in Cardigan Bay. 243 



hand, squeeze the water from it as from a sponge, and was 

 struck with the small space it occupied after such a pressure. 

 It seemed obvious, that if these trunks had been thrown down 

 instead of standing upright, they would have assumed under 

 the weight of superincumbent masses of earth the flat ribbon- 

 like form, which is usual in the native charcoal of the carbo- 

 niferous strata ; and it appeared that the preservation of the 

 delicate structure of the wood with its cells and vessels, not- 

 withstanding the removal of all the vegetable principles con- 

 tained in them, might throw light on the process of petrifac- 

 tion. 



I am informed that the appearances are exactly similar in 

 the northern portion of the submarine forest, which I have 

 not visited. A natural mound or wall of shingles in this, as 

 in the portion above described, separates the sands and sub- 

 marine forest from a tract of marsh and bog, which owes its 

 present condition to the waters of a stream, partly arrested by 

 the mound. 



With regard to the origin of these appearances, I see no rea- 

 son to have recourse to any subterranean movement in order to 

 explain them. The wall of shingles, which now forms a na- 

 tural rampart between the sandy shore with its peat and sub- 

 marine forest on the west, and the tract of peat and marsh on 

 the east, must be supposed, notwithstanding its great dimen- 

 sions, to be liable to changes of position. Placed further sea- 

 ward, it would inclose the tract, which is now submarine ; and 

 if after the growth of the forest, its destruction, and the forma- 

 tion of peat upon its remains, the sea made a breach in the 

 rampart, the present state of things would be in no long time 

 the result, as the difference of level is inconsiderable between 

 the submarine portion and that which is remote from the in- 

 fluence of the waves, and the slight difference of level which 

 exists may, if necessary, be accounted for from the percolation 

 of the same waters which produce the marsh and bog, and 

 which might gradually carry away portions of the substrata. 



The tract which I have been describing is called by the 

 Welch, Cantrev Gwaelod, which means the lowland hundred. 

 According to their ancient memorials it was submerged by an 

 irruption of the sea about the year 520 of the Christian sera. 

 In the Triads of Britain, one of the. most important records, 

 this disaster is imputed to the folly of" Seithenyn, the drunk- 

 ard, who in his drink let the sea over the Cantrev Gwaelod." 

 The representations of the bards respecting the importance 

 and opulence of this district are probably exaggerated ; but 

 I see no reason to dispute their testimony either respecting 



2 12 



