1923. No. 10. DISTRIBUTION OF MATTER IN THE ATMOSPHERE. I I 



to the auroral streamers. The colouring should thus van»- with the degree 

 of dissociation or evaporation from the Nitrogen clusters. 



From obsenations of the auroral spectrum at various height-inter\als, 

 we found that the relative intensity of the green line, as compared with 

 that of the Nitrogen Unes, did not increase as we passed upwards in the 

 atmosphere, but showed a marked diminution. 



On the hypothesis that the upper strata of the atmosphere were in a 

 gaseous state, we concluded that the green line had to be ascribed to 

 Nitrogen. 



If, however, we assume the atmosphere to consist of fairly large 

 electrically charged clusters, we have to consider the possibility that also 

 dust particles of other kinds of matter might be mixed up with the Nitrogen 

 particles. This cosmic dust might even be composed of atoms with a large 

 molecular weight. We might also think of the possibility of Ozon particles 

 being mixed up with Nitrogen. 



The very same particle might also contain different kinds of atoms — 

 and perhaps Nitrogen might enter into chemical combination with some 

 substance. Such an interpretation, however, is ver\- improbable, for, if the 

 green line was due to some element different from Nitrogen, we should 

 expect to find in the auroral spectrum a number of other lines belonging 

 to the same element. But, apart from the green line, only three faint lines 

 have been observed, which are not found among the N-lines. obser\ed in 

 laboratoratories. 



Now the green line is not identical with any known lines. If the other 

 3 unknown lines in the auroral spectrum »4182, 3432, 3208I could be 

 determined accurately enough to show that they are not identical with 

 known lines of any of the elements, I think there can be little doubt that 

 all four lines are emitted from Nitrogen, for it would be a most curious 

 case that an element was brought to emit on/v lines which could not be 

 reproduced in laboratory experiments. 



As also mentioned in my first publications on auroral phenomena, 

 there is nothing strange in the idea that Nitrogen under the conditions 

 present in the upper strata of the atmosphere, when bombarded with cosmic 

 rays, may be brought to emit lines not found under the conditions expe- 

 rimented upon in laboratories. 



In fact, that part of the auroral spectrum which belongs to Nitrogen 

 is of a very singular type. The number of lines is ver^- much smaller 

 than usuallv found in N-spectra observed in laboratory experiments. 

 Although most of the lines belong to the negative and positive band spect- 

 rum of N-, there only appear fairly sharp distinct lines on the plate — and 

 the striations peculiar to these bands are either not visible on the plate 

 or very much reduced in intensity. — Under the conditions under which 

 the auroral light is emitted the formation of the Band-series is suppressed, 

 and only some principal lines appear. 



