193
GEOLOGY OF THE EDWARDS PLATEAU AND RIO GRANDE PLAIN
ADJACENT TO AUSTIN AND SAN ANTONIO, TEXAS,
WITH REFERENCE TO THE OCCURRENCE
OF UNDERGROUND WATERS.
By
ROBERT T. HILL AND T. WAYLAND VAUGHAN.
18 GEOL, PT 2-----13
194
195
CONTENTS
- Page.
- Introduction 199
- Geography of the region ...201
- Rio Grande Plain ...202
- Balcones scarp line...203
- Edwards Plateau ...204
- General principles of artesian waters ...212
- Capacity of rooks-for absorbing moisture...213
- Geology of the region ...215
-
- Table of geologic formations of the region under discussion ...216
- Paleozoic formations ...217
- The Cretaceous ...217
- Comanche series (Lower Cretaceous) ...218
- The basement beds ...218
- Travis Peak formation ...219
- Gillespie formation ...221
- Glen Rose formation ...221
- Walnut formation ...226
- Comanche Peak limestone ...226
- Edwards limestone ...227
- Fort Worth limestone ...235
- Del Rio clays ...236
- Shoal Creek limestone ...237
- Gulf series (Upper Cretaceous) ...238
- Eagle Ford shales ...239
- Austin chalk ...239
- Taylor formation ...240
- Anacacho formation ...240
- Webberville and Eagle Pass formations ...241
- Comanche series (Lower Cretaceous) ...218
- The Eocene ...243
- Alluvial deposits of the Neocene, Pleistocene, and Recent epochs ...243
- Uvalde formation ...244
- Pleistocene terraces ...247
- Terraces of the Colorado River ...247
- The Asylum terrace ...249
- The Capitol terrace ...249
- The present flood plain ...250
- Terraces on the south side of the Colorado River ...251
- Terraces of the Rio Grande ...251
- Onion Creek marl and allied deposits ...252
- Terraces of the Nueces and Leona rivers, and the Leona formation ...253
- The wash ...254
- Fry-pan deposits ...255
- Tepetate and "tierra blanca" ...256
-
- Geology of the region-Continued. Page.
- Igneous rocks ...256
- Kinds of igneous rocks ...256
- Mode of occurrence of the igneous rocks ...256
- Arrangement of the strata ...257
- The main system of dips ...257
- Balcones fault zone ...258
- Capacity of the various rock sheets for water ...260
- Underground water of the region ...264
- Water supply of Edwards Plateau ...264
- Nonflowing wells ...265
- Deep wells of the plateau summit ...265
- Deep wells of the plateau monocline ...266
- Surface wells of the canyon valleys ...266
- Wells near Bee Caves, Travis County ...266
- Wells of Southern Pacific Company in Pecos County ...266
- Gravity springs of the canyons of the plateau ...267
- Kickapoo water hole ...269
- Black water hole ...269
- Artesian wells of the canyons of the plateau ...270
- Morris ranch well ...273
- Nonflowing wells ...265
- Water of Rio Grande Plain ...274
- Nonflowing wells and gravity springs of Rio Grande Plain ...274
- Waters of the Leona and kindred formations ...275
- Water of the Rio Grande terraces below Del Rio ...276
- Wells in the fry-pan deposits ...276
- Water of the terrace formations of the Colorado River ...276
- Water of the Onion Creek marl and similar deposits ...277
- Surface wells and springs of the Cretaceous formations ...277
- Artesian wells ...279
- Austin wells ...279
- TheManor well ...284
- The San Marcos well ...287
- San Antonio wells ...290
- The Del Rio well ...299
- Chemical qualities of the waters ...299
- Fissure springs ...307
- Probable identity of source of artesian and fissure-spring waters ...312
- Source of the underground waters of Rio Grande Plain ...313
- Availability and limitations of the underground waters ...316
- Practical suggestions ...319
- Nonflowing wells and gravity springs of Rio Grande Plain ...274
- Water supply of Edwards Plateau ...264
196
197
ILLUSTRATIONS
- Page.
- PLATE XXI. Map showing the relations of the Edwards Plateau, the Balcones scarp line, and the Rio Grande Plain ...200
- XXII. Sulphur Peak, Uvalde County ...202
- XXIII. The Llano Estacado and Edwards Plateau and contiguous regions ...204
- XXIV. Round Mountain, Uvalde Count ...206
- XXV. Rio Grande Canyon and boca of Pecos Canyon ...208
- XXVI. Hillcoat Cave, Edwards County ...210
- XXVII. Pecos Canyon at the Southern Pacific Railroad bridge ...212
- XXVIII. Upper Glen Rose beds of Mount Bonnel ...222
- XXIX. Bull Creek, Travis County ...224
- X XX. Flints of Edwards limestone ...226
- XXXI. Bluff caverns in Edwards limestone, Rio Grande ...228
- XXXII. Bluff cavern in Edwards limestone, Rio Grande, interior view ...230
- XXXIII. Pyramidal buttes of Edwards limestone, Kinney County ...232
- XXXIV. Shoal Creek limestone, West Fork of the Nueces River, Kinney County ...238
- XXXV. Cross sections in the Austin quadrangle, Texas ...246
- XXXVI. Escarpment and terrace of the Rio Grande at Eagle Pass ...252
- XXXVII. Cap of Pinto Mountains ...258
- XXXVIII. Run-off of Frio (gravity) springs ...268
- XXXIX. Artesian well, City Waterworks Company, San Antonio ...278
- XL. Head of Kickapoo springs, Edwards County ...280
- XLL Sections of artesian wells, Austin and vicinity ...286
- XLII. Artesian well, City Waterworks Company, San Antonio ...290
- XLIII. Sections of artesian wells in vicinity of San Antonio ...294
- XLIV. Barton Creek, above dam, Travis County ...308
- XLV. San, Pedro springs, San Antonio ...310
- XLVI. Map of Austin and vicinity, showing the relations of springs, faults, and wells (this plate should have been numbered XLI) ...282
- XLVII. San Marcos spring ...312
- XLVIII. Natural section of old water channels in limestone bluff below Kickapoo springs ...314
- XLIX. Run-off of San Marcos spring ...316
- L. Karrenfelder, Edwards limestone ...318
- LI-LXIV. Characteristic fossils of the various formations ...322
- FIG. 53. Key map of topographic provinces of Texas ...201
- 54. Diagrammatic profile across the southern margin of the Edwards Plateau, in Edwards and Kinney counties ...207
- 55. Section at Fredericksburg ...221
- 56. Sections at Lohmann's Crossing and Hickory Creek, Colorado River... 223
- 57. Sections of Edwards limestone near Austin ...229
- Page.
- FIG. 58. Section of Shoal Creek limestone, Del Rio clays, and Fort Worth limestone, at Austin ...236
- 59. Section of Del Rio clays near Weymüller's ranch, Kinney County ...337
- 60. Section of Shoal Creek limestone, Kinney County ...238
- 61. Section of Anacacho formation, Anacacho Mountains, Kinney County ...240
- 62. Section (in well) of a part of the Eagle Pass beds, Eagle Pass ...243
- 63. Diagrammatic cross section of the Rio Grande terraces ...252
- 64. Cross section of Rio Grande terraces at Palafox ...252
- 65. Diagrammatic representation of increase in steepness of dip south of the Balcones fault line ...258
- 66. Monoclinal fold, Edwards and Uvalde counties ...260
- 67. Block faults in Edwards limestone, Bee Creek, Travis County ...261
- 68. Sections of artesian wells at Kerrville and Morris ranch ...270
- 69. Wells in gravel beds ...274
- 70. Alluvial deposits around Uvalde, showing wells ...275
- 71. Section across valleys of Nueces and Leona rivers and intervening highland, on line shown in fig. 70 ...275
- 72. Section showing fault at San Marcos springs ...308
- 73. Map showing the localities of artesian wells and springs in the vicinity of San Antonio ...310
- 74. Composite section of the Gillespie formation at Fredericksburg, showing the water-receiving beds ...314
- 75. Honeycombed strata, Edwards limestone ...315
- 76. Displacement at Mount Bonnel, Travis Count ...315
198
199
GEOLOGY OF THE EDWARDS PLATEAU AND 11,10 GRANDE PLAIN ADJACENT TO AUSTIN AND SAN ANTONIO, TEXAS. By ROBERT T. HILL and T. WAYLAND VAUGHAN.
INTRODUCTION
The artesian systems of the eastern half of Texas are numerous and underlie several large areas. The productive areas extend from near Red River,Denton County,to Del Rio,on the Rio Grande, and from near the center of the State to the coastal islands of the Gulf. Collectively they comprise a district 450 miles in length and averaging 300 miles in width. It is doubtful if there has been anywhere a more remarkable development of artesian wells in the last tell years than in this region. At numerous places copious flows of water have been obtained, and districts which a few years ago had only inadequate supplies now possess artesian wells furnishing water in large quantities to cities,ranches, and farms,improving the hygienic conditions, and yielding water for the stock raising,agricultural,manufacturing, and transportation industries
The artesian wells of the eastern half of Texas belong to several distinct systems, the terns "system" including all wells having their source in the same set of rock sheets or strata. It would require a large volume to describe each of these various artesian systems. In the Cretaceous formations alone there are no fewer than five, and two of these the Travis Peak, or Waco, and the Edwards receive consideration in this paper. It is especially proposed to explain as well as possible the principles of the supply of the artesian belt supplied by the Edwards system and the probability of success or failure of wells sunk in different parts of its area.
