Light, Heatf Electricity and Magnetism. 175 



the one surface to the other, as evinced by a current passing 

 through the galvanometer from or towards the other extre- 

 mities of the metals employed. This, to my mind, would 

 evince the origin of radiant heat, the result of friction, in the 

 mechanical processes, drilling, turning, filing, &c., and which 

 on its discovery 1 named " mechanical electricity." 



In the printed report of the British Association for 18'l-5, 

 which met at Cambridge, I find that M. Paul Erman of Berlin 

 presented a paper containing one or two experiments of a 

 somewhat similar nature to the Association, but of which I 

 was not aware until the publication of the report, and the 

 completion of many of my experiments. 



April 2, 18^6, 1 made the following experiment: — 



Upon a mandril of copper a cylinder of bismuth was cast. 

 One end of the mandril was fixed in dry wood, and arranged 

 in a turning lathe ; the other revolved against the point of the 

 "following up head stock," as is usual: the surface of the 

 cylinder or disc was turned smooth, the mandril having been 

 previously soldered to the bismuth, so as to ensure lull metallic 

 conniiunication. 



Instead of a metal rest a wood one was now used, and after- 

 wards a small piece of wood placed under the ordinary rest; 

 to insulate this and the tool from the other portion of the 

 lathe, was found to be all that is necessary in these experi- 

 ments. A spring of brass wire was made to press firmly 

 against the turning mandril, so as to ensure metallic contact, 

 and its other extremity was in communication with the north- 

 ern extremity of the galvanometer*. 



In the following experiments, the direction of the current 

 is simply stated as seen at the galvanometer, which will be 

 found in all cases to be the reverse of what takes place be- 

 tween the opposing metals. Thus, in the experiments in which 

 the zinc robs the copper, as seen at the galvanometer, the cur- 

 rent is progressing towards the copper (see 31, &,c,) ; and yet 

 the actual transfer at the surfaces is from the copper to the 

 /inc. 



Exp. 21. On applying the su)Ooth surface of the end of a 

 piece oi thick rod copper to the turned surface of the cylinder, 

 producing friction, a current was observed from the copper 

 towards the bismuth. The rod copper was soldered to a wire 

 in coiniexion with the southern extremity of the galvanometer. 

 Exp. 22. By accident the rod copp<er was torn away, and 1 



• The galvanometer in tliese cxperitneiUs was not of tlie highest sensi- 

 liility. It consisted of forly-six turns of co|)per wire, the ^'^tli of an inch 

 in diameter. 'I'lie needle was single, and liad ihciefore a northern ten- 

 dency to coiinteniet. 



