19 



Atlantic anticyclone was found to occupy different positions in the corresponding month or months on the 

 charts of each of these years. He found that if BUCHAN'S charts, published in 1869, are compared with 

 those prepared by the same authority for 1870-84, or if MOHX'S charts published in 1879, 1883 and 1903 

 are compared with those prepared by the Meteorological Office, which are based on observations recorded 

 in nearly 3300 logs extending over the period 1855-99, the mean monthly position of tin nfret 



will be found to vary very considerably both in latitude and longitude. Colonel RAWSOX concluded that 

 in addition to the seasonal migration of the high-pressure belt, which has so long been recognised, there 

 is a progressive displacement going on from year to year, in consequence of which the belt is in some 

 years nearer to the Equator than in others. He believed he had found strong evidence of cyclical changes 

 having taken place in the belt's latitude, and of the existence of an interval of 9| years between the times 

 of its passing from its extreme northern to its extreme southern positions, and vice n-rfi't. If Colonel 

 RAWSOX'S theory be correct, the belt should have reached the southern limit of its oscillation in 1903. 

 Now although the southern margin of the high-pressure belt in the several divisions of the ocean, as 

 shown on the monthly charts for that year, appears to be in about its average position as regards 

 latitude, it must be admitted that the monthly charts for both 1901 and 1902 place the margin of the belt 

 in lower parallels. 



SUMMARY OF THE OBSERVATIONS OF WIND. 



On Tables I. to IV. are set forth the number of observations of wind from each of eight cardinal and 

 inter-cardinal points of the compass, and of variable winds and calms recorded in the Southern Ocean 

 during the thirty -months period, October 1901 to March 1904; inclusive also of the number of instances 

 on which these winds attained to gale force. 



For the purposes of comparison the observations are arranged according to (1) the respective seasons, 

 (2) the divisions of the Ocean, and (3) the zones to which they relate. 



In the classification adopted (p. 16), the months of September, October, November are regarded as 

 representing Spring ; December, January, February as representing Summer ; March, April, May, Autumn ; 

 and June, July, August, Winter. The following are the meridional boundaries selected for representing 

 the three divisions of the Southern Ocean, 150 E. to 70 W., Pacific; 70 W. to 20* E., Atlantic ; 20 E. 

 to 150 E., Indian. The grouping of ocean wind and gale frequency into zones of ten degrees of latitude 

 is adopted for convenience. 



Tables V. to IX. state the directions from which the largest and smallest percentages of \vinds of all 

 forces, and of gales only, were recorded in each of the zones and divisions of the Southern Ocenn referred 

 to, in the respective seasons of the year and in all seasons, during the thirty months period. 



Table X. furnishes similar information relating to the Southern Ocean as a whole. 



D 2 



