the High- Tension Magneto. 383 



should vary with c in the manner shown in fig 11. The 

 experimental curves show no such relation ; the maximum 

 peak potential decreases with the number of turns coupled 

 with the primary, even if the coupling is unaltered, and 



Fig. 11. 



o-o 



0-2 c 0-3 



it decreases steadily as the coupling is decreased by the 

 addition of uncoupled secondary turns. The best result 

 is obtained by using as many secondary turns as possible 

 and coupling them as closely as possible with the primary. 



In all particulars, therefore, the theory proves false, and 

 it provides no guidance whatever to the design of the 

 armature. The methods which have been adopted in 

 practice as a result of mere empiricism seem the most 

 satisfactory, and it seems unlikely that any considerable 

 improvement in the efficiency of the magneto can be pro- 

 duced except by reducing the iron losses in the core. 

 Until these losses are made as small as they are in the 

 induction-coil, an application of the theory which neglects 

 those losses is useless. 



But before passing on it is interesting to notice one feature 

 in which the experiments just described agree with those of 

 the previous part. The maximum peak potential calculated 

 by the theory (neglecting damping) for the combination 

 10,000/0 is 4750 volts ; if this is corrected in the manner 

 explained in § 8 for the observed damping, it becomes 

 3090 volts. The observed maximum peak potential is 

 2560 volts, and the ratio of observed to calculated is 0*83. 

 It is worthy of remark that this ratio is not very different 

 from that found for the totally different armatures examined 

 previously, namely 0'7() and 0*78. It is possible that the 

 constancy of this ratio has some theoretical significance. 



Effect of an Air-core Coil in tlie Primary. 



18. In his paper, on which the present work is based, 

 Prof. Jones described measurements of the peak potential 



