Problems, Middle Mississippi River Region, Pleistocene Time 183 
Illinois quadrangle, further emphasizes the fact that the terrace 
fragments found in the tributaries of the Mississippi are deposits 
derived from the master stream rather than from the tributary 
(Fig. 6). Both Rock Creek and Glaize Creek have rather 
straight courses with terrace fragments along them with eleva- 
tions which correlate with the Festus Terrace. The next creek 
south, of approximately the same size, is Sandy Creek. Its 
trough has a straight course parallel to those of Rock and Glaize 
creeks, with the mouth of the valléy at Riverside. Examina- 
tion will show that in the lower mile of the valley there is no 
permanent stream, because at Jerseydale, nearly three miles west 
of the natural mouth of the valley at Riverside, Sandy Creek 
passes through a water-gap and escapes southward into the valley 
of Joachim Creek. The reason is not difficult to find. The 
terrace at Riverside is unus:ally broad and high. It completely 
blocked the course of Sandy Creek, causing it to turn aside 
through its present water-gap. Such a terrace could hardly have 
been constructed by Sandy Creek; it must have been made by 
the Mississippi. The reversal of Sandy Creek by the damming 
of its mouth can be seen on a map or in the field in the course 
of an unnamed tributary to Sandy Creek, which flows west from 
Pevely to Jerseydale along the same course that Sandy Creek 
previously used in flowing eastward. But a somewhat similar 
deflection of a stream-course, due to damming of the mouth of 
the stream by deposits from the Mississippi, can be seen on 
Hubbel Creek. Before discussing this instance, however, men- 
tion must be made of the condition of the terraces south of 
Wittenberg. 
Between Wittenberg and Hubbel Creek, ten miles southwest 
of Cape Girardeau, a set of terraces similar in all respects, 
xcept elevation and slope, to the Festus Terrace may be traced 
on both sides of the Mississippi. The tops of these terraces have 
a lower elevation than the Festus Terrace would be expected 
to have and their slope is also less. But otherwise they resemble 
the Festus Terrace closely, so closely that it does not seem proper 
to separate them from the Festus Terrace. It would seem more 
Probable that the terraces above and below Wittenberg were syn- 
