•158 Prof. Wood and Mr. Moliler on Resonance Radiation 



arrangement -worked quite as well as the more complicated 

 atomizing apparatus employed in the earlier work. 



The sodium bulbs were prepared in the same way as 

 before, except that the sodium was distilled from one side 

 of the bnlb to the other a great many times by waving the 

 flame of a burner around the bulb, the pump working con- 

 tinuously. We found that even after the sodium deposit had 

 been driven across the bulb 170 times, the M c Leod gauge 

 still showed a small trace of hydrogen. 



We have photographed the spectrum of the resonance 

 radiation excited by both D 2 and D 1# In the earlier work 

 we experimented only with the light of the D 2 line on 

 account of its greater brilliancy. 



The efficiency of the polarizing separators was tested by 

 photography, and it was estimated that when the quartz 

 block was set for the extinction of D 1? D 2 was about 50 times 

 brighter than D 1 in the transmitted light. Adjustments of 

 the apparatus were made by coating the bulb with a layer 

 of magnesium oxide, by holding it over burning magnesium 

 wire. The coating was then wiped off with the exception of 

 a narrow vertical strip. The bulb was mounted in a wire 

 frame as before over a chimney heated by a quadruple 

 burner, in such a manner that it could be rotated on a vertical 

 axis. The image of the small window in the sheet-iron 

 chimney surrounding the sodium flame, formed by the 

 polarizing separator, was focussed on the patch of MgO, and 

 an image of this thrown on the slit of the spectroscope. The 

 quartz block was then turned until one only of the D lines 

 appeared in the spectroscope. The bulb was now rotated 

 until the image fell on the clear glass to one side of the 

 oxide patch, and the quadruple burner lighted. 



With the improved apparatus we could give exposures of 

 ten or twelve hours, which was sufficient to yield a fullv 

 exposed spectrogram of the resonance. 



In the case of the first photographs which we made we 

 found that there was always a trace of the D 1 line when the 

 illumination was by D 2 . 



All possible precautions were taken to exclude scattered 

 radiation, and to still further eliminate every chance of error, 

 we added a narrow horizontal strip of dark grey Mg oxide 

 which cut across the patch of resonance, and matched it in 

 intensity. This grey patch we made by heavily smoking the 

 bulb in a gas-flame and then depositing a very light coat 

 of MgO. The bulb with the large white patch and the 

 narrow grey strip is shown in fig. 1 A. Fig. 1 B shows 

 the appearance of the resulting spectrum line as it appeared 



