Philosophy of Storms. 125 



given out in the condensation of vapor into clouds. Therefore, 

 when the air has descended some time in the middle, between 

 two ascending columns, the barometer will fall a little, or at least 

 not stand so high above the mean as it does on the outside of 

 the two clouds, and so the columns will be pressed towards each 

 other. 



If one of two neighboring columns should be greatly higher 

 than the other, its annulus may overlap the smaller one, and of 

 course the current under the smaller cloud will be inverted, and 

 the cloud which may have been formed over the column thus 

 forced to descend, will soon disappear ; for as it is forced down- 

 wards by the overlapping annulus of the more lofty column, it 

 will come under greater pressure, and its temperature will be thus 

 increased, and it is manifest that as soon as its top descends as 

 low as its base, it will have entirely disappeared, and in the mean 

 time the larger cloud will have greatly increased. 



As the air above the cloud formed by an ascending column is 

 forced upwards, if it contains much aqueous vapor, a thin film of 

 cloud will be formed in it by the cold of diminished pressure, 

 entirely distinct from the great dense cumulus below ; but as the 

 cumulus rises faster than the air above it (for some of the air will 

 roll off) the thin film and the top of the cumulus will come in 

 contact ; and sometimes a second film or cap may be formed in 

 the same way, and perhaps a third and fourth. When these caps 

 form, there will probably be rain, as their formation indicates a 

 high degree of saturation in the upper air. 



When the complement of the dew point is very great, (20° and 

 more,) clouds can scarcely form; for up-moving columns will 

 generally either come to an equilibrium with surrounding air, or 

 be dispersed before they rise twenty hundred yards, which they 

 must do in this case, before they form clouds. Sometimes, how- 

 ever, masses of air will rise high enough to form clouds, but they 

 are generally detached from any up-moving column underneath, 

 and of course cannot then form cumuli with fiat bases ; such 

 clouds will be seen to dissolve as soon as they form, and even 

 while forming they will generally appear ragged, thin and ir- 

 regular. 



Moreover, if the ground should be colder during the day, than 

 the air in contact with it, as it sometimes happens after a very 

 eold spell of weather, then ascending columns cannot exist, and 



