﻿Changes in Diffraction Spectra. 33 



image of the line 8 in the original negative, or in one of the 

 enlargements, might be matched against the intensity o£ one 

 of the secondary maxima of the line 1. Knowing the rela- 

 tive intensity of the line 1 and its secondary, a rough com- 

 parison of the lines 1 and 8 could be made. 



A direct Method of r le sting for Aberration. 



The presence of cubic aberration in the echelon spectrum 

 has been confirmed by direct measurements of the shape of 

 the wave-fronts as they leave the echelon. This was done 

 by measuring the very small displacements of a spectrum 

 line, in the focal plane of the telescope, when different parts 

 of the aperture of the echelon were exposed. We could not 

 reduce the aperture to less than one-third of the full width 

 without sacrificing too much the definition of the line to be 

 measured ; so groups of eleven steps were uncovered. 

 Keeping the number of apertures constant, the group was 

 m >ved across from one side of the echelon to the other, 

 while readings of the position of the line 1 in the green line 

 spectrum were made with a micrometer eyepiece. The 

 variations of these readings were plotted against the corre- 

 sponding positions of the centre of the group and a curve 

 drawn which gave a close approximation to the direction of 

 the wave-front for the central two-thirds of the aperture 

 of the echelon, so the shape of the wave -front in this region 

 could be found by integrating. The chief difficulty of ob- 

 taining a satisfactory set of readings was that the temperature 

 generally altered too much while they were being made. 



The results we have obtained show the wave-front concave 

 towards the observer, the curvature at the centre aoreeino; 

 fairly well with the observed shortening of the focus, and 

 also give clear evidence of a cubic aberration. The magnitude 

 of the aberration appears to be about *3\ at the limits of 

 the central 2/3 of the aperture, which would correspond to 

 about one wave-length for the edges of the full aperture. 

 Owing to the temperature difficulty we can only regard this 

 result as confirming the existence of a cubic aberration. 



Experiments on producing Aberration. 



In order to commence with a very simple case, a diffraction 

 grating was made with parallel wires "2 mm. diameter, spaced 

 1 mm. apart by means of two screws of millimetre pitch. 

 No. 1, PI. IV., is a photograph of the spectrum formed by 

 this grating with green mercury light. Eight apartnres 

 were exposed when the photograph was taken, so that between 



Phil. Mag, S, 6, Vol, 23. No. 133. Jan, 1912. D 



