107 
1913-14.] Principia Atmospherica. 
a sudden change of wind between 1100 and 1500 metres height from a 
reasonably steady wind from nearly due south into one almost as steady 
from due north, the change being accomplished within half a kilometre. 
The analysis in this case shows for the layer between 500 and 1100 metres 
a temperature distribution in isotherms nearly north and south with the 
warmer air on the east, and above 1500 metres an entirely different dis- 
tribution with isotherms nearly east and west, and cold to the northward. 
The intermediate layer, 400 kilometres thick, showed a very rapid increase 
of temperature to the west — as much as 7° C. per hundred kilometres. 
The complete arrest of the northerly current and production of a calm 
by the annihilation of the gradient between 1100 and 1500 metres is very 
remarkable, but nevertheless a real fact. The accompanying temperature 
difference is probably due to a strong temperature “ inversion ” at a height 
of about 1500 metres at the place of observation and of 1100 metres at a 
place 100 kilometres distant to the west. On that occasion it lasted for 
some time, as it was found an hour afterwards by a second balloon ; but it 
must be remembered that it was a region of no velocity, and therefore the 
relatively warm and cold airs were not moving. In order to get them 
away, either convection must take place or a gradient must be created. 
Proposition 6 . — The General Circulation of the Atmosphere in the 
Northern Hemisphere. 
The reasoning in this proposition is more general in form than that of 
the foregoing propositions. The extension of our knowledge tends more and 
more to strengthen the conclusion that the proximate cause of the varia- 
tions of pressure in the region of the British Isles must be looked for in the 
layer at a height of about 7 to 9 kilometres ; it is the layer of maximum 
wind- velocity just under the stratosphere, and it is also the layer within 
which must be located a rapid transition of slope of temperature. Below it, 
as set out in Lemma I., the slope of temperature follows the slope of 
pressure ; above it, the slope is in the opposite sense. The mechanism by 
which the changes of pressure are produced is unknown ; but this much is 
apparently true, that within the layer referred to, the relation between the 
pressure and temperature of the air at two places on the same level is that 
of adiabatic expansion. Above the critical layer where this relation holds, 
the air in the high-pressure area is “ too cold,” and below it, for 5 or 6 kilo- 
metres at least, it is “too warm.”* We may suppose that air becomes “too 
warm ’* by the dynamical warming of downward convection, and, perhaps, also 
* See Journal Scottish Met. Soc., 1913, loc. cit. 
