LIGHT MEASUREMENTS 841 



the same position), quantum yield determinations can be made simply by 

 comparing the pressure changes in the two vessels. Average quantum 

 yields can be determined in this way for prolonged periods of illumination, 

 or for illumination with diffuse light, more easily than with instruments 

 which measure momentary light intensity, and require a collimated beam. 

 This actinometer promises to become a very useful tool in the study of 

 photosynthesis; hut despite its convenience it, too, must l)o used with cau- 

 tion, and checked from time to time against a physical light -measuring in- 

 strument such as a thermopile or bolometer. The nature and mechanism 

 of the reaction in the Warburg-Gaffron-Shocken actinometer is unknown, 

 and the exact dependence of its rate on light intensity, wave length, nature 

 of the solvent, presence of impurities, oxygen pressure, concentrations of 

 the reductant and the sensitizer, and rate of stirring, remain to l)e investi- 

 gated. Available measurements indicate that the quantum yield declines 

 slowly with increasing illumination in the "middle range" (3-7 X 10""' 

 einstein/(sec. cm.-)), but changes faster both at the higher light inten- 

 sities (>10 X lO-i*^ einstein/(sec. cm. 2)) and at very low light intensi- 

 ties (<3 X 10-1" einstein/(sec. cm. 2)). It remains to be seen, however, 

 whether the factor determining the change is intensity of the beam — i. e., 

 energy per unit cross section — or its energy (more probably, energy ab- 

 sorbed in unit volume of the liquid). Some decline in quantum yield with 

 increasing illumination may be caused by exhaustion of oxygen in the il- 

 luminated layer; however, this is likely to account only for a part of the 

 observed trend. Another possible cause of this decline is competition of 

 back reactions between the intermediate oxidation and reduction prod- 

 ucts, with the "forward" reaction which leads to the consumption of oxy- 

 gen. If they are bimolecular in respect to the intermediates, the back re- 

 actions must be favored by a higher concentration of the latter and there- 

 fore can become more effective at light intensity increases. 



It must be recalled that photometers with selective spectral sensitivity 

 cannot be relied upon not only in the determination of absolute light inten- 

 sities, but also in the comparison of two light sources or in the determina- 

 tion of the proportion of light absorbed by passage through a colored sys- 

 tem. Unless the absorption is very weak throughout the spectrum, the 

 spectral composition of the transmitted light will be different from that of 

 incident light, and the selectively sensitive instrument will react differ- 

 ently to these two fluxes. It will tend to exaggerate the absorption if the 

 latter takes place in the region of maximum sensitivity, and underestimate 

 it if it occurs in the region of low sensitivity. The steeper the spectral 

 sensitivity curve, the larger will be the errors that occur in absorption 

 measurements in nonmonochromatic light. Selenium barrier layer cells, 

 for example, cannot be used for such measurements even in "monochro- 



