in different Latitudes. 553 



mean terms between several serieses of observations, which 

 always cliffi;reil very little iVoiii each other. It apptars 

 thence ti) result that the action of the Alps has a sensible 

 influence on the intensity of the magnetic force. M. Hun:- 

 boldt observed analogous eflects at the bottom of the 1 •- 

 renecs; for example, at Perpignan. It is not improbable 

 that they arose from the mass of these mountains, or the 

 ferrugineous matters contained in them; but whatever mav 

 be the cause, it is seen by these examples that the general 

 action of terrestrial magnetism is'seusibly modified by local 

 circumstances, the differences of which may be perceived 

 in places very little distant from each other. This tmth 

 will be further confirmed by the rest of this uiemoir. 



It is to causes of this kind, no doubt, that we must 

 ascribe the diminution of the magnetic forces observed in 

 some mountains ; a diminution which, on the first view, 

 might appear contrary to the results obtained during the last 

 aerial voyages. This conjecture is supported by several ob- 

 servations of M. Humboldt. By making his needle to oscil- 

 late on the mountain of Guadaloupe, which rises 6j6 metres 

 (338 toises) above Santa-Fe, he found it in 10 minutes oi' 

 time give two oscillations less than in the plain. AtSilla, near 

 Caracas, at the height of 263'2 metres (l31(j toises) above 

 the coast, the diminution went so far as five oscillations; 

 and, on the other hand, on the volcano of Antlsana, at the 

 lieight of 4934 metres {i!4(37 toises), the number of oscilla- 

 tions in lU minutes was 230; though at Uuito it was only 

 218: which indicates an increase of intensitv. I observed, 

 indeed, a similar etlcct on the suumiit of JViount Genevrc, 

 at the height of I600 or 1600 metres (3 or 9OO toises), as 

 .maybe seen i)y the nunibers uhich I have r'lready given; 

 and it was on this mountain that I found the ijrreatest in- 

 tensity of t,he magnetic force. I saw on the hill of La Su- 

 perga, in the neighbourhood of Turin, an example of these 

 variations equally striking. Observing, with Vassali, on 

 this hill, at the elevation of about 600 metres (300 toises), 

 vvc found 87 oscillations in 10 minutes of time. On the 

 side of the hill we had 88-8 oscillations; and at the bot- 

 tom, on the bank of the I'o, we obtained &7-3, Thouijh 

 these results approach very near to each other, their difit";- 

 tnce is, however, sensible, and fullv shows that their small 

 variations must l)c considered as slight anomalies produced 

 by local circumstances. 



This cxanunation le-ids us to consider the intensity of 

 iTiagnctism on the different points of the surface of the 

 ^lobe, ai subjcet to two sorts of differences. One kind are 



geniral : 



