105 



from fear of their neighbours on the mainland, or they were, themselves, the 

 men of violence who lived partly by plunder, which they conveyed to their refuge 

 in the lake, whence they might, generally with advantage, defy pursuit. 



The latter was probably the character of the dwellers upon this island, 

 otherwise how account for the great variety of bones, not only of wild animals 

 — which formerly abounded here — but pigs, cattle, horses, and sheep. If they 

 were peaceable men, they must have had more contrivance in economising space 

 in their farm-yards than prevails in the present day. 



But to refer again to the old tradition of the submerged city, and the 

 name of the water — "Lake of the Sunken Land," it is very possible there may 

 have been other similar dwelling-places, which, by some catastrophe, may have 

 been engulphed. One other such island, of much smaller dimensions, is known 

 to the writer, though without protection by piles. 



What was before said regarding the undoubted change of the level of the 

 lake, will here be suggestive of an explanation of the old tradition. The raised 

 beach in the meadow below Talyllyn Parsonage shows how completely this island, 

 if it then existed, must have been submerged ; and it may have been at this period 

 that Giraldus visited the lake. The results of tolerably careful dredging and 

 sounding in the narrow southern extremity of the lake— 600 or 700 yards in length 

 —show that the bottom is very level, the depth of water increasing from about 

 12ft. to 15ft., where it may be said to join the larger portion of the lake. 



On examination of the nearly level meadows about Mr. Kaikes' boat- 

 house, the upper soil beneath the turf is found to be a kind of peat, from one 

 to two feet in thickness, composed of the remains of reeds and other aquatic 

 plants. Below this is the shell marl. Along this shore the water deepens, 

 rather abruptly, to the depth above-mentioned, and on examination of the 

 bottom, by dredging, the same formation is found as upon the adjacent meadows. 

 There is, first, the layer of peat materials, but of course, being under water, 

 not compacted, and below is the shell marl. Particular attention is desired to 

 the fact of the bottom here being level, or as nearly so as the meadows above. 

 This was, doubtless, at one time comparatively shallow water, or the materials 

 for the surface layer of peat could not have grown there. There may have been 

 many pile dwellings here, or along its banks. Then, to all appearance, a sudden 

 rise of the water took place, perhaps to the height of 20ft., or nearly level 

 with the spot upon which Talyllyn Church is built. A fresh deposit of shell 

 marl then commenced, and, in due time, the peat formation was deposited. 



The term "sunken land" may thus refer either to the portion thus deeply 

 submerged, or to the shallow shores about Talyllyn Church, where it is ex- 

 tremely probable there may have been dwellings on dry land. A subsidence 

 to the extent of a few feet afterwards took place through natural causes, reducing 

 the level to what it is at present, or within about 18in. of it, by which amount it 

 was reduced, by artificial drainage, a few years back. 



