Fier. 2. 



and an Optical Method of Calibration. 423 



This is obtained by a similar method to that adopted by 

 Lummer and Brodhun. One of the prisms (fig. 2), as A U C, 

 is silvered over the face A D and the silvering then carefully 

 removed except a strip S S parallel 

 to A D. The edges of the strip S S 

 must be as sharp and as regular as 

 possible. The two prisms are then 

 carefully cemented together with a 

 substance of as nearly as possible 

 the same index of refraction as the 

 glass. 



Light entering the prism ADC 

 by the face D C and striking 8 S is 

 then almost totally reflected inter- 

 nally, and passes out through the 

 face A C. This occurs up to the 

 edges, where the rays pass abruptly on through the prism 

 ADB by regular refraction. Those rays which enter the 

 prism A D B by the face B D pass on directly through the 

 entire compound prism except those which encounter the 

 strip S S ; these are reflected and pass out of the face A B. 

 We have thus an abrupt transition between the two sources 

 of light, so that a perfect continuity in the field obtains when 

 the setting is made for a match. 



In order to make comparisons with light as nearly mono- 

 chromatic as possible, glass of great dispersive power should 

 be used, and this requires a cemen ting-fluid of corresponding 

 refractive index. The glass selected is that from the factory 

 of Messrs. Schott and Co., Jena, Catalogue type No. 102, 

 Dense Silicate Flint. n D = l , 6527, mean dispersion C to 

 F, O01950. Canada balsam may be used as the cement, but 

 much better results have been obtained with alpha-mono- 

 bromonaphthaline, ra D = 1*6582 at 20° C. This fluid is very 

 transparent, particularly in the blue and violet, and has so 

 nearly the same index as the glass that there is little internal 

 reflexion at the surface A D. It does not attack the silver 

 strip, and, while rather volatile, can be easily cemented in at 

 the edges by insoluble material such as gelatine, shellac, &c, 

 making it fairly permanent. If necessary, the prism may be 

 readily taken apart, cleaned, and refilled a number of times 

 without injury to the faces or the silvering. 



In order to adjust the instrument for use the collimator T 

 (fig. 3), the prism P, and the telescope R are placed to give 

 minimum deviation for sodium light, the cross-hairs of the 

 eyepiece coinciding with the centre of one of the sodium 

 lines. The collimator T is then screened and T' adjusted so 



