lO 



L. G. Raiib 



not completely eliminated, and after the current had been passed 

 for several hours spectroscopic observations always revealed, more 

 or less faintly, the characteristic flutings of nitrogen, with the 

 helium lines standing out strong and sharp. According to Ram- 

 say,^* one-hundredth of one per cent, of nitrogen is still visible in 



Jso 



y 



/?S 

 /so 



K>/f^ 



/zs 



/6 



I 



J L 



J I I L_j I I 



ys 



.2.0 



.;is 



,30 c/7r 



Fig. 4 



helium, so that it seems probable that in the samples used there 

 were only a few hundredths of one per cent, of nitrogen at the 

 most present as an impurity. 



1* Ramsay, Proc. Roy. Soc, 59, p. 325, 1896. 



