University of Texas Bulletin No. 2645: Decemb~r 1, 1926 THE GUEYDAN, A NEW MIDDLE TERTIARY FORMATION FROM THE SOUTHWESTERN COASTAL PLAIN OF TEXAS By THOMAS L. BAILEY Bureau of Economic Geology J. A. Udden, Director E. H. Sellards, Associate Director PUBLISHED BY THE UNIVERSITY OF TEXAS AUSTIN Publications of the University of Texas Publications Committees: GENERAL: FREDERIC DUNCALF E. K. McGINNIS J. W. CALHOUN H. J. MULLER J. L. HENDERSON G. W. STUMBERG HAL C WEAVER OFFICIAL: W. J. BATTLE R. A. LAW H. Y. BENEDICT F. B. MARSH C. D. SIMMONS The University publishes bulletins four times a month, so numbered that the first two digits of the number show the year of issue, the last two the position in the yearly series. (For example, No. 2201 is the first bulletin of the year 1922.) These comprise the official publications of thfl University, publications on humanistic and scientific sub­jects, bulletins prepared by the Division of Extension, by the Bureau of Economic Geology, and other bulletins of general educational interest. With the exception of special num­bers, any bulletin will be sent to a citizen of Texas free on request. All communications about University publications should be addressed to University Publications, University of Texas, Au3tin. lilHIV!ltllTY 0, Tt1A5 Pltf.:U. AUST\fll University of Texas Bulletin No. 2645: December 1, 1926 THE GUEYDAN, A NEW MIDDLE TERTIARY FORMATION FROM THE SOUTHWESTERN COASTAL PLAIN OF TEXAS By THOMAS L. BAILEY Bureau of Economic Geology J . A. Udden, Director E. H. Sellards, Associate Director PUBLISHED BY THE UNIVERSITY FOUR TIMES A lV; ONTH. AND ENTERED AS SECOND-CLASS MATTER A.T THE POSTOFFICE AT AUSTIN, TEXAS, UNDER THE ACT OF AUGUST 24, 1912 The benefits of education and of useful knowledge, generally diffused through a community, are essential to the preservation of a free govern­ment. Sam Houston Cultivated mind is the guardian genius of democracy. . . • It is the only dictator that freemen acknowl­edge and the only security that free­men desire. Mirabeau B. Lamar TABLE OF CONTENTS PAGE Introduction --------------------------------------------------------------------------------------7 Discovery of the Gueydan Formation_____________ ___ _______ __ ______________ _ 7 Acknowledgments --------------------------------------------------------------_______________ 8 Location and Delimitation of Area Mapped_ ___ __ ___ _ _--·--·--·-··-· 8 Previous Work ---------------------------------------------------------·-·---·--·--------·---9 Name -----------------------------------------------------------------------·----·--·------------15 Physiography ------------------------------·-----------·-----------·----------------------·-----­l':) General Statement ----------------------------------------------------------------------16 Relief --------------------------------------------------------------------------------------------18 Physiographic Divisions -------------------------------------·-·----------------------20 Fayette Hills Belt (Wellborn Plain) ____________________ _______ _____ ___ __ _ 20 Frio Plain --------------------------------------------------------------------------------22 Gueydan Hills Belt__ __ __ ___ ____ ______________ _____ __ ____ ____ ___ __ ______ __ __________ _ 23 Gu-eydan Plain ------------------------------------------------------------------------26 Mesas -----------------------------------------------------------------------------------­ 27 Chalcedony Knobs -----------------·----------·-----------------------------------28 Oakville Hills Belt---------------------------------------------------------·----------29 Reynosa Cuesta Plain·-----·--·----------·-------------·-·-------------------------30 Drainage ___----------------__________________________-----__------------·-----------------------· ---31 Probable Capture or Diversion of Nueces River____ ____ _____________ _ 32 Stratigraphy --------------------------------------------------------------------------------------35 Other Formations Outcropping in the Vicinity of the Gueydan 37 Underlying Formations (Jackson Group)--------------------------·· 37 Fay-ette Formation -------·---------------------------------------------------··­37 Distribution and Stratigraphic Relations .... ·-------------·-­37 Lithologic Character --------------------·---------·-------------38 Ve!n~ in the Fayette ________________________________ ------·--·-------------42 Orann __ ---·---------------------------------------____ ------------------------.. -----43 Frio Formation -------------------------------------------------------------------­ 44 Distribution ___-----------------_______________________________ --------_-------__ . 44 Litholoi;dc Character --------------------------------------·-----------------45 Correlation and Stratigraphic Relations ___ _____________ ____ __ __ ~8 Thickness ---------------------------------------------------------------------·---­51 Origin --------------------------------------------------------------------------··-·-­ 5t Overlying Formations -------------------------------------------------------------­52 Upoer Miocene-Lower Pliocene Seri-es ----------------------·-··-­52 Oakville Formation --------------------------------------------------· ---­ 52 Distribution and Stratigraphic Relations ___________ __ __ _ :):J. 54 ~~1~i~og~~--~~~~~~~e-~__:::::::::::_·:_-_-_-_·_-:_·:_-_-:_·: _-_-_--_-_-_-_-_--_-_-:::::::::::::: 58 Pleistocene Series ----------------------------·----------·-----------·-----------­ 59 Reynosa Formation -----------------------··--·-------··· --·-·---------·---­ 50 Aire and Distribution______ _____ _______ -·------·----------------------·-59 Lit_h?logic Character ---------------·----·--·--------· 60 Oriirrn ______________ -----------____ -... ___ . --__ -·-------. ------------· ---------. --61 The Gueydan Formation ----------------·-·-------··-------·--·-----------------------­ G2 Dip an-d Thickness ---------------------·--·---------·-·--------·---·-·-·----------------t~2 5,; Lithologic Character --------------------------···--·--·------------------------·---­Fant-Member (Lower Gueydan) ________ .. --------·----------------------6t~ General Lithology ________________________ .____________--------------------------66 Lithelogic Variations along the Strike of Fant Member 73 Sections _____________________________________ ._.__ ... ___ ---_-· ---------------------------7b Soledad Member (Middle Gueydan) .. ---·------------------------------80 Volcanic Bould-ers ________ __ ······-·------------·---·--85 Sections ________ ___------------------________ .__ __ .. __ .. ______ ------.. ---·----. --.... --8fl Chusa Member (Upper Gueydan) _ --·---------· -----------89 Sections ----------------------------------------·-·-----··--------··----------------­ 93 Sections of Gueydan and Catahoula in Wells_____________ __ ____ _ 94 Petrology of the Gueydan Rocks..... ··-·--------·--·· ----------·-·-----10'1 Method of Study ------------------------------------· -·----. 10'1 Classification of Pyroclastic Rocks ________ -------------------------------107 Chemical Composition -------· -··--------------·--------·-··------·----110 Conclusions from Chemical Analyses·-·-·--·-···----------llii Table of Contents PAGE Petrography of th-e Gueydan Tuffs______________________ _______________ 110 Texture ------------------------------------------------------------------------·--···-·­ 116 Composition ······----········--·······---···-····--··--------··-··--···-··········-·· 11? Primary Constituents ·-······--····-···· --·····----···-·········· 117 Volcanic Glass -·--·····--··············---············-------············· 117 Feldspar ····---····-····---··--·--··-···········----·-····--··· ················­120 Quartz ---··········--··-· -··············--··-·· ····---····--··· ·······-··········· 121 Accessorv Minerals .. -----····· ····---------·---····················---· 121 Tridyy{iite ·· ·····----··········-· ---··-----------·· ·· ··--· ----··· ·····-··· 122 Secondary Minerals ···---······-·--··········-·--------·---··---·--·-··-··· 12;; C\fontmorillonite (Lcverrierite) ·-········---·················-­126 Opal and Chalcedony --·-----········-------·····---·· ······----····--·· 127 Other Secondary Minerals --·-····-------·············--·-········ 127 Xon-Volcanic Min-erals .... -------········--------···-· ------····--·· lW Petrographic Descriptions of Principal Types of Gueydan Tuffs -······-······ --·---··-······-· ··--------··-·---· ·············--12R Fant )fember ···-····· __ _____ ----··-·· ·--·-·····---------···-················· 128 Soledad :'.\!ember ·-·········---···········-····-------·-----·················· 133 Chusa Member ..... -----··········· ·-···---------····-·····--·-··-··········-134 Petrcgraphy of Lava Pebbles and Boulders ....... ·-··----········· 135 Petrographic Descriptions of Lava Boulders from Fant l\frmber ... ------············-·-·-······-136 Petrographic Desc riptions of Boulders and Pebbles from Soledad Member ·---······--------··········· ·----------··-·-­138 RelatiYe Abundance of Types of Pebbles in Soledad Conglom-erate ······-----················---······-·----······ ·· 141 Petrographic Descrintions of Boulders from Tuff of the Chusa Member ····-------------···---······----···-·-······-·· 141 Petrography of :Non-Pyroclastic Rocks of the Gueydan- Catahoula .. ------········-·····-··--····· ----·---···-··········· 142 Age of the Gueydan --------··--···--··------·-·--··-·····----···---------············ 14() Mineral Veins in the Gueydan (Los Picachos Hills)....... -.. --------148 Origin of Los Picachos Veins. ···········------·---· ---··············-· 151 Siliceous Knobs ··--------····-············-···········-·----·-·--·--·····---·-·-····-······ 153 Intrusions in the Gu-eydan --···-·---·-----··-··-------· -------······--·-------···-l:'i-i Serpentine -------·-···-···········-·---·--·····-· ·--------·-··-·· ·-·----······ ············ 1G4 Origin of the Gueydan . ············------·--········--········------········ ·········· 153 Source of the :\faterial ·--·--··········· ·····--············-------·---·-···· ······-15G Evidence for a Distant Source................ ---·--· -··--·········· ·······--158 Evidence for a Proximate Source ....... _________________ ----·-········-·--· 16:) Evidence for S-everal Volcanoes ......... ______ ____ -----··--···-·-··-··-·· 163 Conclusions as to Location of Yolcanoes __ ·····-·-··-----·······----·· 163 Structural Geology ··········-·--·----------·····-------··· ---········-······· 16-) Faults ···-----·------············-··-·----····---···------··-····· ·····-----· ----··-········--· ·· 1613 Char acter of Displacements.·-····------·-·-····· -··---·--····-·--· 16>l Description of Principal Faults........... --······-----······ ··· 168 Gonzales County --·-··-··· ---·············--------·-·----·-···· ······ 168 Karnes County ---------··· ·-··· ···-·· ·········-· ·-·------·-·-·--···-··· ····· ·---­160 Live Oak County ·--········ ·····-···---·····----· ------·-·· --········-···---· 170 McMullen County ·-------···-·-·· --------·····------·-·-··· ·····-··· ···-···--·· 172 Duval County ·--·········--· ···----·······----......... --·-····--··------· 172 Webb County ---·-···---·····-·· ······· ········ -----·--·········-······ l?!'l Starr County ·-·--··-············-···-·· -····---······ ------········ ----·······-·-173 Folds ··-----···-·-·······-··--·--------········-·· ---·····--·----··· --·-----··--····· 174 Torrecillas Arch ····-·----········---·--· -··--·-·········· -------·····--····· ···-····· 17 ~! Anticlines of Webb, Zapata and Jim H ogg Counties. 17 ! Durnl County Folds ···--·--·······-·----------·-·-·---------··········· 17'3 l\IcMullen Countv Folds ---··-·· . -----·-·-····· --------·-···· ---····-·-····· 170 Wentz Anticline-and Syncline ----------·······--·-· --····· ········ 177 Live Oak County Folds·-··--·------··--·······----------·---··---·············-·· 177 Summary and Conclusions ---··---···-····· ···-· ·-·-············ ··---······--------··· 178 Ind-ex .. 181 LIST OF ILLUSTRATIONS PLATE I. Geologic map of part of southern Texas showing out­crops of Gueydan and adjacent formations. Insert at back. Following Page II to VIII. Views illustrating various features of the forma­tions described in this report _______________________________________________ 179 IX to XII. Thin sections of igneous rock of the Gueydan for­n1ation --------------------------------------------------------------------------------------179 FIGURE 1. Idealized sketch section of Reynosa Escarpment, 1h mile north 20 " east of Mirando City, Webb County._______ _ 54 2a. Diagrammatic sketch showing approximate plan of vein outcrops at Los Picachos Hills. __ __ ____ ___ __________________ __ ___ ____ 152 2b. Diagrammatic sketch of south hill of Los Picachos along line A-A' in Fig. 2a. ________________ --·----------------------------·-------152 DR. EDWARD THEODORE DUMBLE While this bulletin is in press announcement has been made of the death of Dr. Edward Theodore Dumble on Januan· 25. 1927, at Nice, France. Dr. Dumble was the first to" undertake a comprehensfre subdivision of the Ter­tiary and Quaternary se1·ies of Texas; moreoYer, in recent years he and his associates of the Geological Department of the Southern Pacific Railroad ha\·e contributed a val­uable paper on the geology of East Texas to the publications of the Vnfrersity of Texas. It is therefore fitting that in this publication which relates to a part of the area over which Dr. Dumble \Yorked, that special acknowledgment should be made of his contributions to the geology of Texas. Dr. Dumble was born at ::.\Iadison, Indiana, March 28, 1852. At an early age he came \Yith his parents to Hous­ton, Texas, where he resided the greater part of his life. His college \rnrk was taken at \Yashington and Lee Univer­sity, from which he receiYed the degree of Bachelor of Science and later \Yas honored \\·ith Doctor of Science. In his college \York he majo1·ed in chemistry and geology. Although engaged in his early youth in business he did not Iese his interest in science, but privately explored the geology of Texa:::. When the Geological Survey of Texas \ms established in 1888 he became State Geologist and serYed as Director of the SurYey until 1894. Although no State funds \\·ere proYided for this survey beyond February, 1894, Dr. Dumble's connection \Yith the State in an official capacity continued until 1899. The publications issued by this sun·e~· under Dr. Dumble's direction stand as a fitting memorial to his abilih· as a scientist and director of scien­tific actiYities. · In recognition cf his contributions to geology both as an cifficial of the State of Texas and as a private citizen this bulletin of the "Cni\·ersit" of T'exas is dedicated to the memory of Dr. Ed\Yard Theodore Dumble. THE GUEYDAN, A NEW MIDDLE TERTIARY FORMATION FROM THE SOUTHWESTERN COASTAL PLAIN OF TEXAS BY THOMAS L. BAILEY1 INTRODUCTION The announcement of the discovery of an important and extensive deposit of pyroclastic material in the southwestern Texas coastal plain and a brief discussion of its character and stratigraphic relations were given in a short article by the present writer2 in March, 1924. This deposit was designated therein as the "Gueydan formation." The object of the present paper is to describe the general lithology and petrography of the Gueydan formation, or Gueydan tuff; to discuss its stratigraphic relations to the adjacent formations and its partial or complete equivalence in age with the Oligocene rocks of eastern Texas; to discuss the probable mode of origin and source of the volcanic debris; and to outline the structural and physiographic features of that portion of the Texas coastal plain in which the Gueydan tuff outcrops. The recognition of the Gueydan is of importance in this region, where many of the formations are lithologically similar, because it is quite unlike the adjacent formations in character. Certain members of the Gueydan tuff may be useful as key horizons in working out the detailed geologic structure of this district. Discovery of the Gueydan f ormation.-In August, 1923, Mr. Henderson Coquat of Three Rivers, Texas, sent to Dr. J. A. Udden, Director of the Bureau of Economic Geology and Technology of the University of Texas, some 1Manuscript submitted September, 1926, published March, 1927. 2Bai!ey, Thomas L., "Extensive Volcanic Activity in the Middle Tertiary of the South Texas Coastal Plains," Sci. (n. s.) LIX, No. 1526, pp. 299-300, 1924. Unirasity of Te.ras Bulletin small fragments broken from large boulders of a dark vesicular la\'a. Thin sections and crushed portions of these fragments were examined under a petrographic microscope by the \\Titer and were found to be trachyandesite or ande­site. In a note accompanying the samples Mr. Coquat wrote that he had been able to trace boulders of this type across a large part of Live Oak and McMullen counties. With a view to ascertaining the mode of occurrence of this interest­ing igneous material and to obtain an idea of the extent of its outcrop, Dr. Udden asked the ·writer to make a short field study of it. Accordingly, in September, 1923, the writer spent four days in Live Oak and McMullen counties making a reconnaissance examination of the deposits con­taining the boulders and found them to consist largely of volcanic tuff beds. A summary of the results obtained at this time was published in Science as mentioned above. The '"Titer spent the summer of 1924 in making a detailed study of the Gueydan tuff and traced its outcrop for a dis­tance of 240 miles. The present paper is based on the field ·work done at that time. Acknoicledgments.