METEOROLOGY 



533 



because we know from the results for temperature for equatorial 

 regions that convection does go on there up to a level of 1 7 kilo- 

 metres before the stratosphere is reached. 



" We can go further and consider what would be the pressure 

 at the surface if a column of air some 10 or 12 miles in diameter, 

 for example, were replaced by the air which was saturated 

 at the surface and thrust up into the heights. We can compute 

 the pressure-difference between an interior column of air so 

 defined and its environment, neglecting the humidity of the 

 air, in the computation of the density, but allowing for it in 

 the change of temperature. It appears that in these circum- 

 stances the difference of pressure between the exterior 

 column and the environment would be as much as 81 millibars 

 at the surface, and gradually diminish from that to 8 m.b. 

 at the level of 10 kilometres and to nothing at the level of 

 15 kilometres. We can set out all these facts respecting 

 saturated air and its possible relation to its environment in 

 a table as follows : 



NORMAL PRESSURES AND TEMPERATURES IN EQUATORIAL AIR 

 (BAT AVI A) WITH THE TEMPERATURES OF AIR SATURATED 

 WITH WATER VAPOUR AT 300 a., AND REDUCED WITHOUT ANY 

 SUPPLY OF HEAT TO THE PRESSURE AT THE UPPERMOST 

 LEVEL. WITH THE DIFFERENCES OF PRESSURE AT DIFFERENT 

 LEVELS BETWEEN THE NORMAL AIR AND THE COLUMN OF 

 SATURATED AIR. 



" It may be noted that this form of instability is very much 

 dependent upon the temperature of saturation of the air, 

 and it is therefore limited to regions where the air is not only 

 very hot, but also very moist. It should also be noted that 

 the difference of temperature between the rising air and its 

 environment reaches a maximum of 13 a. at 10 kilometres. 

 We can see from this table that under suitable conditions the 



35 



