AND THE PREVAILING WINDS OVER THE GLOBE. 581 



unknown, to attempt to determine the velocity of any general movement of 

 the atmosphere from the observed velocity of the wind at Observatories could 

 lead to no satisfactory result as regards the present inquiry. In the broad 

 results aimed at in this comparison of atmospheric pressure and prevailing 

 winds, it cannot affect the conclusions arrived at to assume, as is here done, 

 that the mean velocities of winds from different directions are equal to each 

 other. 



From the figures given in Table II. the mean direction of the wind has been 

 calculated in the usual way. This direction is represented in the charts by 

 arrows flying in the direction of the wind. In cases where the winds do not pre- 

 ponderate from one quarter, but are nearly equally distributed over different 

 points of the compass, an asterisk (*) is entered on the charts, which thus repre- 

 sents variable winds ; when calms preponderate, a circle with a dot in the centre 

 (0) is used. When two maximum directions are strongly indicated, or when the 

 smaller one is very decidedly marked, instead of resolving the two into one 

 intermediate, which would in many cases represent a wind which scarcely ever 

 occurs at the time, the greater of the two is represented by the ordinary arrow, 

 and the smaller in the more marked cases by a less arrow placed beside it. By 

 this means an important feature in climate is represented. 



Thus two distinct sets of facts are exhibited on the charts, viz., lines showing 

 the mean pressure of the atmosphere, and arrows showing the prevailing winds at 

 the earth's surface, each being independently arrived at by the summing and averag- 

 ing of observed facts. What relation is there between these two classes of facts? 



I. Winds within, or near, a space of Lorn Pressure. — Of this class, the best 

 example is the low pressure which prevails in the north of the Atlantic and adjoining 

 regions in the winter months. This region of low pressure is bounded to the 

 S.W. by the high pressure of North America; to the S. by the high pressure in 

 the Atlantic, about 30° lat. N. ; to the S.E. by the high pressure in the interior of 

 Asia. In January, the difference between the average pressure of Iceland and the 

 interior of Asia is fully an inch. 



It is seen from the charts, that in Baffin's Bay and east of the Rocky Moun- 

 tains, as far south as 40° lat., the winds are N.N.W., N.W., and W.N.W. Cross- 

 ing the Atlantic, winds in the British Islands, in France, and the north of 

 Germany, are from W.S.W. to S.W. ; in Denmark, S.S.W. ; near Bergen, in 

 Norway, S. ; and at Christiansund and Hammerfest, S.S.E. The relation of 

 these winds to the isobaric lines is the same as that which is illustrated by the 

 winds in storms, in their relation to the isobaric lines of these storms. This has 

 been already stated in a Paper by the author, published in the " Transactions of 

 the Society," vol. xxiv. Part i. p. 201, in the following words : — " The wind in 

 storms neither blows round the centre of least pressure in circles, or as tangents to 

 the concentric isobaric curves, nor does it blow directlv towards that centre. It 