Within the last two decades numerous artesian wells have been drilled in the vicinity of San Antonio, Texas, and for many years the source of the water has been a matter of perplexity. Unfortunately it has not been within our power to secure accurate logs of these wells, accompanied by specimens, or even to ascertain with accuracy the number and location of the wells, or to collect other essential data of a statistical "
The terms "Travis Peak" and "Edwards" are formation names used in this paper in place of "Trinity Sands" and "Caprina limestone," hitherto employed. The reasons for the employment of these new names are given on later pages (pp. 216, 227).
200
nature, such as the Government has deemed expedient to collect in other artesian regions. Such facts as are herein presented have been collected incidentally by the writers while making private reconnaissances or official geologic surveys in the region, and through correspondence with citizens. Accordingly, this report is not to be considered as a statistical or engineering paper, but rather one which deals with the geologic side of the underground water question.
The source of the San Antonio water supply has been ascertained by detailed studies of the structure and outcrop of the water bearing beds in the adjacent regions. A detailed map of the Austin quadrangle was made, and the thickness, sequence, paleontology, mineral composition, and water capacity of every portion of the Cretaceous section along the Colorado River were ascertained. A reconnaissance was next made southwestward from Austin via Fredericksburg, Kerrville, and the headwaters of the Frio and Nueces to Fort Clark, Texas, for the purpose of studying minutely the variation of the rock sheets outcropping in that direction, which, owing to the direction of their dip, must necessarily underlie the city of San Antonio. At the southeastern end of this line detailed studies and mapping were resumed upon the Nueces, Brackett, and Uvalde quadrangles. By this method a base line of geologic sections, so to speak, was established along the strike of the outcrop of the Lower Cretaceous formations, the sections at the extremities of which were determined with the greatest accuracy possible, while check sections have been made at numerous intervening points along the line.
The result of this work was the discovery that, while these well waters come from the same series of beds that supply the artesian wells of the Waco, Fort Worth, and Dallas regions north of the Colorado, their occurrence presents some important differences of detail. Instead of having their immediate source in beds of porous sands, like the wells about Waco, they are derived largely from the Edwards limestone, hitherto supposed to be one of the most impervious formations of the whole Cretaceous section.
It became apparent that this hitherto unappreciated water-bearing formation had great possibilities for supplying with either flowing or nonflowing wells a large area of country lying between Austin and San Antonio, extending west of the San Antonio River along the northern margin of the Rio Grande Plain toward the Pecos Piver, and even comprising the extensive summit region of the Edwards Plateau.
There are also many remarkable springs in that portion of the Rio Grande Plains and Edwards Plateau lying between the Colorado and Pecos rivers, which will be discussed in this paper. These belong to two distinct classes, each characterizing one of the geographic subdivisions mentioned, aid each exhibiting a method of escape of the underground water of the region. One line of these springs follows approximately the margin of the Rio Grande Plain, close to the line of
201
the railroad from Austin to Del Rio. The other class embraces the springs of the canyons of the Edwards Plateau.
In order to understand the occurrences of water of the San Antonio system, it is necessary to carry in mind
- (1) The geography of the region.
- (2) The simple laws of the occurrence and distribution of underground water.
- (3) The composition, variation, and arrangement of the rocks underlying the region and affecting the distribution of water.
The general map (Pl. XXI) will illustrate the geographic relations of
GEOGRAPHY OF THE REGION.
The traveler by rail from Austin via San Antonio to Del Rio, on the Rio Grande, sees from the car window two conspicuous regions, each having its peculiar geographic features. On the left is a gently undulating plain whose margin the railroad follows. This is here termed
202
the Rio Grande Plain. On the right is constantly visible a line of low, circular, flat topped hills, the Balcones scarp line, which represents the jagged southeastward front of a higher region which has been called the Edwards Plateau. The Rio Grande Plain, the Edwards Plateau, and the Balcones scarp line are the chief geographic features of the region. Broadly considered, they are a lowland plain inclining gently southeastward to the Gulf of Mexico, an upland plain rising gradually toward the northwest, and a rugged zone of separation which includes a quick ascent from plain to plain.
The Atlantic and Gulf coasts of the United States from the Hudson to the Rio Grande are margined by a broad lowland called the Coastal Plain. The portion of it lying farthest to the southwest is called the Rio Grande Plain. One of the more important geographic divisions of the interior of the continent is the Great Plains. Its most southerly division is the Edwards Plateau. Farther north the Coastal Plain and the Great Plains lie far apart, the Mississippi Valley and the Appalachian belt and other geographic provinces being included between them; but southward they converge, finally meeting in southern Texas, so that the Rio Grande Plain and the Edwards Plateau lie side by side.
RIO GRANDE PLAIN.
In shape this plain is an irregular quadrilateral with angles turned toward the four cardinal points. On the northwest, as just stated, the Balcones scarp line separates it from the Edwards Plateau. On the southeast it is bounded by the Gulf of Mexico, and on the northeast it is arbitrarily limited by the Colorado River. On the southwest it is limited by the folded mountains of Mexico, which lie beyond the valley of the Rio Grande. It also sends a tongue for many miles up the Rio Grande.
The plains drainage, which follows the general slope from northwest to southeast, includes the Nueces, San Antonio, and Guadalupe rivers and their branches, besides various minor streams which join the Rio Grande and Colorado or enter the Gulf direct.
As compared with the adjacent plateau and mountain regions, its characteristic topographic feature is a low relief, but its surface is broken by occasional undulations and in places by hills of considerable height. Some of these are the low scarp lines of drainage valleys; others, like the Dos Hermanos Hills of Webb County, are buttes capped by limestone; others, like Pilot Knob, in Travis County, are old volcanic necks; Pinto, Las Moras, Turkey, and Elm mountains, in Kinney County, are buttes composed below of sedimentary rocks and above of caps of igneous rock; Sulphur Peak and Fort Inge, in Uvalde "
The indentation of this plain up the Rio Grande has been called the "Rio Grande Embayment" by Hill: Bull. Geol. Soc. America, Vol. III, 1891, p. 93.
Hill, ibid., p. 90.
203
County, are masses of solid basalt (see P1. XXII); the Anacacho Hills, extending east and west in southern Kinney County and constituting the most rugose part of the plain, are of still another type, consisting of a monoclinal plateau, or cuesta, sloping southward and presenting a steep scarp to the north. (See geologic profile, fig. 66, p. 260.)
The eastern part of the plain belongs climatically to the humid and subhumid regions; the western part, to the arid. The fertile "black lands" occupy large areas as far southwest as San Antonio; but contlnuous cultivation is limited by increasing aridity west of Bexar County. The plain as a whole is mostly grazing land, and supports one of the greatest stock raising industries of the United States. San Antonio is its commercial center.
In passing westward from comparative humidity to aridity, there is a corresponding decrease in the depth and richness of the residual soils. Soils derived from certain geologic formations which in the Black Prairie and eastern Texas timber regions are fertile, here become progressively more sterile and barren, and are impregnated or coated with a peculiar calcareous encrustation known in the Southwest as "tierra blanca" or "tepetate."
Much of the region is covered by the peculiar flora known as chaparral, consisting of a thorny growth of many species of scrubby acacias-mesquite, guaxillo, and huisatche-between which is an undergrowth of cactus, especially the large Opuntia known as nopal. For this reason the region is sometimes called the "chaparral country." There are also stretches of open prairies covered by nutritious grasses.
It is not within the province of the present paper to deal with the plain in its entirety, but only with a strip along its western margin not exceeding 30 miles wide, a belt underlain by the Cretaceous formations.
BALCONES SCARP LINE.
The Balcones scarp line is the frayed and ragged coastward border of the Edwards Plateau. From the more open and level lower country it appears as a sharp line of timber covered hills, and these are universally called "mountains" by the people of the region. It commences near the northern line of Travis County and continues a little south of west, through Travis, Rays, Comal, Bexar, Medina, Uvalde, and Kinney, to Valverde County, where it meets the Rio Grande. Near Austin its highest summits are about 400 feet above the margin of the lower plain; in Uvalde County, nearly 1,000 feet.
"The "Black Prairie" region is that underlain by the soft Upper Cretaceous chalky limestones and marly clays, where there is considerable rainfall. It comprises the richest agricultural lands of the State. The cities of Austin and Dallas are situated near its western margin; Terrell and Corsicana are near its eastern margin.
A singular fact in the progressive cartography of our country is that notwithstanding the conspicuousness of the Balcones scarp line as a topographic feature and that it was shown upon all the mother maps of the region, such as those made by Bartless, Roemer, and J. De Cordovas (J. H. Colton & Co., New York), it has been omitted from the more recent maps.
