in Colombia hci'iXiecn the Years 1820 and 1830. 15 



from the south. This, however, is more perceptibly the case 

 from latitude 8° to 13°, where the air is cooled to an average 

 of 71°-8. (Humboldt De Distrihutione Geog. PL p. 92). Be- 

 twixt 9° N. lat. and 3° S. lat., if we may trust to observations 

 made at the five points of Panama, Esmeraldas, El Morro, 

 the island of Puna, and Guayaquil, the annual mean is 80°' 1 1, 

 being 2°'A''} less than the mean of the Atlantic coast. A no- 

 table difference also arises from the superior elevation of the 

 Pacific chain of the Andes, and its more immediate vicinity to 

 the coast, while the Venezuelan branch, with the exception of 

 the Santa Marta ridge, is both lower and more inland. A 

 curious exception to the general temperature of the Pacific 

 coast may be found on passing Punta Galera and Cabo San 

 Francisco (lat. 50' N.) to the south. The sky is here almost 

 perpetually clouded, and a drizzling rain falls through the 

 greater part of the year. During a week I passed there I 

 never saw the sun ; and the average temperature was only 

 74;°'l't. This was the more striking, as along the coast, im- 

 mediately to the north of Punta Galera, the weather was con- 

 stantly dry and the sky clear. The miry state of the road 

 across the point of the Cape of San Francisco indicates the 

 line of separation betwixt two distinct climates. It will be 

 seen by the map, that from P. Galera the coast, after run- 

 ning nearly due west, turns abruptly to the south. 



2. On penetrating into the interior of the country, and ex- 

 amining the temperature of small elevations, we may take, as 

 forming an aggregate specimen of the whole country : 1 . The 

 damp wooded valleys of the Orinoco and Magdalena; 2. The 

 forests which border on the Pacific; and 3. The immense 

 plains of Venezuela, alternately flooded and parched with ex- 

 cessive heat. Humboldt assigns to the valley ot the Orinoco 

 a mean temperature of 78°"2. The small number of observa- 

 tions I have made on that of the Magdalena v/ouid give a 

 mean of nearly 83°, which I should scarcely think too high, 

 considering the localities of the river, which, flowing from 

 south to north, affords no channel to the sea-breezes. Its 

 mass of water is also much less considerable than that of the 

 Orinoco; while its numerous sinuosities, and the low ridges 

 which border it in the upper part of its course, contribute to 

 render the air stagnant and suflbcating. The temperature of 

 Honda, at 1200 feet of elevation, is as high as that of any part 

 of the coast, except Maracaybo. The unbroken forests which 

 extend from the roots of the Quitcnian Andes to the shores of 

 the Pacific have a much lower temperature, caused by the 

 proximity of the snow-capped Cordillera, and the humidity 

 which prevails throughout the year. Accurate observations 



