4 Gustaf JV. Elmen 



A thermometer was placed in the Hquid and read before and 

 after observations were taken. The measurements were taken at 

 room temperature. 



The order of the compensator was found by comparing it with 

 a Lj darter-wave plate. The compensator was set for a match, and 

 the quarter-wave plate placed in the field so that it produced no 

 effect. The compensator was now rotated through an angle of 

 five or ten degrees and the quarter-wave plate rotated until the 

 intensities of the two halves of the field were again the same. 

 Then/ if A"', 6' and N. 9 are the orders and the degrees respect- 

 ively through which the two plates are rotated 



(4) or iV=^y-. 



Having thus found the order of the "compensator" for a cer- 

 tain wave-length, its order for any other wave-length was found 

 by means of the curve of differential dispersion of mica.^ ' 



C3 



The wave-length for which the quarter-wave plate produced a 

 retardation of A/4 was determined in the following manner : A 

 beam of sunlight polarized at forty-five degrees to the principal 

 axis of the plate was passed twice through the plate and then 

 through a second Nicol with its plane parallel to the polarizer. 

 The beam was analyzed by means of a spectroscope. The spec- 

 trum contained a black band corresponding to a retardation of 

 half a wave-length. The plate used was found to correspond to 



^D. B. Brace, Phys. Rev., vol. 18, p. 73, 1904. 

 2E. J. Rendtorff, Phil. Mag., May, 1901, p. 545. 



