GRAND CURRENTS OF ATMOSPHERIC CIRCULATION. 
667 
To make this clearer we may notice that if a buoyant central column were for a 
moment existing surrounded by non-rotative air having greater pressure in its lower 
parts than that in the column at the same level, that column could not continue its 
existence. The outer air with its greater pressure would press in on the column, and 
would increase the pressure in its substance instantly, hut the weight of the upper 
portion of the buoyant column would be inadequate to resist the upward thrust so 
produced in the lower part, and so the lower parts would shoot those above them 
upwards with violently accelerating motion. Through the rushing upwards so gene¬ 
rated a breaking up of the column would supervene, and its substance would scatter 
itself in rolling masses among the surrounding air; and the two commingling would 
ascend gradually, and at the same time the pressure of the surrounding air would 
communicate itself to the region where the base of the column had been. 
But now, on the other hand, if the mass of air around the central buoyant column 
be whirling, it will keep itself out by the centrifugal tendency accompanying its own 
rapid revolution, and so will not press in upon and break up that central column of 
air of diminished pressure, and thus the abatement of pressure at the foot of the 
column will be maintained and will become further intensified. 
To Bedfield is due much credit for his able and long-sustained labours in collect¬ 
ing and correlating observed facts as to cyclones and the smaller kinds of whirlwinds. 
He gathered and published # a very interesting collection of accounts of violent 
columnar whirlwinds which formed themselves over large fires of circular masses of 
brushwood, the flame and smoke in each case ascending as a lofty rotating column ; 
and this has had part in suggesting to me some elements in the theoretical considera¬ 
tions here brieffy sketched out. In his remarks on these whirlwinds he emphatically 
brought into contrast the distinction between the flames and smoke ascending with¬ 
out whirling motion from hot furnaces and various ordinary fires, on the one hand, 
and, on the other hand, the revolving columns of flame and smoke often met with in 
those great fires of brushwood in the open air. By his various researches into the 
actions and effects of great storms, Redfield contributed more, perhaps, than any 
other man to the advancement of observationally-derived knowledge of their cyclonic 
character and features. 
Wild and fantastic notions were, however, afloat in those times as to the origin 
of cyclones. Thus Piddington, in his well-known work entitled the ! Sailor’s 
Hornbook,’ even in the edition so late as 1860,t in stating his resultant opinions 
and conclusions, makes such statements as the following :— 
That he considers cyclones to be flat circular disks which may be formed at the sides 
and upper and lower surfaces of clouds, and which, once formed, may either rise 
* “ Some account of Violent Columnar Whirlwinds which appear to have resulted from the action of 
large Circular Fires,” by W. C. Redfield. Read before the Connecticut Academy of Arts and Sciences, 
Jan. 22,1839. Printed in the ‘ American Journal of Science and Arts ” (Silliman’s), 1839, vol. 36. p. 50. 
f The ‘ Sailor’s Hornbook,’ third edition. 
4 q 2 
