223 



agreement proves Ibe justness of a result, which 

 however I could only deduce from altitudes 

 considerably distant from the meridian. A good 

 observation of the stars at Guapasoso* gave me 

 4° 2' for San Fernando de Atabapo. (Gumilla 

 placed the confluence of the Atabapo and the 

 Guaviare in 0° 30' ; D'Anville, in 2° 51'.) I was 

 able to fix the longitude with much more precision 

 in my way to the Rio Negro, and in returning 

 from that river. It is 70° 30' 46" (or 4° 0' west 

 of the meridian of Cumana). The going of the 

 chronometer was so regular d u ring the naviga- 

 tion in a boat, that from the 16th of April to the 

 9th of July it varied only from 27*9", to 28\5". 

 At San Fernando de Atabapo I found the dip of 

 the magnetic needle, rectified with great care, 

 to be 29 # 7° cent. div. ; the intensity of force 219. 

 The angle and the oscillations therefore had 

 diminished considerably from Maypures, in a 

 difference of latitude of 1° IT. The surround- 



copies of Solano's observations which have been circulated ? 

 Gili mentions an instance in the latitude of Atures, which led 

 him into error in all the more southern points. (Saggio, vol. 1 9 

 p. 320 ; and above, p. 12, note.) 



* Obs. Astr. vol. 1, p. 263. The longitude of San Fer* 

 nando has been given in Arrowsmith*s map as I published it 

 (68° 10' from Greenw.), but the latitude is laid down at 

 4° 19'. In this point, as in so many others, the calculations 

 of d'Anville on the longitude have been happier than those of 

 his successors. 



