Earth's Nonadiabatic Atmosphere. 527 



6000 to the isothermal layer, and then becomes negative and 

 probably of increasing amounts. The tendency to negative 

 values sets in high over the equator and then diminishes in 

 elevation towards the poles. 



The pressure P is higher over the middle Tropics (latitude 

 20°) than over the equator or over Europe, and shows the high 

 pressure belt to the level 17,000 meters. The barometric pres- 

 sure is B — Y/gp m , where p m — 13595*8 the density of mer- 

 cury in kilograms per cubic meter, and g = 9'806() the accel- 

 eration of gravity in meters per second. At Lindenburg the 

 pressure difference between computed and observed values 

 shows that the observed values range from mm to l*5 mm larger 

 in the strata up to the highest altitude reached ; the observa- 

 tions in the Tropics are more nearly accordant, the observed 

 pressures being - mm to - 5 mm smaller than the computed. 

 The pressure over Europe is less than over the Tropics in all 

 levels, and hence there is an eastward wind movement over 

 Europe, and a westward movement over the Tropics between 

 the high pressure belt and the equator. The density p increases 

 from the equator to the poles up to 10,000 meters, and about 

 that level becomes higher over the middle Tropics than to the 

 north or south. 



The specific heat Cp is a variable ranging from 993 to 624 

 in these computations. This departure from the normal value 

 993 increases with the height, and by the formula, 



(Cp. - CpJ (T. - T ) = i(q; - q:) + (Q, - Q o ), 



is responsible for the kinetic energy of circulation per unit 

 mass and that of radiation. Since %(q* — q *) depends upon 

 the change of the velocity per 1000 meters, and since the 

 velocities are small in the isothermal layer, it follows that the 

 loss of heat (Q t — Q ) is due to the change in the specific heat 

 at high elevations together with the low pressure and density 

 dependent upon the observed temperatures. The complex 

 thermodynamic adjustment involved in the formulas cannot be 

 easily summarized in a sentence, but it will become clearer to 

 the reader by considering the following paragraphs. The 

 specific heat on the East African plateau was assumed to be 

 993 at the surface, though the elevation is 1110 meters above 

 sea level, and so the computed Cp is not readily compared in 

 the upper levels witli the data of the Tropics and Europe. It 

 is probably about the same in the free air at all levels in all 

 latitudes, when the plateau effect has been eliminated. 



The velocity q increases from the ground up to the iso- 

 tbermal layer, and then rather suddenly drops in amount in 

 this layer. The velocity generally increases from the Tropics 

 towards the poles, including latitude 52°, as is well known. 