204
The drainage flows directly across the scarp line and has cut great canyons backward into the Edwards Plateau. The depth and precipitous character of these increase in the streams successively encountered as one goes westward. The portions of these streams lying within the area of the plateau before they cross the fault line have cut their channels approximately down to the level of the Rio Grande Plain. Their bocas in many places cut the scarp line into a series of tongue like salients projecting toward the plain, and the line is further diversified by plateau remnants in the form of outlying buttes. The position of the scarp is determined by a complex dislocation of the rocks, the Balcones fault, which will be described in a subsequent section (p. 258).
EDWARDS PLATEAU.
The Llano Estacado and the Edwards Plateau together constitute in Texas the Plateau of the Plains. This lies within the area inclosed by the Canadian on the north, the Pecos River on the west, the Balcones escarpment on the south and southeast, and an irregular line of scarps along the headwaters of the eastward flowing drainage of the Colorado, Brazos, and Red rivers of Texas. The general outline of this area is shown in the accompanying photograph of a model, P1. XXIII. It is over 500 miles in length and in places 280 miles broad. It is a vast quadrangular mesa, surrounded on all sides by descending escarpments. In its horizontal geologic structure and its relations to the surrounding region it may be broadly compared to a book laid upon a table and very slightly tilted or raised at one end. Its comonent strata are almost as parallel and regular as the leaves of the volume. While its central portion still presents a general level, its borders are cut by headwater erosion into a fringe of projecting drainage divides, accompanied by many remnantal buttes and mesas, showing the great erosion by which the plateau has been and is being gradually etched away. Its eastern margin in particular has been greatly reduced by this process.
The Llano Estacado and Edwards Plateau merge into each other along the central portion of the summit of the greater plateau, and there is no known line of demarcation between them. There is a great difference between the characters of their surface, their soils, and the underlying geologic strata, which collectively gives to each of the two regions a distinct character. The soil and underlying rocks of the Llano Estacado consist of unindurated loams, marls, and sands. Hardly a stone of building size can be found in its broad extent. The Edwards Plateau is in part a rough limestone country, resembling in some respects the western margin of the limestone country of the Grand "
Boca, a Spanish topographic term, used to indicate the mouth of a canyon valley where it debouches on a plain. Example, Boca del Agna, Jamaica, corrupted by the English into "Bog walk."
205
Prairie region (see Pl. XXIII), with which its, rocks are geologically identical. The hard Cretaceous rocks of the plateau partially underlie the Tertiary formations of the Llano Estacado, and it is probable that the latter have been eroded from the plateau.
It is in the latitude of Upton and Midland counties, or a little north of the thirty first parallel, that the hard rocks of the Edwards Plateau become covered by the marls and sands of the Llano Estacado, and it is here that the line between them is provisionally drawn while awaiting further exploration. In these counties the narrowest width of the whole plateau is found.
The Edwards Plateau occupies nearly the whole of the counties of Crockett,Valverde,Edwards,Sutton,Schleicher,Kimble,Kerr,Bandera,Gillespie,Kendall, and Blanco, and about one-half of the counties of Crane,Upton,Tom Green,Irion,Concho,Menard,Travis,Hays, Comal,Bexar,Medina,Uvalde, and Kinney. The northeastern and western boundaries are cliffs due to erosion. The northeastern overlooks a broad denuded area traversed by the Colorado River and its tributaries. The western margins the Pecos Valley, which at the north is an open plain and at the south a canyon. Beyond the Pecos are folded mountains, but in the region of the canyon these are separated from the river by a broad plateau similar in type to the Edwards.
The southeastern boundary is also a cliff, the Balcones scarp, but, as already stated, was primarily determined by dislocation rather than erosion. The main water parting lies near the western edge, and the greater part of the plateau is drained by streams running eastward and southeastward. These have cut deep canyons, dividing its marginal portions into long, narrow tongues. The drainage westward to the Pecos has accomplished only a moderate amount of erosion, so that the western cliff is comparatively simple in contour.
The part of the plateau which has been most thoroughly studied and with which this report is more especially concerned is the eastern and southern, the portion adjoining the Colorado Valley and the Rio Grande Plain. The characteristics of the plateau are most strongly impressed on the observer who enters it from the Rio Grande Plain, for in crossing the Balcones line he experiences a sudden and complete change of scenery, with accompanying changes in floral, geologic, and cultural conditions. Instead of long, wide sweeps of prairie, void of sharp relief, he finds a region of steep canyons and sloping hillsides. The monotony of deep and dusty soils is replaced by alternate outcrops of cream colored rocks and mils, occurring in long, continuous, and horizontal lines of stratification. A deciduous flora suddenly appears in the canyon valleys, replacing the semiarid chaparral. Rivers of flowing water, fringed by forests, replace the dry and stony stream ways of the plain, the mere trace of which is often lost in times of drought, and they now become fixed features of the landscape, boxed in with steep walled canyons. Rugged evergreen hills succeed the long stretches of low
206
undulating land with the yellow brown adobe soils. Some of the beds of stratification composing the canyon walls are barren of foliage; others are occupied by the dark evergreen shrubs, juniper and Sophora, extending like garlands around the brows of the circular hills. With the increasing altitude the air becomes more exhilarating and the heat of the day is tempered by cooler nights.
The Edwards Plateau, like the whole of the great plateau of which it is a part, presents three simple topographic elements-
- (1) the flattopped summits of the decaying plateau,
- (2) the breaks or slopes of its crenulated borders and canyoned valleys,
- (3) the stream ways
The summit of the plateau is reached by ascending the long canyons of the streams and passing out upon it through their "draws" or caletas. Like that of the Llano Estacado, it is flat and void of constant-running streams. There are a few shallow, pondlike depressions or "sinks," which occasionally contain water immediately after rainfall. In general it is covered with a thick growth of nutritious grass and is without forest. Here and there, however, may be seen thick patches of scrub live oak, known as "shin oaks," growing in dense patches called "shinneries." For miles and miles' the level, grass covered plain stretches before the eye like a great sea, the view broken only at long intervals by the tall shaft of a ranchman's windmill, rising like the sail of a lonely vessel on the level sea. This summit region has an altitude of about 2,400 feet in the northern edge of Edwards County, gradually rising, with an ascent of about 4.5feet per mile, to the west and northwest and falling at the same rate toward the east.
In the dissected border of the plateau are hundreds of remnantal buttes or hills, like Mount Bonnel and Lone Tree Mountain, hear Austin, and bound Mountain, in northwestern Uvalde County (P1. XXIV). Standing upon one of these hills, one can see that its horizontal layers of rock once extended across the valleys to the opposing bluffs and summits and realize that the valleys have been cut out of these once continuous rock sheets. The major stream ways have eroded their wide valleys below the summit level, so that the surface of the old plateau is preserved only in the divides.
Ultimately the headwaters of the streams will all meet, and the general summit level will be cut into innumerable buttes and mesas, like "
This expressive term, denoting the cliff margin of an upland plain where it breaks away to a lower plain or valley, has not yet found its way into the dictionaries, but is in common use in many districts of the far West.
From the Spanish. This is a useful word for the ultimate and smallest headwater ramification of a lateral stream. It is synonymous with the term "draw," used in the middle plains region of the United States, the "coulee " of Montana, and "drain" as used in Colorado. In this paper the local topographic nomenclature of Spanish-America will frequently be used. This has been set forth in a paper entitled Descriptive topographic terns of Spanish-America: Nat. Geog. Mag., September, 1896, Vol. VII.
207
those now found along its border. These in turn will be planed down to lower and lower levels. By this process many layers of rock have been stripped from the plateau in the past, as others will be in the future.
The present plateau summit exists because of the superior hardness and the capacity for resisting erosion of the rock sheet that caps it. This cap rock is the Edwards limestone. It not only everywhere constitutes the summits, but gives character to all the surrounding scarps and hills. The beds which originally overlay it were comparatively soft and yielding, and have almost completely disappeared.
The streams of the plateau have important bearing upon the question of underground water, and as they constitute a group of rivers characteristic of a large region in Texas, they are well worth the serious attention of those interested in the economic and geographic problems discussed in this paper.
The Rio Frio will be described as a type of the rivers flowing through the southern margin of the plateau. From where it passes out of the plateau onto the Rio Grande Plain, just north of the Southern Pacific Railroad, for fully 50 miles toward its mouth its bed is a shallow, usually waterless stream way, meandering across the gravel covered country. Flowing water is usually found in this portion of its
North of the railroad the stream emerges from a boca (like that seen in Pl. XXV) of the Balcones escarpment, and still farther northward it traverses a canyon of the plateau. In this canyon the stream way is in the bottom of a flat valley filled with ancient gravel deposits and bordered on each side by high bluffs of stratified limestone. In its lower portion the valley averages 2 miles in width, and contains some fairly good agricultural lands. Entering the canyon, the road follows the dry, stony stream bed a short distance before flowing water is encountered. Farther up the stream running water again ceases and the dry bed again appears. In this manner the continuity of the river is interrupted and its bed is broken up into alternations of dry and "
The geologic formations here mentioned are described more fully on page 227.
