mountain areas—namely, the central and the circumferential—is for the first time rendered visible.
By adopting this hypothesis, we concede that there is an intimate connection between the glacial period and a predominance of lakes, in producing which the action of ice is threefold; first, by its direct power in scooping out shallow basins where the rocks are of unequal hardness; an operation which can by no means be confined to the land, for it must extend to below the level of high water a thousand feet and more, in such friths as have been described as filled with ice in Greenland (see above, p. 236).
2ndly. The ice will act indirectly by preventing cavities caused by inequalities of subsidence or elevation from becoming the receptacles first of water, and then of sediment, by which the cavities would be levelled up and the lakes obliterated.
3rdly. The ice is also an indirect cause of lakes, by heaping up mounds of moraine matter, and thus giving rise to ponds and even to sheets of water several miles in diameter.
The comparative scarcity, therefore, of lakes of post-pliocene date in tropical countries, and very generally south of the fortieth and fiftieth parallels of latitude, may be accounted for by the absence of glacial action in such regions.
Post-glacial Lake-dwelling in the North of Italy.
We learn from M. de Mortillet that in the peat which has filled up one of the 'morainic lakes' formed by the ancient glacier of the Ticino, M. Moro has discovered at Mercurago the piles of a lake-dwelling like those of Switzerland, together with various utensils, and a canoe hollowed out of the trunk of a tree. From this fact we learn that south of the Alps, as well as north of them, a primitive people having similar habits flourished after the retreat of the great glaciers.
