Short Papers in Geology and Hydrology Articles 60-121 GEOLOGICAL SURVEY RESEARCH 1963 GEOLOGICAL SURVEY PROFESSIONAL PAPER 475-C Scientific notes and summaries of investigations prepared by members of the Conservation, Geologic, and IVater Resources Divisions UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1963UNITED STATES DEPARTMENT OF THE INTERIOR STEWART L. UDALL, Secretary GEOLOGICAL SURVEY Thomas B. Nolan, Director EARTH SCIENCES LIBRARY For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C., 20402FOREWORD This collection of 62 articles is the second of a series to be released in 1963 as chapters of Professional Paper 475. The articles report on scientific and economic results of current work by members of the Geologic, Water Resources, and Conservation Divisions of the United States Geological Survey. Some of the papers present the results of completed parts of continuing investigations; others announce new discoveries or preliminary results of investigations that will be discussed in greater detail in reports to be published in the future. Still others are scientific notes of limited scope, and short papers on methods and techniques. Chapter A of this series will be published later in the year, and will present a synopsis of work of the Geological Survey during the present fiscal year. Thomas B. Nolan, Director. m J 712CONTENTS Page Foreword------------------------------------------------------------------------------------------------------------------ hi GEOLOGIC STUDIES Structural geology 60. Structure of Precambrian crystalline rocks in the northern part of Grand Teton National Park, Wyo., by J. C. Reed, Jr______________________________________________________________________________________________________ Cl 61. Plutonic rocks of northern Zacatecas and adjacent areas, Mexico, by C. L. Rogers, Roger van Vloten, J. O. Rivera, E. T. Amezcua, and Zoltan de Cserna---------------------------------------------------------------------------- 7 Stratigraphy 62. The Ordovician-Silurian contact in Dubuque County, Iowa, by J. W. Whitlow and C. E. Brown___________________ 11 63. Spirorbal limestone in the Souris River(?) Formation of Late Devonian age at Cottonwood Canyon, Bighorn Moun- tains, Wyo., by C. A. Sandberg-------------------------------------------------------------------------------- 14 64. Dark shale unit of Devonian and Mississippian age in northern Wyoming and southern Montana, by C. A. Sandberg-, 17 65. Nomenclature for lithologic subdivisions of the Mississippian Redwall Limestone, Arizona, by E. D. McKee__ 21 66. Mississippian rocks in the Laramie Range, Wyo., and adjacent areas, by E. K. Maughan___________________________ 23 67. Triassic uplift along the west flank of the Defiance positive element, Arizona, by E. D. McKee----------------- 28 68. Revised stratigraphic nomenclature and age of the Tuxedni Group in the Cook Inlet region, Alaska, by R. L. Detter- man_____________________________________________________________________________________________________ 30 69. Redefinition and correlation of the Ohio Creek Formation (Paleocene) in west-central Colorado, by D. L. Gaskill and L. H. Godwin______________________________________________________________________________________________ 35 70. Tertiary volcanic stratigraphy in the western San Juan Mountains, Colo., by R. G. Luedke and W. S. Burbank., 39 71. Fenton Pass Formation (Pleistocene?), Bighorn Basin, Wyo., by W. L. Rohrer and E. B. Leopold___________________ 45 72. Nussbaum Alluvium of Pleistocene(?) age at Pueblo, Colo., by G. R. Scott_______________________________________ 49 Paleontology 73. Age of the Murray Shale and Hesse Quartzite on Chilhowee Mountain, Blount County, Tenn., by R. A. Laurence and A. R. Palmer___________________________________________________________________________________________ 53, 74. Conodonts from the Flynn Creek cryptoexplosive structure, Tennessee, by J. W. Huddle--------------------------- 55 75. Middle Triassic marine ostracodes in Israel, by I. G. Sohn_____________________________________________________ 58 76. Occurrence of the late Cretaceous ammonite Hoplitoplacenticeras in Wyoming, by W. A. Cobban-------------------- 60 77. Paleotemperature inferences from Late Miocene mollusks in the San Luis Obispo-Bakersfield area, California, by W. O. Addicott and J. G. Vedder---------------------------------------------------------------------------- 63 78. Late Pleistocene diatoms from the Arica area, Chile, by R. J. Dingman and K. E. Lohman_________________________ 69 79. Possible Pleistocene-Recent boundary in the Gulf of Alaska, based on benthonic Foraminifera, by P. B. Smith- 73 Geochemistry, petrology, and mineralogy 80. Petrology of rhyolite and basalt, northwestern Yellowstone Plateau, by Warren Hamilton_________________________ 78 81. The Canyon Mountain Complex, Oregon, and the Alpine mafic magma stem, by T. P. Thayer------------------------- 82 82. Modal composition of the Idaho batholith, by C. P. Ross________________________________________________________ 86 83. Solution breccias of the Minnelusa Formation in the Black Hills, South Dakota and Wyoming, by C. G. Bowles and W. A. Braddock________________________________________________________________________________________________ 91 84. Calcitization of dolomite by calcium sulfate solutions in the Minnelusa Formation, Black Hills, South Dakota and Wyoming, by W. A. Braddock and C. G. Bowles________________________________________________________________ 96 85. Apatitized wood and leucophosphite in nodules in the Moreno Formation, California, by R. A. Gulbrandsen, D. L. Jones, K. M. Tagg, and D. W. Reeser________________________________________________________________________ 100 86. Variation in element content of American elm tissue with a pronounced change in the chemical nature of the soil, by H. T. Shacklette........................................................................................... 105 Geochronology 87. Ordovician age for some rocks of the Carolina slate belt in North Carolina, by A. M. White, A. A. Stromquist, T. W. Stern, and Harold Westley____________________________________________________________________________ 107 Geophysics 88. Gravity survey in the Rampart Range area, Colorado, by C. H. Miller__________________________________________ HO 89. Gravity survey of the island of Hawaii, by W. T. Kinoshita, H. L. Krivoy, D. R. Mabey, and R. R. MacDonald. 114 90. Evaluation of magnetic anomalies by electromagnetic measurements, by F. C. Frischknecht and E. B. Ekren----- 117 vVI CONTENTS Sedimentation Fage 91. Glaciolacustrine diamicton deposits in the Copper River Basin, Alaska, by O. J. Ferrians, Jr___________________ C121 92. Competence of transport on alluvial fans, by L. K. Lustig_______________________________________________________ 126 93. Distribution of granules in a bolson environment, by L. K. Lustig_______________________________________________ 130 Marine geology 94. Sediments on the continental margin off eastern United States, by Elazar Uchupi______________________________ 132 Geomorphology 95. Possible wind-erosion origin of linear scarps on the Sage Plain, southwestern Colorado, by D. R. Shawe__________ 138 96. Glacial lakes near Concord, Mass., by Carl Koteff_______________________________________________________________ 142 97. Channel changes on Sandstone Creek near Cheyenne, Okla., by D. L. Bergman and C. W. Sullivan---------------- 145 98. Origin and geologic significance of buttress roots of Bristlecone pines, White Mountains, Calif., by V. C. LaMarche, Jr.........................................................................-.......-........................ 149 Economic geology 99. Bauxitization of terra rossa in the southern Appalachian region, by M. M. Knechtel___________________________ 151 100. An ore-bearing cylindrical collapse structure in the Ambrosia Lake uranium district, New Mexico, by H. C. Granger and E. S. Santos____________________________________________________________________________________________ 156 Engineering geology 101. Formation of ridges through differential subsidence of peatlands of the Sacramento-San Joaquin Delta, California, by G. H. Davis______________________________________________________________________________________________ 162 Analytical techniques 102. Chemical preparation of samples for lead isotope analysis, by J. C. Antweiler________________________________ 166 103. Percent-constituent printing accessory and flow-through cell for a spectrophotometer, by Leonard Shapiro and E. L. Curtis________________________________________________________________________________________________ 171 HYDROLOGIC STUDIES Engineering hydrology 104. Dissipation of heat from a thermally loaded stream, by Harry Messinger_______________________________________ 175 105. Movement of waterborne cadmium and hexavalent chromium wastes in South Farmingdale, Nassau County, Long Island, N.Y., by N. M. Perlmutter, Maxim Lieber, and H. L. Frauenthal_________________________________________ 179 106. Effect of urbanization on storm discharge and ground-water recharge in Nassau County, N.Y., by R. M. Sawyer__ 185 Ground water 107. Mapping transmissibility of alluvium in the lower Arkansas River valley, Arkansas, by M. S. Bedinger and L. F. Emmett______________________________________________________________________________________________________ 188 Surface water 108. Snowmelt hydrology of the North Yuba River basin, California, by S. E. Rantz___________________________________ 191 109. Field verification of computation of peak discharge through culverts, by C. T. Jenkins_________________________ 194 110. Use of low-flow measurements to estimate flow-duration curves, by O. P. Hunt___________________________________ 196 Analytical hydrology 111. Graphical multiple-regression analysis of aquifer tests, by C. T. Jenkins____________________________________ 198 112. Nomograph for computing effective shear on streambed sediment, by B. R. Colby__________________________________ 202 113. Distribution of shear in rectangular channels, by Jacob Davidian and D. I. Cahal_______________________________ 206 Geochemistry of water 114. Sulfate and nitrate content of precipitation over parts of North Carolina and Virginia, by A. W. Gambell_____ 209 115. Differences between field and laboratory determinations of pH, alkalinity, and specific conductance of natural water, by C. E. Roberson, J. H. Feth, P. R. Seaber, and Peter Anderson_____________________________________________ 212 116. Increased oxidation rate of manganese ions in contact with feldspar grains, by J. D. Hem_____________________ 216 117. Solution of manganese dioxide by tannic acid, by Jack Rawson___________________________________________________ 218 118. Effectiveness of common aquatic organisms in removal of dissolved lead from tap water, by E. T. Oborn__________ 220 Experimental hydrology 119. Adsorption of the surfactant ABS35 on illite, by C. H. Wayman, H. G. Page, and J. B. Robertson_______________ 221 120. Biodegradation of surfactants in synthetic detergents under aerobic and anaerobic conditions at 10° C, by C. H. Wayman and J. B. Robertson__________________________________________________________________________________ 224 121. Direct measurement of shear in open-channel flow, by Jacob Davidian and D. I. Cahal__________________________ 228 INDEXES Subject---------------------------------------------------------------------------------------------------------------- 231 Author----------------------------------------------------------------------------------------------------------------- 233Article 60 STRUCTURE OF PRECAMBRIAN CRYSTALLINE ROCKS IN THE NORTHERN PART OF GRAND TETON NATIONAL PARK, WYOMING By JOHN C. REED, JR., Denver, Colo. Abstract.—Metasedimentary gneisses of high metamorphic grade in the northern part of Grand Teton National Park have been subjected to at least two periods of Precambrian deformation. The older was preceded or accompanied by formation of foliated granitic gneiss; the younger was followed by emplacement of quartz monzonite and mica pegmatite. Previously published isotopic age determinations are somewhat anomalous, but they suggest that pegmatite associated with the younger granitic rocks may be as old as 2,660 million years. The Precambrian crystalline rocks of the Teton Range, Wyo., (fig. 60.1) were first noted and briefly described by members of the Hayden survey (F.H. Bradley, 1873; St. John, 1879). Since that time the surrounding Paleozoic and younger rocks have been extensively studied, but the Precambrian crystalline rocks in the core of the range have received surprisingly little attention. Horberg and Fryxell (1942) briefly described some of the metasedimentary rocks in the Precambrian complex, and C. C. Bradley (1956) made a detailed study of structures in the Precambrian rocks in a small area in the central part of the range. Giletti and Gast (1961) made isotopic age determinations on minerals from pegmatite and gneiss from Death Canyon in the southern part of the range, and Eckelmann (1963) studied zircons from a few of the gneisses and granitic rocks. In the summer of 1962 the writer began a study of the Precambrian rocks of Grand Teton National Park. Fieldwork during the summer was concentrated in the northern part, and mapping was completed between Webb Canyon and the divide south of Snowshoe Canyon (fig. 60.2). Tentative conclusions regarding the Precambrian structural history of the crystalline rocks in the area are summarized in this article. No doubt some of these will have to be modified as a result of more detailed petrographic studies and geologic mapping in other parts of the Teton Range. The oldest rocks mapped are a heterogeneous sequence of interlayered medium- and fine-grained amphibole gneiss, biotite-amphibole gneiss, biotite gneiss, and amphibolite, containing thin beds of amphibole schist, biotite schist, biotite-muscovite schist, and locally, sillimanite schist. The sequence also contains layers and lenses of fine-grained light-gray or white quartz-feldspar gneiss and quartzite. In one exposure northwest of Lake Solitude (fig. 60.1), south of the area mapped, a thin layer of dark-gray marble occurs in similar layered gneiss. Pods and sill-like bodies of coarse-grained amphibolite, hornblendite, and altered ultramafic rocks are locally common in the layered gneiss. The gneisses and schists are all of high metamorphic grade, probably in the sillimanite zone. They are conspicuously layered and generally display a strong foliation parallel to layering. Individual layers range from fractions of an inch to tens of feet in thickness, and some can be traced for hundreds of feet. In detail, however, many layers are discontinuous, and some are pulled apart into isolated pods and lenses. Locally, some layers are found in small isoclinal folds, whereas adjacent layers above and below seem undisturbed (fig. 60.3). It seems unlikly that the layering in the gneiss is relatively undisturbed bedding as suggested by Horberg and Fryxell (1942). Rather, the layers appear to represent a sequence of isoclinally folded and intensely sheared beds in which individual folds have been largely sheared out and obliterated, so that no lithologic contact represents an original bedding plane. Thus, inferences as to stratigraphic sequence and original thickness are meaningless. Nevertheless, the character of the layering, the mineralogy of many of the gneisses and schists, and the occurrence of quartzite and marble strongly support Horberg and Fryxell’s conclusion that these rocks were derived from sedimentary rocks; they were probably ART. 60 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C1-C6. 1963 ClC2 STRUCTURAL GEOLOGY 110° 50 110°40' EXPLANATION Quaternary deposits Tertiary volcanic and sedimentary rocks Mesozoic sedimentary rocks / //' Paleozoic sedimentary rocks 1 Precambrian rocks Contact Fault Dashed where approximately located; dotted where concealed i \ montanaI . g ___________1 ''—'"L.-Teton' IDAHO “ Range --T----WYOMING 1 I 10 MILES Figure 60.1.—Generalized geologic map of the Teton Range, showing distribution of Precambrian crystalline rocks and location of the area studied (fig. 60.2). Modified from Love (1956).REED C3 43° 55' 0 I___ 4 MILES Canyon Reynolds* //■A Pf>Pik Bivouac Peak 10 / / Inclined Horizontal Direction and plunge of lineation marked by alinement of mineral grains or elongate aggregates y Strike and dip of axial plane of minor fold /° / Inclined Horizontal Direction and plunge of axis of minor fold or crenulation EXPLANATION Diabase Quartz monzonite and pegmatite Dikes shown by dotted lines Foliated granitic gneiss Containing concordant layers of amphibolite (black) Layered gneiss and schist Generalized trend of layering and foliation shown by lines Contact Dashed where approximately located Fault Dashed where approximately located, dotted where concealed. U, upthrown side; D, downthrown side 'V Strike and dip of bedding 50y x Inclined Vertical Strike and dip of layering Inclined Vertical Strike and dip of foliation Figure 60.2.—Geologic map of the Precambrian rocks in the northern part of Grand Teton National Park. Surficial deposits not shown. Geology mapped in 1962 by J. C. Reed, Jr., assisted by J. H. Dieterich and T. B. Ransom PRECAMBRIAN PALEOZOICC4 STRUCTURAL GEOLOGY derived from a sequence composed largely of shale, calcareous shale, and siltstone, containing some inter-bedded sandstone and a few layers of impure limestone or dolomite. Some of the amphibolite and amphibole gneiss may have been formed from mafic lava or tuff. In addition to the partly obliterated isoclinal folds described above, the layered gneiss displays many sub-isoclinal folds with diverse axial orientations. These differ from the isoclinal folds in being more open and in the continuity of layers in their limbs (fig. 60.4). Locally they have refolded the older isoclines. The diverse orientation of the younger folds suggests that they may be of several generations, but their pattern is complicated by post-Precambrian folding and faulting and has not as yet been satisfactorily worked out. Figure 60.3.—Layered gneiss about half a mile S. 65° E. of summit of Ranger Peak. Interlayered amphibole-biotite gneiss (light gray), amphibole schist (dark gray), and quartz-feldspar gneiss (white). Note the rootless isoclinal fold (upper center) and the many discontinuous layers. Figure 60.4—Subisoclinal fold in thinly interlayered amphibole gneiss and quartzite. About 500 feet north of lower fork of Snowshoe Canyon. Medium-grained granitic biotite gneiss forms large bodies associated with the layered-gneiss sequence and forms smaller sill-like bodies within the layered gneiss. The rock is strongly foliated but is generally without layering, although locally it is crudely layered, especially near contacts with the layered gneiss. Contacts are concordant with the layered gneiss, and foliation in the granitic gneiss is parallel to foliation and layering in the layered gneiss. Near the contacts some finegrained layered gneiss is interleaved with the granitic gneiss. The granitic gneiss contains layers of fine- to medium-grained amphibolite a few feet to several hundred feet thick, some of which have been traced for several miles. The amphibolite in these layers is similar to that which forms thinner layers in the layered-gneiss sequence. The granitic gneiss has nowhere been found in dikes or other crosscutting bodies. The structural concordance between foliation in the granitic gneiss and in the layered-gneiss sequence, the lack of crosscutting, the absence of inclusions, and the continuity of the amphibolite layers suggest that the granitic gneiss may have formed through some sort of transformation in situ of selected parts of the sedimentary sequence from which the layered gneiss was formed, and that the amphibolite bodies represent layers of a composition that was especially resistant to the process. Parallelism between foliation and lineation in the granitic gneiss and similar structures formed during the early deformation in the layered gneiss indicates that the granitic gneiss, whatever its origin, was emplaced before or during the earliest recognized period of deformation. Open folds in the foliation and in some of the amphibolite layers are similar to the second-generation folds in the layered gneiss and show that the granitic gneiss was clearly formed before this deformation. Fine-grained light-gray or white muscovite-biotite-quartz monzonite and associated muscovite- and biotite-bearing pegmatite form large irregular masses in the southern part of the area mapped. Dikes of similar pegmatite and aplite cut the layered gneiss and granitic gneiss. Similar rocks are widespread farther to the south, where reconnaissance shows that they form a major part of the Precambrian complex south of Leigh Canyon. The quartz monzonite locally displays faint flow structure that is discordant with the structures of the enclosing rocks. The larger bodies contain abundant angular blocks of layered gneiss, amphibolite, and granitic gneiss ranging from less than a foot to hundreds of feet in length (fig. 60.5). At the margins of the larger bodies of quartz monzonite and pegmatite the enclosing rocks are crisscrossed by swarms of pegmatite and aplite dikes, so that the contacts are arbi-REED C5 trarily placed where the granitic rocks form about half of the total mass. The dike swarms dwindle out gradually away from the granitic bodies, but a few scattered dikes of mica pegmatite occur throughout the area mapped. Contacts of the dikes of quartz monzonite and pegmatite, and of the inclusions in the granitic bodies, are knifesharp. Undeformed pegmatite and aplite dikes cut through second-generation folds in the layered gneiss, and some of the angular included blocks of layered gneiss contain similar folds, showing that the pegmatite and quartz monzonite were emplaced after the second episode of Precambrian folding. Adjacent to the large granitic bodies some of the layered gneiss contains abundant granitic folia and feldspar porphyroblasts. The metamorphic grade of the layered gneiss, however, seems to remain nearly the same throughout the area. The discrimination between the metamorphic effects associated with the emplacement of the younger granitic rocks and those to be attributed to the older metamorphism will have to await further petrographic study. One dike of fine- to medium-grained gray diabase cuts the Precambrian rocks between Raynolds Peak and Bivouac Peak. The dike ranges from a few feet to about 50 feet in thickness. The dike rock and the trend of the dike are similar to the conspicuous well-known diabase dikes farther south that are truncated by Cambrian rocks, and the dike in the Raynolds Peak area is therefore believed to be of the same age. The dike has chilled borders, and rocks adjacent to it are baked and stained bright red brown. The chilled margins and the sharp truncation of all older structures indicate that the dikes were emplaced long after the quartz monzonite and pegmatite, as has been pointed out by Bradley (1956). Figure 60.5. Angular inclusions of granitic gneiss in fine-grained quartz monzonite about 0.2 mile northwest of Lake Solitude (fig. 60.1). Note rotation of foliation in the blocks, and very faint flow structure in the enclosing quartz monzonite. Dike of light-colored biotite-gamet pegmatite cuts foliation of granitic gneiss in block in upper left, but is truncated by the quartz monzonite. A small dike of younger muscovite pegmatite (below hammer) cuts both quartz monzonite and granitic gneiss, but is inconspicuous in the granitic gneiss in this photograph.C6 STRUCTURAL GEOLOGY The observations and conclusions summarized above suggest the following tentative Precambrian structural history for the northern part of the Teton Range: 1. Deposition of a great but unknown thickness of sedi- mentary rocks composed of shale, calcareous shale, and siltstone, some interbedded sandstone and impure limestone, and possibly some volcanic rocks. 2. Deep burial and regional metamorphism of these rocks, accompanied by intense shearing and isoclinal folding, and probably by emplacement of small bodies of ultramafic rocks. At about the same time, granitization in situ of selected parts of the sequence to form granitic gneiss. 3. One or more periods of folding with less intense shearing. Older isoclinal folds refolded. 4. Emplacement of discordant bodies and dikes of fine- grained quartz monzonite and pegmatite. 5. Long interval of cooling, consolidation, and frac- turing. 6. Emplacement of diabase dike. 7. Long interval of erosion and subsidence followed by deposition of Middle Cambrian Flathead Sandstone. At present it is impossible to assign absolute ages to the various Precambrian events. The only absolute age determinations available from the Teton Range are for pegmatite and gneiss in Death Canyon. The pegmatite is probably the same age as the mica pegmatites associated with the quartz monzonite in the area mapped. Rubidium-strontium age determinations by Giletti and Gast (1961) on muscovite and microcline from the pegmatite give 2,660 and 1,990 million years, respec- tively. Determinations on biotite from the enclosing rocks, however, give 1,360 million years. This anomaly casts considerable doubt on the significance of these determinations. It is hoped that further geochrono-logic work will permit assignment of absolute ages to events recognized on the basis of other geologic evidence. REFERENCES Bradley, C. C., 1956, The Precambrian complex of Grand Teton National Park, Wyoming, in Berg, R. R., ed., Wyoming Geol. Assoc. Guidebook 11th Ann. Field Conf., Teton Range and Jackson Lake, Wyo.: p. 34-42. Bradley, F. H„ 1873, Report of Frank H. Bradley, geologist of the Snake River Division, in Hayden, F. V., Sixth Annual Report of the U.S. Geological Survey of the Territories * * * for the year 1872: Washington, U.S. Govt. Printing Office, p. 189-271. Eckelmann, F. D., 1963, Zircon paragenesis in Precambrian crystalline rocks of the Teton Range, Wyoming, [abs.] : Geol. Soc. America Spec. Paper 73, p. 143. Giletti, B. J., and Gast, P. W., 1961, Absolute age of Precambrian rocks in Wyoming and Montana: New York Acad. Sci. Annals, v. 91, p. 454-^458. Horberg, Leland, and Fryxell, Fritiof, 1942, Pre-cambrian metasediments in Grand Teton National Park, Wyoming: Am. Jour. Sci., v. 240, p. 385-393. Love, J. D., compiler, 1956, Geologic map of Teton County, Wyoming, in Berg, R. R., ed., Wyoming Geol. Assoc. Guidebook 11th Ann. Field Conf., Teton Range and Jackson Lake, Wyo.: in pocket. St. John, Orestes, 1879, Report of Orestes St. John, geologist of the Teton Division, in Hayden, F. V., Eleventh Annual Report of the U.S. Geological and Geographical Survey of the Territories * * * for the year 1877: Washington, U.S. Govt. Printing Office, p. 321-508.Article 61 PLUTONIC ROCKS OF NORTHERN ZACATECAS AND ADJACENT AREAS, MEXICO By CLEAVES L. ROGERS,- ROGER VAN VLOTEN,- JESUS OJEDA RIVERA, EUGENIO TAVERA AMEZCUA, and ZOLTAN DE CSERNA,- Denver, Colo.; Washington, D.C.,- Mexico, D.F. Work done in cooperation with Consejo de Recurtos Saturates no Renovables under the auspices of the Agency for International Development, U.8. Department of State Abstract.—Plutonic rocks ranging in composition from dioritic to syenitic intrude Late Jurassic and Cretaceous marine sedimentary rocks in the central part of the Mexican geosyncline. The sediments were folded and uplifted in Eocene time; the deformed sediments were intruded by igneous masses during the late stages of the early Tertiary orogeny. Faulting and magmatic activity recurred during the middle and late Tertiary. A number of small to moderately large plutons ranging in composition from dioritic to syenitic were examined during a reconnaissance study of phosphate deposits in a large area of north-central Mexico (fig. 61.1). This area, about 26,000 square kilometers in extent is mainly in northern Zacatecas and southern Coahuila States but extends eastward into Nuevo Leon and San Luis Potosr. Most of the igneous rocks are in Zacatecas. The area contains a thick section of marine sedimentary rocks that range in age from Late Jurassic to Late Cretaceous and were folded and uplifted during the Eocene. It has been divided into three geologic belts (Rogers and others, 1962), designated, from north to south, as the valley and ridge belt, plain and range belt, and mineral belt. These belts are characterized, respectively, by folds alone; by folds with some faults; and by both folds and relatively common normal and reverse faults, and igneous plutons. Although igneous rocks are confined mainly to the mineral belt, a few small masses occur in the plain and range belt, as in the Sierra Mesquite del Sur of Coahuila. The mineral belt is a rather broad zone occupying the site of the central part of the Mexican geosyncline (Imlay, 1938). Perhaps because this zone contains a relatively thin sedimentary section, the rocks yielded to the compressive orogenic forces not only by folding but also by extensive faulting. During the late stages of the early Tertiary orogeny these highly deformed sedimentary rocks were invaded by numerous igneous masses, and in middle and late Tertiary time there were several recurrent episodes of faulting and magmatic activity. During these later episodes numerous hypa-byssal bodies were formed, and some of the magmas breached the surface. The igneous activity was accompanied by a period of extensive mineralization during which mineralizing solutions penetrated nearly every anticline in the belt. The central part of the mineral belt has played an important part in the Mexican mining industry since the early days of the Spanish conquest. The intrusive rocks occur as sills, dikes, and small to moderately large stocks. The major areas of plutonic rocks that were observed by the authors are shown in figure 61.1, but there may be other bodies not found in the reconnaissance. The plutonic rocks that were mapped show variations in abundance and a systematic change in composition from west to east, and on this basis the mineral belt and plain and range belt have been divided into three petrographic provinces, labeled western, central, and eastern. To some extent, the older ART. 61 IS U.s. GEOL. SURVEY PROF. PAPER 475-C, PAGES C7-C10. 1963 C7C8 STRUCTURAL GEOLOGY 102°00' 101“30’ 101"00' 100°30' Figure 61.1—Sketch map of the area studied, showing principal occurrences of plutonic rocks. hypabyssal bodies seem to reflect the same changes in composition as the plutonic rocks. Plutonic rocks are relatively scarce in the western province and occur mainly in the vicinity of Pico de Teyra, an isolated mountain peak that forms the most prominent landmark in this part of the mineral belt. The rocks are mainly quartz monzonite, with local monzonitic and granodioritic facies (see Rogers and others, 1961, table 12, p. 125); they also include some albitic pegmatite and aplite, and one albitite pluton measuring several kilometers in diameter. A small isolated mass in the Sierra de San Rafael about 25 kilometers north of Pico de Teyra has been classed as a latite. Mineralization is meager in this province, and there are no mines. The province contains a few small talc deposits whose origin may be related to hydro- thermal activity penecontemporaneous with the intrusion of magmas. The western province extends northward into the plain and range belt, where Imlay (1937, p. 620-622) has described several small masses of monzonite and a dike of light-colored granite, which probably should be classed as an aplite. The greatest concentration of intrusive rocks is in the central province, a major mining district that produces a variety of base and precious metals, including copper, lead, zinc, iron, silver, and gold. The largest mines are in contact-metasomatic deposits lying along or close to the margins of the three largest plutons, which have been designated the Concepcion del Oro, Providencia, and Nochebuena stocks. The rocks of these plutons contain more quartz and plagioclase than the rocks ofROGERS, VAN VLOTEN, OJEDA, TAVERA, AND DE CSERNA C9 the western province and range in composition from granodiorite to diorite. They are characterized by moderately abundant associated aplite but have no associated pegmatite or lamprophyre. The hypabyssal rocks of the central province fall within the same general compositional range and are mainly dacitic to andesitic. The igneous rocks within this province have been previously described by Rosenbusch (in Burckhardt, 1906, p. 23-2§), Bergeat (1910), Imlay (1938, p. 1670-1672), and Rogers and others (1956, p. 35-47). The plutonic rocks of the eastern province are monzo-nite and syenite (see Rogers and others, 1961; tables 13, 14, and 15); the hypabyssal rocks range from trachyte to latite and andesite. The most conspicuous pluton, the Guadalupe Garceron stock, is largely mon-zonitic in composition but is heterogeneous in appearance. It is characterized by abrupt changes in grain size, color, and the ratio of light, to dark constituents, and it is cut by numerous small leucocratic and melano-cratic dikes that range in composition from monzonitic to syenitic. The stock appears to have been formed by successive surges of magma that were increasingly syenitic. The Matehuapil stock is similar to this mass, but the Cerro Pedregoso stock is relatively homogeneous and composed largely of porphyritic syenite. The albite in these rocks occurs as independent grains as well as in microperthitic intergrowth with the orthoclase, and its relative abundance suggests that the rocks may be transitional in composition and related to the alkalic rocks described by Watson (1937) in the Sierra de San Carlos, Tamaulipas, about 150 kilometers to the east. Alkalic rocks also occur in the Tampico embayment of southern Tamaulipas, and have been described briefly by Muir (1936). The eastern province is sparsely mineralized, but it contains numerous prospects and several small mines that have produced copper and other metals. The area contains a little barite, which is being mined in the Sierra de Gomez Farias (fig. 61.1) in the northern part of the province and in the vicinity of Galeana, Nuevo Leon, about 30 kilometers to the east of the mapped area. Galeana is the major barite-producing district in Mexico. The plutons described are believed to be for the most part contemporaneous and to have been emplaced at a relatively shallow depth shortly after the folding of the Mesozoic sediments toward the end of the early Tertiary orogeny.1 They share many of the common characteristics ascribed by Buddington (1959) to plutons of the epizone. The plutons are largely discordant with the country rock which they have intruded, and i The Concepcl6n del Oro stock has been dated recently by Buseck (1962) by the- K-Ar method at about 40 million years, which would place intrusion at the end of the Eocene. some have exerted an active upward thrust that domed and in places faulted the overlying rocks. Faulting of this type can be observed around the Concepcion del Oro and Providencia stocks (fig. 61.2) and in the vicinity of Pico de Teyra (Rogers and others, 1961, pi. 1.). Other bodies, such as the Guadalupe Garceron and Matehuapil stocks, appear to have exerted a lateral pressure which in places strongly folded the adjacent country rock without faulting. EXPLANATION Upper Cretaceous Lower Cretaceous Upper Jurassic formations formations formations Predominantly shale Predominantly limestone Predominantly limestone and sandstone Figure 61.2—Structure sections through Concepcion del Oro granodiorite stock, showing reconstruction of fold. The stock has been emplaced discordantly in the core of an overturned anticline and appears to have domed and faulted the roof by upward magma pressure. The plutons have sharp contacts with the enclosing rock. They have metamorphic aureoles of varying width and intensity, and the Concepcion del Oro and Providencia stocks are bordered by intermittent skarn zones.2 Roof pendants are common in some masses. 2 Most of the plutons have intruded carbonate rocks of Late Jurassic or Early Cretaceous age, but some are in contact with Upper Cretaceous shales or with redbedis and metasediments, largely schists and phyllftes, of the basement complex.CIO STRUCTURAL GEOLOGY Many plutons are composite in character and appear to be the product of successive surges of magma, which generally were increasingly felsic. Late-stage dikes are mostly aplitic, although they include some dark rocks rich in mafic minerals. Associated hypabyssal rocks in part resemble rocks of the plutons in composition. Weakly to strongly developed primary foliation occurs in the marginal zones of some plutons. Primary linea-tion was not observed, but it might be revealed by a more detailed study of the rocks. REFERENCES Bergeat, Alfred, 1910, La granodiorita de Concepcidn del Oro en el Estado de Zacatecas y sus formaciones de contacto: Inst, geol. Mexico Bull. 27. Buddington, A. F., 1959, Granite emplacement with special reference to North America: Geol. Soc. America Bull., v. 70, no. 6, p. 671-748. Burekhardt, Carlos, 1906, Geologic de la Sierra de Mazapil et Santa Rosa: Internat. Geol. Cong., 10th, Mexico City, Guide des Excursions, no. 26. Buseck, P. B., 1962, Contact metasomatic ores at Concepcidn del Oro, Mexico [Abs.j, in Geological Society of America, Abstracts for 1961: Spec. Paper 68, p. 144. Imlay, R. W., 1937, Geology of the middle part of the Sierra de Parras, Coahuila, Mexico: Geol. Soc. America Bull., v. 48, p. 587-630. -------- 1938, Studies of the Mexican geosyncline: Geol. Soc. America Bull., v. 49, p. 1651-1694. Muir, J. M., 1936, Geology of the Tampico region, Mexico: Tulsa, Okla., Am. Assoc. Petroleum Geologists, p. 143-151. Rogers, C. L., Cserna, Zoltan de, Tavera, Eugenio, and Ulloa, Salvador, 1956, General geology and phosphate deposits of Concepcidn del Oro district, Zacatecas, Mexico: U.S. Geol. Survey Bull. 1037-A, 102 p. Rogers, C. L., Cserna, Zoltan de, van Vloten, Rogelio, Tavera, Eugenio, and Ojeda, Jesus, 1961, Reconocimiento geoldgico y depositos de fosfatos del norte de Zacatecas y areas ad-yacentes en Coahuila, Nuevo Leon y San Luis Potosi: Con-sejo de Rec. Nat. no Renovables Bull. 56, 322 p. Rogers, C. L., Cserna, Zoltan de, Ojeda, Jesus, Tavera, Eugenio, and Vloten, Roger van, 1962, Tectonic framework of an area lying within the Sierra Madre Oriental and adjacent Mesa Central, north-central Mexico: U.S. Geol. Survey Prof. Paper 450-0, p. C21-C24. Watson, E. H., 1937, Igneous rocks of the San Carlos Mountains, pt. 2 of The geology and biology of the San Carlos Mountains, Tamaulipas, Mexico: Michigan Univ. Sci. Ser., v. 12, p. 99-156.Article 62 THE ORDOVICIAN-SILURIAN CONTACT IN DUBUQUE COUNTY, IOWA By JESSE W. WHITLOW and C. ERVIN BROWN, Beltsville, Md., and Washington, D.C. Work done in cooperation with the Iowa Geological Surrey Abstract.—The lower Silurian Edgewood Dolomite discon-formably overlies the Upper Ordovician Maquoketa Shale in and near Dubuque County. Although the rocks above and below the contact are similar in lithology, this elusive contact can be recognized by a thin persistent basal conglomerate in the Silurian rocks, and locally by rare erosional remnants of the iron-rich Ned a Member of the Maquoketa Shale. The Mosalem Member of the Edgewood Dolomite of Early Silurian age disconformably overlies the Maquoketa Shale of Late Ordovician age in and near Dubuque County, Iowa. Features of the contact in the Dubuque South quadrangle and the location of outcrops of the Neda Member of the Maquoketa Shale, which occurs as sparse erosional remnants on the surface of disconformity, have been shown previously by us (1060, pi. 2). In that report we also described fragments of the Maquoketa Shale in basal Silurian beds, but did not recognize the regional distribution of the fragments. Exposures of the conglomeratic beds beyond the Dubuque South quadrangle indicate the possible use of the beds regionally as a marker for a gradational contact that is otherwise virtually undetectable. Location of outcrops of both the conglomeratic beds and the Neda Member of the Maquoketa Shale seen in Dubuque County is shown on figure 62.1. The ancient erosion surface on the Maquoketa Shale has 135 feet of relief in the Dubuque South quadrangle; therefore, at most places the thin uppermost unit, the Neda Member, has been removed, exposing the underlying Brainard Member as shown in figure 62.2. The Neda Member consists of interlayered grayish-red soft shale, dolomitic grayish-green shale, and layers of reddish-brown limonitic oolites. Irregularly shaped phos-phatic nodules that have a distinctive yellowish-brown glazed surface and contain embedded oolites are scat- tered in the oolitic layers. A 5-foot section of the Neda at locality 3 is the thickest seen by the authors. The easternmost traces of the member in Dubuque County are at localities 4 and 5, and the northernmost exposure seen is at locality 1. To the west the eroded edge of the Neda is about 2 miles southwest of locality 3, and 1 mile west of locality 2. The member was first discovered in Iowa by J. V. Howell (1916) in road ditches on Lore Hill (locality 1), about 4 miles west of the Dubuque city limits. Agnew (1955), in a study of well logs and drill cuttings from Iowa, reported other occurrences of the Neda in the subsurface. The Brainard Member at the Ordovician-Silurian contact ranges from argillaceous thin-bedded commonly fossiliferous dolomite where the upper beds are preserved, as at localities 3, 4, and 6, to soft unfossiliferous dolomitic shale where the upper beds are eroded away. The argillaceous thin-bedded dolomite of the Brainard at many places resembles the lower beds of the over-lying Mosalem Member of the Edgewood Dolomite; Calvin (1898, p. 142) referred to the lower beds in the Mosalem as “transition beds” and included them in the Maquoketa. The Mosalem Member of Early Silurian age, which consists mainly of wavy-bedded argillaceous dolomite, fills the hollows in the top of the Maquoketa Shale and consequently ranges in thickness from a few feet to as much as 94 feet in the Dubuque South quadrangle. The basal conglomeratic zone is about 1 foot thick and contains fragments of shale, dolomite, and phosphatic nodules from the Maquoketa Shale set in an argillaceous and dolomitic matrix. The matrix also contains iron sulfide, glauconite, and minor quantities of barite. The dolomite and shale fragments are mainly from the Brainard Member, and are as much as 1 inch in maxi- ART. 62 IN U.s. GEOL. SURVEY PROF. PAPER 475-C, PAGES C11-C13. 1963 694-027 0—63-----2 CllC12 STRATIGRAPHY 90°45' 1 0 1 2 MILES I .... I_______I_______I Figure 62.1.—Map showing approximate location of the Ordovician-Silurian contact and outcrops studied in Dubuque County, Iowa. N, Neda Member of the Maquoketa Shale of Late Ordovician age; C, basal conglomerate of the Mosalem Member of the Edgewood Dolomite of Early Silurian age.WHITLOW AND BROWN C13 Figure 62.2—Diagrammatic section showing stratigraphic relations at the Ordovician-Silurian contact in and near Dubuque County, Iowa. Vertical scale, 1 inch—about 200 feet. mum dimension. These fragments are difficult to see megascopically because they are texturally similar to the matrix. Phosphatic nodules and limonitic oolites derived from the Neda Member were recognized at localities 3 and 4, where the Silurian overlies the Neda, and at locality 6, where the Neda is missing. Fragments of phosphatic nodules are present not only in the basal conglomerate but are also scattered throughout the overlying few feet of dolomite. The thickness of the zone containing phosphatic debris increases with increasing thickness of the Mosalem. Locally, as at locality 6, weathered iron sulfide causes iron staining of the basal detrital beds. Glauconite was noted in drill cuttings and in unweathered outcrops of the basal Mosalem, but was not seen in weathered outcrops. Phosphatic nodules were found in the basal beds of the Mosalem in cuttings from holes drilled in the Dubuque South quadrange for the U.S. Geological Survey. Thin-bedded conglomeratic silty dolomite, about 18 inches thick, was found at the base (?) of the Mosalem Member in 1962 by Whitlow at a large roadcut along U.S. Highway 67, a quarter of a mile south of Bellevue, Jackson County, Iowa, about 21 miles southeast of Dubuque. The rock is lithologically similar to the conglomeratic layers in Dubuque County. Except for scattered phosphatic fossil fragments, the detrital character of the rock is difficult to see in outcrop but is evident on freshly broken surfaces. If this thin zone is the basal conglomerate of the Mosalem Member, the Maquoketa at Bellevue is only about 100 feet thick. This is less than reported by Kay (1935), who included in the Maquoketa Shale some of the shaly dolomite beds overlying the conglomerate. Phosphatic material and fragmental debris in the basal beds of the Silurian rocks have rarely been reported in Iowa or in northwestern Illinois. Agnew (1955, p. 1718) reported the phosphatic material in basal Silurian beds at only one place in Iowa, and Trowbridge and Shaw (1916, p. 72) described shale pebbles in basal Silurian beds at two places in northwestern Illinois. Although previous notice of these features is rare, the authors believe that the phosphatic conglomeratic beds or traces of the iron-rich Neda Member can probably be found regionally in Iowa and Illinois and can be used as a marker horizon for the Ordovician-Silurian contact either in outcrop or subsurface studies. REFERENCES Agnew, A. F., 1955, Facies of Middle and Upper Ordovician rocks of Iowa : Am. Assoc. Petroleum Geologists Bull., v. 39, p. 1703-1752. Brown, C. E., and Whitlow, J. W., 1960, Geology of the Dubuque South quadrangle, Iowa-Illinois: U. S. Geol. Survey Bull. 1123-A, 93 p. [1961]. Calvin, Samuel, 1898, Geology of Delaware County, in Annual Report, 1897, with accompanying papers: Iowa Geol. Survey, v. 8, p. 122-199. Howell, J. V., 1916, An outlier of the so-called Clinton Formation In Dubuque County, Iowa: Iowa Acad. Sci. Proc. 23, p. 121-124. Kay, G. M., 1935, Ordovician system of the Upper Mississippi Valley, in Kansas Geol. Soc. Guidebook 9th Ann. Field Conf., Iowa City, Iowa, to Duluth, Minn.: p. 281-295. Trowbridge, A. C., and Shaw, E. W., 1916, Geology and geography of the Galena and Elizabeth quadrangles: Illinois Geol. Survey Bull. 26, p. 61-82.Article 63 SPIRORBAL LIMESTONE IN THE SOURIS RIVER(?) FORMATION OF LATE DEVONIAN AGE AT COTTONWOOD CANYON, BIGHORN MOUNTAINS, WYOMING By CHARLES A. SANDBERG, Denver, Colo. Abstract.—A small outlying deposit 30 miles southeast of the main body of the marine Souris River Formation may have been laid down in the upper reaches of an estuary. Its abundant and unusual biota includes Spirorbis, fish remains, carbonized wood, plant impressions, spores, and megaspores. A thin deposit of carbonaceous spirorbal limestone and silty calcitic dolomite, containing an abundant and unusual Late Devonian flora and fauna, crops out on the steep north wall of Cottonwood Canyon on the west side of the Bighorn Mountains in sec. 34, T. 57 N., E. 93 W., Big Horn County, Wyo. (fig. 63.1). This is the first recognized occurrence of Upper Devonian rocks that are not certainly marine in the Williston basin or in adjacent parts of Montana and northern Wyoming. The deposit tentatively is considered to be part of a small outlier of the Souris River Formation, which is of early Late Devonian (Frasnian) age outside the Williston basin. The outlier extends at least as far southeast as a well drilled 15 miles from Cottonwood Canyon. The Souris River Formation has been traced from its type area in the Williston basin into south-central Montana by Sandberg (1961b, pi. 10). It consists of marine thinly interbedded shaly dolomite, argillaceous limestone, shale, siltstone, and anhydrite, and its age and stratigraphic relations in south-central Montana are the same as those of the deposit in Cottonwood Canyon. Two southward-projecting tongues of the main body of the Souris River lie about 30 miles northwest and 110 miles northeast of Cottonwood Canyon (fig. 63.1). At Cottonwood Canyon, the Souris River( ?) Formation underlies a cliff of massive dolomite, 25 feet high, which is at the base of the Jefferson Formation of early Late Devonian (Frasnian) age, and overlies thin ledges of white dolomite at the top of the Bighorn Dolomite of Late Ordovician age. The Souris River(?) has a maximum observed thickness of 16 feet, of which the lower 11 feet forms a partly covered slope and the upper 5 feet forms weakly resistant ledges. The upper part originally was described by Blackstone and Mc-Grew (1954, p. 39) as a 6-foot bed of “reddish weathering pyritic limestone” and black shale 50 feet above the base of the Jefferson. However, regional correlation of the Jefferson between its type section at Logan, in southwestern Montana, and 12 measured sections in the Bighorn and Pryor Mountains (Sandberg, 1961b, pi. 10) demonstrates that the 25-foot-thick massive dolomite is the basal bed of the Jefferson. Blackstone and McGrew (1954) noted the slight disconformity between it and the underlying pyritic limestone but placed the base of the Jefferson much lower. It is likely, however, that they did not observe the ledges of white Bighorn Dolomite about 10 feet below the pyritic limestone, because the pre-Devonian erosion surface is very irregular and drops sharply eastward from the outcrop of the Souris River(?) Formation. Just 225 feet to the east, the eroded top of the Bighorn is 135 feet lower stratigraphically. The following section of Souris River (?) Formation is exposed for a distance of 75 feet along the canyon wall between talus slides to the east and west. Results of calcium-magnesium analyses of 13 selected samples by James A. Thomas are incorporated in the lithologic descriptions. ART. 63 IN U.S. GEOL. SURVEY PROF. PEPER 475-C, PAGES C14-C16. 1963. C14SANDBERG CIS Devonian: Jefferson Formation. Disconf ormity. Thickness Souris River (?) Formation: (.feet) Dolomite, calcitic, carbonaceous, pale-yellowish-brown, pale-brown, and light-brownish-gray, microcrystalline, slightly silty. Thin bedded to platy but thinly laminated and fissile at base. Contains abundant spores, megaspores, fish plates, and carbonized plant fragments and large flattened stems. Grades eastward to medium-bedded brownish-gray to brownish-black carbonaceous dolomitic limestone containing pyrite concretions and scattered white coiled worm tubes of Spirorbis sp. about 1 mm in diameter. Weathers to yellowish-gray or grayish-orange smooth, powdery surface; forms reentrant ___________________________________________ 1% Limestone, spirorbal, medium-dark-gray, medium to coarsely fragmental, slightly silty, slightly dolomitic, pyritic. Oxidation of pyrite to hematite colors rock pale brown or grayish red to depth of 14 to % in. from weathered surfaces. Composed largely of Spirorbis tubes % to 1 mm in diameter. Interbedded with thin lenses of fissile moderate-yellowish-brown and grayish-brown microcrystalline silty calcitic dolomite containing scattered Spirorbis tubes. Contains carbonized macerated plant remains and large flattened stems as much as 2 in. wide, fish plates and teeth, and rounded pebbles of carbonized wood as much as 1 in. in diameter. Pyrite coats many Spirorbis tubes and partly replaces some carbonized-wood pebbles. Weathers pale brown and yellowish gray; thin bedded to thinly laminated; forms weakly resistant ledges_________________________________________ 2% Limestone, carbonaceous, argillaceous, dark-gray and brownish-gray, microcrystalline to cryptocrystalline. Contains spores, megaspores, carbonized macerated plant remains, and scattered white Spirorbis tubes. Weathers to very-light-gray or yellowish-gray smooth, rounded surface; medium bedded ; forms weakly resistant ledge____________ 1 Dolomite, calcitic, carbonaceous, silty, dark-yellowish-brown, pale-brown, and grayish-brown, micro-crystalline, earthy, porous, friable. Contains fish plates and teeth, carbonized plant remains, spores, megaspores, and scattered Spirorbis tubes. Weathers yellowish gray; thin bedded to laminated ; nonresistant; forms partly covered slope with 1%-foot-thick ledge near middle_____________ 9 Dolomite, silty, pale- to dark-yellowish-brown and yellowish-gray mottled with pale-yellowish-brown; in part carbonaceous. Dark-yellowish-orange jiorous finely to very finely crystalline rhombic sandy dolomite at base. Contains fish plates and teeth, spores, megaspores, and carbonized plant stems. Weathers to yellowish-gray or yellowish-orange smooth surface; thin bedded ; forms weakly resistant ledge______________________________________ 2 Total Souris River ( ?) Formation____________ 16 Unconformity. Ordovician: Bighorn Dolomite. 110° 109° 108° 107° 106° Figure §3.1.—Index map showing distribution of Souris River Formation and equivalents (stippled) in southern Montana and northern Wyoming. The Souris River(?) Formation has a minimum width of 350 feet at Cottonwood Canyon, as indicated by blocks of spirorbal limestone found in the talus slide for 275 feet west of its outcrop. At the east side of the other bounding talus slide, 225 feet east of the outcrop, the Beartooth Butte Formation crops out slightly below the basal cliff of the Jefferson. The Beartooth Butte in Cottonwood Canyon is of Early Devonian age (Black-stone and McGrew, 1954; D. H. Dunkle, written communication, Oct. 14,1959). It consists largely of coarse dolomite conglomerate and fills a deep channel in the underlying Bighorn at this locality (Sandberg, 1961a, p. 1305). The Beartooth Butte occupies most of the 150-foot interval between the Jefferson and Bighorn. However, the upper 20 feet of this interval, into which the Souris River (?) Formation might extend, and the basal 5 feet of Jefferson are covered. Reconnaissance of the canyon for about 4 miles to the east failed to disclose additional exposures of the Souris River (?) Formation. The most abundant fossil in the Souris River(?) Formation at Cottonwood Canyon is Spirorbis sp. Tubes of this polychaete worm, about y2 to 1 mm in diameter, are abundant throughout. Some beds of limestone, particularly those near the top of the deposit, are composed almost exclusively of whole and fragmentary worm tubes and are termed spirorbal limestone (fig. 63.2). The deposit also contains fish remains, rounded pebbles of carbonized wood, impressions of large plant stems, macerated plant remains, spores, and megaspores. Fish remains include the antiarch Both-riolepis cf. B. coloradensis Eastman, palaeoniscoid teeth cf. Rhadinichthys sp., coccosteid plates, and weathered heterostracan carapaces (F. C. Whitmore, Jr., written communication, July 14,1961), and the crossopterygian Holoptychius cf. H. giganteus Eastman (D. H. Dunkle, written communication, Feb. 5,1963). Callixylon-typeClfi STRATIGRAPHY Figure 63.2—Spirorbal limestone near top of Souris River (?) Formation. Cottonwood Canyon. Big Horn County, Wyo. Thin section (XlO) parallel to bedding, in plain transmitted light. Spirorbis tubes are round or crescent shaped and are filled by secondary caleite. Black outlines on some tubes are pyrite coatings. pitting of tracheids was observed in the wood fragments by R. A. Scott (oral communication, Jan. 1961). The microflora consists almost entirely of a single spore species, Punctatisporites cf. P. planus Hacquebard, but it includes a few individuals of many diverse palyno-morph species (R. H. Tschudy, written communication, Feb. 14, 1962). The fish and spores indicate an early Late Devonian age. This biota suggests a marginal marine environment but does not indicate whether deposition was in brackish or fresh water. The diverse microflora is related to land plants and does not include any exclusively marine forms. On the other hand, the fish include both marine and fresh-wrater forms. Holoptychius is believed to have lived in fresh water, although some remains have been found in shallow inshore marine sediments (D. H. Dunkle, written communication, Feb. 5, 1963). The other fish also do not give a clear paleoecologic picture, according to F. C. Whitmore, Jr. (written communication, July 14, 1961), who stated: The Heterostraci lived In the seas or in lower reaches of streams; palaeoniscoids and euarthrodires lived in both marine and fresh water. Botliriolepis was a fresh-water form, but the single plate in this collection could have been washed into a marine deposit. The abundance of Spirorbis sp. accords with a marginal marine environment, for Spirorbis sp. apparently was adapted to fresh water as well as to salt and brackish water (Pruvost, 1930). The main body of the marine Souris River Formation was deposited as the shallow Late Devonian sea transgressed with minor pulsations westward and southward from the Williston basin into southern Montana (Sandberg, 1961b). Both the regional paleogeography and the mixed paleoecology of the fish fauna in the Cottonwood Canyon deposit suggest that rocks of this outlier were deposited in brackish or nearly fresh water in the upper reaches of a long, narrow estuary that extended into the retreating shoreline. REFERENCES Blackstone, D. L., Jr., and McGrew, P. O., 1954, New occurrence of Devonian rocks in north central Wyoming, in Billings Geol. Soc. Guidebook, 5th Ann. Field Conf., Pryor Mountains and northern Bighorn Basin: p. 38-43. Pruvost, Pierre, 1930, La faune continentale du terrain houiller de la Belgique: Inst, royal colonial beige, sec. sci. nat. et med., Mem. 44, p. 103-282. Sandberg, C. A., 1961a, Widespread Beartooth Butte Formation of Early Devonian age in Montana and Wyoming and its paleogeographic significance: Am. Assoc. Petroleum Geologists Bull., v. 45, p. 1301-1309. ------ 1961b, Distribution and thickness of Devonian rocks in Williston basin and in central Montana and north-central Wyoming: U.S. Geol. Survey Bull. 1112-D, p. 105-127. [1962]Article 64 DARK SHALE UNIT OF DEVONIAN AND MISSISSIPPI AGE IN NORTHERN WYOMING AND SOUTHERN MONTANA By CHARLES A. SANDBERG, Denver, Colo. Abstract.—A marine sequence of black and moderate-yellowish-brown quartzose shale and carbonaceous silts tone discon-formably underlies the Madison Limestone (Mississippian) and is separated from underlying Upper Devonian rocks by a regional unconformity. The stratigraphic relations, thickness, lithologic character, and age of this dark shale unit approximate those of the Englewood Formation. A clastic marine sequence of black and moderate-yellowish-brown quartzose shale and carbonaceous silt-stone that unconformably overlies Devonian rocks and disconformably underlies the Madison Limestone of Mississippian age in northern Wyoming and southern Montana is here informally termed a dark shale unit of Devonian and Mississippian age. This dark shale unit is the homotaxial equivalent of the Englewood Formation of the Black Hills, South Dakota and Wyoming. The reference locality of the dark shale unit is in the eastern Beartooth Mountains at the mouth of Clarks Fork Canyon, north of the river, in the NE^NE^ sec. 7, T. 56 N., R. 103 W., Park County, Wyo., in the Deep Lake 15-minute quadrangle. It is approached on Wyoming State Highway 120, which extends north from Cody, Wyo. About 2y2 miles beyond the Clarks Fork bridge, which is 26 miles from Cody, a trail easily traversed by automobile connects the highway to the mouth of the canyon, about 7 miles west. From this trail the reference section is reached by climbing a talus slope between the Jefferson Formation of Late Devonian age and the Madison Limestone, which dip 50° to 55° NE. At the head of this talus slope the following section of the dark shale unit is well exposed along a steep bench below a high cliff formed by the Madison. The dark shale unit at most measured sections consists of dark-gray to black carbonaceous dolomitic quartzose shale and light-olive-gray, yellowish-brown, yellowish-gray, and dark-gray dolomitic siltstone that grade to very shaly and very silty dolomite. Commonly the unit forms partly covered slopes and weakly to moderately Reference section of the dark shale unit of Devonian and Mississippian age in the NE%NEy± sec. 7, T. 56 N., R. 103 TV., Park County, Wyo. Thickness Dark shale unit: (Feet) Siltstone, dolomitic, greenish-gray, light-olive-gray, dark-gray, and dark-yellowish-orange, fissile, partly carbonaceous, nodular, very fine grained; grades to dolomitic quartzose shale. Botryoidal nodules, about 1 in. in diameter, are white crystalline quartz geodes with vugs and cracks partly filled by white calcite and grayish-red hematite. Contains several conodont species that are characteristic of the Lower Carboniferous of Europe (B. F. Glenister, written communication, Mar. 2, 1961). Weathers very dusky red purple, olive gray, light brown, and yellowish brown. Interbedded with fissile siltstone are 2 thick-bedded lenses, 0 to 4 ft thick, of slightly nodular coarser grained limonitic dolomitic siltstone that is light olive gray and moderate yellowish brown mottled with pale red purple, greenish gray, and pale reddish brown. Unit is weakly resistant and slope forming, except for lenses which form moderately resistant ledges, about 4 ft and 11 ft below top______________ 19 Siltstone, dolomitic, limonitic, hematitic; dark yellowish orange to moderate yellowish brown, with grayish-red streaks and mottles in upper half; medium dark gray and olive gray, with yellowish-brown and dark-yellowish-orange laminae in lower half. Weathers moderate yellowish brown, dark yellowish orange, and light brown; upper half massive; lower half medium bedded ; resistant; ledge forming____________ 8 Siltstone, dolomitic, carbonaceous, medium-dark-gray to dark-gray; grades to dolomitic quartzose shale. Weathers yellowish gray and dark yellowish orange; thick bedded to laminated; moderately resistant; ledge forming_______________________________________ 8 Shale, quartzose, dolomitic, carbonaceous, dark-gray to grayish-black. Contains abundant palynomorphs, including hystricosphaeres, Micrhystridium, Tas-manites, and a spore, Leiosphaera sp., which are indicative of marine deposition (R. H. Tschudy, written communication, Feb. 14, 1962). Upper 8 ft weathers dark gray to medium dark gray mottled locally with light olive gray; lower 5 ft weathers light olive gray; weakly resistant; slope forming_________________________ 13 Total, dark shale unit_______________________________ 48 ART. 64 I\ U.S. GEOL. SURVEY PROF. PEPER 475-C, PAGES C17-C20. 1963. C17C18 STRATIGRAPHY resistant ledges, but locally it forms the base of a cliff of Madison. At some localities it consists entirely or partly of microcrystalline to very finely crystalline hematitic slightly silty dolomite that is yellowish gray mottled with grayish red and contains scattered fine to coarse sand grains and fish plates and teeth. Thin partings of dark-gray carbonaceous shale commonly occur in this dolomitic faoies. The base of the dark shale unit locally is a bed of phosphatic quartzitic sandstone, as much as 6 inches thick, containing abundant large cono-donts, fish plates and teeth, black phosphate pellets, and angular pebbles derived from the underlying rocks. The dark shale unit is present in north-central and northwestern Wyoming and south-central and southwestern Montana (fig. 64.1). It crops out in the Bear-tooth and northern Bighorn Mountains, in the Wind River, Teton, Absaroka, Gallatin, and Bridger Ranges, and in the Horseshoe Hills, near Logan, Mont. The dark shale unit is 48 feet thick at the reference locality. From there its thickness increases northeastward to a maximum of about 70 feet in the subsurface north of Billings, Mont. The thickness decreases northwestward from the reference locality to about 12 feet in the western Beartooth Mountains and to about 4 feet in the Gallatin and Bridger Ranges and Horseshoe Hills. The unit also thins southward and eastward; its thickness is about 10 feet in the Bighorn Mountains and in the Absaroka, Wind River, and Teton Ranges. At Logan, Mont. (fig. 64.1), the dark shale unit over-lies the Sappington Sandstone Member at the top of the Three Forks Formation of Devonian and Missis-sippian age. There the dark shale unit previously was correlated by Knechtel and others (1954), with the Little Chief Canyon Member of the Lodgepole Limestone of Early Mississippian age, whose type locality is in the Little Rocky Mountains of north-central Montana. Regional stratigraphic studies by the author indicate, however, that the southern limit of the Little Chief Canyon lies very near the Little Rocky Mountains and 160 miles northeast of Logan, so the continued use of this name at Logan is inappropriate. From Logan southeastward to central Wyoming, the dark shale unit truncates progressively older rocks ranging in age from earliest Mississippian to early Late Figure 64.1.—Outcrops of the dark shale unit. Numbers indicate thickness, in feet, of measured sections. Approximate limit of unit shown by hachured line; unit absent on unhachured side.SANDBERG C19 Devonian (fig. 64.2). These rocks include all three members of the Three Forks Formation and the subsurface Birdbear Formation and upper part of the Dupe-row Formation and their equivalents in the outcropping Jefferson Formation. At its reference locality the dark shale unit overlies the lower, evaporitic member of the Three Forks, but only about 10 miles north, south, and east of this locality it rests on the Jefferson. Figure 64.2.—Age and stratigraphic relations of dark shale unit and adjacent formations. (Ammonoid zones after House, 1962 ; Collinson and others, 1962.) The dark shale unit is disconformably overlain by the Madison Limestone, which truncates it west of Logan in southwestern Montana, in southeastern Montana, and in northeastern and central Wyoming (fig. 64.1). Conodont collections made independently by Gilbert Klapper and the author from the basal few feet of the Madison at 14 measured sections in Montana and Wyoming indicate an Early Mississippian age, equivalent to the Pericyclm-Q t u f e (culla) of the Lower Carboniferous in Europe (Klapper, written communication, Oct. 12,1962). The age of the dark shale unit ranges from very late Devonian, equivalent to the Clynrnenia-Stufe (toY) of the Famennian Stage in Europe, to earliest Mississippian, equivalent to the upper part of the Gattendorficir-Stufe (cul) of the Tournaisian Stage of Lower Carboniferous in Europe (fig. 64.2). This precise determination is based largely on independent conodont collections of Klapper and the author. The conodonts were identified by Klapper (written communication, Oct. 12, 1962), who plans to publish systematic descriptions of the conodont faunas of the dark shale unit and adjacent formations. In the Absaroka and Teton Ranges, Beartooth Mountains, and areas to the northwest (fig. 64.1), the dark shale unit appears to be entirely Early Mississippian. This determination is supported by conodonts collected from the basal sandstone of the dark shale unit at Baker Mountain by the author and previously determined to be of Early Mississippian (Kinderhook) age by W. H. Hass (written communication, Nov. 1,1957). In the Bighorn Mountains (fig. 64.1) the dark shale unit contains beds of very late Devonian as well as of earliest Mississippian age. Conodonts collected from the unit at South Fork Rock Creek by the author were determined to be of Early Mississippian (Kinderhook) age by W. H. Hass (written communication, Nov. 1, 1957). Conodonts from a thin bed directly below the Madison at Little Tongue Canyon were determined to include a mixed fauna of Upper Devonian and basal Kinderhook forms by Koucky and others (1961). ThF bed is here included at the base of the dark shale unit. Conodonts from thin beds of black shale at the top of the Devonian sequence in Cottonwood Canyon were reported to be Upper Devonian by Blackstone and Mc-Grew (1954). Conodonts from these same beds were equated to the Clymenm-Stnie (toY) by Ethington and others (1961). These shale beds are here included in the lower part of the dark shale unit. In the upper beds of the dark shale unit at Cottonwood Canyon, Klapper (written communication, Oct. 12,1962) found Lower Carboniferous (cul) conodonts. In the Wind River Range (fig. 64.1), south of other outcrops of the dark shale unit, the unit appears to be entirely of very late Devonian age. Upper Devonian (toY) conodonts were reported from the top of the Darby Formation at two localities by Klapper (1958). However, Klapper (written communication, Oct. 12, 1962) now recognizes the conodont-bearing beds as the dark shale unit. The dark shale unit of Devonian and Mississippian age is the homotaxial equivalent of the Englewood Formation, formerly Englewood Limestone, of the Black Hills of South Dakota and extreme northeastern Wyoming. Their lithologic character, stratigraphic relations, thickness, and age are very similar.C20 STRATIGRAPHY The Englewood Formation contains variegated silt-stone, shale, and sandstone, some grayish-red to pinkish-gray silty dolomite, similar to the dolomitic facies of the dark shale unit, and a little pinkish-gray and medium- to light-gray limestone. At its type locality near Englewood, S. Dak., the upper part of the Englewood Formation is entirely dolomite, as determined by calcium-magnesium analyses, and the lower part is silt-stone. The Englewood Formation underlies the Pahasapa Limestone of Mississippian age, which is equivalent to a large part of the Madison Limestone, and is separated from the Pahasapa at some localities in the northern Black Hills by an angular unconformity. Like the dark shale unit, the Englewood truncates progressively older beds southward. These range from the Three Forks Formation of Late Devonian age in the subsurface northeast of the Black Hills (Sandberg, 1961) to the Deadwood Formation of Late Cambrian age at the south end of the Black Hills. The thickness of the Englewood Formation ranges from 18 to 54 feet at 6 sections measured by the author in various parts of the Black Hills. The Englewood Formation, like the dark shale unit, contains conodont faunas equated to both the Upper Devonian (toY) and the Lower Carboniferous of Europe (Klapper and Furnish, 1962). On the basis of their age determinations in the Black Hills and the author’s stratigraphic correlation between the Englewood Formation at Coldbrook Canyon in the south end of the Black Hills and equivalent rocks at Sand Canyon in the Hartville area, from which Love and others (1953) reported a Late Devonian brachiopod fauna, the age of the Englewood Formation is here recognized as Late Devonian and Early Mississippian. The dark shale unit of Devonian and Mississippian age and the Englewood Formation are now separate rock bodies. They were deposited penecontempor-aneously in two elongate, arcuate, shallow marine basins, each characterized by a northeast-trending axis, along which a narrow belt of thin sediments accumu- lated. Although these depositional basins are now only about 30 miles apart, it would be difficult to demonstrate that they had once coalesced. Any vestige of very thin deposits, which might have been laid down on a connecting shelf area, has been removed by slight pre-Madison erosion in the intervening Powder River Basin of northeastern Wyoming. REFERENCES Blackstone, D. L., Jr., and McGrew, P. O., 1954, New occurrence of Devonian rocks in north central Wyoming, in Billings Geol. Soc. Guidebook 5th Ann. Field Conf., Pryor Moun-tains-Northem Bighorn Basin, Montana: p. 38-43. Collinson, Charles, Scott, A. J., and Rexroad, C. B. 1962, Six charts showing biostratigraphic zones, and correlations based on conodonts from the Devonian and Mississippian rocks of the upper Mississippi Valley: Illinois Geol. Survey Circ. 328,32 p. Ethington, R. L., Furnish, W. M., and Wingert, J. R., 1961, Upper Devonian conodonts from Bighorn Mountains, Wyoming: Jour. Paleontology, v. 35, No. 4, p. 759-768. House, M. R., 1962, Observations on the ammonoid succession of the North American Devonian: Jour. Paleontology, v. 36, No. 2, p. 247-284. Klapper, Gilbert, 1958, An Upper Devonian conodont fauna from the Darby formation of the Wind River Mountains, Wyoming : Jour. Paleontology, v. 32, No. 6, p. 1082-1093. Klapper, Gilbert, and Furnish, W. M., 1962, Devonian-Mississip-pian Englewood Formation in Black Hills, South Dakota: Am. Assoc. Petroleum Geologists Bull., v. 46, No. 11, p. 2071-2078. Knechtel, M. M., Smedley, J. E., and Ross, R. J., Jr., 1954, Little Chief Canyon Member of Lodgepole Limestone of Early Mississippian age in Montana : Am. Assoc. Petroleum Geologists Bull. v. 38, No. 11, p. 2395-2411. Koucky, F. L., Cygan, N. E., and Rhodes, F. H. T., 1961, Conodonts from the eastern flank of the central part of the Big Horn Mountains, Wyoming, in Paleontological notes: Jour. Paleontology, v. 35, No. 4, p. 877-879. Love, J. D., Henbest, L. G., and Denson, N. M., 1953, Stratigraphy and paleontology of Paleozoic rocks, Hartville area, eastern Wyoming: U.S. Geol. Survey Oil and Gas Inv. Chart OC-44. Sandberg, C. A., 1961, Distribution and thickness of Devonian rocks in Williston basin and in central Montana and north-central Wyoming: U.S. Geol. Survey Bull. 1112-D, p. 105-127 [1962].Article 65 NOMENCLATURE FOR LITHOLOGIC SUBDIVISIONS OF THE MISSISSIPPIAN REDWALL LIMESTONE, ARIZONA By EDWIN D. McKEE, Denver, Colo. Abstract.