130 C. Barus — Interference of Reversed Spectra. 



Experiments were carried out with this consideration in view, 

 by the single grating- and concave mirror method, JV being- 

 reduced from nearly 2 meters to 30 cm , without' any observable 

 change in the breadth or character of the phenomenon. It 

 showed the same alternation on one black and one or two 

 bright linear fringes, or the reverse, throughout. Hence it 

 seems improbable that the phenomenon, or the reverse, i. e., the 

 interference fringes, are referable to such a plan of interfer- 

 ence as is given in figure 11. 



Similarly for the case of two gratings, figure 10, if the data 

 i = 9° 40 v , 2 = 19° 55', D' = 173 X 10~ 6 cm., JST= 162 cm., 

 n = 82 cm., X = 58"9 X 10 _6 cm. be inserted, d6 is about of the 

 same value as above. 



8a. Experiments continued. — With this possible case dis- 

 posed of, it now becomes necessary to inquire into other causes 

 of the phenomenon, as described in paragraph 7. This is con- 

 veniently done with reference to fig. 12, where n and n' are 

 the axes of the pencil of yellow light, reflected from the opaque 

 mirrors JIT and JY, after arrival from the transmitting grating 

 G. It is necessary to consider the three positions of the 

 reflecting grating G' ; viz. G\ G/ and (?,'. In the symmetri- 

 cal position G', the pencils whose axes are n and n! meet at a 

 and are both diffracted along r. In the position G/ they are 

 separately diffracted at o and V in the direction r 1 and r/. They 

 would not interfere, but for the objective of the telescope, or 

 in the other case, of the concave mirror of the grating. In the 

 position 6?/, finally, the pencils n and n' are separately dif- 

 fracted at c and c' into r 2 and r„' and again brought to interfer- 

 ence by the lens or concave mirror, as specified. 



Now it is true that the rays na and n'a (position G'), though 

 parallel in a horizontal plane, are not quite collimated in a ver- 

 tical plane. The pencils are symmetrically oblique to a central 

 horizontal ray in the vertical plane and their optical paths 

 should therefore differ. But fringes, if producable in this way, 

 here, have nothing to do with the rotation of the grating in its 

 own plane and may be disregarded. 



To take the rotation of the fringes first, the interferences 

 obtained by rotation around a normal axis recalls the common 

 phenomenon observed when two picket fences cross each other 

 at a small angle 0, fig. 13. It is interesting to briefly examine 

 the relations here involved when S' and S are two correspond- 

 ing pickets of the grating at an angle $ and the normals D' 

 and D are the respective grating spaces. The intersection of 

 the groups of lines S' and S make the representative parallel- 

 ogram of the figure (S taken vertical), of which B is the large 

 and B' the small diagonal. The angles indicated in the figure 

 are x + y — </> and x' + y' + <f> = 180°. As the bright band in 



