v BELIEF AND EVIDENCE 107 



only be given when the effects of atmospheric dust and 

 moisture as filters of sunlight had been studied. Sunlight 

 is white light, and when we see the sun red, it is because 

 the blue rays in the original beams have been dissipated 

 by the intervening air, or rather by the dust particles in 

 the air and the water droplets which surround them. 



A red sky in the evening indicates that there are plenty 

 of dust particles in the air, and that water-vapour is 

 beginning to be condensed upon them, but relatively few 

 water droplets are being formed, and it is unlikely that 

 they will increase sufficiently to cause rain during the 

 succeeding twenty-four hours. A red sky in the morning 

 shows, however, that the droplets around the dust 

 particles are being protected from becoming smaller by 

 a blanket of overlying moisture, otherwise they would 

 soon be driven into vapour again by the rising sun's 

 rays. The conditions are, therefore, favourable to the 

 growth in size of the droplets, and rain will probably 

 fall during the day. 



The association of a red sky with weather thus admits 

 of complete physical explanation, but it is not at all 

 necessary that the reason should be forthcoming for 

 every fact. It is, however, desirable to distinguish 

 between accurate observation and traditional belief 

 which asks for acceptance without inquiry into the 

 evidence of its truth. 



In the prediction of good and bad seasons, meteoro- 

 logists are to-day little in advance of the priests and 

 astronomers of ancient Egypt or Assyria ; for the only 

 forecasts that can be made are based upon the recurrence 

 of cycles of changes. Unfortunately, this is true to a 

 great extent because, although the knowledge of periodic 

 changes in the earth's meteorology over a cycle of years 

 is of a more exact nature than formerly, the fundamental 



