Ulrich—Paleozoic Systems in Wisconsin, 
103 
occur at appropriate horizons above it. In one of the best of the 
sections in Pennsylvania (at Beliefonte) the Eoubidoux fauna lies 
over 3,000 feet beneath the base of the overlying Ordovician forma¬ 
tions. Moreover, this section shows more than 1,000 feet of yet 
older Canadian limestones before reaching the top of the underlying 
Ozarkian system. 
Without facts like these one could scarcely believe that the break 
between the Shakopee and Oneota in Wisconsin represents time so 
long that it sufficed for the slow marine deposition elsewhere of not 
less than 4,500 feet of limestone and dolomite. Nor is this all, for 
even in the thickest of the Appalachian sections the physical evi¬ 
dence of the break between the Canadian and Ozarkian systems 
appears quite as distinct and impressive as it is in Minnesota and 
Wisconsin between the Shakopee and Oneota dolomites. 
Striking and conclusive evidence of subaerial erosion of the 
surface of the Ozarkian system during the interval between that 
and the succeeding Canadian period was procured recently in south¬ 
eastern Missouri. Namely, in traveling northward through this 
state from Van Buren to Potosi and thence westward toward Mera- 
mec River the Gasconade dolomite, which in Missouri lies at the 
top of the Ozarkian and is always easily recognized by its peculiar 
chert and highly characteristic fossils, w^as found to be generally 
rather thin and in places entirely absent. Indeed the areas in 
which the Gasconade cherts and fossils were not observed must 
aggregate at least 50 square miles. At first this local absence of 
the Gasconade seemed most probably due to nondeposition but 
further investigation, which showed an altogether unusual amount 
of quartz sand and chert pebbles in the overlying Roubidoux forma¬ 
tion, soon tended to change the first suggestion to the belief and 
finally to the conviction that the Gasconade after having been laid 
down over the entire region was then locally removed by erosion 
during the emergent interval between the two periods. The most 
convincing evidence favoring the latter view was the discovery of a 
ring or funnel of steeply inclined and highly fossiliferous Gasconade 
chert in the midst of horizontal beds of an older Ozarkian forma¬ 
tion. As this older formation extended nearly to the tops of the 
surrounding hills and was succeeded in the section by conglomeratic 
Roubidoux sandstone and chert without intervention of the Gas¬ 
conade, no other explanation of the mentioned funnel-shaped out¬ 
crop of Gasconade chert than that it is a remnant of the original 
