﻿and Fluctuations. 349 



the equator. High and low pressure are therefore terms to be 

 applied relatively rather than absolutely ; and the Arctic, with a 

 mean pressure of 29'6, is in reality an area of high pressure ; 

 the Doldrums, with a mean pressure of 29*9, an area of low. 



Owing to the distribution of land and water and the great 

 meteorological disturbances introduced by it, the isobars, in 

 point of fact, differ very widely from such a theoretical view 

 as would make them coincide with parallels of latitude; but 

 where one element decidedly preponderates, and where there 

 are no geographical obstructions, as in the great Southern 

 Ocean, they approach very markedly to this parallelism. 



A ruling temperature, whether high or low, does not appear 

 to exercise any noticeable influence on the pressure ; but a tem- 

 perature that is changing produces a tendency which is often 

 shown rather by its effect on atmospheric movement than by 

 the sluggish barometer ; and this tendency is according to the 

 ordinary laws of the expansion and contraction of gases, and 

 quite in opposition to that which has been proposed by meteoro- 

 logists : an increase of temperature tends to cause a high baro- 

 meter and to give rise to an outward movement ; a decrease of 

 temperature, on the contrary, tends to make the barometer fall 

 and the air move inwards. 



The permanent presence of aqueous vapour in the air has no 

 sensible effect on the barometer ; but in confined districts where 

 aqueous vapour is forming freely the barometer ranges high, 

 and where there is much rain (that is, condensation) it ranges 

 low. 



Where the prevailing winds blow strongly into a corner barred 

 by mountains, as in Siberia, or bifurcate against a line of coast, 

 as against the coast of the Spanish peninsula or California, the 

 barometer stands high in consequence of the mechanical com- 

 pression of the air; and where the prevailing winds circle round 

 any area, or sweep past the opening of a confined district, the 

 barometer stands low in consequence of the abstraction of the 

 air by friction. 



According to these conclusions, the mean barometric pressure 

 at any place, or for any season, depends principally on geogra- 

 phical conditions, on latitude, on the relation between land and 

 sea, on the trend of the coast-line, on the direction and height 

 of mountain-ranges, and on the nature and characteristics of the 

 prevailing winds. Since, however, these geographical condi- 

 tions remain constant, or vary only with the season, they do not 

 appear to account for the deviations from the mean pressure 

 which the barometer is every day showing. But the conclu- 

 sions which I have referred to these conditions are founded, not 

 only on a vast mass of geographical evidence, but on the known 



