GASES IN VACUUM-TUBES. 



313 



curve shows that with constant current the intensity passes through a 

 maximum for a decrease in pressure of the gas, being a maximum at 

 about 5 mm., which agrees with the work of Langenbach (loc. cit.) for 

 the visible spectrum. 



Oxygen (O). 



The first sample of oxygen, made by heating KC10,+MnO„ showed 

 the impurity band at 4.75 /^ and no further examination was made of it. 



A sample of electrolytic-oxygen was then taken from a large gen- 

 erator and washed in KOH and dried over P^O^ on glass wool. Two 

 examinations were made, on different days, but no emission band could 

 be detected at 4.75 /x. In the region of i ^ no deflections greater than 

 1.5 mm. were recorded. As a whole the gas showed no emission lines 

 for the region examined, which was to 5 fi. This is of considerable 

 interest in what follows on CO and CO,, where there is a strong emis- 

 sion line at 4.75 /x. 



Carbon Dioxide (CO2). 



This gas was made from KCIO3 -f H^SO, and dried in P2O5. In 

 figure 144 are given the emission curves of CO, for a pressure of 1.2 

 mm., curve a, and 0.2 mm., curve b. It will be noticed that there are no 



Fig. 144. 



emission lines until we arrive at 4-75/^, curve d, pressure 0.6 mm., 

 which is for a different sensitiveness than a and b. Curve e is for air at 

 pressure of i mm., while the curve c, pressure 0.9 mm., is also for air, 

 found by Drew (loc. cit.), using the old radiometer. 



Curves m m, are for the Lummer form of mercury lamp, in which 

 the window is simply the end of the glass tube, which is ground and 

 polished ; hence the intensity is greatly reduced. 



The current through the gas varies with the pressure. An examina- 

 tion of the 4.75 fx. band was made for constant current and variable 



