26 SECTIONAL ADDRESSES. 
them so that they circulate all round the depression; we find, onthe contrary, 
that they are discontinuous, the stream lines in certain parts meeting 
the stream lines in other parts almost at right angles. Also we find large 
discontinuities in the temperature, each set of stream lines having its own 
temperature. Also we find that the areas of rainfall are not confined to the 
central regions, but are broad bands radiating from the centre like spokes 
in a wheel, showing that the ascending air is not taking place mainly in the 
central region. 
As the result of recent work we now recognise a structure in a cyclone 
which was unknown a few years ago. We owe this new knowledge largely 
to the work of J. Bjerknes and his assistants in the Bergen Geophysical 
Institute. 
If we examine a synoptic chart on which a newly formed rapidly moving 
cyclone is delineated, we can mark out in the southern half of the cyclone 
a region in which the air is definitely warmer than the air in the remainder 
of the field (Fig.5). This air is moving as a south-west wind. If we draw 
the stream lines in this region we find a broad stream of air which comes to 
a sudden end on the chart at a line which starts at the centre and is curved 
in a south-easterly direction. This line is called by Bjerknes the ‘ warm 
front.’ This line obviously marks out where a surface of discontinuity, 
similar to those already described, cuts the surface. Beyond it, at the 
surface, the wind is easterly or north-easterly, and the air is much colder 
than the air in the south-westerly stream. We are here dealing with a 
warm current meeting the flank of a cold current and mounting up over it. 
This supposition is supported by the fact that the rainfall occurs on the far 
side of the line at which the south-westerly current leaves the ground. 
The rainfall is also general and steady, which one would expect if a current 
of air is rising slowly up an inclined plane. Usually a large part of the 
southern half of a cyclone is occupied by the warm south-westerly current 
and the line of discontinuity which I have just described is in the south- 
east quadrant. In the south-west quadrant the flank of the warm current 
is in its turn attacked by cold air flowing from the north or north-west, 
and where the two meet we have another line of discontinuity called the 
‘cold front.’ The cold current cannot flow over the warm current, but 
tends to push its way under. The surface of discontinuity at the cold 
front is very steep and is very unstable. In consequence, we have here 
violent squalls with heavy local showers which are in marked contrast to 
the steady rain at the ‘ warm front.’ 
When we come to trace the origin of the air which meets at the cold 
and warm fronts, we find that the warm current can generally be traced 
back to southerly regions and often to the tropical high-pressure belts ; on 
the other hand, the cold air can equally well be traced back to polar 
regions. Hence it has become usual to describe the air masses which meet 
in cyclones and are separated by the great surfaces of discontinuity at the 
cold and warm fronts as polar and equatorial air. The polar and 
equatorial air have each their own characteristics which they retain for 
a very long time. These characteristics are interesting and important. 
On its long journey from polar regions the polar air starts cold but 
stable; as it comes south over the Atlantic Ocean the lower layers become 
warmed up, because they are always passing over warmer and warmer 
surfaces. Thus the temperature lapse rate increases and the air becomes 
