42 W. Ferrel— Cyclones, Tornadoes and Waterspouis. 
The effect of friction in tornadoes is much less than in cy- 
clones. A cyclone of considerable extent may be regarded as 
a disk, with a diameter many times greater than its depth or 
thickness, and hence the gyrations are very much retarded by 
friction on the earth’s surface ; but a tornado is rather a pillar 
of gyrating air with a very small base in comparison with its 
altitude, and hence the retardation of the gyrations by friction 
on the earth’s surface in this case is comparatively very small. 
The gyration of the air, therefore, except near the earth’s sur- 
face, is very nearly in accordance with the principle of the 
preservation of areas, and hence the lineal gyratory velocity is 
very nearly inversely as the distance from the center, and con- 
sepa upon the earth’s rotation being nearly insensible. 
n account of the rapidity of the gyrations near the center 
ornadoes occur when, fro 
state of unstable equilibrium already referred to. This may 
e near the earth’s surface, but is most usually up in the region 
of the clouds, where the air is saturated with moisture, and 
where consequently this state occurs most frequently, since it 
then requires a rate of diminution of temperature with increase 
of altitude usually less than balf as great as in the case of dry 
air. hen the atmosphere is in this state the air of the lower 
strata, from any slight disturbance, bursts up through the 
upper strata at some point, and the higher it ascends the greater 
is the difference between its temperature and density and those 
of the surrounding strata at the same elevation, and hence the 
greater the tendency to rush up at that point. But, as in the 
ease of the basin of water, if the initial state of the air were 
that of quiescence, there would be only a radial flow of air 
from all sides toward that point without any gyratory motion or 
diminution of tension at the center, and with very little violence 
of motion. The velocity of the ascending current in this case 
would not very great since the column of ascending air 
would soon spread out laterally, and become too great. In 
order to have, therefore, all the conditions of a tornado, it is 
necessary to have, besides the state of unstable equilibrium, 
the other conditions which, as in the case of the water in the 
