158 



humboldt's cuba. 



Climatology advances slowly, because we gather 

 by chance the results obtained at points of the globe 

 where the civilization of man is just beginning its 

 development. These points form small groups, sepa- 

 rated from each other by immense spaces of lands 

 unknown to the meteorologist. In order to attain a 

 knowledge of the laws of nature regulating the dis- 

 tribution of heat in the world 5 we must give to 

 observation a direction in conformity with the needs 

 of a nascent science, and ascertain its most impor- 

 tant numerical data. New Santander, upon the 

 eastern coast of the Gulf of Mexico, probably has a 

 mean temperature lower than that of the Island of 

 Cuba, for the atmosphere there must participate, 

 during the cold of winter, in the effects of the great 

 continent extending towards the northwest. 



On the other hand, if we leave the system of cli- 

 mates of Western America, if we pass the lake, or, 

 more strictly speaking, the submerged valley of the 

 Atlantic, and fix our attention upon the coasts of 

 Africa, we find that in the cis-Atla/ntic system of 

 climates upon the western borders of the old conti- 

 nent, the isothermal lines are again raised, being 

 convex towards the pole. The tropic of Cancer 

 passes between Cape Bojador and Cape Blanco, 

 near the river Ouro, upon the inhospitable confines 

 of the desert of Sahara, and the mean temperature 



