52 



TEMPERATURE. 



Table 19. 

 Difference in Temperature Frnmheim-Cape Evans during Blizzards at Cape Evans. 



Except wth high wnds in December 1911 the temperature was lower at Framheim than 

 at Cape Evans dming blizzards, hence the air in the blizzards at Cape Evans cannot have 

 passed over Framheim as warm northerly winds. 



Because of the relatively high temperature of blizzards during the winter the explanations 

 discussed above have been formulated, but it was forgotten that if the explanations were 

 true for the winter they would be true for the summer also. Thus we should expect high 

 temperatures in summer blizzards as well as in winter blizzards. But we have already shown 

 that this is not the case' for summer blizzards are cold. 



We have now shown that the fohn effect and the circulation of air around low pressure 

 systems do not explain the relationship found between the temperature and wind ; on the 

 other hand an explanation has been given depending on the vertical distribution of tempera- 

 ture which explains all the observed conditions ; it is therefore reasonable to conclude that 

 the latter is the true and sufficient explanation. 



Daily Variation of Temperature. 



In this and the following section use will be made of three terms which have often 

 been used by writers as being almost interchangeable, but to which a special significance 

 will be attached throughout this work, therefore it is desirable that they should be carefully 

 defined : — 



(a) The expression daily variation of temperature %vill be used to express the daily 

 march of the temperature when all temperature changes other than those due 

 to true daily periodic causes have been eliminated as far as possible. The 

 irregular and non-peiiodic changes due to \nnd, cloud, etc., are eliminated by 

 combining a large number of days having a similar periodic temperature variation ; 

 for then the irregular changes more or less completely cancel one another while 

 the true daily variation remains. In addition a correction has to be applied 

 to remove the effect of the yearly variation of temperature. Duiing certain 

 seasons of the year this may be large, for example during September 1911 each 

 midnight was on the average •9°F. warmer than the previous midnight owing 



