368 M. G. Quincke on Diffraction. 



simultaneously with the grating, a mica plate of - is put between 



the Nicol prisms in a suitable azimuth. The diffracted light is 

 then elliptically polarized. For individual gratings the differ- 

 ence of phase of the components polarized parallel and perpen- 

 dicular to the principal diffraction-plane can be determined by a 

 Cabinet compensator. 



Still more striking than in transmitted light are the pheno- 

 mena when, in the azimuth +45°, linearly polarized light is re- 

 flected from a grating, especially a silvered furrow or elevation- 

 grating. In the spectrum of the central image, or the side 

 spectrum of the second class, with the Nicol prisms in a certain 

 position, one or more dark streaks then make their appearance, 

 which when the Nicols are rotated travel from one Fraunhofer's 

 line to another or disappear. Their site varies with the form, 

 distance, and material of the furrows or elevations, the angle of 

 incidence, and the substance in which the diffraction takes place. 

 Furrow- and elevation-gratings of symmetrical form again show 

 the same phenomena. Deviations are to be accounted for by 

 small differences in shape of the furrows or elevations, which 

 very considerably influence the phenomenon. 



Between crossed Nicol prisms, a grating shows in the trans- 

 mitted or reflected light secondary maxima with someFraunhofer's 

 lines which without them are not perceived. They are variously 

 coloured, according to the azimuth of the analyzing prism. 



Different gratings show quantitative, but not qualitative dif- 

 ferences, as I have found by numerous measurements, which 

 will soon be given in another place, where the labours of other 

 observers will also be described. 



Abstracted from all theoretical considerations, the experiments 

 showed : — - 



1. Linearly polarized light gives in general, after diffraction, 

 light elliptically polarized. 



2. The difference of phase and ratio of amplitude of the com- 

 ponents polarized parallel and perpendicular to the principal 

 diffraction-plane vary, the angle of incidence being the same, 

 with the order of the spectrum, so that with an increasing 

 angle of diffraction they may become greater or less. An in- 

 crease or diminution, however, may be succeeded by a dimi- 

 nution or an increase, and so forth. 



3. The increase or diminution is very different for differeut 

 colours ; and, under otherwise similar conditions, one colour 

 may show an increase, another a diminution. 



4. If the difference of phase of the two components polarized 

 parallel and perpendicular to the principal diffraction-plane is 

 small, in the diffracted light a rotation of the plane of polar- 



