1878.] Dr. A. Schuster on the Spectra of Metalloids. 



385 



change from one line spectrum to another is due to the same cause as 

 that of a band spectrum to a line spectrum. 



The chief difficulty in the way of a complete investigation of the 

 spectrum of oxygen consists in the great disturbing influence of the 

 presence of even a small quantity of any carbon compound. Amongst 

 a great many oxygen tubes which were filled by various makers, I only 

 found one which showed the spectrum of pure oxygen ; all the others 

 gave a spectrum of carbonic oxide. It is therefore necessary in filling 

 oxygen tubes, to avoid all greased joints and all india-rubber tubings. 

 I have used a Sprengel air-pump, which communicated with the vacuum 

 tube by means of a ball and socket joint. The joint was kept air- 

 tight solely by means of strong sulphuric acid. The vacuum tube 

 was fused directly to the ball of the joint. To one end of the vacuum 

 tube a piece of hard glass tubing had been fused. This was filled 

 with different substances which, on heating, gave off pure oxygen. 

 The oxygen therefore came only into contact with glass, mercury, and 

 sulphuric acid, and the metal of the electrode. By repeatedly filling 

 and exhausting the tube all extraneous matter, which always is 

 attached to the inner surface of the glass, could be swept away. Per- 

 manganate of potash, oxide of mercury, and chlorate of potash, were 

 used in turn to prepare the oxygen, but no effect was observed which 

 could be traced to the substance used. The effect of the electrodes 

 was eliminated by varying the metals. Aluminium, platinum, silver, 

 brass, and iridium were used as electrodes. Any possible effect of the 

 glass was eliminated by finally repeating all experiments in a glass 

 receiver six inches in diameter, so that no part of the spark came 

 nearer than 2| inches to the glass. In this way it is believed all pos- 

 sibility of error due to the presence of any possible impurities was 

 avoided. 



All measurements were reduced to wave-lengths by interpolation. 

 The lines of iron or any metal whose lines are reversed in the solar 

 spectrum were taken as reference lines. The wave-length of the re- 

 ference lines could therefore always be found on Angstrom's solar 

 map. As a rule the dispersion was equal to that of four heavy flint- 

 glass prisms of a refracting angle of 62°, and it is believed that the 

 accuracy of the measurements equals that of any other spectroscopic 

 measurements. 



Four different spectra of oxygen must be distinguished. At the 

 lowest temperature at which oxygen becomes luminous it gives a 

 continuous spectrum. As the temperature is gradually raised the 

 continuous spectrum is successively transformed into two distinct line 

 spectra, which I call respectively the compound line spectrum and the 

 elementary line spectrum. It' is one of the principal objects of this 

 paper to show that these two line spectra which have been much 

 mixed up together have a separate existence. The generation of one 



