of Induced Electric Sparks. 193 



obtained had never exceeded the quarter of a millimetre in 

 length ? and were quite incapable of producing a luminous 

 discharge through a small vacuum-tube. 



The following experiments led to the method desired : — 

 A rod of soft iron 20 centims. in length was wrapped in 

 folds of vulcanized sheet rubber to a thickness of 2 centims. 

 Around the exterior of this 150 centims. of stoutly insulated 

 copper wire were wound. One end of this wire was connected 

 with one terminal of the secondary wire of a small induction- 

 coil *, the other end was brought within 15 millims. of the other 

 terminal. The intent of this arrangement was to pass, as in 

 the previous experiments, an intermittent current around a 

 metallic conductor. When the coil was set in action the 

 secondary currents passed with rapid discharge across the 

 interval of about 15 millims. between the end of the wire and 

 the terminal ; and it was then found that sparks of several 

 millimetres length could be drawn from the iron rod within 

 the folds of vulcanized rubber by an earth-wire, or by the 

 ringer, or by any conductor held in the hand. It was found 

 impossible to convey any charge from the iron rod by a proof- 

 plane to an electroscope ; and on leading a wire to a sensitive 

 long-coil galvanometer by Colladon no deflection was obtained, 

 though the sparks were drawn from the wire beyond the 

 galvanometer, through which the force thus manifesting itself 

 must therefore have passed. Identical effects followed the 

 substitution of bars of copper, cadmium, and zinc for the 

 iron rod. 



It was next ascertained that no perceptible difference of 

 potential existed between the two ends of the metal rod, and 

 none between the ends and the median portion. The spark 

 could be drawn indifferently at either end, or at the middle, 

 but not from two points at once ; nor could it be led by an 

 insulated wire from one point of the bar to another. Other 

 metallic objects in the neighbourhood of the bar became 

 similarly affected ; and from them sparks could be drawn. 

 These effects were augmented when the distance of 15 millims. 

 over which the secondary current of the coil discharged itself 

 as a spark was increased to 20 millims. They diminished 

 when the distance was reduced to 5 millims., and ceased 

 entirely when the interval was reduced to nothing, so that the 

 secondary current of the coil passed along unbroken metallic 

 connexions. The conclusion to which these results pointed 



* This coil, capable of throwing sparks of 50 millims. length with two 

 cells of Grove's element, was employed in most of the succeeding experi- 

 ments. Each piece of apparatus used, including the cells of the battery, 

 was carefully insulated from the table. 



Phil. Unci. S. 5. Vol. 2. No. 10. Sept. 1876. O 



