1877*1 The Loess of the Rhine and the Danube. 83 
the land rose again, without any evidence of marine life 
having been left behind. 
But is there really no other way of getting water up to 
the heights we require without resorting to this extreme 
hypothesis ? I have in former papers urged that there is — 
that the ice of the Glacial period flowed principally down 
the ocean depressions and blocked up the drainage of the 
continents as far as it extended, causing immense lakes of 
fresh or brackish water. I was first led to believe that the 
ice had effected this in studying the glaciation of North 
America, and in 1866 I advanced the opinion that the 
drainage of the St. Lawrence had been blocked up by the 
ice moving down from the north, and that thus a great 
inland fresh-water sea had been formed, over which icebergs 
floated.* I next, in 1874, applied the theory to explain the 
formation of the Steppes of Siberia, after I had crossed 
them and found that they contained fresh-water shells, and 
I further suggested that the ice that descended from the 
mountains of Scandinavia must have dammed back all the 
rivers of Northern Europe. t Mr. Croll informs us| that 
before this Prof. Geikie had suggested to him that if the 
Straits of Dover were not then cut through, or were they 
blocked up by land ice, <c say by the great Baltic glacier 
crossing over from Denmark, the consequence would be 
that the waters of the Rhine and Elbe would be dammed 
back and would inundate all the low-lying traCts of country 
to the south ; and this might account for the extraordinary 
extension of the loess in the basin of the Rhine, and in 
Belgium and the north of France.” 
Very soon after my paper was published in the “ Trans- 
actions of the Geological Society ” I became dissatisfied 
with the explanation I had offered, because I found that in 
Devonshire there were drift pebbles and transported boulders 
up to heights of 1200 feet above the sea, and also that the 
ice of Scandinavia had actually retired before the boulder 
clay was spread out ; as in Sweden, the latter overlies the 
glaciated rocks and the till left by the land ice. This is also 
the case on the north-east coast of England, as shown seve- 
ral years ago by Mr. Richard Howse. The Scandinavian 
drift of the coasts of Durham and Northumberland was de- 
posited after the land ice, that had scored and grooved the 
rocks below, had melted back, and when the land was 
* Trans. Nova Scotian Institute of Natural Science, 1866, p. 91. 
f Quart. Journ. Geol. Soc., 1874, p. 490. 
I Climate and Time, p. 452. 
