﻿certain Gases in Geissler's Tubes. 41 7 



then, from these observations that the emissive power of a sub- 

 stance may materially alter with the temperature. 



II. Oxygen. 



8. The spectrum which a Geissler's tube filled with pure oxy- 

 gen gives consists, according to PKicker's* description, of a 

 series of bright lines, the least refrangible of which lies in the red- 

 orange, and which, more or less close, extend thence to the violet. 

 Pliicker described only this one spectrum, and in the subsequent 

 investigation, made in common with M. Hittorf, obtained only 

 the same. In communicating these experiments he says f : — 

 "We obtained only one spectrum of oxygen working in the 

 same manner as with nitrogen, with this difference, that under 

 the same circumstances an equally brilliant spectrum was only 

 obtained with a stronger discharge. He further states that, 

 especially in oxygen, the gradual appearance of the bright lines is 

 noticeable — that at first the least refrangible show themselves, and 

 at last as the temperature rises the most refrangible come out — 

 that a drawing, therefore, which represents as simultaneously 

 appearing those lines which are only successively formed, gives an 

 ideal picture of the spectrum rather than accords with nature. 



The experiments previously communicated on the spectra of 

 hydrogen, the observation that the spectrum may materially alter 

 with the density of the gas and the mode of discharge, led me to 

 investigate oxygen in this direction. 



The method of experiment was the same as that described in 

 § 3. The lateral tube of such a spectrum-tube as is there de- 

 scribed was connected with the tube placed over the anode of the 

 apparatus for decomposing water. In order to dry the oxygen 

 supplied, a tube containing phosphoric acid and a bulb filled with 

 strong sulphuric acid were interposed. The production of a per- 

 fectly pure oxygen-spectrum, however, is almost always attended 

 with great difficulties ; the three characteristic lines of hydrogen 

 were almost always observed, arising from the moisture condensed 

 on the inside of the spectrum-tube. It could, however, be ex- 

 pelled by strongly heating the spectrum-tube and then repeat- 

 edly drawing pure oxygen through it. When the tube had thus 

 been dried so that with oxygen under a pressure of from 5 to 10 

 millims. it no longer showed any hydrogen-lines, it was filled with 

 oxygen under the ordinary pressure, then gradually exhausted 

 by the SprengePs pump as described in § 3, and the spectrum 

 investigated which the gas gave at different densities. 



9. It was first found that, using a spectrum-tube of exactly the 



* Vogg.Ann. vol. cvii. 



t Philosophical Transactions for 1865, part 1, p, 23. 



