110 DISPLACEMENT INTERFEROMETRY BY 



is usually accompanied by the presence of the satellites in question. The 

 latter may be made distinct by slightly rotating the plate (fig. 108) on a vertical 

 or horizontal axis, which brings about a more perfect coincidence of the corre- 

 sponding white slit-images. The edges of these may, in fact, sometimes be 

 detected. Using this apparatus, I made a few measurements on the position 

 of seven successive groups, obtained in a specially good adjustment. The 

 micrometer reading AN corresponding to their position was as follows : 



io 4 A/V = 25i, 234, 217, 200, 182, 165, 148 cm., 



the strong fringes being at AN = 0.0200 cm., a normal micrometer as at m' r 

 figure 104, being used. On compressing the plates, figure 108, more tightly, 

 I found io 4 A7V=227, 214, 200, 186, 173 cm. 



The constant differences, 0.0017 cm - i n the first example and 0.0013 in the 

 second, can be nothing more (since both spaces count doubly) than the thick- 

 ness of the air-film inclosed between the plates. If we call this thickness s, 

 and denote the optic paths of the two rays of the interferometer by r and r', 

 the following possibilities of interference present themselves : 



r and r' primary. 



r-\-2S and r' r and r'-\-2s secondary. 



and r' r and r'+^s tertiary. 



and it appears from the above data that three reflections within the air-film 

 still produce an observable effect, even with the present ordinary plate glass. 

 This result, therefore, suggests the design of a peculiar displacement inter- 

 ferometer, using white light and the achromatic fringes ; for it is merely neces- 

 sary to put one of the two parallel plates P and P', figure 108, on a micrometer, 

 in order to specify the distance apart of these plates in terms of the ocular 

 distance of the primary, secondary, etc., achromatics seen in the telescope. 

 They would all coincide when P and P' are in optical contact. Since these 

 groups are here nearly of the same size, it becomes a rather interesting experi- 

 mental question to see how far the separation of P and P' can be practically 

 carried. I made a tentative experiment by inclosing a piece of paper, o.i mm. 

 thick, between the edges of the silver doublet and obtained the two secondary 

 fringes, one of which, however, was much larger and the other smaller than 

 the primary fringe. A number of measurements were made, showing that the 

 large secondary fringes corresponded to a position AN = 0.013, 0.0112 cm., and 

 the smaller to AN = 0.0112 and o.oiu cm. from the primary, agreeing, 

 in view of the preceding measurements, with the surmise. The plates were 

 not good enough to warrant further spacing. 



82. The triangular self =ad justing interferometer. This apparatus is shown 

 in figure 109, in which A is a half- silver reflecting at the 5 face, B and C opaque 

 mirrors. The white light a arrives at L from a collimator and is observed, 

 when the rays are coincident, through the telescope at T. In order that such 

 coincidence may be established the mirror B must be on a micrometer-screw, 



