in highly Rarefied Elastic Fluids. 249 



bristles with greatly elongated little filaments which form a sort of 

 projecting* tuft. These filaments are probably formed by the series 

 of molecules which transmit the discharge. They are much 

 more distinct in the case of hydrogen (the best of gaseous con- 

 ductors) than with other gases. If the knob is employed as 

 positive electrode, it is surrounded by a bright rose-coloured 

 halo (about a centimetre in diameter) which shows very distinct 

 stratifications; next comes a dark annular space terminating at 

 the ring, which is in its turn completely enveloped by a light- 

 violet opalescent sheath. 



Nitrogen presents just the same phenomena as hydrogen, ex- 

 cept that the stratification of the electric light does not begin till 

 a much lower pressure is reached. In a long tube (1 metre in 

 length), the agitation of the stria? under a pressure of 2 millims. 

 is even greater than with hydrogen ; the strise seem to form a 

 helix animated by a movement of rotation about its axis; the 

 light also is brighter, and peach-coloured instead of pale rose; 

 and altogether the phenomenon is very brilliant. In other re- 

 spects, there is the same dark space near the negative electrode, 

 the same faint rose-coloured glow in this dark space under a 

 pressure of 1 to 2 millims., and the same appearance of a kind of 

 mist, and of sharply defined motionless rings of greater bright- 

 ness than the surrounding space. 



Atmospheric air behaves like nitrogen ; I have only found that 

 the stria? are less strongly agitated, and that the light is not of 

 so dark a rose-colour as in the latter gas. 



The appearances that I have described are accordingly, with 

 the exception of slight differences of tint, exactly the same in 

 hydrogen, nitrogen, and atmospheric air; they are the same also 

 whether these gases are dry or whether they contain a greater 

 or less quantity of the vapour of water or of alcohol. The only 

 differences are that the pressures at which the various pheno- 

 mena are obtained, and the tint of the light which accompanies 

 them, differ with the nature of the rarefied elastic fluid. These 

 effects, therefore, cannot be attributed to an electro-chemical de- 

 composition (which could not take place in a well-dried elemen- 

 tary gas), nor to any action depending upon the chemical nature 

 of the elastic fluid. They are evidently the result of a mecha- 

 nical action which accompanies the transmission of electricity, as 

 was first suggested by M, Riess, who showed that an analogous 

 phenomenon is presented, under a slightly different form, by 

 liquids and solids. 



The phenomenon, in the case of elastic fluids, must consist of 

 alternate condensations and rarefactions of the gaseous medium, 

 produced by the always more or less discontinuous series of dis- 

 charges which constitute the electric jet. In fact, whether the 



