0a

 

UNITED STATES DEPARTMENT OF THE INTERIOR
Harold L. Ickes, Secretary
GEOLOGICAL SURVEY
W. C. Mendenhall, Director

GEOLOGY OF THE
MARATHON REGION, TEXAS

BY

PHILIP B. KING

UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON
: 1937
For sale by the Superintendent of Documents, Washington, D. C. Price $2.50 (paper cover)

 

2

 

3

CONTENTS

• Page
• Abstract 1
• Introduction 2
  • Location 2
  • Previous work 2
  • Field work 3
  • Acknowledgments 3
• Geography 3
  • Physical features of trans-Pecos Texas 3
    • Basin and Range province 4
    • Mexican Highlands province 4
    • Great Plains province 5
  • Climate 5
  • Vegetation 5
  • Erosional agencies 6
  • Marathon region 7
    • General features 7
    • Escarpments bordering the Marathon Basin 8
      • Escarpments and plateaus on the east and south sides 8
      • Escarpments on the north side 10
      • Escarpments on the west side 11
      • Relation of escarpments bordering the Marathon Basin to the later tectonic movements 12
    • Ridges of the Marathon Basin 13
      • Novaculite ridges 13
      • Dimple limestone ridges 13
      • Even summit levels on the ridges 14
    • Lowlands of the Marathon Basin 14
      • Rock floors, their nature and origin 14
      • Comparison of rock floors previously described with those in the Marathon Basin 15
      • Rock floors of the Marathon Basin 16
      • Erosion and deposition on the rock floors 16
      • Streams of the Marathon Basin 19
• Stratigraphy 19
  • General outline 19
  • Pre-Cambrian rocks 21
    • Pre-Cambrian rocks north and south of Marathon 21
    • Fragments of crystalline rocks in the Paleozoic sediments at Marathon 21
    • Depth of the pre-Cambrian floor in the Marathon Basin 22
  • Cambrian system 22
    • Dagger Flat sandstone 22
      • General features 22
      • Local features 22
        • Dagger Flat area 22
        • Threemile Hill 23
        • Woods Hollow Tank 23
        • Marathon anticlinorium 23
      • Microscopic character 23
      • Fossils and age 23
      • Stratigraphic relations 24
    • Problem of the Brewster formation 24
  • Ordovician system 25
    • Historical summary 25
    • General features 25
    • Marathon limestone 26
      • General features 26
      • Local features 26
        • Marathon anticlinorium 26
        • Dagger Flat anticlinorium 29
      • Fossils and age 30
      • Stratigraphic relations 30

 

4

• Stratigraphy-Continued.
  • Ordovician system-Continued page
    • Alsate shale 31
      • General features 31
      • Local features 31
        • Marathon anticlinorium 31
        • Dagger Flat anticlinorium 31
        • Relation of Marathon and Dagger Flat areas 31
      • Microscopic character 31
      • Fossils and age 32
      • Stratigraphic relations 32
    • Fort Peña formation 32
      • General features 32
      • Local features 32
        • Marathon anticlinorium 32
        • Dagger Flat anticlinorium 33
        • Jones ranch area 33
      • Microscopic character 33
      • Fossils and age 33
      • Stratigraphic relations 34
    • Woods Hollow shale 34
      • General features 34
      • Local features 34
        • Woods Hollow Mountains 34
        • Simpson Springs and East Bourland Mountains 35
        • Other localities 35
      • Microscopic character 35
      • Fossils and age 35
      • Stratigraphic relations 36
    • Maravillas chert 36
      • General features 36
      • Local features 37
        • Marathon anticlinorium 37
        • Monument Spring and Rock House Gap 38
        • Dagger Flat anticlinorium 39
        • Southeastern exposures 39
      • Microscopic and chemical character 39
      • Fossils and age 41
      • Stratigraphic relations 42
    • General problems of Ordovician stratigraphy 42
      • Correlations and regional relations 42
        • General correlations 42
        • Correlations in trans-Pecos Texas 43
        • Correlations with central Texas 43
        • Correlations with Oklahoma and Arkansas 43
      • Conditions of deposition 43
        • Faunal facies of the Marathon Ordovician 43
        • Source of the sediments 44
        • Depth of water during deposition 45
        • Origin of the chert beds 46
        • Origin of the boulder beds 46
  • Devonian (?) system 47
    • Caballos novaculite 47
      • General features 47
      • Local features 48
        • Northwestern exposures (facies 1) 48
        • Marathon anticlinorium (facies 2) 49
        • Dagger Flat anticlinorium (facies 3 and 4) 50
        • Southeastern exposures 51
      • Microscopic and chemical character 51
        • Novaculite 51
        • Banded chert 51
        • Sandstone 52
      • Fossils and age 52
      • Stratigraphic relations 52

 

5

• Stratigraphy-Continued.
  • Devonian (?) system-Continued. Page
    • General problems of Devonian (?) stratigraphy 53
      • Correlations and regional relations 53
        • Correlations in trans-Pecos Texas 53
        • Correlations with Oklahoma and Arkansas 53
      • Origin of the Caballos novaculite 53
        • Definition of novaculite 53
        • Theories of origin 54
        • Stratigraphic arrangement of the chert and novaculite members 54
        • Significant features in the novaculite 54
        • Significant features in the banded cherts 55
        • Conclusions 55
  • Carboniferous system 55
    • Pennsylvanian series 55
      • Tesnus formation 55
        • General features 55
        • Local features 56
          • Tesnus and Haymond area 56
          • Southeastern part of Marathon Basin 57
          • Pea Colorada synclinorium 57
          • Dugout Creek area 60
        • Microscopic character 60
        • Fossils and age 61
        • Stratigraphic relations 62
      • Dimple limestone 62
        • General features 62
        • Local features 62
          • Eastern part of Marathon Basin 62
          • Western part of Marathon Basin 63
        • Fossils and age 64
        • Stratigraphic relations 64
      • Haymond formation 64
        • Historical summary 64
        • General features 65
        • Haymond area 65
          • General relations 65
          • Lower members of formation 65
          • Boulder-bed member 66
          • Upper member of formation 68
        • Other areas of Haymond formation 68
          • Gap Tank area 68
          • Exposures west of Marathon 68
        • Microscopic character 69
          • Arkosic sandstone 69
          • Thin-bedded sandstone and shale 70
          • Boulder-bearing mudstone 70
        • Fossils and age 71
          • Indigenous fossils 71
          • Fossils of the exotic blocks 72
        • Stratigraphic relations 73
      • Gaptank formation 73
        • Historical summary 73
        • General features 73
        • Local features 74
          • Exposures near Gap Tank 74
          • Exposures east of Gap Tank 75
          • Area west of Marathon 75
          • Black Peak 76
        • Fossils and age 76
          • Faunas of the Gap Tank and Wolf Camp area 76
          • Correlation of the Gaptank strata at Gap Tank and Wolf Camp 80
          • Faunas of the area west of Marathon 80
          • Relation of the strata west of Marathon to the type section 82
        • Stratigraphic relations 82

 

6

• Stratigraphy-Continued.
  • Carboniferous system-Continued.
    • Pennsylvanian series-Continued. Page
      • General problems of Pennsylvanian stratigraphy 83
        • Stratigraphic sequence 83
        • Correlations and regional relations 84
          • General relations 84
          • Correlations in trans-Pecos Texas 84
          • Correlations with central Texas, southern Oklahoma, and Arkansas 85
          • Pennsylvanian-Permian boundary 86
        • Conditions of deposition of the Pennsylvanian series 87
          • Older formations 87
          • Younger formations 88
        • Problems of the Haymond formation 88
          • General features 88
          • Origin of the thin-bedded sandstones and shales 88
          • Origin of the arkoses 89
          • Origin of the boulder-bed member 89
      • Permian series 92
        • Historical summary 92
        • General features 93
        • Wolfcamp formation 94
          • General features 94
          • Local features 94
            • Wolf Camp area 94
            • Exposures in Monument Spring quadrangle 95
          • Fossils and age 95
            • Fossils of the type section 95
            • Age of the Wolfcamp formation in the type section 96
            • Fossils and age of the Wolfcamp formation in the Monument Spring quadrangle 97
          • Stratigraphic relations 97
        • Leonard formation 98
          • Historical summary 98
          • General features 98
          • Exposures in the Monument Spring quadrangle 98
          • Fossils and age 99
            • General features 99
            • Long-ranging fossils in the formation 99
            • Characteristic lower Permian fossils of the formation 99
            • A possible lower Leonard faunule 100
            • Faunal facies of the Leonard formation 100
            • Correlation of the Leonard formation 101
          • Stratigraphic relations 101
        • Word formation 101
          • General features 101
          • Exposures in the Monument Spring quadrangle 102
          • Fossils and age 102
            • General character of the fauna 102
            • Faunas of different parts of the formation 103
          • Stratigraphic relations 104
        • Capitan limestone 104
          • Origin of name 104
          • General features 105
          • Exposures in the Monument Spring quadrangle 105
          • Fossils and age 105
            • Fossils of the Altuda shaly member 105
            • Fossils of the Vidrio massive member 105
            • Fossils of the Gilliam thin-bedded member 105
            • Correlation of the Capitan limestone 106
          • Stratigraphic relations 106
        • Tessey limestone 106
        • General problems of Permian stratigraphy 106
          • Regional relations of the Glass Mountains section 106
          • Sedimentation of the Permian series in the Glass Mountains 107
          • Sources of the clastic sediments in the Permian 109

 

7

• Stratigraphy-Continued. Page
  • Triassic (?) system 109
    • Bissett conglomerate 109
      • General features 109
      • Fossils and age 110
      • Stratigraphic relations 110
  • Post-Bissett and pre-Cretaceous time interval 110
  • Cretaceous system 110
    • General features 110
    • Lower Cretaceous (Comanche) series 112
      • Trinity group Rose formation 112
        • Glen Rose formation 112
          • General features 112
          • Local features 112
            • East side of Marathon Basin 112
            • South side of Marathon Basin 112
            • West side of Marathon Basin 113
        • Maxon sandstone 113
          • General features 113
          • Local features 114
          • Age of the formation 114
      • Fredericksburg group 114
        • Walnut and Comanche Peak formations 114
          • General features 114
          • Local features 114
        • Edwards limestone 115
      • Washita group 115
        • Georgetown limestone 115
        • Del Rio shale and Buda limestone 115
    • Upper Cretaceous (Gulf) series 116
      • Eagle Ford formation 116
      • Possible higher beds 116
  • Quaternary system 116
    • General features 116
    • Gravel deposits 116
    • Alluvium 117
  • Igneous rocks 117
    • Volcanic rocks 117
    • Intrusive rocks 117
• Structural geology 118
  • General features 118
  • Marathon Basin 119
    • General features 119
    • Marathon anticlinorium (including Dugout Creek area) 120
      • General features 120
      • Exposures of the Dugout Creek overthrust 120
      • Structure of the overridden rocks 121
      • Faults of the overriding block 121
      • Folds of the overriding block 122
    • Peña Colorada synclinorium 122
      • General features 122
      • Synclinal area south of Monument Spring 123
      • Simpson Springs and East Bourland Mountain 123
      • Woods Hollow Mountains 123
    • Dagger Flat anticlinorium 123
      • General features 123
      • Structural features of the older rocks 124
      • Folding in the younger rocks 124
      • Folded overthrusts in the younger rocks 124
      • Tear faults in the younger rocks 128
      • Ridges southeast of the anticlinorium 128
    • Synclinorial area between Tesnus and Gap Tank 128
      • General features 128
      • Folds 129
      • Faults 129

 

8

• Structural geology-Continued. Page
  • Marathon Basin-Continued. 130
    • Southeastern part of Marathon Basin
      • General features 130
      • Hells Half Acre fault 130
      • Structural features south of Hells Half Acre fault 130
    • General structural problems in the Marathon Basin 131
      • Theoretical problems of the overthrust faults 131
      • Theoretical problems of the steep thrust faults 132
      • Theoretical problems of the tear faults 132
      • Theoretical problems of the folds 132
    • Age of the deformation 134
    • Regional relations of the structural features at Marathon 136
  • Structural features of the Glass Mountains 138
  • Post-Cretaceous structural features 138
    • Eastern margin of the Marathon dome 138
    • Western margin of the Marathon dome 138
    • Age of the Marathon dome 139
    • Regional relations of the post-Cretaceous structural features 140
• Economic geology 141
  • Ore deposits 141
  • Novaculite 142
  • Building stone 142
  • Water supply 142
  • Oil and gas 142
• Index 145

ILLUSTRATIONS

• Page
• PLATE
  • 1. A. View looking southwest from summit of Horse Mountain; B, Housetop Mountain from south 6
  • 2. Correlated stratigraphic sections of pre-Carboniferous rocks of the Marathon Basin 22
  • 3. A, Upper member of Marathon limestone; B, Monument Spring dolomite member of Marathon limestone; C, Woods Hollow shale; D, Upper cherts, limestones, and shales of Fort Pea formation 22
  • 4. A, Vertical layers of Caballos novaculite; B, Basal conglomerate of Maravillas chert at Rock House Gap; C, Maravillas chert at Rock House Gap; D, Detail of lower novaculite member of Caballos novaculite 22
  • 5. A, Gap on novaculite ridge south of old Fort Pea Colorada; B, Maravillas chert northeast of gap shown in A; C, Novaculite ridge southeast of Hackberry Tank; D, View southwest from Hackberry Tank; E, Axial anticlinal valley in Woods Hollow shale at southeast end of Simpson Springs Mountain 22
  • 6. A, East Bourland Mountain from southeast, from northeast end of Simpson Springs Mountain; B, East Bourland Mountain from northeast; C, West Bourland Mountain, looking northeast 22
  • 7. A, B, Novaculite ridges in Lightning Hills; C, Folding in novaculite on east side of Maravillas Creek at Maravillas Gap 22
  • 8. Correlated stratigraphic sections of Pennsylvanian rocks of the Marathon Basin 54
  • 9. A, Massive sandstone of upper part of Tesnus formation; B, Sandstones and shales of lower part of Haymond formation; C, D, Shales and sandy shales of upper part of Haymond formation 62
  • 10. Geologic map of boulder beds of Haymond formation in eastern part of Marathon uplift 62
  • 11. A, Block of novaculite in boulder-bed member of Haymond formation; B, Waterworn cobbles of Dimple and Gaptank limestones in first conglomerate member of Gaptank formation; C, Exposures of boulder-bed member; D, Rounded boulder of brecciated chert from Caballos formation in boulder-bed member; E,Boulder of novaculite in boulder-bed member; F, Boulder of Pennsylvanian limestone 62
  • 12. Correlated stratigraphic sections of Comanche series on the east and west sides of the Marathon Basin 110
  • 13. A, Black Peak in the Del Norte Mountains, from the northeast; B, Del Norte Gap from summit of Del Norte Mountains to the north; C, Black Peak from the south; D, North end of Housetop Mountain 110
  • 14. A, Limestones of Glen Rose formation intruded by igneous rock; B, Limestones of Glen Rose formation of the Maravillas scarp, dipping south off Marathon uplift 110
  • 15. Structural map of Marathon region 118
  • 16. Map and structure sections of Dugout Creek area In pocket
  • 17. Aerial photograph of northeastern part of Dagger Flat anticlinorium 118
  • 18. Aerial photograph of area of Carboniferous rocks south of Haymond station 118
  • 19. Block diagrams showing folded overthrusts of Dagger Flat anticlinorium 126
  • 20. Hypothetical block diagrams showing progressive development of structural features in Marathon region 126
  • 21. Structure sections of Monument Spring and Marathon quadrangles In pocket
  • 22. Map of trans-Pecos Texas showing principal structural features 134

   

  9

  • 23. General geologic map and sections of Marathon uplift In pocket
  • 24. Geologic map of Monument Spring and Marathon quadrangles In pocket
• FIGURE
  • 1. Index map of trans-Pecos Texas showing location of Marathon region 2
  • 2. Views of the Marathon Basin from the escarpments on the east and west 9
  • 3. Sections of the Cretaceous escarpments on the east and south sides of the Marathon Basin 10
  • 4. Map showing deposits of gravel on the uplands east of the Marathon region 10
  • 5. Block diagrams of the escarpments on the west side of the Marathon Basin 12
  • 6. Block diagrams of three anticlinal mountains in the Marathon Basin, showing forms displayed during the course of their erosion 14
  • 7. Block diagram of the alluvial fan of Antelope Creek, at the base of the Del Norte Mountains 17
  • 8. Section across northeast part of Dagger Flat area showing the four erosion surfaces typical of the MarathonBasin 17
  • 9. Maps showing possible stream history in the south half of the Marathon Basin 18
  • 10. Section through Threemile Hill, showing relation of Dagger Flat sandstone to younger beds 23
  • 11. Sketches showing structural features in the Marathon limestone 27
  • 12. Section in bed of Alsate Creek showing sequence and structure of part of the Ordovician rocks 29
  • 13. Sketches showing structural features in the Maravillas chert 37
  • 14. Section across novaculite ridge at picnic grounds south of old Fort Pea Colorada 38
  • 15. Thin sections of pre-Carboniferous cherts and novaculites 40
  • 16. Hypothetical block diagram showing probable geography of the Marathon region in Ordovician time 44
  • 17. Sketch of ripple marks in Caballos novaculite 49
  • 18. Sections of lower part of Tesnus formation along Rough Creek, in southeastern part of Marathon Basin 58
  • 19. Section across middle part of Devils Backbone, showing structure and relations of white quartzite beds in Tesnus formation 58
  • 20. Sections between East and West Bourland Mountains, showing structural features in Tesnus formation 59
  • 21. Structure section across southern part of Payne Hills 60
  • 22. Sketch showing structural features of the Haymond beds in a bank of San Francisco Creek 67
  • 23. Structure sections of boulder bed of Haymond formation 67
  • 24. Sections showing stratigraphy and structure of the Haymond beds at the fossil-plant locality on Dugout Creek 69
  • 25. Thin section of the mudstone matrix of the boulder bed 71
  • 26. Plan of part of outcrop of boulder bed, showing how large limestone boulders might be interpreted as parts of a single stratum of limestone 90
  • 27. Hypothetical section between exposures of boulder-bed member of Haymond formation and Hells Half Acre fault, showing possible source of boulders 91
  • 28. Stratigraphic diagram of Permian rocks in the Glass Mountains 93
  • 29. Map of trans-Pecos Texas showing structural features of Permian time 108
  • 30. Map of trans-Pecos Texas showing progressive northeastward overlap of Comanche series 111
  • 31. Restoration of structural features in the younger rocks of the Dagger Flat anticlinorium 126
  • 32. Sections of Caballos novaculite in Warwick Hills, showing change in thickness of members and its relation to the Warwick overthrust 133
  • 33. Diagram showing estimates of crustal shortening in different parts of the Marathon Basin 133

 

10

 

1

GEOLOGY OF THE MARATHON REGION, TEXAS

By PHILIP B. KING

ABSTRACT

This report describes the geology of the Marathon region, in trans-Pecos Texas. The Marathon region lies on the edge of the Mexican Highlands province, where that province merges into the Great Plains on the east. Structurally, the region is a broad dome of Cretaceous rocks, from whose central part the Cretaceous cover has been stripped away, leaving an area of low country in the center, the Marathon Basin. Here strongly folded Paleozoic rocks are exposed. The Monument Spring and Marathon quadrangles, described in detail in this report, extend across the basin area.

The Paleozoic rocks exposed in the basin and in the Glass Mountains, which flank it on the northwest, have a thickness of 21,000 feet. The greater part of them were laid down in a subsiding area, the Llanoria geosyncline. The oldest rocks are Upper Cambrian sandstones and shales, whose base is not exposed. Overlying them are 2,000 feet of Ordovician rocks, composed of shaly limestone and shale, with some beds of chert, whose chief fossils are graptolites. The Ordovician is overlain by the Caballos novaculite, possibly of Devonian age, which reaches 600 feet in thickness. This white siliceous rock is the chief ridge maker in the Marathon Basin.

The Caballos novaculite is overlain by a great series of clastic rocks of Pennsylvanian age, as much as 1200 feet thick in the southeastern part of the area but much thinner in the northwest. Two of the lower formations are a mass of arkosic sandstone and shale and are separated by a widespread thinner limestone formation. The two formations contain few fossils other than land plants. The upper of the two contains a remarkable layer of mudstone, in which are embedded large blocks of older rocks. The blocks are believed to have been derived from the erosion of advancing thrust sheets and to have marked the fist strong uplift in the region; they may have been transported to their present positions either by glaciers or by mud streams. The uppermost Pennsylvanian formation consists of conglomerate and sandstone derived from the erosion of rising folds and contains abundant upper Pennsylvanian marine fossils.

The strong deformation to which the Paleozoic rocks of the Marathon Basin have been subjected apparently culminated E after the deposition of this uppermost formation of Pennsylvanian age. The Permian rocks of the Glass Mountains, to the northwest, rest, at least in places, with great angular unconformity on the disturbed older beds. The structural features seen in the basin consist of close folds, trending northeast and overturned to the northwest which are broken by numerous thrust faults. The faulting culminated on the northwest in the nearly flat-lying Dugout Creek overthrust, with a known displacement of more than 6 miles. Farther southeast are other great thrusts, also with miles of displacement, some of which are folded and therefore older than the frontal fault. The folds of the Marathon region are a part of a system of structural features formed from the rocks of the Llanoria geosyncline, which extends northeastward in sinuous courses to the Ouachita Mountains of Oklahoma and Arkansas. Northwest of the geosyncline and folds of the Marathon region, during Paleozoic time, there was a foreland area, which was gently folded at the same time as the movements at Marathon. Southeast of them was a region underlain by pre-Cambrian crystalline rocks. Both these areas are now mostly concealed by Cretaceous and Tertiary rocks.

The Permian rocks of the Glass Mountains, 5,000 feet or more thick, consist of limestones, siliceous shales, clay shales, and sandstones, which interfinger in a most complex manner. The most striking stratigraphic features of the series as exposed in the mountains are limestone reefs, constructed in large part by limesecreting organisms. The reefs apparently had a marked influence on the development of the other lithologic facies. The Permian rocks contain marine fossils, in places very abundantly. Most of the faunas are similar to the Guadalupian fauna originally described by Girty from northern trans-Pecos Texas. The Permian rocks are tilted to the northwest, away from the Marathon Basin, and are apparently in greater part younger than the folds in the basin.

The Cretaceous rocks that surround the Marathon Basin have a maximum thickness of about 1,200 feet and are mostly limestones. They were laid down on the eroded edges of the folded or tilted Paleozoic rocks, whose surface had been reduced to a peneplain during Triassic and Jurassic time. Over the Cretaceous west of the Marathon region lie lavas and tuffs of early Tertiary age. Within the region small masses of igneous rock, in part of alkalic composition, have intruded the Paleozoic and Cretaceous rocks.

The Cretaceous rocks dip gently away from the Marathon dome on its north, east, and south sides. On the west side they are sharply buckled and locally overthrust toward the west. The structural features on the west side of the Marathon dome are in part older and in part younger than the early Tertiary lavas. All the rocks of the dome are broken by normal faults that are younger than post-Cretaceous folds and probably of later Tertiary age.

No rocks younger than the Tertiary igneous rocks exist in the vicinity of the Marathon region except gravel deposits that cover part of the lowlands. These were deposited on various surfaces of erosion. The oldest stands several hundred feet above the present streams, and the gravel on it is probably of Pleistocene age.

The rocks of the Marathon region contain relatively few materials of economic value. Locally there are some metallic minerals, chiefly near the igneous intrusions. The hard siliceous novaculites of the Marathon Basin may be of use for whetstones or road metal. The jointed bedrock of the basin and its cover of gravel contain a supply of underground water. The area does not seem to be favorable for the accumulation of oil or gas.

 

2

INTRODUCTION

Location

-This report deals with the geology of the Marathon region, which lies in the northern part of Brewster County, in western Texas. Particular attention is paid to the geologic features exposed in the Monument Spring and Marathon quadrangles, which extend across the central part of the region and cover an area more than 30 miles east and west by 20 miles north and south. In order to complete the description of the Marathon region, some of the stratigraphic and structural features to the north and south of the two quadrangles are also noted.

The region is crossed from east to west by the main line of the Sunset Route of the Southern Pacific Railroad, upon which, about halfway across the area, is the village of Marathon, the only settlement. (See fig. 1.)

Previous work

-The Marathon region was mentioned in 1890 by Von Streeruwitz, who noted northeastward trending ridges south of Marathon composed of "quartz and quartzite, strongly metamorphosed limestone, and semifused siliceous conglomerations." He mistakenly correlated these with the Cretaceous rocks cropping out to the east and west, which he thought had here been "fused and thrown up by protrusive volcanic rocks."

A good description of the geologic features of the area was given by Hill in 1900. After noting the character "

---

Von Streeruwitz, W. H., Report on the geology and mineral resources of transPecos Texas: Texas Geol. Survey 2d Ann. Rept., for 1890, p. 686, 1891.

Hill, R. T., Physical geography of the Texas region: U. S. Geol. Survey Top. Atlas, folio 3, p. 4, 1900.

 

3

of the Comanche or Glass Mountains, he described the Caballos Ridges, to the south. These were said to be "low ridges * * * rising from the floor of the Marathon plain * * * composed of the degraded vertical edges of Paleozoic limestone, shales, and cherts * * * trending northeast and southwest. The cherts are often white in color and form the backbone of long ridges." On the south and east he found the region to be bordered by scarps and cuestas of "subhorizontal Cretaceous limestone unconformably resting on the subvertical edges of the Paleozoic rocks." He concluded that "the Caballos and Glass Mountains are exposures of ancient post-Paleozoic structure of Appalachian type and age, which have been revealed by the erosion of the Cretaceous sediments that probably once embedded them."

Observations made by Udden in 1905 in the course of a journey to the Chisos country amplified the results of Hill. Udden noted the discovery of Ordovician and Carboniferous fossils in the Marathon region. The comprehensive work in the region by Baker and Bowman in 1915, as a part of their exploration of the southern front ranges of trans-Pecos Texas, revealed the broad outlines of the physiography, stratigraphy, and structure. Rocks of Cambrian age, Ordovician strata with fossils at four horizons, a probable Devonian formation, and four thick Pennsylvanian formations were recognized.

During the same period Udden and Böse studied the Glass Mountains, to the north of the Marathon Basin, and described the upper part of the Paleozoic section exposed there, which includes a great thickness of Permian strata.

The writer began work in the area in 1925. During this and two succeeding summers, associated with R. E. King, he studied the Glass Mountains and the adjoining part of the Marathon Basin for the Texas Bureau of Economic Geology.

Field work:

-The investigation that has provided the results for the present report was an extension of the earlier studies with R. E. King into an area adjoining on the south. The report is based on 8 months' field work in the region in 1929 and 1930 under the auspices of the United States Geological Survey. The season of 1929 was devoted to a reconnaissance of the entire Marathon Basin. In 1930 the Monument Spring and Marathon quadrangles were mapped in detail. Further observations were made during short visits in 1931.

At the beginning of the present investigation only the broader features of the structure and stratigraphy of the region were known. Because of the extreme complexity and small scale of the folds, the wide areas of valley fill, and the confusion that had arisen on some of the stratigraphic problems, many of the minor features of the region were poorly understood. The geologic mapping was therefore done with considerable care. Fortunately, excellent topographic maps were available for most of the area and could be used as a base for plotting geologic observations. The mapping was done partly by an elaborate system of pacing traverses, and partly by recording the observations on enlargements of the topographic sheets. Stratigraphic sections were measured mostly by Brunton compass and by pacing. Where there were exceptionally good exposures, however, a tape measure was used.

Acknowledgments

-A small part of the field observations on which this report is based were taken from notes made by the writer when he was connected with the Texas Bureau of Economic Geology and from data furnished by E. H. Sellards, C. L. Baker, and others. Mr. Baker has also made available many recent fossil collections from the older rocks of the region, for study by paleontologists of the Geological Survey, and has joined the writer on several field conferences that made it possible to correlate the present work with that of Baker and Bowman in 1917. Much information on the fossil plants of the region has been afforded by collections sent to the Geological Survey by Sidney Powers. The report has also been materially improved by visits to the field with various members of the staff of the Survey, including David White, G. H. Girty, Edwin Kirk, G. R. Mansfield, and H. D. Miser.

After this report had been written the writer had the privilege of examining some excellent aerial mosaic maps of part of the Marathon Basin made by the Edgar Tobin Aerial Surveys, of San Antonio, Tex. Through the courtesy of this organization, he has been permitted to reproduce two of the single photographs from which the maps were made. (See pls. 17, 18.)

GEOGRAPHY

PHYSICAL FEATURES OF TRANS-PECOS TEXAS

The Marathon region is in trans-Pecos Texas, the westward-projecting part of the State that lies along the Rio Grande west of the Pecos River (fig. 1). Geographically, this arid and mountainous region is "

---

Udden, J. A., A sketch of the geology of the Chisos country, Brewster County, Tex.: Texas Univ. Bull. 93, pp. 18-21, 76-78, 1907.

Baker, C. L., and Bowman, W. F., Geologic exploration of the southeastern Front Range of trans-Pecos Texas: Texas Univ. Bull. 1753, pp. 67-172, 1917.

Udden, J. A., Notes on the geology of the Glass Mountains: Texas Univ. Bull. 1753, pp. 5-59, 1917.

Böse, Emil, The Permo-Carboniferous ammonoids of the Glass Mountains, West Texas, and their stratigraphical significance: Texas Univ. Bull. 1752, 1917.

King, P. B., The geology of the Glass Mountains, part 1, Descriptive geology: Texas Univ. Bull. 3038, 1931. Ping, R. E., The geology of the Glass Mountains, part 2, Faunal summary and correlation of the Permian formations, with description of the Brachiopoda: Texas Univ. Bull. 3042, 1931.

For a description and classification of the physical features of the trans-Pecos region, based on somewhat different criteria, see Carter, W. T., and others, Soil survey (reconnaissance) of the trans-Pecos area, Texas: U. S. Dept. Agr., Bur. Chemistry and Soils, ser. 1928, no. 35, pp. 1-7, 1928.

 

4

more closely related to Mexico and New Mexico than it is to the rest of Texas. It is a region of rugged sierras, of high plateaus and broad cuestas, and of gently sloping intermontane plains. The mountains have no timber except in sheltered valleys and on the higher summits.

In the clear air of the desert the mountain masses loom with sharp outlines and clear detail from a distance of many miles, and the plains that surround them are deceptively foreshortened. More than half of the region is a lowland. These intermontane areas are either bolsons (structural depressions filled by mountain waste) or destructional plains that slope upward as pediments toward the mountain masses from which they have been carved.

Ephemeral streams, which are dry gravel beds most of the year, discharge from the mountains and flow across the plains. Some of these drain into bolsons with no outlet to the sea, such as the Salt Basin, in the northwestern part of trans-Pecos Texas (fig. 1). Most of the drainage channels, however, lead to the two master streams of the area, the Rio Grande and its major tributary, the Pecos River. Their waters flow to the Gulf of Mexico. The Rio Grande is noteworthy more for its persistence through long stretches of desert land than for its breadth or volume. In its southeastward course across the area the river traverses a succession of desert basins and passes from one to the next through separating mountain barriers in which it has cut narrow and imposing canyons.

The mountains, plains, and plateaus of trans-Pecos Texas have been formed by interaction between various crustal movements of post-Mesozoic age and by the forces of erosion working upon the disturbed crust. The forms thus produced are of varying character, and the area may be divided into several geomorphic and structural provinces.

Basin and Range province

-North of the Texas & Pacific Railway the mountain areas are broad and in part plateaulike, with one side presenting a steep escarpment and the other forming a gentle back slope descending from the crest. Between them are intermontane plains 5 to 15 miles across, whose margins rise as bajada slopes toward the mountains. These mountains are composed of rocks which have been very little folded but which have been broken in the later part of Cenozoic time into numerous fault blocks (pl. 22). Movement along the faults has served to outline the form of the mountain areas, and this form has been modified but slightly by subsequent erosion. The intermontane plains are mostly depressed areas filled by waste carried down from the mountains.

The mountains and desert plains of this part of trans-Pecos Texas resemble those in the adjacent part of central New Mexico, to the north, near the Rio Grande. They are also similar to those in the typical basin and range country farther west, and this part of transPecos Texas is included in the Basin and Range province.

Mexican Highlands province

-South of the Texas & Pacific Railway block mountains of basin and range type are well developed in only a few areas. The mountains and plains are not caused directly by the uplift or depression of blocks of the earth's crust, but mostly by the differential erosion of bedrocks of varied character. The sedimentary rocks and the lava flows have been tilted, flexed, and in places strongly folded by crustal movements older than the block faulting of the basin and range province. In many places there are also masses of intrusive igneous rock. The nonresistant rocks of this region have been worn down into valleys and plains, and such harder rocks as limestones, thick lava flows, and igneous intrusions have been left as ridges, plateaus, and peaks. This area is the northern edge of a region of rugged highlands, whose greatest extent is in Mexico, south of the Rio Grande. It is here termed the "Mexican Highlands province."

The western part of the Mexican Highlands province in trans-Pecos Texas, comprising the Quitman and Eagle Mountains (fig. 1), consists of narrow parallel ridges and mountain chains of resistant, steeply tilted limestone and sandstone, between which are longitudinal lowlands carved from less resistant strata. In places the lowlands are covered to a moderate depth by later Cenozoic lake beds and alluvial deposits, but on the whole they seem to have been formed by erosion rather than by downfaulting or downwarping of the earth's crust. Similar parallel mountain ranges and intermontane lowlands are present southwest of the Quitman and Eagle Mountains, on the Mexican side of the Rio Grande, where they extend from the vicinity of El Paso southeastward past the great bend of the river and on into the interior of the State of Chihuahua (pl. 22, fig. 1). The two mountain ranges in Texas and the similar ranges to the southwest and south of them form the north end of the Sierra Madre Oriental of Mexico.

The eastern part of the Mexican Highlands province, comprising the eastern border ranges, is of greater diversity. Toward the north are the Davis Mountains (fig. 1), a high plateau broken up by canyons and in its more eroded parts separated into mesas, ridges, and isolated peaks. The Davis Mountains are carved from flat-lying or gently flexed lava flows. South of the Davis Mountains are various irregular mountain groups, some of them consisting of sharp peaks, and others of plateaulike blocks or narrow ridges. Between the mountains are lowland areas, some of which are smooth, gently sloping plains, whereas others have been greatly dissected and in part form picturesque

 

5

badlands. The most conspicuous mountain group in the area, the Chisos Mountains, is a group of sharp peaks that stand in the center of a dissected lowland and are composed of masses of intrusive igneous rock and remnants of lava flows.

Southeast of the Davis Mountains and east of the Chisos Mountains are the narrow ridge of the Santiago Mountains and the high broken mountain mass of the Sierra del Carmen. These trend southeast and extend beyond the Rio Grande into Mexico. Northward the Santiago Mountains die out near the line of the Southern Pacific Railroad. The Santiago and Carmen Mountains are composed of folded, resistant limestones. East of the folds of these mountains are several domical uplifts (pl. 22). One of these, expressed topographically as the Serrania del Burro, lies wholly in Mexico, with its northern edges reaching up to the Rio Grande. It is a high dissected plateau, for the limestone cover of the dome is complete over its crest. Farther northwest, on the Texas side of the Rio Grande, is the Marathon dome. Here the limestone cover has been stripped from an extensive area on the dome's crest, and a lowland, the Marathon Basin, has been excavated from the nonresistant underlying beds.

Great Plains province

-East of the Marathon Basin are escarpments of limestone which form the west edge of an extensive plateau area. The plateau summits descend gently eastward from the flanks of the Marathon region and the Serrania del Burro. The 50- to 75-mile belt between the Marathon region and the Pecos River on the east consists wholly of such plateau country, which has been carved into a maze of canyons and low tablelands. The plateau region is the western edge of the Edwards Plateau section of the Great Plains province, which extends far eastward into central Texas.

CLIMATE

Trans-Pecos Texas has an arid or semiarid climate. The average annual rainfall at Marathon and nearby stations is about 17 inches. However, this figure is the average of greatly varying observations of many years, and the amount of rainfall is erratic in both extent and time. Some spots may receive half a dozen rains within a year, whereas others may remain nearly rainless for several years. The yearly rainfall at Fort Stockton, not far north of Marathon, has been as slight as 4 inches and as great as 34 inches, although its average is 15 inches.

One-half or more of the year's rainfall comes during the summer, when most of it is of torrential character. The precipitation during any one of such rains may amount to several inches. Now and then during this time there may be one or more weeks of continuous rain, when the mountains are cloaked in clouds. The entire rainfall of a year may be produced by only a few storms. During the winter some snow falls in the mountains, but as this is the dry part of the year, the amount of such precipitation is not great.

Temperatures at Marathon range from 110° in the summer to below zero in the winter, but ordinarily the variation is not so large. In the summer the diurnal temperature range is as much as 50°. Winds are strongest in the spring, when violent gales, without rain, may persist for a week or more. Violent wind storms of short duration at times accompany the summer thunder-showers.

