WIND. 121 



that the gustiness rose rapidly from May to September, remained high in October and then 

 decreased rapidly and remained nearly constant during November, December, January, and 

 February. This variation from month to month is just as marked with winds from the north 

 as with winds from the south. Also if we fix our attention on any given wind velocity 

 the variation fi-om month to month is exactly the same, thus the variation is not connected 

 with the increased or decreased violence of the wind. It is pmely a seasonal change which 

 affects equally all winds. 



Cause of gustiness. — It is known that the gustiness of the wind depends largely on the 

 character of the country over which the wind blows. For instance the gustiness is much 

 greater in a town than over flat marshy country,* the reason being that obstacles to the 

 wind's motion set up eddies which show themselves by increasing and decreasing the wind's 

 steady motion. 



At first one might be inclined to say that this is the cause of the high gustiness of the 

 wind in McMurdo Sound, for there the wind is sweeping round the great obstacle of Mount 

 Erebus at the foot of which the anemometer was installed. But this explanation fails entirely 

 when we try to apply it to the variations in the gustiness which our investigation has 

 revealed. 



If the gustiness is due to the violence with which the wind strives to pass obstacles 

 why should it decrease as the \vind velocity increases ? Also the obstacles were exactly the 

 same during August, September, and October as during November, December, and January, 

 yet during the former months the gustiness was 33 per cent, greater than during the latter 

 months. 



The following appears to be the most satisfactory explanation. 



We have frequently referred to the layer of abnormally cold air which forma near the 

 ground during the cold months. In the next section we shall see how this cold layer is 

 able to underrun a wind and raise the moving air above the ground so that at ground level 

 the air is calm while above the wind continues to blow. 



The efiect of such a cold layer of air is easily seen. \Vhen the pressure gradient tends 

 to set the whole mass of air into motion, the lower layer of cold air is in contact with the 

 ground and does not readily move while the warm air above slides over the cold layer with 

 little or no friction. Thus the upper warm air is soon moving more rapidly than the lower 

 cold air. 



It is the interaction of these two layers of air which causes the unsteadiness in the 

 resulting wind. One can picture a gusty wind as being analogous to a dirty stream of water 

 entering a clear but stagnant lake. An observer situated at the bottom of the lake would 

 be surrounded at one moment by dirty water and the next by clear, and he would find 

 that the masses of dii-ty water were generally moving more rapidly than the clear water. 

 The dirty water corresponds with the warm upper current, which drags along with it masses 

 of the cold stagnant air like the dirty stream drags along the clear water of the lake. 



In the way that the observer on the bottom of the lake is able to distinguish the origin 

 of the water which pa.sses him by its colour, so we are able to distinguish the successive 

 masses of air by their temperature. The thermographs at Cape Evans were not particularly 

 sensitive nor rapid in action, but their records give conclusive evidence of rapid changes in 

 temperature during winds. Pai-ticularly during blizzards the thermograph trace became very 

 thick, and frequently the trace was thickened over more than five degrees showing that the 

 pen was constantly oscillating by this amoimt, which must have been much less than the 



* Shaw : Reports and Memoranda No. U of the Advisory Committee f c r Aeronautics, page 4. 



16 



