[ 98 ] 



XII. On the Bands formed by the Superposition of Par agenic Spectra 

 produced by the Grooved Surfaces of Glass and Steel. — Part II. 

 By Sir David Brewster, K.H., F.R.S.L.fyE* 

 [With Two Plates.] 



IN the preceding paper, I have described the bands produced 

 by gratings or grooved surfaces with 500 divisions in an 

 inch when the two grooved surfaces are in contact, and the 

 grooves in the one slightly inclined to those in the other. 



The following results were obtained with two gratings, one of 

 which had 2000 and the other 1000 divisions in an inch. 



1. When the surfaces are in perfect contact and the grooves 

 parallel, the bands seen on the united surfaces, either with a lens 

 or by ordinary vision, are very irregular and are parallel to the 

 grooves. They are seen only on the 2nd, 4th, 6th, &c. spectra 

 on each side of the luminous bar or disk. 



By turning the nearest grating slightly to the right from the 

 azimuth 0°, the bands fall back to the left, increasing in num- 

 ber, and descending with their concave sides downwards into 

 distinct serrated black and white bands nearly perpendicular to 

 the grooves. When the nearest grating is turned to the left, 

 the bands descend towards the right, with their concave sides 

 upwards, till they become nearly perpendicular to the grooves. 

 In all these positions the bands are twice as numerous on the 

 fourth spectrum as on the second, and thrice as numerous on 

 the sixth as on the second ; and when the grooved surfaces are 

 perfectly parallel, the bands are immoveable on the grooved sur- 

 faces at all angles of incidence. 



2. When the grooved surfaces are separated by the thickness 

 of one or both of the plates of glass, the bands are very indi- 

 stinctly seen, and they seem to diminish in size with the distance 

 of the grooved surfaces ; but this is not certain, owing to the 

 difficulty of fixing the plates with the grooves at the same incli- 

 nation to each other. 



Similar bands were seen on the united surfaces of gratings of 

 2000 and 2000, 1000 and 1000, 500 and 500, 1000 and 500, 

 and 2000 and 500 divisions in an inch, but always less di- 

 stinctly when the grooved surfaces are separated by the thickness 

 of one or both of the plates. 



The beauty and distinctness of these bands depend upon the 

 skill with which the gratings are ruled. In several of the gra- 

 tings which I possess, the phenomena I have described can 

 hardly be recognized. 



When the combined gratings have the same number of divi- 



* From the Transactions of the Royal Society of Edinburgh, vol. xxiv. 

 part 1 . Communicated by the Author. 



