16 Prof. E. Bdlund on the Electrical 



of electricity. It seems to me that, in drawing from the known 

 impossibility of an electric current traversing the most perfect 

 vacuum between two electrodes the conclusion that a vacuum 

 is absolutely nonconducting, the same mistake has been made 

 as when, from the circumstance that the sun rises in the east 

 and sets in the west, it was believed that one might infer that 

 the sun in reality goes round the earth. The first of these 

 conclusions does not appear more certain than the second. 



I will now cite some observations furnishing a new support 

 for the correctness of what has just beeu said. 



§5. 



Hittorf made a great number of very instructive observa- 

 tions on the conductivity of gases*. In the first series of his 

 researches he employed, as electric source, a powerful Ruhm- 

 korfF induction-apparatus ; the gas, contained in Geissler 

 tubes of different shapes and sizes, was rarefied by means of a 

 mercury air-pump. The electrodes were mostly composed of 

 aluminium, but sometimes of other metals. The current was 

 measured by a mirror-galvanometer. Of the results obtained 

 we shall mention only those which have a connexion with our 

 subject. 



Two cylindrical tubes, equal in width and length, were fur- 

 nished at their extremities with electrodes of platinum wire 

 of 0*07 millim. thickness. The electrodes were in part encom- 

 passed with tubes of glass. The two positive electrodes had 

 each an equal portion of the wire without a glass covering ; 

 but one of the negative electrodes presented a free surface 

 3'87 times as large as that of the other. The distance between 

 the extremities of the electrodes was the same in both tubes. 

 The tubes therefore presented perfect similarity, except that 

 the negative electrode of one had a larger free surface. When 

 they had both been filled with hydrogen and the pressure of 

 the gas was lowered to 1*35 millim. of mercury, they were 

 inserted side by side in the circuit of the induction-apparatus, 

 the current of which traversed them simultaneously ; but it 

 appeared that the current of the tube whose negative electrode 

 had the largest free surface was, on an average, 4*25 times as 

 intense as the current of the other. The intensity of the cur- 

 rent was therefore not far from proportional to the free surface 

 of the negative electrode. The same result was obtained with 

 carbonic acid and with nitrogen, as also with aluminium elec- 

 trodes. Therefore, all other circumstances remaining equal, 



* Pogg, Ann. cxxxvi. pp. 1, 197 (1868), Jubeib. p. 430 (1874) : Wied. 

 Ann. vii. p. 553 (1870). 



