310 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
The larger Norwegian fiords may be instanced as glacial channels 
that present every appearance of having advanced far toward the 
mature stage of a cycle of glacial denudation from an initial or pre¬ 
glacial form not yet well understood. The variation of form 
between the main fiords and their branches gives some indication 
that the glacial work was accomplished in several successive epochs, 
with the interglacial epochs of normal river work between; but this 
is only a suggestion, needing much more field work before it can be 
assured. Not only the deep fiords, but the hanging valleys and the 
uplands also, have been ice-scoured; for hanging valleys frequently 
have a well-defined 17-section, and sometimes receive secondary 
hanging valleys from the enclosing uplands; and the streams of 
the uplands exhibit repeated departures from the forms of normal 
erosion. Although possessing little drift, the uplands frequently 
bear lakes of moderate depth and irregular outline; in spite of the 
breadth to which the upland valleys are opened between the sur¬ 
mounting hills and mountains, their streams frequently change from 
wandering at leisure in split or braided channels along broad floors, 
to dashing down in haste over rocky rapids : a behavior that is mani¬ 
festly inconsistent with that of the mature drainage of a normally 
denuded region. Even the surmounting hills exhibit strong scour¬ 
ing on their up-ice-stream side. It does not therefore seem permis¬ 
sible to conclude that the hanging valleys which open on the walls of 
the greater fiords have not been deepened by ice erosion because 
they escaped the more severe glaciation that scoured out the fiords 
themselves. All the valleys have been glaciated, and all have been 
significantly modified from their preglacial form. The discordance 
of overdeepened main fiord and hanging lateral valley seems to me 
best explained as the result of the mature development of glacial 
drainage, in which the chief trunks and the larger branches of the 
glacial systems had for the most part reached a graded condition. 
Trunk and branch glaciers would then have united at even grade 
as to their upper surface, and the trunk and branch channels would 
have had dimensions satisfactory to the ice currents which flowed 
through them, but the channel beds would have been discordant, as 
they are found to be. 
Review of Previous Writings on Hanging Valleys. 
It has already been stated that hanging side valleys and over¬ 
deepened main valleys have not yet been generally given the impor- 
