254 The Winds. 



winds are a study, and suggestive of marked peculiarities 

 in the climatology of equatorial Africa." 



The trade winds are the permanent, monsoons are the 

 annual, and land and sea breezes are the diurnal features of 

 the general system of the winds. The last are caused by the 

 daily depression of the barometer, which is most conspicuous 

 at the equator. Humboldt, when at Cumana in 1799, affirmed 

 that the hour of the day might be known correctly within 

 fifteen minutes, by the rise and fall of the quicksilver in his 

 barometer, which amounted to nearly the tenth part of an 

 inch daily. In other latitudes, the oscillation, although easily 

 traced, is smaller, and concealed by greater irregular move- 

 ments of the mercurial column. The following table contains 

 the average value of this oscillation, from bi-hourly observations 

 at Greenwich for five years, 1841-45. 



Table VII. — Mean Diurnal oscillation of the barometer at 

 Greenwich, above and below mean height of the barometer, 

 for five years (1841-45) : — 





In. 



In. 



In. 



1 h. a.m. - 



- 0-002 



9 h. a.m. + 0-009 



5 h. p.m. — 0-008 



3h. „ - 



-0-010 



11 h. „ + 0011 



7h. „ —0-000 



5h. „ - 



- 0-012 



lh. p.m. + 0-001 



9h. „ + 0-009 



7h. „ - 



- 0-003 



3h. „ —0-007 



11 h. „ +0-009 



The course of the numbers shows that the barometer stands 

 higher by day than at night, owing to the generation of aqueous 

 vapour. Its lowest minimum is reached about the coldest time 

 in the morning, when the quantity of aqueous vapour is least. 

 Another minimum occurs about the hottest hour of the day, 

 when the heat of the sun causes a portion of the air to flow off 

 from the top of the atmosphere. This minimum is less con- 

 siderable than the great minimum observed at night. 



The existence and character of this oscillation were known 

 from the earliest use of the barometer, and are everywhere the 

 same, wherever the alternation of day and night occurs. The 

 extent and nature of the annual oscillation, on the other hand, 

 depends upon local circumstances, particularly on the distri- 

 bution of land and sea. For other and larger oscillations 

 which return at stated periods no cause whatever can be 

 assigned. It is observed, for example, that about the middle 

 of November a wave of high barometric pressure passes over 

 England, and another about the 25th of December. In this 

 interval the barometer reaches a very low point about the 28th 

 of November. To trace the motion of these atmospheric 

 waves, the ocean has been divided into squares, each number- 

 ing ten degrees of latitude and longitude. A permanent de- 

 pression of the barometer to the extent of fully an inch was 



