4 



these limits pass the tropics eight or nine de- 

 grees ; while in the vicinity of Africa the vari- 

 able winds reign far beyond the parallel of 28 or 

 27 degrees. It is to be regretted, on account 

 of the progress of meteorology and navigation, 

 that the changes, which the currents of the equi- 

 noctial atmosphere in the Pacific Ocean under- 

 go, are much less known, than the variation of 

 these same currents in a sea that is narrower, 

 and influenced by the proximity of the coasts of 

 Guinea and Brazil. Navigators have known for 

 ages past, that in the Atlantic Ocean the equator 

 does not coincide with the line which separates 

 the trade winds of the north-east from the ge- 

 neral winds of the south-east. This line, as 

 Halley* has very well observed, is at the third 

 or fourth degree of north latitude; and if it's 

 position be the effect of a longer abode of the Sun 

 in the northern hemisphere, it tends to prove, 

 that the temperatures of the two hemispheres*^ 

 are in the ratio of eleven to nine. We shall see 

 farther on in this work, when we treat of the 

 part of the atmosphere which extends over the 



* Phil. Trans, vol. xvi, p. 154. Ulloa, Conversaciones, p. 



ioa. 



f Prevost, on the limits of the trade winds. Journ. de 

 Phys. t. xxxviii, p. 369. Supposing with jEpinus, that the 

 southern hemisphere is only one fourteenth colder than the 

 northern, the calculation gives, for the northern limit of the 

 E. S. E. trade winds, the parallel of 1° 28'. 