—New formal names given to the principal subdivisions of the Mississippian Redxvall Limestone in northern Arizona are, in ascending order: Whitmore Wash Member, Thunder Springs Member, Mooney Falls Member, and Horseshoe Mesa Member. The Redwall Limestone of Mississippian age extends across much of northern Arizona and forms in outcrop one of the thickest, most massive cliffs of the region. Although superficially it seems to consist of a single uniform lithology, detailed examination indicates that it is readily divisible into four distinct lithologic units. These units have been recognized in the literature and referred to as informal members for more than 20 years. Various informal designations have been applied to them as follows (ascending order) : I, II, III, and IV (Gutschick, 1943, p. 5; Easton and Gutschick, 1953, p. 3) ; bottom, lower middle, upper middle, and top (McKee, 1958, p. 75) ; and A, B, C, and D (McKee, 1960b). The informal designations A, B, C, and D have been used in well-log descriptions by the American Stratigraphic Co. since 1960. The application of formal names to members of the Redwall Limestone is desirable at this time because the validity of the units has been tested and established through a detailed study of the formation by E. D. McKee and R. C. Gutschick (report in preparation) and also because these units frequently are referred to in the petroleum industry, as a result of recent discoveries of oil and.gas in them. The following names are here given to the four members of the Redwall, listed from oldest to youngest (fig. 65.1) : Whitmore Wash Member, Thunder Springs Member, Mooney Falls Member, and Horseshoe Mesa Member. These names are taken from geographic localities in which the rocks are well exposed and in which detailed sections have been measured; thus the names refer to appropriate type sections. The Whitmore Wash Member, at the base of the formation, is exposed along the east side of the valley of that name in northwestern Arizona. The type section, 101 feet thick, is on the upthrown side of the Hurricane fault, about % mile north of the Colorado River, where the member consists of very thick bedded (4 to 15 feet) dolomite that is verj fine and even grained. In sections to the southeast along the south side of Grand Canyon, the member ranges in thickness from 72 to 85 feet and is also largely dolomite, but farther west it is limestone composed mostly of well-rounded bioclasts and locally of ooids. Medium-scale crossbedding is conspicuous in a few places. The Thunder Springs Member of the Redwall Limestone consists of thin beds of chert alternating with thin beds of carbonate rock. The type section, 138 feet thick, is in the cliff of Redwall Limestone west of the springs at the head of Thunder River, about 2 miles north of the Colorado River in central Grand Canyon. Chert beds are the most conspicuous feature of this member (McKee, 1960); in the western part of Grand Canyon they are associated with limestone, but to the east with dolomite. Grains within the limestone range in size from fine to very coarse and consist of both bioclasts and peloids (intraformational clastic particles), commonly in an aphanitic calcite matrix. Uniformly fine grains are characteristic of the dolomite. The Mooney Falls Member is the thickest, most massive, and most prominent unit of the Redwall Limestone. It forms a sheer cliff throughout the Grand Canyon and from east to west ranges in thickness from ART. 65 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C21-C22. 1963. C21C22 STRATIGRAPHY Figure 65.1.—Section showing stratigraphic relations and thickness of the members of the Redwall Limestone along the Bright Angel Trail, Grand Canyon. about 200 feet to more than 350 feet. Its type section, 312 feet thick, is at Mooney Falls in Havasu Canyon, a major tributary of the Grand Canyon from the south. This member forms most of the walls of the narrow gorge in which the spectacular waterfalls are located, and it is well exposed for many miles in that area. Mostly the Mooney Falls Member is a very pure limestone remarkably free of all detrital matter, but locally, as in the lower part of the type section, it contains dolomite. Detailed study shows that particle size ranges widely from bed to bed and that a well-defined cyclic sequence of texture can be traced across the area (McKee, 1960). The thick, massive beds within this member locally contain abundant coral heads (Syringopora), as at the type section. Zones of certain colonial corals and of endothyrids are well marked and persistent in the upper half. Thin chert beds are common near the top, and large-scale crossbedding is well developed in some areas. Oolitic units occur in the upper part of the member in western Grand Canyon. The uppermost unit, the Horseshoe Mesa Member, is relatively thin bedded and in Grand Canyon commonly forms receding ledges at the top of the great cliff of the Redwall. The thickness of this unit has been much affected locally by pre-Supai erosion, but in general it is thin, ranging from 38 feet at the type section to 125 feet in the west. A type section has been selected at Horseshoe Mesa below Grandview Point on the south side of the Grand Canyon. The section is accessible from the Grandview trail, which crosses Horseshoe Mesa. At this locality, as in most of the region, the member consists largely of aphanitic limestone; encrusting and sediment-binding algal structures are common. The member is characterized by features typical of quiet-water conditions, although crossbedding, ripple marks, and oolite beds occur locally. REFERENCES Easton, W. H., and Gutschick, R. C., 1953, Corals from the Redwall Limestone (Mississippian) of Arizona: Southern California Acad. Sci. Bull., v. 52, pt. 1, p. 1-27. Gutschick, R. C., 1943, The Redwall Limestone (Mississippian) of Yavapai County, Arizona : Plateau, v. 16, no. 1, p. 1-11. McKee, E. D., 1958, The Redwall Limestone: New Mexico Geol. Soc., 9th Field Conf., p. 74-77. ------ 1960a, Cycles in carbonate rocks, in the Bradley volume: Am. Jour. Sci., v. 258-A, p. 230-233. ------ 1960b, Lithologic subdivisions of the Redwall Limestone in northern Arizona—their paleogeographic and economic significance: Art. 110 in U.S. Geol. Survey Prof. Paper 400-B, p. B243-B245. ------ 1960c, Spatial relations of fossils and bedded cherts in the Redwall Limestone, Arizona: Art. 210 in U.S. Geol. Survey Prof. Paper 400-B, p. B461-B463.Article 66 MISSISSIPPIAN ROCKS IN THE LARAMIE RANGE, WYOMING, AND ADJACENT AREAS By EDWIN K. MAUGHAN, Denver, Colo. Abstract.—Mississippian strata in the Laramie Range equivalent to the Madison Limestone are composed, in ascending order, of arkosic sandstone, limestone, and a unit containing fossilifer-ous chert lenses. These rocks are 500 feet thick west of Casper but wedge out southward by convergence of the basal sandy unit and upper elierty unit. This convergence suggests an actively rising positive area farther south in Colorado during Mississippian time. Mississippian strata in the Laramie Range equivalent to the Madison Limestone are composed of a basal unit of arkosic conglomeratic sandstone; a middle unit of limestone; and an upper unit that contains yellowish-gray tabular fossiliferous chert lenses in a matrix of yellow siltstone in the southern part of the range, and limestone in the north (fig. 66.1). The basal sandstone is gradational with the overlying limestone unit, but the upper cherty unit seems to rest sharply upon the limestone. These rocks are 500 feet thick a short distance west of Casper but wedge out southward along both flanks of the Laramie Range. Similar rocks occur also in the subsurface in southeastern Wyoming and adjacent parts of northeastern Colorado and western Nebraska, as indicated on figure 66.2. Thickness measurements given on figure 66.2 were obtained from sections measured by the author south of Marshall and Horseshoe Creek in the Laramie Range, and in the Medicine Bow Mountains; from subsurface logs of the American Stratigraphic Co., Denver, Colo., and Casper, Wyo.; from measured sections in many unpublished theses, mostly from the University of Wyoming; and from measured sections in published reports listed in the references at the end of this report. The southernmost exposure of Mississippian rocks known in the Laramie Range is on the east flank of the range, at the North Fork of Crow Creek, Laramie County, Wyo. (Maughan and Wilson, 1960). On the west flank of the range, the Madison extends southward as far as Garrett, and possibly is continuous as far as the vicinity of Wheatland Reservoir (figs. 66.1 and 66.2). Isolated outcrops are found farther south, in secs. 13 and 9, T. 17 N., R. 72 W. Reworked fragments of Mississippian rock in the basal (Pennsylvanian) part of the younger Casper and Fountain Formations have been reported as far south as Bellevue, Colo., on the east side of the range (Henderson, 1909, p. 158-159), and at Rogers Canyon, central Albany County, Wyo., on the west flank of the range (Knight, 1929, p. 29-30). Fragments of Mississippian rock, mostly yellow-gray chert, are common in the basal part of the Casper Formation north of Rogers Canyon. The basal sandstone unit in the Laramie Range is composed generally of arkosic conglomeratic sandstone that by some geologists has been correlated with the Flathead Quartzite or Dead wood Formation of Cambrian age. Well-rounded to subangular quartz and feldspar pebbles are common in a matrix which is dominantly quartz sand, but which includes silt- and clay-sized particles. Grains of feldspar and ferro-magnesian minerals are also abundant. This arkosic sandstone was derived from weathering of the underlying Precambrian granite and metamorphic rocks. Its thickness at many places is indicated on figure 66.2. The unit commonly grades upward into moderately well sorted quartzose sandstone which, in turn, is increasingly calcareous upward and intertongues and grades into the overlying limestone unit. Laterally the sandstone intertongues and grades into limestone, as illustrated on figure 66.1. Locally, where the sandstone is moderately thick, as in the area a few miles southeast of Marshall, it is very well sorted orthoquartzite. Fossils of Kinderhook age, “notably Chonopectm fischeri, were collected 13 feet above granite at the top of calcareous arkose” (Agatston, 1954, p. 514) in the ART. 66 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C23-C27. 1963. C23C24 STRATIGRAPHY SOUTH A Wheatland Reservoir I-----------16 miles NORTH A Garrett Little Medicine Marshall area area 8 miles------1-—2 miles—|-----13 miles-----1—3 miles—| ro Z o > 0) li. > rc Z o -C KJ T> 0 O sz a (0 O r. iles nort Camero fl * ro «/) o 0) c £ c o CL JC * o ■O 0) W) d> g t o o E 1° 0) > 8 . 1-17. Conant, L. C., and Swanson, V. E., 1961, Chattanooga Shale and related) rocks of central Tennessee and nearby areas: U.S. Geol. Survey Prof. Paper 357, p. 1-91. Conrad, S. G., Elmore, R. T., and Maher, S. W., 1957, Stratigraphy of the Chattanooga Shale in the Flynn Creek structure, Jackson County Tennessee: Tennessee Acad. Sci. Jour., v. 32, no 1, p. 9-18. Hass, W. H., 1956, Age and correlation of the Chattanooga Shale and the Maury Formation: U.S. Geol. Survey Prof. Paper 286, p. 1-47. Pulse, R. R., and Sweet, W. C., 1960, The American Upper Ordovician standard. III. Conodonts from the Fairview and McMillan Formations of Ohio, Kentucky, and Indiana : Jour. Paleontology, v. 34, no. 2, p. 237-264. Sweet, W. C., Turco, C. A., Warner, Earl, Jr., and Wilkie, L. C., 1959, The American Upper Ordovician standard. I. Eden conodonts from the Cincinnati region of Ohio and Kentucky: Jour. Paleontology, v. 33, no. 6, p. 1029-1068. Wilson, C. W., Jr., and Born, K. E., 1936, The Flynn Creek disturbance, Jackson County, Tennessee: Jour. Geology, v. 44, no. 7, p. 815-835.Article 75 MIDDLE TRIASSIC MARINE OSTRACODES IN ISRAEL 1 By I. G. SOHN, Washington, D.C. Work done in cooperation with the Geological Survey of Israel Abstract.—The first Middle Triassic marine ostracodes in the world are recorded from the Makhtesh Ramon, in southern Israel. Their age is established by associated megafossils. This is a preliminary note recording the presence of well-preserved ostracodes in Middle Triassic marine sedimentary rocks of the Makhtesh Ramon, in southern Israel. The ostracodes occur in the vicinity of Har Gevanim, in the same beds from which Parnes (1962) described an ammonite fauna. On the basis of the megafossils, these beds belong to the Anisian and Ladinian Stages (Muschelkalk). The location of the outcrops is shown on figure 75.1. Definite Middle Triassic marine ostracodes are as yet undescribed anywhere in the world, and only a few genera of marine Triassic ostracodes are known (see accompanying table). Range of Triassic marine ostracode genera [Data from Kollmann, 1960; Moore, 1961] Pre- Tri- assic Triassic Post- Triassic Lower Middle Upper Bairdia McCoy, 1844 X X X Bairdiocypris Kegel, 1932 X X ? Bythocypris Brady, 1880 X X Cytherella Jones, 1849 Cytherissinella Schneider, 1956_ X Fabalicypris Cooper, 1946 X ? Gemmanella Schneider, 1956__ X Glyptobairdia Stephenson, 1946 cf. X X X Healdia Roundy, 1926 _ X Monoceratina Roth, 1928 _ X X X Ogmoconcha Triebel, 1941 X X cf. X Paracytheridea Muller, 1894 X Ptychobairdia Kollmann, 1960 - X X X Renngartenella Schneider, 1957. 1 Sponsored by Smithsonian Institution grant NSF-G-24305. The genera Bairdia and Monoceratina are presumed to have lived in the Triassic because species are assigned to them from both pre- and post-Triassic sedimentary rocks. Cryptophyllus Levinson, 1951, is also in this category, but Jones (1962, p. 5) suggests that the Upper Jurassic (Redwater Shale) species may have been reworked from older, possibly Paleozoic, beds. Kollmann (1960, p. 87) also lists small specimens of Kirkby-idae, a Paleozoic family, in sample 214/2 from Lanzing, Austria (Upper Triassic). Preliminary work disclosed well-preserved ostracodes in three samples of weathered and marly fossiliferous limestone. The most common genus is a minute dimorphic sulcate form with a healdiid muscle scar, represented by growth stages, that probably belongs to a new genus. Individuals of several additional genera are present, including one that resembles Monoceratina Roth, 1928, and a second that resembles Ogmoconcha Triebel, 1941. This find is incidental to my work on Lower Cretaceous ostracodes in Israel. I am grateful to M. I. Price, Naphtha Oil Co., for taking me to the outcrops, and to I. Zak, Geological Survey of Israel, for stratigraphic information. REFERENCES Jones, P. J., 1962, The ostracod genus Cryptophyllus in the Upper Devonian and Carboniferous of Western Australia: Australia Bur. Mineral Resources, Geology and Geophysics Bull. 62, no. 3,37 p., 3 pis. Kollmann, Kurt, 1960, Ostracoden aus der alpinen Trias Oster-reichs I. Parabairdia n.g. und Ptychobairdia n.g. (Bairdi-idae) : Jahrb. geol. Bundesanstalt, Wien, Sonderband 5, p. 79-105, pis. 22-27, 3 figs. Moore, R. C., ed., 1961, Treatise on invertebrate paleontology, Pt. Q. Arthropoda, 3. Ostracoda: Geol. Soc. America and Univ. Kansas Press, 442 p., 334 figs. Pames, Abraham, 1962, Triassic ammonites from Israel: Geol. Survey of Israel Bull. 33, 76 p., 9 pis., 12 figs. C58 ART. 75 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C58-C59. 1963.SOHN C59 Figure 75.1.—Sketch map showing the geology of the Makhtesh Ramon (after Parnes, 1962) and the location of ostracode-bearing Middle Triassic sedimentary rocks. K, Cretaceous sedimentary rocks; J, Jurassic sedimentary rocks; "fi, Triassic sedimentary rocks; M, igneous intrusive rocks; X, location of outcrops containing Middle Triassic marine ostracodes. 694-027 0—63-----5Article 76 OCCURRENCE OF THE LATE CRETACEOUS AMMONITE HOPUTOPLACENTICERAS IN WYOMING By W. A. COBBAN, Denver, Colo. Abstract.—Hoplitoplacenticeras, a genus well established in Europe as an index fossil to the upper part of the Campanian Stage, has been discovered in the Baculites asperiformis Range Zone near the top of the Steele Shale in Carbon County. The Wyoming specimen is compared to Hoplitoplacenticeras coes-feldiense (Schliiter) var. schiiteri Mikhailov. While mapping in south-central Wyoming, E. N. Harshman, of the U.S. Geological Survey, collected a fragment of Hoplitoplacenticeras, thus providing the first record of this genus in the western interior of the conterminous United States. Aside from its great rarity, the find is of considerable importance because of the association of the Hoplitoplacenticeras with Baculites asperiformis Meek, one of the western interior zone fossils (Cobban, 1962, p. 708). The fossil was found in a bed of concretionary sandstone near the top of the Steele Shale at USGS Mesozoic locality D2923 in the NW%SE%SW% sec. 35, T. 27 N., R. 79 W., Carbon County, Wyo. Associated fossils included numerous baculites and bits of carbonized wood, and a few fish scales and small pelecypods. The fragment of Hoplitoplacenticeras, 59 mm high and 73 mm wide, consists of the middle part of an adult living chamber, and part of the septate whorls. The suture is not visible. Only one side of the living chamber is preserved, and the younger part of it is broken near the venter and the parts separated. The septate whorls have been pushed into the living chamber, and the height of the outer septate whorl has been shortened on one side along a fracture. A slightly oblique view of the specimen is shown in a drawing prepared by John R. Stacy (fig. 76.1). The cross section of the outer septate whorl is very narrow, and the flanks are flattened. The venter is flat and bordered by nodes. The flanks of the outer septate whorl have broad, straight, slightly prosiradiate ribs. Figure 76.1.—Hoplitoplacenticeras cf. H. Coesfeldiense (Schliiter) var. schliiteri Mikhailov. USNM 132373. Only 4 ribs are visible but, judging from their spacing, about 10 or 11 would be present on a half whorl. Each rib has a low, weak clavate node high on the flank, and each rib terminates in a strong clavate node at the margin of the venter. In the illustration the lateral nodes appear to be located about at the base of the upper third of the flank owing to the shortening caused by the fracture lower on the flank. On the opposite and undamaged side of this whorl (not visible in the illustration), these nodes are located at the base of the upper quarter of the flank. Small sharp umbilical nodes are present; these are widely spaced and only two are visible on the exposed part of the outer septate whorl. The living chamber has flattened flanks and well-defined umbilicus and venter. The older part is ornamented by low, broad, prosiradiate, sinuous ribs that terminate in moderately strong clavate nodes at the edge of the venter. There is no trace of the weak lateral C60 ART. 76 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C60-C62. 1963.COBBAN C61 nodes present on the outer septate whorl. Strong, sharp, bullate umbilical nodes are present on the older part of the living chamber. On the younger part, the ribbing abruptly becomes much denser and the ventral and umbilical nodes weaken and disappear. The fragment from Wyoming very closely resembles the adult from Germany figured by Schliiter (1872, p. 56, pi. 17, figs. 1, 2) as a variant of his Ammonites coesfeldiensis. The only conspicuous difference is the persistence of the lateral nodes onto the part of the living chamber of the German specimen that is the equivalent of the Wyoming specimen. This difference may have little meaning inasmuch as the species of Hoplitoplacenticeras are extremely variable, as demonstrated by Paulcke (1907, p. 181-220) for material from Argentina. Schliiter’s specimen differs from the typical form of II. coesfeldiense by its wider umbilicus and more coarsely sculptured septate whorls. Mikhailov (1951, p. 82), in the course of describing Russian specimens, named Schliiter’s specimen II. coesfeldiense var. schliiteri. Naidin and Shimanskii (1959, p. 193) accepted Mikhailov’s assignment, although Glazunova and others (1958, pi. 55) had earlier regarded Mikhailov’s variety as a species. Schliiter’s Ammonites dolbergensis (1876, p. 159, pi. 44, figs. 1-4) resembles the septate whorls of II. coesfeldiense var. schliiteri, and if these forms prove to be the same, Schliiter’s name will take priority over Mikhailov’s. Aside from the occurrences in Germany, Russia, and Argentina, Hoplitoplacenticeras is known from France (Grossouvre, 1893, p. 116-123), Sweden (0dum, 1953, p. 23), Poland (Nowak, 1909, p. 765), Spain (Basse, 1931, p. 36), Libya (Maxia, 1943, p. 470^73), Madagascar (Basse, 1931, p. 35-37), British Columbia (Usher, 1952, p. 93, 94), and Texas (Young, 1963, p. 63, 64). In all of these areas the genus is restricted to rocks of late Campanian age. Schliiter (1876, p. 245-254) recognized a “zone des Ammonites Coesfeldiensis'''1 in his sequence of Late Cretaceous fossil zones for northwest Germany, and Grossouvre (1901, p. 801, table 35) proposed a “Zone a Hoplites Vari” as marking the lower part of the upper Campanian. Since the time of Schliiter and Grossouvre’s publications, one or the other of these species of Hoplitoplacenticeras has been used as a zonal fossil in the Campanian Stage. Spath (1926, table opposite p. 80) accepted Hoplitoplacenticeras vari as one of the Campanian zones in his compilation of Late Cretaceous ammonite levels. Jeletzky (1951, p. 18) and Seitz (1952, p. 149) have used II. vari as the basal zone of the upper Campanian. Muller and Schenck (1943, text fig. 6), in their proposal for the standard of the Cretaceous, recognized II. coesfeldiense as one of the zones of the Campanian and, in the upper part of this zone, II. vari was used as a subzone. Schliiter (1876, p. 252) noted that II. coesfeldiense and H. vari occurred together in his zone of II. coesfeldiense but that H. vari ranged on up into the next higher zone (Bostrychoceras polyplocum). This longer range of II. vari has been verified by Mikhailov (1951, p. 108). Giers (1934, p. 476) also noted the association of H. coesfeldiense and H. vari, and that H. dolbergense occurred a little below them. In the western interior region, rocks near the boundary of the lower and upper Campanian have been zoned best by means of baculites. Above rocks containing the well-known Scaphites hippocrepis of early Campanian age, the following sequence of baculites (from oldest to youngest) has been determined (Cobban, 1962, p. 705): Baculites obtusus Meek, B. mcleariii Landes, B. asperi-formis Meek, B. sp. (smooth), and B. perplexus Cobban. The discovery of Hoplitoplacenticeras cf. H. coesfeldience var. schliiteri in the B. asperiformis Range Zone suggests that this baculite zone lies at or very near the boundary between the lower and upper Campanian. If H. coesfeldiense var. schliiteri and II. dolbergense are the same, a very low position in the upper Campanian is indicated. The Wyoming specimen is in the U.S. National Museum, Washington, D.C. (USNM 132373). A plaster cast is at the Federal Center, Denver, Colo. REFERENCES Basse, flliane, 1931, Monographic paldontologique du CrCtaco de la Province de Malntirano: Service des Mines, Govt. Gdn. Madagascar et Dependencies, 86 p., 13 pis. Cobban, W. A., 1962, Baculites from the lower part of the Pierre Shale and equivalent rocks in the western interior: Jour. Paleontology, v. 36, no. 4, p. 704-718, pis. 105-108. Giers, Rudolf, 1934, Die Schichtenfolge der Mukronatenkreide der Beckumer Hochflache: Centralbl. Mineralogie, 1934, B, p. 471—476. Glazunova, A. E., Luppov, N. P., and Savel’ev, A. A., 1958, Nadseme stvo Noplitaceae, in Luppov, N. P., and Drushchitz, V. V., Molluski-Golovonogne II [Superfamily Hoplitaceae, in Mollusca-Cephalopoda II] of Orlov, Y. A., ed., Principles of paleontology-Handbook for paleontologists and geologists of the U.S.S.R.: Moscow, State Sci. and Tech. Publishing House of Lit. on Geology and the Preservation of Mineral Resources [Gosgeoltekhizdat], v. 6, pt. 2, p. 112-116, pis. 53-55 [in Russian], Grossouvre, Albert de, 1893, Les ammonites de la craie supdri-eure, pt. 2, Paleontologie, of Recherches sur la craie su-pdrieure: Carte Geol. Prance, Mem., 264 p., pis. 1-39. ------ 1901, Classification des couches supracretaedes, chap. 22 in Pt. 1, Stratigraphie generale, of Recherches sur la craie superieure: Carte G6ol. Prance, Mem., p. 751-830.C62 PALEONTOLOGY Jeletzky, J. A., 1951, Die Stratigraphie und Belemnitenfauna des Obercampan und Maastricht Westfalens, Nordwest-deutschlands und D&nemarks sowie einige allgemeine Gliederungs-Probleme der jtingeren borealen Oberkreide Eurasiens: Geol. Jahrb., no. 1, 142 p., 7 pis. Maxia, Carmelo, 1943, Ammoniti maestrichtiane della Tripoli-tania: Soc. geol. italiana Boll., v. 61, p. 469-487, pi. 8. Mikhailov, N. P., 1951, Ammonity werchnego mela yuzhnoj chasti jewropejskoj chasti SSSR i ich stratigraficheskoe znatschenie [The ammonites of the Upper Cretaceous of the southern part of the European part of the USSR and their stratigraphic significance]: Akad. Nauk SSSR, Inst. geol. nauk Trudy, no. 129, geol. ser. no. 50, 143 p., 19 pis. [in Russian], Muller, S. W., and Schneck, H. G., 1943, Standard of Cretaceous System: Am. Assoc. Petroleum Geologists Bull., v. 27, no. 3, p. 262-278. Naldin, D. P., and Shimanskii, V. N., 1959, Golovonogne Mol-luski, in M. M. Moskvin, (ed.), Atlas verkhnemelovoi fauny Severnogo Kavkaza i Kryma [Cephalopod Mollusca, in Atlas of the Upper Cretaceous fauna of the Northern Caucasus and Crimea] : Trans. All-Union Sci.-Research Inst. Natural Gases (VNIIGAZ), [Moscow] p. 166-220, 23 pis. [in Russian]. Nowak, Jan, 1909, O kilku glowonogach i o charakterze fauny z Karpackiego Kampanu [On some cephalopods and the character of the fauna of the Carpathian Campanian] : Kosmos, v. 34, p. 765-787, 1 pi. [Polish with German summary]. 0dum, Hilmar, 1953, The macro-fossils of the Upper Cretaceous, pt. 5 of De geologiska resultaten frftn bormingama vid iHollviken: Sveriges geol. undersokning, ser. C. Arsbok 46 (1952) N :o3, no. 527, 37 p., 4 pis. Paulcke, W., 1907, Die Cephalopoden der oberen Kreide Stid-patagoniens: Freiburg Naturf. Gesell. Berichte, v. 15, p. 167-244, pis. 10-19. Schliiter, Clemens, 1871-1876, Cephalopoden der oberen deut-schen Kreide: Palaeontographica, v. 21 (1871-1872), p. 1-120, pi. 1-35; v. 24 (1876), p. 121-264, pis. 36-55. Seitz, Otto, 1952, Die Oberkreide-Gliederung in Deutschland nach ihrer Anpassung an das internationale Schema: Deutsche geol. Gesell. Zeitschr., v. 104, p. 148-151. Spath, L. F., 1926, On new ammonites from the English Chalk: Geol. Mag., v. 63, no. 740, p. 77-83. Usher, J. L., 1952, Ammonite faunas of the Upper Cretaceous rocks of Vancouver Island, British Columbia : Canada Geol. Survey Bull. 21, 182 p., 31 pis. Young, Keith, 1963, Upper Cretaceous ammonites from the Gulf Coast of the United States: Texas Univ. Pub. 6304, 373 p., 82 pis.Article 77 PALEOTEMPERATURE INFERENCES FROM LATE MIOCENE MOLLUSKS IN THE SAN LUIS OBISPO-BAKERSFIELD AREA, CALIFORNIA By W. O. ADDICOTT and J. G. VEDDER, Menlo Park, Calif. Abstract.—Tropical molluscan taxa in upper Miocene strata near Bakersfield, Calif., suggest a marine hydroclimate warmer than existed in the San Luis Obispo area, 100 miles to the west. A late Miocene temperature regime analogous to that of the present-day outer coast of southwestern Baja California, Mexico, is indicated. Certain faunal data critical to the interpretation of late Miocene water-temperature distribution in California apparently have been overlooked in Hall’s (1960, 1962) study and definition of molluscan provinces of this age along the San Andreas fault. Approximately 60 molluscan taxa of late Miocene age from the southern San Joaquin basin not listed by Hall include several species and genera suggestive of a tropical marine climate. Most of the additional taxa are listed in published reports that discuss subsurface occurrences of late Miocene mollusks in the southern San Joaquin basin; a few are previously unreported species from surface localities in this area. The modification of Hall’s paleotemperature interpretation required by these data invalidates the use of his inferences as definitive evidence for large lateral displacement along the San Andreas fault. A critical area in Hall’s interpretation (1960) is the San Luis Obispo-southern San Joaquin Valley region (fig. 77.1). According to Hall, a composite subtropical fauna from 5 areas on the west side of the fault is opposed by a composite warm-temperate fauna from 2 widely separated areas east of the fault. Temperature inferences were made by comparing certain species and genera with the modern faunas of northern and southern Baja California. The living mollusks of these geographic areas along the outer coast of Baja California are considered sufficiently distinct by Hall (1960, map 1) to include them in a newly recognized tripartite subdivision of the California province (lat 23° N. to 3414 ° N.) of Dali (1909) and others. The boundary ART. 77 IN U.S. GEOL. SURVEY PRC 120" Figure 77.1—Late Miocene molluscan fossil localities and marine deposition in southern San Joaquin Basin. separating Hall’s Magdalenan (subtropical) and Ense-nadian (warm-temperate) subprovinces is placed near Cedros Island (lat 28° N.). Provincial boundaries can be conveniently expressed by mean winter surface-water isotherms which are generally normal to the coastline of western North America. Northern end points of range for tropical (Panamic) and subtropical PAPER 475-C, PAGES C63-C68. 1963. C63C64 PALEONTOLOGY taxa measured in these terms have been used to interpret Cenozoic paleotemperatures (Durham, 1950a; Hall, 1960; Valentine, 1961). The principal indicator for a subtropical marine environment in the San Luis Obispo region seems to be the occurrence of an undescribed pelecypod referred to the subgenus Area s.s., which today is limited, in the eastern Pacific, to latitudes south of the 18°C minimum winter surface-water temperature isotherm (Hall, 1960, p. 298). Supporting evidence presented by Hall for interpreting subtropical water temperatures west of the fault was the diversity of species of Turritella (four species) and the abundance of Anadara and Lyropeeten. In the- southern San Joaquin basin, Hall (1960, p. 286) lists 20 mollus-can taxa which he regards as a warm-temperate fauna. All but one of these forms occur in the Comanche Point area, where 27 taxa previously have been reported (Clark in Merriam, 1916; Nomland, 1917; and Hoots, 1930). Other published occurrences of marine mollus-can fossils considered were from the San Emigdio area (2 species listed by Pack, 1920, p. 42, which were regarded as Pliocene forms by Hoots, 1930, p. 258); the Crocker Flat landslide area (3 species listed by Simonson and Kreuger, 1942, p. 1620); and 11 taxa from the Lebec quadrangle identified by Corey {in Crowell, 1952, p. 13). Faunal data from east of the fault not included by Hall consist of an assemblage of 55 taxa of mollusks identified by H. R. Gale {in Preston, 1931, p. 15, 16) from well cores in upper Miocene strata west of Bakersfield (see table 77.1). This assemblage includes 48 taxa previously unreported from surface localities in the southern San Joaquin basin and contains a number of tropical and subtropical species and genera. The mollusks identified by Gale {in Preston, 1931, p. 15, 16) are from the Fruitvale oil field 3 miles west of Bakersfield, Calif, (fig. 77.1). Northwest of the Fruitvale field, Grant and Gale (1931, p. 408) list 15 species of mollusks from the Santa Margarita Formation in the Superior Ansolabehere No. 1 well (sec. 9, T. 29 S., R. 27 E., M.D.B. and M.). All of these species are recorded in Gale’s larger list. Other subsurface occurrences of megafossils of probable late Miocene age are from the Mountain View oil field southeast of Bakersfield, where the Wharton sand (local usage) has been compared with the type Santa Margarita Formation on the basis of the molluscan assemblage (Miller and Ferguson, 1943). In the southeast part of the Mountain View field, Miller and Bloom (1937, p. 10) describe “marine fingers” in the lower part of the Chanac Formation with marine fossils including “dementia per tenuis, large variety.” Miocene mollusks identified by Gale from the Fruitvale field are from the upper 500 to 600 feet of the Santa Margarita Formation, which is described as a gray, fine- to medium-grained sand with abundant fossil remains (Preston, 1931, p. 12). Beck (1952) assigns the Santa Margarita Formation penetrated by wells in the area southwest and southeast of Bakersfield to the upper Mohnian Stage (Kleinpell, 1938). A recent correlation section (Church, Krammes, and others, 1957) places the upper part of the Santa Margarita Formation at the Fruitvale field within the upper Mohnian Stage. As stated by Gale {in Preston, 1931, p. 15), the molluscan assemblage is indicative of “post-Temblor Miocene” or a late Miocene age in the usage of the Pacific coast larger invertebrate chronology. Recent collecting at Comanche Point has nearly doubled the number of taxa from that locality (see table 77.2) and further increases the known late Miocene fauna from the southern San Joaquin basin to about 80 taxa. Analysis of the supplementary faunal data strongly suggests that interpretation of temperature affinities of late Miocene molluscan assemblages, as they are now known, cannot be used as definitive evidence for large lateral displacement along the San Andreas fault as was proposed by Hall (1960). The tropical faunal element near Bakersfield indicates water temperatures higher than those suggested by the San Luis Obispo fauna west of the San Andreas fault (table 77.3), in contradistinction to Hall’s (1960, p. 288) conclusion that “cooler temperatures must have prevailed [in the southern San Joaquin basin] because of the total absence of distinct Magdalenan faunal [=subtropical] elements.” Further evidence for subtropical to tropical marine water temperatures east of the fault is found in the Coalinga area, where two taxa characteristic of the Pan-amic molluscan province have been described from the Santa Margarita Formation (Nomland, 1917, p. 301). Discussion.—For the purpose of this discussion the species identified by Gale and other workers are considered to have been competently determined. Information on the geographic ranges of living species and genera is from Grant and Gale (1931), Burch (1944-46), and Keen (1958). Although definitive criteria for Hall’s (1960) subdivision of the Californian molluscan province into Magdalenan, Ensenadian, and Southern Californian subprovinces are not readily apparent, particularly along the unprotected outer coast, these units are used provisionally in the following discussion and are referred to as provinces as a matter of convenience. A large lucinid, Miltha xantusi (Dali), one of the most abundant species in the well cores, is reported living off Cape San Lucas, Baja California, near the boundary of the Panamic and Magdalenan provinces and has been collected intertidally at La Paz (Pilsbry and Lowe, 1933). A similar form is present at Coman-ADDICOTT AND VEDDER C65 che Point. The extinct gastropod “Phos” dumbleana Anderson in Hanna is an indicator of tropical hydro-climate at the generic level. Although this species is probably better assigned to the genus Tritiaria, it is here referred to “Phos” in deference to the most recent treatment of living eastern Pacific species by Strong and Lowe (1936). Another extinct species listed by Gale, Ficus (Trophosycon ocoyana (Conrad), supports the interpretation of subtropical to tropical marine waters in the southern San Joaquin basin during late Miocene time. In the eastern Pacific the genus Ficus is restricted to the Panamic province; its northernmost occurrence along the outer coast of Mexico is off Cape San Lucas, at or south of the tropical-subtropical provincial boundary. This species occurs in well cores from the Reef Ridge Shale of late Miocene age at Kettleman Hills oil field (Barbat and Johnson, 1934). Ficus also has been reported from the Santa Margarita Formation northeast of Coalinga by Nomland (1917, p. 301) as “Ficus, cf. nodiferous Gabb.” The genus Turricula, represented in the Santa Margarita fauna from the Fruitvale field by the extinct species Turricula ochsneri (Anderson and Martin), is characteristic of the Panamic province. This species seems to be very closely related to Turricula (Turricula) libya Hall, which is living in the Panamic province off Cape San Lucas. Calyptraea mamillaris Broderip, the second most abundant gastropod in cores from the Santa Margarita Formation at the Fruitvale field, is not known to be living farther north than Magdalena Bay along the southern coast of Baja California (Keen, 1958, p. 311). A similar form is present at Comanche Point. Turritella cf. T. vanvlecki Arnold, also listed by Gale, is very closely related to the living T. gonostoma Valenciennes. Both species have been placed in a group of approximate subgeneric rank, the Turritella broderipi-ajna stock, by Merriam (1941, p. 51). In the eastern Pacific, living representatives of this stock range throughout the Panamic province. Twrritella gonostoma, which typifies this stock, is the most northerly ranging species occurring in the living fauna as far north as San Juanico Bay (lat 26° N.) on the outer coast of southern Baja California (Keen, oral communication, 1962) J Another species referable to this stock, Turritella freya Nomland, is described from the Santa Margarita Formation north of Coalinga. Other species of Turritella occurring east of the San Andreas fault in strata of this age are Turritella cf. T. cooperi Carpenter, identified by Merriam (1941, p. 49, fig. 8; p. 118) from the San Pablo “formation” (Cierbo Sandstone) on the 1 A more northerly occurrence reported from Scammons Lagoon (lat 28° N.) In the “Hemphill collection” (Grant and Gale, 1931, p. 773) is questionable inasmuch as Phleger and Ewing (19620 list T. gonostoma as a characteristic element in late Pleistocene deposits that lie a. few feet above sea level. north side of Mount Diablo, and an unnamed “Turritella,, sp. (large)” in the Santa Margarita Formation north of Coalinga (Nomland, 1917, p. 301). There are thus at least 3 and possibly 4 species of Turritella in upper Miocene strata on the east side of the San Andreas fault. Cancellarids commonly found in middle Miocene strata of California also are listed from the subsurface section of upper Miocene rocks near Bakersfield. Using the classification of Marks (1949), Cancellaria lickana Anderson and Martin is best assigned to Cancellaria s.s., which, in the eastern Pacific, is restricted to the Magdalenan and Panamic provinces. Cancellaria cf. C. pacifica Anderson is representative of the subgenus Euclia, which lives in tropical seas. Cancellaria joa-qumensis Anderson, a form that seems to have no Recent analogs in the eastern Pacific, may be intermediate between Cancellaria s.s. and Euclia. Cancellaria joaquin-ensis closely resembles Cancellaria pabloensis Clark from the Mount Diablo, Coalinga, and Comanche Point areas. Two other species referable to Cancellaria s.s. have been collected recently at Comanche Point (see table 77.2). Corey (in Crowell, 1952, p. 13) lists an Oliva sp. from the Santa Margarita Formation of the Lebec quadrangle, about 20 miles south of Comanche Point on the east side of the San Andreas fault. Living representatives of this genus are characteristic of the Panamic molluscan province, but one species ranges northward into the Magdalenan province. Gale (in Preston, 1931) lists a taxon as lOlivella biplioata Sowerby but indicates parenthetically that the specimen might be an Oliva, 0. futheyana Anderson. The late Miocene Lyropectens are related to the Recent Panamic-Magdalenan species Lyropecten (Nodipecten) subnodosus (Sowerby), which is assigned to Lyropecten s.s. by some authors (Hertlein [as subgenus of Pecten], 1935, p. 317; Durham, 1950b, p. 65; Keen, 1958, p. 74). The common occurrence of late Miocene species of Lyropecten s.s. in the southern San Joaquin basin further suggests water warmer than that postulated by Hall. It should be noted that Lyropectens are among the most abundant fossils at nearly all late Miocene localities in the Coalinga area, along the west side of the Temblor Range from Recruit Pass south to Bitterwater Creek, at Comanche Point, and in the Lebec area. A point to be considered in paleoecologic interpretation is the difference in opinion of various workers in ascribing temperature ranges to certain critical molluscan taxa. Dosmia ponderosa Gray, for example, is closely related to late Miocene species. Grant and Gale (1931, p. 352-354) consider late Miocene forms as variants of the living Dosinia ponderosa, which ranges fromC66 PALEONTOLOGY Scammons Lagoon, near the latitude of the Magdale-nan-Ensenadian provincial boundary, southward to Peru. Valentine and Meade (1961) treat this species as a Panamic form which extended its range north of the provincial boundary by inhabiting protected, warm-water environments. Hall (1960), however, identifies the northern end point of range of this taxon with the Ensenadian molluscan province. Doubtless, arguments can be made for either interpretation, but the former seems to be more acceptable in light of Emerson’s (1956) discussion of the latitudinal coexistence of tropical species in protected embayments and cooler water assemblages in outer coast environments in the Magda-lenan province. It is apparent that use of the indicated minimal temperatures of tropical (Panamic) or warm-temperate (Ensenadian) provinces in interpreting late Miocene occurrences of this taxon will yield widely divergent estimates of paleotemperature. Conclusion.—The distribution of tropical and subtropical marine species and genera is not compatible with Hall’s (1960) interpretation of late Miocene surface-water temperatures from which he has postulated large right-lateral movement along the San Andreas fault. The relationship of the critical San Luis Obispo-southern San Joaquin basin late-Miocene faunas suggests a complex distribution of marine surface-water temperatures. The pattern probably was related more to the configuration of an irregular, embayed coastline protected to the west and to the north by large insular or peninsular blocks separated by narrow straits rather than to the present latitudinal distribution of winter surface-water temperatures along a fairly regular, open coast. If large right-lateral slip has occurred along the San Andreas fault since deposition of upper Miocene strata, a paleogeographic restoration would place the molluscan fauna of the San Joaquin basin with its tropical element considerably farther north than the presumed subtropical fauna of the San Luis Obispo region. Whether or not the fault has had relatively great lateral movement, the presence of tropical and subtropical molluscan species at the same latitude in late Miocene time seems to be indicated. This can be satisfactorily explained by reference to the modem occurrence of many Panamic species in shallow embayments along the west coast of southern Baja California. If the southern San Joaquin basin was a protected embayment during late Miocene time, it seems likely that relicit faunal elements may have persisted there long after more temperate faunas had spread southward along the open outer coast. The intent of this discussion is not to conclude whether paleoecologic interpretation can be used to prove or to disprove lateral movement along the San Andreas fault; rather, it is to point out that the faunal assemblages, as they are now known, do not necessarily support the contention that late Miocene molluscan provinces were alined from north to south through the present area of California in a manner which approximated the present latitudinal distribution of winter oceanic surface-water temperatures. Instead, a fairly complex temperature regime, possibly analogous to that of the outer coast of southern Baja California but with much larger protected embayments, existed in the central California region during late Miocene time. Tabu: 77.1.—Miocene mollusks from the Fruitvale field [Gale in Preston (1931). Generic designations as listed by Gale; arrangement revised] Gastropods Calliostoma sp. Tegula sp. IMelanella californica (Anderson and Martin) Turritella cf. T. vanvlecki Arnold Calyptraea mamillaris Broderip Poliniees (Poliniccs) diabloensis (Clark) Polinices (Neverita) reclusianus (Deshayes) Ficus (Trophosycon) ocoyana (Conrad) Forreria wilkesana (Anderson) Mitrella tuberosa (Carpenter) Phos dumbleana Anderson in Hanna Nassarius (Uzita) antiselli (Anderson and Martin) Nassarius (Uzita) cf. N. amoldi (Anderson) lOlivella hip Heat a Sowerby Olivella pedroana (Conrad) Cancellaria joaquinensis Anderson Cancellaria lickana Anderson and Martin Cancellaria cf. C. Pacifica Anderson Cancellaria cf. C. nevadensis Anderson and Martin IDaphnella sp. IMangelia cf. M. kernensis Anderson and Martin Turricula ochsneri (Anderson and Martin) Pseudomelatoma penicillata (Carpenter) Terebra pedroana Dali Turbonilla sp. Pelecypods Clycymeris septentrionalis (Middendorff) Ostrea cf. 0. lurida Carpenter Pecten (Aequipecten) discus Conrad Anomia sp. Lucina (Here) richthofeni Gabb Lucina (Miltha) xantusi (Dali) Lucina (Myrtea) acutilineata Conrad f Lucina (Myrtea) calif ornica Conrad Lucina (Myrtea) nuttalli Conrad Taras harfordi (Anderson) Cardium sp. Dosinia ponderosa longidens Grant and Gale Irus lamellifer perlamellifer Grant and Gale Chione sp. Venerupis (Protothaca) staminea (Conrad) Amiantis callosa stalderi (Clark) Petricola carditoides (Conrad) Mactra (Spisula) hemphill Dali Mactra (Spisula) sp. Anatina (Raeta) plicatella longior Grant and GaleADDICOTT AND VEDDER C67 Table 77.1.—Miocene mollusks from the Fruitvale field—Continued Pelecypods—Continued Tellina idae Dali Apolymetis biangulata (Carpenter) Macoma identata (Carpenter) IMacoma inquinata (Deshayes) Macoma wilsoni (Anderson and Martin) Oari edentula (Gabb) Solen sicarius Gould Corbula (Lentulium) luteola Carpenter IMyra sp. Panope generosa Gould Table 77.2.—Mollusks from the Santa Margarita Formation at Comanche Point [Combined lists of previous collectors. Clark in Merriam, 1916; Nom-land, 1917, Clark in Hoots, 1930. Generic designations revised in part] Gastropods Calyptraea sp. “Natica" sp. Nassarius pabloensis (Clark) “Fusinus” fabulator Nomland Conus sp. Bulla sp. Pelecypods Pinna alamedensis Yates Ostrea titan Conrad Ostrea cf. O. vespertina Conrad Chlamys hastatus (Sowerby) Aequipecten raymondi (Clark) Lyropecten crassieardo (Conrad) Lyropecten crassieardo biformatus (Nomland) Lyropecten estrellanus (Conrad) Lucirna excavata Carpenter [=Phacoides richthofeni (Gabb)] Miltha sanctaecrucis Arnold [possible misidentification of Miltha xantusi (Dali)] Cardium sp. Dosinia arnoldi Clark Dosinia sp. Amiantis stalderi (Clark) Saxidomus nuttalli Conrad Clementia (Egesta) pertenuis (Gabb) Chione sp. Apolymetis biangulata (Carpenter) [=Metis alta (Conrad)] Solen sp. Siliqua cf. S. lucida (Conrad) Teredo sp. [ Supplementary list from USGS loc. M1619 (near bead of second eastdraining gully due north of hill 1039, NW(4SW% sec. 24, T. 32 S., R. 29 E., coUected by J. G. Vedder, 1962. Most of the taxa listed above are present at this locality] Gastropods Astraea cf. biangulata (Gabb) [?=A. raymondi (Clark)] Littorina mariana Arnold? Calyptraea mammilaris Broderip? Crepidula cf. C. adunca Sowerby Crepidula princeps Conrad? Polinices sp. [may be Natica of Clark] Neverita reclusiana (Deshayes) [may me Natica of Clark] Table 77.2.—Mollusks from the Santa Margarita Formation at Comanche Point—Continued Gastropods—Continued “Trophon" sp. Pterynotus sp. Mitrella aff. M. tuberosa (Carpenter) Olivella aff. O. pedroana (Conrad) Cancellaria (Cancellaria) cf. C. sanjosei Anderson and Martin Cancellaria (Cancellaria) n. sp.? cf. C. decussata Sowerby Cancellaria (Cancellariaf) cf. C. pabloensis Clark Crassispira n. sp. ? cf. C. ericana Hertlein and Strong Conus n. sp.? cf. G. purpurascens Sowerby [presumably Conus sp. of Clark] Strioterebrum aff. S. martini (English) Scaphander? sp. Pelecypods Anadara? sp. Glycymeris aff. G. subobsoleta (Carpenter) H innit es? cf. H. giganteus (Gray)1 Pododesmus cepio (Gray) ? Lucinisca nuttalli (Conrad) Miltha cf. M. xantusi (Dali) Dosinia cf. D. merriami Clark [may be Dosinia sp. of Clark] Protothaca? cf. P. staminea (Conrad) Protothaca cf. P. tenerrima (Carpenter) Tellina sp. Corbula (Caryocorbula) n. sp. 1 University of California Museum of Paleontology localities A—9417, A-9418. Table 77.3.—Late Miocene mollusks from the San Luis Obispo area [Compiled by Hall (1960, table 7)] Gastropods Astraea raymondi (Clark) Bulla sp. Calyptraea martini Clark Crepidula sp. N ep tunea cierb oensis (Clark) Nucella rankini Eaton and Grant Ocenebra selbyensis (Clark) Trophon (Forreria) carisaensis (Anderson) Trophon (Forreria) carisaensis mirandaensis Eaton and Grant Trophon clarki cuyamanus Eaton and Grant Trophon gillulyi Eaton and Grant Trophon pabloensis Clark Turritella carrisaensis Anderson and Martin Turritella cooperi Carpenter Turritella freya Nomland Turritella margaritana Nomland Pelecypods Anadara obispoana (Conrad) Anadara trilineata (Conrad) Apolymetis biangulata (Carpenter) Area s.s. n. sp. Chione sp. Dosinia sp. Gari sp. Glycymeris sp. Lucinisca nuttallii (Conrad)C68 PALEONTOLOGY Table 77.3.—Late Miocene mollusks from the San Luis Obispo area—Continued Pelecypods—Continued Lucinmna acutilineata (Conrad) Ostrea bourgeoisii Rem on d Ostrea titan Conrad Ostrea titan eucorrugata Hertlein Panope generosa Gould Pecten (Lyropecten) crassicardo Conrad Pecten (Lyropecten) estrellanus Conrad Pecten (Chlamys) hodgei Hertlein Pecten (Aequipecten) raymondi Clark Saxidonws nuttallii (Conrad) Schizothaerus nuttallii (Conrad) Solen sp. Tivela sp. Trachycardium quadragenarium (Conrad) Yoldia sp. REFERENCES Barbat, W. F., and Johnson, F. L., 1934, Stratigraphy and Foraminifera of the Reef Ridge shale, upper Miocene, California : Jour. Paleontology, v. 8, p. 3-17. Beck, R. S., 1952, Correlation chart of Oligocene, Miocene, Pliocene, and Pleistocene in San Joaquin Valley and Cuyama Valley areas: Am. Assoc. Petroleum Geologists, Soc. Econ. Paleontologists and Mineralogists, Soc. Exploration Geophysicists Guidebook, Joint Ann. Mtg., Los Angeles, Calif., p. 104. Burch, J. Q., ed., 1944-46, Distributional list of the west American marine mollusks from San Diego, California, to the Polar Sea : Conchological Club Southern California Minutes, nos. 33-63. Church, H. V., Jr., and Krammes, Kenneth, (chairmen) and others, 1957, Cenozoic correlation section, south San Joaquin Valley: Am. Assoc. Petroleum Geologists, Geol. Names and Correlations Comm., San Joaquin Valley Sub-comm. on the Cenozoic. Crowell, J. C., 1952, Geology of the Lebec quadrangle, California : California Div. Mines Spec. Rept. 24, 23 p. Dali, W. H., 1909, Report on a collection of shells from Peru, with a Summary of the littoral marine Mollusca of the Peruvian zoological province: U.S. Nat. Museum Proc., v. 37, p. 147-294. Durham, J. W., 1950a, Cenozoic marine climates of the Pacific Coast: Geol. Soc. America Bull., v. 61, p. 1243-1264. Durham, J. W., 1950b, Megascopic paleontology and marine stratigraphy, pt. 2 of The 1940 E. W. Scripps cruise to the Gulf of California: Geol. Soc. America Mem. 43, 216 p. Emerson, W. K., 1956, Pleistocene invertebrates from Punta China, Baja California, Mexico, with remarks on the composition of the Pacific Coast Quaternary faunas: Am. Mus. Nat. Hist. Bull., v. Ill, art. 4, p. 317-342. Grant, U. S., IV, and Gale, H. R., 1931, Catalogue of the Marine Pliocene and Pleistocene Mollusca of California : San Diego Soc. Nat. Hist. Mem., v. 1,1,036 p. Hall, C. A., Jr., 1960, Displaced Miocene molluscan provinces along the San Andreas fault, California: California Univ., Dept. Geol. Sci. Bull., v. 34, no. 6, p. 281-308. ------1962, Reply [to a review of Hall (1960) by J. W. Durham and S. R. Primmer, in the same issue] : Am. Assoc. Petroleum Geologists Bull., v. 46, no. 10, p. 1953-1960. Hertlein, L. G., 1935, The Templeton Crocker expedition of the California Academy of Sciences, 1932—No. 25, The Recent Pectinidae: California Acad. Sci. Proc., 4th ser., v. 21, p. 301-328. Hoots, H. W., 1930, Geology and oil resources along the southern border of San Joaquin Valley, California : U.S. Geol. Survey Bull. 812-D, p. 243-332. Keen, A. M., 1958, Sea shells of tropical west America; marine mollusks from Lower California to Columbia: Stanford, Calif., Stanford University Press, 624 p. Kloinpell, R. M., 1938, Miocene stratigraphy of California: Tulsa, Okla., Am. Assoc. Petroleum Geologists, 450 p. Marks, J. G., 1949, Nomenclatural units and tropical American Miocene species of the gastropod family Cancellariidae: Jour. Paleontology, v. 23, no. 5, p. 453-464. Merriam, C. W., 1941, Fossil Turritellas from the Pacific coast region of North America: California Univ., Dept. Geology Sci. Bull., v. 26, no. 1, p. 1-214. Merriam, J. C., 1916, Mammalian remains from the Chanac Formation of the Tejon Hills, California: California Univ., Dept. Geology Bull., v. 10, no. 9, p. 111-127. Miller, R. H., and Bloom, C. V., 1937, Mountain View oil field: California Div. Oil and Gas, Summary of Operations, California Oil Fields Ann. Rept. 22, no. 4, p. 5-36. Miller, R. H., and Ferguson, G. C., 1943, Mountain View oil field: California Div. Mines Bull. 118, p. 565-570. Nomland, J. O., 1917, Fauna of the Santa Margarita beds in the North Coalinga region of California: California Univ., Dept. Geology Bull., v. 10, no. 18, p. 293-326. Pack, R. W., 1920, The Sunset-Midway oil. field, California, pt. 1, Geology and oil resources: U.S. Geol. Survey Prof. Paper 116,179 p. Phleger, F. B., and Ewing, G. C., 1962, Sedimentology and oceanography of coastal lagoons in Baja California, Mexico: Geol. Soc. America Bull., v. 73, p. 145-182. Pilsbry, H. A., and Lowe, H. N., 1933, West Mexican and Central American mollusks collected by H. N. Lowe, 1929-1931: Acad. Nat. Sci. Philadelphia Proc., v. 84, p. 33-144. Preston, H. M., 1931, Report on Fruitvale oil field: California Div. Oil and Gas, Summary of Operations, California Oil Fields Ann. Rept. 16, no. 4, p. 5-24. Simonson, R. R., and Krueger, M. L., 1942, Crocker Flat landslide area, Temblor Range, California: Am. Assoc. Petroleum Geologists Bull., v. 26, no. 10, p. 1608-1631. Strong, A. M., and Lowe, H. N., 1936, West American species of the genus Phos: San Diego Soc. Nat. Hist. Trans., v. 8, p. 305-320. Valentine, J. W., 1961, Paleoecologic molluscan geography of the Californian Pleistocene: California Univ., Dept. Geol. Sci. Bull., v. 34, no. 7, p. 309-442. Valentine, J. W., and Meade, R. F., 1961, Californian Pleistocene paleotemperatures: California Univ., Dept. Geol. Sci. Bull., v. 40, no. 1, p. 1—46.Article 78 LATE PLEISTOCENE DIATOMS FROM THE ARICA AREA, CHILE By ROBERT J. DINGMAN and KENNETH E. LOHMAN, Washington, D.C. Work done in cooperation with the Instituto de Investigaci6nes Geoldgicat, Santiago, Chile, under the auspice» of the Agency tor International Development, U.S. Department of State Abstract.—The extensive Tertiary (?) sequence of ash-flow and detrital deposits in the Arica area was studied. Samples from within the Lluta Formation and from strata underlying it were barren of microfossils. Samples from two deposits of diatomaceous earth overlying the formation or interbedded in the upper part indicate a probable late Pleistocene age. A thick sequence of ash flows and interbedded fan-glomerate locally associated with deposits of diatomaceous earth mantles much of northern Chile, western Bolivia, northwestern Argentina, and southern Peru. In general the sediments are coarse grained, poorly sorted, and poorly stratified. The crossbedding of the water-laid sediments indicates deposition from areas that correspond to the present areas of high relief. For example, in northern Chile the direction of deposition was westward from the Andes Mountains, and the sequence is preserved in the attitude in which it was deposited, with few exceptions. Along much of the western slope of the Andes the uppermost ash-flow bed is a moderately welded tuff that forms the present land surface and slopes uniformly 3° to 5° W. Below an altitude of approximately 2,000m the tuff bed is mantled by a thick series of detrital material. Bowman (1909) misinterpreted this depositional surface in the Pica area of Chile as a peneplain, and upon this interpretation based his still widely accepted theory of Pleistocene uplift of the Andes. The volcanic-clastic sequence has been described by many geologists and has been assigned many names. The most complete description was made by Galli and Dingman (1962) in the Pica area, where the deposits were named the Altos de Pica Formation. The Altos de Pica Formation in the type locality is 735 m thick and consists of 3 coarse-grained sedimentary members interbedded with 2 intercalated welded-tuff members. Doyel andHenriquez (writtencommunication, 1962) proposed the name Lluta Formation for the same sequence in the Arica area. The presence of the pyroclastic members within the sequence suggested the name given to the formation by Bertrand (1885), “Traquitica”; and by Bruggen (1918), “Liparitica” and “Riolitica”. The last two formation names are still in general use, although they are misnomers inasmuch as volcanic rocks make up less than 10 percent of the formation in many areas. The determination of the age of this sequence, which extends over many thousands of square kilometers and may be 2,000 m thick in some localities, has been one of the main problems of Chilean geology. The sequence is generally considered to be of Tertiary or possibly Quaternary age (table 78.1). It is of continental origin and contains no diagnostic vertebrate or invertebrate fossils, nor is it known to interfinger with marine deposits. All the ages assigned in table 78.1 are, therefore, only estimates by the various authors. The only fossil evidence is the classification by Douglas (1914) of fragments of a jawbone from the Stratos de Rio Mauri as being from a Nesodon similar to that of a Miocene species from the Santa Cruz Formation of Argentina. In June 1960 an attempt was made to date the Lluta Formation of the Arica area by means of diatoms. Samples were collected from the shales of the (marine?) Arica Formation, which underlies the Lluta Formation with marked angular unconformity; from dolomitic beds intercalated within the Lluta For- AKT. 78 IN U.S. GEOL. SURVEY PROP. PAPER 475-C, PAGES C69-C72. 1963. C69C70 PALEONTOLOGY Table 78.1.— Tertiary and Quaternary formations in northern Chile and adjoining areas Period Epoch Bolivia (Douglas, 1914) Bolivia (Ahlfeld, 1946) Peru (Jenks, 1948) Chile (Bruggen, 1950) Argentina (Groeber, 1957) Chile (Galll and Ding- man, 1962) Chile (Doyel and Hen- riguez, written com- munica- tion, 1962) Chile (Ding- man, 1963) Quaternary Pleistocene Riolltica Formacion 1 (Chile) Altos de Pica Formation S'lSZiT 3 Lluta Formation1 Unnamed formation of ash-flow deposits 1 Tertiary Pliocene Estratos del Rio Mauri1 Araucaniano Formacion Miocene Mauri Volcanic Series 1 ? Liparltica Formacion 1 Riolltica Formacion 1 (1950) n Pedro Formation imbores Formation Oligocene Corcoro System Chacani Volcanics 9 Eocene § San Pedro Formacion 03 Ph ■ Stratlgraphically and lithologically equivalent to the Altos de Pica Formation. mation; and from deposits of diatomaceous earth that overlie or are intercalated in the uppermost beds of the Lluta Formation (table 78.2). The locations of the sampled localities are indicated in figure 78.1. All 17 of the samples listed in table 78.2 were disaggregated, concentrated for diatoms, and systematically examined under the microscope. Five samples (5272, 5273, 5283, 5287, and 5288,) proved to be devoid of diatoms. Because many of the remaining 12 samples had very similar diatom assemblages, 7 were selected for intensive study. The diatoms found in them are listed in table 78.3. It will be noted that several were selected from each of the two productive localities. Unfortunately, two of the barren samples (5272 and 5273) came from the older marine(?) sediments that unconformably underlie the Lluta Formation; thus no inferences can be made concerning these beds except that they are older, according to the field evidence. All samples containing diatoms were obtained from the two deposits of diatomaceous earth listed in table 78.2. The larger of these deposits is 7y2 km north- northeast of Arica and 4 km east of the coast on a headland approximately 240 m above sea level. The diatoms from this outcrop were studied by Frenguelli (1938), who gave a different location for the deposits. However, the samples examined by Frenguelli were obtained from Dr. Humberto Fuenzalida, present director of the School of Geology at the University of Chile, who had been given them by Senor Tomas Vila. Senores Fuenzalida and Vila have agreed (oral communication, 1963) that the earlier samples were obtained from the same deposit that was sampled during this investigation. This deposit probably formed in a shallow freshwater lake that may have been a sag pond along one of the north-trending faults of the area. The deposit overlies an ash-flow bed of the Lluta Formation and may be overlain by the upper detrital member of the formation. Unfortunately, the upper part of the Lluta Formation is lithologically identical with the more recent colluvial deposits of the area so that it is possible that the coarse sediments that overlie the eastern partDINGMAN AND LOHMAN C71 Table 78.2.—List of samples collected for diatom determination ZJSGS diatom Location locality on fig. 78.1 5272 1 5 Location and description of samples 5 4 km northeast of Arica, Chile; clay pit in the Arica Formation 200 m east of Esso oil tanks. Sample from unconsolidated marine (?) sediments of probable early to middle Tertiary age which are overlain with marked angular unconformity by the Lluta (upper Tertiary to Recent) Formation. 5273 1 5 Same locality as 5272. 5274 1 Approximately 7% km northeast of Arica, Chile; 3 km east of Pan-American highway. Samples from deposit of diatomaceous earth overlying or intercalated In the Lluta Formation. Samples 5274-5279 were taken by channeling the exposed vertical face of the deposits. Sample 5274 from 0-1.6 m above base of sampled Interval. 5275 1 Same locality as 5274. Interval 1.6-2.25 m. 5276 1 Same locality as 5274. Interval 2.25-4.45 m. 5277 1 Same locality as 5274. Interval 4.45-5.85 m. 5278 1 Same locality as 5274. Interval 5.85-7.1 m. 5279 1 Same locality as 5274. Interval 7.1-7.65 m. 5280 1 Same locality as 5274. Layer of very pure diatomaceous earth. 5281 1 Same locality as 5274. Layer of very pure diatomaceous earth. 5282 1 Same locality as 5274. Caprock overlying deposit of diatomaceous earth. 52831 3 Arica, Chile. Dolomite quarry in Quebrada Diablo, 20 km east of Arica. Dolomite is intercalated in the Lluta Formation. 5284 2 Boca Negra, Chile. Deposit of diatomaceous earth in the Lluta Valley. Deposit overlies or is intercalated in uppermost part of the Lluta Formation. 5285 2 Same locality as 5284. 5286 2 Same locality as 5284. 52871 4 5.1 km south of the Chaca, Chile, police station. Impure dolomite bed intercalated in the Lluta Formation, on south side of Quebrada Chaca. 52881 4 Same locality as 5287. 1 No diatoms found. of the diatomaceous deposit may be younger in age than the Lluta Formation. The same stratigraphic situation exists at Boca Negra, where the diatomaceous-earth deposit overlies the uppermost ash flow and is overlain by coarse sediments. The extensive assemblage of diatoms obtained contains three extinct species, Mctstogloia atacamae, Melo-sira spinigera and Navicula fuemalida, the first two of which were originally described from a lacustrine limestone in the Calama basin, Chile. The limestone has been called early Pleistocene by Frenguelli (1936) on the basis of somewhat tenuous stratigraphic correlations. These two species occur rarely in the present material. The last species was described from the Figure 78.1.—Map of the Arica area, Chile, showing the locations from which samples were collected. 1, Arica diatom deposit; 2, Boca Negra diatom deposit; S, Dolomite beds; 4, Chaca diatom deposit; 5, Tertiary sediments. diatomite near Arica by Frenguelli (1938), who considered it to be late Pleistocene. Because the last species, Navicula fuemalida, occurs commonly in the Lluta Formation, more importance can be attached to it. The balance of the species found, with the exception of two thought to be new, are represented in living assemblages. The diatoms indicate a late Pleistocene age for the diatomaceous earth. Some species of diatoms from the Lluta Formation (marked by the footnote reference in table 78.3) also occur in the Calama basin in the limestone considered by Frenguelli (1936) to be early Pleistocene. However, the most abundant species in all the collections studied for this report is Denticvla elegam, a Recent form that goes back to late Tertiary. It is rare in the Calama basin. Although the diatomaceous deposits in which these collections were made may be early Pleistocene in age, the evidence for a late Pleistocene age is much more compelling. The diatom assemblages indicate deposition in a nonmarine, though somewhat saline, shallow lake, whose waters were cool to cold.C72 PALEONTOLOGY Table 78.3.