in different Latitudes. 



15 



present a still more certain confirmation of it in the variations 

 observed in Russia, and on the frontiers of China, by the cele- 

 brated astronomer Schubert, who has been so kind as to com- 

 municate them to me. For, having gone in the summer of 1 805 

 from Casau to Tobolsk, and from Tobolsk to Irkutsk, he met the 

 two branches we have just described, one after another in their 

 northern parts, wheie they are widest asunder. This is clearly 

 pointed out by his observations, which I feel it my duty to re- 

 cord here. 



Table. 



The two branches which compose this line either never move, 

 or move very slowlv. It is certain that the variation has not 

 altered at New Holland during the last 140 years. 



Some indications are also to be found of a fourth line, without 

 anv variation. It was observed by Captain Cook in the South 

 Sea, near the point of the greatest inflexion of the magnetic 

 equator. This line has not been discovered by navigators 

 fiirther towards the north ; but it is extremely probable that it 

 is continued there : for, according to a very just remark made 

 by M. Humboldt, since on both sides of each line of no varia- 

 tion, the variation changes its sign from east to west, it neces- 

 sarily follows, that on the whole surface of the globe the num- 

 ber of lines without variation must be even, tiiat it may fall again 

 on the same sign, after all the alternations from plus to minus f. 



Having 



* The minutes were destroved by the seal of the letter, 



t Tiiis will he easily understood by inspectiiijf fig. 5. The circle A is 

 divided inlo three parts: and if we mark the first arc CD with the 

 si^ii + ; the second DEwith — ; and the third EC with +; the first and 

 last arcs will have the same sign, which is not according to nature. On 

 the contrary, in the circle B, which is divided iuj'oui- parts, if wc mark the 

 first FG with + ; the second with — ; the third with + ; and the fourth 

 with — ; (here will he a regular alternation of the hijjiis in proceeding round 



the 



