490 Dr. Norman Campbell on the "Break " 



observations are directed is that the speed of the separation 

 of: the terminals is a relatively unimportant matter in the 

 prevention of sparking. The speed doubtless exerts some 

 influence, and a reduction of the speed much below that 

 usually employed would probably be harmful ; on the other 

 hand, an increase, if sufficiently large, would doubtless be 

 beneficial. But there is a wide range of speeds, including 

 those usually employed, over which the speed has practically 

 no effect, [f the use of a smaller primary capacity is to be 

 made possible, the improvement must be sought by reducing 

 the heating at, the contact and preventing the development 

 locally of high temperatures which permit the formation of 

 ;m arc. 



The exact conditions which tend to prevent arcing have 

 not been studied in detail : it would be interesting, for 

 example, to inquire whether the use of material of high 

 thermal conductivity for the electrodes would be beneficial. 

 But in the magneto the characteristics of the break are fixed 

 rather by mechanical than electrical considerations, and it 

 is likely that, even if alterations which would prevent arcing- 

 could be suggested, they would be found impracticable. 

 However, it is important to realize that, if arcing can be 

 prevented, the break, which will be sparkless even when 

 there is no primary condenser, is attainable. It is often 

 thought that, unless enormous velocities of separation of the 

 contacts are available, it is quite impossible to break so 

 highly inductive a circuit as the primary of a magneto or 

 induction-coil without the occurrence of a spark at the 

 break in the absence of a condenser of considerable capacity 

 connected across the terminals. Such an impression is 

 quite erroneous. It is quite easy to construct a break 

 which will carry currents up to 1 amp. without any con- 

 siderable heating of the electrodes at all ; such a break will 

 be sparkless if used for such currents in the primary of any 

 magneto or induction-coil of ordinary construction, even if 

 there is no primary condenser and the speed of separation 

 of the contacts is as low as 20 cm. per sec. 



But there is one point which needs further attention. 

 Is the absence of a visible " flash " at the break evidence 

 that the break is really perfect, and that the interruption 

 of the current occurs instantaneously and without dissipation 

 of energy ? 



A perfectly definite answer could be given to this question 

 only if it could be shown that the oscillations excited by the 

 break of the primary current agreed with those calculated 



