MR MILNE ON THE GEOLOGY OF ROXBURGHSHIRE. 4(55 



There are, in many places, indications that the rivers, in this country, flowed 

 formerly at a much higher level than they ever now reach. On the north side 

 of the Tweed, for about a mile above its confluence with the Teviot, there is an 

 extensive flat, about 70 feet above the ordinary level of the river, extending back 

 probably half a mile from the river, and there bounded by an abrupt bank, which 

 runs for some distance parallel with the river. On this elevation, flat, or terrace, 

 Floors Castle stands ; as also, a part of the town of Kelso. The terrace, at its 

 side nearest the river, has a steep face or front, about 15 or 20 feet high, at the 

 foot of which there is another and lower terrace, intervening between it and the 

 Tweed. 



At Castleton, there is a steep bank, from 50 to 70 feet high, which runs on 

 the north side of the town for about two miles nearly parallel with the Liddel, and 

 which, at some former period, has evidently been the north bank of the river, but 

 which is now, for a considerable distance, more than a mile distant. The base 

 of this cliif is from 30 to 40 feet above the ordinary level of the river. 



It is proper here to take some notice of those curiously shaped stones in the 

 valley of Allan Water, known by the popular name of Fairy Stones. They are most 

 commonly in the form of flattened spheres, and, though generally separate, they 

 are sometimes united together. They consist of a brownish- white clay, hard in 

 the mass, though easily scratched with a knife. They effervesce very briskly 

 with acids, and they appear from their colour also to contain a small proportion 

 of iron. They are found in the channel, but more frequently on the west bank of 

 the river, at the edge of the stream. 



Various opinions have been offered, to account for these stones ; some sup- 

 posing that they are formed, like stalactites, by the dropping of water ; others, 

 that they are fragments of some hard rock, worn by aqueous attrition. My own 

 opinion is, that they are mere concretions in finely laminated clay, of which 

 there is a large bed on the west bank of the river. But deferring till next part of 

 this Memoir, the grounds of this opinion, I may only here mention, that the bed 

 of clay from which these fairy stones are derived, is overtopped by a mass of 

 gravel, the weight of which, aided by the infiltration of water, causes constant 

 slippings into the river. The clay thus exudes or is squeezed out into the stream. 



The clay is extremely tough and plastic. It resembles exactly the clay found 

 near Berwick, on the left bank of the Tweed, where stones of the same lenticular 

 shape are found. In both places the clay is finely laminated, and equally tena- 

 cious, indicating originally deposition in still waters. 



In the channel of Kale water, near Morebattle, stones, having characters in 

 many respects similar to those just described, occur, though of a very different shape. 

 They are not spherical but elongated sometimes as long as 14 to 18 inches, and 

 with a transverse diameter of only 3 or 4 inches. They have precisely the same 

 colour as the fairy stones of Allan Water, and present like them laminae of strati- 



VOL. XV. PART III. 6 K 



