PRECIPITATIOxN 27 



direction brought a desirable increase of temperature, and ou some days they had a 

 fair imitation of the mild southerly blizzards which were such a conspicuous feature 

 at the ship {i.e. at Hut Point, S. end of Ross Island. — Authors). 



These observations suggest two possible sources of snow supply for the plateau 

 W. of the Fen-ar Glacier : 



(1) Moisture-laden air which has worked inland at a high level off the Ross Sea. 



The chilling of the air on the high plateau from below upwards makes the layers 

 of air next the plateau surface denser than before, but in so doing necessarily with- 

 draws material from the air above, and, consequently, sooner or later a height will be 

 arrived at in the atmosphere overlying the plateau where the pressure is lower than 

 it is for air at a corresponding altitude over the Ross Sea. Thus, while a surface air 

 current, "the plateau wind," blows down towards the sea and sweeps over Erebus 

 in a current about 9000 to 10,000 feet deep between altitudes of 6000 feet and 

 15,000 feet, at a still higher altitude air will flow in to fill up the low pressure at 

 high levels above the plateau. The air to replenish the cold masses which are 

 continually gliding off the plateau can only be derived fi-om the sea, and thus 

 moisture, in other words, snow, is conveyed inland by what may be termed a high- 

 level sea breeze. This movement of the atmosphere, of such importance for con- 

 sidering snow supply in the Antarctic, is illustrated in Plate VI. 



The main source of supply of the moisture to the " high-level " sea breeze in the 

 Antarctic is, of course, the Southern Ocean lying north of the main boundary of the 

 continent. 



The main high-level sea breeze probably flows towards the "Temperature Pole" 

 (near the South Geographic Pole) in a grand high-level cyclone overlying the cold air 

 masses which spiral out fitfully from the " Temperature Pole." But Ross Sea, when 

 open, sets up a secondary high-level anticyclonic area with its " low " situated over 

 the plateau. Air flowing in from over Ross Sea towards the Magnetic Pole Plateau 

 would have the trend of its current governed by the nearest areas of lower pressure 

 in this upper atmosphere, and the lowest pressure at high atmospheric levels lies 

 over the coldest land. On the Magnetic Pole Plateau the land is colder the further 

 one proceeds south towards the " Temperature Pole." Hence in this part of the 

 Antarctic, near the west shore of Ross Sea, where the local high-level secondary 

 cyclone perhaps dominates the main cyclone, one would expect the trend of the 

 landward inflowing high-level air currents to assume a general N.E. to S.W. 

 direction. This was the observed direction of the inflowing high-level air currents 

 near the coast side of the Magnetic Pole Plateau on December 27.* 



* On December 28, 1908, the weather was calm but very thick. At the surface of the plateau, later 

 the same day, it was observed that a breeze from ofl' the Ross Sea covered the top of Mount Nansen with 

 cloud, but further inland this high-level breeze off the sea was deflected by the high-level N.W. air 

 currents blowing from the South Ocean. Snow seemed to be falling heavily this day over the great horst 

 to our east. We were at the time 20 miles west of the western side of the horst, and ."lO miles distant from 



