Mr. W. Brown on the Oscillations of the Barometer. 259 



rows at ABCE, Plate V. fig. 4*. Now the upper current 

 by which these storms begin within the tropic is from north- 

 east, but after this change in their course the upper current 

 is from south-west and the lower one north-east; let us then 

 suppose A the station at which the storm first arrives after the 

 change in the direction of the atmospheric currents, the upper 

 one being now S.W., the rarefaction of the atmosphere is first 

 produced here by the flowing of the air from it towards south, 

 according to the recession of tropical storms, and into this 

 rarefaction the south current descends, but instead of restoring 

 the atmospheric pressure, it still further increases its diminu- 

 tion by its momentum, and extends northwards and eastwards 

 to B. But it is evident that the rarefaction will extend to a 

 much greater distance in front of the storm than on the west- 

 ern side of itf, hence the pressure on the line A C, fig. 3 (but 

 now less inclined from C B), overcomes at length the force of 

 the south wind at A where the depression of the barometer is 

 the greatest and the wind sets in from N.W., as shown in the 

 figure, and restores the pressure, so as to maintain by again 

 raising the height of the atmospheric columns at A, the velo- 

 city of the upper current now flowing to more northerly lo- 

 calities. 



But whilst this is going on, the north-east wind is blowing 

 at C and E, being produced there by two causes, its recession 

 from A, and the flow of air produced by the rarefaction at the 

 limit of the south wind, as at E, fig. 1, Plate V.; and as the 

 direction of the upper current is from south-west, the diminu- 

 tion of the atmospheric pressure is of course carried on in that 

 direction ; and hence a rarefaction is maintained, into which 

 the south current flows from columns at a point eastward of 

 A, as B, which as yet is not subject in so great a degree as 

 A, to the opposition of the air on the line A C of fig. 3. But 

 the south wind meeting the north, the wind is S.E. But as the 

 line A C advances, its pressure prevails both at B and C, and 



* I would just remark in passing, how well these positions would accord 

 with the hypothesis of the wind moving in a whirl, could the fourth quarter 

 wanting, when the wind should be south-west, be found,- but scarcely one 

 observation of tlie wind from south-west occurs in the direct path of the 

 storm, for when the wind is stated as blowing from south-west, it is either 

 previous to the change in the progressive motion, or to the west of the 

 " hurricane tract of the storm." The occurrence of the southerly current 

 as a south-east wind in front of the north-east, as exhibited by the dia- 

 grams, is a grand illustration of § 11. - 



f This is evident from the direction of the progressive motion, but an 

 observation given in the data of the storm of 1821 (Law of Storms, p. 16) 

 shows how abruptly the storm terminates on the west, for ''at Wilmington 

 there was no gale," but " a severe gale was experienced thirty miles out* 

 side of the American coast off Wilmington (N. Carolina)." 



T2 



