80 DISPLACEMENT INTERFEROMETRY BY 



identical glass-paths. The tube cc is exhausted to a known degree through the 

 tubulure a, similarly c' through a'. The compensator is shown in place in 

 figure 82. Brass tubes about 15 cm. long are usually of convenient size and 

 will incline the horizontal spectrum fringes about 30, or restore them from 

 30 of inclination, if exhausted. 



60. Observations of refraction. The methods of procedure have been elu- 

 cidated in the preceding paragraphs relating to the self-adjusting interferom- 

 eter. It is there inferred (in connection with figure 82) that we may either 

 make a solution like mercury potassic iodide (for instance), optically identical 

 with that of the glass at any given wave-length of light, by diluting the con- 

 centrated solution; or that we may interpolate between two given solutions, 

 or from one given solution up or down, by aid of the Billet or the air compen- 

 sator. In case of a plate about 3 mm. thick, an air compensator (fig. 85) about 

 15 cm. long will throw the center of spectrum ellipses from the extreme red 

 into the yellow and correspondingly for the other colors. A second solution 

 would thus admit of interpolation from the green toward the blue, etc., by 

 aid of the same compensator. Three solutions should suffice. A longer air 

 compensator could of course be used ; but this would correspondingly elongate 

 or else complicate (if introduced on both sides, in rays 2 i and 3 4, fig. 82) the 

 apparatus and thus militate against easy manipulation. The range of the 

 Billet compensator is longer and its control much easier. 



To bring the center of ellipses into the field of the spectrum, it suffices to 

 use the trough containing the solution 5 with or without the body g. The 

 trough need merely be rotated on a horizontal axis until centers of ellipses 

 appear or (if centers are beyond the limits of the visible spectrum) the fringes 

 are vertical. In such a case the center of ellipses is not displaced longitudin- 

 ally (red to violet) in the spectrum and there is no interference with the 

 measurement based on longitudinal displacement. This easy possibility of 

 adjustment is a great convenience in practical work. Allowance, however, 

 must be made for the virtually increased thickness of the plate submerged. 



When the mercury potassic solution is concentrated, almost any glass plate 

 submerged and penetrated by but one beam will show only very fine spectrum 

 interference fringes, hair-lines, which may be made vertical, as just specified. 

 It requires considerable dilution (more than one-half) to obtain large fringes. 

 These eventually show curvature, and finally the center of ellipses appears 

 in the extreme red. On further dilution (which must now be made more 

 cautiously) the center passes through the spectrum from red to violet. 



If, in place of plate glass, ground glass is substituted, or in any case of 

 window glass, the ellipses come out remarkably well, quite adequately for 

 measurement. With very coarse ground glass and irregular fragments of 

 glass this is not apt to be the case. 



Lenses of several diopters, so long as the index of the glass is identical with 

 that of the solution within the spectrum, show clear fringes throughout the 

 height of the spectrum; but the lines are eventually apt to be doubly inflected. 