VEGETATION

The Marathon region and surrounding parts of trans-Pecos Texas have a vegetation adapted to the semiarid climate. The smooth plains of the Marathon Basin are grass-grown, but in the low places, where ground water is nearest to the surface, there are expanses of creosote bushes (Covillea) and dense thickets of mesquite (Prosopis juliflora) and catclaw (Acacia greggi). Low rocky ledges in the plains and terraces of limestone gravel that fringe the mountains support clumps of sotol (Basylirion wheeleri), lechuguilla (Agave lecheguilla), and other yuccas. Prickly pear or nopal (Opuntia) and ocotillo (Fouquieria splendens) grow on the low foothills. Higher in the mountains a sparse growth of juniper and pinon spreads over the exposed surfaces and summits and gathers in groves on the northern shaded slopes. Small clusters of live oak and manzanita (Arctostaphylos pungens) grow in protected valleys. Near water holes and stretches of flowing water in the stream channels is a lush growth of reeds and alders, shaded by cottonwood trees. The giant cacti that characterize the Sonoran Desert farther west are lacking in trans-Pecos Texas, but otherwise there is much similarity in the vegetation of the two regions.

The region is most attractive in the spring when numerous small plants come into blossom, covering the hillsides with a mat of brightly colored flowers. After the summer rains also, the brown and barren hills turn green as the vegetation comes to new life. Some of the plants, by reason of desert adaptation, show an immediate and wonderful rejuvenation after these unexpected downpours. The leafy clumps of resurrection plants (Selaginella pringlei ?), matted over many of the limestone surfaces, are dry and brown most of the year, but unfold and turn green within an hour after a rain.

"

---

For a useful discussion of the vegetation of this part of Texas see Bray, W. L., The vegetation of the sotol country in Texas: Texas Univ. Bull. 60, 1905. A concise summary is also given in Carter, W. T., and others, op. cit., pp. 7-11.

 

6

EROSIONAL AGENCIES

Agencies observed in the Marathon region.-The sedimentary rocks of the Marathon region, especially the greatly deformed strata of Paleozoic age, have been prepared for weathering by previous jointing. Where they are least fractured and have the fewest bedding planes they stand as bold cliffs and hogbacks. The dominant rocks of the region, which are limestones of various sorts, weather chiefly by solution, despite the low humidity. Solution widens joints and fractures, makes channels, pits, and shallow depressions on the exposed surfaces, and undermines ledges. In places, some of the granular limestones of Pennsylvanian age and crystalline dolomites of Permian age show a well- developed exfoliation of undetermined origin. The older Paleozoic cherts and novaculites are not affected greatly by either solution or exfoliation, but in most places they break down readily along closely spaced joints. Rock breakage by diurnal temperature change does not appear to be an important agent of weathering in any of the rocks of the region.

The fractured blocks of limestone and chert are loosened from their parent ledges by frost action. Gravity and rain wash help carry them down the steep slopes below the outcrops. Many rock masses may also be broken from the cliffs by lightning, for scars apparently produced by its impact may be seen on the faces of the steeper bluffs at many places in the area.

Because of the dry climate, there is but a thin cover of vegetation and, in comparison with humid regions, a small amount of rock decay. The soils on the hillsides and mountain slopes are therefore thin and are full of angular rock fragments. Rock ledges are abundant on the slopes, except in the shaly formations, and even here gullies only a few feet deep lay bare the underlying strata. Talus is generally lacking. In the plains the surficial material forms a thicker cover over the bedrock, but most of this has been washed in from the surrounding hillsides.

The surfaces of all the mountains, hills, and plains are covered with a network of watercourses, ranging from small gullies to broad, gravel-covered creek channels. This strongly suggests that the dominant erosional agent of the region is running water. In the semiarid climate, however, the work of water is spasmodic, and the drainage channels are dry most of the year. After rains the water runs rapidly down the mountain slopes, discharging into rocky gorges in the mountains or directly onto the plains. There is so little vegetation and soil that not much rain water is absorbed where it falls. The drainage channels leading away from the storm area become rushing turbid rivers, with the flood waters at times advancing down the hitherto dry channel like a wall. Sometimes the writer has heard these torrents emit a rumbling sound, doubtless from the impact of boulders against each other while in movement. Nearly all the erosion accomplished by the streams of the region takes place during the flood periods. Banks are deeply undercut at the stream bends, depressions are hollowed out in the channels, and gravel bars are shifted downstream. The undercutting is a phase of lateral corrasion, which, according to Blackwelder, is "geologically * * * rapid, apparently more so than most other processes in the desert."

On the level plains the flood waters may spread far beyond the insignificant swales on the surface and flow down the slope as a mass of interlacing rivulets, or even as sheet floods a few inches to a few feet deep and several miles in width. In the path of sheet floods the writer has seen on the steeper slopes closely spaced shallow gullies and on the gentler slopes small heaps of sticks, rubbish, and fine mud. Erosion and deposition of this sort, accomplished by sheet floods, appears to be of minor consequence, and in the Marathon region at least sheet floods are not the important agent of erosion that McGee suggested.

The flood waters eventually disappear into the gravel channels of the streams or in the alluvium of the plains. Very little of the run-off reaches the Pecos River or the Rio Grande by surface flow. However, there is much continuous underflow within the gravel beds of the channels. In the larger creeks there are stretches of permanently flowing water where the underflow is raised to the surface by sills of bedrock.

During years of normal climate the wind is not an important agent in the erosion of the Marathon region. The most striking wind storms are the great gusts that precede summer thundershowers. These carry great quantities of dust into the air and sometimes even across the lower mountain ridges, but they are local in extent and erratic in direction. In dry years wind storms may occupy several weeks of the spring and may carry much suspended matter into the air. During the exceptionally dry winter and spring of 1933-34 such dust storms were more prominent than usual in trans-Pecos Texas. Many storms were observed by "

---

Udden, J. A., Etched potholes: Texas Univ. Bull. 2509, 1925.

Blackwelder, Eliot, Exfoliation as a phase of rock weathering: Jour. Geology, vol. 33, pp. 793-806, 1925; Insolation hypothesis of rock weathering: Am. Jour. Sci., 5th ser., vol. 26, pp 97-113, 1933.

Blackwelder, Eliot, Talus slopes in the Basin Range province [abstract]: Geol. Soc. America Proc., 1934, p. 317, 1934.

The same criterion has been used by Bryan, Kirk, Wind erosion near Lees Ferry, Ariz.: Am. Jour. Sci., 5th ser., vol. 6, pp. 303-305, 1923.

Shuler, E. W., A rise down canyon [Davis Mountains]: Sci. Monthly, vol. 31, pp. 129-133, 1930.

Blackwelder, Eliot, Desert plains: Jour. Geology, vol. 39, p. 138, 1931.

McGee, W J, Sheetflood erosion: Geol. Soc. America Bull., vol. 8, pp. 87-112, 1897.

Baker, C. L., and Bowman, W. F., Geologic exploration of the southeastern front range of trans-Pecos Texas: Texas Univ. Bull. 1753, p. 163, 1917.

 

 

6b

 

7

the writer during this period in the Guadalupe Mountains, northwest of the Marathon region. Most of then were accompanied by sharp changes in temperature some were associated with strong winds, but in other: the movement of air currents did not exceed 10 mile,. an hour. In all of them there was a persistent movement of air in some one direction, so that transportation of dust from one region to another apparently tool place on a grand scale.

Such dust storms are exceptional under present climatic conditions, but they suggest that a relatively slight decrease in annual rainfall might permit muck fine material to be carried from the region by deflation and that this process may have been active during dry times in the past.

The only features that can be definitely attributed to wind work are the "charcos" found in low places on the plains, generally in areas of fine-grained alluvium. They are steep-banked circular or oval depressions, 10 to 25 feet in diameter, and 2 or 3 feet deep. They are probably caused by cattle in search of water trampling a wet place after a rain. The vegetation is thus destroyed, leaving mud exposed, and during dry seasons this is carried away as dust by the wind. More conspicuous features ascribable to wind work, such as sand-blasted rocks and sand dunes, are entirely lacking.

Comparison of erosional agencies in and and humid regions -In regions of arid climate, as a general rule, mechanical weathering dominates over chemical weathering, and on the steeper slopes rill wash is more effective than soil creep. Because of the lack of creep, steep mountain slopes tend to endure in the unconsumed areas of such regions until well along in the cycle of erosion, whereas in regions of humid climate they have at this stage changed to subdued forms. In arid regions the streams of both mountains and plains areas are intermittent rather than permanent, but because of the lack of vegetation the subdivision of the more steeply sloping areas into watercourses is much more minute than in humid regions. When the streams flow the material carried by them is larger in amount and coarser in texture than that of streams in humid regions. The profile of such streams, even at grade, is therefore relatively steep.

The erosional agencies of arid regions, being unlike those of humid regions, produce unlike land forms. Davis notes that "the rocky and boulder-clad slopes of maturely dissected mountains in arid regions, together with the barren pediments below them", contrast strongly with "the soil-cloaked and forested slopes of maturely dissected mountains in humid regions, together with the fertile valley floors below them." He believes, however, that "their unlikeness is rather a matter of degree than of kind" and that "the unlike features are really homologous."

According to Davis, a very baffling problem is concerned with the relative rates of erosion and degradation in humid and arid regions. It seems as if humid stream erosion must * * * be more rapid than arid stream erosion in the early stages of an erosion cycle; also that, in a much later stage, degradation may be more rapid on the bare slopes of an arid region than on the plant-covered slopes of humid regions.

MARATHON REGION

GENERAL FEATURES

The Marathon Basin, on the crest of the Marathon dome, is 30 miles wide and 40 miles long and consists of plains, hilly lowlands, and low mountain ridges, carved from folded Paleozoic strata. The basin is surrounded by limestone escarpments, which stand higher than any of the ridges in the basin. On the east, south, and west sides the limestones of the escarpments are of Cretaceous age and are mostly gently tilted away from the uplift. On the north the Paleozoic rocks beneath the Cretaceous contain a resistant mass of limestone and form the broad cuesta- like upland of the Glass Mountains.

The Cretaceous rocks now found on the escarpments bordering the Marathon Basin at one time extended entirely over the crest of the Marathon dome. They have been stripped off the higher parts of the dome by recession of their cliffs and by the excavation of the weak underlying Paleozoic beds to form the Marathon Basin. These processes are still going on.

The northern part of the Marathon region slopes northward and northeastward toward the Pecos River, but the greater part slopes southward and is drained by Maravillas, San Francisco, and smaller creeks, which flow into the Rio Grande (fig. 9, B). The maximum relief in the Monument Spring and Marathon quadrangles is 2,700 feet. The lowest point, 3,450 feet above sea level, is on San Francisco Creek where it leaves the southeast corner of the Marathon quadrangle, and the highest summit is an unnamed peak in the Del Norte Mountains, 6,151 feet high, in the northwestern part of the Monument Spring quadrangle. Horse Mountain, the summit of one of the ridges of Paleozoic rock, is the highest peak in the Marathon Basin. Its crest, 5,010 feet high, is lower than the summits of any of the limestone escarpments on the rim. Most of the ridges in the basin are not more "

---

Blackwelder, Eliot, Yardangs: Geol. Soc. America Bull., vol. 45, pp. 164-165,1934.

This and other possible origins of charcos are discussed by Kirk Bryan (The Papago country, Arizona: 1". S. Geol. Survey Water-Supply Paper 499, pp 121-123, 1925).

This subject has been treated at some length by W. M. Davis (Rock floors in arid and humid climates: Jour. Geology, vol. 38, pp. 146-149, 1930). It is also discussed in less technical form in his Physiographic contrasts, east and west: Sci.

Davis, W. M., Rock floors in arid and in humid climates: Jour. Geology, vol. 38, p. 145, 1930.

Idem, p. 158.

 

8

than 700 feet above their surroundings, but the escarpments that encircle the basin rise 1,000 to 1,500 feet above the floor.

The physical features of the Marathon region have been formed by the differential erosion of resistant and nonresistant rocks by streams. Among the more resistant rocks are limestones, which, because of the semiarid climate, stand as hogbacks, steep-sided plateaus, and high mountains. The region as a whole has reached a mature stage of erosion. There are, however, wider areas of sloping plains and much steeper and more rugged unconsumed areas than would be present in similar areas at the same stage of erosion in a humid climate.

Superb views of the Marathon Basin are to be had from the high escarpments on its east and west sides (fig. 2). The panorama is particularly impressive from Housetop Mountain (fig. 2, A), a projecting tongue of the Cretaceous plateau on the east edge of the basin, whose western face rises as a bold cliff 1,500 feet above the basin floor. From this eminence, in the clear air of the desert, the whole basin and its surroundings appear spread out like a map.

On the sky line to the west and south rise the mountains of the Mexican highlands, which lie beyond the Marathon Basin. To the southwest are the rugged peaks of the Chisos Mountains, to the south the domelike mass of the Sierra del Carmen, gashed by the canyon of the Rio Grande. Farther east the Serrania del Burro and other ranges stretch far away into Mexico until they are lost in the bright haze of the horizon.

Between the observer and the mountains of the horizon are ridges and plateaus of lesser order. To the southeast are Cretaceous tablelands, sloping toward the east, intricately carved into canyons, whose sides are banded by limestone ledges, as straight as if drawn by a rule. Westward the tablelands rise and project in long promontories into the Marathon Basin. The limestone ledges at the ends of the promontories are broken into disconnected tables and conical buttes, perched on reddish rounded slopes and hillocks of Paleozoic rock. Here and there ledges are discernible in the lower beds, but these run at a steeper angle than those of the Cretaceous limestones.

In the middle distance, between the tablelands and the observer, are the hills and plains of the Marathon Basin. The flats, streaked in places by white gravel deposits, are covered by a lacy network of drainage channels, with fringes of dark vegetation. Between the flats are bare rocky ridges and miniature mountains, each of which assumes a color and form determined by the nature of the rock from which it has been carved. Near at hand are reddish hills and ragged ledges of sandstone and narrow hogbacks of limestone. Farther away, in the center of the basin, is a broad cluster of hills, streaked by white ledges of novaculite. From this distance, most of the hills have no evident plan or arrangement, and they are apparently turned and twisted in greatest confusion. Here and there, however, the eye can distinguish sharp ridges and chains of knobs in the white rock, and in places a spine of vertical strata projects above the rest.

Such a distant view, in a region of great complexity, can only reveal the outlines of the geography and geology and serve to arouse the imagination of the observer. If he should wish to untangle the geologic history of the land, he must descend into the plains, in order to analyze its many features.

ESCARPMENTS BORDERING THE MARATHON BASIN

Escarpments and plateaus on the east and south sides

- The Cretaceous limestone escarpments on the east and south sides of the basin stand 500 to 1,500 feet above the basin floor. These are parts of* the high western dissected margin of the Edwards Plateau. East of the basin the plateau surface inclines gently eastward on the stripped bedding planes of resistant layers (pl. 1, B, and fig. 3, A). To the south the inclination of the strata is steeper, and here two prominent parallel limestone cuestas, called the Maravillas scarp by Hill, face northward toward the basin (pl. 14, B, and fig. 3, B).

The rocks of the plateau consist of an alternation of resistant limestone layers with weaker beds of marl. Near the basin the cap rock of most of the escarpments is the Edwards limestone (pl. 13, D). The resistant beds of the Edwards and other limestones resemble "huge stone walls of so ancient a date that they have crumbled into ruins. The less resistant beds form less steep slopes covered with debris from the overlying more resistant layers, and the whole escarpment or canyon wall gives a buttress affect like that of ruined Gothic architecture."

Drainage in the plateau country adjacent to the Marathon region is prevailingly consequent; the streams follow the slope of the Cretaceous surface and radiate from the Marathon dome (fig. 9, B). Several of the consequent streams head in the Marathon dome and flow eastward or southward into the plateau country. These have broken the escarpments at the edge of the Marathon Basin into separate segments and in places have reduced the segments to narrow promontories. The streams in the plateau have carved innumerable canyons, which have broken the originally continuous plateau surface into small areas of tableland.

In one of the canyons east of the Marathon Basin there is evidence of a relatively recent drainage change. "

---

Hill, R. T., Physical geography of the Texas region: U. S. Geol. Survey Top. Atlas, folio 3, p. 4, 1900.

Baker. C. L., and Bowman, W. F., Geologic exploration of the southeastern front range of trans-Pecos Texas: Texas Univ. Bull. 1753, p. 133, 1917.

 

9

 

10

A broad valley, followed by the Southern Pacific Railroad, branches from Dry Canyon about 20 miles east of the edge of the Marathon Basin (fig. 9, B, and preliminary topographic map of the Longfellow quadrangle). The upper part of this valley, near the Marathon Basin, drains, not into Dry Canyon, but into Maxon Creek, which leaves the broad valley and flows southward through a narrow gorge into San Francisco Creek. It would seem that Maxon Creek, cutting

headward from San Francisco Creek in relatively recent geologic time, captured the headwaters of a stream flowing eastward into Dry Canyon (fig. 9, A).

Most of the upland surfaces of the plateau country stand at accordant levels. These are not ancient uplifted peneplains, but plains that follow the upper surface of some resistant stratum of limestone, from which the overlying softer beds have been removed by erosion. The effect of the resistant, flat-lying beds on the topography is so great that it is difficult to prove the existence of any former levels of erosion higher than the present one in the plateau. In the country southeast of the Marathon Basin, however, are some remnants of old gravel deposits on the divides between the present streams. These have been mapped by N. H. Darton in the region between the lower course of San Francisco Creek and Dryden, 50 miles to the east (fig. 4).

One such deposit was examined by the writer on the highway between Dryden and Sanderson. It consists of well-rounded cobbles, 2 feet in maximum diameter, of chert, novaculite, and sandstone from the Maravillas, Caballos, and Tesnus formations of the Marathon Basin. There are also some cobbles of Cretaceous limestone and of volcanic rock like that now seen only northwest of the Marathon Basin. The eastward trend of the remnant gravel areas diverges at a wide angle from the courses of the present streams and the slope of the country, and none of the streams near the eastern gravel areas now head in the Marathon Basin, from which most of the material was derived. The gravel may have been deposited by one stream or several streams, but in any event it suggests a very different topography at the time of deposition from that of the present.

Escarpments on the north side

-On the northern border of the Marathon Basin are the Glass Mountains, an asymmetric cuestalike upland trending northeast, carved from northwestward tilted limestones, sandstones, and shales of Permian age. The southward-facing front of the range is dissected into cuestas capped by resistant limestone beds, which are separated by lowlands or slopes carved from sandy and shaly beds. In

"

---

Blackwelder, Elliot, Desert plains: Jour, Geology, vol. 39, pp.124-135, 1931.

 

11

the northeastern part of the mountains (as shown on the map of the Hess Canyon quadrangle), the face of the escarpment that borders the Marathon Basin is remarkably straight and steep and simulates a fault scarp.

An examination of the structure and stratigraphy proves that there are no faults here. * * * The strike of the strata in the scarps is parallel with the direction of the scarps [and] the dip of the strata is * * * in the opposite direction from that toward which the escarpments face.

The crests and back slopes of the Glass Mountains are composed of massive dolomite. The back slope differs from that of a true cuesta in that it is not controlled by the dip of the beds that underlie it but by the stripped surface of the ancient peneplain on which the overlying Cretaceous strata were deposited. This surface also declines northwestward, but at a slighter angle than the Permian dolomites. From it, in the mountain area, most of the Cretaceous rocks have been carried away by erosion.

Most of the streams that drain the north and northwest slopes of the Glass Mountains are consequent streams that follow low places in the folded surface of the Cretaceous rocks. Most of them do not show a close relation to the numerous northwestwardtrending normal faults that disturb the rocks of the mountains; they cross them at right or oblique angles, and although most of them flow from upthrown to downthrown blocks, some flow from the downthrown to the upthrown. There is, however, one conspicuous exception. Hess Canyon, a straight, narrow gorge in the northeastern part of the mountains, follows the northwest trend of the faults and is structurally a narrow graben. "When the long, narrow wedge of Hess Canyon sank to form a deep graben, the previous consequents crossing its site could no longer maintain their flow and were diverted into a new consequent course along the bottom."

Escarpments on the west side

-The escarpments on the western margin of the Marathon Basin are known as the Del Norte Mountains in the north, and the Santiago Mountains in the south (pl. 23). The two ranges are separated near the boundary between the Monument Spring and Santiago Peak quadrangles by Del Norte Gap, but structurally they are a continuous chain. The limestones that compose them are in some places as flat as the strata on the east margin of the basin, but in others they are considerably folded and faulted. The mountains are cut by several gaps. In the north is Doubtful Canyon (pl. 24), followed by a tributary of Maravillas Creek, which flows eastward across the mountains into the Marathon Basin. Next to the south is Del Norte Gap (pl. 13, B, and map of Santiago Peak quadrangle), which lies on a divide and is not now followed by any stream. South of the Marathon Basin the Santiago Mountains are also cut by the wind gap of Persimmon Gap and the water gap of Dog Canyon (both in the Bone Spring quadrangle).

The Del Norte Mountains are an upland, several miles broad and about 20 miles long, formed of limestones gently tilted toward the west. Eastward, they face the Marathon Basin in a steep, fairly straight escarpment, indented by few valleys. At the foot of the escarpment is a fault that has raised nonresistant Paleozoic beds on the east against Cretaceous limestones on the west (fig. 5, A). The escarpment has been caused by the wearing away of the nonresistant up-faulted beds, leaving the resistant down-faulted beds relatively undissected (fig. 5, B). It is thus an obsequent fault-line scarp.

The Santiago Mountains are about 40 miles in length and extend for some distance south of the southwest corner of the Marathon Basin. (See maps of Santiago Peak and Bone Spring quadrangles.) For most of their length they are a steep-sided ridge scarcely 2 miles across, carved from a narrow belt of vertical limestones (fig. 5, D). To the north, high mesas of flat-lying limestone, the Cochran Mountains (pl. 23, sec. Q-Q'-Q"; pl. 21; and map of Santiago Peak quadrangle) lie between them and the Marathon Basin. The Santiago Mountains owe their height partly to a large normal fault that has downthrown the Cretaceous beds on the east (fig. 5, C), but erosion has in most places leveled off the up-faulted beds for several miles west of its trace.

The escarpments of the Del Norte, Cochran, and Santiago Mountains are flanked on the east by pediments, or rock-cut plains, which slope down to Maravillas Creek on the east. The plains are covered by a thin layer of limestone gravel washed down from the mountains. In the Del Norte and Cochran Mountains, where the resistant limestone beds lie flat, recession of the cliffs has taken place by sapping of the nonresistant Paleozoic shales beneath. Considerable recession has apparently taken place in fairly recent time, when Maravillas Creek and its tributaries trenched the gravel of the pediments. As a result, the mountains north and south of Del Norte Gap are flanked by a lowland carved from shale, beyond which, at a distance of about a mile, are cuestalike remnants of gravel, sloping to the east and with their steepest "

---

Baker, C. L., and Bowman, W. F., Geologic exploration of the southeastern front range of trans-Pecos Texas: Texas Univ. Bull. 1753, p. 160, 1917.

King, P. B., Geology of the Glass Mountains, part 1, Descriptive geology: Texas Univ. Bull. 3038, p. 22, 1931.

Idem, p. 29.

Blackwelder, Eliot, The recognition of fault scarps: Jour. Geology, vol. 38, pp. 305-306, 1928.

 

12

inclination near the mountains. The cuestalike areas of gravel probably extended up to the bases of the escarpments when they stood farther forward than now.

Relation of escarpments bordering the Marathon Basin to the later tectonic movements.

-The writer believes that most of the physical features to be seen in the escarpments bordering the Marathon Basin have been formed by the stripping of the Cretaceous cover from the crest of the Marathon dome. He therefore considers it probable that they have resulted from the erosion of rocks of different composition and structure and that they are not caused directly by doming, folding, or faulting.

A somewhat different interpretation, however, has been made by Baker, who believes that "from the standpoint of physiographic development the doming of the Cretaceous rocks rimming the Marathon Basin probably occurred long subsequent to the Laramide movements, and very possibly as late as the Lafayette", an interpretation which implies that the last folding was sufficiently recent to have had a direct influence on the aspect of the present land forms. The surface rocks of the Del Norte and Santiago Mountains are thought by him to be "the surface rocks at the time the latest deformation began", and the streams in the water gaps of Doubtful and Dog Canyons are interpreted as "almost certainly antecedent."

In order that the reader may understand the problem, it seems desirable that the Tertiary structural history of the Marathon region as worked out by the writer and as more fully described on later pages be summarized at this place. Since the withdrawal of the seas, at the end of Cretaceous time, the Marathon region has remained as a land area. After the Cretaceous period two groups of rocks were laid down on parts of the surface in trans-Pecos Texas-the lava flows and associated sediments of the Davis Mountains country and the bolson and lacustrine deposits of the basin and range country. No remnants of these rocks are now found within the Marathon region, and it seems improbable that either of them were ever laid across the region in any great thickness. After the emergence of the Marathon region from the Cretaceous sea, it was subjected to three periods of movement. The movements in the first two periods, one of which was older and the other younger than the lavas to the west, caused the broad doming of the Marathon region and produced local sharp folding and thrust faulting in the Del Norte and Santiago Mountains on the west. The last movement, considerably later than the other two, broke the rocks of the region into fault blocks; its effects are particularly marked in the Glass, Del Norte, and Santiago Mountains.

It seems probable that the Marathon region has been more or less actively eroded during the whole time since the last doming and that nearly all the present surface features have resulted from that erosion. The steepness of many of the escarpments, particularly on the west side of the basin, appears to be a normal feature of the desert landscape, even in a region of as "

---

Compare the description of similar features along the base of the Book Cliffs by W. S. Glock (Premonitory planations in western Colorado: Pan-Am. Geologist, vol. 57, pp. 32-33,1931).

Baker, C. L., Date of the major diastrophism and other problems of the Marathon Basin, trans-Pecos Texas: Am. Assoc. Petroleum Geologists Bull., vol. 12, p. 1115,1928.

Baker, C. L., and Bowman, W. F., Geologic exploration of the southeastern front range of trans-Pecos Texas: Texas Univ. Bull. 1753, p. 168, 1917.

Idem, p. 148.

 

13

mature dissection as this one. In places, it is true, the ridges follow the structure of the rocks within the mountains, but these ridges are composed of resistant Cretaceous limestone. That they have been produced by the stripping of nonresistant beds from their surface seems probable, and a thick sequence of such rocks, of later Cretaceous age, is exposed on the west flanks of the Del Norte and Santiago Mountains.

Moreover, these two mountain ranges, at least in the Marathon region, are not simple anticlines, but are the overturned and much faulted western limb of the Marathon dome (fig. 5 and pl. 21). The Marathon Basin, to the east of them, is structurally much higher than these mountains and has been reduced to its present low altitude by the erosion of the weak rocks that underlie it. In one of the places where the structurally greater altitude of the basin area has been caused by faulting the down-faulted rocks now rise above the up-faulted rocks in an obsequent fault-line scarp. As the region east of the mountains has a greater structural height than its surroundings, it seems rather improbable that such streams as that in Doubtful Canyon, which flow eastward across the mountains and into the basin, could have existed before the doming and maintained their courses in the face of such adverse conditions. On the contrary, it seems probable that the gaps in which these streams flow might at one time have been occupied by streams that drained westward down the structural surface of the dome, and that such streams were afterward captured by subsequent streams actively cutting in the weak rocks of the basin (fig. 9, A and B).

The normal faults of the Glass, Del Norte, and Santiago Mountains are considerably younger than the last time of doming and apparently had a much more direct influence on the topography. Thus in the Glass Mountains most of the streams seem to follow the structural lines produced by the folds but not the faults, as if they had come into existence before the time of faulting. Moreover, in at least one place, Hess Canyon, the faulting seems to have formed a new consequent stream, which cuts across the older drainage lines.

RIDGES OF THE MARATHON BASIN

The strongly folded and faulted Paleozoic rocks of the Marathon Basin have been revealed by the stripping of the cover of Cretaceous limestones from the crest of the Marathon dome. They are a fragment of the denuded roots of a widely extended mountain system, formed in the later part of Paleozoic time. The folds strike northeast, at right angles to the trend of the post-Cretaceous uplifts, and the extensions of the folds on each side are concealed by the cover of younger rocks.

Two rock formations, more resistant than the rest, stand as ridges in the Marathon Basin. The lower of these stratigraphically is the Caballos novaculite, at the top of the pre-Carboniferous succession; the upper is the Dimple limestone, lying within the Carboniferous. As the resistant beds are in all places vertical or steeply tilted, the ridges are narrow and are breached at many places by gaps or sags.

In their setting amidst the grander features of trans-Pecos Texas they must be regarded as mountains in miniature, resurrected of late to a mere shadow of their one-time glory (compare pl. 24, D and F) by the fortuitous circumstance of being denuded of the mantle accumulated on them by the sea, which had formerly entirely buried their ancient summits.

Novaculite ridges

-The novaculite rises in white ridges of bare rock, mostly monoclinal, supported on the inner side by beds of chert and limestone (pls. 1, A; 7, B; fig. 2, B). Soft shales lie on both sides of the resistant beds, and their nonresistance to erosion accentuates the sharpness of the ridges. Novaculite hogbacks enclose the excavated cores of two broad anticlinoria in the western part of the basin (pl. 23). Between the anticlinoria, shorter but no less conspicuous novaculite ridges are carved from lower anticlines and thrust blocks. On the flanks of the anticlinoria the hogbacks run nearly straight and unbroken, save for water gaps of superimposed streams. Where the older rocks pitch beneath the surface on the northeast and southwest ends of the anticlinoria, the novaculite hogbacks pass into convoluted zigzag ridges, which wind across the axes of the plunging folds. The Warwick and Lightning Hills, 8 miles east of Marathon, are a maze of such winding hogbacks (pl. 19, C).

The lesser folds, between the anticlinoria and to the south of them, present all stages of degradation, from broad-backed mountains well expressing the doubly plunging anticlinal structure to mere chains of knobs projecting above the plain. The crest of Horse Mountain is still sheeted over by novaculite (fig. 6, A), as this stratum is thicker here than elsewhere. Some other mountains, stripped of this resistant member, still show their anticlinal form in the cherts and limestones beneath. Many more, like East Bourland Mountain (fig. 6, B, and pl. 6, A, B), are partly breached by axial anticlinal valleys that penetrate weak shales in the core. One of them, the largest of the Woods Hollow Mountains (fig. 6, C), has a wide depression down its center excavated from the shales, from which projects a low ridge of the next resistant member below. This axial depression is drained by a water gap scarcely 100 feet wide in the encircling novaculite ridge.

Dimple limestone ridges

-The limestone ridge maker does not rise as high as the novaculite, and its ridge "

---

Baker, C. L., and Bowman, W. F., op. cit., pp. 162-163.

 

14

crests are breached more widely by superimposed streams. Its hogbacks surround synclinal areas and are most extensive in the eastern and northeastern parts of the basin (pl. 23). Those near Haymond, in the eastern part of the basin, extend in great curves around broad synclinal plains. One of these is shown, just below Horse Mountain, in figure 2, A. In the western part of the basin the limestone ridges are less extensive. The most prominent one, West Bourland Mountain, rises out of a synclinal lowland between novaculite ridges. It is a remnant patch of limestone 2 miles long, presenting steep faces outward on all sides but with a synclinal valley hollowed out in the center (pl. 6, C). In the southern part of the basin the Carboniferous sandstones also stand at considerable height and are carved into rugged tracts of broken ridges and ledges. These areas are locally known by such descriptive titles as Hells Half Acre and Devils Backbone.

Even summit levels on the ridges

-Many of the hogbacks, of both the limestone and the novaculite, are conspicuously even-crested (fig. 8). Near the center of the basin most of the summits stand at about 4,500 feet, but toward the south, east, and west they gradually decline in height. Two or three ridges, such as Horse Mountain and Simpson Springs Mountain, rise conspicuously above this general level.

Unlike the even summits on the Permian dolomites in the Glass Mountains, the even summit levels in this area probably do not represent the resurrected base of the Cretaceous, for in the surrounding escarpments the base of the overlying Cretaceous beds lies at a higher altitude. There is a possibility that the even summits may represent the last remnants of one or more former high-level surfaces of erosion, now preserved only on the harder rocks of the region; if this is true, too little now remains to tell much about their character.

There is a stronger possibility that the regularity of the summit levels was produced during the last or next to the last cycle of erosion by backward cutting of the ridge slopes from the rock floors on each side. The adjacent rock floors slope in a similar direction to the crests of the ridges and at about the same angle. If the hard rocks were approximately uniformly jointed everywhere, they would tend to produce ridges having nearly the same angle of slope at all places. As the hard rocks are of small but nearly constant thickness, the ridge crests might therefore have a nearly constant height above the rock floors.

LOWLANDS OF THE MARATHON BASIN

Rock floors, their nature and origin

-Under the influence of the semiarid climate erosional agencies have worn down the nonresistant rocks of the Marathon region into rock floors of wide extent. These closely resemble the worn-down surfaces which observers of desert land forms have variously termed "pediments", "graded plains", "suballuvial benches" and "subaerial platforms", "rock floors", and "planes of lateral corrasion." The term "pediment", which is most widely used for these features, is not entirely appropriate for the Marathon region, because it implies a sloping rock-cut plain fringing a mountain base, whereas the surfaces in the Marathon region may be unrelated to any mountain area. For them, Davis' term "rock floor" seems preferable, but the term "pediment" will be used for rock floors of steeper inclination near the mountain areas.

"

---

King, P. B., The geology of the Glass Mountains, part 1, Descriptive geology: Texas Univ. Bull. 3038, p. 22, 1931.

McGee, W J, Sheetflood erosion: Geol. Soc. America Bull., vol. 8, pp. 92, 110, 1897. Bryan, Kirk, Erosion and sedimentation in the Papago country, Arizona: U. S. Geol. Survey Bull. 730, p. 54, 1922.

Udden, J. A., Sketch of the geology of the Chisos country: Texas Univ. Bull. 93, pp. 10-14,1907.

Lawson, A. C., The epigene profiles of the desert: -California Univ., Dept. Geology, Bull., vol. 9, p. 34, 1915.

Davis, W. M., Granitic domes of the Mohave Desert: San Diego Soc. Nat. History Trans., vol. 7, p. 223,1933. The term was also used without special definition in a paper published by Davis in 1930.

Johnson, D. W., Planes of lateral corrasion: Science, new ser., vol. 73, pp. 174-177, 1931.

 

15

Rock floors and pediments have in the past been widely mistaken for surfaces of deposition, such as bolson plains and bajadas, but as Blackwelder has pointed out, "the pediment, and not the bajada, is the normal and inevitable form developed in arid regions under stable conditions. It is not exceptional * * * but is dominant, widespread, and characteristic." According to this author ; pediments may be distinguished from bajadas by their low and uniform gradient, by their rambling and braided rather than outward-forking streams, and by the absence of a convex fan form opposite canyon mouths. Some of these criteria have recently been questioned by Johnson, who believes that rock-cut and constructional surfaces in arid regions may be more nearly alike than has been supposed. He brings forth theoretical and field evidence which suggests that "bedrock surfaces near canyon mouths * * * [may] possess the form of alluvial fans."

The manner in which rock floors may be produced has been variously interpreted. Davis has given a useful summary of the theories that were suggested before 1930, many of which need not be repeated here. The first carefully worked out sequence of events was presented by Lawson, following a briefer statement by Paige." Lawson's conclusions have been more or less closely followed by Bryan, and to a certain extent by Davis. The theory is summarized as follows:

A steep-sloping mountain front is worn back at a constant declivity by the ordinary subaerial processes of mountain recession as controlled by arid weathering and washing. Below the mountain front the worn-down surface of the rock floor-the pediment-is at first narrow, rather steep, and covered with a graded embankment of alluvium. But the embankment gradually rises * * * because its lower end reaches the rising detrital floor of an infilling intermont basin; and as it rises it overlaps farther and farther on the growing pediment, which, as it broadens, becomes less steep and almost bare.

In this theory chief emphasis is placed on the recession of the mountain slopes by weathering, and the pediment is considered to be "a slope of transportation * * * determined by the grade necessary to transport debris away from the mountains."

The problem is approached from a different viewpoint by Johnson and Blackwelder, who attribute the chief work of pediment cutting to the lateral corrasion of streams flowing across it. Johnson calls attention to an early deduction by Gilbert that downward wear of streams ceases when the load equals the capacity for transportation. Lateral corrasion then becomes relatively and actually of importance and carves an even surface covered by a thin deposit of alluvium. By cutting laterally into each other's valleys and consuming all remnants of the intervening divides, neighboring streams cooperate to carve a single plain of broad extent.

As noted above, however, Johnson and Blackwelder are not in agreement as to the forms produced. Johnson believes that rock fans, similar to alluvial fans, must be produced by lateral corrasion because the "inclined stream [is] relatively * * * fixed in position at the point of issuance from the canyon mouth and shifts more and more widely below that point.." He follows Paige in ascribing the steepness of the mountain front behind the pediments, not to weathering, but to the undercutting of streams swinging laterally on the surface of rock fans.

Most students of desert erosion believe, with Blackwelder, that the rock floor or pediment "is the desertinhabiting species of the genus peneplain * * * and the higher gradient which distinguishes it is conditioned by aridity." Johnson, however, suggests that this surface is "developed rapidly without any necessary relation to base level and may normally be trenched by streams without any change in the attitude or altitude of the areas affected."

