244 
STRATIGRAPHICAL GEOLOGY 
and allow elastic usage of old terms, until gradually without con¬ 
troversy, they sink into innocuous desuetude. 
Keyes. 
Derivation of the Peter Sandstone. Because of the remarkable 
purity of the sandrock, the small size of the components, and the 
peculiarly perfect rounding of the grains the Peter Sandstone 
is often regarded as having been formed by eolian means. Despite 
these features the formation is now demonstrated to have been 
laid down in the sea. Marine fossils are found abundantly in 
some localities. One place alone furnishes no less than fourteen 
genera, embracing 13 species of pelecypods, 7 species of gastero- 
pods, 3 cephalopods, 3 brachiopods, 1 bryozan and 1 sponge. 
Although in the north the entire Peter section is massive sand¬ 
stone, in the south there are important limestone beds intercalated. 
The history of the components of a sandstone is complex. It 
takes into account transportation successively by winds, streams 
and waves. Textural criteria alone afiford uncertain proofs of 
the exact nature of the transportation even to the latest deposit in 
which the sand-grains are found. Complexity of this history is 
further increased when the sand passes through several cycles 
of migration, from loose beds to solid rock and then through 
loose sediment to solid rock again. 
Compared with average marine sandstones, in content of clay, 
iron, mica, heavy minerals, and carbonates, the purity of the 
Peter sandstone does not depart sufficiently from that of associ¬ 
ated marine sandstones to demand any essentially different ex¬ 
planation of origin. The degree of purity actually existing, as 
determined by special chemical analyses, is satisfactorily accounted 
for by assuming a derivation from older already well-sorted sand¬ 
stones. 
After weighing the evidence on the various possible sources of 
the materials composing the Peter sandstone, the Pre-Cambrian 
occurrences of the Texas Llano region, of the Rocky Mountains, 
and of the Ouachita Mountains, all seem to be out of the question. 
This leaves only the Canadian Shield in the north to consider. 
At the time the Peter formation was laid down this land-mass 
embraced not only the Pre-Cambrian crystallines, but also broad 
belts of so-called Potsdam sandstones of Cambric age. This for- 
