﻿Method for examining Optical Qualities of Glass Plates. 615 



The upper part of the two beams, just before entering the 

 telescope, is displaced downwards by the glass plate, B,with 

 the result that, if the latter is properly adjusted, there will 

 appear in the field of view a sharp horizontal line separating 

 the upper and lower interfere ace bands. 



The lower part of the beams passes through the two thin 

 glass plates, G and Gi ; the former is fixed, and the latter is 

 capable of rotating about a horizontal axis through its upper 

 edge. The rotation of G l7 actuated by a long lever and 

 micrometer screw, causes a change in the optical path of the 

 lower right-hand beam and thus produces a bodily shift of 

 the lower set of bands ; the upper bands, as in Lord Rayleigh's 

 original instrument, are merely a set of reference marks. 



Testing the Plate. 



The plate is placed at 45° to the optic axis as shown in the 

 diagram (this we will call position " a ""), with its upper edge 

 as near as can be to the horizontal separating line. The 

 upper and lower bands are now brought into coincidence and 

 the reading of the micrometer head noted ; at the same time 

 the path of the two beams through the plate is located. The 

 plate is now rotated through 180° (position "6"), and placed 

 so that the light still passes through the same parts of the 

 glass, the bands are again brought into coincidence and the 

 reading noted ; the difference between the readings is a 

 measure of the difference in the optical paths in the two 

 positions. 



The optical arrangements were such that the central bright 

 area of the interference pattern was traversed by four sharp 

 black bands, and settings were always made on a certain one 

 band. After a little practice it was found that the error of 

 setting for either observer did not exceed half a scale-division 

 from the mean of a number of settings. Since 50 scale- 

 divisions correspond to one band interval, the shift of the 

 lower band system can be measured to the 1/100 of a band ; 

 this corresponds to a retardation of one beam on the other 

 of the 1/100 of the wave-length of the light used. 



The following are a set of readings made in this manner : — 



Diff. 



Observer. 



Position- 

 of plate. 





Headings. 





Moans 





"■ 





"> 





B 



.... a 



188-0 



1880 



187-8 



187 9 



B 



.... h 



192-0 



192-0 



192-2 



192-1 



T 



.... a 



187-0 



187-0 



18fi-5 



186*8 



T 



.... b 



191-1 



1911 



191-1 



1911 



