85 G The American Geologist, May, i897 
The Cave Creek Formation. — The development of the Cave Creek 
formation between southern Kansas and northern Texas is practically 
continuous. The data obtained during the April reconnaissance settled 
clearly the fact that the Shiiner gypsum is no mere local bed, but, as 
the highest member of the Cave Creek formation, is second only to the 
Medicine Lodge member and nearly coextensive with the latter within 
the limits of the formation as at present known. Thus the cave forma- 
tion is for the most part, as on Cave creek, the southern Kansas locality 
from which it takes its name, tripartite in character. It is magnifi- 
cently displayed in the great i-ange of bluffs and buttes which I have 
called the Marcy range, and which constitutes one of the most striking 
features of the plains topography, stretching almost uninterruptedly 
from southern Kansas, apparently at least, to the eastern foot-slope of 
the Llano Estacado, and of which I have elsewhere described the north- 
eastern portion. From Okeene, Oklahoma, this range of bluffs may be 
seen leaving the Cimarron river in the vicinity of Glass mountains and 
slowly retreating toward the North Canadian, which the gypsum reach- 
es near Darlington. In one of its canyon recesses, nestles the head of 
Salt creek. East of Watonga and north of the principal heads of 
Kingfisher creek, the escai'pment gives place to an acclivity of such a 
grade as to allow a somewhat ready passage for the Kingfisher- Waton- 
ga stage-road. With unimportant interruptions due to erosion and 
blanketing, the outci-op of this great gypsum group extends thence 
across the Wichita Indian reservation, Washita county, and the south- 
eastern part of Roger Mills county, and is thence continued across 
either the northern and western parts of Greer county, Oklahoma, and 
Collingsworth and Hall counties, Texas, or across Gypsum creek of Greer, 
and Groesbeck creek of Hardeman county (near Quanah); till, reaching 
the foot-slope of the Llano Estacoda, it pursues for an unknown dist- 
ance, a course roughly parallel with the eastern escarpment of the lat- 
ter.* The lowest member of the formation, the Medicine Lodge gypsum, 
whose western and southern limits are unknown, but which is probably 
the most extensive bed of gypsum in North America, is usually more 
conspicuous than the Shimer, owing partly to its greater thickness and 
partly to the fact that the presence of the Shimer gypsum, at a short 
remove above it, often continues the cliff-front up to the summit of the 
formation, so that the Medicine Lodge member is seen cut squarely off 
in the face of the bluff, while the Shimer is more frequently rounded off 
or reduced by erosion or concealed by detritus from the overlying Dog 
Creek formation. 
At the head of Salt creek in Blaine county, Oklahoma, the Cave Creek 
formation has a thickness of 70 or 80 feet, the Medicine Lodge and Shi- 
mer gypsum-beds being respectively 18 to 24 and 15 to 18 feet thick, and 
the interval of Jenkins clay nearly equal to the combined thickness of 
the gypsums. I was a little surprised to find that the gypsum caves and 
*On the April journey time failed iu which to ascertain whetlier wliat may be 
called the Quanah gypsum, or what may be called the Collingsworth gypsum forma- 
tion represents the Cave Creek formation in Texas. But it seems certain that the 
Quanah is a lower gypsum than the Collingsworth, 
