876 
MESSRS. W. E. WTLSOX AND P. L. GRAY 
Or, omitting tv.m obviously bad observations, the mean difference between “ calcu¬ 
lated ” and “ observed ’’ 
+ 24-81 +3 
26 ~ 26 
+ 0 - 1 . 
The latest work on this subject is that of Paschen,* who gives full references to 
the papers of other experimentalists. His method of working is very complicated. 
* ‘Wiedemann’s A.niialen,’ rot 49, 1893, p. 60. 
