Mr. W. Brown on the Oscillations of the Barometer. 253 



a previous storm, and not restored on the arrival of the advan- 

 cing storm, which approaches from atmospheric columns on 

 the south but little affected by the preceding one; which differ- 

 ence indeed is apparently the cause of the storm; it occurs 

 however after the returning or northerly current has set in in 

 the north, and a little raised the height of the barometer there; 

 and on its termination, the height of the barometer in the south 

 is reduced as low as that in the north, in some parts of which 

 it is depressed to a greater degree than before, the fall begin- 

 ning in the south and advancing towards north as the wind 

 itself. But not only this, the minimum height to which the 

 barometer is reduced, and the setting in of the north wind 

 (though always from north-west), occur first in the south, but 

 on the western side (by the south being now meant the parts 

 along the line H A of fig. 2, Plate V., where the upper current 

 is supposed, with regard to localities north of them, first to de- 

 scend); so that, in the south, the north wind is blowing and 

 the barometer rising, whilst the south wind is blowing and 

 the barometer falling in the north. 



The advance of the south wind northwards after its cessa- 

 tion on the south, admits of very easy solution on the suppo- 

 sition of the storm being carried forwards by means of por- 

 tions of air descending from an upper current flowing in its 

 proper position, and which the reduction of the height of the 

 atmospheric columns towards which it is moving, allows to 

 flow with its velocity little checked ; thus the south wind con- 

 tinues below, not by the force of the original impulse received 

 at its outset from the line H A, but by that of successive im- 

 pulses received during its course: moreover, the rapid in- 

 crease of the heights of the atmospheric columns in the south 

 by the influx of air from north-west, still maintains the upper 

 current of the atmosphere. 



The setting in of the northerly current from north-west in 

 the south, and the consequent rising of the barometer before 

 these changes occur in the north, may be explained by a consi- 

 deration of the form of the space, on which the depression of 

 the barometer previous to the occurrence of the second one 

 produced by the advancing storm, exists. Suppose the greatest 

 depression of the barometer to be produced along the line 

 A B, fig. 3, Plate V. (H A of fig. 2), according to § 15, and 

 the full effect of the first storm in depressing the barometer 

 to have taken place ; as that storm moved from A to B, the 

 pressure will be in some degree restored on the parts of this 

 line towards A; hence the depression will decrease from B to 

 A; but it also decreases towards C (§ 2) ; hence the form of 

 this space on the west will be somewhat like that bounded by 



