SURVEY OF THE BRITISH ISLES FOR THE EPOCH JANUARY I, 1886. 85 
the smaller value occurs at the western station. Between them lie Hereford, Ross, 
and Chepstow, and the rates obtained from observations at these in April, 1889, are 
— l'’85, — l'’83, and — l'‘84, which are in all cases less than those at Malvern and 
Clifton. These observations then support the view that there is a real maximum 
followed by a minimum rate of change of Dip in this neighbourhood. 
Fig. I. 
Secular change of Hip, 1858-86. 
This induces us to show in fig. 1 that if all stations at which there is a maximum 
rate of change are joined, we get two long lines running from Edinburgh to Exeter, 
and from Stornoway to Limerick. The stations are not sufficiently numerous to 
enable us to draw any certain conclusions about them, but the point ought not to be 
lost sight of when the survey is repeated. 
For the present we note only that of the three pairs of neighbouring abnormal 
stations at which observations were made in all three surveys, all three are abnormal 
(in the sense tliat the rate of change is greater at the easterly station) whether we 
take the interval 1837-87 or 1857-87. In both cases the rate at Inverness is greater 
than at Fort Augustus, at Glasgow than at Cumbrae, and at Berwick than at 
Edinburgh. 