-The writer wishes to express his sincere appreciation to Mr. Henderson Coquat for conduct­ing him to some of the best exposures of the igneous rock in Live Oak and McMullen counties. In fact Mr. Coquat had suspected the igneous origin of much of this material before the writer had seen it. Location a11d delimitation of the area mapped.-The area discussed embraces a belt 240 miles long and from two to fourteen miles wide near the center of the southwestern half of the Gulf Coastal Plain of Texas, approximately parallel­ing the shore line at a distance of from 80 to 110 miles northwest and west of the Gulf of Mexico. This belt includes the surface outcrop of the above-mentioned Gueydan forma­tion which was traced from a point near the town of Moul­ton, Lavaca County, midway between the Louisiana and Mexican boundaries of Texas, to a point twenty-two miles north of the lower Rio Grande at Rio Grande City, Starr Gueydan Formation County, and seventy miles south-southeast of Laredo. The northwestern edge of the Gulf Coastal Plain is produced by the Balcones fault scarp at the edge of the Edwards Plateau. In the northern part of its length the belt is situated mid­way between the Balcones scarp and the Gulf coast. The southern part, however, is nearer to the coast than to the Balcones scarp because of the fact that the main Balcones fault zone changes its southwest trend north of San Antonio to a pronounced westward course south of San Antonio, while the Gueydan outcrop continues approximately parallel to the shore line. The discussion of the physiography is not restricted to the rather narrow belt outlined above, but includes also a general description of the physiography of much of the southwestern coastal plain of Texas. Several of the geological formations outcropping adjacent to the Gueydan are also discussed, although less intensively. Previous work.-The published accounts of the geology and the geologic maps of that portion of the Texas coastal plain outlined above are strictly of a reconnaissance nature. The stratigraphic units have not been clearly defined nor fully described for the most part. This has led to consider­able confusion, inasmuch as the same beds have been included in a certain formation by one writer and in an entirely different formation by another. Identical beds are not uncommonly placed in various formations at differ­ent localities along their strike, even by the same writer. This confusion has been due partly to a lack of extensive outcrops and partly to lateral variation in the lithology of some of these formations. It has also been caused by the fact that the writers did not give sufficient attention to the petrology of the district. A more thorough field study of the rocks would have prevented many wrong designations. For example, many beds of coherent white tuff have been identified as limestone, although they contain either no lime carbonate at all or very small traces of it. A careful examination of this particular rock with the hand lens would have revealed the presence of many pumice fragments and angular glass grains. University of Te:ras Bulletin The first comprehensive attempt to subdivide the Tertiary and Quarternary rocks of the southwestern coastal plain of Texas into formations was made by Dumble~ in a paper published in 1894. The type localities of the Frio, Oakville (Lapara), and Lagarto formations which were named and described in that paper by Dumble are found in Live Oak and McMullen counties near the center of the area here discussed, where the writer did his most detailed work. The Frio of Dumble includes both the Frio and Gueydan formations of the present writer. No pyroclastics are reported from the Frio, although he does mention beds of volcanic dust and ash in the Fayette, which underlies the Frio. His type localities of the Frio clay on the Frio and Nueces rivers are within the outcrop of the Gueydan tuff of the present writer. In 1903, Dumble' published another paper in which he gave much additional information about the geology of southvvestern Texas, including some detailed sections. Accompanying this report is a sketch map showing Dumble's interpretation of the distribution of the various Tertiary formations in the portion of southwestern Texas traversed. According to this map the Frio in Live Oak County wouid include most of the Gueydan as discussed in the present paper and a little of the Oakville, while the Frio proper of the present report is included for the most part in the Fayette by Dumble. In McMullen County, between Tilden and the southeast corner of the county, Dumble maps the Frio so that it includes the Frio proper and a small part of the lower Gueydan as described in this paper. His Frio thus includes a different set of beds in these two sections, as is proven by actually tracing the lower Gueydan tuff beds from one section to the other. The upper half of the Gueydan is included by Dumble in his Oakville and Lapara forma­tions. The better consolidated white tuff beds of the lower 3Dumble, E. T., "The Cenozoic Deposits of Texas," Jour. Geo!., Vol. 2, No. 6, pp. 549-567, 1894. 4 Dumble, E. T., "The Geology of Southwestern Texas," A. I. M. E. Trans., Vol. 33, pp. 913-987, 1903. Gueydan Formation Gueydan have evidently been mistaken by Dumble for Rey­nosa limestone of Pleistocene age although they are inter­bedded with the strata which he calls Frio. Thus in the last paragraph on page 953 of his paper Dumble states: At the mouth of Comanche Creek, 15 feet of interbedded brown sand" and sar;dy clays were found, but from that point to 0.5 mile south of the falls of the Atascosa everything seen belonged to the Reynosa and overlying beds. On page 978 Dumble describes the following sections under the heading "Reynosa beds" : At a hill half a mile north of the (Weedy or San Cristoval) creek we found the tuffaceous limestone, interbedded with yellow clays and sand, passing into Iavendar clays at base.... (Section given). Following Weedy Creek to its junction with the Atascosa and that river to the "falls," three miles below, the Reynosa was seen to have an extensive development, forming precipitous banks from 20 to 30 feet high along the creek ... The writer has studied these sections and found excellent outcrops of the consolidated, non-calcareous white tuff, con­glomeratic tuff and pumice-pebble conglomerates belonging to the lower part of the Gueydan formation. In 1906 Deussen and Dole" published a paper which is devoted mainly to a discussion of the water supply of part of this region, accompanied by a brief account of the geology of McMullen and La Salle counties. During a period of sixteen years following the preceding articles nothing of importance was published on the region. During this interim two commercial gas fields and one com­mercial oil field were discovered in the southern portion of the area, thus attracting the attention of geologists to that part of the district. Accordingly, in 1922, Sellards6 pub­ 5Deussen, Alex., and Dole, R. B., "Ground Water in La Salle a1~d McMullen Counties, Texas," U. S. Geo!. Surv. W. S. Paper, 375g, pp. 141-177, 1916. 6Sellards, E. H., "Notes on the Oil and Gas Fields of Webb and Zapata Counties," Univ. Texas Bull. No. 2230, pp. 5-29, 1922. Uni1·ersity of Te:i:as B11lletin lished a short bulletin on these oil and gas fields. The stratigraphy is briefly described. Sellards noted outcrops of light-colored volcanic ash containing plant impressions near .Aguilares, and stated that they probably represent the Fayette formation. During the next two years, 1923 and 1924, three papers accompanied by geologic maps appeared. Tro\vbridge' made a reconnaissance report on the geology of the coastal plain of Texas adjacent to the Rio Grande, in which he mentioned prominent beds of volcanic ash both in his Frio and Fayette (Eocene) formations. The writer does not agree with Trowbridge's location of the contact between the Frio and the Fayette, nor with his statement that the Oakville (Upper Miocene) and Lagarto (Lower Pliocene) formations do not outcrop in the vicinity of the Rio Grande. In 1923, Dr. J. L. Wortman of Brownsville, Texas. dis­covereds in the calcareous tuft' or "ashy marl" deposits near Rio Grande City some determinable Upper Miocene fossils, including Protohippus sejunctus Cope, which prove that deposits of Upper Miocene age do outcrop. Moreover, mas­sive beds of characteristic Oakville sandstone outcrop prominently six miles north-northwest of Rio Grande City. Trowbridge apparently included most of the Gueydan tuft' in his Frio formation, but it is probable that some of his Fayette volcanic ash beds also belong in the Gueydan as here defined. The greater part of Jones's paper9 on the Webb and Zapata county oil fields is devoted to the structure of the region. He places considerable emphasis upon a widespread occur­rence of "siliceous knobs, veins and bedded quartzites" as related to faulting and as evidence of it. In the discussion of Jones's paper Dilworth Hager men­ 7Tro\\"bridge, A. C., "A Geological Reconnaissance in the Gulf Coastal Plain of Texas near the Rio Grande," 1J. S. Geo!. Surv. Prof. Paper 131D, pp. 97-98, 1923. SPersonal communication, November 14, 1923. r·Jones, R. A., "The Relation of the Reynosa Escarpment to the Oil and Gas Fields of Webb and Zapata Counties, Texas," Bull. Am. Assoc. Petr. Geo!., Vol. VII, No. 5, pp. 532-545, 1923. Gueydan Formation tioned the presence of volcanic ash beds in the area and sug­gested that the widespread occurrence of opal and other forms of silica was due to the leaching of silica from these beds. Alexander Deussen discussed the paper as follows: In answer to Mr. Hager's statement that the opal may be due to leaching of silica from volcanic ash beds, I wish to state that the volcanic ash beds continue clear across the state to the Sabine River. No opal occurs east of the Colorado River. It would seem that if volcanic ash were responsible for opal such knobs would occur wherever ash beds occur ... [The writer would suggest in this connection that the more arid climate, causing a higher alkalinity of the ground water west of Colorado River, has made the ground water a more efficient solvent for the silica in the tuffs.] Deussen also mentioned "rhyolite" boulders in the "Oak­ville sandstones" in Duval County and asked Jones what he considered to be their origin. Jones replied as follows : ... I think the rhyolite boulders and blocks of vesicular basalt in the vicinity of the Government Well, northwestern Duval County, may have been transported from the west, for instance, Uvalde County. By far the most complete geologic map and report on this area is by Deussen.10 On page 91 he states : In the eastern part of the area under consideration, the Frio clay lies conformably beneath the Catahoula sandstone, but in the western part of the area it lies unconformably beneath the Oakville sandstone. The present writer found that the Catahoula (Oligocene) does not lens out as stated by Deussen but passes by a gradual lithologic change to the southwestward into the Gueydan formation which continues to outcrop nearly to the Rio Grande. The Frio clay as mapped by Deussen includes the Gueydan (Oligocene) and the Frio (Eocene) of the present writer. Much of Deussen's "clay, sandy limestone 10Deussen, Alex., "Geology of the Coastal Plain of Texas \Vest of Brazos River," U. S. Geol. Surv. Prof. Paper 126 (1924). U11il'ersity of Te.ras B11lleti11 and sandy clay" in the Live Oak and McMullen county sec­tions is found by microscopic examination to be mainly tuff or argillaceous tuff. The light-colored platy bed shown above the clay or bentonite in his Plate XXVI is recognized by the present writer, who has studied this outcrop, as being a bed of trachyte or trachyandesite tuff. A section of so-called Frio clay is described by Deussen on page 93 as follows : In the bed of Atascosa Creek at the crossing of the Oakvi!le­Crowther road about eight miles northwest of Oakville, there is exposed 6 inches of green argillaceous sandy limestone, over­lying 7 inches of green, calcareous, sandy clay. The outcrops at this locality have been determined as tuff and argillaceous tuff by the present writer. Near this locality are found good outcrops of pumice-pebble con­glomerate composed largely of rounded glassy pumice pebbles in a matrix of argillaceous tuff. Deus sen makes two short references to volcanic ash in his Frio formation. On page 92 he states: "Fragments of agatized wood, beds of volcanic ash and nodules and masses of white chalcedony may also be seen in places." Also on page 93 : "Extensive beds of volcanic ash are found in the western part of Karnes County, about ten miies west­southwest of Karnes City." In 1924 Dumble11 published another paper which is based largely on field work by Paul L. Applin and Lyman D. Reed and on laboratory work by Mrs. Esther R. Applin and Miss Grace Newman. In this paper Dumble corrects several of references made in his earlier publications. The beds of white tuff which were formerly confused, at least in part, with the Reynosa are placed in the Frio and designated as "clays, sands, and sand-rock" in spite of the facts brought out by the following statement made on pages 434 and 435: 11Dumble, E. T., "A Revision of the Texas Tertiary Section with Special Reference to the O!l \\'ell Geology of the Coast Region," Bull. Arn. A~soc. Petr. Geo!.. Vol. 8, No. 4, pp. 421-H-L 1924. Gueydan Formation Under the microscope the differences of Frio materials from those of underlying beds are strongly marked.... The sand grain.:; vary in fineness but arc usually pla1l'y and angular fragments of volcanic ash. The quartz also appears in splin­tery needle-like fragments and in groups of hair-like filaments scattered through the sandy clays. [The so-called quartz is evidently glassy feldspar in part and in part filamentous vol­canic glass or pumiceous fragments.] In the same paper the Whitsett beds of Dumble include the Frio and the upper part of the Fayette formations of the present writer. The Frio as defined by Dumble is appar­ently identical with the Gueydan formation of the present writer. Dumble considers his Frio (the Gueydan) as Jack­son (Eocene) in age, although he states that only one fossil, an undetermined leaf impression, was found in this forma­tion. This seems inconclusive as will be indicated by later discussion. Name.-Dumble's 1924 paper,'~ in which he redefined his Frio formation, appeared subsequent to the short paper in which the Gueydan formation was named and defined by the writer.1'' The formational name "Gueydan" has been introduced for the following reasons: ( 1) The term "Frio clay" has been used for so many years that "Frio" immediately suggests to students of Texas stratigraphy a formation composed mainly of clay. (2) The Frio is included in the Eocene by previous writers. Less confusion will result if the term "Frio" is retained for the lower portion of the old "Frio formation" as mapped by Deussen, which is composed principally of clays and is presumably of Eocene age. (3) If the pyroclastics and associated rocks which are called in this paper the "Gueydan formation" are to be placed in one of the previously recognized stratigraphic divisions of the Texas coastal plain section, they should be placed in the c2~tahoula formation, their partial or 120p. cit., pp. 424-444. 