Igg PHENOMENA ACCOMPANYING 



rci-istancc once surmountod, we may interrupt the passage of the discharge with- 

 out iucurring the necessity of waiting more than an instant for the transmission 

 to recommence, when we close the circuit anew, provided the interruption does 

 not exceed an hour or two. The luminous current presents, with liydrogen 

 under a pressure of 5 or C""^, very neat and distinct striai of a rose color; at a 

 pressure of 2"^', they become much larger and less distinct ; the color is also 

 paler. The same occurs Avith air and with nitrogen, but the effects are more 

 striking with hydrogen. A remaikable ajjpearance presented by the current in 

 the interior of the spiral is, that it seems to undergo a very distinct movement 

 of rotation, in a direction which appears to vary with the direction of the dis- 

 charge ; but this last result is not very constant, which has led me to think that 

 the rotation is only apparent, and that it is the effect of the discontinuity of the 

 discharges which constitute the current, a discontinuity which produces the 

 illusion of a displacement. This point, however, deserves to be studied anew. 



In order to observe the action of magnetism on the spiral current, I place the 

 spiral of glass between the two poles of the electro-magnet in such a way that 

 its plane shall be the same with that of the two polar surfaces, the two prolon- 

 gations being thus rendered vertical, the one above, the other below, that plane. 

 The magnetization, according to its direction, either condenses the current 

 towards tlie interior walls of the spiral tube, or, on the contrary, repels it towards 

 the exterior walls, rendering it very difi'use. In the first case, it becomes highly 

 brilliant, and the stratifications are very distinct; in the second case, they are 

 but slightly visible, and the current itself is much larger and quite dim. It 

 aj)pcars to undergo, in even a more sensible manner, the movement of rotation, 

 of which we have spoken. A quite curious fact is, that in the vertical branch 

 of the tube which is below the spiral, and consequently between the two 

 branches of the electro-magnet, the current divides itself, under the influence of 

 the magnetism, into two streams or filaments, which tend, respectively, to one 

 and the other side of the tube. Of these two filaments, one is very small, and 

 of little brilliancy, in comparison with the other. The cause of this separation 

 consists, very probably, in the fact that the inductive current of the Ruhmkorff 

 apparatus is really composed, as Ave have already taken occasion to say, of two 

 successive and opposite inductive currents, one having much more tension, and 

 passing almost exclusively through the gas, while the other is transmitted with 

 much difficulty, but yet passes, in very small proportion, it is true, since the 

 action of the magnet separates it from the principal current, Avhich is the only 

 one in general that it is requisite to consider in this kind of phenomena, because 

 it is by much the strongest. 



I have sought to determine in the case of the spiral tube, as I had done Avith 

 the large rectilinear tube, the influence of magnetization on the resistance of the 

 gas to the transmission of the discharge, and 1 have obtained a rather curious 

 result. The tAvo points of platina of the apparatus of derivation being at a dis- 

 tance of ten millimetres from one another in the distilled water, I obtained a 

 derivative current of 20°, the spiral tube being filled Avith hydrogen under the 

 pressure of 2""". The spiral Avas placed A'ertically between the tAvo horizontal 

 armatures of the electro-magnet Avhich Avere exactly in contact Avith its tAvo 

 faces. As soon as the magnetization took place, the derivative current Avas re- 

 duced to 15°, Avhen the discharge Avas repelled, and driAcn tOAvards the exterior 

 Avails of the spiral Avith an apparent movement of rotation, and it Avas raised, on 

 the contrary, to 25°, Avhen the discharge Avas condensed towards the interior 

 Avails of the spiral. Does this influence of the direction of the current or of the 

 magnetization de])end on the particular form given to the stream, or to the small 

 diamctoi- of the tube, in comparison Avith its dcA'clopmeut in length ? It is a 

 point for future elucidation. 



I pass uoAv to the case Avhere the magnetic pole is in the midst of the gas 

 which transmits the discharge. I have first operated with a spherical globe. 



