164 



humboldt's cuba. 



snow ; and frost is known only on the crests of these 

 hills and of the Copper Mountains. We must suppose 

 that other conditions than the rapid fall of tempera- 

 ture in the upper regions of the air are needed for it 

 to hail and snow. 



I have stated elsewhere that it has never been 

 known to hail at Cumana, and very rarely in 

 Havana, happening only once in fifteen years, during 

 violent electrical explosions and S.S.W. winds. At 

 Kingston, Jamaica, the fall of the thermometer at 

 sunrise to 20°.5 (69° F.), is cited as an extraordinary 

 phenomenon. In that island we must ascend the 

 Blue Mountains to the height of 7,325 feet, to see 

 it fall to 8°.3 (47° F.) in the month of August. At 

 Cumana, 10° U. lat., I have not known the ther- 

 mometer to to fall to 20°.8 (69°.4F.). 



Changes of temperature occur very suddenly in 

 Havana. In April, 1804, the variation of the ther- 

 mometer in the shade, within the space of three 

 hours, was from 32°.2 to 23°.4C. (89° to 74°.1F.), 

 that is, 9° C. (16°.2 F.) 3 which is very considerable 

 in the torrid zone, and twice as great as the varia- 

 tion found on the coast of Colombia, further south. 

 The inhabitants of Havana (K lat. 23° 8') complain 

 of cold when the temperature falls rapidly to 21° C. 

 (69°.8 F.), and in Cumana (N. lat. 10° 28'), when it 

 falls to 23° C. (73°.4 F.) In April, 1804, water sub- 



