THE PROPAGATIOX OF ELECTRICITY. 181 



current, and consequently that of the resistance, is much greater when the 

 sounds are immersed in the obscure space near the negative electrode than in 

 the himinous part of the stream near the positive one. Thus, under the pi'css- 

 ure of 2'^°\ it is impossible in hydrogen to perceive the least derived current 

 in the black space, while this derived current is at the same time 35° in the 

 luminous space ; at a pressure of 15"*™ it was 90° in the nei^-hborhood of the 

 two electrodes alike, but there Avas as yet no formation of the obscure space, 

 and consequently the state of density of the gas was the same at the two ex- 

 tremities of the tube. Tlie resistance of the obscure space is also very feeble 

 in nitrogen under a pressure of 2™"', since the derived current is only 3°, while 

 it is 18° in the luminous space ; but the difference between the two derived 

 currents is less than in hydrogen. This difference results from the fact that 

 hydrogen, having a conductibiliry much superior to that of nitrogen on the one 

 hand, the absolute intensity of the current is gi-eater, which explains why we 

 have 35° instead of 18° in the luminous space ; on the other hand, the derived 

 portion must then be less where rarefaction still more augments the conducti- 

 bility of the gas, which accounts for our having 0° in place of 3° in the obscure 

 space. 



Let us here remark, that all the results which show the unequal resistance 

 presented by differeiit parts of the same gaseous column to the propagation 

 of electricity are readily comparable with one another, since it is the same elec- 

 tric stream which successively traverses these different and unequally conducting 

 parts. 



If we place the sounds in a portion of the stream which is ^ of the distance 

 from one of the electrodes, and consequently § from the other, we have for 

 the intensity of the derived current, under a pressure of 2™™ in air or nitrogen, 

 8"^ when the negative electrode, 12° when the positive, is nearest to the sounds. 

 In hydrogen we have 20° and 36°. Thus, the couductibility of the gaseous 

 column goes on diminishing gradually from the obscure space, where it is at its 

 maximum, to the space near the positiA'c electrode, where it is at its minimum. 



By placing the sounds always in the same portion of the stream, we can 

 find in the intensity of the derived current a sufficiently exact expression of the 

 degree of resistance of different gases at different degrees of pressure, provided 

 we take care, by means of a rheostat, to give to the principal current in each 

 case the same degree of absolute force. This is an investigation Avith v,'hich I 

 am at present occupied, and which is not yet finished. • 



We see, then, that the obscure space near the negative electrode offers much 

 less resistance to the passage of the current than does the luminous part near 

 the positive electrode. It thence results that, for the same reason that the less 

 conductive portion of the gaseous column is more luminous than that Avith 

 greater conducting capacity, Avhich remains nearly dark, the temperature of the 

 first should be higher than that of the second — an inference Avhich experiment 

 has fully confirmed. 



Two thermometers of mercury, Avith cylindrical reserA'oirs, were placed in the 

 interior of the tube, which is 16 centimetres in length and 4 in diameter, at the 

 respective distance of one centimetre from each of the electrodes — a distance 

 sufficient, as Avas ascertained, to annul the cooling or heating influence of those 

 electrodes. That the influence would rather liaA'e been refrigerant, Avas found 

 susceptible of A^erification by bringing them nearer the reservoir of the ther- 

 mometers, which is not surprising, in view of their dimensions, (full metallic 

 balls one centimetre in diameter.) 



By causing the electric stream to traverse the rarefied nitrogen or hydrogen, 

 a great difference Avas at once perceptible between the temperature acquired by 

 the thermometer placed in the dark space near the negatiA-e electrode and that 

 acquired by the thermometer placed in the luminous space near the positive 

 electrode. These differences observe nearly the same ratio between the press- 



