174 PHENOMENA ACCOMPANYING 



In two other comparative expcriiiieuts, the pressure and the distance of the 

 electrodes were made to vary alike for the nitrogen and hydrogen. The two 

 eases had been successively introduced into the same ovoid globe. The fol- 

 lowing table gives the intensity of the derived current for three different dis- 

 tances' of the electrodes under diftcrent pressures, when nitrogen is in the ball: 



Pressure. Distance of the electrodes. 



14 cent. 7 cent. 1 cent. 



20""" 10° 23° bb"" 



gm"' 40° 47° bb° 



5mm 50° bb° 57° 



3mm 550 55° 57° 



When ibr nitrogen we substitute hydrogen, the results differ somewhat, 

 especially in the lower pressures, and when the electrodes are in close proximity 

 with one another, which proceeds probably from the circumstance that the 

 gaseous medium, in view of the form of the vessel, may be regarded as having 

 an almost indefinite breadth. In a large receiver, in effect, where the current 

 passes between a central ball and a concentric ring having a diameter of 12 

 centimetres, the intensity of the derived current is very little increased by 

 diminishing the pressure beyond 10"''^\ That intensity, measured by the de- 

 rived current, amounts, under a pressure of 15""™, to 35° ; under a pressure of 

 10™'", it attains 45"; then augments gradually as far as 5"'™, when it reaches 

 its maximum of 50°, which it never exceeds, manifesting rather a slight ten- 

 dency to become less under 2'"'^. By raising the central ball so as to give to 

 the electric sheet a conical instead of circular form, the conductibility is not 

 sensibly diminished. Under the same circumstances the atmospheric air does 

 not present a resistance much greater than the hydrogen ; thus the intensity 

 of the derived current is 35° at 5'"'" instead of 50°, and at 2™™ is 45°. How- 

 ever, with the tube of one metre in length, hydrogen must be subjected to a 

 much weaker pressure in order to transmit the discharge, but its conductibility 

 increases very raj)idly with the diminution of that pressure. Thus, the appa- 

 ratus of derivation being placed in the circuit, Ave have : 



Pressure. InteDsity of the derived current. 



5'"'" , 0° 



4mm ]^2° 



3'^™ 22° 



Sni'" 30° 



2mm 52° 



Finally, with the same tube, one metre in length and 5 centimetres in 

 diameter, a sensible and regular augmentation in the intensity of the derived 

 current has been obtained, for the same pressure and in the two gases alike, by 

 a corresponding diminution of distance between the electrodes. The compari- 

 son of the numbers indicates, that, when the gas is sufficiently rarefied to be a 

 good conductor, that is, to permit the discharge to become, so to say, continu- 

 ous, it follows, like liquids and solids, in its conductibility, the law of the inverse 

 of the length. It has been already seen that the influence of the section 

 and of the volume is very considerable ; but I have not been able to determine 

 its law in a j)recise manner. 



Thus far I have considered the |iropagation of electricity in gaseous sub- 

 stances only in relation to the resistance they oppose to it — a resistance variable 

 with their nature, their density, and their dimensions. I have but glanced 

 at this part of my subject, to which I shall return at an early occasion, as I 

 propose to extend my researches to a much larger number of gasd^us substau- 