Comparison of rock floors previously described with those in the Marathon Basin

-The rock floors of the Marathon Basin show certain differences from the typical examples farther west that have been described. These differences are more in degree than in kind and are dependent on the local geology, the rainfall, and the disposition of the drainage.

At Marathon the rock floors are carved from steeply tilted sedimentary rocks of varying resistance, rather than from massive rocks such as granites. Most of the rock floors of the region are therefore cut on belts of weak rock. The sedimentary rocks are not as subject to spalling and granular decay as the granites in the pediment areas of southern Arizona and the Mojave Desert. Their weathered fragments are therefore of different size and shape, and this influences the form of the graded slopes that must be carved to transport them.

The base level of the rock floors is controlled by numerous streams which have access to the master "

---

Blackwelder, Eliot, Desert plains: Jour. Geology, vol. 39, p. 138, 1931.

Idem, pp. 136-137.

Johnson, D. W., Rock fans in arid regions: Am. Jour. Sci., 5th ser., vol. 23, pp. 389-420, 1932.

Johnson, D. W., Planes of lateral corrasion: Science, new ser., vol. 73, p. 175, 1931.

Davis, W. M., Rock floors in arid and humid climates: Jour. Geology, vol. 13, pp. 14-19, 1930.

Lawson, A. C., The epigene profiles of the desert: California Univ., Dept. Geology, Bull., vol. 9, pp. 23-48, 1915.

Paige, Sidney, Rock-cut surfaces in the desert ranges: Jour. Geology, vol. 20, pp. 442-450, 1912.

Davis, W. M., Rock floors in arid and in humid climates: Jour. Geology, vol. 38, p. 15, 1931.

Bryan, Kirk, Erosion and sedimentation in the Papago country, Arizona: U. S. Geol. Survey Bull. 730, p. 57, 1922.

Johnson, D. W., Planes of lateral corrasion: Science, new ser., vol. 73, p. 174, 1931.

Gilbert, G. K., Geology of the Henry Mountains, pp. 126-133, U. S. Geol. and Geog. Survey Rocky Mtn. Region, 1877.

Johnson, D. W., Rock fans in arid regions: Am. Jour. Sci., 5th ser., vol. 23, p. 392, 1932.

Paige, Sidney, op. cit., pp. 449-450.

Blackwelder, Eliot, Desert plains: Jour. Geology, vol. 39, p. 138, 1931.

Johnson, D. W., Planes of lateral corrasion: Science, new ser., vol. 73, p. 177, 1931.

 

16

drainageways of the Rio Grande and the Pecos, rather than by an interior basin, whose surface is rising slowly by aggradation. This condition favors the development of rock floors of wide., extent, as suggested by Bryan and Davis, and the alluvial apron assumed by Lawson and Paige either does not exist or is thin.

The rock floors do not encircle mountain areas but lie in a basin between the highlands. As the resistant rocks of the basin form a relatively small part of the whole, the ridges of unconsumed material are relatively narrow and are penetrated in all directions by arms of the rock floors. The rock floors of the Marathon region thus resemble in many of their features the intramontane canyons and headwater basins in southern Arizona described by Bryan.

Rock floors of the Marathon Basin

-Stratigraphically, there are three groups of weak rocks in the basin. The lowest group, consisting of pre-Carboniferous limestone and shale, lies at the surface in the two anticlinoria in the western part of the basin. The upper two groups, consisting of sandstone and shale of Carboniferous age, lie respectively below and above the ridge-making Dimple limestone and form broad expanses of plain surrounding the anticlinorial areas. Erosion in the Marathon Basin has proceeded with greatest ease along these belts of weak rock, thus favoring the development of subsequent streams, at the expense of consequent streams superimposed on the Paleozoic surface from the former Cretaceous cover. Rock floors have been cut back laterally from the superimposed or subsequent master streams, as far as the limiting belts of hard, ridge-making rocks on both sides.

Along the escarpments bordering the Marathon Basin are pediments with gradients of 200 or 300 feet to the mile, but these flatten outward and have a concave upward profile. Within the basin itself the rock floors have gradients of 75 feet or less to the mile. The rock floors have a general inclination southward toward the Rio Grande, but in detail they are a complex group of sloping surfaces, each drained by an axial stream tributary to one of the master drainage channels. The rock floors in the different drainage basins have an unlike form and gradient because of the varying character of the bedrock, amount of run-off, and relation to each master stream. Some are also controlled by local base levels, determined by sills of hard rock in the stream beds." Though the rock floors of each drainage area are accordant at their lower ends with those along the master stream, they may be discordant at their heads and margins with the floors drained by other streams. The coalescing floors of adjacent drainage areas are thus not likely to meet at the same level, and the junction in some places is marked by a low dissected escarpment that faces the lower rock floor." The discordance is most marked where the coalescing floors drain into far separated master streams. Thus the heads of rock floors drained by tributaries of the Pecos in the northeastern part of the Marathon region stand above W B Flat on the south (pl. 23 and map of Longfellow quadrangle), at the top of an escarpment 500 feet high.

In the Marathon region, so far as the writer's observations go, there are no well-developed rock fans of the type described by Johnson. It is possible, however, that some of the fanlike areas at the bases of the higher mountains, which the writer has interpreted as due to deposition on the pediments, may be underlain by fans carved from bedrock.

Erosion and deposition on the rock floors

-In most places the rock floors of the Marathon Basin are either covered by a thin layer of gravel or dissected to a moderate depth.

Undissected rock floors covered by gravel are most extensive in the north half of the basin (pl. 23). The alluvial deposits here may be of some antiquity, for cobbles of Cretaceous and Dimple limestone occur in some of them where no outcrops of those formations now remain in the drainage area. They may also have been the source of the elephant bone reported to have been found near Marathon in 1930.

Most of the rock floors in the area are covered by 10 to 20 feet of gravel, which effectively masks the bedrock over wide areas, although valleys that incise the surface reveal the small thickness of the deposit. At some places the cover is thicker. Water wells in the northern part of the basin, near the Glass Mountains, penetrate 100 feet or more of gravel, and some of the streams that discharge from these mountains have deposited low alluvial fans on the rock floor." Alluvial fans have been built also along the northeast base of the Del Norte Mountains by Antelope Creek (fig. 7) and the stream in the large canyon 2 miles south of Altuda. Another large fan exists south of the Marathon Basin, in the Hood Spring quadrangle. The outer edges of the two fans in the Del Norte Mountains have been built across the courses of other streams and have ponded some of them (fig. 7), thereby interrupting the normal cycle of down cutting. Clearly these two fans cannot be "

---

Bryan, Kirk, op. cit. (Bull. 730), p. 56.

Davis, W. M., Rock floors in arid and in humid climates: Jour. Geology, vol. 38, p. 18, 1931.

Bryan, Kirk, op. cit., pp. 47-48.

Baker, C. L., and Bowman, W. F., Geologic exploration of the southeastern front range of trans-Peces Texas: Texas Univ. Bull. 1753, p. 164, 1917.

Davis, W. M., Granitic domes of the Mohave Desert: San Diego Soc. Nat. History Trans., vol. 7, pp. 237-239, 1933. King, P. B., Geology of the Glass Mountains, part 1; Texas Univ. Bull. 3038, p. 27, 1930.

King, P. B., op. cit. (Texas Univ. Bull. 3038), p. 20.

Blackwelder, Eliot, Desert plains: Jour. Geology, vol. 39, p. 139, 1931.

King, P. B., op. cit., fig. 9 C.

 

17

of the rock-cut type described by Johnson. The outer edges of the fan in the Hood Spring quadrangle are dissected by tributaries of Maravillas Creek, suggesting that the building of this fan antedated the dissection of the rock floors described below.

Deposition of alluvium on the surface of the rock floors indicates that the balance of conditions which permitted their erosion has been modified, either by a change in the base level of the master streams or by a decrease in the amount of rainfall. The widespread occurrence of the deposit in all drainage basins suggests that it was caused by a change in climate toward aridity.

Dissection of the rock floors and pediments and of their alluvial cover is of several sorts. At many places along the bases of the mountains the drainage channels "cut deep, steep-walled trenches with few lateral tributaries in the heterogeneous materials of the debris fans, and lower down spread out in broad, almost imperceptible channels in the * * * materials beyond the foot of the debris slopes." Similar features have been described by Bryan in the Papago country and are explained by him as caused by a relative increase in the amount of rainfall, but Baker has suggested that they may be produced as a normal feature after maturity is reached in the cycle of erosion.

In the central and southern parts of the Marathon Basin most of the rock floors are dissected by Maravillas and San Francisco Creeks and their tributaries. Toward the north the streams are cutting headward into the broad area of gravel-covered lowland of the northern part of the basin, and in places the south edge of the undissected area is marked by a low scarp.

Typical dissected country is found along Peña Blanca Creek, in the southeastern part of the Marathon Basin (pl. 24), where the originally smooth rock floor is trenched to a depth of 50 or 100 feet by valleys of trellis pattern, which have been carved along belts of shale of Carboniferous age, leaving the intervening belts of sandstone as low, even-crested ridges. The tops of the ridges are in places covered by patches of gravel, which are remnants of deposits formerly covering the whole of the old rock floor. These patches are accordant with the undissected gravel-covered floors of the northern part of the basin. Similar relations are also found in the lowland of pre-Carboniferous rock in the Dagger Flat area (fig. 8). Here most of the remnants of the old rock floor are preserved on the edges of a resistant limestone bed, the Port Peña formation, which now crops out in low ridges. Some recession of the Cretaceous escarpments bordering the Marathon Basin has apparently taken place at the same time as the down cutting of the rock floors in the basin.

In the southern part of the Marathon Basin, 25 to 50 feet below the higher rock floor, is another floor of smaller extent (fig. 8). Because of the large contour interval of the topographic maps, it was not possible to map the two surfaces separately. Along the streams a still lower surface has been developed, which is covered by clay and gravel of finer texture than those found on the two older surfaces. This surface has its widest extent along the lower course of Maravillas Creek; elsewhere it fringes the streams in belts rarely more than 2 miles in width and vanishes entirely where the streams cross areas of hard rock.

The dissection of the rock floors apparently resulted from headward cutting of the tributaries that drain by the shortest route to the Rio Grande. The process is therefore probably caused by a lowering of the base level of that stream, which may have resulted either from slight regional uplift or from the lowering of a "

---

Baker, C. L., and Bowman, W. F., Geologic exploration of the southeastern front range of trans-Pecos Texas: Texas Univ. Bull. 1753, p. 161, 1917. For a map of one of these features see King, P. B., op. cit., fig. 9 A.

Bryan, Kirk, Erosion and sedimentation in the Papago country, Arizona: U. S. Geol. Survey Bull. 730, pp. 60-65, 1922.

Baker, C. L., Notes on the later Cenozoic history of the Mohave Desert region in southeastern California: California Univ., Dept. Geology, Bull., vol. 6, pp. 374-377,1911.

 

18

temporary base level by the destruction of a rock dam in the river.

There is some suggestion that the dissection of the rock floors may have been aided by slight recurrent uplifts of the Marathon dome. Tilting to the south is suggested on the southeast side of the dome by the anomalous position of upland gravel deposits (fig. 4) and by the capture of an extensive drainage system by lower Maxon Creek (fig. 9). Within the basin also dissection is most active in the central and southern parts, where southward-flowing streams would be accelerated by such an uplift. In the northern undissected part of the basin, which also drains mostly to the south, erosion processes do not seem very active and may have been retarded by an uplift with its center farther south. Most of the features in the basin, however, may be interpreted in the other ways.

In many of the valleys the calys of the lowest surface are now in process of dissection by steep-walled arroyos. According to John Bennett, those near his apiary south of Haymond have been cut since about 1920. According to C. L. Baker, those in Green Valley, southwest of the Marathon Basin, have been cut since about 1890. This cutting is similar to that in New Mexico described

 

19

by Bryan. It may have been brought about by heavy grazing of the country by stock, which has perhaps facilitated run-off and soil erosion.

Streams of the Marathon Basin

-The lowlands of the Marathon Basin are drained by several streams with intermittent flow, the largest of which are Maravillas and San Francisco Creeks. In places these streams pass from one lowland area to another by crossing the ridges of hard rock in narrow water gaps. The broader features of the stream pattern have been acquired from the consequent drainage that was formed on the now removed cover of tale Marathon dome and that has been superimposed on the Paleozoic rocks beneath. The pattern has been considerably altered by later cutting along weak rock belts by subsequent streams, as may be seen by comparing on the topographic maps the dendritic stream pattern of the Cretaceous areas surrounding the Marathon Basin with the trellis pattern within the basin. Other modifications of the original pattern have been caused by the advantage possessed by streams flowing southward, off the dome, to the nearby, low-lying Rio Grande over those flowing northward to the higher, more distant Pecos River.

Maravillas Creek is apparently consequent to a post- Cretaceous syncline as far north as the latitude of Del Norte Gap. Subsequent streams, such as Woods Hollow Creek, enter it from the east in this part of its course (fig. 9, B). They are cut in weak-rock belts in the Paleozoic folds of the Marathon Basin. The water gaps by which they cross some of the hard-rock belts may have been inherited from a former southwardflowing consequent drainage system (fig. 9, A). The upper course of Maravillas Creek is apparently subsequent and is cut out along the disturbed zone at the west margin of the Marathon uplift. Its two northeastern tributaries, Peña Colorada and Dugout Creeks, have large headward extensions that drain the southern foothills of the Glass Mountains, where the strata dip north off the Marathon uplift. These creeks may possibly be the dismembered fragments of an original consequent stream that flowed southwestward through gaps in the Del Norte Mountains to join Chalk Draw, which is a consequent stream flowing in a Cretaceous syncline west of the Marathon uplift, in the Santiago Peak quadrangle (fig. 9). One such gap, Doubtful Canyon, is now followed by the eastward-flowing Maravillas Creek. Another, Del Norte Gap, is now a wind gap.

San Francisco Creek is a consequent stream which has acquired the original headwaters of the eastward- flowing Maxon and Dry Creeks by headward cutting along belts of weak rock (fig. 9, B). The two water gaps by which it crosses the limestone hogbacks near Haymond were probably inherited from the consequent headwaters of Maxon Creek. Maxon and Dry Creeks now head near the edge of the Marathon Basin, in broad valleys that are broken off to the west by dissected country draining into San Francisco Creek. The main western tributary of San Francisco Creek, Peña Blanca Creek, is a subsequent stream that flows in a weak-rock belt of northeastward-dipping sandstones and shales, between hogbacks of novaculite and of Carboniferous limestone.

STRATIGRAPHY

GENERAL OUTLINE

Most of the bedrock that crops out in the Marathon region consists of consolidated sedimentary rocks. Such rocks also underlie the wide areas of Quaternary gravel and clay in the plains. Intrusive igneous rocks occupy only small areas, and extrusive rocks are found only along the west flank of the Marathon region.

The stratified rocks in the region are of many ages and represent nearly the whole span of geologic time from the Cambrian to the Tertiary. The base of the Paleozoic is not exposed, for the lowest beds raised in the anticlines are a part of the Upper Cambrian. These oldest rocks form the base of a thick Paleozoic succession that includes formations of Ordovician, Devonian (?), and Carboniferous age. The Carboniferous strata attain a great thickness and include in the Marathon Basin thick formations of Pennsylvanian age and in the Glass Mountains a series of Permian strata. Above the Paleozoic on the flanks of the Marathon uplift are rocks of Lower and Upper Cretaceous age. Above them on the west are Tertiary lavas and tuffs.

Nearly all the strata contain fossils, and in some places the fossils are very abundant. Marine invertebrate fossils are found in the rocks of Cambrian, Ordovician, upper Pennsylvanian, Permian, and Cretaceous age. In the lower part of the Pennsylvanian and in the Tertiary tuffs west of the Marathon region some of the strata contain plant remains.

The stratified rocks of the region were laid down under progressively changing geographic conditions. The Paleozoic rocks below the Permian were laid down in a subsiding area, the Llanoria geosyncline (fig. 16 and pl. 20). The geosyncline did not have the same form as the modern, nearly circular Marathon Basin. At the edges of the basin the rocks of geosynclinal facies strike northeast beneath the Cretaceous cover and are encountered again to the east in deep wells and to the southwest in the small uplift of the Solitario. "

---

Bryan, Kirk, Date of channel trenching (arroyo cutting) in the arid Southwest: Science, new ser., vol. 62, pp. 338-344, 1925.

Sellards, E. H., Pre-Paleozoic and Paleozoic systems, in The geology of Texas, vol. 1, Stratigraphy: Texas Univ. Bull. 3232, p. 23, 1933.

 

20

The geosyncline did not extend as far to the northwest or southeast. To the northwest the Paleozoic rocks in the nearest exposures are thinner and more calcareous and were laid down on a sea bottom that did not either rise or subside greatly. To the southeast the sediment: in the geosyncline suggest that there was a land area Llanoria, which at times rose to a considerable height In Mesozoic times the Paleozoic geographic feature: had disappeared. The Cretaceous rocks were laic down on the surface of an extensive peneplain carved from the older rocks, and the greatest area of subsidence was southwest of the Marathon region.

The rocks of the region stand in a variety of structural attitudes, and some of them have been greatly deformed. The earlier Paleozoic rocks of the Llanoria geosyncline were folded and faulted before Permian time, and the Permian rocks lie unconformably upon them along the south side of the Glass Mountains. The Cretaceous rocks in turn truncate both the folds of the Marathon Basin and the tilted rocks of the Glass Mountains. The Cretaceous rocks have themselves been deformed, both before and after the early Tertiary volcanic eruptions.

The total thickness of the Paleozoic rocks is about 21,000 feet, of which the pre-Carboniferous strata comprise 2,500 feet; the Pennsylvanian, 12,000 feet; and the Permian, 6,500 feet. The Cretaceous rocks in the area have a maximum thickness of about 1,200 feet. The following table summarizes the formations exposed in the Marathon region:

 

21

PRE-CAMBRIAN ROCKS

No rocks of pre-Cambrian age crop out in the Marathon Basin. The floor of basement rocks in the region has been deeply buried by Paleozoic strata of the Llanoria geosyncline. To the north and south of the geosynclinal area pre-Cambrian rocks lie nearer to the surface and have been discovered in a few exposures and a few deep wells.

Pre-Cambrian rocks north and south of Marathon

About 50 miles northeast of Marathon, near Fort Stockton, a well drilled by the Shell-Humphreys Companies on university land penetrated old rocks below the Permian at a depth of 4,750 feet. The basement rocks are considered to be granites by J. T. Lonsdale, who found that the abundant minerals in the cuttings are quartz, microcline, albite, hornblende, and biotite. Accessory minerals include magnetite, zircon, apatite, and calcite. This is probably a part of the foreland area north of the Llanoria geosyncline. The granite probably has the same relation as the pre-Cambrian rocks that crop out at the surface near Van Horn, 100 miles northwest of Marathon. At that place strata of Permian age overlap across the older Paleozoic rocks and rest locally on the pre-Cambrian on the crests of pre-Permian uplifts.

About 80 miles south of Marathon C. L. Baker has found schists beneath the Cretaceous on the crest of an anticline in the Sierra del Carmen, east of the village of Boquillas. This exposure is probably a part of the land that bordered the geosyncline on the southeast.

Fragments of crystalline rocks in the Paleozoic sediments at Marathon.

-The land area of Llanoria, southeast of the geosyncline, appears to have been composed largely of crystalline rocks and probably stood as a highland or mountain area during a large part of Paleozoic time (fig. 16 and pl. 20 A). For the most part the former highland is now buried beneath Cretaceous and younger strata, and the hypothesis of its former existence is based largely on evidence supplied by the composition of the Paleozoic sediments in the geosyncline.

"

---

Van der Gracht (The Permo-Carboniferous orogeny in the south-central United States: K. Akad. Wetensch. Amsterdam Verb., Afd. Natuurk., deel 27, no. 3, table Va and elsewhere, 1931) notes a reported discovery of pre-Cambrian rocks in the similar district of the Solitario uplift, southwest of Marathon. This report has proved to be erroneous, according to a letter from E. H. Sellards, June 1932.

Sellards, E. H., op. cit. (Texas Univ. Bull. 3232), p. 52. The well has also been noted by E. L. Jones and R. C. Conkling (Basement rocks in the Shell-Humphreys well, Pecos County, Tex: Am. Assoc. Petroleum Geologists Bull., vol. 14, pp. 314-316, 1930) and by P. B. King (Geology of the Glass Mountains, part 1: Texas Univ. Bull. 3038, p. 117, 1930). Jones and Conkling regarded the rocks as metamorphosed sandstones, penetrated by igneous dikes.

Quoted by Bose, Emil, Vestiges of an ancient continent in northeastern Mexico: Am. Jour. Sci., 5th ser.. vol. 6, p. 133, 1923. See also Kellum, L. B., Imlay, R. W., and Kane, W. G., Evolution of the Coahuila Peninsula, Mexico; Part 1, Relation of structure, stratigraphy, and igneous activity to an early continental margin: Geol. Soc. America Bull., vol. 47, pp. 972-977,1936.

 

22

Many of the Paleozoic strata contain fine fragmental material derived from crystalline rocks. Conglomerate beds in various parts of the Ordovician section, but particularly in the Maravillas chert, contain fragments of vein quartz. The Tesnus and Haymond formations (Pennsylvanian) contain beds of arkose, which increase in number and thickness toward the south. Their source perhaps lies in highlands in that direction. The arkose contains grains of microcline and other minerals, probably derived from the break up of granite, as well as small chips of slate and phyllite.

In the boulder-bed member of the Haymond formation near Haymond station there are fragments of ancient rocks of larger size. Most of these are well-rounded cobbles, some of which reach a foot in diameter. They consist of granite, aplite, pegmatite, vein quartz, rhyolite, quartz conglomerate, and possibly of schist. Several thin sections of the rocks have been examined. Some of the specimens consist of fine-grained granite and some of porphyritic granite. Others consist of rhyolites of various types that contain large phenocrysts of quartz and plagioclase, whose edges are rounded by resorption. The groundmass of the rhyolites is a finely crystalline aggregate, probably a devitrified glass. In some specimens it shows a well-developed sinuous flow structure. In the exposures of the boulder-bed member north of Haymond station cobbles of the igneous rocks are few, a fact which suggests that they came from a southern and probably distant source.

Depth of the pre-Cambrian floor in the Marathon Basin

-There is no clear evidence to show at what depth the pre-Cambrian floor lies beneath the Marathon Basin. The oldest exposed parts of the Paleozoic section are of Upper Cambrian age. The character of the structural features of these and the overlying Ordovician rocks suggests that they are underlain by a considerable body of strata that is probably of incompetent character. Unfortunately, paleogeographic evidence furnishes no good clues to the age of these oldest members of the Paleozoic. They may be Middle or Lower Cambrian. The writer considers it possible that the thrust sheets in the Marathon Basin originated in wedges of crystalline rocks that lie at an unknown depth beneath (pl. 20, B, C, and D).

CAMBRIAN SYSTEM

DAGGER FLAT SANDSTONE

GENERAL FEATURES

The Dagger Flat sandstone was named by the writer in 1931 for exposures in Dagger Flat, 13 miles south of Marathon. The sandstones were first described by Baker and Bowman in their section on Threemile Hill, 3 miles northeast of Maravillas Gap, and were designated, without age assignment, as member 2 of their † Marathon series.

The formation is the oldest rock found in place in the Marathon region, and its base is nowhere exposed. Its strata are in all places so intricately contorted that the exposed thickness is not exactly known, but a probable maximum of 300 feet is found on the south side of Dagger Flat. It is exposed in long, narrow belts in the center of the anticlines in both the Marathon and Dagger Flat anticlinoria, with the most extensive exposures in the south, on the Buttrill ranch and near Threemile Hill.

The Dagger Flat sandstone consists of thick ledges of saccharoidal buff sandstone interbedded with shale. These pass toward the top into shales and thin flaggy sandstones, with some calcareous beds that contain a few Upper Cambrian fossils. The formation is overlain without apparent break by thin flaggy graptolite-bearing limestones of the Marathon formation (Lower Ordovician).

LOCAL FEATURES

Dagger Flat area

--The Dagger Flat sandstone is well displayed for a distance of 4 miles northeast from the Buttrill ranch (pl. 23) in a series of narrow belts on the crests of anticlines. The southeasternmost of these belts is the broadest and has a width of 1,500 feet. The lowest beds are conspicuous ledges, each 3 or 5 feet thick, of white to buff sugary, moderately coarse- grained sandstone, weathering pale brown (sec. 12, pl. 2). In places these beds pass into a fine conglomerate of rounded pebbles of vein quartz, some of which show a notable secondary regrowth of quartz crystals. The massive sandstones are overlain by flaggy and thinly laminated brown and greenish micaceous sandstones weathering to angular blocks and flags, with much interbedded shale, particularly toward the top. In the upper part there are also several layers of laminated brown calcareous sandstone. Their bedding surfaces are strewn with fragmental fossils, including the cephalons of various trilobites, such as Agnostus and many shells of Lingula and Obolus. There are also several layers of conglomerate composed of small black chert, gray limestone, and clear quartz pebbles in a matrix of brown sandy limestone. The formation is "

---

King, P. B., Pre-Carboniferous stratigraphy of the Marathon uplift: Am. Assoc. Petroleum Geologists Bull., vol. 15, pp. 1064-1065, 1931.

Baker, C. L. and Bowman, W. F., Geologic exploration of the southeastern front range of trans-Pecos Texas: Texas Univ. Bull. 1753, p. 83, 1917.

A dagger (†) preceding a geologic name indicates that the name has been abandoned or rejected for use in classification in publications of the U. S. Geological Survey. Quotation marks, formerly used to indicate abandoned or rejected names, are now used only in the ordinary sense.

The Buttrill ranch referred to in this report is the unnamed house east of the Marathon road in the northeast corner of the Santiago Peak quadrangle. Dagger Flat is in the southwest corner of the Marathon quadrangle. These two places should not be confused with the Buttrill ranch and Dagger Flat farther south, shown on the map of the Bone Spring quadrangle.

 

 

 

 

 

 

 

22g

 

23

overlain by flaggy limestones, shales, and flagstone conglomerates of the Marathon formation. The greatest thickness measured is about 300 feet, but no base is exposed. Estimates of thickness of the formation are difficult to make because of the complexity of the structure.

Threemile Hill

-A short distance to the southwest of the exposures on the Buttrill ranch are the original outcrops described by Baker and Bowman, on Threemile Hill (pl. 23). Here the rocks are of similar character but are so intensely contorted that the massive sandstones have been sliced into lenses or rolled into boulderlike masses, around which the shales have been squeezed and indurated nearly to slates. At this locality the Dagger Flat sandstone is overlain in places by a thin layer of Marathon limestone, followed by little-deformed Woods Hollow shales and earthy limestones and by the Maravillas chert (fig. 10). Similar intense contortion is found at Maravillas Gap, where the Dagger Flat sandstones are overlain, locally with a marked difference in dip, by the Maravillas chert (sec. J-J', pl. 21). The writer considers these relations to be tectonic rather than stratigraphic. It is thought that the Dagger Flat is separated from the rocks above it by an overthrust, whose trace extends along the northwest slope of the hills in which the exposures lie (fig. 10). Several large areas of overthrust rocks 2 miles northwest of Threemile Hill, now isolated from their roots by erosion, are probably the northern extension of the upper strata on Threemile Hill (sec. I-I'-I" and J-J', pl. 21).

Woods Hollow Tank

-In the northeast end of the Dagger Flat anticlinorium, 1½ to 5 miles northeast of Woods Hollow Tank, the Dagger Flat sandstone lies at the surface in two long anticlinal belts (pl. 24). The highest beds of the formation are the only part exposed here. The outcrops consist largely of shale and sandy shale, with 6-inch to 1-foot beds of mediumgrained calcareous brown sandstone, ripple-marked in places. There are also some brown limestone nodules which contain Agnostus and other trilobites. Some of the sandy beds weather to a peculiar chocolate-brown velvety surface, apparently from the leaching of the calcareous constituents, so as to leave a mat of fine quartz grains on the surface.

Marathon anticlinorium

-The Dagger Flat sandstone is found in small outcrops along narrow anticlinal belts southwest of Marathon, between Fort Peña Colorada and Monument Spring, as well as 2 miles northwest of the fort, on the south side of the road to the Roberts ranch (pl. 24). The formation here consists of much crumpled and indurated greenish shale with several layers of fine- to coarse-grained sandstone, in part calcareous. There are some arkosic pebbly layers and a few nodular layers of very fine grained dark-gray or black limestone, weathering chocolate brown. These beds contain scattered fragments of brachiopods and trilobites.

MICROSCOPIC CHARACTER

Thin sections of sandstones from the Dagger Flat formation Of Threemile Hill and the region northeast of the Buttrill ranch show that the rock consists mostly of well-rounded large quartz grains in a matrix of finely crystalline calcite. Many of the quartz grains show regrowth of the crystal faces by the addition of secondary quartz. There are also a few grains of calcite and chert. Much of the quartz shows strain shadows, and some of the specimens are traversed by irregular veinlets of coarsely crystalline calcite. A thin section of a more pebbly phase from the Buttrill ranch consisted of large, well-rounded grains of sandy limestone and chert in a calcareous matrix, with a few grains of feldspar, calcite, and quartz.

FOSSILS AND AGE

The fossils of the Dagger Flat sandstone are poorly preserved and difficult to collect. The largest collections were made in the Dagger Flat anticlinorium, and nearly all are characterized by Agnostus, Lingula, and Obolus. They have been studied briefly by Edwin Kirk and C. E. Resser, who consider the material to be of Upper Cambrian age. The formation is probably almost equivalent to the Bliss sandstone of the El Paso region.

STRATIGRAPHIC RELATIONS

The base of the Dagger Flat sandstone is not exposed, and its possible relation to older Cambrian or pre-Cambrian

 

24

formations beneath is not known. The contact between the formation and the overlying Marathon limestone is not a sharp one and is further complicated by local folding. In some places a thin conglomerate appears to form the base of the overlying formation and rests on shales assigned to the Cambrian. Elsewhere the basal beds of the Marathon formation are thin limestone flags containing Dictyonema.

PROBLEM OF THE ‡ BREWSTER FORMATION

In 1915 Baker and Bowman collected fossils from two anticlinal mountains 8 to 12 miles southwest of the town of Marathon-one locality 6 miles northeast of the junction of Peña Colorada and Maravillas Creeks, on East Bourland Mountain (pl. 6, B), and the other 1½ miles northeast of the creek junction, on Simpson Springs Mountain (pl. 5, E). The higher fossil collections from both places came from the Maravillas chert, but below this fossils were found in nodular masses of dense or crystalline gray limestone, in part glauconitic, embedded in clay shales and brown sandy flagstones. The base of the lower series was not exposed. In collections from both localities Ulrich distinguished an Upper Cambrian fauna "like * * * that found in the Dunderberg shale of the Nevada * * * section", and in addition, at the second place, a fauna characterized by Symphysurina, to be correlated with "a well-marked zone in the lower part of the Pogonip of Nevada" of which "the evidence permits of only one conclusion * * * namely, that it is Ozarkian, and most probably Upper Ozarkian." On the basis of these determinations, Baker and Bowman set off the lowest beds at the Simpson Springs Mountain locality as the ‡ Brewster formation, of Upper Cambrian age, named for Brewster County. The similar clays and flagstones lying above them, containing the so-called "Ozarkian" fossils, they considered to be member 1 of their ‡ Marathon series. The direct superposition of Maravillas (Upper Ordovician) on supposed Ozarkian of Ulrich at this place and the absence of the intervening Ordovician zones of other localities were explained by a marked erosional unconformity at the base of the upper formation.

Recent collecting at these localities has disclosed some puzzling anomalies not explained by the first interpretations. The sandy flagstones and the matrix of many of the thin layers of conglomerate are profusely fossiliferous at different levels below, between, and above the limestone "nodules." The flagstone fossils are largely species of Diplograptus, Climacograptus, and Glossograptus, and the conglomerate fossils are various bryozoans, brachiopods, trilobites, and crinoids. In the surrounding region these fossils characterize the Woods Hollow shale, of Middle Ordovician age. In addition, from the shales F. A. Bush and B. H. Harlton have obtained a microfauna which they state is of Black River age. Furthermore, southwest of Garden Springs the Woods Hollow shale, here plainly overlying the earlier Ordovician, also contains large "nodules" of the mottled and glauconitic limestone. Additional collecting from the "nodules" of Simpson Springs and East Bourland Mountains has given plentiful evidence that the fossils are of Cambrian and early Ordovician age.

It is here suggested that the "nodules" are in fact large erratic boulders of Cambrian and Ordovician limestones embedded in the shaly and flaggy strata of the Middle Ordovician Woods Hollow shale. As it is clear that the type locality of the ‡ Brewster formation is a part of the Woods Hollow shale, it has been recommended that the name " ‡ Brewster" should be abandoned. The name "Dagger Flat sandstone" is used here for the indigenous Cambrian strata.

Two faunas appear to be represented in the boulders of the Woods Hollow shale. The oldest of these is of Upper Cambrian age. In a collection made by C. L. Baker on East Bourland Mountain E. O. Ulrich has identified the following species:

• Lingulella manticula (White).
• Lingulella desiderata (Walcott).
• Acrotreta idahoensis (Walcott).
• Acrotreta sp.
• Alokistocra aff. A. aoris (Walcott).

In a collection by the writer on Simpson Springs Mountain Ulrich has recognized Agnostus and several new species of Upper Cambrian trilobites. Ulrich notes that the first three species in the list from East Bourland Mountain are typical Upper Cambrian fossils which have been found in Nevada and elsewhere, and that the fauna is decidedly like that found in the Dunderberg shale of the Nevada Upper Cambrian section.

"

---

Baker, C. L., and Bowman, W. F., op. cit. (Texas Univ. Bull. 1753), p. 83.

Communication to the writer, 1931. Van der Gracht, in referring to the Dagger Flat sandstone (op. cit., table Vb), makes the following comment: "There is controversy whether the Cambrian is represented at Marathon. The above version is contested by B. H. Harlton and F. A. Bush, who advised the writer that the type locality yielded a profusive microfauna of Black River age." The collections of Bush and Harlton came from the type locality of the ‡ Brewster and not of the Dagger Flat formation.

King, P. B , Pre-Carboniferous stratigraphy of the Marathon uplift: Am. Assoc. Petroleum Geologists Bull., vol. 15, p. 1064, 1931.

 

25

A younger fauna is of early Ordovician age. In a collection made by C. L. Baker on Simpson Springs Mountain Ulrich has identified the following species:

• Obolus rotundatus (Walcott).
• Lingulella pogonipensis (Walcott).
• Schizambon typicalis (Walcott).
• Eoorthis desmopleura (Meek).
• Symphysurina spicata n. var. (Ulrich).
• Symphysurina 2 n. sp.
• Conokephalina inexpectans (Walcott).
• Apatokephalus finalis (Walcott).
• Hungaia? sp.

From East Bourland Mountain Josiah Bridge has obtained various species of Symphysurina and other fossils similar to those in the preceding list. Ulrich notes that the fauna is like that in a well-marked zone of the lower part of the Pogonip limestone of Nevada. The Pogonip limestone is classified by the Geological Survey as of Lower Ordovician age, but this part of the formation is assigned by Ulrich to the upper part of his Ozarkian system.

It is probable that the erratic blocks in the Woods Hollow shale are of a facies foreign to the Marathon geosyncline. Their lithology is quite unlike that of indigenous rocks of the same age, and their faunas have not been revealed by the most diligent collecting from ledges in place. Massive limestones of Upper Cambrian and early Ordovician age are not known to crop out at any place in trans-Pecos Texas. They are found farther east, in the Llano-Burnet uplift (Central Mineral Region). Probably a western extension of these limestones lies beneath the surface not far north of the Marathon folded belt.

ORDOVICIAN SYSTEM

HISTORICAL SUMMARY

Rocks of Ordovician age were reported to occur in the Marathon Basin by J. A. Udden, who collected a few fossils, considered to be of Trenton age by Schuchert, during the course of his journey to the Chisos country, "along the wagon road near Ridge Spring and at different points south from this place for a distance of 10 miles."

Important stratigraphic work in the area was done by Baker and Bowman in 1915. They collected fossils and made brief studies of the Cambrian and Ordovician succession, and their fossil collections were studied by E. O. Ulrich. During field work in 1929 and 1930 the writer restudied the Ordovician rocks of the Marathon Basin and divided the section into five formations.

GENERAL FEATURES

The Ordovician system in the Marathon Basin, with a maximum thickness of about 2,000 feet, is brought to the surface in the Marathon and Dagger Flat anticlinoria (pl. 23). The higher members also appear in anticlines of less height between the anticlinoria and to the south of them. The system is now divided, in ascending order, into the Marathon limestone, the Alsate shale, the Fort Peña formation, the Woods Hollow shale, and the Maravillas chert. The Lower, Middle, and Upper Ordovician are all represented. All the formations contain fossils in greater or less quantity, but collecting is nearly always difficult, as it entails the splitting of great quantities of slaty shales and thin-bedded limestones that bear little or no surface indication of the presence of organic remains. The best fossils are obtainable in finely granular limestone, where there has been little compression of the shells.

