through Rarefied Gases. 415 



The eye perceives, for instance, in one part of the tube red, in 

 another violet, and in the middle cylinder a fainter colour ; so 

 that one would be inclined to imagine that the ponderable matter 

 which becomes luminous is differently distributed through the tube. 

 In addition to this, it happens in many cases that the colour of 

 the electric light undergoes a change in its passage through the 

 narrow tube on the excitation of the great electro-magnet (upon 

 the approximated armatures of which we shall suppose the tube 

 placed, in equatorial position) ; so that if the light were, for in- 

 stance, in the first place of a beautiful red or violet, it becomes 

 of a fine green or blue colour (59). But in all cases, whatever 

 may be the colour-impression produced on the eye, the distribu- 

 tion of the colours in the spectrum remains for the same gas 

 entirely of the same kind; it is the intensity of the colours alone 

 which changes in different degrees in different portions of the 

 spectrum : so that when the eye (whose judgment is, moreover, 

 considerably influenced by the external illumination) is at fault, 

 still the nature of the gas or vapour contained in the tube is un- 

 failingly determined by means of the spectrum. 



68. If, for instance, we take a narrow tube with platinum 

 electrodes whose capacity is only a fraction of a cubic centi- 

 metre, and, having filled it with hydrogen, so far exhaust it that 

 the tension is reduced to two or three millims., then the traces 

 of gas are so small that they may be said to be inappreciable by 

 chemical reaction, and the most delicate chemical balance is 

 unable to recognize their weight, being, as it is, a fraction of the 

 thousandth of a milligram. On passing an electric current 

 through the tube, however, these traces of the gas becoming- 

 luminous, are rendered visible to and recognizable by the 

 almost dazzled eye by means of a beautiful characteristic 

 spectrum. 



69. As yet I have only observed these spectra by means of a 

 telescope (such as that employed by Fraunhofer in the obser- 

 vation of the lines of the spectrum), without angular measure- 

 ment. The refracting angle of the flint-glass prism, which was 

 placed immediately before the object-glass of the telescope, was 

 45°. The eye-glass of the telescope was, however, generally 

 replaced by one of less magnifying power. The narrow tube 

 through which Ruhmkorff's apparatus was discharged, was sus- 

 pended vertically at a distance of 4 millims. 



70. Hijdroyen yas gave a comparatively simple spectrum, in 

 which five bright bands of almost equal breadth were especially 

 prominent : — A, a bright violet band beyond the limit of the 

 spectrum; three bands in the green, of which the one, B, bounded 

 the green towards the violet, D forming the boundary on the 

 other side, and C between these last about twice as far from B 



