480 Mr Filon, On the variation with the wave-length 



The scale s, which consisted of a network of squares of 5 mm. 

 side, and carried besides one horizontal and one vertical millimetre 



> 



w 



D* 



^N- 



IT 



Fig. 2. 



scale, was ruled for me on glass with a dividing engine by Mr Edwin 

 Wilson, of Mill Lane, Cambridge. It was viewed through a 

 telescope IT, in front of which was placed an analysing Nicol i\T. 



The slit T could be illuminated, either by a powerful arc lamp, 

 or by a spark between metallic poles in air. An alloy of Lead, 

 Tin and Cadmium was employed for this purpose. The use of 

 this alloy was suggested to me by Mr H. C. Ramage, of St John's 

 College, to whom I am also indebted for finding out for me the 

 wave-lengths of a large number of characteristic lines in the spark 

 spectrum of this alloy. This spectrum has a large number of 

 bright lines fairly uniformly distributed from the red to the 

 extreme violet and is accordingly very useful for comparison pur- 

 poses. In the diagram (Fig. 2) the spark holder is in position 



