500 



NICHOLS— COLOR-PHOTOGRAPHS OF 



that of liquid air there is a return to green. The distribution of 

 colors and approximate temperatures are indicated in Fig. 4 (A). 



That this remarkable distribution of colors is to be interpreted 

 as the result of the killing off of the " slow " red band by cooling 

 and its subsequent recurrence at about — 160°, is beautifully con- 

 firmed by the appearance of the tube after the entire cessation of 

 excitation. When the spark circuit is broken, the green vanishes at 

 once but the red regions above and below continue to glow for many 

 seconds as indicated in Fig. 4 (B). 



In 1910 I measured the change in the phosphorescence of this 

 substance^ effected by temperature. The curve, which is reproduced 

 in Fig. 5, applies to the red band only, since the green band would 

 have vanished long before observations of brightness could be made 



140° -120° -100°- -eo" 



Fig. S- 



by the method then employed. It describes exactly the appearance 

 noted in Fig. 4B, i. e., a rapid diminution on cooling, complete dis- 

 appearance at about — 120° C, a recurrence of measurable intensity 

 below — 140° with a maximum at — 160°. 



All BaCu combinations, so far as examined, have a red-yellow 

 color of phosphorescence at room temperature with the red band of 

 longer duration, but the relative intensity of the red varies greatly 



^Proc. of Am. Philosophical Sac, XLIX., 275, 1910. 



