of a Magneto or Induction- Coil. 48 ( .) 



it likely that the true limiting speed is not really much 

 greater. 



On the other hand, it should he noted that a sufficient 

 increase in the speed of break may prevent arcing in another 

 way ; the heating of the electrodes may not begin until the 

 separation begins, and a reduction of the time of separation 

 may reduce the time available for heating them to the 

 requisite temperature. A reduction of arcing due to this 

 cause is most likely to occur when, as in Lord Rayleigh's 

 experiments, the circuit is broken by catting a solid con- 

 ductor ; but it will also occur with n. contact such as thit 

 employed in a magneto, where the terminals are held 

 together by considerable pressure. 



It is probably for this reason that it has been found that, 

 if, by the use of special cams, the contacts of a magneto 

 are opened very slowly, flashing can be made to occur 

 when it will not occur with the normal rate of opening. In 

 the contacts used in these experiments, the pressure between 

 the contacts before opening was much less and the resistance 

 •of the contact much greater ; heating began before the 

 contacts separated. In such circumstances the limiting 

 current appeared quite independent of the speed of 

 separation within the range which could be observed (20 to 

 250 cm./sec). 



Again, it is possible that the increase in the secondary 

 potential which Lord Rayleigh found may be merely due 

 to the fact that the time that the arc lasted was greatly 

 reduced ; for if the interval between the first opening of the 

 contacts and the total cessation of the current could be made 

 very much less than the period of the oscillations excited, 

 the loss of efficiency due to "sparking" would be greatly 

 reduced. But this explanation is improbable. For, if it 

 were correct, there would be a gradual increase in the 

 efficiency as the speed of break was increased, whereas 

 Lord Rayleigh found no increase whatever until the speed 

 of a rifle bullet was reached. However, in the experiments 

 described here there was some evidence that an increase in 

 the speed of break, when there is sparking, produces some 

 slight increase in the secondary potential developed ; but 

 when there is sparking the secondary potential varies so 

 greatly at successive breaks that accurate measurements are 

 difficult. In all the cases examined the secondary potential 

 when there was sparking was at least 30 per cent, less than 

 that obtained with a slightly smaller current which produced 

 no sparking. 



(5) The Perfect Break. — The conclusions to which these 



