E. Desor on the Drift of Lake Superior. 103 
_ Ina geological point of view, the terraces are the most import- 
ant, since they afford direct evidence of the changes of level 
which have occurred since the deposition of the drift. There 
can be no doubt that, wherever terraces of stratified materials 
are found above each other, the waters have once stood at so 
many levels. It might be, and indeed it has been, inferred from 
this, that when terraces occur along a shore, they ought to be 
found everywhere of the same size; and hence, that when their 
level is irregular, it is a proof that the upheaval was not uniform: 
This view, although correct in principle, is, however, apt to lead 
'0 mistakes when appligd without discrimination. 
he error in this respect arises from the fact that terraces have 
been too often mistaken for, or confounded with, mere beaches. 
It'should not be lost sight of, that terraces and bluffs are the re= 
sult of the undermining action of the waves. Their size and 
shape must therefore be determined by the force of this agency. 
If a basin of water is so situated as to have one of its shores ex- 
Posed to the full force of gales, while the opposite shore is shel- 
tered by high lands, we may easily conceive of a subsidence of 
the waters from a higher to a lower level, without at all-altering 
the slope of the coast. In the mean time, the right shore, not 
being Protected, will be so acted upon as to occasion a succession 
Tra Again, the destructive action may be so effective in 
ertain places as to wash off, in the course of time, even the ter- 
faces of former levels, and to leave only a single bluff. 
There are many places along the lake shore where the peculiar 
Pe of the terraces and their diversity are to be ascribed to 
sch a process. The 9 
» ) %#, and. that, 5 
While Stationary, the upper bluff was formed. Afterwards, the 
vel sank, and another bluff was formed at , and st 
later, another at o. The subsidence of the water must have 
been iMtermittent—the epochs of subsidence, which are indicated 
by the between the terraces, being followed by intervals of 
par »during which the terraces were formed. Bat this regular 
ton of terraces does not extend far. It is limited to @ 
Posed. nmost he kind of materials of which the bluff is com 
Povedl "Thos, bias of cicy see steeper wlan Livi of loam, and binfts bf loait 
steeper ag luffs of clay are ste than oO app when several 
such of or gravel. The term 
ce," Seen one above, ne other, so as to app 
teal hij. : e in having a double 
