704 • LECTURE LVI. 



ciency of rotatory motion, a south east wind, as usual, which is found to 

 prevail between Madagascar and New Holland, as far as the equator. In 

 consequence perhaps of friction in its passage, it gradually loses its impetus 

 towards the west, and at the equator is nearly a south wind : but in proceed- 

 ing north from the equator, it becomes, from an excess of rotatory motion, a 

 south west wind, which blows into the Arabian gulf, and the bay of Bengal. 

 Both these winds are however variously modified by the particular situations 

 of the islands and continents. From October to March, on the contrary, 

 the sun having south declination, the south east trade wind stops at 10° south 

 latitude; the trade winds on the north side of the equator are as usual north 

 east; and beyond the equator they become for some degrees north west, the 

 circumstances being the reverse of those which happen in the summer months, 

 at greater distances, on the o»her side of the equator. (Plate XLII. XLIII. ) 



The last fact is the simplest of all. The land and sea breezes are produced 

 by the ascent of the air over the land in the day time, while the land is 

 hotter than the sea; and jits descent at night, when the land is become 

 colder: hence the breeze comes from the sea by day, and from the land by 

 night. 



The violent agitations of the air, which constitute hurricanes and whirl- 

 winds, occur more commonly in tropical climates than in others. The 

 causes of these storms are little understood: their course is said to be 

 generally opposite to that of the trade winds ; but tornados, which are less 

 regular hurricanes, originate indift'erently from every quarter. 



The variations of the weight of the air, which occasion the winds, and 

 other changes in its density, which are the effects of the winds themselves, 

 are indicated by the height of the barometer, which is in general the more 

 variable as the winds are more liable to sudden changes. Hence in the 

 neighbourhood of the equator the height of the barometer is scarcely ever a 

 quarter of an inch more or less than 30 inches, which is very nearly its mean 

 height on the level of the sea in every part of the globe: in Great Britain 

 it is sometimes as low as 28 inches, but never higher than 31. We have 

 already seen that the elevation of any place above the lea reduces the height 

 ^f the barometer according to a law which is determined by the general 



