236 
MESSRS. W. DE LA RUE AND H. W. MULLER ON THE 
He introduced a “ rheostat,” consisting of a U tube filled with fluid into which wires could he inserted to 
a less or greater depth, and having adjusted the resistance so as to obtain a given deflection when the 
tube was in circuit, he subsequently put the tube out of circuit and again adjusted the rheostat so as to 
obtain the same deflection ; the difference between the two readings was taken by him as the resistance 
of the tube in terms of the rheostat. 
Although he interprets his results differently, and is inclined to consider that gases at high temperatures 
really conduct electricity, they seem to ns to support onr hypothesis that the discharge is a disruptive 
one in all cases, even at the high temperatures he experimented with. For his measurements show 
that when he employed a definite potential and adjusted the external resistance to produce different 
currents, the resistance of the gas appeared to decrease with the increase of current, this agreeing 
so far with our results, But we are quite at a loss how to reconcile the following experiments with 
ours : — 
“ La seconde serie de determinations experimentales met en evidence nn fait assez curieux : c’est que 
pour la meme intensity electrique (the same amount of current), la resistance de l’air chauffe au rouge 
est d’autant plus grande que le nombre des elements de la pile est aussi plus grand.” Being unable to 
render an account to ourselves of this very remarkable result, we have, May 28th, 1878, made some 
experiments with tube 31, employing 1200, 2400, 3600, 4800, and 6300 cells, and by the introduction of 
wire resistance maintained a constant deviation of the galvanometer of 29°, we found that the apparent 
resistance of the tube was represented in each case by about 30,000 ohms. Our experiment differed from 
that of Becquerel, in that we employed rarefied gas at ordinary temperature instead of heated gas at 
atmospheric pressure, and our rheostat consisted of wire and not of a liquid electrolyte. 
Wiedemann and Ruhlmann (Poggend. Ann., cxlv., pp. 235-259, 364-399) by very different means 
have apparently arrived at the same result as our own. They started from the ground that the 
observations of one of them (G. Wiedemann) on the amount of heat generated in a Geissler-tube had 
demonstrated that rarefied gases did not behave like metallic conductors as had been asserted by other 
experimentalists (Moreen, Ann. de Chim. et de Phys., 4 e serie, iv., p. 325 ; Poggend. Ann., cxxx., p. 612 ; 
De la Rive, Compt. Rend., vi., p. 669: Archives de Geneve, nouv. ser., xvii., p. 53; Hittorf, Pogg. Ann., 
cxxxvi., p. 201, &c.) ; for the heat generated in the tube was far from varying as the square of the 
current. Wiedemann and Ruhlmann selected for their source of electrification the Holtz machine, 
which was driven by a hydromotor, so as to give a constant current, and had so great an internal 
resistance as to render negligible the resistance of the external part of their circuit. By ingenious 
contrivances they succeeded in measuring the time-interval between the rapidly-succeeding flashes of 
which the discharge in rarefied gases, under the circumstances of their experiments, was proved to consist. 
They found that, when other things were kept constant and the current alone varied, the time-interval 
between a pair of successive discharges was inversely proportional to the current : in other words, the 
quantity which flowed across the tube at each discharge was constant for all the values of current 
investigated by them. Hence, as fast as a certain definite accumulation of electricity (density) was attained 
on the terminals, these discharged themselves, became recharged to the same accumulation and discharged 
themselves again, and so on. So far, therefore, as their observations extend, they seem entirely accordant 
with our own result stated above, for it is obvious that as the interval between the discharges in their 
experiments did not exceed 0'002 second, the effect on any electrometer as yet devised would be practically the 
same as if the terminals were kept constantly charged to that potential which they have when the accumu- 
lation is sufficient to cause discharge; a, fortiori, would this be the case when a battery of small internal 
resistance is substituted for the Holtz machine. The range of Wiedemann and Ruhlmann’s observations 
on this point appears, however, to have been very limited; in their illustrations of the constancy 
of the quantity flowing across at each discharge when the current is varied, the relation between the 
maximum and minimum currents employed is 40 : 24’5. 
