﻿Changes in Di fraction Spectra. 25 



on the wave-length and on the material. This requires a 

 long preliminary investigation for every glass used. Jn the 

 second place, it' the field is to he adequately covered, a 

 sufficient number of isochromatic lines must be brought into 

 view. This involves the production of a very high stress, 

 sometimes dangerously near the breaking-point. This must 

 cause some plastic flow and permanent set, and so renders 

 the method unsuitable for testing the results of the theory of 

 elasticity. 



On the other hand, the use of the isoclinic lines does away 

 with these objections. The same results are obtained, what- 

 ever the photoelastic constants of the material, and a very 

 moderate stress will show the effects. Indeed, to avoid con- 

 fusion in the greater number of the experiments described 

 above, the load was kept such that the isochromatic lines 

 generally did not appear in the field of view. 



In conclusion, I have to express my most hearty thanks to 

 Sir Alfred Ewing, K.C.B., F.R.S., then Professor of Engi- 

 neering in the Univei>ity of Cambridge, for all his kind help 

 and encouragement during the course of the work. 



II. Changes in Dijjraction Spectra produced by Aberration 

 and the Aberration of an Echelon Grating. By Herbert 

 Stansfield, D.Sc, Assistant Lecturer and Demonstrator 

 in the Physical Laboratories, Manchester University, and 

 H. P. Walmsley, B.Sc, Graduate Scholar*. 



[Plates III. & IV.] 



| N the course of some previous experiments with an 

 JL echelon grating, it was found that a bright line in the 

 mercury green line spectrum was accompanied by a series 

 of faint fine lines on the longer wave-length sidef. Experi- 

 ments made to test the origin of these faint lines proved that 

 they were secondary maxima of the same wave-length as the 

 bright line, for they moved farther away from the bright 

 line when the number of apertures of the echelon was 

 reduced. 



Lord RayleighJ has shown that a cubic aberration of a 

 continuous wave-front increases the first secondary maximum 

 on one side of the image and decreases the one on the other 

 side. We find that similar results hold for a diffraction 



* Communicated hv the Authors. 



t Phil. Mag. vol. iviii. p. 371 (PJ09). Proceedings, Physical Soc. o( 

 London, vol. xxii. p. 822 (1910). 



\ Published Papers, vol. i. p. -1-30 ; Phil. Mag, vol. viii. p, 406 (1870). 



