MO 



WIND. 



diagram of frequence of motion of lower clouds is useless for giving the mean motion of the 

 lower atmosphere. 



The difference in the diagrams for the Erebus smoke as observed from Hut Point and 

 Cape Evans is large, the chief discrepancies being in the north-west and south-west directions. 



SURFACE 



tV/NDS. 



LOWER 



CLOUDS. 



UPPER 



CLOUDS. 



M^ EREBUS 



SMOKE. 



52 



% 10% 



rt 



?o% 



—I — 



30% 



Fio. 49. Motion of atmosplieie. Percentage fieiiuency, Hut Point 



It is probable that the cause is the difference in the point of view of the mountain 

 from the two stations. It is quite possible that the observations of south-west direction at 

 Cape Evans were ailiectcd for the reason already stated, namel_y, that the mountain itself, 

 being to the north-east, hid much smoke travelling away in that direction, hence the south- 

 west directions in the Cape Evans diagram very likely need to be increased. It is not 

 clear why the north-westerly directions were so few as scon from Hut Point and so many 

 as seen from Cape Evans, but that this was due also to the point of view is supported 

 by the observations made on sledge journeys from Hut Point which gave 10 per cent, of 