The Ordovician section at Marathon is composed of relatively thick beds of shale, muddy limestone, and chert. These beds are intercalated with thinner layers of conglomerate, boulder beds, and sandstone. The faunas are mostly of a specialized facies, with plentiful floating and attached graptolites, associated with linguloid and oboloid brachiopods, pteropods, and trilobites. A very different contemporaneous facies is found in exposures 100 miles to the northwest, where the rocks are nearly all dolomitic limestones (fig. 16) and contain faunas characterized by orthoid brachiopods, cephalopods, corals, and sponges. The differences between the two sections are, however, more apparent than real, for several fossils and a few faunal groups are found in both regions. There is no suggestion that the strata were deposited in separated seaways. It is probable that the more or less clastic Ordovician strata at Marathon were deposited on or near muddy shores, in agitated water, and that the limestones with the gastropod-cephalopod assemblages were deposited farther from shore, in clean and quiet water.

During the course of the present field work 10 stratigraphic sections of Ordovician rocks were measured, of which only 5 extend entirely from the Cambrian to the probable Devonian (pl. 2). At four places in the area the stratigraphic sequence is particularly well exposed and is relatively free from structural complications :

• (1) On the south side of the road to the Roberts ranch, 3 miles west-southwest of old Fort Peña Colorada;
• (2) between Woods Hollow and Little Woods Hollow, on the old Louis Granger place, 6 miles southeast of Marathon;
• (3) on the south side of the Woods Hollow Mountains, 3 miles northeast of "

  ---

  Bridge, Josiah and Dake, C. L., Faunal correlation of Ellenburger limestone of Texas: Geol. Soc. America Bull., vol. 43, pp. 725-748,1932.

  Udden, J. A., Sketch of the geology of the Chisos country: Texas Univ. Bull. 93, pp. 18-20, 1907.

  Baker, C. L., and Bowman, W. F., Exploration of the southeastern front range of trans-Pecos Texas: Texas Univ. Bull. 1753, pp. 79-101, 1917. A preliminary statement, containing the first published descriptions of their formations, is given in Review of the geology of Texas: Texas Univ. Bull. 44, 1st ed., 1916.

  King, P. B., Pre-Carboniferous stratigraphy of the Marathon uplife: Am. Assoc. Petroleum Geologists Bull., vol. 15, pp. 1066-1076, 1931.

 

26

Woods Hollow Tank;
• and (4) on the south side of Dagger Flat, 4 miles northeast of the Buttrill ranch.

These localities were studied with the greatest care, and at each place a detailed section was measured.

MARATHON LIMESTONE

GENERAL FEATURES

The term "Marathon limestone", as here used, is a restriction of the term "†Marathon series" of Baker and Bowman to the limestones and associated rocks that crop out within the town of Marathon. These strata, which are a well-marked unit of Deepkill (Beekmantown) age, are exposed in the streets and vacant lots of the town. The basal beds, resting on the Dagger Flat sandstone, crop out a mile southwest of the railway station, on the north side of the old road to Alpine, and the highest layers, dipping beneath the Alsate shale, occupy a northeastward-trending belt of outcrop that crosses the Boquillas road 1¼ miles south of the town.

Other good exposures are found on the south side of the road to the Roberts ranch, 3 miles west-southwest of old Fort Peña Colorada. Part of the formation is also beautifully revealed in the bed of Alsate Creek, on the opposite side of the road (fig. 12). There are extensive exposures of the formation in the northeast end of the Dagger Flat anticlinorium, northeast of Woods Hollow Tank.

In aerial photographs the outcrops of Marathon limestone are lighter-colored than, those of adjacent formations but are streaked with faint light and dark bands that mark the outcrops of individual beds (pl. 17).

The Marathon limestone generally ranges between 500 and 1,000 feet in thickness but thins to only 350 feet in the southernmost exposures (pl. 2). The most conspicuous parts of the formation are beds of flaggy limestone that weather to an ashen-gray or bluish color, some of which contain graptolites. Partings of shale separate most of the limestone layers (pl. 3, A), and there are a few thick members of greenish clay shale. The argillaceous parts of the formation probably make up one-third or one-half its total thickness. Between the limestones are a few layers of sandstone and many beds of intraformational conglomerate. Near the middle of the formation is the Monument Spring dolomite member, which reaches 90 feet in thickness in the Marathon anticlinorium, where it has a wide and persistent development, but it thins and disappears southeastward in the Dagger Flat anticlinorium. The member is a massive mottled dolomitic limestone that contains fossils like those of the El Paso limestone in the region to the northwest.

LOCAL FEATURES

Marathon anticlinorium

-The Marathon limestone has a wide exposure between the town of Marathon and the Roberts ranch along the crest of the Marathon anticlinorium, where it crops out in nearly level plains or in rolling hills covered with ashen-gray outcrops of its limestone flags. The formation is intricately folded and crumpled, and in places the weaker beds are cut out by squeezing or faulting, so that measurements of thickness are not altogether trustworthy, and thicknesses of individual members are quite different in closely adjacent sections. The competent Monument Spring dolomite member near the middle of the formation is broken and shattered, so that its outcrop is characteristically a chain of disconnected boulders. As a result of deformation the layer has been repeated in a most bewildering manner, and only careful field mapping does away with the impression that there are many beds of this sort in the section instead of one. The limestone flags of the formation in many places are traversed by veins of granular calcite and are cut at oblique angles to the bedding by cleavage planes that are strongly slickensided and coated with small calcite crystals. The interbedded shales are somewhat indurated but are not otherwise altered.

The Monument Spring dolomite member, near the middle of the formation, is named for its exposure half a mile west of Monument Spring, 12 miles southwest of Marathon. Its maximum thickness in the area near Fort Peña Colorada and Alsate Creek is 94 feet, but it thins to 25 feet 8 miles to the southwest, near Monument Spring. It consists of dense mottled dolomitic limestone, breaking with conchoidal fracture, and is constructed of small nodular masses of blue-gray dolomite, closely packed in a yellowish dolomitic matrix. In places it contains small angular calcareous fragments of similar appearance to the matrix. It weathers to light-colored, rounded boulders or disconnected ledges, which appear white from a distance (pl. 3, B). In places the rock is strongly silicified; the yellowish matrix is changed to brown chert, and the bluish nodules are relatively unaltered, so that the rock takes on a ribbed or spongy appearance. In other places the yellowish matrix has yielded more readily to decay, and the nodular portions remain as hard lumps. The exposures then resemble the nodular marls of the Comanche Peak and Georgetown formations in the Cretaceous. Near the top and base of the member the beds are less mottled and thick-bedded and are intercalated with thin layers of compact dark-gray or brown limestone. Fossils are common in parts of the dolomite, "

---

For a discussion and interpretation of limestones of this type see Twenhofel, W. H., Treatise on sedimentation, 2d ed., pp. 334-335, 1932.

 

27

and there are many obscure reef-like structures, in which heads of Cryptozoon and various cuplike sponges such as Calathium are embedded.

The beds above and below the Monument Spring dolomite member are identical in appearance and cannot be distinguished in the field. They consist of flaggy limestones with much interbedded shale (pl. 3, A) and scattered layers of conglomerate and sandstone. The flaggy limestones are dense, are gray or black, and weather to ashen-gray or bluish surfaces. They break with conchoidal fracture. Shale partings occur between most of the layers. Some of the dense limestones contain seams or lenses which are spotted with bituminous matter and which consist of mats of tangled graptolite stipes and crushed linguloid and pteropod shells. There are a few thin layers of globular silicified sponges (fig. 11, A), and the weathered surfaces of some of the beds are thickly strewn with the separated spicules of hexactinellids. Some of the dense limestones are traversed by vertical cylindrical tubules filled with calcite, probably the traces of algae, and some are penetrated by small tubes and pockets of comminuted organic material (fig. 11, B), perhaps the excrement of some burrowing animal. The limestone bedding surfaces show channel markings and small pockets filled by cross-bedded sand. In the upper part, on Alsate Creek, thin argillaceous limestones have well-developed mud cracks (fig. 11, C). Near Monument Spring some of the lower limestones contain nodules of black chert and reddish silicified seams.

Each section of the formation contains five or six layers of intraformational conglomerate 1 to 5 feet thick. Locally these beds reach 8 or 10 feet in thickness but thin gut along the strike. They are composed of shattered limestone flags, turned this way and that and cemented by granular limestone. The flags resemble the limestones that lie beneath the conglomerate layers and were probably derived from them. Some of the conglomercate layers also contain more distantly derived, well-rounded pebbles of chert, limestone, and vein quartz.

 

28

The best section in the area is that on the south side of the road to the Roberts ranch, 6 miles southwest of Marathon, where the formation dips to the southeast with only minor duplications by folding. The following section (sec. 4, pl. 2) is a composite of several closely adjacent measurements on the hills south of the road, in which thicknesses have been averaged so as to give as nearly an accurate view of the sequence as possible. The areal relations of the locality are shown on plate 16.

Nearby, on the east side of the road, the upper part of the same sequence is clearly exposed in the bed of Alsate Creek (fig. 12 and pl. 3, A; for location see pl. 16). The following section was measured at this place by Josiah Bridge and C. L. Dake and is reproduced from their field notes. It illustrates the complexity of the structure revealed by good exposures of the formation.

A well drilled in 1935 by King & Franklin near the old Louis Granger place, on the anticline between Woods Hollow and Little Woods Hollow, penetrated

 

29

over 1,000 feet of Marathon limestone (pl. 2). The following is a condensed record of the boring, based on a study of samples by C. L. Baker and Mrs. D. O. Carsey. The correlations are by Mr. Baker:

The log thus indicates a thickness of 1,148 feet of beds below the assumed top of the Marathon limestone, which is a greater figure than any measured thickness of the formation in the area. No strata of Dagger Flat type were encountered, and it is quite probable that the lower beds in the boring are either greatly thickened by squeezing of shales from the flanks into the crest of the anticline, or that their apparent thickness is caused by their steep dip.

Dagger Flat anticlinorium

-In the Dagger Flat anticlinorium measured sections of the Marathon limestone range between 950 feet in the northeast and 350 feet near the Buttrill ranch, in the south (pl. 2). The Monument Spring dolomite member is only a few feet thick in the northern part of the area and is absent on the Buttrill ranch. Its exposures are broken and not persistent, as a result of shattering during deformation. Along the south side of the Woods Hollow Mountains, where the formation is 950 feet thick, the member is 450 feet below the top. Higher up, 350 feet below the top, is a 6-foot bed of black or purple granular thin-bedded chert. Between this and the dolomite member are numerous layers of brown sandstone and several beds of indurated greenish siliceous shale. Near Garden Springs the formation contains a few ledges of light-gray limestone 2 or 3 feet thick.

On the south side of Dagger Flat, near the Buttrill ranch, the Marathon limestone is intensely folded and occupies a narrower belt of outcrop than elsewhere. This may be due to a thinning out of the sediments but possibly also results from squeezing during deformation. Most of the formation here is indurated shale, and there are only a few thin flags of limestone. The limestone increases in prominence in successive strips of outcrop to the north. In the basal part there is much dull-lustered splintery black chert in thin layers. There are also three or four massive conglomerate beds 1 to 2 feet thick.

The following section illustrates the character of the formation in this region (sec. 12, pl. 17). The intense folding of the rocks at the locality may have sheared out some of the beds and makes the determined thickness approximate at best.

 

30

The first reported discovery of the fauna of the Marathon limestone was made at Threemile Hill by Baker and Bowman. This collection appears to have come from thin slices along a thrust plane between the Woods Hollow shale and the Dagger Flat sandstone (fig. 10). Except for a few loose blocks, no typical limestones of the Marathon formation were found at the locality by the writer, and the Fort Pena and Alsate formations are wanting entirely. Most of the limestone assigned to the Lower Ordovician in Baker and Bowman's original section is earthy limestone intercalated in the lower part of the Woods Hollow shale. Dicellograptus has been collected from it by the writer.

FOSSILS AND AGE

The faunas of the upper and lower members of the Marathon limestone are similar in character, although they represent distinct zones in the Deep Kill section of New York. The most common graptolite genera are Tetragraptus, Phyllograptus, and Didymograptus, but at some localities Goniograptus and Loganograptus are also found. Various dendroid graptolites, including Dictyonema, are found throughout both the lower and upper members. Associated with the graptolites are many linguloid and oboloid brachiopods, a few pteropods and trilobites, and in some zones many sponges. As these faunas are to be correlated with part at least of graptolite beds 1 to 5 of the section at Deep Kill, they belong to the Beekmantown.

Large collections of graptolites have been made by the writer from the lower and upper members of the Marathon limestone, but so far only a few of these have received detailed study. The following are lists of graptolites and associated fossils from the region that have been specifically identified:

• 1. About 8 miles southwest of Marathon on road to Roberts ranch, below Monument Spring dolomite member, Monument Spring quadrangle; R. E. King, collector; identified by Rudolf Ruedemann:
  • Didymograptus nitidus (Hall). U.S.N.M. 85363.
  • Didymograptus patulus (Hall).
• 2. Thrust slice on north slope of Threemile Hill, Santiago Peak quadrangle; C. L. Baker, collector; identified by E. O. Ulrich:
  • Didymograptus cf. D. extensus (Hall).
  • Tetragraptus aff. T. fruticosus (Hall).
  • Phyllograptus cf. P. ilicifolius Hall.
  • Phyllograptus cf. P. angustifolius Hall.
  • Paterula sp.
  • Acrotreta sp.
• 3. Three-fourths mile northwest of hill 4360 of Peña Blanca Mountains, apparently below Monument Spring dolomite member, near axis of small anticline, Marathon quadrangle; C. L. Baker and J. Hosterman, collectors; identified by Rudolf Ruedemann.
  • Didymograptus cf. D. extensus Hall (fragment).
  • Loganograptus logani (Hall) (fragment).
  • Phyllograptus typus Hall.
  • Phyllograptus anna Hall.
  • Diplograptus (Glyptograptus) dentatus (Brongniart).
• 4. Same horizon and locality as 3; P. B. King, collector; identified by Edwin Kirk: (U. S. Geological Survey locality 2411)
  • Phyllograptus cf. P. ilicifolius Hall.
  • Didymograptus cf. D. nicholsoni var. planes Elles and Wood.
  • Tetragraptus sp.
• 5. Three-fourths mile northwest of bench mark 4015 on Terlingua road, 50 feet above Monument Spring dolomite member, Monument Spring quadrangle; P. B. King, collector; identified by Edwin Kirk (U. S. Geological Survey locality 2412):
  • Acrotreta sp.
  • Didymograptus bifidus (Hall).

The fauna of the Monument Spring dolomite member is of a distinct facies. It contains plentiful masses of Cryptozoon, at least three species of sponges, including Calathium cf. C. formosum, undetermined orthoid brachiopods, and various poorly preserved gastropods. A single cystoid calyx was found, but stem segments occur in the rock in great numbers. Cephalopods are plentiful and well preserved; the most characteristic are Piloceras and the endoceratoid Colpoceras. According to Edwin Kirk, this fauna is nearly identical with that in the lower half of the El Paso limestone at El Paso, but particularly in the beds several hundred feet above its base.

STRATIGRAPHIC RELATIONS

The top of the Marathon limestone in the Marathon anticlinorium is sharply separated from the Alsate shale by coarse conglomerate. In the bed of Alsate Creek the limestone below the conglomerate is channeled and broken, and the conglomerate is deposited in cavities and interstices. In the Dagger Flat anticlinorium the contact is also drawn at conglomerate, but the intercalated limestones above and below it are only slightly different in character.

"

---

Baker, C. L., and Bowman, W. F., op. cit. (Texas Univ. Bull. 1753), p. 89.

Many of the identified species from the Marathon limestone and higher formations of the Ordovician are figured in Sellards, E. H., Pre-Paleozoic and Paleozoic systems, in The geology of Texas, vol. 1, Stratigraphy: Texas Univ. Bull. 3232, pl. 4 and pp. 235-236, 1933.

 

31

ALSATE SHALE

GENERAL FEATURES

The Alsate shale was named for Alsate Creek which joins Peña Colorada Creek from the west at old Fort Peña Colorada. The shale is well exposed in cut in the creek 2½ miles west-southwest of the fort (for location, see pl. 16), and it is widely distributed in narrow convoluted belts of outcrop in both the Marathon and Dagger Flat anticlinoria. In most places the shale occupies a covered interval between the ledges of Marathon and Fort Pena limestones. The Alsate shale is a thin but distinctive formation overlying the normal Deepkill zones of the Marathon formation, set off below and above by conglomerates and characterized by the exotic graptolite genus Oncograptus. The formation consists of two facies. In the northern exposures it is mostly shale, but in the south there are many limestone ledges (pl. 2). The faunas in both areas are similar, and the two facies probably intergrade.

LOCAL FEATURES

Marathon anticlinorium

-Near the type locality the formation consists of 5 to 18 feet of conglomerate, overlain by 20 to 40 feet of shale (fig. 12 and sec. 4, pl. 2). Elsewhere in the Marathon anticlinorium the shales are much thinner, but on the hills north of the road to the Roberts ranch and west of Alsate Creek they are 100 feet or more in thickness. As this area is intensely folded and is broken by many small thrust faults (sec. C-C; pl. 16), some of the thickening may be of tectonic origin. The formation here appears to be thickest along anticlinal and synclinal axes, and the folding in the competent Fort Pena formation above is more open than in the Marathon limestone below. This difference suggests that the shales acted as a plastic cushion between the two more competent members, in much the same manner as the Woods Hollow shale, higher in the section.

The conglomerate at the base of the formation is massive and lenticular. It is made up of rounded or subrounded fragments of limestone and chert in a matrix of buff sandy limestone. In the bed of Alsate Creek the conglomerate rests on a channeled surface of flaggy limestone of the Marathon formation. The overlying shale is indurated and greenish and breaks into small hard chips. It contains some lenses of black chert. In many places there are thin nodular beds of dense gray or greenish limestone, weathering yellow and locally containing fossils. The northern exposures include also lenses of gray or buff saccharoidal to 4 quartzitic quartz sandstone. In some exposures are iron-stone nodules that weather to brown limonite but on fresh surfaces consist of radiating marcasite crystals.

Along several of the narrow anticlinal axes 1 to 1½ miles southeast of Marathon there are, in the lower part of the shales, several ledges of granular laminated e limestone with brown siliceous seams. These contain tain Oncograptus and resemble lithologically the limestones of the Alsate formation in the Dagger Flat anticlinorium.

A thickness of 351 feet of hard dark-gray shale was penetrated below the Fort Pñena formation by the King & Franklin well in the Woods Hollow Mountains, and part or all of it may belong to the Alsate shale.

Dagger Flat anticlinorium

.-In the Dagger Flat anticlinorium the formation is from 100 to 145 feet thick (sec. 8, pl. 2). Here ledges of limestone like those in the Fort Pena formation are a conspicuous part of the formation, but each is separated from the next by 3 to 8 feet of poorly exposed shale. Most of the limestones form 6-inch to 1;2-foot beds, are gray, and are finely granular or sandy. In part they are thinly laminated and are seamed by brown siliceous matter. Some are cross-bedded. Some of the granular layers contain abundant graptolites, including Oncograptus, and oboloid and linguloid brachiopods. There are also thinner layers of dense gray limestone, in part laminated, weathering pale buff and yellow. The formation is locally separated from the underlying Marathon limestone by thin layers of conglomerate.

Relation of the Marathon and Dagger Flat areas

- The rocks of the Alsate formation in the Marathon anticlinorium are of somewhat different character from those in the Dagger Flat anticlinorium. Most of the Marathon area has none of the finely granular laminated limestones that are conspicuous in the Dagger Flat area, and no specimens of Oncograptus have been collected there. However, in both areas the rocks contain graptolites and other fossils of late Deepkill age. Moreover, in the eastern part of the Marathon anticlinorium, southeast of Marathon, a few ledges of granular limestone with Oncograptus have been found near the base of the Alsate shale, so that it is probable that the sets of strata in the two areas are the lateral facies of a single formation.

MICROSCOPIC CHARACTER

A thin section of saccharoidal sandstone from a lens in the Alsate shale north of the road to the Roberts ranch southwest of Alsate Creek consisted of rounded to angular quartz grains of medium size with a few small grains of feldspar, calcite, and chlorite, in a matrix of crystalline calcite. The rock contains a few "

---

Alsate Creek is named for Alsate, chief of the Chisos Apache Indians, who once occupied this territory. Kokernot Spring, near Alpine, was once called Charco de Alsate. See Raht, C. G., The romance of Davis Mountains and Big Bend country, pp. 166-170, 273-279, El Paso, Rahtbooks Co., 1919. According to M. B. Arick (Early Paleozoic unconformities in trans-Pecos Texas: Texas Univ. Bull. 3501, p. 119, 1936) the name is more properly spelled Alcate, but the writer has not been able to verify this spelling in any reference works at his disposal.

 

32

curved fragments of shells. The quartz grains show strain shadows and some shattering.

FOSSILS AND AGE

The limestones of the Alsate formation in the Dagger Flat anticlinorium contain graptolites in thin zones nearly everywhere including the remarkable genus Oncograptus, hitherto known only in Australia, British Columbia, and Idaho. Ruedemann interprets the genus as an aberrant offshoot of Didymograptus in the Pacific realm at a horizon probably corresponding t graptolite beds 3 to 7 of the Deep Kill section. Four collections from outcrops lying between Woods Hollow Tank and a point 3 miles to the northeast are typical of the formation in this region; the Oncograptus was identified by Ruedemann, and the remaining fossils by Kirk:

• Oncograptus upsilonT. S. Hall.
• Didymograptus cf. D. extensus (Hall).
• Didymograptus sp. (with large stipes).
• Tetragraptus sp.
• Various dictyonemids.

In the Alsate Creek area fossils are found in some o the limestone nodules. A collection from a locality northwest of the road to the Roberts ranch and 1 mile west-southwest of point 4240, identified by Kirk, is typical of the fossils in this region:

• Tetragraptus sp.
• Didymograptus sp.
• Phyllograptus sp.
• Orthis (small brachiopod probably referable to this genus in a restricted sense).
• Maclurites sp. (apparently identical with a new species occurring high in the Beekmantown member of the Pogonip limestone in Nevada).
• Seleneceme sp. (small trilobite).
• Large trilobites (fragments).

The Alsate graptolites are of late Deepkill age, and the Maclurites indicates that the formation is also to be correlated with some part of the latest Beekmantown.

STRATIGRAPHIC RELATIONS

The Alsate shale is everywhere separated from the overlying Fort Peña formation by layers of conglomerate, which are particularly prominent in the Marathon anticlinorium. In this area also the change in lithology from the shales of the Alsate to massive sandy limestones and intercalated conglomerates of the Fort Peña is very striking. From this evidence it is supposed that the two formations are separated by a considerable unconformity, which probably represents the later part of Lower Ordovician time.

FORT PEÑA FORMATION

GENERAL FEATURES

The Fort Peña formation is the chief ridge maker in the Paleozoic succession below the novaculite and rises in low hogbacks out of the generally level country of the Marathon and Dagger Flat anticlinoria. The type locality is on one of the hogbacks directly north of old Fort Peña Colorada (pl. 24). On aerial photographs the outcrops of the Fort Pena formation stand out as a narrow band that is darker than the outcrops of adjacent formations (pl. 17).

Superficially the formation somewhat resembles the Dimple Dimple limestone, of Pennsylvanian age, for which in 1 some places it has been mistaken. The formation ranges from 125 to 200 feet in thickness and consists mostly of alternations of thick-bedded limestone, in part sandy, with bedded bluish and purplish chert (pl. 2). There are some thin partings of shale, and near the base one or more beds of coarse conglomerate. The limestones contain a few fossils.

LOCAL FEATURES

Marathon anticlinorium

-The Fort Peña formation in the Marathon anticlinorium between Marathon and the Roberts ranch stands in low but prominent ridges. One of these parallels the novaculite hogback on the southeast flank of the anticlinorium for long distances (pl. 24) and stands but prominently north of old Fort Pena Colorada, the type locality. Farther north the formation crops out in hills of synclinal structure. In the Marathon anticlinorium the base of the formation is marked by a layer of coarse conglomerate 5 to 15 feet thick, of subrounded pebbles and cobbles of chert, limestone, and sandstone, a quarter of an inch to 6 inches in diameter, in a matrix of gray sandy limestone. Above are gray granular sandy limestones. In places there are fine pebbly seams and some partings of indurated drab shale. The shales increase in prominence toward the top. In the upper part are many thin beds of reddish or bluish granular chert. The uppermost limestones change gradually to a drab color, and the shale beds increase in thickness near the contact with the overlying Woods Hollow shale. Graptolites and oboloid brachiopods are found in some of the granular limestone layers, and some graptolites have been collected from the uppermost interbedded shales on Alsate Creek (pl. 3, D).

At a locality "1½ miles south of Marathon, where the dip is about 75° S. 55° E.," Baker and Bowman found an exposure of Dimple limestone containing Pennsylvanian fossils. Their description corresponds to exposures of the Fort Peña formation which cross the road at this point and which contain Ordovician graptolites and brachiopods. The Pennsylvanian fossils "

---

Ruedemann, Rudolf, Fossil evidence of the existence of a Pacific Ocean in early Ordovician time [abstract]: Geol. Soc. America Bull., vol. 39, pp. 299-300, 1928. For a valuable discussion of the anatomy and phylogeny of this genus, based on specimens from the Marathon region, see Bulman, O. M. B., The structure of Oncograptus T. S. Hall: Geol. Mag., vol. 73, pp. 271-278, 1936. The specimens certainly did not come from the El Paso limestone, as stated by this author, and are probably from the Alsate shale.

Baker, C. L., and Bowman, W. F., op. cit. (Texas Univ. Bull. 1753), p. 106.

 

33

reported from this place may have come from a loose block on the surface.

The following section (sec. 4, pl. 2) was measured 3 miles west-southwest of Fort Peña Colorada on the south side of the road to the Roberts ranch (see pl. 16 for location):

Dagger Flat anticlinorium

-In the Dagger Flat anticlinorium the Fort Peña formation is divisible into several members, which are well developed near Garden Springs and Woods Hollow Tank. At the base is a thin conglomerate, not everywhere present, of large and small limestone and chert pebbles, with some angular flags closely set in a granular sandy gray-brown limestone matrix (member Aof sec. 8, pl. 2). These are followed by (B) about 75 feet of thick-bedded gray-brown granular limestone, with some thin-bedded chert. The succeeding beds (C), about 15 feet thick, consist mainly of shales and occupy a sag in the hogback ridge. Some dense, thinly laminated limestone, weathering yellow, is interbedded in the shale, as well as a few layers of gray granular fossiliferous limestone. The upper member (D) consists of bluish-gray to reddish, very massive chert in several 3- to 4-foot ledges, weathering into sharp splinters. It is succeeded by thin limestones and interbedded shales that grade into the Woods Hollow shale.

The following section (sec. 8, pl. 2) is typical of the sequence in the Dagger Flat anticlinorium:

Jones ranch area

-In the southeastern part of the Marathon Paleozoic area, southeast of the Jones ranch, in the Dove Mountain quadrangle (pl. 23), are strata that are tentatively correlated with the Fort Peña and Woods Hollow formations. They crop out to the south of and apparently up the dip from exposures of Maravillas and Caballos. The lower rocks consist of fine-grained pink quartzite and brown limestone with embedded grits of chert and quartz. There is also some flaggy dark-gray granular limestone in beds a few inches thick, interbedded with indurated green shale. These strata are overlain by the equivalent of the Woods Hollow shale, which here consists of greenish shale and sandy shale, with some indurated sandy flagstones.

MICROSCOPIC CHARACTER

Two thin sections of cherts from the Fort Peña formation in the Marathon anticlinorium show similar characteristics (fig. 15, C). The rocks are laminated

 

34

and consist of a fine-textured mat of radiating fibers of chalcedony. Embedded in the matrix are large and small rhombic crystals of calcite and scattered rounded grains of quartz. There are also numerous hollow spicules, probably of sponges. Most of these are composed of crystalline calcite, but a few have been wholly or partly altered to chalcedony. In places the fibers of the chalcedonic matrix radiate from the embedded grains. A few minute veins of calcite traverse the rocks.

FOSSILS AND AGE

With the exception of fossils in a few thin seams in the limestones, the Fort Peña formation contains little evidence of life. From limestone beds beneath the upper chert member near Garden Springs were collected Diplograptus, Ceraurus, Bucania, and a rafinesquinoid probably allied to Ptychoglyptus. Southwest of Marathon, in the Marathon anticlinorium, Diplograptus and Climacograptus are plentiful. In some of the more coarsely granular limestones they are not compressed, and all the minute structures of the stipes are beautifully preserved. Associated with them are oboloid and linguloid brachiopods and a small Orthis of the type 0. tricenaria. In several of the collections southwest of Marathon there are also large stipes, seemingly of Didymograptus, and specimens of a recurved Tetragraptus.

Most of this fauna is suggestive of the Black River, but the occurrence here and there of the two primitive genera last named, reminiscent of the fossils collected by Ruedemann at the Ashill quarry, near Mount Moreno, N. Y., suggests that the formation is older and possibly Chazyan. The field relations of the Fort Pena formation suggest that it is of Middle rather than Lower Ordovician age, as its massive sandy limestones rest with coarse basal conglomerate on dissimilar Lower Ordovician strata and appear to grade up into the Woods Hollow shale.

STRATIGRAPHIC RELATIONS

The Fort Peña formation apparently passes conformably upward into the Woods Hollow shale. The massive cherty gray limestones of the formation change near the top into drab, thinly laminated limestones, with much interbedded shale, and near the top into shales with a few flaggy limestone layers. The fossils also suggest that the rocks are not greatly different in age.

WOODS HOLLOW SHALE

GENERAL FEATURES

The Woods Hollow shale is best exposed between Woods Hollow and Little Woods Hollow, where it crops out in an anticlinal valley on the former Louis Granger ranch (fig. 6, C). The formation nearly everywhere is worn down to a valley between the hogbacks of the Fort Peña formation and the higher ridges of the Maravillas chert and Caballos novaculite. In most places it is covered by soil and crops out only in gullies and creek banks. The formation is the same as Baker and Bowman's members 4 and 5 of their Marathon series. The strata at Garden Springs, Horse Mountain, and Fort Peña Colorada mentioned by them are also a part of the formation. The differences in lithology noted by them at these localities are at most the local facies of a single stratigraphic unit.

The Woods Hollow shale consists of greenish clay shales, with interbedded thinly laminated gray or yellowish sandy limestone and limy sandstone (pl.3, C). There are also some beds of nodular coarsely granular conglomeratic limestone, crowded with fragmental fossils. In several localities in the southwestern part of the Marathon Basin the shales include large embedded boulders that contain Cambrian and Lower Ordovician fossils (p. 24). The formation has a thickness of 300 or 400 feet (pl. 2).

The Woods Hollow shale has a variable width of outcrop (pl. 24). In places it forms a narrow strip, but elsewhere it crops out over wide areas. In the northeastern part of the Dagger Flat anticlinorium it occupies an area of several square miles, but the strata here have probably been piled up and repeated by the squeezing out of beds on the nearby flanks of the fold. The strong differences in structure between the Fort Pena and Marathon formations below and the Caballos and Maravillas formations above suggest that the incompetent shales of the Woods Hollow formation have had a cushioning effect, which has allowed the two groups of strata to be deformed in a different manner. The complexity of these larger structural features is reflected by the intense contortion of the shales in all the local exposures of the formation. North of Peñna Blanca Spring a cut in a creek bed shows shales and flagstones dipping steeply southward but standing in isoclinal folds. Exposures in a small arroyo near Monument Spring contain limestone beds that have been broken and rolled and their ends rounded. The limestone fragments are embedded in shales, which have been kneaded and contorted around them. In spite of this deformation, the shales are only slightly indurated and the formation is not metamorphosed.

LOCAL FEATURES

Woods Hollow Mountains

-At the type locality the Woods Hollow shale, 470 feet thick, is clearly revealed in an anticlinal basin and dips in regular order off the highest Fort Peña limestones, which crop out in the "

---

Ruedemann, Rudolf, Graptolites of New York, part 1: New York State Mus. Mem. 7, pp. 499-500, 1907.

 

35

center (sec. D-D'-D", pl.21). The lower part of the formation consists of flaggy, thinly laminated gray or yellowish sandy limestone or calcareous sandstone, with shale partings. This grades up into greenish clay shale with a few interbedded flaggy limestone layers (pl.3, C). Some of the sandy beds are rill-marked on bedding surfaces, and some of the lower flaggy layers contain graptolites. There are five or six nodular beds of coarsely granular and conglomeratic limestone, weathering yellowish, which are crowded with comminuted remains of bryozoans, trilobites, brachiopods, and crinoids. There are a few beds of black shale near the top of the formation.

The following section (sec. 7, pl. 2) was measured by the writer. Some additional notes by C. L. Baker have been incorporated in the section.

Fort Peña formation: 1. Dark-gray, finely granular, thinly laminated limestone, in beds a few inches to several feet thick, interbedded with some shale and fine sandstone. The flaggy layers contain Diplograptus. Forms low anticlinal ridge in middle of basin 30

Base of section not exposed. Lower beds have been penetrated in King & Franklin well. (See p. 29.)

Simpson Springs and East Bourland Mountains

- The shales, sandy and limy flagstones, and conglomeratic limestones that lie beneath the Maravillas chert on Simpson Springs and East Bourland Mountains are of Woods Hollow age, but at these places, as well as near Garden Springs, they contain rounded masses of hard gray limestone as much as 3 feet in diameter, in part mottled and in part glauconitic (pls. 5, E, and 6, B). The limestone masses lie in soft clay shales and flaggy sandstones in the upper part of the formation. Cambrian and early Ordovician fossils have been collected from the limestones by both C. L. Baker and the writer, and the interbedded strata contain Middle Ordovician fossils. There is thus no doubt that the limestone masses are boulders which have been transported to the positions they now occupy. The lower part of the exposures on Simpson Springs Mountain was named the †"Brewster formation" by Baker and Bowman and considered to be of Cambrian age, but this interpretation is now discarded.

Other localities

-A few miles southeast of Marathon, near the junction of the Boquillas and Terlingua roads, and also near Peña Blanca Spring, a 30-foot member of sandstone crops out prominently in the lower part of the formation. The sandstone is in 1- to 2-inch layers, with a few thicker beds and is mostly thinly laminated, ferruginous, and quartzitic. It weathers to small angular iron-stained chips and blocks. On the south side of Dagger Flat, in the upper part of the formation, many layers of dense gray laminated limestones are interbedded in the shale. They weather to pale-buff earthy surfaces and form rounded ledges. In Baker and Bowman's section on Threemile Hill these beds were apparently grouped with rocks containing Deepkill graptolites. However, the writer has collected Dicellograptus from them at this locality.

MICROSCOPIC CHARACTER

A thin section of sandstone from the lower part of the Woods Hollow shale in the Marathon anticlinorium southeast of Marathon was examined. The sand grains are arranged in laminae a few millimeters thick, which stand out as dark and light bands in the hand specimen. The light layers are composed almost entirely of quartz, most of whose subangular grains are interlocked, though there are small amounts of interstitial crystalline calcite. Some of the quartz grains show strain shadows. There are a few minute grains of minerals with a high refractive index. In the dark layers the quartz grains are more dispersed and are set in an opaque argillaceous and ferruginous matrix.

FOSSILS AND AGE

Fossils are plentiful in the Woods Hollow shale but, are nearly all comminuted and poorly preserved.. Diplograptus is common in the lower flags at the type locality. Higher up, in the more granular limestones, are ramose and massive bryozoans, including Phaenopora, Nicholsonella, Rhinidictya, and Anolotichia; the trilobites Illaenus and Asaphus; the mollusks Modiolopsis and Holopea; and the brachiopod Orthis near 0. tricenaria. Along the road between Ridge Spring

 

36

and Garden Springs were collected Dicellograptus, Diplograptus, Glossograptus, Ceraurus, and a large Hormotoma. Collections on East Bourland Mountain, not far from limestone boulders of Cambrian and early Ordovician age, contain Diplograptus and numerous Glossograptus. F. H. Bush and B. H. Harlton have informed the writer that microfossils obtained by them from this same locality are like those of Black River age in Oklahoma. At a locality 3¾ miles northeast of the Roberts ranch, about 100 feet below the Maravillas chert, Sidney Powers collected a graptolite identified by Ruedemann as Glossograptus echinatus Ruedemann (U.S.N.M. 85371).

Two miles southwest of the Lightning ranch, Boüse collected the following fossils, which were identified by E. O. Ulrich:

• Anolotichia aff. A. revalensis Bassler.
• Nicholsonells sp. ( ramose, branches slender).
• Phaenopora cf. P. incipiens Ulrich.
• Stichtoporella cf. S. exigua Ulrich.
• Rhinidictya sp.
• Sowerbyella ("Plectambonites") aff. S. quinquecostata (McCoy).
• Eurychlinia sp.
• Aparchites sp.

Most of the fossils in the Woods Hollow shale seem clearly to be of Middle Ordovician age and suggest that it is to be correlated with the Trenton. Some of the graptolites, however, such as Glossograptus echinatus, suggest a correlation with the Normanskill (Chazy), so that there is a possibility that the formation is older than Trenton. For the present the formation is classified as of Middle Ordovician age.

