480 Dr Hartndge, An apparatus for projecting spectra 



An Apparatus for Projecting Spectra. By H. Hartridge. 

 [Read 2 May 1921.] 



If a Thorpe replica celluloid diffraction grating of about 14,000 

 lines to the inch be mounted in optical contact with the hypotenuse 

 of a right angled glass prism, and if now a parallel beam of white 

 light be caused to enter normally at one of the other faces, then it 

 is found, while the direct beam is totally internally reflected and 

 occupies the same position as it would if the prism had no grating 

 mounted on it, that the 1st and 2nd order spectra of one side are 

 found approximately in their normal relationship in regard to the 

 direct beam, but that the spectra of the other side pass out through 

 the hypotenuse of the prism without suffering reflection. If the 

 prism be made of crown glass the yellow region of the 1st order 

 spectrum is in a straight line with the incident beam, and the 

 dispersion of this spectrum is about 54 % greater than that of the 

 diffraction grating alone. It is also found that (1) the resolving 

 power for a given aperture is increased about 41 % , (2) the purity 

 factor for a given collimator slit width is also increased by about 

 half, (3) the intensity of this spectrum for a given purity and dis- 

 persion is also considerably greater than that of the grating alone. 

 It is to this latter feature that the arrangement owes its value. 

 To make use of this spectrum for projection purposes the green, 

 blue and violet regions of the 2nd order spectrum are screened off 

 by a deep orange colour filter. 



The transmission of the spectra through the hypotenuse of the 

 glass prism, in apparent contradiction to the laws of geometrical 

 optics, is almost certainly due to the inclination of the sides of the 

 separate strip elements of the replica, which when the cast was 

 made filled the grooves of the original grating. These sides are then 

 roughly normal to the incident light rays and a number of ele- 

 mentary wavelets would therefore, on Huyghen's hypothesis of 

 diffraction, pass out of the surface, but since they would retain the 

 phase relationships of the original beam they would be in a con- 

 dition to interfere with one another, and thus form the spectra that 

 are observed. 



This prism grating would therefore seem to be in closer 

 theoretical relationship with the echelon of Michelson, or the 

 echelette of R. W. Wood, than with the ordinary diffraction 

 grating. 



