GeoJoyy of the St. Croix Dalles. — Berkeii. 379 
AdditioiKit Data and Discussion. Vertical or basaltic 
cracks and fissures are more numerous in some outcrops than 
in others. Outcrops west of the river and especially those in 
the immediate vicinity of Taylor's Falls, show this character 
in its highest development. Certain parts of the flows exhibit 
these joints to a more marked degree than others, the supposed 
upper portions of certain flows being singularly free from 
them. It is possible that the vicinity of the falls has suffered 
from additional disturbance, for jointing planes are especially- 
abundant at the Dalles accompanied by large fissures. One 
of these large fissures has exerted a powerful influence upon 
the course of the St. Croix river at this place. 
If one proceeds up the river toward the elbow without 
change of course so far as to the enclosing wall on the east 
side, it can be observed that there is an immense fissure in 
that wall extending without change of direction, but with 
converging sides back a considerable distance from the river. 
The vertical sharpl}^ cut walls of this fissure stand apart 
about 20 feet and widen in the direction of the river's course. 
At present the St. Croix turns almost a right angle at the el- 
bow in the Dalles and flows westerly in a straight line between 
walls of solid rock for a distance of almost half a mile, while 
a continuance of its original southerly course in a channel 
through the same kind of rock for only 20 rods would have 
accomplished the same purpose. Such freakishness can be 
explained onl^^ on the ground of some structural aid in the 
misdirection of the river. The columnar structure of the 
rocks has made it possible to preserve almost vertical cliffs 
throughout the history of this river. The heat of summer 
and the frosts of winter levy contributions from year to year 
for the accumulating piles of debris at the foot of the cliffs. 
Some of these blocks drop into the river whose maximum 
depth, now 100 feet, is gradually but certainly diminishing. 
In view of the erosion effect above the falls it is scarcely pos- 
sible to believe anj-^ such deptli attainable in this extremely 
hard dial)ase rock without the acceptance of one of the fol- 
lowing conditions : either the complete undermining of the 
igneous rock at its edge by the erosion of the underlying for- 
mation, or the occupancy of an original fissure line and plane 
of weakness by tlie river, making it necessary only to main- 
