18 



seco and St. Anne*. It would have been per- 

 haps as prudent to have continued our course. 

 The old charts are filled with rocks, some of 

 which really exist, but the greater part are ow- 

 ing to those optical illusions, which are more 

 frequent at sea than in inland countries. The 

 position ofthe real dangers is generally indicated 

 by chance; they have been seen by pilots who 

 were several degrees out in their longitude, and 

 we might be certain of meeting neither with 

 rocks nor breakers, if we directed our course to- 

 ward the points where they are laid down on the 

 maps. As we approached this pretended Maal- 

 stroom, we observed no other motion in the 



* JefFery's and Van-Keulen's charts indicate four islands, 

 which are only imaginary dangers : the islands Garca and 

 St. Anne, to the west of the Azores $ the Green island (la- 

 titude 44° 52', longitude 28° 30') and the isle of Fonseco (la- 

 titude 13° 15', longitude 57° 10'). How is it possible to be- 

 lieve in the existence of four islands in latitudes crossed by 

 thousands of vessels, when of so many small rocks and shoals* 

 announced by credulous pilots for a century past, there are 

 scarcely above two or three that are real ? As to the general 

 question, what is the degree of probability we may admit, that 

 an islet visible at the distance of a league may be discovered 

 between America and Europe, we might submit this to a ri- 

 gorous calculation, if we knew the number of ships that have 

 annually crossed the Atlantic these three centuries past, and 

 if regard be had to the unequal distribution of those vessels 

 in different latitudes. If the Maal-stroom were, as Van 

 Keulen admits, in 16° of latitude and 39° 30' of longitude , 

 we should have crossed it the 4th of June. 