—Diatoms identified from the Lluta Formation in the Arica area [A, abundant; C, common; F, frequent; E, rare. Numbers in boxheads are USGS diatom localities] Diatom 8 km northeast of Arica Boca Negra 5274 5276 5279 5280 5282 5284 5286 Achnanthes cf. A. micro- F Amphora acutiscula Kiitzing.. R lineata Ehrenberg _ R F cf. A. salina Wm. Smith R F R sp _ _ _ _ _ R Anomoeoneis sphaerophora (Kiitzing) Pfitzer F F Caloneis formosa (Gregory) Cleve _ - R Cocconeis placentula var. euglypta (Ehrenberg) Cleve 1 _ __ F F R R F Cymbella cf. C. gracilis (Rabenhorst) Cleve.. F turgida (Gregory) Cleve *_ R c Denticula elegans Kiitzing kittoniana Grunow A A R A F A A F tenuis var. crassula (Naegli) Hustedt F F Epithemia turgida var. granulata (Ehrenberg) R F Fragilaria construens var. F Gomphonema intricatum F lanceolatum Ehrenberg. . lanceolatum var. insignia (Gregory) R F longiceps var. gracilis R Mastogloia atacamae R Mastogloia elliptica F R C elliptica var. dansei (Thwaites) Cleve F F C R Melosira spinigera Hustedt *. R F cryptocephala Kiitzing. . cryptocephala var. veneta (Kiitzing) F F R F F R lanceolata (Agardh) R pupula var. capitata F F cf. M. simplex Krasske . R R R Nitzschia cf. N. sublinearis R SP R See footnotes at end of table. Table 78.3.—Diatoms identified from the Lluta Formation in the Arica area—Continued [A, abundant; C, common; F frequent; E, rare. Numbers in boxheads are USGS diatom localities] Diatom 8 km northeast of Arica Boca Negra 5274 5276 5279 5280 5282 5284 5286 Pinnularia major (Kiitzing) Cleve. _ _ _ _ _ _ R viridis (Nitzsch) KUtzing ... . R sp _ _ _ R Rhopalodia gibberula F gibberula var. rupestris R musculus (Kiitzing) Muller _ _ R Synedra tabulata (Agardh) Kiitzing R ulna (Nitzsch) Ehrenberg 1 C C C R c ■ Also occur in lacustrine limestone of the Calama basin. REFERENCES Ahlfeld, Friedrich, 1946, Geologia de Bolivia: La Plata, Argentina, Mus. La Plata Rev. (nueva serie), Seccidn Geol., v. 3, p. 3-370. Bertrand, A., 1885, Memoria sobre las Cordilleras del Desierto de Atacama i rejiones limitrofes: Santiago, Chile, Imprenta Nac. Bowman, Isaiah, 1909, The physiography of the central Andes: Am. Jour. Sd., 4th ser., v. 29, no. 165, p. 197-217. Bruggen, Juan, 1918, Infrome sobre el agua subterr&nea de la rejion de Pica: Santiago, Chile, Soc. Nac. de Mineria. ------1950, Fundamentos de la geologia de Chile: Santiago, Chile, Instituto Geogrdfico Militar. Dingman, R. J., 1963, Quadrangulo Tulor, Antofogasto Province, Chile: Santiago, Chile, Instituto Inv. Geol., Carta Geol. de Chile, v. 4 [In press] Douglas, J. A. 1914, Geological sections through the Andes of Peru and Bolivia, pt. I—From the coast of Africa in the north of Chile to La Paz and the Bolivian “Yungas”: Geol. Soc. London Quart. Jour., v. 70, p. 1-53. Frenguelli, Joaquin, 1936, Diatomeas de la Caliza de la Cuenca de Calama: Mus. la Plata Rev. (nueva serie), Seccion paleontologia, v. I, p. 3-34. ------1938, Analisis microscopico del tripoli de Arica: Santiago, Chile, Dept. Minas y Petroleos, Ministerio de Fomento. Galli, C. O., and Dingman, R. J., 1962, Quadrangulos Pica, Alca, Matilla y Chacarilla: Santiago, Chile, Instituto Inv. Geol., Carta Geol. de Chile, v. 3, nos. 2, 3, 4, and 5, 125 p., 11 pi. Groeber, Pablo, 1957, Chile, pt. 7 of Amerique Latine, in HoflE-stetter and others, I.exique Stratigraphique International: Paris, France, Comm. Stratigraphy, Intemat. Geol. Cong., v. 5, p. 195. Jenks, W. F., 1948, Geologia de la Hoja de Arquipa: Lima, Peru, Direccidn de Minas y Petroleo Bull. 9.Article 79 POSSIBLE PLEISTOCENE-RECENT BOUNDARY IN THE GULF OF ALASKA, BASED ON BENTHONIC FORAMINIFERA By PATSY B. SMITH, Menlo Park, Calif. Abstract.—Of 10 cores from sediments in the Gulf of Alaska, 7 contained boreal faunas throughout, similar to those living in the area today, and 3 contained boreal foraminiferal faunas at the top and Arctic faunas in the lower part. It is inferred from the cores that the change in faunas marks a possible Pleistocene-Recent boundary in the sediments. In the spring of 1961,10 cores were taken by scientists on the U.S. Coast and Geodetic Survey ship Pioneer from sediments in the Gulf of Alaska, west and southwest of Kodiak Island (fig. 79.1). The coring was done under the direction of Lt. Comdr. H. P. Nygren, oceanographer of the Pioneer, and the samples were examined by G. W. Moore, U.S. Geological Survey, Menlo Park, Calif. Eight cores were obtained with a Phleger coring tube and two with a modified Ewing piston coring tube. Five of the cores are from the Continental Shelf (at depths of 76 to 240 meters) and five are from the north scarp of the Aleutian Trench (at depths of 810 to 5,540 meters) (table 79.1). The top centimeter of each core was preserved in ethanol so that living Foraminifera could be recognized Table 79.1.—Location and length of cores and description of the top centimeter of sediment Location Core Core No. Lati- tude (north) Longi- tude (west) Depth (m) length (cm) Sediment Color 1 1__ 57°18' 155°20' 230 56 Dark greenish gray. Olive gray. Grayish olive. Do. 3 54°33i 157°24' 2,070 6 4 55°31' 156°16' 240 122 5 65°17' 155°09' 1,950 33 6.. 55°52' 154°25' 810 54 Clayey very fine sand. Grayish olive green. Dark greenish gray. Olive gray. Dark greenish gray. Do. 7 56°25' 155°36' 76 15 8 55°36' 158°23' 146 101 9 54°55' 157°59' 117 72 Pebbly medium sand. 10 54°51/ 155°24' 4,170 13 11 54°27' 155°23' 5,540 43 Olive gray. 1 Color of the wet sediment follows the convention of Goddard and others (1948) 158°00' ______156°00'__________154°00' Figure 79.1.—Map of part of the Gulf of Alaska, showing station locations. Depth contours in meters. by a stain test for protein. Generally, 1-centimeter samples were taken at 10-centimeter intervals from the remainder of each core. From these samples, distribution of successively older faunas was determined and the age and depositional environment of the cored sediments were interpreted. ART. 78 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C73-C77. 1963. C73C74 STRATIGRAPHY Living and dead faunas (table 79.2) from the tops of all the cores are similar to those from the boreal waters of the Continental Shelf and Slope off Washington, Oregon, and California. A few species characteristic of arctic waters are present in the samples from the Continental Shelf and Slope in the Gulf of Alaska, but most forms characteristic of Arctic water are not present (Phleger, 1952; Loeblich and Tappan, 1953; Green, 1960). The shallow (76-240 meters) benthonic faunas of the shelf samples (cores 9, 8, 7, 4, 1) are similar to those of the Continental Shelf off Washington, Oregon, and northern California (Enbysk1; Bandy, 1953). Characteristic species are most of the Nonions and Nonionel- 1 Enbysk, B. 1960, Distribution of Foraminifera in the northern Pacific: Univ. Washington, Ph. D. thesis. las, Bulirwinella elegantissima, Uvigenna hollicki, Virgulina pauciloculata, and Epistonvinella pacifica (generally a bathyal species off California). The present bathyal faunas in the Gulf of Alaska (cores 6,5,3) (810-2,070 meters) have a wider distribution than the shallow faunas; many of the species occur at similar depths not only off Washington, Oregon, and California, but also off the coast of Central America (Smith, 1963). Included in these faunas are Bolivina argentea, B. spissa, BvZimm/i marginospinata, B. mba-cwninata, and most commonly, Uvigerina peregrina. The present abyssal faunas (cores 10, 11) (4,170 and 5,540 meters) are composed almost entirely of arenaceous species, very similar to faunas found in deep north Pacific waters from the Aleutian Trench to Hawaii. Table 79.2.—Foraminifera in the top centimeter of each core [Abundance given as percentage of dead faunas. Number of living specimens shown in parentheses] Depth zone Shelf Bathyal Abyssal Core No. 7 8 1 4 6 5 3 10 11 Species Depth (m) 76 117 146 230 240 810 1,950 2,070 4,170 5,540 Arenaceous Adercotrema glomeratum (Brady) 1 2. 5 < 1(4) <1 1 Alveophragmium nitidum (Goes) __ 4. 5 1 1 15 8 A. ringens (Brady) __ __ <1 1 A. subglobosum (G. O. Sars) 10 A. weisneri (Parr) _ _ _ <1 9 Ammobaculites agglutinans (d’Orbigny) A. cf. A. americanus (d’Orbigny) __ 3 A. agglutinans filiformis Earland- _ . 6. 5 Ammoglobigerina globigeriniformis (Parker and Jones) _ __ <1 10 2 Ammo mar ginulina foliacea (Brady) _ 5 2 Bigenerina minutissima (Earland) <1 <1 2 <1 C. trullissata (Brady) __ _ <1 5 <1 5 Eggerella bradyi (Cushman) 7 10 30 1 1 1 2(1) 2 7 1. 5 <1 2 1 Frag. A 2 <1 <1 <1 Placopsilina bradyi Cushman and McCulloch- 7 6 1 <1 R. difflugiformis Brady __ _ _ _ _ <1 2 10 16 2 2. 5 1 2 5. 5 Spiroplectammina biformis (Parker and Jones). 1 5 <1 1 1 5 i <1 <1 Trochammina grisea Heron-Alien and Earland. <1 <1 1 <1 KD 2 T. cf. T. malovensis Heron-Alien and Earland. 1 See footnote at end of tableSMITH C75 Table 79.2.—Foraminifera in the top centimeter of each core—Continued [Abundance given as percentage of dead faunas. Number of living specimens shown in parentheses] Depth zone Shelf Bathyal Abyssal Core No. 7 9 1 8 1 4 6 6 3 10 11 Species Depth (m) 76 117 146 230 240 810 1,950 2,070 4,170 5,540 Calcareous benthonic (1) 5(4) 2 5. 5 1 <1(D 1 2 <1 10 (3) 2 <1(1) 9(4) 2. 5 <1 1 (3) 1(2) <1 - •S 5; { EXPLANATION TS ■§ IS,® <0, S o 5.2 il-2 s;a o. 1 . Qag Alluvium and glacial deposits Rhyolite lava flows of Madison Plateau Showing form lines Qrri3, youngest flow Qm2, intermediate flow Qm i, oldest flow Rhyolite lava flows of Central Plateau , vent area II!!! Obsidian dome Qb Basalt flows and basaltic andesite Rhyolite tuff of Yellowstone Plateau Tp-C Oligocene, Cretaceous, and Precambrian rocks > O. < Z I- < D o Contact STRUCTURES OF LATE QUATERNARY AGE D ........ Normal fault Dotted where concealed. 0, doumthrown side Syncline 6 x Analyzed specimen (see table) Fiqube 80.1.—Geologic map of the northwest part of the Yellowstone Plateau, Idaho, Montana, and Wyoming. Geology by Warren Hamilton in 1959, with additions from Boyd (1961) and W. Bradley Myers (written communication, 1962).C80 GEOCHEMISTRY, PETROLOGY, AND MINERALOGY Chemical analyses of hasalt and rhyolite of the northwest part of Yellowstone Plateau [Major oxides determined by rapid colorimetric methods by Paul Elmore, I. Barlow, S. Botts, and O. Chloe, Washington, D.C., 1961. Fluorine and chlorine determined by Vertie C. Smith, Denver, Colo., 1961. Other minor elements determined by semiquantitative spectrographic methods by Paul R. Barnett, Denver, Colo., 1961; values reported as midpoints of logarithmic-third divisions] Basalt and basaltic andesite Rhyolite ash-flow tuff Rhyolite lava-flow and dome rocks Averages of column 4-11 Range of columns 4-11 1 2 3 4 5 6 7 8 9 10 11 12 13 Field number Laboratory numbers YS 6 H 3303 158059 YS 20C H 3310 158055 YS 5 H 3302 158047 YS 4 H 3301 158046 YS 20A H 3309 158054 YS 8 H 3305 158050 YS 24 H 3311 158056 YS 26A H 3313 158058 YS 7A-2 H 3304 158049 YS 3 H 3300 158045 YS 27 H 3314 158048 Major oxides, in weight percent SiO, 46.7 48.1 51.5 75.8 76.1 77.3 77.3 75.7 75.8 76.4 76.7 76.4 75.7-77.. All(>3 15.7 15.7 15.5 12.4 12.7 12.5 12.2 11.9 12.0 12.2 11.9 12.2 11.9-12.' FeiOj 1.2 4.2 2.0 1.2 1.7 .8 1.3 .5 .8 .5 .5 .9 \0.9-1.9 as FeO 13.0 9.4 9.0 .40 .10 .17 .08 1.0 1.2 .84 1.2 .6 /FeO MgO 6.8 6.8 6.0 .15 .16 .07 .08 .06 .18 .21 .13 .13 0.07-0 CaO 9.3 9.2 8.5 .27 .34 .24 .22 .34 .57 .57 .38 .4 0. 2-0. Na,0 3.2 3.3 3.0 3.3 3.3 3.3 3.6 3.3 3.4 3.0 3.4 3.3 3. 0-3. K,0__ .46 .56 1.1 5.0 5.0 4.8 4.6 5.0 4.9 5.0 5.0 4.9 4. 6-5. HiO .30 .18 1.2 .72 .30 .58 .30 1.1 .32 .68 .60 .6 0. 3-1. TIO, 2.4 2.2 1.8 .14 .14 .10 .10 .12 .20 .13 .14 .13 0.1-0. P.Oj .39 .33 .28 .02 .02 .02 .02 .02 .02 .04 .01 .02 0. 01-0.O' MnO .22 .21 .17 .03 .04 .02 .04 .04 .06 .03 .04 .04 0.02-0. 0( 99.7 100.2 100.1 99.4 99.9 99.9 99.8 99.1 99.5 99.6 100.0 Minor elements, in weight percent Ba 0.03 0.03 0.07 0.07 0.07 0. 003 0.007 0.015 0.03 0.07 0. 015 0.04 0. 007-. 07 Be <. 0001 <.0001 <.0001 .0003 .0003 .0003 .0003 .0003 .0003 .0003 .0003 .0003 .0003 Ce <■01 <.01 <.01 .015 .015 <.01 <01 .015 .015 .015 .015 ~. 01 <. 01-. 015 Cl .01 .01 .02 .01 .03 .02 .02 .09 .08 .08 .09 .05 . 01-. 09 Co .007 .007 .007 <.0002 <.0002 <.0002 <. 0002 <.0002 <.0002 <.0002 <. 0002 <. 0002 Cr .015 .015 .015 <•0001 .00015 <■0001 <.0001 <. 0001 .00015 <. 0001 <.0001 <■0001 <. 0001-. 00015 Cu .007 .007 .003 .00015 .003 .00015 .00015 .00015 .0003 .0003 .00015 ~. 001 . 00015-. 003 F .03 .03 .03 .01 .01 .03 .03 .14 .14 .12 .14 .08 . 01-.14 Ga .0015 .0015 .0015 .0015 .0015 .0015 .0015 .003 .0015 .003 .0015 .002 . 0015-. 003 La <. 002 <. 002 <. 002 .007 .007 .007 .003 .007 .015 .007 .015 .008 . 003-. 015 Mo <.0002 <.0002 <. 0002 .0003 <. 0002 .0003 .0003 .0003 .0003 .0003 .0003 .0003 <. 0002-. 0003 Nb <.0005 <.0005 <.0005 .003 .003 .003 .003 .003 .003 .0015 .003 .003 . 0015-. 003 Nd___ <. 005 <. 005 <. 005 .007 .007 <005 <.005 .007 .015 <. 005 .015 ~. 004 <•005-. 015 Ni .007 .015 .007 <.0002 <. 0002 <.0002 <. 0002 <.0002 Locations of springs shown on fig. 83.1 of Article 83 as numbered water-sampling localities. > Equivalents per million = parts per million X 1/combinlng weight.BRADDOCK AND BOWLES C99 mosphere) of carbon dioxide, dolomite was many times more soluble than calcite. Dolomite and gypsum were unstable, and the reaction CaC03 • MgC03 + CaS04±^2CaC03 + MgS04 (+ water) occurred, causing calcitization of dolomite and the accumulation of magnesium sulfate in solution. REFERENCES Gott, G. B., and Schnabel, R. W., 1963, Geology of the Edge-mont NE quadrangle, South Dakota: U.S. Geol. Survey Bull. 1063-E, p. 127-190. Yanat’eva, O. K., 1956, The nature of the solubility of dolomite in water and in calcium sulfate solutions at different partial pressures of CO*: Zhur. Neorg. Khim. Trans., v. 1, no. 7, p. 1473-1478.Article 85 APATITIZED WOOD AND LEUCOPHOSPHITE IN NODULES IN THE MORENO FORMATION, CALIFORNIA By R. A. GULBRANDSEN, D. L. JONES, K. M. TAGG, and D. W. REESER, Menlo Park, Calif. Abstract.—Nodules containing apatitized wood fragments occur in the Moreno Formation of Cretaceous age. Pyrite, deposited in open space in the wood, and apatite are primary minerals formed in a marine environment; leucophosphite and gypsum are secondary minerals formed during weathering. X-ray data for leucophosphite are presented for this new occurrence of the mineral. Apatitized fossil wood in itself is rare, but in association with leucophosphite in gypsum-encased nodules, as described here, it is unique. Simpson (1912) reviewed early reports of phosphatized fossil wood, and since then only a few occurrences have been described (Simpson, 1920; Read, 1936; Hofmann, 1944; Maslennikov and Kavitskaya, 1956; Goldberg and Parker, 1960). The wood in the samples described here has been replaced by carbonate fluorapatite, hereinafter referred to as apatite. Leucophosphite, a hydrous potassium ferric phosphate, is new in this mode of occurrence and previously has been found only in a pegmatite (Lindberg, 1957) and in deposits formed by reactions of iron-rich materials with bird guano (Simpson, 1931-32) and bat dung (Axelrod and others, 1952). The Moreno Formation, in which the fossil wood occurs, is of latest Cretaceous and earliest Paleocene (?) age. It crops out in the foothills along the western side of San Joaquin Valley, mainly in Fresno and Merced Counties, Calif. It consists of several thousand feet of dominantly purple and chocolate-brown organic-rich mudstone with intercalated sandstone lenses. Marine fossils occur throughout the section, although preservation of megafossils is generally poor. Foraminifera are locally abundant, and several rich floras of diatoms have been described (Hanna, 1927, 1934; Long and others, 1946). The specimens of fossil wood were obtained from the lower part of the Moreno at the head of Escarpado Canyon, Fresno County (figs. 85.1 and 85.2), where Payne (1941; 1951) divided the formation, in ascending order, as follows: Dosados Sand and Shale, Tierra Loma Shale, Marca Shale, and Dos Palos Shale Members (fig. 85.3). The fossil wood was obtained from the upper part of the Dosados Sand and Shale Member and the lower part of the Tierra Loma Shale Member. 124° 00' 120° 00' Figure 85.1.—Index map of California showing location of described area. C100 ART. 85 IN U.s. GEOL. SURVEY PROF. PAPER 475-C, PAGES C100-C104. 19«3.GULBRANDSEN, JONES, TAGG, AND REESER C101 R. 11 E. R. 12 E. EXPLANATION **. © 111 S'* Km Moreno Formation Kp Panoche Formation (/) D O l UJ f O < \— LU cr o Contact Mapped by Payne (1951) X2 Fossil-wood locality Figure 85.2.—Map showing location of fossil-wood localities in the Moreno Formation, Fresno County, Calif. The wood is not ideal for fossil identification, as it is poorly preserved. Two of the better preserved specimens, examined by R. A. Scott, U.S. Geological Survey, could be identified with certainty only as dicotyledons; one is provisionally assigned to the family Sapotaceae, genus Chrysophyllum, a form now found only in tropical and subtropical regions. The fossil wood occurs in gypsum-encrusted nodules in brown mudstone, and the nodules have considerable range in size and shape. The largest found was a slab 1X11/2 feet in largest dimensions; the longest, a branch, was 1 y2 feet long. A typical nodule (fig. 85.4) is composed of three parts: a core of fossil wood, an earthy zone, and a crust of gypsum. Nodules exposed to surface weathering usually lack all or part of the gypsum crust. The core of a nodule is composed principally of gypsum, pyrite, and apatite. Apatite generally preserves in detail the cell structure of the wood (fig. 85.5) and apparently is an initial replacement mineral. Under the microscope most of it is seen as brown submicro-crystalline particles that are isotropic in aggregate; a very small amount is a colorless anisotropic filling of microscopic fractures. Some brown apatite masses reveal no cell structure. Pyrite is associated with the apatite and is regarded as contemporaneous; it occupies cell space and larger openings. Gypsum occurs as light-brown fine-grained crystals in veins and larger replacements of the apatitized wood, rarely preserving the cell structure. In some nodules, bassanite occurs in place of gypsum. Leucophosphite in the core is rare. Although the dissolution of apatite and pyrite is complete in many cores, a gross wood structure is still evident. The earthy zone partly encases the core material and differs markedly from the enclosing mudstone in that it characteristically is a tan soft aphanite composed principally of cristobalite-opal, quartz, leucophosphite, and gypsum. The composition ranges widely, and either cristobalite-opal or brown gypsum may constitute most of the earthy matrix in some nodules. Montmoril-lonite and jarosite are present in small amount in the earthy zone of some nodules. Leucophosphite, which is regarded as a secondary mineral, occurs as spherulites and pellets that commonly are concentrated in a layer immediately beneath the gypsum crust but also are found in fractures and other areas within the earthy zone. Two unidentified minerals, one common but not abundant, appear to be confined to the earthy zone. The crust is composed of white coarsely crystalline gypsum. The crystals are elongate perpendicular to the surface of the earthy zone, or wood core where the earthy zone is absent, thus delineating a sharp discontinuity between the crust and the rest of the nodule. The mudstone in which the nodules were found is composed principally of mixed-layer montmorillonite-illite, quartz, plagioclase, K-feldspar, and mica; traces of cristobalite-opal and gypsum also are present. Gypsum occurs in veins and is strikingly abundant in surface exposures. Samples of mudstone from drill cores are similar in composition to surface samples except that pyrite is present and gypsum is absent. All mineral identifications have been confirmed by X-ray analysis. X-ray data for the leucophosphite in the Moreno Formation are listed in the accompanying table along with data of the synthetic mineral and data from well-formed crystals in the Sapucaia pegmatite mine. The leucophosphite of the Moreno Formation is submicrocrystalline and produces a generally diffuse X-ray pattern. The d values and intensities of the principal peaks check well with the values measured on the other samples, despite the lack of sharp resolution and the partial interference, due to impurities, that is indicated for a few peaks.C102 GEOCHEMISTRY, PETROLOGY, AND MINERALOGY System Series Formation and member (after Payne, 1951)________ Lithologic description >- cr < 0) c a) o o _o> Q_ cn D O LlI O < H UJ cr O a> a a 3 Cima Sandstone Lentil Dos Palos Shale Member Marca Shale Member Tierra Loma Shale Member Dosados Sand and Shale Member Gray fine-grained massive friable silty sandstone; 80 feet thick Brown organic-rich shale; 800-900 feet thick Siliceous light-gray, creamy-white, or light-brown shale, varies from soft and punky to very hard. Contains abundant Siphogenerinoides whiiei; diatomaceous; 300 feet thick Brown shale with local sandstone dikes. Contains abundant fossil fish scales and other organic matter,- 1,150 feet thick Interbedded massive thick-bedded fine-grained light-yellow sandstone and brown shale. Sandstone dikes common; 200 feet thick Massive brown-weathering sandstone Figure 85.3.—Diagrammatic columnar section of Moreno Formation in Escarpado Canyon, Fresno County, Calif. Modified from Payne (1951).GULBRANDSEN, JONES, TAGG, AND REESER C1G3 I_____ 1 cm Figure 85.4.—Section of nodule showing apatitized wood, A; earthy zone with leucophosphite pellets, B; and gypsum crust. C. 'Figure 85.5.—Apatitized wood (family Sapotaceae?, genus Chry sophy Hum?) showing preservation of intricate wood structure. The origin of the nodules may be postulated from the nature and distribution of the minerals within the nodules themselves and from evidence in the sedimentary rocks in which the nodules occur. The presence of pyrite in the nodules and the abundant organic matter as well as pyrite in the surrounding sediments indicates that the wood was deposited in a reducing environ- X-ray diffraction data for leucophosphite [Lines less than 2.629 not included] Moreno Formation i Synthetic2 Sapucaia pegmatite mine, Brazil * d I d I d I 7. 55 6. 72 25 100 7. 50 6. 77 0. 31 1. 00 7. 60 6. 79 6. 09 5. 99 4. 76 4. 28 4. 21 4. 08 3. 79 3. 65 3. 54 3. 37 3. 25 3. 23 3. 06 3. 017 2. 990 2. 956 2. 916 2. 829 2. 685 2. 655 3 10 1 7 3 2 2 2 3 1 1 »3 X X 7 1 1 1 4 4 1 3 5. 94 4. 70 4. 22 4. 06 3. 76 95 30 1 35 « 40 20 5. 92 4. 73 f 4. 23 l 4. 20 4. 05 3. 78 . 76 . 28 . 21 . 28 . 14 . 21 3. 34 « 65 3. 34 3. 20 3. 09 3. 03 / 3.00 \ 2. 97 . 35 . 14 . 14 . 59 . 28 . 28 3. 03 2. 970 55 40 2. 869 2. 798 -2. 672 2. 629 25 35 30 30 2. 90 2. 81 2. 66 2. 64 . 37 . 41 . 35 . 37 i Escarpado Canyon, SW^SWt* sec. 6, NWHNWM sec. 7, T. 15 S., R. 12 E., Fresno County, Calif. X-ray powder analysis on wide-range diffractometer, Ni-flltered Cu Kai radiation, (X=1.54050A). * Haseman and others, 1950, p. 80. Product 1. s Lindberg, 1957, p. 219. 4 Peak broad and intensity enhanced by cristobalite-opal and quartz impurities. 8Intensity enhanced by quartz impurity. • Broad. ment that had a pH lower than that of ordinary sea water. In addition, the wood itself probably created a local environment that favored deposition of apatite. Limited data from a modem occurrence of apatitized wood on the Pacific sea floor (Goldberg and Parker, 1960, p. 631) show that the bottom water contains almost no oxygen and is rich in phosphate. The cristobalite-opal composition of the earthy zone probably was derived from diatoms or other siliceous organisms that were attached to the wood before it sank or that drifted against it on the bottom. The abundance of such organisms is a feature of oceanic areas of modern phosphate deposition (McKelvey, 1959, p. 1783). Leucophosphite and gypsum represent in large part only a recombination of the elements already present in the phosphatized wood when uplift of the enclosing Moreno Formation exposed it to weathering. Oxidation of the pyrite locally produced sulfuric acid, which dissolved the apatite. Ensuing reactions produced the new solid phases, leucophosphite and hydrous calcium sulfate.C104 GEOCHEMISTRY, PETROLOGY, AND MINERALOGY REFERENCES Axelrod, J. M., Carron, M. K., Milton, Charles, and Thayer, T. P., 1952, Phosphate mineralization at Bomi Hill and Bambuta, Liberia, West Africa: Am. Mineralogist, v. 37, p. 883-909. Goldberg, E. D., and Parker, R. H., 1960, Phosphatized wood from the Pacific sea floor: Geol. Soc. America Bull., v. 71, p. 631-632. Hanna, G. D., 1927, Cretaceous diatoms from California: California Acad. Sci. Occasional Papers, no. 13, 48 p. ------1934, Additional notes on diatoms from the Cretaceous of California : Jour. Paleontology, v. 8, p. 352-355. Haseman, J. F., Lehr, J. R., and Smith, J. P., 1950, Mineralogieal character of some iron and aluminum phosphates containing potassium and ammonium: Am. Soil Science Soc. Proc., v. 15, p. 76-84. Hofmann, Elise, 1944, Pflanzenreste aus dem Phosphoritvorkom-men von Prambachkirchen in Oberdonau: Palaeonto- graphica, v. 88, pt. B, p. 1-81. Lindberg, M. L., 1957, Leucophosphite from the Sapucaia pegmatite mine, Minas Gerais, Brazil: Am. Mineralogist, v. 42, p. 214-221. Long, J. A., Fuge, D. P., and Smith, James, 1946, Diatoms of the Moreno Shale: Jour. Paleontology, v. 20, p. 89-118. McKelvey, V. E., 1959, Relation of upwelling marine waters to phosphorite and oil [abs.] : Geol. Soc. America Bull., v. 70, p. 1783. Maslennikov, B. M., and Kavitskaya, F. A., 1956, O fosfatom veshchestve fosforitov [The phosphate substance of phosphorites] : Doklady Akad. Nauk SSSR, v. 109, no. 5, p. 990-992 [in Russian], Payne, M. B., 1941, Moreno shale, Panoche Hills, Fresno County, California [abs.] : Geol. Soc. America Bull., v. 52, p. 1953-1954. ------1951, Type Moreno Formation: California Div. Mines Spec. Rept. 9,29 p. Read, C. B., 1936, The flora of the New Albany shale, pt. 1, Diichnia kentuokiensis, a new representative of the Calamopityeae: U.S. Geol. Survey Prof. Paper 185-H, p. 149-162. Simpson, E. S., 1912, Unusual types of petrifaction from Dan-darragan: Nat. History and Sci. Soc. Western Australia Jour., v. 4, p. 33-37. ------1920, On gearksutite at Gingin, Western Australia: Mineralog. Mag., v. 19, p. 23-39. ------1931-32, Contributions to the mineralogy of western Australia—series VII: Royal Soc. Western Australia Jour., v. 18, p. 61-74.Article 86 VARIATION IN ELEMENT CONTENT OF AMERICAN ELM TISSUE WITH A PRONOUNCED CHANGE IN THE CHEMICAL NATURE OF THE SOIL By HANSFORD T. SHACKLETTE, Denver, Colo. Abstract.—Chemical analysis of ash of an American elm tree whose base was flooded by galena and sphalerite tailings indicated that phosphorus and zinc were concentrated in the living wood, that calcium was depleted, and that potassium was concentrated in both living wood and nonliving cells. Two large American elm (Vlmm americana L.) trees growing under sharply contrasting conditions were analyzed to detect differences in the element content of their tissues. The trees, near Tennyson, Grant County, in southwestern Wisconsin, were growing on a slope within 150 feet of each other and were 3(4 feet in diameter at breast height. One was growing in normal soil, whereas the other had been enclosed on the lower side of the slope by an earth and rock dam to make a settling basin for tailings from an ore mill located a few hundred yards upslope. The dam was constructed and the basin put in use in 1952, 10 years before the field study reported here. Fine tailings suspended in water, resulting from the crushing of lead (galena) and zinc (sphalerite) ores and from their separation in the milling process, had been poured into the basin. The water had evaporated and the “fines” had settled into a firm deposit composed principally of dolomite, estimated to be 5 feet thick around the trunk of the enclosed tree. A partial chemical analysis of this soil material is given in table 86.1. The change in soil level over the roots of the tree and the periodic flooding is slowly killing the tree, probably because of the effect of reduced soil aeration on the roots and not because of the heavy-metal content of the deposit. At the time of examination the last 2 years of growth had been greatly retarded, as determined by the length of branch-growth increments and the width of annual-growth rings. In addition, the leaves were markedly sparse and chlorotic; more so in 1962 than in 1961. The adjacent elm tree on unaffected soil was vig- Table 88.1.—Chemical analyses of American elm samples and soil samples [J. B. McHugh, J. H. Turner, and S. L. Power, U.S. Geological Survey, analysts] Sample Ash (percent of dry weight) Ca K P Cu Pb Zn Percent of ash, by weight Wood: 1.5 20 23.0 1.50 0.008 0.0050 0.070 1.3 16 35.0 .60 .012 .0075 .050 2.5 41 9.4 1.20 .006 .0025 .060 1.0 19 25.0 .60 .015 .0100 .050 Branches (1-4 years old): 4.5 25 12.0 2.40 .008 .0150 .400 5.0 32 7.0 1.20 .004 .0150 .050 4.8 .009 .0080 .073 Leaves: 7.0 16 16.0 2.40 .006 .0050 .070 12.0 20 16.0 1.80 .004 .0025 .010 13.3 .004 .0025 .021 Percent of dry weight 16 2.0 0.15 0. 015 0.0200 1.000 Soil near normal tree, Aj 6 1.4 .06 .001 .0150 .020 * From 22 other American elm trees growing under normal soil conditions in Wisconsin. orous and healthy, judging from its growth increments and leaf color. Samples of the tree trunk at breast height were taken separately of the wood formed during the last 10 years and that formed during the 10 years before flooding to ascertain whether the change in chemical composition of the soil was reflected in the element content of the wood. An increment borer was used to determine, by ring count, the thickness of the wood formed during these 2 periods (214 And 2 Indies, respectively). A 1-inch-diameter wood auger was then used to remove the wood for chemical analysis from that part of the trunk that grew during these two periods. The same procedure was followed in sampling the normal tree, the wood increments for the 2 periods being 2(4 and 2 inches. In addition, samples of the leaves and of current to 4-year-old branches were collected from the 2 trees. The ART. 86 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C105-C106. 1963. 0105C106 GEOCHEMISTRY, PETROLOGY, AND MINERALOGY A2 soil horizon from near the normal tree was also sampled. The tissue samples were air dried and then burned to ash in an electric oven in which the heat was increased 50°C per hour to 550°C and then held constant for 14 hours. The ash was analyzed by colorimetric methods (Ward and others, 1963) for calcium, potassium, phosphorus, copper, lead, and zinc. The sample of fines from the tailings and the sample of normal soil were sieved without grinding, and the minus-80-mesh fractions were analyzed by colorimetric methods. Results of these analyses are presented in table 86.1, as are also the averages of tissue analyses of 22 American elm trees which are growing under normal soil conditions in Wisconsin and which are cited to demonstrate the validity of using the normal tree of this study as a control. Results.—The following results were noted in the study: 1. Flooding the American elm tree with ore-mill tailings resulted in a greater percentage of phosphorus and zinc in the ash of wood formed after the flooding occurred, but had no effect on the percentage of these elements in ash of the wood formed before flooding, as compared with the percentage in wood ash of the normal tree (table 86.2). Table 86.2.—Increase or decrease in percentage of elements in flooded tree samples as compared with normal tree samples [Percentage of ash by weight. +, increase; —, decrease] Sample Ca K P Cu Pb Zn Wood, 1952-61 Wood, 1942-51 -21 -3 -7 -4 +13.6 +10.0 +5.0 0 +0.30 0 +.12 +.60 +0.002 -.003 +.004 +.002 +0.0025 -.0025 0 +.0025 +0.01 0 +.35 +.60 Leaves - 2. The percentage of calcium in the ash of all tissues was lower in the flooded tree (table 86.2), although the amount of this element in the soil was increased almost threefold (table 86.1). No explanation is offered for the reduction in calcium content of the tissue. 3. The percentage of potassium in the ash of both recently formed and older wood was significantly greater in the flooded tree (table 86.2). This indicates that the transport and deposition of this element is not exclusively related to metabolic processes of the tree, for the gain in potassium of the older wood presumably took place after the wood cells were dead. Wood cells of this age (10 years and older) are generally assumed to be nonliving, and in this tree the older wood sample was entirely “heartwood.” 4. The differences in the copper and lead content of the ash appear to have no clear-cut relation to the increased amounts of these elements in the soil. The indicated increase of these elements in recently formed wood of the flooded tree compared with the amount in the normal tree is about the same as their decrease in the older wood (table 86.2). The apparent reduction of these elements in the ash of recently formed wood of the flooded tree as compared with the older wood of the same tree (table 86.1) is likewise probably within the limits of experimental error. 5. An increase in the zinc content of the soil resulted in an eightfold increase in the percentage of zinc in ash of the branches and a sevenfold increase in ash of the leaves (table 86.1). This is the most pronounced change in element content of any of the plant tissues. Conclusions.—The addition of zinc and phosphorus to the soil in which the tree was growing resulted in an increase of these elements only in wood formed subsequently. Therefore, chemical analysis of tree rings may be a useful method of determining the year when the trees were exposed to increased amounts of these elements in the soil. In geochemical exploration it is important to know if a chemical anomaly in the soil is caused by naturally occurring element concentration or by manmade soil contamination; if the date of this increase in element content can be determined, it may be possible to determine the cause. If anomalies in certain elements at a site can be attributed to habitation by man, one may suspect that anomalies in some other elements are due to the same cause. REFERENCES Ward, F. N., Lakin, H. W., Canney. F. C., and others, 1963 Analytical methods used in geochemical exploration by the U.S. Geological Survey: U.S. Geol. Survey Bull. 1152, 100 p.Article 87 ORDOVICIAN AGE FOR SOME ROCKS OF THE CAROLINA SLATE BELT IN NORTH CAROLINA By A. M. WHITE; ARVID A. STROMQUIST; T. W. STERN, and HAROLD WESTLEY; Washington, D.C.; Denver, Colo.,- Washington, D.C. Work done in cooperation with the North Carolina Department of Conservation and Development, Division of Mineral Resources Abstract.—Two lead-alpha age determinations on zircon from felsic crystal-lithic tuff in part of the Carolina slate belt indicate an Ordovician age for these rocks. The results add to growing evidence of a Paleozoic rather than a Precambrian age for much of the southeastern Piedmont province. Lead-alpha ages have been determined for zircon from two localities in the Carolina slate belt in the eastern part of the North Carolina Piedmont. Both samples of zircon are from the southeastern part of the Albemarle quadrangle (fig. 87.1), and the indicated age for each is Ordovician according to the Holmes time scale (Holmes, 1959, p. 204). Results of analyses and age calculations for the zircon are given in the accompanying table. Lead-alpha age determinations of zircon from rocks of the Carolina slate belt Alpha counts per milligram per Sample No hour Pb(ppm)1 Calculated age 2 (m.y.) ZU-1___________ 52 9.6 440±60 ZU-2___________ 201 40 470±60 1 Average of duplicate determinations ZU-1 analyzed by Charles Annell and Harold Westley ; ZU-2 by Harold Westle.v and I. H. Barlow. 2 Ages (rounded to nearest 10 million years) calculated from the equations: t = C Pb/a where t is the calculated age, in millions of years; C is a constant based on the U/Th ratio and has a value of 2,485 ; Pb Is the lead content, in parts per million ; and a is the alpha counts per milligram per hour ; and T=t— where T is the age, In millions of years, corrected for decay of uranium and thorium; and k is a constant based on the U/Th ratio and has a value of 1.56 XlO-A The Carolina slate belt of low-grade metasedimentary and metavolcanic rocks extends from Virginia south- ward across the Carolinas into Georgia, where the rocks are known as the Little River Series (Crickmay, 1952, p. 31). To the east it is bounded by rocks similar to those of the Charlotte belt (King, 1955, p. 346-350), by discontinuous outcrops of the Newark Group of Triassic age, and by overlap deposits of the coastal plain of Cretaceous and younger age. To the west, rocks of the slate belt are truncated by faults and metamorphosed along the border of a complex of plutons of the Charlotte belt. Only rocks of the greenschist facies are commonly considered slate-belt rocks. They comprise a sequence of alternating rhyolitic to basaltic volcanic rocks inter-bedded with argillite, siltstone, sandstone, and their tuffaceous equivalents. The sequence is intruded by granitic to gabbroic sills, stocks, and batholiths of probable Paleozoic age (King, 1955, p. 349), and by Triassic diabase dikes that trend northwest to northeast. Felsic flows and tuffs were selected as the most likely rocks to yield zircon, the sole mineral in these rocks on which age determinations are possible. Only 1 of 12 units sampled, however, yielded enough zircon for analysis. This unit, which is in the southeastern part of the Albermarle quadrangle, is part of a sequence of acid lithic tuffs in the lower volcanic unit as mapped by Stromquist and Conley (1959, p. 15, 16, and map) and by Conley (1962). It is the lowest stratigraphic unit in the Albemarle and Denton quadrangles and consists of felsic lithic tuff and crystal tuff with inter-bedded rhyolite flows. The tuffs are cut by several rhyolite dikes. The unit is traceable from the southern part of the Albemarle quadrangle to the area north of 694-027 O—63- -8 ART. 87 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C107-C109. 1963. C107C108 GEOCHRONOLOGY 80° 78° cp Deposits of the coastal plain (Cretaceous and younger) vu Volcanic and sedimentary rocks of the Carolina slate belt (Paleozoic) Sedimentary rocks of the Newark Group (Triassic) cb Rocks of the Charlotte belt (probably Precambrian and Paleozoic) ZU-2X Sampled locality Figure 87.1.—Index map of central North Carolina showing localities sampled for zircon. Geology generalized from geologic map of North Carolina (Stuckey and Conrad, 1958) and from King (1955). Asheboro, and from the sedimentary rocks of the Newark Group on the east to the area near the Pee Dee River on the west (fig. 87.1). The structure appears to be that of a large southwest-plunging anticline. In the upper part of the unit, poor bedding and a lack of sorting of the constituent fragments and crystals suggest subaerial deposition. Therefore, this unit probably represents an old landmass built up from the sea floor by a series of volcanic eruptions and flows. The unit sampled is overlain by a younger sequence of volcanic and sedimentary rocks extending at least to the bound- ary of the slate belt with the Charlotte belt to the west. Eastward, the younger unit may extend to the sedimentary rocks of the Newark Group (fig. 87.1). The tuff was sampled at exposures in roadcuts along North Carolina Highway 27, 3.4 miles east of the Pee Dee River and 1 mile north of the junction of Clarks and Dumas Creeks (sample ZU-1), and 1.3 miles east of the Pee Dee River and 0.7 mile southeast of White Crest Church (sample ZTT-2). Both localities are in Montgomery County. About 700 pounds of saprolite of felsic crystal-lithic tuff was panned to obtain zircon for sample ZU-1, and about 300 pounds for sample ZU-2. The zircon obtained from the tuffs is euhedral, very fine grained, and pale pink. Detrital zircon of more than one generation and from more than one source is present in some slate-belt tuffs as well as in alluvial deposits in the area (White and Stromquist, 1961), but the zircon reported here is believed to be primary because of its euhedral crystal form and uniform grain size and color. Rocks of the Carolina slate belt heretofore have been dated only by inference and by speculation, for no definitive fossil material has been reported from them. Ordovician fossils have been found in the Arvonia and Quantico slates of Virginia, but the geologic relations of these rocks to the Carolina slate belt are not known. For many years the slate-belt rocks were considered Precambrian, and they were assigned a Precambrian age on the geologic map of the United States (Stose and Ljungstedt, 1932). Stuckey and Conrad (1958, p. 26-29; map) designated them as Precambrian or lower Paleozoic (?) on the geologic map of North Carolina. An increasing amount of evidence favors a Paleozoic age for the slate-belt rocks as well as for much of the piedmont terrane west of the slate belt (King, 1951, p. 119-144; 1955, p. 332-373; Crickmay, 1952, p. 31; Over-street and others, 1961, p. B104; Overstreet and others, 1962, p. C81). Moreover, stratigraphic equivalence of some units within the piedmont, such as the Carolina slate belt and the Kings Mountain belt to the west (King, 1955, map), has been suggested (Keith and Sterrett, 1931, p. 4; Kesler, 1936, p. 34; Overstreet and Bell, 1960, p. B199). The Ordovician age reported here for rocks of the Carolina slate belt supports the opinion that many of the rocks in the southeast piedmont are Paleozoic rather than Precambrian in age, and it permits for the first time the assignment of a definitive age to part of the Carolina slate belt.WHITE, STROMQUIST, STERN, AND WESTLEY C109 REFERENCES Conley, J. F., 1962, Geology of the Albemarle quadrangle, North Carolina: North Carolina Dept. Conserv. and Devel., Div. Mineral Resources Bull. 75, 26 p. Crickmay, G. W., 1952, Geology of the crystalline rocks of Georgia: Georgia Geol. Survey Bull. 58, p. 1-54. Holmes, Arthur, 1959, A revised geologic time scale: Edinburgh Geol. Soc. Trans., v. 17, pt. 3, p. 183-216. Keith, Arthur, and Sterrett, D. B., 1931, Description of the Gaffney and Kings Mountain quadrangles [South Carolina-North Carolina] : U.S. Geol. Survey Geol. Atlas, Folio 222, p. 1-13. Kesler, T. L., 1936, Granitic injection processes in the Columbia quadrangle, South Carolina: Jour. Geology, v. 44, no. 1, p. 32-42. King, P. B., 1951, The tectonics of middle North America: Princeton, N.J., Princeton Univ. Press, p. 3-203. ------1955, A geologic section across the southern Appalachians—An outline of the geology in the segment in Tennessee, North Carolina, and South Carolina, in Russell, R. J., ed., 1955, Guides to southeastern geology: Geol. Soc. America Guidebook, 1955 Ann. Mtg., p. 332-373. Overstreet, W. C., and Bell, Henry, 3d, 1960, Geologic relations inferred from the provisional geologic map of the crystalline rocks of South Carolina: Art. 87 in U.S. Geol. Survey Prof. Paper 400-B, p. B197-B199. Overstreet, W. C., Bell, Henry, 3d, Rose, H. J., Jr., and Stern, T. W., 1961, Recent lead-alpha age determinations on zircon from the Carolina Piedmont: Art. 45 in U.S. Geol. Survey Prof. Paper 424-B, p. B103-B107. Overstreet, W. C., Stern, T. W., Annell, Charles, and Westley, Harold, 1962, Lead-alpha ages of zircon from North and South Carolina: Art. 88 in U.S. Geol. Survey Prof. Paper 450-C, p. C81. Stose, G. W., and Ljungstedt, O. A., 1932, Geologic map of the United States: U.S. Geol. Survey. Stromquist, A. A., and Conley, J. F., 1959, Geology of the Albemarle and Denton quadrangles, North Carolina: Carolina Geol. Soc. Field Trip Guidebook, Oct. 24, 1959, North Carolina Div. Mineral Resources, Raleigh, N.C., 36 p. Stuckey, J. L., and Conrad, S. G., 1958, Explanatory text for the geologic map of North Carolina: North Carolina Dept. Conserv. Devel., Div. Mineral Resources Bull. 71, p. 3-51, map. White, A. M., and Stromquist, A. A., 1961, Anomalous heavy minerals in the High Rock quadrangle, North Carolina: Art 118 in U.S. Geol. Survey Prof. Paper 424-B, p. B278 and B279.Article 88 GRAVITY SURVEY IN THE RAMPART RANGE AREA, COLORADO By CARTER H. MILLER, Denver, Colo. Abstract.—Gravity interpretation indicates that the Rampart Range is bounded on the east by a high-angle reverse dip fault and that an outlier of sedimentary rocks on the west is about 2,000 feet thick. A gravity survey was made across the Manitou Park graben, Rampart Range, and the eastern foothills belt of Colorado (fig. 88.1) to obtain the depth and two-dimensional configuration of the sedimentary rocks in the graben and in the foothills belt. The configuration of the rocks in the foothills belt and in the graben is primarily controlled by the Rampart Range fault and the Ute Pass fault, respectively. The Rampart Range extends about 40 miles northwest from Colorado Springs, Colo., to the South Platte River. The range is about 10 miles wide and is bordered on the west by the Manitou Park graben, which is approximately 3 miles wide, and on the east by the eastern foothills belt (fig. 88.1). GEOLOGY The Rampart Range is an anticlinal horst of part of the Precambrian Pikes Peak batholith. The range is bounded by the Rampart Range fault on the east and the Devils Head fault zone on the west (Boos and Boos, 1957, p. 2657-2661). The Manitou Park graben is bounded by the Devils Head fault zone on the east and the Ute Pass fault on the west. The southern part of the graben contains outcropping Cambrian to Pennsylvanian sedimentary beds, which dip westward from the Rampart Range and are terminated at the Ute Pass fault. Some of the Pennsylvanian section, of which the valley floor of the Monitou Park graben is composed, has been removed by erosion. An anticline, consequent upon the Rampart Range, is indicated by projecting the attitude of the beds southward around the end of the range.1 A syncline in the graben is suggested by westward-dipping beds on the east side of the graben and, in places, by near-vertical beds on the west side. The eastern foothills belt of the Front Range of Colorado is composed of a thick section of sedimentary rocks that dip steeply to the east, away from the Rampart Range. The foothills belt is bounded on the west by the Rampart Range fault. In parts of the belt the beds are overturned, which indicates that the fault is a high-angle reverse fault. The fault has cut out most of the sedimentary rocks older than Tertiary in that part of the belt covered by this survey. THICKNESS AND DENSITY OF UNITS IN THE STRATIGRAPHIC SEQUENCE The Pikes Peak Granite, of Precambrian age, is homogeneous in mineral content and density. In the Manitou Park graben, Cambrian to Mississip-pian formations, which include the Sawatch Quartzite, Ute Pass(?) Dolomite (Peerless Formation of current usage), Manitou Limestone, Williams Canyon Limestone, and Madison (?) Limestone have a total thickness of approximately 240 feet.2 The thickness of two sections of the Pennsylvanian and Permian Fountain Formation, including the Glen Eyrie Shale Member, averages approximately 3,900 feet. One section was measured at Perry Park by Aslani3 and the other was measured in the valley of Fountain Creek, near Colorado Springs (McLaughlin, 1947, 1944, 1948). The stratigraphic section between the Fountain Formation and the Arapahoe Formation of Late Cretaceous 1 Fowler, W. A., 1952, Geology of the Manitou Park area, Douglas and Teller Counties, Colorado : Colorado Univ. unpub. Ph. D. dissert., p. 24. » Op. cit., p. 6-19. 3 Aslani, Morad-Malek, 1950, The geology of southern Perry Park, Douglas County, Colorado: Colorado School of Mines unpub. thesis, no. 686, p. 30-32. C110 ART. 88 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C110-C113. 1963.MILLIGALS MILLER cm R. 69 W. R. 68 W. Fountain Formation (Pennsylvanian and Permian) Cambrian through Mississippian rocks D V________ Fault Dashed where approximately located. U, upthrown side; D, downthrovm side Strike and dip of beds X Strike and dip of overturned beds X Strike of vertical beds — 205 —________ Gravity contour Dashed where inferred. Contour interval 5 or 1 milligal Observed Bouguer anomaly X Computed Bouguer anomaly 2.39 Density, in grams per cubic centimeter Pikes Peak Granite (Precam brian) -.DENVER I °Coloradol Springs COLORADO I Area of report Figure 88.1.—Geologic and gravity map of part of the Rampart Range area (geology adapted from Boos and Boos, 1957, figs. 9 and 10). Sections show a two-dimensional analysis across Manitou Park and the eastern foothills belt. MILLIGALSC112 GEOPHYSICS age consists of shale and sandstone. At Perry Park the section is about 7,300 feet thick.4 The Arapahoe, Denver, and Dawson Formations range in age from Late Cretaceous through Paleocene. The formations are principally poorly cemented sandstone and conglomerate and have a total thickness of about 1,400 feet east of the Rampart Range. This thickness was estimated from a section measured by Reichert (1956, pi. 2, p. 108-111) in the Castle Rock quadrangle, Colorado. The total thickness of the stratigraphic sequence east of the Rampart Range averages 12,800 feet. This thickness is presumed to be accurate within about 10 percent. An average density of 2.58 g per cm3 for the Pre-cambrian granite of the Rampart Range was determined from six fresh samples. An average density of 2.59 g per cm3 for the Cambrian to Mississippian sedimentary rocks was determined from eight samples picked at random from the section at Manitou Park. The density for the Pennsylvanian to Tertiary rocks is considered to be the same as the average density for shale and sandstone. The bulk density of shale is taken to be 2.31 g per cm3 and that of sandstone 2.28 g per cm3 (Birch and others, 1942, p. 22,23). GRAVITY DATA A total of 130 gravity stations were established, and the data from 71 stations were used in this article. The survey was made with a Worden gravimeter that has a scale constant of approximately 0.227 milligals per scale division. The survey was tied to base stations that are part of a network established in 1961 by D. J. Stuart, of the U.S. Geological Survey. Complete Bouguer anomalies were computed with elevation corrections of 0.062 mgal per ft and with terrain corrections that were computed or estimated through zone L of Hammer’s tables (Hammer, 1939). These corrections were made using a density of 2.50 g per cm3. Latitude corrections were based on the international gravity formula of 1930. Elevations of the gravity stations were taken from bench marks and photogrammetric points shown on topographic maps published by the U.S. Geological Survey. The station elevations are thought to have an average error within ± 5 feet of the true value, which corresponds to a gravity error of less than ± 0.3 milligals. The latter value may be considered to be the largest error in the gravity computations. INTERPRETATION The two-dimensional graticule interpretations (fig. 88.1, A-A', B-B') were aided by independent control. In the Manitou Park graben, control was furnished by the known angle of dip of the beds on the east side of the graben and by the limited range of density of the Fountain Formation. The interpretation across the foothills belt was controlled by the known average thickness of the rocks (12,800 feet) and by theoretical limits of dip imposed by the observed gravity curve and density contrast. Because the average density of the Cambrian to Mississippian formations is the same as that of the underlying Precambrian granite there is no gravimetric distinction between the two. Therefore, gravity computations were made to the base of the Fountain Formation. A density contrast of 0.3 g per cm3 between the Fountain Formation and Mississippian to Cambrian formations was used in the interpretation of the geology in the graben. In the foothills belt, density is assumed to increase, as a result of compaction of the sedimentary rocks, at a rate of 0.025 g per cm3 per 1,000 feet of depth. Values for density contrast were chosen at 0.30, 0.20, and 0.05 g per cm3, corresponding to three intervals of depth of about 4,000 feet each. The intervals extend from the ground surface down to the base of the Fountain Formation. Because the Pikes Peak batholith is extensive and is constant in density and mineral content it was assumed that the regional gradient is flat over a relatively small distance. Therefore, the regional gradient along cross section B-B' is taken as a straight line of zero slope and of the same magnitude as the greatest gravity value along the cross section. This value is quite consistent with gravity values of other stations on the long axis of the Rampart Range. The regional gradient along cross section A-A' is taken as a straight line connecting the two greatest gravity values along the cross section. The maximum computed depth to the base of the Fountain Formation in the Manitou Park graben is about2,000 feet (fig. 88.1, A-A’). 4 Op. clt, p. 41-70.MILLER C113 CONCLUSIONS The sedimentary rocks in the Manitou Park graben form a syncline, the west limb of which is truncated by the Ute Pass fault. Overturned beds, which might indicate reverse faulting, are not predominant in the exposed parts of the foothills belt. However, the computed dip of the Rampart Range high-angle reverse fault, as defined by the observed gravity curve is limited to a reversed dip of 50° to 75° (fig. 88.1, B-B'). The magnitude of the computed anomalies in the undulating basement rocks (fig. 88.1, B-B') approaches the maximum error in the gravity computations and, therefore, the validity of these anomalies is questionable. REFERENCES Boos, C. M., and Boos, M. F., 1957, Tectonics of eastern flank and foothills of Front Range, Colorado: Am. Assoc. Petroleum Geologists Bull., v. 41, no. 12, p. 2603-2676. Birch, A. F., Schairer, J. F., and Spicer, H. C., eds., 1942, Handbook of physical constants: Geol. Soc. America Spec. Paper 36, 325 p. Hammer, Sigmund, 1939, Terrain corrections for gravimeter stations: Geophysics, v. 4, no. 3, p. 184-194. McLaughlin, K. P., 1947, Pennsylvanian stratigraphy of Colorado Springs quadrangle, Colorado: Am. Assoc. Petroleum Geologists Bull., v. 31, no. 11 p. 1936-1981. Reichert, S. O., 1956, Post-Laramie stratigraphic correlations in the Denver basin, Colorado: Geol. Soc. America Bull., v. 67, pi. 2, p. 108-111.Article 89 GRAVITY SURVEY OF THE ISLAND OF HAWAII By W. T. KINOSHITA,1 H. L. KRIVOY,2 D. R. MABEY,1 and R. R. MacDONALD,2 1 Menlo Park, Calif.,2 Hawaiian Volcano Observatory, Hawaii Abstract.—Large gravity highs are centered over 4 of the 5 major volcanoes on the island of Hawaii. A lower amplitude gravity high occurs near the fifth volcano, Hualalai. These highs are attributed to intrusive rocks and denser parts of flows that are near the surface beneath the volcanoes and are convergent at depth. Local gravity surveys in the vicinity of Kilauea volcano and its two major associated rift zones (Krivoy and Eaton, 1961) have been supplemented by a regional survey of the island of Hawaii (fig. 89.1). Measurements along the main roads and trails provided generally good coverage at elevations below 6,000 feet, but at higher elevations large areas are inaccessible and the gravity map is necessarily generalized. An average density of 2.3 g per cc was used for all above-sea-level material in the gravity reductions and terrain corrections. The reductions were made to sea level and the terrain corrections were made for above-sea-level terrain to a distance of about 100 miles from the station. Terrain corrections were made at about half the stations and interpolated at the remainder because the topography is such that a near-linear relation exists between station altitude and amount of correction in many parts of the island. A few corrections for submarine topography were made, and it was found that their effect on a map with a 10-mgal contour interval was negligible. Because the total relief on Hawaii is more than 13,000 feet, the selection of a density to compute the Bouguer corrections determines to a large extent the character of the Bouguer anomaly map. The density of 63 dry samples, collected by R. R. Doell from the denser parts of numerous flows for remanent-magnetization studies, ranged from 1.8 g per cc to 3.0 g per cc but averaged 2.3 g per cc. Although the probable average density of the intrusive and nonvesicular flow rocks as shown by a few measurements is about 2.8 g per cc, these rocks constitute only a small part of the exposed rock on the island. Woollard (1951) found by a gravity-profile method that a density of 2.3 g per cc was the most applicable for his gravity study on Oahu. The Bouguer gravity-anomaly map (fig. 89.1) shows pronounced gravity highs over Kohala Mountain, Mauna Kea, Mauna Loa, and Kilauea, 4 of the 5 major volcanoes which make up the island of Hawaii. Hualalai, the fifth volcano, lies at the north end of an elongate high of much lower amplitude than those over the other four. Other features of the gravity map include an east-trending gravity nose on the northeast slope of Mauna Kea, southwest-trending and east-trending gravity noses from the summit of Kilauea, a south-trending gravity nose from the summit of Mauna Loa, and low-gravity fields at Hilo and at the northwestern part of the island between Hualalai and Kohala Mountain. The general correlation between the gravity highs and the topographic highs suggests that the density assumed in making the Bouguer corrections is too low, and that some discussion of these reductions is needed. There is little question that 2.3 g per cc is a reasonable approximation of the dry density of the exposed flows, and there is no reason to expect significant compaction of these flows between the surface and sea level. When this density is used in making the Bouguer correction, the anomalies obtained reflect the presence of masses with densities different from this value. In making the Bouguer reductions, no effort has been made to remove the part of the anomaly caused by density contrasts above sea level. An examination of the gravity map reveals several areas where substantial gravity differences do not correlate with surface topography. The most striking example is the +251-mgal value on the 8,000-foot summit of Hualalai, which is about the same as the gravity values at sea level along the southeast and east shores of the island. C114 ART. 89 IN U.s. GEOL. SURVEY PROF. PAPER 475-C, PAGES C114-C116. 1963.KINOSHITA, KRIVOY, MABEY, AND MacDONALD Cl 15 156°_____________________________________________________________________________________________________________155° Figure 89.1.—Bouguer gravity-anomaly map of the island of Hawaii.C116 GEOPHYSICS The four gravity highs over the individual volcanoes are part of a more extensive high that covers the central part of the island, so that although the total gravity relief is about 120 mgals, the gravity decreases only about 50 mgals between volcanoes. In this respect the island of Hawaii is quite different from Oahu, where Woollard (1951) found maximums of about 110 mgals over the volcanoes and no connecting high between volcanoes. Horizontal distances between the adjacent volcanoes on both islands are about the same. Quantitative analyses of the gravity highs are incomplete, but some qualitative conclusions on the cause of the anomalies can be reached. The amplitude of the highs and steepness of the gradients indicate that the highs are probably produced by mass anomalies above the sea floor and maybe even above sea level. The central high on Hawaii indicates the presence of relatively dense rocks at depth between the 4 volcanoes and suggests the presence of one or more of the following rock configurations: (1) a continuous intrusive mass, extending from Kohala Mountain to Mauna Loa and Kilauea, with cupolas beneath the 4 volcanoes having gravity maximums, (2) flow rocks between volcanoes, with a bulk density approaching that of intrusive rocks, or (3) interfingering of intrusive rocks with flows from the 4 volcanic centers, the whole mass of which has a bulk density approaching that of intrusive rocks. East-west asymmetry of the gravity anomalies requires the denser rocks to have a greater lateral extent to the east than to the west in all three possibilities. The third possibility is favored by the writers because there is no evidence of interconnection between volcanoes above the sea floor. Furthermore, although some massive parts of flows have measured densities of 3.0 g per cc, there is no surface evidence that such flows accumulate only in certain areas. The absence of a pronounced closed high over Huala-lai is surprising because Hualalai is the third highest volcano on the island, and it is similar in most respects to the other volcanoes on Hawaii. Hualalai’s summit lies at the northern end of an elongate gravity high that reaches a maximum value at least 8 miles to the south. This high has very small closure and its maximum value is about 85 mgals less than the maximums over Mauna Kea or Mauna Loa, but the local relief of about 55 mgals is about the same as the relief between the other volcanoes on the island. An interesting possibility is that Hualalai lies on the north rift zone of an older volcano buried by Mauna Loa lavas. The elongate, rather than bull’s-eye, gravity-contour pattern could be produced by diametrically opposed north- and south-trending rift zones of the buried older volcano. The gravity effect of the southern rift zone could be masked by its proximity to the larger Mauna Loa anomaly. A recorded offshore eruption near the south end of this high in 1877 (Dana, 1891) occurred at about the same time as a Mauna Loa summit eruption, and possibly the two were related. Relatively narrow, low-amplitude gravity noses occur over or near the more conspicuous rift zones of Mauna Loa and Kilauea. The east-trending high on the northeast slope of Mauna Kea lies between a rift zone pointed out by Stearns and MacDonald (1946) and an east-trending submarine ridge off the northeast coast of the island. These highs are probably produced by the greater abundance of intrusive rocks in the rift zones, although it is possible that ponded flows in parts of the rifts are responsible for at least a part of the gravity maximums. The gravity lows at Hilo and at the western part of the island in an area bounded by Kohala Mountain, Mauna Kea, and Hualalai are over areas probably covered with thick accumulations of low-density flow rocks. REFERENCES Dana, J. D., 1891, Characteristics of volcanoes: New York, Dodd, Mead and Co., 399 p. [1890]. Krivoy, H. L., and Eaton, J. P., 1961, Preliminary gravity survey of Kilauea Volcano, Hawaii: Art. 360 in U.S. Geol. Survey Prof. Paper 424-D, p. D205-D208. Steams, H. T., and MacDonald, G. A., 1946, Geology and ground-water resources of the island of Hawaii: Hawaii Div. Hydrography Bull. 9, p. 168. Woollard, G. P., 1951, A gravity reconnaissance of the island of Oahu: Am. Geophys. Union Trans., v. 32, no. 3, p. 358-368.Article 90 EVALUATION OF MAGNETIC ANOMALIES BY ELECTROMAGNETIC MEASUREMENTS By F. C. FRISCHKNECHT and E. B. EKREN, Denver, Colo. Abstract.—Good agreement was obtained between field data taken over magnetic nonconductive iron-formation in the Cu-yuna Range, Minn., and laboratory data using a scale model made of powdered magnetite. Field data are presented also for a more complex situation in the Gogebic Range, Wis., in which the iron-formation is highly conductive electrically, as well as magnetic. Electromagnetic measurements were made over parts of unoxidized iron-formation in the Cuyuna Range, Minn., and the Gogebic Range, Wis., to determine the practicability of estimating magnetite content by electromagnetic techniques. Parts of the iron-formation in both areas are highly magnetic and contain magnetite disseminated in a quartz- or chert-rich matrix. The iron-formation in Wisconsin overlies quartzite, siltstone, and argillite and underlies slate that commonly contains conductive graphitic layers. Both electromagnetic and magnetic methods respond to the magnetic susceptibility of a geologic body, and the relation between magnetic susceptibility and magnetite content is well known (Werner, 1945). The dominant cause of many anomalies obtained by magnetic methods is remanent magnetization, which is not predictably related to magnetite content. Because the electromagnetic methods are unaffected by remanent magnetization, they offer, in principle at least, a better means of estimating magnetite content than do the magnetic methods. The chief problem in using electromagnetic methods to evaluate high-amplitude magnetic anomalies is that deposits of the Lake Superior type containing a high percentage of magnetite often are good electrical conductors, and, at the frequencies ordinarily used, electromagnetic response is due to the combined effects of susceptibility and conductivity. Ward (1959, 1961) has outlined a theory for determining the conductivity, susceptibility, size, and depth of a buried sphere and suggests the extension of the technique to bodies of other shape. Ward shows that it is possible to determine the permeability contrast between the sphere and the surrounding rock by using the ratio of the response extrapolated to zero frequency to the response extrapolated to infinite frequency, without determining any of the other parameters. The permeability contrast may also be obtained readily from the zero frequency response alone, by comparison with model data. As an example of the latter approach (fig. 90.1), a turam model profile was compared with a turam field profile and with a vertical-component magnetometer profile taken over the thin-bedded facies of the main iron-formation of the North Cuyuna Range, Minn. (Schmidt, 1959). In the turam method (Frischknecht and Ekren, 1961), the complex ratio of the voltages induced in two receiving loops is measured along traverses normal to a long wire carrying an alternating current. The profiles in figure 90.2 are the normalized vertical component of the alternating magnetic field calculated from the measured ratios. A 200 X 200-foot induction loop was used as the source, rather than a grounded wire, to avoid the effects of galvanic currents. The turam anomaly is typical of that obtained over a magnetic nonconducting body (Tornquist, 1958). Other data obtained by the turam and slingram methods substantiate the indication that conductivity effects were negligible. The model was made of powdered magnetite having a susceptibility of about 0.08 cgs units. The model and loops were scaled so that the model represents a prism having dimensions of about 70X300X750 feet buried at a depth of 100 feet. There is good agreement between the model profile and the field profile; therefore, the susceptibility, width, depth of burial, and dip of the scale model are probably a fairly accurate replica of the actual magnetic bed involved. The depth and the extent along strike of the iron-formation are no AKT. 90 I\ U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C117-C120. 1963. C117Cl 18 GEOPHYSICS doubt greater than in the model, but due to the relatively short distance between the transmitting loop and the iron-formation, only the nearest part of both the iron-formation and the model contributed significantly to the electromagnetic response. Figure 90.1.—Comparison of turam model profile with a turam field profile and a magnetometer profile taken over the Cuyuna Range, Minn. In contrast with the amplitude of the electromagnetic field profile, which is largely the result of magnetic susceptibility, the amplitude of the magnetometer profile is greatly affected by remanent magnetization. From other studies, Gordon Bath, of the U.S. Geological Survey (oral communication, 1958), has deduced that the contribution of remanent magnetization to the anomaly is 3 to 5 times the contribution of induced magnetization. Because the anomaly represents about 25 percent .of the earth’s normal vertical field, the contribution of induced magnetization is about 6 percent of the earth’s normal field. This agrees well with the magnitude of the electromagnetic anomaly, even though the inducing electromagnetic field did not have exactly the same direction as the earth’s magnetic field. In tracing the Iron wood Iron-Formation in the Gogebic Range, Wis., by means of slingram traverses, it was found by Frischknecht and Ekren (1961) that the response of the iron-formation depended partly on susceptibility, although conductive effects were usually predominant. Variable frequency measurements were made at a locality where the contact between the Iron-wood Iron-Formation and the Palms Quartzite is exposed. A 50 X 50-foot horizontal transmitting loop was used. A small receiving coil, a fixed reference coil, and a ratiometer were used to measure the field at various frequencies and distances from the transmitting coil. Conventional turam equipment was used to verify the results obtained with the experimental apparatus operating at a single frequency of 500 cycles per second. Figure 90.2 shows profiles obtained at 3 different frequencies: 40, 200, and 1,000 cps. Three anomalies, labeled A, B, and C, are recognized; only part of anomaly C was measured. There is, no doubt, some interdependence among these anomalies because of their proximity to each other. Anomaly A is characterized by in-phase values greater than 100 percent and out-of-phase values that are essentially 0. These results are indicative of a member of the iron-formation having fairly high magnetic susceptibility and low electrical conductivity. The anomaly increases slightly at the higher frequencies, showing a small conductivity effect, but at 40 cps the anomaly is almost entirely caused by the high susceptibility of the formation. It is probably caused by iron-formation having considerable magnetite in the form of discontinuous grains which do not increase significantly the overall conductivity of the rock. The susceptibility and dimensions of the unit could be readily estimated by comparison with model experiments as in the preceding example. Anomaly B is practically nonexistent at 40 cps. The out-of-phase component is positive at 200 cps and negative at 1,000, whereas the in-phase component is increasingly greater at 200 and 1,000. Anomaly B must therefore be caused by a part of the iron-formation having low magnetic susceptibility and intermediate electrical conductivity. A graphitic bed containing little magnetite probably is the cause. Anomaly C is more complex than A or B. To facilitate the study of anomaly C, figure 90.3 shows the maximum in-phase and out-of-phase values of anomaly C and A plotted against frequency. At frequencies below about 400 cps the in-phase component is greater than 100 percent, becoming larger as the frequency is decreased. Except for the inflection in the curve of the in-phase component at about 50 cps, both the in-phase and the out-of-phase curves resemble theoretical curves for a conductive permeable sphere (Ward, 1959). The disturbing body certainly is not spherical, but data are not available for other shapes.IN-PHASE COMPONENTS, OUT-OF-PHASE COMPONENTS, IN PERCENT IN PERCENT FRISCHKNECHT AND EKREN C119 Figure 90.2.