3i6 



INFRA-RED EMISSION SPECTRA. 



of the alcohol vapor into H (red line) and CO, or CO, just as am- 

 monia shows nitrogen bands. 



Fig. 147. 

 Nitrogen (N). 



The first sample of nitrogen examined was made by heating a satu- 

 rated solution of equal parts of sodium nitrate, NaNOg, and ammonium 

 chloride, NH4CI, washed in KOH, and dried in H2SO4 and in P2O5. 



The emission spectrum of this sample of N is given in figure 148, 

 curve b. In addition to strong lines just at the end of the red we have 

 the usual band at 4.75 /x, which is unusually intense. In fact, for a 

 pressure of 5 mm. the deflections from this band were from 30 to 35 

 centimeters. 



One difficulty in making nitrogen by this method is that the oxides 

 are also formed. The pressure of the 4.75 /u, band showed that it is due 

 to CO, or traces of N in CO or COo. The problem then was to prepare 

 a small quantity of nitrogen which was free from oxides of nitrogen, 

 and also free from O and CO,. The presence of a trace of an inert 

 gas like helium did not enter the question of the origin of the 4.75 ju 

 band. Hence atmospheric nitrogen was most serviceable. To this end 

 a glass tube containing PoOg on glass wool was connected in series 

 with a gas pipette containing a strong KOH solution and a pipette 

 containing sticks of red phosphorus under water. The phosphorus 

 of course was used to remove the O, while the KOH removed the COo. 

 The air was washed back and forth for some time. The KOH pipette 

 was then replaced by a mercury gas pipette, and the gas was washed 

 back and forth for about half an hour to dry it and remove the last 

 traces of oxygen. 



