184 cosMus. 



ern coast of Africa, was 188^° distant from the node in the 

 J:50uth Sea, close to the little islands of Gilbert, nearly in the 

 meridian of the Viti group. In the beginning of the present 

 century, at an elevation of 11,936 feet above the level of the 

 sea, I made an astronomical determination of the point (7° 1' 

 south lat., 48° 40' west long, from Paris), where, in the in- 

 terior of the New Continent, the chain of the Andes is inter- 

 sected by the magnetic equator between Quito and Lima. To 

 the west of this p^int, the magnetic equator continues to trav- 

 erse the South S3a in the southern hemisphere, at the same 

 time slowly drawing near the terrestrial equator. It first pass- 

 es into the northern hemisphere a little before it approaches 

 the Indian Archipelago, just touches the southern points of 

 Asia, and enters the African continent to the west of Socotora, 

 almost in the Straits of Bab-el-Mandeb, where it is most dis- 

 tant from the terrestrial equator. After intersecting the un- 

 known regions of the interior of Africa in a southwest direc- 

 tion, the magnetic equator re-enters the south tropical zone in 

 the Gulf of Guinea, and retreats so far from the terrestrial 

 equator that it touches the Brazilian coast near Os Ilheos, 

 north of Porto Seguro, in 15° south lat. From thence to the 

 elevated plateaux of the Cordilleras, between the silver mines 

 of Micuipampa and Caxamarca, the ancient seat of the Incas, 

 where I observed the inclination, the line traverses the whole 

 of South America, which in these latitudes is as much a mag- 

 netic terra uicognita as the interior of Africa. 



The recent observations of Sabine* have shown that the 

 node near the island of St. Thomas has moved 4° from east to 

 .west between 1825 and 1837. It would be extremely im- 

 portant to know whether the opposite pole, near the Gilbert 

 Islands, in the South Sea, has approached the meridian of the 

 Carolinas in a westerly direction. These general remarks wil] 

 be sufficient to connect the different systems of isoclinic non- 

 parallel lines with the great phenomenon of equilibrium which 

 is manifested in the magnetic equator. It is no small advant- 

 age, in the exposition of the laws of terrestrial magnetism, that 

 the magnetic equ.ator (whose oscillatory change of form and 

 whose nodal motion exercise an influence on the inclination 

 of the needle in the remotest districts of the world, in conse- 

 quence of the altered magnetic latitudes)! shpuld traverse the 



* See the remarkable chart of isoclinic lines in the Atlantic Ocean 

 for the years 1825 and 1837, in Sabine's Contributions to Terrestrial, 

 Magnetism, 1840, p. 134. 



t HuTuboldt, Ueber die seculd'e Verdnderung der Magnetischen Inr 



