REFRACTION OF LIGHT, REFRACTOMETRY AND INTERFEROMETRY 



is obtained by successive reflections on the (85). Small size and portability are achieved 



front and the rear face of a thick mirror, by folding the Rayleigh interferometer and 



The writer had the opportunity to work with utilizing an auto-collimating optical system, 



a reference mstrument kept at the Paris Despite the apparent simplicity, the meas- 



Institute of Optics, in which the separation urements are subject to numerous causes of 



is 50 mm between beams. Despite the ap- error controllable only with great difficulty, 



parent simplicity, the adjustment of this ap- One drawback of Rayleigh 's interferome- 



paratus is extremely laborious. If the mirror ter is its very low luminosity. Svenson found 



faces are not parallel to better than a small that considerable light gains accrue if the 



fraction of a wavelength, the fringes seen are single entrance slit is replaced by a grating, 



due to defects in the mirrors, and the desired To understand the production of interfer- 



non-localized fringes vanish. ences under these conditions, it must be re- 



The interferometer built by Lord Rayleigh membered that all trajectories represented 

 in 1896 is a direct adaptation of Young's ex- by the light rays are reversible. Thus, illu- 

 periment, to which was added an Arago minated slits may take the place of the 

 Compensator. In the path of light were dis- bright fringes in the plane of the normal in- 

 posed narrow gas tubes, 40 cm long, termi- terference pattern. One such slit in the posi- 

 nated with parallel faced glass windows, for tion of the achromatic fringe produces an 

 the study of gases. It is with this apparatus image in the place of the usual entrance slit, 

 that Rayleigh discovered Argon (83). Another slit in the position of the fringe of 



The standard Rayleigh interferometer has order One will produce an image displaced 



proved its value in a voluminous literature, laterally in the plane of the entrance slit. 



It has also proved its inherent encumbrance The required spacing of the entrance slits is 



and difficulty of use. In order to obtain easily calculated. The fringe width, L, is 



sharp fringes, this instrument requires an given by the fundamental formula: 



extremely powerful source of Hght and a very = f.\/w (A6) 

 narrow entrance slit. Spectral dispersion 



measurements are difficult on account of the where d is the spacing of Young's slits, and 



low luminosity. / is the focal length of the collimator lens. 



A very important improvement was made The proper spacing S of the slit is 

 by W. E. Williams (84), basing his work on S = k-L\ (47) 

 the design of the Michelson star interferome- 

 ter (an instrument of gigantic proportions when k is a small integer, 

 with a 30-meter base, or interfering slit dis- Another variant of the Rayleigh inter- 

 tance, used to measure the diameter of dis- ferometer is the Linnik instrument, recently 

 tant stars). In the Williams instrument, the manufactured in the USSR. This is a large 

 interfering slits are extremely close together, size industrial device in which the phase 

 thus permitting the use of a rather large, separation system is ingeniously adapted to 

 very luminous entrance slit. Increased sep- the verification of extended flat surfaces, a 

 aration between the two beams is obtained feat heretofore considered the prerogative of 

 by means of an Albrecht rhomb. A similar, amplitude-splitting types of instruments, 

 but reversed rhomb, placed before the tele- 

 scope objective, re-combines the two beams 

 and produces the interference pattern. The laws of interference of polarized light 



Another variant of the Rayleigh device is had been formulated by Fresnel and Arago. 



the Haber-Lowe interferometer, manufac- To interfere, it is essential that the two rays 



tured by Zeiss, and popularized by Hirsh of light, in addition to being synchronous, 



510 



Polarization Interferometers. 



