THE PROPAGATION OF ELECTRICITY. 191 



of the dry and pure gas was in the commencement 5™™ ; at this pressure, as we 

 shall forthwith see, the hydrogen transmits the discharge only under the form 

 of a luminous sheet. The pressure was afterwards augmented solely by means 

 of the vapor of alcohol, with the following results : 



Pressure. Number of revolutions in a minute. 



Ring', positive. Ring, negative. 



7™™ Luminous sheet 92 



lO-"™ 80 52 



ISni"" 64 48 



IS'"" 48 38 



ISmni 40 32 



22°im 30 25 



27">™ 24 18 



36'°'° 12 10 



38™™ 12 10 



The division into distinct currents, more or less numerous, was manifested 

 when the ring was the positive electrode. 



When pure and dry hydrogen is adopted as the medium in which the dis- 

 charges take effect, the phenomena of rotation are obtained with giVat difficulty. 

 At rather strong pressures, such as that of 128™™, we have a number of currents, 

 but these currents are too intermittent to allow of the magnet's acting upon them. 

 At 90"'™ I have obtained a small stream under the form of a bluish-white 

 iilament, which, the ring being positive, turned at the rate of thirty-five times 

 per minute ; but, at the lapse of some instants, it became subdivided into a 

 multitude of small, irregular streams, and rotation was no longer perceptible. As 

 far as 40™™, the action of the magnet was indistinct. At 30™™, the negative 

 ring was covered with small violet sheaths, at equal intervals, which seemed to 

 experience, at the moment it was magnetized, a tendency to move in one direc- 

 tion or the other, according to the direction of the magnetization. The same 

 was the case with the small brilliant points, likewise distributed at minute in- 

 tervals, with which the ring, Avhen positive, is covered. At 5™™, and still more 

 at three and at two, the ring is entirely covered, when it is negative, with aline 

 violet-colored sheath, which becomes contracted under the influence of the mag- 

 net. The top of the iron rod, which is then positive, is surrounded by a beau- 

 tiful white aureole, slightly tinged with rose, three centimetres in breadth, and 

 stratified in a very marked degree. Magnetization sensibly contracts this au- 

 reole, and compresses its striae without diminishing their number, elevating it, 

 and, at the same time, giving it the form of a pear resting with its base on the 

 magnetic pole. When this pole is the negative electrode, there issues from it, 

 as we have seen, a brilliant tuft of a violet color, which conforms itself to the 

 action of the magnet. 



All the phenomena just described show, in a striking manner, the molecular 

 differences which various elastic fluids present, as regards one another, even at 

 an advanced degree of rarefaction. Thus in hydrogen, although that gas is a 

 very good conductor of electricity, electric currents can, wuth difficulty, and, 

 indeed, scarcely at all, obey the action of the magnet, probably by reason of the 

 slight density of the gas. In air, and in nitrogen, it is quite otherwise, and still 

 more when these gases are humid. The singular property possessed by the 

 electric current of dividing itself into several small and distinct streams, instef.d 

 of diffusing itself, under the influence of magnetization, when the medium which 

 transmits it contains a more or less quantity of vapor, would seem to indicate 

 in the vapor a greater cohesion than in the gases properly so called, if, indeed, 

 we may employ the term cohesion when the question relates to elastic fluids so 

 much rarefied. It might also be possible that this division into streamlets is 



