34 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
ledges defend and protect the higher terraces. Only a small portion 
of the flood plains, however, is left back of such defended cusps. 
There may extend down-stream a long trail of the same flood plain 
to mark its more widespread extent; and the trail only exists there 
because of the up-stream and possible down-stream cusp which has 
prevented the river from swinging more widely and so consuming the 
entire terrace. The extent of such terraces from the point of protec¬ 
tion will vary from, a few feet to several hundred. If the ledge dis¬ 
covered lies near the apex of a down-sweeping meander, the meander 
may succeed in slipping past it by withdrawing slightly toward the 
center of its valley. The obstacle is thus avoided, and the river 
impinges deeply on the down-valley side. An example of this type is 
seen where the West River slips by the rock on its right bank at the 
pumping station, and is now sweeping broadly to the right. By 
slipping past this and other rock barriers the river succeeded formerly 
in making the entire sweep of the meadow (plate 1). 
To the west of the meadow and south of the farm buildings there 
is a strong rock control. At three succeeding stages, when the meander 
reached this point, the rock ledge was discovered below the apex of the 
meander. In such a position the meander could not slip by, but was 
forced to bend outward and cut a curved re-entrant in the terrace 
front next up-stream from the ledge. The meanders farther up-stream 
continued their advance down-stream, which compressed the meander 
next to the ledge to a relatively strong curvature. This sharpened 
the defended cusp and caused it to point up-stream. The features 
which represent this control are the rock ledges to the south, the 
sharp bend or the deep concavity of the terraces to the north of the 
defining ledge, and the sharp cusp of the ledge pointing up-stream. 
The area of the West River just described is of special interest since 
it shows that the form and outline of the terrace front have there at least 
been wholly controlled by the definite rock ledges and the accidental 
position which they held with reference to the meanders of the river. 
Three times it happened that the defining ledge was on the tangent 
between two meanders or at the point of river inflection; and the last 
time as the river was sweeping around the meadow, the defining ledge 
chanced to be near the apex of the meander. In the first instance the 
stream could not slip past, whereas at another time, slipping past, it 
did its greatest work below the defining ledge. Examples of rock 
control on the left side of the valley have been cited. It frequently 
