Evaluating the Surface-Temperatures of the Planets. 751 

 whence on putting r 1 = l — a 1 — t 1 we obtain 



a 

 R= -S. 



2~ 



h+% 



The values of t and a depend upon the glass. By way of 

 illustration let us take t = 0'6, a = 0*3. For radiation from a 

 surface under 100° 0. Melloni found that even ihm glass is 

 quite opaque. We have then t x = 0, and if we neglect 

 reflexion, probably small, ai = l. 



Then B=j£s = l-5S. 



oU 



If the glass were removed we should have 

 R = S. 



The temperature of the ground is therefore ^1*5 = 1*1 

 times as high under the glass as it is in the open. If, for 

 instance, it is 27° 0. or 300° A. in the open, it is 330° A. or 

 57° 0. under the glass. 



If the glass re fleets some of the radiation R then a x is less 

 and the ground temperature is still higher. 



If the ground, instead of being black, reflects a fraction p 

 of the incident sunlight, or has total albedo p, the formula 

 must be modified. If we take into account merely the first 

 reflexion from the ground and assume that the glass has 

 absorption a for it ; then we easily find 



a [a \ , 



"1 



w 



If we take p = 0'l the numerator is 0*78 instead of 0*75, 

 and if we assume the fourth power law for the low-tempe- 

 rature radiation emitted by the surface, the temperature is 

 about 1 per cent, higher. But the ground will probably 

 reflect a much smaller fraction of the whole spectrum, and 

 the correction for total albedo becomes inconsiderable. 



If we replace the sun by cloud the radiation is, on the 

 average, of much lower temperature, and / and a are much 

 nearer to ^ and a x . The value of R/S is then much nearer 

 to 1, and the covered ground has a temperature much less 

 raised above that of the open ground. This agrees of course 

 with common experience. 



