294 MR. A. W. RUCKER AND DR. T. E. THORPE ON A MAGNETIC 
at many stations. He apparently had chiefly the terrestrial isomagnetics in view, 
and did not consider that for the discovery of the local peculiarities of a station, it is 
essential that all three elements should be determined at it. In part this omission 
was due to misadventure. A number of Declination observations had to be rejected 
as the mirror was found to have been out of adjustment. If, therefore, the calculated 
magnitudes and directions of the disturbing forces depend to any great extent on the 
distribution of tlie stations or on the method of reduction, conclusions drawn from 
Welsh’s survey and our own could not lie in harmony. 
In the case of the Declination comparison is possible, not only with the survey 
of 1857, but with the observations collected by Sir F. Evans in 1872. We have 
measured the distances of the stations from the isogonals given by him (‘ Phil. Trans.,’ 
1872, vol. 162, p. 319), and have deduced the disturbances. In the accompanying 
map (fig. 22) we have combined the results of all three surveys. 
The direction of a short line drawn through the stations, shows whether the north 
pole of the needle was deflected to the east or west. Observations made by Mr. 
Welsh are indicated by a dotted line, those due to the naval oflicers by two 
short lines with a dot between them, and our own by a continuous line. 
A discrepancy between any two of the surveys is thus indicated by a cross. 
Lines of no disturbance are drawn, separating districts in which the regional 
disturbance is of opposite signs, and passing, when possible, through stations at which 
different results have been obtained in different surveys. 
Without insisting on the accuracy of the details of these curves, we think that a 
study of the map can only lead to the conclusion that the districts which they bound 
are affected by some common cause which produces a similar effect upon the needle. 
It can hardly be doubted for instance that the magnet is deviated in opposite direc¬ 
tions in the neighbourhood of Elgin and Banff. All three surveys agree as to this 
fact. Again, the evidence is very strong that on the mainland to the north of the 
Caledonian Canal, except, perhaps, near Cape Wrath and the Pentland Firth, the 
needle is deflected by regional forces to the east, while in the Islands on the West 
Coast and the Mull of Cantyre it is deflected to the west. Of course it would be 
possible to find in these districts places, such as Canna or the Cuchullin Hills in Skye, 
where large deviations in both directions could be obtained within a few yards, but 
the great majority of survey stations are chosen too carefully for the local error to 
become so overwhelmingly predominant. We have omitted Portree because the 
ground there is known to be unfavourable to our purpose. The differences between 
the three results obtained are too great to make a mean value trustworthy. At 
Canna on the other hand, where it will be remembered we observed at twenty-three 
places, the mean value cleared of the larger disturbances agrees with the results 
obtained at neighbouring stations. 
The following facts are also important. 
Seven stations were common to the surveys of 1857, 1872, and 1886. One of 
