56 



Mr Filon, On the variation with the wave-length 



P 1 and P 2 , and the spectrum viewed in the telescope T of a 

 spectroscope. 



The block B was placed in a straining apparatus (Fig. 2) made 



t — i — i — i — i — i — i — i — i — i — i — i — i — r 



w 



TAT 



Fig. 2. 



for the purpose by Messrs Harling of Moorgate Street. A steel 

 lever L, the top surface of which was graduated, carried a weight 

 W whose distance from the fulcrum F was adjustable, and rested 

 on a knife-edge block K which in turn pressed on the glass block B. 

 The horizontal distance between F and K was constant and equal 

 to 3 cms., the knife edges fitting into fine notches cut into the 

 steel. 



The pressure P on the block B was then 



Wx 

 3 ' 



P = 



x being the distance from F, in cms., of the knife-edge hanger 

 which carried the weight. 



The instrument was designed to carry safely a fairly large 

 weight. The limiting values actually used were roughly 



W = 40 lbs., x = 45 cms., 



so that the pressure on the block was about 600 lbs. 



In order to ensure the uniform distribution of this pressure 

 throughout the block, which was too short in proportion to its 

 length to make it safe to assume that the stress was uniform, even 

 over the middle section, sheets of lead were introduced between 

 the block B and the compressing blocks. 



Even this, however, did not suffice to make the stress suf- 



