yOL. XIII.] PHILOSOPHICAL TRANSACTIONS. ^2/ 



piemen's Land, in south latitude, there arises another easterly variation, which 

 seems not so great as the former, nor of so large extent ; for that at the island 

 Rotterdam it is sensibly less than on the east coast of New Guinea ; and, at the 

 rate it decreases, it may well be supposed that about 20° farther east, or 225° 

 east longitude from London, in the latitude of 20*^ south, a westerly variation 

 begins. — 7. That the variations observed by the honourable Sir John Nor- 

 borough at Baldivia, and at the west entrance of the Straits of Magellan, 

 plainly show, that that east variation noted in our third remark is decreasing 

 apace; and that it cannot reasonably extend many degrees into the South Sea 

 from the coast of Peru and Chili, leaving room for a small westerly variation 

 in that tract of the unknown world that lies in the midway between Chili and 

 New Zealand, and between Hounds Island and Peru. — 8. That in sailing north- 

 west from St. Helena by Ascension, as far as the equator, the variation con- 

 tinues very small east, and as it were constantly the same; so that in this part 

 of the world the course, wherein there is no variation, is evidently no meridian, 

 but rather north-west. — 9. That the entrance of Hudson's Straits and the mouth 

 of the river of Plate, being nearly under the same meridian, at the one place 

 the needle varies 29-^° to the west, at the other 20^° to the east. This plainly 

 demonstrates the impossibility of reconciling these variations by the theory of 

 Bond : which is by two magnetical poles and an axis, inclined to the axis of the 

 earth ; from whence it would follow, that under the same meridian the variation 

 should be in all places the same way. 



These things being premised may serve as a sure foundation to raise the su- 

 perstructure of a theory upon. But first it would not be amiss to show hereby 

 the mistake of Gilbert and Descartes; the first whereof supposes, that the 

 earth itself being in all its parts magnetical, and the water not, wheresoever the 

 land is, thither also should the needle turn, as to the greater quantity of mag- 

 netical matter. Now this in many instances is not true, but most remarkable 

 on the coast of Brasil, where the needle is so far from being attracted by the 

 land, that it turns the quite contrary way, leaving the meridian to lie NbE, 

 which is just along the coast. As to the position of Descartes, that the iron 

 and loadstones hid in the bowels of the earth, and the bottom of the sea, may 

 be the causes that the needle varies; if we consider for how great a part of the 

 earth's surface, ex. gr. in the whole Indian sea, the needle declines the same 

 way, and that regularly ; it will follow, that the attracting substance that occa- 

 sions it must be very far distant. Now by experience we find the little force 

 that iron guns have on the compass in ships, and the experiments now before 

 the Royal Society do plainly show, how little a magnetism there is in most 

 crude iron ores, what quantity thereof must be then supposed to make so 



4L 2 