208
watered segments. In the flowing segments are clear and swift bodies of water, supplying as much as 1,000 gallons per minute and competent to irrigate many acres of land.
Wide, flat bottomed canyons of this type continue from the Balcones scarp line via Van Pelts and Leakey to the "water hole" near the head of the West Fork, and the number of springs and quantity of living water progressively increase in that direction. Above the headwater "hole" a different type of canyon sets in and permanent water ceases. The fall or gradient of the meandering stream way in the canyon is about 23 feet per mile in the Nueces, and is probably the same in the Frio. The fall of the bottom of the canyon, without considering the stream meanderings, is about 60 feet per mile.
The scarps which make the outer borders of these flat bottomed valleys are nearly always steep, usually consisting of alternations of bluffs, benches, and slopes corresponding with the horizontal lines of stratification and giving to the landscape a terraced topography. Between the widely separated escarpment walls are the broad, level second bottoms, standing some 50 feet above the level of the present stream ways. These bottoms consist of an alluvial deposit (the Uvalde formation) and record a peculiar event in the history of the region an event to be discussed in subsequent pages. Nearly all the living or perennial water of the stream ways occurs in the flat bottomed canyons, and they are the seat of the chief agricultural population of the region. The beds of their stream ways are usually composed either of the smooth surface of some horizontal limestone stratum, over which the water may flow for a great distance, or of clean washed flints and limestone bowlders, often bleached to a chalky white color in the glaring sunshine.
In places the waters make wide and deep pools having the peculiar light sea green color characteristic of all the spring rivers breaking from the Cretaceous limestone of southwest Texas. Elsewhere they flow for long distances over a single stratum of horizontal limestone. In the latter case the water spreads out in a thin sheet only an inch or two in depth and sometimes 100 feet wide.
It is almost impossible for the traveler who has seen the continuation of this same stream in the dry region of the Rio Grande Plain to recognize it in the beautiful flowing river now before him. Forests of ash, pecan, and elm fill the valley, while gigantic cypresses border the water. If he should chance upon one of these water holes without having traced the continuity of the stream course, he would believe that lie stood upon the banks of a large and continuously flowing river. He would soon find, however, that after flowing a short distance the water would disappear, either by absorption into the bed of the gravelfilled stream way or through fissures in the solid underbed. These running water holes are constant, and do not depend upon the local rains, but are supplied by perennial springs draining the rocks underlying the plateau.
209
Above the springs at the head of the flat bottomed valleys the limestone bluffs become more vertical and close in toward each other, changing the cross profile of the canyon from U shape to V shape. In this, which may be called the cataract portion of the stream way, the gradient is much steeper, ranging from 100 to 300 feet per mile. here the stream way carries no water except in time of flood, but then it is a series of cataracts leaping from one stratification plane clown to another. The cataract portion of the stream way is usually through the upper part of the Edwards limestone, the rock which makes the scarp rock or mesa edge of the plateau. The flat summit, or "divide," is finally reached after climbing the rough and rocky slopes; and here the drainage is represented only by a few long, shallow caletas.
In resume, each of these stream courses may be said to present four well marked aspects:
- (1) the summit caletas, or draws;
- (2) the cataract portion, where they plunge from the summit through V shaped canyons (tijeras) to the flat bottomed canyons;
- (3) the flat bottomed canyons (plazas), and
- (4) the continuation of the dry stream bed across the lower plain.
The caletas and upper canyons are usually dry and waterless arroyos except in time of storm. The flat bottomed canyons contain permanent pools of flowing water, fed by springs, and on the lower plain the running water disappears entirely or for a considerable distance.
The chief laterals of these major canyons of the plateau are also peculiar and consist entirely of dry arroyos, carving the marginal canyon walls into thousands of entrant and reentrant curves, producing circular, flat topped hills, with slopes composed of alternations of small scarps and minor slopes. Toward their heads these laterals sometimes expand into amphitheatral basins, from which minor laterals ramify in palmate arrangement. These basins are covered with the wash soil of the adjacent hillsides, and are locally known as "grass valleys." So evenly is the debris scattered over them that it is often difficult to trace the stream way. Water is found in these laterals for only a few minutes or hours after each ordinary storm, and is evaporated or imbibed before reaching the main stream, but in time of sudden and very heavy cloudburst rainfall they carry gigantic and dangerous torrents. Each torrent moves the " wash" of the land a step nearer the major stream ways, and rolls the gravel in the beds of the latter onward toward the lower country of the Rio Grande Plain. Thus it is that an intermittent stream of gravel is and has been for a long time gradually flowing away from the Edwards Plateau and spreading over the Rio Grande Plain.
There are many interesting caverns in the Edwards Plateau, and inasmuch as their occurrence, together with the general question of limestone solution, has great bearing upon the distribution of underground water, it is essential that they be briefly mentioned. They are "
A V-shaped canyon, from the Spanish for scissors.
210
of three general types:
- (1) small cavities within individual limestone strata, giving them what is locally termed a honeycombed structure;
- (2) open caverns occurring in certain bluff faces along the stream valleys;
- (3) underground caverns of vast extent dissolved out of many strata
The flora of the Edwards Plateau presents many peculiar variations from that of the adjacent regions, especially the Rio Grande Plain. It shows three distinct phases, viz: the phase of the stream bottom, that of the breaks, and that of the summit. The low, alluvium filled valleys of the rivers present conditions of loose soil and constant moisture favorable to the growth of trees; hence narrow ribbons of forest are found along the streams and extending up into the semiarid region, far west of the limits of the upland forests of the humid region and dissociated from them. These embrace many species, such as the elm, chestnut oak,walnut,sycamore,cypress,live oak, and pecan. The live oaks and pecans attain great size and beauty. The occurrence of the cypress is a peculiar anomaly. This tree, which ordinarily grows only in the swamps and bayous of the low subcoastal regions, attains an enormous size at the edge of the deeper holes near the heads of permanent water of the Pedernales,Blanco,San Marcos,Guadalupe, Cypress,Onion Creek, and other streams. These localities are at altitudes from 1,000 to 1,750 feet above the sea, hundreds of miles west of the great cypress swamps of the eastern tier of Texan counties, with which they have no possible continuity. We have not noticed the cypress, however, in the Nueces, which flows front the still more arid western edge of the south side of the plateau, although the other trees usually accompanying it grow at Kickapoo water hole as large and as luxuriantly as elsewhere. These ribbons of valley forest near the headwaters of the Nueces and Frio terminate with the southern edge of the Edwards Plateau, practically ceasing at the margin of the Rio Grande Plain, and heave no connections with any other forest region whatever. Accompanying them are numerous ferns, maidenhair and
211
other delicate species, all plants requiring continuous moisture, while even the familiar mullein and Jamestown weed occur. This flora of the valleys is of interest, inasmuch as it is a modified representative of that of the great Atlantic timber belt, occurring as an isolated outlier in the semiarid region, preserved and nurtured in these valleys by the presence (due to geologic causes) of water and soil.
While the valleys support this modified flora of the humid region, the rocky slopes of the breaks between the streams and the summit present another group of vegetation shrubby trees which prefer the crevices of rocks, or small scrubby plants which develop large, strong, and hearty roots, out of proportion to the size of the growth above ground. These plants are found along the ledges of limestone wherever their roots can find a hold, or on the almost soilless outcrop of the interstratified chalky marls. Among them are dwarf oaks, the pinon Sopliora, and mountain juniper, which seem to prefer to follow the ledges of loosely jointed rock, besides many coriaceous perennials, including the agarita (Berberis) and the eastern yucca. There are also many small species of Compositor, Liliaceae, the wild poppy (Argemone mexicana), and other plants growing on these slopes.
West of the Frio, in the breaks of the south end of the Edwards Plateau, where the rocks are hotter and more arid than to the eastward, the remarkable and unique resurrection flora of the limestone mountains of Mexico is found. This is characterized by plants growing upon the hot and sterile rocks, and adapted to irregular rainfall, long drought, and scarcity of soil by their thick, coriaceous parts, which ordinarily look dry and dead, but which rapidly unfold and revive after a rain, taking advantage of every drop of rainfall in order to store sufficient moisture to enable them to survive the long periods of intervening drought. Even the ferns, mosses, selaginellas, and kindred plants, which ordinarily constitute our ideals of delicate and tender herbage, have in this region a thick, leathery texture. On the almost barren surface of these rocks grows the melon shaped, edible cactus the devil's pincushion.
Here also, for the first time in proceeding westward across Texas or northward across the Rio Grande Plain from Mexico, one meets the peculiar plants (agaves and yuccas) which become so marked a feature farther westward, the serrated sotol, with its flower stalk rising to a height of 15 feet; the dagger like lechuguilla or ixtle plant, and several species of yucca not seen farther eastward.
