ADAMS AND WIIvLIAMSON : BIRKFRINGENCE; AND STRESS 



613 



on the Newton scale, the color depending on the amount of 

 birefringence present, but if the load be uneven the color will 

 vary from place to place in bands or streaks or irregular blotches. 

 After numerous unsuccessful attempts to secure uniform 

 stress throughout the blocks of glass, the following method was 

 found to be satisfactory. The thrust was applied through two 

 hardened steel blocks {A, A in Fig. i) the curved surfaces of 

 which were in contact with the curved surfaces of the blocks 

 B, B. The adjacent surfaces of steel and glass were ground so 

 as to be as nearly plane as possible and were then ground together 

 with very fine emery. Above and below the block of glass was 

 placed a piece of thin drawing paper. With this arrangement 

 we were able to load the glass uniformly up to a pressure of 200 

 kg. per sq. cm., which was sufficient for the purpose. 



TABLE 2 

 Birefringence Produced in Glass by a Thrust of i Kg. per Sq. Cm. 



Kind of glass 



Ordinary Crown 



Borosilicate Crown. . . 

 Light Barium Crown. 

 Heavy Barium Crown, 



Barium Flint 



Light Flint 



Medium Flint 



Heavy Flint -. . , 



Extra Heavy Flint. . . , 



—h X 10~ D/W. 



Birefringence due to 



1 kg. per sq. cm. 



-2.57 X 10-7 

 -2.85 X 10-7 

 -2.81 X 10-7 



-2 .15 X 10"' 

 -3 .10 X lO'"' 

 -3 .20 X 10"' 

 -3.13 X io~^ 



-2 .67 X IO~" 

 -I .22 X IO~' 



Modulus of 



rigidity. 



R 



0.28 X 10^ 

 0.29 X 10^ 

 0.30 X 10^ 

 0.29 X 10^ 

 0.26 X lo^ 

 0.24 X 10* 

 0.22 X 10^ 

 0.22 X 10* 

 0.20 X 10^ 



Table i shows the chemical composition and optical properties 

 of the glasses used, and in table 2 are given the results of our 

 measurements. If W is the load applied to the block of glass 

 by the testing machine and D the corresponding optical path- 

 difference as determined by the graduated quartz wedge, the 

 ratio D/W was found to be constant for each specimen to within 

 the limit of accuracy of the measurements. This result is in 

 accord with the observation by Brewster^ in 181 4 that the 



* Brewster, Phil. Trans. 1814, 1815, 1816. 



