Liquefaction of Gases. 199 



bluish colour might be derived, it was carefully tested for that 

 substance. Paper moistened with potassium iodide and starch, 

 exposed for several hours to the action of the oxygen used, 

 was not coloured at all ; and when the oxygen was made to 

 pass through a solution of potassium iodide and starch the 

 result was the same. It remained for several weeks in an 

 iron flask, in contact with solid potassium hydroxide, and 

 was by this means completely purified from carbonic acid, 

 vapour of water, and chlorine. After these experiments, 

 there is no doubt that liquid oxygen, seen in layers of about 

 30 millim., possesses a distinctly bluish colour. This colour is, 

 moreover, quite in agreement with the absorption spectrum 

 of oxygen. It was rather strange that a colourless liquid — as 

 it was hitherto thought to be — should give such a pronounced 

 absorption spectrum, in which the absorptions in orange, 

 yellow, and red are preponderant ; but after the bluish 

 colour of liquefied oxygen was proved, this apparent contra- 

 diction no longer exists. 



I may conclude with a word or two about the colour of 

 the sky. There exist so many hypotheses on that point, that 

 I scarcely venture to add one more. But the simplest theory, 

 in my opinion, would be to ascribe that colour to the principal 

 component part of our atmosphere, which — at least in a 

 liquid state — is blue. 



On the Critical Pressure of Hydrogen. 



[These researches were published in Polish, in the Reports of the 

 Cracow Academy 1891, vol. xxiii. p. 885 ; a short German abstract 

 therefrom was printed in the Bullet. Intern, of the same Academy. The 

 following description is taken from the first-named source, and is 

 explained by figs. 2 and 3.] 



In my former researches, undertaken in 1884 and 1885, I 

 showed that hydrogen cannot be liquefied even by employing 

 the lowest obtainable temperatures and a high pressure, 

 reaching to 190 atm. ; and that it is only during the sudden 

 expansion from a high pressure that a greater or less trace 

 of liquefaction can for an instant be seen. This depends on 

 the temperature of the frigorific medium, as well as on the 

 initial pressure of expansion. As cooling agents there were 

 employed : — oxygen boiling under atmospheric pressure {t — 

 — 181°*4) and m vacuo, reaching to 9 millim. (£= — 211°*5) ; 

 also air boiling under atmospheric pressure (t= — 191 0, 4) and 

 invacuo at 10 millim. (£= —220°), as well as nitrogen boiling 

 under atmospheric pressure (£= — 194°'4) and in vacuo at 

 60 millim. (£= -213°). 



