of Electricity through Rarefied Air. 1 3 



The deflections, varying only 2 per cent, between the pres- 

 sures 31*1 and 0*58, may be considered approximately con- 

 stant. If this invariable deflection be designated by k, and if 

 a is a constant, we can therefore put e+Tf—ok for the pres- 

 sure between these limits. If x be the invariable deflection 

 between the same limits of series 1, and if b is another con- 

 stant, we can write e^r^—bx, from which we get 

 ak-\-bx , 7 ak — bx 

 e= — ^ — and T ^~ — 2 



The observations show that x is rapidly augmented with the 

 rarefaction; the following results are obtained from these two 

 series : — If the density of the air at which the induced current 

 presents its maximum be taken for the starting-point, the elec- 

 tromotive force e increases with the ulterior increase of the rare- 

 faction, while, on the contrary, the electric resistance of the air 

 undergoes continual diminution during the same time. In the 

 rarefied-air space here employed the resistance of the air dimi- 

 nished by a quantity approximately equal to the increase of 

 the electromotive force when the rarefaction of the air was 

 augmented. 



If the distance between the electrodes be made greater than 

 it was in the preceding experiments, the diminution of r-J,, all 

 other conditions remaining equal, will, when the density is 

 reduced, evidently be greater than when I is less. It may 

 therefore easily happen, if I is sufficiently great, that rj, un- 

 dergoes, when the rarefaction is augmented, a diminution 

 exceeding the increase of e, and, consequently, that the sum 

 e + ril continues to diminish with the pressure. In that case 

 the maximum of the induced current, if a maximum be pro-- 

 duced, will be found at a pressure below that of the foregoing 

 observations. In Morren's experiments the distance between 

 the electrodes amounted to 240 millim. ; and it is very likely 

 that we must attribute to that cause the circumstance that he 

 found the maximum of the current at so low a pressure as 

 1 millim. or slightly under. 



For determining the variations of the sum e + r x l at different 

 air-pressures another process was employed, which, however, 

 was not capable of giving so exact a result as the experiments 

 with the induced current. A spark micrometer was used, 

 composed of two brass balls introduced into a wider glass 

 tube, one of which balls could be moved nearer to or further 



current similar to that of experiment 1. But it will without difficulty be 

 understood that in experiment 2 the disjunction-current produced in the 

 tube cannot be attributed to any oscillatory discharge ; and therefore there 

 is nothing to authorize us to see in a discharge of that nature the cause 

 of the great deflections given by experiment 1. 



