338 M. A. J. Angstrom's Optical Researches. 



17. Nitrogen. — It follows from what has gone before, that the 

 bright hues in the air spectrum beloug almost exclusively to the 

 nitrogen. To submit this conclusion to a direct test, I enclosed 

 a stick of phosphorus in the apparatus and ignited it by means 

 of a heated copper wire, introduced through the small opening 

 which permitted the issue of the gas. The aperture was then 

 closed. The nitrogen thus obtained is not pure, but is mixed 

 with a white smoke of phosphoric acid ; this, howevei', falls and 

 leaves the nitrogen alone behind. The electric spark showed the 

 same appearances as in atmospheric air. This result not only 

 corroborates the conclusion already referred to, but it also shows 

 that the presence of the phosphoric acid does not produce any 

 change, and that the air spectrum cannot be regarded as the 

 result of the combustion of oxygen and nitrogen. 



18. Another question follows : — Is the glowing of the gaseous 

 particles produced by the propagation of the electricity, or is it 

 a purely mechanical consequence of the motion of the metallic 

 particles ? The latter appears to be principally the case. When, 

 for instance, the sparks pass between two points in the same 

 horizontal plane, the metallic particles will nevertheless be scat- 

 tered vertically, and the lines 7, E, 8 appear as in a, fig. 5. 

 Hence we see that the glowing of the air particles is strongest 

 in the direction in which the metallic particles are scattered, and 

 that both directions commingle, as it were, in the centre. It is 

 remarkable, also, that the metallic particles show no tendency 

 whatever to combine with each other. Thus, for example, on 

 one occasion the zinc lines appeared on the blue, as at b, fig. 5, 

 without any sensible change being produced in the air spectrum. 

 The latter observation seems, however, to be opposed to the 

 foregoing, and hence to leave the question in respect to glowing 

 unanswered. A further contribution to its solution is furnished, 

 I believe, in what follows. 



In experiments with the sulphurets, for example with sul- 

 phuret of silver and sulphuret of antimony, when the same charge 

 was used, two kinds of sparks showed themselves alternately, one 

 stronger than ordinary, the other much weaker and of a dull red 

 appearance, but so that at the surface of the metal itself it pos- 

 sessed its ordinary character. Regarded through the prism, the 

 spark produced at the edge of the spectrum some bright points, 

 but beyond this the field was dark. I placed the prism at a 

 distance of only 4 feet from the spark, but in this case also 

 the field was so dark that it was not possible to determine directly 

 the position of the bright points. Only by the circumstance 

 that now and then stronger sparks showed themselves, was such 

 a determination possible, and it was thus found that the bright 

 points corresponded exactly to the lines peculiar to the metal, 



