( -<^>2 ) 



exainiiie it while the light passes or does nut |)ass through an 

 absorbent snbstanee. The apparent displaeenienl of tlie bajid near 

 the limit of absor[)lion must tlieji immediately a[)pear. 



For my measurements with [)otassium ferrieyanide I iiave made 

 use of rotations of IP and higher. A quartz j)late 0.4 m.ni. thick, 

 out at right angles to the optical axis was now nsed and with it a 

 similar rotation is obtained near the limit of absorption. This plate 

 preceded and followed by a nicol was placed between the collimator 

 and the experimental tubes, \vhich moreover were monnted in precisely 

 the same way as they were for measurements of the rotation in 

 potassium ferrieyanide. A large nnmher of adjustments have been 

 made l\v rotating one of the nicols, one set where the experimental 

 tube was fdled witii a V/., "/„ solution of potassium ferrieyanide, and 

 one with water instead of the salt-sobition. The calibration of the 

 spectrum was made as before with a mercury arc lamp. The following 

 values have thus been obtaijied, as means of pairs of adjustments: 



(limit of absorption, about 481) 



