214 MM. A. de la Rive and E. Sarasin on the Action of 



meter is closed, the electromagnet magnetized, and the current 

 passed. In the compartment which is placed between the two 

 magnetic poles the luminous jet is very much condensed and. 

 driven against the side of the tube. In the other compartment 

 scarcely any effect is apparent, except the diminution of bright- 

 ness resulting from the considerable augmentation of resistance 

 produced in the compartment submitted to magnetization, and 

 consequently the diminution of the intensity of the discharge. 

 The magnetism is permitted to act during from 10 to 20 se- 

 conds; then the large glass cock is closed, while the current is 

 still propagating; at the same moment the interruptor of the 

 Ruhmkorff is turned, and thus the current is stopped. Finally, 

 the manometer is observed while the cock which separates the 

 first compartment from the manometer is opened and then the 

 large glass cock. By operating thus it is found that the gas 

 contained in the compartment on which the magnet has acted 

 presents a very sensible increase of pressure, while in the other 

 an equivalent partial vacuum has been produced. By observing 

 the manometer, a great oscillation is seen to be produced there 

 the moment the cock is opened which puts it in communication 

 with the compartment on which the magnet has acted ; the level 

 of the mercury falls several hundredths of a millimetre. Then, 

 when we open the glass cock (which puts the second compart- 

 ment in communication with the first and with the manometer), 

 we observe a second oscillation, in direction opposite to the first ; 

 and when the motion of the mercury has ceased, we see that its 

 level has returned exactly to what it was before the experiment. 

 Thus, then, under the action of the magnetism, a certain quan- 

 tity of gas has passed from the compartment which is outside of 

 that action into that which was submitted to it, and has conse- 

 quently increased the density of the gas contained therein. 



This effect, of course, varies with the force of the magnet, with, 

 the intensity of the induced current, and the initial pressure of 

 the gas operated on. It evidently increases with the intensity 

 of the magnetism and with that of the discharge ; it increases 

 also very notably with the initial density of the gas, provided the 

 discharge is strong enough for its intensity not to be too much 

 weakened by this augmentation of density. For a given inten- 

 sity of the current of the induction-coil, and a determined dis- 

 tance of the electrodes, there is, then, for each gas a pressure at 

 which the effect observed is the maximum. This pressure is the 

 lowest for air (the worst conductor of the three gases on which 

 we operated), high for carbonic acid, and higher still for hy- 

 drogen. For example, in the large tube with the glass cock, 

 and with the Ruhmkorff of medium size, the three pressures 

 corresponding to the maximum effect were 7-8 millims. for air, 



