R. M. Deeley — Erosion bij Rivers and Glaciers. 395 



ill contact with water scarcely subjected to any stress whatever. 

 Such conditions, however, are not stable ones, and must give rise 

 to continuous thawing and refreezing. 



Between ice in this condition and other rocks, when tbe water 

 can escape through tbe porous ground, and is therefore not subjected 

 to any great hydrostatic pressure, it has a freezing-point above that 

 of tbe ice, and the conditions are favourable for the adherence of the 

 glacier to its bed. The exact nature of the freezing process may be 

 seen by considering the case of a wire loop passed over a piece of 

 ice and weighted at its ends. In such a case the ice under pressure 

 melts, the water moves round the wire and again freezes on its top. 

 ]n this way the wound is healed as fast as the wire cuts its way 

 down, and the sling finally passes right through the mass without 

 separating it into two portions. When resting upon a porous rock, 

 such as sandstone or gritstone, each particle of quartz pressed into 

 the glacier causes it to melt, and a portion of the low-temperature 

 water, as it escapes, freezes again in the sandstone. Indeed, rough 

 porous surfaces must imbed themselves into the glacier and resist 

 sliding, even should they not actually freeze to the glacier. The 

 tendency of ice is really rather to adhere to the bottom, and drag 

 along with it, and break up, the rocks upon which it rests. In all 

 glaciated districts, as a matter of fact, porous rock-surfaces beneath 

 Boulder-clays, or over which ice has passed, are contorted, torn, and 

 trailed, and but seldom striated or polished. 



Soil-cap motion has been largely pressed into service to explain 

 many of the features which have really resulted from the drag of 

 a glacier along its bottom. However, a critical examination of the 

 phenomena presented by some forms of "trail" and " underplight " 

 reveals peculiarities which soil-cap motion, resulting from gravity, 

 could not have caused, the direction of flow indicated by it frequently 

 being uphill, or sufficiently powerful to bend and displace rocky 

 beds. This tearing and breaking-up action of glaciers also comes 

 into play when they pass over hard but jointed and bedded rocks. 



There are few geologists who have visited glaciated districts who 

 have not been struck by the great difference in the appearance of 

 vallej's viewed from the two aspects, up and down. Looking down, 

 all the surfaces and outlines are seen to be rounded, whereas looking 

 wp, rough, rocky, and irregular surfaces everj'where meet the eye. 

 According to the teaching of one school, these rough surfaces were 

 formed before the advent of the glaciers, and represent portions of 

 the old Pre-Glacial weathered surfaces ; the subsequent work of ice 

 having merely served to round and smooth one side of them. Such 

 a view affirms that the erosion afi'ected by the glaciers must have 

 been very small. But the angular and nnweathered appearance 

 of these craggy surfaces, and the presence of large quantities of 

 angular unweathered rock in the great moraines and sheets of 

 boulder-clay, clearly negative this view, and prove that glacier 

 erosion has really been extremely active. 



Now when a glacier moves down a valley it presses heavily 

 against each boss or hummock of rock it touches. Here melting 



