Temperature on the Colour of Pigments. 407 



new faces thus formed were rubbed together until they had 

 acquired complete identity of surface-texure. One of the 

 pieces was placed in the position usually occupied by the 

 lamp L, the lamp itself being set up about ten centimetres 

 away. The newly prepared surface was illuminated by the 

 lamp, and it made an angle of 45 degrees with the axis 

 of the polarizing prisms. The other piece of the carbonate 

 was mounted in place of the platinum strip, and in the same 

 manner as the first one. It was illuminated by rays from I/, 

 which had been removed to a sufficient distance to restore the 

 ratio of brightness of the two spectra to a convenient value. 

 The spectra of the rays reflected by these two identical sur- 

 faces were compared, wave-length for wave-length, from red 

 to violet. The difference, which was indeed most striking, 

 could be due only to selective absorption by the condensing- 

 lens and the Nicol prisms. Subsequent measurements, made 

 when both pieces of magnesium carbonate were illuminated 

 by the same lamp, showed that L and I/, which were lamps 

 of the same type carefully adjusted to the same temperature, 

 emitted the same quality of light, so that no correction for 

 difference of illumination was necessary. It was not essential 

 to our purpose to determine what portion of loss by absorp- 

 tion was due to the action of glass and what to calcite, but 

 the effect of the two taken together was so surprisingly large 

 that it seemed well worth while to separate them. This was 

 readily accomplished by repeating the comparison of the 

 spectra with the lens removed, a condition under which 

 absorption within the calcite alone remained. The data 

 obtained in these experiments upon the selective absorption 

 in glass and calc-spar have no direct bearing upon the subject- 

 matter of this paper. As they are in themselves of some 

 interest, however, they will be communicated under a distinct 

 title. 



This preliminary investigation being completed and the 

 correction-factors for absorption fixed upon, we proceeded to 

 study the spectrum of magnesium oxide at 25° 0. This was 

 not one of the substances from which we expected marked 

 effects as the result of heating, but we were anxious to deter- 

 mine in what manner it differed from true whiteness. The 

 brightness of its spectrum in the region of the D-line, more- 

 over, was the value to which all subsequent measurements 

 were to be referred. Measurements were made at seven 

 points in the visible spectrum, the wave-lengths of which are 

 given in Table I. The brightness of each region is expressed 

 in terms of that of the corresponding region in the spectrum 



