342 Prof. Wood and Mr. rvimura on Scattering and 



1 mm. by a diaphragm which could be raised by a micro- 

 meter-screw. An exposure of one minute was given: the slit 

 diaphragm was then raised 1 mm., the quartz bulb raised to 

 a red heat hy a Bunsen burner, and a second exposure 

 of fifteen seconds made. Fig. 2 (Plate VI.) shows the 

 result of the experiment. The reversed 2536 line appears 

 as a doublet and is indicated by an arrow, the faint com- 

 panion on the short wave-length side appearing to its left. 

 This was the exposure made by light reflected from the 

 cold bulb. Above it we have the exposure made with the hot 

 bulb. The light reflected from the hot bulb is seen to consist 

 solely of the short wave-length component of the doublet 

 (widened and reversed 2536 line), for which the reflecting 

 power of the quartz-mercury vapour surface is very slight. 

 The long wave-length component has disappeared entirely, 

 owing to the very low value of the reflecting- power for 

 this frequency. The width of the doublet is about 0'8 A.U. 

 It is perhaps worthy of mention that we have here a rather 

 efficient method of isolating from the total radiation of a 

 quartz-mercury arc running at a moderately high tempe- 

 rature, a single line of wave-length about 04 A.U. less than 

 that of the 2536 line of a similar lamp running at a low 

 temperature. 



This might be very useful in certain special investigations. 

 One could of course make the difference even less than 

 0'4 A.U. by operating the lamp at a lower voltage. 



The reflecting power of the dense vapour for the light 

 forming the continuous background of the spectrum of the 

 cadmium spark was also investigated. It was found that 

 the spectrum of the reflected light showed a bright line in 

 coincidence with the 2536 line of mercury. No cadmium 

 line appears at this point, and the continuous background 

 is absent or very faint on the long wave-length side of the 

 bright line, due to the low value of the reflecting power 

 for this region. An enlargement of the photograph is re- 

 produced on Plate VI. fig. 8. The quartz bulb was at a full 

 red heat during this experiment. It is evident from the 

 photograph that the spectroscope lacked sufficient resolving 

 power to show the minimum to advantage. The bright line 

 with its dark border on the right, due to the powerful and 

 feeble reflexion of these wave-lengths in the light' of the 

 continuous background of the spark, is indicated by an arrow. 

 Above is the 2536 line of mercury for comparison. 



