DAILY VARIATION OF CLOUD. 



155 



;iu(l light wiuds are iu all luoutlis less tliau during blizzards, but the diti'erence is very much 

 greater from May to July than from October to March. During the former of these periods 

 there is no direct sunlight and radiation is great, hence as soon as the wind drops the clouds 

 disappear and the average amount of cloud during the intervals between the blizzards is only 



10 



miles/hour 



Jan. 



Feb. 



Mar.- 



Apl. May 



Jun. 



Jul. 



Aug. Sep. 



Oct. 



Nov. 



Dec. Jan 



Fig. 51. Annual variation of cloud according to wind, 

 half that during the blizzards. On the other hand, during the period of nearly constant sun- 

 light this force does not act, but the reverse holds, the radiation received by the clouds 

 from the sun tends to increase rather than diminish them. Hence during these months there 

 is only a small decrease in cloud amount in the intervals between the blizzards. 



We now see that the yearly variation of the cloud amount is a complex phenomenon, 

 depending on two variables. 



(a) The frequency of blizzards which if acting alone would give a maximum of cloud 



in the winter and a minimum in the summer. 



(b) The amount of radiation received and emitted by the cloud particles which if it 



acted alone would give a maximum of cloud in the summer and a minimum 

 in the winter. 

 The combination of these two efEects is that minima of cloud occur in summer and winter 

 and maxima iu the months near to the equinoxes. 



Daily Variation of Cloud. 

 Observations of the amount of cloud were made at Cape Evans every four hours at 

 midnight, 4 a.m., 8 a.m., midday, 4 p.m. and 8 p.m. As however station time was used, these 

 observations must be put one hour earlier in local time. The record is not quite complete, 



