278 NEOZOIC TIME. [CHAP. XIII. 



due to the action of floe-ice carried in a definite direction 

 by a steady current. In the next place, he assumes that 

 all Boulder-clays are the " ground-moraines " of confluent 

 glaciers or ice-sheets, and that they have invariably been 

 formed between such ice- sheets and the land surfaces over 

 which the ice passed. That such confluent glaciers existed 

 during the Ice Age in Britain I would not deny, but it by 

 no means follows that the various Boulder-clays were each 

 and all formed on the surface of the land. There is great 

 difficulty in understanding how a country could be glaciated 

 and at the same time covered with an almost universal 

 mantle of Boulder-clay if both processes were effected by 

 an ice- sheet moving over dry land. For if the thickness 

 and weight of the ice were such that its base conformed 

 to the surface of the country, and the pressure it exercised 

 so great that every slight prominence was scratched, 

 grooved and moulded by the stones frozen into the ice, how 

 could there be room at the same time for a layer or pad of 

 compacted mud to accumulate between the ice and the 

 rock? Professor Geikie has been confronted with this 

 difficulty, and he admits that it is impossible for the two 

 operations to take place at the same place and at the same 

 time, but he suggests that the clay was formed under 

 those parts of the ice where the pressure was least, and 

 that the rock was only scratched where the pressure was 

 greatest. If Boulder-clay was found to fill lake-like 

 hollows this explanation might be accepted, but since it 

 behaves as if it had originally formed an almost universal 

 mantle over all the lower parts of the country, the suppo- 

 sition cannot be admitted. Moreover, Professor G-eikie 

 often disregards his own suggestion and refers the for- 

 mation of a Boulder-clay overlying a glaciated surface to 

 the action of the ice which scratched that surface. To 

 most minds I think the obvious inference would be that 

 the scratching must have been done first, and that the clay 



