222 44 



usually sufficient to pass the gas twice through the agent in order to ohtain the 

 result aimed at by the experiment. 



With respect to the glass apparatus K, it was a hard glass tube containing 

 the before mentioned Mg-CaO-Na mixture when nitrogen was to be absorbed. 



When hydrogen and carbon monoxide were to be destroyed, the combustion 

 tube K contained copper oxide; in this case, moreover a glass tube containing 

 solid hydroxide of potassium was inserted to absorb the carbon dioxide formed 

 by the combustion of carbon monoxide. The temperature of the copper oxide was 

 generally allowed to fall below red heat before the gas was pumped out of the 

 tube for the last time. 



To separate argon and helium, the tube K had the form indicated in Fig. 11. 

 The little bulb contained charcoal, made by bringing fragments of cocoanut shells 

 to a great heat. The bulb was environed by liquid air before the gas was admitted. 

 Then argon, xenon and krypton are, as Dewar has found, retained by the charcoal, 

 while helium and neon are only slightly absorbed by the charcoal at the temper 

 ature of liquid air. 



By means of porous contact in the mercury cup Q, the gas contained in the 

 tubes P, K, R could be introduced into the spectrum tube S, so that it could be 

 investigated in the spectroscope. As expected, the investigation with the spectro- 

 scope showed that the gas not absorbed by the charcoal contained helium. Besides 

 this spectrum of helium, the spectrum of neon was in many cases very conspicuous, 

 showing that the spring gases in question contain neon in quantities comparable 

 with the amount of helium. 



In some of the spectra investigated, some lines due to mercury and hydrogen 

 also appeared, but they were only visible in the beginning when the pressure of 

 the gas in the spectrum tube was very small, and disappeared under increased 

 pressure. They were always much fainter than the strong lines due to helium 

 and neon. I concluded from this that the amount of hydrogen contained in the 

 remaining gas was so small, in proportion to the amount of helium and neon, that 

 it could be considered as a negligible quantity in the consecutive measurement of 

 the volume of the helium-neon mixture. Having collected the helium-neon mixture 

 in a small collecting tube, the liquid air was removed, and by the consequent 

 heating of the charcoal the greater part of the gas absorbed became free. The last 

 remnant of the gas was freed by heating the charcoal to some two hundred degrees 

 by means of the spirit lamp. 



To measure the small volumes of the gases in question, I employed a volumeno- 

 meter, — originally designed by Professor K. Prytz, — in a slightly modified form. 

 The apparatus as I used it is illustrated by Fig. 12. The two vertical glass tubes 

 A and B are provided with divisions, and communicate underneath; but their upper 

 ends are closed by the porous plugs M and N, which are covered by mercury. By 

 raising the mercury reservoir Q, which is connected by the indiarubber lube L to 

 the measuring tubes A and B, the air is expelled from the tubes. The remaining 



