ANNEALING STRAINS IN GLASS AND THEIR DEGREE OF PERMANENCE*

A bstract A statistical comparison has been made of 13 sets of measurements of optical path difference, taken over a seven‐year period, on 24 samples of the standardized disks distributed in 1938 to the manufacturers of glass containers. These data show that the original optical retardation of about 22.8 millimicrons has not changed by more than =1.3%, equivalent to =0.30 millimicron or =0.10° on the scale of the measuring instrument. Five tests were made in 1938, five in 1945, and three between. The standard deviation of the combined errors for each of the 13 averages of 24 measurements is =0.0338° and is composed of an accidental error of observation of =0.005° and an error of adjustment of the instrument of =0.0333°. The range of =1.3%, or =0.10°, beyond which the disks could not have changed over seven years, represents the 3‐sigma probability limit of combined errors. While this possible change of =1.3% would be insignificant in relation to the utility of the standardized disks, it represents a possible experimental error too great to permit broad conclusions concerning the ultimate stability of birefringence in strained glass.

Hali tarjima qilinmagan