The chemie composition gives a hint as to the origin of the St. Louis 
cayes, and bears out the conclusions here presented. Prof. John RR. 
Proctor says* that in the vicinity of Mammoth Cave the subearboniferous 
limestone is “a massive, remarkably homogeneous rock with no interven- 
ing strata of shale or sandstone, conditions most favorable for the forma- 
tion of caverns.” In the main his statement is true of the equivalent 
strata in Indiana, but does not take into consideration certain beds of 
limestone that weather to a calcareous shale or the variable chemical 
structure of the Indiana stone, both important elements in studying the 
relations of the strata to the caves they bear. Probably more to the point 
is the statement of Prof. W. H. Wheeler,; who, in writing of the topog- 
raphy of St. Louis County, Missouri, says: “The limestones of the St. 
. Louis area are very hard, tough, and resist mechanical disintegration, but 
on account of the prevalent purity, they are very susceptible to chemical 
dissolution.” “If the upper portion of the limestone is impure, and es- 
pecially if high in magnesia, it is much more resistent to chemical dissolu- 
tion, and the sinkhole method of drainage is frequently absent. In this 
ease the drainage is by surface channels, which are abrupt and irregular 
and vary sharply from gentle to heavy slopes.’’ But, while it is conceded 
that homogeneity and purity largely determine whether the dissolution is 
chemical or mechanical, they dot not appear to fulfill all the required con- 
ditions. The Bedford odlitic and Paoli limestones, by chemical analyses, 
are shown to be from 95 to 98 per cent. calcium carbonate, and the Mitch- 
ell limestone less rich in lime by 10 per cent., yet the first two forma- 
tions have but few caves, while the last is undermined with cavernous 
openings. That the surface exposure of the Mitchell limestone contributes 
greatly to its disintegration has already been mentioned, but this does 
not explain its inherent susceptibility to chemical dissolution. If the 
number of analyses of the St. Louis limestones above the Bedford o6litic 
are not near so many as one would wish, those which are available seem 
to be suggestive. Dr. G. M. Levette, under direction of Prof. E. T. Cox, 
made a number of analyses of hydraulic cement rock from the lower Mitch- 
ell limestone strata of Harrison County, and as equivalent beds of ce- 
ment rock are found at Becks Mill, Clifty, and many other places, one 
of them is here given.* 
*The Century Magazine, March, 1898, p. 643. 
+ Keyes’ Mo. Geol. Sur. XI, p. 249. 
* Geol. Sur. Ind., 1878, p. 75. 
