T. Mellard Reade — Glacial Geology. 317 



The divergence of flow from Criffel will thus have been as great 

 as 40 degrees, a very improbable thing to have happened with an 

 ice-sheet traversing a basin fed from three sides. But when we 

 consider that Lake District granites are found all along the same 

 base, the improbability becomes to my mind an impossibility ; for 

 it would involve, first, a concentration and mixing of the two 

 differently-derived rocks on or in the ice-sheet, and their after dis- 

 tribution in a fan-like form. This is omitting from consideration 

 the further difficulty of the Antrim flints and other erratic rocks 

 found over the same area. 



High-Level Sands and Gravels. 



Not only does the land-ice theory fail to explain the areal distri- 

 bution of erratics, but we have the further difficulty to contend with 

 that they are found at all levels up to 1400 feet above the sea. 

 They are, as already pointed out, proportional!}' greater in numbers 

 at the top of Tryfaen than in the Drift of the coastal plain of Car- 

 narvonshire. As a question in dynamics it has never been shown 

 ■what gradient would be required for an ice-sheet from the North 

 thus to overpower the native glaciers of Snowdonia. The Welsh 

 mountains are higher than any in the South of Scotland, so that to 

 overpower the thrust of the ice northwards from a Snowdonian 

 centre, and to deflect it south-westwardly, an enormous pile of snow 

 would have to be concentrated in Scotland. A surface slope of half 

 a degree would mean a depth of snow of over a mile at Crifi'el, plus 

 the depth of snow on Tryfaen. The gradient of the Welsh ice-sheet, 

 even if the top of Snowdon were bare, would be about 4 degrees 

 from Snowdon to the top of Tryfaen. Putting it in the most favour- 

 able light for the land-ice theory, it would appear that even had 

 the snow been two miles thick at Criffel, and in Wales reached only 

 to the top of Snowdon, the Snowdonian glacier would have overcome 

 the northern ice-sheet and have eff"ectually prevented the landing 

 of sea-bottom on a flanking spur of Snowdon, were such an event 

 otherwise possible. 



But there yet remains another difficulty ; the deposits in question 

 are admittedly aqueously deposited, so that on the ice-sheet theory 

 the ice containing these sea-bottom remains must have melted and 

 in such a manner as to have deposited the sands and gravels and 

 boulders in a stratified mass on the summit of a hill 1400 feet above 

 the sea-level. No one has grappled with this hydraulic difficulty. 

 Similar reasoning applies to the Gloppa drift, which is 1100 feet 

 above the sea and over 60 feet thick, probably 100 feet, as it has 

 never been bottomed, and bears in its curi'ent-bedded stratification 

 indubitable marks of aqueous deposition. There still remains a most 

 important fact which must not be lost sight of in this connexion, 

 namely, the presence of perfect and delicate shells. Since Mr. 

 Nicholson, F.G.S., made his splendid collection of glacial shells on 

 Gloppa, this difficulty has been emphasized. The fragmentary con- 

 dition of most of the Drift-shells had been triumphantly pointed to 

 as a convincing proof of the passage of an ice-sheet over them. 



