114 Mr. H. Jeffreys on Periodic 



I£ we neglect the effects of the velocities and put k constant 

 for each medium, we have 



<g-£V 2 V=/. (2) 



At a boundary we have the condition 



['££.-'■ •■•••» 



where F is the rate of supply of heat from external 



sources to unit area of the surface, 

 the numbers and 1 refer to the two media, 



K is the product of k, the density, and the 

 specific heat, 

 and dn is measured along the normal from to 1. 



Further, V is continuous at every boundary. 



The supply of heat to the atmosphere takes place in several 

 ways, of which the chief are : 



(1) The solar radiation is largely absorbed as it comes 



down. 



(2) Radiation is directly reflected from the surface of 



land and sea, and is again absorbed as it rises. 



(3) When the surface of the earth is heated, it radiates 



heat, which is absorbed by the air. 



(4) The heat of the earth is redistributed by convection, 



the mechanism of which is by eddies, and by 

 radiation between masses of air. 



The supply of heat from external sources to the boundary 

 between land and air is limited to the portion of the solar 

 radiation that reaches it. This amount is, on the whole, 

 probably about -J of what is absorbed by the air. The 

 redistribution of heat within the solid earth is to be 

 attributed to conduction alone. 



With sufficient accuracy for our purpose we can also 

 suppose that all the radiation supplied to the ocean is 

 absorbed at the free surface. It is well known that a very 

 small thickness of water is required to stop the dark heat 

 radiation ; light is more penetrating, but H. Wild gives * 

 for distilled water and white light the transmission co- 

 efficient 0*98835, the unit of length being one Parisian 

 inch ; so that, after passing through a layer of distilled 

 water 6 metres thick, the intensity is diminished to 3-A^ of 



* Poggendorff's Annalen, vol. xcix. 1856, p. 272. 



