F. W. Very— On the Solar Constant. 201 



Art. XIII. — On the Solar Constant ; by Frank W. Very. 



In his article in the September (1914) issue of this Journal,* 

 Professor Bigelow employs two different methods for obtaining 

 the solar constant and obtains by each a value of four gram 

 calories per square centimeter per minute. This value may 

 not be seriously in error, but the arguments on which it rests 

 are not entirely convincing. 



By the first method, " the pyrheliometer at any station 

 observes the ; effective '. radiation I, and never the solar con- 

 stant, which is A = 21 on the upper level where J a =r0" {op. 

 oit., p. 281). Here J a denotes the selective radiation of the 

 atmosphere, and where the atmospheric radiation is zero, the 

 atmospheric absorption must also be lacking; but why must 

 the solar constant "on the upper level," that is, at the outer 

 limit of the absorbent atmosphere, be twice as great as the 

 radiation measured by a pyrheliometer in this situation ? 

 Apparently because the pyrheliometer is supposed to be a very 

 thin, flat, black body which experiences a certain thermal 

 effect from solar radiation received normally on only one sur- 

 face, but whose own heat is radiated from two such surfaces, 

 which alike possess the quality of the black body, whence the 

 extra heat of the black body, derived from solar radiation, is 

 only half as great as it would be were the radiation issuing from 

 the rear surface to be suppressed. This argument, though not 

 stated in so many words, is implied by the form of the radia- 

 tion equation and shows a complete misconception of what we 

 mean by a pyrheliometer. The ideal pyrheliometer is an 

 instrument having only one efficient surface, which is that 

 exposed to the sunshine. Where, as in some actinometers, the 

 wholly blackened spherical bulb of a thermometer is used as a 

 receiver, the factor 4 is always introduced in the reductions in 

 allowing for the four-fold radiating surface of the receiver. 

 The original pyrheliometer of Pouillet was an approximation 

 to a single-surfaced, flat radiator, because, while the front sur- 

 face was blackened, the rear was of silver, polished as bril- 

 liantly as possible. As a radiant body, this very nearly fulfilled 

 the ideal specification. If the instrument could be used in a 

 vacuum, or outside the atmosphere, it would leave little to be 

 desired ; but convection losses, which are large, affect both the 

 black and polished surfaces, so that the theory is not as simple 

 as we have supposed. 



* " The Determination of the ' Solar Constant' by Means of Computations 

 Based on the Data of Balloon Ascensions," by Frank H. Bigelow ; this Jour- 

 nal (4), vol. xxxviii, pp. 277-281. 



Am. Jour. Sci.— Fourth Series, Vol. XXXIX, No. 230.— February, 1915. 

 14 



