COMPARISON OF THEORY WITH OBSERVATIONS. 



231 



On comparing the curves and winds shown in figure 74 with those shown on the plates 

 in Volume II, we see a most striking resemblance. Examining first the bottom curves, giving 

 the pressure diiierence between Cape Evans and Framheim and the winds at Cape Evans, 

 we see that on both the blizzard winds occur when the curves rise, and the northerly winds 

 when the curves fall. We also see that the blizzard commences on figure 74 just when the 

 pressure difference between the two stations is zero. From the actual observations as shown 

 on the plates it appears that a blizzard commenced whenever the curve approached its zero 

 and seldom at any other time. On the other hand northerly \\ands, as a general rule, com- 

 menced when the difference curve reached its maximum. 



Turning now to the relationship between the winds and actual pressure at each stations 

 we shall also find important resemblances between the theoretical and the actual conditions. 



According to the theoretical curve, figure 74, S.E., S. and S.W. winds at Framheim 

 are associated with a falling barometer at that station, while easterly to northerly winds 

 occur when the pressure is rising. The actual observations plotted on the plates show the 

 same relationship. There are many exceptions, and we have seen that we are to expect 

 them, but there can be little doubt of the general relationship, see in particular plates 4, 

 5, 10, 11 and 22. According to Mohn S.E., S. and S.W. winds occurred with a rising 

 barometer in 13.5 cases and with a falling barometer in 144 cases ; while N.E. and £• 

 winds occurred in 181 and 143 cases respectively. The differences are not large but they are 

 in the right direction. If one considers only periods during which the pressure changes are 

 large the result is much more satisfactory. 



At Cape Evans, according to the theoretical curves, northerly winds occur with a falling 

 barometer and southerly with a rising barometer. This relationship can be seen on the plates 

 particularly with the northerly winds. It is, however, clearly brought out by calculating the 

 average rise and fall of the barometer with different winds. The barometer changes during 

 four hours about each wind observation have been tabulated for 17 months with the follow- 

 ing result : — 



Table 125. 

 Gcqje Evans change of barometer and ivind. 



We thus see that the theoretical relationship between the winds and pressure does agree 

 in the main with the actual observations.* 



The pressure curves on the plates show that the actual waves which travel across the 

 area vary greatly in amplitude and phase and we have considered only one in which 

 a = -3" and <p = 45°. If we investigate other waves in the same way we find that the 

 theoretical northerly and southerly winds at Cape Evans commence and end at very nearly 

 the same positions on the difference curve, but vary much more on the actual pressure curves. 

 This agrees with the fact which we have just discussed, that the actual v.-inds are not so 

 closely related to the pressure as to the pressure difference between Cape Evans and Framheim. 



* See, however, paras, (4) and (5) on page 237. 