STRATIGRAPHIC RELATIONS

The Woods Hollow shale is separated from the overlying Maravillas chert by a sharp lithologic break, which represents a change from shale and sandy limestone deposition to the deposition of chert and massive limestone. At some localities, as near the junction of the Boquillas and Terlingua roads south of Marathon, the shales and cherts are interbedded for a few inches below the contact. At some other places, particularly at Monument Spring and Rock House Gap, at the southwest end of the Marathon anticlinorium, there is at the base of the Maravillas a coarse conglomerate containing fragments of Woods Hollow limestones and older rocks, which indicates at least a local uplift and time of erosion between the Woods Hollow and Maravillas epochs of deposition.

Such a break, however, is not as great as has previously been supposed. At no place is the Woods Hollow formation truncated by the Maravillas or even markedly reduced in thickness. The previous supposition that the Maravillas rests on the Cambrian in Simpson Springs and East Bourland Mountains has been shown to be erroneous, for the Cambrian fossils at these localities are in transported fragments of limestone. The only locality where the Maravillas chert lies in contact with rocks older than the Woods Hollow shale is at Maravillas Gap, in the Santiago Peak quadrangle, where it rests on the Dagger Flat sandstone. This relation is thought to have resulted from tectonic action. Differences in dip between the Maravillas and Woods Hollow formations are likewise probably of tectonic origin and result from the gliding of competent over incompetent beds during deformation. At Peña Blanca Spring and the picnic grounds at Fort Peña Colorada the competent Maravillas chert lies on much crumpled soft Woods Hollow shale. At these places the shale is traversed by large veins of fibrous calcite. which probably have filled tension fissures.

MARAVILLAS CHERT

GENERAL FEATURES

The Maravillas chert was named by Baker and Bowman for exposures at Maravillas Gap, 20 miles south of Marathon, in the Santiago Peak quadrangle. Excellent exposures are also found on the northwest slope of Threemile Hill, 3¼ miles northeast of Maravillas Gap, and near Fort Peña Colorada, east of the picnic grounds (pl. 5, A and B). As originally described, the formation was said by Baker and Bowman to reach 800 feet in thickness and to contain rocks of both Trenton and Richmond ages. However, on a reexamination of the exposures by the writer in company with Baker, it was concluded that the measurements of thickness were not corrected for local duplication, and Edwin Kirk believes, after a new study of the faunas, that the formation is entirely of Upper Ordovician age. The name "Maravillas" is therefore applied here, with a different age interpretation, to the same strata for which it was used by Baker and Bowman.

The Maravillas chert ranges from 100 to 200 feet in thickness in the Marathon anticlinorium but thickens southward to 400 feet (pl. 2). It crops out on the inner slopes of the novaculite hogbacks and consists of interbedded limestone and black bedded chert (pl. 4, C), the chert predominating toward the top. In the northwestern part of the area it has a thick coarse basal conglomerate, and there are other conglomerate beds higher in the section. Along the southeast flank of the Dagger Flat anticlinorium, limestones near the middle of the formation contain reef-like aggregates of bryozoans.

"

---

Baker, C. L., and Bowman, W. F.. op. cit. (Texas Univ. Bull. 1753), p. 85.

Baker, C. L., and Bowman, W. F., op. cit., p. 87.

 

37

LOCAL FEATURES

Marathon anticlinorium

-In the northeastern part of the Marathon anticlinorium the lower or calcareous part of the formation is sharply set off from the upper or cherty part. In the region south of the town of Marathon, near the junction of the Boquillas and Terlingua roads, the lower division is 90 feet thick. In this region it consists of thick ledges of finely granular dark-gray limestone with thin interbedded layers of brown and gray chert. These are interbedded with thin-bedded dense bituminous limestones, which give off a fetid odor when struck. Graptolites are very plentiful on the bedding surfaces of the denser limestones. The granular limestones contain brachiopods and trilobites. In the exposures near Marathon there are, in the lower beds, three or four lenticular conglomerate layers, 4 feet or less in thickness, of limestone and chert 'pebbles in a granular-limestone matrix. Rarely also the conglomerates contain well-rounded pebbles and grits of translucent quartz. Most of them contain brachiopods, cup corals, and colonial corals bearing marks of attrition, but of approximately the same age as the fossils in the associated beds.

The upper part of the formation south of Marathon is about 170 feet thick. It consists of little else than black, dull-lustered, bedded chert. This is mostly in thin beds, but there are some massive layers 3 or 4 feet thick which are aggregates of pillowlike nodules (fig. 13, B). On weathered surfaces the chert breaks into small blocks, similar to lumps of coal. There are a few thin interbedded layers of bituminous limestone, and many of the chert layers are separated by shale partings. About two-thirds of the way up in the formation is a persistent layer of buff siliceous shale 5 feet thick, which makes a narrow bench on the hillsides that can be followed for long distances by the eye.

The sequence described above is well exposed east of the picnic grounds in the gap south of Fort Peña Colorada. The Maravillas chert here stands in a cliff as a result of the undercutting of Peña Colorada Creek, so that nearly every bed in the formation is exposed (pl. 5, A and B, and fig. 14). The following section (sec. 5, pl. 17) was measured at this place:

 

38

Monument Spring and Rock House Gap

-In the southwestern part of the Marathon anticlinorium, near Monument Spring and Rock House Gap, the lower calcareous beds that are well developed to the northeast are nearly all replaced by conglomerate (sees. 2 to 6, pl. 2). At the base of the section is 25 feet of conglomerate, composed of rounded to subangular fragments as much as 3 feet across in a matrix of gray sandy limestone. The fragments consist of various older rocks of the region. They include buff laminated limestone from the Woods Hollow shale, red granular chert from the Fort Pena formation, obscurely mottled limestone similar to the Monument Spring member of the Marathon limestone, and quartzitic sandstone with angular

 

39

quartz fragments from the Dagger Flat sandstone. The conglomerate also contains small angular pebbles of black chert, like that in the Maravillas itself, and blocks of fine-grained calcareous gray sandstone of an unknown horizon. Near Rock House Gap a boulder of the calcareous sandstone 5 feet long is embedded in the conglomerate (pl.5, B). At both localities the basal conglomerate contains abundant abraded corals and brachiopods. At Monument Spring fine conglomerates in the upper part of the lower member are interbedded with lenses and layers of black chert a few inches thick, which are clearly not transported and are probably not due to replacement of some previous rock (fig. 13, A). It is likely that they are chert deposits of primary origin.

The limestone layers above the conglomerates at Rock House Gap are very fossiliferous (pl. 5, A). The dense platy limestones contain orthoid brachiopods and graptolites, and the more granular layers, which form lenses, contain orthoid brachiopods and such trilobites as Cryptolithus. The shale zone near Monument Spring is an ashen-gray soft shale on which the dark-colored imprints of graptolites stand out with sharp contrast. A few feet from the top of the Maravillas at Monument Spring, in hard black shale partings between nodular cherts, a large graptolite fauna was collected.

Dagger Flat anticlinorium

-In the Dagger Flat anticlinorium the conglomerate beds of the Maravillas chert are fewer and thinner, and the differentiation between the lower calcareous and the upper cherty parts of the formation is not sharply marked. Near the middle of the formation along the southeast flank of the anticlinorium there are reef-like aggregates of bryozoans, some of which form massive beds with a maximum thickness of 10 feet. These are absent in the northwest flank of the anticlinorium and farther to the northwest. There are also some massive limestones with very irregular bedding, evidently having a channel structure. They contain a few comminuted fossils. The following section (sec. 13, pl. 2) of the Maravillas chert is characteristic of the southeastern exposures of the formation:

Southeastern exposures

-In the southeasternmost exposures of the formation chert greatly predominates over limestone. In the Santiago Mountains, 8 miles south of Maravillas Gap (Santiago Peak quadrangle), it is nearly all black chert with a few thin lenticular limestone beds. On Rough Creek, in the Dove Mountain quadrangle, it consists of black chert with bands of gray chert in 3-inch to 1-foot beds and with two or three layers of black limestone. In this region there is a few feet of dark-green shale between the uppermost cherts and the base of the Caballos.

MICROSCOPIC AND CHEMICAL CHARACTER

Thin sections of the cherts from the Maravillas formation show them to consist of a fine-textured mass of

 

40

cryptocrystalline quartz. In the more compact types there is no very notable structure. Some of the cherts which in hand specimen show a spotted texture, contain fine fragments of bryozoans, brachiopod shells, and

spicules (fig. 15, B). Many of these are composed of crystalline calcite, but some of them have been partly or wholly altered to chalcedony. One rock from the Buttrill ranch, which in hand specimen is a laminated granular chert, is seen in thin section to contain much quartz sand (fig. 15, A). The sand grains are fine and well rounded and are widely scattered. They are arranged in laminae, separated by layers of chalcedony, in some of which are embedded calcareous spicules. Some of the limestones in the Maravillas chert are phosphatic. No tests were made during the present investigation, but two samples

 

41

collected by Udden near Ridge Spring "were analyzed and one yielded 6.03 and the other 1.70 percent of phosphorus pentoxide. This is clearly in excess of the usual phosphatic content of limestones."

FOSSILS AND AGE

Fossils are plentiful at various levels in the Maravillas chert, but particularly in its lower limestone beds. A few graptolites were obtained in the interbedded limestones of the chert series and in some of the shale partings a few feet from the top of the formation.

The conglomerates at and near the base of the Maravillas contain many abraded fossils, which include Columnaria, Halysites, Paleofavosites,and Streptelasma among the corals and Platystrophiaand Hebertella among the brachiopods (U. S. Geological Survey localities 2417 and 2395). According to Kirk, this fauna is like that in the basal part of the Montoya limestone at El Paso and is of nearly the same age as that in the overlying interbedded limestones.

The limestones, between the conglomerates and overlying them, contain Cryptolithus, Harpes, Strophomena, and various graptolites, including Dicellograptus nicholsoni, Climacograptus antiquus, and Diplograptus amplexicaulis. The most extensive collection from these beds is that of Baker and Bowman at Rock House Gap, in which the following species have been determined by E. O. Ulrich:

• Diplograptuscf. D. amplexicaulis (Hall).
• Lingula sp.
• Leptobolus sp.
• Scenidium cf. S. anthonense Sardeson.
• Strophomena? (two species).
• Sowerbyella ("Plectambonites") aff. S. sericeus (Sowerby).
• Platystrophia n. sp.
• Bythocypris sp.
• Krausella arcuata Ulrich.
• Aparchites sp.
• Eurychilinaaff. E. ventrosa Ulrich.
• Dicranella cf.D. spinosa Ulrich and D. simplex Ulrich.
• Cryptolithus sp.

According to Kirk the genus Cryptolithus occurs in Nevada in beds overlying beds containing a coral fauna like that in the basal Maravillas and is also found in the Viola limestone of Oklahoma.

Several collections of graptolites from the Maravillas chert have been studied by Ruedemann. One of these, obtained by R. E. King three-quarters of a mile east-southeast of bench mark 4361, in the southwestern part of the Hess Canyon quadrangle, contains the following species:

• Glyptograptus amplexicaulis (Hall). (U.S.N.M. 85358.)
• Glyptograptus amplexicaulis var. pertenuis Ruedemann. (U.S.N.M. 85361.)
• Glossograptus quadrimucronatus (Hall).
• Dicellograptus nicholsoni Hopkinson (fragment).
• Climacograptus sp.

At a locality 2 miles to the southeast, about 4 miles northeast of Marathon, R. E. King collected Glossograptus quadrimucronatus var. angustus Ruedemann (U.S.N.M. 85365). A locality about 2 miles south of this one and 2½ miles northeast of Marathon yielded

• Climacograptus antiquusLapworth. (U.S.N.M. 85364.)
• Leptobolus cf. L. walcotti Ruedemann.

The faunas of the first two localities are considered by Ruedemann to be of Trenton or Utica age, but the Climacograptus antiquus is found in New York only in older beds.

The middle part of the Maravillas in the southern part of the Marathon Basin contains lenticular bryozoan reefs, built of such genera as Constellaria and Pachydictya, in which are embedded a few shells of Platystrophia, Rafinesquina, and Hebertella. Kirk has also collected Cryptolithus in the reef beds on Threemile Hill. The following collection by Baker and Bowman from the reef beds at Maravillas Gap was identified by E. O. Ulrich:

• Hallopora aff. H. elegantula (Hall).
• Anaphragma mirabile Ulrich and Bassler?
• Hemiphragma imperfectum (Ulrich).
• Bythopora cf. B. delicatula (Nicholson).
• Crepipora hemispherica Ulrich?
• Rhombotrypa subquadrata (Ulrich).
• Lioclema wilmingtonense Ulrich.
• Constellaria n. sp.
• Favositella cf. F. epidermata (Ulrich).
• Pachydictya sp.
• Rhinidictya sp.
• Eurydictya cf. E. sterlingensis Ulrich.
• Phaenopora wilmingtonensis Ulrich.
• Helopora sp.
• Arthroclema angulare Ulrich.
• Sceptropora facula Ulrich.
• "Dalmanella" aff. D. testudinaria (Dalman).
• Dinorthis cf. D. subquadrata (Hall).
• Glyptorthis insculpta (Hall).
• Platystrophia sp.
• "Plectambonites" elegantula Foerste.
• Sowerbyella ("Plectambonites") cf. S. saxeus (Sardeson).
• Rhynchotrema manniense Foerste.
• Rhynchotremacf. R. anticostiense (Billings).

The higher beds of the Maravillas chert contain little else than undetermined species of Climacograptus and Diplograptus and a few linguloid shells.

The Maravillas faunas, when studied by Ulrich in 1916, were divided, like those of many other western formations (such as the Montoya limestone and Bighorn dolomite), into a lower group, correlated with the "

---

Udden, J. A., Sketch of the geology of the Chisos country, Brewster County, Tex.: Texas Univ. Bull. 93, p. 100, 1907.

Baker, C. L., and Bowman, W. F., op. cit. (Texas Univ. Bull. 1753), p. 91. The locality as given by them is "one-half mile north of Payne's ranch on the east side of Maravillas (Dugout) Creek, 300 yards from the creek." According to Baker (letter, November 1929), the Payne ranch referred to is the locality marked "Rock House" on the Monument Spring topographic map.

Baker, C. L., and Bowman, W. F., op. cit. (Texas Univ. Bull. 1753), pp. 89-90.

 

42

Trenton, and an upper group, correlated with the Richmond. The corals and bryozoans from each locality were considered to be the younger fauna, and the trilobite Cryptolithus and its associated diplograptids were called Trenton. Further field work has shown that the coral-bearing beds are interbedded with and in many places lie beneath those containing the Cryptolithus fauna and that the Cryptolithus fauna in turn extends upward and mingles with the bryozoan fauna. It is therefore clear that the formation is a faunal as well as a lithologic unit.

Kirk has stated his views on the Trenton-Richmond problem in the West in a discussion of the problematical lower Bighorn, as follows:

[The lower Bighorn] has been variously correlated with the * * * Trenton and Ricnmondian. Equivalent beds in Manitoba and the far north have been correlated by Foerste * * * with the Richmond. Several years ago * * * I suggested that this horizon was probably of Cincinnatian age (used in the sense of being pre-Richmondian and post-Trenton), and * * * a considerable amount of corroborative evidence has since come to hand. One of the most striking pieces of evidence has been the finding of Cryptolithus sp. in abundance in the Eureka district, Nevada, and the Marathon Basin, Texas. In the Marathon Basin * * * [the Cryptolithus zone] occurs in a series of limestone carrying a fauna of post-lower Bighorn age * * * Foerste's objections to the assignment of * * * [the lower Bighorn horizon] to the Trenton appear valid to me. I do not see the necessity, however, of skipping the lower Cincinnatian, as he does, and correlating with the Richmond.

Kirk therefore believes that the Maravillas chert is of Upper Ordovician age, probably to be correlated with a part of the Montoya limestone of the El Paso section. There is a strong lithologic resemblance between the cherts and limestones of the Maravillas and the cherty limestones of the Montoya, but the Montoya lacks thick members of bedded chert in which limestone is absent. Like the Montoya, the Maravillas consists of thick-bedded, sparingly cherty limestone in the lower half and very cherty beds above

STRATIGRAPHIC RELATIONS

The contact between the Maravillas and Caballos formations was carefully studied at many places. At one or two places it is marked by conglomerate, but as a rule the vitreous buff, brown, or gray chert of the lower member of the Caballos novaculite overlies the dull coaly-black chert of the Maravillas with sharp contact. Some evidence of post-Maravillas erosion is afforded by the irregular thickness of beds between the base of the Caballos and the "shale zone" of the Maravillas, a circumstance which is further emphasized by the greater constancy in thickness of members below the shale zone.

In 1917 Baker and Bowman noted "a small amount of light-brown arenaceous shale, weathering pinkisih", at a locality 8 miles northeast of the junction of Peña Colorada and Maravillas Creeks. Shale of this description, about 5 feet thick, was seen by Baker at the writer at a place near Maravillas Gap. From similar pinkish shale at the top of the Maravillas at Fort Peña Colorada, Josiah Bridge has collected Ordovician graptolites. At the crest of the anticline on lower Rough Creek and at several other places south of Horse Mountain, the writer has seen thin layers of green slaty shale between the two formatiormations. Sellards has recently observed a similar shale, 25 to 50 feet thick, in the Solitario uplift, southwest of the Marathon region.

These shales may be an argillaceous phase of the Maravillas chert or are perhaps the remnants of a distinct formation laid down in the post-Maravillas and pre-Caballos time interval. If the latter alternative is true, it is possible that formerly they had a wider, extent and have since been removed by pre-Caballos erosion or by tectonic squeezing.

GENERAL PROBLEMS OF ORDOVICIAN STRATIGRAPHY

CORRELATIONS AND REGIONAL RELATIONS

General correlations

-The 2,000 feet of Ordovician rocks of the Marathon region are of Lower, Middle and Upper Ordovician age. Strata representing the Beekmantown (Deepkill), Black River (?), Trenton and Cincinnatian of the New York section are found in the region. It is also suggested that the Fort Pena may be as old as the beds at Mount Moreno N. Y., which Ruedemann considers to be Chazy Field evidence seems to show, however, that then is a considerable hiatus between the Fort Peña and the Alsate, so that the Fort Peñna is more probably Middle Ordovician. Further collecting at this horizon is desirable.

In addition to this probable hiatus, there is evidence of marked uplift and diastrophism in the region near the Middle and Upper Ordovician boundary. There are no observable overlaps, discordances, or channeled contacts, but the Middle Ordovician flagstones, in which are embedded Cambrian and Ordovician exotic limestone blocks, are succeeded by Upper Ordoviciain conglomerates that are peculiar aggregates of boulders cobbles, and abraded fossils, in which some of the fragments are derived from beds as old as the Cambrian. This disturbance is similar to the Taconian movements of Schuchert in the eastern United States and to that represented by a widespread pre-Cincinnatian unconformity studied by Kirk in the Great Basin.

"

---

Kirk, Edwin, Harding sandstone of Colorado: Am. Jour. Sci., 5th ser., vol. 20, pp. 464-465,1930.

Baker, C. L., and Bowman, W. F., op. cit. (Texas Univ. Bull. 1753), p. 89. Bake in a letter to E. O. Ulrich, Oct. 23, 1916, gives the shale locality as Maravillas Gap the place where it was seen by the writer later.

Sellards, E. H., op. cit. (Texas Univ. Bull. 3232), p. 79.

 

43

Correlations in trans-Pecos Texas

-The Marathon Basin section of Ordovician rocks is duplicated in the exposures of the smaller uplift of the Solitario, to the southwest. These have recently been studied by Sellards, and fossil collections from them have been identified by Edwin Kirk and Rudolf Ruedemann. Representatives of the Dagger Flat sandstone, Marathon formation, Woods Hollow shale, and Maravillas chert are recognizable from similarities of lithology and fossils, but the Marathon formation of the Solitario area is more shaly and the Maravillas more cherty than in the Marathon Basin. The Alsate and Fort Pena formations have not been recognized in the Solitario area but probably are present. A layer of shale 25 to 50 feet thick lies between the cherts of the Maravillas and the Caballos novaculite.

The sections at Marathon and in the Solitario area are very different from those exposed farther northwest in trans-Pecos Texas, in the El Paso and Van Horn quadrangles (fig. 16). Here the rocks are nearly all dolomitic limestones which contain orthoid brachiopods, cephalopods, gastropods, corals, and sponges. There is, however, sufficient evidence to indicate general relations. The Dagger Flat sandstone is probably to be correlated with the Bliss sandstone of the El Paso district. The Marathon limestone is correlative with the El Paso limestone, and the Maravillas chert with the Montoya limestone. The Monument Spring dolomite member is the exact replica in lithology and fauna of the middle part of the El Paso limestone, and the Maravillas chert is lithologically similar to the Montoya limestone. The beds intervening between the Marathon and Maravillas are not represented at Van Horn or El Paso.

Correlations with central Texas

-In central Texas no formations of a geosynclinal facies like that at Marathon are exposed or have been penetrated in wells. In the Llano area there is an extensive exposure of Ordovician rocks, the upper part of the Ellenburger limestone, which shows a facies similar to that of the El Paso limestone farther west and like it is of Beeknmantown age. Rocks equivalent to the Marathon limestone, but of different facies, are probably to be found in the 2,000 feet of thickness of these rocks. The Ellenburger limestone has been penetrated by numerous borings in central Texas and has been found in deep wells as far west as Irion, Reagan, and Crockett Counties, within less than 100 miles of the Marathon Basin. Overlying it, in Reagan County, are sandy and shaly strata of Chazy age, which may be correlative with some of the beds above the Marathon limestone in the Marathon Basin.

Correlations with Oklahoma and Arkansas

-The Ordovician formations in the Marathon Basin exhibit some interesting resemblances to Ordovician rocks in various parts of Oklahoma and Arkansas. The rocks of the Ouachita Mountains, which were also deposited in the Llanoria geosyncline, contain similar graptolite faunas, but they are largely slates and sandstones and lack the limestone beds found at Marathon. According to Miser, "the Marathon section represents deposits that were formed somewhat nearer the inner or northwest border of the geosyncline than the Ouachita facies, [which] * * * seems to have been formed closer to the old Paleozoic land of Llanoria." The Deepkill graptolites of the Mazarn shale are almost certainly of the same age as those of the Marathon limestone. The fauna of the Blakely sandstone that resembles that at Mount Moreno, N. Y., is perhaps comparable to the Fort Pena fauna, and there is a suggestion that the Normanskill graptolites of the Womble shale will be found in the Woods Hollow shale. Overlying the Womble formation is the black bedded Bigfork chert, which closely resembles the upper or predominantly cherty part of the Maravillas formation.

There is also some resemblance between the Marathon section and that in the Arbuckle Mountains, although the general facies of the two are different. The Beekmantown portion of the Arbuckle limestone contains mottled dolomitic beds and masses of cryptozoans like those in trans-Pecos Texas in the El Paso limestone and in the thin layer of the Monument Spring dolomite at Marathon. The strata of the Woods Hollow shale have a slight resemblance to some of the shales, sandstones, and limestones of the Middle Ordovician part of the Simpson formation in the Arbuckle Mountains. The platy bituminous limestones in the lower part of the Maravillas chert closely resemble graptolite-bearing limestones in the lower part of the Viola limestone in this same region.

CONDITIONS OF DEPOSITION

Faunal facies of the Marathon Ordovician

-The Ordovician rocks at Marathon are not highly fossiliferous. In the fossiliferous layers the variety of species "

---

Sellards, E. H., op. cit., pp. 76-79, fig. 9, p. 119.

Richardson, G. B., U. S. Geol. Survey Geol. Atlas, El Paso folio (no. 166), pp. 3-4, 1909; Van Horn folio (no. 194), pp. 4-5, 1914. Further descriptions of these formations, based on recent observations by P. B. King, are given by E. H. Sellards (op. cit., pp. 74-75).

Dake, C. L., and Bridge, Josiah, Faunal correlation of the Ellenburger limestone of Texas: Geol. Sec. America Bull., vo.. 43, pp. 725-748, 1932.

Sellards, E. H., op. cit. (Texas Univ. Bull. 3232), pp. 80-81.

Sellards, E. H., Bybee, H. P., and Hemphill, H. A., Producing horizons in the Big Lake oil field, Reagan County, Tex.: Texas Univ. Bull. 3001, pp. 149-203, 1930.

Miser, H. D., in discussion of King, P. B., Pre-Carboniferous stratigraphy of Marathon uplift: Am. Assoc. Petroleum Geologists Bull., vol. 15, p. 1083, 1931.

Miser, H. D., and Purdue, A. H., Geology of the De Queen and Caddo Gap quadrangles, Arkansas: U. S. Geol. Survey Bull. 808, pp. 24-42, 1929.

The literature on the Ordovician stratigraphy of the Arbuckle Mountains is extensive. Among recent papers see Decker, C. E., and Merritt, C. A., Physical characteristics of the Arbuckle limestone: Oklahoma Geol. Survey Circ. 15, 1928; The stratigraphy and physical characteristics of the Simpson group: Oklahoma Geol. Survey Bull. 55, 1931. Decker, C. F., Viola limestone, primarily of Arbuckle and Wichita Mountain regions, Oklahoma: Am. Assoc. Petroleum Geologists Bull., vol. 17, pp. 1405-1435,1933.

 

44

is not large, and there is a constant association of certain groups in such a manner as to produce two well-marked faunal facies.

From the base to the top of the Ordovician section at Marathon there are numerous layers that contain floating and attached graptolites. Associated with the graptolites are linguloid and oboloid brachiopods, pteropods, and certain specialized types of trilobites (Cryptolithus and others). Fossils of this sort are found in shales and in dense, thinly laminated bituminous limestones. The graptolite beds at Marathon are more calcareous than those of other regions, but their persistence through the whole of the system confirms Ruedemann's axiom that "graptolite shales, as a rule, are deposited in the same region for longer intervals than most other fossiliferous rocks."

Associated with the layers containing graptolites are a lesser number that contain corals, sponges, bryozoans, orthoid brachiopods, cephalopods, and gastropods. Some of the layers with these fossils, such as the Monument Spring dolomite and the bryozoan beds in the Maravillas chert, are thick and persistent and are sharply set off from the graptolite beds above and below. Other layers are thinner. In the Maravillas and Woods Hollow formations the fossils of this facies occur in lenses and thin beds of granular and even pebbly light-colored limestone. These are interbedded with thinly laminated bituminous graptolite-bearing limestone, but the fossils of the two facies seldom occur in the same beds.

Conditions during the time of deposition of the graptolite beds do not appear to have been very favorable to life, for the fossils consist only of a few specialized groups. Some, such as the graptolites and pteropods, appear to have been adapted to a planktonic habitat. Others, such as the linguloid brachiopods and the trilobites, appear to have been adapted to life on muddy bottoms. The bituminous and argillaceous character of the graptolite beds suggests that the unfavorable environment was caused by a muddy sea bottom and by the rarity of bacteria so that organic matter accumulated on it faster than it could be decayed. The alternation between the two facies implies considerable fluctuations in the character of the sea bottom and circulation of the water. These were probably brought about by changes in the character of sediments brought from nearby lands and by changes in currents rather than by an oscillation in the depth of the sea.

The fossils of the coral-bryozoan-cephalopod facies are found throughout the Ordovician limestones of the El Paso and Van Horn districts, where graptolites are entirely absent. The layers at Marathon containing these fossils are probably tongues extending from the northwest. In other parts of North America such fossils are not commonly found in the same region as rocks that contain graptolites, and in some places the rocks of the two facies are supposed to have been laid down in different troughs, separated by land barriers. The close association between fossils of the two types at Marathon and the obvious similarity between some of the fossil zones at Marathon and those in the Van Horn and El Paso districts, to the northwest, make it unlikely that such a barrier ever existed in trans-Pecos Texas.

Source of the sediments

-The greater part of the sediments of the Ordovician system in the Marathon region were apparently derived from the southeast, presumably from the land mass of Llanoria (fig. 16), for the formations change in character in that direction.

In the northwestern exposures of the Marathon formation limestones are very prominent. Most of these are thin-bedded and flaggy, but in the middle is a "

---

Ruedemann, Rudolf, Graptolites of New York, part 2: New York State Mus. Mem. 11, p. 61, 1908.

See, for example, Ruedemann, Rudolf, Alternating oscillatory movement in the Chazy and Levis troughs of the Appalachian geosynclines: Geol. Sec. America Bull., vol. 40, pp. 409-416, 1929.

 

45

member of dolomitic limestone that resembles rocks of Beekmantown age extensively developed in the foreland to the northwest. The sporadic but wide distribution of sponges in all parts of the formation in this area also suggests a northwestward approach to a clear-water environment. In the Dagger Flat anticlinorium limestone flags decrease in number and the dolomitic limestone bed thins and disappears. Most of the formation in the southeastern part of the anticlinorium is shale. In the upper part of the formation in the northeastern part of the anticlinorium are thin beds of gritty sandstone.

Similar but less pronounced stratigraphic changes can be observed in the Fort Peña and Woods Hollow formations. The granular bedded cherts of the Fort Peña, which are in part fine, well-cemented sandstones, increase in thickness in a southeastward direction. Similarly, sandstone members are more prominent in the exposures of the Woods Hollow shale on the southeast flank of the Dagger Flat anticlinorium than they are in the exposures farther northwest. In the little- known outcrops on the Jones ranch (Dove Mountain quadrangle), in the extreme southeastern part of the Paleozoic area, the beds below the Maravillas, correlated with the Woods Hollow and Fort Peña formations, consist of fine-grained quartzite, gritty and pebbly limestone, and micaceous flaggy limy sandstones.

Toward the southeast the Maravillas formation thickens and changes from predominant limestone to predominant chert. The origin of the chert is uncertain, but it may have had its source in that direction.

By contrast to the great bulk of the strata, a few layers in the Ordovician system appear to have had a northwestern source. The Alsate formation appears to change from a shale and limestone succession in the southeast to a shale with few limestone beds toward the northwest. In its northwesternmost exposures, toward Dugout Creek, it contains lenses of coarse sugary sandstone. The conglomerates at the base of the Maravillas and the boulders in the upper part of the Woods Hollow shale also appear to have come from the northwest. Their occurrence is confined to a triangular area between Garden Springs, Del Norte Gap, and Monument Spring, suggesting a nearby local uplift to the west. The fragments in the Maravillas all resemble those in the underlying formations, but those in the Woods Hollow are of a limestone facies foreign to the geosyncline and may therefore have come from farther away, from some part of the foreland area.

Depth of water during deposition

-The character of the Ordovician strata in the Marathon region suggests that they were nearly all laid down in relatively shallow water above the level of wave base, with some parts subjected to intermittent subaerial exposure.

The flaggy limestones of the Marathon formation are interbedded with numerous conglomerate layers, made up of shattered limestone flags, mostly derived from the beds next beneath. The conglomerates were apparently produced by wave action, which shattered and heaped up fragments of the flaggy limestones shortly after their deposition. The limestones themselves in places contain small pockets cut out by wave current action and filled by cross-bedded sand and also tubes and cavities excavated by burrowing animals (fig. 11, B), which are comparable to features seen in the littoral zone today. Limy shales in the formation at one place show well-developed mud cracks (fig. 11, C). Graptolites occur in close association with the beds that show evidence of shallow-water origin. Some were collected a few feet below the layers that contain the mud cracks.

The Woods Hollow shale contains layers of coarse pebbly limestone in which numerous fossil fragments are embedded. These are suggestive of agitated waters and of shallow sea bottoms rich in life.

The Maravillas chert is mostly free from elastic admixtures. However, in the basal part of the formation at Monument Spring and some other places coarse conglomerates are intercalated with and are overlain by bedded chert (fig. 13, A). Lenses of chert fill pockets in the upper part of the conglomerate beds. There are also fine conglomerate layers interbedded in the higher parts of the chert succession. The abraded fossils in the conglomerates, of about the same age as those in the undisturbed layers between, and the numerous pebbles of black chert like that indigenous in the formation indicate approximately contemporaneous destruction of the Maravillas deposits by the waves and their immediate incorporation into new sedimentary rocks. The bryozoan reefs and irregularly bedded, channeled limestones of the southern exposures were probably laid down near a shore line.

These evidences for a shallow-water origin of the Ordovician deposits at Marathon are in agreement with the interpretation of Grabau and O'Connell for the graptolite beds of southern Scotland, which they consider to be "mud deposits on the flood plain and the lagoons of a large delta or series of deltas, where periodic high tides washed in the planktonic graptolites and stranded them on the flats." It is also comparable to the occurrence of recently deposited black muds in bays and lagoons where tide and storm "

---

Grabau, A. W., and O'Connell, M., Were the graptolite shales, as a rule, deep- or shallow-water deposits?: Geol. Soc. America Bull., vol. 28, pp. 959-964, 1917.

 

46

action, is weak, on the east shore of the Baltic and the north shore of the Black Sea.

Origin of the chert beds

-The Ordovician rocks at Marathon contain a large amount of cherty material. This is particularly true of the Maravillas formation in which a large proportion of the strata consist of bedded chert. Lower down in the section considerable amounts of bedded chert are found in the Fort Peña formation, and locally there are also some chert beds in the Marathon limestone. It is noteworthy that other systems of the Paleozoic at Marathon also include layers of chert and siliceous shale, which suggest that conditions favorable for the accumulation of siliceous deposits were remarkably persistent in the area.

Part of the chert in the Ordovician of the Marathon region is of secondary origin. The original calcareous shells of fossils in some of the limestones of the Maravillas chert have been silicified, and some of the thicker limestone ledges have been replaced along the bedding planes by seams and irregular knots of siliceous material. Some of the thinner limestones can be traced along the outcrop into chert beds. It is probable that these features of obvious secondary origin are a minor part of the whole mass of siliceous deposits. Their formation must have been materially aided by the large amount of silica available in the primary chert deposits.

That a considerable part of the chert is of primary origin is indicated by much evidence. The intraformational conglomerates in the Maravillas formation contain pebbles of both chert and limestone. The cherts are identical with those in place in the formation- a fact which indicates that they assumed their present character at least shortly after deposition. Moreover, thin limestone beds in the chert and thin chert beds in the limestone are sharply set off from the enclosing strata, without gradation between them. The stratification of the cherts is quite different from that of the limestones. Most of them are more thinly bedded, some are nodular, and a few of the thicker beds are aggregates of large pillowlike masses. These features have not been seen in the associated limestone deposits.

Some of the chert layers are laminated and even cross-bedded. These parts are seen under the microscope to contain seams of fine embedded sand grains and various fossil fragments. Some of the cherts also contain siliceous and calcareous spicules.

The source of the material for the cherts lay to the southeast. Those in the Maravillas formation thicken markedly in that direction at the expense of the limestone beds, even within the 40-mile breadth of the Marathon Basin. To the northwest the equivalent of the Maravillas chert is a limestone, the Montoya formation. However, the Montoya contains a great deal more chert than the limestone beds that underlie and overlie it.

The silica that formed the cherts was derived from a land area that was able to supply a large quantity of material through a long period of geologic time. This may have been due to some peculiarity of the rocks of which the land was composed, or it may have resulted from volcanic activity, although direct evidence for such activity in Ordovician time is lacking. That the land from which the siliceous materials was derived was not far distant is suggested both by the rapid increase in volume of cherty material to the southeast and by changes in character in the associated sediments in a similar direction. The thickening of the chert beds to the southeast suggests that whatever process carried the silica into the sea was incapable of transporting it for any great distance.

It is possible that the cherts were derived from colloidal silica that was flocculated directly on the sea bottom. Most of the material thus flocculated was deposited in regular strata, but the nodular parts (fig. 13, B) may have been original gel masses of silica. Organisms with siliceous skeletons, such as sponges, must have found encouragement in this environment, as their spicules are found in the rock, but it is unlikely that they played more than an incidental part in the formation of the chert. The fine clastic materials present in parts of the chert represent an occasional deposition of sediments by mechanical processes but do not show that the chert as a whole had this origin.

Origin of the boulder beds

-Boulders as much as 5 feet across are embedded in the upper part of the Woods Hollow shale and the basal part of the Maravillas chert. The origin, mode of transportation, and deposition of these masses present many puzzling features.

The boulders of the Maravillas formation are associated with coarse, poorly sorted cobbles in its basal conglomerate. The intercalation of thin layers of chert and limestone in the conglomerate and the water-rounded character of many of the fragments indicate that the rock is a marine rather than a terrestrial deposit and that it is not a breccia of tectonic origin. As the boulders and cobbles include rocks from nearly every older formation in the region down to the Cambrian, they must have been derived from an uplift several thousand feet in height. As the rocks are like those indigenous to the region, the uplift was probably within the Llanoria geosyncline, although "

---

Twenhofel, W. H., Treatise on sedimentation, 2d ed., pp. 262-263, 1932. Trask, P. D., Origin and environment of source sediments of petroleum, pp. 149-150, 1932.

Twenhofel, W. H., op. cit., pp. 541-542.

For a description and interpretation of the very similar cherts in the Bigfork formation see Hendricks, T. A., Knechtel, M. M., and Bridge, Josiah, Geology of Black Knob Ridge, Okla.: Am. Assoc. Petroleum Geologists Bull., vol. 21, pp. 6-7, 1937.

