Intelligence and Miscellaneous Articles. 447 



ON THE ABSORPTION-SrECTRUM, AND THE COLOUR OF LIQUID 

 OXYGEN. BY K. OLSZEWSKI. 



In an earlier research the author found four absorption-bands in 

 the spectrum of liquid oxygen corresponding to the wave-lengths 

 628, 577, 535, and 480. Liveing and Dewar (Pliil. Mag. [5] 

 vol. xxvi. p. 286), who observed the absorption-spectrum of gaseous 

 oxygen in a long steel tube under high pressure, found the same 

 four absorptions in the visible part of the spectrum, and moreover 

 in the extreme red the bands corresponding to the Fraunhofer 

 lines A and B, which have also been observed by Egeroif and 

 Janssen. An apparatus recently constructed by the author for 

 liquefying a large quantity of oxygen enabled him to repeat his 

 former experiments, and to examine more closely the absorp- 

 tion-spectrum of a thicker layer of liquid oxygen in the extreme 

 red. 



The liquid oxygen was poured from the generating vessel into a 

 thin glass tube closed at the bottom, which as a protection against 

 external warmth was closely fitted in a set of three beakers. The 

 thickness of the column of oxygen was 30 millim., and its height 

 about 50 millim. The liquid oxygen maintained itself for half an 

 hour under the pressure of the atmosphere, and at its boiling- 

 point ( — 181°'4) in a sufficient quantity for making the experiment, 

 although a considerable quantity of heat was imparted to it, 

 especially by the limelight concentrated by a condensing lens, which 

 the author used for preparing the absorption-spectrum. A uni- 

 versal spectroscope by Kriiss, with a Rutherford's prism, was used 

 for investigating the absorption-spectrum. Besides the four 

 absorptions already observed the experiments gave a fifth shadowy 

 band, corresponding to the Fraunhofer line A, which was especially 

 distinct when a red glass was interposed between the source of 

 light and the slit of the spectroscope. This band appeared feebler 

 than the three absorptions corresponding to the wave-lengths 628, 

 577, and 480, but stronger than the absorption 535. With this 

 comparatively small dispersion the band A could not, of course, be 

 resolved into lines. An absorption corresponding to Fraunhofer's 

 line B could not also this time be observed. 



Liquid oxygen has been described as a colourless body on the 

 basis of the experiments made in 1883, in which only small 

 quantities could be obtained. Since then the author has observed 

 that whenever oxygen was liquefied in rather wide glass tubes, it 

 showed in incident light in layers of about 15 millim. a bluish 

 colour. In the experiments described above, in which for the first 

 time a relatively large quantity of liquid oxygen was collected in a 

 glass vessel, its bright blue colour was decidedly apparent. In 

 order to be certain that the oxygen prepared from chlorate of 

 potassium and manganese did not contain traces of ozone from which 

 the colour might arise, it was carefully tested in this direction. 

 Iodide of potassium and starch-paper was not coloured when kept 