—Electromagnetic profiles at 3 frequencies across iron-formation on the Gogebic Rnnge, Wis.IN-PHASE COMPONENTS, IN PERCENT C120 GEOPHYSICS FREQUENCY, IN CYCLES PER SECOND Figure 90.3.—Variation of electromagnetic anomalies with frequency on the Gogebic Range, Wis. The body probably has a relatively high magnetic susceptibility and electrical conductivity. More detailed field work and model work are needed to explain the cause of the inflection in the in-phase curve and to inter- pret the anomaly further. In particular the measurements need to be extended to lower and higher frequen-z cies. There is little difficulty in making measurements k at higher frequencies than those used in these studies, LU but conventional electromagnetic equipment for use at lower frequencies is very bulky. One other possible z approach to identify the effect of magnetic susceptibil- UJ . . # t | lty is to use a very sensitive magnetometer and utilize 1 the natural time-variant magnetic field (War d and Rud- 8 dock, 1962). S> REFERENCES < - Vertical or inclined shaft Adit Open pit Figure 100.1.—Map showing location of Doris No. 1 mine within the belt (shaded) of primary deposits in the Poison Canyon Sandstone of Zitting and others (1957).C158 ECONOMIC GEOLOGY Mudstone^ :onglomerate in back MAP SECTION EXPLANATION Contorted sandstone Showing direction of dip Inclined workings Chevrons point downslope 0 Raise Sandstone mmumxnr o o Mudstone or conglomerate layer Carbonaceous ore Fault, showing dip; dashed where inferred Todorokite in fractures ore in fault zone Oranopilite and >peite-like mineral Sample suite 30G62 Native gray, selenium Incline cuts mudstone layer 40 FEET A' Figure 100.2.—Map and section of the Doris No. 1 mine.GRANGER AND SANTOS C159 of a black ore layer. The uranopilite is bright yellow and has a moderately strong greenish-yellow fluorescence. It forms small aggregates with a finely felted texture made up of short disoriented bladed or needlelike crystals. The zippeite-like mineral forms minute rounded aggregates of powdery yellow crystals. The same mineral has been noted in several mines throughout the district but it generally forms minute “pincushions” of radiating needlelike crystals. Sparse coatings of the same or, perhaps, other yellow efflorescent uranium minerals also occur elsewhere on the mine walls. Minute scattered flakes of greenish-yellow to green nonfluorescent uranium (?) minerals occur in the partly oxidized black ore layers, but no attempt has been made to identify them. Limonitic iron oxides and partly oxidized pyrite are disseminated throughout both the mineralized and barren sandstone. Where it has been protected from oxidation by the abundant organic gangue in the ore, pyrite occurs as minute cubes and irregularly shaped masses less than 1 mm across. Calcite is scarce and may have been largely removed by acid solutions created by the oxidizing pyrite. Elsewhere in the district, native gray selenium is common at the interfaces between oxidized and unoxidized rock. In the Doris No. 1 mine, however, the rocks are pervasively oxidized, and the only selenium noted was in fractures in mudstone above the ore and in thin gray streaks less than 2 inches long in partly oxidized sandstone adjacent to black ore layers. Tiny selenium needles are disseminated in the interstices throughout the gray streaks. Kaolinite “nests” (Granger, 1962) are distributed throughout the host rock, forming aggregates as much as several millimeters across. They are most abundant in the coarser grained barren sandstones and are either sparse or extremely small in the finer grained sandstones and ore layers. Walls of open fractures in the cylindrical fault zone north of the foot of the shaft (fig. 100.2) are locally coated with dull velvety-black todorokite, 2(Mn,Ca) 0.5MnO2'IH2O. It forms tiny radiating crystalline pisolitic aggregates less than 0.2 mm in diameter which have grown, one upon the other, to form microbotry-oidal textures and minute knobby columns. The individual crystals are fibrous platelets which are dark brown in transmitted light under a microscope. This is the only occurrence of todorokite yet known in the Ambrosia Lake area and is one of few localities known in the United States. The results of analyses of a suite of samples (No. 30G62) that were taken across an ore layer in the central part of the mine are shown graphically in figure 100.3. The content of organic carbon, lead, molybdenum, vanadium, and selenium correlate, to some extent, with the uranium concentration. Pervasive weak oxidation of the host rock may have resulted in some redistribution of these elements, but it is believed that the effect has been minor. A sample of ore from the cylindrical fault north of the shaft has a composition similar to the samples within the ore layer described. The character and thickness of the rock sampled in suite 30G62 are given below. Description of sample in suit 30026 Thickness [Letters refer to zones shown In figure 100.3] (inches) a. Medium-grained barren sandstone 20 to 16 inches above ore layer; yellowish gray with scattered kaolinite nests 1.0 to 1.5 mm across. Limonite occurs as disseminated specks, clay stain, and thin films on quartz grains. Quartz overgrowths are sparsely present on sand grains and form minute triangular crystal faces. Calcite is extremely sparse___________ 4 b. Fine-grained moderately well sorted barren sandstone 16 to 4 inches above ore layer; ranges from very pale orange to grayish orange with small white kaolinite nests. Pyrite is mostly oxidized, leaving specks and films of limonite. Sand grains commonly coated with a film of white clay. A few thin lenticular streaks containing needles of gray selenium were noted------------------------------------------------ 12 c. Fine-grained barren sandstone 4 to 0 inches above ore layer ; yellowish gray with local pale reddish-brown hematitic specks and stain. Kaolinite may be present but doesn’t form nests. A loose claylike material fills many intertices. Alinute disseminated flakes of a greenish-yellow uranium mineral are present_____ 4 d. Fine-grained to very fine grained dark-gray ore in upper 2 inches of ore layer; well cemented by carbonaceous material. Pyrite is sparse and largely oxidized to limonitic stain. Kaolinite aggregates are present but very small__________________________ 2 e. Middle 2 inches of ore layer identical with 30G62d_____ 2 f. Very fine grained grayish-black ore in lower 2 inches of ore layer. Carbonaceous material fills nearly all interstices. All interstices not completely filled by carbonaceous material are filled with kaolinite. A few scattered flakes of an unidentified greenish nonfluorescent uranium( ?) mineral are present___ 2 g. Fine-grained barren sandstone 0 to 4 inches below the ore layer; yellowish gray with local moderate yellowish-brown limonite specks and stain. Sand grains all coated with a thin film of white clay_ 4 h. Fine-grained barren sandstone 4 to 16 inches below the ore layer; yellowish gray with local grayish-orange limonite stain. Some clay disseminated throughout the rock but doesn’t form either nests or coatings on the sand grains___________________ 12 i. Fine- to medium-grained barren sandstone 16 to 20 inches below the ore layer; yellowish gray with local grayish-orange limonite stain and specks that resemble corroded pseudomorphs. A few small scattered kaolinite nests are present, and each sand grain is coated with a film of white clay________ 4C160 ECONOMIC GEOLOGY CLASSIFICATION OF THE DEPOSIT Unoxidized uranium ore of two types is recognized in the Ambrosia Lake district: a prefault type, considered to be primary, and a postfault type, believed to be redistributed (Granger and others, 1961). The prefault ore is stratigraphically controlled and invariably associated with an abundant gangue of authigenic carbonaceous material; it generally contains concentrations of lead, molybdenum, and vanadium. In contrast, postfault ore is partly structurally controlled and ordinarily contains much less, if any, carbonaceous material, lead, and molybdenum but more vanadium than the prefault ore. The faults that separate these deposits in time formed long after the Morrison was deposited and displace the primary ore, the Dakota Sandstone, and at least the lower part of the overlying Mancos Shale. Chemically, the ore in the Doris No. 1 deposit (fig. 100.3) is similar to the ores in prefault deposits elsewhere in the district and contrasts with the postfault ores. The organic carbon, lead, and molybdenum content is typical of prefault ore and much higher than in most postfault ore. Although the vanadium content is greater than in most prefault ore, it is not abnormally high for prefault ores in the Poison Canyon Sandstone, and the uranium-vanadium ratio in the ore layer is higher than in much of the postfault ore in the district. L£> O O D ) 0 Cu 4 5.3 23 30.7 52 69.3 61 81.3 As 4 5.3 11 14.7 w 0 0) 0 Ag 2 2.7 33 44.0 55 73.3 62 82.7 Ca 2 2.7 22 29.3 55 73.3 56 74.7 Ba 1 1.3 6 8.0 10 13.3 41 54.7 Mg 0 0 15 20.0 36 47.8 60 80.0 A1 0 0 14 18.7 45 60.0 70 93.3 Cd 0 0 3 4.0 23 30.7 0) 0 Te 0 0 2 2.7 w 0 (>) 0 Ni 0 0 2 2.7 7 9.3 (>) 0 U» 0 0 2 2.7 3 4.0 5 6.7 Na 0 0 2 2.7 3 4.0 to 0 Sn 0 0 1 1.3 8 10.7 24 32.0 V 0 0 1 1.3 5 6.7 p) 0 Mn 0 0 0 0 22 29.3 45 60.0 Co 0 0 0 0 8 10.7 (0 0 Sr 0 0 0 0 7 9.3 25 33.3 Mo 0 0 0 0 5 6.7 12 16.0 T1 0 0 0 0 4 5.3 20 26.7 B 0 0 0 0 3 4.0 4 5.3 Cr 0 0 0 0 2 2.7 6 8.0 La 0 0 0 0 2 2.7 0) 0 T1 0 0 0 0 1 1.3 0) 0 Zr 0 0 0 0 1 1.3 3 4.0 Y 0 0 0 0 0 0 9 12.0 Sc 0 0 0 0 0 0 7 9.3 Au 0 0 0 0 0 0 4 5.3 In 0 0 0 0 0 0 2 2.7 Nb 0 0 0 0 0 0 1 1.3 K 0 0 (') 0 (•) 0 w 0 i Below detection limit by method used. * Chemical determination. A method for preparing high-purity ore-lead samples for isotopic analysis was desired that would be simpler and shorter than the sulfate separation method (Scott, 1939) used by the U.S. Geological Survey for several years. The sulfate method is generally satisfactory when 10 milligrams or more of Pb is available, but it is time consuming and requires the use of a number of ART. 102 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C166-C170. 1963. C166ANTWEILER C167 lead-free reagents. Most of the other lead-separation methods are useful in certain circumstances, but they have one or more of the following disadvantages: they are lengthy, require the use of a number of lead-free reagents, fail to achieve complete separation from other elements, are limited by interferences, or result in a compound not amenable to isotopic analysis. The method described below overcomes most of these disadvantages. It is simple and comparatively fast, requires only two lead-free reagents (water and nitric acid), separates lead from most other elements (barium and strontium excepted), is not limited by the elements commonly found in galena, and results in a compound (the nitrate) that is readily converted to the sulfide, the iodide, or to other compounds for mass analysis. The procedure is based on the fact that lead is selectively precipitated with strong HN03. This selectivity was used by Baxter and Grover (1915) to purify lead for atomic-weight determination, by Schrenck and Delano (1931) to purify lead for electrodeposition studies, by Neumann and Perlman (1950) to separate radioactive bismuth from lead, and by the Oak Ridge National Laboratory (W. C. Davis, written communication, 1961) to remove all spectrographically detectable impurities from separated lead isotopes. EXPERIMENTAL WORK The first experiments showed that all spectrographically detectable impurities were removed by repeated precipitations of lead nitrate with concentrated (sp gr 1.42) HN03. The procedure was satisfactory except that lead loss was significant. Although quantitative recovery is not essential for samples prepared for isotopic analysis, small samples could be lost before being purified to the desired degree. A study was made, therefore, of the solubility of lead nitrate in various concentrations of nitric acid at room temperature and at 0°C. This was done as follows: a saturated solution of lead nitrate in nitric acid was prepared by adding to 100 ml of HNOs an excess of finely ground Pb(N03)2; the mixture was warmed and stirred, allowed to reach the desired temperature, and filtered. Lead in the filtrate was determined colorimetrically by the dithizone method, or gravimetrically as the sulfate or chromate. Figure 102.1 and table 102.2 show the rapid decrease in lead solubility as nitric acid concentration is increased and temperature is decreased; less than 1 microgram of Pb dissolves in 1 ml of 100-percent HN03 at 0°C. It will be noted that the solubility of lead nitrate in fuming HN03 is greater than in 100-percent HN03; this is attributed to the action of nitrous acid and nitrogen oxides present in fuming HN03 As one would expect, several nitric acid precipita- LEAD DISSOLVED, IN MICROGRAMS PER MILLILITER OF ACID Figure 102.1.—Solubility of lead as a function of nitric acid concentration at 24°0. and 0°C. Table 102.2.—Solubility of lead in nitric acid of various concentrations at 24°C and 0°C Approximate HNO3 (weight percent) Specific gravity of HNO3 * 20° _ at^C Solubility of Pb (micrograms per milliliter) 24° C 0°C 67 1.40- 230 85 72 1.42 50 17 8 2. 9 1. 4 1. 0 . 9 . 9 10 81 1.454 84 1.465--. 88 1.478. - . 10 93 1.489 95 1.494 4 98 1.503 100 1.510 Fuming_ _ 1.59 18 tions are required to remove all impurities from lead nitrate, because any precipitate may occlude small amounts of other elements in the solution. To determine the number of precipitations required, the National Bureau of Standards lead isotope reference sample (NBS-200, galena from Ivigtut, Greenland) was decomposed with nitric acid. The lead nitrate obtained was precipitated with 100-percent HN03, and the precipitate was collected on a filter disk, sampled, and redissolved. This procedure was repeated until a total of five precipitations had been made. Samples taken after the first, fourth, and fifth precipitations were analyzed spectrographically. The analyses, given in table 102.3, show considerable purification after just one nitric acid precipitation. Only calcium was detectable after the fourth precipitation, and it was removed by one more precipitation. The purity obtained by 4 precipitations is thought to be adequate for isotopic analysis, and therefore the recommended procedure for most ore-lead samples consists of sample decomposition with dilute HN03 followed by 4 consecutive precipitations with 100-percent HN03. ACID, IN WEIGHT PERCENTC168 ANALYTICAL TECHNIQUES Table 102.3.—Semiquantitative spectrographic analyses, in percent, of the National Bureau of Standards lead isotope reference sample (NBS-200) and of lead compounds prepared from it [Only elements found are shown. The “less than” «) values were generally reported as “looked for but not found,” but estimated visual detection limits have been substituted to show amounts that mieht be present but escape detection. Silicon was not specifically looked for. Analyses in Pblj column by Nancy Conklin; all others by R. O. Havens] Element Raw galena PbL from 1 HNOs precipitation Pb(N03)j from 4 HNO3 precipitations Pb(N03)s from 5 HN03 precipitations A1 0. 015 <0. 001 0. 003 <0.001 Fe . 15 . 0007 <. 0007 <. 0007 Mg .00015 . 00015 <.00015 <. 00015 Ca . 0007 . 0005 . 0007 <. 0007 Mn . 0015 <. 0002 <. 0002 <. 0002 Ag . 7 . 015 <. 0001 <. 0001 Bi 1. 5 . 01 <. 001 <. 001 Cu . 07 <. 0001 <. 0001 <. 0001 Sb . 03 <.01 <.01 <.01 Sn _ _ . 015 <. 001 <. 001 <. 001 Reagents.—To prevent the possibility of isotopic contamination, only lead-free reagents were used. Water was prepared by double distilling demineralized distilled water in an all-quartz still. “Concentrated” HNOs (sp gr 1.42) was prepared by distilling analytical reagent-grade HN03 under reduced pressure. Dilute HN03 (1:1 v/v) was prepared by diluting a volume of concentrated HN03 (sp gr 1.42) with an equal volume of water. Very strong HN03 (100 percent, sp gr 1.50-1.51) was prepared by carefully distilling equal volumes of analytical reagent-grade HN03 and H2S04 at 0.01-mm Hg pressure in a still equipped with a Vig-reaux fractionating column. (Rapid distillation results in carryover of some sulfate ion, which would form insoluble lead sulfate.) Very strong HN03 should be handled only in a well-ventilated hood, and only with protective gloves. Procedure.—Dissolve about 0.1 g galena or other lead mineral by heating with 30 ml 1:1 (v/v) HN03 until decomposition is complete. If samples are small, decrease quantity of reagents accordingly. Filter and wash residue 5 times with hot 1:1 HN03. Collect filtrates and washings in a vycor or quartz dish and evaporate to incipient crystallization of lead nitrate. Precipitate lead nitrate by adding 15 ml of 100-percent HN03 (sp gr 1.50), warm, and stir to maximize solution of impurities. Chill to 0°C, and decant through a fine-porosity fritted disk. Wash precipitate with two 5-ml portions of cold HN03 (sp gr 1.50); discard filtrate and washings. Dissolve lead nitrate in dish and on filter with 1:1 HXO,. Evaporate solution to incipient crystallization of lead nitrate, and precipitate lead nitrate as before. Repeat filtration,solution, and precipitation at least twice more. Dissolve lead nitrate from fourth precipitation with water, and evaporate solution to dryness. Dry purified lead nitrate crystals for 2 hours at 100°C. Note: Some sulfuric acid is obtained by decomposing galena with nitric acid. When the lead nitrate solution is evaporated, some lead may be precipitated as lead sulfate; repeated precipitation and solution as described usually eliminates the sulfate interference by the end of the second or third precipitation. Results.—The procedure was used to purify a large number of galena samples from Colorado for isotopic analysis. The following five of these samples were used to evaluate the procedure: 1. Specimen Co-P-Bl-1. Galena from the Boomer mine, Lake George district, Park County, Colo. Collected by W. N. Sharp and C. C. Hawley, 1959. 2. Specimen Co-L-Hg. Galena from the Homestake mine, Lake County, Colo. Collected by Ogden Tweto, 1951. 3. Specimen Co-J-Og. Galena from the Augusta lode near Evergreen, Jefferson County, Colo. Collected by R. S. Cannon, Jr., and J. C. Antweiler, 1961. 4. Specimen Co-Ch-Tg. Galena from Turret district, Chaffee County, Colo. Given to J. C. Antweiler by George Corley, Corley Mining Co., 1961. 5. Specimen Co-Ko-Vgc. Galena from Victory mine, Kokomo district, Summit County, Colo. Collected by S. C. Creasey, 1949. Evaluation should be based upon a comparison of isotopic analyses and the chemical purity of samples prepared by this and by other methods. The isotopic analyses have not yet been made, but spectrographic analyses of the five samples are given in table 102.4. Methods of preparation used were (1) the sulfate method (Scott, 1939), (2) galena-to-lead-iodide method (Cuttita and Warr, 1960), (3) method of this article with just 1 nitric acid precipitation, and (4) method of this article wfith 4 nitric acid precipitations. All samples were converted to lead iodide for uniformity. The nitrate method is superior to the others in terms of product purity. Even one nitric acid precipitation results in a product with fewer impurities than is obtained by the lengthy sulfate procedure. Four nitric acid precipitations reduced all impurities to less than 0.1 percent (in the extreme example) ; impurities in the samples from Turret and Kokomo were reduced to a few parts per million. Comparatively little purification was obtained by use of the galena-to-lead-iodide procedure. Similarly, one would expect little purification by the use of other sample-preparation procedures, such as acid decomposition of galena followed directly by reconstitution of lead sulfide, that do not selectively isolate lead from impurities inherent in the sample. Discussion.—The alkaline earth metals are the most serious interferences in the nitrate method. Strontium and barium, like lead, are quantitatively precipitated by strong HX03, and precipitates of these metals carry calcium (Meinke, 1955). The effect of the alkaline earth metals on lead isotope analysis is no better knownANTWEILER C169 than the effect of other impurities. Although it might seem desirable to make certain that samples submitted for isotopic analysis are free of these metals, it is doubtful in most cases whether such effort is justified. Less than 10 percent of the samples summarized in table 102.1 contained 0.01 percent or more Sr, and less than 15 percent contained 0.01 percent or more Ba. Further-more, strontium or barium, even though present in the purified nitrate, would be greatly reduced by subsequent conversion to either the sulfide or iodide for isotopic analysis. Table 102.4.—Semiquantitative spectrographic analyses, in percent, of galena samples and of lead iodide prepared from them by different methods [Analyst: Nancy Conklin. The “less than” «) values were generally reported as “looked for but not found,” but estimated visual detection limits have been substituted to show amounts that might be present but escape detection) 1. Galena sample Co-P-Bl-1 Raw galena Pblj from H2SO41 Pbla from HI2 Pbla from 1 HNOa precipitation Pbla from 4 HNO3 precipitations Si 0. 03 (*) 0. 007 0. 007 <0. 0015 Fe . 03 <0.001 . 02 . 03 <. 001 Mg . 0002 <. 001 <. 00015 . 002 . 005 Ca . 0007 . 001 . 0005 . 0015 . 05 Ag— . 5 . 01 .07 . 02 . 003 Bi . 7 . 003 . 015 .007 <. 001 Cu . 3 . 0005 . 007 . 01 . 0005 Nb . 003 <. 001 <.001 <. 001 <. 001 Sb . 05 <.01 <.01 <.01 <.01 Sn . 007 <. 001 <. 001 <. 001 <. 001 Y . 002 <. 001 <. 001 <. 001 <. 001 Yb <. 0003 <. 0003 <. 0003 <. 0003 <. 0003 Zn . 03 <. 02 <. 02 <. 02 <. 02 2. Galena sample Co-L-Hg Si 0. 07 (3) 0. 015 0. 003 0. 03 Fe 2. 0. 1 . 007 . 002 <. 001 Mg . 3 . 01 . 002 . 0007 . 0007 Ca._. . 015 . 007 . 0015 <.0005 . 0007 Mn . 05 . 0015 <. 0002 <\ 0002 <. 0002 Ag . 07 . 001 . 02 . 0015 . 007 Bi . 02 <. 001 . 002 . 003 <. 001 Cu . 15 . 01 . 007 <. 0001 . 0003 Sb . 02 <.01 <.01 <.01 <.01 Zn . 07 <. 02 <. 02 <. 02 <. 02 Table 102.4.—Semiquantitative spectrographic analyses, in percent, of galena samples and of lead iodide prepared from them by different methods—Continued [Analyst: Nancy Conklin. The “less than” «) values were generally reported asS “looked for but not found,” but estimated visual detection limits have been substituted to show amounts that might be present but escape detection] 3. Galena sample Co-J-Og Raw galena Pbl2 from H2SO41 Pbla from HI2 Pbl2 from 1 HNO3 precipitation Pbl2 from 4 HNO3 precipitations Si 0. 07 0. 07 0. 007 0. 003 0. 003 Fe . 05 . 05 <. 001 . 002 . 015 Ca . 015 . 0015 <.0005 <. 0005 <. 0005 Ag . 02 . 001 . 007 . 0015 <. 0001 Bi . 02 <. 001 . 0015 . 003 <. 001 Cu . 7 . 0005 . 02 <.0001 . 0005 Mo . 003 <. 0005 . 0015 <. 0005 <. 0005 4. Galena sample Co-Ch-Tg Si 0. 03 0. 3 0. 03 <0. 001 <0. 001 A1 . 007 <. 001 . 007 <. 001 <. 001 Fe . 001 . 003 . 005 . 0007 <. 0007 Mg._ ... . 002 . 0005 . 0015 . 00015 .00015 Ca . 002 . 0007 . 0005 <. 0005 <. 0005 Mn . 01 <. 0002 . 005 <. 0002 <. 0002 Ag._ -- . 05 . 001 . 03 . 005 <.0001 Bi . 01 <. 001 . 003 <. 001 <. 001 Cd . 015 <. 005 . 01 <. 005 <. 005 Cu . 3 <. 0001 . 15 <. 0001 <. 0001 In . 002 <. 001 . 003 <. 001 <. 001 Ni . 0007 <. 0003 <. 0003 <. 0003 <. 0003 Sb . 03 <.01 <• 01 <.01 <. 01 Zn . 7 <. 02 1. <. 02 <. 02 5. Galena sample Co-Ko-Vgc Si 0. 007 0. 007 0. 07 <0. 001 <0. 001 Fe . 005 <. 001 <. 001 <. 001 <. 001 Mg . 0005 . 00015 . 00015 <. 00015 <. 00015 Ca . 00015 <. 0005 <. 0005 . 0005 . 0005 Mn . 0005 <. 0002 <. 0002 <. 0002 <.0002 Ag__ . 1 . 003 . 03 . 015 <.0001 Bi . 005 <. 001 . 001 . 0015 <. 001 Cd . 005 <• 005 <. 005 <. 005 <• 005 Cu . 0015 . 00015 . 001 . 003 <. 0001 Sb <. 07 < 01 . 03 <.01 <• 01 Sn _ _ _ . 005 <. 001 . 003 <. 001 <. 001 1 PbS04 transformed to Pbla; PbSO« prepared for gravimetric Pb determination as described by Scott (1939). 2 Method described by Cuttita and Warr (1960). 3 Not specifically looked for.C170 ANALYTICAL TECHNIQUES REFERENCES Baxter, G. P., and Grover, F. L., 1915, A revision of the atomic weight of lead. The analysis of lead bromide and chloride: Am. Chem. Soc. Jour., v. 37, p. 1027-1061. Cannon, R. S., Jr., Pierce, A. P., Antweiler, J. C., and Buck, K. L., 1961, The data of lead isotope geology related to problems of ore genesis: Econ. Geology, v. 56, p. 1-38. Cuttita, Frank, and Warr, J. J., 1960, Preparation of lead iodide for mass spectrometry: Art. 221 in U.S. Geol. Survey Prof. Paper 400-B, p. B487-B488. Ehrenberg, H. Fr., 1953, Isotopenanalysen an Blei aus Miner-alen: Zeitschr. Physik, v. 134, p. 317-333. Meinke, W. W., 1955, Nuclear chemical research and radiochemical separations, progress report 4, Nov. 1954r-Oct. 1955, Project No. 7, Contract No. AT (11-1) 70: Michigan Univ. 81 p.; Atomic Energy Comm. U-3116. Neumann, N. M., and Perlman, I.. 1950, Isotopic assignments of bismuth isotopes produced with high energy particles: Phys. Rev., v. 78, p. 192. Nier, A. O., 1938, Variations in the relative abundances of the isotopes of common lead from various sources: Am. Chem. Soc. Jour., v. 60, p. 1571-1576. Richards, J. R„ 1962, Interpretation of lead isotope abundances: Nature, v. 195, p. 590-591. Schrenck, W. T., and Delano, P. H., 1931, Electrolytic determination of lead as lead dioxide: Indus. Eng. Chemistry, Anal. Ed., v. 3, p. 27. Scott, W. W„ (N. H. Furman, ed.), 1939, Standard methods of chemical analysis, 5th ed., v. 1, The elements: New York, D. Van Nostrand Co., Inc., p. 504—505. Zykov, S. I., and Stupnikova, N. I., 1957, Izotopny analiz svintsa bez predvaritel’noy khimicheckoy podrotovki minerala r Isotopic lead analysis without preliminary chemical preparation of the mineral] : Geokhimiya, 1957, no. 5, p. 430-434. [See also Geochemistry, a translation of Geokhimiya, 1957, no. 5, p. 506-510.]Article 103 PERCENT-CONSTITUENT PRINTING ACCESSORY AND FLOW-THROUGH CELL FOR A SPECTROPHOTOMETER By LEONARD SHAPIRO and EDWARD L. CURTIS, Washington, D.C. Abstract.—An accessory has been designed which converts an ordinary spectrophotometer to one which uses the color density of a solution to automatically calculate and print on tape the concentration of the constituent in the solution. No computations are required, and the results are obtained more rapidly and are more dependable than by conventional methods. A newly designed simple flow-through cell provides greater convenience and speed of operation. An accessory has been designed to convert an ordinary spectrophotometer to one which automatically calculates and prints on tape the percent constituent of a solution. The accessory, which can be built with commercially available components, is designed for use with a Model B Beckman spectrophotometer. An additional convenience in operating the accessory-equipped spectrophotometer is a newly designed flowthrough cell for holding the solution to be tested. The cell, which is made of Lucite and glass, is relatively inexpensive and easy to make in the average machine shop. PRINTING ACCESSORY General description The sensitivity of the spectrophotometer is varied so that the reading on the spectrophotometer meter is always 100-percent transmission. This is done by rotating a logarithmically wound potentiometer automatically with a small servomotor. The angular rotation required to vary the sensitivity to maintain 100-percent transmission is directly proportional to the change in concentration from one solution to another. A gear on the shaft of the logarithmic potentiometer drives a linear potentiometer which serves as a read-out device. An external voltage is applied across the ends of the read-out potentiometer, and the voltage between one end and the movable contact is measured and printed ART. 103 IN U.S. GEOL. SURVEY PROF. 694-027 O—63--12 with a digital-voltmeter-printer. Any change in this measured voltage is directly proportional to the change in angular rotation, which in turn is directly proportional to the difference in concentration of the different solutions. The external voltage is adjusted so that the voltage being read out for a known standard solution is numerically equal to the known concentration expressed as percent constituent. All subsequent readings are then printed automatically in terms of percent constituent. Components The components for the printing accessory are shown in the semischematic diagram of figure 103.1. All components except the digital voltmeter and printer can be placed into the instrument box which serves as a support for the spectrophotometer (fig. 103.1). The on-off toggle switch below the spectrophotometer turns on the line and battery power. The automanual toggle switch is a 4-pole double-throw type which in the manual position causes the spectrophotometer to operate in the normal manner. When the switch is thrown to the auto position a logarithmically wound potentiometer is switched into the sensitivity circuit of the spectrophotometer, and the pre-existing set of resistors which normally controls the sensitivity in three steps is switched out. The same switch also turns on the servomotor, which drives the potentiometer and introduces a 3-volt battery and resistor across pins 11 and 13 at the back of the spectrophotometer. Normally, when the meter needle indicates 100-percent transmission, the voltage at pins 11 and 13 of the Model B Beckman spectrophotometer is 3 volts. This voltage varies in the same direction as the percent transmission. When the accessory is switched in, the 3-volt battery and resistor are placed across these pins, as PAPER 475-C, PAGES C171-C174. 1963. C171C172 ANALYTICAL TECHNIQUES shown in figure 103.1. A current flows through the resistor when the voltage is other than 3 volts at pins 11 and 13. The input leads of a servoamplifier are connected across the resistor, which causes the servomotor to rotate whenever a current flows through the resistor. Whenever a new solution is poured into the spectrophotometer, the servoamplifier causes the logarithmically wound potentiometer to restore 3 volts at pins 11 and 13 and to give a meter reading of 100-percent transmission. The read-out potentiometer, which is geared to the logarithmic potentiometer, provides a measurement of the rotation required to restore equilibrium from one solution to another. By manually adjusting the readout potentiometer across a 45-volt battery when a standard solution is in the spectrophotometer, the operator can make the digital-voltmeter reading numerically the same as the percent-constituent reading for the known standard. (A percent-constituent knob is provided as shown in figure 103.1 for this adjustment.) Then when other solutions are placed into the spectrophotometer the digital voltmeter will automatically read percent constituent, provided that the color system follows Beers’ law. Two other controls are shown in figure 103.1. One is a servo-sensitivity knob, which varies the servoamplifier sensitivity by changing the value of R in the 3-volt circuit to a point where the servo does not hunt excessively. The other is a zero-adjustment knob, which turns another potentiometer to balance the 3-volt battery exactly with the spectrophotometer output, so that the meter reading will be right at 100-percent transmission when it is to be automatically maintained at balance. These last two controls, once they have been adjusted, should rarely need readjustment. Servoamplifier Digital voltmeter Solution in )—Sensitivity feedback ■ Log-wound potentiometer ' Read-out potentiometer Figure 103.1.—Semischematic diagram of percent-constituent printing spectrophotometer. The log-wound potentiometer, read-out potentiometer, servomotor, and servoamplifier are contained in the instrument box under the spectrophotometer.SHAPIRO AND CURTIS C173 Circuitry Two separate circuits are used. One is wired into the spectrophotometer through the 4-pole double-throw switch (auto-manual switch, figs. 103.1 and 103.2) so that it may be introduced into or eliminated from the normal existing circuit, as shown in figure 103.2; the other is a read-out circuit (fig. 103.3) not connected electrically with the previous circuit. The toggle switches used for placing the 3-volt and 45-volt batteries into operation, and for turning on the 115-volt a-c for the whole instrument, may be placed on one 3-pole double-throw toggle switch (on-off switch, fig. 103.1). The special components purchased from commercial sources included the log-wound potentiometer, special Helipot model 7603, function y= 10299.9 percent) by passing 30 to 40 ml of a new solution through the cell. Figure 103.3.—Circuitry of read-out and print-out system.C174 ANALYTICAL TECHNIQUES Figure 103.5.—Mounting bracket for flow-through cell. PROCEDURE The following procedure is suggested for operating the spectrophotometer with the percent-constituent printing accessory and flow-through cell: Turn on the instrument and allow it to warm up for at least 30 minutes. Prepare a blank, a known standard, and a series of unknown solutions in the usual manner for spectrophotometry. Pour about 40 ml of the blank solution into the funnel. When the solution stops flowing, adjust the slit to any position so that the meter needle is automatically resetting itself. If this position is not 100 percent transmission ±1 percent, turn the zero adjustment knob to bring the transmission within this limit. Note the exact meter reading. Open the slit to a point where the meter reading goes well above 100 and then close it carefully to bring the needle just to the previously noted balance position. It is now at the zero position of the log-wound potentiometer. Pour about 40 ml of the standard solution into the funnel. When the overflow of the liquid has stopped, adjust the reading on the digital voltmeter with the percent-constituent knob to read the same value as the known standard. Press the print button to print the value of the standard. Pour each of the unknown solutions into the funnel in sequence, and each time the liquid stops flowing, press print button to print the percent constituent of the solution. DISCUSSION The use of the electronic arrangement described provides a uniform sensitivity across the full scale of the instrument because the amplifier receives the same voltage change for the same change in absorbance. A deviation of 0.001 in optical density is sufficient to initiate corrective action. Theoretically, the servoamplifier-servomotor combination can provide a more sensitive response to color change than is obtained by normal manual operation. In practice the accuracy of the system may be limited by any deviation of the log coil from its designed curve and any mechanical lag in the links between the servomotor and the two potentiometers. A special pair of gears cut to a logarithmic relation is feasible and might be an improvement over the existing log coil plus gearing. Results obtained with the printing spectrophotometer for the determination of Si02, A1203, Fe203, Ti02, MnO, and P205 in silicate rocks do not differ significantly in accuracy and precision from those obtained with a manual spectrophotometer. The benefits of this instrument include not only time saved by eliminating computations but also more dependability by eliminating human errors that arise in the computations.Article 104 DISSIPATION OF HEAT FROM A THERMALLY LOADED STREAM By HARRY MESSINGER, Washington, D.C. Abstract.—Energy-budget analysis of a heated stream failed to account for the observed rapid cooling. Discrepancies between predicted and actual downstream temperatures are attributed mainly to inadequacies in existing methods for measuring the effective solar and atmospheric irradiation of partially shaded water surfaces. Effective planning for industrial and recreational utilization of streams commonly requires foreknowledge of water-temperature variations that may be expected throughout the year. Where water temperatures are raised significantly above the natural level, as by steam powerplants using water-cooled condensers, a knowledge of the rate at which water temperatures are reduced downstream by natural cooling processes is important. A theoretically sound approach to the problem is the energy-budget concept, which has been used extensively to calculate evaporation rates from lakes and reservoirs (Harbeck and others, 1958). THEORY In applying energy-budget methods to streams, the various inflow, outflow, and stored-heat terms are evaluated for successive short subreaches. In each of the subreaches, temperature change is assumed to be linear with distance. The energy budget for each subreach, over any arbitrarily selected time period, may be written as: Qr+Q^Qb-VQb+Qh+Qo+Qc-^-Qs, (l) where Qfi=net incoming atmospheric and solar radiation, Q(=heat content of water flowing into subreach, QB=back radiation (long-wave) from water surface, Qe=energy lost by evaporation, Qh=energy lost by convection to atmosphere, (20=heat content of water flowing out of subreach, (2c=heat conducted to streambed, and Qs=increase of heat storage in subreach. ART. 104 IN U.S. GEOL. SURVEY PROF. Each of the preceding terms, except Qe, may be evaluated, either by direct measurement or by calculation, by standard procedures developed for determining evaporation losses (U.S. Geological Survey, 1954). Conductive loss to the streambed can be computed, with reasonable accuracy, by multiplying the measured temperature gradient in the bed at a representative point in the subreach by the average thermal conductivity of the bed material. Equation 1 may be rearranged thus: (Qt~ Qo) = Qb-\-Qe-\-QbjtQc-\-Qs— Qr> (2) where the left-hand term denotes the loss in sensible heat of the water as it passes through the subreach. If the subreach originally selected is just long enough to produce a small, known temperature drop (for example, 1°C) and if all the right-hand terms are expressed in units of heat flow per unit area, then the required length, l, may be expressed as: , KFM (3) where F= average flow rate in subreach; AT= temperature drop in subreach; W= average width of subreach; and K= a constant, whose value depends on the units employed in the other terms. APPLICATION The energy-budget method was used to predict the temperature profile of a heated section of the West Branch of the Susquehanna River below Shawville, Pa. Use of the stream water for condensing steam from the turbines of the Shawville powerplant, which operates at a constant' load of 600,000 kw, raised the temperature of the river about 15°C during the period of study (October 17 and 18, 1962). Over a 24-hour period beginning at 0600 hours on October 17,1962, temperature measurements were made PAPER 475-C, PAGES C175-C178. 1963. C175C176 ENGINEERING HYDROLOGY approximately bihourly at several sections downstream and one upstream from the powerplant. Ten measurements were made at equal distances across the width of each section. Only one measurement was necessary at each point because vertical mixing was virtually complete. Calibrated thermistor probes, accurate to 0.1°C, were used in the measurements. Average temperatures at each of 4 sections downstream and 1 upstream from the powerplant are shown in figure 104.1. Station 1, about 0.4 mile downstream from the powerplant diversion dam, included the flow from a small, unheated tributary (Trout Run). No other appreciable surface inflow occurred in the 4.9-mile reach between stations 1 and 4 during the period of study. Streamflow was virtually constant at 370 cubic feet per second, by actual measurement, and ground-water inflow in the reach was insignificant. Wind velocity was measured with sensitive 3-cup totalizing anemometers that were installed in midstream at stations 2 and 3, exactly 2 meters above the mean water surface, and were read at 2-hour intervals. Figure 104.1.—Variation of stream temperature with time in the West Branch of the Susquehanna River, near Shawville, Pa. Bed temperatures were measured by means of a hollow plastic probe driven into the streambed at station 1. Three thermistors which were embedded 6 inches apart in the surface of the probe were connected to a portable wheatstone bridge circuit. The bed-temperature readings were taken simultaneously with the stream-temperature readings. Total radiation, solar radiation, and humidity measurements were made in a well-exposed area on the roof of a local school about 0.1 mile north of the river, between stations 3 and 4. In calculating heat conduction to the streambed, Qc, an average thermal conductivity of 0.004 cal per sec cm °C was used, corresponding to a value about midway between that of wet mud and sandstone (Ingersoll and others, 1948, p. 244). The bed loss in each successive subreach below station 1 was estimated by assuming that it was approximately proportional to the dilference between the average water temperature in the subreach and the natural water temperature, as measured upstream from the powerplant. No really serious errors can result from this assumption, because the actual value of the Qc term was only about 5 percent that of the sensible heat loss, (Qi—Qo), in each subreach. Conduction and convection losses to the atmosphere, Qh, were calculated by multiplying the evaporation losses, Qe, by a modified Bowen ratio, Rr. The method has been used in previous energy-budget studies. The modified Bowen ratio was calculated to fit a recently developed theory (W. D. Sellers, oral communication, 1962) that involves the ratio of the heat and momentum transfer coefficients (KH and KM)- In his original work, Bowen (1926) assumed that KH/KM was virtually unity, whereas Sellers considers it to be a function of the Richardson number Ri. For the large differences between water and air temperatures measured in this study, the ratio KH/KM had an average value of about 1.30. The effect of applying this factor to the Bowen ratio was to increase Qh by an amount approximately equal to 7 percent of the (Qi—Qo) term. The actual rate of cooling of the stream, as determined by the average temperature at each measuring station over the 24-hour study period, was considerably greater than that predicted by theory (fig. 104.2). The average stream temperature fell 3°C in slightly less than 2 miles of flow, whereas the predicted length of reach, to effect the same temperature decrease was 3.3 miles. Several possible explanations for the discrepancy, in decreasing order of probability, are: 1. Excessive values were assigned to QB in the calculation. The total radiant energy falling on the stream surface may be considerably less than that measured by the radiometer, because of shading by nearby hills and trees. The degreeMESSINGER C177 of shading is a complex function of the sun angle, time of year, direction of flow, type and amount of adjacent vegetation, and surrounding topography. 2. Errors were made in windspeed measurements. The average windspeed measured at station 3, during the last 10 hours of the study period, was less than 0.4 miles per hour. Accuracy of the measurement at such low velocities is questionable, because of friction losses in the bearings and gear trains. 3. Knowledge of water-surface temperatures, rela- tive to the measured bulk temperatures, was in- Figure 104.2.—Variation of average stream temperature with distance downstream from station 1 for entire 24-hour study period. adequate. Surface temperatures directly affect the Qbi Qe-, and QH terms of the energy budget. If, for example, during periods of high thermal radiation, the surface temperature were only 1°C higher than the bulk value, the (Qi~Qo) term would be increased about 7 percent. In an attempt to determine the causes of the discrepancy, separate energy-budget computations were made for each 4-hour period. The results, shown in the six curves of figure 104.3, suggest the explanation for the discrepancy. The curves show that the greatest deviation from theory occurred during the daylight hours; especially, between 1000 and 1400 hours, when the net incoming radiation was highest (288 cal per cm2). During the next 4-hour period (Qr=211 cal per cm2), the deviation was only slightly less. Conversely, the smallest deviations are noted for the two 4-hour nighttime periods beginning at 1200 hours (Qr=78 and 81 cal per cm2, respectively). These observations give greater weight to the possibility of errors in the QR terms and in the assumed water-surface temperatures, as discussed above. On the other hand, windspeed inaccuracies become somewhat less important, because the lowest speeds occurred during the nighttime hours when deviations from the theory were smallest. Future energy-budget studies of streams will center about the development of more reliable techniques for evaluating the Qr term. The use of noncontacting, infrared thermometry should establish whether or not temperature differences at the water surface are of sufficient magnitude to affect the energy budget materially. REFERENCES Bowen, I. S., 1926, The ratio of heat losses by conduction and by evaporation from any water surface: Phys. Rev., v. 27, p. 779-787. Ingersoll, L. R., Zobel, O. J., and Ingersoll, A. C., 1948, Heat conduction, 1st ed.: New York, McGraw-Hill Book Co., 278 p. U.S. Geological Survey, 1954, Water-loss investigations—Lake Hefner studies, technical report: U.S. Geol. Survey Prof. Paper 269, 158 p.TEMPERATURE, IN DEGREES CENTIGRADE C178 ENGINEERING HYDROLOGY Figure 104.3.—Variation of stream temperature with distance downstream for various 4-hour study periods, measured temperature; dashed line, predicted temperature. Solid line,Article 105 MOVEMENT OF WATERBORNE CADMIUM AND HEXAVALENT CHROMIUM WASTES IN SOUTH FARMINGDALE, NASSAU COUNTY, LONG ISLAND, NEW YORK By N. M. PERLMUTTER; MAXIM LIEBER,1 and H. L. FRAUENTHAL;2 Mineola, N.Y.; and Hempstead, N.Y. Work done in cooperation with Nassau County Department of Health and Nassau County Department of Public Works Abstract.—A slug of contaminated ground water moving through glacial outwash and Containing as much as 3.7 parts per million of cadmium and 14 ppm of hexavalent chrominum was investigated by test drilling and sampling in 1962. The slug extends about 4,200 feet from recharge basins at a plating plant to MaSsapequa Creek, where part of the waste discharges naturally and part moves a short distance downgradienit beneath and into the stream. A slug of contaminated ground water containing high concentrations of cadmium and chromium from plating wastes, which have been discharged during the past 20 years by an industrial plant in South Farmingdale in east-central Nassau County, was defined in detail by test drilling in 1962 (figs. 105.1 and 105.2). The slug has been mapped four times since 1949 to determine changes in its size and to follow its course downgradi-ent. Hexavalent chromimum in excess of 0.05 parts per million and cadmium in excess of 0.01 ppm are considered objectionable in public-water supplies (U.S. Public Health Service, 1962), although the cumulative toxicity of these metals as a result of long-term consumption by humans and animals is uncertain. Short-term ingestion of the metals may produce some ill effects, but fortunately much of the contamination seems to be eliminated from the body by natural processes. 1 Assistant Director, in charge of Sanitation Laboratories, Division of Laboratories and Research, Nassau County Department of Health. 2 Hydraulic Engineer, Division of Sanitation and Water Supply, Nassau County Department of Public Works. ART. 105 IN U.S. GEOL. SURVEY PROF. 740 73° 72° Figure 105.1.—Map of Long Island, N.Y., showing location of South Farmingdale, Nassau County. We wish to thank all the Nassau County and U.S. Geological Survey personnel who participated in the field and office work, particularly Mr. Charles Kirsner, Department of Public Works who supervised the welldrilling program, Commissioner Eugene F. Gibbons and Deputy Commissioner John H. Peters, Nassau County Department of Public Works, and Dr. Joseph H. Kinnaman, Commissioner, Department of Health, for their support and interest. Hexavalent chromium was detected first in a supply well at the industrial plant in South Farmingdale in 1942, but no further study was made of the problem for several years owing to a .shortage of personnel during PAPER 475-C, PAGES C179-C184. 1963. C179C180 ENGINEERING HYDROLOGY Recharge basins Spi x TOME®] 1000 o 1000 2000 FEET 1 > ... I--------------1__________| Aerial photograph of South Fanningdale area showing extent of contamination in 1962 (ruled area), and location of sections X-X' and Y-Y'. World War II. In 1945, seven shallow test wells were installed, as far as 900 feet south of the plant, by the hiew York City Department of Water Supply, Gas, and Electricity to investigate the potential hazard to the city’s infiltration gallery at Massapequa, about 3y2 miles south of the source of contamination. The test wells, which were bottomed only a short distance below the water table, showed little or no contamination in 1945, but a resampling of several of the wells by the New York State Department of Health in 1948 showed as much as 3.5 ppm of hexavalent chromium in the water. On the basis of these early studies, a chromium treatment plant recommended by the State Health Department was put into operation at the industrial plant in 1949. Because of continued concern over the effectiveness of the treatment for chromium and the fact that cadmium and other metals were being removed from the waste water only incidentally, if at all, the Nassau County Departments of Health and of Public Works made a series of surveys of the contaminated slug, including test drilling, in 1949,1953,1958, and 1962. The results of the first two surveys were reported by Davids and Lieber (1951), Lieber and Welsch (1954), and WelschPERLMTJTTER, LIEBER, AND FRAUENTHAL C181 (1955). Their reports showed that the slug extended at least 3,600 feet from the plant by 1949. The 1958 survey was incomplete and the results were never published. INVESTIGATION IN 1962 The 1962 investigation of the geologic, hydrologic, and chemical aspects of the contaminated water was the most intensive to date. About 90 test wells were drilled in and near the slug by the Nassau County Department of Public Works, and a series of water samples were collected from Massapequa Creek. The test wells were drilled in the vicinity of the recharge basins at the plating plant, along Motor Avenue, Lambert Avenue, Fallwood Parkway, Plitt Avenue, Spielman Avenue, EaSt Drive, and in the bed of Massapequa Creek and along its banks between Spielman and Second Avenues (fig. 105.2). Most of the wells were driven with 1 Clinch drive points, and ranged in depth from about 8 to 75 feet. The final depth of the driven wells was sufficient to determine the full thickness of the slug at most locations. Wells were drilled by cable tool to depths of 97 to 140 feet at 3 sites to check for contamination in the deeper 'beds and to collect geologic samples. Water samples collected at 5-foot intervals were analysed mainly for hexavalent chromium and cadmium by the Nassau County Health Department, Division of Laboratories. Hexavalent chromium was determined by the use of s-diphenylcarbazide and colorimetric comparison, and cadmium was determined by dithizone extraction (American Public Health Association and others, 1960). The chromium determinations are believed to be reliable to the nearest 0.01 ppm. Because the test for cadmium is still considered to be tentative owing to problems in the analytical procedure that are probably caused by interfering metallic ions, some doubt exists about the reliability of cadmium determinations below concentrations of about 0.05 ppm. Results of chemical and spectrographic analyses of several water samples by the Geological Survey were not available at the time of preparation of this article. GEOHYDROLOGIC ENVIRONMENT The land surface in the South Farmingdale area (fig. 105.2) is a gently sloping outwash plain notched by several south-flowing shallow tributaries of Massapequa Creek. It ranges in altitude from about 40 to 70 feet above sea level and is underlain by 2 geologic units of significance in this study (fig. 105.3A) : an upper unit of glacial outwash, called the upper Pleistocene deposits, and a lower unit of stream deposits of Late Cretaceous age, called the Magothy (?) Formation. The upper Pleistocene deposits consist chiefly of highly permeable beds of brown fine to coarse sand and gravel containing a trace of silt, and some subordinate thin scattered lenses of silt and silty clay. The lower 20 to 30 feet of the unit generally contains more clay than the upper part. The total thickness of the upper Pleistocene deposits ranges from about 80 to 130 feet. The Magothy (?) Formation consists chiefly of beds and lenses of gray fine sand and sandy and silty clay containing lignite, some beds of brown fine sand, and, in a few places, thin beds of coarse sand containing some gravel. The strata of the Magothy (?) Formation have a lower average permeability than those of the upper Pleistocene deposits, because of their higher clay content and poorer sorting. Neither of the geologic units in the report area contains lenses of silt and clay that are either thick or extensive enough to act as significant hydraulic barriers to the movement of water. However, the lenses of silt and clay doubtless have an important geochemical influence on local concentrations of cadmium and chromium ions. About half the average annual rainfall of 45 inches percolates down to the upper Pleistocene deposits, where the ground water occurs under unconfined conditions. The water table, which is the upper surface of the zone of saturation in the ground-water reservoir, ranges in depth from about 15 feet below land surface in the northern part of the area to within a foot of the land surface at Massapequa Creek in the southern part. It has an average gradient of about 12 feet per mile. Water in the underlying Cretaceous deposits is connected hydraulically with the water in the Pleistocene deposits but is increasingly confined at depth by layers of silt and clay. However, the two geologic units form a single aquifer having a wide range in permeability. Water enters the aquifer beneath the South Farming-dale area by (1) direct recharge from rainfall, (2) underflow from the adjoining area to the north, (3) return of industrial waste water through recharge basins, and (4) seepage from hundreds of domestic cesspools. The ground water moves generally southward (fig. 105.3/1); some discharges into Massapequa Creek, and some moves as underflow beneath the creek and the adjoining area. Most of the report area is supplied with water from deep wells owned by the South Farmingdale Water District. A small number of home-owners maintain shallow wells that are used for lawn sprinkling only. EXTENT AND MOVEMENT OF CONTAMINATION Sampling of the test wells and Massapequa Creek shows that the contaminated slug of ground water is a cigar-shaped body in plan view, and extends south from the industrial recharge basins to the valley of Massapequa Creek, where it is moving slowly downgradientC182 engineering hydrology Figure 105.3.—Sections showing geology and direction of movement of ground water (A), and lines of equal chromium (B) and cadmium (O) content, in parts per million, along line X-X' (fig. 105.2) in 1962. Contaminated water body shaded. into and beneath the stream (figs. 105.2 and 105.3 B, C). The slug was about 4,200 feet long and had a maximum width of about 1,000 feet in 1962. It is elongated generally in the direction of natural ground-water flow and is widest near Spielman Avenue. The width of the slug actually- doubles in a distance of 2,700 feet between Motor and Spielman Avenues, owing largely to dispersion. Further widening of the slug is apparently deterred by Massapequa Creek, which acts hydraulically as a line sink or line discharge and hence deflects the flow from its original path. The slug narrows considerably at its southern extremity as it moves beneath the valley toward points of discharge in Massapequa Creek. Section X-X' (fig. 105.3 /?, C) shows the vertical extent and concentration of hexavalent chromium and cadmium in a plane oriented along the approximate direction of flow in the center of the slug. The maximum concentration of hexavalent chromium detected in 1962 was 14 ppm at well 23 near Spielman Avenue, and the maximum cadmium content was 3.7 ppm at well 67, a short distance south of the recharge basins. Figure 105.3/? shows that the concentration of hexavalent chromium is decreasing at the north end of the slug, and comparison of figures 105.37? and 105.3/7 shows that the chromium ions have migrated farther south than the cadmium ions. Chemical analyses made in 1962 of the treated effluent from the plating plant show that the effluent is generally relatively free of hexavalent chromium but at times contains as much as 3.5 ppm of cadmium.PERLMTJTTER, LIEBER, AND FRAUENTHAL C183 Section Y-Y' (fig. 105.4) shows the concentration of . hexavalent chromium beneath Lambert Avenue in a vertical plane approximately perpendicular to the direction of ground-water flow. Two centers of contamination are illustrated on the section. The main center is near the intersection of Woodward Parkway and Lambert Avenue. A second and smaller center of contamination at the east end of the section may be due to relatively minor leakage from one of the plating buildings east of the main line of flow from the recharge basins. The 1962 survey showed for the first time that the contamination had not only reached Massapequa Creek but was present in the stream as well as in the beds beneath it. The maximum concentration of hexavalent chromium in Massapequa Creek was 2.1 ppm, about 300 feet north of Tomes Avenue (fig. 105.2), and a concentration of 0.01 ppm was found in a sample from the creek about a mile south of Southern State Parkway. The maximum load of hexavalent chromium carried by the stream was calculated to be about 4.5 pounds per day at 2d Avenue. The cadmium concentration in the stream ranged from 0 to 0.07 ppm north of Southern State Parkway; it was not detectable south of Southern State Parkway. The public-supply wells nearest the contaminated slug are those of the South Farmingdale Water District (fig. 105.2). The wells are screened to a depth of about 150 to 600 feet in the Magothy (?) Formation and, owing to their depth and the direction of movement of the slug, are in no danger of contamination now or in the foreseeable future under present hydraulic conditions. No general-purpose domestic wells are known to be in use in the present path of the slug. The Massapequa infiltration gallery, which serves as a standby source of water for New York City, is about 2y2 miles south of the southern extremity of the slug. No water has been drawn from the gallery since February 1958. Because of the great dilution of the slug after it discharges into Massapequa Creek, no significant concentrations of metallic contaminants have reached or are likely to reach the vicinity of the gallery via the stream. Also, the opportunity for dilution, by millions of gallons of fresh ground water, of any remnant of the slug which might move to the gallery as underflow suggests that potential operation of the gallery in the future should not be seriously influenced by this source of contamination. Because of the slow rate of ground-water flow, the age of the contaminated water in different parts of the WEST Y Q a: P EAST Y‘ Water table <0.01 <0.01 <0.01 <0.01 '' 4.00 <0.01 __ < W “ 0.01 • 2.25 2.40 190 ~ ,1 20- xObserved 1959 / /„ ' '\ /--^'Synthesized i i i i i I i I I I 1 1 1 _J l_ 5 10 15 20 25 APRIL 5 10 15 20 25 MAY - ..... ■ ■ - - 5 10 15 20 25 JUNE - '■ *•. •«*'. v. . 6000 4000 2000 0 Figure 108.1.—Synthesized and observed daily runoff during the snowmelt seasons, North Yuba River below Goodyears Bar, Calif. DAILY MEAN DISCHARGE, IN CUBIC FEET PER SECONDArticle 109 FIELD VERIFICATION OF COMPUTATION OF PEAK DISCHARGE THROUGH CULVERTS By C. T. JENKINS, Denver, Colo. Abstract.—Tests on a battery of five 58- by 36-inch pipe-arch culverts show that, for the range tested, agreement is good between measured discharges and those computed indirectly from studies of small-scale models. Tests made at a battery of five pipe-arch culverts indicate good agreement between results of indirect methods for computation of peak discharge at culverts and the results of current-meter measurements made under the most favorable conditions. The tests were made on the Colorado River about a mile downstream from Lake Granby, Colo., during March and April 1962. The methods of indirect computation are those outlined by Carter (1957), based on laboratory model tests made by the U.S. Geological Survey at Georgia Institute of Technology and on the results of previous tests by other investigators. Carter (1957, p. 2 and pi. 1) identified six types of flow, classified according to the factor or factors that control the discharge through the culvert, and defined coefficients of discharge for each type for a large range in entrance geometry. Generally, flow conditions in a culvert at the time of peak discharge are not observed in the field, and the flow type must be deduced from criteria based on culvert geometry and relative elevations of high-water marks. However, during the tests below Lake Granby it was possible to classify the flow types by observation as Carter’s types II and VI. Hie control for type II is critical depth at the outlet; for type VI, entrance and barrel geometry. Carter’s coefficients for type-II flow are based generally on tests on models with barrel diameters of 4 to 6 inches, and those for type VI on tests on models 18 to 36 inches in diameter. The small models were either circular or rectangular in cross section; the larger models included corrugated pipe arches. Carter (1957, p. 19) specified that his type-VI coefficients apply to pipe arches, but did not state whether or not his type-II coefficients apply to shapes other than those tested. In practice, they are used for pipe arches as well as for circular sections. Each of the culverts below Lake Granby was a 58-by 36-inch standard corrugated metal pipe arch 51 feet long. The culverts were about 9 feet apart center-to-center, and were laid on slopes ranging from 0.004 to 0.01. The entrance projections ranged from 1.2 to 2.8 feet from the embankment. The openings differed only slightly from the standard dimensions and shape given by the manufacturer. During the test period, four discharges were held steady long enough for flow to stabilize through the culverts. Each flow was measured with a current meter at an excellent measuring section just below the lake. Surveys were made of the geometry of the culverts and the approach section and of water-surface elevations for each flow. Values of the discharges computed as outlined by Carter (1957, p. 2-10, 14) are shown in the accompanying table. Carter (1957, pi. 1) indicated that type-VI flow occurs only when the ratio of the height (Ai—z) of water above the upstream invert to the diameter (D) of the culvert is greater than 1.5; type-V flow was observed in the field in all five culverts when (h±—z)/D ranged from 1.34 to 1.45. However, Carter implied that the {h1—z)/D criterion is not inflexible, because he listed coefficients for “high-head” flow for (h1—z)/D ratios as small as 1.3. Agreement between computed and measured discharge is good, indicating that, within the range tested, C194 ART. 109 IN U.s. GEOL. SURVEY PROF. PAPER 475-C, PAGES C194-C195. 1963.JENKINS C195 no serious error is introduced by applying the results of model tests to prototypes much larger than the model, and in some instances, substantially different in shape. However, similar field tests on other prototypes, especially at larger (hj-z)/D ratios, would be required to confirm completely the results of the model study. During such tests, the type of flow should be noted so that criteria for flow classification, as well as discharge coefficients, can be tested. REFERENCE Carter, K. W., 1957, Computation of peak discharge at culverts : U.S. Geol. Survey Circ. 376. Comparison of measured and computed discharge for five culverts Range of (hi—z)/D Type of flow i Discharge Observed in field From Carter’s criteria Measured (cfc) Computed (cfs) Difference (percent) 0.92-1.03 n ii 244 246 + 0. 8 .99-1.10 ii ii 275 268 -2. 5 1.34-1.45 VI 373 374 + 3 i.34-1.45 ii 373 354 — 5 1 1.58-1.69 VI VI 413 445 + 7. 8 1 Types of flow described by Carter (1957).Article 110 USE OF LOW-FLOW MEASUREMENTS TO ESTIMATE FLOW-DURATION CURVES By OLIVER P. HUNT, Albany, N.Y. Work done in cooperation with New York State Department of Public Works Abstract.—Flow-duration curves for streams where only a few low-flow measurements are available can be estimated on the basis of nearby gaged streams. The low-flow measurements must be made under base-flow conditions. A new technique is described for estimating low flows under these conditions. Flow-duration curves for streams where only a few low-flow measurements have been made can be estimated by relating the discharge of these measurements to the concurrent discharge at gaging stations on one or more comparable streams for which flow-duration curves are already available. A method of estimating the flow-duration curves is described by Searcy (1959). A shorter method that gives practically the same results is described in this article. In both methods, the flow measurements must be made, during periods when the flow consists largely of ground-water effluent (Lang-bein and Iseri, 1960, p. 5), and measurements must be made under several fates of base flow. The records for three recording gages in Dutchess County, N.Y., were selected to illustrate the use of this method. Sites with actual gaging-station records were selected so that the duration curve developed by this method could be compared with the actual duration curve defined by many years of record. The stations used, with period of record available, are listed below: Fishkill Creek at Beacon, N.Y. (1944-55). Tenmile River near Gaylordsville, Conn. (1929-55). Wappinger Creek near Wappingers Falls, N.Y. (1928-55). As the records for the above stations were for different periods of time, all were extrapolated to a common base period, 1926-55. (See Searcy, 1959, p. 12-14). On the assumption that Wappinger Creek is the unknown, a duration curve of daily flow was developed on the basis of nine selected base-flow discharge measurements. The dates and discharges used are shown in columns 1 and 2 of table 110.1 Columns 3 and 4 are corresponding daily discharges from Tenmile River and Fishkill Creek, respectively. Columns 5 and 6 show the percentage of time that discharges in columns 3 and 4 were equaled or exceeded, obtained from base-period duration curves of daily flow for the Tenmile River and Fishkill Creek gaging stations. Two independent duration curves of daily flow for Wappinger Creek were developed using logarithmic-probability paper (fig. 110.1). One of these curves was based on the Tenmile River record and one on the Fishkill Creek record. For the curve based on the Tenmile River record, discharges shown in column 2 of table 110.1 were plotted at percentages listed for the same date in column 5 of table 1. The smooth curve (solid line) on figure 110.1 was then drawn on the basis of these plotted points. A curve based on Fishkill Creek record (dashed line) was drawn in the same manner, using percentages shown in column 6 of table 110.1. The hypothetical flow of Wappinger Creek estimated as described above is compared in table 110.2 with the actual flow as recorded at the gaging station near Wappingers Falls, N.Y., for the period 1926-55. The discharges shown in column 2 of table 110.2 were taken from the correlation curve based on the Tenmile River record (solid line) on figure 110.1 for selected percentages of time. Similarly, the discharges in column 3 of table 110.2 were taken from the correlation curve based on the Fishkill Creek record (dashed line) on figure 110.1. The average of columns 2 and 3 is listed in column 4. Column 5 lists the discharges taken from the actual duration curve of daily flow for Wappinger Creek for the base period 1926-55. Columns 6 and 7 indicate the difference (in cubic feet per second and in percent) between the mean of the duration curves C196 ART. 110 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C196-C197. 