DIRECTION OF PREVALENT WINDS 23 



So far the winds follow the direction which they might be expected to take on 

 the theory of streams of heavy cold air blowing spirally outwards down the slope of 

 the inland ice from the summit of the main " ice divide." 



The normal direction for such winds should, theoretically, in the latitude and 

 neighbourhood of Boss Sea be from S.E. to N.W. or E.S.E. to W.N.W. There are 

 four notable exceptions : 



First, the easterly winds at Framheim. A disturbing factor in the simple 

 anticyclonic circulation is the warm gulf of Ross Sea, in considerable areas of which 

 water is open throughout the year. 



This gulf encourages the further westerly deflection of the northerly moving air 

 masses coming from off the land to the east of Ross Sea, and bends them around into 

 easterly winds.* 



The next departure from the prevalent direction is in the case of the winds which 

 blow out of Barne Inlet and Mulock Inlet on the west side of the Ross Barrier. 

 These winds have a general W.S.W. to E.N.E. dii'ection. 



They owe their direction probably to three factors : 



(1) A sag in the horst before it reaches that " horst within a horst " the Royal 

 Society Range. This allows the cold air of the plateau near the Britannia Range, 

 dammed against the high bluffs and bastions of Mount Huggins, to spill over this sill 

 in the horst, and stream into the low-pressure ai'ea of Ross Sea. This sag is further 

 accentuated by deep inlets, the Barne and Mulock Inlets. 



(2) The steep isobaric as well as isothermal grade towards Ross Sea. 



(3) The general northerly movement of the Antarctic surface air from the 

 Temperature Pole of the inland plateau. 



The next exception to the general rule is that of the plateau winds west of the 

 Royal Society Range. These blow to E. by N., or E.N.E. into Ross Sea. Apparently 

 they are in part responsible for the deflection of the steam column of Mount Erebus, 

 which usually spreads away in an E. by N. to E.N.E. direction. The cold air from 

 this plateau appears to flow down into Ross Sea, following the steepest grade which 

 is also the shortest route to base level. Along the coast, from McMurdo Sound to 

 the Reeves Glacier near Mount Nansen, the blizzard winds have a S. to N. direction, 

 blowing parallel to the steep western fault scarps of the gi-eat meridional horst. 



These blizzard sastrugi are crossed, opposite each of the outlet glaciers, by strong 

 sastrugi trending nearly W. to E. and marking the direction of the plateau wind 

 when it sweeps, often with hurricane force, through the portals of the glaciers of the 

 western mountains down on to Ross Sea. 



The fourth exception, and a very remarkable one, is that of the winds between 

 the coast near the Drygalski Barrier and the parting of the winds on the Magnetic 

 Pole Plateau. 



* Amundsen gives the following percentage of his total number of wind observations: N. 1-9; N.E. 

 7-8; E. 31-9; S.E. 6-9; S. 123; S.W. U-3 ; W. 2C,; X.W. M; Calm, -'1-3, 



