PETER SANDSTONE 
291 
absence of fossils is no more striking than that of the Rubidoux 
and Jefferson sandstone beds, but little more than that of associ¬ 
ated limestones, fossils being limited in all these formations almost 
wholly to the beds replaced by chert. The Peter sandstone shows 
a less ratio of cross-bedding to bedding than the Roubidoux for¬ 
mation. Indeed, there is no valid reason for assuming any dif¬ 
ferent source of origin for the sands of the Peter formation than 
for those of the older beds, which are undoubtedly marine. 
The Joachim dolomite, which rests with apparent conformity on 
the Peter sandstone, seems to occupy the same stratigraphic posi¬ 
tion as the top of the Peter formation farther north. Both lie 
next below beds believed to be the Lowville-Black River equiva¬ 
lents of the East. Sand-grains of the Peter type, and even con¬ 
siderable lenses of sandstone, are common in the Joachim for¬ 
mation. Fossils are very rare in the formation, and are reported 
from but one or two localities. This is probably significant in 
explaining the rarity of fossils in the Peter sandstone. The sur¬ 
face of the formation underwent notable erosion before the 
Bryant [Plattin] beds were laid down over it, and the Jasper beds 
of Arkansas seem to occupy this interval. The Joachim dolomite 
is not known to occur north of Calhoun County, Illinois, and is 
believed by some to be the deep-water equivalent of the Peter 
sandstone. 
■ Careful comparisons of the purity, degree of rounding, uni¬ 
formity, and size of grain of the Peter sandstone and the sand¬ 
stones in the section immediately beneath show conclusively that 
these beds do not differ in any appreciable particular, certainly 
not enough to demand any different explanation of origin. In 
fact, the suggestion is repeatedly forced upon the observer, that 
these sands were derived from a common source, and underwent 
practically the same history. 
None of these sands, of which the Roubidoux and the Peter 
sandstones are the most important, could possibly have been de¬ 
rived from within the area in which they now lie. To the east, 
the fact that the Roubidoux equivalent, the Beekmantown, and the 
Peter equivalent, the Chazy section, are limestones with little 
sand, seems to eliminate Appalachia as a source. In the Ozarks, 
only a very small pre-Cambrian area, if any at all, was exposed 
as late as the time when the Beekmantown beds were first laid 