1963.HUNT C197 based on nine selected discharge measurements and that based on the 30-year base period 1926-55. The mean duration curve of daily flow for Wappinger Creek developed by use of the nine discharge measurements compares very favorably. This close comparison is believed due to the fact that (1) the centroid of each drainage area used for correlation is within 20 miles of the Wappinger Creek site, (2) the precipitation pattern is similar for all three basins, and (3) discharge measurements, column 2 of table 110.1, were of base flow. In any study involving the derivation of duration curves from base-flow measurements it is essential that precipitation patterns be comparable in the basins considered. A substantial saving in time (about 50 percent) has been achieved by using the method described in this article in low-flow duration studies in New York. The results compare very closely with those obtained by using the procedure described by Searcy (1959). Figure 110.1—Estimated duration curve of daily flow, Wappinger Creek, near Wappingers Falls, N.Y. Table 110.1.—Base-flow measurements for Wappinger Greek and concurrent daily discharges and percentage of duration for Tenmile River and Fishkill Creek Date Discharge measure- ment, Wappinger Daily discharge (cfe) Percent of time discharge equaled or exceeded that shown Creek (cfs) Tenmile River Fishkill Creek Tenmile River Fishkill Creek 1 2 3 4 5 6 1956: May 10 . _ _ 380 492 519 18 16 June 30 49 80 79 73 71 Aug. 18-_ -- 20 25 24 95. 8 91 1957f May 10 118 139 130 60 60 July 11 14 17 11 98. 9 97 Sept. 13 4. 8 10 6. 0 99. 96 99. 3 1958: June 20 71 113 118 65 62 Aug. 20_ 17 31 18 92. 8 93. 7 Sept. 25_ 58 89 44 71 82. 5 Table 110.2.—Comparison of estimated discharge for selected percent duration with discharge from base-period duration curve for Wappinger Creek Percentage of time discharge equaled or exceeded that shown Discharge from correlation curve (cfs) Discharge of Wappinger Creek, from 1926-55 duration curve (cfs) Difference between correlation curve and 1926-55 curve Tenmile River Fishkill Creek Average of columns 2 and 3 cfs Percent 1 2 3 4 5 6 7 20 355 340 348 390 -42 -10. 8 30 265 255 260 280 -20 -7. 1 40 200 190 195 200 -5 -2. 5 50 148 138 143 142 + 1 + • 7 60 102 92 97 98 -1 -1. 0 70 61 57 59 63 -4 -6.3 80 38 36 37 39 -2 -5. 1 85 30 28 29 30 -1 -3.3 90 23 22 22. 5 22. 5 0 0 95 17 15 16 16 0 0 99 10 8. 4 9. 2 10. 2 -1.0 -9. 8 REFERENCES Searcy, J. K., 1959, Flow-duration curves, in Manual of hydrology, pt. 2, Low-flow techniques: U.S. Geol. Survey Water-Supply Paper 1542-A., p. 1-33. Langbein, W. B., and Iseri, K. T., 1960, General introduction and hydrologic definitions, in Manual of hydrology, pt. 1, General surface-water techniques, U.S. Geol. Survey Water-Supply Paper 1541-A, p. 1-29.Article 111 GRAPHICAL MULTIPLE-REGRESSION ANALYSIS OF AQUIFER TESTS By C. T. JENKINS, Denver, Colo. Abstract.—The graphical method of multiple regression described is well adapted to estimating the permeabilities of various layers of an aquifer from well-log and pumping-test data. Results of graphical analysis of a set of hypothetical data agree closely with the results of algebraic analysis of the same data. Multiple regression has been used to solve several problems in hydraulics and hydrology. Among these are determination of stage-fall discharge relations, relations between basin characteristics and index floods, and low-flow characteristics of streams. Although, to the knowledge of the writer, the technique has not been widely used in analysis of well logs and of the results of pumping tests to estimate the permeabilities of various layers of an aquifer, it seems to be particularly well adapted to that purpose. The procedure in no way enhances the data. The results are subject to errors introduced by departure of the aquifer from the ideal and to errors in the data, but the “best” answers can be obtained quickly and simply. The method used may be either algebraic or graphical. A graphical method probably is faster, if there are only a few tests and only 2 or 3 layers to be considered, and it has the advantage of revealing anomalous data readily. Curvilinear relations between the thickness of a material and its permeability hardly would be expected, but if such relations exist, they can be detected much more readily by the graphical method than by the algebraic method. On the other hand, the dependability of the results can be calculated if the algebraic method is used. Both graphical and algebraic methods are described in many textbooks (for example, Ezekiel, 1941). The illustrations presented here are specific applications of long-known standard methods. Selection of a “model” that the data are expected to follow is essential if the algebraic method is used, and highly desirable if the graphical method is used. The ART. Ill IN U.S. GEOL. SURVEY PROF. C198 model for pumping-test analyses is obvious. Mathematically, the relation expected is expressed by the equation Z1=bZa + cXz + d2T4 • • -, (1) where Xx is the transmissibility of the aquifer (the dependent variable), the coefficients b, c, d, • • • are the permeabilities of the various layers, and X2, X3, Xi, • • • are the corresponding thicknesses (the independent variables). The expected relation is the simplest form of multiple regression. The terms on the right side of the equation can be described graphically by straight lines through the origin with slopes b, c, d, ... if X2, X3, Xt, . . . are plotted as abscissas against the parts of Xx that are due to the transmissi-bilities of the corresponding layers as ordinates. Rigorous solution of the equation generally will show an intercept; that is, the calculations will indicate a finite value, either positive or negative, for transmissibility of the aquifer when all thicknesses are zero. A large value of the intercept is one indication that the data are not consistent. The graphical method of evaluating the intercept and the coefficients b, c, d, • • • for the hypothetical data shown in the accompanying table is illustrated in figure. 111.1. The steps are: 1. Plot the transmissibility of the aquifer as the ordinate against the independent variable that is estimated to have the greatest effect (in this case, the thickness of the gravel, X2). Draw curve 1 as the average of the points. 2. Draw curve 2 through the origin parallel to curve 1. 3. Measure the vertical deviation of each point from curve 2 with dividers and plot it as the ordinate against the variable that has the second greatest effect (in this case, the thickness of sand and gravel mixed, X3) as the abscissa. Draw curve 3 as the average of these points. PAPER 475-C, PAGES C198-C201. 1963.DEVIATION FROM CURVE 4 DEVIATION FROM CURVE 2 TRANSMISSIBILITY, IN 10,000 GPD PER FT JENKINS C199 4. Draw curve 4 through the origin parallel to curve 3. 5. Plot the vertical deviations from curve 4 of each point against the thickness of sand. Draw curve 5 as the average of these points. This line generally will not pass through the origin. Because it is known that there should be no intercept for the curve showing the estimated transmissibilities of various thicknesses of sand, it might seem logical to draw a line through the origin parallel to curve 5. This can be done, but it is not necessary. At this point in the analysis, the posi- tions of the curves make little difference; it is their slopes that are important, because the slopes represent the first estimates of the permeabilities. 6. The next step is to remove the estimated effects of the last two variables from the plotted position of the points that defined curve 1. For example, well 9 penetrates 20 feet of sand and 24 feet of sand and gravel (see table). From curve 5, the estimated effect for 20 feet of sand is 80,000 gallons per day per foot and for 24 feet of sand and gravel, 110,000 gpd per ft—a total of 190,000. The transmissibility is 750,000 gpd per ft, Curve 8 5000 Curve 9 Slope SAND AND GRAVEL Curve 10 1000 SAND Curve GRAVEL Curve 7 Curve I'j, Curve 6 Curve 2 Curve Curve 4 SAND AND GRAVEL Curve 5 0 10 20 30 40 50 60 70 80 90 THICKNESS OF LAYER, IN FEET 20 30 40 50 60 70 80 90 THICKNESS OF LAYER, IN FEET ?. 30 p: 20 50 Figure 111.1.—Graphical determination of multiple-regression equation from pumping-test data and well logs.C200 ANALYTICAL HYDROLOGY so the adjusted ordinate value in units of 10,000 gpd per ft is 75 —19=56. This value represents the new estimate of the transmissibility of 61 feet of gravel, and the plotted position is indicated by a plus sign on the first diagram. Similar treatment of the other points results in a new scatter diagram. The excessive deviation of the point for well 11 indicates that the data from well 11 are anomalous. Either the data are in error, or the aquifer differs from that penetrated by the other 19 wells. Data from well 11 were not considered in the rest of the analysis. 7. Draw curve 6 as the average of the plotted plus signs. 8. Draw curve 7 through the origin parallel to curve 6. 9. Plot the vertical deviations from curve 6 of the originally plotted points against the thickness of the sand and gravel mixture. Draw curve 8 as the average of the plotted points. Draw curve 9 through the origin parallel to curve 8. 10. Draw curve 10 through the origin if possible. If the intercept of curve 10 differs appreciably from zero, or if the remaining scatter seems excessive, steps 6 through 10 can be repeated; otherwise, the slopes of curves 7, 9, and 10 are the “best” estimates of the permeabilities of the materials. The scatter about curve 10 and its intercept, if any, are related to the scatter and intercepts of all three curves, 7, 9, and 10, and are measures of the consistency of the data and the completeness of the solution. Hypothetical data -from pumping tests and well logs Well No. Transmissibility, (gpd per ft) Xi Thickness of gravel (ft) X, Thickness of sand and gravel (ft) Xj Thickness of sand (ft) x4 1 _ _ _ 210, 000 220, 000 10 20 22 2 __ 5 30 9 3 460, 000 130, 000 30 24 18 4 5 10 42 5 560, 000 20 60 38 6 - 490, 000 41 20 0 7 360, 000 520, 000 18 32 16 8 39 10 85 9 _ 750, 000 61 24 20 10 500, 000 11 80 30 11 600, 000 790, 000 60 40 20 12 60 32 15 13 480, 000 31 28 31 14 270, 000 20 0 82 15 320, 000 0 50 59 16 380, 000 20 34 25 17 - 440, 000 500, 000 10 62 18 18 -- 40 16 40 19 _ 470, 000 280, 000 35 20 11 20 - - - 20 10 31 The lack of an intecept by curve 10 and the small scatter about the curve probably would satisfy most investigators, but the procedure outlined in step 6 has been repeated, resulting in curve 11, to illustrate that a further reduction in scatter can be attained. The slope of curve 11 is nearer to the correct permeability of gravel than that of curves 6 and 7. The regression equation resulting from the graphical analysis is X,= 10,000 gpd/ft2X2 +5,000 gpd/ft2X3+1,000 gpd/ft2X4 (2) The coefficients of X2, X3, and X4 are, respectively, the permeabilities of gravel, sand and gravel mixture, and sand. The technique illustrated has been outlined in great detail in the interests of clarity, but several variations that entail a little less labor can be used. The two most important involve steps 1 and 6. In many analyses, it may be possible to make a better first estimate of the slope of curve 1 than that of a line through the average of the points. The investigator may know intuitively the approximate value of the permeability of the gravel. If so, he can eliminate curve 1 and draw curve 2 initially. In some analyses, a good estimate of the slope of curves 1 and 2 can be made by considering only those data that have equal, or nearly equal, A-alues of the second (and the third, if possible) most important independent variable. For example, in the illustration given, wells 2, 7, 12, and 16 penetrate between 30 and 34 feet of the sand and gravel mixture, and the range both in the transmissibilities of the sections penetrated and thicknesses of gravel is wide. If curve 1 had been drawn through these 4 points, the resulting slope would have been 9,400 gpd per ft2, much closer to the final value of 10,000 than the slope of 8,600 that resulted from drawing curve 1 as the average of all the points. The arithmetic involved in step 6 can be eliminated by applying, with dividers, the vertical deviation of the plotted points from curve 5 to curve 2, thus determining the position of the points that define curve 6. For example, point 9 is about 40,000 gpd per ft above curve 5, and its position as a plus sign in the first diagram is about 40,000 gpd per ft above curve 2, plotted against 61 feet, the thickness of gravel in well 9. Similarly, point 20 is about 20,000 gpd per ft below curve 5, and it is plotted as a plus sign 20,000 gpd per ft below curve 2.JENKINS C201 The best estimate of the permeabilities can be obtained without using curves 2, 4, 7, and 9, but if curves 7 and 9 are not used, it will not be possible to measure the intercept, if any, of curve 10. All variations of the graphical method will give the same results within the limitations of plotting accuracy and individual judgment as to what represents acceptable scatter. The mechanics of the algebraic method will not be discussed herein (see Ezekiel, 1941, or other texts), but the resulting equation (excluding data from well 11) is Xx= —800 gpd per ft+10,000 gpd per ft2X2 + 5,020 gpd per ft2Ar3 + 900 gpd per ft2AT4 (3) The close agreement between equations 2 and 3 illustrates that the graphical method, used with a reasonable degree of care, gives results consistent with those computed algebraically. REFERENCE Ezekiel, Mordecai, 1941, Methods of correlation analysis 2d ed.: New York, John Wiley and Sons, Inc., 531 p.Article 112 NOMOGRAPH FOR COMPUTING EFFECTIVE SHEAR ON STREAMBED SEDIMENT By BRUCE R. COLBY, Lincoln, Nebr. Abstract.—A nomograph for computing effective shear on bed-sediment particles from known mean velocities is given and explained. It is based on a slightly different velocity equation from the one that is commonly used, and it eliminates the need for trial-and-error computations of effective shear. Unless a streambed is approximately plane, only part of the total shear on the bed is effective in moving sediment on the bed, and this part is related to mean velocity and to bed configuration (Einstein, 1950, p. 9-10). Because bed configuration may vary widely and somewhat unpredictably in alluvial streams, the effective shear frequently is computed from known mean velocity. This article presents a nomograph for computing one measure of the effective shear from known mean velocity. Probably the most commonly used equation for computing effective shear is one based on work by Keulegan (1938) and given by Einstein (1950, p. 10) as u in which ^[gR7s' /12.27 R'x\ -5.75logi0^ y (1) 12=mean velocity for a cross section; ^gR'S= shear velocity with respect to sediment particles (a specific form of effective shear on bed sediment); <7=acceleration caused by gravity; R' = hydraulic radius with respect to particles; S= energy gradient; £=parameter for transition from a hydraulically rough to a hydraulically smooth boundary; and fcs=grain roughness, usually assumed equal to diameter for which 65 percent of bed sediment by weight is finer. The two numerical constants in this equation need further explanation. The constant 5.75 equals 2.30, the natural logarithm of 10, divided by the turbulence constant, k, which is here assumed to equal 0.40. The constant 12.27 is based on an equation given by Keulegan (1938, p. 722, eq. 60) and should be about correct for the particular form correction that applies to a rectangular channel whose hydraulic radius is 0.10 of the width. The constant would be about 11.6 rather than 12.27 for a rectangular channel whose hydraulic radius is 0.05 of its width. For such a channel the ratio of hydraulic radius to average depth, R/d, is 0.90, which is lower than the ratio for many natural channels. For an infinitely wide rectangular channel, the constant is the antilog of 6.00/5.75 (Keulegan, 1938, p. 717, eq. 38) or 11.1. The effect of a change from 11.1 to 12.27 under the log sign is usually small, so that a precise constant is not required, but a logical constant might be 11.4 or 11.5 for many natural channels. However, the use of average depth rather than hydraulic radius is convenient and is, in general, sufficiently accurate for most natural streams. Hence, in the following discussion and definition of a practical nomograph, 11.1c?, a quantity that is theoretically indicated for infinitely wide rectangular channels, has been used as follows: u wm 5.75 (2) The use of the hydraulic radius R under the log sign is consistent with Keulegan’s (1938) equations, but Einstein (1950, p. 10) changed the R under the log sign to R', presumably when he changed the shear velocity w* (equals \gR'S) to the shear velocity with respect to the particles w* (equals VgRS). Two items should be noted with respect to the use of R rather than R' under the log sign. One item is that the equation 2 can be derived directly by integration of the point-velocity equation that was used by Einstein (1950, p. 8, eq. 3) from the streambed to a distance d C202 ART. 112 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C202-C205. 1963.COLBY C203 above the bed. Hence, equation 2 correctly gives the average velocity at a stream vertical whose point velocities conform to the point-velocity equation used by Einstein. The second item is that the measure of depth, whether R or d, under the log sign has the nature of an upper limit for the integration of point velocities along a vertical. If It' is used under the log sign, the mean velocity u presumably should represent a mean for only that part of the cross section that is within a distance R' from the channel boundary, and this is a difficult mean velocity to determine by measurement. Because of the change under the log sign from R' to R or to d, a change that is usually far greater than the relatively insignificant change from 12.27 to 11.1, -JgR'S or «* computed from equation 1 may differ appreciably for some flows from w*, computed from equation 2. Several other uncertainties may arise with respect to the applicability of equation 2, but the only one of these to be discussed in this short article is the variability of k. Einstein and Chien (1954, fig. 12) showed a decrease of k with an increase in a measure of the energy that is required to transport suspended sediment over relatively smooth sand beds. Hubbell and Matejka (1959, p. 71, 72, and 74) reported values for k much higher than 0.40 for some flows over dune beds. Sayre and Albertson (1961, p. 144r-147) also found k much higher than 0.40 for flows over relatively widely spaced rigid baffles. (The term k is used here although some of the constants probably should not be called turbulence constants.) An experimentally determined turbulence constant for a given flow may be used, if it is considered to be more suitable than 0.40, by replacing 5.75 in equation 2 with 2.30/k. The mean-velocity equation in the form u V(M) r=5.75V~9 logic =34.0+32.6 (log,„ |+log.o *) (3) can be solved for ^(RS)m by trial and error, either arithmetically by estimating a value of x and repeating the computation for V(RS)m or graphically according to a method given by Colby and Hubbell (1961, p. 3-5 and pi. 1). The trial-and-error'feature of the computation can be eliminated by use of the nomograph shown on figure 112.1. The principles involved in the nomograph are fairly simple. Equation 3 shows that the ratio u/-,2= —1.372+ (7/8) log (pt—p0)d2/4pv2, with its limits of applicability, 4.50.6 are infrequent. The trend for proportionately larger wall shears along with larger Froude numbers (bottom curve, fig. 113.1) suggests the possibility that banks in a natural stream will erode until the Froude number and wall-floor shear ratio are decreased again to values associated with stream stability. For any given discharge and a given “stable” value of F for a particular stream, an increase in width due to erosion will tend to bring about a decrease in depth, and a larger width-depth ratio. Leopold and Maddock (1953, p. 29) bring attention to the works of others who claim that natural streams do indeed tend to develop large width-depth ratios: ... To be stable, the channel carrying bed loads, therefore, should have a higher velocity along the bed, but the same velocity along the banks, and this could only occur with a wider, shallower section. . . . Indian rivers tend to adopt broad, shallow sections, and . . . this type of cross section is best adapted for the transportation of heavy silt. The rigid test flume cannot adjust its shape, of course, but at low Froude numbers (and also at large depths or small width-depth ratios), it is evident from figures 113.1 and 113.2 that the values of shear around the perimeter tend to be more peaked or less uniform on the wall (top curve, fig. 113.1), higher on the floor than the average shear (middle curve, fig. 113.1; fig. 113.2), and smaller on the wall than on the floor (bottom curve, fig. 113.1).C208 ANALYTICAL HYDROLOGY An alluvial stream will react to nonuniform shear stresses by erosion and deposition. Such a natural stream, which has had opportunity to adjust and react to the shear patterns of figures 113.1 and 113.2 imposed by the flow of water through it and which has achieved a reasonably stable shape for the low Froude-number range at which it usually flows, could develop a cross section that is broad, has a large width-depth ratio, and is gently curved upwards towards the shores from a slightly deeper middle region. REFERENCES Leopold, L. B., and Maddock, Thomas, Jr., 1953, The hydraulic geometry of stream channels and some physiographic implications : U.S. Geol. Survey Prof. Paper 252, 57 p. Hsu, E. Y., 1955, The measurement of local turbulent skin friction by means of surface pitot tubes: The David W. Taylor Model Basin Research and Devel. Rept. 957.Article 114 SULFATE AND NITRATE CONTENT OF PRECIPITATION OVER PARTS OF NORTH CAROLINA AND VIRGINIA By A. W. GAMBELL, Washington, D.C. Work done in cooperation with the U.S. Weather Bureau Abstract.—Preliminary data from a precipitation-sampling network covering a 34,000-square-mile study area suggest that the atmosphere is a major source of sulfate and nitrate in the stream waters of southern Virginia and eastern North Carolina. The chemical composition of atmospheric precipitation has been the subject of several investigations in recent years. Studies by Junge and Werby (1958) and Eriksson (1959; 1960) are only two of the more recent investigations indicating that the atmosphere is a potential source of large quantities of water-soluble material. To learn more about the origin of this material, the processes by which it is introduced into precipitation, and, ultimately, its role in determining the quality of water, a network of precipitationsampling stations has been established covering a well-defined drainage area. This article presents data obtained from the first month of network operation. Average concentration and approximate total load of the major ionic constituents in precipitation have been computed. Some general comparisons with river-quality data illustrate the magnitude of atmospheric contributions to stream waters. Interpretation of the data is, of necessity, quite limited. Detailed interpretation will be reserved for a later time when data for a longer period are available. NETWORK DESIGN AND EQUIPMENT In July 1962, 28 precipitation-sampling stations were established in North Carolina and Virginia. These stations form a network covering approximately 34,000 square miles, drained by 5 major streams: the Chowan, ART 114 IN U.S. GEOL. SURVEY PROF. Roanoke, Tar, Neuse, and Cape Fear Rivers. The area included in the network is shown in figure 114.1 and location of the stations is shown in figures 114.2 and 114.3. Each sampling station is located at a U.S. Weather Bureau cooperative observer site. With one or two exceptions, sampling stations were purposely located outside the larger metropolitan areas. This was done to minimize contamination from local sources. The principal exception is the sampling station at Norfolk, Va. It is in the center of the city; accordingly, the precipitation has a high sulfate content. Each sampling station is equipped with a standard 8-inch precipitation gage and a special collecting device. The collecting device consists of a 5-inch-diameter glass funnel mounted in the top of an insulated enclosure. The funnel drains through a polyethylene tube into a polyethylene bottle in the interior of the enclosure. The entire device is mounted with the rim of the collecting funnel 5 feet from the ground. The funnel is continuously open and therefore collects a certain amount of dry fallout as well as precipitation. Composite samples are collected monthly from each station and analyzed for sulfate, chloride, nitrate, sodium, calcium, and potassium. MINERAL CONTENT OF RAINFALL AND STREAM WATERS Precipitation over the network area during August 1962 averaged slightly less than 4 inches. The heaviest precipitation fell in the southeastern part of the network; however, the amounts were reasonably uniform throughout the area. The areal variations of the major ionic constituents in precipitation are of considerable interest because they furnish valuable clues as to the source of these materials. PAPER 475-C, PAGES C209-C211. 1963. C209C210 GEOCHEMISTRY OF WATER Although only sulfate and nitrate are discussed here, the distribution of sodium, chloride, and calcium also displayed discernible patterns during August. Figure 114.2 illustrates two important features of the areal variation in sulfate concentration. It shows first that sulfate is widely prevalent in precipitation on the study area in August 1962 and, second, that the sulfate concentration increases inland from less than 1.5 parts per million along the Atlantic Ocean to more than 4.5 ppm in the northwestern part of the area. This trend of increasing concentration landward is contrary to what would be anticipated if the ocean were the principal source of atmospheric sulfate. The occurrence of sulfur in the atmosphere was discussed by Junge (1960). In unpolluted areas sulfur occurs primarily in three forms: as SO*"2 in aerosols, and as S02 and H2S gas. Sea-salt aerosols, soil dust, and the oxidation of S02 and H2S are sources of S04'2. A large part of the S02 in the atmosphere originates from the combustion of fossil fuels. Most of the H2S in the atmosphere is probably of natural origin, resulting from the decay of organic material. Although the role played by these substances is not completely understood, previous studies suggest that industrial activity and sea-salt aerosols are the predominant sources of atmospheric sulfur. Because there is little industry in the western part of the network area, a source other 78° 0 100 200 MILES 1 -------------1--------------1 Figure 114.1.—Precipitation-sampling network (shaded area). 78° A I Location of river-quality Drainage area represented sampling stations by river-quality sampling stations Figure 114.2.—Sulfate content of precipitation (parts per million) on study area, August 1962. Isopleth interval, 1.0 ppm. 78° EXPLANATION A |_____| —------“ Location of river-quality Drainage area represented Boundary of drainage sampling stations by river-quality sampling area stations Figure 114.3.—Nitrate content of precipitation (parts per million) on study area, August 1962. Isopleth interval, 1.0 ppm. Boundary of drainage areaGAMBELL C211 than local industry must account for the trend observed in figure 114.2. Either sulfur has been brought in by advection from a nearby area, or some presently unexplained natural source is responsible. The lines of equal sulfate content (fig. 114.2) clearly indicate that the ocean was not the major source of atmospheric sulfur in this area during August 1962. Figure 114.3 shows a distinct pattern in the areal variation of nitrate. Precipitation occurring over the coastal plain is almost devoid of nitrate. Farther inland, however, the precipitation contains nitrate in relatively large concentrations. Several explanations might account for this striking difference. The fixation of atmospheric nitrogen by lightning discharges is one known source of nitrate. Sixteen thunderstorm-precipitation samples collected by the author contained nitrate concentrations ranging from 0.7 to 8.1 ppm; the average concentration was 2.9 ppm. Because August is a month of frequent thunderstorm occurrence, at least part of the nitrate shown in figure 114.3 is probably the result of thunderstorm activity. The fact that convective storms are more frequent along the mountains than near the coast further strengthens this contention. The soil may be an indirect source of atmospheric nitrate, as ammonia is produced in the later stages of decay of most organic matter. If the pH of the soil is relatively high, some ammonia will escape to the atmosphere. Junge (1958) proposed that the oxidation of this ammonia may produce much of the nitrate found in precipitation. Therefore, areal differences in soil composition may account for differences in the nitrate content of precipitation. The addition to the soil of commercial fertilizers containing nitrate may also be significant. Another possibility is that much of the sulfate and nitrate in the western part of the area may have a common source. Oxides of both sulfur and nitrogen are produced in the combustion of most fuels. Data for the winter months should help greatly to clarify the picture. A summary of the data concerning sulfate and nitrate, as well as the other major ionic constituents for August 1962, is presented in table 114.1. The values in table 114.1 are based on the total volume of precipitation on the 34,000-square-mile, five-basin area. One important purpose of the network study is to determine what fraction of the dissolved solids in the natural water of the area is of direct atmospheric origin. Although data for a single month are insufficient to make such an evaluation, it appears that the atmos- pheric contribution will prove to be substantial. Table 114.2 illustrates this point with respect to sulfate and nitrate. The river values are time-weighted averages. The precipitation values are based on the volume of precipitation on the respective drainage areas. The shaded sections of figures 114.2 and 114.3 correspond to the drainage areas represented. Although the averages in table 114.2 cannot be used for direct comparison, the general implications are evident. In an area where the natural water is of relatively low dissolved-solids content, such as that covered by the network, the atmosphere may be the principal source for a variety of constituents. Table 114.1—Average concentrations and approximate total loads of major ionic constituents in precipitation on the study area during August 1962 Ion Average concentration (ppm) Total load (tons) SOr2 2. 8 25, 000 Ca+2 1. 1 10| 000 Na+1 . 8 7', 500 5, 300 3, 800 1, 200 NOs-' . 6 CT> . 4 K+1 . 1 Table 114.2.—Comparison of sulfate and nitrate content of river waters with that of precipitation in August 1962 Drainage area Sulfate (ppm) Nitrate (ppm) Sampling station (figs. 114.2 and 114.3) upstream from station (sq mi) River water Pteeipi- tation River water Precipi- tation A. Chowan River at Wlnton.N.C. 4,198 5.6 2.5 0.9 0.1 B. Tar River at Tarboro, N.C 2,140 5.1 2.9 2.3 .5 C. Neuse River at Cowen Landing near Vanceboro, N.C 4,027 7.4 2.7 4.6 .4 D. Cape Fear River at Navassa, N.C 7,060 9.4 3.1 .8 .5 REFERENCES Eriksson, Erik, 1959, The yearly circulation of chloride and sulfur in nature; meteorologic, geochemical and pedalogieal implications, pt. 1: Tellus, v. 11, p. 375-403. ------ 1960, The yearly circulation of chloride and sulfur in nature; meteorologic, geochemical and pedalogieal implications, pt. 2: Tellus, v. 12, p. 63-109. Junge, C. E., 1958, The distribution of ammonia and nitrate in rain water over the United States: Am. Geophys. Union Trans., v. 39, no. 2, p. 241-248. ------1960, Sulfur in the atmosphere: Jour. Geophys. Research, v. 65, p. 227-237. Junge, C. E., and Werby, R. T., 1958, The concentrations of chloride, sodium, potassium, calcium, and sulfate in rain water over the United States: Jour. Meteorology, v. 15, p. 417-425.Article 115 DIFFERENCES BETWEEN FIELD AND LABORATORY DETERMINATIONS OF pH, ALKALINITY, AND SPECIFIC CONDUCTANCE OF NATURAL WATER By C. E. ROBERSON, J. H. FETH; PAUL R. SEABER; and PETER ANDERSON, Menlo Park, Calif.,- Trenton, N.J.,- and Philadelphia, Pa. Work done in part in cooperation with the New Jersey Division of Water Policy and Supply Abstract.—Determinations of pH and alkalinity of ground and surface water from the Sierra Nevada, Calif., and ground water from the coastal plain of New Jersey indicate appreciable differences between laboratory and field measurements. Special caution is required for geochemical interpretations based on laboratory determinations of these properties. Scientific literature cites few if any data supporting the widely held belief that field measurements of some water-quality properties are more representative than laboratory measurements. Because of this lack of data, the TT.S. Geological Survey has made two widely separated studies, comparing field and laboratory measurements of alkalinity (as bicarbonate) and pH of water samples collected in the Sierra Nevada of California (Roberson and Feth) and in the Atlantic Coastal Plain of New Jersey (Seaber and Anderson). The New Jer-sey study also compared field and laboratory measurements of specific conductance. This article presents some of the data obtained in these studies. CALIFORNIA STUDY In the California study, field and laboratory determinations of pH and alkalinity were made of 88 samples of water. Of these samples, 73 were from springs and 15 were from streams. Most of the samples (76) were from areas of granitic rock, and the rest were from other geologic environments such as volcanic rocks, alluvium, and serpentine. Most of the samples had a low dissolved-solids content, only 3 exceeding 350 parts per million in calculated dissolved solids. Samples for laboratory determinations were stored in 350-ml pressure-sealed soft-glass (“citrate”) bottles filled virtually to capacity to minimize contact with air during transport and storage. ART. 115 IN U.S. GEOL. SURVEY PROF. C212 Field determinations of pH and alkalinity were made in a truck-mounted laboratory that provided electrostatic shielding and protection from the wind. These determinations were made immediately after the samples were collected. Field pH of samples was measured by a portable battery-operated meter (Beckman model N or model G). The meter was standardized using buffer solutions whose temperatures had been brought to that of the water to be tested. Field alkalinity of samples was determined by potentiometric titration to pH 4.5 (Rainwater and Thatcher, 1960, p. 94). Laboratory determinations of pH and alkalinity were made in a temperature-controlled room where temperatures of buffer solutions and samples were allowed to equilibrate with the ambient air temperature of the room. The time interval between field and laboratory determinations ranged from 5 to 120 days. Laboratory pH of samples was measured with a line-operated (Beckman model H) pH meter. Laboratory alkalinity of samples was determined by potentiometric titration to pH 4.5 (Rainwater and Thatcher, 1960, p. 94). Results obtained for the 88 samples studied are shown in table 115.1. The difference between field and laboratory determinations of pH (ApH) ranged from 0.0 to 2.8 pH units. The average difference is 0.3, and the standard deviation is 0.5 pH units. For alkalinity (as bicarbonate), the difference between field and laboratory determinations ranged from 0 to 26 ppm. The average difference is 3 ppm, and standard deviation is 5 ppm. In general, the changes in alkalinity between field and laboratory determinations are less pronounced than the changes in pH. This difference in the two sets of results indicates that field pH is a more critical determination than field alkalinity. PAPER 475—C, PAGES C212-C215. 1963.ROBERSON, FETH, SEABER, AND ANDERSON C213 Tabue 115.1.—Comparison of field and laboratory determinations of water samples collected in the Sierra Nevada, Calif. [O, ground water; SW, surface water] Laboratory No. Source pH Alkalinity as HCO3-1 (ppm) Time interval (days) Temperature of sample at source (°F) Calculated dissolved solids (ppm) Labo- ratory Field Labo- ratory Field 1204 G 8.1 8.1 200 202 60 58 242 1205 G 6.7 7.1 54 56 60 52 97 1207 G 7.2 7.5 64 62 60 52 106 1216 G 8.0 8.2 109 107 21 58 135 1218 G 7.4 7.5 31 30 15 57 50 1224 G 7.1 7.0 66 64 19 47 82 1225 SW 7.5 7.8 57 58 18 38 71 1226 G 6.9 8.2 42 37 13 51 73 1228 SW 6.0 8.1 3 6 18 60 9 1229 SW 6.3 7.0 4 4 18 53 11 1230 SW 6.0 8.8 3 6 18 36 8 1232 G 6.1 6.3 5 5 17 53 12 1233 G 5.8 5.9 20 19 12 47 35 1234 SW 6.9 7.0 11 11 12 41 23 1236 SW 5.8 6.1 2 2 12 40 3 1245 G 6.2 6.8 23 26 30 40 54 1246 G 7.7 7.9 93 96 29 64 98 1247 G 7.7 7.5 67 68 48 55 72 1248 G 6.8 7.4 35 38 29 52 49 1249 G 7.3 7.7 75 78 28 53 88 1250 G 7.6 7.9 76 80 28 52 87 1251 G 6.9 7.3 70 72 28 61 79 1252 G 7.4 7.4 88 89 28 69 93 1255 G 6.2 6.2 17 19 28 50 35 1256 G 6.2 6.3 19 21 27 47 32 1257 G 6.6 6.9 10 13 27 49 21 1258 G 5.9 6.2 8 8 27 59 18 1259 G 5.7 5.6 10 14 27 41 23 1261 G 7.3 7.5 83 84 60 50 93 1287 G 5.9 5.9 9 10 13 56 23 1290 G 7.0 7.2 100 103 12 70 154 1291 G 7.1 7.8 76 80 19 54 91 1292 G 6.7 6.4 148 152 11 67 203 1293 G 6.9 6.4 120 119 11 53 154 1295 - G 6.1 5.5 26 24 11 50 54 1296 G 6.2 5.5 48 46 11 52 91 1297 G 6.4 6.0 54 57 11 48 99 1304 SW 7.6 8.1 63 65 45 55 78 1309 SW 7.8 8.1 74 64 45 58 89 1324 G 8.1 7.8 59 55 90 126 2,990 1325 G 7.7 7.6 93 86 19 48 119 1326 G 7.6 7.3 81 79 19 40 177 1328 G 6.8 6.8 740 738 90 86 1,870 1330 G 7.2 7.1 67 71 16 51 84 1331 G 7.4 7.2 105 107 16 49 100 1478 SW 7.7 8.2 273 280 120 48 347 G 7.1 7.3 108 103 120 38 1481 SW 7.3 7.4 47 49 6 46 59 1482 SW 7.3 7.4 47 49 6 41 57 1500 G 5.8 5.8 11 14 6 43 28 1507 SW 6.3 6.8 5 8 20 48 11 1508 SW 6.2 6.4 6 8 20 48 13 1510 SW 7.2 7.2 21 23 20 42 34 1526 G 6.6 6.8 101 103 7 48 139 1527 G 6.7 6.7 72 75 7 60 120 1528 G 6.5 6.6 77 67 6 47 128 1529 G 7.4 7.4 84 95 6 53 127 1530 G 5.6 5.8 14 14 6 51 29 1531 G 6.8 6.9 34 32 6 62 53 1532 — G 7.0 6.9 97 100 6 51 133 1534-- G 5.4 5.8 12 14 5 52 25 1535 G 5.6 5.8 17 20 5 44 32 1539—<- — G 7.5 6.7 36 40 90 50 60 1541 G 7.5 7.9 88 90 90 50 101 1542 G 6.6 6.6 91 92 90 45 101 1543 G 6.6 6.6 73 76 90 45 90 1544 SW 7.6 7.9 80 91 90 62 96 1559 G 6.6 6.8 33 35 75 54 65 1560 G 8.0 8.2 190 216 75 56 172 1561 G 7.1 7.6 109 93 75 56 150 1564. G 7.5 8.0 143 168 60 60 149 1565. G 6.2 6.5 31 34 60 44 54 1566. G 7.7 7.6 59 64 60 50 71 1567. G 5.8 6.2 19 21 60 44 53 1568. G 7.4 7.1 68 70 60 50 82 1569. G 6.3 6.7 40 41 60 41 64 1570. G 6.3 6.5 49 52 60 46 78 1571. G 7.4 7.5 63 63 60 48 79 1572. G 7.4 7.7 103 108 60 48 106 1573. G 6.5 7.3 13 15 60 44 32 1574. G 6.8 7.2 37 38 60 44 54 1575. G 7.8 7.4 108 111 60 83 1,190 1576. G 7.6 7.4 88 81 60 44 102 1577. G 5.8 6.2 18 18 60 48 41 1578. G 7.4 7.5 146 150 60 65 335 1579. G 5.9 6.1 42 45 60 50 78 1580.- G 5.8 6.0 28 30 60 54 57 1581. G 6.8 7.1 119 113 60 70 162 Water samples having a low dissolved-solids content showed, in general, slightly greater change in pH than water having a higher mineral content. Similarly, samples having a low total dissolved carbonate-species content generally showed appreciable variation between pH determined in the field and in the laboratory. Although the time interval between field and laboratory determinations ranged from 5 to 120 days, comparison of pH with time of storage shows little correlation. However, as a group, laboratory determinations made within 1 week of the time that the samples were collected showed somewhat smaller changes than determinations made after longer storage. Some possible reasons for the differences between determination of pH in the field and in the laboratory are worth noting. Determinations made in the laboratory would be lower than those made in the field if, during transport and storage, the water sample absorbed C02 from the atmosphere. This could happen if the pressure stopper of the sample bottles leaked or if the samples stood open in the laboratory for a significant length of time before pH measurements were made. The latter possibility is ruled out because readings were made as soon as the samples were removed from the bottles, but the possibility of leakage cannot be entirely eliminated. Reactions within the water samples during storage, such as release of C02 by microorganisms, would also lower the pH. Laboratory pH values higher than those measured in the field would result from loss of C02 in water taken in the field at low temperature and stored at a higher temperature in a bottle having some air space above the water surface. Reaction between the water samples and the soft glass of the bottles might also raise the pH. The change in pH between field and laboratory analyses with the logarithm of total dissolved carbonate-species content calculated from the field pH and alkalinity measurements is shown in figure 115.1. The points scatter, but the trend suggests that exchange of C02 between water sample and atmosphere may be largely responsible for observed changes in pH. Water having a low content of carbonate species (larger negative value of the logarithm) apparently tends to gain C02 and to show a decrease in pH, whereas water having a high total carbonate-species content may lose C02 to the atmosphere and show an increase in pH. Absence of persistent trends in study graphs of other variables, however, suggests that more than one factor is involved and that two or more may interact to cause changes in pH and alkalinity in some samples. The changes observed in bicarbonate content and in pH, when field and laboratory determinations are com-C214 GEOCHEMISTRY OF WATER 0.0 Q z < 3 z UJ O li. h-< z z UJ = UJ 2 li! t “ Q ul v Z g —1-0 Si K§ u. O s3 -2.0 1 1 1 a) 1 V • • "is ! J 1 m • • • • • • • • • • • • :j ' s 1 ? • • • • • • • • • M» • • • |<5 1 o !i i % • • • • • II • i a i ® I £ l w I ra l* I l -4.0 -3.0 -2.0 log [2 CO,! species] Figure 115.1.—Graph comparing change in pH with calculated total dissolved carbonate-species content in sample. pared, indicate that laboratory determinations of these variables are not reliable for use in geochemical equilibrium calculations. The lack of correlation between pairs of variables, such as pH and temperature, pH and time of storage, and pH and HCCV1, suggests that projection of laboratory-determined data to calculated data indicative of conditions prevailing in the field cannot be made with confidence. Therefore, pH, and preferably also bicarbonate, should be determined in the field if the data are to be used in thermodynamic calculations. Field determinations of pH can be made to close limits of tolerance (±0.05) if considerable care is exercised. NEW JERSEY STUDY In the New Jersey study, field and laboratory determinations of pH, alkalinity, and specific conductance were made on water samples collected from 38 wells tapping the Englishtown Formation. At each wellhead, samples were collected in 1-gallon polyethylene bottles, which were kept capped as much as possible during the field determinations. For laboratory determinations, two citrate bottles were filled from each polyethylene bottle. One sample was used for determinations of pH, alkalinity, and specific conductance. The other sample was acidified and used for the determination of iron. Field determinations of pH, alkalinity, and specific conductance were made on all samples immediately after the samples were collected at the well site. The pH was determined by a battery-operated pH meter (Beckman Model M). Although this meter can be read to the nearest 0.02 pH unit, pH data in table 115.2 were rounded to the nearest 0.1 pH unit. Alkalinity (as bicarbonate) was determined by titrating the samples with a standard solution of sulfuric acid using phenolphthalein and methyl orange as acid-base indicators (American Public Health Association, 1960, p. 44-48). On several water samples, alkalinity was determined by both this indicator method and a modification of the potentiometric method described by Rainwater and Thatcher (1960, p. 94-95). Results obtained by the two methods agreed within 1 percent. Alkalinity was reported to the nearest part per million. Specific conductance was measured to the nearest micromho (at 25°C) by a battery-operated instrument (Industrial Instruments Solu-Bridge). All instruments were calibrated before and after fieldwork and were found to be within the accuracy of the determination as described by Hem (1959). Laboratory determinations were made by the standard methods of the U.S. Geological Survey (Rainwater and Thatcher, 1960), which also are the ones used in the California study. All laboratory determinations were made within 1 month of date of collection. The data obtained in the New Jersey study are shown in table 115.2. Except for one sample, field pH was greater than laboratory pH. The maximum difference between field and laboratory determinations of pH was 2.5. For all samples, field alkalinity was greater than laboratory alkalinity. The maximum difference between field and laboratory alkalinities was 33 ppm, but most differences (29 samples) were in the range from 5 to 15 ppm. The difference between field and laboratory determinations of alkalinity was greater than 5 percent for all but 1 of the 38 ground-water samples. According to Hem (1959, p. 97), agreement closer than 2 to 5 percent cannot be expected for duplicate alkalinity determinations. For all but five samples, field specific conductance was greater than that determined in the laboratory. The greatest differences between field and laboratory determinations of pH, alkalinity, and specific conductance were found in samples whose iron precipitated during storage of the sample. Samples having a visible precipitate of iron had average differences of 0.9 pH units, 19 ppm of alkalinity, and 28 micromhos of specific conductance. Samples having no visible precipitate of iron had average differences of 0.6 pH units, 11 ppm of alkalinity, and 17 micromhos of specific conductance. Iron precipitated in all 7 samples in which the field pH was 7.0 or less and in 3 of the 5 samples in which the field pH ranged from 7.1 to 7.3. Iron did not precipitate in any of the remaining 26 samples, which had a field pH of 7.4 or greater. Average differences be-ROBERSON, FETH, SEABER, AND ANDERSON C215 tween field and laboratory pH, alkalinity, and specific conductance for these pH ranges are shown in table 115.3. Table 115.2.—Field and laboratory determinations of physica and chemical properties of water from the Englishtown Formation in New Jersey [Chemical analyses, in parts per million] Well No. pH Alkalinity (as HCOs) Specific conductance (in micromhos at 25° C) Iron (Fe) I recipitate in sample bottle Field Lab. Field Lab. Field Lab. Lab. 1___ 4.8 3.8 3 0 148 166 11 Fellow. 2 5.0 4.4 5 0 480 446 11 ted. 3 6.6 5.2 22 3 67 37 7.8 fellow. 4 6.6 7.0 70 49 175 134 9.8 Do. 5 - 6.7 4.2 32 0 82 43 11 : led. 6 6.9 5.8 48 16 110 59 11 Do. 7 7.0 6.3 94 74 181 159 8.6 fellow. 8 7.1 6.6 72 65 145 130 3.0 21ear. 9 7.2 6.8 106 93 203 193 2.0 fellow. 10 7.3 6.7 156 132 285 270 .14 Do. 11 7.3 6.8 56 52 141 133 .36 Jlear. 12 7.3 7.0 152 129 260 242 3.4 fellow. 13 7.4 6.9 98 88 215 195 1.6 Slear. 14.... 7.4 6.9 60 49 130 104 1.6 Do. 15 7.6 7.1 104 94 181 175 .68 Do. 16 7.7 7.3 138 129 230 214 1.5 Do. 17 7.7 7.2 106 93 182 182 1.9 Do. 18 — 7.8 7.3 106 988 218 228 .62 Do. 19 7.8 7.2 108 97 184 173 .47 Do. 20 7.8 7.5 112 104 202 175 .45 Do. 21 7.9 7.3 106 99 180 178 .46 Do. 22 8.0 7.4 164 150 255 242 .65 Do. 23 8.0 7.5 132 121 235 211 2.0 Do. 24 8.0 7.4 96 91 183 171 1.7 Do. 25 8.0 7.4 99 84 158 153 .26 Do. 26 —. 8.0 7.5 160 149 270 244 .35 Do. 27.. 8.0 7.3 164 151 265 267 .45 Do. 28 8.0 7.3 106 100 167 177 .22 Do. 29 8.0 7.5 138 129 238 228 .64 Do. 30. — 8.0 7.2 114 106 185 184 .31 Do. 31 8.1 7.5 96 83 180 159 .59 Do. 32 8.1 7.2 126 116 219 198 .24 Do. 33 8.1 7.6 118 111 208 193 .30 Do. 34 8.2 7.7 122 120 208 191 .20 Do. 35 8.3 7.9 1 149 140 235 214 1.0 Do. 36. 8.3 7.9 2 184 166 244 271 .24 Do. 8.5 7.5 3 246 218 352 351 Do. 38 8.5 8.0 4 251 218 348 233 .19 Do. 1 Includes equivalent of 2 ppm of carbonate (CCh). 2 Includes equivalent of 6 ppm of carbonate (COs). a Includes equivalent of 10 ppm of carbonate (CO3). 4 Includes equivalent of 8 ppm of carbonate (COs). Table 115.3.—Average differences, ignoring sign, between field and laboratory determinations of pH, alkalinity, and specific conductance of water samples collected in the Englishtown Formation of Nev) Jersey Range in field pH No. of samples Average differences between field and laboratory determinations pH Alkalinity (as HCOs) Specific conductance (in micromhos at 25°C) 4.8-7.0 7 1. 1 19 34 7.1-7.3 5 . 5 14 13 7.4-8.5 26 . 6 11 17 4.8-8.5 38 . 7 13 19 The differences between field and laboratory data possibly can be explained by the precipitation of iron compounds, changes in temperature of the samples, and changes in chemical equilibrium between the time of field analysis and laboratory analysis (Rainwater and Thatcher, 1960, p. 31-32). Because precipitations of iron compounds and changes in temperatures would result in a decrease in the bicarbonate concentration, the pH and the alkalinity concentration would be expected to decrease; possibly the specific conductance would decrease also. The following observations are based on a comparison of the analytical data for water samples from 38 wells tapping the Englishtown Formation in New Jersey: 1. The field determinations of pH, alkalinity, and specific conductance generally are higher than laboratory determinations. 2. Samples in which iron precipitated between the time of the field determinations and the laboratory determinations showed the greatest differences between the field and laboratory values of pH, alkalinity, and specific conductance. 3. Iron precipitates appeared in all samples having a field pH of 7.0 or less, in some samples having a field pH of 7.1 and 7.3, and in no samples having a field pH of 7.4 or more. 4. The highest concentrations of iron and the greatest differences in pH, alkalinity, and specific conductance characterized samples collected in areas where little or no calcium carbonate is present in the natural aquifer materials. Because field determinations of pH, alkalinity, and specific conductance probably are more representative of the actual state of the water in the Englishtown Formation, they should be preferred for geochemical interpretations of this aquifer system. On the other hand, if water is being rated as to its suitability for various uses after storage, laboratory determinations probably are more useful. CONCLUSIONS In summary, both studies showed that field determinations of pH and alkalinity are generally higher than laboratory determinations of these variables, and that field determinations of pH and alkalinity are more representative of water in its natural environment than are laboratory determinations. REFERENCES American Public Health Association, 1960, Standard methods for the examination of water and waste water: New York, Am. Public Health Assoc., Inc., 11th ed., 626 p. Hem, J. D., 1959, Study and interpretation of the chemical characteristics of natural water: U.S. Geol. Survey Water-Supply Paper 1473,269 p. Rainwater, F. H., and Thatcher, L. L., 1960, Methods for collection and analysis of water samples: U.S. Geol. Survey Water-Supply Paper 1454, 301 p.Article 116 INCREASED OXIDATION RATE OF MANGANESE IONS IN CONTACT WITH FELDSPAR GRAINS By JOHN D. HEM, Denver, Colo. Abstract.—Solutions of Mn4! in contact with feldspathie sand lose manganese by ion-exchange adsorption on the feldspar grains and by oxidation followed by precipitation of manganese oxide. Removal of manganese is substantially faster in solutions at pH 8 than at pH 7. Laboratory studies reported here indicate that feld-spathic sand moderately catalyzes the oxidation of Mn+2 in water and that pH strongly influences the rate of oxidation. In these studies the effect of feldspathie sand on the rate of loss of Mn+2 from solution was studied by means of two series of batch-type experiments. In the first series of experiments, 500-ml volumes of aqueous solution containing 10 ppm of Mn+2 and adjusted to pH 7.70 with sodium hydroxide were placed in each of 6 polyethylene bottles. Five of these bottles contained feldspathie sand, in amounts ranging from 10 to 100 g, that had been pretreated with sodium chloride solution and washed so that ion-exchange sites were occupied by sodium ions; the sixth contained no sand. The sand used in these studies consisted of rounded grains in the range 0.10 to 0.80 mm. Microscopic examination of a representative sample of the sand showed that it consisted of approximately 45 percent quartz, 21 percent orthoclase feldspar, and 34 percent plagioclase feldspar. The cation-exchange capacity of the sand, determined by equilibration with 1.0 molar solution of manganese chloride, was 0.66 milli-equivalents per hundred grams. The solutions were kept in the laboratory at a temperature of 25°±1°C. Periodic sampling and analysis of the solutions indicated that both the pH and the manganese concentration of the test solutions decreased with time. A second series of experiments was made with solutions at a higher and more constant pH. The six solutions of this series differed from those of the first series only in that they contained 100 parts per million of HC03_1 added as sodium bicarbonate. The pH of these solutions was adjusted to 8.03 and remained near 8.0 throughout the experiments. In the second series, as in the first, the solutions were sampled and analyzed periodically. Results of some of the experiments are shown in figure 116.1. Test solutions that were not exposed to feldspathie sand showed little change in concentration of Mnt2 at either pH near 7 or near 8, even after more than 350 hours. At pH near 8.0, the oxidation rates of Mn+2 were too slow to be reliably measured. Test solutions that were in contact with feldspathie sand Figure 116.1.—Rate of loss of manganese from solutions near pH 7 (dashed lines) and pH 8 (solid lines) in presence and absence of feldspathie sand. C216 ART 116 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C216-C217. 1963.HEM C217 showed a rapid initial decrease in Mn+2 concentration, followed by a slower, steady decrease. The decrease in Mn+2 concentration was approximately proportional to the weight of sand in the test bottles and was greater at pH near 8 (solid line, fig. 116.1) than at pH near 7 (dashed line). The results of these studies are most readily explained by the hypothesis that some Mn+2 is adsorbed on the sand and some is oxidized to form a precipitate of manganese oxide. Manganese is removed rapidly from solution at first, mostly by exchange for sodium on the sand. As the solution and the exchange surfaces approach equilibrium, the effect of ion exchange disappears and the manganese-loss curve (fig. 116.1) becomes a straight line as the manganese is removed only by oxidation catalyzed by the sand-grain surfaces. Several lines of evidence support this hypothesis. For example, when other conditions were comparable, the slopes of the straight segments of the curves were steeper for solutions in contact with larger amounts of sand. Also, the slopes were steepened by increasing the pH and were flattened by adding bicarbonate and sulfate ions, as observed in earlier work with manganese oxidation rates (Hem, 1963). The catalysis mechanism was not closely studied, but such an effect of solid surfaces on reaction rates is not particularly unusual. In order to compare the behavior of Mn+2 with a similar ion that is not subject to oxidation, a series of experiments was run using calcium solutions in place of manganese solutions. The ionic radius of Mn+2 is slightly smaller than that of Ca+2, and manganese is a little more strongly adsorbed by the sand than calcium. The calcium experiments showed an initial period of rapid loss, but after 4 days the solutions stabilized and no further loss occurred, regardless of the amount of sand present. Natural mineral surfaces probably catalyze the formation and precipitation of manganese oxide from weathering solutions. The effect helps to explain the chemical behavior of manganese in natural water. REFERENCE Hem, J. D., 1963, Chemical equilibria and rates of manganese oxidation: U.S. Geol. Survey Water-Supply Paper 1667-A. Article 117 SOLUTION OF MANGANESE DIOXIDE BY TANNIC ACID By JACK RAWSON, Austin, Tex. Abstract.—Although manganese dioxide is nearly insoluble in distilled water, tannic (digallic) acid may bring considerable amounts of manganese into solution. The study suggests a possible mechanism for the solution of insoluble manganese from soils and sediments by natural waters containing tannic acid or similar organic extracts. Manganese is usually present in soils and sediments in an oxidized, insoluble state. Most of the manganese found in natural waters probably results from the solution of manganese from soils and sediments aided by organic matter and’bacterial action (Hem, 1959, p. 66). Manganese in excess of 0.3 parts per million is objectionable in public water supplies and may require removal for some industrial processes. Manganese dioxide is nearly insoluble in distilled water but is moderately soluble in tannic acid, which is common in organic soils and sediments. The action of aqueous solutions of tannic acid on manganese dioxide was studied with a view to a more thorough understanding and explanation of phenomena that occur when natural waters containing organic extracts come into contact with soils and sediments containing insoluble manganese. Tannins are a large group of complex substances found in many plants. Gallotannin, the common tannic acid of commerce, is obtained principally from gallnuts (Conant and Blatt, 1947, p. 506). This particular tannin is a mixture of gallic acid esters of glucose. According to Hem (1960, p. 77) the gallotannin molecule, in water solution, hydrolyzes to give glucose and digallic acid. Digallic acid has the following semistruc-tural formula: (HO) 3C6H2 • C02 • C„H2 (OH) 2C02H The tannic acid used in the experiments reported here is represented by the above formula. Solutions containing 10,100, and 1,000 ppm of tannic acid were prepared, and each solution was divided into six aliquots. The pH of each aliquot was adjusted with carbonate-free potassium hydroxide to a predetermined ART. 117 IN U.S. GEOL. SURVEY PROF. C218 value ranging in unit steps from 4.5 to 9.5. The aliquots were poured into polyethylene bottles, and 1 g of reagent-grade manganese dioxide was added to each. The bottles were tightly stoppered and were placed in a constant-temperature room at about 25 °C. Periodically for 56 days, the Eh (redox potential) and pH of samples from each bottle were determined. The samples were then passed through plastic-membrane filters (average pore diameter of 0.45 micron) to remove precipitated or colloidal manganese. The concentration of the manganese in the filtrates was then determined by the permanganate method as described by Rainwater and Thatcher (1960, p. 205-207). After 56 days of storage, the 100-ppm tannic acid solutions were separated from the solid manganese dioxide and were filtered through plastic-membrane filters. To determine whether or not collodial manganese passed through the filters, portions of these filtrates were centrifuged at 3,000 revolutions per minute for 30 minutes and then refiltered. No appreciable loss of manganese from the filtrates occurred. Though not conclusive, the retention of manganese in a filterable state suggests that the manganese was not colloidal. The remaining portions of the filtrates from the 100-ppm tannic acid solutions were stored in tightly stoppered plastic bottles to determine whether or not the manganese would be retained in solution in the absence of excess solid manganese dioxide. After 111 days of storage, no significant loss of manganese occurred from the solutions. The final portions of the 100-ppm tannic acid solutions were aerated by passing air, presaturated with water, through them for 10 hours. Two weeks later, the pH of the samples was raised to 9.0 with carbonate-free potassium hydroxide. The samples were filtered and the concentration of manganese in the filtrates was determined. The results indicated that the manganese content had decreased only slightly. This retention of manganese in solution indicates a manganese-tannic acid complex that strongly resists oxidation. PAPER 475-C, PAGES C218-C219. 1963.RAWSON C219 The table summarizes the effect of tannic acid on the solution of manganese dioxide. The results indicate that tannic acid, even in dilute solutions, may bring considerable amounts of manganese into solution. Figure 117.1 shows the relation between storage time, pH, and manganese concentration in test solutions containing tannic acid. The results indicate that the rate of solution of manganese is a function of both the initial pH and the tannic acid concentration of the test samples. In test solutions having the same initial pH, the rate of manganese solution increases with the tannic acid concentration. In test solutions having the same tannic acid concentration, however, the reaction rate is greater in those samples having the lower initial pH. The exact mechanism of manganese solution by tannic acid was not determined in this study. The permanganate method, though quite sensitive and specific for manganese, does not permit distinction of the forms in which manganese was originally present in the solu-tons. Manganese dioxide, however, is a strong oxidizing agent. Tannic acid, on the other hand, is a reducing agent. In his study of complexes of iron with tannic acid, Hem (1960, p. 84-85) demonstrated that the reduction of ferric iron to ferrous iron in acid solutions was effected by tannic acid. Manganese is chemically related to iron. The solution of manganese by tannic acid, therefore, probably results from chemical reduction. If this assumption is correct, complexing of tannic acid with divalent manganese also occurs. The retention of manganese in solution in the 100-ppm tannic acid solutions supports this argument. The experiments suggest a mechanism for manganese solution from soils and sediments by natural waters containing tannic acid or similar organic extracts leached from decaying organic debris. Figure 117.7—Solution of manganese by tannic acid. Effect of tannic acid concentration on solution of manganese dioxide [Eh in volts, Mn concentrations in parts per million] 10 ppm 100 ppm 1,000 ppm Initial pH After 3 hours After 66 days After 3 hours Alter 56 days After 3 hours After 56 days pH Eh pH Eh Mn pH Eh pH Eh Mn pH Eh pH Eh Mn 4.5 5. 3 0. 577 6. 2 0. 598 3. 6 5. 6 0. 539 6. 9 0. 565 21 4. 9 0. 547 6. 4 0. 512 112 5.5 6. 2 . 527 6. 6 . 586 3. 5 6. 0 . 521 6. 8 . 568 14 5. 9 . 478 6. 2 . 515 90 6.5 6. 6 . 492 6. 6 . 586 2. 4 6. 5 . 490 6. 9 . 563 13 6. 5 . 477 6. 4 . 537 80 7.5 7. 0 . 471 6. 6 . 579 1. 8 7. 0 . 490 6. 9 . 578 10 7. 1 . 477 6. 4 . 547 54 8.5 7. 3 . 465 6. 6 . 579 2. 0 7. 4 . 483 6. 9 . 576 5. 2 8. 3 . 428 6. 6 . 545 21 9.5 8. 7 . 389 6. 7 . 562 . 79 8. 6 . 407 7. 0 . 571 5. 2 9. 2 . 357 6. 8 . 531 15 REFERENCES Conant, J. B., and Blatt, A. H., 1947, The chemistry of organic compounds: New York, Macmillan Co., 640 p. Hem, J. D., 1959, Study and interpretation of the chemical characteristics of natural water: U.S. Geol. Survey Water-Supply Paper 1473,254 p. Hem, J. D., 1960, Complexes of ferrous iron with tannic acid: U.S. Geol. Survey Water-Supply Paper 1459-D, p. 75-94. Rainwater, F. H., and Thatcher, L. L., 1960, Methods for collection and analysis of water samples: U.S. Geol. Survey Water-Supply Paper 1454, 297 p. 5* 694-027 O—6i ■15Article 118 EFFECTIVENESS OF COMMON AQUATIC ORGANISMS IN REMOVAL OF DISSOLVED LEAD FROM TAP WATER By EUGENE T. OBORN, Denver, Colo. Abstract.—Lead sorbed by four kinds of aquatic plants is approximately in proportion to the area of the plant-body surface in contact with the water. Symbiotic bacteria were the most active of the organisms studied. Small amounts of lead taken in drinking water are not easily eliminated by the body, but accumulate, instead, until irreparable body damage has been done (Offner, 1944, p. 284). In a study of this danger the author investigated the effectiveness of aquatic vegetation and associated microbiological symbionts in removing lead and other dissolved ions from tap water. Experiments for this study were made during the summer of 1961. Limnologists generally agree that soil-rooted aquatic plants, although wholly submerged, obtain their mineral salts from the substratum and not from the surrounding water. Floating plants that are not rooted in the soil, however, obtain their mineral salts directly from the surrounding water (Butcher, 1933). The plants used in the experiments reported here were all of the water-rooted variety. Submerged horn wort (C eratophyllum demersum L.) has an abbreviated root system. Common bladderwort (Utricvlaria vulgaris L.) likewise has an abbreviated root system, but additional sorption may take place through the many bladders present. With common pond scum (Cladophora glomerata [L.] Kiitz) direct sorption is possible over most of the plant surface except for the part floating above water. Rust-colored bacterial water slime used in this study was a very mucoid symbiotic growth on which sorption could take place through any part of the completely submersed irregular mucous envelope. The bacteria1 consisted of 53 percent gram-negative rods, unable to reduce nitrates to nitrites, but capable of growing both at room temperature and at 37°F. The remaining 47 percent were gram-positive cocci, able to reduce nitrates to nitrites and capable of growing at 37°F but not at room temperature. The plants (each 1.0-g blotter-dry weight) were placed in 250-ml Erlenmeyer flasks containing 200 ml of tap water. The water contained 10 to 20 micrograms of lead tagged with 0.1 microcurie of carrier-free Pb210 (equivalent to 1.2X 10~3 /xg of radioactive lead) as lead nitrate. The flask tops were covered with cellophane, secured with rubber bands, and the flasks were allowed to stand undisturbed in natural light at the ambient temperature of the laboratory for 2 weeks. The plants were then removed from the water, and the radioactive lead content of plants and water was determined. The accompanying table, which lists the plants in increasing order of surface area per unit weight, shows the relative effectiveness of the bacteria and the three species of aquatic plants in removing radioactive lead from tap water. (Determinations by Division of Industrial Medicine, School of Medicine, University of Colorado, with the cooperation of Dr. R. F. Bell.) Radioactive lead (percent) In or on In the Plant the plant water Hornwort_____________________________________ 41 59 Bladderwort__________________________________ 58 42 Pond scum____________________________________ 79 21 Bacterial slime______________________________ 87 13 Evidently, removal of radioactive lead by the plants was approximately proport ional to the amount of body sorptive surface of the aquatic plants in contact with the surrounding water. REFERENCES Butcher, R. W., 1933, Studies on the ecology of rivers, pt. 1, On the distribution of macrophytic vegetation in the rivers of Britain: Jour. Ecology, v. 21, p. 58-91. Offner, M. M., 1944, Fundamentals of chemistry: New York, Barnes and Noble, Inc., 408 p. 11 Determined by Bacteriology Laboratory, School of Medicine, University of Colorado. ART. 118 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGE C220. 1963. C220Article 119 ADSORPTION OF THE SURFACTANT ABS35 ON ILLITE By C. H. WAYMAN, H. G. PAGE, and J. B. ROBERTSON, Denver, Colo Work done in cooperation with the Federal Housing Administration Abstract.—A radiochemical-tracer technique was used to study the adsorption of the surfactant ABS35 on illite. The length of the alkyl chain, presence of phosphate ion, pH, and the amount and type of ionic salt present influence this adsorption. Illite and other clay minerals are inefficient adsorbents for ABS in comparison with synthetic materials (colloidal alumina and activated charcoal). Incompletely degradable surfactants injected into the ground in household waste water have in many areas polluted ground-water supplies. One of the natural means by which these surfactants are removed from waste water is by their adsorption on soil minerals. The ability of illite to adsorb alkylbenzenesulfonate (ABS), an essential incompletely degradable surfactant in detergents, is described in this article. As in similar studies of the adsorption of ABS on kaolinite (Way-man, Robertson, and Page, 1963) and on montmoril-lonite (Wayman, Page, and Robertson, 1963) the ABS was tagged with S35 (ABS35). The adsorption of ABS on clay minerals is also compared with that on synthetic materials (activated charcoal and colloidal alumina). A radiochemical-tracer technique identical to that described for the kaolinite study was used in this study. The effects of such variables as alkyl-chain length (Ci2 and Ci5), pH, phosphate ion, and type and amount of dissolved ionic salts (NaCI, CaCl2, and A1C13) on ABS adsorption by illite were studied. ABS was added as the sodium salt and can be represented in solution as the anion (A-CeH.