798 THE GEOGRAPHICAL SIGNIFICANCE OF FLOODS, 



Causes leading to jyresent statement. 



The writer was led to apply Gilbert's principle as the result of 

 a short visit to the Lakes and Sounds of South- Western New 

 Zealand in 1902-1903. During that trip a number of topographic 

 forms, totally unlike those of non-glaciated Australia, had been 

 observed. For a time no solution could be found to the problem. 

 At this juncture a glacial note(9,a) and a letter were received 

 from Prof. W. M. Davis. So simplystated were these notes that no 

 doubt existed in the writer's mind as to " glaciation of preglacial 

 valley contours " being the key to the problem. At a later date 

 the writer(l,a) was led to announce the " Flood H3^pothesis " to 

 account for the peculiar forms exhibited by the sounds and 

 associated lakes and canons. The stagnation of present-day 

 glaciers in Alaskan fiords; the overriding of glacial debris by the 

 same ice-masses, and the lack of corrasive power exhibited by 

 Alpine and other glaciers, all seemed explicable on this assump- 

 tion of former ice-flood action. 



Stream-studies were then entered upon in some little detail. 

 Along gutters, floods were observed to excavate holes alnaost 

 identical in shape with those of typical cirques, sound and lake- 

 basins, as also to glacial canons. Pronounced stagnation during 

 drought or interstorm (flood) periods was noted. From these 

 attention was directed to incipient cailon-contours, then to 

 brooklets, brooks and rivers. In all, similar and similarly 

 situated forms were observed. In every instance basins were 

 seen to have been excavated at points of marked convergence; 

 at points of weakness; and also other points of heavy thrust in 

 the large or small canons. The valley-basins were observed to 

 possess reversed grades downstream : in short, they were minia- 

 tures of the fiord- and lake-basins existing in regions of recent 

 intense glacial action. 



One was thus led, step by step, to reason from flood-eftects in 

 small channels such as roadside gutters, incipient caiions and 

 brooklets, to those in basins and spurless caiions of intensely 

 glaciated regions. In a word, knowing that certain gigantic and 

 remarkable "facts of form" — to wit, cirques, fiord- and associated 



