TEMPERATURE AND WIND. 47 



.seldom to show the eii'eot so clearly, and have therefore not been includetl in the table. 

 In column VIII the differences in temperature between northerly and southerly winds of 

 velocity between ] I and 30 miles an hour aie shown. With the exception of June the 

 southerly winds were the colder by two to four degrees. The June observations were taken in 

 1911 and 1912; in the latter year the southerly winds obeyed the rule and were 4-6°F. 

 colder than the northerly winds ; it is therefore probable that the June discrepancy would 

 disappear if more observations were available. It is to be noted that although in the wdnter 

 the Barrier is probably on the average nearly 30°F. colder than the Ross .Sea, the southerly 

 winds which blow from the Barrier are not quite 5°F. colder than the northerly winds which 

 blow from the Ross Sea. This is entirely due to the fact that the temperature of the winds 

 is governed by the temperatm-e of the upper atmosphere, which has bean sliown to be only 

 slightly colder over the Barrier than over the Ross Sea during the winter (see page 44). 

 This result confirms, therefore, the assumptions used in determining the probable temperature 

 of the upper atmosphere. 



The layer of cold air near the ground, on the removal of which depends the sudden 

 rise of temperature when a wind commences, cannot come into existence if suiBcient solar 

 radiation is received to keep the temperature of the ground above that of the air in contact 

 with it. This appears to take place in October and from then on to the corresponding 

 month in the autumn, Febniary, the solar radiation is sufficient to cause convexion currents 

 which maintain a sensible temperature gradient, and prevent the formation of a cold ground 

 layer. In the absence of solar radiation in March, April, and May the same effect is pro- 

 duced by the relatively warm water of McMurdo Sound, for it is not until the end of May 

 that the ice is sufficiently thick to prevent an appreciable warming of the air by the under- 

 lying water. From October to May, then, the cold layer cannot form over McMurdo Sound 

 and the figures in columns VI and VII for these months show only a small difl"erence of 

 temperature between calm weather and windy weather. It will be noticed, however, that the 

 northerly winds are two or three degrees warmer than calms, while the southerly winds are 

 colder than calms except in the extreme months of the period. 



This is exactly what would be expected from geographical considerations when there is 

 no cold surface layer to be removed ; for southerly winds are cold and northerly winds are 

 warm. It will also be noticed that with the exception of December the difference in tem- 

 perature between noi-therly and southerly winds (column VIII) is greater in the summer 

 months, November to February, than in the winter. May to August, although the actual tempera- 

 tures of the Ross Sea and Barrier have the reverse relationship. This again confirins our 

 conclusion tliat the temperature of the upper air determines the temperature of the winds, 

 for we have already shown that the difference in temperature in the upper air over the 

 Ross Sea and Barrier is greater in summer than in winter. The small difference of only 

 1'6°F. in December is a direct consequence of the uniformity of temperature found over 

 the whole Ross Sea area in this month. 



Examples of the Effect of the Wind on Temjierature. — In the foregoing wo have discussed 

 the influence of the wind on the temperature by talcing the observations in the bulk, and 

 have found that the average conditions are explained by taking into consideration the vertical 

 as well as the horizontal distribution of temperature. We will now examine a few typical 

 examples which will show the process in action in individual cases. For this purpose figures 

 16 to 19 have been prepared from the records of the thermograph and the Dines' anemo- 

 meter. The thin continuous curve shows the temperature according to the scale on the left 

 of the diagram, and the lower ragged curve the wind velocity according to the scale on the 

 right of the diagram. The wind direction is entered along the bottom of the diagrams. 



