Royal Society. 571 



In order to use the plate, it is to be placed on the table of a com- 

 pound microscope having a magnifying power of 40 to 50 times, the 

 table of the microscope being previouslj^ covered with black cloth 

 to prevent the intrusion of all foreign light. The small covering 

 plate bemg placed upwards, the polished face forming an acute 

 angle with the upper side, is to be turned towards the window sup- 

 plying light. The illumination must take place in a plane at right 

 angles to the length of the plate, and at an angle of 18° inclination 

 to the plane of the plate. The light of strongly illumined clouds, or 

 ot a point of the heavens near to the sun, is to be employed : direct 

 sunlight is not recommended. A greater angle of incidence is of no 

 consequence when making a preliminary observation ; but it is neces- 

 sary, m order to obtain strongly luminous colours, to turn the plate 

 a little in azimuth while examining the lines, thereby producing the 

 bnghtest illumination. If these preliminary precautions be taken, 

 there will appear in the ruled space, — 



for A, deep red ; 



B, orange ; 



C, yellow ; 



D, green ; 



E, light blue ; 



F, indigo ; 



G, violet. 



If the foregoing numerical values for the systems A, B, C, D, E, 

 F and G, be multiplied by cos 18°, the length of the undulations of 

 light in air for those colours which are seen in the rulings will be 

 obtained. This method is considered to admit of great exactness 

 with this small angle of incidence, because its cosine varies but 

 little. 



As in the following five systems, H, I, K, L and M, no colour is 

 perceived, the distances between the lines being there smaller than 

 the lengths of the violet light-waves in air, the prediction of the un- 

 dulatory theory is confirmed. 



The plate is next to be turned so that the small covering plate, 

 and, therefore, also the side of the glass upon which the lines are 

 drawn, is downwards, and the polished narrow face towards the light, 

 which, as before, is incident at an angle of 18° on the plate. In this 

 position, the light falling perpendicularly on the narrow face, passes 

 through it unbroken, and being within tiie glass reflected from the 

 rulings, the coloured pictures of all the twelve systems are now ex- 

 hibited in the microscope. If the colours which these systems now 

 show, and which have been formed in the glass, be compared with 

 the former colours in the air, the system F, which is deep red, har- 

 monizes with the system A in tiie air spectrum, and the systems G, 

 II, I, K, L and M in the glass spectrum, according to their order 

 and colour, corresijond to the systems B, C, D, E, F and G in the 

 air spectrum ; and the author considers that the comparison of the 

 foregoing values in the systems — 



2Q2 



