493 
it is extremely difficult to observe the atmosphere in the manner thus 
indicated. The barometer presents manifestations of the physical state 
of the entire column of air situated vertically over it; and although such 
indications are limited in their character, they are those on which we 
are obliged principally to rely. Rapid changes in atmospheric pressure in 
any locality, as well as unequal simultaneous pressures in places but little 
apart, are now regarded as very frequent precursors of storms; but if 
we inquire into the causes of such inequalities of pressure, we shall find 
that other indications of atmospherical disturbances may be expected. 
When two masses of air of unequal density are in contact, the ten- 
dency towards equilibrium will not be so immediate or so simple as most 
writers on the theory of winds appear to assume. Resisting forces of 
various kinds exist, among which some arise out of the friction of the 
currents among themselves, and against terrestrial objects; but the most 
important will result whenever it happens that the less dense mass of 
air possesses greater elasticity than the denser mass. Instable equilibrium 
may thus exist for a short time before either of the two masses could 
give way ; but while such a state of comparative equilibrium temporarily 
subsists between the two great masses, minor disturbances will arise 
wherever their strata came into immediate contact. The mingling of the 
heterogeneous strata will take place, as in all fluids, by a-process of con- 
vection. 
I have already shown the important influence exercised by this pro- 
cess in the mode of heating of the atmosphere, and I have attempted to 
point out its relations with the most ordinary horizontal and vertical 
movements of the air.* In order to observe such movements, I devised 
a species of vane, which shows the existence of an upward or downward 
motion in the air, as well as the horizontal direction of the wind. When 
the air is still, or moving parallel to the earth’s surface, the vane shows 
no upward or downward inclination, In the presence of two great 
heterogeneous masses of air, yet in a state of instable equilibrium, we 
would usually not have any indications of the convections of their par- 
ticles, except the occasional ascent of light objects from the ground, or 
the sudden indraught of smoke in chimneys. But such phenomena 
are necessarily vague. If we are entitled to assume that much vertical 
convection of the air, with comparatively little horizontal motion, fore- 
shadows an approaching disturbance, it follows that a better mode of 
observing vertical currents is desirable. During the three winter months, 
I have been occasionally observing vertical currents of the atmosphere, 
with the aid of such a vane; my observations have not been made at 
fixed hours, nor on every day ; but they have usually arisen whenever 
abnormal disturbances occurred. As a general rule, I have found that 
most of the storms which we experienced during the past winter have 
been preceded by violent vertical movements of the atmosphere. In most 
* “Reports of the British Association for 1858 ;” “Atlantis,” No.V.; and “ Philoso- 
phical Magazine” for May, 1860. : 
