L86 



The most conspicuous feature of the Antarctic wind which is exhibited by the figures in this table is the 

 predominance of east winds throughout the two years over which the observations extended. East-north- 

 east and north-east winds come next in frequency ; but north winds were conspicuous only during the 

 months of November and December, 1902, and January, 1903; south-easterly winds were, on the whole, 

 less common than north-easterly winds; southerly winds were comparatively rarely felt; and west winds 

 were never felt at all. 



Generalizing the observations still further by grouping them under eight points only and omitting the 

 calms, as in Table II., we get for the entire series 44 per cent, of easterly winds, 20 per cent, of north- 



TABLE II. Distribution of Observations of Wind Direction under Eight Points. 



easterly, and 16 per cent, of south-easterly winds; so that no less than 80 per cent, of all the winds 

 observed blew from the quadrant between north-east and south-east. Of the remaining 20 per cent., 

 11 per cent, fall under north and 6 per cent, under south, leaving only 3 per cent, to be divided between 

 all the westerly points. 



An examination of the values for the individual months fails to show any clear indication of a monthly 

 change in the distribution of wind frequency. The observations extended over two complete years, so 

 that each month is twice represented ; but a variation of frequency shown by the observations of one 

 month is not, as a rule, supported by the observations made in the corresponding month of the other year. 

 For example, in November and December, 1902, and January, 1903, one-fifth of all the winds observed 

 were from north, but in the corresponding months of 1903-4 there was no such preponderance of winds 

 from that point, December having but 8 per cent, and January only 2 per cent, of winds from north. 



The observations were then combined, in order to see whether they indicated any distinctive seasonal 

 distribution, and for this purpose they were arranged in four groups, as in Table III. 



The first of these groups embraces the four consecutive months May -August during which the sun was 

 continuously below the horizon ; the second group the four months November-February, during which the 

 sun was continuously above the horizon ; the third and fourth groups comprise the autumnal and vernal 

 equinoxes respectively, two months, March-April and September-October, in each group. 



The table gives the mean percentage of frequency under each of sixteen points for the several seasons 

 first for each of the two years separately, and then for the two years combined. 



In the winter and spring a somewhat marked frequency of north-east winds is shown in both years, and 

 in the winter there was also a greater number of winds from the west of south than is shown by the 

 figures for the other seasons. 



In summer the average of northerly winds for the two years combined is rather large owing to the 

 great frequency of those winds in the summer months of 1902-3. This frequency was not repeated in 

 the following year, and whether the distribution observed in the first or that observed in the second 

 summer most nearly indicates the normal for the season must remain a matter of doubt. 



The distribution of winds in the spring months appears to approximate somewhat to that of the winter 

 in the greater frequency of north-east winds and also of winds from points to the west of south ; whilst in 



