280 Mr. W. Brown on the Storms of Tropical Latitudes. 



most violent therefore where this is the greatest ; and thus 

 arises the great violence of the storms on the American coast, 

 compared with that in general of British storms. 



If then I have been successful in my attempt to establish the 

 views set forth in this essay, all the more important phasno- 

 mena of the wind may be simply classed with those of the 

 southerly and northerly breezes of our own climate, the 

 various directions of the wind arising either from the simple 

 deflection of one of the currents by the rotation of the earth, or 

 from the collision of the two currents after being so deflected ; 

 the barometer rising or falling, as the density of the one or 

 the momentum of the other prevails. But there yet remains the 

 ultimate step in this inquiry, to discover the immediate or 

 proximate cause of the descent of the upper current at any 

 particular time. 



The principal difficulty to be overcome seems to be, that the 

 descending air is warmer, and therefore lighter than that pre- 

 viously occupying its place ; there must therefore exist some 

 cause capable of producing a descent of air in opposition to 

 its density. Such a cause may perhaps be found in the dif- 

 fusive tendency of aqueous vapour. There seems no reason 

 to deny this property of perfect gases to the vapour existing, 

 as it is now admitted, as an independent constituent of the 

 atmosphere; and as the upper current contains the largest 

 quantity of vapour, there will exist a tendency in the air of 

 the higher and of the lower currents to intermix : according 

 to the conclusions deduced by ThomasThompson (Phil. Mag. 

 vol. iv. p. 321) from Prof. Graham's law of the diffusive force 

 of gases, this force remains throughout all stages of the inter- 

 mingling of the gases to be inversely as the square root of their 

 densities. It would seem therefore in the present case to re- 

 quire, in order that the upper air should be made to descend 

 and the lower to ascend, that the quantities of vapour in the 

 two currents (supposing the air of each to be of the same tem- 

 perature) should be such, as to make the square root of the 

 number expressing the density of the mixture of air and 

 vapour of the lower current relatively to that of the higher, 

 equal to the number expressing the density of the dry air of 

 the same current relatively to that of the higher, at their real 

 temperatures. This however would require too great a dif- 

 ference to be allowed; but if the air itself does not at first 

 descend, the vapour, by infiltrating through the particles 

 of air, and arriving amongst comparatively cold air* will be 

 partially condensed (perhaps the cause of the rain which so 



* The difference of temperature arising from the difference of elasticity 

 of air of unequal elevations is here left out of consideration. 



