824) Capt. Kater’s Experiments fm^ Deterrmning tlie Varlatim 
treme stations should comprise an arc of sufficient length to ren- 
der the effect of irregular attraction insensible ; and this effect 
might be diminished, if not wholly prevented, by selecting sta- 
tions of similar geological character, and which should differ as 
little as possible in elevation above the level of the sea. 
If, however, some deduction be made for the superior densi- 
ty which it has been remarked exists at Portsoy, the compres- 
sion deduced from that station and Unst, may perhaps be 
considered as not far distant from the truth, both being situated 
on rocks of a similar nature ; Unst consisting chiefly of serpen- 
tine, and Portsoy of serpentine, slate, and granite ; and as 
the ellipticity given by the experiments at Unst and Arbury 
Hill, is nearly the same as that resulting from Urist and Port- 
soy, it would be no improbable conjecture that the sudden in- 
crease of gravitation observed at Arbury Hill, may be occasion- 
ed by a rock of primitive formation, approaching the surface of 
the earth in the vicinity of that station 
These facts appear sufficient to explain the anomalies which 
have been remarked in the Trigonometrical Survey of Great 
Britain. For if the disturbing force in the neighbourhood of 
Arbury Hill, were supposed to be situated to the north of tliat 
station, the plumb line would be attracted northward, the ob- 
served latitude would be less than the true, and the length of 
the degree deduced from the arc between Dunnose and Arbury 
would be in excess, and that derived from the arc between Cliff 
ton and Arbury in defect. This last error will be augmented, 
if we suppose the attraction of the matter near Arbury Hill to 
be felt at CJifton, and the plumb line at that station to be drawn 
towards the south. 
M. Biot, by a comparison of his numerous experiments at 
Unst with those made at Formentara and Dunkirk, in conjunc- 
tion with M. Arago, obtains for the resulting compression. 
But if the allowance for the elevation of Formentara above the 
level of the sea, be corrected in the manner suggested by Dr 
Young, the ellipticity should be about The details of 
* Since the above was written, I find the conjecture I have hazarded remarks 
ably supported by fact ; for on consulting Smith's Geological Map of England, it 
appeal's that Mount Sorrel, a mass of granite, is situated, together with othep 
recks of piimitive formation, about 30 miles to the north of Arbury HilL 
