﻿26 



Dr. H. Stansfield and Mr. H. P. Walmsley on 



grating, and that with a suitable amount of aberration four 

 or five secondary maxima may be visible on one side of a 

 bright line and none on the other. 



In the case of the echelon grating we have employed 

 the aberration is apparently produced by the clamping of the 

 glass plates. We find that the echelon produces a slight 

 curvature in wave-fronts or surfaces of equal phase passing- 

 through it. The light which issues from the small end of 

 the echelon, and has passed through the parts of the plates 

 which are most compressed, is slightly in advance of that 

 which has passed through the plates farther from the 

 clamping rods, as indicated diagrammatically in fig. 1, the 

 position of the rods being indicated by the line CC. 



Fi<r. 1. 



If the curvature of the surface of equal phase were con- 

 stant it would only produce a shortening of the focus of the 

 observing telescope, which is a well-known feature *. The 

 curvature increases, however, towards the side on which the 

 plates are clamped, so that the dotted circle in fig. 1, coin- 

 ciding with the middle of the surface, is in advance of the 

 surface on the observer's left, and falls behind it on the 

 right. 



We find that the deviation of the surface from the circle 

 can be nearly represented by a cubic aberration. 



Calculation of the Effects of Cubic Aberration. 

 In order to find the effects which would be produced by 

 cubic aberrations of the wave-front the distribution of light 

 * Phil. Mag-, vol. xviii. p. 372 (1909). 



