XIII. ACTION SPECTRA AND ABSORPTION SPECTRA 441 



for scotopic vision which, as shown in Figure 5, agrees nicely with the 

 absorption spectrum of visual purple when proper account is taken 

 of absorption by nonparticipating substances. In most cases, how- 

 ever, there is departure from such agreement, although the action 

 spectrum reflects the character of the absorption spectrum to a greater 

 or less extent. Obviously, all the assumptions employed in the 

 analysis cannot always apply. For example, the light absorber 

 probably absorbs much more than 10% in many instances — hence 

 thin layers should be used when feasible — and, even if all the as- 

 sumptions are fulfilled, there are still numerous other sources of error, 

 some physical, some biological. The same conditions applying to 

 rigid layers may apply to inadequately stirred suspensions of cells, 

 leading to errors in the calculation of photochemical efficiency and in 

 the interpretation of action spectra. 



Numerous reasons have been mentioned above for expecting the 

 absorption spectrum of a substance when measured outside the living 

 system not to agree accurately with the absorption spectrum of that 

 substance when inside the living system. There is also the possi- 

 bility that a part of the incident light is absorbed by nonparticipating 

 substances that absorb selectively with respect to wavelength. A 

 good example of this is found in the erythema of sunburn, where the 

 photochemical reaction takes place in the Malpighian layer of the 

 epidermis beneath the horny layer (corneum), which strongly ab- 

 sorbs some of the wavelengths that elicit the erythema (5). Ab- 

 sorption by the corneum shows a strong maximum at 280 niju, which 

 no doubt accounts for the sharp minimum in the action spectrum at 

 this point (see Fig. 3) . A similar case is that of the action of ultra- 

 violet radiation on pollen grains, where the pollen wall absorbs 

 strongly the wavelengths that affect the living protoplasm within 

 (35). Still another example is the absorption of shorter wavelengths 

 by the lens of the human eye, which is illustrated in Figure 5. It is 

 also possible for a nonparticipating substance inside the cell, in close 

 association with the light absorber, to act as an internal filter; a still 

 more difficult effect to estimate quantitatively. 



It is obvious that the scattering of light by structures in the living 

 organism may affect the absorption spectrum, since the proportion 

 of the light scattered may vary greatly with the wavelength. Fluoies- 

 cence is another possible source of difficulty, since not only does it 

 serve to diffuse the light but introduces chromatic impurity, as the 

 fluoresced light may have a considerably different wavelength from 