-SOs)-1, where R represents the alkyl chain. The preparation of the illite and the experimental procedures were identical with those used in the study of kaolinite. The illite used was the standard clay mineral (No. 35) of the American Petroleum Institute. The composition of the mineral can be represented in a broad sense (Warshaw, 1958, p. 304) by [Ko.5_o.7 (H3O ) O-O.3Jo.5-l.otAl, 7_2.rAlg0_0.3j2.0 [Si, . 3-3.6 Alo. 4-0.7 b. oD]0 (OH )2 or in the form (Grim and others, 1937, p. 823) (OH)4Ky(Al4Fe4Mg4Mg6)(Sis_yAly)OI0. Microscopic and X-ray diffraction examination indicated that the clay contained about 12 percent impurities in the form of sericite, quartz, plagioclase, calcite, pyrite, and a trace of carbonaceous material. The cation-exchange capacity of the clay is 25.0 milliequiva-lents per 100 g, and the pH of a suspension of 200 mg of the clay in 35 ml of distilled water is about 6.8. Figure 119.1 shows that both pH and length of the alkyl chain influence adsorption in solutions containing 5.0 parts per million of ABS. Dodecyl (CJ2) ABS is only slightly adsorbed (maximum of 50 /tg of ABS per g of clay) even at pH 4. Pentadecyl (Ci5) ABS is adsorbed most readily at pH 4, but desorption seems to occur with time. Significant amounts of negative adsorption are indicated for both dodecyl and pentadecyl ABS at pH 10; this effect can be explained by repulsion of anionic ABS by the potential-determining hydroxyl ions (OH)'1 on the clay surface at pH 10 and by disparities in the electrical double layer at the clay-solution interface. The increased adsorption of both types of ABS at low pH can be explained by significant numbers of positive sites on clay. In addition, pentadecyl ABS exhibits a greater adsorption than dodecyl ABS because the longer chain material (Ci5) has a higher free energy of adsorption. Detailed supporting data for these effects have been given previously (Wayman, Page, and Robertson, 1963). ART. 119 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C221-C223. 1963. C221ABS ADSORBED PER GRAM OF ILLITE, IN MICROGRAMS EXPERIMENTAL HYDROLOGY C222 Figure 119.1—Influence of pH, length of alkyl chain, and time on adsorption of ABS by illite. Test solutions contained 5 ppm of pentadecyl ABS and 5 ppm of dodecyl ABS. Figure 119.2 shows the results of adsorption of ABS on illite when 10 ppm of phosphate and either 5 ppm of dodecyl or 5 ppm of pentadecyl ABS are in solution. Maximum adsorption of ABS takes place for both forms of ABS at pH 4; apparent equilibrium adsorption is attained in about 6 days. For dodecyl ABS, adsorption at pH 7 is similar to that at pH 10. For pentadecyl ABS, adsorption-desorption are indicated. However, after lengthy periods of equilibration, the amount of either type of ABS adsorbed at a specified pH seems to approach the same value. In acid solutions (pH 4), phosphate seems to increase the amount of ABS adsorbed by illite. The effect of enhanced adsorption with phosphate in solutions might be attributed to adsorption of ABS in the form of an aluminum-phosphate-ABS complex; reasons for such an adsorption mechanism have been given elsewhere (Wayman, Page, and Robertson, 1963). Figure 119.3 shows the influence of salt concentration, valence, and pH on the adsorption of ABS by illite. The data shown represent apparent equilibrium adsorption measured for as long as 10 days. At any pH the amount of ABS adsorbed increases with increase in salt concentration. The optimum conditions for removal are at high salt concentrations and in acid solutions (pH 4). Trivalent (A1C13) and divalent (CaCl2) salts influence ABS adsorption to a greater extent than NaCl, except in alkaline solutions (pH 10). These adsorption phenomena can most readily be explained as being due to the combined effects of availability of positive sites on clay at low pH, lowering of the critical micelle concentration of surfactants by ionic salts, relation of valency to action of cations as described by the Schulze-Hardy rule, and adsorption by an aluminum-hydroxy-ABS complex. Details of these considerations have been described previously (Wayman, Page, and Robertson, 1963). Figure 119.4 shows the relative adsorption capacity of kaolinite, illite, and montmorillonite in comparison with synthetic adsorbents (colloidal alumina and activated charcoal). Colloidal alumina is positively charged in acid solution. It consists of fibrils of boehmite (AlOOH) approximately 5m/i in diameter and 100 to 150m/x in length, a surface area of 274 m2 per g, a pore diameter of 77 A, and a pore volume of 0.53 cc per g. Activated carbon is quite variable in physical dimensions. Particle diameters range from less than 100 A to 5,000 A; average open-pore volumes range from about 0.1 to 0.3 cc per g, total surface areas are Figure 119.2.—Influence of pH, phosphate, and time on ABS adsorption by illite. Test solutions contained 5 ppm of dodecyl or pentadecyl ABS and 10 ppm of phosphate. MOLAR CONCENTRATION OF AICI3 Figure 119.3.—Influence of salt concentration, valence, and pH on adsorption of ABS by illite. Test solutions contained 5 ppm of dodecyl ABS.WAYMAN, PAGE AND ROBERTSON C223 as much as 1,000 m2 per g, and average pore diameters are about 20 A. The experimental method for the synthetic adsorbents was identical with that used for the clays. Figure 119.4 shows that, compared with the synthetic adsorbents, natural clays are inefficient adsorbents for ABS. These disparities in adsorption of ABS can be attributed to the exclusively positive surface and large pore diameter of colloidal alumina and to the high external and internal surface area of activated carbon. REFERENCES Grim, R. E., Bray, R. M., and Bradley, W. F., 1937, The mica in argillaceous sediments: Am. Mineralogist, v. 22, p. 813-829. Warshaw, 0. M., 1958, Experimental studies of illite, v. 7 of Clays and clay minerals: London, Pergamon Press, p. 303-316. Wayman, C. H., Robertson, J. B., and Page, H. G., 1963, Adsorption of the surfactant ABS35 on kaolinite: Art. 238 in U.S. Geol. Survey Prof. Paper 450-E, p. E181-E183. Wayman, C. H., Page, H. G., and Robertson, J. B., 1963, Adsorption of the surfactant ABS 35 on montmorillonite: Art. 59 in U.S. Geol. Survey Prof. Paper 475-B, p. B213-B216. Figure 119.4—Adsorption of ABS by clay minerals as compared to synthetic adsorbents.Article 120 BIODEGRADATION OF SURFACTANTS IN SYNTHETIC DETERGENTS UNDER AEROBIC AND ANAEROBIC CONDITIONS AT IO C By C. H. WAYMAN and J. B. ROBERTSON, Denver, Colo. Work done in cooperation with the Federal Housing Administration Abstract.—Studies of bacterial degradation of ABS at 10°C under aerobic and anaerobic conditions indicate significant reduction of ABS in samples containing 1,450,000 bacteria per ml. Straight-chain ABS is degraded to a greater extent than branched-chain ABS, and the aerobic environment is more effective than the anaerobic in ABS reduction. Since World War II, more than 500,000 new synthetic chemicals have been placed at public disposal, and this amount is increased by about 10,000 per year. Many of these products are discharged directly into surface water or into the ground. Unless these compounds can be chemically degraded (by dissolution or hydrolysis) or oxidized by aerobic bacteria, pollution of ground water or surface water may result. Anionic surfactants in synthetic detergents resist biochemical degradation. Thus, when sewage effluents containing these refractory substances are dumped into rivers or streams, pollution of ground water may occur from natural infiltration from the streams. Likewise, septic-tank effluents containing surfactants may pollute ground water unless the surfactants are adsorbed by soils during infiltration. In commercial detergents, alkylbenzenesulfonate (ABS) is the anionic surfactant that is responsible for pollution problems. The alkyl chain in this compound probably contains 12 to 15 carbon atoms and apparently is highly branched (Continental Oil Co., 1955), as indicated by the following structure: CH3—CHi—CHs CHi -CH-CEj- CHa CH3 -CH—CHs—C—CHs A NaSOa (1) Other forms of ABS branched structures may be represented as follows: CHs CHs CHs—CHs—CHi—CHs—A—CHs—CHa—C---- Ahs CHs Y NaSOa CHa CHa CHa ^C—CHs—i—CHs—CH*— A—/\ Ahs Ahs Ahs NaSOs (2) (3) Because ABS is not completely degraded by bacteria in water, pollution of surface water and ground water by ABS has been reported from many localities throughout the Nation (Lauman, 1959; Nichols and Koepp, 1961; Walton, 1960). McKinney and Symons (1956) indicated that common soil or water bacteria, such as Alcaligenes, Pseudomonas, Aerobacter, Escherichia;, and Flavobacterium are capable of surviving and growing in ABS agar when ABS is the only source of carbon for metabolism. Wayman and others (1963) have shown also that concentrations of ABS normally found in sewage or surface water are not toxic to bacteria, especially Escherichia coli. These data suggest that bacteria can metabolize ABS, but not completely. The inability of bacteria to degrade ABS completely seems related to the structure of ABS. For activated-sludge systems in sewage plants, McKinney and Symons (1959) point out that ABS metabolism starts with the terminal methyl group (CH3) on the hydrocarbon side chain. The terminal methyl group is oxidized by enzymes to a carboxyl group. Then the hydrocarbon side C224 ART. 120 IN U.S. GEOL. SURVEY PROF. PAPER 475-C, PAGES C224-C227. 1963.WAYMAN AND ROBERTSON C225 chain is metabolized until it reaches the quaternary carbon. The quaternary carbon is probably located at the point where the hydrocarbon side chain joins the benezene ring. This oxidation scheme would apply to the type 1 ABS structure shown above. In the normal mechanism of oxidation, two hydrogen atoms are removed at a time and oxidized to water. When the oxidation reaches the quaternary carbon, only one hydrogen is available. This effect blocks ABS metabolism and terminates the oxidation. Hence, the complex branching of ABS prevents complete degradation by bacteria. Other investigators (Hammerton, 1955; and Truesdale, 1962) also attribute incomplete degradation of ABS to the branched structure. To remedy or to ameliorate this effect of incomplete degradation of ABS, the development of ABS with a straight-chain configuration as shown below has been suggested: CHa—(CHj)b—CH Ah» V CHa CHa—(CHa) 8—i /\ dma \/ -NaSOa —NaSOa (4) (6) The ABS structures represented by types 1 to 3 are referred to as “hard” substances, whereas those for 4 and 5 are called “soft.” Studies of ABS degradation by activated sludge indicate that about 67 percent of the “hard” type of ABS and 94 percent of the “soft” type of ABS in solution can be degraded and thus be removed from solution. In surface water or in water infiltrating the ground, degradation of ABS can be achieved only by bacterial activity. For surface water, the process involves an aerated system, and for infiltrating water, especially from septic tanks, an anaerobic system. Because previous studies on ABS degradation by activated-sludge treatment cannot be applied directly to surface water or to ground water, the authors studied the effects of aerobic and anaerobic systems on ABS in the temperature range 10° to 35 °C. This article summarizes experimental results for aerobic and anaerobic systems at 10°C. The source of bacteria utilized in the study consisted of the effluent from the Denver sewage treatment plant. The composition of selected constituents of this effluent over a 1-year period indicated the following ranges: bacteria, 100,000 to 4,000,000 per ml; detergent (ABS), 3 to 8 parts per million; dissolved solids, 450 to 1,000 ppm; dissolved oxygen, 0 to 5 ppm; biological oxygen demand (BOD), 125 to 250 ppm; and pH, 6.2 to 8.1. Solutions were prepared by either no dilution or dilution of the sewage effluent with distilled water to give starting concentrations of total or coliform bacteria that ranged from 6,000 to 1,500,000 per ml. Then three types of ABS were added to various solutions, giving starting concentrations of as much as 13 ppm: sample 1, branched-chain ABS; sample 2, domestic straight-chain ABS; and sample 3, German straight-chain ABS. All samples were prepared in 1-liter volumetric containers and reduced to 500 ml for testing purposes. Samples were placed in a constant-temperature bath (10° ± 0.1°C). Periodically, samples were removed from solution and analyzed for ABS, total bacteria, and coliform bacteria. ABS was determined by the methylene-blue method. Total bacteria and coliform bacteria were determined by the pour-plate technique using Nutrient and MacConkey agars, respectively, at an incubation temperature of 37° for 72 hours. Figure 120.1 shows the rate of degradation of both branched-chain ABS and domestic straight-chain ABS under anaerobic conditions at 10°C. The starting concentrations of ABS and bacteria studied were: Sample 1: Branched-chain ABS______________ 13 ppm. Coliform bacteria________________ 2,000 per ml. Total bacteria___________________ 6,000 per ml. Sample 2: Domestic straight-chain ABS-----12 ppm. Coliform bacteria________________ 2,000 per ml. Total bacteria___________________ 6,000 per ml. The data indicate that about 6 days is required for bacteria to become acclimated to the environment. After 6 days, bacteria thrive more readily on straight-chain ABS than on branched-chain ABS. The bacteria were unable to metabolize significant amounts of either form of ABS even after 50 days. Bacteria probably exist for long periods of time on other constituents in the diluted sewage. The dying off of bacteria with time might be attributed to the toxic effect of constituents concentrated in solution from bacterial metabolism. Figure 120.2 shows the effect of degradation of ABS under the conditions indicated in figure 120.1, but the starting concentrations were 15,000 per ml and 60,000 per ml for coliform and total bacteria, respectively. No significant degradation of ABS was observed; with increased concentration of bacteria, the period for acclimation to the environment increases. Both figures 120.1 and 120.2 indicate that in moderate anaerobic environments coliform bacteria die off more rapidly when branched-chain ABS is in solution. Thus, even in anaerobic environments coliform bacteria may persist for significant periods of time through metabolism of sewage constituents other than ABS.C226 EXPERIMENTAL HYDROLOGY © Branched-chain ABS Total bacteria □ Straight-chain ABS Coliform bacteria Figure 120.1.—Rate of degradation of ABS under anaerobic conditions at 10° C for initial concentrations of coliform bacteria (2,000 per ml) and total bacteria (6,000 per ml). © Branched-chain ABS Total bacteria □ Straight-chain ABS ^ ^ Coliform bacteria Figure 120.3 shows the results of experiments on undiluted sewage effluent for starting concentrations of 1,450,000 total bacteria per ml. These solutions had an original content of branched-chain ABS of 4 ppm in a somewhat degraded condition. Additional branched-chain or straight-chain ABS was added to raise the initial ABS concentrations to between 10 and 11 ppm. Data are given for both anaerobic and aerobic systems. Aerobic conditions were maintained by bubbling filtered air through solutions at a constant flow rate. Figure 120.3 shows that the time required for equilibrium degradation of ABS in both aerobic and anaerobic systems is more than 10 days. In anaerobic environments, about 50 percent of both branched-chain and straight-chain types of ABS added to the undiluted sewage waste can be degraded. There seems to be little difference between degradation of the German straight-chain ABS in aerobic as compared to anaerobic environments; significant amounts of the German EXPLANATION © Branched-chain ABS A No added ABS O Domestic straight-chain ABS Anaerobic conditions O German straight-chain ABS /Aerobic conditions Figure 120.2.—Rate of degradation of ABS under anaerobic Figure 120.3.—Rate of degradation of ABS under anaerobic conditions at 10°C for initial concentrations of coliform bac- and aerobic conditions at 10°C for initial total bacterial con- teria (15,000 per ml) and total bacteria (60,000 per ml). centrations of 1,450,000 per ml.WAYMAN AND ROBERTSON C227 straight-chain ABS can be degraded in both environments. After about 19 days in an aerobic environment, as much as 75 percent of the domestic straight-chain ABS added to the undiluted sewage effluent can be degraded. The significant and rapid growth curves of bacteria might be attributed to combined effects of metabolizing both ABS and other nutrients in the sewage solutions. Results for coliform bacteria can be explained on a basis identical to that for figure 120.2. This study suggests the following conclusions: 1. For both aerobic and anaerobic environments, at 10°C and having initial total bacterial counts of 1,450,000 per ml, the time required for equilibrium degradation of ABS is more than 10 days. 2. In anaerobic environments described in (1), about 50 percent of both branched-chain and straight-chain types of ABS added to undiluted sewage waste can be degraded under conditions of equilibrium. 3. In aerobic environments described in (1), about 40 percent of the branched-chain ABS, 75 percent of the domestic straight-chain ABS, and 50 percent of a German straight-chain ABS can be degraded under conditions of equilibrium. 4. As much as 50 percent degradation of domestic straight chain occurs after 2 days in the aerobic environment described in (1). This may be compared with degradations of 25 percent or less for the other ABS types in both anaerobic and aerobic environments for 2 days. REFERENCES Continental Oil Co., 1955, Structure of alkyl benzene sulfonates: Continental Oil Co. Central Research Labs, Rept. 139-55-503. Hammerton, C., 1955, Observations on the decay of synthetic anionic detergents in natural waters: Jour. Appl. Chemistry, (London), v. 5, p. 517-524. Lauman, H. E., 1959, Study of synthetic detergents in ground water, Suffolk County Area, Long Island: Prepared for the Long Island Home Builders Inst., Inc., 9 p. McKinney, R. E., and Symons, J. M., 1959, Bacterial degradation of ABS, I. Fundamental biochemistry: Sewage and Indus. Wastes Jour., v. 31, p. 549-556. Nichols, M. S., and Koepp, E., 1961, Synthetic detergents as a criterion of Wisconsin ground water pollution: Am. Water Works Assoc. Jour., v. 53, p. 303-306. Truesdale, G.A., 1962, Pollution by synthetic detergents: Chem. Products, Jan. 1962, p. 1-8. Walton, Graham, 1960, ABS contamination: Am. Water Works Assoc. Jour., v. 52, p. 1354-1362. Wayman, C. H., Robertson, J. B., and Page, PI. G., 1963, Factors influencing the survival of Escherichia coli in detergent solutions : Art. 57 in U.S. Geol. Survey Prof. Paper 475-B, p. B205-B208.Article 121 DIRECT MEASUREMENT OF SHEAR IN OPEN-CHANNEL FLOW By JACOB DA VIDIAN and D. I. CAHAL, Washington, D.C. Abstract.—A floating-plate and knife-edge beam-balance device designed for the direct measurement of the shear along the boundary of an open rectangular channel was built and tested. The instrument was found to be unsatisfactory; its accuracy was inadequate and its operation was erratic. The standard method for the determination of shear in open-channel flow utilizes a pitot tube as described by Hsu (1955). In an effort to make direct measurements of shear in open rectangular channels, a floating-plate beam-balance instrument was designed and built by the U.S. Geological Survey. Tests of the instrument in a laboratory flume indicated that its operation was unsatisfactory and its accuracy was inadequate. A schematic sketch of the instrument is shown in figure 121.1. The shear-plate device was tested at the National Bureau of Standards Hydraulic Laboratory in a smooth tiltable flume 140 feet long and 18 inches wide. Tests were made at normal depths of flow of 5 and 8 inches. The shear plate, 1 inch high, 6 inches long, and %6 inch thick, floats within an opening in the wall of the flume 118 feet from the entrance and has a clearance of 0.015 inch all around. The plate is set with its bottom 2 inches off the floor of the flume. Behind the plate, on the outside of the flume, is a water-filled plastic compartment so arranged that the plate is completely surrounded by fluid. Four vertical fins are attached to the back side of the plate, allowing free circulation of water between the flume and the outside well, and serving as hydraulic dampeners to lessen the tendency of the plate to flutter. The shear plate is mounted on the vertical arm of a knife-edge beam balance. The upstream end of the horizontal arm of the balance has suspended from it the core of a linear variable-differential transformer. The shear plate is positioned free and flush with the flume wall in still water, then exposed to the moving water in the flume. Any devia- tion from the null or centered position of the plate is indicated on the dial of a transformer indicator. On the downstream end of the horizontal arm is a pan to hold the weights necessary to maintain the null position of the plate. Because the ratio between the plate arm and the pan arm is unity, and because the knife edge of the balance is practically frictionless, the shear can be computed as the ratio of the weight in the pan Figure 121.1—Schematic sketch of shear-plate knife-edge beam-balance device. ART. 121 IN U.S. GEOL. SURVEY PROF. PAPER 475-C. PAGES C228-C229. 1963. C228DAVIDIAN AND CAHAL C229 to the area of the plate. The variation of the friction factor /, within a Keynolds number range from R = 6 X 1()4 to 6 X 105, is well described by the equation 1/V?=2.03 log IR V/-1-21. The relationship between the velocity of the flow and the plate shear as computed from the balance weights is shown in figure 121.2 (top curve). All the Q tests at 8 inches water depth and most of the tests at g 5 inches depth show a relation well described by the „ solid line. The dashed line shows that at 5 inches £ Q. depth and between velocities of 2 and 5 feet per sec- h ond (corresponding to Froude numbers, F, of about if 0.6 to 1.2) the measured plate shear is too high. Al- \ though it was not possible to achieve critical flow in the | 8-inch depth series, there seems to be no shear increase at velocities in the 8-inch series which are critical in the 5-inch series. This apparent reaction of the shear plate in the neighborhood of the critical Froude number has not been checked at other depths of water, but has been affirmed in a repetition of the measurements at the 5-inch depth. Figure 121.2 (bottom curve) shows the straight-line relationship between velocity and the plate shear as determined by measurements of local surface friction made by the Preston-tube method, using a round pitot at the wall surface in the vicinity of the plate. The 5-inch-depth measurements do not show an increase in shear at the critical Froude number. Additional studies of boundary shear in this flume, using a Preston tube are reported in Article 113. Although the reason for the high indicated shear values at F = 1.0 is not known, it is not this phenomenon alone which makes the shear-plate beam-balance device an unreliable instrument. The difficulty in visually adjusting the plate and in setting and keeping it free within its niche and flush with the wall becomes greater with increase in velocity and associated increases in opacity and turbulence. Furthermore, although it presents no difficulties if regularly removed, mineral matter deposited from the water and from the aerosol wetting agent vised, gradually builds up on various parts of the mechanism; however, it is easily cleaned off. In some tests, the weight required to bal- Figube 121.2.—Variation of shear with velocity. Top, shear computed with shear-plate beam-balance device; bottom, shear computed with Preston tube. Circles, 8-inch depth; dots, 5-inch depth. ance the shear force varied as much as ±20 percent. It is doubtful that an accuracy much better than ±10 percent can be ascribed to computations of shear based on the shear-plate data, particularly for higher Froude numbers, even though repeatability was excellent in some runs. The shear-plate device in its present form is an unsatisfactory means of measuring boundary-wall shear directly in water. Its accuracy is limited by the characteristics of the medium in which it is intended to operate. REFERENCE Hsu, E. Y., 1955, The measurement of local turbulent skin friction by means of surface pitot tubes: The David W. Taylor Model Basin Research and Devel. Rept. 957. i SUBJECT INDEX [For major subject heading such as “Economic geology/' “Ground water,” “Stratigraphy,” see under State names or refer to table of contents] A ABS. See Surfactants. Page Alaska, glacial geology, Copper River Basin... C121 paleontology, Gulf of Alaska............... 73 stratigraphy, Cook Inlet area........... 30 Alkalinity, of natural water, differences in determination..................... 212 Alluvial fans, competence of transport on___ 126 Alluvium, transmissibility, ground water.... 188 Alpine mafic magma stem, definition of term.. 82 Alteration, dolomite, by ground water....... 96 Alumina, theory of concentration.............. 161 Ammonites, Cretaceous, Wyoming................. 60 Anhydrite, solution............................ 91 Apatite, replacement of wood................. 100 Appalachian region, origin of bauxite deposits. 151 See also under State names. Aquifer tests, analysis by graphical multiple regression........................ 198 Arica Formation, Chile, paleontology........ 69 Arizona, economic geology, Ambrosia Lake area.............................. 166 stratigraphy, Defiance positive element... 28 Grand Canyon........................... 21 structural geology, Defiance positive element............................ 28 Arkansas, ground water, Arkansas River valley............................ 188 B Bacteria, effect on degradation of surfactants in detergents..................... 224 sorbtion of lead in water................. 220 Bauxite, theory of origin..................... 151 Bed sediment, computation of effective shear.. 202 Bolsons, distribution of granules............. 130 Bowser Formation, Alaska, definition........... 33 Bristlecone pines, origin of buttress roots. 149 Burns Formation, Colorado, redefinition______ 41 Buttress roots, as indicators of erosion rate_ 149 c Cadmium, in ground water...................... 179 Calcitization, dolomite, by ground water.... 96 California, engineering geology, Sacramento- San Joaquin Delta................. 162 geomorphology, north-central part........ 162 White Mountains....................... 149 mineralogy, San Joaquin Valley............ 100 paleontology, San Luis Obispo-Bakersfield area______________________________ 63 quality of water, Sierra Nevada........... 212 sedimentation, Deep Springs Valley------126,130 surface water, Yuba River basin........... 191 Cambrian, Tennessee, paleontology-------------- 53 Canyon Mountain Complex, Oregon, definition........................................... 82 Carbonate rocks, alteration by ground water.. 96 Carboniferous. See Mississippian, Pennsylvanian. Chile, paleontology, Arica area............... 69 Chromium, in ground water..................... 179 Collapse structures, as control for uranium deposition......................- 166 Page Colorado, geomorphology, Sage Plain........... C138 geophysics, Rampart Range area.............. 110 stratigraphy, northeastern part............. 23 Pueblo area.............................. 49 San Juan Mountains....................... 39 west-central part........................ 35 Conodonts, Ordovician and Devonian, Tennessee........................................ 55 Continental shelf, submarine geology, east coast ............................. 132 C retaceous, Idaho, petrology------------------ 86 Montana, petrology------------------------- 86 Wyoming, paleontology...................... 60 Culvert coefficients, field vertification of laboratory computation methods. 194 Cynthia Falls Sandstone, Alaska, redefinition. 33 Cyprus, petrology, Troodos Complex........... 82 D Detergents. See Surfactants. Devonian, Montana and Wyoming, stratig- raphy............................... 17 Tennessee, paleontology.................... 55 Wyoming, stratigraphy...................... 14 Diamicton, in glaciolacustrine deposits........ 121 Di .toms, Pleistocene, Chile..................... 69 Discharge, stream, effect of urbanization on.. 185 Dolomite, calcitization by calcium sulfate__ 96 E Edgewood Dolomite, Iowa, stratigraphy_______ 11 Elm, American, effect of soil chemicals..... 105 Energy-budget analysis, of heated river_____ 175 Englewood Formation, South Dakota and Wyoming, correlation................. 19 Erosion, vegetation as an indicator of rate_ 149 Eureka Tuff Colorado, redefinition_______________ 41 F Feldspar, effect on oxidation of manganese ions in water....................... 216 Fenton Pass Formation, Wyoming, definition. 45 Fitz Creek Siltstone, Alaska, definition____ 33 Flow-duration curves, estimated by low-flow measurements________________________ 196 Flow-through cell, for spectrophotometer.... 173 Foraminifera, Pleistocene, Alaska--------------- 73 G Gaikema Sandstone, Alaska, redefinition_____ 30 Galena, in mine tailings, effect on vegetation.. 105 Gilpin Peak Tuff, Colorado, definition_____ 43 Glacial lakes, Massachusetts............. 142 Granules, origin in sediments.................. 130 Gravity surveys, Colorado, Rampart Range area_________________________ 110 Hawaii, volcanoes................... 114 H Hawaii, geophysics, i sland of Hawaii......... 114 Henson Formation, Colorado, redefinition___ 41 Hesse Quartzite, Tennessee, paleontology... 53 Page Hoplitoplacenticeras, Cretaceous, Wyoming... C60 Hydraulics, direct measurement of shear in open-channel flow_................. 228 distribution of shear in rectangular channels....................................... 206 I Idaho, petrology, Idaho batholith.............. 86 petrology, Yellowstone Plateau............. 78 Idaho batholith, modal composition.............. 86 Illite, adsorption of surfactants on........... 221 Indiana tennesseensis, Cambrian, Tennessee 54 Iowa, stratigraphy, Dubuque County.............. 11 Isotope analysis, preparation of samples..... 166 Israel, paleontology, Makhtesh Ramon......... 58 J Jurassic, Alaska, stratigraphy................. 30 L Lead, removal from water by aquatic organisms........................................ 220 Lead-alpha age determinations, Carolina slate belt........................... 107 Lead isotopes, preparation of samples for analysis......................... 166 Leucophosphite, occurrence in nodules...... 100 Limestone, spirorbal......................... 14 Lluta Formation, Chile, paleontology........... 69 M Madison Limestone, Colorado, Nebraska, Wyoming, stratigraphy............. 23 Magnetic anomalies, evaluation by electromagnetic measurements........................ 117 Magnetite, estimation of content by electromagnetic methods............................ 117 Manganese, effect of feldspar on oxidation_ 216 Manganese dioxide, effect of tannic acid on solubility....................... 218 Maquoketa Shale, Iowa, stratigraphy. ......... li Massachusetts, glacial geology, Concord area. 142 Mexico, structural geology, northern Zacatecas............................................ 7 Mine tailings, effect on vegetation.......... 105 Minnelusa Formation, South Dakota and Wyoming, petrology................ 96 South Dakota and Wyoming, stratigraphy.................................... 91 Minnesota, geophysics, Cuyuna Range........ 117 Miocene, California, paleontology............ 63 Mississippian, Arizona, stratigraphy.......... 21 Colorado, stratigraphy.................... 23 Montana and Wyoming, stratigraphy------ 17 Nebraska, stratigraphy.................... 23 Wyoming, stratigraphy..................... 23 Moenkopi Formation, Arizona, structural geology........................... 28 Mollusks, Miocene, California................. 63 Montana, petrology, Idaho batholith___________ 86 petrology, Yellowstone Plateau............ 78 stratigraphy, southern part............... 17 C231C232 SUBJECT INDEX Page Moreno Formation, California, mineralogy... C100 Morrison Formation, New Mexico, economic geology.............------------ 156 New Mexico, structural geology............ 156 Mudflow deposits, subaqueous, in proglacial lakes............................. 121 Multiple-regression analysis, graphical, of aquifer tests...................... 198 Murray Shale, Tennessee, paleontology....... 53 N Nebraska, stratigraphy, western................ 23 New Jersey, quality of water, coastal plain... 212 New Mexico, economic geology, Ambrosia Lake.............................. 156 structural geology, Ambrosia Lake.......... 156 New York, ground water, Long Island--------- 179,185 surface water, Dutchess County............. 196 Long Island............................ 179 Nitrate, in precipitation.................... 209 Nodules, leucophosphite....................... 100 North Carolina, geochronology, Carolina slate belt............................... 107 quality of surface water, eastern region... 209 Nussbaum Alluvium, Colorado, redefinition.. 49 o Ogallala Formation, Colorado, stratigraphy.. 60 Ohio Creek Formation, Colorado, redefinition... 35 Oklahoma, geomorphology, Sandstone Creek. 145 Ordovician, Iowa, stratigraphy.................. 11 North Carolina, geochronology............. 107 Tennessee, paleontology.................... 55 Oregon, petrology, Canyon Mountain.............. 82 Ostracodes, Triassic, Israel................... 58 Oxidation, of manganese ions in water---------- 216 P Paleotemperature studies, Miocene, California......................................... 63 Peak discharge, field verification of laboratory computation methods............... 194 Peatland, subsidence.......................... 162 Pennsylvania, quality of water, Susquehanna River............................. 175 Pennsylvanian, South Dakota and Wyoming, stratigraphy....................... 91 Percent-constituent determination, by printing spectrophotometer......................... 171 Permeability, graphical multiple-regression analysis.......................... 198 Permian, South Dakota and Wyoming, stratigraphy....................... 91 pH, of natural water, differences in determinations......................................... 212 Picayune Formation, Colorado, redefinition. 41 Page Pleistocene, Alaska, diamicton deposits_____ C121 Alaska, paleontology........................ 73 C hile, paleontology...................... 69 Colorado, geomorphology.................... 138 stratigraphy........................... 49 Massachusetts, glacial lakes............... 142 Wyoming, stratigraphy....................... 45 Pliocene, Colorado, stratigraphy................ 49 Idaho, Montana, Wyoming, petrology______ 78 Potosi Volcanic Group, Colorado, redefinition. 43 Precambrian, Wyoming, structural geQlogy... 1 Precipitation, content of sulfate and nitrate... 209 Q Quaternary, Idaho, Montana, Wyoming, petrology...................... 78 See also Pleistocene, Recent. R Radioactive tracers, in study of removal of dissolved ions from waters..... 220 Recent, Alaska, paleontology_________________ 73 Recharge, effect of urbani zation on....... 185 Red Glacier Formation, Alaska, definition_ 30 Redwall Limestone, Arizona, definition of new members............................ 21 s Sand, feldspathic, effect on oxidation of manganese ions in water.......................... 216 San Juan Formation, Colorado, redefinition.. 39 Scarps, origin................................. 138 Sediments, distribution on east coast continental shelf................................ 132 transport on alluvial fans................. 126 Shear, direct measurement in open-channel flow............................... 228 distribution in rectangular channels..... 206 on bed sediment, computation............... 202 Silurian, Iowa, stratigraphy.................... 11 Silverton Volcanic Group, Colorado, redefinition------------------------------------------ 41 Snowmelt, determination by study of stream- flow............................... 191 Solution breccia, in Minnelusa Formation----- 91 Souris River Formation, Wyoming, stratigraphy.......................................... 14 South Dakota, petrology, Black Hills.........91,96 stratigraphy, Black Hills.................. 91 Specific conductance, of natural water, differences in determinations...................... 214 Spectrophotometer, accessory for calculating and printing percent constituent. _ 171 Sphalerite, in mine tailings, effect on vegetation. ........................................ 105 Spirorbis, Devonian, Wyoming................... 14 Steele Shale, Wyoming, paleontology.------------ 60 Stream channels, geometry.............,...... 145 Page Streamflow, computation of effective shear on bed sediment.................... C202 estimation of flow-duration curves______ 196 Subsidence, of peatland...................... 162 Sulfate, in precipitation.. 209 Sunshine Peak Rhyolite, Colorado, redefinition ......................................... 43 Surfactants, adsorption on illite.............. 221 biodegradation in synthetic detergents___ 224 T Tannic acid, effect on solubility of manganese. 218 Tennessee, paleontology, Blount County______ 53 paleontology, Flynn Creek area-------------- 55 Terra rossa, as bauxite parent material..... 151 Tertiary, Colorado, stratigraphy............35,39 Mexico, plutonic rocks...................... 7 Thermal loading, heat dissipation from streams............................ 175 Transmissibility, alluvium.................. 188 Triassic, Arizona, structural geology___________ 28 Israel, paleontology..................... 58 Oregon, petrology......................... 82 Troodos Complex, Cyprus, petrology.............. 82 Turbidity-current deposits, in proglacial lakes 121 Tuxedni Group, Alaska, definition--------------- 30 Twist Creek Siltstone, Alaska, definition___ 33 u Uranium, structural control of deposition___ 156 Urbanization, effect on storm discharge and ground-water recharge-------------- 185 v Vegetation, as an indicator of rate of erosion.. effect of mine tailings................ effect on stream-channel geometry...... sorbtion of lead in water.............. Virginia, economic geology, Valley and Ridge province........................ quality of surface water, southern region.. Volcanism, Colorado, San Juan Mountains... Volcanoes, Hawaii, gravity survey.......... w Wisconsin, biogeochemistry, Grant County. _ 105 geophysics, Gogebic Range................. 117 Wood, apatitized, occurrence.................. 100 Wyoming, oaleontology, Bighorn Mountains. 14 paleontology, Carbon County............... 60 petrology, Black Hills................... 91,96 Yellowstone Plateau................... 78 stratigraphy, Bighorn basin................ 45 Bighorn Mountains.. ................... 14 Black Hills........................... 91 Laramie Range__________________________ 23 northern part......................... 17 structural geology, Teton Range............. 1 149 105 145 220 152 209 39 114AUTHOR INDEX A Addicott, W. 0_. Anderson, Peter. Antweiler, J. C_- Page C63 212 166 B Bedinger, M. S........................... 188 Bergman, D. L___........................ 145 i Bowles, C. Q.._.........................91,96 Braddock, W. A........................ 91,96 Brown, C. E...........................- 11 Burbank, W. S.......................... 39 c Cabal, D. L.............................- 206,228 Cobban, W. A------------------------------- 60 Colby, B.R............................... 202 Cserna, Zoltan de........................... 7 Curtis, E. L............................... 171 D Davidian, Jacob—---------------------- 206,228 Davis, G. H----------------------------- 162 Detterman, R. L...............-........- 30 Dingman, R. J--------------------------- 69 E Ekren, E. B........... Emmett, L. F.......... 117 188 F Ferrians, O. J............................- 121 Feth, J. H................................. 212 Frauenthal, H. L.....................--.... 179 Frischknecht, F. C......................... 117 Gambell, A. W.... Gaskill, D. L____ Godwin, L. H..... Granger, H. C.... Gulbrandsen, R. A. C233 209 35 35 156 100 H Hamilton, Warren_______ Hem, J. D.............. Huddle, J. W........... Hunt, O.P.............. J Page C78 216 55 196 Jenkins, C. T.............................- 194,198 Jones, D. L.................................. 100 K Kinoshita, W. T............................. 114 Knechtel, M. M............................... 151 Koteff, Carl................................. 142 Krivoy, H. L................................ 114 L LaMarche, V. C., Jr....................... 149 Laurence, R.A.............................. 53 Leopold, E. B..........................-- 45 Lieber, Maxim-------------------------- 179 Lohman, K. E_____________________________ 69 Luedke, R. G----------------------------- 39 Lustig, L. K........................ 126,130 M Mabey, D. R...........................- 114 MacDonald, R. R_________________________ 114 McKee, E. D_...........................21,28 Maughan, E. K............................ 23 Messinger, Harry....................... 175 Miller, C. H............................ 110 o Oborn, E.T...............................- 220 Ojeda, J................................. 7 P Page, H. G.................................. 221 Palmer, A. R_______________________________ 53 Perlmutter, N. M--------------------------- 179 Rantz, S. E------------------------------- C191 Rawson, Jack------------------------------ 218 Reed, J. C., Jr............................... 1 Reeser, D.W................................ 100 Roberson, C. E.............................. 212 Robertson, J. B......................... 221,224 Rogers, C.L................................... 7 Rohrer, W. L--------------------------------- 45 Ross, C. P----------------------------------- 86 s Sandberg, C. A............................. 14,17 Santos, E. S............................... 156 Sawyer, R. M.......-........................ 185 Scott, G. R.................................. 49 Seaber, P. R................................ 212 Shacklette, H. T............................. 105 Shapiro, Leonard............................ 171 Shawe, D. R................................ 138 Smith, P. B................................... 73 Sohn, I. G.................................... 58 Stern, T. W.................................. 107 Stromquist, A. A............................ 107 Sullivan, C. W---------------------------- 145 T Tagg, K. M................................ 100 Ta era, E................................... 7 Thayer, T. P............................... 82 u Uchupi, Elazar.......................... 132 V Vedder, J. G_____________________________63 Vloten, Roger van.......................... 7 w Wayman, C.H......................... 221,224 Westley, Harold....................... 107 White, A. M............................. 107 Whitlow, J. W........................... 11 i \ U.S. GOVERNMENT PRINTING OFFICE : 1963 O—694-0277 DAY USE 1M -» T -CV T~ ■> 'If '?4 / $ 1 IShort Papers in Geology and Hydrology Articles 122-1 72 GEOLOGICAL SURVEY RESEARCH 1963 GEOLOGICAL SURVEY PROFESSIONAL PAPER 475-D Scientific notes and summaries oj investigations prepared by members of the Conservation, Geologic, and IVater Resources Divisions UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1964UNITED STATES DEPARTMENT OF THE INTERIOR STEWART L. UDALL, Secretary GEOLOGICAL SURVEY Thomas B. Nolan, Director For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402FOREWORD This collection of 51 articles is the last of the chapters of Professional Paper 475. The articles report on scientific and economic results of current work by members of the Geologic and Water Resources Divisions of the U.S. Geological Survey. Some of the papers present results of completed parts of continuing investigations; others announce new discoveries or preliminary results of investigations that will be discussed in greater detail in reports to be published in the future. Still others are scientific notes of limited scope, and short papers on techniques and instrumentation. Chapter A of this series presents a summary of investigations covering a wide range of work done during the 1963 fiscal year. It also includes maps showing the status of topographic mapping, a list of publications by Survey authors, a list of cooperating agencies, addresses of principal offices, and a list of investigations in progress. Thomas B. Nolan, Director.CONTENTS Page Foreword______________________________________________________________________________________________________________ hi GEOLOGIC STUDIES Stratigraphy and structural geology 122. Comparison of late Paleozoic depositional history of northern Nevada and central Idaho, by R. J. Roberts and M. R. Thomasson_____________________________________________________________________________________________________ D1 123. Thrust-fault relations in the northern Coast Ranges, Calif., by R. D. Brown, Jr_______________________________________ 7 124. Upper Pliocene marine strata on the east side of the San Joaquin Valley, Calif., by R. L. Klausing and K. E. Lohman _ 14 X125. Miocene vertebrates of the Barstow Formation in southern California, by G. E. Lewis__________________________ 18\ 126. The Thirsty Canyon Tuff of Nye and Esmeralda Counties, Nev., by D. C. Noble, R. E. Anderson, E. B. Ekren, and J. T. O’Connor______________________________________________________________________________________________________ 24 127. St. Kevin Granite, Sawatch Range, Colo., by Ogden Tweto and R. C. Pearson____________________________________________ 28 128. Reinterpretation of the late growth of the Gypsum Valley salt anticline, San Miguel County, Colo., by F. W. Cater. 33 129. A Pleistocene section at Leonards Cut, Burke County, N.C., by J. C. Reed, Jr., Bruce Bryant, E. B. Leopold, and Louise Weiler_______________________________________________________________________________________________________ 38 130. Surface and subsurface stratigraphic sequence in southeastern Mississippi, by D. H. Eargle_________________ 43 Mineralogy and petrology 131. Magmatic differentiation in a volcanic sequence related to the Creede caldera, Colorado, by J. C. Ratte and T. A. Steven_________________________________________________________________________________________________________ 49 132. Revised Tertiary volcanic sequence in the central San Juan Mountains, Colo., by T. A. Steven and J. C. Ratte. _ 54 133. Valleriite and the new iron sulfide, mackinawite, by H. T. Evans, Jr., Charles Milton, E. C. T. Chao, Isidore Adler, Cynthia Mead, Blanche Ingram, and R. A. Berner____________________________________________________________ 64 134. Eddies as indicators of local flow direction in rhyolite, by David Cummings_______.________________________ 70 Geochemistry 135. Cadmium in samples of the Pierre Shale and some equivalent stratigraphic units, Great Plains region, by H. A. Tourtelot, Claude Huffman, Jr., and L. F. Rader_____________________________________________________________ 73 136. Analysis of geochemical prospecting data from the Rocky Range, Beaver County, Utah, by J. J. Connor and A. T. Miesch______________________________________________________________________________________________________________ 79 137. Investigation of sampling-error effects in geochemical prospecting, by A. T. Miesch and J. J. Connor_______ 84 138. Effect of cation exchange on the thermal behavior of heulandite and clinoptilolite, by A. O. Shepard and H. C. Starkey_____________________________________________________________________________________________________________ 89 139. Determination of the ion-exchange capacity of a zeolitic tuff, by H. C. Starkey____________________________ 93 140. Geological and geochemical reconnaissance, southern part of the Smyrna Mills quadrangle, Aroostook County, Maine, by Louis Pavlides and F. C. Canney_________________________________________________________________________ 96 Geochronology 141. Age of basement rocks from the Williston basin of North Dakota and adjacent areas, by Z. E. Peterman and C. E. Hedge_____________________________________________________________________________________________________ 100 142. Isotopic ages of glaucophane schists from the area of Cazadero, Calif., by D. E. Lee, H. H. Thomas, R. F. Marvin, and R. G. Coleman._______________________________________________________________________________________ 105 Geophysics Xl43. Crustal structure in the vicinity of Las Vegas, Nev., from seismic and gravity observations, by J. C. Roller_ 108 X. 144. Hawaiian seismic events during 1962, by R. Y. Koyanagi______________________________________________________ 112 145. Seismic investigations on Cape Cod, Mass., by R. N. Oldale and C. R. Tuttle_________________________________ 118 Economic geology 146. Geologic setting of the Spar City district, San Juan Mountains, Colo., by T. A. Steven______________________ 123 147. Thorium and uranium in monazite from Spokane County, Wash., by J. W. Hosterman, W. C. Overstreet, and J. J. Warr, Jr___________________________________________________________________________________________________ 128 148. Rich oil shale from northern Alaska, by I. L. Tailleur______________________________________________________ 131 vVI CONTENTS Sedimentation Page 149. Dissimilarity between spatial and volocity-weighted sediment concentrations, by H. P. Guy and D. B. Simons__ D134 150. Temporary storage of fine sediments in islands and point bars and alluvial channels of the Rio Grande, New Mexico and Texas, by C. F. Nordin, Jr., and J. P. Beverage________________________________________________________ 138 151. Fluvial sedimentation in Mammoth Cave, Ky., by C. R. Collier and R. F. Flint________________________________ 141 152. Quaternary mudflow deposits near Santiago, Chile, by Kenneth Segerstrom, Octavio Castillo U., and Eduardo Falcdn M________________________________________________________________________________________________ 144 Paleontology 153. New occurrences of the rugose coral Rhizophyllum in North America, by W. A. Oliver__________________________ 149 Glacial geology 154. Glacial chronology of Ullsfjord, northern Norway, by G. W. Holmes and B. G. Andersen________________________ 159 Analytical techniques 155. Effluent collector for gas chromatography, by D. F. Goerlitz and W. L. Lamar________________________________ 164 156. Use of sodium-sensitive glass electrodes for solubility determinations, by A. H. Truesdell and C. L. Christ_ 167 157. Semimicro X-ray fluorescence analysis of tektites using 50-milligram samples, by H. J. Rose, Jr., Frank Cuttitta, M. K. Carron, and Robena Brown__________________________________________________________________________ 171 158. Determination of total iron in hematitic iron ores by X-ray fluorescence spectrometry, by W. W. Niles____ 174 159. A gas jet for d-c arc spectroscopy, by A. W. Helz________________________________________________________ 176 HYDROLOGIC STUDIES Geochemistry of water 160. Dolomite solubility in ground water, by Ivan Barnes and William Back..___________________________________ 179 161. Effect of tree leaves on water quality in the Cacapon River, W. Va., by K. V. Slack______________________ 181 162. Relation of percent sodium to source and movement of ground water, National Reactor Testing Station, Idaho, by F. H. Olmsted________________________________________________________________________________________ 186 163. Relation of fluoride content to recharge and movement of ground water in Oasis Valley, southern Nevada, by G. T. Malmberg and T. E. Eakin___________________________________________________________________________________ 189 Analytical hydrology 164. Reservoir storage on streams having log-normal distributions of annual discharge, by C. H. Hardison_________ 192 Marine hydrology 165. Environmental factors affecting attached macro-organisms, Patuxent River estuary, Maryland, by R. L. Cory___ 194 Surface water 166. Anomalous streamflow-ground-water regimen in the Mad River basin, near Springfield, Ohio, by W. P. Cross and A. J. Feulner________________________________________________________________________________________ 198 167. Height-frequency relations for New Jersey floods, by D. M. Thomas------------------------------------------- 202 Ground water 168. Seasonal temperature fluctuations in Surficial sand near Albany, N.Y., by R. C. Heath-------------------- 204 169. Hydrogeology of the Santiago area, Chile, by W. W. Doyel, R. J. Dingman, and Octavio Castillo U__________ 209 170. Ground water in the Arica area, Chile, by W. W. Doyel____________________________________________________ 213 Hydrologic instrumentation 171. “Lazy” thermometers and their use in measuring ground-water temperatures, by R. C. Heath_________________ 216 172. Television apparatus for borehole exploration, by J. E. Eddy.----------------------------------------- 219 INDEXES Subject---------------------------------------------------------------------------------------------------------- 221 Author------------------------------------------------------------------------------------------------------------- 223Article 122 COMPARISON OF LATE PALEOZOIC DEPOSITIONAL HISTORY OF NORTHERN NEVADA AND CENTRAL IDAHO By RALPH J. ROBERTS and M. RAY THOMASSON,1 Menlo Park, Calif., and Midland, Tex. Abstract.—Three major sequences of upper Paleozoic rocks have been mapped in northern Nevada: one deposited east of the Antler orogenic belt, one within the belt, and one west of the belt. Recognition of comparable sequences in central Idaho suggests similarity in depositional history of the two areas. In 1960, sections of Paleozoic rocks in the area between Arco, Challis, and Hailey, Idaho (fig. 122.1), were visited by Roberts in company with C. P. Ross and other geologists of the U.S. Geological Survey. Thomasson had previously studied sections of upper Paleozoic rocks in central Idaho in 1956, 1957, and 1958 during fieldwork on his dissertation for the University of Wisconsin. The authors independently noted similarities between upper Paleozoic rocks and tectonics of north-central Nevada and those of central Idaho, which suggest that the two areas have related depositional histories. Although rock units cannot be traced from one area to the other because of the intervening volcanic rocks of the Snake River Plain, the stratigraphic succession, lithology, and structural histories are comparable. Comparison of the sequences in the two areas is shown in the accompanying table and on figure 122.2. ROCK SEQUENCES In north-central Nevada, three contemporaneous major sequences 2 have been described (Roberts and others, 1958) (fig. 122.2A): (1) the Eureka-Carlin sequence, a combination of the Eureka sequence, 1 Shell Oil Co. 2 The term “sequence” is used here in an informal sense to denote a geographically discrete succession of major rock units that were deposited under related environmental conditions (Silberiing and Roberts, 1962). A single name is applied to a sequence in order to avoid repeating all the formation names. In the same way, comparable sequences in Idaho are given informal names of their own. Figure 122.1.—Index map of report area in Nevada and Idaho, showing part of the Antler orogenic belt and the Snake River Plain (stippled pattern). ART. 122 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D1-D6. 1964. D1D2 STRATIGRAPHY AND STRUCTURAL GEOLOGY Descriptions of late Paleozoic rocks in northern Nevada and central Idaho Sequence as shown in fig. 122.2 Name, age, and thickness of formation Lithology Facies Northern Nevada Havallah sequence (Roberts and others, 1958; Silberling and Roberts, 1962) Havallah Formation, Pennsylvanian and Permian, 10,000+ ft Upper part: Calcareous sandstone, limestone, shale, minor conglomerate Lower part: Sandstone, shale, chert Upper part: Offshore but shoaling toward the top. Lower part: Offshore. Pumpernickel Formation, Pennsylvanian (?), 7,000+ ft Argillite, chert, sandstone, greenstone Eugeosynclinal, mostly offshore and moderately deep water. Edna Mountain Formation, Permian, 700+ ft Sandstone, shale, shaly limestone Nearshore, shallow water. Antler sequence (Roberts and others, 1958) Antler Peak Limestone, Upper Pennsylvanian and Permian, 625 ft Limestone, shaly and sandy limestone Nearshore, shallow water. Battle Formation, Middle Pennsylvanian, 730+ ft Conglomerate, sandstone, shale, limestone Terrestrial to shallow-water marine. Buckskin Mountain Formation, Permian, 1,200+ ft Sandstone, limestone Nearshore, shallow water. Carlin sequence (AfterDott, 1955; Fails, 1960) Strathearn Formation, Upper Pennsylvanian and Permian, 1,500+ ft Limestone, fine conglomerate Shallow water. Moleen and Tomera Formations, Lower and Middle Pennsylvanian, 3,200+ ft Limestone, fine conglomerate, sandstone Shallow water. Tonka Formation, Lower Mis-sissippian to Pennsylvanian, 2,500+ ft Conglomerate, sandstone, shale, graywacke Offshore, foredeep, moderately deep water. Central Idaho Ketchum sequence (Umpleby and others, 1930; Bostwick, 1955; Thomasson ■) Wood River Formation, Pennsylvanian and Permian, 8,000 + ft Sandstone, calcareous sandstone, limestone, chert Basal conglomerate, 40-260 ft Mostly offshore, moderately deep to shallow water. Milligen Formation, Mississip-pian(?), 7,000+ ft Argillite, siliceous argillite, chert, limestone, quartzite Eugeosynclinal, offshore, moderately deep water. Summit Creek sequence (Modified after Thomasson1) Conglomerate on Summit Creek, Upper Pennsylvanian, 1,910 ft Conglomerate, sandstone, shale Shallow water. Unnamed unit Shale Shallow water. Star Hope sequence (Modified after Thomasson;1 Ross, 1960; Skipp, 1961a) Conglomerate in Muldoon area, Lower Pennsylvanian, 1,324 ft Conglomerate, shale, sandstone, limestone Shallow water. Muldoon Formation: Wildhorse Member, Upper Mississippian or Lower Pennsylvanian, 2,586 ft Iron Mine Member, Upper Mississippian, 3,650 ft Garfield Member, Upper Mississippian, 2,587 ft Copper Creek Member, Lower Mississippian, 861 ft Unnamed shale member, Lower Mississippian, 4,000 ft Shale, minor conglomerate, limestone Conglomerate, sandstone, silt-stone, mudstone Sandstone, siltstone, mudstone Graded graywacke, black shale Siliceous argillite, minor chert, sandstone, limestone Foredeep, offshore: Moderately deep water. Do. Do. Do. Moderately deep to deep water. 1 See footnote 3, p. D4.ROBERTS AND THOMASSON D3 ANTLER OROGENIC BELT A. Northern Nevada, Mississippian through Early Permian. Following the Antler orogeny, sediments of the Antler sequence were deposited near the orogenic belt and those of the Havallah and Carlin sequence farther offshore B. Northern Nevada, Late Permian. During the Sonoma orogeny, the Havallah sequence moved eastward on the Golconda thrust, and clastic rocks were shed into eastern Nevada Summit Creek C. Central Idaho, Missi ssippian through Early Permian. Following the Mississippian orogeny, sediments of the Summit Creek sequence were deposted near the orogenic belt and those of the Ketchum and Star Hope sequences farther offshore EXPLANATION Conglomerate Sandstone Shale and graywacke Bedded chert Limestone Shaly limestone Shale or argillite Figure 122.2.—Diagrammatic sections showing // Volcanic rock Dolomite a comparison of late Paleozoic depositional history in northern Nevada with that in central Idaho.D4 STRATIGRAPHY AND STRUCTURAL GEOLOGY (composed of the Chainman Shale, Diamond Peak Formation, Ely Limestone, and the Carbon Ridge and Garden Valley Formations, exposed in the Eureka area) (fig. 122.1), and the Carlin sequence exposed farther north (composed of the Tonka, Moleen and Tomera, Strathearn, and Buckskin Mountain Formations of Dott, 1955, and Fails, 1960); (2) the Antler sequence, composed of the Battle Formation, Antler Peak Limestone, and Edna Mountain Formation; and (3) the Havallah sequence, composed of the Pumpernickel and Havallah Formations (Ferguson, Muller, and Roberts, 1951; Silberling and Roberts, 1962). The Eureka and Carlin sequences were deposited east of the Antler orogenic belt; the Antler sequence was deposited within and near the belt; and the Havallah sequence was deposited west of the belt in a separate basin. The clastic rocks that predominate in these sequences were derived from the emergent Antler orogenic belt. Carlin and Star Hope Sequences Rocks comparable to the Carlin sequence have been described in central Idaho (Ross, 1934, 1960, 1962a; Skipp, 1961a, 1961b; Scholten, 1957; Thomasson, 1959a, 1959b, Thomasson; 3 Shannon, 1961) and are here called the Star Hope sequence. In the Muldoon area, between Hailey and Mackay, the sequence is 15,340 feet thick (Thomasson 3) and is composed of an unnamed shale at the base, the Muldoon Formation, and an upper conglomerate. Correlative units have been recognized in the Copper Basin area, 10 miles west of Mackay (Ross, 1960, 1962a, 1962b, 1962c; Skipp, 1961a, 1961b). The Carlin and Star Hope sequences are both characterized by shale and siliceous shale at the base of the section and by conglomerate, graywacke, sandstone, shale, and limestone in the middle and upper parts. In detail, the sections differ, but in general the rock types are similar. The conglomerates in both sequences are composed predominantly of fragments of chert and quartzite derived from eugeosynclinal rocks; these conglomerates coarsen westward, indicating derivation from a western source. Eastward, both sequences intertongue with finer clastic rocks and limestone of shelf facies. Antler and Summit Creek sequences Formations of the Antler sequence, which were deposited within and near the orogenic belt in northern Nevada, are much thinner and more lenticular than offshore correlative units on both sides of the belt (figs. 122.1 and 122.2A). Locally, uplift during the late Paleozoic caused erosion of parts or all of the sequence. In central Idaho, comparable rocks thus far have been 3 Thomasson, M. R., 1959, Late Paleozoic stratigraphy and paleotectonics central and eastern Idaho: Wisconsin Univ., Ph. D. thesis, 244 p. recognized only at one place, on the ridge at the head of Summit Creek. They are here called the Summit Creek sequence (figs. 122.2Cand 122.3). This sequence consists of about 1,900 feet of conglomerate, sandstone, and shale of Pennsylvanian age (Thomasson 3). The basal unit, like the Battle Formation of the Antler sequence, rests unconformably on folded Ordovician rocks. Havallah and Ketchum sequences In the Hailey area the Milligen and Wood River Formations, together here called the Ketchum sequence, are respectively similar to the Pumpernickel and Havallah Formations of the Havallah sequence in Nevada (Silberling and Roberts, 1962). The Milligen Formation, at the base, is mostly dark shale and chert; the Pumpernickel is mostly shale and radiolarian chert, but locally contains a good deal of volcanic rock; the Milligen apparently lacks volcanic rock. The Wood River Formation contains bedded chert, sandy limestone, and shale that are strikingly similar to comparable units of the Havallah Formation; the Wood River differs in that it contains coarser clastic rocks near the base. The source of the clastic rocks in the Ketchum sequence has not yet been resolved. Thomasson3 measured current directions in the Wood River Formation; north of Hailey these suggest derivation from the southwest; near Hailey, current directions suggest derivation from the northwest. No diagnostic current marks were noted in the Milligen Formation. However, Roberts believes that some detritus in the Milligen and Wood River Formations must have been transported westward from the orogenic belt, as indicated in figure 122.2, but agrees that other sources may also have supplied detritus. Contrasts of the Ketchum with other Idaho sequences The sharp contrasts in lithology and stratigraphy between the Ketchum, Summit Creek, and Star Hope sequences in Idaho are due to differences in their depositional environments. The Ketchum sequence represents mostly an offshore and deep-water facies in the lower part and shallow-water facies in the upper part, whereas the Summit Creek represents a nearshore facies that grades eastward into offshore facies of the Star Hope sequence. Ross (1962b, p. 385, 1962c, pi. 1) shows a depositional contact between rocks here assigned to the Ketchum and Star Hope sequences. On the geologic map of the Hailey quadrangle these rocks are shown in thrust-fault contact (Umpleby and others, 1930, pi. 1); moreover, the Ketchum sequence has been highly deformed (figs. 122.2, 122.3), in sharp contrast to the relatively simple deformation of the Summit Creek and Star Hope sequences. These con-43° 50' 114° 20' INDEX MAP 0 2 I_____________L HAILEY QUADRANGLE Area shown in A and B iKetchum I Hailey EXPLANATION Tertiary volcanic rocks Cretaceous (?) intrusive rocks Cwr Ccg; Wood River Formation Ccg, basal conglomerate jcm^ Milligen Formation o Contact Fault Thrust fault Teeth on upper plate Ordovician rocks Algonkian rocks KETCHUM SEQUENCE B EXPLANATION Tertiary volcanic rocks Cretaceous (?) intrusive rocks SUMMIT CREEK SEQUENCE U4'10' PIPw Ewe; Wood River Formation Pwc. basal conglomerate Milligen Formation 4 MILES _l Ordovjcian rocks m Algonkian rocks STAR HOPE SEQUENCE Conglomerate on Summit Creek ANGULAR UNCONFORMITY Undifferentiated rocks l in Boulder Creek and f Muldoon areas J z o Figure 122.3.—Geologic map of part of the Hailey quadrangle, Idaho, showing original and alternative interpretations. A, Original interpretation: generalized after Umpleby and others (1930, pi. 1). All Pennsylvanian clastic rocks are assigned to the Wood River Formation. B, Alternative interpretation: Pennsylvanian clastic rocks are differentiated and assigned to separate sequences; the Ketchum sequence is thrust over Algonkian (?) and Ordovician rocks and the M Summit Creek sequence. Age assignments are according to Bostwick (1955) and Thomasson.3 O ROBERTS AND THOMASSOND6 STRATIGRAPHY AND STRUCTURAL GEOLOGY trasts are interpreted as evidence that the Ketchum sequence moved into the Hailey area from the west or southwest on a major thrust (here named the Hailey thrust), and is therefore allochthonous like the Havallah sequence in Nevada (Roberts and others, 1958, p. 2849; Silberling and Roberts, 1962). CONCLUSION The striking similarities in major sequences of upper Paleozoic rocks in northern Nevada and central Idaho indicate similarity in depositional and orogenic history. Nolan and others (1956, p. 54) suggest Carboniferous uplift of a landmass to the west of Eureka, Nev., which supplied coarse detritus to the Diamond Peak Formation; Thomasson,3 Ross (1960, p. B232; 1962a), and Skipp (1961a, p. 387-388) also recognized that coarse clastic rocks in the Mackay quadrangle, Idaho, indicate Mississippian to Pennsylvanian orogeny to the west. The orogenic movements in Nevada took place along the Antler orogenic belt which probably extends northward into central Idaho (fig. 122.1). This belt now may be largely covered in Idaho by younger thrust plates. Remnants of the orogenic belt in the Hailey area show characteristic structural features and allochthonous rock units like those of the orogenic belt in Nevada. The picture presented here is no doubt oversimplified, and additional fieldwork in central Idaho should be undertaken to test these suggestions in the field. The terms Milligen and Wood River should be restricted to the formations in their type areas and to continuations of these lithologic units within the Milligen-Wood River depositional basin. New terms should be used for correlative units deposited within the orogenic belt, in the Muldoon trough, and on the shelf east of the trough. REFERENCES Bostwick, D. A., 1955, Stratigraphy of the Wood River formation, south-central Idaho: Jour. Paleontology, v. 29, no. 6, p. 941-950. Dott, R. H., 1955, Pennsylvanian stratigraphy of Elko and northern Diamond Ranges, northeastern Nevada: Am. Assoc. Petroleum Geologists Bull., v. 39, p. 2211-2305. Fails, J. G., 1960, Permian stratigraphy at Carlin Canyon, Nevada: Am. Assoc. Petroleum Geologists Bull., v. 44, no. 10, p. 1692-1703. Ferguson, H. G., Muller, S. W., and Roberts, R. J., 1951, Geology of the Winnemucca quadrangle, Nevada: U.S. Geol. Survey Geol. Quad. Map GQ-11. Nolan, T. B., Merriam, C. W., and Williams, J. S., 1956, The stratigraphic section in the vicinity of Eureka, Nevada: U.S. Geol. Survey Prof. Paper 276, 77 p. Roberts, R. J., Hotz, P. E., Gilluly, James, ano Ferguson, H. G., 1958, Paleozoic rocks of north-central Nevada: Am. Assoc. Petroleum Geologists Bull., v. 42, no. 12, p. 2813-2857. Ross, C. P., 1934, Correlation and interpretation of Paleozoic stratigraphy in south-central Idaho: Geol. Soc. America Bull., v. 45, p. 937-1000. ------- 1960, Diverse interfingering Carboniferous strata in the Mackay quadrangle, Idaho: in U.S. Geol. Survey Prof. Paper 400-B, p. B232-B233. ------- 1962a, Upper Paleozoic rocks in central Idaho: Am. Assoc. Petroleum Geologists Bull., v. 46, no. 3, p. 384-387. ------- 1962b, Stratified rocks in south-central Idaho: Idaho Bur. Mines and Geology Pamph. 