13Bailey, Thomas L., "Extensive Volcanic Activity in the Middle Tertiary of the South Texas Coastal Plain," Sci. (n. s.), LIX, No. 1526, pp. 229-300, 1924. University of Texas Bulletin complete equivalent. However, the most. characteristic lithologic type in the Catahoula is a quartzose sandstone, whereas the sandstone beds in the Gueydan are volcanic sandstones poor in quartz. PHYSIOGRAPHY GENERAL STATEMENT The region under consideration in this paper is included in the Texas portion of the Gulf Coastal Plain physiographic and geographic province of the United States. Over one­third of the State of Texas belongs in this province. That third or fourth of the coastal plain which borders the Gulf of Mexico is an almost featureless, level plain sloping seaward at the gentle rate of two feet per mile. Farther inland the coastal plain slopes seaward at an average rate of from three to six feet per mile, and is so dissected by the action of streams that in many localities the land surface is quite uneven and hilly. The general geologic structure of the coastal plain is homoclinal and the dip is toward the Gulf at a rate ranging from about 20 feet per mile for the youngest (Reynosa Pleistocene) formation, to 50 and even 120 feet per mile for the older formations (Miocene to Upper Eocene). The rocks underlying the part of the coastal plain which is discussed in this paper vary considerably in their resistance to erosion, both as to entire formations and as to members of the same formation. Rain and stream action has eroded the less resistant strata more deeply than the more resistant beds; the present land surface is consequently divided into a series of low, roughly parallel cuestas with steeper slopes or, in some places, even pronounced erosional scarps on their western or northwestern sides, and gentle, often nearly imperceptible, dip slopes toward the east and southeast. The scarps are of course produced by the outcropping edges of relatively resistant beds or a series of such beds. The ftat dip-slope surface of a cuesta will be called a "cuesta­ plain." Where there is rapid alternation of softer and Gueydan Formation harder beds the cuesta-plains are narrow and not very prominent. On the other hand, where the resistant beds are relatively thick or, if thin, very resistant to erosion, and where they are also separated from each other by a thick series of easily eroded beds, the individual cuesta­plains are much broader and the scarps are more prominent. The most pronounced erosional scarps are produced by thin but very resistant beds, such as siliceous-cemented sand­stones, especially where such beds are underlain by a thick series of unconsolidated soft clays. The condition last stated has been responsible for the development of imposing mesas, such as Loma Alto, La Chusa, and San Caja, in south­eastern McMullen and northwestern Duval counties. On these mesas is a comparatively thin cap of very hard, quartzite-like sandstones underlain by a thick series of unconsolidated, dust-like, tuffaceous clay deposits. These mesas are portions of the Oakville cuesta-plain which have been isolated from the main body of the cuesta by erosion. Erosion of the series of weak strata which separate the parallel cuesta ridges gives rise to flatter areas which may assume the form of valley-like depressions. For these the term "inter-cuesta-plain" is proposed. The continuity of the cuestas is broken in many places by rather broad stream valleys which cut across them trans­versely so that the cuesta topography is not generally evident except in the more or less semi-arid southern fourth of the Texas coastal plain or where a series of very weak strata, such as soft clays, lie next to a series of resistant strata, such as well indurated sandstones or limestones. Thus the general topographic appearance of much of the district is that of a slightly uplifted and submaturely dissected plain or low plateau. Nearly the whole of that part of the region southwest of San Antonio River in Karnes County is covered with a thick growth of various kinds of thorny bush or chaparral. This chaparral-covered district has a semi-arid, warm temper­ate climate and an average yearly rainfall of 25 inches or less. In the vicinity of the Rio Grande the annual rainfall is about 13 inches. North of the San Antonio the climate Uni'.v ersity of Te.ras Bulletin is more humid and portions of the district are sparsely to thickly wooded, but stretches of open prairie are very com­mon in some belts. The rainfall in this more northerly part ranges from 25 to 35 inches a year. RELIEF The relief of the portion of the coastal plain with which we are concerned is nowhere great. The highest point is located in southeastern Webb County, one mile south­southeast of Mirando City, and according to the United States Army topographic surveys has an elevation of 1,004 feet above sea level. The lowest point is on the Nueces RiYer three miles south of Three Rivers. The elevation of this point is 180 feet. The maximum relief for the whole area is thus 824 feet. The average relief of the entire district taken by topographic quadrangles (112 degree lat. and long.) is much less, only about 280 feet. The general difference in elevation between the tops of divides and the bottoms of adjacent larger stream valleys is from 200 to 400 feet for most of the district. Within a strip embracing both the outcrop of the Gueydan formation, as shown on the geologic map (Pl. I.) and two parallel belts, five miles wide, bordering the Gueydan outcrop on either side, some interesting differences in relief are noted. If the strip just defined be subdivided into segments intercepted by parallels of latitude 30 minutes apart, and the highest and lowest points in each segment be determined, the following data11 are obtained: 14These data are taken from the "Progressive Military Maps" based on topographic work by the "Corps of Engineers, U. S. Army, Southern Division." Some of the elevations are only approximate, l'ut those near railroads and rivers are fairly accurate. Gueydan Formation Highest Lowest Latitudes Point Point Relief Between 30 00 and 29 30 575 feet 325 feet 250 feet Between 29 30 and 29 00 525 feet 190 feet 335 feet Between 29 00 and 28 30 510 feet 220 feet 290 feet Between 28 30 and 28 00 640 feet 180 feet 460 feet Between 28 00 and 27 30 850 feet 350 feet 500 feet Between 27 30 and 27 00 1004 feet 400 feet 604 feet Between 27 00 and 26 30 700 feet 300 feet 400 feet Between 26 30 and 26 00 370 feet 150 feet 220 feet The most significant facts brought out by the above list are: (1) There is a gradual decrease in elevation of 500 or 600 feet along the line of "highest points" both toward the northeast and toward the south from the point of maximum elevation. In other words, along the strike of the strata there is a distinct arching of the surface in northwestern Duval, eastern Webb, northeastern Zapata, and northwestern Jim Hogg counties, with the high point on the arch near Mirando City in Webb County. (2) The greatest relief (604 ft.) to be found in any of the segments of the belt is in that segment in which the point of maximum elevation also occurs, in spite of the fact that no large streams are found in this locality at the present time. The lowest point in this segment is also higher than the lowest point in any other segment. (3) The highest elevations in each segment are prac­tically all situated to the southeast or east of the Gueydan outcrop, on the summits of inland-facing escarpments pro­duced by the outcropping lower beds of the Oakville (Upper Miocene) sandstone and of the Reynosa (Lower Pleistocene) limestone. An exception is Tornillo Hill, located on the Karnes-Atascosa county line, eleven miles southwest of Falls City. This knoll of silicified sandstone is situated near the coastward margin of the outcrop of the Fayette (Upper Eocene) sandstone. ( 4) The maximum elevations listed above are, in the southern half of the area, higher than any points situated twenty to fifty miles farther inland, up the general slope of Unfrersitu of Te.ras Bulletin the coastal plain. This fact does not hold for much of the northern half of the district, which is situated northeast of Live Oak County. These facts point to a geologically recent structural dom­ing of the southern portion of the region ·with the center of uplift located near Mirando City. This arching was evidently so late that it is still reflected in the present topography, although the relatively small streams in this portion of the area have been rejuvenated sufficiently to produce the maximum relief near the center of this uplift. Deussen1 0 has called this the Torrecillas arch and has discussed most of the evidence for it. PHYSIOGRAPHIC DIVISIONS Deussen16 has divided this part of the coastal plain into belts which are, beginning "·ith the one nearest the coast, as follows: (1) Reynosa Plain, (2) Oakville Plain, (3) Frio Plain, (4) Wellborn Plain. Each is given the name of the geologic formation on the outcrop of which it is situated, with the exception of the Wellborn Plain, which occupies the belt coincident with the Fayette outcrop. The writer agrees with these divisions except for the "Frio Plain." Deussen uses "Frio Plain" as the geographic division corresponding to the outcrop of his "Frio formation" or the Frio and Gueydan formations of the present writer. The writer would divide Deussen's "Frio Plain" into three physio­graphic belts south of the southwestern border of Karnes County. Beginning at the western or northwestern margin of this region the writer recognizes six physiographic divisions. These are briefly described below. Fayette Hills Belt (Wellbom Plain) .-This belt is from five to fifteen miles wide and is coincident ·with the outcrop of the Fayette formation. Its western margin is marked "·Deussen, Alex., "Geology of the Coastal Plain of Texas West of Brazos River," U. S. Geo!. Surv. Prof. Paper 126, pp. 6, 124 0 nd 126, 1924. IGQp. cit., pp. 8-10. Gueydan Formation by a more or less prominent escarpment which is well devel­oped near Smiley, Gonzales County, at Tilden and at Kings Hills, six miles southwest of Tilden, McMullen County, near Roma on the Rio Grande and at many other localities. This escarpment is not distinct in southern Webb and most of Zapata counties. Here the belt is a rolling plain interrupted along its length by a number of low cuesta ridges. The surface of the Fayette belt is rather maturely dissected by consequent streams transverse to the strike, but the major divides are broad and very gently convex upward, or nearly level in places. The steepest slopes in this, as in all the other belts in the region, occur close to the main streams. Because of the fact that the Fayette is composed of dif­ferentially resistant sandstone and clay or shale beds, a number of parallel large and small cuesta ridges are pro­duced, which vary in number and prominence from place to place because of the lenticular character of the beds. The soil of the Fayette Hills Belt is mainly a light buff or drab sandy loam, sand or loam, becoming finer textured in the southern part of the district where the rocks are more highly argillaceous. Some strips of dark gray clay­loam also occur. In many places, especially at the western margin, rocky sandstone slopes and cliffs are present. In the semi-arid district south of Atascosa County the Fayette Hills are almost everywhere covered with a dense growth of thorny chaparral from five to fifteen feet in height, and of various genera. In many places prickly pear cactus abounds. Some of these shrubs have no leaves, but have chlorophyll-bearing thorns up to three inches in length. This thick covering of thorny brush makes geologic work difficult. In Atascosa and western Karnes counties the vegetation is mostly mesquite, but other kinds of thorny brush are present. Trees, such as cottonwood, elm, and live oak, occur along many of the larger streams. In Karnes County there is a fairly rapid transition from the brush characteristic of the more arid portion of the central coastal plain into comparatively open woods composed mainly of post oak. Along the river bottoms in the northern half of Unh'ersitu of Te.ras Bulletin the district forests of large dicotyledonous trees of several kinds are found. F1·io PZai11.-This division, as considered here, is a some­\Vhat irregular, narrow belt which extends from the south­western boundary of Karnes County to the Rio Grande and ranges in width from a mile or less in much of the northern part, to five miles or more in the southern part of McMullen County. It is coincident in extent with the out­crop of the Frio formation as shown on the geologic map (Pl. I). Although it is a rather narrow belt, it is never­theless one of the most distinctive physiographic divisions of the area, because the Frio is more easily attacked by erosion than any other of the formations. This belt is a typical, nearly flat, rolling inter-cuesta-plain. In general its surface lies from 100 to 300 feet lower than that of the adjacent physiographic belts, giving the impression of a broad-bottomed valley. The Nueces River has taken advan­tage of the weakness of the Frio beds and has established its present course along the outcrop of these soft clays across McMullen County, a distance of more than thirty miles. In this portion of its course the Nueces becomes a subsequent stream. Even in portions of its extent not occupied by stream valleys the Frio Plain may exhibit a valley-like form. This feature can be noted on the Tilden-Three Rivers road from bvo to four miles east of the McMullen-Live Oak county line. Here is a lo\Y stretch containing, in the vicinity of the Frio River, pools of stagnant ·water, around which are found numbers of both dead and living mesquite trees. Since the trend of this low strip is almost transverse to the course of the Frio River, it can hardly be a second bottom of the Frio Ri\·er. A much larger, although similar northward­trending topographic depression having a width of two miles or more, is located in northeastern Webb County ten or twelve mi!es northwest and north-northwest of Moglia and three miles \vest of the Gates ranch house. A few small \vet-"·eather streams empty into the depression and there is a small i~termittent stream, Prieto Creek, ft;wing northward along its bottom and finally draining into the Nueces River. Notwithstanding the ease with which the Gueydan Formation Frio formation is eroded, this depression is much too large a feature to have been produced by Prieto Creek. More­over, the surface of the depression is covered by a thin veneer of gravel containing many pebbles as much as four inches in diameter which could not have been transported by Prieto Creek with its present volume and gradient. This depression is believed to have been a former channel of the Nueces River at a time when it probably emptied into Baf­fins Bay in Kleberg County. The other evidence in support of this conclusion will be discussed under "Drainage." In Zapata County this belt, like the Fayette Hills Belt, is not very distinct topographically on account of the more argil­laceous, and therefore similar, character of the adjoining formations. The soil of the Frio Plain is a dark gray or black, sticky clay which cracks along numerous small intersecting joints. Some of the low-lying portions are slightly saline and sup­port a growth of salt-loving sedges. Patches of sandy loam soil occur in places. The vegetation is predominantly thorny chaparral including much mesquite. In Live Oak County and in parts of McMullen County open grassy prairies are found, though stretches of bad-lands are developed on a small scale. Gueydan Hills Belt.-This physiographic division is pro­duced by the outcrop of the lower and middle beds of the Gueydan formation, namely, the Fant and Soledad members of that formation. The approximate outcrops of these members are shown on the geologic map. This belt extends from western Karnes County to northern Zapata County and ranges from an average width of three miles in Live Oak and eastern McMullen counties to five or six miles in southern McMullen and northern Duval counties. South of Duval County it gradually narrows and is only about two miles wide at the south line of Webb County. The western margin of the belt is marked by a faint to pronounced escarp­ment which is produced by the outcropping edges of strata of consolidated white tuff considerably more resistant to University of Te~i;as Bulletin erosion than the underlying Frio clays. The escarpment is very conspicuous for a distance of thirty-five miles in the neighborhood of the Nueces River in McMullen County, for here the river has carved out a fairly deep strike valley just northwest of the outcrop of the hard tuff beds. In places the scarp is marked by a row of picturesque, glaring white cliffs of tuff as much as fifty feet high. The best example of such cliffs is found at "Chalk Bluffs" on the Ray ranch, five and one-half miles south of Wentz in east­central McMullen County (Pl. III, Fig. 1). North of Duval County the white tuff scarps at the western margin of the Gueydan Hills Belt are the most prominent of all the cuesta scarps in this belt. But there are a number of other cuesta ridges showing westward-facing scarps developed east of the white tuff scarp, especially southwest of Live Oak County where the middle Gueydan consists of alternating beds of differential hardness. Most of the latter cuesta ridges are topographically less important than the nearly flat-topped ridges and narrow valleys produced by conse­quent or obsequent streams tributary to the Nueces and flowing at right angles to the strike of the cuesta ridges. The general impression one forms of this belt, especially when he looks from the eastern margin up the gentle slopes of the cuestas lying toward the west, is that of a low plateau. In southern Webb and northern Zapata counties the Guey­dan Hills become very narrow and are almost completely wedged out by the Oakville and Reynosa escarpments. In most of Zapata County this physiographic division loses its identity and practically merges into a rolling plain formed by the outcrop of several formations, all of which are pre­dominantly clays. This hilly belt is also only locally recognizable northeast of western Karnes County, because the lower Gueydan is composed of less indurated tuff and because there is no strip of Frio between it and the Fayette Hills Belt. Also, the whole Gueydan and the Catahoula into which it grades north of Karnes County have so narrow an outcrop in this part of the district that they do not have a marked effect upon the topography. Gueydan Formation In northwestern Duval County are found a number of very striking, flat-topped