The flora of the summit of the plateau is radically different from that of the breaks or of the valleys. All trace of shrubs or trees disappears, save here and there a patch of shin oaks and dwarf evergreens, and in time of verdure the eye beholds apparently a never ending sea of grass, "
In the lower slopes, around the headwaters of the Frio (Frio water hole), the writers last year discovered a large area of the edible Berberis swaseyi Buckley, a species which Coulter says in his Flora of West Texas is known only from the canyons of the Pedernalis. The fruit of this plant is a large edible berry well worthy of cultivation. We have never seen it elsewhere.
212
through which appear many bright colored flowers. This is the southern end of the flora of the Great Plains region, which continues far northward.
Although there is much agriculture in the wide, fertile, plaza canyon valleys indenting the plateau, especially in Blanco, Gillespie, Comal, and Kendall counties, the slopes and summits constituting the larger part of the area are not adapted to agriculture, owing to the rocky character of the soil, the semiaridity of the country, and the impossibility of irrigation. So little are they fitted for agriculture that the extent of the summit of the Edwards Plateau can almost be traced upon the map by the scarcity of post offices and other evidences of population. They constitute, however, good grazing country and support many large sheep and cattle ranches.
GENERAL PRINCIPLES OF ARTESIAN WATERS.
The rocks of the earth forma system of natural works by which water is collected, stored, and distributed. They constitute the basins, reservoirs, conduits, and other portions of the plant for retaining and distributing underground water.
The details of natural waterworks differ in different places in the same way that the details of artificial systems differ in different plants, but the distribution in both is governed by the common principles of hydrostatics. The efficiency of a natural system is determined by the texture of the rocks and the geologic structure of the region, so that an understanding of the availability of underground water in any region necessitates a knowledge of the elementary geology of that region.
It is neither convenient nor advisable here to discuss minutely the source, storage, and distribution of underground water, but in order that the subject may be understood, we shall. give a brief explanation of the elementary principles governing its occurrence.
- (1) The primary source of all underground water is the rainfall.
- (2) Rocks imbibe water. Imbibition may take place by percolation and by absorption. By percolation is meant the process by which water proceeds through cavities, cracks, fissures, or other breaks in the continuity of the underlying rocks. Absorption takes place when water enters the small interstices of the rock.
- (3) Water can flow in rocks, and the rapidity of this flow is known as the capacity for transmission.
- (4) Different kinds of rock have different capacities for imbibing and transmitting water. These capacities are not proportional to one another, but vary independently, according to the kind of rock.
- (5) All water entering the earth tends to gravitate downward along lines of least resistance (easiest transmission).
- (6) The only ordinary agency by which rock sheets may be naturally
drained of their water is gravity. Gravity drainage may be of two
kinds, direct or artesian.
- (a) Direct drainage is that by which water
- (b) Artesian drainage is that in which the water rises above or escapes from a level higher than the bottom of the hydrostatic column. The simplest illustration of this principle is the equilibrium maintained by water in the two limbs of a U-shaped tube. Water from high reservoirs carried by pipes to the upper stories of buildings also illustrates this principle. Artesian water will not rise higher than its head or source, and its maximum pressure varies with the difference between the altitude of the head and that of the outlet.
213
escapes at the bottom of a hydrostatic column. If sheets, beds, or masses of rock containing water be cut into by a well hole, a ditch, erosion, or other means, to a lower level than that at which the water entered, all the water contained in the rocks above the level of the incision will be drained from the rocks by the simple process of seepage, just as water escapes from the bottom of a moist sponge or from the faucet at the bottom of a barrel. This is the method by which water occurs in ordinary nonartesian or dug wells, and in seepage springs so frequently found along the low banks of rivers, valleys, or other natural incisions into the strata. Modern agricultural field subdrainage depends upon the same principle.Water which falls upon the surface of the earth as rain is disposed of by surface run off, by evaporation, and by absorption into the underlying rocks. The water forming the run off passes over the surface to form streams and lakes; that which is evaporated passes again into the atmosphere; the remainder sinks below the line of evaporation into the rock mass beneath, supplying wells and springs. The water of the last class forms the subject of this paper.
CAPACITY OF ROCKS FOR ABSORBING MOISTURE.
Rocks of open texture, such as loose sands and sandstones, gravels, and chalk, have a sponge like capacity for imbibing water. Water poured upon sand will quickly disappear by imbibition. If we wish to filter water, we run it through beds of sand or gravel. Bricks are sprinkled before they are put into buildings, and they absorb from 20 to 60 per cent of their weight of water. On the other hand, if one wishes to shed water or otherwise prevent its percolation, one constructs roofs of tile, makes tables of marble, and builds tanks and cisterns of cement or clay. Few stop to consider, when thus using rocks, that they are making practical application of the broad principles which control the occurrence of underground water. By careful and accurate experiments, such as anyone can make, the capacities of all the known rocks for the imbibition and transmission of water have been determined, and it is shown that sandstone or chalk will absorb many times as much water as slate, marble, or granite. These facts can be observed after every rain. If the rain falls upon loose, sandy soils, like those of the two Cross Timber regions of Texas, it quickly disappears by absorption; if, on the other hand, it falls on the clay
214
soils of the prairies, it stands for some time in pools. such as those called "hog wallows" in Texas.
The capacity of rocks for the transmission of water is entirely different from their capacity for imbibition. If one could construct of sand, clay, slate, granite, chalk, and close textured limestone filtering vessels of equal capacity, and then fill them with water, one would find diverse results, illustrating the capacity of these rocks for transmission of water. Water would pass so slowly through the close textured limestone, slate, and granite that the quantity filtered would be practically imperceptible. At first the sand and chalk would drink in the water equally fast, but after complete saturation it would require longer time for the water to percolate through the chalk than through the sand.
The distribution of underground water is dependent upon the arrangement of rocks in sheets or strata. By a simple arrangement of the porous sands between impervious materials, nature has constructed reservoirs and conduits for the retention and distribution of water, and it is by the character of the arrangement of the rock sheets that the negative or positive conditions for the procurement of artesian waters are determined.
A stratum usually consists of two related parts the outcrop and the embed. That portion exposed at the surface of the earth is the outcrop, and that portion which is concealed underground beneath and between the other rocks may be termed the embed. The outcrop of a water bearing stratum constitutes its main catchment or receiving area, and the embed constitutes the storage reservoir.
If the water bearing rock sheet is inclosed between impervious beds, and inclines beneath the surface, the water will be conducted to a lower level than the outcrop and will remain stored in the earth under hydrostatic pressure until an outlet is provided for it. A water impregnated stratum embedded in this manner is an artesian reservoir. The water bearing strata, together with the impervious strata beneath and above them, constitute an artesian system.
When water is conducted downward to a lower level by an embedded stratum, it acquires a tendency, due to hydrostatic pressure, to rise higher than the overlying retaining bed. When the beds overlying such a water bearing stratum are penetrated by an opening, artificial or natural, the contained water will rise through the overlying bed. Such earth waters which rise under the influence of hydrostatic pressure through an opening to a higher level are known as artesian waters. Hence an artesian well may be defined as one in which the water rises by means of hydrostatic pressure above the top of the embed. If artesian waters rise to the surface, they are known as flowing artesian wells; if the water fails to rise to the surface in a well, such a well is known as a nonflowing artesian well. The height to which the water will rise depends upon several conditions. If the water bearing stratum is embedded and incised at only one place down the dip from its surface outcrop, the water at the place of incision will rise to the same level as
215
the line of complete saturation of the bed. This level is somewhat lower than that of the lowest surface outcrop. If there are several incisions, the question is not entirely one of static equilibrium. For instance, if water is naturally received by the stratum in one part of its outcrop and discharged from another lower part, the conditions in the intervening area are those of dynamic rather than static equilibrium, and the water will rise in a well to a level intermediate between the receiving and discharging levels.
The artesian water bearing strata of the State east of the Pecos River are composed mostly of extensive sheets of sands, clays, and limestones, succeeding one another in orderly arrangement, except along the Balcones zone of faulting, and in general having a gentle inclination toward the sea, so that intraveling northwestward, although constantly ascending in altitude, one encounters the outcropping edges of rock sheets of lower and lower stratigraphic position. This produces the simple arrangement of a tilted plain built up of a series of alternately impervious and pervious layers. The rain falling upon the outcropping edges of the latter, sinks into the embed, and by gravity is conducted seaward down the plane of its inclination to lower levels beneath the surface. Each different stratum, including any particular water-bearing stratum, becomes embedded deeper and deeper to the southeastward of the point where it outcrops at the surface. This structure is very simple and its detail can be traced out, measured, and mapped as accurately as that of the successive layers of stone in a building. To do this is the work of geologists, and it is not a matter of speculation and hypothesis, but simply the application of specially acquired knowledge, similar to that which one must possess to be proficient in any profession.
The foregoing principles are all applicable in explaining the numerous artesian wells and springs of the Rio Grande Plain and the headwater springs of the Edwards Plateau, but the explanation requires also an understanding of the order and arrangement of the strata, and we therefore proceed to consider the geology of the region.
GEOLOGY OF THE REGION.
