of a Magneto or Induction- Coil. 485 



thought worth while to state it, is readily found if the 

 influence of the nature of- the contacts at which the primary 

 current is broken is examined. According to the theory we 

 have been examining, the nature of the contacts should be 

 almost without influence ; for the spark potential is almost 

 independent of the form and material of the terminals when 

 the distance separating them is very small. But the crudest 

 experiments will indicate that the nature of the contacts is a 

 most important feature in determining the limiting current. 

 In one matter indeed the influence is familiar to all who 

 have used a magneto or induction-coil : we have onlv to 

 roughen the surface of the contacts and sparking immediately 

 occurs at the break which was sparkless before; the limiting 

 current may be reduced 50 per cent, by scratching the 

 surface and restored to its original value by polishing. 



More systematic experiments indicated that, if the 

 constants of the circuits remained unchanged, the limiting 

 current which could be broken without sparking always 

 increased as the resistance of the contacts decreased. Some 

 observations on the resistance of the contacts are described 

 in a note at the end of this paper, but in general it may be 

 said that the resistance of a contact can be decreased either 

 by increasing the area of the surfaces in contact or the 

 pressure between them. Thus in a certain circuit the contact 

 fitted in a magneto would break without sparking a current 

 of 4 amp. ; a contact consisting of a ball £ in. in diameter 

 pressed with a force of 50 grams weight against a steel 

 plate would break only 2*2 amp. ; if force was reduced to 

 1 gm. weight, the limiting current was barely -§- amp. In 

 no case could any relation be found between the limiting 

 current and the speed of break, which in these observations 

 could be varied from 300 to 62 cm./sec. The influence of 

 the material of which the contact was composed was not 

 investigated. 



(4) Formation of an arc. — The great variation of the 

 limiting current with the resistance of the contact gives at 

 once the clue to discrepancies which have been noted. The 

 "" spark " which is seen to occur at the broken contact is not 

 a true spark, but an arc*. The discharge only occurs if 

 the passage of the current between the contacts before the 



* This conclusion might almost have been deduced from the mere 

 appearance of the u spark." So far as we have observed, a luminosity 

 confined to a narrow path between the contacts more or less straight, is 

 never seen; what appears when a primary circuit is broken is a "flash,*' 

 such as occurs when steel strikes flint. It may be mentioned that 

 several writers have called the spark at break an " arc," but they do not 

 seem to have realized the implications of that term. 



