47

no evidence of it has been found in the outcrops in the region.

The transportation of blocks as much as 5 feet across from such an uplift to their present position is difficult to explain. It is true that streams at times carry boulders 10 feet in diameter, and streams may have carried the fragments now embedded in the Maravillas to the edge of the sea. However, the deposit itself seems to be far from any shore existing at that time. Large blocks might have been rafted from the shore by floating ice, although their association with corals and similar fossils would seem to argue against a climate sufficiently cold for this.

The boulders in the Woods Hollow shale are stratigraphically not far beneath those in the Maravillas chert, but they do not occur at the same localities and are of somewhat different character. They consist of limestone blocks of Cambrian and early Ordovician age, which are not like indigenous rocks of the same age in the Marathon region. They are not embedded in conglomerates but lie, with no associated smaller fragments, in shales and flaggy sandstones. As their source is obviously foreign and probably lay to the northwest of the Llanoria geosyncline, they must have been transported for longer distances than those of the Maravillas chert. Like those in the Maravillas, they might have been transported by floating ice.

Unlike the Maravillas boulders, however, they fulfill many of the conditions postulated by Van der Gracht for a tectonic origin. They occur in much crumpled and contorted shales. The shales lie between two competent series of beds and are known to have acted as a gliding plane during the folding and thrusting of the region. The boulders are derived from rocks of foreland facies, over which, at least in part, the Marathon geosynclinal rocks have probably been thrust. The boulders are found in the shale not far south of the outcrop of the Dugout Creek overthrust, which is perhaps the frontal thrust of the Marathon folded belt.

If the boulders are of tectonic origin they have been derived from Cambrian and Ordovician limestones of foreland facies, which lie beneath the thrust sheets of the region. Blocks of such underlying rocks may have been plucked off during Pennsylvanian time by thrust sheets advancing over them from the south and thus incorporated in incompetent layers in the overriding mass by intraformational shearing.

There is some evidence, though not proof, that the boulders are of true sedimentary origin rather than of tectonic origin. They are closely associated stratigraphically with the boulders of obvious sedimentary origin in the Maravillas chert. The blocks are all of rocks older than the strata of the Woods Hollow formation, although if their origin were tectonic they might be of much later age. Moreover, other incompetent layers in the section in this part of the Marathon Basin contain no boulders, though some of them are almost as well adapted for such a process of tectonic intercalation as the Woods Hollow shale.

The Caballos novaculite was named by Baker and Bowman for exposures on Horse Mountain, sometimes called Caballos Mountain, the highest summit of the folded beds of the Marathon Basin. The resistant white novaculites of the formation are the chief ridge makers in the area, and the structure of their outcrops is clearly revealed in the desert environment of sparse vegetation (pls. 1, A; 5, C, D; 6, A; 7). They also stand out conspicuously on aerial photographs.

The Caballos formation reaches a maximum measured thickness of 600 feet in the southern part of the area and is only 200 feet thick in the extreme northwestern part (pl. 2). The novaculite beds, which constitute a prominent part of the formation in the south, give place in the northwest to bedded chert, which has many shale partings and a few limestone intercalations in the Dugout Creek area. In the north a lower novaculite member is very prominent but southward this is subordinated to an upper novaculite. The chert and novaculite beds are divisible into five members-a lower chert member at the base, a lower novaculite member, a middle chert member, an upper novaculite member, and an upper chert member at the top. The members change in thickness from northwest to southeast across the area, so that various facies of the formation can be distinguished. These characterize northeastward-trending belts. The stratigraphy and facies of the formation are summarized in the following table:

Van der Gracht, W. A. J. M. van Waterschoot, Permo-Carboniferous orogeny in the south-central United States: K. Akad. Wetensch. Amsterdam Verh., Afd. Natuurk., deel 27, no. 3, pp. 58-61, 1931.

Baker, C. L. and Bowman, W. F., Geologic exploration of the southeastern Front Range in trans-Pecos Texas: Texas Univ. Bull. 1753, p. 93, 1917.