308 G, W. Bulman — Formation of the Boulder-clay. 



stones while carrying others for long distances without taking off 

 their sharp edges ; it must bold firmly embedded in its lower surface 

 stones witb which to striate and groove the rocks below, while at 

 the same time it leaves a large number behind to form the moraine 

 profonde. 



And when we turn to the glaciers themselves we see them accom- 

 plishing a series of actions as apparently incompatible. For while the 

 ice floios along its bed more or less like a river, it yet polishes and 

 striates the rocks below, apparently by holding stones firmly grasped 

 in its lower part ; and while grinding down the rocks over which it 

 jDasses to the finest powder, it has yet been observed to rise gently 

 over loose deposits of clay and stones without disturbing them ; at 

 times it allows stones to sink into it, so that its surface moraines 

 disappear, while at other times it works up the stones from below 

 to the surface, and they appear again. 



While the anomalies of Glacial action ai'e so great it would be 

 rash to assert that Boulder-clay cannot be formed beneath the ice, 

 but the great fact of constantly running sub-glacial streams renders 

 it improbable. These streams laden with fine rock-flour issue from 

 the termination of glacier and ice-sheet summer and winter. How 

 far such streams ascend the glacial valleys it is difficult to say. 

 Probably they extend at least as far as the point where the neve 

 gives place to the ice-river. Such a conclusion, at least, seems to 

 follow from certain considerations as to the probable temperature 

 beneath a considerable sheet of ice or snovi^. 



Ice and snow are bad conductors of heat, and hence the lower 

 surface of a thick ice-sheet will tend to take the temperature due to 

 the surface of the earth — say, about 50° F. The pressure, moreover, 

 of the superincumbent ice will tend to liquify the lower portion. 

 Such streams running beneath the entire extent of the glacier 

 proper would readily account for the number of rounded stones 

 found in the Boulder-clay, as well as for intercalations of sand and 

 gravel ; but do not permit the supposition of a sufficient quantity of 

 rock-flour being left behind to form more than such fragmentary 

 fringes and patches as a river leaves along its course. The greater 

 part of this finely ground rock matter will be carried beyond the 

 limit of the ice, and if the glacial streams spread out over level 

 plains deposits of clay of considerable extent might be formed on 

 land. Most of this sediment, however, will be laid down in water. 

 If we study the distribution of glacial mud in Switzerland at the 

 present day we find it is mostly laid down in lakes. Many of its 

 glacial mud-laden streams fall into lakes from which they emerge 

 clear and pm-e, having deposited their sediment thei'ein. Thus the 

 Ehine flows into Lake Constance and the Unter Sea, the Rhone 

 into tlie Lake of Geneva, the Aar into the Lakes of Brienz and of 

 Thun, the Ticino into Lake Maggiore, the Reuss into the Lake of 

 Lucerne. 



This fine sediment would pi'ovide the clay, and if boulders could at 

 times be carried in, the conditions for the formation of boulder-clay 

 would be present in these Swiss lakes to-day. And, whatever may 



