SUNLIGHT AND ITS MEASUREMENT 157 



actual prismatic scale {i.e., the dispersions encountered) must 

 be given. It is to be hoped that the uninterpretable prismatic 

 spectra that are usually presented in biological exposition will 

 be abandoned and that normal energy spectra will be published. 



Because selective deductions of energy occur in the outer 

 envelopes of the sun, thereby producing the Frauenhofer lines, 

 it is obvious that measurements of the energy distribution in 

 the spectrum will not yield a smooth curve. Instead, the 

 graph will exhibit many irregularities and where the Frauen- 

 hofer lines are very close together the intensity will be so re- 

 duced that the curve will exhibit great depressions, as is the 

 case in the short-wave portion. 



A smoothed graph prepared from averages of a large number 

 of observations is, however, useful in exhibiting the general 

 form of the energy distribution curve without the distractions 

 consequent upon the simultaneous presentation of the irreg- 

 ularities caused by absorption in the outer envelopes of the 

 sun. Such a curve 9 is presented in figure 2, drawn to normal 

 scale. It is evident that the maximum intensity falls very 

 nearly at X = 0.47^, which is very close to the G line and 

 about where blue becomes changed to indigo. Since the "mean 

 solar constant" is given as 1.93 15° — calories, calculations from 

 a measurement of the area enclosed between the curve and its 

 base show that each of the small rectangles in the figure corre- 

 sponds under the conditions for which the solar constant is 

 defined to approximately 0.0294 calories. From this value 

 and measurements of the proper ordinates in the graph, the 

 energy values between chosen wave lengths may be rather 

 readily estimated. As has been noted this is a curve of averages, 

 the actual curve obtained on any given day differs from it to 

 some extent both in total area enclosed and in form. 



The "solar constant" 10 fluctuates irregularly, values of from 



9 Abbot, C. G., The sun's energy spectrum and temperature. Astrophys. 

 J. 34: 197-208. 1911. 



10 The term "solar constant" has obviously proved to be unfortunate and 

 before a larger volume of literature has further sanctioned its use a more 

 accurately descriptive term should be substituted. 



