238 PRESSUEE, WINDS AND WEATHER. 



temperature should have the lower mean pressure, and also that this apparently inverted 

 pressure gradient should increase as the temperature difference increased. 



The cause is now obvious. Framheim is on the east of the Barrier and shares in the 

 temperature of the Barrier. The low temperature of the Barrier is the principal cause of the 

 high winds over the west of the Barrier, and the high winds over the west of the Barrier, 

 being constrained to travel parallel to the Western Mountains, produce the pressure difference 

 between Cape Evans and Framheim ; at the same time the high winds through McMiurdo 

 Sound constantly remove the layer of cold air which forms during calm weather and so 

 raises the mean temperature at Cape Evans. Thus it is the very fact that Framheim is 

 cold which causes the relative high pressure and temperature at Cape Evans. 



Daring the winter the action is intensified, the temperature difference between the Barrier 

 and Ross Sea is increased producing greater air motion and with it a greater difference in 

 temperature and pressure at the two stations. 



Weatlier. — It remains now for us to consider the type of weather associated with each 

 type of pressure distribution. 



We have seen from our previous discussion of the actual data that southerly winds are 

 associated with cloud and snowfall while northerly winds are associated with cloudless skies 

 and absence of jjrecipitation (pages 11 and 151). These are the conditions which we should expect 

 from the pressure distribution associated with the two types of wind. The southerly type of 

 pressure distribution is practically that of a cyclone with its centre situated over the south 

 of the Ross Sea. Also the crowding together of the stream lines of air-flow over the north- 

 west corner of the Barrier mast cause forced ascensional air motion in that region. Thus 

 cloud and precipitation would be the natural consequence of such a pressure distribution and 

 of such constrained air motion. Thus the weather associated with blizzards is explained. 

 When northerly winds occur at Cape Evans there is a high pressure area over the Ross Sea. 

 In such a case the whole of the Ross Sea area is practically under anticyclonic conditions. 

 Also the high northerly wind itself indicates that considerable outflow of air is taking place 

 from the region of high pressure. This air can only be supplied by descending currents which 

 would effectively prevent the formation of cloud and the precipitation of snow. 



An important characteristic of the blizzards is the suddenness with which they frequently 

 commence. 



It appears to me that two factors are responsible for this feature. When the pressure 

 difference which is ultimately to result in a blizzard becomes established gradually the air 

 over the Barrier slowly gets set into motion, but the cold surface layer tends to remain 

 stationary while the upper layers slide over it. Thus the upper air may be in rapid motion 

 while a calm continues on the ground ; the cloud observations frequently showed this to be 

 the case. This condition cannot continue indefinitely, the cold surface layer gets disturbed, 

 broken up locally and finally swept away. Then the blizzard commences with a sudden burst 

 of wind and an appreciable rise of temperature. The pressure difference, however, is not always 

 established gradually. When a pressure wave travels across the country at about 30 miles 

 an hour it sets up its pressure gradient faster than the air can get into motion. In these 

 conditions the air behind is moving faster than the air in front, and so no air motion takes 

 place at each point until this mass of moving air arrives with a sudden burst. Similar effects 

 are seen when tidal waves enter an estuary and are impelled forward by the water behind 

 faster than the water in front can be set in motion, the result being the well-known pheno- 

 menon of the tidal bore. 



Now both these factors, the cold layer and the travelling pressure wave, are much more 

 developed in the winter, and a study of the records shows that nearly all the sudden com- 

 mencement of blizzards occurred in the winter and early spring months. In the summer 



