ADAMS AND WILLIAMSON: BIREFRINGENCE AND STRESS 



621 



vestigated by Pockels (see table 3) g is greater than p and there- 

 fore this glass when acted upon by a thrust shows positive 

 double refraction. ^^ Since the next flint glass in the series 

 (PbO = 67.5 per cent) exhibits negative birefringence, it is evi- 

 dent that there must be some intermediate flint glass for which 

 the birefringence would be zero and which therefore would re- 

 main isotropic under the influence of any elastic deformation. 

 In attempting to make such a glass the composition 74.6 per 

 cent PbO, 23.6 per cent Si02, 0.6 per cent (KoO -|- Na20) was 



l-saej 0x1 ds fsr CenT 



Fig. 4. 



Graphical representation of the results for the flint glasses. The abscissa is % 

 PbO and the ordinate is the birefringence caused by a thrust of i kg./cm^. 



tried by Pockels, and was found to give positive birefringence 

 of 0.2 X 10 ~^. This datum has also been plotted in Fig. 4, 

 from which it can be seen that a glass containing about 74.0 

 per cent PbO would, no matter how poorly annealed, show no 

 birefringence. 



APPLICATION TO PROBLEMS DEALING WITH GLASS 



As an example of the way in which the results given in this 

 paper may be used, suppose that a sheet of ordinary crown 

 glass be observed between crossed nicols and illuminated with 

 white light, and suppose that in looking through 10 cm. of the 

 sheet the interference color in a certain part of the cross-section 

 is seen to be a red of the first order. In order to ascertain the 



15 It should be remembered that in Equations (3), (4) and (5), P is a thrust. If 

 P be a tension most glasses will show positive double refraction. 



