SUNLIGHT AND ITS MEASUREMENT 203 



The area between line I) and its base (total radiation for 15° 

 sun) is 0.58 of the area between curve B and its base (total 

 radiation for 90° sun). Since this last ratio is more than three 

 times the first one given it is plain that the darkening is no 

 measure of the total radiation, as might have been predicted. 



The area between line D and its base in the portion of visible 

 radiation is 0.42 of that of the corresponding portion of the 

 area below line B, so the actinometer reading is likewise no 

 measure of the visible radiation; nearly 55% error. 



The area between line I) and its base in the effective portion 

 of the spectrum is 0.23 of the corresponding area below line B 

 so that the actinometer does not even measure this portion of 

 the spectrum with any degree of accuracy; nearly 18% error. 



All this merely emphasizes the usually unappreciated facts 

 that the actinometer readings are only valid when the radiation 

 is either monochromatic or invariable in its distribution of 

 energy, and that sunlight does not fulfill these requirements. 

 The illustration also indicates that if a sufficiently small portion 

 of the spectrum were transmitted to the paper by an a ppropriate 

 screen, this portion might be measured satisfactorily for many 

 purposes although no conclusion could be reached regarding the 

 energy transmitted in other portions of the sun's spectrum. 



To complicate the case still further a very large share of the 

 ultra-violet (radiation of wave lengths shorter than about 0.4 /x) 

 reaching the earth comes from the sky and, as will be seen from 

 the sensitivity curve in figure 5, these waves are the most ef- 

 fective in darkening silver chloride. Furthermore, when the 

 sun is close to the horizon, the sky is more effective than the 

 direct beam 45 so that if the actinomter is to measure either sky 

 light or sun light is must be protected from radiation from the 

 other. 



There are of course other types of actinometers besides those 

 depending upon the darkening of silver chloride paper. Photo- 

 sensitive preparations that must be chemically treated after 

 exposure to render the image visible ("plates" are examples) 

 have been used and while the simple reciprocity law is not 

 applicable to their action, Schwarzschild's rule is said to be, 



THE PLANT WORLD, VOL. 22, NO. 7 