125, 126 p. ------- 1962c, Paleozoic seas of central Idaho: Geol. Soc. America Bull., v. 73, p. 769-794. Scholten, Robert, 1957, Paleozoic evolution of the geosynclinal margin north of the Snake River plain, Idaho-Montana: Geol. Soc. America Bull., v. 68, no. 2, p. 151-170. Shannon, J. P., 1961, Upper Paleozoic stratigraphy of east-central Idaho: Geol. Soc. America Bull., v. 72, no. 12, p. 1829-1836. Silberling, N. J., and Roberts, R. J., 1962, Pre-Tertiary rocks in northwestern Nevada: Geol. Soc. America Spec. Paper 72, 58 p. Skipp, Betty A. L., 1961a, Interpretation of sedimentary features in Brazer Limestone (Mississippian) near Mackay, Custer County, Idaho: Am. Assoc. Petroleum Geologists Bull., v. 45, no. 3, p. 376-389. -------1961b, Stratigraphic distribution of endothyrid Forami- nifera in Carboniferous rocks of the Mackay quadrangle, Idaho: in U.S. Geol. Survey Prof. Paper 424-C, p. C239-C244. Thomasson, M. R., 1959a, Late Paleozoic stratigraphy and paleotectonics in central and eastern Idaho [abs.]: Geol. Soc. America Bull., v. 70, no. 12, p. 1687-1688. ------- 1959b, Paleocurrent and sedimentary structure studies in upper Paleozoic rocks in central Idaho labs.]: Geol. Soc. America Bull., v. 70, no. 12, pt. 2, p. 1687. Umpleb.v, J. B., Westgate, L. G., and Ross, C. P., 1930, Geology and ore deposits of the Wood River region, Idaho: U.S. Geol. Survey Bull. 814, 150 p. • See footnote, p. D4.Article 123 THRUST-FAULT RELATIONS IN THE NORTHERN COAST RANGES, CALIFORNIA By ROBERT D. BROWN, JR., Menlo Park, Calif. Abstract.—A klippe of volcanic rock which covers about 50 square miles in Glenn, Colusa, and Lake Counties provides evidence of a folded thrust fault that can be traced eastward into the Stony Creek fault zone. Rocks involved in the thrusting are of Late Jurassic and Cretaceous age. The structural relations described in this article involve strata in the northern Coast Ranges of California and in the foothills along the western border of the Sacramento Valley (fig. 123.1). The structural 0 25 50 75 100 MILES 1 _______I________I-------1________I Figure 123.1—Index map of part of northern California, showing location of Stonyford (S) and Lodoga (L) quadrangles. interpretations and the map data shown in figure 123.2 are based upon detailed mapping by the author and Ernest I. Rich in the Lodoga 15-minute quadrangle (Brown and Rich, 1961) and in the Stonyford 15-minute quadrangle (unpublished), which cover the area of figure 123.2 east of long. 122°45' W. (fig. 123.1), and upon reconnaissance mapping by the author in areas to the north, west, and south. STRATIGRAPHIC NOMENCLATURE The existing nomenclature for the rocks described is avoided in the discussion of structural relations. This is in part because the application of existing names would involve cumbersome stratigraphic discussions that are not germane to this article, but also because some of the existing names involve stratigraphic and structural concepts that are incompatible with the field relations described here. The following brief discussion is intended to demonstrate some of the difficulties inherent in the use of the existing names. Along the foothills west of the Sacramento Valley, Sandstone, siltstone, and conglomerate totaling more than 40,000 feet in thickness have been subdivided into formations chiefly on the basis of contained faunal elements. Although some formational units are valid rock-stratigraphic units (for example, the Venado Formation of Kirby, 1943), those with which this paper is most concerned, the Knoxville, Paskenta, and Horse-town, are defined in age terms and the definitions differ slightly from author to author. Redefinition in strict lithologic terms is difficult because of complex interfingering of lithologies and because many of the rocks of differing age are lithologically alike. Where lithologic distinctions are apparent and extensive, they generally do not coincide with boundaries based on geologic age, so that naming new lithologic units would seem preferable to redefining existing biostratigraphic units. However, for a rock sequence so widespread and so variable in lithologic detail as that along the ART. 123 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D7-D13. 1964. D7D8 STRATIGRAPHY AND STRUCTURAL GEOLOGY western border of the Sacramento Valley, new formation-name proposals should be based upon detailed mapping throughout several 15-minute quadrangles, and detailed geologic mapping of so large an area is not yet available. A more complex problem is involved in the sheared and deformed rock sequences that form the core of the Coast Ranges. These rocks have been assigned by previous workers to the Franciscan Formation, the Franciscan Group, or the Franciscan Series; by some the name has been used as a general term, but most recent workers have referred to the Franciscan Formation, thereby implying that the unit is a stratigraphic entity. Regardless of the rank of classification of the Franciscan, geologists who use the term agree that it includes a diverse suite of lithologies, characterized by graywacke, siltstone, semischist, phyllonite, chert, basic pillow lava, and flow breccia that are locally altered to greenstone, minor hypabyssal intrusive rocks, and gla-ucophane schist; serpentine and ultramafic bodies are commonly found in association with all these rocks. The thickness and internal stratigraphy of the gross unit are unknown, and even estimates of thickness are subject to errors of several orders of magnitude because of the lack of structural and stratigraphic detail in Franciscan terranes. Neither the upper nor the lower age limit of the Franciscan is known, but some of the glaucophane schists have been radiometrically dated at 130 to 150 million years ±5 percent (Art. 142; and Lee and others, 1963), and fossils indicative of an age range from Late Jurassic to Late Cretaceous have been described from Franciscan sedimentary rocks (Irwin, 1957). The Franciscan must therefore range in age from Late Jurassic or older to as young as Late Cretaceous. The contact relations of the Franciscan with younger, older, or contemporaneous, less deformed strata are generally obscured by major faults, and in areas where younger strata are found resting upon Franciscan rocks, the two are separated by major unconformities representing long periods of geologic time. In spite of the diversity in primary lithology and in degree of metamorphism, in spite of the uncertain age limits and little-known contact relations, and in spite of the general absence of internal stratigraphic and structural detail within the Franciscan, these rocks have long been considered a discrete lithologic unit. The concept of the Franciscan as a structural entity is a useful one for broad regional studies, but most workers have used the term in a stratigraphic sense (that is, Franciscan Formation) which does not seem to be justified by the field evidence that is now available. A further and more important objection to the use of the term Franciscan in local, detailed studies is that its use tends to compartmentalize the geologic problems into two fundamental units: the complex lithologic and structural assemblage of rocks referred to the Franciscan, and the more orderly rock sequences of the west border of the Sacramento Valley. The objections to this division are made obvious by the relations described in this paper, for in the part of northern California described here, the deformed rocks of the Coast Ranges (Franciscan) and those of the Sacramento Valley are intimately interrelated, both stratigraphic ally and structurally. Because of these interrelations, a twofold classification into Franciscan and Sacramento Valley rocks yields an oversimplified basic premise with which it is extremely difficult to reconcile the observed field relations. SUMMARY OF STRUCTURAL RELATIONS Evidence of major thrusting is found throughout a large part of the Stonyford quadrangle of the northern Coast Ranges. The complexly folded and crushed terrane involved in the thrusting is separated from the simple structures and uniform stratigraphy of the western border of the Sacramento Valley by the Stony Creek fault zone, a major north-trending tectonic feature that extends for more than 80 miles and is invaded by serpentinized peridotite along much of its length. East of the Stony Creek fault zone, siltstone, sandstone, and conglomerate are more than 40,000 feet thick and range in age from Late Jurassic (Tithonian) to Late Cretaceous (Campanian). These rocks form a north-trending outcrop belt as much as 15 miles wide, which extends along the entire western border of the Sacramento Valley. Within this belt the major structures trend northward, and although folding and faulting locally reverse the dip and repeat parts of the section, an eastward dip predominates and the oldest rocks crop out to the west along the Stony Creek fault zone. West of the Stony Creek fault zone, a more varied suite of rocks underlies the Coast Ranges. These consist in part of sandstone and siltstone that are indistinguishable from part of the rocks along the western border of the Sacramento Valley, but most of the sedimentary rock has been cataclastically metamorphosed to phyllonite or to cataclastic sandstone or semichist. The cataclastically deformed rocks may in places be partly recrystallized, but this is not generally apparent in the field; their most obvious lithologic characteristic is the crushing and shearing of mineral grains and bedding features along shear planes that parallel the bedding. Both the unmetamorphosed sedimentary rocks and their cataclastic counterparts contain one or more thin flows of basaltic pillow lava whichBROWN D9 grades locally to flow breccia and in places is overlain by manganiferous chert. Fossils of Early Cretaceous (Valanginian) age are found a few hundred feet strati-graphically below the basalt, and fossils of Late Jurassic (Tithonian) age are found a little lower in the section. A much thicker sequence of volcanic rock crops out on the summit and slopes of St. John Mountain and Snow Mountain and in the area between them (fig. 123.2). These rocks consist of pillow lava and flow breccia with interbeds of chert and volcanic-rich sedimentary rock and sills of diabase. Because they form a mountainous area of rugged topography with as much as 5,000 feet of local relief, they can be studied in three dimensions. Detailed (1:48,000) geologic mapping has disclosed that they form a klippe which rests in different places upon: (1) a friction carpet of chaotically sheared sedimentary debris; (2) bodies of serpentinized peridotite, sill-like in form and emplaced along a basal thrust plane; or (3) phyllonite and cataclastic sandstone, derived by bedding-plane shearing from siltstone and sandstone. Similar thrust sheets of sedimentary rock or phyllonite can also be recognized but are less obvious than the klippe of volcanic rock. EVIDENCE OF THRUSTING The volcanic klippe on Snow Mountain and St. John Mountain consists of several thousand feet of basaltic lava, breccia, and tuffaceous sedimentary rock that are folded along west-trending axes. Except near the basal fault plane, primary textures and structures in the volcanic rocks are well preserved, the rocks are little altered, and they contrast sharply with the altered and intensely sheared phyllonitic sedimentary rocks that crop out elsewhere in the area west of the Stony Creek fault zone. Both the structural trends and mapped lithologic units within the klippe are truncated at a folded thrust fault that underlies the volcanic rocks and separates them from intensely deformed sedimentary rocks below. Additional evidence for the thrust fault is found in the relationship of the volcanic rocks to intensely sheared chaotic sedimentary rocks northwest, north, and northeast of St. John Mountain. From Upper Nye Camp eastward, the fault surface at the base of the volcanic sequence traces a sinuous course along the northern face of St. John Mountain (fig. 123.3), decreasing in altitude from about 5,200 feet at Upper Nye Camp to about 2,000 feet at the eastern boundary of the klippe. Below the projection of this surface, a broad relatively low area that extends north and northeast for several miles is underlain by crushed and sheared siltstone and sandstone, so thoroughly faulted and mixed that they resemble shear-zone debris. These intricately faulted and folded rocks are cut by syntectonic quartz veins, and they contain scattered masses of resistant rock (fig. 123.3), a few feet to several hundred feet in diameter; many of the resistant masses show rounded altered borders and internal deformation features that suggest transport in a zone of tectonic movement. The resistant rock masses are predominantly volcanic rocks or chert and, except for the alteration and tectonic borders, they resemble rocks found on St. John Mountain and Snow Mountain; some of the masses, however, are of conglomerate, hard sandstone, or glaucophane schist. Although many of the resistant masses have clearly rotated or slumped downhill for short distances, others are clearly unaffected by landsliding and exhibit tightly sealed sinuous contacts with the enclosing crushed and sheared sedimentary rocks (figs. 123.4 and 123.5). In a few places, syntectonic quartz veins can be seen crossing the contacts between the country rock and the resistant masses. Most of the large resistant masses of volcanic rock and chert are found in topographically high positions within a few tens of feet beneath the projection of the fault surface that forms the base of the volcanic sequence on St. John Mountain. A few of these are so large and so near the projection of the thrust surface that they may actually be small klippen, but most appear to be tectonic inclusions that have been ripped from the upper plate and incorporated into the shear zone immediately below. True klippen probably are represented by several basalt outliers north and northeast of Upper Nye Camp, for some of these lie at or above the projected thrust surface, and in a few places thin zones of serpentine crop out at the base of the outliers. A larger klippe forms the upper part of Elephant Hill, where bedded chert, volcanic-rich sedimentary rock, and minor basalt with a relatively uniform strike and dip crop out. These rocks overlie chaotically deformed sedimentary rocks that crop out at lower elevations around the base of the hill, and the base of the chert-sediment-basalt sequence is marked, on the southwest side of Elephant Hill, by saline springs, abundant landslides, and in at least one place by lenses of serpentine. The chaotically deformed sedimentary rocks, together with the masses of chert, volcanic rock, and other resistant rock types, are interpreted as a flat-lying or gently east-dipping friction carpet, interposed between massive competent volcanic rocks above and less competent cataclastic sedimentary rocks below. The friction-carpet rocks superficially resemble landslide debris, and because they are weakly resistant to erosion and are lithologically and structurally heterogeneous they are subject to landslides. A few characteristicsDIO STRATIGRAPHY AND STRUCTURAL GEOLOGY WBBMM _ Upper ~ Nye Camp JElephant Hill SNOW MTN A A' Snow Mtn STONY CREEK FAULT ZONE Rocks below (west of)— Stony Creek fault zone Phyllonite and semischist with minor basaltic volcanic rocks Phyllonite and semischist EXPLANATION Rocks above (east of) Stony Creek fault zone 111 L) ^ j Sandstone, siltstone, and O l conglomerate r ml Siltstone and sandstone SSh Js, Tuffaceous siltstone Jb, Basaltic sandstone CE3 1 »£ Basaltic volcanic rocks with chert f (CU and minor sedimentary rocks J Rocks emplaced or formed along stony Creek fault zone or along related fault zone Peridotite and serpentine ES Friction-carpet debris Js Jv Sedimentary and volcanic rocks Occurs in tectonic lenses along peridotite-invaded parts of Stony Creek fault zone Contact Dashed where concealed -i—--------- ■ * Syncline _J)_____________ Fault D, down thrown side Thrust fault Mapped fault traces show location of eastern or upper limit of fault zones Figure 123.2.—Generalized geologic map and structure section showing thrust-fault relations near Stonyford, Calif. For simplicity: landslides, alluvium, and terrace deposits are not shown.BROWN Dll Figure 123.3—North face of St. John Mountain, showing approximate trace of thrust, with volcanic rocks above the fault trace and friction carpet below. Resistant mass in foreground is altered diabase. Ridge on right skyline, underlain by friction carpet, is drainage divide between coastal (Eel River) drainage and interior (Stony Creek-Sacramento River) drainage. of the friction carpet that help distinguish it from landslides are: 1. Abundance of resistant masses of exotic lithology, for which no local source is available (that is, intensely altered volcanic rock, glaucophane schist, conglomerate, and sandstone masses of distinctive lithology). 2. Intense sublaminar shearing and quartz veining in sedimentary rock that encloses the resistant masses (fig. 123.5). 3. Map distribution independent of topography: friction-carpet rock several hundred feet thick is found on major drainage divides where no higher slopes are available to supply landslide debris (fig. 123.2). 4. Map distribution indicative of a widespread tabular unit. In the area north and northeast of St. John Mountain, no clearly defined lower plate can be mapped, and in some parts of this area the friction carpet appears to be as much as 800 feet thick. In other areas, however, it is thinner and can be seen to grade down into relatively undeformed sedimentary rocks of the lower plate. Accurate thickness estimates of the blanket are difficult to make because these rocks are relatively weak and slump into topographic lows, because the lower contact is gradational, and because 716-626 O—64---2 Figure 123.4—Resistant mass of altered volcanic rock (light gray), in North Fork of Stony Creek. Mass is about 50 feet by 30 feet across and is completely enclosed in chaotically sheared sedimentary rock (dark gray). Sealed contacts, shearing, and quartz veining of both rock types suggest a tectonic origin. Figure 123.5—Detail of shearing and contact relations of resistant volcanic mass (light gray) shown in figure 123.4. Width of photographed area about 5 feet. through-going faults or folds are unrecognizable within the terrane underlain by friction carpet. Nevertheless, the thickness estimate of 800 feet in the area north and northeast of St. John Mountain appears to be valid, for stream bottoms 800 feet beneath the projected surface of the thrust fault are underlain by chaoticallyD12 STRATIGRAPHY AND STRUCTURAL GEOLOGY sheared, quartz-veined, sedimentary rock containing resistant masses with healed borders and crosscutting quartz veins. This abnormal thickness of friction-carpet debris may be related to a broad southeasttrending anticlinal flexure of the thrust surface. This flexure is reflected both in the mapped distribution of the thrust planes and in the plunging nose of serpentin-ized peridotite northwest of Stonyford. If the flexure predated or accompanied thrusting, tne lower confining pressure along its crest would account for the abnormal thickness of friction carpet there. Thrust relations are more difficult to recognize on the eastern and southern sides of the St. John Mountain-Snow Mountain klippe because on those sides the friction carpet is relatively thin and it, as well as other details along the base of the klippe, is obscured by landslide debris. Nevertheless, the tectonic nature of the basal surface of the volcanic sequence is still apparent, for where it can be examined the surface truncates structures and lithologies in the overlying rocks and is marked by intense shearing, saline springs, and sill-like bodies of serpentinized peridotite emplaced along the thrust zone. The western boundary of the klippe, also much obscured by landslides, has not yet been mapped in detail, but reconnaissance west of Snow Mountain has shown that the thrust surface dips gently westward for about 6 miles and extends nearly to the Eel River. The base of the klippe at its western boundary has a northward component of dip of about 200 feet per mile and is found at altitudes ranging from about 3,250 feet on the south to about 2,500 feet on the north. Along the base, numerous landslides issue from a shear zone containing lenses of serpentine and finely comminuted serpentine debris. Bold exposures of volcanic rock are found immediately above the shear zone, and cataclastically deformed sedimentary rocks crop out below. The Snow Mountain-St. John Mountain klippe therefore extends from the North Fork of Stony Creek nearly to the Eel River and covers an area of about 50 square miles. The basal thrust plane is gently arched along a north-trending axis that passes approximately through Snow Mountain, and the thrust surface exhibits about 4,000 feet of structural relief. RELATIONS OF THRUST FAULTING TO STONY CREEK FAULT ZONE Although the volcanic klippe is the most obvious, several other thrust sheets, involving chiefly sedimentary and cataclastically deformed sedimentary rocks, have also been mapped. These are less apparent because the rocks of both the upper and lower plates are lithologically similar, and they can be mapped best where friction-carpet debris or peridotite bodies can be traced along the fault zone. The amount and direction of dip of the mapped thrust surfaces varies locally, but all have an average inclination of a few degrees toward the east, and all appear to converge eastward and to join the Stony Creek fault zone. Detailed relations at the lines of juncture with the Stony Creek fault zone are difficult to interpret because this fault zone, like many of the thrust faults in the area, is invaded by peridotite and serpentinized peridotite. In most places the fault zone is vertical and trends northward, parallel to the bedding in the sedimentary rocks to the east, but near Stonyford the fault zone and the peridotite bodies along it swing sharply westward to outline an east-plunging structural basin. The stratigraphic and structural relations of the rocks within this basin and those to the west of it afford a key to the relation between the Stony Creek fault zone and the tectonic surface beneath Snow Mountain and St. John Mountain. West of Stonyford and east of the Stony Creek fault zone, an eastward-dipping volcanic sequence, identical with that on Snow Mountain and St. John Mountain, crops out. The upper contact of these volcanic rocks is concealed by alluvium and terrace deposits, but their mapped distribution, their structural attitude, and lithologic similarities of parts of this sequence to parts of the Late Jurassic sedimentary rock sequence to the east indicate that the volcanic rocks conformably underlie the sedimentary rocks, and locally are probably inter-bedded with them. On this basis, the Stonyford volcanic rocks are tentatively assigned a Late Jurassic age and are considered a part of the predominantly sedimentary rock sequence that underlies the western border of the Sacramento Valley. The volcanic sequence near Stonyford and that forming the klippe farther west are lithologically alike; both are floored by major fault surfaces, and where the two volcanic sequences are in closest proximity, serpentinized peridotite is emplaced along the fault boundaries of each. Mapping of the serpentinized peridotite shows that these thin sill-like bodies were originally a single continuous sheet that has either been folded along north-trending axes into a syncline and anticline or was emplaced along a pre-existing undulating shear surface. Erosion along the “anticlinal” axis has destroyed much of the original continuity, but in at least one place this continuity is preserved, and even where it is missing the map data are adequate to reconstruct it. The continuity of the serpentine together with the other stratigraphic and structural relations demonstrates that the thrust fault beneath Snow Mountain and St. John Mountain is at least a branch of the Stony Creek fault zone, and that it may be the structurally highest branch of that zone.BROWN D13 SUMMARY The described relations show that the Stony Creek fault zone is a bedding-plane thrust, near vertical for most of its mapped trace but flattening and branching into a complex set of thrust surfaces toward the west. In the Stonyford quadrangle, thrusting seems to be localized along the contact between a competent sequence of volcanic rocks and a thick overlying sequence of clastic sedimentary rock; locally, however, the volcanic rocks may be found in the upper plate, as they are in the two areas west of Stonyford. The rocks below, or west of, this thrust zone are predominantly cataclastically metamorphosed; those above, or east of the thrust are little deformed. Aside from the superimposed effects of cataclastic deformation the sedimentary rocks on opposite sides of the Stony Creek fault zone exhibit no marked differences in either lithology or age, but conglomerate, relatively common in the Upper Jurassic and Lower Cretaceous rocks of the upper plate, is rarely found below the thrust. Volcanic rocks are thickest and most widespread east of, or above, the Stony Creek fault zone, but at least one basalt flow of Early Cretaceous age is found in rocks west of, or below, the fault zone. Field evidence is still inadequate to establish the mechanism of thrusting or to prove either its amount or direction, but drag folds immediately above some of the thrust surfaces suggest that the upper-plate rocks have moved northward. This direction of movement agrees with the westerly trend of fold axes and faults in the Snow Mountain-St. John Mountain klippe, but additional data are needed before the history of thrust movement can be reconstructed. REFERENCES Brown, R. D., Jr., and Rich, E. I., 1961, Geologic map of the Lodoga quandrangle, Glenn and Colusa Counties, California: U.S. Geol. Survey Oil and Gas Inv. Map OM-210. Irwin, W. P., 1957, Franciscan group in Coast Ranges and its equivalents in Sacramento Valley, California: Am. Assoc. Petroleum Geologists Bull., v. 41, no. 10, p. 2284-2297. Kirby, J. M., 1943, Upper Cretaceous stratigraphy of the west side of the Sacramento Valley south of Willows, California: Am. Assoc. Petroleum Geologists Bull., v. 27, no. 3, p. 279-305. Lee, D. E., Thomas, H. H., Marvin, R. F., and Coleman, R. G., 1963, Isotope ages of glaucophane schists from Cazadero, California [abs.]: Geol. Soc. America 59th Ann. Mtg. Cordil-leran Sec., Berkeley, Calif., 1963, preprint, p. 42-43.Article 124 UPPER PLIOCENE MARINE STRATA ON THE EAST SIDE OF THE SAN JOAQUIN VALLEY, CALIFORNIA By R. L. KLAUSING and K. E. LOHMAN, Sacramento, Calif., and Washinrjton, D.C. Work done in cooperation with the California Department of Water Resources Abstract.—Recent drilling on the east side of the San Joaquin Valley in Tulare County has revealed marine Pliocene strata that are about 1,050 feet thick and contain diatoms. The diatom assemblage indicates a late Pliocene age and suggests a correlation with part of the San Joaquin Formation exposed on the west side of the valley. The rocks of Pliocene age occurring in the subsurface along the southeast side of the San Joaquin Valley have generally been assigned to the Etchegoin and Chanac Formations of Pliocene age (Weaver and others, 1944, and Sperber, 1952). Evidence reported in this article indicates that at Richgrove, just north of the Tulare -Kern County line 8 miles east of Delano (fig. 124.1), the entire marine Pliocene sequence should be assigned to the upper Pliocene and may be the equivalent of a part of the San Joaquin Formation which is exposed on the west side of the valley. In 1959 a core hole was drilled to a depth of 2,200 feet at Richgrove to obtain information on the lithology and the physical and engineering properties of the sediments in the depth penetrated by water wells. This core hole, 24/26-36A2, is in the NE^NE^ sec. 36, T. 24 S., R. 26 E., Mt. Diablo base line and meridian. As shown by the composite logs (fig. 124.2), the core hole penetrated unconsolidated continental deposits from the Sierra Nevada to a depth of 744 feet, a thick section of siltstone and claystone 1 containing a few sand interbeds from 744 to 1,900 feet and underlain by 300 feet of sand from 1,900 to 2,200 feet. Thirty core samples spanning the depth interval from 754 to 2,101 feet were examined by Lohman, who identified an assemblage of marine diatoms in the interval between 1,141 and 1,641 feet (see accompanying table). Most of the diatoms listed in the table have a wide stratigraphic range, occurring throughout the Pliocene and some even into the Miocene. However, two short-ranging species, Hemidvicus ovalis and Rhaphoneis iatula (Lohman, 1938, pi. 22), have not been reported from strata older than late Pliocene. Furthermore, 51 percent of the Richgrove species also occurs in the San Joaquin Formation in the Kettleman Hills. Therefore, Lohman concludes that the diatom assemblage in the Richgrove cores indicates a late Pliocene age for the interval from 1,141 to-1,641 feet and that it may be the equivalent of the late Pliocene San Joaquin Formation of the Kettleman Hills area. The core interval between 744 feet and 1,141 feet, though unfossiliferous except for a few unidentifiable fragments of marine diatoms at 1,003 feet, is fine grained and lithologically similar to the underlying diatomaceous zone. On the basis of lithologic similarity, this upper unit is also tentatively assigned to the upper Pliocene. A few marine invertebrate fossils were observed in the section between 1,641 and 1,900 feet. E. J. Moore (written communication, Nov. 10, 1960) identified 2 marine pelecypod genera, Macoma sp. and Cryptomya sp., in the interval between 1,733 and 1,767 feet. No definite age determination could be made. This marine section is lithologically similar to the overlying diatomaceous unit and is tentatively assigned to the Pliocene (?) in this article. The sand below 1,900 feet is correlated with the Santa Margarita Formation of Diepenbrock (1933), 1964. 1 Claystone as logged from core inspection at well, but particle-size analyses indicate mostly siltstone. ART. 124 IN U.S. GEOL. SURVEY PROF. PAPER. 475-D, PAGES D14-D17. D14KLAUSING AND LOHMAN o o g> o o D15 Figure 124.1.—Geologic section along north line of Kern County through the Richgrove core hole. *Santa Margarita Formation of Diepenbrock (1933).D16 STRATIGRAPHY AND STRUCTURAL GEOLOGY Generalized lithologic description (from core descriptions and electric logs) Sand, clean to silty, yellowish-brown; silt and sandy silt interbeds; several gravelly layers. Sand, silty and clayey, fine to coarse, yellowieh-brown; calcareous toward bottom,carbonaceous material. Sand, clean to silty, medium, loose, dark-yellowish-brown. Sift, sandy, fine to coarse, moderate-yellowish-brown; some gravel, calcareous zones. Sand, silty and clayey, fine to coarse, yellowish-brown to olive-gray. Clay and silt, sandy, firm to lithified, brownish- to olive-gray._________________ Sand, silty, greenish- to bluish-gray; thin interbeds of silt, clay, and claystone; some cementation. Claystone, siltstone, and silty sand, calcareous streaks, dark-bluish- to _________________________greenish-gray._______________________________________ Claystone and clay, silty, with sandy layers, greenish-gray. Claystone, silty to sandy, dark-greenish- to bluish-gray; thin interbeds of clay, silt, and loose sand, plant remains; rhyolitic ash zone at 1,058 ft. Sand, silty, clayey, fine to medium, dark-greenish- to bluish-gray. Claystone, silty, sandy, dark-greenish- to bluish-gray; thin interbeds of silt, sand, and fine gravel, diatomaceous, plant remains. Sand, silty, clayey, fine to medium, bluish-gray. Claystone and sand interbedded, firm to loose, dark-bluish- to greenish-gray; 10-ft sand layer at base, diatomaceous. Claystone, silty, sandy, bluish- to greenish-gray; interbeds of silt, sand, and fine gravel, more sand toward bottom, carbonaceous material, mollusk shells, crab remains, diatomaceous above 1,641 ft. Sand, silty, clayey, bluish-gray; claystone interbeds, mollusk shells and fish remains. Claystone, sandy, silty, bluish-gray; sand interbeds, mollusk shells, fish scales and teeth, carbonaceous material. Sand, slightly clayey, some gravel, fine to coarse, greenish-gray; carbonaceous material, shell fragments. Figure 124.2.—Composite logs of test hole 24/26-36A2, Richgrove, Calif.KLAUSING AND LOHMAN D17 Diatoms from U.S. Geological Survey core hole 24126-36AS! [C, common; F, frequent; R, rare; *, fresh and brackish-water adventitious forms. Identifications by K. E. Lohman] Diatoms Actinocyclus octonarius Ehrenberg____ Actinoptychu8 maculatus Grove and Sturt............................. senarius Ehrenberg.............. sp.............................. Aulacodiscu8 probabilis Schmidt...... Biddulphia aurita (Lyngbye) Brebis- son and Godey..................... Caloneis sp.......................... Coscinodiscus asteropmphalus Ehrenberg................................ cf. C. elegans Greville. ....... excentricus Ehrenberg........... kurzii Grunow................... cf. C. lineatus Ehrenberg....... vodulifer Schmidt............... obscuru8 Schmidt................ oculus-iridis Ehrenberg......... Coscinodiscus vetu8tis8imu8 Pan- tocsek............................ spp. (fragments)................ Diploneis ovalis var. oblongella (Naegli) Cleve*................... smithii (Brebisson) Cleve*...... Endictya robusta (Greville) Hanna and Grant......................... Fragilaria sp....................... Frickea lewisiana (Greville)* Heiden. Frustulia interposita (Lewis) DeToni* lIemidi8CU8 ovalis Lohman .......... Lithodesmium undulatum Ehrenberg. Melosira recedens Schmidt............ sulcata (Ehrenberg) Kiitzing.... Navicula pennata Schmidt............ placentula Cleve*............... sp.............................. Nitzschia etchegoinia Hanna and Grant............................. granulata Grunow................ cf. N. marginulata Grunow....... navicularis Grunow______________ Podo8ira montagnei Kiitzing......... Rhaphoneis amphiceros Ehrenberg_____ amphiceros var. rhombica Grunow. angulans Lohman................. fatula Lohman................... Stephanopyxis sp.................... Syndendrium sp...................... Trachyneis aspera (Ehrenberg) Cleve. Xanthiopyxis ovalis Lohman.......... R F R F F F F F R F F C R F R C F F F R F F C F R R F R C C R R C R F F F F R Depth (feet) San Joaquin Formation f Etchegoln 1 Formation ' i » 1,458 S3 3 § F F F F F — X X R F F F K X X R X R X X X F X X F F X X R .... X X F R X X x F F c c F R .... R R X R R X F F F X F F F F C X X R X R x F F F .... F .... X X R .... R R X X R X X F F F X X X X X' X x R R R X X of Miocene age, on the basis of electric-log correlations north from the Mt. Poso oil field (Park and Weddle, 1959, pi. II). In the Richgrove core hole the marine diatomaceous-siltstone unit is differentiated from the overlying continental deposits by a marked change in lithology. This lithologic change, which is clearly shown on the electric log for the core hole (fig. 124.2), can be recognized on electric logs as far west as well 24/26-31L2 (fig. 124.1). Westward from this point the top of the upper Pliocene is tentatively correlated with the upper Mya zone (Woodring and others, 1940, p. 28), which has been identified in numerous oil test wells. REFERENCES Diepenbrock, A., 1933, Mount Pozo oil field: Calif. Dept. Nat. Resources, Div. Oil and Gas, Summary of Operations, California Oil Fields, v. 19, no. 2, p. 12-29. Lohman, K. E., 1938, Pliocene diatoms from the Kettleman Hills, California: U.S. Geol. Survey Prof. Paper 189-C, 22 p., 4 pis. Park, W. H., and Weddle, J. R., 1959, Correlation study of southern San Joaquin Valley: Calif. Dept. Nat. Resources, Div. Oil and Gas, Summary of Operations, California Oil Fields, v. 45, no. 1, p. 33-34, 3 pis. Sperber, F. H., 1952, Oilfields of central San Joaquin Valley province in AAPG-SEPM-SEG Guidebook field trip routes, Joint Ann. Mtg. Am. Assoc. Petroleum Geologists, Soc. Econ. Paleontologists and Mineralogists, and Soc. Exploration Geophysicists, p. 153-155. Weaver, C. E., and others, 1944, Correlation of the marine Cenozoic formations of Western North America: Geol. Soc. America Bull., v. 55, p. 569-598. Woodring, W. P., Stewart, Ralph, and Richards, R. W , 1940, Geology of the Kettleman Hills oil field, Calif.; stratigraphy, paleontology, and structure: U.S. Geol. Survey Prof. Paper 195.Article 125 MIOCENE VERTEBRATES OF THE BARSTOW IN SOUTHERN CALIFORNIA By G. EDWARD LEWIS, Denver, Colo. Abstract.—The U.S. Geological Survey has adopted the name Barstow Formation and abandoned Rosamond Series for the rock unit that crops out from Alvord Mountain to Black Mountain, San Bernardino County, Calif. Between 2,000 and 2,800 feet of sedimentary and pyroclastic rocks make up the Barstow Formation. About the upper third of the Barstow constitutes one of two unnamed members; it yields late Miocene vertebrates (.Eucastor, Hemicyon, Merychippus sumani, Parucosoryx, Para-moceros). The upper half of the lower member yields middle Miocene vertebrates (Protohippus tehachapiensis, Brachycrus, Merriamoceros). No fossil vertebrates have been reported from the lowest third of the Barstow Formation. In the Mojave Desert of southern California, fossil vertebrates of Miocene age are found in the Barstow Formation, which includes at least two (perhaps more) members and local faunas. The two known faunas from the Barstow range in age from middle to late Miocene; some authorities (Schultz and Falkenbach, 1940, 1941, 1947, 1949) would extend this range to early Pliocene. Outcrops identified as the Barstow Formation have been mapped in the Mojave Desert, San Bernardino County, Calif., from lat 34°40' to 35°15' N., and from long 116°30' to 117°20' W. The belt of discontinuous outcrop stretches from Alvord Mountain on the east, through the Calico Mountains and Mud Hills, to Black Mountain on the west. NOMENCLATURE OF THE BARSTOW FORMATION More than a half a century ago, O. H. Hershey (1902a, facing p. 2) published a “Geological Reconnaissance Sketch Map of Southern California” that showed the Quaternary of the Hinckley-Barstow-Daggett area of the Mojave Desert and the underlying Tertiary Rosamond Series. This map clearly showed the Quaternary of the Mojave River valley at Barstow and for several miles eastward, where it overlies Tertiary rocks along the Santa Fe Railway. Two months later, Hershey formally named, defined, and designated the FORMATION type section for the Rosamond Series, and identified outcrops north of the railroad “from Hinckley station eastward . . . Barstow . . . Daggett” as “typical of the Rosamond Series” (Hershey, 1902b, p. 365-368). These are the rocks mistakenly called Barstow formation by Merriam (1915, p. 252-253; 1919, p. 441-442). But Hershey had already pre-empted the name Barstow for one of the Quaternary formations shown on his map; in the same publication where he named the Rosamond, he also named, defined, and designated the type section of the Barstow, spoken of both as formation and series, “in Mohave River valley . . . about one and one-half miles east of Barstow ... in ... a bluff just north of the railroad . . . this formation is . . . thin, overlies unconformably the earlier series [Rosamond], and remains generally in a horizontal position” (Hershey, 1902b, p. 369). For many years after Hershey’s original definition, geologists working in the Mojave Desert and specifically in the dissected, folded upper Miocene rocks of the Mud Hills and Rainbow Basin about 10 miles north of Barstow, consistently applied the name Rosamond to these sedimentary and pyroclastic rocks. In one recent published account, Durrell (1953, p. 24) says: “The Tertiary rocks of the Mud Hills, the higher beds of which are upper Miocene (Merriam, 1919, p. 454) in age, are generally known as the Rosamond Series, a name first applied by Hershey (1902, p. 349-372).” During this span of time, the same usage was followed by those later workers who were concerned with the mappable, physical, stratigraphic rock units, as exemplified by Baker (1911, p. 339-352, 357), Pack (1914, p. 146), Knopf (1918, p. 258), Hulin (1925, p. 42, 47, 48; 1934, p. 419-420), E. C. Simpson (1934, p. 395-396, 400), and Gardner (1940, p. 278, 281). Miller (1944) applied no formal name to this Miocene rock unit; he used Tertiary volcanics in his text and on his map of the area. Baker (1911, p. 341), in addition to ART. 125 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D18-D23. 1964. D18LEWIS D19 following Hershey’s usage of the Rosamond, also recognized the overlying Barstow of Hershey and used the term correctly. Wilmarth (1938, p. 1843) cited Hershey’s map legend (1902a) for the first usage of Rosamond, gave it an age assignment of “Tertiary (Miocene?),” and added (under his second contribution, 1902b) “uncon. underlies Barstow Series.” She (1938, p. 119) cited Hershey’s second contribution (1902b, p. 369-370) for the first usage of the term Barstow, complete with type section, and added “uncon. overlies Rosamond,” thereby recognizing the validity of Hershey’s Quaternary Barstow Formation. But, not realizing that two different formations were involved, she also cited Merriam (1915, 1919) in the seeming belief that his usage of Barstow was identical with Hershey’s. Merriam saw Hershey’s (1902a) map but evidently did not see Hershey’s formal description of the Rosamond and Barstow (1902b, p. 365-369) because Merriam (1911, p. 167-168) said: “According to a sketch map published by Hershey the point at which the collection was made would fall within what is designated by Hershey as the Rosamond Series. This series has not, however, been characterized in any way so that the nature of the formation is unknown. As geographic location is one of the important factors concerned, the horizon at which this collection was obtained may be referred to . . . as the Mohave beds.” Merriam 2 years later referred to “the fauna . . . representing a stage near Upper Miocene . . . from the typical Mohave beds north of Barstow,” (1913, p. 435-436). Then in 1915 he—inadvertently, it seems clear—began the misapplication of “the term Barstow . . . used for the beds containing the Upper Miocene fauna” (1915, p. 251). This mistaken application of the unavailable term Barstow was followed by those later workers who were not concerned with stratigraphy or other physical geology but with the morphology of the faunal elements and the correlation of the fauna collected from the area about 10 miles north of Barstow. Some of these paleontologists used the term Barstow formation: Merriam (1919, p. 441), Simpson (1933, p. 81, 90), Wood and others (1941, p. 12, 14), and Colbert (1942, p. 1494). Most have written of ages and faunas—not formations: Frick (1926, p. 27, 34) and Stirton (1930, p. 218) discussed fossils from the “Barstow Miocene.” Frick also used “Barstow Mio-Pliocene” (1933, p. 607); and Osborn (1936, p. 742) and Stock (1937, p. 398) referred to fossils from the “Barstow beds.” Stirton (1940, p. 178, 181-182) and Stirton and Teilhard de Chardin (1934, p. 279-282, 284) used Barstow with the stipulation that such names refer to a fauna and not to a formation. Schultz and Falkenbach did not use the term Barstow to refer to either a fauna or a formation, but spoke of the “Miocene deposits, north of Barstow” (1940, p. 223; 1947, p. 244) and the “Pliocene deposits, north of Barstow” (1941, p. 32). In short, when writing of the Tertiary rocks about 10 miles north of Barstow, many geologists, particularly the earlier writers, have avoided the use of Merriam’s Barstow Formation. Many paleontologists since Merriam have not used the term Barstow Formation for these same rocks, but have applied the name Barstow to the fauna contained therein. Others, including the Wood (1941) Committee, have used the name Barstow Formation for these rocks of Miocene age. The confusion regarding the use of these names caused the U.S. Geological Survey, in 1924, formally to abandon the name Rosamond Series and formally to adopt the name Barstow Formation for the Miocene rocks in the Barstow syncline. It has been so used by all Survey geologists since that time. Hewett (1954) has published a summary of the up-to-date knowledge of the general geology of the area involved. FAUNAS OF THE BARSTOW FORMATION The mammalian genera known from all zones of the Barstow Formation, according to published reports, are listed below, with tentative synonymy, and authors’ initials in parentheses according to the following scheme: Frick, 1926, 1933, 1937—(F); Hall, 1930— (H); Lewis in Byers, 1960—(L); Macdonald, 1949— (MD); Merriam, 1919—(M); Schultz and Falkenbach, 1940, 1941, 1947—(SF); Stirton, 1930, 1935—(ST); and Stock, 1937—(S). Talpa (L) Limnoecus (ST) Hypolagus (H) (L) = Lepust (M) Eutamias (H) Diprionomys (H) Peromyscm (H) Eucaslor (L) = Monosaulax (ST) Tomarctus (L) = Tephrocyon (M) Leptocyon (L) = Cants? (M) Hemicyon (F) Amphicyon (L) Aelurodon (L) (M) Bassariscus (L) Brachypsalis (L) Martes (L) Pseudaelurus (M) machaerodont (M) Gomphotherium (L) = Trilophodon (F) = Tetrabelodon? (M) Archaeohippus (M) Hypohippus (L) (M) Merychippus (L) (M) Dyseohyus (S) = Prosthennops? (M) Brachycrus (L) (SF) = Merycoehoerus? (M) Merychyus CL) (SF) Ustatochoerus (SF) Procamelus (L) (M)D20 STRATIGRAPHY AND STRUCTURAL GEOLOGY Aepycamelus (L) = Alticamelus (M) Hesperocamelus (L) (MD) Rakomeryx (F) = Dromomeryx (M) Ramoceros (F) = Merycodusl (M) Cosoryx (F) Meryceros (F) (L) = Merycodus (M) H. S. Mourning, J. R. Suman, C. L. Baker, and J. P. Buwalda carried on fieldwork north of Barstow between 1911 and 1913; their collections were those that Merriam first assigned to the Mohave Beds (1911) and later to his Barstow Forrnation (1915). These collections are at the Museum of Paleontology, University of California, Berkeley. By far the largest collection was made by the Frick Laboratory of the American Museum of Natural History, between 1920 and 1940. For these specimens no identifiable locations, geographic or stratigraphic, have been published, so that no comparisons can be made with specimens from other known horizontal (geographic) and vertical (rock unit) locations. For this reason, it is only possible to quote from Frick (1937): “It has been deemed advisable to consider as broadly contemporaneous the . . . ‘Middle Miocene’ to ‘Upper Pliocene’ . . . tentatively united as . . . the ‘Late’ Tertiary” (p. 5). “Merycodont remains are of frequent occurrence in the main portion of the Late Tertiary . . . Certain collections have been secured, from . . .restricted and well-defined horizons of . . . California” (p. 276). “The First, Second, Third and Fourth Divisions ... in the Barstow area” (p. 436) are mentioned on numerous pages of this publication, but with no explanation of the sequence or stratigraphic significance of these “Divisions.” Merycodonts are recorded as having been collected from many distinctively—but not geographically— named localities whose stratigraphic position in the formation is not mentioned: they are only said to have come either “from Barstow, California” (Frick, 1937, p. 331-333, 347-348, 367-368) or “Barstow localities” (p. 435-443). Although some few may be synonymous, there are 24 distinctively named localities: “bottom layer” (p. 368), “Green Hills deposit” (p. 332), “Lower Green Hills” (p. 443), “Green Hills, indefinite locality” (p. 333), “Hemicyon Stratum” (p. 368, 436, 438), “Hidden Quarry”* (p. 367), “Hidden Hollow Quarry” (p. 368, 440), “Leader Quarry”* (p. 367-440), “Mayday Quarry”* (p. 347), “New Year Quarry”* (p. 348, 436, 438), “North End”* (p. 347, 439), “North End, lower layer”* (p. 347), “Old Mill” (p. 368), “Skyline Quarry”* (p. 331, 347, 436, 437), above “Skyline Quarry” (p. 331), “Starlight Quarry”* (p. 347), “Steepside Quarry” (p. 332, 333, 347), “Valley View” (p. 368, 441), “Valley View Quarry” (p. 436), “White Layer” (p. 368), “White Operation” (p. 368, 436, 439), “Yermo” (p. 333), and “Yermo Quarry” (p. 442). The only information relative to stratigraphy is that the “Steepside Quarry” correlates with the “Green Hills horizon” (p. 332), and that the above listed localities marked with an asterisk (*) correlate with the “First Division” (Frick, 1937, p. 347-348, 367). Frick also published on “the occurrence in one well-defined stratum of two forms of Hemicyon” from the “Hemicyon Stratum, Barstow Miocene” (1926, p. 34). Schultz and Falkenbach described oreodonts from the Frick collections, and added the following information: “From the Miocene deposits, north of Barstow . . . Brachycrus remains in the Barstow area are recorded only from the . . . Steepside Quarry . . . Ness Quarry . . . Green Hills . . . Green Hills horizon or Second Division which underlies the later deposits including the ‘Hemicyon Stratum’ of the First Division” (Schultz and Falkenbach, 1940, p. 223-228); “Mery-chyus (Metoreodon) relictus fletcheri . . . from the Miocene deposits, north of Barstow . . . from ‘Red or Third Division’ ” (1947, p. 244-245); and“ Ustatochoerus medius mohavensis . . . from the Pliocene deposits, north of Barstow . . . from the ‘Hemicyon Stratum’ of the First Division” (1941, p. 32). Schultz and Falkenbach (1949, p. 80) also published a stratigraphic distribution chart. This work gives the best idea of the stratigraphic succession of some elements of the Frick collections. From 1950 to 1955 the U.S. Geological Survey was engaged in mapping much of the Mojave Desert; during this fieldwork a collection was made, chiefly by R. H. Tedford and R. L. Shultz, Jr. I studied this collection; one of my preliminary reports is a part of T. W. Dibblee, Jr.’s, report in preparation on the Opal Mountain and Fremont Peak quadrangles, California. Another of my reports is incorporated in F. M. Byers, Jr’s, report (1960) on the Alvord Mountain quadrangle, California. The specimens discussed in these reports are a part of the U.S. Geological Survey collections at the Denver Federal Center. Both Byers (1960) and Dibblee (written communication) have mapped several carefully defined Tertiary formations, and have made illustrated sections to clarify the stratigraphic relations of the fossil-bearing beds. One of the best continuous sections of the Barstow Formation has been measured and described by Durrell (1953, fig. 12, sections A-A' and B-B', pi. 4) in the Barstow syncline area about 10 miles north of Barstow. Section A-A' shows 1,955 feet of sedimentary and subordinate pyroclastic rocks within the formation, which is here overlain unconformably by Pleistocene sediments. Section A-A' does not include the basal beds of the formation, which are included in section B-B' about 4,000 feet west of section A-A'. The basal 139 feet ofLEWIS D21 granitic conglomerate overlies with angular unconformity an older formation of granitic breccia and rhyolitic tuff breccia. The total thickness of the formation included in Durrell’s two measured sections is 2,149 feet, much less than in the fossil-yielding area of outcrop farther west in the Barstow syncline. The total thickness is far from constant; it and the thickness of each individual bed have a considerable range from one part of the Barstow syncline to another. Most of the fossil vertebrates have been collected from 2% to 5 miles west-northwest of Durrell’s sections, where the aggregate thickness of outcrops of different parts of the formation adds up to at least 2,500 and perhaps as much as 2,800 feet. The total original thickness is not known; the highest beds, having been eroded, are either still being eroded or are overlain unconformably by Pleistocene sediments. There are at least two distinct faunal assemblages from the strata that have been called the Barstow Formation, but the lack of published stratigraphic data about the Frick collections makes a conclusive statement impossible. The U.S. Geological Survey collections from the Barstow syncline area all come from the uppermost third of the section exposed there, where many old diggings and large quarries were seen. These localities are all either above or only a few feet below (stratigraphically) a conspicuous bed of white tuff as much as 15 feet thick that crops out on the skyline as one travels northward up the canyon (EM sec. 23, T. 11 N., R. 2 W.) on Fossil Bed Road (USGS Opal Mountain, Calif., quadrangle). Dibblee (written communication) considers this tuff as marking the base of the upper, unnamed member of the Barstow Formation. Large quarries along the strike of this, Dibblee’s lower marker tuff, may represent Skyline and other localities of Frick’s (1937, p. 331, 347) First Division that yielded Meryceros (Meryceros), Cosoryz (Para-cosoryx), and Ramoceros (Paramoceros). Specimens referable to the first of these subgenera and to the Merychippus (Protohippus) species described by Mer-riam (1919, p. 479-507) are by far the commonest fossils from the uppermost third of the formation, where the U.S. Geological Survey collected specimens of 15 genera. Eucastor, Merychippus (Merychippus) su-mani, Paracosoryx, and Paramoceros are key fossils in this upper faunal assemblage. Merychippus sumani and Eucastor were collected, for example, at one locality 300 feet above the lower marker tuff. The age seems to be latest Miocene, most nearly equivalent to that of the Tonopah local fauna of Nevada and comparable to the oldest fauna from the Ogallala of Nebraska. Schultz and Falkenbach (1941, p. 32; 1949, p. 80) also assigned this fauna to the same stratigraphic position on the basis of a single specimen of Ustatochoerus medius mohavensis. A second distinct faunal assemblage seems to come from the middle third of the section that crops out in Coon Canyon of the Barstow syncline area (center NWM sec. 23, T. 11 N., R. 2 W., Opal Mountain, Calif., quadrangle). Here several old diggings in greenish rocks were seen. They may represent Steep-side and other localities of Frick’s Green Hills (1937, p. 99-105, 332-333) or Second Division (Schultz and Falkenbach, 1940, p. 223-228) that yielded Brachycrus buwaldi, Ramoceros (Merriamoceros), and Rakomeryx. This faunal assemblage is considered to be of the same age as the Sheep Creek local fauna according to Schultz and Falkenbach (1949, p. 80). This view is supported by the U.S. Geological Survey discovery (SEM sec. 34, T. 12 N., R. 4 E.) of Merychippus teha-chapiensis (Buwalda and Lewis, 1955, p. 148-150) 1,200 feet above the base of the formation, and of Brachycrus buwaldi (NW% sec. 30, T. 12 N., R. 4 E.) 1,700 feet above the base of the formation in the Alvord Mountain quadrangle, California. To be sure, the Merychippus tehachapiensis occurs 500 feet stratigraphically below the Brachycrus buwaldi and could possibly represent a fauna even older than the Brachycrus. But Buwalda and Lewis (1955, p. 150) believed that Merychippus tehachapiensis is close both in age and morphology to M. primus of the Sheep Creek local fauna of Nebraska. Perhaps the Frick collections coptain evidence either for or against a common stratigraphic position for the two species found in the Alvord Mountain area. With the evidence now available, we must assume that Brachycrus is the least rare fossil in the fauna from the middle third of the formation, and is a key fossil together with Ramoceros (Merriamoceros) and perhaps Merychippus tehachapiensis. Insofar as published information and known experience go, the lowest third of the formation has yielded no identifiable fossil vertebrates. Merychippus tehachapiensis, as already stated, could possibly represent a third fauna older than either of the two that are definitely recognized. A fourth fauna, intermediate in age between those of the Tonopah and Sheep Creek faunas, may conceivably be present. Schultz and Falkenbach (1947, p. 244-245) record one subspecies of oreodont whose age they consider intermediate between that of B. buwaldi and that of U. m. mohavensis: It is “the holotype . . . the only specimen known” of “ Merychyus (Metoreodon) relictus flet-cheri . . . from the Miocene deposits, north of Barstow . . . Red or Third Division.” They show the “stratigraphic distribution of . . . each genus and subgenus . . . species and subspecies” on a chart withD22 STRATIGRAPHY AND STRUCTURAL GEOLOGY corresponding epochs and Great Plains formations as follows (Schultz and Falkenbach, 1949, p. 80-81): “pliocene Valentine Ustatochoerus medius mohavensis miocene ‘Lower Merychyus (Metoreodon) Snake relictus fletcheri Creek’ ‘Sheep Creek’ Brachycrus buwaldi." This opinion of these authorities must be given great weight. But it is based on a solitary fragment of maxilla with broken and badly worn cheek teeth of a subgenus whose stratigraphic range is shown to be exactly equivalent to that of Brachycrus on the same chart. The “Red or Third Division” of Frick (1937) is a stratigraphic nomen nudum, and of no help in the description of the formation; presumably it is lower than the lower marker tuff but higher than the rocks that yield Brachycrus, and thus within the middle third of the formation. In resume, then, we have: (1) undisputed evidence that the uppermost third of the formation corresponds to a zone that contains an assemblage of fossil vertebrates usually thought to occupy an upper Miocene stratigraphic position; (2) suggestive evidence that the middle third of the formation corresponds to a zone that contains an assemblage of fossil vertebrates usually thought to occupy a middle Miocene stratigraphic position; and (3) empirical negative evidence that the lower third of the formation contains no identified fossil vertebrates (there are undetermined algae near the base) and is of unknown age not younger than middle Miocene. Differences of opinion do not exist as to the correlation with local faunas of the Great Plains, but only as to whether the uppermost fauna should be referred to the uppermost Miocene according to the usual practice (Merriam, 1919; Simpson, 1933; Wood and others, 1941) or to the lowermost Pliocene according to different authoritative opinion (Frick, 1933; Schultz and Falkenbach, 1941). Most of the diagnostic key fossils— equines, merycoidodontids, and antilocaprids—are exclusively New World forms that cannot be used for intercontinental correlation. Hemicyon and Gompho-therium are two genera that give direct evidence for intercontinental correlation of the youngest fauna; most authorities consider them to represent a late Miocene migration from the Old World to the New as part of “the most extensive faunal interchange between the early Oligocene and the latest Pliocene” (Simpson, 1947, p. 639). REFERENCES Baker, C. L., 1911, Notes on the later Cenozoic history of the Mojave Desert region in southeastern California: California Univ., Dept. Geol. Sci. Bull., v. 6, p. 333-383. Buwalda, J. P., and Lewis, G. E., 1955, A new species of Mery-chippus: U.S. Geol. Survey Prof. Paper 264-G, p. 147-152. Byers, F. M., Jr., 1960, Geology of the Alvord Mountain quadrangle, San Bernardino County, California: U.S. Geol. Survey Bull. 1089-A, 71 p. [1961]. Colbert, E. H., 1942, The geologic succession of the Proboscidea, in Osborn, H. F., Proboscidea; a monograph of the discovery, evolution, migration, and extinction of the mastodonts and elephants of the world, v. 2: New York, Am. Mus. Press, p. 1421-1521. Durrell, Cordell, 1953, Geological investigations of strontium deposits in southern California: California Div. Mines Spec. Rept. 32, 48 p. Frick, Childs, 1926, The Hemicyoninae and an American Tertiary bear: Am. Mus. Nat. History Bull., v. 56, p. 1-119. ------- 1933, New remains of trilophodont-tetrabelodont mastodons: Am. Mus. Nat. History Bull., v. 59, p. 505-652. ------- 1937, Horned ruminants of North America: Am. Mus. Nat. History Bull., v. 69, 669 p. Gardner, D. L., 1940, Geology of the Newberry and Ord Mountains, San Bernardino County, California: California Jour. Mines and Geology, v. 36, p. 257-292. Hall, E. R., 1930, Rodents and lagomorphs from the Barstow beds of southern California: California Univ., Dept. Geol. Sci. Bull., v. 19, p. 313-318. Hershey, O. H., 1902a, The Quaternary of Southern California: California Univ., Dept. Geol. Sci. Bull., v. 3, p. 1-29, pi. 1. ------- 1902b, Some Tertiary formations of Southern California: Am. Geologist, v. 29, p. 349-372. Hewett, D. F., 1954, General geology of the Mojave Desert region, California [Pt.] 1 in Chap. 2 of Jahns, R. H., ed., Geology of southern California: California Dept. Nat. Res., Div. Mines Bull. 170, p. 5-20. Hulin, C. D., 1925, Geology and ore deposits of the Randsburg quadrangle, California: California Div. Mines Bull. 95, 152 p. ------- 1934, Geologic features of the dry placers of the northern Mojave Desert: California Jour. Mines and Geology, v. 30, p. 416-426 [1935]. Knopf, Adolf, 1918, Strontianite deposits near Barstow, California: U.S. Geol. Survey Bull. 660-1. p. 257-270. Macdonald, J. R., 1949, A new Clarendonian fauna from northeastern Nevada: California Univ., Dept. Geol. Sci. Bull., v. 28, p. 173-194. Merriam, J. C., 1911, A collection of mammalian remains from Tertiary beds on the Mohave Desert: California Univ., Dept. Geol. Sci. Bull., v. 6, p. 167-169. ------- 1913, New protohippine horses from Tertiary beds on the western border of the Mohave Desert: California Univ., Dept. Geol. Sci. Bull., v. 7, p. 435-441. ------- 1915, Extinct faunas of the Mohave Desert, their significance in a study of the origin and evolution of life in America: Popular Sci. Monthly, v. 86, p. 245-264. -------1919, Tertiary mammalian faunas of the Mohave Desert: California Univ., Dept. Geol. Sci. Bull., v. 11, p. 437-585. Miller, W. J., 1944, Geology of parts of the Barstow quadrangle, San Bernardino County, California: California Jour. Mines and Geology, v. 40, p. 73-112.LEWIS D23 Osborn, H. F., 1936, Proboscidea; a monograph of the discovery, evolution, migration, and extinction of the mastodonts and elephants of the world, v. 1: New York, Am. Mus. Press, 802 p. Pack, R. W., 1914, Reconnaissance of the Barstow-Kramer region, California: U.S. Geol. Survey Bull. 541, p. 141-154. Schultz, C. B., and Falkenbach, C. H., 1940, Merycochoerinae, a new subfamily of oreodonts: Am. Mus. Nat. History Bull., v. 77, p. 213-306. ------ 1941, Ticholeptinae, a new subfamily of oreodonts: Am. Mus. Nat. History Bull., v. 79, p. 1-105. ------ 1947, Merycbyinae, a subfamily of oreodonts: Am. Mus. Nat. History Bull., v. 88, p. 157-286. ------ 1949, Promerycochoerinae, a new subfamily of oreodonts: Am. Mus. Nat. History Bull., v. 93, p. 69-198. Simpson, E. C., 1934, Geology and mineral deposits cf the Elizabeth Lake quadrangle, California: California Jour. Mines and Geolog3', v. 30, p. 371-415. Simpson, G. G., 1933, Glossary and correlation charts of North American Tertiary mammal-bearing formations: Am. Mus. Nat. History Bull., v. 67, p. 79-121. Simpson, G.G., 1947, Holarctic mammalian faunas and continental relationships during the Cenozoic: Geol. Soc. America Bull., v. 58, p. 613-687. Stirton, R. A., 1930, A new genus of Soricidae from the Barstow Miocene of California: California Univ., Dept. Geol. Sci. Bull., v. 19, p. 217-228. ------- 1935, A review of the Tertiary beavers: California Univ. Dept. Geol. Sci. Bull., v. 23, p. 391-458. -----— 1940, Pbylogeny of North American Equidae: California Univ., Dept. Geol. Sci. Bull., v. 25, p. 163-197. Stirton, R. A., and Teilhard de Chardin, Pierre, 1934, A correlation of some Miocene and Pliocene mammalian assemblages in North America and Asia with a discussion of the Mio-Pliocene boundary: California Univ., Dept. Geol. Sci. Bull., v. 23, p. 277-290. Stock, Chester, 1937, A peccary skull from the Barstow Miocene of California: Natl. Acad. Sci. Proc., v. 23, p. 398-404. Wilmarth, M. G., 1938, Lexicon of geologic names of the United States: U.S. Geol. Survey Bull. 896, 2396 p. Wood, H. E., 2d, and others, 1941, Nomenclature and correlation of the North American continental Tertiary: Geol. Soc. America Bull., v. 52, p. 1-48.Article 126 THIRSTY CANYON TUFF OF NYE AND ESMERALDA COUNTIES, NEVADA By D. C. NOBLE, R. E. ANDERSON, E. B. EKREN, and J. T. O’CONNOR, Denver, Colo. Work done in cooperation with the U.S. Atomic Energy Commission Abstract.—Rhyolitic ash-flow and air-fall tuffs assigned to a new formation, Thirsty Canyon Tuff, of Pliocene age crop out from Beatty to Goldfield. The formation is subdivided into five newly defined formal members—the Spearhead, Trail Ridge, Dry Lake, Gold Flat, and Labyrinth Canyon Members—and an informal upper member. A sequence of rhyolitic ash-flow and air-fall tuffs here named the Thirsty Canyon Tuff crops out over a large area in the general vicinity of Beatty and Goldfield, Nev. The presently known extent of the formation, based on recent geologic mapping in twelve 7'(-minute and three 15-minute quadrangles and on reconnaissance in parts of four additional 15-minute quadrangles, is shown on figure 126.1. The Thirsty Canyon Tuff is best exposed and attains its maximum thickness north-northeast of Beatty in the general vicinity of Thirsty Canyon (fig. 126.1), for which the formation is here named. The formation is here subdivided into five formal members and an informal upper member. The formal members are, from bottom to top, the Spearhead, Trail Ridge, Dry Lake, Gold Flat, and Labyrinth Canyon. A complete stratigraphic section of the Thirsty Canyon Tuff is nowhere exposed; therefore, only a type area located in the headward part of Thirsty Canyon (fig. 126.1) is designated. Type localities are designated for the members. In addition to being practical cartographic units, the five formal members of the Thirsty Canyon Tuff are genetic rock units. Welding and cooling characteristics indicate that each member was deposited within a relatively restricted interval of time, and that each member cooled to a greater or lesser degree as a single unit. The Spearhead Member is a composite sheet; the Trail Ridge Member a multiple ash-flow simple cooling unit; the Dry Lake Member a simple cooling unit, probably a single flow; the Gold Flat Member a compound cool- ing unit; and the Labyrinth Canyon Member a single ash-flow simple cooling unit. (The terminology of ash-flow cooling units is that of Smith, 1960.) Although in most places the members are conformable or paraconformable, locally they are separated by lenses of conglomerate and sandstone and by intertonguing lava flows. These rocks are excluded from the Thirsty Canyon Tuff. The Thirsty Canyon Tuff shows much lateral variability, chiefly because of the marked topographic irregularity of the surface on which it was deposited. The formation as a whole and the individual units within the formation thin and thicken laterally in a complex and often unsystematic manner. Pronounced lateral variation in degree of welding and in the amount and character of devitrification and vapor-phase crystallization are associated with changes in thickness of individual members and of smaller units within members. Preliminary studies indicate that the rocks of the Thirsty Canyon Tuff have similar mineralogic and chemical compositions. Phenocrysts, which comprise 5 to 30 percent of the rock, include soda-rich sanidine, pigeonite, a green clinopyroxene probably containing appreciable amounts of aegirene and (or) heden-bergite, fayalite, strongly absorptive brown amphibole, and zircon. Rare quartz phenocrysts have been found only in the Gold Flat Member. Chemically the rocks are soda rhyolites and pantellerites. The chemical and mineralogical similarities strongly suggest that the members of the Thirsty Canyon Tuff are comagmatic. Despite these similarities, the members differ appreciably both in general appearance and petrography. Also, there is significant lateral and vertical variation in phenocryst, lithic-fragment, and pumice content within the individual members. D24 ART. 126 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D24-D27. 