NOTES ON A LARGE CRYSTAL SPHERE. 379 



Consider the case of a ray falling normally on the surface of a 

 double-refracting sphere. Then whatever be the course of the ex - 

 traordinary ray within the sphere, it must clearly emerge normally 

 also. Hence the apparent deviation of the extraordinary ray from 

 the original direction of the incident ray (a quantity very easily 

 measured) will be twice the true deviation due to one refraction. 

 The greatest possible deviation of the extraordinary ray from its 

 course at normal incidence was, in this way, found to be 19' 45". 

 This gives for the ratio of the two refractive indices 1.00576; Rud- 

 berg gives for quartz 1.006. 



The possibility of obtaining such a result from optical measure- 

 ments which, at the time, were made from simple curiosity with no 

 studied attempt at scientific accuracy, is I think an indication of the 

 remarkable purity and exquisite finish of the mineral which it fell to 

 my good fortune to have in possession for a few days. 