The rocks of this region consist of various beds of marl, clay, limestone, sandstone, etc., a few igneous rocks (proportionally a very small part of the whole), and some alluvial or surface deposits derived by erosion from the older formations. With the exception of the igneous and alluvial rocks, the strata are all composed of material which was laid down beneath the surface of the ocean, and embedded in them are found many marine fossils, the remains of animals which inhabited the waters of the old ocean. These fossils are of the greatest value in determining the geologic position of the beds containing them, and therefore we give from each of the principal beds a few illustrations of the most important and most abundant forms, by the aid of which the layman can readily recognize the geologic horizons. (Pls. LI-LXIV.)
216
The nomenclature used in describing these rock sheets is as follows: The geologic unit is the bed or stratum. Series of lithologically and paleontologically related strata are grouped into formations, as the "Glen Rose formation." When the strata are all of one kind of rock the name of the rock is sometimes substituted for the word formation, thus: "Edwards limestone" or "Edwards formation." Groups of this class may be embraced in still larger groups called divisions, as the "Fredericksburg division," and these may be grouped into series, as the "Comanche series," and these compose the still larger geologic groups, such as the Cretaceous or the Eocene.
As the work of geologic investigation. has progressed new facts have made necessary the revision of classification, and new names have from time to time been substituted for those first employed. In the classification herein presented the previous nomenclature has been refined by substituting appropriate geographic names (mononyms where possible) for all formations and abandoning the use of paleontologic and mineralogic names. Thus the terms "Dinosaur sands","Caprina limestone," "Exogyra texana beds","Exogyra arietina clays","Fish beds","Exogyra ponderosa marls","Hippurites limestone," etc., have all been replaced by appropriate geographic names. Even some geographic names originated in earlier writings have been abandoned because found to duplicate names previously used for different formations elsewhere.
The sequence and systematic classification of the formations of the region tinder discussion are presented in the table below, and in the following pages the several formations will be individually described. They will be taken up in the order in which they were deposited, the oldest and lowest first.
 |
217
PALEOZOIC FORMATIONS.
Below the Cretaceous rocks which constitute the surface of the Edwards Plateau and Rio Grande Plain lies a foundation of the older and different rocks of the Paleozoic system. These are exposed by erosion only along the foot of the northern scarp of the plateau in Burnet, Blanco, Gillespie, and Mason counties. So far as known, no drill has as yet penetrated them beneath the Rio Grande Plain, and as they are so deeply buried they need little consideration in the presemt discussiom. In the Edwards Plateau, however, the well drill has reached the Paleozoic rocks beneath the Cretaceous in several places, notably at Kerrville and at Morris ranch, in Gillespie County. From data derived from these drillings we are led to believe that the horizontal Cretaceous beds of the Edwards Plateau lie upon an uneven floor of the older rocks. The probable relation of these Paleozoic rocks to the overlying Cretaceous is shown in the figure of the wells at Kerrville (fig. 68, p. 270). One hundred feet of Carboniferous rocks form the base of the section exposed at the month of Hickory Creek on the Colorado River (see p. 220).
THE CRETACEOUS.
The Cretaceous formations of Texas are by far the most important in the State, in both areal extent and economic value. They are seamade rocks, and rest unconformably upon the older Paleozoic rocks, which formed the ocean floor when their deposition began. They constitute almost the entire surface of the Edwards Plateau and much of that of the Rio Grande Plain. Their southern margin is covered by the Eocene overlap, and they are overlain in places by superficial deposits of gravel of Neocene and Pleistocene age. They contain all the artesian water described in this paper, and for this reason it is important that their sequence and occurrence be well understood. They likewise supply the most valuable building material stone, lime, and cement-and some of them. contain oil and gas.
They are mostly limestones and clays, sometimes containing slight admixtures of very fine silica, but there are also great beds of sand and sandy mixtures, occurring principally at the base, middle, and top of the group. The limestones are of many kinds, predominantly light colored, more or less chalky in texture and composition, and rarely of the hard, ringing, close textured character, such as is usually met in American Paleozoic limestones. Some of them are very porous, even cavernous, in texture; others are more massive; some have slight admixtures of magnesia, clay, or silica; others are almost pure carbonate of lime. In some instances they are soft, often marly; in others, quite indurated. The clays are likewise very chalky, so that on exposure to air they readily crumble and weather into soils. The many
218
beds of clay and limestone resemble one another so much that it is difficult for one not accustomed to making geologic observations to discriminate them. The geologist who has studied them carefully can readily recognize them, not only by slight lithologic differences, but by the peculiar fossils which characterize them.
The many formations which make up the Cretaceous system are classified into two great series, the lower of which has been termed the Comanche and the upper the Gulf series. They have also in various parts of Texas been classified into smaller groups called divisions. As most of the formations are of limited extent as compared to the whole system, these divisions are of importance for the indication of the chronologic relations of formations occurring in different regions, and they have therefore been indicated in the table on page 216, but they are not essential to the discussion of the economic questions of this paper.
COMANCHE SERIES (LOWER CRETACEOUS).
In southern Texas this series includes the greater part of the Cretaceous system, and its total thickness is somewhat more than 1,500 feet. At the base the beds are of feebly coherent sand, then come marls and limestones in alternation, and then a heavy body of limestone, followed by other marls and shales.
THE BASEMENT BEDS.
The basement beds of the Comanche differ from the overlying formations of the series by the fact that they are usually composed of sands instead of calcareous marls and limestones. These are more or less fine grained and slightly compact, accompanied in places by small pebble conglomerate, and locally varying in composition according to the material of the adjacent rocks from which they were derived. Thin beds and lamina, of clay occur in the sands, and the residual surface usually has a reddish color. Furthermore, the outcrop of these beds usually bears a growth of timber such as is found in the region known as the Upper Cross Timbers. While the basal beds of the local Cretaceous section, wherever exposed, are usually of this sandy nature, these beds are not everywhere of exactly synchronous deposition, for they represent the littoral of a sea which was progressively transgressing northwestward across an uneven land during the whole Comanche epoch.
So far as the region considered in this paper is concerned, the lower beds and their relations to the overlying and underlying rocks are exposed chiefly in the slopes of the Colorado drainage in western Travis, eastern Burnet,northern Hays,Gillespie, and other counties along the northern edge of the Edwards Plateau, and especially along a line between the town of Burnet and Travis Peak post office, via Smithwick
219
Mill. There is no doubt, however, that these formations occur embedded beneath most of the Edwards Plateau and Rio Grande Plain, and that if a drill could penetrate far enough it would reach them anywhere beneath these regions. They exhibit such variation from place to place that the precise correlation of different occurrences is difficult, and it has seemed best to give separate names to the rocks at the two most important localities.
TRAVIS PEAK FORMATION.
These beds are especially well displayed. on both slopes of the valley of the Colorado River, between the mouths of Sycamore and Cypress creeks, in Burnet and Travis counties. The name they bear was given them because they are well exposed in the vicinity of Travis Peak post office.
While they are arenaceous in composition and porous in texture, like the basement beds of the Comanche in general, they differ considerably from the allied beds to the northward. They consist of conglomerate, composed of coarse rounded pebbles of Silurian and Carboniferous limestones, granite, Llano schists, quartz derived from the adjacent Paleozoic rocks, beds of finely cross bedded pack sand, white siliceous shell breccia resembling the Florida coquina, and some clay.
At the base is usually conglomerate. Succeeding this is coarse, angular, cross bedded sand, which becomes more finely triturated until it reaches the condition known in Texas as "pack sand" i.e., a very fine grained, loosely consolidated sand, cemented by carbonate of lime. In the sands are occasional patches of red and greenish white clays, resembling very much the characteristic colors of the Potomac beds of the Atlantic coast, and they are sometimes accompanied by lignite and fossil bones.
The following section by Mr. J. A. Taff will give an idea of the sequence and composition of the formation as exposed in the valley of the Colorado, in the locality between Travis Peak post office and Smithwick Mill, Burnet County:
 |
See Annual Report Geol. Survey Texas, Austin, 1890, p. 118.
All sections given in this paper are described from the top down and are numbered from the hottom up.
Third Annual Report Geol. Survey Texas, Austin, 1892, p. 295.
220
 |
The sandstone contains grains of silica from the size of a pea to the most minute particles, and small subangular fragments of clay in the cement of lime.
Fossils occur in these beds as low down as the contact conglomerates, but they are neither plentiful nor distinct. The upper or coquinalike beds are full of casts and molds, among which are undetermined species of Trigonia, Pholadomya, and Cyrena, and Ammonites justinae.
In these beds also appears the first of the several oyster agglomerates of the Comanche series. This is composed of a solidified mass of large oyster shells, forming a stratum 7 or 8 feet in thickness, just below the junction of Post Oak and Cow creeks.
Accompanying the oyster breccia another noteworthy feature of the Trinity division appears i. e., an excess of epsom salts, or magnesium sulphate. The oyster shell bed effloresces into a powdered earthy substance accompanied by the epsom salts. Magnesian and pyritiferous layers occur in other horizons higher in the division, and their presence is no doubt in part the cause of the mineral character of some of the artesian waters, especially those wells which are not drilled into the basement sands below these layers.
