Lesley.] - 3g2 [April. 



the southwest ; and at the close, from the northwest ; but in neither 

 case from more than a few points to the north or south of the mag- 

 netic west. There were but two hours during the whole voyage 

 during which the wind blew from any point east of south. At the 

 extreme of violence for each gale the wind blew dead ahead. During 

 the intervals of a few hours between the gales, the sky would clear, 

 and the wind come quietly in from the west, until its shifting to the 

 southwest gave the signal for the opening of the next blow. 



The barometer fell rapidly and low at the beginning of each gale, 

 and rose more slowly afterwards. If Captain Galton's hypothesis of 

 a reversed descending cyclone to accompany a rising barometer after 

 the direct ascending cyclone with a falling barometer, be intended to 

 apply to all atmospheric disturbances, small as well as large, some 

 exhibition of this supposed phenomenon should have been made by 

 this series of gales. (Phil. Mag. No. 174, p. 225, Proceed. R. 

 S., December 18th, 1862.) " It is hardly possible," he justly urges, 

 " to conceive masses of air rotating in a retrograde sense in close 

 proximity, as cyclonogists suppose, without an intermediate area of 

 direct rotation, which would, to use a mechanical simile, be in gear 

 with both of them, and make the movements of the entire system 

 correlative and harmonious." But we have this very conception 

 realized before our eyes, habitually, in every series of eddies on the 

 surface of a stream. If the cyclonal columns were stationary and 

 contiguous, some intermediate disturbance, analogous to a pinion be- 

 tween two cogs, must occur; and if the interval have a diameter 

 equal to that of each cyclonal column, the disturbance might per- 

 haps assume a simple reverse columnar form and motion at its cir- 

 cumference, opposed as they would be by the vertical stability of its 

 axis. But if the two supposed stationary vorticals were nearer than 

 one diameter, the disturbances of the interval must become very 

 complicated, and hardly recognizable on a chart by any simple system 

 of curves. 



But as, in fact, cyclonic columns, such as those under discussion, 

 have a forward slip, like the eddies in water, the line of motion of 

 any particle of one column (or of its observed ba,se) is by no means 

 in the curve of a volute; but, like the path described by a particle 

 of the earth's surface in its course round the sun, approximates a 

 straight line, oscillating from side to side without an epicycle. If, 

 therefore, in a series of cyclones of retrograde projection, the inter- 

 vals were filled with other cyclones of direct projection, there could 

 be no concealing the fact; and at least traces of such an interval 