hills which, if viewed from the west, resemble mesas. They occur near the eastern border and in the center of the Gueydan Hills Belt. The most impressive of these are the Soledad Hills on which the large new Soledad ranchhouse is located and which will serve as an illustration for a number of similar hills in this belt. The Soledad Hills are about two miles long by one-half mile broad and are elongated-elliptical in plan. Their long axis extends north-northwest, making an angle of approximately 60 degrees with the strike of the strata. On all sides except the east and southeast the hills rise steeply from 50 to 125 feet above the level of the surrounding rolling plains. The steepest slopes as well as the highest portions of these hills are located on the western side. This is due to the fact that they are produced by the outcrop of very resistant, thick, volcanic conglomerate beds which dip to.vard the southeast at the rate of approximately 100 feet to the mile. Because of the gentle dip of these conglomerates the surface of the hills appears nearly flat, except where small intermittent streams have made reentrants into them. Some of these streams have in fact cut through part of the hard con­glomerate and have partially dissected the mass of the hills, dividing the more elevated part into three mesa-like pro­jections. The dip of the beds of which the hills are com­posed produces a cuesta slope on the eastern side so gradually that the surface of the hills practically merges into the lower plains area to the east and southeast. How­ever, this cuesta slope has been somewhat trenched at its northern end by Soledad Creek, a stream flowing into the Nueces. Thus the hills have a mesa-like appearance on all sides except the southeast, furnishing the most distinctive land mark in Duval County. Similar but somewhat less prominent hills are fairly numerous in the western corner of Duval County, where this conglomerate or other beds of conglomerate are well developed. Small rounded hills are also produced by a more advanced stage in the erosion of such hills. Unil:ersity of Te.ms Bulletin The soil produced by the weathering of the white tuff is a pale gray loam which is porous and light in weight. The beds of clay, which are commonly interbedded with volcanic tuff, give rise to a dark clay soil much like that of the Frio. The sandstones and conglomerate of the middle Gueydan produce gravelly or sandy loam soils. No stretches of deep sandy soil were observed in this belt. The Gueydan Hills are in general thickly covered with thorny shrubs much like those ·which occur in the southern part of the Fayette Hills. On the higher hills near the Reynosa escarpment in southern Webb County small patches of grease"·ood, typical of the more arid regions of Texas, are found. Much mesquite is present, especially along the courses of intermittent streams where trees of it thirty feet high are found. Gueydan Plain.-This strip of rolling to nearly flat plains country coincides, for the most part, with the outcrop of the upper Gueydan very friable, loess-like tuff and tuffaceous clay. Like the Frio Plain, this is an inter-cuesta-plain, but it is much broader than the Frio in most of the region. The Gueydan Plain is a distinct physiographic feature all the way from southern Gonzales County to Moglia in eastern Webb County. In the neighborhood of Moglia it narrows abruptly and ends at the foot of the Oakville and Reynosa escarpments, because the younger Oakville and Reynosa formations completely overlap the upper Gueydan. It ranges in \Yidth from less than a mile in Lavaca and Gon­zales counties to five or six miles in most of Live Oak, Mc­Mullen, and Duval counties. Northeast of southeastern Atascosa County practically the entire outcrop of the Guey­dan formation is included in the Gueydan Plain, on account of the fact that only a few thin beds of resistant rock occur, even in the lower Gueydan, in this portion of the district. A number of the tributary streams which flow across the Gueydan Plain have developed vertical-sided, flat-bottomed, inner valleys or channels very much like those developed by the streams flowing across loess plains. The tendency to form such valleys is evidently due to the well-developed vertical jointing in the tuffaceous clays of the upper Guey­ Gueydan Formation dan similar to that found in loess. Plate VIII, Figure 1, shows such a valley produced by White Creek in west-central Live Oak County. The soil of the Gueydan Plain is a light gray to pale buff, rather porous loam or clay-loam. It supports through most of the area a chaparral growth similar to that found in the Gueydan Hills. In that portion northeast of Live Oak County the soil is generally black and rather fertile. Many open, prairie-like stretches occur, but some parts are cov­ered with a scattered growth of small oak, mesquite, and other small to medium-sized trees. Along the major streams north of Duval County are fringes of timbered country. Mesas.-Interrupting the monotony of the Gueydan Plain in southeastern McMullen and northwestern Duval counties are eight abrupt mesas or groups of mesas rising 100 to 160 feet above the adjacent plain. The largest of these, Loma Alto Mesa, is the highest point in McMullen County. San Gaja is the most northerly of the mesas. The mesas owe their existence to the presence of a cap of siliceous-cemented, very hard Oakville sandstone and conglomeratic sandstone. This cap rock, although usually only from two to twenty feet thick, is so much more resistant to erosion than the underlying pulverulent tuff and clay that a surprisingly large amount of the latter has been removed, compared with the volume of the cap rock eroded during an equal period of time. The cap rock is nearly flat-lying, with the exception of San Caja where a definite tilt of 5 degrees or more toward the southwest is evident even in a photograph. The flat tops of most of these mesas have given rise to such fanciful names as "Devils Center Table," which is applied to the central mesa of the La Chusa group. A photograph of La Chusa is found in Plate IV of Deussen's paper. The slopes of all these mesas are practically covered with loose slabs of the cap rock left thus by removal of the soft under­lying material. Overhanging cliffs are quite common at the base of the sandstone cap. With the exception of Loma Alto, all of these mesas are situated on the eastern portion of the Gueydan Plain, as is to be expected since they are salients of the Oakville Uni1.'ersity of Te:ras Bulletin escarpment which have been separated by stream erosion from the main mass of the sandstone escarpment which forms the eastern boundary of the Gueydan Plain. In some places long, peninsula-like salients are in the process of being segmented into island-like mesas by headward erosion of streams into their sides. Loma Alto is situated fully two miles back of the main Oakville escarpment and owes its greater height to the fact that its cap rock is located higher up the regional dip of the basal Oakville. Chalcedony Knobs.-ln addition to the flat-topped mesas several much smaller conical or rounded knob-like hills are found in McMullen, Duval, and eastern Webb counties on the Gueydan Plain and in the western edge of the Oakville Hills Belt. Some of these knobs when seen from a distance are suggestive of small volcanic necks. They are only from 30 to 75 feet higher than the country immediately surround­ing them, and 100 to 300 yards in diameter, but they furnish striking land-marks on account of their peculiar form and steep sides. These knobs look as though perched upon low circular eminences which have slopes that are slightly con­cave in profile. Such knobs owe their existence to the fact that they are composed of massive opal and chalcedony, or very hard quartzite-like sandstone which in places shows a good desert varnish. The principal knobs of this type are Tendita, Paint Hill, and the Seven Sisters, all of which are situated near the south line of McMullen County, from three to six miles east of Loma Alto Mesa, and Sernosa Hills, located about three miles southeast of the Soledad Hills (south end) in Duval county. The Picachos Hills, in northern Duval County five miles southwest of Loma Alto, consist of three elliptical ridge­like elevations from forty to sixty feet high and a few much smaller and lower ridges trending North 42° East. The individual ridge-like hills do not form a continuous line but are arranged en echelon, with the general trend of the group likewise in a northeast direction. They owe their elevation to the resistance to erosion of steeply-dipping veins of chalcedony and opal. Gueydan Formation Oakville Hills Belt.-Situated immediately southeast and east of the Gueydan Plain is a wide strip of rolling hilly country, of which the western portion has a general eleva­tion of from 50 to 150 feet above that of the Gueydan Plain. The elevation of the Oakville Hills Belt gradually decreases toward the southeast or east following the regional slope of the coastal plain. Probably the boldest escarpment in the region north of Duval County is that which separates the Oakville Belt from the Gueydan Plain. This is known as the Oakville escarpment, because it is capped by the Oakville sandstone. The sandstone has a much greater resistance to erosion than the underlying Gueydan tuff. The prominent mesas, such as Loma Alto, are portions of this scarp which have been disconnected from the main body by erosion. The Oakville Hills Belt is coextensive with the outcrop of the Oakville formation, northeast of McMullen County, and has an average width of seven miles. In this county it narrows rapidly and its breadth averages only two or three miles in McMullen and Duval counties. In eastern Webb County the overlying Reynosa limestone so nearly completely overlaps the Oakville that the escarpment formed by the western edge of the Reynosa merges into the Oak­ville scarp. The Oakville Belt is thus practically absent as a separate physiographic unit in Webb and in most of Zapata and Jim Hogg counties, but emerges from beneath the Reynosa again in northwestern Starr County. In southern and central Starr County it attains a breadth of from three to five miles. The most dissected part of the Oakville Hills Belt is near its western margin, while the surface near the eastern margin is gently undulating. The soil of this belt is usually a deep mantle of light buff to light gray, medium-grained sand. It supports a thick growth of thorny brush like that found on most of the other physiographic divisions in this part of the district. In Live Oak and Karnes counties many patches of post oak, live oak, and other hardwoods alternate with the brush. In eastern Karnes and the countries farther northeast the belt is gen­erally a treeless, hilly prairie. Unfrersity of Te.ras Bulletin Reynosa Cuesta Plain.-The Reynosa Plain is one of the largest physiographic divisions of the coastal plain of Texas, attaining a width of forty miles in the vicinity of Hebbron­ville, Jim Hogg County, and extending from Guadalupe Rh-er to the Rio Grande. The numerous large outliers of Reynosa limestone capping many of the higher hills between its present western border and the Balcones escarpment suggest that this plain ·was formerly much more extensive than it is at present. The time of its greatest extent was not later than the Pleistocene Epoch, for the Reynosa lime­stone is of lo\ver Pleistocene age. Since forming it has been \Yarped considerably as shown by a discordance in the elevation of its surface of o\-er 500 feet between a point on its \vestern edge near Mirando City, Webb County, and a similarly situated point near Green, in southern Karnes County. Deussen,17 in discussing this warping of the Tor­recillas Plain, the higher of the two principal terraces into which he dh-ides the Reynosa Plain, calls the upwarped part of this plain the "Torrecillas Uplift," as has been previously mentioned. The western border of the Reynosa Plain comes into the area under discussion only in that part \vhich is situated south of Duval County, but the ·writer has crossed the entire width of the plain at se,·eral places. Its \vestern margin is marked by a very abrupt erosional scarp. Streams have cut reentrants into the scarp at numerous localities, but in other places, notably in eastern ·webb County, the scarp forms a nearly continuous line of cliffs from 20 to more than 100 feet high. Some of the salients from this scarp have been disconnected from the main Renoysa Plain, form­ing mesas similar to those capped by Oakville in McMullen County. These Reynosa-capped mesas are practically all situated less than one-fourth mile from the main Reynosa scarp and are generally much smaller and rarely so prominent as those capped by the Oakville. 17 Deussen, Alex., l.". S. Geol. Surv. Prof. Paper 126, pp. 5, 6, 124 and 126, 1924. Gueydan Formation On account of its cuesta character the surface of the Rey­nosa Plain is progressively more dissected toward the west, being quite rugged close to the escarpment parts of it. The broad, flat-topped interfluves and narrow, steep-sided valleys indicate that the youthful stage in the cycle of erosion has not been passed. There is thus a greater area of nearly flat surface than of distinctly uneven surface in this belt, so that the term "plain" is thought to be the most appropriate designation for this physiographic division. In portions of the Reynosa Plain the soil is thin and the country is rocky with many irregular patches of white lime­stone and limy conglomerate outcropping. Other parts are w;ell covered with a thick mantle of red clay or red sandy loam soil and locally by deep yellowish sand or coarse gravel. South of the Nueces, the more dissected westerly part of this belt is usually covered with a close growth of chaparral and cactus, forming in many places impenetrable thickets. The rest of the plain is an open grassy prairie country with occasional clumps of mesquite dotting its surface here and there. DRAINAGE The principal river systems, beginning at the northeast, are Guadalupe, San Antonio, Nueces, and Rio Grande. Only Nueces River receives important tributaries in this region. With the exception of the well-known Rio Grande, all of these streams head in the Edwards Plateau from 190 to 250 miles from the coast. Except for the Nueces and its con­fluent, the Frio, these rivers flow across the district in a definite southeasterly direction and are normal consequent streams. These rivers and some of the larger creeks have developed silt-covered flood plains from one-quarter mile to more than two miles wide. In addition to the flood plain and the high Uvalde terrace, which represents the inland extension of the surface of the Reynosa Plain mentioned above, one to three silt or gravel-covered terraces are usually present on one side or the other of the important streams. These concordant terraces represent successive stages in University of Texas Bulletin the uplift of the coastal plain. Since Deussen18 has treated these terraces very thoroughly, they are merely mentioned by the present writer. Probable capture or diversion of the Nueces River.