40 



ture to add, before the astronomical observa- 

 tions I made at Cumana, might have become 

 dangerous to navigators, were not the sea uni- 

 formly calm in those regions. The errors in la- 

 titude were still greater than those in longitude, 

 since the coasts of New Andalusia stretch to the 

 westward of Cape Three Points fifteen or twenty 

 miles more to the north, than appears in the 

 charts published before the year 1800. 



Toward eleven in the morning, we perceived 

 a very low islet, covered with a few sandy downs ; 

 and on which we discovered with our glasses no 

 trace of habitation or culture. Cylindrical cac- 

 tuses rose here and there in the form of cande- 

 labra. The soil, almost destitute of vegetation, 

 seemed to have a waving motion, in consequence 

 of the extraordinary refraction, which the rays 

 of the sun undergo in traversing the strata of 

 air in contact with plains strongly heated. Un- 

 der every zone, deserts and sandy shores appear 

 like an agitated sea, from the effect of looming. 



The appearance of so flat a country scarcely 

 corresponded with the ideas we had formed of 

 the island of Margaretta. While we were busy 

 in laying down our bearings on the charts, and 



publish charts of America ; the new determinations deserv- 

 ing so much the more confidence, as the positions have been 

 verified by very different astronomical methods, and by ob- 

 servers who did not communicate their results to each other, 

 till long after they had terminated their labours, 



