270 



BULLETIN OF THE UNIVERSITY OF WISCONSIN. 



trated upon the slit of a large, direct vision spectroscope (C),. 

 the eye-piece of the latter being removed, and in its place is 

 fixed the slit of a Geneva Society Spectrometer (D), the eye- 



*-C 



Fig. 3. 



piece of which is provided with a filar micrometer. The glass- 

 plate on which the cyanin prism (E) had been formed was cov- 

 ered with black paper, in which two small apertures had been 

 cut, one coming over the cyanin prism and the other slightly to 

 one side of it. The latter, or clear aperture, gave a direct image 

 of the slit of the spectrometer ; the former, a deviated image, due 

 to the bending of the rays by the prism. Both images being in 

 the field of the telescope at once, the distance between them could 

 be measured by the filar micrometer, and these readings when 

 reduced to the circular scale, furnished the data from which the 

 refractive indices were calculated. Any slight deviation due to- 

 the possible prismatic form of the glass plate used was thus elim- 

 inated. By turning a tangent screw, operating on the prism 

 system of the direct vision spectroscope, the focused spectrum 

 could be made to traverse the slit of the spectrometer, which was 

 thereby illuminated with monochromatic light of any desired 

 wave length. 



To determine the wave length of the light corresponding to 

 any particular reading of the cyanin prism the following meth- 

 ods were used. First, by opening the slit rather wide a consid- 

 erable portion of the solar spectrum was allowed to enter the- 

 instrument in which the Fraunhofer lines were distinctly visi- 



