210 



Mr. G. J. Burch. On the Relation of 



(1.) Place a coloured screen over the end of the scale tube of an 

 ordinary spectroscope, the scale being removed, and cover the slit 

 with a card. After looking at the coloured light for some seconds, 

 cover the scale tube, and simultaneously uncover the slit. This ex- 

 periment is easily tried, but is open to the objection that the first 

 stimulus is not perfectly monochromatic. 



(2.) Illuminate the scale tube with monochromatic light from a 

 prism, and proceed as before This plan obviates the difficulty 

 referred to, but does not afford sufficient light to produce the ful 

 -effect. 



(3.) A single prism A, fig. 2, with collimator B, and slit C, is fixed 



ria. 2. 



near the back surface of the last prism D of the large spectroscope in 

 such a position that the rays from it are reflected into the telescope E 

 of the large spectroscope. This second spectrum is of course much 

 fainter than the one observed directly through the instrument, but 

 that is an advantage rather than otherwise. It is only necessary to 

 arrange two black cards with slits in them in such a way that when light 

 passes through the first spectroscope the second is obscured, and on 

 touching a spring the conditions are reversed. 



In order to make the effects more marked, a short slit should be 

 used for the large spectroscope, so that it may give a band of mono- 

 €hromatic light across the middle of the field, fairly bright but rather 

 narrow. The eye should be fixed on the centre of this band. After 

 a few moments, by the action of the spring referred to, a black card 

 is suddenly brought over the slit of the spectroscope, shutting off" the 

 light, while at the same moment the screen is removed from the other 

 spectroscope, and the complete spectrum appears, filling the entire 

 field of view. For an instant, a dark shadow is seen, not extending 