 

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.

DEVONIAN (?) SYSTEM

CABALLOS NOVACULITE

GENERAL FEATURES

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.

 

48

The marked changes in facies led in the earlier studies of the area to some confusion in the interpretation of the stratigraphy. In the preliminary announcement of their work, Baker and Bowman proposed the Caballos novaculite, with limits not clearly defined, and the overlying †Santiago chert, from 20 to 450 feet thick. The chert was said to have been named "from a locality at the east base of the Santiago Range, east of the range's summit." The †Santiago was rejected as a formation unit by them in 1917, because "later work by the senior author appears to indicate that they [Caballos and Santiago] are really one formation, two members of both the original Caballos and Santiago being included in the section at some localities." The writer's work has afforded complete confirmation of this later interpretation. (See foregoing table.)

The typical Caballos of 1916 (on Horse Mountain) apparently comprised the lower chert member, the lower novaculite member, the middle chert member, and the upper novaculite member of the present paper. On the other hand, the type †Santiago of 1916 (facies 2 of the northwestern Marathon Basin) consisted of the middle and upper chert members. As shown in the table, the upper novaculite member between the two chert members is thin or absent in this area and is not easily recognized. The type section of the †Santiago thus includes part of the type Caballos and suggests that the two formations should be included in one unit, as stated by Baker and Bowman in their final report. For this reason, in the present report only one formation, the Caballos novaculite, is recognized.

LOCAL FEATURES

Northwestern exposures (facies 1)

-In the northwesternmost exposures of the Caballos formation, in the Dugout Creek area (pl. 16), the formation is about 250 feet thick. Near the base is 10 or 20 feet of white vitreous novaculite, which is probably the equivalent of the lower novaculite member farther southeast. The remainder of the formation consists of layers of chert 1 to 8 inches thick, of dull to vitreous luster, banded by various dull colors, such as white, black, brown, green, and pale blue. Near the top are some dull-black cherts. The banded cherts in some exposures are crumpled, broken, and contorted by deformation. Most of the bedding surfaces of the cherts are smooth and regular, but some are wavy or hummocky. Many of the layers are separated by thin partings of hard greenish siliceous shale. At several localities thin lenticular layers of coarsely crystalline buff or gray siliceous limestone and dense laminated limestone are interbedded with the chert. Some of these contain angular chert fragments. A mile west of the Roberts ranch one of the limestone beds contains plentiful linguloid shells.

The following section, measured in the ridges in the southern part of the Payne Hills, 5 miles north-northeast of the Roberts ranch, shows the general character of the formation in this region (sec. 1, pl. 2; fig. 21).

"

---

Udden, J. A., Baker, C. L., and Böse, Emil, Review of the geology of Texas: Texas Univ. Bull. 44, 1st ed., p. 41, 1916.

Baker, C. L., and Bowman, W. F., op. cit., (Texas Univ. Bull. 1753) p. 100.

Idem. p. 94.

 

49

Marathon anticlinorium (facies 2)

-On the southeast flank of the Marathon anticlinorium, at such localities as Monument Spring and Fort Peña Colorada, and on Simpson Springs and East Bourland Mountains, the lower novaculite is very prominent and is succeeded by a thick member of bedded chert. The formation in this region is in general 300 to 400 feet thick but reaches a maximum of 520 feet at Monument Spring.

The lower chert member, which rests on the Maravillas chert and underlies the lower novaculite, is thin and discontinuous. It is a brown, vitreous, splintery, thick-bedded chert. The lower novaculite, averaging about 125 feet in thickness, forms the crests of the Caballos hogbacks. Its lower and upper parts are white and vitreous and stand out in ledges or bouldery outcrops with indistinct bedding. The middle or main mass of the member is dull-lustered, white or creamcolored, and of porcelainlike texture and forms wellmarked layers several inches thick (pl. 4, D). These are split by innumerable little vertical joints which cause the rock to break into small flat-faced, sharp-angled fragments few of which are more than 4 inches across. These fragments were utilized by the Indians in the manufacture of arrowheads at the gap south of Fort Peña Colorada. Partly trimmed fragments of novaculite are found at numerous places in this vicinity. On the novaculite ridge 7 miles southwest of Marathon and on the southeast side of Simpson Springs Mountain some of the bedding surfaces in the lower novaculite are marked by closely set ripple marks (fig. 17).

The upper part of the formation consists mostly of banded cherts and is the original †Santiago chert of Baker and Bowman. At the base, directly on the lower novaculite, are some thin local quartzitic sandy beds and in places fine conglomerate. The cherts crop out in several low cuestas, separated by saddles occupied by thinner-bedded or shaly members. They form beds a few inches to more than a foot thick. The chert is vitreous to dull-lustered and is banded by laminae of various colors a fraction of an inch thick. The predominant color is green, but there are also gray, black, and brown bands and some of bright jasper red. A 15-foot layer of siliceous green shale 180 feet above the lower novaculite is found at many places near Monument Spring. In the upper part of the formation on East Bourland Mountain there is considerable interbedded red and green siliceous shale, from which H. D. Miser has collected conodonts. The joint surfaces of the cherts are much stained by iron and manganese at several places, as at the Clark manganese prospect, on the east side of the Boquillas road 2 miles south of Marathon. In the northeastern part of the Marathon anticlinorium there is a 5-foot member of white massive vitreous novaculite about 100 feet above the base of the formation; this disappears to the southwest near Fort Peña Colorada (sets. 4, 5, and 6, pl. 2).

The following section, typical of this phase of the formation, was measured at the Clark manganese prospect (sec. 6, pl. 2).

 

50

Dagger Flat anticlinorium (facies 3 and 4)

-Along the northwest flank of the Dagger Flat anticlinorium the lower novaculite member is prominent and is about 60 feet thick (pl. 3, C), but the upper novaculite thickens from 5 feet in the northwest to about 25 feet in the Woods Hollow Mountains. The novaculite members have a similar thickness along the northwest flank of the anticlinorium from Ridge Spring northeastward to the Warwick Hills.

Farther southeast, on the southeast flank of the Dagger Flat anticlinorium, the relative thicknesses of the two novaculite members are reversed, the lower member being about 15 feet thick and the upper about 100 feet (fig. 32; sets. 9, 10, and 11, pl. 2). The change in thickness takes place on the crest of the anticlinorium and may be traced out in the convoluted novaculite outcrops in the Warwick Hills. In the central part of the hills the thickness of the two members is approximately equal. The formation in these hills is only 150 to 200 feet thick, as compared with 300 or 400 feet to the northwest and southeast. This thinning is shared by all members equally and is probably original in the deposit. Perhaps it was caused by slight positive movements along the anticlinorium during Caballos time.

Both the upper and lower novaculites in the Dagger Flat anticlinorium are white and vitreous. The lower member near Woods Hollow Tank is ripple-marked. Directly above the lower novaculite at several places is an angular chert conglomerate in a siliceous matrix. Just under the upper novaculite in the Woods Hollow Mountains, on Horse Mountain, and elsewhere is a 15-foot bed of green siliceous shale, and in places there is some red shale. At the top of the upper novaculite some 1-foot beds of white chert are interbedded with those of darker color.

The following section (sec. 12, pl. 2) was measured 4 miles northeast of the Buttrill ranch:

At Maravillas Gap and Horse Mountain the formation attains a thickness of 500 or 600 feet but with the same members as to the northwest. The novaculite on Horse Mountain is more brecciated and veined than in other places and the joint surfaces are in many places coated with red hematite. The upper novaculite member forms the surface and slopes of Horse Mountain (pl. 24). The upper cherts on the mountain are light-colored, thick-bedded, and vitreous below and pass upward into dull black, brown, and green argillaceous cherts in beds a few inches thick. Near Maravillas Gap the upper cherts have undulatory or

 

51

hummocky bedding surfaces. In the ridges southeast of the Dagger Flat anticlinorium the cherts of this member are thick-bedded and light-colored and are similar in aspect to novaculite, into which they appear to be changing (pl. 5, C).

Southeastern exposures.

-In the southeastern part of the Marathon Basin the novaculite is exposed at several places. On Rough Creek (Dove Mountain quadrangle) about 200 feet of beds are exposed in the center of an anticline, surrounded by basal Tesnus shales, which appear to overlap the formation unconformably. At the base is 25 feet of bluish or white novaculite in 1-foot beds, much shattered and brecciated. Above is 175 feet of bluish, black, or greenish dull lustered chert in 2- to 6-inch beds with thin partings of siliceous shale, which are most abundant toward the top. There is also a small exposure of novaculite 1½ miles south of the Jones ranch, in the same quadrangle.

MICROSCOPIC AND CHEMICAL CHARACTER

Novaculite.

- A thin section of cream-colored porcelainlike novaculite from the lower novaculite member of the formation at Fort Peña Colorada is of uniformly fine-grained siliceous material, with no stratification (fig. 15, D). The rock shows a few round or irregular spots, which appear dark under crossed nicols. They consist of siliceous material like that surrounding them but of very fine texture. There are a few widely dispersed minute, well-rounded quartz grains. Chips of the novaculite have a more porous texture than those of chert. This is probably the cause of the dull luster of the novaculite when compared with the chert in hand specimens. The specimen from Fort Peña Colorada is of finer grain than the typical novaculite of Arkansas. Baker and Bowman have studied some specimens of Caballos novaculite that are coarser and consist of a holocrystalline aggregate of fine interlocking granules of quartz.

A specimen of thinly laminated or banded novaculite from the base of the upper novaculite member of Threemile Hill showed in thin section fine-grained siliceous material, interbedded with bands of coarse fibrous chalcedony.

Baker and Bowman have studied some thin sections of Caballos novaculite which have a very uneven texture, with knots of coarse chalcedonic fibers, of undulatory extinction. They interpret the knots as altered tests of Radiolaria. Similar features have been noted by the writer in the banded cherts associated with the novaculite but have not been seen in the novaculite itself.

Specimens of novaculite collected by Baker and Bowman from the hills south of Warwick siding were analyzed by J. E. Stullken in the laboratory of the Texas Bureau of Economic Geology. The specimens are probably from the lower novaculite member of the formation. The following analyses are given by Baker and Bowman:

Tests made by Stullken show that Caballos novaculites contain from 5.2 to 7.0 percent of soluble silica by digestion in sodium hydroxide for 30 minutes, as compared with 3.5 percent for the Arkansas novaculites, and 2.0 to 2.5 percent by digestion in potassium carbonate for 6 hours, as compared with 0.3 to 1.1 percent for the Arkansas novaculite.

Banded chert

-The banded cherts are of various textures and colors and are arranged in laminae from a few millimeters to 1 centimeter in thickness. In thin section the lighter bands are seen to be massive and of a fine, even texture, not unlike some of the finer-grained varieties of novaculite. The darker bands are fine- to moderately coarse-grained and are mostly well stratified (fig. 15, F). They contain small angular to well-rounded detrital grains of quartz, a few grains of minerals of high refractive index, and some small blades of mica. Some of the dark bands have a faint uniform extinction in one direction, produced by minute blades of clay minerals oriented parallel to the stratification. The clay minerals also probably cause the varying colors of the cherts in different bands. Thin sections of some of the chert contain minute opaque crystals of pyrite, and these have also been seen on the surfaces of polished specimens.

Some of the layers, particularly those of coarser texture, contain round bodies, with some evidence of an original outer test wall and with an interior filled by "

---

Baker, C. L., and Bowman, W. F., op. cit. (Texas Univ. Bull. 1753), pp. 99-100.

Idem, p. 100 and pl. 2.

Idem, p. 96.

 

52

radiating chalcedonic fibers (fig. 15, F and F'). According to L. G. Henbest, these are probably the outer capsules of Spumellaria, a group of Radiolaria. One of the capsules he found to bear uncertain traces of porous structure. There are also some rodlike bodies which Henbest says resemble sponge spicules. One specimen of chert from the Payne Hills contained numerous round tests of varying size, a millimeter in maximum diameter, filled with radiating chalcedonic fibers (fig. 15, E).

Remains of Radiolaria have been reported by other investigators from cherts similar to those examined by the author. Baker and Bowman report that in some thin sections examined by them Radiolaria make up as much as 50 percent of the rock. Specimens collected by H. D. Miser have been examined by Henbest, who reports as follows:

Radiolarian skeletons are very numerous in this collection but belong to only a few species. * * * The skeletons are symmetrical, radially spinose, and have an outer and a single inner capsule. The outer capsule appears to have a honeycomb structure. They belong therefore to the Spumellaria. * * * The skeletons are in part concentrated in association with the varves, and a large number show varying degrees of flattening, but a few of the better-preserved individuals appear to have a distribution that is not so closely associated with the varves, a condition that turned out to be more apparent than real. The larger part of the organic detritus is zoned, however, and is probably the chief cause for the varved appearance of the rock from a macroscopic viewpoint.

Sandstone

-The granular and sandy cherts that overlie the lower novaculite member are fine sandstones composed of chert and quartz grains with a considerable cherty matrix. A specimen from the old Clark prospect, 2 miles south of Marathon, contained rounded quartz grains and large angular chert and novaculite fragments. Some of the quartz grains in this specimen have been enlarged by secondary growth to form an interlocking mass of crystals.

FOSSILS AND AGE

The Caballos novaculite contains a few fossils, but none have so far given very definite indications as to the age of the formation. Edwin Kirk and the writer have collected abundant linguloid brachiopods from a thin limestone lens in the formation 1 mile west of the Roberts ranch. H. D. Miser collected conodonts from siliceous shales in the upper part of the formation on East Bourland Mountain. Baker and Bowman, Henbest, and the writer have seen Radiolaria in thin sections of the cherts. The linguloids are not sufficiently diagnostic to furnish precise evidence for the age of the formation. Conodonts have given valuable evidence on the age of the cherts and novaculites of the Woodford and Arkansas formations in Oklahoma and Arkansas. Those collected in western Texas have, however, not yet been studied. According to Henbest, the Radiolaria are similar to those in the Arkansas novaculite, but they have not been specifically identified.

The Caballos novaculite is so strikingly similar to the Arkansas novaculite of Oklahoma and Arkansas, not only in lithology but in the character of the members and their stratigraphic behavior, that there is a strong presumption that the two are of the same age. The Arkansas novaculite has yielded fossils of Middle Devonian and Upper Devonian age. Until further evidence is obtained, the Caballos novaculite may best be classified as Devonian (?).

STRATIGRAPHIC RELATIONS

The Caballos novaculite is overlain by the Tesnus formation with considerable unconformity, and the break probably represents most if not all of Mississippian time. On Rough Creek, in the southeastern part of the Marathon Basin, the Tesnus formation overlaps the whole thickness of the Caballos, which is here folded into a steep anticline (fig. 18, B). Elsewhere such sharp overlaps and differences in folding are not evident. The slightly variable thickness of the upper cherts of the Caballos novaculite and also the thin chert conglomerates found nearly everywhere at the base of the Tesnus formation are indicative of an erosional break between the two. In some places, however, the boundary between the two formations is not easy to draw, because the uppermost dull cherts and siliceous shales of the Caballos closely resemble the basal indurated shales of the Tesnus. The Tesnus formation thins greatly to the northwest across the Marathon Basin, as explained beyond. This may be the result of a gradual northwestward overlap of its strata against the unconformable upper surface of the Caballos novaculite. Such an overlap is suggested by the appearance of thick and coarse conglomerates in the Tesnus formation of the northwestern exposures.

There is also a suggestion of an unconformity within the Caballos novaculite. At many places in the region a layer of conglomerate and sandstone occurs at the top of the lower novaculite member. The fragments consist of chert and novaculite derived from the underlying beds. This conglomerate occupies a similar position to the conglomerate beds at the base of the middle member of the Arkansas novaculite in Arkansas, and the two occurrences may represent the same depositional break.

"

---

Baker, C. L., and Bowman, F. W., op. cit., p. 101.

Henbest, L. G., Radiolaria in the Arkansas novaculite, Caballos novaculite, and Bigfork chert: Jour. Paleontology, vol. 10, p. 77, 1936.

Miser, H. D., and Purdue, A. H., Geology of the DeQueen and Caddo Gap quadrangles, Ark.: U. S. Geol. Survey Bull. 808, pp. 58-59, 1929.

Idem, p. 52.

 

53

GENERAL PROBLEMS OF DEVONIAN (?) STRATIGRAPHY

CORRELATIONS AND REGIONAL RELATIONS

Correlations in trans-Pecos Texas

-The Caballos novaculite is exposed at one place in trans-Pecos Texas outside the Marathon Basin. This is in the Solitario uplift, southwest of Marathon, where the strata are very much like those described above. It is reported that they consist of a prominent lower novaculite member, overlain by bedded cherts. The sequence is thus of the same facies (facies 2, p. 48) as that in the northwestern part of the Marathon Basin.

Northwest of the Marathon and Solitario uplifts, in the Van Horn and El Paso districts, strata that may be equivalent to the Caballos novaculite are exposed. In the Sierra Diablo, north of Van Horn, platy bituminous shales, shaly limestones, and white, buff, and green cherts lie between the Fusselman limestone (Silurian) and strata of Mississippian age. In the Hueco Mountains, east of El Paso, the Fusselman limestone is overlain by cherts, which are succeeded by strata containing Mississippian fossils. In the Franklin Mountains, north of El Paso, shales overlying similar cherts contain Devonian fossils and are correlated by Darton with the Percha shale of New Mexico.

Correlations with Oklahoma and Arkansas

-There are many similarities between the Caballos novaculite and the Arkansas novaculite of the Ouachita Mountains in Oklahoma and Arkansas. The thicknesses of the two formations are similar. It is true that the Arkansas novaculite attains a maximum thickness of 900 feet, which is greater than any observed thickness of the Caballos. This is in its southernmost exposures, however; in most of the Ouachita Mountains the thickness does not exceed 600 feet, and in the northern part of the mountains it is only 250 feet. The members are also similar in the two regions. The lower member of the Arkansas novaculite is a massive white novaculite. It is overlain by a middle member of bedded chert with much shale. At the top of the formation is another member of massive novaculite. The massive novaculites may be equivalent to the two novaculite members of the Caballos. They thin out to the north, so that the northern exposures consist mostly of shales and bedded cherts. The upper member disappears before the northern edge of the Ouachita Mountains is reached. On top of the lower member of the Arkansas novaculite there is a layer of conglomerate that occupies a similar position to the conglomerate bed at the top of the lower novaculite member in western Texas.

The upper member of the Arkansas novaculite differs from the novaculite of the Caballos in several of its features but particularly in that it contains crystals of calcite and other carbonates. The novaculites and cherts of the other members of the Arkansas are very similar to those in the Caballos, both in hand specimen and in thin section.

The lower member of the Arkansas novaculite contains Leptocoelia flabellites and is of Middle Devonian age. Conodonts found in the middle member are the same as those found in the Chattanooga shale. This member and the unfossiliferous upper member are therefore to be correlated with the Chattanooga formation, which the United States Geological Survey classifies as of Devonian (?) age.

In the Arbuckle Mountains, west of the Ouachita Mountains, is the Woodford chert, composed of siliceous shale, chert, and thin limestone. This formation occupies a position similar to that of the Arkansas novaculite and also contains conodonts of Chattanooga age. It is correlated by Ulrich with the middle member of the Arkansas novaculite. There is thus a northwestward change in the character of the strata in Oklahoma, like that in trans-Pecos Texas.

ORIGIN OF THE CABALLOS NOVACULITE

Definition of novaculite

-The term "novaculite" has long been used for a fine quality of whetstone or razor hone, and in 1890 it was applied by Griswold to rocks of that type in Arkansas. According to Griswold, the novaculites of Arkansas are composed almost entirely of silica, contain little or no soluble silica (chalcedony), tend to be translucent, and have a gritty rather than a glassy texture. Novaculite appears to be a variety of chert, for chert has been defined as including "all forms of finely crystalline, nonfragmental silica, including opaline, semicrystalline, and completely crystalline varieties." Moreover, "

---

Sellards, E. H., personal communication.

King, P. B., Possible Silurian and Devonian strata in the Van Horn region: Am. Assoc. Petroleum Geologists Bull., vol. 16, p. 96, 1932.

King, P. B. and R. E., Stratigraphy of outcropping Carboniferous and Permian rocks in trans-Pecos Texas: Am. Assoc. Petroleum Geologists Bull., vol. 13, p. 910, 1929.

Darton, N. H., Devonian strata in western Texas [abstract]: Geol. Soc. America Bull., vol. 40, p. 116, 1929.

Miser, H. D., and Purdue, A. H., Geology of the De Queen and Caddo Gap quadrangles, Ark.: U. S. Geol. Survey Bull. 808, p. 50, 1929.

Griswold, L. S., Whetstones and the novaculites of Arkansas: Arkansas Geol. Survey Ann. Rept. for 1890, vol. 3, pp. 193-194, 1892.

Miser, H. D., and Purdue A. H., op. cit., p. 52.

Honess, C. W., Geology of the southern Ouachita Mountains of Oklahoma Oklahoma Geol. Survey Bull. 32, p. 117, 1923.

Cooper, C. L., Conodonts from the Arkansas novaculite, Woodford formation, Ohio shale, and Sunbury shale: Jour. Paleontology, vol. 5, p. 145, 1931.

Miser, H. D., and Purdue, A. H., op. cit., p. 58. In this paper the occurrence of Genesee fossils is reported from the middle member. Ulrich (Fossiliferous boulders in the Ouachita "Caney" shale and the age of the rocks containing them: Oklahoma Geol. Survey Bull. 45, p. 33, 1927, and other papers) correlates the unfossiliferous upper member with the Boone formation, of Mississippian age.

Ulrich, E. 0., op. cit., p. 33.

Griswold, L. S., op. cit., p. 2.

Idem, p. 187.

Van Hise, C. R., A treatise on metamorphism: U. S. Geol. Survey Mon. 47, p. 816, 1904.

 

54

Ross states that there is no essential microscopic difference between novaculite and more normal varieties of chert.

In the Caballos formation of the Marathon Basin, two beds are set off as novaculite members, which are considered to be separated, as well as underlain and overlain, by chert members. The novaculites are distinguished by their white color from the cherts, which exhibit a variety of colors, mostly darker. The novaculites are massively bedded and crop out in ledges as much as several feet thick, whereas the layers of chert are rarely more than several inches thick. The massive quality of the novaculite may also be observed in thin section, for the rock is uniformly fine grained, with no evidence of lamination, whereas the banded cherts are marked by laminae of different colors and textures. The novaculite appears to be of a somewhat porous texture, whereas the fracture surfaces of the cherts are smooth and glassy.

Theories of origin

-The novaculite of Arkansas has been interpreted as a replacement product of limestone or dolomite and as a metamorphosed chert, but neither of these suggestions seems adequate to explain the observed features of the rock. It has also been considered to be a fine siliceous clastic sediment and a colloidal precipitate. A possible relation to volcanic activity and to the growth of Radiolaria is suggested by the observations of some geologists. The literature on the origin of novaculite has been summarized by Griswold, Derby, and Miser and Purdue.

Studies of the Caballos novaculite by the writer have failed to establish definitely the validity of any of these interpretations, but they have furnished some new evidence bearing on the origin of the rock. This evidence is discussed below.

Stratigraphic arrangement of the chert and novaculite members

-The two novaculite members of the Caballos formation are most prominent in the southeastern exposures, whereas in the northwestern exposures banded cherts predominate. This change is accomplished by a northwestward thinning of the novaculite members. The upper novaculite member extends no more than halfway across the Marathon Basin and pinches out. The lower novaculite extends farther, but in the northwesternmost exposures it is thin and has assumed more the character of chert than of novaculite. This member also thins, however, to the southeast, from a maximum near the center of the area. The conglomerates immediately above it suggest that the southeastward thinning is possibly caused by an erosional break.

The banded cherts contain a few partings, as well as thicker layers, of siliceous shale, even in the southeasternmost exposures, but in the northwestern exposures such shaly layers are very prominent. In the northwestern part of the area, but nowhere else in the region, they also contain thin lenses of limestone. Moreover, in the Van Horn and El Paso districts, to the northwest, the possible equivalents of the Caballos are shales and limestones, with a subordinate number of chert beds.

These stratigraphic changes appear to be related to the position of the shore lines and land areas during Devonian (?) time. Clastic rocks, both overlying and underlying the Caballos novaculite, show clear evidence of being derived from a land area to the southeast, and it is therefore probable that the changes in the Caballos in the same direction are related to the same cause. If this is so, the novaculites may have been deposited nearest the shore, the cherts farther out, and the shales and limestones farthest of all. A similar explanation has been offered by Griswold for the novaculites of Arkansas.

Significant features in the novaculite

-As noted in the local descriptions, at several places in the region the bedding surfaces of the novaculites are covered by fine corrugations that have all the appearance of ripple marks (fig. 17). The ridges are spaced about a quarter of an inch apart but may be as far apart as half an inch. They are sharp-crested, and the troughs between are broad and rounded. In places the ridges bifurcate. Ripple marks have also been noted by Miser and Purdue in the Arkansas novaculite. They are described as large and uneven and are therefore probably different in character from those in trans-Pecos Texas.

The form of the marks suggests that they were oscillation ripples, produced "by the to-and-fro motion of the water, occasioned by the passage of wind waves." They suggest that the novaculite was laid down in shallow water and also imply that the sediments which were afterward consolidated into novaculites were deposited in finely granular rather than gelatinous or colloidal form.

The scattered rounded and angular quartz grains and the tests of Radiolaria seen under the microscope in the novaculites of the Caballos formation indicate "

---

Ross, C. S., personal communication, 1932.

Rutley, Frank, On the origin of certain novaculites and quartzites: Geol. Soc. London Quart. Jour., vol. 50, pp. 377-392, 1894. Derby, 0. A., Notes on the Arkansas novaculite: Jour. Geology, vol. 6, pp. 366-368, 1898.

Branner, J. C., Arkansas Geol. Survey Ann. Rept. for 1888, vol. 1, p. 49 (footnote), 1888.

Griswold, L. S., op. cit., pp. 191-192.

Honess, C. W., op. cit., p. 138. Miser, H. D., and Purdue, A. H., op. cit., p. 57.

Honess, C. W., op. cit., pp. 121-128.

Baker, C. L., and Bowman, W. F., op. cit., p. 100. Henbest, L. G., op. cit., pp 76-78.

Griswold, L. S., op. cit., pp. 169-187.

Derby, 0. A., op. cit., pp. 366-368.

Miser, H. D., and Purdue, A. H., op. cit., pp. 55-57.

Griswold, L. S., op. cit., pp. 192-194.

Miser, H. D., and Purdue, A. H., op. cit., p. 51.

Grabau, A. W., Principles of stratigraphy, p. 713, 1924.

 

 

55

that a small part of the rock is detrital and organic origin. Most of the rock, however, is fine and structureless, and if it were originally either detrital or organic, it must have been very different from the sand grains or the tests. Moreover, the content of silica in the novaculite is much greater than that of any tuff, which would seem to show that the rock is not directly of volcanic origin.

Significant features in the banded cherts

-The banded cherts contain more undoubted elastic and organic material than the novaculites. The rock is divisible into alternating bands and laminae of different textures, which apparently represent the original stratification of the deposit. Some bands consist of fine structureless silica very similar to novaculite. Other bands contain a large percentage of sand and clay particles, and tests of Radiolaria and other organisms. These bands appear to grade into siliceous shales by a slight increase in the elastic components. The different colors in the bands may have been caused by variations in the organic and ferruginous material associated with the clays, and the bands may have been caused by periodic fluctuations in conditions of sedimentation in a region of relatively quiet water.

Conclusions

-With. this inconclusive evidence, no definite interpretations can be made as to the origin of the novaculite. A relation between novaculite deposition and the secretion of silica by such organisms as Radiolaria, with volcanic activity, is possible but remains to be proved. To the writer the northwestward thinning of the novaculite members, their relation to the banded chert members, and the ripple-marked bedding surfaces suggest that the novaculite of the Caballos formation may have been laid down as a fine elastic sediment, rather than as a precipitate.

CARBONIFEROUS SYSTEM

PENNSYLVANIAN SERIES

The Pennsylvanian strata of the Marathon region were first studied by Baker and Bowman in 1915 and divided in ascending order into the Tesnus, Dimple, Haymond, and Gaptank formations. Later work has not modified this classification, though it has furnished more information on the ages of the different parts of the succession and has altered the interpretation of some of the exposures.

The formations are predominantly elastic, made up of sandstone and shale, with some beds of limestone and conglomerate. They reach a thickness of at least 12,000 feet in the southeastern part of the area but are much thinner to the northwest (pl. 8). The lowest formation, the Tesnus, is a mass of shales and fine sandstones 6,500 feet or more thick; the Dimple formation is a limestone as much as 1,000 feet thick; the Haymond formation, 3,000 feet thick, is again sandstone and shale, with a remarkable boulder bed in the upper part. The upper 1,800 feet of the succession, the Gaptank formation, is an alternation of limestones and conglomerates, with sandy and shaly beds. Marine fossils are found sparingly in the Dimple limestone and parts of the Haymond formation but are most abundant at numerous horizons in the Gaptank formation. Fossil plants, mostly abraded and poorly preserved, have been found at several places in the Tesnus and Haymond sandstones.

The folded and faulted lower formations crop out in rugged hills and ridges over wide tracts in the Marathon Basin. The higher parts of the succession are found only on the north side of the folded belt, where they have been partly overridden by thrust sheets that advanced from the south. The succeeding Permian strata to the north of them, in the Glass Mountains, are tilted away from the uplift and overlap across the folded and eroded edges of the Pennsylvanian.

TESNUS FORMATION

GENERAL FEATURES

The Tesnus formation was named by Baker and Bowman for exposures near Tesnus station, on the Southern Pacific Railroad east of Haymond, in the eastern part of the Marathon Basin. The Tesnus is the oldest Carboniferous formation in the Marathon region and is extensively exposed around the southeast, east, and northeast edges of the basin, where it flanks the central area of pre-Carboniferous rocks. It also crops out in narrow synclines between the anticlinoria of older strata. Because of the generally nonresistant character of its sandstones and shales, it occupies low places on the plains. In the northern part of the basin it is widely mantled by wash, but toward the south recent dissection has broken the old surface into an intricate topography of cockscomb sandstone ridges and shale valleys.

The Tesnus formation is a great mass of interbedded sandstone and shale, in thin and thick beds, nearly barren of fossils except for a few plant remains in the upper part. In most places it attains a thickness of several thousand feet, but its thickness is variable (pl. 8). In the northwestern part of the basin it is about 300 feet thick and is nearly all black shale, with few sandstone beds. In the southeastern part it exceeds 6,500 feet in thickness and is predominantly sandstone, with many arkose layers and several prominent massive layers of white quartzite. In this part of the area the basal part of the formation is predominantly shaly and "

---

Baker, C. L., and Bowman, W. F., Geologic exploration of the southeastern Front Range of trans-Pecos Texas: Texas Univ. Bull. 1753, pp. 101-112, 1917. Preliminary descriptions of the formations are found in Review of the geology of Texas: Texas Univ. Bull. 44, 1916.

Baker, C. L., and Bowman, W. F., op. cit., p. 101.

 

56

has been called †"Rough Creek shale member" by Baker and Bowman, but that name is preoccupied in the Pennsylvanian of central Texas.

LOCAL FEATURES

Tesnus and Haymond area

-Along the east side of the Marathon Basin, east of the novaculite ridges that bound the Dagger Flat anticlinorium on the southeast, the Tesnus and other strata of Carboniferous age are folded into broad open anticlines and synclines (sets. B-B', C-C', and D-D'-D", pl. 21). In this region the formation is about 6,500 feet thick. Only the upper part is exposed near Tesnus station, the type locality.

The basal shale member, the lower fourth of the formation, is clearly separable from the upper three- fourths, in which sandstone predominates. The basal member crops out in rolling hills along the flanks of the novaculite ridges and appears on aerial photographs as a dark-colored, featureless surface (pl. 18). It is mostly a soft greenish shale, with interbedded layers and lenses of pale-green argillaceous sandstone. There are also some thin beds of hard platy dark-blue or black shale. The shales are less competent than the sandstones above and the novaculite below and are irregularly folded and crumpled. In places they are sheared and macerated by faulting. The interbedded sandstones are most prominent near the top of the member, and it grades into the main body of the Tesnus formation above. In view of the lenticular character of the sandstone beds, it is doubtful whether the boundary as drawn is at the same level at all places.

The upper part of the formation is predominantly sandstone in massive ledges that stand in low parallel ridges, separated by shallow valleys carved from the interbedded shales. In aerial photographs the outcrops of the sandstones and shales stand out in a striking manner as alternating narrow light and dark bands, which clearly reveal the structure of the formation (pl. 18). The sandstone beds are mostly buff or green and are friable and somewhat arkosic. Some of the members are thick-bedded, and there are a few massive layers as much as 50 feet thick (pl. 9, A). Other sandstone beds, mostly fine-grained, are thinly laminated and flaggy, with numerous shale partings. Some of the layers near the middle of the upper member southwest of Haymond station are coarsely ripple-marked. Near Tesnus station sandstones in the upper part contain the pinnules of ferns. At several places there are layers of thin-bedded black, dull-lustered chert. The upper 300 or 400 feet of the Tesnus is predomi nantly black indurated splintery shale, with subordinate sandstone layers. The contact with the overlying Dimple formation is drawn at the lowest limestone layer interbedded in the shale.

The following section of the Tesnus formation, 6,520 feet in thickness, was measured between Peña Blanca Spring and the Haymond Mountains (sec. 6, pl. 8). The structure of the region is simple and is made plainly evident by excellent exposures, so that there is little chance for duplication by folding and faulting. The writer was aided in the instrument work on this section by A. G. Nance and John Bean.

 

57

Southeastern part of Marathon Basin

-Along lower San Francisco Creek, south of the great fault that bounds Hells Half Acre on the north, the folds of the Tesnus formation are obscure and closely packed and, except near the northern border, the succeeding Dimple limestone is wanting (sec. D-D'-D", pl. 21). In places the strata have been so much crumpled and broken that there are no continuous ledges. Many of the massive sandstones are traversed by lines of shear and fracture and are veined by quartz and calcite. The thinner beds show a large development of secondary mica and other evidences of incipient metamorphism.

The lowest beds are exposed along Rough Creek for several miles above its entrance into San Francisco Creek in the Dove Mountain quadrangle (fig. 18, A, and pl. 23). The Caballos novaculite crops out on the crest of an anticline a mile above the mouth of the creek and is overlain uncomformably by green indurated clay shales, forming a basal shale member of the Tesnus. These constitute the type †SRough Creek shale of Baker and Bowman. The banded cherts included in this member by Baker and Bowman appear to belong to the upper part of the Caballos formation and are exposed only below the unconformity on the north flank of the anticline (fig. 18, B). Interbedded sandstones are abundant in the upper part of the shalemember of the Tesnus on Rough Creek, and it grades, into the main body of the Tesnus formation above. The basal shale member on Rough Creek has a similar stratigraphic position and thickness to that in the Tesnus and Haymond area to the north, and it is likely that the two are of the same age.

Farther west, near the Indian Creek ranch, in the Hood Spring quadrangle (pl. 23), the Caballos novaculite is overlain by thin-bedded argillaceous sandstone and the basal shale member is not recognizable as such.

The upper part of the Tesnus formation, which crops out in rugged ridges in the region of Hells Half Acre and Devils Backbone, consists of compact greenish quartzitic sandstone and coarse-grained friable buff arkose, with small amounts of interbedded shale. There are several massive members of white or cream-colored quartzitic sandstone in beds 25 to 50 feet thick, which stand in jagged hogbacks. These sandstones resemble the Jackfork sandstone of the Ouachita Mountains. On Devils Backbone there are two such white members, separated by several hundred feet of shales and sandstones (fig. 19), but the stratigraphic relations of the quartzites are not everywhere evident, and their outcrops are broken by obscure faults and folds.

Peña Colorada synclinorium.

-On the opposite or northwest side of the Dagger Flat anticlinorium the Tesnus formation consists predominantly of shale and is scarcely more than 2,000 feet thick. These strata probably represent only the upper part of the succession exposed in the Tesnus and Haymond area, and it is believed that the lower part, including the basal shale member, has passed out by overlap against the Caballos novaculite.

Several layers of conglomerate a few feet thick occur near the base, composed of angular chert and novaculite fragments in a siliceous matrix. The conglomerates 3 miles south of Marathon also contain spherical calcareous pisolites as much as an inch across. The conglomerate beds increase in number and thickness toward the northwest. At several places, as near Sunshine Springs and in the Woods Hollow Mountains, there are beds of maroon-red shale near the base of the formation.

The greater part of the formation is composed of olive-green clay shales, in part sandy, and of blue-black indurated slaty shales, which weather to gray chips and splinters. There are some interbedded layers and "

---

Baker, C. L., and Bowman, W. F., op. cit., p. 103, pl. 1 b.

