Light emitted by Incandescent Zinc Oxide* 19 



Postsckipt {October 11). 



Solutions of saponine and gelatine, substances which con- 

 fer the foaming property, have been examined. With very 

 small quantities the difference of readings may amount to a 

 degree, not to be diminished by repeated skimming. The value 

 of k is thus -'005. 



The suspicion above suggested that the true value of k for 

 clean water may be numerically higher than is indicated by 

 the results obtained with the aid of the brass hoop has been 

 verified by some observations upon surfaces cleansed by heat. 

 The water, as clean as possible, was contained in a large 

 shallow tin tray. By the application of gentle heat to the 

 part of the tray under optical examination any residual grease 

 is driven off, in consequence of the smaller tension of the 

 warmer surface. If the whole surface is fairly clean to begin 

 with, a very moderate difference of temperature suffices to 

 keep the grease at bay. The difficulties of the experiment 

 have so far prevented a complete series of readings ; but the 

 following, obtained on October 2, seem sufficient to establish 

 the fact : — 



+ 30 c 

 -30° 



+ 30° 

 -30° 



43, 42, 42, 42 



47, 47, 48, 47 



52, 51, 51, 50 

 47, 46, 47, 45, 48 



At the conclusion of the second set the contamination was 

 evidently returning. It would seem that on the cleanest 

 surfaces the difference of readings may amount to 5', the 

 necessity of readjustment on passing between /3= ±30° being 

 obvious on simple inspection. Corresponding to this 



k=+ -00042. 

 Terling Place, Witham. 



II. On the Character of the Light emitted by Incandescent 

 Zinc Oxide. By Edwakd L. Nichols and Benjamin W. 

 Snow*. 



IN a recent paper on the influence of temperature on the 

 colour of pigments f, we described the singular changes 

 which rise of temperature produces in the character of the light 



* Communicated by the Authors, 

 t Phil. Mag. [5] vol. xxxii. p. 401. 

 2 



