296 
PETER SANDSTONE 
water than the Roubidoux sands, since no sun-cracks have ever 
been noted in the shaly layers in Missouri. Such cracks are 
very common on Roubidoux surfaces, in similar layers, and at 
many horizons. Their absence in the Peter formation, then, 
indicates not only that the materials were laid down under water, 
but that it probably was not subject to as frequent emergence 
as in the case of the Roubidoux beds. Farther north, and in closer 
proximity to the shore-line, it is more than likely that there were 
occasional emergences which brought local shoals within the 
range of wind-action for considerable periods of time, and this 
may well account for the fact that some samples of the Peter 
sandstone show a much greater degree of rounding than others, 
even when the effects of secondary enlargement are eliminated 
from consideration. 
The character of the land-mass from which the Peter sands 
were derived is a point of moment. This, as already shown, 
was in general to the north of the present belt of the Peter for¬ 
mation, and largely occupied the area commonly known as the 
pre-Cambrian, or Canadian Shield, extending for unknown dis¬ 
tances to the north, the northeast, and the northwest. 
The region was probably not arid. The rainfall of the Upper 
Mississippi Valley at the present time, is controlled very little by 
the matter of elevation, and over large areas practically not at 
all. The important control is the cyclonic storms of the prevailing 
westerly winds, drawing their chief moisture from the Gulf of 
Mexico. Since the area under consideration exhibited when the 
Peter sands were being deposited, a relation of land and water 
very similar to that now existing, there is certainly no reason to 
doubt adequate precipitation. The fact that the rock materials 
are very completely weathered also suggests at least moderate 
precipitation. Even the Potsdam sandstone derived from this 
land-mass carries very little undecomposed feldspar, except at 
its very base. 
Throughout Potsdam deposition, the land-mass probably never 
stood very high, and this is evidenced by the fact that weathering 
seems always to have kept well ahead of erosion, so that no coarse 
conglomerate material was developed, except in the immediate vi¬ 
cinity of old monadnocks like the Baraboo Range. The lack of 
feldspathic constituents in the resulting sediments also indicates 
