and the evidence, of their having existed in Britain. 573 



diverging at the outlets of the valleys, and their being oblique, and 

 never horizontal on the flanks, which they would be, were they due 

 to the agency of water, or floating masses of ice. The cause of this 

 obliquity the author assigns to the upward expansion of the ice, and 

 the descending motion of the glacier. 



The most remarkable striated rocks in the Alps are near Handeck, 

 and near the cascade of Pissevache ; and the best examples M. 

 Agassiz has seen in Scotland, are those of Ballahulish, and in Ireland 

 those of Virginia. 



If the analogy of the facts whicli he has observed in Scotland, 

 Ireland, and the north of England, with those in Switzerland, be 

 correct, then it must be admitted, M. Agassiz says, that not only 

 glaciers once existed in the British Islands, but that large sheets 

 (nappes) of ice covered all the surface. 



The author then details the proofs that glaciers did not descend 

 from the mountain summits into the plains, but are the remaining 

 portions of the sheets of ice which at one time covered the flat countrj*. 

 It is evident, he says, if the glaciers descended from high mountains, 

 and extended forward into the plains, the largest moraines ought to 

 be the most distant, and to be formed of the most rounded masses ; 

 whereas the actual condition of the detrital accumulations is the re- 

 verse, the distant materials being widely spread, and true moraines 

 being found only in valleys connected with great chains of lofty 

 mountains. 



It must then be admitted, the author argues, that great sheets of 

 ice, resembling those now existing in Greenland, once covered all 

 the countries in which unstratified gravel is found ; that this gravel 

 was in general produced by the trituration of the sheets of ice upon 

 the subjacent surface ; that moraines, as before stated, are the effects 

 of the retreat of glaciers ; that the angular blocks found on the sur- 

 face of the rounded materials were left in their present position at 

 the melting of the ice ; and that the disappearance of great bodies of 

 ice produced enormous debacles and considerable currents, bj- wiiich 

 masses of ice were set afloat, and conveyed, in diverging directions, 

 the blocks with which they were charged. He beheves that the 

 Norwegian blocks found on the coast of England have been correctly 

 assigned by Mr. Lyell to a similar origin. 



Another class of phajnomena connected with glaciers, is the form- 

 ing of lakes by the extension of glaciers from lateral valleys into a 

 main valley ; and M. Agassiz is of opinion, that the parallel roads of 

 Glen Roy were formed by a lake which was produced in consequence 

 of a lateral glacier projecting across the glen near Bridge Roy, and 

 another across the valley of Glen Spcane. Lakes thus formed natu- 

 rally give rise to stratified deposits and parallel roads, or beds of 

 detritus at diff'erent levels. 



The connexion of very recent stratified deposits Avith glacier- 

 detritus, j\I. Agassiz observes, is difficult to explain ; but he conceives 

 that the same causes which can bar uj) valleys and form lakes, like 

 those of Brienz, Thun and Zurich, may have formed analogous 

 barriers at tiic point of contact with the sea sufficiently extensive to 



