L. A. Bauer— Variation of Terrestrial Magnetism. 201 
refer the secular variation to the secular shift of these secondary 
oles. 
4 I shall now show that these secondary magnetic dip poles 
are as truly magnetic poles as those which we believe to be in 
the arctic and antarctic regions. The following main condition 
must be fulfilled: 
In going along a magnetic meridian of the secondary system 
with a dip needle having its plane in that meridian the north 
end of the needle must point vertically downwards over the 
secondary north pole (i.e. over the west focus), at the magnetic 
equator (the zero isapoclinic), the needle must be horizontal 
and over the secondary south pole (east focus), the south end 
must point vertically downwards. 
Let us assume for simplicity, that the secondary magnetic dip 
poles are exactly on the equator, as they are in fact very nearly. 
The equator will then be a magnetic meridian of the secondary 
system. The formula for reducing a dip observed in the 
actual magnetic meridian to the plane passing through the 
secondary magnetic meridian will be on the assumption made: 
tan I, = tan I,,. cosee D 
Ie being the dip in the equatorial plane and Im the dip in the 
actual magnetic meridian and D the observed declination. 
From this it follows that the points where D is zero and Im 
not zero, tan Je=o or Je= 90°. In other words, the intersec- 
tions of the agonic lines with the equator mark the places of 
vertical dip in the equatorial plane. The dotted lines on the 
diagram are the agonics. It will be noticed that they pass 
nearly through the secondary magnetic poles. 
The points where the dip needle would be horizontal in the 
equatorial plane are found with the aid of the above formula, to 
be in long. 188° E. and 2° W. See how closely these points come 
to the intersections of the secondary magnetic equator (zero 
isapoclinics) with the geographical equator. 
Since the horizontal component of intensity directed equa- 
torially, He, can be derived from the formula 
EL: = H,, sin D 
Hm being the observed horizontal component in the actual 
magnetic meridian, it follows that where D is zero He is zero. 
We have already seen that these points fall in the neighbor- 
hood of the secondary poles.* 
Again, let us suppose that we have really two magnetic fields 
instead of one and that they are at right angles to each other, the 
*TIf we construct the isapodynamics with the aid of values of the horizontal 
intensity scaled from Neumayer’s isodynamic chart for 1885, we shall find that 
the departure of observed intensity from computed intensity is zero for the fol- 
lowing two points along the equator, 73° W. and 60° E. 
