168 



Mr. T. Royds on the Reflective 



in exactly the same position relative to the hemisphere. 

 When the external optical arrangement was altered, the 



Fisr. 1. 



thermopile was placed in its first position to receive the light 

 directly, and the external concave mirror so adjusted that 

 the slit image tell on the thermopile slit which could be seen 

 from behind through a window. In this way, it was not 

 necessary to alter the two positions of the thermopile. These 

 two positions were determined once for all by experiment 

 and checked at intervals. The continuous line in fig. 2 

 shows the deflexions for a matt reflecting surface as the 

 thermopile was gradually displaced in the great circle of the 

 hemisphere. The distance between the incident slit image 

 and its image reflected in the hemisphere was made as small 

 as possible in order to have good definition ; this distance 

 amounted to 2'5 mm., the diameter of the hemisphere being 

 5 cm. The thermopile slit was 1 mm. wide, the image of 

 the illuminated slit falling just within it. A series of 

 Langley diaphragms was placed in front of the opening in 

 the hemisphere in order to avoid strong convection currents. 

 The lampblack surfaces were deposited on polished silver, 

 ^ mm. thick, from the fine flame of a small petroleum 



