104 Prof. T. R. Lyle on the Variations of 



source of loss exists when iron is subjected to alternating 

 magnetizing forces has already been shown by Steinmetz and 

 by Siemens by other means. 



An interesting case of transformation and what is called 

 reflexion of energy is drawn attention to and discussed. 

 If the E.M.F. impressed on the magnetizing circuit on the 

 iron ring be sinusoidal, the flux-wave produced contains 

 third, fifth, &c. harmonics. These higher flux-harmonics 

 induce currents in the magnetizing circuit which are dissi- 

 pated as heat in it. Thus we have a transformation of electric 

 energy due to alternating currents of frequency n to energy 

 of currents whose frequencies are 3ra, 5n, &c, which is 

 reflected back into the magnetizing circuit. This the author 

 believes is a hitherto unnoticed source of transformer loss. 



2. Fig. 1 (PL II.) shows the arrangement of the apparatus. 

 The magnetizing current was obtained from a four-pole rotary 

 converter T supplied with direct current from storage-cells. 

 By means of rheostats placed both in the armature and field- 

 circuits, the speed could be varied and adjusted. One end of 

 the spindle of the commutator G was directly connected to the 

 spindle of this converter, and on the other end was a screw 

 thread which worked a tangent wheel. On this wheel an 

 ebonite stud is fixed which momentarily breaks the circuit of 

 an electric chronograph once every revolution, thus recording 

 the time of every 200 periods. 



The magnetizing current, drawn from the slip-rings of the 

 rotary converter, passed through a regulating resistance, a 

 Kelvin balance, the primary of the air-circuit transformer M, 

 and the primary coil on the iron ring B. 



One end of the secondary of M is joined to one end of 

 that of B, and the junction connected to one of the fixed 

 brushes of the commutator. The other ends of the secondaries 

 are connected, as shown in fig. 1, to the three-way key K, 

 from the lever of which connexion is made to the other fixed 

 brush. By this means either secondary can be joined to the 

 commutator, and thence from the movable brushes of the 

 latter through a reversing key and a high resistance to the 

 galvanometer. 



A Clark's cell with a megohm is arranged so that it may 

 be used at any time for the purpose of determining the 

 reducing factor of the galvanometer. The deflexions of the 

 latter were so nearly proportional to the currents producing 

 them over the part of the scale used that no calibration was 

 necessaiy. 



3. The details of the rings, called ring I. and ring II. 

 respectively, are as follows : — 



