386 RESEARCH IN CHINA. 
especially prominent in those varieties (No. 40) which have an ocher- 
colored matrix—the ocher being produced by the oxidation of the iron 
carbonates. Quartz grains and bits of carbon may also be found in several 
of the specimens. 
Wherever the pebbles exhibit corroded surfaces, as mentioned above, 
the cavities are filled with relatively coarse crystals of the cement. Siderite 
is especially apt to occur in such situations, probably in consequence of 
reactions between the calcareous cement and the ferruginous material con- 
centrated in the peripheries of the pebbles. 
Origin of the rocks.—Several hypotheses have been proposed to account 
for the peculiar structure of these rocks and of others similar to them 
which have been found in the United States. It has been suggested (a) 
that the nodules are autoclastic fragments produced by differential move- 
ments in thin-bedded, shaly limestones; (b) that they are concretionary 
bodies which have been formed im situ during or after the building of the 
rock; (c) that they are fossil organisms, or detached pieces or aggregations 
of them; (d) that they are pebbles derived from the breaking up of older 
limestones and cemented together by calcareous material; and (e) that they 
represent lenses of incipient solidification in the calcareous mud, this mud 
being deposited in shallow water and subject to partial disturbance and 
rearrangement by waves and currents. 
When all of the facts are considered it is evident that the first two of 
these hypotheses will not stand. The fragments which compose an auto- 
clastic, or “‘ reibungs-breccia,’’ are mutually similar, and the strata in which 
they occur show evidence of having been strongly deformed. In these 
conglomerates, however, the nodules are conspicuously rounded, usually 
having a definite position with reference to the stratification of the rock, 
and occur frequently in the strata which have suffered no considerable 
distortion. Moreover, it is inconceivable that autoclastic rocks in thin 
continuous layers should recur at approximately the same horizons over 
a territory hundreds of miles in extent. 
It is equally evident that the bodies are not of a concretionary nature. 
They have been produced neither by the filling of cavities nor by the aggre- 
gation of material around nuclei, both of which processes tend to develop 
concentric structures.* On the other hand they are in all respects similar 

* The local cementation of loose particles by CaCO3 sometimes forms concretions in clays. ‘This 
type of concretion often lacks the concentric structure, and the original stratification may not be dis- 
turbed. (Sheldon, J. M. Arms: Concretions from the Champlain Clays of the Connecticut Valley.) 
The production of such bodies seems to be dependent upon the circulation of ground-water in rocks elevated 
above the sea. The conglomeratic limestones were formed beneath the sea, as indicated by marine fossils 
in the matrix, and, therefore, seem to have had a different origin. 
