no 
MR. H. X. DICKSON ON THE CIRCULATION OF THE 
Bureau mdteorologique de France.’ Even this is a troublesome matter, as it is not 
easy to get all the corrections necessary for rendering the observations fully compar¬ 
able, and many of the data are only published after long delays. 
The simplest method of obtaining an approximate view of the atmospheric con¬ 
ditions prevalent during 1896 and 1897 seemed to be to use the differences of the 
monthly means from the long-period averages at a number of stations, thus avoiding 
all the troublesome reductions to sea-level, and to plot the differences on charts. 
The data for the two years were partly obtained from the publications of the 
meteorological services concerned, l)ut through the kindness of the Director I was 
furnished with the as yet unpublished means for a large number of the stations of 
the Danske Meteorologisk Institut. The long-period means used were those in the 
“ Challenger ” Report on Atmospheric Cii'culation, and the differences are given in 
Table III. 
The anomalies shown by the charts are not, of course, to be regarded as having 
sufficient local accuracy to be worth detailed quantitative discussion ; even if they 
were it would not be possible to deal with them rigorously, for the relation between 
a drift current and the wind which causes it is still quite uncertain. It is necessary 
to look merely at the lu'oader outlines, and to seek for differences which occur con¬ 
sistently over considerable areas and continue for successive mouths. 
The most marked departure from the average distribution of pressure in 1896 istbe 
excess in the lower latitudes during the first half of the year. With the single excep¬ 
tion of the month of June there is continued high pressure from January to August, 
and the excess is greatest to the south and south-west of the British Isles. This 
indicates an unusual extension of the Atlantic anticyclone north-eastwards, and 
consequently stronger and steadier westerly and south-westerly winds, which would 
produce an unusually large easterly surface drift south of the 50th parallel, and 
excessive banking-up of water south-west of Great Britain. Hence we should 
expect all the branches of the northerly current to exceed their usual strength, and 
later in the year to find large supplies of Polar water making their appearance, the 
residt of excessive melting of the Polar ice : this is precisely what the observations 
have shown. 
In January 1896 the area of high barometric pressure extended over Iceland and 
Southern Greenland, the least excess being in the south-western area round Newfound¬ 
land ; but during Febi'uary, March, and April pressure was l^elow the average in the 
north, and the deficiency increased eastwards to form a belt of specially low pressure 
lying along lat. 60° to the Norwegian coast in March. This would lead to abnormal 
easterly drifts from tbe Labrador-Ginenland region during spring, which would 
ultimately join the nortlierly currents on the eastern side. The waters of the 
northerly streams are therefore chiefly derived in winter from the Gulf Stream 
area, while in spring there is an increasing admixture of water from the Labrador 
currents. 
