THE PROPAGATION OF ELECTRICITY. 175 



ces. I pass now to phenomena of quite anotbev order, and which relate to the 

 mode itself in Avhich the propagation of electricity is effected in gases — a mode 

 which manifests itself under the form of stratification of the electric light. 



^ II.— INVESTIGATIONS REGARDING THE STRATIFICATION OF THE 

 ELECTRIC LIGHT. 



It is known that at a certain degree of diminution of the elastic force of 

 a gas which transmits the electric current, that current becomes stratified — that 

 is to say, is decomposed into strata alternately obscure and luminous. The 

 stratification commences hy the appearance of certain slight striae or furrows on 

 the side of the positive electrode ; then gradually, as the elastic force dimin- 

 ishes, the current, which Avas at first very narrow, dilates, and the strioe grow 

 larger. Next appears an obscure space, separating the extremity of the lumi- 

 nous column from the negative electrode, which is itself surrounded with a bluish 

 atmosphere. This atmosphere continues to dilate, and the obscure space to 

 lengthen, in proportion as the rarefaction of the gas increases. 



In order to obtain the stratification of the electric light, it is necessary to di- 

 minish the pressure of a gas in proportion as the gas offers more resistance to 

 the transmission of electricity. Thus in hydrogen, under a pressure of 18"'™, 

 the electric stream, which consists as yet of but a small rose-colored filament 

 from three to four millimetres in diameter, is seen to divide into very distinct cir- 

 cular sheets, alternately obscure and luminous, the breadth of which is one-fourth 

 of a millimetre. These striae, at first more distinctly marked at the positive elec- 

 trode, become general throughout the whole electric current, whatever bo its 

 length ; and, in proportion as the pressure diminishes, the stream becomes en- 

 larged, so as even to occupy the whole interior of a tube five centimetres in di- 

 ameter. At the same time the breadth of the alternately obscure and luminous 

 divisions so increases that, under a pressure of 2^^, it is about 5™™. These di- 

 visions are themselves annular, as I have satisfied myself by closing the tube 

 which contains the rarefied gas, at one of its extremities, with a glass disk, 

 which permits the whole interior of the tube to be seen in the direction of its 

 length. 



When the striae begin to appear, an obscure space, as has been said, is seen 

 to form in front of the negative electrode, increasing in proportion as the press- 

 ure diminishes, so far as finally to occupy a length of ten centimetres— ^a length 

 which is independent of that of the gaseous column. However, by oh^orving 

 with attention this obscure space, Ave discover, beyond an interval which is per- 

 fectly black, and of a well-defined length of from 2 to 3™™, a palish, rose-colored 

 gleam, which is only visible in utter darkness. This gleam, which has the form 

 of a cone, whose base is the last section of the luminous column, only appears 

 when the pressure has become very slight and quite inferior to that under which 

 the obscure space is manifested. It is accompanied by the appearance, in the 

 same obscure space, and at unequal intervals, of several still more luminous 

 rings, (I have counted as many as four,) which contrast, by their immobility and 

 their well-defined outlines, with the agitated striae or divisions of the rest of the 

 current. Let us add, that the luminous and stratified part of the current, which 

 is much the longest, is so much the more distinctly and sharply separated from 

 the obscure or palish part, as the electric discharge is more intense. 



The bluish atmosphere which surrounds the negative electrode is also en- 

 larged in proportion as the pressure diminishes, and nearly in the same ratio 

 as the striae. At the same time, its brightness becomes less vivid, and its exte- 

 rior outline less sharply defined. This bluish atmosphere, which at first 

 enveloped only the negative ball, at last, and in proportion as the pressm-e 

 diminishes, equally envelops, in all its length, the metallic rod which supports 

 the ball ; at least, if this be not covered Avith an isolating coat, Avhich indicates. 



