a Non-Sparking Key. 313 



When a t and a't are small, as in German silver, the per- 

 centage increase of resistance per degree equals the percentage 

 decrement in conductivity per degree ; but if a t is large, as 

 in copper, this is not the case ; so that while '388 is the per- 

 centage increase of resistance of copper per degree, '293 is the 

 percentage decrement of conductivity per degree. 



Non-sparking Key. — Various so-called non-sparking keys 

 have been devised during the last year or two ; but, as far as 

 we are aware, they by no means completely fulfil their object. 

 On breaking a circuit through which a currrent is flowing, a 

 spark is produced the brilliancy of which depends on the 

 coefficient of self-induction of the circuit, and on the current 

 flowing through the circuit just before the circuit was broken. 

 To completely avoid this spark, the energy stored up in the 

 circuit must be utilized to send a current through an unbroken 

 circuit, and this can be accomplished if the resistance of the 

 circuit, instead of being suddenly varied from a small value 

 to infinity, is increased steadily by the insertion of a resist- 

 ance increasing at a proper rate. If it increases too quickly, 

 a spark will certainly take place ; if too slowly, the wire of 

 the resistance-coil which has been added to the circuit will be 

 dangerously heated. 



Fig. 7 (Plate XII.) shows an arrangement which we have 

 devised for introducing this resistance at a proper rate, inde- 

 pendently of any control on the part of the person stopping 

 the current. To start the current, the handle H is turned 

 moderately quickly until a solid piece of metal comes into con- 

 tact with the brush B, when a catch holds the key in that 

 position, and which is the position shown in the figure. The 

 current now passes from one of the binding-screws to the 

 other through practically no resistance. To stop the current 

 the little handle, h, is slightly pulled, when a spring, S, which 

 was coiled up when the handle was turned round to put the 

 current on, causes the cylinder to turn in the opposite direc- 

 tion, the various metallic pins on the right-hand side coming, 

 in succession, into contact with the brash B, until finally the 

 pins M made of ebonite, and not of brass, come into contact 

 with the brush B, and the circuit is entirely broken. Between 

 each of the metallic contact-pins there is a resistance varying 

 from a small fraction of an ^ohm for the first coil to some 

 hundreds of ohms for the last. The effect is, then, that resist- 

 ance at a perfectly definite rate is added to the circuit before 

 breaking it. 



These resistances are adjusted partly by calculation and 

 partly by trial ; so that with the particular rate of rotation 

 which the coiled spring S gives to the barrel of the key, the 



Phil. Mag. S. 5. Vol. 17. No. 106. April 1884. Y 



