1921] 



on Cloudland Studies 



433 



are obtained than from dark clouds. A bright cloud in early June 

 gave 2 ' 2 times the black dome displacement at earth temperature, 

 while even in February a Z/E ratio of 1*8 was obtained. The 

 results were generally higher at or just before noon than a few hours 

 later (direct sunlight, if any, being screened off). The proportion 

 due to short wave lengths, excluding those known to be absorbed by 

 liquid air,* was investigated by filtering the radiation through cooled 

 glass (as for example, a small beaker containing 2 or 3 cm. depth of 

 clean liquid oxygen, in the neck of the thermoscope flask, or a thin 



PROPORTION of 

 BLACK BODY E 



16 MINUTES 



glass disc in a hinged ring covering when required the aperture of 

 the lowest diaphragm above the cell shutter). 



The responses obtained when continuous exposures with the cold- 

 glass screen were made while variable broken clouds crossed the 

 zenith (on July 13, earth temperature 23*7° 0.) are shown in 

 curve 1 of Fig. 12, the scale of ordinates being in terms of unity taken 

 for the value given by the black dome at the time. The passage of 

 bright portions of cloud corresponded to the peaks of the curve, 

 while values near or below zero (negative reading) resulted from the 

 clearer blue spaces. Corresponding curves from observations without 

 any cold glass screen are shown in curves 2 and 3 (taken March 21, 



Proc. Roy. Inst., xv. p. 556. 



