592 Lord Rayleigh on the Induction- Coil. 



14 mm., and it made no perceptible difference whether or not 

 the falling weight was employed. Considering that the 

 velocity of the weight at impact must have been about 30 feet 

 per second and that its mass was large compared with that of 

 the spring, these results were far from promising. With a 

 stronger primary current the advantage gained from the con- 

 denser was much greater, and the utility of the quicker break, 

 with or without condenser, seemed to be nil. 



But, in spite of the failure of the quick break, one or two 

 observations presented, themselves which seemed worthy of 

 being followed up. It was noticed that, with one Grove cell 

 in the primary, the spark, although very inferior when no 

 condenser at all was employed, was improved when the usual 

 condenser (of large capacity) was replaced by a single sheet 

 of coated glass (Franklin's pane). And, what was perhaps 

 more instructive still, when the already weak primary current 

 was further reduced by the insertion of one or two ohms extra 

 resistance, the spark-length (now very small) was less with 

 than without the usual coil condenser. This observation was 

 repeated, with like result, upon another coil (by Apps) and 

 its associated condenser. At any rate in the case of very weak 

 primary currents, the usual condenser did harm rather than: 

 good. 



The view, suggested by the foregoing results, that while 

 the ordinary break was quick enough in the case of weak 

 currents to allow a condenser to be dispensed with, the 

 superior arcing power of strong currents demanded a much 

 more rapid break, encouraged further efforts. An attempt to 

 secure suddenness by forcibly breaking with a jerk a length 

 of rather thin copper wire, forming part of the primary 

 circuit, failed entirely, as did also, perhaps for want of suffi- 

 ciently powerful appliances, an attempt to blow up a portion 

 of the primary circuit by electric discharge. Another method,, 

 however, at once allowed an advance to be secured. This 

 consisted in cutting the primary circuit by a pistol-bullet ;. 

 and it was found that this form of break without condenser 

 was about as efficient as the usual platinum break w T ith con- 

 denser, although the primary current was increased to that 

 supplied by three or four Grove cells and the spark-length to 

 40 mm., that is, under about the ordinary conditions of 

 working. 



A further improvement was effected by cutting away about 

 half of the bullet with the intention of raising its velocity. The 

 following results were recorded with an Apps' coil excited by 

 three Grove cells. The spark-gap being 50 mm., the usual 

 platinum break' and condenser were not able to send a 

 spark across. Even with the somewhat more efficient break: 



