84

the lateral facies of the other. The few plant remains and invertebrate fossils found in the Haymond formation and the abundant fauna in the transported blocks of the boulder-bed member are definitely older than anything found in the Gaptank formation. Moreover, the Haymond formation possesses constant lithologic characters over the entire Marathon Basin, including those places where it underlies the Gaptank, and these characters are unlike those of the later formation.

A gap of unknown extent lies within the Haymond formation, for at no place is there a continuous succession from its base to its top. However, the boulderbed member east of Haymond station, which lies near the top of the first of the four sequences mentioned above, may be the same as that southeast of Gap Tank, which lies near the base of the third sequence. This boulder bed probably provides a link between the two areas.

The Gaptank formation in the third sequence contains many fossil zones, by means of which it can be correlated approximately with formations in the upper part of the central Texas and northern midcontinent sections. The fossil zones in the formation show that it embraces equivalents of several formations, and even groups, of the sections farther east. Under more favorable circumstances the formation would be broken into many subdivisions. Such a division could not, however, be attempted here because the formation is exposed over too small an area and its structure is too complex.

Pennsylvanian marine fossils are found in the disturbed strata of the fourth sequence west of Marathon. They indicate an age no greater than that of the Gaptank in the third sequence, and a few of them appear to be equivalent to those in its upper part. Baker's conception of the beds in the fourth sequence as a separate entity thus has little justification in fact, although they show some differences in lithology from the type Gaptank.

.-The sequence of Pennsylvanian rocks at Marathon presents many problems of correlation. Its great thickness and the predominance of clastic material make it unique among the sections in trans-Pecos Texas, and the scarcity of fossils in most parts of it makes a comparison with regions of similar stratigraphy difficult.

The fossils that have been found indicate that the section at Marathon represents the greater part of Pennsylvanian time. Its lower parts probably extend into the early Pottsville, and its upper parts are equivalent to the higher members of the succession in the better-known sections farther east. Both plant and invertebrate fossils indicate that the three lower formations, comprising by far the greater part of the whole section, are of Pottsville age. The upper divisions of the Pennsylvanian, comprising equivalents of the greater part of the northern midcontinent section, all appear to lie within the relatively small thickness of the Gaptank formation. The Pottsville equivalents are thus greatly expanded, whereas the post-Pottsville equivalents have a normal or even a reduced thickness.

.-Pennsylvanian rocks of the same facies as those in the Marathon Basin are exposed at only one other locality in transPecos Texas, the Solitario uplift. Here only the Tesnus formation is found. East of the Marathon Basin, however, such rocks have been penetrated by wells. Water wells a few hundred feet in depth not far east of the Marathon Basin encounter such rocks below the Cretaceous, and deep wells in the southeastern part of Terrell County and in Val Verde County, to the east, have entered sheared and talcose shales, probably also of Pennsylvanian age, after passing through a thick Cretaceous section.

In the northwestern part of trans-Pecos Texas, 100 miles or more from the Marathon Basin, the Pennsylvanian is exposed, but it is here of very different facies. In the Hueco and Franklin Mountains the Magdalena limestone, 1,500 feet thick, lies between rocks of Mississippian and Permian (?) age. This formation is the southward extension of the group of that name in New Mexico. Field observations by J. Brookes Knight indicate that there is an upward progression of faunas in the formation. The lower part in the Hueco Mountains is characterized by the primitive fusulinid genera Wedekindellina, Fusulinella and Fusulina, which are replaced upward by the more advanced genus Triticites. In the lower part there are abundant Chaetetes milleporaceus Milne-Edwards and Haime and Spirifer rockymontanus Marcou. The zone of Chonetes mesolobus Norwood and Pratten occurs near the middle of the section, and the upper part includes abundant fossils of upper Pennsylvanian aspect, including Enteletes near the top. The faunal succession in the Magdalena limestone of the Hueco Mountains is thus comparable to that in the Gaptank formation, and whether it includes equivalents to any of the older beds of the Marathon Basin section is doubtful.

Nearer to Marathon, in the Sierra Diablo, north of Van Horn, there are less continuous exposures of Pennsylvanian "

Sellards, E. H., Pre-Paleozoic and Paleozoic systems, in The geology of Texas, vol. 1, Stratigraphy: Texas Univ. Bull. 3232, p. 118, 1933.

Christner, D. D., and Wheeler, 0. C., The geology of Terrell County: Texas Univ. Bull. 1819, pp. 11-12,1918.

Sellards, E. H., op. cit., fig. 10, p. 128, and pp. 190-191.

King, P. B., and King, R. E., Stratigraphy of outcropping Carboniferous and Permian rocks of trans-Pecos Texas: Am. Assoc. Petroleum Geologists Bull., vol. 13, pp. 910-911, 1929. Most of the information in this paragraph is derived from field observations made later by J. Brookes Knight and the writer.

Identifications by C. O. Dunbar.