of the double refraction in strained glass. 838 



composition of the two specimens supplied to Herr Pockels and 

 to myself was not by any means identical. Indeed the curve 

 given by Pockels in the paper referred to, showing the variatiou of 

 G with the percentage of PbO, indicates that for glasses such as 

 S 57 which contain near 80 per cent, of PbO the variation in G is 

 extraordinarily rapid. 



Thus the amount of PbO in my glass was stated by Messrs 

 Zeiss to be about 79 per cent. The amount in the specimen 

 examined by Pockels was 80'05 per cent. It seems therefore that 

 this may quite well account for the difference, especially when we 

 remember that C was found to vanish for a glass containing about 

 74 or 75 per cent, of PbO. 



If we turn to the rate of change of C with the wave-length, 

 we find 



Here the agreement is considerably better, which would seem 

 to suggest that the dispersion of double-refraction does not here 

 vary quite so fast with the composition as the double-refraction 

 itself. Again Pockels' results and my own agree in making C 

 vary with X far more rapidly in the green than in the red ; a 

 result still more marked when we take observations in the blue, 

 as is shown in Fig. 6. 



10. In the observations of the other glasses, the curvature is 

 not so marked. 



In the case of O 41 and O 152, more especially the latter, a 

 straight line gave a decidedly poor fit, and the following formulae 

 for G were adopted. 



O 41, 

 G-- 



O 152, 



(7= 1-745 + -0156 ( X ^°)- -0236 /; 



iooo ; • 



X - 5400Y- 



^VKTootr; 



