200 THE PASS OF DON CRISTOVAL. 



sandy plain, to the trunks of trees, and to the rocks that 

 projected into the sea like promontories. When the sun 

 appeared these inert masses seemed suspended in air; 

 and on the neighbouring beach, the sands presented the 

 appearance of a sheet of water gently agitated by the 

 winds. A train of clouds sufficed to seat the trunks of 

 trees and the suspended rocks again on the soil ; to render 

 the undulating surface of the plains motionless ; and to 

 dissipate the charm which the Arabian, Persian, and 

 Hindoo poets have celebrated as 



" The sweet illusions of the lonely desert.'' 



They doubled Cape Matahambre very slowly. Hum- 

 boldt determined, "as they sailed, as they sailed," the 

 positions of Cayo de Don Cristoval, Cayo Flamenco, 

 Cayo de Diego Perez, and Cayo de Piedras. He also 

 employed himself in examining the influence which the 

 changes at the bottom of the sea produce on its tempera- 

 ture at the surface. 



Notwithstanding the small size of their bark, and the 

 boasted skill of their pilot, they often ran aground. The 

 bottom being soft, there was no danger ; but, nevertheless, 

 at sunset, near the pass of Don Cristoval, they preferred 

 to lie at anchor. The first part of the night was beauti- 

 fully serene : they saw an incalculable number of falling- 

 Btars, all following one direction, opposite to that from 

 whence the wind blew in the low regions of the atmo- 

 sphere. The most absolute solitude prevailed in this 

 spot, which, in the time of Columbus, was inhabited and 

 frequented by great numbers of fishermen. The inhabit- 

 ants of Cuba then employed a small 6sh to take the 

 great sea-turtles. The " fisher-fish," formerly employed 



