4 Of) 



of the drift of the smoke from Mount Erebus fire of special value, because, owing to the distance from the 

 mountain of the observer at Winter Quarters and also to his position relative to the line followed by the 

 smoke, it was sometimes difficult for him to determine with precision the direction of the latter. Under 

 these circumstances an independent set of observations, made sometimes at nearly the same instant, but 

 always from a different point of view, becomes very useful. The two sets of observations were, therefore, 

 in the first place dealt with separately, and then combined, and the results are given in Table X. 



TABLE X. Relative Distribution under Eight Points of the Direction of Movement of the Air Currents 

 at High Levels, as obtained from Observations of () Lower Clouds, (b) Smoke from Mount Erebus, 

 and (c) Upper Clouds. 



The same table also gives information respecting the movement of the upper air currents as shown by 

 the directions in which clouds entered as "upper" and "lower" clouds respectively were observed to 

 travel. It is, perhaps, fairly safe to assume that the clouds entered as lower clouds were correctly classified 

 as such, but it may not be an equally safe assumption that all those entered as " upper " clouds really 

 belonged to the highest cloud levels. Possibly the dividing line between the two classes of cloud may not 

 always have been easy of recognition, and, in consequence, some of the so-called " upper " clouds, although 

 doubtless liirjh clouds, might, perhaps, with more accuracy have been assigned to the lower levels. The 

 point is ,of some importance, in so far as it may have influenced the result obtained as regards the 

 prevailing direction of movement, and it ought also to be kept in tnind when attempting to assign an 

 approximate average height for the strata dealt with. As regards the height to which the results given in 

 the table may be considered to apply, we shall probably not be far wrong if we assume it to be from 

 5000 to 7000 feet for the lower clouds, and at least 20,000 feet for the upper clouds, in which case the 

 drift of the Erebus smoke would represent the motion of the air at a level approximately midway between 

 the two. 



The figures in the table are very striking. It will be remembered that throughout the whole time the 

 ship lay in Discovery Bay the dominant winds blew from the eastward, and that whilst 80 per cent, of all 

 the winds observed blew from between north-east and south-east, the number of winds from west did not 

 amount to even 1 per cent But at the level of the lower clouds this distribution becomes materially 

 changed and gives place to one which is far more equable, but which, nevertheless, shows a decided 

 tendency for the air current to shift more to the southward and westward, nearly one half of all the 

 observations being included between south-east and south-west. Of the individual points, north and 

 south are those which have the largest percentages, each having 18 per cent, of the total number of 

 observations, whilst the percentages under north-east, east, south-east and south-west vary from 10 per 

 cent, to 15 per cent. West, which is still the least frequently observed direction, has a percentage of 4, 

 and north-west 8 per cent., so that, even at this comparatively low level, a westerly current begins to 

 show itself. 



