86

unfavorable to abundant marine life in the Marathon region. The faunas of the exotic blocks in the Raymond formation have a much more direct relation to those of the Marble Falls and Wapanucka limestone and contain many identical or comparable species.

The Haymond formation probably corresponds to part of the second of the three main groups of clastic rocks in central Texas, southern Oklahoma, and western Arkansas (p. 85). The occurrence of fossiliferous blocks in the formation indicates clearly that it is younger than the Marble Falls limestone. The plants and fusulinid indigenous to the formation show, however, that it is still of lower Pennsylvanian age. As such, it may be correlative with the clastic rocks in the lower part of the Strawn group of central Texas and with the Atoka formation of Oklahoma and Arkansas. The occurrence of large boulders in part of the Haymond formation suggests the possibility that this part may be of the same age as the boulder-bearing beds in the Johns Valley shale of the Ouachita Mountains . Miser now considers these beds to be a little younger than the Wapanucka limestone, which would make them approximately correlative with the Raymond boulder-bearing beds.

The overlying Gap tank formation occupies a considerable span of later Pennsylvanian time and its num errors fossils permit approximate correlations with formations of other regions. The Gaptank contains many of the fossils cited by Moore as characteristic zone fossils of the upper part of the Pennsylvanian. The fossils indicate that the formation contains equivalents of the upper part of the Strawn group, the Canyon group, and perhaps a part of the Cisco group of central Texas. It is also equivalent to a considerable part of the Des Moines and Missouri groups of the northern midcontinent area. The conglomeratic lower part of the Gaptank formation may be related to the clastic rocks of the upper Strawn, which also contain a conglomerate, the Brazos River member. The more calcareous upper part of the Gaptank formation is similar to the Canyon group, in which there are many thick limestone beds.

-The boundary between the Pennsylvanian and Permian series in the south-central United States has been subject to much difference of opinion. The eventual solution of the problem will probably be made by a regional study of all the areas of exposure in this part of the country. Such a study, or even a review of the existing evidence, is outside the scope of this paper.

In the Marathon region the boundary is assumed to lie at the horizon of the great unconformity that separates tilted Carboniferous rocks of the Glass Mountains from greatly disturbed older Carboniferous rocks of the Marathon Basin. Paleontologically, this unconformity is supposed to lie between the Uddenites zone in the upper part of the Gaptank formation and the Pseudoschwagerina zone in the lower part of the Wolfcamp formation. An unconformity is also found below the Hueco limestone, which contains the Pseudoschwagerina zone in northwestern trans-Pecos Texas. In the Hueco Mountains the upper part of the Magdalena limestone, which underlies the Hueco, contains fossils of approximately the same age as those in the upper part of the Gaptank formation.

In the beds immediately below and above the boundary as now drawn in the Marathon region there is a gradual replacement of fossils of Pennsylvanian character by those of Permian character. Permian elements first appear in the brachiopods of the Uddenites zone. In the overlying Wolfcamp formation, where there are still many species of upper Pennsylvanian type, there are also some species that continue unchanged into higher beds of the Permian and other genera and species that appear to be ancestral to the characteristic Permian forms above. The faunas of the Leonard formation above the Wolfcamp are definitely of Guadalupian character, and only a few Pennsylvanian species persist. The faunal changes encountered on ascending the section in the Glass Mountains are probably in large part the result of progressive evolution. Whether they are related to changes in environment caused by the late Pennsylvanian and Party Permian diastrophism has not been proved.

Marked diastrophism took place in late Pennsylvanian time in the Arbuckle Mountains and the Wichita Mountain system of central Texas and southern Oklahoma. The Ouachita Mountains may also have been disturbed at the same time, but as no sedimentary rocks of this age are found near them it is not possible to date the later part of their movements. The late Pennsylvanian movements caused the deposition of the third group of clastic sediments found in that series. The disturbances in central Texas and southern Oklahoma probably correspond roughly with the great disturbance at the end of Gaptank time in the Marathon region. In those areas, however, instead of a single great unconformity, there are several lesser ones, and in some places where the unconformities are not large, strong movement in nearby areas is indicated by the character of the sedimentary rocks. This is particularly true in the Cisco group of central Texas. Recently Sellards has reported the occurrence of "

Miser, H. D., op. cit., p. 1007.

Moore, R. C., Correlation of Pennsylvanian formations of Texas and Oklahoma: Am. Assoc. Petroleum Geologists Bull., vol. 13, pp. 896-900, 1929.

The characteristic fossil of this zone has been generally known as Schwagerina, but as a result of recent taxonomic revisions by C. O. Dunbar that name is now applied to another genus, and the names Pseudoschwagerina and Paraschwagerina substituted for it. See Dunbar, C. O., and Skinner, J. W., Schwagerina versus Pseudoschwagerina and Paraschwagerina: Jour. Paleontology, vol. 10, pp. 83-91, 1936.

Sellards, E. H., op. cit., p. 172.