A CHEMICAL PHOTOMETER 237 



the north sky together with some reflected Hght from a low white 

 wall opposite the north window in which the last-mentioned 

 exposure was made. At the end of the five hours the three 

 flasks were titrated and the decomposition of oxalic acid (0.229 

 gram) in the fully insolated mixture was taken as 100%. The 

 decomposition in the shaded flask (0.121 gram) showed that in 

 passing through the cloth the photochemical effect of the light 

 as measured by this reaction had been reduced to 47.2%. The 

 decomposition in the flask exposed to the north light was 0.029 

 gram, or 12.6% of that which took place during the same time 

 in the flask in direct sunlight. The figure for the transmission 

 of the shade cloth is nearly the same as its coefficient, as deter- 

 mined by Professor KimbalP with the use of the Smithsonian 

 silver disk pyrheliometer (42.7% at normal incidence) when it 

 is considered that the flask under the cloth was subjected to 

 considerable reflected light from a white wall during the expo- 

 sure. The same is true of the flask exposed in a north window — 

 that is, the percentage of total light (12.6%) also includes some 

 reflected light from the wall to the north of the flask. 



Experiments to determine the reliability of the solutions and 

 the accuracy of titration of the oxalic acid showed, by the use 

 of series of duplicate flasks exposed simultaneously, that there 

 were no differences beyond an experimental error of ±1%. The 

 results of further experiments conducted for the purpose of 

 measuring the intensity of the sunlight on clear days, hour by 

 hour, produced a curve very similar to the records obtained with 

 the pyrheliometer. This was true whether separate exposures 

 of an hour each were made or aliquots were taken at the end 

 of each hour from a continuously exposed volume of the oxalic- 

 acid-uranium-acetate mixture. 



From the tests outlined above, this chemical photometer 

 seems to be affected by light in a degree comparable to the 

 pyrheliometer in spite of the fact that the tw^o instruments 

 doubtless are influenced by different portions of the solar spec- 



' See Kimball, H. H., The shading effect of wire insect cages. Monthly 

 Weather Review, September, 1916, 44: 501-506, for a description of his method 

 of determining shading effect. 



