JIOTION OF THE UPPER ATMOSPHERE. 



139 



Thus at the ground levul the traiisfercuco of air in almost entirely from the south, at 

 the height of Erebus smoke the motion is from the west and at th.o greatest height observed 

 the motion is from the north. 



Motion of the Upper Atmosphere at Cape Adare. — The above conchisions are borne out 

 by the observations made f)f cloud motion at Cape Adare, the results of which have been 

 included in table 72, page 133. In figure 18 three diagrauLS are shown, for the winds, the 



II 10°/, 20% 30% 40%, 50% 



Fig. -IS. Motion of atmosphere. Percentage frequency at Cape Adare. 



low clouds and the high clouds. At the level of the low clouds the chief motion is from 

 the S.E. and the mean motion is 8. 51° E. showing an excess of motion from the south 

 over the north. The clouds described as high at Cape Adare were in all probability lower 

 than the high clouds observed by means of the camera obscura at Cape Evans, and therefore 

 their mean height was probably between the high clouds and Erebus smoke as observed at 

 Cape Evans. It will be noticed that the diagram is intermediate between those for Erebus 

 smoke and high clouds at Cape Evans, and the mean direction N. 23° W. is also inter- 

 mediate. Thus at Cape Adare also the transference of air at the ground level is from south 

 to north and in the upper atmosphere from north to south. 



Rui Point Observations. — The observations of upper air motion made^ at Hut Point (within 

 15 miles of Cape Evans) by the Di.scovery Expedition need to be reviewed in the light of 

 the new observations. For this purpose the diagram from page 496 of the Meteorological 

 Results of the National Antarctic Expedition 1901-01 has been reproduced here, figure 49. 



With regard to the surface winds we have already explained that the east and north-east 

 winds at Hut Point were really the southerly winds of the Barrier deflected by the foot- 

 hills of Erebus and therefore correspond to the south-east winds at Cape Evans. Comparing 

 the lower clouds with the surface winds we see nothing to correspond with the prevailing 

 direction of the surface winds. Now if the clouds were really lower clouds their directions 

 should not depart largely from the directions of the surface winds for they are largely carried 

 along in the lower current. Further there can be no doubt that the frequency with which 

 the lower clouds moved from the south must have been much greater than the frequency 

 with which they move from the north, for all over the west of the Ross Sea area the 

 lower atmosphere moves mainly from the south, yet on this diagram the frequency is ^jracti- 

 cally the same from the north as from the south. On looking up the actual meteorolo- 

 gical register the reason is obvious. The direction of the motion of the clouds was never 

 entered when the sky was completely overcast with Itnv clouds ! Hence all the cloud motion 

 during blizzards which is exclusively from southerly directions has been neglected, thus the 



