258 



ing lately received excellent barometers, con- 

 structed by M. Fortin, may direct the geodesic 

 levellings, which are always slow and expensive, 

 to be preceded by barometric levellings, which 

 in the torrid zone are extremely exact. I am 

 assured that in those countries correspondent 

 observations may be dispensed with, on account 

 of the marvellous regularity of the horary varU 

 ations, without fearing errors of 4 or 5 toises. 



The points which ought to be carefully ex- 

 mined are the following :—the Isthmus of Hua- 

 sacualco, between the sources of the Rio Chima- 

 lapa and the Rio del Passo ; the Isthmus of Nica- 

 ragua *, between the lake of that name, and the 



This cape not having been carried to the north in the same 

 measure as the bottom of the gulf, near the mouth of the Rio 

 Mandinga, it thence results, that, according to the first map* 

 the gulf enters 24', and according to the second, 7'. It is 

 probable that the changes of latitude which result from the 

 last expedition of M. Fidalgo, must be attributed to the want 

 of artificial horizons, and to the difficulty of observing the sun 

 with instruments of reflexion, amidst a group of islands, and 

 above a sea where the horizon is not clear. More to the west 

 the mean breadth of the isthmus, between Castillo de Cha- 

 gres, Panama, and Portobello, is 14 marine leagues ; the 

 minimum of its breadth (8 leagues) is two or three times less 

 than the breadth of the isthmus of Suez, which M. Le Pere 

 finds to be 59,000 toises. 



* If the question here agitated related only to canals of 

 small navigation, fit solely to enliven inland trade, I should 

 also have named the coast of Verapaz and Honduras. The 



