the High- Tension Magneto. 393 



with the total field, and therefore with the primary current ; 

 whereas observation shows that the peak potential developed 

 is very closely proportional to the current broken. If eddy- 

 currents are the main source of loss, finer laminations might 

 produce the desired result ; but it is impossible to say at 

 present whether any lamination which is mechanically possible 

 would produce a marked effect. 



In the present state of the investigation hardly more light 

 is thrown on matters of theoretical interest than on those of 

 practical importance. But it may be urged that several 

 of the relations found are extremely suggestive, and may 

 help to indicate the path to a more complete theory. It is 

 perfectly clear that the effect of the interaction between the 

 circuits and the iron is to change profoundly the relation 

 between the two components of the oscillation which are 

 predicted by the simple theory; it is possible that other 

 components are introduced. If the effect of the iron core is 

 merely to alter the relation between the amplitudes, the 

 phase angles, and the damping coefficients of the two com- 

 ponents, it seems not wholly impossible that a theory might 

 be produced which would be as readily applicable to numeri- 

 cal computation and practical design, as is the simpler theory 

 which Prof. Jones has applied with such success to the 

 induction-coil. 



Further progress might doubtless be made on the experi- 

 mental side by determining more carefully the exact 

 wave-forms of the oscillations, by analysing them into their 

 harmonic components, and inquiring in what manner the 

 relations between their amplitudes, phase angles, and damping- 

 differ from those predicted by the theory. Experiments in 

 this direction had been begun when the time came to put an 

 end to the investigation. They were by no means complete, 

 and few conclusions could be drawn. So far as could be 

 discovered, there was no sign of more than two different 

 periods concerned in the oscillations, but whether the two 

 periods which could always be detected were those of two 

 harmonic components had not been fully ascertained. 



Summary. 



11. The paper is a continuation of the one in the March 

 number. The object of the experiments is to discover how 

 far the relation between the peak potential, the primary 

 capacity, and the coupling of the circuits which is predicted 

 by the theory is found in experiment. 



12. The machine investigated is described. 



