through Rarefied Gases. 409 



of the maguet. Let ua fai-ther suppose that any number of such 

 curves of any form spread out in all directions from one given 

 point of space, then, on exciting the magnet, all such chains would 

 collect together into one and the same curve — in that magnetic 

 cm-ve, namely, which is completely determined by the given 

 point. 



48. The hypotheses conditioning such a phaenomenon are such 

 as can scarcely be realized ; so that the phaenomenon itself will 

 probably remain a merely imaginary one. If, however, in place 

 of the linked iron chain, we suppose rays of magnetic light, the 

 phaenomenon is converted into one which actually exists. 



49. On leading an electrical discharge through a Geissler's 

 tube provided at its extremities with bulbs to the centres of 

 which the electrodes penetrate, such magnetic light radiates from 

 every point of the negative electrode, and spreads through the 

 whole interior of the bulb containing this electrode. If all the 

 negative electrode except its extremity be isolated by a fused 

 coating of glass, then the above radiation is confined to the free 

 point. The rays proceeding from this point collect in one single 

 line of light, which coincides with the magnetic curve passing 

 through the end of the negative electrode, and which by its lu- 

 minosity renders such magnetic curve visible. 



Thus every ray which is bent in this magnetic curve, forming 

 a portion of the arc of light, behaves exactly as if it consisted of 

 little magnetic elements placed with their attracting poles in con- 

 tact. In other words, such a ray behaves as a magnetic thread 

 of perfect suppleness, and which accordingly, on being rigidly 

 held in any one point (the extremity of the negative electrode), 

 assumes the form of the magnetic curve passing through this 

 point, or, what comes to the same thing, as an electrical current 

 twisted in an infinitely thin spiral. 



50. By the above illustrations I have merely sought to make 

 the nature of the phaenomenon intelligible, without in the least 

 attempting to describe the nature of the magnetic light itself*. 



* I have nothing further to add to what I have aheady remarked (38) 

 concerning the formation of the luminous surfaces which take the place of 

 the magnetic curves of hght when the electrode, no longer isolated, radiates 

 magnetic light throughout the whole of its length. I may take this opportu- 

 nity of mentioning a wider tube of about 32 centims. in length, into which 

 the one electrode penetrated a distance of 10 centimetres. If, in passing 

 the electricity through, this electrode was made the negative one, the tube 

 was filled, as far as the electrode penetrated, with magnetic hght. When 

 the tube was placed in an equatorial position upon the approximated arma- 

 tures, this light contracted to a beautiful, single, dome-shaped light-sur- 

 face. Ou placing the tube in uu axial jjositiou, the magnetic light collected 

 in the |ilauc pussiug through the wire, and formed in this plane a splendid 

 arch upon the glass, extending from one armature to the other. The inner 



