determination was checked. 



4. The method by which the current and flux waves were 

 determined was practically the same as that described in the 

 paper already quoted, except that readings of the galvano- 

 meter which give the ordinates for a definite phase were 

 only taken for every six degrees on the divided circle which 

 carries the movable brushes of the commutator. This gave 

 15 ordinates per half wave, and from these, without plotting, 

 the first, third, and fifth harmonics were easily obtained 

 by an arithmetical method of analysis when the seventh and 

 higher harmonics could be neglected. Whether or not the 

 latter assumption was legitimate would appear during the 

 analysis, and in all cases in which the amplitude of the 

 seventh harmonic was greater than one per cent, of that of 

 the first, it was determined, and though its values are not 

 given in the tables that are to follow, its effect was allowed 

 for. 



5. In order to explain the procedure by which the results 

 arrived at were deduced, the galvanometer readings for a 

 pair of a>>ociated waves will be given, and the treatment to 

 which these readings were subjected will be indicated. 



The following table give- four times the galvanometer 

 deflexions for a pair of associated current and flux waves, 

 with the corresponding divided-circle readings. The latter, 

 when doubled, give the corresponding phase-angles, as one 

 complete revolution of the divided circle corresponds To two 

 full waves, the rotary converter and the commutator both 

 being four-pole. The readings Bor the current-wave are 

 represented by y. and those lor the flux-wave by /3. The 

 other necessary details of the experiment are also given. 



