CLIMATE—THE SYNTHESIS OF WEATHER 
relation of the circumpolar pressure (or wind) distribu- 
tion averaged over a short period (five days) with 
subsequent developments of the pressure pattern. 
In the first three of these methods the choice of ele- 
ment and the position on the globe are purely at the 
caprice of the investigator. They are in fact almost 
entirely blind speculation—hit or miss, if miss, try 
again. It may be said that, in the past, many of the 
discoveries in physics were based on such procedure, 
but in a subject such as climatology covering so wide a 
geographical field and with so many physical variables 
the chances for success by such methods are very small. 
The fourth method based on the movement of long- 
wave patterns has a dynamical basis and as such is ina 
different category. The movement of these waves and 
the meridional flux of air profoundly influence spells 
of weather, therefore in some manner these long waves 
and in particular the blocking anticyclones must be 
taken into account in any concept of physical and 
dynamic climatology. Nevertheless the path which 
should be pursued in the construction of such a clima- 
tology is not clear. 
Following the line of thought in the previous part of 
this article, we are led to consider the climatological 
map of any element as the representation of an integra- 
tion of the values of that element derived from tra- 
jectories which have their origins in certain source 
regions. Now the appropriate value of the element on 
any occasion is influenced firstly by the state in the 
source regions and secondly by any modification of that 
state introduced during its travel, whether by diffusion 
or radiation or change in pressure or by any other 
cause. 
Clearly we may consider the map for a single month 
in the same way, and we may review the change in that 
map by consideration of the change in the source regions 
or the change in the modifications introduced in the 
travel of the air masses, as well as by the frequency 
with which the air travels from the different sources. 
Such an approach to the problem of climatology has 
the merit that it endeavours to link the mechanism of 
climatic change with physical properties, and it is clear 
that if any advance were possible from this approach, 
the horizon of usefulness would extend greatly. 
The source regions of the Northern Hemisphere are 
enumerated in considerable detail by Petterssen [29]. 
They are generally recognised as being regions in which 
the surface of the earth has a homogeneous covering; 
there is the additional proviso that the air motion must 
be sluggish in order that the air mass may acquire 
uniformity of character. It is known from the synoptic 
charts that these source regions are liable to consider- 
able movement from one period to another, some indeed 
may be swept entirely away at times. Their boundaries 
in any case are rather indeterminate, yet the very fact 
of the use that is made of the air-mass characteristics in 
synoptic meteorology emphasises that there is a recog- 
nisable continuity in them and gives hope that the 
examination of the variation with time of the character 
of source regions is a practicable step. That such an 
973 
examination has not been made before now, so far as 
the present writer knows, is probably because a number 
of the source regions, and those the most important 
ones, are in locations that are by no means readily 
accessible, particularly to upper-air observation. How- 
ever, this difficulty is becoming less with the extension 
of information. 
Now that the drawing of synoptic charts of the whole 
hemisphere is a practicable process, not only daily but 
two or even more times each day, there would seem to 
be an urgent need for an analysis of the variations from 
day to day in the position, the imtensity, and the 
growth (or decay) of the major sources. Admittedly 
any such examination is liable to a measure of sub- 
jectivity, since the definitions of the source regions are 
elastic, but even with that limitation such an analysis 
would teach the meteorologist much. 
The next problem is that of the modification of air 
masses in translation over various surfaces. 
Belasco’s calculations of the modification of polar air 
travelling from Iceland to Britain and of tropical air 
travelling from the Azores are given in Tables II and 
III. From these figures it would seem that in summer 
the air in travelling from Iceland to Britain (say 1000 
miles) is warmed by about 5C or 6C all the way up 
from 900 mb to 500 mb, whereas in winter it is warmed 
by about 9C at 900 mb, but by only about 7C at 500 mb. 
On the other hand tropical air, in travelling from the 
Azores (say 1200 miles), is cooled by about 3C all the 
way up in winter, but by only 1C in summer at 900 mb 
and by about 3C at the higher levels. In these calcula- 
tions no allowance is made for change in direction of 
the wind streams at the greater heights, but the figures 
would seem to show the magnitudes of the changes. 
We still do not know to what extent these changes in 
temperature arise from convection of heat from the 
surface, from radiation, or from vertical movements; 
fields of deformation will have to be considered and the 
consequential effects on frontogenesis as suggested by 
Bergeron [7]. However, with the more complete modern 
maps of the upper air a survey of the changes of tem- 
perature (and possibly of humidity) along the tracks 
of the air masses becomes possible. By these means we 
may be able to set out more exactly the manner in 
which the air masses tend to be changed in relation to 
the surfaces over which they move, and the atmospheric 
processes affecting them. 
After these basic data have been determined, an ex- 
amination could be made of the life history of the air 
masses which combined to form the monthly mean 
values for individual years. A summary of such life 
histories in a suitable form would undoubtedly enlarge 
our knowledge. 
Severe winters, for example, are in many places 
chiefly due to the advection of air from intensely cold 
sources, but we do not know whether they are greatly 
affected by the intensity of the source, or whether the 
severity is the result merely of the persistence of the 
flow; on the other hand, the source may have moved 