-The courses of the Nueces and its confluent, the Frio, have cer­tain distinctive features which have never been fully ex­plained. Northwest of southeastern La Salle County the Nueces flows southeastward, practically paralleling the courses of the other major streams of the coastal plain. From southeastern La Salle County to Oakville, Live Oak County, a distance of fifty-six miles, measured in a straight line, the river flows northeast, almost at right angles to its upper course. Near Oakville it joins the combined Atascosa and Frio, makes a sharp, right-angled turn to the southeast and continues in this direction to the Gulf of Mexico at Corpus Christi Bay. A few miles farther north the Frio practically repeats the course of the Nueces except that the bends are not quite so pronounced, and it is evident that the same agencies caused the change in the course of both rivers. Deussen19 explains the abnormal courses of these rivers as follows: The map shows that no streams cross the Torrecillas uplift in its highest part. The upper course of Nueces River heads directly for it, but instead of crossing the uplift the stream has apparently been deflected by it more than 50 miles to the north­east. Frio River shows a similar though not so pronounced deflection before it joins the Xueces near Oakville. The minor drainage lines on the uplift also seem to have been affected by the uplift, for they radiate from it in all directions. This ar­rangement suggests that the uplift not only diverted Nueces and Frio riYers but established a new system of consequent streams upon its slopes. Deussen uses the peculiar bends in the Nueces and Frio rivers in support of the efficacy of his Torrecillas uplift but gives no definite evidence that these rivers were actually diverted from former channels, nor does he by any means 1 8 0p. cit., pp. 13, 14, 15 and 114-119. IDQp. cit., p, 126. Gueydan Formation rule out the alternative that the changes in the courses of these rivers were caused by stream piracy. The writer would submit the following data in support of the existence of a former channel for the Nueces : (1) In northeastern Webb County, southeast of the upper large bend in the Nueces is found the very wide, valley­like depression described on page 23 which may be traced for ten miles or more in a north-south direction. The floor of the depression is nearly covered with loose chert gravel of the same type as that found on the gravel terraces of the Nueces in McMullen County, and is apparently incapable of being distributed over such a wide belt by the small inter­mittent stream, Prieto Creek, which now occupies the center of this depression. The gravel does not furnish such satis­factory evidence as do the size and form of this depression, for the gravel may be residual from a former capping of Reynosa, which commonly carries abundant chert pebbles. (2) The Reynosa escarpment makes a very pronounced eastward bend two miles north of Moglia, near the east line of Webb County, and there is a very large southeastward-extending reentrant in this scarp in western Duval County, six miles northeast of Moglia at Parilla Creek. This reentrant is suggestive of a water-gap pro­duced by a large stream. If the upper course of the Nueces were projected it would pass through this gap and to the Gulf via Baffins Bay. (3) The elevation of the Nueces River in the south­eastern corner of La Salle County where it makes the bend to the northeast is only 320 feet. That of the Rio Grande at Laredo, which is the point on the latter river nearest the bend in the Nueces, is about 420 feet. Also the elevation of the highest points on the present surface is 500 feet greater in eastern Webb County than in Karnes County east of the Nueces, indicating a general lowering of the land surface toward the northeast so that the river located farther to the east has the advantage. ( 4) Southwest of Karnes County the Reynosa and Oak­ville escarpments are distinctly higher than at any points in the territory situated twenty to fifty miles farther inland, University of Te;rns Bulletin although this is not true for Karnes County and the country northeast along this belt. These escarpments thus consti­tute a barrier which is difficult for rivers to cut across. (5) Most of the large streams in the Texas coastal plain empty into irregular shallow bays which have apparently been produced by a recent slight depression of the coast or by scour of these rivers accompanying a recent uplift. Such a bay is Baffins which does not now receive any large stream. If the Torrecillas uplift is responsible for the change in the course of the Nueces, a reasonable assumption, the change necessarily occurred after the deposition of the Reynosa, which is involved in the arching. Therefore this change must have taken place in Pleistocene or Recent time, as the lower Reynosa is Pleistocene in age. Since this is recent enough for portions of the former channels of the river to be preserved, the valley-like depression mentioned above could be an old channel of the Nueces. This change in course may have been produced (1) by river piracy, (2) by diYersion, or (3) by a combination of the two. Deussen, in the preceding quotation, states that the Nueces and Frio were "deflected." This is interpreted to mean that these rivers left their channels of their own accord, because they were unable to cut down as fast as the uplift proceeded, although he gives no evidence for the exclusion of piracy. Deussen states further that "the ver­tical movement at these places (Torrecillas and Ojuelos) was probably not less than 200 feet." An uplift of this magnitude is quite insufficient to divert the Nueces and Frio rivers at their present degree of intrenchment, unless it were exceedingly rapid or \Vere associated with a depressive movement in the region of the present subsequent portions of these rivers. That the whole district has been recently uplifted to some extent is shown by the accordant terraces found along these rivers at present. From the evidence at hand the writer is unable to rule out the hypothesis that the rivers were actually diverted. It seems just as logical, if not more so, for the major por­tions of these streams to have been captured by the head­ Gueydan Formation ward erosion of subsequent tributaries of the present Atascosa along a belt of weak strata occupied by the present Frio Plain. The surface of the main divides in the district has already been shown to have a definite northeastward plunge. The Atascosa would thus have been in a favorable position to effect such a capture. The remaining alternative isi that such a subsequent tributary to the Atascosa did exist before these rivers changed their courses but did not actually tap the rivers. Instead, the rivers at flood time may have suddenly found a more favorable channel to the northeast than that across the Torrecillas arch and, coming to the subsequent valleys formed by tributaries to the Atascosa, followed these val­leys to their junction with the Atascosa. Even the hypothe­sis just stated would not hold unless the rivers were very low grade streams and much less intrenched than at present. Unfortunately, no conclusive evidence as to which of the three hypotheses is correct can be gained by a study of the comparative width of the Nueces Valley both above and below the point where it makes the northwestwardly swing. This inability to reach a definite conclusion is due to the fact that throughout most of the subsequent stretch the val­ley is situated in the very soft and easily removed Frio clay. In western McMullen County where the river cuts obliquely across the harder lower Gueydan tuff beds several steep bluffs are present; the inner valley is somewhat narrower than the portion just above the subsequent stretch. There are rapids in the river here, but all these features may be simply due to the greater resistance to erosion of the tuft's, which would prevent the river from widening its new valley so quickly here as it would higher up the course. STRATIGRAPHY The geologic formations exposed in the coastal plain of Texas consist principally of marine and fresh water sand­stones, clay, limestones, conglomerates, and mixtures of these types ranging in age from lower Cretaceous (Co­manchean) to Recent. Beds of pyroclastics-tuffs, volcanic University of Te:ras Bulletin conglomerates, bentonites, and tuffaceous sands and clays­are present both in the Cretaceous and the Tertiary. Near the Balcones fault zone, at the northwestern margin of the coastal plain, ten or more volcanic necks and small dikes, sills and laccoliths are found. Most of the intrusives in the coastal plain are situated in Uvalde County from sixty to seventy-five miles west of San Antonio. Duessen~0 has given a table of the formations of the southwestern coastal plain showing their general lithology and thickness; the writer will merely list the formations in order, for the reader's convenience. This list is in agree­ment with Deussen's division except that the "Frio clay" of Deussen is separated into two formations, the Frio (corre­sponding to the lower part of the Frio clay of Deussen) and the Gueydan tuff. The Lapara sand is included in the Lagarto. The sequence and general lithology of the Tertiary formations of the southwestern coastal plain of Texas are as follows: 15. Alluvium and wind-blown sand .. Recent 14. Beaumont clay. f Disconf?r~ity? Upper to Lower 13. L1ss1e gravel. Pleistocene Disconformity? ( Fluviatile) 12. Reynosa limy conglomerate and limestone I Unconformity. J 11. Lagarto clay, sandstone and sand. l Lower Piiocene­Unconformity? UJ?per Miocene 10. Oakville sandstone. r (Fluviatile) Unconformity. 9. Gueydan luff, volca?i~ conglomerate, sand-1 Oligocene or (and} stone and bentomt1c clay. (Equivalent. Lower Miocene ? at least in part, to Catahoula sand-r (Continental pyro­stone.) elastic and fiuvia­ tile) 8. Frio clay. I Jackson } Upper Eocene 7. Fayette sandstone, ( (Marine,palustrine clay and shale. J Group _ and flu viatile) 2oop. cit., pp. 20-23. Gueydan Formation 6. Yegua clay. 5. Cook Mountain sand­stone. 4. Mt. Selman clay and sandstone. Unconformity. I Claiborne Group (Marine and some fluviatile) Middle Eocene I r-(Marine and 3. Carrizo sand and sand-1 stone. continental)Wilcox GroupDisconformity. (Mainly2. Indio sandstone, shaleJ fluviatile)and lignite. Unconformity. j 1. Midway clay. 1 Lower Eocene (Marine) Unconformity. Upper Cretaceous clay. Of the above formations only the Fayette, Frio, Gueydan, Oakville, and Reynosa outcrop in the portion of the coastal plain with which this paper deals. These formations have a regional dip to the southeast ranging from 80 to 120 feet per mile in the oldest or lowest and about 20 feet per mile in the youngest or highest. The truncation of their dipping strata by erosion causes the formations to outcrop in belts which roughly parallel the present shore line of the Gulf of Mexico. The Gueydan is the formation with which this paper is particularly concerned and the formations associated with the Gueydan will be described mainly with the purpose of bringing out their lithologic dissimilarity to the Gueydan and their stratigraphic relations to it. OTHER FORMATIONS OUTCROPPING IN THE VICINITY OF THE GUEYDAN Underlying Formations (Jackson Group) FAYETTE FORMATION Distribution and stratigraphic relations.-The Fayette outcrops in a belt which averages 4 miles in breadth in Gonzales County, gradually increases in width to 7 miles Unive1·sity of Te:ras Bulletin in Karnes and Atascosa counties and to 8 or 10 miles or more in parts of McMullen and northeastern Webb counties. In southern Webb County it narrows rather abruptly to 5 miles or less near Aguilares, widens to 10 miles or more in northern Zapata County and narrows again to 7 or 8 miles near the Rio Grande in western Starr County. Southwest of the Atascosa-Karnes county line the Fayette outcrop is separated from that of the Gueydan by a rather narrow strip of Frio clay but northeast of that line no Frio clay can be recognized; here the Gueydan is found to rest directly on the Fayette. The contact between the two latter is apparently conformable or slightly dis­conformable. The term "Fayette" is used by the writer with the same significance as Deussen uses it in Professional Paper 126. Lithologic character.-There is considerable difference between the prevailing lithologic character of the Fayette in northeastern Karnes County and in Gonzales County and its character in the portion of the area farther southwest discussed in this paper. In the more northerly district much of the Fayette consists of very light grayish-cream to white, non-calcareous, platy-bedded, sandy to silty tuffaceous shales or very fine-grained shaly sandstones, which very commonly carry leaf impressions. This rock is generally more or less cemented ·with opal although some of the more friable samples of it break down readily into a doughy mass. Under the microscope the argillaceous substance which constitutes most of the shale is found to have a mean index of refraction of about 1.51, an index far too low for kaolin. It is identified as montmorillonite, the principal mineral of bentonite, which is commonly formed by the alteration of volcanic glass. The light weight of this rock suggests a tuff although only a few angular, much altered grains which suggest volcanic glass are noted. No fresh glass grains, which are abundant in the Gueydan, were seen in four typical fresh samples taken from as many localities in Atascosa, Karnes, and Gonzales counties. When a me­chanical separation of the samples is made, grains coarser than 1/ 16 millimeter in diameter are found to comprise less Gueydan Forrnatfon than 1 per cent to 10 per cent of the entire sample. Most of these grains are between 1/8 and 1/ 16 millimeter, and very few grains coarser than 14 millimeter are present. The sort­ing is considerably better than rocks of somewhat similar appearance found in the Gueydan formation. A large num­ber of sub-rounded grains of various minerals including nu­merous chert grains are noted, indicating that the material of which the tuffaceous shale is composed has been reworked or mixed with non-volcanic material, probably by water, and therefore probably is not a true volcanic tuff or dust. The mineral composition of fine sand grains found in this rock with their estimated percentages, taken from a typical sample collected on Dewee's ranch, seven miles southwest of Falls City, is as follows: Plagioclase (mostly oligoclase and andesine)-32% ; opaline crusts (from cement)-20%; orthoclase and sanidine-10% ; quartz-25'/o ; chert-10% ; chalcedony-2% ; augite-trace; hypt:rsthene-trace; brown hornblend-rare; green hornblend-rare; magnetite-trace; tourmaline (brown and blue varieties)-rare; apatite-tract:; altered volcanic glass(?)-trace; chlorite-trace; limonite-trace. Most of the grains of heavy minerals are rounded. Casts of small Corbula-like pelecypods % inch in length are occasionally found in the dark-colored beds o.f this shale. Similar light-colored tuffaceous shales and, also, extremely friable, flaky and thinly laminated, light buff to grayish­pink, tuffaceous shales carrying a great abundance of leaf impressions are found in several horizons in the Fayette farther southwest, in western Karnes and in Atascosa, McMullen and other counties. These beds also contain a number of worn grains and are evidently reworked volcanic dust deposits. Beds of "volcanic ash" have been reported from the Fayette by several writers as mentioned pre­viously, but such beds are generally thin and do not comprise much of the formation. In the wells west of Calliham a bed of surprisingly hard, white, kaolin-like bentonite is encountered near 600 feet and practically at the base of the Fayette. This bed is very hard to drill through on account of the swelling of this substance. Universitu of Te:ras Bulletin In the upper part of the Fayette (or possibly lower Guey­dan with a Catahoula facies) in Gonzales County are found prominent beds of light gray or brownish-gray, medium­grained, indurated, quartzose sandstone, the weathered sur­faces of which show abundant small criss-cross, vein-like ridges approximately an inch in width. On breaking this rock the vein-like masses are found to extend through the rock and to be composed of harder sandstone of a nearly pure white color, while the areas between them are more friable and more or less stained with iron oxide. The vein­like masses are evidently due to the leaching and cementing action of silica-bearing solutions migrating along the joint cracks in the sandstone. Since the rest of the rock does not possess so much cement it is less resistant to weathering and erosion. The white color is due in part to the sinter-like opaline cement. Some horizons of fine-grained, light buff, argillaceous sandstone and cavernous silty to sandy clay or shale carry a fair abundance of marine pelecypods and gastropods of Eocene age. The cavernous spaces in this sandstone are due to the solution of numerous fossils. Such a bed out­crops four miles southwest of Sample, Gonzales County. These fossils together with the numerous plant remains from the tuffaceous beds have fairly well established the Jackson age of the Fayette.21 In La Salle, McMullen, Atascosa, and western Karnes counties the Fayette is developed in its more typical facies and consists largely of sandstones. The most common lithologic type, as especially well illustrated in numerous well sections in Live Oak and McMullen counties, is a light gray, friable, massive to platy-bedded, or laminated, non­·calcareous to slightly calcareous, fine-grained, silty, arkosic, sandstone. Cross-bedding is very common, though it is not .always present. On weathering most of the sandstone be­'comes light buff in color. In the majority of the sandstone beds fossils are rare but several horizons carry an abundance 21Deussen, Alex., U. S. Geol. Surv. Prof. Paper 126, pp. 82-84, 1924. Gueydan Formation of generally poorly preserved, chalky-looking pelecypod shells, including a large number of Tellina eburniopsis? Conrad. Lignitic streaks or partings and thin beds of olive­green clay are also common in many of the sandstones. The microscopic analysis of a core sample of light gray, friable, imperfectly laminated, slightly calcareous, very fine­grained silty sandstone from a depth of 648 feet in the Black Panther Oil Company's Nicols No. 1 well (located in northern Live Oak County just west of the Atascosa River, five miles north of Three Rivers) will illustrate the physical and mineralogical character of much of the Fayette sandstone. The description is as follows: Sorting.-Above % mm. (only pyrite concretions)-1%; %-14 mm.-3%; 14-Vs mm.-30%; Vs-1/16 mm.