NINETEENTH ANNUAL MEETING. 



39 



passing the ray through the two nicols in front of the slit, the first of which was free 

 to revolve, while the second was fixed with the longer diagonal of its faces parallel 

 to the slit. By revolution of the first nicol this spectrum could be given any desired 

 intensity, from that which it possessed when the polarizing planes of the two nicols 

 were parallel, to zero, when the nicols were in position of extinction. The intensity 

 of this spectrum was indicated by a finger attached to the movable nicol, which finger 

 moved over a graduated circle. The ray producing the other spectrum, reflected 

 from the surface of the pigment to be tested, entered the slit by total reflexion from 

 the surface of a small, right-angled prism, the edge at one of the acute angles of 

 which was at right angles to and in contact with the middle of the slit. In other re- 

 spects, the instrument was an ordinary, one-prism spectroscope, into the eye-piece 

 of which had been inserted a diaphragm containing a small, narrow, vertical slit. 



If such an instrument be set up in front of a gas flame, so that the direct rays 

 of the flame will form the polarized spectrum, and the pigment, the whiteness of 

 which is to be tested, be placed above the reflecting prism at the slit so that its sur- 

 face is illuminated by the same flame, and by that alone, certain of the rays reflected 

 by the pigment will enter the upper half of the slit and form the unpolarized spec- 

 trum. 



Were this pigment a true white, the reduction of the direct spectrum necessary 

 to render these spectra equally bright would be precisely the same throughout their 

 entire length. Any deviation from true whiteness manifests itself by undue bril- 

 liancy or weakness of some components of the light reflected by the pigment. 



The amount of this variation may be measured, moreover, by going through the 

 spectrum from end to end and noticing for each wave-length the position of the 

 movable nicol at which that particular ray from the direct flame equals in intensity 

 the corresponding ray reflected by the pigment. The spectrum of the pigment 

 under these circumstances was very faint, but by taking the precaution to guard the 

 retina throughout the experiment from the fatigue which follows exposure to even 

 moderately bright objects, it was found possible to obtain a series of comparisons 

 of the reflexion sjjectrum of a block of magnesium carbonate with the direct spec- 

 trum of the flame illuminating it. 



To avoid even the loss of sensitiveness of the eye involved in reading the posi- 

 tion of the movable nicol, it was necessary for two persons to take part in the meas- 

 urements, one observ- 



(1). 

 (2)\\ 



ing at the eye-piece 

 of the instrument and 

 keeping his eyes con- 

 stantly in the rested 

 condition upon which 

 accurate settings so 

 much depended, while 

 the other moved the ro- 

 tating nicol to the po- 

 sition indicated by the 

 observer, read off that 

 position and the wave- 

 length to which the 

 reading corresponded, 

 made the necessary 

 notes, and kept the ap- 

 paratus in proper ad- 

 justment. This'plan served another very important end 



B C D Eh F G R 



Figure I. — Intensity curves of two specimens of Magnesium Carbonate 



The movement of the Nic- 



