PETER SANDSTONE 
293 
the Roubidoux sands, a clearly marine formation, and therefore 
afford no proof. The same is true of the size and degree of 
uniformity of the sand-grains. 
The chert conglomerate at the base of the formation shows no 
evidence of wind-action. Bedding is more prominent than cross¬ 
bedding, and nothing like dune-structure is anywhere noted, 
even in the more protected valleys of the old erosion surface, 
where waves could not have worked it out. Marine fossils have 
been found in Arkansas in the basal Everton beds, the first de¬ 
posit above the old erosion surface, as well as in the main body 
of the typical Peter sandstone in Minnesota. 
The conditions of deposition of the formation, then, seem to be 
about as follows: After the post-Canadian [Beekmantown] 
emergence and erosion, the sea advanced from the south, spreading 
over the southern edge of the continental platform from the 
Ouachita geosyncline, from which it may never have completely 
retreated. The first advance found a low limestone land-mass 
with a mantle of ordinary residual-clay soil. This is partly 
worked over; but owing to the character of the relief, which was 
obviously not completely planed down by wave attack since it 
still exists, some of this residual clay was preserved in places. 
Across this area from the north flowing rivers delivered some 
sands from northern sources, but the earlier submergence at the 
south was accompanied by the formation of considerable lime¬ 
stone, representing the lower Simpson beds of Oklahoma, and, 
with the still further encroachment of the sea, the Everton beds 
of Arkansas and southern Missouri; the sand being locally and 
irregularly distributed. During this stage it is quite probable 
that along the valleys of the main streams emerging from the 
supply ground to the north there may have been local dune 
areas, just as there are now along the valleys of rivers generally 
that carry much sand. 
After the shore had progressed northward at least past the 
central region of Missouri, and possibly far up into Iowa and 
Illinois, closing the period of Everton deposition, there appears 
to have come a warping which caused the emergence of the Ozark 
region, allowing the erosion of the surface of the Everton lime¬ 
stones. 
It is entirely possible that this emergence may have brought 