-36%; silt annes on account of the great quantity of glassy plagioclase feldspar and worn fragments of more Unicersity of Te:ras Bulletin or less altered andesite. It might be conceived that large deposits of volcanic debris were extant in Fayette time a short distance to the northwest of its present outcrop, but the Yegua formation which underlies the Fayette in appar­ent conformity contains little volcanic tuft' or ash, while in the formations outcropping still farther inland volcanic debris is even less common. In order to explain the large amount of plagioclase feldspar in the sandstones which comprise most of the Fayette formation, an enormous vol­ume of igneous, possibly volcanic, debris must have been removed. It is impossible to tell, without a petrographic study of the older sedimentary formations outcropping farther inland just how much of this feldspar might have been derived from older sediments outcropping in the coastal plain of Texas. The chert, tourmaline, probably much of the quartz, the secondary foraminifera, and some of the other components of the Fayette show that the volcanic debris was mixed with a large amount of non-volcanic detritus, probably derived from the erosion of Cretaceous and Lower Eocene sedimentary rocks which were outcrop­ping in the Texas coastal plain to the west and north. FRIO FORMATION Distribution.-As was mentioned before, the name "Frio formation" is used here with a different significance from its usage by any previous writer, so far as known. The Frio formation here designates those predominantly argil­laceous strata which lie conformably or disconformably beneath the light-colored or white tuft's of the Gueydan formation and conformably on the Fayette formation, which consists largely of sandstones in most of the area under discussion. The writer has followed the outcrop ot the Frio along a belt paralleling the Fayette outcrop from a point in extreme southeastern Atascosa County near the Karnes county line, to the Rio Grande. In Atascosa and Live Oak counties the belt is quite variable in width because it is partly or almost completely overlapped by the Gueydan formation. Its width in these two counties and in eastern Gueydan Formation McMullen County ranges from less than a mile to four miles, as shown on the geologic map. The Frio-Fayette contact south of central McMullen County is only approximately located on the map. This lower contact is very difficult to follow accurately on account of the dearth of outcrops. However, the Gueydan-Frio contact, or upper contact, is believed to be fairly accurate, because the relatively hard beds of the basal Gueydan are apt to outcrop and are there­fore more easily traced, except in Zapata County where few outcrops of any kind are seen, save along the Oakville­Reynosa escarpment. The Frio formation fails to outcrop northeast of Atascosa County although it is quite typical in appearance up to the point of its disappearance. North­east of this point the Gueydan rests on the Fayette so that it is quite unlikely that the Frio beds pass into lithologically dissimilar beds along their strike. Their disappearance is possibly due in part to the thinning of the Frio toward the northeast but they are more probably overlapped by the Gueydan, because some unfossiliferous, pink and green clays, lithologically similar to the Frio, were encountered in the Lavaca County Oil Company's Laas No. 1 well in eastern Lavaca County between 2,070 and 2,392 feet and in wells in Karnes County. The log of the Laas well is found in a publication by the present writer.22 Lithologic character.-The Frio consists almost entirely of very soft and plastic, gypsiferous, creamy-green to light grayish-green and purplish-pink clays or variegated pink and green clays. The clay is generally unstratified and breaks up into minute angular fragments along the closely spaced joint cracks so that good outcrops are very difficult to find. It is quite uniform in character throughout the region of its outcrop. Many beds contain numerous creamy-white, calcareous concretions and a few beds carry cream-colored opaline and chalcedonic concretions up to six inches in diameter. One of the best exposures of the upper Frio beds is found in a cut along the San Antonio, Uvalde & " 2Bailey, T. L., "The Geology and Natura.I Resources of Colorado County," Univ. Texas Bull. No. 2333, pp. 150-153, 1923. University of Te:i·a,s Bulletin Gulf Railway, one mile northwest of Fant City, Live Oak County. This section is of considerable interest because it illustrates the general lithology of the Frio and the appar­ently conformable character of the Gueydan-Frio contact, and also shovvs a marked, reversed dip of the strata due to either a small anticlinal fold or a fault with the northwest side downthrown. SECTION IN CUT ON S.A.U. AND G. RY. I MILE NORTHWEST OF FANT CITY Gueydan (Fant me;nber): 2. Grayish-white, very friable, rathe!' massive-bedded tuff or bentonitic volcanic sand (exposed only at north­west end of cut on account of reversed dip)_--·· _________12' Frio: 1. Creamy-gray to greenish, plastic, gypsiferous clay con­taining thin lenticular streaks of sandy material. The gypsum occurs in clusters or in irregular streaks of acute-edged, platy crystals from 1/ 16 to 1/ 4 inch long distributed rather abundantly through the clay. The surface of the clay is nearly covered with these selenite crystals. Approximate thickness exposed at southeast end of cut ---------------------------------------------------------15' Strike N 30° E; Dip 10° ~W (Taken on No. 2). A complete section of the Frio in Live Oak County was obtained in H. Coquat and Associates' Hicks No. 1 well and is given on pages 94-100. The contact between the greenish-gray Frio clay and the brownish-gray, medium-grained Fayette sandstone is ex­posed at the bridge on the Three Rivers-Whitsett road one mile south of Whitsett. The basal beds of Frio exposed here consist of creamy-green very soft and plastic, bentonitic clay which contains a great abundance of flattish gypsum crystals. These crystals are generally in irregular seams but may be scattered through the rock. The clay is non­calcareous and weathers to a tawny color. Gueydan Formation PETROGRAPHIC DESCRIPTION OF FRIO CLAY, 1 MILE SOUTH OF WHITSETT, LIVE OAK COUNTY Sorting.-About 15% of washed material (coarser than 1/ 16 mm.) is present in the sample and fully 90% of this washed material is secondary gypsum crystals or fragments of them. The sand grains that occur are subangular to angular, and are between 1J1 and 1/16 mm. in diameter. Composition of Washed Material.-Gypsum-90% ; plagioclase (mostly oligoclase, but occasional albite, andesine and labradorite grains)-3%; orthoclase and sanidine-1o/o ; quartz-5% ; chert 1o/o ; volcanic g·lass (often pitted as if by corrosion)-trace; microcline­rare; barite-trace; magnetite-trace; green hornblend (well rounded)-trace; calcite-trace; titanite (rounded)-rare; forami­fera (fragments of Globigerina, probably derived from Cretaceous rocks) -tr ace. A sample of Frio clay from four miles west of Rio Grande City was found to have a composition almost identical to that of the above sample except that the sample from near Rio Grande City contains a number of composite concretions of limonite, goethite, and gypsum. Ostrea georgiana, an Eocene oyster, was also found here. A similar analysis was made of a purplish-pink and light grayish-green, mottled, gyps if erous, marly clay, containing some bentonite and numerous microcrystalline calcareous concretions, collected from a small gully near the northwest corner of Live Oak County, six miles southeast of Crowther. PETROGRAPHIC DESCRIPTION OF FRIO CLAY FROM NORTHWESTERN CORNER OF LIVE OAK COUNTY, 6 MILES SOUTHEAST OF CROWTHER Sorti.ng.-Washed material comprises only about 1h% of the sam­ple. About 90% of this washed material consi3ts of secondary calcite and gypsum. Even when the gypsum and calcite are excluded the sorting of the washed material is poor, the sand grains ranging from 1 mm. (rare) to 1/ 16 mm. with the maximum separate between %, and% mm. Composition of Washed Material.-Calcareous concretions (up to 2 mm. in diameter)-55%; transparent granular calcite-6%; gyp­sum crystals-30% ; plagioclase-2%; sanidine-trace; quartz-trace; chert-I%; volcanic glass (partly altered)-trace; opal-1% ; green ho~r.blend-1% ; hypersthene-2%; epidote-1 % ; barite-trace; zircon-trace; titanite-trace; magnetite-trace; pyrite-trace; Un·ii:ersity of Te:ras Bulletin limonite-trace. No fossils are noted in this sample, but T extularia globulosa and Globigerina sp., which are probably derived from the Cretaceous marls, are quite numerous in an associated bed of creamy­green calcareous clay. Beds of flaggy, cross-bedded, mottled brownish-pink and grayish-green, friable, argillaceous, fine-grained sandstone are locally present near the top of the Frio. The most prominent outcrop of this rock is found in the southwestern corner of McMullen County, two miles northwest of Walker ranch. These sandstones are associated with clays carrying opaline concretions. Thin beds of concretionary limestone are of very local occurrence in the Frio. The only indigenous, aquatic fossils seen in the Frio clay by the author were a few specimens of Ostrea georgiana Conrad found in the vicinity of the Rio Grande. These oysters are much more abundant in the Fayette formation. Portions of silicified trees are found in the Frio also. Correlation ancl stratigraphic relations.-The Frio clay was named for the Frio River by Dumble23 in 1894 and the type locality then described occurs in the upper Gueydan formation of the present writer. However, in a later paper/1 he maps the Frio in Live Oak County in such a way that it includes most of the Gueydan and practically none of the Frio (as recognized in the present paper), while in McMullen County it includes the Frio and also most of the lower Gueydan but excludes the beds of his type locality in Live Oak County. The same beds found at this type locality are placed in the Oakville and Lapara in southeastern Mc­Mullen County, and a number of the exposures of Gueydan capped by Oakville on mesas, such as Loma Alto and La Chusa, are found in his report on page 969 under the head­ing "Lapara Sections." The Frio, as defined in this latter paper, is thus not a stratigraphic unit. In a recent paper, 23Dumble, E. T., "The Cenozoic Deposits of Texas," Jour. Geol.. Vol. 2, No. 6, pp. 549-567, 1894. 2·1Dumble, E. T., "The Geology of Southwestern Texas," Trans.. A.I.M.E., Vol. 33, pp. 913-987, 1903. Gueydan Formation Dumble25 defines his Frio so that it is practically coincident with the writer's Gueydan, but believes it to be of Jackson, Eocene (not Oligocene or younger) age. His reason for nlaciug these beds in the Eocene is given in the following auotation taken from page 435. No f ossils except a leaf impression were found in the beds here assigned to the Frio. The Jackson age of the Frio is indicated by the fact that we have traced what we consider to be its continuationzsa southward to the Conchos RivH in Mexico where it forms a good part of the Pomeranes Mountains, and on the eastward-facing slopes of these mountains we find this Frio overlain by beds of fossiliferous sands and clays belonging to the lower Oligocene. The Gueydan beds (Dumble's Frio) were not found to outcrop on the Texas bank of the Rio Grande. Since Dumble does not describe the Mexican beds, it is impossible to know to which strata he referred in the expression "continuation southward" of his Frio. Opposed to the Eocene age of the Gueydan are the following facts : (l) The same strata were traced by the writer from Dumble's type locality of his Frio in Live Oak County to southern Gonzales and northern Lavaca counties where they overlie beds of typical Catahoula (Oligocene) sandstone and underlie typical Oakville (Upper Miocene) sandstone. The relations in these counties are quite clear. Since the dis­tance is much shorter from Dumble's type locality to Gon­zales County than the Conchos River in Mexico, there is less chance of error in mapping. With the exception of a few offsets due to dip faults there is no marked discontinuity in the outcrop of these strata between the type locality and Gonzales County such as exists for twenty miles north of the Rio Grande and for an unknown distance south of that river. 25Dumble, E. T., "A Revision of the Texas Tertiary Section with Special Reference to the Oil-Well Geology of the Coast Region," Bull. Am. Assoc. Petr. Geol. Vol. 8, No. 4, pp. 424-444, 1924. 2sartalicized by the present writer. Uniuersity of Te:cas Bulletin (2) The marked variation in width of the Frio clay belt, as defined by the writer, and its disappearance north­east of Atascosa County, probably because it is overlapped by the Gueydan, strongly suggests an unconformity or disconformity between the Gueydan and the Frio although their contact in most of the places where it is exposed appears to be conformable or nearly so. As has been previously mentioned on pages 6-7, Trow­bridge"'' has included beds ranging in age from Upper Eocene to Lower Pliocene in his "Frio Formation" and placed this "formation" in the Eocene. The author has found it quite possible to subdivide the Frio of Trowbridge into the Frio, Oakville, and Lagarto formations along the north side of the Rio Grande. Deussen's Frio embraces both the Gueydan tuff and the Frio clay. In regard to the age of this "formation" Deussen:" makes the following statement on page 91: The sparse fauna of the Frio clay makes its correlation diffi­cult. The lo\Yer beds in southwest Texas carry Ostrea georgiana Conrad, which may indicate a deposit of either late Eocene or early Oligocene age. The upper beds may be in part of early Oligocene age, but until further evidence is available the forma­tion is classified as of late Eocene (J3ckson) age. In ea·.'tern Texas b2ds of clay that are in general of similar litholog·ic composition oYerlie the Catahoula sandstone. These cla:ity of Tc.ras Bulletin irregularities on their surfaces and frequently give only a suggestion of a rounded or elliptical form. They are sepa­rated from each other by ramifying dendritic masses and filn's of a yellovvish or pale greenish substance which con­sists of a mixture of bentonitic and opaline material. Some of the tuff beds are composed almost entirely of such round­ish lumps, which give to the rock a conglomeratic appear­ance. In a few localities veins of light bluish barite occur in the joints or cracks between the lumps instead of the opal and bentonite. The large lumps are composed of the same kind of tuff as the matrix in which they are included and they possibly represent fragments of a dehydrated and somewhat solidified crust which was formed on the upper surface of the mud flow and later engulfed; these fragments were subsequently comminuted and partly rounded by rolling over and over in the more liquid volcanic mud below. Also, the spheroidal masses may have been formed near the ends of mud flows by a process similar to that which produces a mass of irregular blocks of the solidified crust in the ends of lava flows. In other words, the more liquid mud may have pushed forward, incorporated fragments of the partially solidified mud at the front of the flow, and rounded them by rubbing them against one another as the mud flow proceeded. It is impossible to ascertain which of these plausible explanations is the true one, and it is probable that several processes have contributed to the for­mation of the larger lumps. These larger lumps are com­monly composed of a number of smaller lumps. Lacroix·10 has described and photographed tuffs composed of such pisolite-like bodies of various sizes up to nearly 1 cm. in diameter. He states that they have been formed by the cohesive action of rain drops. Many of the smaller pisolite­like bodies in the Fant member of the Gueydan tuff have probably originated in this manner. Some of the smaller lumps are pumice pebbles, rounded, no doubt, by the attri­tional action of finer particles and other pumice fragments 40La<:roix, A., "La ::\Tontagne Pelee et ses Eruption«" Pa1•1·s M . .., , asson et C1l', Jl. 4~0 and fig:. 