178 PHENOMENA ACCOMPANYING 



ncativc electrode, strice of a fine rose color arc immediately seen to form in the 

 obscure space, their diameter being that of the stratitied column— that is, of 

 the tube, while they are at the same time very narrow and well defined. They 

 are gradually proijagated in the tube, confounding themselves with the original 

 stria?, which arc much larger and less distinctly limited; then, as soon as the 

 entrance of the gas is arrested, the luminous column is seen to recede slowly 

 from the negative electrode, and resume gradually its primitive appearance. 

 "When the introduction of the gas takes place on the side of the positive elec- 

 trode, in place of the stria) occupying the whole cavity of the tube, we see a 

 brilliant stream of very small diameter {2 to 3""") distinctly striated, and quite 

 similar to u minute spiral spring (rcssort a hovdin,) advance along the axis of 

 the tube in the relatively obscure interior of the luminous column, which itself, 

 as soon as the gas begins to enter by the positive as well as by the negative 

 electrode, immediately advances so as to occupy almost entirely the obscure 

 space up to the negative electrode, from which it is only separated by the in- 

 terposed stratum, 2'""' in thickness, which it cannot surmount. Then, the intro- 

 duction of the gas once stopped, everything returns quickly to the normal state. 

 By whichever of the two extremities of the tube the gas is made to penetrate, 

 we see, on the entrance of the gas, a very subtle mist of a roseate white color 

 make its appearance, and diffuse itself in the tube ; but this, as soon as the in- 

 troduction of the additional quantity of gas has ceased, passes over from the 

 negative to the positive electrode, leaving the obscure space to form itself anew, 

 and momentarily hiding in its passage, by enveloping them as it were with a 

 light cloud, the successive stratifications of different parts of the column ; then 

 this mist disappears, and the luminous column resumes its primitive appearance, 

 which it maintains so long as nothing is changed either in the electric current 

 or the state of tin? gas traversed by it. The appearance of this mist, which per- 

 fectly resembles that I have mentioned as existing in the dark space of the 

 column in a state of repose, well denotes the agitation into which the introduc- 

 tion of a small additional quantity of gas throws the whole column — an agitation 

 so conspicuously manifested by the progression of the striaj and their encroach- 

 ment on one another. The phenomenon presents this further feature : that the 

 definitencss and brightness of the stria; in the gaseous portion introduced, which 

 make them so plainly distinguishable from the gas which was already in the 

 rtube, enable us to follow the progressive movement of that portion from one end 

 ■of tlie tube to the other. The experiment may be repeated several times in 

 succession by successive introductions of additional quantities of gas, provided 

 that each time the pressure be not increased more than ^ of a millimetre, and 

 that the total ])ressure do not in all exceed 5 or C""". 



With nitrogen and atmospheric air the incidents arc the same, only avc can- 

 not push the experiment so far, the pressure, at which the phenomenon ceases 

 to take place with these gases being much less than it is with hydrogen. The 

 narrow strife Avhich display themselves at the moment of the entrance of the gas 

 on that side where the entrance takes place are also less distinct and less bril- 

 liant, but there is equally a momentary disappearance of the obscure space, the 

 production of a roseate mist, and progression of this mist, on the cessation of 

 the introduction of gas, from the negative electrode to the positive. With the 

 three gases alike, we see, when the introduction is effected on the side of the 

 negative electrode, the mist advance at first like the slender striated thread 

 which follows the axis of the tube from the positive electrode to the negative ; 

 then, having arrived at this extremity of the tube, it turns back, passing over, 

 as has been said, from thp negative to the positive electrode. 



This mist evidently proceeds from a portion of the gas which, in entering the 

 tube, is excessively dilated, and becomes visible by the electricity which trav- 

 erses it. From the slowness Avith which the mist is propagated we may judge 

 of the feeble degree of elastic force in the gas. It is to the same cause probably 



