THE AURORA BOREALIS 215 



AMOUNT OF GAS DETECTABLE IN A MIXTURE 



1. Helium in hydrogen, 10 per cent, of helium barely visible. 



2. Hydrogen in helium, O'OOl ,, of hydrogen visible. 



3. Nitrogen in helium, O'Ol of nitrogen almost invisible. 



4. Helium in nitrogen, 10 of helium difficult to detect. 



5. Argon in helium, 0'06 still visible. 



6. Helium in argon, 25 invisible. 



7. Nitrogen in argon, 0'08 just visible. 



8. Argon m nitrogen, 37 barely visible. 



9. Argon in oxygen, 2 '3 ,, difficult to distinguish. 



This table shows the enormous differences which exist 

 between the behaviour of different gases. To take the 

 extreme cases while it is possible to detect 1 part of 

 hydrogen in 100,000 of helium, it is barely possible to 

 recognise 1 part of argon in 2 of nitrogen. 



Similar experiments with krypton showed that 

 In air, 1 part of krypton is visible in 7,100 parts. 



In oxygen, 1 1,250,000 



In hydrogen,! 67 



In argon, 1 7,150 



In helium, 1 ' 2,860,000 



The pressure of krypton, too, in the case of air is almost 

 inconceivably low ; it amounts to only one thirty-millionth 

 of the usual atmospheric pressure. This shows the enor- 

 mous persistency of the krypton spectrum that is, of the 

 most conspicuous line, the auroral green, for that was the 

 one observed in all instances. If, then, an electric dis- 

 charge passes through the upper and rarefied strata of the 

 atmosphere, the probability of detecting the green line of 

 krypton will be much greater than that of detecting the 

 spectrum of any other element, even though the latter 

 be present in enormously greater proportion. Hydrogen 

 alone has any marked power of extinguishing the spectrum 

 of krypton. 



It is possible to calculate the maximum height of the 

 aurora, on the supposition that the krypton line is no 



