F W. Very— On the Solar Constant 203 



these levels ; nor is there any certain evidence of proportional- 

 ity between depletion and atmospheric mass at still higher 

 altitudes. If a distinction were to be made between reflective 

 depletion and absorption, the former must come principally 

 from the dust and air molecules of the lower air, while the 

 principal selective absorption of solar rays is found at higher 

 levels. Granting the equality of spectrobolometric and thermo- 

 dynamic ratios, since the Mt. Wilson observations are the best, 

 they favor a lower value than 4 for the constant, or about 



•258 



-^r— X 4 = 3*6, which agrees with the value obtained by 



Savelief to which I have ascribed exceptional merit for reasons 

 given in my paper, " A Criterion of Accuracy in Measurements 

 of Atmospheric Transmission of Solar Radiation."* 



I am of the opinion that the sums of various thermodynamic 

 and radiant data for the entire atmospheric column which pur- 

 port to be given in gram cal./cm 2 min., where the numerical 

 value of the sum in each case approximates to 4, have nothing 

 whatever to do with the solar constant ; because, in the first 

 place, the conversion from M. K. S. mechanical values into 

 C. G. Min. thermal values has been incorrectly made. For 

 example, values which were originally obtained in M. K. S. 

 mechanical units have evidently been converted into C. G. 

 Min. (thermal) units by multiplying by 0'000014337. But 



10 7 

 1 joule/nrsec. = 10 7 ergs/m 2 sec. (mechanical) == _ — T 



gram cal./m 2 sec. (thermal), which is equivalent to 



10 7 x 60 

 10- X 4-185 X 10' = °-°° 14337 



for the C. G. Min. System, and the factor should be 100 times 

 greater than that actually used. The mistake has arisen because 

 in the original equations certain quantities were stated in 

 values per unit mass, which was entirely permissible, but the 

 calorie is a mass unit, and when quantities are given in ther- 

 mal measure it is not allowable to eliminate mass, except where 

 the mass is always unity, which is not the case here. Conse- 

 quently, the mass of each specified section of the air column 

 should be restored in making the thermal conversion. Elim- 

 ination of nonessentials is a very convenient procedure, but it 

 must never be forgotten that the eliminated quantities are 

 only in abeyance, and that they must be restored whenever 

 the conditions under which their omission was permissible no 

 longer hold good. 



In the next place, the passage of radiation through the 

 atmosphere is a flux of a fairly definite amount of radiant 

 * Astrophysical Journal, vol. xxxvii, pp. 31-47, January, 1913. 