181, 1904. Gueydan Formation while the mud flow was in motion. A few of these rounded pumice lapilli may have had an original rounded form. Lacroix41 states that the pumice lapilli from Mont Pelee, both from vertical projections and inclined fiery clouds, are ordinarily rounded on account of their slight hardness and their fragility due to their porous structure. It is, there­fore, more probable that the Gueydan pumice lapilli were rounded after their expulsion from the volcano. Many of the very small rounded bodies are hollow and are evidently glass bubbles. These smallest spherules aTe abundant in tuff beds where only a few of the larger tuff lumps occur. The vesicular cavities, which are common in this tuff, are very irregular to rounded or flat ellipsoidal in shape and may be lined or completely filled with soft waxy, cream­colored to light gray, or pink bentonite. Crusts of trans­parent, botryoidal hyalite opal or mixtures of bentonite and opal are present in other cavities. In a few of the tuff beds in Live Oak, McMullen, Webb, and Starr counties some of the cavities are lined with tiny glittering plates of tridy­mite or with a mixture of botryoidal hyalite and tridymite. These minerals are discernible with a hand lens, but the crystal form of the tridymite cannot be distinguished with­out a microscope. The tubular cavities probably represent channels along which steam and other vapors-originally included in the mud flow-escaped. In certain beds these tubular cavities, now filled with opal and bentonite, are quite prominent and resemble small smooth-surfaced rootlets crossing each other at various angles. Some dendrites of manganese and limonitic concretions are seen in this tuff. No fossils have been found in the mud-flow tuff beds. (2) A number of beds of light gray to white or grayish­pink, very dense textured, chert-like silicified tuff or "por­cellanite" are present at many localities in the Fant mem­ber, especially near its base. Although this rock has a hardness of 5 or 6, and breaks with a conchoidal fracture like that of chert, it is cut by closely spaced, irregular, con­choidal joint cracks which cause it to break down readily 11op cit., p. 371. U11il'Nsity of Tc.ms Bulletin into such small angular fragments that it is difficult to obtain a large hand specimen of the rock. Some of this rock occurs in platy beds like those of chert, while other samples are difficult to distinguish megascopically and without chem­ical and hanlness tests from a dense-textured limestone. All gradations between this rock and slightly silicified, typi­cal, vesicular tuff :ue present. Some of the best exposures of this silicified tuff are found near Wentz in McMullen County and on the Chapote ranch, six miles northwest of Loma Alto, in· McMullen County and at several localities near Simrn01rn in western Live Oak County. The beds are commonly most completely silicified in their lower portion. Adjacent beds above and below these may be only slightly silicified. Harker'" states that silicification is very common in the acid tuffs. (3) Friable, regularly bedded, white to light gray or yellmvish-green, medium-grained tuffs, which on first glance resemble sandstones, are commonly interstratified with the mud-flow tuffs. These tuffs are entirely different in appear­ance from the mud-flow tuffs. They have evidently reached their place of deposition by settling through the air. The air-deposited tuffs are so friable that they may easily be crushed by a slight pressure of the fingers into a powder of angular grains, many of which are platy. The individual grains are predominantly of volcanic glass and may be identified with a hand lens. In many samples over half the grains are of fine to coarse sand grade; therefore, this type of tuff >Vil! Le classified as sand-tuff. The remainder of each sample is fine material, largely bentonite, which causes a sample of the rock to break down quickly into a doughy mass when placed in water. An individual bed of this sancl-tuff attains a thickness of ten feet in the bluff on the northwest side of the Nueces River, three-quarters mile northwest of Simmons. The tuff in this bed tends to break roughly parallel to the bedding but the rock seems to be homogeneous from top to bottom and is the product either of a single explosiYe volcanic outburst or of several outbursts in 4"Harker, Alfn•d, " Petrnlog-y for Students," 5th Ed., p. 259, 1919. Gueydan Formation such rapid succession that material of the same type fell continuously to a depth of ten feet or more. This type of sand-tuff probably comprises one-fourth of the Fant member in western Live Oak and eastern McMullen counties. (4) Many beds of soft, creamy-gray to grayish-green argillaceous (bentonitic) tuff are inte1·calated between beds of tuff of other types. These are probably formed in part by the nearly complete alteration by hydration of original tine-textured tuff or volcanic dust beds into clay-like benton­ite and in part by the reworking of some of the other types of tuff by streams. The reworked tuffaceous clays can sometimes be distinguished from the bentonitic tuffs by current-bedding, poorer sorting and the presence of chert grains in the former, although the two types are very similar in appearance. Lying loose on the outcrops of certain beds of pinkish or white, argillaceous Fant tuff in the southwestern corner of McMullen County, are found a great number of water­worn, reddish-brown to dark gray boulders and pebbles of non-porous to vesicular and amygdaloidal, porphyritic tra­chyandesite or acid andesite. These boulders commonly range from 3 cm. to 30 cm. in diameter and resemble many of the boulders which occur so abundantly in the middle Gueydan (Soledad) conglomerate. A photograph of a group of these boulders is shown in Plate VII, Figure 1. The boulders appear to have weathered out of the tuff beds in this region and, in fact, many are coated or have their vesicles filled with the tuff. (5) Several thick beds of purplish-pink and grayish-green mottled, non-calcareous to marly, unlaminated, bentonitic clay are found in the lower Gueydan; these are rarely ex­posed except where capped by resistant tuff which protects them from rapid disintegration into soil. From well records it appears that about one-third of the Fant member is composed of this type of clay. The clay carries many ellip­tical, pinkish to whitish, dense-textured to granular cal­careous and siliceous concretions up to 10 cm. in diameter. Dark splotches and dendrites of manganese oxide ai·e com­mon along fracture surfaces throughout these concretions. Uni re rsitu of Te;ras Bulletin Similar manganese dendrites are also present in the joint cracks of the surrounding clay. These clays are evidently fluviatile-having been derived from the erosion of older Gueydan tuff beds and of Frio clay and older formations. These clays are very similar in appearance to those of the Frio except that the former contain very little gypsum, which is quite abundant in the Frio. (6) Beds of creamy-gray to light pinkish-brown, fine to medium-grained, argillaceous, arkosic, and generally tuf­faceous sandstone of ftuviatile origin are also present in the Fant tuff. This sandstone is characteristically cemented with opal but a mixture of opal and calcite cement is quite common. The creamy-gray sandstone is thick in parts of Duval, Webb, and Zapata counties. It is evidently derived in part from the erosion of tuff beds. (7) Coarse to fine conglomerates composed largely of bluish-gray, rounded to subrounded pumice pebbles or lapilli are locally present in the Fant tuff. The matrix of these pebbles is a medium-grained sandy tuff, tuffaceous sand or bentonitic clay. Such conglomerates are rather uncommon in the Fant member, being much more characteristic of the Soledad, or middle member. The best exposure of pumice­conglomerate (which may belong to the basal Fant beds since it appears to be on the upthrow side of a fault here) is located on the Armstrong place, one-quarter mile west of the Atascosa River in the northeast corner of the Thomas Henry survey in northern Live Oak County. Lacroix43 has described beds composed almost entirely of pumice lapilli in his section of the pyroclastic material covering Pompeii. The Gueydan pumice pebble or lapillus beds may have been formed by a process of gravitative sorting similar to that described by Lacroix in Vesuvian eruptions, but it is quite as probable that running water concentrated this coarse material. A conglomerate composed of large water-worn boulders or pebbles of white mud-flow tuff and a few chert pebbles ·•"Lacroix, A., "La Montagne Pelee apres ses Eruptions," Paris, l\Iasson et Cie, p. 120, 1908. Gueydan Formation is exposed at the "Falls" of the Atascosa River. It rests upon a slightly irregular, eroded surface of a yellow, gyp­siferous clay which is probably Frio. The beds in this region commonly show dips of 10 to 20 degrees northwest, apparently having been dragged into this attitude by fault­ing. If the clay underlying the conglomerate at the "Falls" of the Atascosa is really Frio, then a slight unconformity exists between the Gueydan and the Frio in this vicinity. Lithologic variations along the strike of the Fa,rit rneni­ber.-In Zapata County the tuff beds of the Fant member are thinner and the member as a whole becomes more argil­laceous in character than at its type locality in Live Oak County. The clays in the Gueydan south of McMullen County commonly carry some gypsum and are therefore difficult to distinguish from the Frio clays. The white tuff beds which outcrop prominently from six to twenty miles north and somewhat west of Mirando City in Webb County are mapped as part of the Fant member of the Gueydan, although they may represent a different horizon of tuffs and belong in the Fayette formation. These tuffs have evidently been placed in the Fayette by Trowbridge, as shown by his map.44 The uncertainty in regard to the exact position of the white tuff beds, which are traceable through eastern Webb and most of Zapata counties, is due to the fact that the white tuff beds in the lower Gueydan of north­western Duval County appear to be rather thin and are interbedded with gray, buff, and brownish conglomeratic sandstones and bentonitic clays, whereas, in Webb County, similar white tuff beds are again thick and outcrop promi­nently. The tuff beds in Webb County are non-vesicular, platy-bedded and generally more silicified than farther northeast, thus differing from the typical Gueydan. On the other hand there are some beds of grayish-white, friable sand-tuff of the same character and composition as the Fant sand-tuff in Live Oak County. Interbedded with these tuffs in Webb County are beds of light gray, somewhat silicified, 44Trowbridge, A. C., U. S. Geo!. Surv. Prof. Paper 131D, Plate xxvn, 1923 U11i1·ersitu of Te:ras Bulletin tuffaceous sandstone, containing a large number of biotite flakes and rounded to angular grains of reddish-brown or dark gray, hyalopilitic andesite or trachyandesite. Under­lying the series of tufFs and tufi'aceous sandstones are greenish-gray and pink and green mottled, gypsiferous clays containing a fe\v lenses of friable sandstone. Similar clays outcrop ove1· a strip three to five miles wide lying west of the base of the tuff, and are believed by the writer to belong in the Frio formation instead of the Fayette. These clays and some of the bentonitic day beds in the tuff series con­tain logs of silicified wood up to ten inches in diameter. The logs show distinct angular rings and contain a number of large pores which resemble resin ducts. It is probable that they ~tl'e logs of gymnospermous trees, but they have not been identified. Similar logs are abundant in the Cata­houla formation in eastern Texas. For the several reasons given above, these white tuffs and associated rocks are mapped \.Vith the Gueydan formation. Outcrops of lumpy, vesicular, white, tridymite-bearing, mucl-tlow tuff of appearance and petrographic character identical to those of the mud-flow tuff of Live Oak and McMullen counties are found in the hills west of the Mirando City oil field, one mile southwest of Mirando City, under­lying the middle Gueydan conglomerate and tuff. Similar tuff is also found six miles north of El Sauz, Starr County, twenty feet west of the Hebbronville-Rio Grande City road. Here it is surrounded by Reynosa limestone. These two outcrops indicate that the Fant trachyte tuff extends south as far as northern Starr County, even though the strati­graphically lower, platy-bedded tuffs of Webb and Zapata counties should later be proven to belong in the Fayette formation. Northeast of Atascosa County no mud-flow tuffs are noted in the Fant member which here consists of slightly con­solidated sand-tuff and finer bentonitic tuffs, as well as tuffaceous days and sandstones. In Gonzales County and eastward, many beds of coarse-grained, friable to extremely well-indurated, quartzose sandstone, of a character typical of that of the Catahoula sandstone in East Texas appear Gueydan Formation in the Fant and Soledad members. These beds are dis­similar to any rocks found elsewhere in the Gueydan. Thus, the Gueydan grades eastward along its strike into the Catahoula (Oligocene) sandstone. Sections.-The following sections will illustrate the character of the Fant tuff at its outcrop. SECTION ON CHARLIE YORK CREEK, 4 MILES WEST OF THREE RIVERS, BETWEEN THREE RIVER.S-TILDEN ROAD AND FRIO RIVER, LIVE OAK COUNTY Ft. In. 9. Alternating harder and softer, massive-bedded and wavy-bedded, creamy to yellowish-white, somewhat lumpy, mud-flow tuff and argillaceous tuff 10-12 8. lndurated, creamy-white, vesicular tuff with a xposed sur­ f aces. This bed shows a few minor crumplings. It is cut by many thin, selenite veins, especially along bedding-planes. Some satin-spar is present in vertical lines _ 2 3 1. Rather massive, pinld>:e mbles beds seen in the upper Frio in southwestern McMullen County. This sandstone becomes sun-cracked on exposure like clay. Contains many seams of sele­ nite alOng bedding and vertical point planes. Base not exposed. Thickness :"! G+ Total Thickness 1!) Strike doubtful, hut apparently nearly E. and W. Dip approximately % 0 or less N. Uufrersity of Tea:as Bulletin This section may possibly be in the Fayette. SOLEDAD MEMBER (Middle Gueydan) The Soledad member of the Gueydan tuff is characterized by beds of volcanic conglomerate, the pebbles of which con­sist partly or entirely of trachyandesite, andesite, trachyte, and similar effusive rocks, bluish-gray pumice and pink to light green tuff; by brown, pink and conspicuous, yellowish­green, volcanic sandstone composed of volcanic debris simi­lar to that of the conglomerate; and by beds of pink and light green, powdery, argillaceous tuff, bentonite and tuf­faceous clay. Silicified tuff lenses are small but fairly common. This member is thickest and most typically devel­oped at Soledad Hills and Government Wells in w€stern and northwestern Duval County. It appears to thin out gradually along the strike both northeast and south of these localities. However, the member is traceable to the north­east as far as central Karnes County and to the south as far as northeastern Zapata County, perhaps farther. (1) The most unique and distinctive rock in the Soledad member is that found at the Soledad Hills and numerous other localities in Duval, Webb, and Zapata counties. (See Plate VII, Figure 2.) As seen at Soledad Hills, the type locality, it is a fine conglomerate to a very coarse cobble or boulder conglomerate of a dark brownish-gray color which grades into conglomeratic sandstone in numerous horizons. The rock is cemented with a great abundance of milky-white to bluish, or translucent opal and chalcedony. The light­colorecl cement contrasts sharply with the dark pebbles, giv­ing the rock a very striking appearance. Some of the opal is of the fire-opal variety, showing a faint to distinct play of colors. The chalcedony is finely banded. In many places nearly half of the cementing material consists of pinkish to buff, chert-like, silicified tuff or silicified tuf­faceous clay. The pebbles which are unmistakably water­·worn, range in size from 1 mm. up to large boulders. One of these boulders measures 2 feet by 11/: feet by 114 feet. The larger pebbles, cobbles, and boulders are usually concen­ Gueydan Formation trated in certain strata or lenses. Other strata or lenses consist of fine conglomerates, grits, or even coarse sand­stones. The common range in the diameter of the pebbles is between 3 cm. and 20 cm. The majority of grains and pebbles above 2 mm. are well rounded to subrounde"d. Most of these are elliptical in outline and a number are quite flat with rounded edges; their parallel orientation causes the rock to appear well stratified. Practically all of the boulders, cobbles, and larger pebbles in the conglomerate are composed of porphyritic to aphanitic volcanic rocks representing several petrologic types, including trachyandes­ite, trachyte, andesite, and a few other types. The com­monest colors are reddish-brown, chocolate-brown, and dark gray, but green, red, light gray, purple, and occasionally white pebbles are found. Most of these pebbles and boulders are dense, but some are amygdaloidal or vesicular. The large boulder mentioned above has a spongy-looking texture, and is no heavier than many boulders half its size. The smaller pebbles and sand grains also consist of the same kinds of effusive rocks, but many small pebbles of pink or light green, fine-textured tuff or silicified tuff occur. Less than 10% of the rock consists of small, rounded pebbles or coarse sand grains of light brown or yellowish chert and vein quartz. Interbedded with the conglomerate beds are a number of lenses or beds of pink argillaceous tuff and silicified tuff. This conglomerate is fully 75 feet thick at its type locality. West of Government Wells in north­western Duval County it is represented by several thick conglomerate beds separated by beds of tuff, clay, and sand­stone. At localities a few miles northeast and a few miles south of the Soledad Hills the volcanic rock pebbles are mixed with a considernble number of large chert, quartz, and silicified tuff pebbles; and at Parilla Hills in extreme western Duval County, five miles northeast of Moglia, Webb County, the volcanic pebbles comprise only 10'j,, or less of the conglomerate, with only occasional large boulders of volcanic rock. Much opal and chalcedony cement is present here but probably half of the matrix and cement of the University of Te·;ras Bulletin conglomerate consists of pinkish or light green silicified clay like that which is also common in the Soledad Hills conglomerate and at Government Wells. In Webb and Zapata counties the Soledad conglomerate carries only 5% or less "to 20 j;. of volcanic rock pebbles and boulders but