100 DISPLACEMENT INTERFEROMETRY BY 



orientation of the spectro-telescope revolving around its axis may be deter- 

 mined by the appearance and evanescence of fringes. On the other hand, the 

 spectro-fringes, particularly if large, remain clearly enough in the field for 

 the observation of the motion of a large number ('i.e., for interferometry) 

 before they vanish. 



75. Reversed spectra. If the linear interferences are produced by placing 

 a grating between m and m' (fig. 95) and suitably adjusting the apparatus, little 

 effective modification is possible. As the breadth of this phenomenon is 

 independent of dispersion, a prism (preferably 30) may be placed between 

 m and m', as in figure 92, Chapter VIII. The narrow but very luminous 

 spectra, on superposition, etc., then show an intense string of interference 

 beads at the line of symmetry if the slit is also narrow. If M and M', figure 95, 

 are quite coplanar and the fringes therefore horizontal, the slit may now be 

 broadened with a corresponding effect on the interferences, but eventually they 

 are lost in the glare of light. When viewed through the spectro-telescope, 

 gT, however, they come out distinctly and the slit may now be broadened 

 indefinitely. The striations are found to be very sensitive to the degree of 

 vertically of the slit and slight departures from the vertical throw the stria- 

 tions into opposite inclinations to the horizontal. This accounts for frequent 

 occurrence of arrow-headed forms, as these correspond to a vertical slit. 



If the mirrors M and M' are separated as in figure 95, the fringes are no 

 longer horizontal, as a rule. The spectro-telescope is ineffective and the slit 

 can not be broadened. On displacing M or its micrometer-screw s, however, 

 the strong duplicated fringes are soon found, and when viewed through gT 

 any breadth of slit is permissible. In this way the linear phenomenon may 

 be expanded laterally to an indefinite area and the character of its individual 

 fringes (which were originally point-like) observed in detail. If the reversed 

 spectra are passed through each other, the fringes undergo marked changes, 

 such as from horizontal maxima in a blue field of coincidence to fine vertical 

 lines in a red field, for instance. Each color, moreover, requires a particular 

 adjustment, AN, to secure the maximum sharpness of the design. If the fringes 

 are not duplicated they are nearly invisible in the spectro-telescope. 



For a given line of coincidence the fringes admit of but little displacement, 

 AN. When this is varied the fringes appear in one size and vanish in a mark- 

 edly different size and inclination. 



As a whole the advantages gained in duplicating and enlarging the linear 

 phenomenon are not as striking as is the case with non-reversed spectra, a 

 result to be anticipated from the increasing presence of non-interfering light. 



76. The same, continued. I have already instanced that the roof -shaped 

 or arrow-headed forms of interference patterns also occur with homogeneous 

 light along the line of contact when a cylindrical lens is placed on a plate. 

 One may therefore suppose that the effect of cylindrical variation of thickness 

 normal to the line of contact when wave-length is constant is (formally) closely 



