through Rarefied Air, 317 



Current froni the surface to the rod : heat 100. 



Current from the rod to the surface : heat 122. 

 Lastly, the discharge circuit was elongated by a platinum 

 wire, 16 feet long and 0*057 line thick; the knob was again 

 placed in the cylinder at a distance of |^ths of an inch from the 

 platinum surface, and the charge of the battery was inci'cased to 

 a quantity of electricity equal to 16. The mean results were : — 



Current from the surface to the knob : heat 100. 

 Current from the knob to the surface: heat 105. 



In all the numerous experiments which were instituted, and 

 in which the discharge current of the battery passed through air 

 rarefied to 1 or 2 lines of pressure between a small and a com- 

 paratively great, but arbitrarily- shaped metallic surface, a differ- 

 ence was observed in the heating effect upon the discharging 

 wire according to the direction of the current ; this difference, 

 however, although sometimes greater and sometimes less, was 

 always in favour of one and the same direction. From the above 

 experiments, it will be seen that the greatest heat was produced 

 when the direction of the current was the same as that which, in 

 the first-mentioned experiments with a magneto- induction cur- 

 rent, produced a very small, or no deflection of the magnetic 

 needle. When the discharge current of a Leyden battery passes 

 through very rarefied air between a very small and a comparatively 

 large metallic surface, the heat produced in the remaining part of 

 the discharge circuit is greatest when the current passes f)'om the 

 large to the small surface. The meaning of a very small surface 

 will be hereafter explained. 



This remarkable and hitherto unknown change in the action 

 of an electric discharge is easily explained by known facts. I 

 satisfied myself that the battery was equally well discharged 

 whatever might have been the direction of the current. As the 

 discharge circuit had the same constitution in both cases, the 

 difference in the heating power of the current could only be 

 occasioned by the method of discharge. It is well known that 

 a change in the heating effect, occasioned by a change in the 

 method of dischai'gc, can be shown in the clearest possible man- 

 ner. When the discharge circuit is interrupted by a thin stratum 

 of distilled water placed between metallic plates, and when the 

 strongest charge of the battery which will pass noiselessly through 

 the water i.s employed, a delicate electric thermometer is neces- 

 sary in c)rder to detect the small amount of heat which is thereby 

 produced in the discharge wire. If the plates immersed in water, 

 however, are brought inappreciably nearer to one another, the 

 discharge passes through with a spark, and at the same time the 

 heating effect iu the discharge wire is so great that the thcrmo- 



