e— x. 



42 



e. 



96 



83 



169 



42 



116 



176 MM. Elster and Geitel on 



When the wire c was brought half a millimetre below the 

 wire a b, there was no constant orientation of the electrometer- 

 needle. The following are the maximum values obtained : — 



x+e. x — e. e. 



129 -9 138 



Series 2 (e>x). Conduction to earth in the point u. De- 

 pendence on the state of incandescence of the wire a b. 



Incandescence of the wire a b. e+x. 



Dull red .... 54 



Bright red .... 86 



White 74 



The last series of experiments show that, with the same 

 position of the wire c Q millim. above ah), the value of e is 

 lower at a white heat than at a bright red heat — a very sur- 

 prising circumstance, but confirmed by many control expe- 

 riments. 



If the wire c be covered with a coat of potassium chloride, 

 the electromotive force sinks considerablv; in one experiment 

 it fell from 188 to 34. 



The wire c was now replaced by a water electrode (fig. 3) 

 or a wet string. The experiments showed indubitably that 

 liquids also, separated from glowing platinum by a stratum 

 of hot air, become electrically excited ; only this excitation, 

 exactly as with the flame, is much less. The results were, 

 with 



c a platinum electrode 

 c a water electrode . 

 c a wet string . . . 



e+x. 



i — j. 



i. 



85 



84 



169 



26 



15 



41 



26 



■2± 



50 



Accordingly, from the fact that fame* show themselves di- 

 stinctly electric even when all contact with metals is avoided, it 

 mud not at once be concluded that they hove an electricity pecu- 

 liar to them. 



Let it be further remarked that, both when the wet string 

 and when the water electrode was employed, the amount of 

 the potential-difference between a b and the wire in contact 

 with the liquid of the electrodes, when directly connected by 

 water, was determined before the definitive experiments. The 

 deflection of the electrometer-needle amounted for it to only 

 a few divisions of the scale; so that no source of error could 

 spring from this. 



The phenomena here discussed, which had not, to our know- 

 ledge, been before observed, stand evidently in the closest con- 



