TTTEORETIOAL DISCUSSION. 257 



Wliatever may be the actual distribution of high and low land within the Antarctic, there 

 are two extremes between which it must lie: (a) the whole area at sea-level, and (b) the 

 whole, or the greater part, at some unknown elevation. 



I therefore propose to show by means of diagrams similar to the one prepared by 

 Meinardus and reproduced as figure 75 above, the vertical distribution of pressure in each 

 of these two extreme cases, and then combine them along a section of the Antarctic the 

 configuration of which we laiow to some extent. 



Figures 78, 79 and 80 represent these hypothetical vertical sections of the Antarctic 

 south of latitude 50° S. In these diagrams I have chosen a much more contracted vertical 

 scale than the one used by Meinardus. This is in order to reduce as far as possible a 

 fallacious impression whi(h one receives on examining Meinardus's drawing. Any one looking 

 at figure 75 cannot help associating the great mass of the continent, as there represented, 

 with an isolated mountain rising from a level plain. In the latter case the moimtain would 

 have very little effect on the pressure distribution, which would be very nearly the same 

 near to the mountain as at some distance from it in the free air. Hence one uncon- 

 sciously accepts Meinardus's drawing without reaUsing that the great Antarctic highlands 

 will affect the pressure distribution in a very different manner fi'om that of an isolated 

 mountain peak. If, as we have good reason to believe, the Antarctic Continent is an 

 elevated tableland many hundreds of miles across, it is necessary to construct our 

 diagram to give this impression and this can best be done by keeping the vertical scale 

 as small as possible. When this is done one is no longer tempted to run isobars 

 up to and over the surface without pausing to consider how the tableland will affect the 

 pressure. 



In these diagrams an attempt has been made to represent the pressure changes in the 

 atmosphere by means of Unes. If one calculated the air pressure at each point of the 

 diagrams and joined all pomts at a given pressure by means of a line, we should have a 

 series of isobars similar to those with which we are so famiUar on weather charts, except 

 that they would represent pressures in a vertical instead of a horizontal plane. Such lines 

 would i-ise and fall as one passed through regions of high and low pressure. Unfortunately, 

 however, the actual change in height of such lines is far too small to be represented on our 

 diagrams. For instance, if the sea-level pressure at the Pole were 10 mm. higher than at 

 60° S., the isobar which touches the sea-level at 60° S. would be raised less than 90 metres 

 at the Pole. This amount is much too small to be shown on any diagram, which extends 

 to 8,000 metres in the vertical direction. In order, therefore, to let the eye easily take in the 

 changes in pressure, the vertical variation of the isobars is greatly exaggerated. This has 

 the disadvantage that the lines drawn to represent the pressure changes are no longer true 

 isobars, for they do not give the true pressure at each height. All that one can say is 

 that the lines on these diagrams show by their rise and fall how the true isobars rise and 

 fall. 



It must also be clearly understood that the position of the lines in these diagrams is 

 not calculated, they are simply sketched and made to rise and fall according to the conditions 

 which they are drawn to represent. The vertical distance between the lines is constant 

 over each place, but it varies from place to place according to the mean tempera- 

 ture of the air over that place. Thus the vertical distance between the lines in all the 

 diagrams is greatest at the edges on account of the relatively high temperature at 

 50° S. latitude, and least over the Pole where the temperature is supposed to be 

 least. Again the vertical separation of the lines has not been calculated, but convenient 



33 



