SURVEY OF THE BRITISH ISLES FOR THE EPOCH JANUARY 1, 1886. 261 
postulate only the possibility of the existence of points or surfaces which exert 
magnetic forces. They will be called attractive or repulsive accordiug as they attract 
or repel a north-seeking pole. If the existence or apparent existence of such centres 
is established the cause of the pheDomenon will be a proper subject for enquiry. 
The distortion produced by a symmetrical mass of magnetic matter on the isogonals 
depends upon the angle which these curves make with the magnetic meridian. 
If we are dealing with a sufficiently small area, both the isogonals and the magnetic 
meridians may be represented by straight lines. If these lines are mutually 
perpendicular, and if the declination is westerly and increases with the latitude, the 
forms of the true isogonals in the neighbourhood of a symmetrical attracting mass, 
the centre of which is below the surface at A, will be of tlie type shown in Fig. 3, in 
which NS is the magnetic meridian through A. 
Fig. 3. 
N 
-Teri’estrial Isogonals ruuuiug approximately east and west. 
-True Isogonals produced by centre of attraction at A. 
In Japan the distribution of the isogonals and meridians approximates to this simple 
arrangement, and in the neighbourhood of the Fossa Magna according to Dr. Naumann 
(‘Die Erscheinungen des Erdmagnetismus,’ Stuttgart, 1887), a very remarkable 
disturbance of this kind is produced. 
If the isogonals and meridians are very nearly coincident, and if the Declination 
increases with the longitude, the effect on the isogonals of a weak attractive centre 
will be of the type shown in Fig. 4. If, however, the disturbing force is sufficiently 
powerful to make the Declination, at some point to the east of the centre, greater than 
its undisturbed value at the centre, the form of the lines must approximate to that 
shown in Fig. 5. The Declination will increase rapidly with the longitude, attain a 
maximum value, fall to its normal value at points on the magnetic meridian which 
pass through the centre—on each side of which the attracting matter is supposed 
to be symmetrically situated—then fall to a minimum, and finally resume its normal 
rate of increase. 
If, retaining all the other assumptions, we suppose that the isogonals and meridians 
are inclined at an angle of less than 90°, the distorted curves will assume forms 
similar to those in Fig. 6. It will be noticed that if the undisturbed isogonals be 
