70 
FIFTH REPORT- 1835. 
It further appears that A and B recede from, and a and b , on 
the contrary, approach, the geographical poles respectively ad¬ 
jacent to each. 
We may conceive the poles corresponding to A and B (which 
are the points of more powerful directive influence in each he¬ 
misphere) to belong to one magnetic axis, and the poles corre¬ 
sponding to a and h to another magnetic axis, subject to the fol¬ 
lowing conditions; viz. that the two northern poles have an 
eastward motion, and the two southern a westward motion; that 
in each axis the north pole has a considerably quicker motion 
than the south pole; that both poles A and B of the one mag¬ 
netic axis recede from, and that both poles a and b of the other 
axis approach, the adjacent geographical poles ; and lastly, that 
the axes prolonged are not diameters, but chords .; they could 
only be diameters for a moment, on account of the opposite mo¬ 
tion of the opposite poles; and then only if it should also hap¬ 
pen that each end of the axis is at the same time at equal di¬ 
stance from its adjacent pole of the earth*. 
In introducing the supposition of two magnetic axes, M. Han- 
steen carefully guards against being understood to affirm the 
existence of such, or as having any purpose beyond that of sug¬ 
gesting a convenient mode of connecting together and represent¬ 
ing the facts made known by observation. So far from deciding 
on, he does not even discuss, the question, whether the causes 
of terrestrial magnetism are to be sought within the globe, or 
externally; though it is apparent that his own opinion inclines 
to the latter supposition. But with this understanding, he af¬ 
firms that all the phenomena hitherto made known by observa¬ 
tion admit of representation upon the supposition of two such 
magnetic axes as are above described; and he proceeds to ex¬ 
emplify this by showing the explanation such an hypothesis 
affords of the changes of variation and dip in the maps of those 
phenomena corresponding to different epochs. 
Reckoning back the positions of the points of convergence to 
the year 1600, and viewing them in connexion with the map of 
the variation for that year, we should have b in 83° N. and about 
30° E., i. e. to the north-east of Spitsbergen ; and a to the south 
west of Terra del Fuego ; at which time the axis ah was conse¬ 
quently much nearer the Atlantic than the Pacific. B, requiring 
to be placed more westerly than its later situation, would be 
* We may imagine in each axis two principal epochs; one when the two poles 
should be in opposite meridians, and the other when they should both be in one 
and the same meridian; but in the first case the prolonged axes would still not 
be diameters, because the opposite poles are at unequal distances, each from its 
adjacent pole of the earth. 
