Resonance Radiation of Sodium Vapour. 457 



Other resonance phenomena show that the centres of radia- 

 tion in the sodium molecule are not entirely independent. 

 One of us * showed in 1905 that the vapour, when excited by 

 blue-green light, in the region of its band-absorption spectrum 

 emitted a narrow yellow band coincident with the D lines. 



More recently Strutt | has found that excitation by light 

 of wave-length equal to that of the ultra-violet doublet 

 (second member of the principal series, of which the D lines 

 form the first member) caused the emission of the 1) lines. 



In view of results which will be mentioned presently it 

 seems possible that the effect observed by Strutt may have 

 been due to traces of hydrogen in his bulbs. 



The polarization method of separating close spectrum lines, 

 which was described by one of us t in 1914, has been em- 

 ployed in the present work. The apparatus has been improved, 

 however, in a number of respects. Larger polarizing prisms 

 have been emploj^ed, and the entire polarizing system was 

 enclosed in a box and kept at a constant temperature by 

 means of an electrical heater controlled by a thermostat. 

 The thick quartz plate was re-figured and polished, as some 

 slight errors had been found which affected the purity of 

 the light delivered by the apparatus. 



The spectroscope was built up with two large portrait 

 objectives of 3 inches aperture and 24 inches focus, and two 

 5-inch prisms of flint glass. 



The plates were clamped against the tube which held the 

 eyepiece of the filar micrometer, and by moving this slightly 

 with the screw between exposures, two or more images could 

 be recorded on the same plate side by side. 



A very great advance resulted from the discovery that the 

 new." pyrex " glass made by the Corning Co. (Corning, N.Y.) 

 resists the action of sodium vapour far better than any other 

 glass in the market.. Bulbs blown from tubes of this glass 

 showed scarcely any discoloration after ten hours' heating, 

 which is sufficient to colour bulbs of soft-glass a dark brown. 



As a source of light we employed a Meker burner fed with 

 sodium by a large disk of asbestos board charged with sodium 

 chloride and rotated once in twelve hours by a clock. It is 

 important that the disk touches only the edge of the flame 

 on the side opposite to that which faces the illuminating 

 system, otherwise there is a reversal of the 1) lines, which is 

 very detrimental to the excitation of the resonance. This 



* R. W. Wood, Phil. Mag. [6] x. p. 408 (1905). 

 t Strutt, Proc. Roy. Soc. Series A, xci. p. 511. 

 X Wood', Phil. Mag. [6] xxvii. p. 524 (1914). 



