THE PROPAGATION OF ELECTRICITY, 180 



After these first experiments, which served as my introduction to this sort of 

 researches, I resumed the study by availing myself of the jar of twenty by six- 

 teen centimetres, described above. The following are the results obtained with 

 dry atmospheric air : 



Pressure. Number of revolutions iu a minute. 



Ring', positive. Ring, negative. 



16'"°^ 55 oG 



iy""Ji 83 55 



9'"'" 99 63 



G'"« „ 100 



3"m „ 128 



At 9°^™, with the ring serving as a positive electrode, there is no longer a 

 stream, but a dilatation of the discharge, forming a sector of from 30° to 45'^ ; 

 and this sector obeys the movement of rotation as the stream before obeyed it. 

 But it enlarges, in proportion as the pressure diminishes, and at 6™'» forms a 

 complete circular sheet, and it is then that the rotation, which, up to this point, 

 had increased iu rapidity, becomes no longer sensible. When the ring serves 

 as a negative electrode it is covered with a violet sheath, whose size likewise 

 increases in proportion as the pressure diminishes, but which occupies only half 

 the circumference of the ring under a pressure of 4™"^. It is seen to turn very 

 rapidly, but at a pressure of 2""™ it occupies the whole circumference of the 

 ring, and there is no longer any sensible rotation. At the summit of the mag- 

 netized iron rod there is a roseate aureole, from which, as has been said, emanates 

 at one point a very short jet iu shape of a comma, which turns with the violet- 

 colored sheath, from which it is separated by a very considerable obscure interval. 



It should be remarked that, at a pressure of 6, of 4, and sometimes of even 

 3™"", it most often happens, when the ring serves as positive electrode, that at 

 the first moment of the circuit being formed there issues a stream which turns 

 too rapidly to allow its velocity of rotation to be measured, but which quickly 

 expands so as to form, for so ne instants, a sector which continues to revolve, 

 and soon after a complete circular Sheet, which no longer manifests any movement. 



It does not follow that the action of magnetism is annulled when the gas is 

 too much rarefied for the continuance of a sensible rotation. That action is 

 manifested under another form, as is shown by experiments made under a 

 pressure of from 3 to 2^^^™. Thus, if the ring serves as negative electrode, the 

 violet sheath which surrounds the soft iron is seen, at the moment when this 

 last is magnetized, to subside sensibly, and to rise at the instant of its being 

 demagnetized. If, on the contrary, the ring serves as positive electrode, the 

 rose-colored sheet Avhich tills the interval between the ring and the summit of 

 the central rod of iron is raised, as well as the violet sheet which issues from 

 that summit, at the moment of magnetization, and depressed at the instant of 

 demagnetization. 



The following is a more complete experiment with dry nitrogen, and shows 

 that rotation begins to manifest itself at stronger pressures when the ring is 

 positive than when it is negative : 



Pressure. Number of rotations in a minute. 



Ring, positive. Ring, negative. 



35™°i 12 , 



29°''" 27 



2lmm 45 3G 



16'™ 67 51 



121121 99 ,'39 



8«i™ 115 70 



6mm „ 115 



5™™ ,, 150 



