Temperature on the Colour of Pigments. 



421 



luminescent surface of extremely evanescent and unstable 

 character. At the time that our measurements of the re- 

 flecting-power were made, our knowledge of the conditions 

 to be met was very incomplete. It is not possible by means 

 of the data obtained to trace accurately the changes of colour 

 which the oxide undergoes between ordinary temperatures 

 and the red heat, to say nothing of the rapid time-changes 

 which must take place at some of the higher temperatures. 

 It is easy, however, to detect certain typical changes which 

 the colour-curves undergo. The data given in Table XIII. 

 have been selected with a view to illustrating these types. 



The readings at 25°, 578°, 750°, and 815° were made with 

 the same coating and in the order named. A set of measure- 

 ments made at 884° have been discarded because they were 

 to some extent vitiated by the presence of light directly 

 radiated. An attempt was made to estimate the amount of 

 such light and to apply a correction. The result was un- 

 satisfactory. The four sets just mentioned have been plotted 



Fig. 14 



¥ - 



2. - 



. — -, 



Zu.o (*) 



' J^^» 











^\\25° 









\\\ 



- 





\r V 



, ?r 



■ £/*- 



_&- 



















a c 



£ F 



together in fig. 14. The curves show, for the cold oxide, a 

 spectrum weak both in the red and violet, with marked 



