258 GENERAL AIR CIRCULATION. 



distances taken in order to bring out clearly tlie general effect of raising and lowering the 

 temperature. 



Figure 78 has been drawn to represent what would be the pressure distribution if the 

 whole area within the Antarctic were at sea-level. The lowest pressure at sea-level is shown 

 at 60° S. and the pressure rises from this point to the Pole. The vertical distance between 

 the isobars depends on the temperature, hence this vertical distance decreases from the edges 

 of the diagram to the centre. The consequence is that the rise of pressure south of 60° S. 

 at sea-level is compensated in the upper atmosphere where the pres.sure decreases all the way 

 from 50° S. to the Pole. In accordance with Meinardus's calculations I have made the 

 height at which anticyclonic conditions over the Polar area give place to cyclonic to be 

 between 2,000 and 3,000 metres. The pressure distribution in this diagram is practically the 

 same as that shown in figure 75 prepared by Meinardus and that it represents the condi- 

 tions which would hold if the whole area withLir the Antarctic were at sea-level no one will 

 doubt. But what does it signify ? Simply that the ice surface causes the sea-level pressure 

 at its centre to be higher than at its edges. 



This comes about from two causes : (o) the geographical position of the ice surface, which 

 being concentric with the Pole has a lower temperature at its centre than at its margin ; 

 (6) the formation of a glacial anticyclone due to excessive radiation which Hobbs' work 

 shows forms over any large ice surface whether it is concentric with a pole or not. These 

 two causes acting together would produce a relatively intense anticyclone over the Antarctic, 

 if it were at sea-level. 



We will now turn to the second extreme case and consider the consequences of a high 



tableland within the Polar area. Figure 79 has been constructed to represent this case. The 



tableland has been represented as occupying the whole area within 70° S., and to be at a 

 uniform height of 2,000 metres. 



What is the pressure distribution likely to be over such a vast tableland ? Meinardus 

 assumes that it would be the same as if there were no tableland there at all. With this con- 

 clusion I cannot agree. The pressure distribution in the upper air without the tableland 

 shown in figure 78 depends on the presence of heavy cold air in the lowest atmosphere, 

 when this is removed by the land mass the upper air pressure distribution must be altered. 

 I can see no reason why the general pressure distribution over a level surface at 2,000 

 metres should be different from that of a similar surface at sea-level, so long as the raised 

 surface is of sufficient extent to be the governing factor in the pressm'e distribution. The 

 causes which produce the high pressm-e over the Antarctic specified in (a) and (b) above 

 both come into play on this raised tableland and it is therefore reasonable to conclude that 

 the pressure will be higher over the centre of the tableland than over its margin. In other 

 words that an anticyclone will exist over the tableland. This has been represented in figure 

 79. Comparing the isobars sho\vn in figures 78 and 79 we see that the lowest is similar 

 in both except that in the latter it only extends as far as the edge of the tableland near 

 to which it takes a shghtly greater upward turn. The second Hne in figure 78 is nearly 

 horizontal within the Polar region. In figure 79, however, as it approaches the upper margin 

 of the tableland it rises and then passes over the surface indicating the increased pressure 

 towards the centre. If we assume that the free-air temperatures are approximately the same 

 whether the tableland is present or not, the vertical distance between the isobars will be 

 the same at each latitude in figures 78 and 79. The lines have been drawn on this assump- 

 tion and it is seen that at about 2,000 metres over the tableland the isobars are again 

 nearly horizontal and above this height cyclonic conditions agaiji hold. The difference between 