 

58

lenses of argillaceous green sandstone, much sheared and fractured, which in places have a cone-in-cone structure. Near West Bourland Mountain the argillaceous sandstones weather to spherical cannon-ball concretions (fig. 20, B). At wide intervals in the succession are ledges of compact dark-green quartzitic sandstone, and in the upper 500 feet several thick layers of arkose. These arkose layers are conspicuously developed around the flanks of West Bourland Mountain, where they are separated from the Dimple limestone, which caps the mountain, by several hundred feet of indurated shale and bedded chert (fig. 20, A). Each arkose layer at this locality rests on a channeled surface of the bed beneath, with the pockets in the channels filled by coarse sandstone crowded with water-worn plant fragments. Several casts of large logs of Calamites were also noted.

The following section was measured across the valley of Peña Colorada Creek between the novaculite outcrops of East Bourland Mountain and the Dimple limestone in the syncline of West Bourland Mountain (fig. 20, A; sec. 5, pl. 8). Baker and Bowman report a thickness of 3,370 feet at the same place but probably did not correct their measurements for minor structural features. The beds change from an overturned position below to a gentle inclination above, with several local folds and reversals of dip between (fig. 20, B). The thicknesses given here for the lower part of the section are scarcely more than approximations. "

---

Baker, C. L., and Bowman, W. F., op. cit. (Texas Univ. Bull. 1753), pp. 101-102.

 

3

The lower shaly portion is believed not to be of the same age as the basal shale member of the Tesnus and Haymond area.

 

60

Northwest of this section the formation thins out rapidly and near Monument Spring is only 500 feet thick (sets. 2-5, pl. 8). The plant-bearing arkoses of West Bourland Mountain disappear along the outcrop, and the whole formation to the northwest ;;onsists of bluish indurated shale with several interbedded layers of conglomerate. This northwestward thinning is probably chiefly the result of overlap of the Tesnus strata on the surface of the Caballos novaculite beneath.

Dugout Creek area.

-Exposures of the Tesnus formation are nearly continuous from Monument Spring around the southwest end of the Marathon anticlinorium to the Dugout Creek area, on its northwest flank. The formation here is nowhere more than 300 feet thick and consists of greenish clay shales and indurated blue-black shales, with interbedded conglomerate and some beds of dull-lustered chert. In places the Dimple limestone lies directly on cherts of the Caballos formation, but whether from the actual disappearance of the Tesnus formation by overlap or from later tectonic squeezing cannot be said. The following section was measured in the southern part of the Payne Hills, 5 miles north-northeast of the Roberts ranch (sec. 1, pl. 8, and fig. 21).

MICROSCOPIC CHARACTER

Thin sections of the Tesnus sandstones show that most of them are made up of fine, subangular to subrounded quartz fragments in a chloritic matrix. The matrix imparts the characteristic greenish color to the fresh exposures of the rock, and its decomposition gives rise to the characteristic brown weathered surfaces.

Arkosic sandstones are fairly common in the formation. Megascopically they are greenish, of dull earthy appearance, and spotted by small light and dark grains. Such rocks are most common in the southeastern exposures of the formation. Three thin sections of arkose were examined, one from Devils Backbone, one from West Bourland Mountain, and one from a layer 1,000 feet below the top of the formation southeast of the Dimple Hills, in the northeastern part of the Marathon Basin. The three are of similar character. Quartz grains, somewhat shattered and nearly all showing strain shadows, make up 75 percent of the rock. Most of them are angular, but a few are well rounded. Most of the other grains are angular chips of sericitic slate and quartz-chlorite phyllite. There are also a few grains of potash feldspar, sodic plagioclase, chert, biotite in bleached twisted plates, and muscovite. A few small grains of high refractive index may be seen. The arkoses from West Bourland Mountain, in the northwestern part of the area, are

 

61

of finer grain than the other two and contain more quartz, but fragments of slaty and cherty rocks are present, as well as feldspar and mica.

Other sandstones in the formation contain more quartz and approach quartz sandstones and quartzites. Nearly all, however, contain a few grains of metamorphic rocks. The matrix of most of the Tesnus sandstones, including the arkoses, is an iron-rich chlorite, possibly chamosite. Honess suggests that the chlorite matrix in similar sandstones of the Stanley shale is derived from the alteration of an original ferruginous and argillaceous cement. Of all the sections of sandstone and arkose examined, including those from the strongly deformed southeastern part of the area, none could be classified as a truly metamorphosed rock. However, all the quartz grains show strain shadows, and many are crushed, granulated, and sheared.

The layers of white quartzite in the southeastern part of the area are quite different from the other sandstones of the Tesnus formation. They consist almost entirely of quartz, and there is no chloritic or argillaceous matrix. Some of the quartz grains are rounded and others angular. There is some variation in the size of the fragments, so that the smaller grains fill the interstices between the larger ones. There has also been some secondary regrowth of crystals. The whole rock is therefore an interlocking mass of quartz grains with very little matrix.

FOSSILS AND AGE

The Tesnus formation contains a few plant remains and Foraminifera in the upper part, but otherwise no fossils have been found in it. The largest collections have been obtained from the north end of West Bourland Mountain, 10 miles southwest of Marathon, from arkosic sandstones 400 feet below the top of the formation (bed 17 of section on p. 59; U. S. Geol. Survey localities 7937 and 8093). Material was gathered here by Sidney Powers in 1929 and by David White and the writer in 1930. Dr. White examined both collections and reported as follows:

All specimens are badly water-worn, fragmentary, and even abraded. The identifications are therefore largely tentative.

• 1. Neuropteris gigantea?Nearly complete pinnule, certainly belonging to the Neuropteris gigantea type. In the Appalachian trough this is mainly confined to the middle Pottsville.
• 2. Neuropteris sp. Tip of a narrow pinnule with extremely oblique nervation, suggesting N. obliqua.
• 3. Alethopteris. Indeterminate fragment of very narrow compact form close to a species in the middle Pottsville of the Appalachian trough.
• 4. Numerous fragments of fern petioles, very badly abraded. One specimen strongly suggests Heterangium.
• 5. Calamites node and portion of two internodes tentatively referred to Calamites ramifer.
• 6. Calamites cf. C. cistii.
• 7. Asterophyllites stems.
• 8. Leaves probably belonging to an Asterophyllites, like Asterophyllites gracilis.
• 9. Lepidodendron clypeatum? A few detached and deformed bolsters probably belonging to this species or included under the comprehensive name Lepidodendron obovatum.
• 10. Lepidostrobus sp. Fragment of axis from which bracts and sporangia have been stripped. It suggests Lepidostrobus variabilis, though the axis is rather thick for that species.
• 11. Stigmaria sp. Scattered cicatrices left by rootlets.
• 12. Cordaites sp. Fragment of narrow leaf not specifically determnable.
• 13. Artisia sp. Pith of Cordaites stem.
• 14. Cardicocarponsp. Nucleus indeterminable.
• 15. Trigonocarpum cf. T. ampulliforme.
• 16. Trigonocarpum. Very small ovate form, possibly undescribed.

In discussing the age of this flora, Dr. White stated that

The characters exhibited by the calamarian forms, the pteridosperms, and the Lepidodendron show that the beds from which the plants were obtained are undoubtedly Pennsylvanian. The gigantea type of Neuropteris, the Alethopteris, and to some extent the Lepidodendron point toward a middle Pottsville age, though the evidence is not absolutely conclusive. Viewed broadly, the flora is Westphalian; certainly it cannot be older than Namurian.

He also stated that the flora is clearly younger than that of the Jackfork sandstone of Oklahoma but probably older than that of the Atoka formation.

Plants have also been collected by Sidney Powers from the upper part of the Tesnus formation south of Tesnus station. From his collection at this place, Dr. White identified the tips of some leaflets as either Neuropteris or an elongated Cardiopteris. "I think it is Neuropteris and referable either to Neuropteris gigantea of the middle Pottsville or to Neuropteris capitata of the upper Pottsville."

Foraminifera have been obtained by Bruce Harlton from shale samples collected near the top of the Tesnus formation 18 miles east of Marathon. They are said to be of "lowermost Pennsylvanian age" and to be identical with "microfossils from the Caney shale in Oklahoma."

In a calcareous sandstone on the north slope of West Bourland Mountain below the plant horizon were collected what appear to be sponge spicules. According to G. H. Girty, they "consist of slender rods, circular in section and showing no appreciable taper. They have all the characters of a not uncommon type of sponge spicule except that they are black and shining, as if the material of which they are composed was phosphatic instead of siliceous." These fossils afford no evidence as to the age of the sandstone.

"

---

Honess, C. W., Geology of the southern Ouachita Mountains of Oklahoma: Oklahoma Geol. Survey Bull. 32, p. 189, 1923.

Letter from David White to Sidney Powers, July 1930.

Powers, Sidney, Age of the folding in the Oklahoma Mountains: Geol. Soc. America Bull., vol. 39, p. 1066, 1928.

 

62

STRATIGRAPHIC RELATIONS

The Tesnus formation is probably conformable with the Dimple limestone, which overlies it. In nearly every locality the two formations are separated by a transition zone of interbedded limestone and shale. In the northwestern part of the area the lower beds of the Dimple are conglomeratic and the Tesnus is only a few hundred feet thick. This is probably not the result of a break between the two formations and appears to be caused by a northwestward overlap of the Tesnus formation on the Caballos novaculite. The shales at these northwestern localities are not the basal shale member as recognized in the southeastern part of the basin, for they can be traced from section to section till they interfinger with the plant-bearing sandstones of the upper Tesnus along Pena Colorada Creek. They do not resemble the basal shales in lithology.

DIMPLE LIMESTONE

GENERAL FEATURES

The Dimple limestone was named by J. A. Udden or exposures in the Dimple Hills, in the northeastern part of the Marathon Basin, 20 miles northeast of Marathon. The massive and resistant limestones of the formation rise as low monoclinal ridges above the waste-mantled plains of the basin. The formation is widely exposed in narrow sinuous belts of outcrop in the eastern part of the area and extends from the Dimple Hills southward beyond Haymond station. It is also found in scattered synclinal areas along Dugout and Peña Colorada Creeks, in the western part of the Marathon Basin. The southeasternmost exposures are small patches surrounded by outcrops of the Tesnus formation in the faulted complex of Hells Half Acre and Devils Backbone.

The formation is composed of limestone in moderately thick beds. Most of the limestone beds are gray, granular, and somewhat sandy, with scattered seams of chert pebbles. Other beds are dense and very bituminous. The limestone contains a sparse fauna of marine invertebrates. In the eastern part of the Marathon Basin there is much shale in the upper and lower parts of the formation, so that it grades by transition zones into the clastic rocks of the Tesnus and Haymond below and above. The boundaries of the formation are drawn at the highest and lowest limestone beds.

LOCAL FEATURES

Eastern part of Marathon Basin.

-At the type locality, in the synclinal area of the Dimple Hills, the Dimple limestone is over 1,000 feet thick (sec. 8, pl. 8). The top is not exposed here, though the highest beds evidently belong to the upper transition zone.

The formation is thinner toward the south (sec. 6-8, pl. 8). Baker and Bowman report 925 feet in the water gap traversed by the Southern Pacific Railroad 132 miles west of Haymond station, and the writer measured only 497 feet at the water gap in the Haymond Mountains 4 miles south of the station. The lower 50 feet and the upper 150 feet in this region consist mostly of dark indurated shales, with a few thin beds of limestone. The main mass of the formation consists of dark-gray, finely granular limestone, weathering gray or yellow, in beds 1 to 4 feet thick, with many partings of dark indurated shale. Some of the limestones are banded by seams of brown chert, part of which are arranged in domelike concentric structures, possibly of organic origin. There are some layers of thinly laminated flaggy gray or bluish dense limestone and many thin interbedded layers of gray dull chert. In the lower part some of the yellow-weathering limestones contain granular seams crowded with fragments of brachiopods and bryozoans. A few crushed and flattened shells of ammonoids were collected from the flaggy layers. Many of the limestone beds contain fine sand, but there is very little chert conglomerate in the southern part of the area. In the excellent exposure on the Sanderson road, 15 miles east of Marathon, there are four or five thin beds of gray clay between the limestone layers.

The following section was measured in the water gap 4 miles south of Haymond station (sec. 6, pl. 8).

"

---

Udden, J. A., Baker, C. L., and Bose, Emil, Review of the geology of Texas: Texas Univ. Bull. 44, 1st ed., p. 46, 1916.

King, P. B., Geology of the Glass Mountains, pt. 1: Texas Univ. Bull. 3038, pp. 36-38,1930.

Baker, C. L., and Bowman, W. F., op. cit. (Texas Univ. Bull. 1753), p. 105.

King, P. B., op. cit., p1. 2, A.

 

 

62b

 

 

 

62e

 

63

Western part of Marathon Basin

-In the western part of the Marathon Basin the Dimple limestone is only about 100 feet thick (sets. 1-5, pl. 8), but it is similar to the beds exposed farther east. It consists mainly of dark-gray granular limestones, which crop out in ledges several feet thick (pl. 6, C) and contain seams of fine chert pebbles and crinoidal fragments. There are many layers of thin-bedded or flaggy sandy brown limestone and siliceous limestone, and some beds of indurated greenish shale. Some of the bedding surfaces of the flaggy limestones are marked by a net of winding trails or fucoidlike markings.

Along Peña Colorada Creek the limestones are succeeded by shales belonging to the Haymond formation, but farther northwest, along Dugout Creek, they are overlain by about 200 feet of indurated greenish shales and thin beds of green compact sandstone in which there are half a dozen ledges of brown sandy and pebbly limestone (fig. 21). Although these beds are classed with the Dimple limestone, they may well be equivalent to the lower part of the Haymond formation farther south.

A short distance north of the Roberts ranch there is at the base of the Dimple limestone a lenticular layer of conglomerate as much as 10 feet thick, composed of subangular cobbles and pebbles of Caballos chert and novaculite and of Maravillas limestone and chert in a siliceous matrix. About 3 miles north of the Roberts ranch two other thinner conglomerate beds are found higher in the formation. The fact that the Tesnus formation beneath is almost cut out by overlap suggests that these conglomerates were derived from an exposed surface of the pre-Carboniferous rocks farther north, on which the Tesnus formation was never deposited. In the outliers of the Dugout Creek overthrust sheet northeast of Lenox the Dimple contains thick beds of angular chert breccia cemented by silica, and nearly all the limestones are conglomeratic.

The following section was measured in the southern part of the Payne Hills, 5 miles north-northeast of the Roberts ranch (sec. 1, pl. 8, and fig. 21).

 

64

FOSSILS AND AGE

Fossils are not abundant in the Dimple limestone. Many of the layers appear to be composed of the fragments of shells, but these are so finely comminuted that no identifiable material is preserved. Thin sections of the limestone show many such shell fragments.

Foraminifera (other than fusulinids) are reported by Harlton from shales interbedded with the limestone layers of the Dimple Hills and of the exposures 18 miles east of Marathon. They were also noted in a thin section from the basal layers south of Iron Mountain. Harlton correlates the microfauna studied by him with that of the Marble Falls and Wapanucka limestones.

Megafossils have been found at several localities in the formation, but only a few specimens have been found at each place, and none of these are well preserved. The largest collection was obtained by G. H. Girty and the writer in 1929 from the lower part of the formation, 131~ miles west-northwest of Haymond station, to the north of the Southern Pacific Railroad (U. S. Geological Survey locality 6707). From this collection Mr. Girty has identified the following fossils:

• Orbiculoidea sp.
• Chonetes aff. C. arkansanus Girty.
• Chonetes n. sp.
• Productus aff. P. cora D'Orbigny.
• Pustulan. sp.
• Pustula aff. P. globosa Mather.
• Pustula sp.
• Hustedia multicostata Girty?.

In commenting on this collection, he states:

It is not reminiscent of the more common Pennsylvanian faunas and is almost equally lacking in definite alliance to the more common Mississippian faunas. Very tentatively I am considering it as of Pottsville age, although one of the Producti is almost identical with a Mississippian species. The new species of Pustula is very close to an undescribed but distinctive productid that occurs in the earlier Mississippian faunas of Oklahoma and Arkansas. The remainder of the fauna does not very well bear out the relation suggestec' by that resemblance but appears to be more closely affiliated with Pottsville faunas.

From the southwest side of the Dimple Hills, 1 mile east of the Arnold ranch (U. S. Geological Survey locality 7084), the following fossils have been collected:

• Clisiophyllum sp.
• Cystodictyasp.
• Crinoid stem.
• Minute gastropods.
• Coeloconus sp.

At three localities poorly preserved ammonoids have been found in the Dimple limestone. A collection from the basal layers of the formation on West Bourland Mountain (U. S. Geological Survey locality 6919) and one from a similar horizon 6 miles south of Haymond station (U. S. Geological Survey locality 6918) each contain large crushed shells which not improbably belong to the same species. According to Mr. Girty, "they have a rather wide umbilicus, suggesting Gastrioceras, but the form of the suture cannot be determined." A collection from bed 24 of the section 4 miles south of Haymond station (U. S. Geological Survey locality 6917) (p. 62) contains two specimens of a much smaller ammonoid, which is completely flattened in stiff shale. These specimens "show the sculpture but afford no evidence that the shell was originally chambered. The sculpture, which consists of strong, rather coarse revolving costae, together with such other characters as are preserved, recalls Gastrioceras caneyanum (Girty)." This same collection also contains small three-rayed sponge spicules.

The other fossil collections each contain only a few poorly preserved specimens. One from exposures west of Iron Mountain contains sponge spicules, crinoid stems, and a species of Spirifer. Another, 2½ miles north-northeast of the Roberts ranch, contains a zaphrentoid coral, crinoid stems, Streblotrypa sp., and Productus sp. A collection from the lower part of the formation 15 miles east of Marathon on the Sanderson road contains Rhombopora? sp., and one from a similar horizon 5 miles southwest of Haymond station contains Amplexus? sp.

In summarizing the collections from the Dimple limestone, Mr. Girty writes:

The fossils in the Dimple suggest a Pottsville age. At all events, they apparently must represent an unusual facies if the horizon from which they came is regarded as post-Pottsville.

STRATIGRAPHIC RELATIONS

The contact between the Dimple and Haymond formations, as exposed north and south of Haymond station and near Peña Colorada Creek southwest of Marathon, is apparently gradational, and the Dimple limestone is separated from the shales and sandstones of the Haymond formation by several hundred feet of transition beds.

HAYMOND FORMATION

HISTORICAL SUMMARY

The name "Haymond formation" was applied by Baker in 1916 to exposures in the synclines near Haymond station, in the eastern part of the Marathon Basin, but there has been considerable doubt as to the status of the unit, because Baker later suggested that it might be a part of the Tesnus formation that "

---

King, P. B., Geology of the Glass Mountains, pt. 1: Texas Univ. Bull. 3038, p. 38, fig. 13 (middle left), 1930.

Powers, Sidney, Age of the folding in the Oklahoma Mountains: Geol. Soc. America Bull., vol. 39, p. 1066, 1928.

Udden, J. A., Baker, C. L., and BOse, Emil, Review of the geology of Texas: Texas Univ. Bull. 44, 1st ed., p. 46, 1916.

Baker, C. L., and Bowman, W. F., Geologic exploration of the southeastern front range of trans-Pecos Texas: Texas Univ. Bull. 1753, p. 107, 1917. Baker, C. L., Date of major diastrophism and other problems of the Marathon Basin: Am. Assoc. Petroleum Geologists Bull., vol. 12, p. 1114, 1928.

 

66

had been overthrust across the Dimple. Field work by R. E. King and the writer in 1927 indicated that the Haymond formation overlay the Dimple without such a structural break, and its validity was demonstrated beyond question in 1930 with the discovery of boulders of Tesnus and Dimple rocks in the boulderbed member of the younger formation. In the last few years there has been much interest in the formation on account of the large boulders which are found in it.

GENERAL FEATURES

The Haymond formation is a mass of shales and sandstones whose thickness locally exceeds 3,000 feet. Although sections have been found which extend through part of the Haymond up to the Gaptank formation, or down through part of it to the Dimple limestone, no exposure is known where there is a complete and uninterrupted sequence from the base to the top of the formation (pl. 8). The Haymond beds crop out in disconnected synclinal areas in various parts of the Marathon Basin but are most extensively exposed between Haymond station and Gap Tank. Because of their generally nonresistant character, they form low rolling hills or plains and are extensively mantled by gravel and wash.

Most of the formation consists of layers of sandstone and carbonaceous shale a fraction of an inch to several inches in thickness, in regular, rhythmic alternation (pl. 9, B, C, D). There are a few thicker sandstone beds at long intervals, and near the base some thick bodies of shale. The peculiar stratification of the Haymond formation is quite unlike anything seen in the Tesnus, with which it has previously been confused. At a few places in the region the Haymond sandstones and shales have thin intercalations of sandy limestone, some of which contain marine invertebrate fossils.

The upper part of the formation contains thick layers of massive arkose, and in the syncline east of Haymond station there are several members of boulder-bearing mudstone as much as 150 feet thick. The boulders in the mudstone beds include erratic blocks of the older formations as much as 130 feet across (pl. 11, C-F). Southeast of Gap Tank a thinner boulder bed lies at about the same stratigraphic position.

HAYMOND AREA

General relations

-The Haymond formation is exposed in the two synclines northwest and southeast of Haymond station. Both synclines are broad and .open but are overturned and faulted on the southeast. In the southeastern syncline the Haymond formation crops out in a belt 1Y, miles in width, which is terminated on the southwest up the pitch of the fold by outcrops of the Dimple limestone. These curve around the axis and are cut off by a fault on the southeast. The strata on the northwest flank of the syncline dip southeastward in regular order at angles of 30° to 45° in the south (sec. A-A'; pl. 21), but with the inclination increasing gradually northward to about 70° near the Sanderson road. The strata on the southeast flank of the syncline have steep or overturned dips and are cut off a short distance southeast of the synclinal axis by a steep thrust fault that raises Tesnus sandstones against those of the Haymond. Southeast of Haymond station the outcrops of the boulder-bed member of the formation appear to lie along the trace of the synclinal axis, but farther north, near Housetop Mountain, the axis is in strata overlying the boulder bed to the east. Here the boulder bed and the rocks that enclose it are a part of a steeply tilted normal sedimentary sequence, without duplication by folding or faulting.

In the syncline east of Haymond station the sequence is roughly as follows (sec. 6-7, pl. 8):

Lower members of formation

-The lower four members of the Haymond formation crop out in a belt of low rolling hills about a mile wide east of Haymond station (pl. 24), where there are several excellent exposures in railroad excavations and in the cut banks of San Francisco Creek. This part of the formation is also exposed in the narrower syncline northwest of Haymond station.

"

---

King, P. B., and King, R. E., The Pennsylvanian and Permian stratigraphy of the Glass Mountains: Texas Univ. Bull. 2801, p. 113, 1928.

King, P. B., Baker, C. L., and Sellards, E. H., Erratic boulders of large size in the west Texas Carboniferous [abstract]: Geol. Soc. America Bull., vol. 42, p. 200, 1931. Sellards, E. H., Erratics in the Pennsylvanian of Texas: Texas Univ. Bull. 3101, pp. 1-17, 1931. King, P. B., Large boulders in the Haymond formation of west Texas [abstract]: Geol. Soc. America Bull., vol. 43, p. 148, 1932. Baker, C. L., Erratics and arkoses in the middle Pennsylvanian Haymond formation of the Marathon area; trans-Pecos Texas: Jour. Geology, vol. 40, pp. 577-607, 1932. Carney, Frank, Glacial beds of Pennsylvanian age in Texas [abstract]: Geol. Soc. America Proc., 1934, p. 70, 1935.

 

66

In the large cut on San Francisco Creek 3 miles south of Haymond station the rocks dip steeply eastward with some local folding (fig. 22). They consist of a rhythmic alternation of thin beds of shale and sandstone (pl. 9, B). The arenaceous beds are flaggy fine-grained, greenish sandstones or hard pale-greenish sandy shales, in ¼-inch to 2-inch beds. The argillaceous beds are black or dark-blue carbonaceous shales in layers of similar thickness. At intervals of 5 to 15 feet are beds of compact sandstone as much as a foot thick, some of which are current-marked and cross-bedded on a small scale. In other exposures the light-colored layers are sandy shales rather than sandstones. The sandstone beds break down into thin brown plates and flags, which cover the hillsides. The rhythmically bedded nature of the formation is not well revealed except on fresh exposures.

Boulder-bed member

-The term "boulder-bed member of the Haymond formation" is applied to a complex group of interstratified, thin-bedded sandstones and shales, massive arkose, and boulder-bearing mudstone, lying in the upper part of the formation east of the Marathon quadrangle. The southwesternmost exposures of the member are 3 miles southeast of Haymond station, and the northeasternmost are west of the summit of Gap Peak and about half a mile south of the Sanderson road (pl. 10). Between these points, a distance of 8 miles, the boulder-bed member crops out in a linear belt, interrupted only by strips of surficial deposits. The arkose layers in the member rise in low knobs and hogbacks, but the boulder-bearing mudstones are mostly worn down to valleys and lowlands. At one locality, due west of the summit of Housetop Mountain, large erratic blocks are so numerous in the boulder bed that the member stands up in a group of rugged hills (pl. 11, C; sec. D-D', fig. 23).

The boulder-bed member is about 900 feet thick west of the summit of Housetop Mountain, where it consists of five mudstone layers, from 25 to 150 feet thick, interbedded with thin-bedded sandstones and shales and ledges of massive arkose. Farther north two boulder beds at the base and one at the top merge into sandstones and shales along the strike of the rocks, and only the two middle layers persist to the northernmost exposures (sec. A-A', fig. 23). There are two layers of massive arkose at the top of the member in this region, and it is underlain by thinly bedded sandstones and shales. South of the latitude of Housetop Mountain the boulder-bed member is not well exposed for a distance of 3 miles, and the next continuous outcrop is south of the railroad. Here the main arkose layers underlie the boulder bed and there is only one mudstone layer, which pinches out between sandstones and shales 2 miles south of the railroad. This layer is probably on the axis of the syncline, and whatever higher beds once existed have been removed by erosion. Probably the mudstone of the southwestern exposures, like that of the northeasternmost, is correlative with the middle part of the boulder-bed member near Housetop Mountain, and the lower arkoses connect with the thin arkose layers between the lower mudstone members farther north.

The fragments in the boulder-bed member include a great variety of rocks. Among them are pre-Cambrian crystalline rocks, such as granite, aplite, pegmatite, vein quartz, rhyolite porphyry, quartz conglomerate, and possibly schist. No recognizable masses from the Cambrian or early Ordovician have been found, but there are a few blocks from the Maravillas chert. Numerous fragments from the Caballos formation include massive white novaculite, thinly laminated novaculite, varicolored banded cherts, in part contorted, and coarse chert breccias cemented by silica (pl. 11, C, E). The breccias resemble those seen along some of the thrust faults in the novaculite area. There are many fragments of dense, fine-grained greenish quartzitic sandstone from the Tesnus formation, as well as one or two blocks of its characteristic indurated green shale. Boulders from the Dimple limestone are rare except for a single large slab near the Bennett place, a thick-bedded, compact dark-gray limestone, with seams of chert conglomerate and thick shale partings. There are also boulders of black, yellow-weathering dense limestone, containing chert in irregular knotty masses (pl. 11, F). These boulders contain Pennsylvanian fossils, but the rock and its fauna are quite unlike any found in place in the Pennsylvanian succession of the Marathon Basin.

The most striking feature of the boulder-bed member is the large size of some of these embedded rock fragments. Some of the novaculite blocks are 50 feet across, and those of the fossiliferous Pennsylvanian limestone reach 100 feet in longer dimension. The largest block is the mass of Dimple limestone east of the Bennett place, which is 130 feet across (sec. E-E', fig. 23). The novaculite blocks are mostly angular, but some are spherical (pl. 11, D) ; all are nearly equidimensional. Most of the limestone blocks are angular, and many of them are slablike (pl. 11, F), being two or three times wider than thick. Nearly all the slabs lie parallel to the bedding of the matrix. Associated with the large blocks are innumerable smaller well-rounded to subangular pebbles and cobbles. Baker has found striations on a number of the smaller stones, particularly on the dense quartzites. One boulder of Maravillas black chert 18 inches long, which had numerous intercrossing striae on a flattened surface, was found by him near the Bennett place. The pre-Cambrian "

---

Baker, C. L., letter, February 1931.

 

67

stones have smoothed and rounded surfaces, but a large number have one or more flattened faces, and a few are striated. Many of the cobbles are broken and shattered by subsequent deformation of the rocks (pl. 11, E), and some of the larger novaculite blocks are slickensided.

The smaller fragments are evenly distributed along the outcrop of the boulder-bed member, but the larger masses are found only in groups and clusters (pl. 10). The clusters are separated by a mile or more of outcrop without any large masses. The largest cluster of boulders is west of the summit of Housetop Mountain,

 

68

where the boulder-bed member reaches its greatest thickness and has the greatest number of mudstone members (pl. 11, C). The large blocks at this and other places are found in all the mudstone layers but are slightly more abundant in the middle ones.

Except for a few small blocks embedded in the arkose layers, all the boulders lie in a matrix of massive dark greenish-gray mudstone, containing fine flakes of mica, grains of rotten feldspar, and a few wood fragments. The pebbles and boulders are sparsely set in the matrix, like plums in a pudding. Bedding planes are obscure or lacking in the mudstone, but in places lenses of well-bedded sandstone and arkose as much as several feet in thickness are enclosed. Some of these lenses show irregular dips, which may be original in the deposit.

In sharp contrast to the mudstone layers are the beds of stratified clastic rocks intercalated with them. There are several layers of thin-bedded sandstones and shales as much as 15 feet thick and many thin to thick ledges of coarse friable arkose. In places the arkose is pebbly, and rarely small boulders are embedded in it. The arkose contains a few imprints of Cordaitesstems. In places the arkose layers merge into lenses of chert and limestone breccia 5 to 10 feet thick, consisting of closely packed angular fragments 6 inches or less across with a small amount of interstitial arkosic matrix.

Upper member of formation

-The strata above the boulder-bed member are exposed only on the west side of Housetop Mountain. They show the rhythmic bedding characteristic of the lower part, but most of the light-colored beds are sandy shales, with sandstone .layers only at considerable intervals. They extend up to the axis of the syncline west of Housetop Mountain and are the highest Paleozoic strata exposed in the region. By what interval their highest parts are separated from the Gaptank formation, if this was ever laid down over them, is not known.

OTHER AREAS OF HAYMOND FORMATION

Gap Tank area

-In the vicinity of the old Clark ranch, 4½ miles southeast of Gap Tank (pl. 23), there are 1,800 feet of Haymond beds in a downward succession below the Gaptank formation, with the base of the formation not exposed (sec. 9, pl. 8). The lower part consists of flaggy sandstone and shale, with two thick members of massive arkose, of which the lower contains fossil plants.

In the upper part of the exposure, 400 feet below the basal beds of the Gaptank, is a boulder bed 10 to 25 feet thick which extends for 4 miles along the outcrop. It rests on a bed of coarse arkose and fine chert conglomerate. The fragments in the boulder bed are closely packed, and the interstices are filled by coarse arkosic grit. Most of the fragments consist of Caballos novaculite and bedded chert, some of which are contorted and brecciated. One slab of novaculite near the southeast end of the exposure measures 10 by 6 by 2½ feet (pl. 11, A), and angular masses as much as 3 feet across are common. There are also fragments of fossiliferous limestone from the Maravillas formation, some of which contain reef-like masses of bryozoans like those seen on the outcrop only in the southern part of the Marathon Basin. The bed also contains a few cobbles of Tesnus quartzite, rhyolite porphyry, and blocks of limestone with Pennsylvanian fossils. The boulder bed at this locality is notable for its isolation in a succession of finer-grained deposits. The fragments are somewhat different from those in the boulder bed near Haymond. It is true that novaculites and cherts are abundant in both, but this locality has only a few fragments of pre-Cambrian and fossiliferous Pennsylvanian limestone, whereas they are abundant in the southern exposure, which has only a few fragments of Maravillas limestone.

About 8 miles south of Gap Tank, in the syncline south of the Dimple Hills anticlinorium, there are extensive exposures of the lower part of the Haymond formation, but no outcrops of boulder beds. Near the anticlinal axis, on the Pecos-Brewster County line 3 miles southeast of the Dimple Hills, the thin-bedded sandstones and shales contain two intercalated layers of brown sandy and pebbly limestone from which E. H. Sellards and C. L. Baker have collected a few specimens of Fusulina.

Exposures west of Marathon

-Thin-bedded shales and sandstones belonging to the lower part of the Haymond formation are exposed in several synclinal areas southwest of Marathon and south of the Dugout Creek overthrust. These grade up from the Dimple limestone below, with thick beds of dark shale near the base. At the foot of the escarpment of the Cochran Mountains, 3 miles east of Del Norte Gap, are thick ledges of arkosic sandstone, with obscure plant remains.

Beneath the Dugout Creek overthrust, near Dugout Creek (locality A, pl. 16), are scattered exposures of the upper part of the Haymond formation, surrounded by the basal beds of the Gaptank formation (pl. 9, C, D). The rocks of this region are much folded and contorted, so that the detailed relations cannot everywhere be made out. At the best-known locality, on Dugout Creek 2½ miles south of the summit of Dugout Mountain, 250 feet of beds are exposed below the plane of the Dugout Creek overthrust. Most of the succession consists of dark carbonaceous shale and greenish sandy shale in alternating layers a fraction of an inch thick, with thin sandstone beds at intervals of 3 or 4 feet (fig. 24, B). Careful measurements of part of the exposure show an average of 185 alternations for each

 

69

10 feet of section (fig. 24, C). Near the top of the exposure are two thin layers of granular limestone containing fragments of brachiopods and crinoid stems, and lower down is a limestone lens that contains fusulinids. In the lower part, on the banks of Dugout Creek, are several thick beds of sandstone, some of which contain plant remains. The fossil evidence from this locality, some of which is conflicting, is considered on pages 71-72.

MICROSCOPIC CHARACTER

Arkosic sandstones

-Three thin sections of the coarse arkosic sandstones of the Haymond formation were examined. One of these was from the arkose beds below the boulder-bearing mudstone southeast of Haymond station. The other two were from layers southeast of Gap Tank, one from immediately below the boulder-bed member and the other from a horizon 1,000 feet lower in the section. These sections are all of similar character. They consist of about 60 percent of quartz grains; the remaining constituents are mostly small chert fragments and granitic detritus. The quartz grains in the thin section from the Haymond station locality are shattered and show strain shadows, but those in the sections from localities farther north where the rocks are not so greatly folded show less of the effects of deformation.

The section from the beds southeast of Haymond and the one from the lower horizon southeast of Gap Tank both show moderately coarse-grained sandstones, made up of subangular grains of various minerals, with a few well-rounded grains. Many of the quartz grains show small included blades of apatite. Biotite and muscovite plates are abundant and are in places bent and twisted about the other grains. Feldspar grains are common and are fresh and unaltered except for a small amount of sericite along the cleavage planes. A few rectangular cleavage fragments were noted. The most abundant feldspar mineral is plagioclase, but the section from the northern locality contains several grains of microcline. There are a few fragments of slaty rock and of chert and a few grains of unaltered hornblende and minerals of high refractive index. The matrix makes up only a small part of the rock and is chloritic and argillaceous.

The rock from the higher horizon southeast of Gap Tank is of coarser grain and contains many large angular fragments of cryptocrystalline chert. Otherwise its -constituents are much the same as those of the other two rocks; it consists predominantly of quartz, with lesser amounts of biotite, muscovite, microcline, and sodic plagioclase. A few minute grains of colorless garnet and a single larger fragment of hornblende were noted.

The constituents of these sandstones suggest that they were derived from the weathering of granitic

 

70

rocks. The apatite-bearing quartz and the microcline are characteristic of granites, and the occurrence of sodic plagioclase, biotite, and muscovite tend to sustain this conclusion, though they may also occur in metamorphic rocks. These characteristic constituents are somewhat different from those of the Tesnus sandstones, in which fragments derived from metamorphic rocks predominate over those of igneous rocks.

Thin-bedded sandstone and shale

-A thin section several inches square was cut from a specimen of thinbedded sandstone and shale collected in the cut on Dugout Creek 2½ miles south of Dugout Mountain. As seen in the section, the rock is composed of alternating layers of dark shaly and light sandy material from an eighth to half an inch in thickness. Most of these layers are continuous across the slide and through the hand specimen from which it was cut, but in detail there is considerable variation in their form and thickness.

Several dark argillaceous bands 4 millimeters or less in thickness cross the slide. These consist largely of argillaceous material, but there are a few embedded quartz silt particles as much as 0.015 millimeter in diameter. The argillaceous bands grade through a small thickness into the sandy beds at the top and base.

The lighter sandy bands contain a large number of subangular quartz particles. These range from 0.045 to 0.120 millimeter in diameter, or from "coarse silt" to "fine sand" in Udden and Wentworth's classification. Most of the fragments are nearer the minimum than the maximum, but there are a few larger grains. The quartz grains are embedded in a small amount of argillaceous matrix.

Within the sandy layers there are many thin bands and streaks of argillaceous material, some of which are no thicker than the diameter of the larger quartz grains. In parts of the slide they form sharply marked, regular layers, which alternate with sandy beds several millimeters thick. A larger number are not at all regular but form streaks and shreds, embedded in sandy material, and are in part considerably contorted. The irregularity of these parts may have resulted from secondary disturbance of an original laminated deposit, perhaps by gliding or compaction shortly after deposition.

