28 SECTIONAL ADDRESSES. 
and Novaya Zemlya. These deflected currents move southwards into 
the region of the warm westerlies as tongues of cold air set at right angles 
to the prevailing westerly winds. These tongues of cold air act like barriers 
in a stream; the warm air is forced to rise over them, with a consequent 
increase of pressure on the windward side and a decrease on the lee side. 
Around the region of low pressure thus formed the air circulates ; the 
tongue is deformed, the tip being bent eastwards and a cyclonic depression 
is produced. 
There is probably a portion of the truth contained in each of these 
theories, but they both fail because they make cyclones phenomena of 
the lower atmosphere, while we have good reason to believe that the air 
in the stratosphere plays a large part in the formation of cyclones. The 
reasons for this belief are many, but two will suffice. The pressure 
changes associated with cyclones are absolutely as great at the base of 
the stratosphere as they are at the earth’s surface, thus indicating that as 
large masses of air move within the stratosphere as move below it. And 
secondly, Goldie has shown that the stratosphere over polar air, even in 
middle latitudes, has the same height and temperature as the stratosphere 
in polar regions, while the stratosphere over equatorial air in the same 
way retains the characteristics of the stratosphere in the low latitudes 
from which the air started. 
This has led to the idea that the movements in the atmosphere which 
transport polar and equatorial air to their meeting-place in middle 
latitudes are not confined to the troposphere, but that the air moves as 
a whole, troposphere and stratosphere travelling together. 
It must be admitted that we are still far from a complete understanding 
of the mechanism of cyclonic depression ; on the other hand, we do now 
know some features which are common to all depressions, and we have a 
much clearer idea of the source of the energy and the conditions necessary 
to their production. We have to imagine that in polar and tropical regions 
the air is relatively stagnant, and so has an opportunity to reach the state 
of thermal equilibrium appropriate to those regions. As already stated, 
the atmosphere is only a thin film, and we picture large areas or slabs of 
this film breaking away from their proper locality and moving into middle 
latitudes. Apparently the detached films move as a whole, at least to a 
considerable distance within the stratosphere. When two such portions 
of the atmospheric film come into juxtaposition they are not in equilibrium 
relative to one another and readjustment must take place. The surface 
of contact remains more or less intact, but the cold air tends to sink and 
undercut the warm air, while the warm air slides up the surface of dis- 
continuity. The whole motion takes place on the revolving surface of the 
earth, and the forces called into play by this revolution result in the air 
movement taking place in what appears to be a great vortex. The energy 
of the winds is derived mainly from the readjustment of the centre of 
gravity of the air mass considered as a whole, although the latent heat of 
condensation provides some additional energy by supplying heat to the 
warmer air as it ascends the slope of the surface of discontinuity. It will 
be admitted, I think, that this is a radically new idea regarding the 
mechanism of a cyclone. 
I might add a word here about tropical cyclones. All that I have said 
so far refers to the cyclonic depressions of middle latitudes. As to whether 
