Magnetism on Gases traversed by Electric Discharges. 217 



as the gaseous column on which we operated was very short and 

 not sufficient, especially in the case of hydrogen, to eliminate the 

 induced current of feeble tension given by the Ruhmkorff coil at 

 the same time as the current of strong tension, in the opposite 

 direction, which was used in the experiment, we introduced into 

 the induced circuit a cylindrical Geissler tube of 49 centims. 

 length, 30 millims. diameter, and containing rarefied hydrogen. 



To measure the influence of magnetism on the electric conduc- 

 tivity of a rarefied gas traversed by the induction-current, we 

 reduced each of the gases operated on to a given pressure, the 

 same for all three ; then we regulated the intensity of the indu- 

 cing current in such manner as to have always, in each series of 

 experiments, the same intensity of the induced current. Having 

 thus, at the commencement of each experiment, placed the gas 

 in identical initial conditions, we magnetized ; then we sought 

 the degree of rarefaction to which it was necessary to reduce it 

 in order that, under the action of the magnetism, the intensity 

 of the traversing current might become the same as before the 

 magnetization. This method is more delicate than that which 

 would consist simply in measuring the diminution produced in 

 the deflection of the galvanometer when the rarefied gas is sub- 

 mitted to the action of magnetism. That deflection may, in fact, 

 under the action of magnetism, be reduced from 60° or more to 

 a very slight one — from 6° to 8°, for example, and even less, 

 about the value of which it is very easy to make a mistake. 

 There is, besides, a great advantage in varying the pressure, of 

 which a very precise measurement can be obtained — rather than 

 the intensity, which cannot be accurately measured by the de- 

 flections of the galvanometer-needle, to which it is not propor- 

 tional when they exceed 30°. 



We commenced the experiment by making a vacuum through- 

 out the system ; then we introduced dry and pure the gas to be 

 operated on, exhausted again, reintroduced the gas, and repeated 

 this two or three times in succession, so as to expel foreign gases. 

 By afterwards suitably regulating the cocks of the air-pump, we 

 succeeded in making its action perfectly regular, and thus ob- 

 tained, very exactly, the degree of rarefaction desired (30 mil- 

 lims. for example). The moment this pressure was reached,, the 

 cock of the glass tube was closed ; then the intensity of the in- 

 ducing current was regulated so as to obtain the required deflec- 

 tion of the galvanometer. This being done, we magnetized, and 

 gradually exhausted the tube until the galvanometer-deflection 

 had returned to exactly what it was before. We then again in- 

 terrupted the communication of the glass tube with the pump, 

 verified again that the induced current had really resumed pre- 

 cisely its initial intensity, and, lastly, took the reading of the 