some pebbles of this type seem to occur throughout the conglomerate. In these counties the Oakville sandstone and conglomerate generally rests directly on top of the Soledad conglomerate, causing difficulty in determining at what point the Soledad ends and the Oakville begins along the face of the Reynosa or Bordas escarpment. It is also im­possible to know without borings to what extent this Soledad conglomerate is overlapped by the Oakville and Reynosa. Conglomerate of this type has not been seen north of the Duval-McMullen county line, but some of it probably extends into McMullen County before lensing out. (2) Below the Soledad conglomerate is a persistent horizon of a peculial' green volcanic sandstone. At some localities similar beds occur also above the conglomerate or interstratified with it. Green sandstone of this type has been seen as far northeast as northern Live Oak County where a small outcrop of the sandstone appears on the Naves ranch, three and three-quarter miles southeast of Fant City along the main Three Rivers-Pleasanton road. Other good exposures are found six and one-half miles north of Loma Alto Mesa and two and one-quarter miles north­west of Loma Alto schoolhouse in McMullen County. Sev­eral beds of this sandstone are seen in the road at Government Wells, Duval County, and one mile west of that settlement. The rock is a fairly intense to light yellowish­green, friable, generally somewhat laminated and cross­bedded, medium-grained, gritty, argillaceous sandstone. The green color is produced by a greenish, argillaceous, pos­sibly chlorite-bearing matrix and coating on the sand grains. In this sandstone a few to a considerable number of rounded grains are found, which indicate that the grains have been subjected to erosion. Many grains in the sandstone are composed of fine-grained andesite, trachyandesite, and other Gueydan Formatfon volcanic rocks; the remainder is composed of feldspar, chert, and volcanic glass. The rock in Live Oak County consists very largely of sharp-edged volcanic glass frag­ments. It is possible that this green sandstone may be used as a key horizon because it is rather thin and has a distinctive appearance. (3) Massive to cross-laminated beds of brownish-gray, grayish-cream, or brownish-pink, tuffaceous sandstones constitute one of the most prominent rock types in the Sole­dad member. These sandstones are friable to well indurated in character and range from coarse to fine-grained in tex­ture. They occur at several horizons in the middle Gueydan, forming a series of cuesta ridges in McMullen County be­tween the Ray ranch and San Caja Mesa. Most of this rock is medium-grained in texture, and the usual brownish color is due in part to the large number of red, yellow, brown, gray, and black grains of volcanic rock, such as trachyandesite, and grains of chert, and in part to the brownish-pink or light pink argillaceous material which occupies some or all of the spaces between sand grains. This argillaceous material seems to be tuff aceous and is in places silicified like that material which forms part of the matrix or cement for the Soledad conglomerate. Both the brownish (3) and the green (2) sandstones of the Soledad member are evidently derived from the same general source as the Soledad conglomerate, because the volcanic rock grains and the matrix of the sandstones are of the same types as the pebbles and the matrix of the conglomerate. Angular to somewhat worn volcanic glass grains and small pumice pebbles occur abundantly in some localities in this brownish sandstone. Many of the grains in the coarser textured sandstones are well worn and rounded by stream action and even in the finer grained type most of the grains are sub­angular instead of angular. In some beds a number of rounded, brownish-pink, often silicified, argillaceous tuft' or clay pebbles are noted. In addition to the argillaceous material, which may act as a cement, a large or small amount of granular calcite cement is found. Some specimens con­tain opaline cement also. In Karnes County, three miles University of Tex~as Bulletin south of Falls City, a horizontal outcrop of this argillaceous sandstone shows well developed, vertical cracks which ap­pear to be sun-cracks. They are filled with stringers of soft white caliche, although the sandstone itself contains very little calcareous cement. This is the most northerly outcrop of the Soledad sandstone that was found. (4) An exceedingly interesting and unusual rock con­tained in the middle Gueydan is a light brownish-pink, greenish or whitish pumice-conglomerate. The pumice pebbles, of a subrounded to well rounded shape, are cemented by a scanty matrix of rather friable, somewhat argillaceous, porous, vitric tuff of a pinkish or light greenish color. On account of the remarkable rarity of any extra­neous, non-volcanic material in this rock and from the marked vesicular character of the matrix it is possible that it is of purely eruptive origin and that the pumice frag­ments are lapilli which became rounded by grinding against one another during transportation through the air or in a mud flow, or in a volcanic rock slide. However, the pumice pebbles show a most striking variety of colors-bluish-gray, greenish-gray, cream-colored, yellowish-green, and dark gray or nearly black-which suggests that the pumice came from more than one volcanic vent or from several eruptions from the same vent, and may have later been concentrated by running water. Conglomerates of this type outcrop six and one-half miles south of Wentz, eastern McMullen County; four miles southwest of Simmons; and also on the south bank of Weedy Creek, two miles east of Fant City, Live Oak County. Other beds containing pumice pebbles have an argillaceous or sandy matrix and are quite evidently stream-deposited. ( 5) Associated with the pumice-conglomerate beds are friable, brownish-pink tuffs with a vesicular structure. These beds are practically never white in color and are usually much more friable than the somewhat similar beds in the Fant member. They also contain less glass and more bentonitic material than the Fant beds and commonly carry small angular grains of brownish trachyandesite or andesite. Many brownish-pink, extremely friable or pulverulent beds Gueydan Formation of very fine-grained argillaceous or bentonitic tuff and tuffaceous clay are intercalated with the tuff beds. Volcanic BouldeJ·s.-The tuffaceous clay beds and the purer tuff beds in the Soledad member carry a few rounded, apparently water-worn boulders of red-brown to dark gray, very vesicular andesite or trachyandesite. These are found in many parts of Live Oak, McMullen, and Duval counties. They range from 1" to 1' in diameter. Most of the boulders are very spongy and light in weight. The largest boulders found anywhere in the area are exposed in a gray loam soil on a chaparral-covered plain on Section 480, J. E. Mur­phy survey, two miles northeast of Loma Alto Mesa, where fifty or more of the boulders were seen. Several of them are subangular to angular in form, as much as two feet in diameter, and each weighs several hundred pounds. A few do not seem to have been worn at all. They consist of dark gray to nearly black, vesicular, acid andesite or tra­chyandesite carrying many partly resorbed platy pheno­crysts of plagioclase up to three centimeters, or one inch, in length. The outcrop of the boulders forms a fairly definite line, probably because the lens or bed in the Soledad member which carries them has been tilted and then truncated by erosion. The only rock that outcrops near these boulders is a pink, friable, argillaceous tuff belonging to the upper part of the Soledad member or to the lower part of the Chusa member of the Gueydan formation. In addition to the volcanic boulders a few angular to rounded cobbles and pebbles of hard siliceous sandstone resembling that of the Fayette are locally found in the tuff beds just mentioned. It is difficult to conceive of a method by which such a great number of large boulders could have been concentrated here unless an old volcanic vent is located within a few miles of this locality. (6) Beds of whitish or pinkish clay are interbedded with the previously described rocks. These clays, as well as the sandstones and most of the conglomerates, are evidently of fluviatile origin. (7) Underlying the pumice conglomerate on the bank of Weedy Creek, two miles east of Fant City, is a dense­ Unive1·sity of Texas Bulletin textured, rather soft, somewhat conglomeratic, unfossilifer­ous limestone. A number of sun-cracks are seen in this limestone. They are filled with a tuff which contains pumice pebbles. That this is a fresh-water limestone seems prob­able from these relationships. A fairly coarsely crystalline limestone containing euhedral rhombs of siderite was pene­trated in the Hawley well located one mile south of the Weedy Creek outcrop and two miles southeast of Fant City. Below this conglomeratic limestone, twenty feet of serpen­tine was encountered. Sections.-The following are given as the most representa­tive sections of the Soledad member exposed. SECTION 2 MILES NORTH OF GOVERNMENT WELLS, DUYAL CO., IN BLUFFS ON BOTH SIDES OF COTULLA ROAD Feet Gueydan (Soledad member): 4. Coarse-grained, grayish-brown sandstone or grit and conglomerate with vertical and lateral grada­ tions from sandstone to conglomerate. These rocks are fairly hard, and thin to massive-bedded. The pebbles consist mainly of buff, gray, or light brown chert but many pebbles of brown, red and purplish trachyandesite (?) and pink argilla­ ceous tuff and clay are also present. The pebbles are generally rounded and range in size from coarse sand to boulders 1 foot in diameter. The largest boulders consist of vesicular volcanic rocks ------------------------------------------------------------------------------20 3. Brownish-pink, dense-textured, sandy, much jointed silicified tuff or tuffaceous clay containing numerous irregular lenses, streaks, and pockets of coarse sandstone or fine conglomerate. This grades into No. 4. Thickness___________ __ _______ ____________ 1 2. A yellowish-green, very friable, arkosic, pebbly grit or coarse conglomeratic sandstone. Many of the sand grains and pebbles in this sandstone consist of volcanic rocks, apparently andesite and trachyandesite. The cement is a mixture of greenish argillaceous material and calcite. The base of this bed has a 5° inclination to the south__ _l2 Gueydan Formation Feet Disconformity Fant member (?) : 1. Profusely jointed, more or less silicified, dense­ textured, white tuff grading upward into a soapy, cream-colored bentonite which breaks into spheroidal joint fragments and irregular to tetra­ hedral pieces. Some glass grains containing bub­ bles can be seen in the upper portion with the naked eye ---------------------------------------·--------------------------------15 Total Thickness ____ _____ __ _ _________ _ _______ __ _________.47 Dip uncertain, beds approximately horizontal. SECTION AT SOUTH END OF SOLEDAD HILLS, NORTHWEST DUVAL COUNTY Gueydan (Soledad member): 2. Well-indurated to friable, brownish-gray, mas­ sive-bedded conglomerate consisting almost en­ tirely of pebbles, cobbles and boulders of porphy­ ritic volcanic rocks of several types (mostly trachyandesite) and some of tuffacE:ous clay. The cement is white chalcedony, opal and si.Jicified tuffaceous ( ?) clay. Some lenses and angular, boulder-like masses of pink silicified tuffaceous clay up to 2 feet in thickness are present in places....50 Disconformity (?) : 1. Soft, powdery, finely pisolitic, pink altered tuff or bentonitic clay containing streaks and lenses of conglomerate like No. 2. Base not exposed. Thick­ ness exposed --------------------·_---------------------------------------------25 Total Thickness -----------------__ ____ __ _______________________________ 75 Approximate strike N. 24°E; Dip 1/2°-1° S. E. SECTION % MILE EAST OF THE McMULLEN COUNTY LINE, 4 MILES SOUTHWEST OF SIMMONS, LIVE OAK COUNTY Soledad member: 9. Brownish-pink, friable, porous, andesite ( ?) tuff composed mainly of volcanic glass fragments and containing a few rounded boulders (possibly in part volcanic bombs) of red-brown vesicular andesite? ________-----------------------------------------------------------.. 4 Ft. In. 8. Hard to friable, thin-bedded, light brownish-pink. feldspathic sandstone containing man~· angular vclcanic gla:os fragments and subangular grains of andesite? Grades into Xo. 9 at top.. ......20 - CoYered . .... ---------.. 1-! '· Fant member: 6. \Yhite, friable and powdery, fine-grained, argilla­ceous tuff or dust-tuff.. .. 5 5. Hard, green and pink mottled, closely jointed, si­ licified clay or shale 5 ,I, Fine-grained, friable, sand-tuff of light gray color 1 3. Like Xo. 6__ __ _________ -4 :2. Hard, oli,·e-green. silicified clay shale 6 1. Like Xo. 6. 6 Total Thicknes::: 2 SECTION ON WEEDY CREEK, LIVE OAK COUNTY, 2 MILES ABOVE FANT CITY BRIDGE Soledad member: 3. Conglomerate composed almost entire!~· of pumice pebbles in an argillaceous tuff matrix of a pink color. In the lower part are boulders and pebbles of soft white limestone... . ···-------------·----2 2. An irregularly bedded lens of sun-cracked, rather soft, white limestone. The sun-cracks are filled with tuff canying pumice pebbles. This lens grade upward into limestone-tuff conglomerate 1 Fant member: 1. Somewhat induratecl. lumpy, Yesicular. mud-fl.o\\­tuff to \\·ater's edge in \\'eedy Creek 6 Total Thickness 9 The rocks here are considerably crumpled. In one place a small northea~t :;triking fault with a dis­ placement of I feet is ~een. The north\Yest is down- t hrown. Dips on these beds Yariable, z:_5: S.\Y., s. and S.E. Gueydan Formation SECTION ON BRANCH OF ROCK CREEK, 1/s MILE NORTH OF NOPAL­ NIXON ROAD, 8 MILES SOUTH OF SMILEY, GONZALES COUNTY ·Gueydan (Soledad member(?) with Catahoula facies): Feet 6. Fairly friable, porous, wavy-bedded, medium-grained, ar­gillaceous, tuffaceous (?) sandstone which is cemented with white, waxy opal. Rock shows many red splotches of ferru­ginous material formed from the weathering of tiny, trigonal or hexagonal plates of vitreous, orange-yellow jarosite which commonly occurs in some of the vesicle-like cavities in the rock. The beds of this rock are commonly 6 to 8 inches thick. They resemble tuff beds in the gnarly appearance of their outcrop. Thickness of No. 6____________________ 10 Fant Member: 5. Unstratified, creamy-white, rather lumpy, argillaceous tuff 12 4. Very much cross-bedded and laminated, very friable, fine­grained, white, tuffaceous sandstone, composed largely of volcanic glass ----·-------· ------------·------·------------.. -------------·--------_____________ 7 3. Lamellar, white, tuffaceous shale containing lenses of pure white, powdery, bentonitic tuff 8 inches thick. Thickness of No. 3 -----------·­----­----.. ....·------­----­------·--------------· ·-----------------·---------------·· 4 .2. Like No. 4______________________ ____________ ______ ___________ ·------------------­------------·---­ 6 1. Lamellar, creamy-white, fairly hard, shaly, vitric tuff, interstratified with a few 1-foot beds of grayish white, friable sand-tuff. Base not exposed. Thickness exposed ..... 10 Total Thickness _______ .. ---­---·-·· _____ ________ _ ________ -------------­------··· 49 Beds which appear to be identical to No. 6 in physical char­ acter and mineralogical composition have been observed % mile east-northeast of Rockland, Tyler County, about 100 miles northeast of Houston in East Texas, in strata of practically unquestionable Catahoula age. These beds can also be traced from the locality of this section into north­ western Lavaca County, near Moulton, where they are mapped with the Catahoula by Deussen.-"; CHUSA MEMBER (Upper Gueydan) The Chusa member consists of non-indurated, very tuf­faceous clays and argillaceous tuffs which are generally unstratified. In a few localities they contain small, angular to rounded pumice lapilli or pebbles. This member is thus ''"Deussen, Alex., U. S. Geo!. Surv., Prof. Paper 126, Plate VIII, 1924. V11frersity of Te.fflS Bulletin composed of the same kind of material which is commonly interbedded with the Soledad sandstone and conglomerate strata, and grades downward into the latter member. The Chusa contains some thick, fine-grained, vitric tuff or vol­canic dust beds, but it seems to be essentially composed of the finer debris from underlying tuff beds which has been mingled "·ith variable amounts of extraneous, non-volcanic material by stream action. l\Iost of it is, therefore, not tuff but should be more appropriately termed bentonitic or tuf­faceous clay. This member, as noted previously, is com­pletely owrlapped by younger formations south of Duval County. but northeast of McMullen County it appears to be the thickest portion of the Gueydan. The thickness ha.;:; not been determined because a complete section of this member is nowhere exposed in one vertical outcrop, and because the beds are practically unstratified, which prevents recognition of definite horizons by ·which to correlate out­crops. Some borings beginning above the top of the Chusa member haYe passed through it and other members of the Gueydan into the un-~iO