1024 Separate Excitation of Emission of D Lines of Sodium. 



resonance radiation from the central patch was often no 

 brighter than that from the lateral patches which were 

 excited by D x light. It was found, however, that i£ the 

 amount of sodium in the exciting flame was reduced, the 

 lateral spots of resonance radiation diminished in intensity, 

 while the central one changed scarcely at all, retaining a 

 brilliancy of about double that of the lateral ones. This will 

 be easily understood from what has just been said about the 

 greater intensity of the D x line in the resonance radiation 

 excited by a brilliant flame. 



Photographs showing this phenomenon are reproduced on 

 PL XVII. figs. 5, 6, and 7. Fig. 5 is a photograph of the 

 three patches of exciting light thrown on a bulb covered with 

 magnesia. The central one, which contains the D 2 light, is 

 tw T ice as bright as the lateral ones (Dx). 



In fig. 7 we have the photograph of the resonance radiation 

 from the bulb under the same conditions of illumination. All 

 three strips have approximately the same intensity. Pig. 6 

 was made under the same conditions, except that the amount 

 of sodium which the atomizer fed to the flame was greatly 

 reduced. Here we have practically the same intensity ratio 

 in the case of the resonance radiation as obtains in the case 

 of: the magnesia. All of these results are easily explained by 

 the circumstance that D 2 is more easily reversed than D : . 

 The white spots of light in figs. 6 and 7 are due to regular 

 reflexion from the glass walls of the bulb. 



The light of the exciting flame was examined with a very 

 powerful echelon of 20 plates in optical contact, each plate 

 15 mm. in thickness. This instrument was loaned by the 

 kindness of Mr. Twyman of the Adam Hilger Co. The D 

 lines were examined separately by interposing the polarizing 

 separator between the flame and the instrument. It was found 

 that D 2 showed a faint trace of reversal, even with the mini- 

 mum quantity of sodium in the flame. D 1? however, reversed 

 only when the amount of sodium was considerably increased. 



The resolving power of the echelon was about 300,000, and, 

 judging from the ratio of the width of the lines to the distance 

 between the spectra of adjacent orders, the total observable 

 width was about 0*13 Ang. 



The absorption of the vapour in the glass bulbs was also 

 examined with the echelon, employing the flame as a source. 

 A distinct increase in the reversal of the D 2 line was observed, 

 when the temperature of the bulb reached 120°. The 

 diameter of the bulb was only 5 cm., and it is probable that 

 with an absorption-tube one metre in length the absorption 

 could be detected at a temperature not very much above the 

 melting-point of the metal. 



