408 Messrs. Nichols and Snow on the Influence of 



of the ideal white (already defined) ; a method which was 

 followed in the case of all subsequent determinations. The 

 results are shown graphically in fig. 3. In this diagram and 

 in all that follow it, abscissae are wave-lengths and ordinates 

 are intensities. The spectrum of the normal white in each 



Fisr. 3. 



h- 



.2 















AlcjO 



- 



\ \ 5 ° 

 $*> 4* >*/- 



















B d 



case would be represented by a horizontal line, the ordinate 

 of which is l'OO. It will be seen that the magnesium oxide 

 differs from the ideal white in that it reflects the less refran- 

 gible rays more freely than it does those of shorter wave- 

 length. The deficiency increases steadily towards the violet, 

 where, in the region *425/-t, the pigment reflects only sixty-one 

 one hundredths as much light as the ideal white would have 

 done under the same illumination. The strip of foil, still 

 bearing the smoke-film of magnesium oxide, was afterwards 

 brought to 758° C, and the measurements were repeated with 

 the pigment at that temperature. The values obtained (see 

 Table I.) show a very slight decrease in reflecting-power but 

 no change of colour. 



