As mentioned in the first comniunicati'm the pressure of any electri- 

 cally neutral gas present in the atmosphere will vary according to the 

 equation: 



Mo- 

 dp = — ?gdh = — ^ dh. 



(I) 



The variation of pressure depends on the molecular weight (M). and 

 if several gases were present the composition of the atmosphere should 

 vary rapidly as we pass upwards. 



Assuming the temperature above lo km. to be 220" Abs., the pres- 

 sures at different heights for the gases H, Ke. X, and O are given in 

 table I\ of the first communication, and from these numbers we find 

 at different heights the following ratios between the pressures of H, He, 

 O, and that of X: 



If the upper strata are electrically charged, the variation of pressure 

 — as we saw in the first paper — will be different from that of a neutral 

 gas; but for a given difference of molecular weight the charge should 

 increase the rate at which the ratio between the densities of the two gases 

 changes with variation of altitude. 



If the auroral region of the atmosphere. alx)ve 100 km. say, 

 were composed of gases with a different molecular weight, and if these 

 gases were present in such quantities that, through bombardment with 

 electric rays, they were able to give out light of sufficient intensity 

 to be observed in the auroral spectrum, the intensities of the lines emitted 

 from different gases would vary largely with variation of height. 



Now we found that a great number of auroral lines are emitted from 

 Nitrogen, but the strongest of all auroral lines (the green line) and three 

 faint lines in blue and ultraviolet cannot be found among the lines of 

 Nitrogen given in the tables of spectral lines. 



