630 PHILOSOPHICAL TRANSACTIONS. [aNNO I683. 



Towards the southern pole the effect is much the same, only that here the 

 south point of the needle is attracted. Hence it will follow, that the variation 

 on the coast of Brasil, at the river of Plate, and so on to the Straits of Ma- 

 gellan should be easterly, as in our third remark ; if* we suppose a magnetical 

 pole situated about 20° more westerly than the Straits of Magellan. And this 

 easterly variation extends eastward over the greatest part of the Ethiopic Sea, 

 till it be counterpoised by the virtue of the other southern pole, as it is about 

 midway between the Cape of Good Hope and the isles of Tristan d'Acuntia. 

 From thence eastward, the Asian south pole, as I must take the liberty to call 

 it, becoming prevalent, and the south point of the needle being attracted by it, 

 there arises a west variation very great in quantity and extent, because of the 

 great distance of this magnetical pole of the world. Hence it is, that in all the 

 Indian Sea, as far as HoUandia Nova and farther, there is constantly west va- 

 riation, at that under the equator itself it arises to no less than 18°, where it is 

 most. About the meridian of the island Celebes, being likewise that of this 

 pole, this westerly variation ceases, and an easterly begins, which reaches, ac- 

 cording to my hypothesis, to the middle of the South Sea, between Zelandia 

 Nova and Chili, leaving room for a small west variation governed by the Ame- 

 rican south pole, which I showed to be in the Pacific Sea, in the 6th and 7 th 

 remark. 



What I have now said, plainly shows the sufficiency of this hypothesis, for 

 solving the variations that are at this time observed in the temperate and frigid 

 zones, where the direction of the needle chiefly depends on the counterpoise of 

 the forces of two magnetical poles of the same nature ; and I suppose I have 

 shown how it comes to pass, that under the same meridian the variation should 

 be in one place 2g^ west, and in another 204- east, as I noted in my gth re- 

 mark. 



In the torrid zone, and particularly under the equinoxial, respect must be 

 had to all the four poles, and their positions well considered, otherwise it will 

 not be easy to determine what the variations shall be, the nearest pole being 

 always the strongest; yet not so as not to be counterbalanced sometimes by the 

 united forces of two more remote, a notable instance whereof is in our 8th re- 

 mark, where I took notice, that in sailing from St. Helena by the isle of Ascen- 

 sion to the equator, on a NW course, the variation is very little easterly, and 

 in that whole track unalterable. For which I give this reason, that the South 

 American pole, which is considerably the nearest in the aforesaid places, re- 

 quiring a great easterly variation, is counterpoised by the contrary attraction of 

 the North American and the Asian south pole, each whereof singly is in these 

 parts weaker than the American south pole, and upon the NW course, the 