At the top of the sandy beds in the Colorado section a yellow, arenaceous, fossiliferous limestone appears. This marks the first or lowest appearance of the peculiar fossils Monopleura (Caprotina) and Requienia, "
See artesian water discussion, p. 300.
221
and indicates the beginning of the conditions which finally produced the Glen Rose formation.
The Travis Peak formation records a subsidence of the land which was taking place during its deposition. As the waters deepened the deposits changed from coarser to finer material, becoming more comminuted and calcareous at the top of the beds, until the sand grains are so fine as to be almost imperceptible to the eye, the whole mass becoming quite chalky and "magnesian" in appearance.
GILLESPIE FORMATION.
In Gillespie County are other exposures of the basement beds of the Cretaceous, resting upon the Paleozoic. These appear in the valley of Barron Creek at Fredericksburg and along the Pedernalis below that town. They are composed of sandy grits and clays of a more brilliant vermilion hue than is ordinarily met with in similar deposits elsewhere. As shown in fig. 55, amid in fig. 74, p. 314, they rest on the Paleozoic and grade up into the upper beds of the Glen Rose formation. These might possibly be correlated with the basal sands of the Travis Peak formation, but they are more probably the stratigraphic equivalent of the lower portion of the Glen, Rose formation where it rests on the Travis Peak.
 |
GLEN ROSE FORMATION.
This consists largely of beds of flaggy argillaceous or massive chalky limestones, alternating with thin strata of marly clay, white and yellow in color. From this alternation of limestone and clay the beds were provisionally called the "Alternating beds" when first differentiated by the senior author.
The indurated beds consist of white and yellow limestones of either brecciated, crystalline, arenaceous, magnesian, or chalky structure. They are of varying thickness and of great uniformity in extent. They
222
are separated by softer, unconsolidated, slightly argillaceous marls of oolitic structure, sometimes white and sometimes yellow.
The Travis Peak formation grades upward into the Glen Rose without break in the sedimentation, as can be seen in the high bluffs of Cow Creek, immediately below Mr. Heasel's house at Travis Peak post office, in the western part of Travis County.
The basement beds of the Comanche series have been described as being predominantly of an arenaceous character. The Glen Rose formation may be distinguished as being essentially calcareous. Each of the subdivisions, however, is accompanied by the material of the other as accessory. Thus the calcareous Glen Rose formation is slightly arenaceous at its base, the arenaceous material being siliceous grains so triturated that they are reduced to an almost impalpable powder. This gradually diminishes as we ascend in the beds, and the lime and clay proportionately increase.
The lowest Glen Rose beds are marked by the appearance of strata of homogeneous texture, such as "magaesian" marls and hard layers in which the fossil Requienia² occurs. The name "Caprotina Horizon No. 1" has been applied to these beds, because in the earlier geologic literature the fossils now called Requienia were termed Caprotina. The top of the Glen Rose formation is just below a bed of yellow marl which is persistent over a great area in central Texas, and is the culminating horizon of the oyster Exogyra texana, after which this bed has been called.
In nearly all complete sections the Glen Rose formation shows three marked subdivisions. The lower and upper thirds are composed of thinly bedded and alternating marl and flags, usually weathering into terraced slopes; the middle third is made up of thicker and more massive beds which constitute bluffs. Some of the beds near the base of the thicker layers are quite chalky in texture and carry many peculiar fossils, especially noteworthy being a large foraminifer (Orbitolina texana), besides many large casts of mollusks. The lower portion of the formation carries much fine arenaceous material accessory to the calcareous material. and its indurated and unindurated beds do not occur in such uniform alternations as do those of the upper third. For instance, there will be 10 or 12 feet of soft, friable material and then a thin layer of less than a foot of indurated stone. In weathering this results in wide terraces with steep slopes.
The yellow magnesian strata also increase in thickness in ascending series, and become very conspicuous in the middle portion, often being from 5 to 15 feet in thickness, as seen in the bluffs of Mount BonNel, near Austin. These magnesian limestones are soft and of a cream or brownish yellow color, and alternate with strata of marls similarly constituted, "
The term magnesian has long been applied to certain yellow strata in these beds. Whether they are or are not magnesian in composition we can not state positively.
A species of this genus of shells from the Edwards limestone is figured on P1. LIV, fig. 1 a,1 b.
223
and are sometimes accompanied by pockets or nodules of calcite, aragonite, strontianite, celestite, and epsomite.
The upper third of the formation, as seen at the top of Mount Bonnel, presents alternations of friable marls and hard limestone strata. The limestone strata usually average less than a foot in thickness. These alternations occur with great regularity and persistence. Clay is the chief accessory of the calcareous beds. The marls are soft and laminated and are composed largely of minute shell fragments, giving the beds a distinctly granular, oolitic character. They have little clay and imbibe the moisture very freely.
While possessing no great agricultural possibilities, the basal or alternating beds are capable of producing valuable building material, among which are building stones. Some of these have rich "magnesian" buff yellow colors, while the limestones often resemble the stones of Caen, France, which are imported into this country. Some of the beds are also valuable for the manufacture of hydraulic cements, although at present they are not utilized. These rocks also contain undeveloped beds of epsom salts, strontianite, and other materials.
The alternations of horizontal beds of soft marls and hard limestones above described produce the bench and terrace topography of the slopes of many of the canyons and along the margin of the Edwards Plateau from the East Fork of the Nueces to the Colorado, where the streams have cut downward through
224
the Edwards limestone. Of this character are the beautiful terraces and bluffs of the Colorado, as seen from Mount Bonnel westward to the Burnet County line, as well as those of the Pedernales, Guadalupe, Comal, Medina, Hondo, Frio, and all the numerous streams indenting the southern margin of the plateau. In addition to the localities already mentioned the beds are exposed west of San Antonio along the line of railway between Aue and Boerne and in the valley of the Guadalupe above Kerrville as far as Via. They are also exposed in the lower slopes of the valley of the East Nueces and its tributaries below Vance to a point a few miles south of Montell. The channel of the West Nueces has barely cut down to their top and exposes them at only two places along the stream bed at Kickapoo springs and in the north bend of the river in northern Kinney County.
The accompanying detailed section of the entire thickness of the beds of the bluffs of the south side of the Colorado in the vicinity of Round Mountain, Travis County, is typical of these beds. It coincides almost exactly with Mr. Taff's Sandy Creek section, previously measured, on the opposite side of the river.
 |
Mr. Taff's section measured 447 feet, or 8 feet less than ours. Geol. Survey Texas, Third Annual Report, 1891, Austin, 1892, pp.298-299.
225
 |
The rocks of the middle of the Glen Rose formation are the oldest exposed in the Nueces Valley. The thickness of the formation in this region, estimated from studies in the vicinity of Kerrville, is approximately 500 feet, which if true would indicate a uniform thickness along the entire line of strike across the region treated in this paper. This uniformity of thickness is not maintained along the line of dip, however, as will presently be shown. Taff's measurement of the Travis Peak formation makes it about 213 feet in thickness at its outcrop in the Colorado Valley.
It is difficult to determine the thickness of the embedded portions of these strata, especially the Travis Peak. The careful measurements
226
of the surface outcrop do not coincide with the artesian well borings made to the east, the latter showing much greater thickness. For instance, the borings of the San Marcos artesian well (see p. 287) show a thickness of 904 feet of the Glen Rose and Travis Peak penetrated by the drill at that place. At Austin also the well records indicate that at least 1,215 feet of the two formations are penetrated. Neither of these wells has as yet reached the bottom of the Travis Peak. These data indicate that the aggregate thickness of the formations increases rapidly coastward away from the line of outcrop. The beds were deposited on the eastern slope of an old, subsiding, preexisting Paleozoic upland, against which the formations of the Cretaceous in general and of the lower beds in particular were cumulatively deposited until it was buried, and toward which the calcareous beds changed into an arenaceous character.
WALNUT FORMATION.
If the classification of the Cretaceous formations had originally been made in the region here discussed, the Walnut formation would probably not have been given separate status, for, though well developed farther north, it is here thin and unimportant. In its typical development along the Brazos River it consists of laminated clays alternating with limestone flags, and both clays and flags are accompanied by great quantities of the two peculiar species of oyster, Exogyra texana and Gryplaea marcoui sp. nov. H. & V., Pl. LIII. Along the Colorado River at the northern limit of the Edwards Plateau the formation is only 10 or 12 feet thick, consisting of thin, friable, yellow, arenaceous marls, in which are great numbers of Exogyra texana. It determines a distinct bench near the summit of the high mesas west of Austin. Southwestward, toward the Nueces, it becomes less and less distinguishable, until at that stream there is only a foot or two of yellow clay, accompanied by the characteristic fossils.
COMANCHE PEAK LIMESTONE.
This is a persistent bed of white chalky limestone, presenting a shattered reticulated appearance on weathering. It is partly characterized by an abundant fossil fauna containing a large number of Exogyra texana, which is especially abundant in its basal portion. It is from 40 to 50 feet thick, thinning toward the Rio Grande. Although it is insignificant as regards thickness, and lithologically might be considered the base of the Edwards limestone, it is one of the most persistent paleontologic horizons of the Texas Cretaceous section.
