110 



Messrs. F. L. Usher and J. H. Priestley. [Apr. 13, 



had become even more inefficient as an absorber of heat than the first. 

 Again, if the films are shaded for an interval at a time when the tem- 

 perature difference is diminishing, this difference begins to increase again 

 when the exposure to light is renewed, probably because the carbon dioxide 

 and water undergoing photolysis are used up faster than they can diffuse 

 into the films. 



A final series of measurements, in which every precaution was taken to 

 avoid known sources of error, was made in April, 1909, with the apparatus 

 already described. The two films were illuminated through a large ground- 

 glass window and the vessels containing them were carefully protected from 

 draughts. The galvanometer was a dead-beat instrument, and a movement 

 of the spot of light over 25'4 scale-divisions corresponded to a temperature 

 difference of 1°. The error in the readings maybe taken as +0 o- 01. The 

 results are as follows— 



Film A — In tube containing potassium hydroxide solution 

 Film B — „ carbon dioxide „ 



Exposed to light at 12.15 p.m. 



Time. 



*A-*B. 



Time. 



tK-tB. 



12.16 | 1 -30 



12.17 1 -74 



12.18 1 -78 



12.19 1 -86 



12.20 1 -81 



(At 12.30 the solutions were interchanged, 

 and the films were exposed to light again 

 at 2 p.m.) 



2.1 -20 



2.2 -00 



2.3 -0-20 



2.4 

 2.5 

 2.6 

 2.7 

 2.8 



(Here the carbon d 



2.14 

 2.16 

 2.18 



-0-31 

 -0-47 

 -0-51 

 -0-71 

 -0-79 



oxide solution was 



wed) 



-0-67 

 -0-91 

 -1 -03 



From the figures given above it is evident that the temperature difference 

 observed depends on the composition of the atmosphere surrounding the 

 films, and that, apart from any want of symmetry in the films themselves, 

 the one in air containing moist carbon dioxide keeps at a lower temperature 

 than the one in air free from that gas. The transitory difference in the 

 wrong direction observed when the films were first exposed to light after 

 interchanging the solutions is doubtless due to a little residual carbon 

 dioxide in film B. A possible source of error, due to a difference in the 

 thermal properties of the gases in the two tubes, was examined by washing 

 the chlorophyll films off their metal supports with benzene, and taking 

 readings when the bare metal was exposed to light, the solutions being 



