NO. 3 ATMOSPHERIC SCATTERING OF LIGHT FOWLE II 



As has been stated in former communications, this strongly con- 

 firms the accuracy of our estimations of the atmospheric losses affect- 

 ing the radiation reaching us from the sun. 



There is to be expected above the altitude of Mount Wilson (1^730 

 meters) a certain amount of what has been called " dry haziness " to 

 distinguish it from a similar haziness associated with aqueous vapor. 

 Before the Mount Katmai eruption of 1912, during 1910 and 191 1. 

 this caused a loss of only about -i of one per cent from the in- 

 coming solar radiation when the sun was in the zenith. The mean 

 of the coefficients for these two years ( table 1 ) . given in the lower 

 line of that table, may be taken as a close approximation to the 

 transparency of dry. dust-free air. During 1913, this loss due to dry 

 haziness decreased from its enormous value of 25 per cent just sub- 

 sequent to the Mount Katmai eruption to about 3 per cent and during 

 1914-15 to about 1 per cent, but it increased again to 3 per cent 

 during 1916. 



Within the same spectrum region, the transmission coefficients for 

 atmospheric aqueous vapor (a l0x ) also apparently vary with the 

 inverse fourth power of the wave-length. The scattering of radia- 

 tion when passing through liquid water is shown to be the same as 

 would be expected from the number of (H 2 0) molecules present if 

 the same quantity of water existed in a gaseous state. But the same 

 amount of water in the form of atmospheric water vapor should give 

 50-fold less absorption than that observed. This may be due to 

 some combination (HX))., of a portion, at least, of the vapor. In- 

 creasing the effective diameter of the scattering particle may be far 

 more effective in scattering the radiation than is compensated by the 

 resultant decrease in their number ; for the scattering varies with 

 the sixth power of the diameter and only directly with the number. 

 This peculiar molecular condition might be supposed connected 

 with some ionization phenomenon, and possibly, like the aurora 

 (Stormer), in some way might be dependent on charged particles 

 coming from the sun. As shown in figure 2 there does seem to be a 

 connection between this phenomenon, curve a, the solar radiation 

 intensity, curve b, and the sun-spot numbers, curve c. This amounts 

 to saying that the smaller the average solar radiation or the sun-spot 

 number, the greater is the absorptive power of atmospheric water 

 vapor. This result requires further testing. It is, however, con- 



