82 SURFACE GEOLOGY. 
near East Concord, the river has changed its entire width from south- 
west to north-east, and a third of a mile to the east it has changed more 
than its width in an opposite direction. On the east side of the “Fan” 
or broad interval opposite the north part of the city, the river flowed in 
1804 by a very circuitous route 460 rods, which was shortened to 150 
rods by great freshets in 1826, 1828, and 1831, which cut a direct course 
across two peninsulas then known as Sugar Ball point and Hale’s point. 
Ponds already mentioned occupy portions of the old channel. Ten years 
later, Dr. Prescott reported the rapid undermining of Sugar Ball bluff, 125 
feet high, of which the river had carried away, between 1853 and 1863, a 
mass 80 rods long and 40 rods wide. . This erosion is still going forward, 
being aided by springs near the foot of the bluff. At Davis’s bluff, about 
a mile south, a width of three rods was swept off in 1863 in three days. 
Erosion at this point has continued thirty years, requiring a dwelling near 
the edge of the bluff to be several times moved, and the road changed. 
The same undermining of the high plains by the river is also going on 
at several places north and south of Fisherville. One mile south-east 
from Boscawen bridge, the plain, 110 feet above the river, is fast wear- 
ing away, and portions of it 10 feet wide and 150 feet long had fallen 
in 1875 10, 20, and 40 feet, remaining nearly level, so that their sapling 
pines, 10 to 30 feet high, were still upright and growing on the side of 
the steep sand-bluff. These would be carried away, to be followed by 
new slides during the next high flood. One mile farther south, and at 
other points below Fisherville, a similar rapid erosion was observed. A 
quarter of a mile north-east from Fisherville bridge, a bluff, which has 
been so recently undermined that it is not yet grassed, is now separated 
from the river by a wide area which does not exceed five feet above the 
ordinary height of water. These recent incursions of the river upon the 
plains, and the rapid changes in its channel upon the intervals, washing 
away yearly from one bank and adding to the side opposite, leave no 
doubt that the river has flowed at the foot of the bluffs along their whole 
extent, occasionally making a deep excavation beyond its ordinary bounds, 
as on the east side south of Sugar Ball bluff; that the high plain once 
filled the whole valley; and that the river has swept many times from 
side to side over the space occupied by its lower terraces and interval. 
Important changes in the channel of the Merrimack have also been 
