August 17, 1893] 



NATURE 



373 



The mirrors b and f were maintained in position • by 

 the conical foot a (Fig. 2) standing in a cylindrical holein 

 a blank which was stuck to the glass bed-plate by pitch, b 

 resting simply on the surface of the glass, whilst the foot c 

 stood in a V-groove in a brass blank, also stuck by pitch 

 to the glass bed-plate. The fine adjustments required are 

 for c (Fig. i) a motion in the direction «:/, and for (/, 

 adjustments in altitude and azimuth. These were re- 

 spectively obtained by placing the two feet b and c 

 (Fig. 2) (which should be rounded) of the mirror c (Fig. i) 

 in a long Vgroove (a piece of angle brass was used), 



Fig. 2. — The dotted lines indicate tile position of the mirror. 



the third foot resting on the glass surface ; the foot c was 

 held against a screw passing, in the direction of the groove, 

 through a brass blank soldered at the end thereof, which 

 gave the longitudinal motion required for that mirror. 

 By means of a lever of 1 8 " or so in length attached to it, a 

 piece of steel wire, with a thread cut by means of stocks 

 and dies of40tothe inch, wasfound capable of adjusting 

 to a quarter of a wave-length of light. The adjustments 

 of the 1 fmaining mirror d (Fig. i) were obtained by 

 allowing me conical foot a (Fig. 2) to rest in a cylindri- 

 cal hole, whilst the foot b rested on the glass as in the 



F'g- 3- 



mirrors i^ and/ (Fig. i). The leg c (Fig. 2) was formed 

 of a piece of steel wire with a screw thread cut as 

 described above, with a large brass blank soldered to 

 its upper end ; this gave the adjustment in altitude. 

 The adjustment in azimuth was obtained by hold- 

 ing the horizontal piece of band brass b c by means 

 of a piece of elastic against the end of a screw of 

 similar pitch to that last described, passing through a 

 vertical pillar attached to the baseplate. The whole 

 arrangement is shown in Fig. 3. 



To avoid the effect of vibrations the whole may be 

 supported on a block of stone, resting, in its turn, on 



^' For icme remarks on the general principles of these "geometrical 

 illdcs" ard "clamps," fee Thomson and Tait'" "Nat. Philosorhy.i' 

 part 1. p. 150. 



NO. 1242, VOL. 48! 



hollow india-rubber balls ; a plan adopted successfully by 

 Dr. O. Lodge. When mounted in this manner the bands 

 may be shown in a room possessing only an ordinary 

 wooden flooring. 



The bands are obtained as follows : A bat's-wing 

 burner, or other source of white light, is placed at the 

 focus of the lens L (Fig. 3) and arranged so as to illumine 

 the mirror A. A card with a pinhole in it is then placed 

 in front of the lens L, and, on looking in the direction 

 M A, two images of this will be seen. By means of the 

 screws B and C these two images are superimposed ; 

 and the distance a e having been adjusted by means of 

 the screw and lever F and a steel scale, to be equal to 

 A D, a sodium flame is placed in the focus of L and the 

 pinhole card removed ; the sodium bands will at once 

 appear. By means of the screws B and C these are 

 adjusted to a convenient width, and then, the bat's-wing 

 burner having been replaced in the focus of L, the lever F 

 is turned very slowly till the coloured bands appear. 

 This can be done much more easily by placing a piece 

 of platinum wire holding some sodium into the flame. 



Fig. 4.- Photograph of interference band showing cold match. 



when the bands due to the sodium will be faintly out- 

 lined on the white background, thus giving a guide as to 

 whether or no you are turning the screw f too fast. The 

 bands appear on the surface of the mirror E, and if an 

 electric arc or a mi.xed gas limelight jet be substituted as 

 the source of light, they can be projected on a screen so 

 as to be visible to a large audience. 



The forms of these interference bands, supposing each 

 of the four pieces of glass to be perfectly plane and 

 parallel, is given by Michelson {Phil. Mag. April 1882). 

 The peculiar form of the bands obtained in my appa- 

 ratus is shown in Fig. 4 ; this form is due to the curvature 

 of the surfaces of the various glasses. A thin piece of 

 glass or a soap film may be introduced into one of the 

 paths and the displacement of the bands exhibited. But 

 perhaps the prettiest experiment is to introduce the 

 glowing end of a match into one of the rays. Suppose 

 this ray to be A E (Fig. 3) ; then, the appearance pre- 

 sented is exhibited in Fig. 5, where the bands are seen 

 to curve round the end of the match as if it were pushing 

 them inwards. A cold body, such as a piece of copper 

 wire, cooled in a freezing mixture, has an opposite effect, 

 attracting the bands into it. These effects are, of course, 

 due to the heating or cooling of the air near the hot or 

 cold body. Now it will be found on slowly turning the 



