ON THE INDUCTION-COIL. 



207 



auxiliary of the utmost importance,, mcreasing the sj)ark-length 5 or 

 10 times, even when the break is made at pièces of platinum. And, 

 although it has been customary to say, no donbt correctly, that the 

 condenser acts by absorbing into itself the primary spark, and so in- 

 creasing the snddenness of break, it is usnal to attribnte to it a further 

 virtue, and not unnaturally when it is remembered that the effect may 

 be not merely to stop, bnt actnally to reverse, the primary current. If, 

 however, the theory of the foregoing pages is correct, we shall be con- 

 strained to take a différent view. 



The action of the condenser, and especially the most advantageons 

 capacity, has been stndied experimentally by Walteu and by Mizuno. 

 That there mnst be a most advantageous capacity is évident before- 

 hand, inasmuch as a very small capacity is continuons with no con- 

 denser at ail, and a very large capacity is continuons with an uninter- 

 rupted flow of the primary current. It is more instructive that the 

 former observer found the most advantageous capacity to vary with 

 the manner of break (whether in air or under oil), and that the latter 

 found a dependence upon the strength of the primary current, a larger 

 current demanding a larger condenser. 



When a condenser is employed it is important that it be connected 

 as directly as possible with the points between wich the break is made 

 to occur. A comparatively small electromagnet, included between one 

 of the break-points and the associated condenser- terminal, suffices to 

 diminish, or even to annul, the advantage which the use of the con- 

 denser otherwise présents *). The explanation is, of course, that the 

 current in an electromagnet so situated tends to flow on across the 

 break- gap, and so to establish an arc, with a force which the condenser 

 is powerless to relie ve. 



Eeturning to the theoretical aspect of the question, and inquiring 

 whether there is any reason for expectiug a condenser to give an ad- 

 vantage as compared with an absolutely sudden cessation of the primary 

 current, it is difficult to see ground for other than a négative answer. 

 In the case of no magnetic leakage, somewhat closely approached, one 

 would suppose, in practice, an instantaneous abolition of the primary 

 current throws the wliole available energy into the secondary circuit, 



*) Phil. Mag. Vol. 2, p. 282, 1901. 