"Gryphaea pitcheri(in part) of previous writings; described in Bull. U. S. Geol. Survey No. 151.
227
EDWARDS LIMESTONE.
This formation is the most conspicuous and extensive in the Texas Mexican region. It is composed mostly of limestone, but there are some marly layers. It shows slight variation in color, composition, texture, and mode of weathering. In general the beds are whitish, although layers of buff, cream, yellow, or dull gray are frequent. These colors depend much upon weathering. In composition most of the beds are as nearly pure carbonate of lime as can be found in nature, but some have small admixtures of silica, epsomite, chloride of sodium, and perhaps other salts as yet undetermined. Clay is absent except as a minor constituent in the few marly layers. Iron is sparingly present as pyrites, and is revealed by the red color of the clay that weathers out of a few beds. Exceedingly fine siliceous particles occur in the so called "magnesian beds " light brown porous beds which appear southward from Comanche County but no pebble, bowlder, lignite, or other undoubted piece of land derived debris has ever been found.
The limestones vary in degree of induration from hard, ringing, durable strata to soft pulverulent chalk that crumbles in the fingers and resembles very much the prepared article of commerce. Some of the beds are coarsely crystalline, with calcitized fossils, and are susceptible of high polish. The beds also vary in texture. Some of them are quite porous and pervious, while others are close grained and impervious. Some are homogeneous throughout; others have bard and soft spots, the latter dissolving by the percolation of underground water and constituting what is popularly termed "honeycombed" rocks. The harder spots in some cases seem to be in process of induration, suggesting a step in the formation of flints. The holes in the honeycombed layers often represent what were once spots containing soluble salts of iron and other accessory minerals.
South of the Paluxy River the formation can always be distinguished by the immense quantity of flint nodules which are embedded in and between the limestones and which lie scattered over the surface everywhere. These are of many shapes (Pl. XXX); some are fusiform, like elongated roots; others are knotty, like warty potatoes; others are parts of extensive sheets or very flat lenses. They vary in size from that of a lien's egg to a foot or more in diameter. They also vary greatly in color. Upon fresh fracture some are almost jet black; others light blue, gray, or opalescent; still others are delicate pink in color. There is some evidence that each particular kind occupies a definite horizon, but we are not prepared to state this as a positive fact.
"The geographic name Edwards is here substituted for the "Caprina limestone" of Shumard, the latter being abandoned because it is a paleontologic term; also for the term "Barton Creek limestone" of Hill, abandoned by him and revived by Cragin. The term "Barton Creek" is objectionable, and was abandoned by its author because it was not a good locality name and because it is a two word name. The name Barton has been applied to a division of the English Eocene for many years, but we doubt if, in the absence of any defined code of geologic nomenclature, such an objection may be considered valid.
228
In most cases the Edwards limestone may also readily be distinguished by the peculiar aberrant mollusks of the genera Monopleura, Requienia, and Radiolites bivalve fossils which have cornucopiate form, suggesting a resemblance in shape to the horns of cows, goats, and sheep.
The formation is stratified into a succession of massive beds accompanied by very few flaggy and marly layers. Some of the strata are harder than others and project beyond the softer layers in the profile of the hills as overhanging shelves; others are soft and erode very rapidly. South of the Colorado, where the Walnut formation becomes insignificant, the Edwards limestone is almost inseparable from the underlying Comanche Peak, since both are composed chiefly of carbonate of lime. The Comanche Peak strata are less consolidated, and, as they are somewhat argillaceous, possess a more marly texture than the Edwards limestone, which is usually a firm, white, ringing limestone of great hardness and durability; so that the Edwards weathers into cliffs, while the Comanche Peak is wrought into lower lying slopes; but in most cases reliance must be placed upon paleontologic determinations to distinguish the two formations.
Neither is the Edwards limestone always sharply defined from the overlying Fort Worth, except by paleontologic criteria. It is true that the Fort Worth limestone is slightly more arenaceous, but the differences are so slight that their detection requires the trained eye of the geologist. As the upper limestone is less than 75 feet in. thickness, the layman or well driller unversed in paleontology can nearly always be sure that any rock occurring 75 feet below the Del Rio clays belongs to the Edwards formation.
The Edwards limestone, being more purely calcareous than any other of the Comanche series, probably corresponds to the deepest and most extensive submergence of the Comanche epoch. It is true that in the Glen Rose formation occasional thin beds of chalk are met with. and that some of these are composed almost entirely of foraminifera, but such chalks usually contain a considerable percentage of clay, recognized as an offshore deposit.
Occasional bands of soft brownish yellow stone are intercalated with the limestone. These bands are popularly called "magnesian," and are composed largely of all exceedingly fine grained siliceous element like tripoli. As these beds often contain flints, the silex may be of organic origin.
Topographically, the Edwards limestone is one of the most important formations in Texas. In fact, it is the determining factor in the topography of the whole of the Edwards Plateau and Grand Prairie regions. "
See Plates LIV and LV, of typical fossils.
It is doubtful if the interior of the Edwards limestone is always as hard as the surface outcrop, for many of the chalky rocks of Texas harden or set on surface exposure.
229
Its hardness being superior to that of the overlying and underlying beds, its consequent resistance to erosion has preserved it as the capstone of the innumerable round II mountains" (buttes) and mesas of the State and of the extensive Edwards Plateau and Grand Prairie regions. Not only are most of the buttes and mesas capped by it, but these are accompanied by scarps overlooking the lower lying valley prairies which follow the stream. The walls of the canyons which many of the streams have cut are also composed largely of the Edwards limestone, especially the higher and headwater portions of those rising in the Edwards Plateau. To its hardness is also largely due the topography of the limestone mountains of Mexico.
It shows many types of weathering. Some of the strata make bold cliffs nearly 50 feet in height, the faces of which, although apparently of homogeneous texture, weather into small open caverns (Pls. XXXI and XXXII). This weathering sometimes brings out a thinly laminated structure associated with white efflorescence. The bottoms of caverns of this character are filled with a layer of white, pulverulent earth. The residual products of other massive ledges weathering into caverns are vermilion colored clays, in which are beautiful fossils composed entirely of crystallized calcite.
The hard limestones weather into vertical, square cut bluffs, while the soft and more homogeneous beds of marly or chalky texture form slopes. Where these hard and soft beds occur in alternation there is a corresponding alternation of scarps and slopes in the topographic profile. Some of the beds of homogeneous texture having great thickness weather into pyramidal hills, as seen along the monoclinal fold at the south edge in the northern part of Kinney County (Pl. XXXIII).
Near the summit of the Edwards Plateau, where the flaggy layers prevail, the slopes of the stream ways are gentle and are characterized by low, vertical steps, from 2 inches to 2 feet high, over which the stream descends from one rock layer to another. This slope is usually interrupted by vertical bluffs, composed of thick strata, which constitute a cornice in the profile of the canyon of the plateau. These slopes and scarps alternate until the base of the Edwards limestone is reached, beneath which the Comanche Peak bed weathers out in concave profile.
Where the Edwards formation forms extensive stretches of level country, such as that between Manchaca and Oak Hill in Travis County, in the western part of Williamson County, and the summit of the plateau, and the surface stratum is of homogeneous texture, it weathers into millions of miniature ridges, crests, and drainage lilies, illustrating the whole process of erosion and mountain carving. These minutely eroded limestone surfaces are technically known as "karrenfelder" (see Pl. L, p. 318), and they are formed by the solvent effect of the rainfall upon the sun heated limestone surfaces. The crevices in these level areas of Edwards limestone country are usually grasscovered,
230
with occasional patches of scrub oak. The surface is very rocky, the karrenfelder protruding in jagged points through the rich but, scanty soils. Sometimes residual flints occur in such immense quantities over these surfaces that one is apt to mistake them for a water rolled gravel formation.
In Kinney County the high hills north of Fort Clark, with the exception of Las Moras Mountain, are made up largely of the Edwards limestone. The beautiful canyon of the Pecos, near its mouth, as shown in Pl. XXVIII is one of the finest 39 examples of its outcrop.
A small, narrow belt also outcrops on the downthrown side of the Balcones fault. In Travis County an area of it occurs, as shown on the map (see Pl. XLVI, p. 282), along a belt lying between the International and Great Northern Railway and a line drawn through Oatmanville (Oak Hill) and Mount Bonnel, and many other places as far west as the Frio. Aransas Pass Railway crosses it just south of Leon Springs.
It is well displayed in the banks on the south side of the Colorado in the western part of the city of Austin between McDonald's brick yard and the city dam. Owing to faulting, this section is somewhat complicated and not continuouly exposed at any single locality. The details as made out at three localities at Austin are shown in the accompanying sections and figures (fig. 57). It will be noticed that in the lower portion of the Bee Creek section, which is still above the base of the whole of the formation, arenaceous marls and limestones are quite numerous. These play an important part in the artesian conditions from the Colorado southwestward.