SUPERPOSITION OF PARAGENIC SPECTRA, ETC. 231 



The bands are of the same size as with a single grating when the grooved 

 surface is uppermost, but they are very much larger when the grooved surface 

 is in contact with the steel. 



When the grooved surface is slightly inclined to the steel surface, as in fig. 

 11, and the grooves parallel to the plane of reflexion, a double system of hyper- 

 bolic bands is seen, as in fig. 12, having one asymptote coincident with the bar 

 of light and the other at right angles to it. One of the systems of hyperbolas 

 is on one side of the bar and the other system on the other side, the number of 

 bands on the second spectrum being double those on the first. 



When the grooves are inclined to the plane of reflexion by turning them to 

 the left or to the right, the double system of hyperbolas moves to the left or to 

 the right, the curves of each system crossing the spectra, as in fig. 13, and being, 

 as before, twice as numerous on the second as on the first spectrum on both sides 

 of the bar. By increasing the inclination of the grooves to the plane of incidence, 

 the system of hyperbolas moves farther to the left or to the right. 



When the bar of light is placed at E and the eye at R, fig. 11, the system of 

 hyperbolas is inverted, as in fig. 14. 



It is curious to observe the passage of the parallel rectilineal bands into 

 hyperbolas, when the inclination of the grooved to the steel surface commences. 

 The parallel bands open at their lower end, as in fig. 13, or at their upper 

 end, as in fig. 14, and change into hyperbolas. When the light was strong, I 

 observed a second but fainter system of hyperbolas lying between the principal 

 system and the luminous bar, and caused probably by reflexion from the second 

 surface of the grating. The effect produced by the crossing of the bands arising 

 from these two systems of hyperbolas was remarkable, and similar to what I had 

 observed in combining two gratings of 500 divisions in an inch. This second system 

 of hyperbolas was most distinct when the plane of reflexion from the surface of the 

 steel was coincident with the plane of reflexion from the glass ; and the double sys- 

 tem was seen with grooved surfaces of 500, 1000, and 2000 divisions in an inch. 



In using accidentally a steel surface that was not perfectly flat, I was surprised 

 to observe that the bands were not hyperbolas, but circular rings varying in form and 

 size with the angle which the grooves formed with the plane of reflexion. In order 

 to examine this new and beautiful phenomenon, I placed the grooved surface of the 

 grating, AB, upon a convex surface of steel, MN, as in fig. 1 5, so that the rays from 

 the luminous body might reach the eye at E, after reflexion from the convex surface, 

 MN. The reflected image of the grating is thus superposed upon the direct image, 

 and two systems of concentric rings are seen upon the surface of the grating. At 

 the point of contact, C, and around it, are seen the rings of thin plates described 

 by Newton, and increasing in size with the radius of the surface MN. Around and 

 concentric with these as shown at a b, fig. 16, is seen a beautiful system of serrated 

 rings formed upon the paragenic spectra, as in fig. 16, the number of rings upon the 



VOL. XXIV. PART I. 3 E 



