68 On the Rotary Action of Storms. 
He presents a theory to account for such motion, which it is not 
necessary now to examine. Dr. Hare has proposed another me- 
thod of accounting for tornadoes—a truly brilliant suggestion— 
of which it is only to be remarked, at present, that it proceeds on 
the assumption of a rush of air from all quarters to a central 
point. It has been attested also, that at large clearing fires in 
calm weather, creating centripetal currents, the whirlwind and 
mimic tornado have been produced. In accounting for the whirl- 
ing motion therefore, the central tendency of the air will be pre- 
supposed. 
In the case of a large fire kindled in an open plain on a calm 
day, a small circle about the fire is first acted on by the abate- 
ment of pressure on the side next the fire, and thus receives an 
impulse towards the common centre. As this moves in, the next 
outer circle loses support and begins to move. Each particle of 
air is moved at first by an impulse towards the centre, and during 
its approach to the central region it receives fresh impulses of the 
same direction; and if it comes from some distance its velocity 
is in this way accelerated, until it reaches the space where the 
horizontal is broken by the upward motion. It is obvious that 
particles propelled by such impulses would seek the common 
centre in the lines of its radii, and their horizontal forces would 
be neutralized by impact, if no cause for deviation was at hand. 
But the great law of deflection which affects the course of the 
winds, applies to the movements of these particles. The parti- 
cles which seek the centre from the northern points are deflected 
west, while those from southern points are deflected east. The 
whole rush of air from the northern side of the centre, coming like 
a breeze, bears west of the centre, while an equal breeze from the 
southern side bears east of the centre. ‘The consequence is that 
the central body of air, including the fire, is acted upon by two 
forces which combine to make it turn round to the left. These 
forces are aided by the deviation of the currents from the east- 
erly and westerly parts of the circle. 'The breeze from the west 
extreme inclines to the tangent of the parallel of latitude at its 
original place of repose, and therefore strikes south of the centre, 
into which the impulses it receives would otherwise carry it. 
The air from the east side also inclines toward the tangent of the 
parallel of latitude there, which is oblique to the north from the 