It is perhaps desirable at this place to summarize the character of the thin-bedded sandstones and shales as seen on the outcrop, in hand specimen, and in thin section. Rocks of this type showing great thicknesses occupy wide areas in the Haymond formation in the Marathon Basin. The specimen studied in thin section has many of the characters of the strata found elsewhere in the region but differs somewhat from most of them in that the sandy beds are finer-grained and more argillaceous.

On the outcrop rocks of this type are found to consist of layers, several inches in maximum thickness, of sandy and shaly material in repeated rhythmic alternation. Each layer persists through the whole length of the exposures that have been studied. Measure ments at the locality 2½ miles south of Dugout Mountain show an average of 185 alternations in each 10 feet of section, or an average thickness of 0.27 inch for each sandy or shaly layer. The thickness of the individual beds, however, is far from regular. In parts of each exposure there may be groups of four or more thick shaly beds, separated by thin sandy beds, and in other parts there may be similar groups of thick sandy beds. At intervals of 5 to 20 feet in each exposure there are also sandstone beds as much as a foot thick, which stand out prominently as ledges.

In the hand specimens and under the microscope the thin sandy beds seen on the outcrop contain shaly laminae only a fraction of a millimeter in thickness. So far as these could be observed, they are rarely continuous, and many of them appear to be contorted and broken.

The sandy layers in the rock are noncarbonaceous and consist of fine sand or coarse silt, composed of quartz fragments, with a small amount of argillaceous material. The texture varies in different parts of the section and different parts of the region, so that in some places the beds are fine sandstones, and in others sandy shales. The grain size appears to be somewhat greater in the thicker sandstone beds, for even where the thin sandy layers are argillaceous, the thicker layers are true sandstones. None of the sandstones, however, attain a noticeably coarse or gritty texture. The shaly layers consist chiefly of argillaceous material, though under the microscope a few grains of quartz silt can be seen. They are prevailingly dark-colored and apparently contain a certain amount of carbonaceous material.

Boulder-bearing mudstone

Thin sections of the mudstones of the boulder-bed member were cut from fresh specimens collected on the dump of a water well half a mile south of the Sanderson road, near the north end of the exposures of the member (fig. 25). Specimens of the mudstone were also collected by C. L. Baker from a locality several miles farther south and were studied by F. J. Pettijohn of the University of Chicago.

The greater part of the rock consists of fine angular mineral and rock particles, which are evenly dispersed in an opaque argillaceous matrix. The matrix makes up about 30 percent of the rock. Most of the fragments "

---

Twenhofel, W. H., Treatise on sedimentation, 2d ed., p. 202, 1933.

Baker, C. L., op. cit. (Jour. Geology, vol. 40), p. 585.

 

71

have right- or acute-angled corners. There are a few rounded grains, which contrast strongly with the angularity of the rest. The fragments are poorly sorted. Most of them range from 0.075 to 0.225 millimeter in diameter, or from "very fine sand" to "fine sand" in Udden and Wentworth's classification. Other fragments are larger. There are a great many as much as 1 millimeter in diameter, and one slide also contains four subangular pebbles 5 to 10 millimeters across.

About half of the grains in the rock are quartz, a few of which contain apatite inclusions. Most of the quartz shows faint strain shadows, but none of the grains have been shattered by deformation. About a quarter of the grains are feldspar, most of which is plagioclase. The feldspar is very little altered, except for some sericite along the cleavage planes. There are a few chips of slaty and schistose rock and of chert. The larger pebbles consist of finely crystalline fossiliferous limestone. There are also a few grains of muscovite, biotite, garnet, and calcite.

At one place in the slide is a lens or streak of argillaceous material about an inch in length, which tapers at the ends into a thin seam that passes between the granular parts of the rock. The lens contains only a few fine quartz grains and shows a faint extinction in one direction as a result of the parallel orientation of its clay particles. Aside from this argillaceous lens, there is no evidence of stratification in the mudstone.

FOSSILS AND AGE

Indigenous fossils

-The Haymond formation contains few indigenous fossils. At some localities fossil plants have been collected, though so far only two, places have afforded much material.

In the middle part of the formation, about 1,200 feet below the boulder-bed member, 3 miles southeast of Gap Tank, near the old Clark ranch, David White and the writer collected plants from ledges of arkosic sandstone (U. S. Geological Survey locality 8089). Dr. White reported as follows on this collection:

The fossil-plant fragments were very badly ground up before indiscriminate mixture in sand. No fragments of ferns or fernlike leaves are present. All plant debris is decorticated. The examination of the material permits the recognition of the following forms:

A fragment of a calamarian stem with open nodes, alternating ribs, and well-developed leaf scars. It is possibly referable to Calamites suckowii.

Another incomplete internode of a calamarian stem is comparable to Calamites cistii.

Also there is present a slender stem apparently belonging to a defoliatedAnnularia.

Two seeds, one without envelope and one lacking an apex, represent the nuclei of Cardiocarponhi. These nuclei agree in size and shape with a type common in the Pennsylvanian and especially in the Pottsville. They are rather nearly comparable to the nuclei of Cardiocarpon fluitans.

The alternate arrangement of the costae at the nodes in the calamarian and annularian fragments leaves no doubt as to the Pennsylvanian age of the formation. The aspect of the Cardiocarpon remains strongly suggests a stage somewhere in the Pottsville.

At a locality about a mile farther east and at about the same stratigraphic level Sidney Powers collected Artisiasp., Cardiocarpus sp., Lepidodendron sp., and Calamitessp., which were identified by C. B. Read.

From arkoses near the top of the boulder-bed member west of Housetop Mountain Sidney Powers collected Calamites sp., Artisia sp., Lepidodendron sp., Trigonocarpus sp., and Cordaites sp., which were also identified by Mr. Read.

From a layer in the bed of Dugout Creek, 2½ miles south of the summit of Dugout Mountain, David White and the writer made a rather extensive collection of plants (U. S. Geological Survey locality 8094). Dr. White reported as follows on this material:

The material is extremely fragmentary. It is comminuted in current-bedded sand. Following is a preliminary list of the species in hand:

• Neuropteris, part of a broad pinnule.
• Neuropteris, point of a narrow pinnule.
• Dictyoxylon stem.
• Stigmaria scars.
• Cordaites, fragment of leaf; probably new species.
• Carpolithus transsectus?
• Cardiocarpon seeds, including a remnant belonging probably to Cardiocarpon elongatum.
• Trigonocarpum, slender triangular form; deformed.
• Lepidostrobus, part of slender axis.

The fragments are for the most part too incomplete for specific determination, and the collection therefore has little

 

72

diagnostic value. It is fairly evident, however, that the collection is of Pottsville age, and probably it belongs in the upper Pottsville.

In a limestone bed 30 feet higher the writer collected a fossil identified by C. O. Dunbar as a small species of Triticites, whose "age should be younger than the oldest Gaptank and younger than the Strawn." This fossil suggests that a part of the exposure at this place is of Gaptank rather than Haymond age. As shown in figure 24, A, the structure is complex, and the limestone with Triticites may somehow have been included tectonically in older rocks. A few fragmentary remains of Productus and Spirifer have been found in thin limestones several hundred feet above the plant horizon on Dugout Creek, but these layers have yielded no specifically identifiable material.

Marine invertebrate fossils are rare in the Haymond formation. Fusulinids were collected by E. H. Sellards and C. L. Baker from a thin bed of brown sandy lime-stone in the middle part of the formation 3 miles southeast of the Dimple Hills, on the Pecos-Brewster County line. These have been determined by C. O. Dunbar as a Fusulina, probably belonging to a new species. The genus Fusulina ranges no higher than the Strawn group of central Texas and the upper part of the Des Moines group of the northern midcontinent area, and its occurrence in the Haymond indicates an early Pennsylvanian age for that formation.

Fossils of the exotic blocks

-The exotic blocks in the boulder-bed member of the Haymond formation east of Haymond station consist of rocks of a great variety of ages and lithologic types. Most of these are easily identified with formations that crop out in the Marathon Basin, but the numerous masses of limestone that contain Pennsylvanian fossils are unlike anything seen in place in the region.

The collections from the exotic blocks were made by J. Brookes Knight and the writer and were identified by G. H. Girty. The largest collection was obtained from a limestone block 50 feet long about 1 mile northwest of the summit of Housetop Mountain (U. S. Geological Survey locality 6999) and included the following fossils:

• Cristellaria sp.
• Foraminifera undet.
• Sponge spicules.
• Sponge?
• Productus coloradoensis Girty.
• Productus aff. P. pertenuis Meek.
• Spirifer aff. S. opimus Hall.
• Nucula anodontoides Meek?
• Leda bellistriata Stevens?
• Parallelodon sp.
• Aviculopecten, 2 sp. undet.
• Myalina swallowi McChesney.
• Astartella sp.
• Bellerophonaff. B. incomptus Gurley.
• Worthenia aff. W. speciosa (Meek and Worthen).
• Pleurotomaria aff. P. fisheriSayre.
• Pleurotomaria aff. P. ornatiformis Morningstar.
• Pleurotomaria, 2 sp. undet.
• Goniasma aff. G. lasallensis Worthen.
• Goniasma n. sp.
• Aclisina aff. A. swallowana (Geinitz).
• Naticopsis aff. N. nana (Meek and Worthen).
• Trachydomia n. sp.
• Orthonema aff. 0. carbonarium (Worthen).
• Holopea? n. sp.
• Euomphalus sp.
• Flemingia? n. sp.
• Cyclonemasp.
• Meekospira aff. M. peracuta (Meek and Worthen).
• Pseudozygopleurao aff. P. peorense(Worthen).
• Rotalina? n. sp.
• Zygopleura aff.Z. parva (Cox).
• Griffithides sp.
• Ostracodaundet.

The following collection was obtained from a block about 1 mile west of the summit of Housetop Mountain (U. S. Geological Survey locality 6909):

• Lithostrotion? sp.
• Fenestella sp.
• Cystodictya? sp.
• Productus coloradoensisGirty?
• Productus sp.
• Euomphalus sp.
• Trachydomia n. sp.

The following collection was obtained not far from the last, from an exotic block 70 feet in length (U. S. Geological Survey locality 7073):

• Chaetetes milleporaceus Milne-Edwards and Haime.
• Triplophyllum sp.
• Glyptopora sp.
• Productus aff. P. altoensis Norwood and Pratten.
• Pustula aff. P. wallaciana (Derby).
• Edmondia sp.

Mr. Girty makes the following comments on these collections:

These three collections rather clearly represent two different faunal facies, the first two belonging together and the last containing a different fauna. Although different in facies, the two faunas are not necessarily widely different in geologic age, and, so far as the facts are known to me, both faunas find their nearest allies in the Pottsville. It should be understood that neither of them is an exact duplicate of any Pottsville fauna that I have seen; instead, they merely contain a few of the same or comparable species.

The first fauna is by far the most extensive and consists largely of gastropod shells. Brachiopods, which in most faunas of Pennsylvanian age excel in abundance most other classes, are rare, and pelecypods scarcely less so. The gastropods, on the other hand, are abundant, and by a somewhat arduous process a large number of individuals have been recovered. They are also richly varied, though mostly ranging in size from small to minute. Certain collections from the Marble Falls limestone resemble this fauna in general makeup and also in a community of similar or identical species. How many species are identical "

---

Dunbar, C. 0., letter, August 1934.

 

73

or merely similar I am not prepared to say, for the material from the Marble Falls limestone is not well preserved, and the specimens fail to show clearly the sculpture, which, in its scale, corresponds to the diminutive size of the specimens themselves.

Such comparisons as are suggested by the faunas of these three collections reach out, not to the fauna of any one Pottsville formation, but now to the fauna of the Marble Falls, now to that of the Smithwick shale, now to that of the Wapanucka limestone, or more remotely to that of the Morrow formation. Thus the most abundant gastropod in the first collection is a small Trachydomia, which has not been observed in the other Pottsville formations mentioned but which may be represented by a few specimens from the Morrow formation. The fact that the gastropods of all the Pottsville faunas are as yet largely unstudied snakes this evidence difficult to evaluate. On the other hand, the richly varied brachiopods of these Pottsville formations, which are fairly well known, do not appear in the collections under consideration in sufficient variety to afford much evidence in correlation.

A smaller collection, made by C. L. Baker and E. H. Sellards, has been examined by F. B. Plummer, of the Bureau of Economic Geology, University of Texas. Plummer notes its resemblance to a collection obtained from cores in the Big Lake oil field, northeast of the Marathon Basin. Baker, on seemingly meager evidence, states that the fauna "is definitely middle Pennsylvanian" and suggests its correlation with some part of the Strawn or Canyon groups of central Texas. There is thus a possibility that rocks of more than one age are represented in the limestone blocks. The writer has the impression, however, that most of the collections made by Baker and Sellards came from the same fossiliferous blocks as those that yielded his own collections; and until more definite evidence is obtained, it is thought more likely that all the collections are of early Pennsylvanian age, as suggested by Mr. Girty.

STRATIGRAPHIC RELATIONS

The contact of the Haymond with the Gaptank formation is evidently conformable. The Chaetetes-bearing limestone at the base of the Gaptank contains no basal conglomerate. The contact between the two formations is best exposed along the road between Marathon and Fort Stockton, 2 miles south-southwest of Gap Tank. A slight divergence in dip between the thin-bedded sandstones and shales of the Haymond and the Chaetetes-bearing limestone of the Gaptank can be seen here, but this appears to be the result of slipping of competent limestone over incompetent sandstones and shales during deformation. The conglomerate beds that are a conspicuous part of the middle layers of the Gaptank formation nowhere extend to its base. The first appearance of conglomerates in the Gaptank is several hundred feet above the base.

GAPTANK FORMATION

HISTORICAL SUMMARY

The name "Gaptank formation" was applied by Udden in 1916 to exposures of upper Pennsylvanian strata at Gap Tank, in Stockton Gap, 23 miles northeast of Marathon (pl. 14). The original definition evidently included the strata now known as the Wolfcamp formation, which were separated in the following year as a result of studies by Böse of the fauna of the Uddenites zone and higher beds at Wolf Camp, in the Glass Mountains, which he considered to be of Permian age. Further knowledge of the fossils near the Pennsylvanian-Permian boundary obtained in recent years has led to the conclusion that the Uddenites zone is of Pennsylvanian age, as first suggested by Keyte, Blanchard, and Baldwin, and the top of the Gaptank formation is now placed above this zone.

West of Marathon are exposures of strata, here considered to be a part of the Gaptank formation, which were variously placed in the Tesnus, Haymond, and Gaptank formations by Baker, Böse, and Udden. A separate formation name ("Dugout beds") has also been proposed for them. In 1927 Schuchert expressed the opinion that these beds were "unmistakably lower Gaptank", an interpretation for which much confirmatory evidence has been gained by the writer and others.

GENERAL FEATURES

The Gaptank formation crops out only in the northern part of the Marathon Basin. In the area to the south only the older formations remain, and if the Gaptank formation was ever laid down there it has since been removed by erosion. The type area of the Gaptank lies north of the region described in this report, but the problematic strata exposed on Dugout Creek "

---

Sellards, E. H., Erratics in the Pennsylvanian of Texas: Texas Univ. Bull. 3101, pp. 15-16, 1931. See also Baker, C. L., Erratics and arkoses of the middle Pennsylvanian Raymond formation: Jour. Geology, vol. 40, p. 590, 1932.

Baker, C. L., op. cit., p. 590.

Udden, J. A., Baker, C. L., and Bose, Emil, Review of the geology of Texas: Texas Univ. Bull. 44, 1st ed., p. 47, 1916.

Udden, J. A., Notes on the geology of the Glass Mountains: Texas Univ. Bull. 1753, pp. 38-43,1917.

Bose, Emil, Permo-Carboniferous ammonoids of the Glass Mountains: Texas Univ. Bull. 1762, p. 16, 1917.

Keyte, I. A., Blanchard, W. G., and Baldwin, H. L., Gaptank-Wolfcamp problem of the Glass Mountains, Texas: Jour. Paleontology, vol. 1, pp. 175-178, 1927.

Sellards, E. H., Pre-Paleozoic and Paleozoic systems, in Geology of Texas, pt. 1, Stratigraphy: Texas Univ. Bull. 3232, p. 148, 1933. King, P. B., Permian stratigraphy of trans-Pecos Texas: Geol. Soc. America Bull., vol. 45, pp. 727-729, 1934.

Baker, C. L., and Bowman, W. F., Geologic exploration of the southeastern front range of trans-Pecos Texas: Texas Univ. Bull. 1753, p. 104, 1917.

Bose, Emil, op. cit., p. 17.

Udden, J. A., op. cit., p. 17.

Baker, C. L., Date of the major diastrophism and other problems of the Marathon Basin, trans-Pecos Texas: Am. Assoc. Petroleum Geologists Bull., vol. 12, p. 1114, 1928.

Schuchert, Charles, Pennsylvanian-Permian systems of western Texas: Am. Jour. Sci., 5th ser., vol. 14, p. 386, 1927.

King, P. B., and King, R. E., The Pennsylvanian and Permian stratigraphy of the Glass Mountains: Texas Univ. Bull. 2801, pp. 117-119, 1928. King, P. B., The geology of the Glass Mountains, pt. 1: Texas Univ. Bull. 3038, pp. 45-48, 1930. Miller, A. K., A new ammonoid fauna of late Paleozoic age from western Texas: Jour. Paleontology, vol. 4, pp. 383-386, 1930.

 

74

and here considered to be a part of the Gaptank formation crop out over a wide area in the northwestern part of the Monument Spring quadrangle. These exposures are bounded on the south by early Paleozoic strata along "the trace of the Dugout Creek overthrust and on the north by Permian rocks in the foothills of the Glass Mountains. The best-exposed section is in the vicinity of the type locality.

The Gaptank is the youngest Pennsylvanian formation in the area and the last to be involved in the Marathon disturbance. It is the only member of the series that contains fossils in any abundance. It is somewhat more variable in lithology than the formations below and consists of sandstones and shales, with interbedded conglomerates and limestones. The conglomerate fragments are derived from the Maravillas chert, the Caballos novaculite, and the Dimple limestone, which are thousands of feet lower in the section than the base of the Gaptank. They indicate the rise of local folds in the Marathon geosyncline in the middle part of Gaptank time.

LOCAL FEATURES

Exposures near Gap Tank

-The type locality of the Gaptank formation, in the region near and to the south of Gap Tank (pl. 23), has been described in detail in a previous publication. At that locality there is exposed a section about 1,800 feet thick, considerably folded in the south but passing conformably, with a dip of about 30°, beneath the Wolfcamp beds on the north. Along the crest of the broad anticline south of Gap Tank the basal limestones of the formation, which contain large masses of Chaetetes milleporaceus, rest on thin-bedded sandstones and shales of the Raymond formation. The basal limestone is succeeded by 900 feet of sandstone and shale, with five interbedded layers of coarse conglomerate, of which the lower ones reach 50 feet in thickness and contain rounded limestone cobbles as much as 2 feet in diameter (pl. 11, B). Within the small area of exposure near Gap Tank all the conglomerate beds thin out to the north. This suggests a nearby southern source for the material. The upper 800 feet of the formation has no conglomerate and consists of thick layers of limestone, interbedded with sandstone and shale. Fossils have been collected at various horizons in the section at Gap Tank but particularly in the lower part. These make possible a correlation of the formation with the Pennsylvanian sections of central Texas and the northern midcontinent area.

The following section of the Gaptank and Wolfcamp formations (sec. 9, pl. 8) is a composite of various short sections measured near Gap Tank and is presented for reference, so that the exposures west of Marathon in the Monument Spring quadrangle may be compared with it.

"

---

King, P. B., op. cit. (Texas Univ. Bull. 3038), pp. 44-45.

 

75

Exposures east of Gap Tank

.-The Gaptank formation is also exposed at a locality 7 miles east of Gap Tank (pl. 23), where it lies between steeply inclined flaggy sandstones of the Haymond formation and gently tilted thin-bedded limestones of the Hess member of the Leonard formation. The Gaptank exposures at this place were mapped with the Hess in the previous work on the Glass Mountains, but since that time abundant Triticites irregularis (identified by C. O. Dunbar) and various brachiopods of Pennsylvanian aspect have been collected from the limestones. The formation here consists of thickbedded gray limestones and sandy marls, with subordinate conglomerate beds, made up of rounded pebbles and cobbles of Dimple limestone. This exposure, with its predominance of calcareous beds, illustrates the variability of the strata of the Gaptank formation and confirms the suggestion that the thick conglomerate beds of the type locality are derived from a nearby local source.

Area west of Marathon

-The area of Gaptank formation west of Marathon has been overridden by a great mass of pre-Carboniferous rocks along the Dugout Creek overthrust, and the formation is now exposed by the erosion of the overthrust sheet (pl. 16). Here and there small hills of novaculite overlie the formation, and in places blocks several feet to 10 feet across of novaculite, bedded chert, and chert breccia are scattered over the surface or partly embedded in the incompetent sandy or shaly beds of the formation. The structural complexity produced by the overthrusting, coupled with the scattered nature of the exposures, makes it impossible to obtain a complete section of the formation, but about 1,500 feet of beds are probably present.

In some of the closely folded anticlines and anticlinoria are exposures of the upper part of the Haymond formation, with its characteristic thin alternating beds of carbonaceous shale and greenish sandy shale or sandstone. The Haymond age of strata of this sort is indicated by the fossil plants collected on Dugout Creek 2½ miles south of Dugout Mountain (pl. 16). Similar beds, much crumpled and broken by small thrusts (pl. 9, D), were seen farther north along Dugout Creek and in the anticlinorial area on the south side of Dugout Mountain. These beds are not mapped separately from the overlying Gaptank formation in detail, on account of the complexity of the structure, The Gaptank strata lack the peculiar thin-bedded alternation of the lower formation, their sandstones are coarser-grained and of buff or rusty-brown color, and they include numerous limestone and conglomerate beds.

The lowest Gap tank strata are exposed in the region south of Dugout Mountain and consist of soft drab shales with scattered ledges and ridges of fine- to medium-grained rusty-brown sandstone, in part calcareous, and several layers of gray limestone from a few feet to 15 feet in thickness. The limestone layers "

---

King, P. B., Geology of the Glass Mountains, pt. 1: Texas Univ. Bull. 3038, geologic map, 1930.

 

76

are remarkably folded and broken and do not form continuous ledges for any great distance. Each one is more coarsely granular toward the base and in places even conglomeratic. Toward the top each is finegrained or compact. The lowest limestone bed contains a few fossils, which include Chaetetes milleporaceus and Fusulina meeki. These suggest a correlation with the Chaetetes-bearing limestone at the base of the formation in the type section. Near the Wilcox & Anderson well, at the foot of Dugout Mountain, there are in the lower part of the formation some layers of conglomerate, which contains well-rounded chert pebbles.

Farther south, at the northeast end of the novaculite ridges of the Payne Hills and 4 miles south-southeast of Lenox, are other exposures, evidently of younger beds (pl. 16). These are overlain by overthrust masses of older strata, which form the capping of the Payne Hills. At the northeast corner of the hills dark-blue shale has been cut into by several ravines and contains thin ledges of granular fossiliferous gray-brown limestone and limy sandstone. In a small gully southeast of the Payne Hills, 4½ miles south-southeast of Lenox and 1¾ miles south of the Arnold ranch, is a great block of limestone, lying in shale and seemingly a part of an indistinct steeply tilted limestone ledge, striking east (pl. 16). The dense gray limestone of the block is packed with ammonoid shells, from which a large fauna has been collected.

Farther east, on the Decie and Hargus ranches, are other exposures of the high part of the Gaptank formation, here much obscured by wash and terrace gravel (pl. 24). Near milepost 580, 4 miles west of Marathon on the Southern Pacific Railroad and along the highway to the north of it, 400 or 500 feet of the formation is exposed. There are 8 or 10 ledges, each as much as 3 feet thick, of hard, finely granular limestone, weathering rusty brown and containing angular fragments of green chert, rounded limestone pebbles, and a few small chips of green shale. All the calcareous layers are more or less fossiliferous. They are separated by beds of olive-green clay shale and sandy shale, with some interbedded slabby rusty-brown sandstones. Similar beds are exposed near the old Wedin oil boring, north of the highway on the Decie ranch. South of the old well are some ledges of coarse conglomerate, with subrounded limestone and chert fragments.

Black Peak

-At Black Peak, north of Doubtful Canyon, in the Del Norte Mountains, are rocks that are tentatively mapped as part of the Gaptank formation (pl. 24). The exposures are isolated from the other areas of Paleozoic rock to the east and northeast, and the rocks of the peak are thrust to the west against the Cretaceous limestones of the Del Norte Mountains (sec. M-M', pl. 21). A much sheared gray crystalline limestone, several hundred feet thick, forms the main mass of the peak and dips steeply to the east (pl. 13, C). It is overlain by thin-bedded limestone, shale, and some sandstone. The highest beds exposed are darkgray limestones, from which Pennsylvanian fossils have been collected by R. E. King.

FOSSILS AND AGE

FAUNAS OF THE GAP TANK AND WOLF CAMP AREA

The Gaptank formation at the type locality contains fossils in various beds of the stratigraphic section. They show a progressive change in character from the base to the top. The faunas are described below in ascending order.

Bed 1.-The basal layer of the formation at Gap Tank is known as the Chaetetes-bearing limestone. From the collections of R. E. King and the writer at several localities 2 miles south of Gap Tank (U. S. Geological Survey locality 6703), G. H. Girty has identified the following fossils:

• Chaetetes milleporaceus Milne-Edwards and Haime.
• Campophyllum torquium Meek?
• Clisiophyllum n. sp.
• Fistulipora sp.
• Productus n. sp.
• Productus boonensis Swallow var.
• Pustula semipunctata (Shepard)?
• Squamularia perplexa (McChesney) var.
• Streblopteria sp.

From the same collections C. O. Dunbar has identified Fusulina haworthi (Beede) and Wedekindellina euthysepta (Henbest).

According to Mr. Girty:

The most distinctive form in bed 1 is Chaetetes milleporaceus. * * * As this genus of corals is especially abundant in the lower part of the Pennsylvanian sections of Kansas and Missouri, if indeed it is not restricted in its upward range to the Des Moines group, and as the fauna, so far as it is known, contains nothing that would indicate a Pottsville age, the tentative conclusion seems to be warranted that the first Gaptank deposits were laid down in early post-Pottsville time.

According to Dunbar and Condra, Fusulina haworthi ranges from the Fort Scott limestone to the Altamont limestone of the northern midcontinent area, and Wedekindellina euthysepta is widely distributed in the Cherokee shale of the northern midcontinent area and its equivalents in Illinois and Ohio. The latter species is also found in association with Fusulina meeki in the lower part of the Strawn group in the section on the Brazos River in central Texas.

Bed 3.-Thin limestone layers at the base of the first conglomerate member (bed 3) on the north flank of the "

---

Cf. Moore, R. C., Correlation of Pennsylvanian formations of Texas and Okla-homa: Am. Assoc. Petroleum Geologists Bull., vol. 13, p. 896, 1929.

Dunbar, C. 0., and Condra, G. E., The Fusulinidae of the Pennsylvanian system of Nebraska: Nebraska Geol. Survey, 2d ser., Bull. 2, p. 84, 1927.

White, M. P., Some Texas Fusulinidae: Texas Univ. Bull. 3211, p. 21, 1932.

 

77

Gap Tank anticline, 1 mile southwest of the tank, have yielded a small fauna, of which the following have been identified by Charles Schuchert:

• Chaetetes milleporaceus Milne-Edwards and Haime.
• Campophyllum torquium Meek.
• Chonetes mesolobus Norwood and Pratten.
• Productus gratiosus var. occidentalis Schellwien.
• Productus cora D'Orbigny.
• Spirifer cameratus Morton.
• Composita subtilita (Hall).

The most distinctive fossil in the fauna is Chonetes mesolobus, which is found in the Mineral Wells formation of the Strawn group in central Texas and the upper part of the Des Moines group in the northern midcontinent area. It is a characteristic zone fossil of the lower Pennsylvanian.

Most of the cobbles in the conglomerate members of the Gaptank formation are derived from the erosion of the Dimple limestone and older formations. In the lower members, however, there are some fragments of coarsely crystalline light-gray crinoidal limestone. From such cobbles in the first conglomerate member the following fossils, which have been identified by C. O. Dunbar, have been collected:

• Large cup corals.
• Crinoid stems.
• Platyceras sp.
• Derbya crassa (Meek and Hayden).
• Productus (fragment).
• Pugnax cf. P. osagensis (Swallow).
• Squamularia perplexa (McChesney) (small type).

Schuchert collected a fragment of Chaetetes from this conglomerate. The fauna of the fossiliferous cobbles is decidedly of lower Pennsylvanian aspect. This fact and the lithology of the rock make it probable that the cobbles were derived from the erosion of the Chaetetes- bearing limestone, at the base of the formation.

Bed 6.-In the shale between the second and third conglomerate members (bed 6), there is locally a nodular ferruginous limestone that contains various pyritized fossils. From collections at several localities 2 miles southeast of Gap Tank (U. S. Geological Survey localities 6691 and 7095) G. H. Girty has identified the following fossils:

• Enchostoma .
• Lophophyllum profundum(Milne-Edwards and Haime).
• Orbiculoidea sp.
• Productus aff. P. insinuatus Girty.
• Productus cora D'Orbigny var.
• Pustula sp.
• Marginifera muricata (Norwood and Pratten).
• Pugnax rockymontana (Marcou).
• Composita subtilita (Hall).
• Allerisma sp.
• Edmondia sp.
• Nuculopsis ventricosa (Hall).
• Leda bellistriata Stevens.
• Bellerophon crassus var. wewokanus (Girty).
• Euphemites carbonarius (Col).
• Bucanopsis sp.
• Worthenia tabulata (Conrad).
• Trepospira depressa (Cox).
• Phanerotrema grayvillense (Norwood and Pratten) var.
• Phanerotrema sp.
• Meekospira? sp.
• Sphaerodoma sp.
• Orthoceras sp.
• Nautilus? sp.
• Various species of goniatites.

Mr. Girty remarks that this fauna "rather pointedly resembles the fauna of the Wewoka formation of Oklahoma." The most characteristic shell of the collection, Pugnax rockymontana, is abundant in the upper part of the Des Moines group of the northern midcontinent area and is found very rarely in the Kansas City group. In the central Texas section it is found only in the Graford formation (lower Canyon).

Bed 10.-From calcareous sandstones between the fourth and fifth conglomerate members (bed 10) a large fauna has been collected. From this horizon, at several localities 2 miles southeast of Gap Tank (U. S. Geological Survey localities 6705, 7085, and 7088), Mr. Girty has identified the following species:

• Textularia sp.
• Wewokella aff. W. solida Girty.
• Mesandrostria n. sp.
• Coelocladia n. sp.
• Amblysiphonella prosseri Clarke?
• Lophophyllum profundum (Milne-Edwards and Haime).
• Axophyllum sp.
• Delocrinus sp.
• Fistulipora sp.
• Stenopora sp.
• Fenestella sp.
• Polypora sp.
• Rhombopora sp.
• Rhipidomella carbonaria (Swallow).
• Meekella striaticostata (Cox).
• Derbya aff. D. bennetti Hall and Clarke.
• Chonetes verneuilianus Norwood and Pratten.
• Productus cora D'Orbigny.
• Productus calhounianus Swallow.
• Productus pertenuis Meek.
• Productus portlockianus Norwood and Pratten.
• Marginifera wabashensis Norwood and Pratten.
• Teguliferina armata (Girty).
• Rhynchopora illinoisensis (Worthen).
• Dielasma bovidens (Morton).
• Spirifer triplicatus (Hall).
• Spiriferina kentuckyensis (Shumard).
• Squamularia perplexa (McChesney).
• Composita subtilita (Hall).
• Cleiothyridina orbicularis (McChesney).
• Edmondia? sp.
• Parallelodon carbonarius (Cox).
• Parallelodon aff. P. sangamonense (Worthen).
• Schizodus? sp.

"

---

Schuchert, Charles, The Pennsylvanian-Permian systems of western Texas: Am. Jour. Sci., 5th ser., vol. 14, p. 386, 1927.

Moore, R. C., op. cit., p. 896.

 

78

• Aviculopecten interlineatus (Meek and Worthen).
• Streblopteria sp.
• Lima retifera Shumard.
• Aviculipinna peracuta (Shumard).
• Astartella varica McChesney.
• Bellerophon stevensianus McChesney.
• Pleurotomaria n. sp.
• Pleurotomaria aff. P. subcalaris (Meek and Worthen).
• Yvonia aff. Y. subconstricta (Meek and Worthen).
• Worthenia perizomata White?
• Porcellia peorensis Worthen.
• Murchisonia aff. M. copei White?
• Anomphalus? sp.
• Trachydomia wheeleri (Swallow) ?
• Trepospira depressa (Cox).
• Platyceras parvum (Swallow).
• Diaphorostoma peoriense (McChesney)?
• Naticopsis aff. N. nana (Meek and Worthen).
• Zygopleura aff. Z. plebeia (Herrick).
• Bulimorpha sp.
• Orthoceras sp.
• Metacoceras aff. M. perelegans Girty.
• Gastrioceras sp.
• Griffithides? sp.
• Ostracoda undet.

From this same horizon Parenteletes cooperi King has been identified by R. E. King, Schistoceras smithi Böse by Emil Böse, and a new species of Shumardites by F. B. Plummer. The beds also contain great numbers of small Triticites, which at one time were considered by Dunbar to be T. irregularis (Schellwien and Staff), but which he "now regards as a new species of about the same age."

Mr. Girty says:

This extensive fauna contains many novel features and to that extent is not comparable to any of the Pennsylvanian faunas already known. It is related to certain faunas of the Cisco, but especially those that occur in the lower part of the formation. Insofar as there exist grounds for comparison, there seems to be no reason why this fauna might not represent a horizon well up in the Pennsylvanian of Kansas, without, on the other hand, the existence of much reason why it should represent a horizon so late.

The fusulinid in this fauna is found in the lower Canyon and upper Strawn groups of central Texas and the Kansas City group of Kansas. According to F. B. Plummer, the ammonoid Schistoceras smithi is found in the Palo Pinto limestone at the base of the Canyon group in central Texas.

Bed 13.-From the first limestone member (bed 13) specimens of a small, poorly preserved, silicified fusulinid have been collected. C. O. Dunbar states that it resembles Triticites cullomensis var. pygmaeus Dunbar and Condra, which is found in the Shawnee formation of the northern midcontinent area.

Bed 19.--Nodular limestones in the lower part of the fourth limestone member (bed 19) a quarter of a mile west of Gap Tank contain a small fauna, from which the following fossils have been identified by R. E. King:

• Heterocoelia beedei Girty.
• Spirifer triplicatus (Hall).
• Composita subtilita (Hall).
• Rhipidomella carbonaria (Swallow).

Beds beneath the Uddenites zone.-Farther west, a mile northeast of Wolf Camp, the strata immediately beneath the Uddenites zone of the Gaptank formation contain a few fossils. These beds may be in about the same part of the formation as bed 19. The following fossils have been identified from the locality near Wolf Camp by R. E. King:

• Cladopora sp.
• Echinocrinus spines and plates.
• Platyceras sp.
• Derbya bennetti Hall and Clarke

.

From limestones directly beneath the Uddenites zone at this locality C. O. Dunbar has identified Triticites cullomensis Dunbar and Condra, a fusulinid that characterizes the Jacksboro limestone member of the Graham formation (upper Canyon) of central Texas and the Shawnee and lower part of the Wabaunsee groups of the northern midcontinent area.

Uddenites zone.-The Uddenites zone at Wolf Camp and between this place and Gap Tank (pl. 23) contains a large and abundant fauna. Immediately north of Wolf Camp, in the saddle between an outlying butte and the main escarpment capped by the overlying Wolfcamp formation (R. E. King locality 88), the fossils listed below have been identified. The fossils occur in shale beds, or loosely embedded in thin brown limestone layers.

• Triticites ventricosus (Meek and Hayden).
• Rhipidomella carbonaria (Swallow).
• Orthotichia kozlowskii King.
• Orthotetella wolfcampensis King.
• Parenteletes cooperi King.
• Streptorhynchus pyramidale King.
• Meekella irregularis var. texana King.
• Chonetes granulifer Owen (small type).
• Chonetes verneuilianus Norwood and Pratten.
• Productus semireticulatus var. hermosanus Girty.
• Productus semistriatus Meek.
• Linoproductus cora(D'Orbigny).
• Linoproductus villiersi (D'Orbigny).
• Pustula semipunctata(Shepard).
• Waagenoconcha montpelierensis (Girty).
• Overtonia cristato-tuberculata (Kozlowski).
• Marginifera capaci (D'Orbigny).
• Marginifera lasallensis (Worthen).

Dunbar, C. 0., letter, August 1934.

White, M. P., op. cit., p. 19.

Dunbar, C. 0., and Condra, G. E., op. cit., p. 95.

Fusulinid identified by C. O. Dunbar (King, P. B., Geology of the Glass Mountains, pt. 1: Texas Univ. Bull. 3038, p. 55,1931); brachiopods by R. E. King (Geology of the Glass Mountains, pt. 2; T