Vol. 54.] IN SOUTH WALES, DEVON, AND COENWALL. 253 



than J mile distant. The whole of the Haven would be laid dry if 

 the land were raised enough to bring the bottom of the rock-valley 

 at Milford Docks above low-water spring- tides. 



There are indications of similar submerged valleys farther north 

 on the Welsh Coast. A survey made of the river Tivy, 25 miles 

 north of Milford Haven, shows that at from 4 to 5 miles from the 

 sea, above the lower Cilgerran slate-quarries, where the river has 

 not been partly filled in with quarry-refuse, the bottom is 20 feet 

 below the level of low water as it was in 1851. Above Cilgerran 

 slate-quarries there are depths of as much as 25 feet below the same 

 level. Low water is now several feet higher, in consequence of the 

 accumulations of quarry-refuse in the river. 



Still farther north, at Barmouth,^ it is on record that the rock- 

 bottom of the Mawddach was traced by the cylinder-foundations of the 

 railway- viaduct to 40 feet below the level of low water, where it was 

 shelving rapidly downward beneath a bed of peat of unascertained 

 thickness, but too deep to be penetrated for foundations. 



Along the south coast the next indication of a submerged valley 

 is at Loughor, where the piles of the bridge across the river were 

 driven 37 feet below the level of low water without reaching rock. 



Rivers Tawe and Neath. 



There are evidences of submerged rock-valleys in the rivers Tawe 

 and Neath, though complete sections have not been made out. 



At Landore, 2 miles above Swansea, the river Tawe is crossed by 

 the South Wales Railway. When the first viaduct was built, a 

 boring was put down to a depth of 30 feet, showing, in descending- 

 order blue clay, sand, gravel, and boulders. The viaduct has been 

 reconstructed, and the foundations of the main piers (according to 

 information given me by Mr. Inglis, the engineer of the Great 

 Western Railway Co.) are in hard Eoulder Clay, which was pene- 

 trated by piles for 30 to 40 feet before a hard bearing was reached. 

 Over the Boulder Clay, in ascending order, are soft blue clay with 

 small boulders, then silt, then hard clay and shingle. In a section 

 at Swansea Docks described by Mr. M. Moggridge in 1856,^ the 

 lowest bed, below the level of low water, and not passed through, is 

 described as brown clay and gravel with boulders, some of the 

 latter large enough to require blasting. This is no doubt the 

 Boulder Clay of the Landore section. The material most abundant in 

 it was Coal Measure sandstone, next Millstone Grit, Old Red Sand- 

 stone nearly as abundant, and Carboniferous Limestone less so, 

 though a large boulder was of this material. Assuming that these 

 rocks came from the northern outcrop, they must have travelled 18 

 or 20 miles down the valley. Above this bed was marine clay with 

 Scrohicularia piperata^ in which were peaty beds containing remains 

 of oak, beech, birch, alder, and hazel. 



1 H. Conybeare, Proc. Inst. C. E. vol. xrsii (1871) p. 139. 

 ^ Quart. Journ. Greol. Soc. vol. xii, p. 169. 



