8 On the Laws of Terreslrial Magnelum 



have recourse to the detached observations of more modera 

 navigators. But the needles which thev have made use of not 

 admitting of comparison one with another, and their methods 

 of observing not being the same, it is conceived that these dif- 

 ferences must produce many apparent anomalies in the results; 

 and, theiefore, we can at most only hope to find out the ceneral 

 circumstances of these phaenomena, without being able to esti- 

 mate them in detail. Lastly, what increases these difficulties 

 still further is, that we are wholly in want of observations, made 

 in a great part of the globe, where they are so much the more 

 neeessar}', as the whole of the facts appear to indicate there the 

 action of some extremely remarkable local causes, of which it is 

 impossible to form an idea, except from the experiments them- 

 selves. For this reason, I must here confine myself to point 

 out merely what is at present known of the general facts of these 

 phaenomena, without undertaking to connect them by calcula- 

 tions, for which the most essential data are wanting. This will 

 be sufficient to inform navigators of the places on the globe 

 Avhere it would be most useftil for them to direct their attention 

 and to increase their observations. 



I shall first consider the inequalities of the magnetic dip in 

 different climates of the earth ; because this phenomenon ap- 

 pears to vary with time, much less than the variation. The first 

 thing to be done, in order to di;icover ,'onie of its laws, is to deter- 

 mine the points on the globe where the dip is nothing ; or wliere 

 a needle, that is perfectly horizontal before being magnetized, 

 still keeps the same position afterwards. The series of these 

 points form a curve line on the surface of the earth, which is 

 called the magnetic equator, and which all authors have hitherto 

 considered as a great circle, inclined in an angle of about twelve 

 degrees to the terrestrial equator. This, indeed, is what is in- 

 dicated l)y all the observations made on an extent of more than 

 180° of longitude in the Atlantic Ocean, the Indian Sea, and that 

 part of the South Sea which washes the coasts of South America. 

 To explain this, let M' M" (Plate I. fig. 1) be two points on the 

 globe, where observations have shown that the dip is nothing; 

 draw the great circle A E'E" to represent the terrestrial eciua-- 

 tor, and let AM be another great circle perpendicular to A E'E'', 

 representing the terrestrial meridian, of which the longitude is 

 reckoned on the equator. Then, if from the places of observa- 

 tion M'M" we draw other ])ortions of meridians M' E', M"E" 

 terminated also in the equator, the arcs AE', AE" which I call 

 V and /", are the longitudes of the places M' M", and the arcs 

 E'M', E"M", which I call a' and \", are their geographic lati- 

 tudes. This being done, if by these points we draw an arc 



M" M' N' 



