28 Mr. Donovan on Galvanometric Defkctmis 



of the same class as those called voltaic? or, in other words, is electricity the 

 agent in all these phenomena? 



In the "Philosophical Magazine" for 1852 I published an essay, in which 

 were adduced reasons for believing that electricity is not a simple elementary 

 fluid, but a compound of several elementary constituents, one of which is the 

 deflecting agent. As soon as I became acquainted with the fact that brisk de- 

 flections are producible by attrition of heterogeneous metals, it struck me that, 

 perhaps, the power thus developed might be the deflecting agent, either in a 

 separate form, or at least in a less complicated state than it usually occurs. 

 In order to study this question it was necessary to construct an instrument by 

 which the attrition of metals could be carried on energetically, and for a great 

 length of time, without much labour. I therefore caused the following appa- 

 ratus to be made ; it answered also for some of the experiments already 

 described. 



A wooden wheel, four feet in diameter, with a winch for turning it, was 

 adjusted in a heavy, solid frame. The wheel carried a band which passed 

 round a pulley 2-33 inches in diameter, and so mounted in an iron frame that 

 by one revolution of the wheel the pulley revolved twenty times. As the 

 wlieel could be made to revolve once in a second, the pulley would revolve 

 twenty times in that period. The axle of the pulley was adjustable to various 

 thick, circular plates of wood, the peripheries of which were each shod with a 

 large, heavy ring of a different metal, which had been turned in a lathe. A 

 socket was placed in such a situation relatively to these rings, that a cylinder 

 of a metal, always Of a different kind from that on the circular plate of wood, 

 could by the pressure of a spring be made to rub against the metallic ring as 

 it revolved. Thus, attrition between any two metals could be efiected with the 

 greatest rapidity for any required period of time. The socket was lined with 

 ivory. On the frame of the machine were affixed two binding screws, one of 

 which communicated with the revolving ring of metal, the other with the rub- 

 bing metal fixed in its socket. 



My first trial was made on iodide of potassium dissolved in water, with a 

 little starch also in the solution. The attrition was between bismuth and 

 antimony. With this liquid a small tube Avas filled, into each end of which 



o 



