in different Latitudes, 99 



necessarily change the chemical state and the natural arrange- 

 ment of the ferruginous parts, in places where it exercises itself; 

 which changes cannot be made without affecting the direction 

 of the magnetic needle, and modifving the general action of the 

 globe in those points. There are, indeed, several examples of 

 these alterations which have happened suddeidv; and M. Hum- 

 boldt has observed similar ones at Peru, after a great earth- 

 quake*. It is therefore possible, that the particular magnetic 

 centre belonging to the South Sea is owing to similar causes. 

 There exist, without doubt, analogous ones in other countries: 

 and is it not their alterations which during the last two 

 hundred years have produced these changes of variation of the 

 compass : Avhich changes are so irregular, that it has hitherto 

 been impossible to find any law for them, but which, by this 

 very irregularity, ap])ear to announce that they are not the 

 effect of an uniform and constant cause? According to this idea, 

 nothing in Europe would oblige the compass to return towards 

 the east; and indeed, since it has ceased to decline to the west, it 

 has not been observed to retrograde bv any sensible cjuantity : 

 so that, according to the observations hitherto made, it is im- 

 possible to decide whether it will ever return or not. 



The magnetic action of the terrestrird globe is not confined 



these will act on each other, and gradually change the forces with whicli 

 they attract, and thereby influence the vaiiatiun of tlie compass. This 

 action will however, in time, hring the « hole to an equilibiium ; and as two 

 poles appear to be the natural and most simple state, to chis it will always 

 have a tendency to return, although the internal chanjics that take place in 

 the bowels of tiie earth, arising; tVorn chemical causrs, niay frequently coun- 

 teract and retard this operation. I'lie metallic substances wliich the earth 

 contains near its surface are well known to be contiiuially altpriiiif their 

 nature: and if the variati.in of the compass depend at all upon the changes 

 of these substances, as is commonly supposed, it is evident, the;;e changes 

 take place much more rapidlv than is sieiierallv belitvi d. 



* Father de la Tone observed in Italv, during the great eruption of 

 Mouiit\'esuvius, that the variation was 10" in the niornin<i; at noon it was 

 14'; and in the eveuins: it was 10", and it continued in that state t;ll the 

 lava grew so cold as no lonucr to emit any lijiht lu the night, after which it 

 slowly increased to liij deureffS, where it remained. 



Daniel Ijernouilli I'ound the needle change its position 45' by an earth- 

 quake. 



IVofessor Midler, at Manhcini, observed also that the variation of the 

 peedie in that place was greatly aflccted by the earthquake in Calabria. 

 Such streams of lava as (lowed from Hecla in the last dn^adful eruption, 

 must have uiadc a transfereii<;eof magnetic niattor that would considerably 

 affect the needle: but no (jbservations seem to hiivo been made on the oc- 

 casion ; for we know that common ii()n-st<jne, *\liich has iu> effect on the 

 needle, will, by mere cenieniafion vrith iiiiy inllammable substance, bccoiiie 

 iiiaunctic. In this way Dr. Knight sometimes made artilicial loadstones. 

 —EiiC2/. ISrit. 



G 2 to 



