Dodge.] 268 [March 2.. 
v:illey until it has luulcrcut and removed all the deposits remain - 
ing upon the sides of the valley down to that Lase-level ? If it 
should do this work in every case it is obvious that there would 
never he at any one time more than one terrace in a river valley, 
and after a sufhciently long time there would be none at all, for 
all the land would be cut (h)wn to the level of the stream itself. 
This idea is well stated hy Prof. .lames Geikiei when he says, 
•'Should the stream continue to ilow'with the same volume and 
under the same conditions, the newer flats would eventually come 
to occupy as broad a space as that formerly covered by tlie older 
terraces, the latter in fact would be completely demolished." In 
order to account for two or more terraces in a river valley it is 
necessary to deteiinine why the width of the flood plain of the 
stream should become less and less as the valley is cut deeper and 
deeper into the deposits. In other words, why lateral cutting 
should decrease under these conditions. 
Previous study of the physics of river erosion has shown that 
with a land nearly reduced to base-level, the tendency of a river 
is to cut laterally rather than to deei)en its channel. Under the 
opposite conditions of a high or rising land, the tendency would 
be for the streams to cut down their channels deei)er and deeper 
until they had come as close to base-level as they could. Keejj- 
ing these facts in mind it becomes an easy matter to account for 
numerous terraces at any one place in so many of the streams of 
the northern United States and Canada, if we remember that, as 
far as we know, the land in these regions began to rise at the close 
of the glacial period and has been rising ever since. Consequently 
a stream of constant volume cutting down into the deposits 
previously dropped in the valley, would tend to leave many 
terraces, for each terrace plain would represent the temporary 
alluvial plain level of the stream. If the present geographical 
cycle continues until the stream has cut nearly to base-level, 
lateral wearing will undoubtedly be stronger than now, and the 
present terraces will be successively undercut and disaj)pear. 
No conditions that we know at i)resent, that do not allow the 
river to decrease the radius of its swings as it cuts downward, 
could form more than one te'rrace ; although one teriace might 
1 rrchistoiic Eun)[»t', [). -109. 
