570 



Notes. 



the same pitch, let the ray A B (fig. 135) impinge on the grating G 1 , and let 

 B C be the dioptric or normally diffracted ray, and B D one of the diffracted 

 rays of the first order to which the grating gives rise. When the ray BC 

 meets the second grating, it again splits up into several proportions, one of 

 them (C E) proceeding in the original direction. When B I) meets the second 

 grating G- this also is split up, the dioptric portion continuing in the direc- 

 tion D H, and the first diffracted rays proceeding in the direction D K and 

 D K 1 . But as the grating is of the same pitch as the other one, the angle 

 between D H and D K must be the same as the angle between B C and B D, so 

 that DK is parallel to A B or C E; that is to say, part of the incident Light 

 which was diffracted off a particular angle by the first grating has been again 

 rendered parallel to the incident ray, and consequently also parallel to the 

 transmitted dioptric ray which has not had its direction changed. 



Fig. 136. 



Fig. 137. 



And as this reasoning applies equally for rays of all colours and for 

 diffracted rays of any order, it is clear that they all issue parallel to one 

 another, the only difference being in their distance from the central or dioptric 

 ray. This distance from the central ray for diffracted rays of different colours 

 is, as will be seen, strictly proportional to their wave-length, and the peculiarly 

 interesting feature presents itself that this proportionality is independent 

 of the distance separating the two gratings ; for. as may be seen in figs. 136 

 and 137 (in which the violet rays V are represented by ordinary and the 

 red rays B by dotted lines), the ratio of C V to C B does not depend on the 



