27 



perfection of annealing, in a state of compression in the 

 outer parts and of tension near the centre, and these stresses 

 affect the refractive index of the glass, causing it to assume 

 a condition of pseudo-heterogeneity, with index of refraction 

 diminishing towards the centre. 



By means of the application of the interferometer, as 

 developed by Messrs. Hilger, in figuring prisms, etc., with a 

 view to the correction of departure from homogeneity, this 

 effect may in many cases be cured, and the remaining effect 

 of double-refraction, unless unusually severe, is then not 

 important but it is naturally desirable to have a glass which 

 has a proper degree of perfection in annealing, so that correc- 

 tion for any departure from homogeneity caused by stress is 

 unnecessary. 



The most sensitive way of detecting stress in a plate of 

 glass is to pass approximately parallel polarised light through 

 it and examine it by means of an analysing nicol. If the 

 light is intense, no block of glass ever produced will appear 

 free from stress though it may be so free from this defect 

 that the greatest amount of double-refraction observed is 

 that due to pressure on the block of glass at the points at 

 which it is supported. Obviously, such perfection in an- 

 nealing is a refinement which is unnecessary, even in the finest 

 instruments. 



For general purposes, a less sensitive test is more useful. 

 The polarised light, after passage through the specimen, 

 passes through a wave-plate (giving a relative retardation of 

 one wave for sodium light) and then through the analysing 

 nicol. The specimen thus viewed against a background of 

 the sensitive first-order violet colour, any departure from 

 perfection in annealing is made evident by a change in colour 

 towards red to yellow or blue to green. The particular tint 

 observed is a direct indication of the extent of double- 

 refraction.* 



2. It is essential that the glass employed in an optical 

 instrument should exhibit a proper degree of homogeneity, 

 so that it may not vitiate the performance of the instrument. 



For certain purposes, where, from the nature of the 

 optical system, great perfection in homogeneity is not essen- 

 tial, such as the manufacture of lenses of eyepieces, very 

 cheap object-glasses, etc., selected plate glass may be used, 

 provided light passes in a direction perpendicular to the 



* The apparatus for this test was shown by Messrs. Chance 

 Brothers, at the Optical Society's Exhibition, January, 1917. 



