480 On the Resonance Spectra of Iodine Vapour. 



may be. The Savart fringes were formed as before, cutting 

 vertically across the horizontal cone of fluorescent light 

 excited by the polarized light from the mercury arc. A 

 screen of black paper, perforated at its centre with a horizontal 

 aperture about 3 cm. in length by 2 mm. in width, Mas now 

 mounted between the bulb and the Savart plate in such a 

 position that the aperture covered the brightest part of the 

 fluorescent cone. On looking through the analysing nicol 

 the aperture was seen brightly illuminated except for the 

 dark regions where it covered the Savart fringes. A direct- 

 vision prism of high dispersion was now held before the eye, 

 oriented in such a way as to give vertical dispersion. A 

 vertical spectrum, or a series of images of the aperture 

 corresponding to the brightest lines 

 of the resonance spectrum was clearly Fig. 2. 



seen, and the fringes cut across all 



of these images (tig. 2) and were 

 quite as distinct in the images 

 corresponding to wave-lengths not 

 present in the mercury arc as in 

 those due to light which was re- 

 emitted from ihe vapour without 

 change of wave-length (resonance 

 radiation). This method was far 

 more satisfactory than the one em- 

 ployed for the same purpose in the 



case of sodium vapour, and gave results about which there 

 could be no question. The addition of helium appears to 

 diminish the percentage of polarization, for in the bulb 

 containing helium at a pressure o£ 10 mm., the percentage 

 of polarization was found to be only 13, against 17 when the 

 iodine vapour was in vacuo. In helium at 10 mm. the 

 resonance spectrum is completely replaced by the band 

 spectrum, and the fact that polarization is still present 

 indicates that the polarized vibration is passed over to the 

 other electron systems. 



Resonance Spectra with the Echelon. 



Through the courtesy of Mr. Twyman, who placed a 

 number of very fine echelons made by the Hilger Company 

 at my disposal, I was able to make a preliminary study of 

 the appearance of resonance spectra under high resolving 

 power. The first inspection was very discouraging, for the 

 width of the lines turned out to be as great as, or greater 

 than, the distance between the orders. 



