Wave-length of Light. 359 



The micrometer-constant here used was that of the new 

 micrometer, where 1S = 0'257. 



In precisely the same way grating II. was rerneasured, the 

 six series giving the following relations : — 



cm. S. 



5G = 15 S a 2 + 157-4"] 

 5G = 15 S a 2 + 154-9 | 



5G = 15 S a 2 + 154-9 

 5G = 15 S a 2 + 162-4^ 



Combining and reducing these results as before, we have the 

 equation 



42640 spaces=5G = 15 centim. S B ,+39*9 at 20°. 



The temperature-variation in the two days of observation was 

 only 0°-2. 



Gratings III. and IV. were then measured. In this case a 

 large number of comparisons were obtained at both high and 

 low temperatures with the object of detecting any differences 

 which might exist between the coefficients of expansion of 

 the gratings and those of the speculum-metal standards. III. 

 and IV. being a little over a decimetre in length were very 

 easy to measure, particularly since the lines were very sharp 

 and of approximately the same width as those on the 

 standards. 



III. proved to have sensibly the 'same coefficient as the 

 standards. I subjoin the comparisons made at or very near 

 20°:— 



G = Dm^ 



) a 2 + 32-9 



G = „ 



-+33-0 



G= „ 



+ 32-7 



G= „ 



+ 33-2 



G= „ 



+ 32-3 



G= „ 



+ 32-6 



G= „ 



+ 34-5 



G= „ 



+ 33-4 



G = „ 



+ 34-2 



G= „ 



+ 32-6 



Combining these and other series of observations gives 

 finally 



28418 spaces = G = DimS^-hS^ at 20°. 



It should be noted that the extreme variation in the above 

 2 B2 



