184 A. P. COLEMAN DRY LAND IN GEOLOGY 



The central areas were most effectively scoured, and in many places the 

 rocks beneath, owing to unequal hardness, have been shaped into roches 

 moutonnees, forming hills well rounded on the side from which the ice 

 advanced. Boulder-clay is a highly specialized product of land ice ; float- 

 ing ice, such as floes or bergs, is not known to produce it, the materials 

 dropped through the water when melting being necessarily more or less 

 stratified. The "soled boulders" or "striated stones" from boulder-clay 

 have special characters not caused by any other agency, such as mudflows 

 or torrential action. They are manufactured articles, easily recognized 

 by one familiar with glacier work, and not to be confounded with stones 

 scratched or smoothed in other ways. These familiar features are re- 

 called because they serve as criteria for the recognition of the ancient 

 glaciations to be mentioned later. 



The hummocky, moutonnees surfaces left by the Pleistocene glaciers 

 on Archean rocks which have disordered structures and vary in dura- 

 bility are very characteristic and were once looked on as the direct handi- 

 work of the ice-sheets themselves. The clean and polished surfaces of 

 fresh rock, generally well striated and often deeply scored, are eloquent 

 of the stripping and grinding of the glacier, but the original surface 

 forms have not been greatly changed, as will be shown later. 



Most of the great Pleistocene ice-sheets gathered on comparatively low 

 ground and reached sealevel, often occupying large areas of shallow sea- 

 bottom as well as the land. Few of them began in mountain regions, and 

 the flow of those on level ground was caused by the slope of the upper 

 surface of the ice-mass and not by the inclination of the floor beneath. 

 They could even move uphill for thousands of feet, when the ice-sheet 

 was thick enough in the center, and their flow took place outward in all 

 directions. 



Doubtless conditions were similar in earlier glaciations, and it is not 

 necessary to assume great mountain ranges to account for them, as some 

 geologists have done. 



Peemocaebonieeeous Ice Age 



The first undoubted proofs of ancient glaciation seem to have been 

 found by the Blandfords in India, and the first Memoir of the Indian 

 Survey (1859) contains a brief account of the Talchir tillite in central 

 India, illustrated by a rough sketch. Soon after South African and Aus- 

 tralian tillites of the same age were described. There was at first a good 

 deal of skepticism expressed by European and American geologists as to 

 the reality of the discoveries. Ramsay's interpretation of certain English 

 boulder conglomerates as glacial a few years before had' been disputed, 



