M. G. Quincke on Diffraction. 367 



elevations of a grating have a very considerable influence on the 

 difference of intensity of the light polarized parallel and perpen- 

 dicular to the principal diffraction- plane. The phenomenon 

 changes with the colour of the flame, the substance in which the 

 diffraction takes place, and the angle of incidence of the rays. 



Further, in front of the object-glasses of a collimator and an 

 astronomical telescope, I placed two Nicol prisms, the azimuth 

 of which could be determined with accuracy to minutes on ver- 

 tical circles. The grating was placed between the prisms. In 

 some instances the telescope was laid aside, and the eye looked 

 directly through the analyzing Nicol at the grating. The slit 

 of the collimator was usually illuminated with daylight. 



When the Nicol prisms were crossed, the illuminated slit in 

 the eyepiece of the telescope appeared black ; on inserting the 

 grating, it was illuminated, and the maxima or spectra of the 

 second class became visible. The central image of the slit ap- 

 peared variously coloured, according to the position of the Nicol 

 prisms. Viewed with one eyepiece-prism, it mostly shows a 

 dark streak in the spectrum, parallel to the Fraunhofer lines ; 

 and on rotating the analyzing Nicol to greater azimuths, with 

 some gratings this streak travels towards the red, with others 

 towards the blue. The latter case, where the component pola- 

 rized parallel to the principal plane of diffraction is greater fox 

 the red than for the blue, is the more frequent. 



The amount of rotation of the analyzer which carried the 

 dark streak through the entire spectrum varied with the angle 

 of incidence, the nature of the material of the grating-bars or 

 furrows, the fineness of the grating, and the substance in which 

 the diffraction took place. It varied from a fraction of a minute 

 to |° in transmitted light. 



In lateral spectra likewise, parallel to Fraunhofer's lines ap- 

 pear dark streaks, which with the rotation to greater azimuths, 

 according to the grating and the spectrum, go from the red to 

 the blue, or from the blue to the red. The rotation is very dif- 

 ferent with different gratings, and with different lateral spectra 

 with the same grating, and may amount to 5° or more. With 

 greater angles of diffraction the superposition of spectra of dif- 

 ferent orders disturbs the observation. 



When several parallel gratings are inserted one behind another, 

 very complicated phenomena enter, which have been partly in- 

 vestigated by Brewster* and Grovaf. With a suitable arrange- 

 ment of the gratings, the turning of the polarization-plane for a 

 determined maximum of the second class can be increased. 



Often the dark streaks do not appear in the spectrum until, 



* Phil. Mag:S. 4. vol. xxxi. pp. 22 & 98 (1866). 



t Comptes Rendus, vol. lxxii. p, 855 (1871); vol. lxxiv.p.932 (18/2). 



