190 Physical Geography of 



swamps, having become so shallow, through the enormous quantity of 

 glacier silt deposited in them, that they also may soon disappear under 

 the continually advancing masses of debris. 



The extent of these flats, and of the lakes in them, stand in almost 

 all cases in exact proportion to the extent of the present glaciers at 

 the end of the valley, and, therefore, of course to the height, extension, 

 and other orographical conditions of the alpine chains. The form and 

 width of the valleys above the alpine lakes show in the most striking 

 way that they must have once been the bed of great glaciers, to the 

 action of which they principally owe their present form. They are 

 frequently, even up to the present glaciers, of the same width as the 

 lakes. On both sides of them, several thousand feet above the level of 

 the valley, enormous moraines are found stretching along the moun- 

 tains, so that one can often follow the terminal moraine at the lower 

 end of the lake for twenty miles upwards. Hoches moutonnees occur 

 everywhere. However, where the colossal glaciers of the Ice Period 

 have pierced through to the Canterbury plains, the secondary ridges 

 are also rounded of£ and the valleys widened. 



In the valleys several miles wide, the present rivers flow in numbers 

 of branches, uniting and separating a hundred times, and changing 

 their bed after each great fresh. At the same time they are often so 

 straight, that one can see the glaciers from the lakes, as is the case, 

 for example, at Lake Pukaki, from which the Tasman glacier, and 

 Lake Takapo from which the Classen glacier can be seen. Indeed it 

 does not require much power of imagination to bring before one the 

 time when the glaciers were often fifty to sixty miles long and six to 

 ten miles wide. In one of my official reports " On the formation of the 

 Canterbury Plains," I have given a detailed description and explana- 

 tion of these interesting and important physical geological conditions. 



In order to understand the former occurrence of the great post- 

 pliocene glaciers, we have to assume no change in the climate, but 

 simply to consider the existence of the plateau-like mountain systems 

 towering above the snow-line, the enormous snow masses of which 

 might be sufficient to explain the existence and extension of the 

 glaciers. A comparison with the Dovrefield in JNTorway might perhaps 

 not be out of place here. I have also tried to show that the terrace 

 formation which we meet with everywhere, even in the valley of the most 

 unimportant streamlet, has not been brought about by an upheaval of 

 the land, but only and entirely by the retreating of the sources and 