1964.NOBLE, ANDERSON, EKREN, AND O’CONNOR D25 Figure 126.1—Generalized map showing presently known surface and subsurface distribution of the Thirsty Canyon Tuff (stippled). The formation may extend beyond queried contacts. Type area of the Thirsty Canyon Tuff is enclosed by dashed line. Rectangle shows area covered by figure 126.2. The Thirsty Canyon Tuff unconformably overlies the Rainier Mesa Member of the Piapi Canyon Formation (Poole and McKeown, 1962). Although it postdates most of the Tertiary faulting of the area, it is cut by numerous north-south trending normal faults and, south of Goldfield, has undergone strong hydrothermal alteration. The formation is deeply incised by canyons and has been eroded from large areas. In most areas it is the youngest sequence of volcanic strata, but locally it is overlain by intermediate to mafic lavas. The Thirsty Canyon Tuff is dated, primarily on isotopic evidence, as Pliocene in age. A potassium-argon age of approximately 7.5 m.y. (R. W. Kistler, written communication, 1963) has been obtained on sanidine from the lower part of the Spearhead Member. Potassium-argon ages of less than 13 m.y., the Miocene-Pliocene boundary of Kulp (1961), have also been obtained on six units stratigraphically below the Thirsty Canyon Tuff. In addition to the potassium-argon age of 7.5 m.y., the structural involvement, hydrothermal alteration, and degree of erosion that theD26 STRATIGRAPHY AND STRUCTURAL GEOLOGY Thirsty Canyon Tuff has undergone indicate that the formation is of Pliocene rather than Pleistocene age. SPEARHEAD MEMBER Ransome (1909) applied the name Spearhead Rhyolite to an ash-flow tuff that crops out west of Goldfield. The unit is “. . . well exposed at Rabbit Spring . . . and in Pozo Canyon” (Ransome, 1909, p. 71), about three-quarters of a mile southwest of the center of Goldfield (fig. 126.1), where it is overlain by basalt. Regional mapping shows that Ransome’s Spearhead Rhyolite is part of a large composite sheet forming the lower part of the Thirsty Canyon Tuff. Where described by Ransome, the Spearhead Rhyolite is a simple cooling unit 20 to 80 feet thick. However, traced to the south, this cooling unit incorporates additional ash flows and in a short distance becomes a compound cooling unit several hundred feet thick. South of Gold Flat this cooling unit is underlain by another compound cooling unit of similar composition that overlies a sequence of rhyolitic lava flows and minor tuffs informally termed the rhyolites of Pillar Spring. The break between the 2 units becomes progressively less distinct to the south, and disappears completely approximately 8 miles south of Black Mountain. Because of this merger, the lower compound cooling unit is also included in the Spearhead Rhyolite, which is here reassigned as the Spearhead Member of the Thirsty Canyon Tuff. Where the two compound cooling units of the Spearhead Member are distinguishable, they are informally designated the upper and lower parts of the member. The upper part is distinguished by numerous red to brown lenses of scoriaceous to massive devitrified glass that are commonly several feet in diameter and as much as a foot thick. The rocks of the Spearhead Member are highly variable in lithology, ranging from shard tuffs to tuffs containing abundant pumice, lithic fragments, and phenocrysts, and from air-fall tuffs and poorly welded ash-flow tuffs to densely welded vitric and devitrified ash-flow tuffs. Very commonly the tops of both the upper and lower parts of the Spearhead Member are composed of lithic-, pumice-, and crystal-poor shard tuff. The present volume of the Spearhead Member exceeds that of the other members of the Thirsty Canyon Tuff combined. The lower part of the member reaches a maximum thickness of about 400 feet in the lower part of Thirsty Canyon. East of Stonewall Mountain, rocks of the upper part of the Spearhead Member locally reach a thickness of almost 300 feet. TRAIL RIDGE MEMBER The Trail Ridge Member is here named for Trail Ridge (fig. 126.2). In most outcrops the member in- cludes a basal bed of air-fall pumice, generally less than 5 feet thick but locally as much as 30 feet thick. In the vicinity of Black Mountain and Thirsty Canyon the bidk of the member consists of lithic-rich moderately to densely welded ash-flow tuff, which is capped by a thin unit of densely-welded shard tuff containing very few pumice fragments or lithic inclusions and few phenocrysts. To the north the lithic-rich part of the member is absent, and the member consists entirely of shard tuff. At the type locality (fig. 126.2), north of Black Mountain (lat 37°21' N., long 116°39' W.), about 7 miles west of Trail Ridge, the Trail Ridge Member is well exposed for several miles along the walls of a north-south-trending canyon. Here the member, approximately 100 feet thick, overlies the upper part of the Spearhead Member and is paraconformably overlain by the Gold Flat Member. Both the lower lithic-rich zone and the upper shard tuff zone are well developed. The Trail Ridge Member locally reaches a thickness of almost 200 feet in the general vicinity of Thirsty Canyon, but in most places is much thinner. The member is essentially coextensive with the Spearhead Member; it is generally densely welded and forms the caprock of a large part of Pahute Mesa. DRY LAKE MEMBER The Dry Lake Member, as here named for Dry Lake, crops out northeast and southwest of Dry Lake and north of Tolicha Peak (fig. 126.2). The member, which nowhere exceeds 30 feet in thickness, is lithologically distinct from both the underlying Trail Ridge Member and the overlying Gold Flat Member. Approximately 25 percent of a typical outcrop is composed of black to brown coarsely vesicular blocks of pumice as much as 3 feet in diameter. Only the uppermost part of the member is devitrified. At the type locality (lat 37°26' N., long 116°41' W., fig. 126.2) a section extending from the Trail Ridge Member through the Gold Flat Member is exposed. GOLD FLAT MEMBER The Gold Flat Member is here named for outcrops near Gold Flat. However, its type locality is about 10 miles south of Gold Flat in the upper reach of Thirsty Canyon (lat 37°15' N., long 116°36' W., fig. 126.2). Here approximately 170 feet of the member is exposed on the walls of Thirsty Canyon and its tributaries. The base of the member consists of several inches to about 8 feet of densely to partially fused light-brown tuff probably of air-fall origin. This unit is succeeded by densely welded generally devitrified dark-green to bluish-gray tuff as much as 20 feet thick which passes abruptly upward into densely to moderately welded yellow-brown or brownish-red devitrified tuff.D27 NOBLE, ANDERSON, EKREN, AND O’CONNOR 116°45' 116°30' EXPLANATION Labyrinth Canyon and Gold Flat Members miiiiiiiiiiiiiiiiiiH Gold Flat and Dry Lake Members Labyrinth Canyon Member Gold Flat Member < Dry Lake l. Member Figure 126.2.—Map showing the distribution of the Dry Lake, Gold Flat, and Labyrinth Canyon Members and type localities of the Trail Ridge (TR), Dry Lake (DL), Gold Flat (GF), and Labyrinth Canyon (LC) Members. The Gold Flat Member is characterized by abundant lithic fragments and by complexly twinned eubedral phenocrysts of alkali feldspar as much as 3 cm long, in addition to the soda-rich sanidine crystals 1 to 5 mm in diameter that are typical of the Thirsty Canyon Tuff. A distinctive ash flow, characterized by very abundant phenocrysts and by numerous cognate ejecta as much as several feet in diameter and composed dominantly of phenocrysts, forms the top of the member in the vicinity of the type locality. The member reaches a maximum thickness of about 200 feet in several places in the Thirsty Canyon-Black Mountain area. LABYRINTH CANYON MEMBER The Labyrinth Canyon Member, here named for Labyrinth Canyon, crops out west, north, and east of Black Mountain (fig. 126.2). At the type locality (lat 37°17' N., long 116°41' W.) in the headward part of Labyrinth Canyon (fig. 126.2), the Labyrinth Canyon Member reaches its maximum thickness of approximately 50 feet. The lower 5 to 20 feet of the member is composed of slightly to moderately welded light-gray or light-brown vitric ash-flow tuff containing about 20 percent white pumice fragments. The basal vitric zone is sharply overlain by pink to bluish-gray devitri-fied tuff. UPPER MEMBER The upper member consists of two beds of white to gray pumice-rich air-fall(?) tuff 1 to 8 feet thick that crop out on the dip slope of Boulder Cuesta (fig. 126.2), approximately 2 miles southeast of the summit of Black Mountain. The upper member overlies the Gold Flat Member. Although the Labrinth Canyon Member and the upper member are nowhere in contact, colluvial gravels below the upper member appear to overlie the Labyrinth Canyon Member. REFERENCES Kulp, J. L., 1961, Geologic time scale: Science, v. 133, p. 1105-1114. Poole, F. G., and McKeown, F. A., 1962, Oak Spring Group of the Nevada Test Site and vicinity, Nevada: Art. 80 in U.S. Geol. Survey Prof. Paper 450-C, p. C60-C62. Ransome, F. L., 1909, Geology and ore deposits of Goldfield, Nevada: U.S. Geol. Survey Prof. Paper 66, 25S p. Smith, R. L., 1960, Zones and zonal variations in welded ash flows: U.S. Geol. Survey Prof. Paper 354-F, p. 149-159. 716-626 0-64-Article 127 ST. KEVIN GRANITE, SAWATCH RANGE, COLORADO By OGDEN TWETO and ROBERT C. PEARSON, Denver, Colo. Abstract.—Granitic rocks of a batholith in the Sawatch Range were in part intrusive and in part generated in place. They constitute a unit that originated independently of the Silver Plume Granite in the Front Range, to which they were once speculatively assigned. The name St. Kevin Granite is here given to the rocks of a small Precambrian batholith and related plutons in the northern Sawatch Range and part of the adjoining Mosquito Range near Leadville, Colo. In the past, these rocks have been called Silver Plume Granite (Stark, 1935) or Silver Plume(?) Granite (Behre, 1953, p. 20-22). This correlation with the well-known Silver Plume Granite of the Front Range assumes an identity that has not been proved. The two granites have in common only the facts that they are both biotite-muscovite granites that have conspicuous trachytoid porphyritic facies, and that each is the youngest major Precambrian granite in its respective area. They differ in minor degree in petrographic details, range of composition, and characteristic accessory minerals, and, in part, in form of the intrusive bodies and conditions of emplacement. Most importantly, however, they seem to have been derived from separate bodies of magma. Isotopic-age data currently available suggest that the two granites are of the same general age, but data for precise comparisons are not yet available. The St. Kevin Granite occurs principally in a batholith about 25 miles long and 12 miles wide centered near the crest of the Sawatch Range west of Leadville. About a third of the batholith lies in the Holy Cross quadrangle (fig. 127.1), where it has been mapped in detail; the remainder has been examined only in reconnaissance. Many satellitic bodies of the granite lie in the mapped areas north and east of the batholith; others probably exist in the unmapped areas to the south and west. The granite takes its name from St. Kevin Lake and the adjoining St. Kevin mining district (fig. 127.1). 106°45' 106° 15' Figure 127.1.—Generalized outline of the St. Kevin batholith. Where it has been mapped in detail, the batholith is characterized by extremely intricate borders, as indicated in figure 127.2. Similar complex interfingering of granite and the wallrocks of metasedi-mentary gneisses is known to exist in the southern part of the batholith, near Mount Massive, and in places, at least, in the western part. The only part of the batholith relatively free of “islands” and projecting fingers of gneiss is an area 3 or 4 miles in diameter near its center. Although many of the bodies of gneiss within the batholith are clearly xenoliths, many are in structural continuity and thus appear to D28 ART. 127 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D28-D32. 1964.TWETO AND PEARSON D29 39° 106°21' Figure 127.2.—Generalized map of part of the St. Kevin batholith. be parts of large, attached bodies only incipiently sundered by the granite. Contacts between the granite and the metasedimentary gneisses are predominantly concordant but in detail are discordant in many places. Some are of knife-edge sharpness, particularly in crosscutting bodies, but typically the contacts are gradational. In places, sill-like bodies of granite a few inches to tens of feet wide alternate with strips of gneiss of the same general dimensions through contact zones hundreds of feet wide. Elsewhere the transition takes place in a broad pegmatitic zone that consists on one side of pegmatite and granite in complex mixture, and on the other, of pegmatite and gneiss in equally complex arrangement. In many places the gradation from granite to gneiss is almost imperceptibly gradual for distances of tens or hundreds of feet and then is completed abruptly in distances of only inches or a few feet. In the abrupt parts of such zones, gneiss is peppered with feldspar and muscovite crystals typical of the granite. In the remainder of the zone, the granite is streaked with ghosts or delicate screens of partly granitized gneiss. The screens outline structural features that commonly are in strict continuity with those in the adjoining gneiss. Locally, streaky granite of this type loses its organized fold structure in a mass of complex swirls that presumably mark the site of liquid flow. The St. Kevin Granite comprises many textural and compositional varieties of granitic rocks. Four main facies—(1) trachytoid hybrid, (2) normal, (3) grano-dioritic, and (4) fine grained—were distinguished in mapping, but each of these has variants, and all intergrade. A fifth major facies, called Hell Gate Porphyry by Stark and Barnes (1932, p. 34), occurs in the unmapped western and southern parts of the batholith. As recognized later by Stark and Barnes (1935, p. 474), this rock is merely a textural facies of what is here called the St. Kevin Granite, and it therefore does not warrant a formal name. We shall refer to the rock hereafter as the coarse porphyritic facies of the St. Kevin Granite. The trachytoid hybrid facies is a coarse-grained inhomogeneous rock that appears at first glance to consist principally of closely packed parallel crystals of microcline %-1 inch long and coarse flakes of muscovite. The feldspar crystals lie in a matrix that ranges from little-altered biotite gneiss through partly reconstituted gneiss to fine- or medium-grained granitoid rock. The composition of the hybrid rock varies with the content of gneiss, but most of it has the composition of quartz monzonite (see accompanying table and fig. 127.3). The hybrid facies typically contains abundant inclusions of gneiss in addition to the films and fine laminae between the feldspar crystals. Irregular biotite- and sillimanite-ricli clots sheathed in coarse muscovite are especially characteristic. The hybrid facies is largely confined to the border zone of the batholith, where it forms long streaks ranging from a few inches to hundreds of feet in width. It characteristically lies between gneiss on one side and some other facies of the St. Kevin Granite on the other although it occurs also as streaks within other facies. Many bodies of gneiss within the batholith have selvages of the trachytoid hybrid granite, which clearly is a reaction product. In such bodies, the foliation of the hybrid granite generally parallels that of the gneiss, but it parallels the contacts wherever they transect the foliation of the gneiss. The normal facies of the St. Kevin is a light-gray to light-pink even-grained to markedly porphyritic biotite-muscovite granite or quartz monzonite (see table and fig. 127.3). The microcline in it is generally in prismatic or tabular subhedral crystals, even in the nonporphyritic facies; in many places, but especially near contacts and in small bodies, these crystals are oriented in trachytoid texture. In many places the normal facies has border zones a few feet to half a mile wide characterized by abundant thin mica-rich wisps and streaks, many of which are clearly remnants of earlier gneiss. Elsewhere, the granite appears com-positionally uniform and uncontaminated, but the widespread occurrence of accessory sillimanite, and locally, of garnet and packets of metamorphic(?) biotite, suggests a derivation in part at least from older rocks. Nowhere has the granite been found to containD30 STRATIGRAPHY AND STRUCTURAL GEOLOGY Chemical analyses, norms, and modes of St. Kevin Granite from Holy Cross quadrangle, Colorado Normal Trachy- Fine- Grano- Grano- Normal facies toid grained dio- dio- facies 1 composite 2 hybrid facies3 facies 4 ritic facies 8 ritic facies« Chemical analyses (weight percent) Si02 71. 15 72. 41 71. 20 67. 67 66. 75 66. 46 AI2O3 14. 54 14. 40 14. 71 15. 16 15. 18 15. 02 F02O3 . 95 . 42 . 64 1. 32 1. 29 2. 53 FeO 1. 35 1. 44 1. 72 2. 41 3. 10 1. 91 MgO - . 52 . 48 . 73 . 85 1. 12 1. 04 CaO 1. 02 . 77 . 75 1. 86 2. 14 2. 13 Na20 2. 99 2. 76 2. 56 2. 84 3. 59 3. 15 K20 5. 94 5. 66 5. 22 5. 75 3. 95 4. 91 H20+ . 40 . 71 1. 09 . 43 . 63 . 62 h2o- . 02 . 06 . 11 . 00 . 03 . 10 Ti02 . 34 . 23 . 31 . 70 . 83 . 84 p2o5 . 12 . 20 . 18 . 27 . 35 . 36 MnO . 03 . 03 . 04 . 04 . 05 . 05 C02 . 22 . 07 . 07 . 03 . 30 . 27 Cl . 02 . 01 . 01 . 02 . 03 . 02 F . 13 . 06 . 05 . 34 . 19 . 24 S . 01 . 01 . 00 . 01 . 08 . 08 Sr . 02 . 02 . 00 BaO . 05 . 05 . 07 . 13 . 11 . 09 Subtotal 99. 80 99. 79 99. 48 99. 83 99. 72 99. 82 Less O . 06 . 03 . 02 . 15 . 13 . 15 Total 99. 74 99. 76 99. 46 99. 68 99. 59 99. 67 Norms Quartz 28. 56 31. 32 33. 18 23. 70 25. 14 25. 44 Orthoclase 35. 03 33. 36 30. 58 34. 47 23. 35 28. 91 Albite 25. 15 24. 12 21. 48 24. 10 30. 39 26. 72 Anorthite 3. 06 3. 34 3. 34 6. 67 5. 56 5. 56 Corundum 2. 04 2. 65 3. 67 1. 73 2. 96 2. 45 Magnetite 1. 39 . 70 . 93 1. 86 1. 86 3. 71 Ilmenite . 61 . 46 . 61 1. 37 1. 52 1. 52 Enstatite - - 1. 30 1. 20 1. 80 2. 10 2. 80 2. 60 1. 19 1. 56 2. 11 2. 11 3. 30 . 67 1. 01 1. 01 Fluorite . 23 . 62 . 16 . 39 . 50 . 20 . 20 . 70 . 60 . 36 . 36 Modes (volume percent) Quartz 27. 7 29. 3 26. 0 28. 5 28. 4 38. 7 26. 1 33. 2 17. 7 23. 2 Plagioclase 25. 0 25. 4 29. 5 34. 9 36. 6 Biotite and 3. 6 9. 1 7. 0 13. 9 8. 1 Muscovite 4. 3 9. 1 3. 1 3. 2 1. 9 Magnetite and 0. 6 Tr. 1. 1 1. 3 Apatite Tr. 1. 0 Tr. . 2 . 3 0. 1 Tr. Tr. . 2 . 1 Tr. Tr. Tr. Tr. Tr. Tr. Tr. Tr. Tr. Tr. Tr. 1 From dump at west portal TIagcrman tunnel, southwest corner of quadrangle. Analyst, V. C. Smith. (Field No. 48T57, analytical laboratory No. 1)1773.) 2 Composite sample from head of Longs Gulch. Analyst, M. Seerveld. (Field No. 69T55, analytical laboratory No. D1417.) 3 From 0.05 miles southwest of Deckers Lake. Analyst, M. Seerveld. (Field No. 72T55, analytical laboratory No. D1419.) « From west side St. Kevin Lake. Analyst, M. Seerveld. (Field No. 142T56, analytical laboratory No. D1418.) * From elevation 10,500 feet on old Colorado Midland Railroad grade on east side of Busk Creek. Analyst, Paula Montalto. (Field No. 25T57, analytical laboratory No. E-2208.) 6 From prospect pit at elevation 11,025 feet on east face of Bald Eagle Mountain. Analyst, Paula Montalto. (Field No. 30T57, analytical laboratory No. E-2209.) Quartz EXPLANATION Quartz B Figure 127.3.—A, Plot of modal quartz, K feldspar, and plagioclase in Sr. Kevin Granite. B, Plot of normative quartz, orthoclase, and plagioclase in St. Kevin Granite. monazite, a characteristic accessory mineral of the Silver Plume Granite in the Front Range. The normal facies is the principal rock of the batholith; it occurs at the borders as well as in the interior of the batholith, and also as the sole facies in many isolated bodies of granite. The granodioritic facies is darker than any of the other facies and is porphyritic to seriate porphyritic, fine to medium grained, and weakly foliated. Composition-ally it is on the borderline between granodiorite and quartz monzonite (see table and fig. 127.3). It is characterized by sparsely and irregularly scattered white to buff rounded feldspar grains that give it aTWETO AND PEARSON D31 spotted appearance. The granodioritic facies is confined largely to the interior of the batholith, where it forms vaguely to sharply defined streaks a few tens of feet to several hundreds of feet wide in granite of the normal facies. The fine-grained facies is a gray, buff-weathering, fine-grained, equigranular, massive to weakly foliated biotite-muscovite quartz monzonite (see table and fig. 127.3). Except for scattered large flakes or clusters of muscovite that weather in relief, the rock is homogeneous. It is restricted to the border zones of the batholith, where it forms sharply defined bodies as much as 2 miles long and 1 mile wide. The coarse porphyritic facies of the St. Kevin Granite is biotite-muscovite granite or quartz monzonite characterized by abundant blocky microcline pheno-crysts about 1 inch long. The groundmass is almost identical to nonporphyritic granite of the normal facies, but the large and thick microcline phenocrysts give the rock an appearance different from that of any of the other facies of St. Kevin Granite. Relicts of gneiss are locally abundant in the rock. The coarse porphyritic facies seems to be restricted to the western and southern sides of the batholith, but it is not nearly so widespread as the map of Stark and Barnes (1935, pi. 1) showed the Hell Gate Porphyry to be. Contact relations and gradations between the various facies of the St. Kevin Granite suggest that all the facies are of the same general age. In places, granite of the normal facies clearly cuts the trachytoid hybrid facies, but as the hybrid facies is almost certainly a reaction front generated by magma of the normal facies, there can be little difference in age at any one place. Although some contacts between the normal and the granodioritic facies are sharp and show a sequence of intrusion, the time relations are inconsistent from body to body, or even from outcrop to outcrop; moreover, the two rocks intergrade imperceptibly in many places, and they evidently crystallized simultaneously. Similar almost imperceptible gradation between the normal and the coarse porphyritic facies is displayed conspicuously along the west side of the batholith, where the two facies alternate many times in a distance of a few hundred feet, and where no clear contacts exist. Only the fine-grained facies shows in part a definite and consistent age relation to the other facies; in places it cuts both the trachytoid hybrid and the normal facies and contains inclusions of these rocks. In other places, however, -the fine-grained facies seems to grade both into normal and granodioritic facies, and we therefore consider it to be of the same general age as the other facies. The St. Kevin Granite evidently was in part generated in place and in part crystallized from a mobile melt that was a magma in every sense of the word. Metasomatism or granitization in place is shown by the reaction selvages of trachytoid hybrid granite, and by the faithful preservation of folds and other structural features of the gneissic wall rocks in granite half a mile or more from the edge of the batholith. On the other hand, evidence of intrusion is compelling. Many dikes and sills as well as parts of the border of the main batholith have sharp contacts that show no sign of granitization; foliation in the granite faithfully follows the contacts wherever they crosscut the gneiss; inclusions of unaltered wallrocks that can be traced with certainty to a nearby source lie within the granite; metasomatic rocks identified by their ghost gneissic structure are disrupted by local zones of flow, some of which extend as dikes into the wallrock gneisses; and the long, straight or arcuate, diffuse contacts between different varieties of rocks in the interior of the batholith suggest boundaries between streams of magma rather than any structure inherited from pre-existing metamorphic rocks. We conclude that (1) from the widespread evidence of liquid movement, most of the St. Kevin batholith was, intruded as magma at the crustal level now exposed; (2) from the widespread evidence of metasomatism, an additional substantial fraction of the St. Kevin Granite was generated at the level now exposed; (3) from the evidence of local mobilization of metasomatic rock, and from the intricate, generally concordant, form of the batholith borders, the wallrocks were at practically the same temperature as the invading granite and in a relatively plastic state; (4) from the occurrence of probable relict metamorphic minerals such as sillimanite and garnet in the intrusive part of the batholith, the batholithic magma was generated by melting of gneisses similar to those which it intrudes; and (5), from conclusions 2, 3, and 4, the intrusive magma may have been generated at a crustal level little deeper than the one now exposed. Thus we believe we see the St. Kevin batholith at about the top of its root zone, or the zone of transition from one of magma generation by crustal melting below to one of intrusion above. If this is so, and the batholith is a product of local melting, the granite is a unit genetically distinct from the Silver Plume Granite and its cognates in the Front Range, even though it is probably of the same general age.D32 STRATIGRAPHY AND STRUCTURAL GEOLOGY REFERENCES Behre, C. H., Jr., 1953, Geology and ore deposits of the west slope of the Mosquito Range: U.S. Geol. Survey Prof. Paper 235. Stark, J. T., 1935, Migmatites of the Sawatch Range, Colorado: Jour. Geology, v. 43, p. 1-26. Stark, J. T., and Barnes, F. F., 1932, The structure of the Sawatch Range: Am. Jour. Sci., 5th ser., v. 24, p. 471-480. ------- 1935, Geology of the Sawatch Range, Colorado: Colorado Sci. Soc. Proc., v. 13, p. 467-504.Article 128 REINTERPRETATION OF THE LATE GROWTH OF THE GYPSUM VALLEY SALT ANTICLINE, SAN MIGUEL COUNTY, COLORADO By FRED W. CATER, Denver, Colo. Abstract.—Reinvestigation of outcrops in the southeast end of the anticline indicates that the relation of Cretaceous beds to Paleozoic rocks is explained best by faulting and landsliding rather than by sedimentary pinchouts and thinning as proposed previously. Reexamination of the Gypsum Valley area, San Miguel County, Colo., (fig. 128.1) has suggested changes in an earlier stratigraphic and structural interpretation of the area by the author. In an earlier report (Cater, 1955), the author stated that although up-thrusting of salt into the core of the salt anticline underlying the valley seemed generally to have ceased by the beginning of Morrison (Late Jurassic) time, inconclusive field relations in the southeast end of the valley indicated that Paradox beds (Pennsylvanian) had intruded rocks as high as the Mancos Shale (Late Cretaceous). Earlier, Stokes and Phoenix (1948) reached virtually the same conclusion. Landis and others (1961) stated, however, that: “The overall distribution of the Mancos and Mesaverde relative to the salt structures appears to be best explained by a pinchout of pre-Mancos Cretaceous rocks and a thinning of the Mancos on the salt structures.” Inasmuch Figure 128.1.—Index map showing location of area described in this article. as this interpretation differs radically from that of the earlier writers and because the scientific and economic implications of a correct explanation of the observable geologic relations are of considerable importance, the author reexamined part of the area. This reexamination indicates that neither of the preceding explanations is entirely consistent with the mappable relations. It now seems that Stokes, Phoenix, and the author were wrong in postulating possible late intrusion of Paradox beds into the Mancos, and that Landis and others were wrong in explaining the relations as a thinning and pinchout of Cretaceous rocks on salt structures. Mancos that rests on Paradox is explained best by landsliding, whereas the pinchout proposed by Landis and others is explained best by faulting, as indicated in an earlier paper (Cater, 1955). The differences between the interpretation presented herein and that of Landis and others (1961) arise from the identification of formations at some outcrops and from evaluation of evidence of faulting and landsliding. These differences are shown in figure 128.2. The map in figure 128.2.4 is reproduced from part of figure 197.2 of Landis and others (1961); that in figure 128.21? is of the same area, showing this author’s revisions. The differences, though minor, are critical. The geologic section in figure 128.24. is reproduced from part of their figure 197.3, and that in figure 128.21? is the author’s revision of the same geologic section. In a gully in the southeast corner of the area (fig. 128.24) the Salt Wash Member of the Morrison Formation is shown by Landis and others as a thin sliver adjacent to a fault between the 7,180-foot and the 7,240-foot contours, whereas according to the author’s interpretation (fig. 128.2S) the Salt Wash is relatively flat lying and much more extensive and does not extend as far northwest down the gully. To the northwest, Landis and others have mapped outcrops in three localities as Burro Canyon(?); the author interprets them differently. At locality A (fig. 128.21?), in ART. 128 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D33-D37. 1964. D33D34 STRATIGRAPHY AND STRUCTURAL GEOLOGY R. 16 W. R. 16 W, CONTOUR INTERVAL 100 FEET A B 0 500 1000 FEET 1 1_J I 1 I________________I Figure 128.2.—Geologic maps and sections of part of southeastern Gypsum Valley, San Miguel County, Colo. A, mapping and interpretation by Landis and others (1961, parts of figs. 197.2, 197.3); B, mapping and interpretation by the author. Circled letters in B are localities discussed in text. Numbers in sections refer to subdivisions of the Rico Formation and the upper member of the Hermosa Formation shown by Landis and others (1961, fig. 197.4).Middle Middle and Upper Lower Upper Pleistocene Pennsylvanian Pennsylvanian Upper Triassic Upper Jurassic Cretaceous Cretaceous and Recent CATER D35 EXPLANATION Qal Alluvium Qg Gravel deposits UNCONFORMITY > > or < z or u H < D O < < { { Kmv Mesaverde Formation Km Mancos Shale UNCONFORMITY Kbc Burro Canyon(?) Formation Shown only on Map A U) D o LU Jms Salt Wash Sandstone Member of Morrison Formation Je Entrada Sandstone Shown only on Map A UNCONFORMITY ■Rc Chinle Formation UNCONFORMITY Pc Cutler Formation UNCONFORMITY Pr Rico Formation Upper member of Hermosa Formation \ U lb (/) < or D “l y ib ib < or P z < J aL -> UJ 0. l i z < > lb z z lb D o or UJ u. O CD or < (J Contact Dashed where approximately located; dotted where concealed Fault Dashed where approximately located; dotted where concealed. U, upthrown side; D. downthrown side 3y Strike and dip of beds *>/ Strike of vertical joints Prospect Figure 128.2.—Continued the gully at 7,060 feet altitude, an outcrop only 3 feet wide and 5 feet long is shown by them on figure 128.2.4 as Burro Canyon(?) Formation. It is fine-grained buff sandstone, and although it is conceivable that a bed of similar material thick enough to have produced a block of this size could be found in the Burro Canyon, the author has never seen one in the course of several field seasons of mapping in the region Furthermore, the outcrop possesses none of the diagnostic features characteristic of the Burro Canyon. It is, on the contrary, almost identical with sandstones of the Mesaverde Formation on the east side of the gully and apparently is a slice of that formation caught in a fault subsidiary to the main fault zone of the area. About 1,200 feet north-northwest of this outcrop, at an altitude of 7,050 feet (locality B, fig. 128.21?), the outcrop called Burro Canyon(?) by Landis and others resembles conglomerates of that formation, but on close examination the rock is seen to be a conglomeratic arkose. As the Burro Canyon is nonarkosic, this outcrop is interpreted by the author as Permian Cutler Formation. Less than a hundred feet northwest of this outcrop is another, also shown by Landis and others as Burro Canyon(?). This second outcrop at locality B is fine-grained sandstone identical with the Mesaverde concealed beneath gravel deposits uphill to the east and is almost certainly a slumped block of Mesaverde resting on the Cutler. The actual contact with the Cutler is not visible. In summary, the Burro Canyon does not crop out in the area; thus it cannot be used as evidence in an argument favoring deposition of Cretaceous rocks on formations older than the Morrison. Other questionable identifications in this area are strata shown as Salt Wash on the 7,000-foot contour, where it crosses section A-A' (fig. 128.24.), and almost certainly the rock shown as Entrada Sandstone about 200 feet to the north at an altitude of 6,990 feet (locally C, fig. 128.25). The rock shown as Salt Wash is Mesaverde. That shown as Entrada resembles the Entrada superficially but differs from it in that it contains subangular pebbles of clear quartz % inch in diameter, it is less well sorted, and, although it contains large round frosted grains of quartz, the grains are scattered haphazardly through the rock rather than deposited in distinct layers. In all probability this rock also is Mesaverde, for there is similar material uphill in nearby outcrops of Mesaverde. At the northwest corner of the area, between the 6,700-foot and 6,800-foot contours (locality E), Mancos is mapped as resting on Rico where, in fact, there are no exposures. There is little doubt, however, that beneath the soil cover and overburden, Mancos does rest on Rico, but it is a mass of Mancos that has slumped downhill.D36 STRATIGRAPHY AND STRUCTURAL GEOLOGY A few feet north of the pit that Landis and others show at an altitude of 6,690 feet on the map (locality D) is a second pit that they do not show; this pit exposes a contact of Mancos and Cutler. Bedding in the Mancos is not badly disturbed and is roughly parallel to the surface of the underlying Cutler to within about 15 inches of the contact. Bedding in this lower part, however, is completely destroyed; fragments are slickensided and crushed, and the basal 4 to 6 inches contains stretched out lentils of pulverized Cutler as much as an inch thick and a foot long— suggesting a fault on the base of a landslide. All other isolated masses of Cretaceous rocks that rest on older rocks mentioned in the foregoing paragraphs are almost certainly masses of similar material that have moved downhill. In fact, were this an area where Cretaceous rocks wedged out against Paleozoic rocks, it would be reasonable to expect to find at least a few outcrops of Mesaverde to the west of the gully that generally follows the west edge of the Cretaceous rocks in the northwest part of the mapped area; however, none are found there. Furthermore, the Mesaverde on the east side of this gully is fractured and broken into jumbled but still juxtaposed blocks, a fact hinted at by Landis and others on their map (fig. 128.2A), which shows a strike-and-dip symbol east of the gully near the lower contact of the Mesaverde at 6,910 feet. This symbol shows beds striking east and dipping 38° S., an attitude unrelated to the mapped contact immediately below. The jumbled nature of the Mesaverde results from slumping of the underlying Mancos, a formation unsurpassed in the region for its ability to form landslides, some of them of enormous size. Although the trace of the fault that separates Cretaceous rocks from older rocks is covered by Mancos and Mesaverde rocks and by gravel of Quaternary age, north of where Landis and others mapped the fault in the southeast corner of the area of figure 128.2, fairly direct evidence for its north-northwestward continuation shows in the face of the previously mentioned pit at 6,960 feet (locality D). In the face of this pit is a north-northwesterly trending fault which is not shown on their map. Although this fault is in the Cutler Formation, it is parallel to and only a few feet from the edge of the Cretaceous rocks and is probably subsidiary to a much larger fault. Considerable quantities of copper minerals (malachite) have been deposited in and adjacent to this fault, and it is perhaps significant that other similar copper occurrences in the area are closely associated with large faults (Vogel, 1960). To the southeast, beyond the area shown in figure 128.2, the fault swings to the southeast and for 6 miles forms the southwest margin of the synclinal graben occupying the collapsed crest of the Gypsum Valley anticline. Vogel (1960) shows the throw on this fault to exceed 3,500 feet immediately southeast of the mapped area. It seems unlikely that a fault of this magnitude could die out completely along the strike in a distance that is less than the throw. At a number of localities shown by Landis and others (1961, fig. 197.2), but not shown on figure 128.1A of this article, Mancos rests directly on Paradox beds. Landis and others consider the contact to be sedimentary, whereas Stokes and Phoenix (1948) and Cater (1955) believed that the salt was intruded into the Mancos. It is doubtful whether either of these explanations is valid. With collapse and erosion of the crestal rocks in the anticline, conditions became ideal for landsliding of the Mancos from the partly surrounding uncollapsed masses of Mancos and Mesaverde that formed relative highlands. It seems most likely that these masses of Mancos that now rest on the Paradox are but the partly eroded remnants of former landslide blocks. In this connection it is interesting to note that Vogel (1960) described a butte in Disappointment Valley, a few miles south of this area, that is capped by a landslide consisting of the upper part of the Mancos and the Mesaverde; this slide rests on Mancos that belongs stratigraphically 1,300 feet below the top of the formation. The slide is several miles from the nearest outcrops of the Mesaverde. A more spectacular structural history than that shown by the evidence is implied if the thesis of Landis and others is accepted. For the Mancos to wedge out against Rico and Cutler beds would require a spire of the Rico and Cutler less than a mile across to rise more than 4,000 feet above the base of the surrounding Salt Wash during Mancos time. Field observation does not support this supposition; in fact, one of the reasons why the history of development of the salt anticlines is so well preserved in the southeast end of Gypsum Valley is that the rise of salt was less pronounced in this area. Upper Paleozoic beds are preserved as are pre-Morrison Mesozoic formations, and the unconformities between these formations preserve the record of anticlinal growth. In other places the salt rose more vigorously; overlying beds were turned up sharply and removed from the crest of the salt cores, and pinchouts of pre-Morrison beds are abrupt and mostly buried under younger rocks in the flanks of the anticlines. Yet here in the very area where the rise of salt had been relatively slow during growth of the salt cores, the interpretation of Landis and others requires that a spire of Rico and Cutler beds be thrust upward several thousand feet, tens of millions of years after substantial flowage of salt had ceased (at the beginning of Morrison time). Then, by their interpretation, this same spire would have had to subside to a position lower than the one itCATER D37 originally occupied, and all without appreciable faulting. REFERENCES Cater, F. W., Jr., 1955, Geology of the Gypsum Gap quadrangle Colorado: U.S. Geol. Survey Geol. Quad. Map GQ-59. Landis, E. R., Shoemaker, E. M., and Elston, D. P., 1961, Early and late growth of the Gypsum Valley salt anticline, San Miguel County, Colorado: Art. 197 in U.S. Geol. Survey Prof. Paper 424-C, p. C131-C136. Stokes, W. L., and Phoenix, D. A., 1948, Geology of the Egnar-Gypsum Valley area, San Miguel and Montrose Counties, Colorado: U.S. Geol/ Survey Oil and Gas Inv. Prelim. Map 93. Vogel, J. D., 1960, Geology, and ore deposits of the Klondike Ridge area, Colorado: U.S. Geol. Survey open-file report, 205 p.Article 129 A PLEISTOCENE SECTION AT LEONARDS CUT, BURKE COUNTY, NORTH CAROLINA By JOHN C. REED, JR., BRUCE BRYANT, ESTELLA B. LEOPOLD, and LOUISE WEILER, Denver, Colo. Abstract.—An exposure of peat described by Kerr in 1875 at Leonards Cut probably accumulated on a former flood plain of the Catawba River well below the Piedmont Upland surface. Pollen assemblages and field relations show that the deposit is nonglacial and probably represents an interglacial episode of Pleistocene age. The Piedmont Upland surface in the area may therefore be pre-Pleistocene. Organic remains in surficial deposits are very unusual in the Piedmont Province in North Carolina. We were therefore surprised and interested when R. A. Laurence of the U.S. Geological Survey (written communication, 1961) called attention to the following description of Leonards Cut by Kerr (1875, p. 157): The railroad cuts through the piedmont region, especially from Morganton to the foot of the Blue Ridge furnish many admirable sections of hill side drift. The most interesting is 9 miles beyond Morganton, known as Leonards cut. On the upper slope of a high hill a cut of 80 feet exposes a bed of peat and drift wood 15 feet thick with the underlying soil filled with stumps and roots. Above the peat is a bed of rudely stratified gravel and sand, 10 to 15 feet deep, on which grows the present forest. In the peat are numerous shining wing-covers of beetles and seeds of many species of plants, cones of several kinds of pine, and of hemlock, squirrel-gnawed hickory nuts, seed pods of kalmia, etc. No species have been observed which are not found at present living in the region . . . With the aid of Mrs. M. R. McVey of the Morganton-Burke County Public Library in Morganton, we located Leonards Cut 2.45 miles west of the Glen Alpine railroad station along the Southern Railroad (figs. 129.1 and 129.2). Leonards Cut is in the southwestern part of the Morganton basin, a northwestern extension of the Piedmont physiographic province enclosed by the Blue Ridge on the northwest, the Hickorynut Mountains on the southwest (outside the area shown on fig. 129.1), Figure 129.1.—Map of part of western North Carolina, showing location of Leonards Cut (X). The outlined area surrounding Leonards Cut is shown on figure 129.2. the South Mountains on the southeast, and the Brushy Mountains on the northeast (outside the area shown on fig. 129.1). The basin is drained by the Catawba River and its tributaries, which have excavated steepsided flat-bottomed valleys 100 to 300 feet deep separated by flat even-crested interfluves. Farther east the interfluves are less dissected and form the flat surface of the Piedmont Upland. D38 ART. 129 Iff U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D38-D42. 1964.REED, BRYANT, LEOPOLD, AND WEILER D39 81 *49' 1000 0 3000 FEET CONTOUR INTERVAL 40 FEET Figure 129.2.—Topographic map of the area surrounding Leonards Cut (X). Near Leonards Cut the interfluve surfaces are at altitudes of about 1,300 feet; the surface of the Catawba River, 0.6 mile north of Leonards Cut, is at an altitude of about 1,040 feet. Since Kerr made his visit the cut has slumped and become much overgrown with honeysuckle, but a small gully (fig. 129.3) south of the tracks exposes a fairly complete section of surficial material, including Kerr’s peat bed (table 129.1, fig. 129.4). The peat bed does not extend more than 50 feet east of the erosion gully, and it is absent on the north side of the tracks opposite the exposure in the gully. It may extend several hundred feet farther west. The deepest part of the cut, several hundred feet east of the gully, exposes only 10 to 15 feet of red colluvial clay overlying 30 to 40 feet of saprolitized mica schist and gneiss. In the north side of the cut, less than 40 feet north of the gully, yellow-gray alluvial silt and clay rest directly on saprolitized bedrock. The uppermost surficial deposits are exposed in cuts along the north side of a dirt road parallel to and about 200 feet south of the railroad (fig. 129.2). This material is rudely sorted angular gravel and coarse brown sand Figure 129.3.—Gully on south side of Leonards Cut in which peat and overlying unconsolidated deposits are exposed. U.S. Geological Survey paleobotanical localities D1995A and D1995B indicated by A and B, respectively. containing saprolitized fragments of gneiss, schist, and pegmatite arranged in crude foreset beds that dip 30°-40° W. To the west it grades into finer grained material that overlies the red clay layer at the top of the measured section. The sand and gravel deposit is absent in cuts on the south side of the road and in the railroad cut to the north. The peat probably accumulated in a shallow pond or swampy area at the edge of a former flood plain of the Catawba River. Similar swampy areas are common along modern streams in the area. The overlying stratified silt and clay is similar to material underlying the present flood plain of the Catawba. The red clay is typical of the residual and colluvial clay that mantles most slopes in the area. The rudely stratified sand and gravel probably filled a gully cut into the colluvial clay before the Catawba assumed its present course. The field relations of the gully-fill deposit indicate considerable modification of the topography since its deposition. The deposits at Leonards Cut offer the only known clue to the minimum age of the Piedmont Upland surface in the Morganton basin. The organic material lies well below the level of the upland surface, but more than 120 feet above the present master drainage. A radiocarbon age of 2,270 ±200 years for organic material near the base of the alluvial deposits on the present flood plain of Muddy Creek about 7 miles southwest of Leonards Cut (Suess, 1954) indicates that the masterD40 STRATIGRAPHY AND STRUCTURAL GEOLOGY Figure 129.4.—Diagrammatic section of unconsolidated deposits exposed in gully on south side of Leonards Cut D1995A and D1995B, locations of pollen samples described in table 129.2. Table 129.1.—Stratigraphic section of unconsolidated deposits exposed in a gully on the south side of Leonards Cut [Base of section 2 feet below level of top of railroad tracks; elevation about 1.160 feet] Type of deposit (fig. 129.4) Distance above base (feet) Lithology Remarks Colluvial 28-33 Unstratified red clay; 6-inch gray layer at base. Basal contact irregular and gradational. Similar red clay rests directly on saprolitized bedrock 200 feet east of section. Colluvial or flood plain 19-28 Unstratified yellow-gray silty clay containing scattered mica flakes and feldspar grains. Flood plain 12-19 Stratified gray silt and clay containing thin sandy layers. A few organic layers as much as 2 inches thick becoming less common near top. Scattered fragments of carbonized wood near top. 10-12 Dark, brown silt and clay containing thin sandy layers. Swamp 5-10 Dark-gray organic silt and peat containing scattered wood fragments. USGS paleobot. loc. D1995B at top of interval. 0-5 Dark-gray organic silt and peat containing wood fragments, flattened logs as much as 6 inches in diameter, and a few nuts and hemlock (?) cones. Base not exposed. USGS paleobot. loc. D1995A 3 feet above base of section.REED, BRYANT, LEOPOLD, AND WEILER D41 drainage has been at about the present level for at least that length of time. Pollen and spores from samples of organic silt and peat from 3 feet (USGS paleobot. loc. D1995A) and 10 feet (USGS paleobot. loc. D1995B) above the base of the exposed section in Leonards Cut have been examined and the pollen assemblages compared with those from samples of organic muck from nearby modern bogs (table 129.2). All identified pollen and spores in the Leonards Cut fossil material are representative of plants that grow in the area today; no relict Tertiary genera and no northern elements were found. The main difference between the assemblages of fossil and modern pollen is that spruce and fir, which make up between 1 and 4 percent of the fossil material, Table 129.2—Pollen types identified in organic layer at Leonards Cut and in nearby muck deposits of Recent age (percent of total pollen) [X, present] Leonards Cut fossil samples Modern samples D1995A D1995B D1991 D1992 Gymnospermae: 52.5 61.5 76.0 62.0 3.9 .4 .4 1.3 1. 5 .5 .5 1. 5 .8 .8 Angiospermae: .4 7.8 1.7 1.0 4.2 7.0 4.6 1.0 11.2 .4 .8 .4 .8 .5 .5 1.3 1.3 1.5 4.2 3.1 2.9 3.0 4.7 X 1.0 .4 .4 X .4 .5 1.9 8.4 9.5 7.0 1.5 4.6 .4 1.3 5.0 1.7 .4 4.7 3.4 2.0 2.8 5.0 3.4 1.0 X 4.5 1.0 Total number pollen grains counted Lower plants: Spores: 258 X 238 207 214 X Algae: x X x i “ U.S. Geological Survey paleobotanical localities. D1995A. Leonards Cut locality, Burke County, N.C. Peat 3 feet above base of section exposed in gully on south side of railroad. Location indicated on figures 129.2 and 129.3. D1995B. Leonards Cut locality, Burke County, N.C. Peat 10 feet above base of section exposed in gully on south side of railroad. Location indicated on figures 129.2 and 129.3. D1991. Organic muck from depression on flood plain of Catawba River immediately north of bridge at Bridgewater (fig. 129.1). 2.05 miles N. 85° W. of Leonards Cut (collected by C. T. Sumsion). D1992. Organic muck from spring or seep on hillside about 200 feet south of Leonards Cut (collected by C. T. Sumsion). are absent in the modern material. Since these genera now grow at higher elevations 25 miles west of the fossil locality, this suggests that these gymnosperms were growing at least in small numbers closer to the Leonards Cut locality at the time of deposition than at the present time, but the climate was probably not much different than the present one. Deposits similar to those at Leonards Cut are unusual in the Piedmont province, although logs associated with organic clays have been discovered in exhumed gully fills in Cleveland County, N.C., and Spartanburg County, S.C. (Whitehead and Barghoorn, 1962). The radiocarbon age of a log from near the base of one deposit north of Shelby in Cleveland County is greater than 30,000 years (Suess, 1954). A pollen profile from this locality (Whitehead and Barghoorn, 1962) records little change within the 4-foot section, and contains an assemblage very similar to that at Leonards Cut. However, it contains two elements not found at Leonards Cut: Ephedra (Mormon’s tea), a western shrub now exotic to the eastern United States, and Arceuthobium (dwarf mistletoe) now occurring in the northeastern United States. This locality is about 40 miles from the nearest present day occurrence of spruce and fir; it is about 100 feet lower than Leonards Cut and is about 30 miles to the southeast. Eargle (1940) has described deposits of organic material deeply buried by soil in Spartanburg County, S.C. The radiocarbon age of logs imbedded in two of these deposits is greater than 34,000 years (Rubin and Alexander, 1958) or greater than 35,000 years (Whitehead and Barghoorn, 1962). Cain (1944) studied pollen from four sections of these buried soils. Assemblages near the top of one of his sections closely resemble pollen rain one would expect at the localities today, but near the base of two of his sections the abundance of spruce and fir (as much as 25 percent of the total pollen) suggests a climate cooler than the present. Whitehead and Barghoorn (1962) studied three other localities in Spartanburg County about 30 miles due south of Leonards Cut. They found relatively little change within two sections and a pronounced shift from hemlock to pine in the third. Although the nearest spruce and fir trees are now from 30 to 45 miles away, fossil spruce pollen comprises as much as 8 percent of the tree-pollen count, and fir pollen is present. Northern and montane elements recorded in all three sections (Whitehead, 1963) strengthen the interpretation that these localities record a climate somewhat cooler than now.D42 STRATIGRAPHY AND STRUCTURAL GEOLOGY Frey (1953) found that pollen from Singletary Lake near the coast of North Carolina indicates a climate consistently cooler than the present from more than 20,000 to about 10,000 years ago. The absence of relict Tertiary genera in the Leonards Cut material is indicative of Quaternary age. Frey (1953) has shown that cool, glacial climates during the Pleistocene extended as far south as the coastal plain of North Carolina. The similarity between the fossil pollen assemblages from Leonards Cut and modern assemblages in the same area shows, however, that the deposit is not glacial. The present elevation of the deposit above the modern master stream suggests that it is not Recent; it is therefore believed to be of interglacial or interstadial Pleistocene age. The surface of the Piedmont Upland in the Morganton basin is older than the Leonards Cut deposit and may be pre-Pleistocene. REFERENCES Cain, S. A., 1944, Pollen analysis of some buried soils, Spartanburg County, South Carolina: Torrey Bot. Club Bull., v. 71, p. 11-22. Eargle, D. H., 1940, The relations of soils and surface in the South Carolina Piedmont: Science, v. 91, p. 337-338. Frey, D. G., 1953, Regional aspects of the late-glacial and postglacial pollen succession of southeastern North Carolina: Ecol. Mon., v. 23, p. 289-313. Kerr, W. C., 1875, Report of the Geological Survey of North Carolina, v. 1, Physical geography, resume, economical geology: Raleigh, North Carolina, 325 p., map. Rubin, Meyer, and Alexander, Corrine, 1958, U.S. Geological Survey radiocarbon dates IV: Science, v. 127, p. 1476-1487. Suess, H. E., 1954, U.S. Geological Survey radiocarbon dates I: Science, v. 120, p. 467-473. Whitehead, D. R., 1963, “Northern” elements in the Pleistocene flora of the Southeast: Ecology, v. 44, p. 403-406. Whitehead, D. R., and Barghoorn, E. S., 1962, Pollen analytical investigations of Pleistocene deposits from western North Carolina: Ecol. Mon., v. 32, p. 347-369.Article 130 SURFACE AND SUBSURFACE STRATIGRAPHIC SEQUENCE IN SOUTHEASTERN MISSISSIPPI By D. HOYE EARGLE, Austin, Tex. Work done on behalf of the Atomic Energy Commission Abstract.—The stratigraphic sequence in southeastern Mississippi includes outcropping units of Miocene to Recent age; many subsurface units of Tertiary, Cretaceous, and Jurassic age down to the Jurassic(?) salt; and possibly older sedimentary rocks that are underlain by the basement crystalline rocks. Two newly named units are described—the Tatum Limestone Member, a basal unit of the Catahoula Sandstone of Miocene and 01igocene(?) age; and the Andrew Formation of Early Cretaceous age. A detailed knowledge of the stratigraphic section of southeastern Mississippi was required for use in geologic and geophysical studies of the Tatum salt dome, Lamar County, Miss., for the U.S. Atomic Energy Commission. The deepest well in the southeastern Mississippi region penetrated only a few feet of bedded salt believed to be the Louann; consequently, the rocks below the salt are not known. Because detonations at the site, however, are expected to give geophysical data on rocks beneath the salt, the units that are found below the salt in adjoining areas are shown in the accompanying table. Subsurface sections were constructed radiating outward from the salt dome at least 25 miles in all directions in which velocity surveys of wells were available. Study of the continuous-velocity logs of wells shows, as expected, a close correlation between lithology and zones of relatively uniform velocity characteristics. The zones compare closely with named stratigraphic units. The stratigraphic nomenclature is here brought up to date and, for several units, names that more nearly conform to the rules of stratigraphic nomenclature are herein introduced. Some names in common usage have been adopted, and some lithologic names have been changed to indicate more specifically the lithology of the subsurface units. Type sections in wells are designated and briefly described. The accompanying table shows the names and stratigraphic relations of the units.1 As described by Matson (1916) the Citronelle Formation (Pliocene) (1, in table) is the sequence of sediments, chiefly nonmarine, above the Pascagoula Formation (Miocene) and below stream-terrace deposits. Its type locality at the town of Citronelle (Mobile County) in nearby southwestern Alabama includes (a) sand and gravel, considered to be Pleistocene in age (Roy, 1939; Carlston, 1951), that caps the uplands and lies uncon-formably on the deposits below, and (b) littoral, estuarine, and shallow marine deposits considered to be Pliocene in age (Berry, 1916; Stringfield and La-Moreaux, 1957). In Mississippi, only the sand and gravel deposits have been mapped as Citronelle, and they are considered by some geologists to be Pleistocene in age (Brown and others, 1944; Fisk, 1938; Doering, 1956). More recently Brown is of the opinion that the Citronelle is Pliocene in age (written communication to G. V. Cohee, June 23, 1959). The Citronelle is several hundred feet thick along the coast of southeastern Mississippi and contains important aquifers. Inland it is thinner than along the coast and forms extensive upland plains and caps interstream divides. Remnants are found as far north as 140 miles from the Gulf of Mexico. i Formal use of these names was considered at a conference in Jackson, Miss., on September 17, 1962, arranged by Esther R. Applin, of the U.S. Geological Survey, and attended by representatives of the U.S. Geological Survey, the Mississippi Geological Survey, the Mississippi Geological Society, and the Trowbridge Sample Service. Thicknesses given in the accompanying table were obtained from well logs of southeastern Mississippi and from data from Andrews (1960), Nunnally and Fowler (1954), Rainwater (1960, 1961), and Scott and others (1961). The author thanks his colleagues of the U.S. Geological Survey and Jules Braunstein, Shell Oil Co.; E. H. Rainwater, Shell Development Co.; and Eleanor Caldwell, Humble Oil and Refining Co.; for discussions and advice as to the use of these names. 716-626 0-64- -4 ART. 130 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D43-D48. 1964. D43D44 STRATIGRAPHY AND STRUCTURAL GEOLOGY The Pascagoula and Hattiesburg Formations (2) together form an indivisible stratigraphic unit. The Pascagoula Formation was described near the coast, along the Pascagoula River, where the beds are chiefly marine, and the Hattiesburg was described about 50 miles or more inland at Hattiesburg, Miss., where the beds are less marine and consist in large part of sands that are important fresh-water aquifers. The Catahoula Sandstone is divided into an upper part, chiefly nonmarine sand and clay, and a lower part termed the Tatum Limestone Member (3), a new name introduced to replace “Limestone of the Heterostegina Surface and subsurface units in southeastern Mississippi and adjoining areas [Numbers in parentheses keyed to text discussion] Sys- tem Series Group Unit Approximate thickness (feet) Lithologic character Quater- nary Pliocene to Recent Coastal terrace deposits 300 ± Silt, sand, gravel. (1) Citronelle Formation 150 ± Gray to mottled red and orange silty clay, sand, gravel. Miocene (2) Pascagoula and Hattiesburg Formations, undifferentiated. 350-2, 300 + Greenish-gray silty clay, sand, and gravelly sand. Catahoula Sandstone Upper part 250-700 Gray to olive sand, silt, and silty clay. Miocene(?) and Oligocene (?) (3) Tatum Limestone Member. 90-300 White to gray sandy limestone and marl; glauconitic calcarenite. T ertiary Oligocene (4) Chickasawhay(?) Limestone 60-470 Gray to white sandy limestone and fossiliferous sandstone and clay. Vicksburg Bacatunna Clay Member of Byram Formation. 50-215 Calcareous clay. Glendon Limestone Member of By ram Formation and Marianna Limestone, undifferentiated. White to gray sandy limestone and marl. Red Bluff Clay 30 ± Gray, fine sand and clay interbedded, and soft fossiliferous limestone. 1 1 Paleocene Eocene 1 Jackson Yazoo Clay 0-253 Olive to gray calcareous clay. (5) Moodys Branch Limestone (Ocala Limestone to south). 30-300 White sandy limestone, fossiliferous glauconite, and fossiliferous calcarenite. Claiborne Cockfield Formation 25-215 Lignitic clay and fine sand. (6) Cook Mountain Limestone 115-235 Hard to soft white calcarenite (lime sand) and glauconitic; bentonitic clay. (7) Sparta Sand and Zilpha Clay. 90-625 Gray shale and thin siltstone, interbedded. (8) Winona Marl 90-200 Glauconite, marl, green sand, and shale. (9) Tallahatta Siltstone Hard gray siltstone, glauconite. (10) Wilcox Undivided (Salt Mountain Limestone equivalent near base). 2, 250-3, 200 Gray fine-grained sandstone and green to gray shale, interbedded. Chalky white fossiliferous limestone. Midway Porters Creek Clay 650-1, 050 Gray shale. (11) Clayton Limestone 10-25 Limestone.EARGLE D45 Surface and subsurface units in southeastern Mississippi and adjoining areas—Continued [Numbers in parentheses keyed to text discussion] Sys- tem Series Group Unit Approximate thickness (feet) Lithologic character w S3 o (12) Selma Group and Eutaw Formation, undifferentiated. 925-1, 500 White chalk to gray marl, shale, and calcareous sandstone at base. O o3 +3 02 i- O Gordo Formation (upper Tuscaloosa of oil geologists). 300-900 Sandstone and gray shale. Upper Tuscaloosa (13) Coker Formation (includes marine Tuscaloosa and lower Tuscaloosa of oil geologists with “massive sand” at base). 620-1, 160 Shale and some sandstone, mostly marine (includes thick sand, containing few shale lenses at base, and some gravel). GG p (14) Dantzler Formation 575-1, 150 Red to gray mottled shale, buff, red, and green sandstone and siltstone. o V o m P (15) Andrew Formation 1, 000-1, 880 Limestone, sandstone, and gray to green shale. P (17) Haynesville Formation (includes Buckner Member) 1, 500-3, 450 ± Sandstone and red beds (Buckner Member contains salt in Alabama). P 02 ft Smackover Formation Oolitic limestone. a & Norphlet Formation Sand and shale in Mississippi and Alabama, red shale and conglomerate in Arkansas and Louisiana. i 2 & as —✓ (18) Louann Salt 0-5, 000 ± White coarsely crystalline halite. £ ® Werner Formation 200 ± Anhydrite, red shale, sandstone. Per- | mian(?) (19) Eagle Mills Formation (restricted) . 7, 000 ± Red beds, alluvial, continental flood-plain; some gray and green beds, gray to white siltstones and sandstones, in association with diabase. Pennsyl- vanian(?) (20) Morehouse Formation 1, 200 ± Dark-gray shale. Pennsylvanian and older Unknown. Sedimentary and metasedimentary rocks. Precambrian Basement rocks Unknown Crystalline igneous and metamorphic rocks. zone” or “Heterostegina limestone.” The following is a description of the Tatum Limestone Member. Type section.—U.S. Atomic Energy Commission hydrologic test well 1, Tatum dome area, Lamar County, Miss., is the type section. Well 1 is in the NE^SWtf sec. 12, T. 2 N., R. 16 W., and the datum is 321 feet above sea level. Hydrologic test well 2 in the SW^SW^ sec. 14, T. 2 N., R. 16 W.,for which the datum is 302 feet, is a reference section.D46 STRATIGRAPHY AND STRUCTURAL GEOLOGY Areal extent.—Southern Mississippi to northern Florida and the northern part of the Florida parishes of Louisiana. Thickness.—In well 1, 163 feet, from a depth of 1,373-1,536 feet; in well 2, 170 feet, from 1,470-1,640 feet. The Tatum thins to about 90 feet in the southeastern part of the area and thickens to 300 feet toward the west. Lithology.—Sandy limestone, marl, glauconitic calcarenite, and calcirudite (made up chiefly of fossil and limestone fragments). Electrical-log characteristics.—High resistivity; generally very low self-potential, which is characteristic of a fresh-water aquifer. Sonic-log characteristics.—High in sonic velocity. It is the first zone of high velocity found in drilling and in geophysical exploration of southern Mississippi. Paleontology.—Larger foraminifers are chiefly Heterostegina sp., Miogypsina sp., and Sorites sp. Many species of smaller Foraminifera have been identified from cores from well 1 by Ruth Todd and Doris Low. Age.—Miocene(?) and 01igocene(?). Larger foraminifers are said to be definitely Oligocene in age (Cole and Applin, 1961); smaller foraminifers suggest Miocene age (Ruth Todd and Doris Low, personal communication, 1962), although an Oligocene age is not necessarily precluded. The Tatum is tentatively correlated with fcssiliferous limestones and marls of the Paynes Hammock Sand (Miocene) and the upper bed of the Chick-asawhay Limestone (Oligocene) as described by MacNeil (1944). Chickasawhay Limestone (4) has been defined from surface outcrops only, and use of the name for a unit of the subsurface is questioned because exact correlation with surface outcrops is uncertain. Moodys Branch Limestone (5) is used in southeastern Mississippi instead of Moodys Branch Formation to indicate the characteristic subsurface lithology of the unit. The Ocala Limestone, in the southern part of the report area, as in western Florida and southern Alabama, is equivalent to the Moodys Branch Limestone and the Yazoo Clay farther updip. Cook Mountain Limestone (6) is used instead of Cook Mountain Formation to indicate the characteristic subsurface lithology of the unit. It has been called the Camerina limestone by some geologists, but this genus is also present in other formations. The terms Sparta Sand and Zilpha Clay (7) are used collectively in this article for the unit that underlies the Cook Mountain Limestone. “Upper part of Cane River equivalent” (of Louisiana) has been used for this unit in southwestern Mississippi. Some sandy beds in the middle of the chiefly clayey and silty unit are the downdip equivalents of the Sparta Sand. Winona Marl (8) is used for the green sand and shale and glauconitic marl whose equivalent is the Winona Sand of the outcrop. The term “Cane River Marl” (or, as used by some geologists, “Lower part of Cane River equivalent”) is commonly used by oil geologists for this subsurface unit in southwestern Mississippi. Tallahatta Siltstone (9) is used in the report area instead of Tallahatta Formation to indicate more definitely the characteristic subsurface lithology of the unit. Recent studies in Mississippi show that the different units in the Wilcox section (10) are closely correlative with the various formations of the Wilcox Group in Alabama. However, in this subsurface study, the Wilcox Group has not been subdivided. Clayton Limestone (11) is used instead of Clayton Formation of the outcrop to indicate the characteristic subsurface lithology of the unit. Some geologists consider that the shale and sandstone section below the chalk of the Selma Group and Eutaw Formation, undifferentiated (12), correlates with the basal part of outcropping Eutaw Formation equivalents in Arkansas. “Massive sand” (13) for the basal beds of the Upper Cretaceous is used informally (a formal name is not recommended) since the beds are not of uniform lithologic character and they vary in stratigraphic position. The Dantzler Formation (14) was originally described by Hazzard, Blanpied, and Spooner (1947). The following is a description of the formation. Type section.—Humble Oil and Refining Co. No. B-l Dantzler Lumber Co. well, Jackson County, Miss. The well is in sec. 30, T. 5 S., R. 8 W., and the derrick floor elevation is 108 feet above sea level (datum, 1 foot above rotary, or about 110 feet). Depth 8,905-9,910 feet (Nunnally and Fowler, 1954, p. 25). Areal extent and thickness.—Southeastern Mississippi, extending into the southwest corner of Alabama; 1,200 feet thick (Nunnally and Fowler, 1954, p. 28) but thins rapidly to the west. Western limit, Franklin and Amite Counties; northern limit, from northern boundaries of Lincoln and Covington Counties southeast to Greene County. Beyond these limits the formation cannot be distinguished from the underlying Andrew Formation. Lithology.—Nonmarine sands, fine- to medium-grained, white to dull-red, and green; shales, dark-purplish-red, generally mottled with white, yellow, ochre, and gray; some shales are dark gray; some are micaceous, slightly chloritic, silty. Some beds are carbonaceous and lignitic, others are calcareous and contain gray, red, or white limestone nodules Electrical-log characteristics.—Generally lower resistivity and self-potential than beds above and below. Sonic-log characteristics.—Generally lower sonic velocity than beds above and below. Paleontology.—F. W. Rolshausen, of the Humble Oil and Refining Co., has found oyster shells, ostracods, and Chara in some cores from the type well, and one specimen of Haplophrag-moides in one core from a depth of 9,779-9,789 feet. Age.—May include beds of Cenomanian age at top, as originally correlated by Hazzard, Blanpied, and Spooner, but is unconformably overlain by beds of known Cenomanian age, the Tuscaloosa Group. The Andrew Formation (15) is here named and described as generally marine rocks previously called “Pre-Dantzler rocks of Washita and Fredericksburg Groups, undifferentiated” (Nunnally and Fowler, 1954,EARGLE D47 p. 22-25). The following is a description of the Andrew Formation. Type section.—Gulf Oil Co. No. 25 J. M. Andrew well, sec. 6, T. 1 N., R. 16 W., Baxterville oil field, Lamar County, Miss. Elevation of the derrick floor is 233 feet above sea level (datum is 1 foot above rotary, about 2 feet above derrick floor, or about 235 feet). Cuttings and cores were examined by Esther R. Applin (written communication, October 1962), who found that the formation extends in depth from 9,800 feet (electrical-log point, or the 9,810-9,820-foot sample) to 11,360 feet (electrical-log point, or the 11,380-11,390-foot sample). Areal extent and thickness.—Across southern Mississippi and into adjoining States. Northern limit is defined by a change in facies to nonmarine sands and shales whose lithology is similar to overlying and underlying formations. Northern limit of recognizable Andrew Formation extends from Claiborne County on the northwest to southeastern Greene County on the southeast (Nunnally and Fowler, 1954). Lithology.-—According to Mrs. Applin’s sample and core descriptions the Andrew Formation consists, toward the top, of dull- to dark-red, gray, and olive-gray shale containing beds of brownish-gray finely sandy limestone, some shell fragments, and some beds of olive-gray dolomite and light-cream limestone. This grades down into gray and greenish-gray to dull-red micaceous shale alternating with limestone, minor beds of fine-grained sandstone containing some carbonaceous matter, and grayish-green siltstone. Much of the lower part is dark-gray shale. Electrical-log characteristics.—Generally low self-potential and alternating zones of moderately high and low resistivity. Sonic-log characteristics.—Generally high sonic velocity, but thin zones of strongly contrasting velocities. Paleontology.—According to Mrs. Applin, the upper part of the Andrew Formation contains abundant bivalves and ostracods. A few specimens of Lituola inflata were found in samples from depths of 10,280-10,300 and 10,360-10,370 feet. Ostrea fragments were common throughout the section; ostracods were abundant and scattered throughout. Quinqueloculina and other Miliolids were found at 10,596-10,601 and at 10,626-10,628)4 feet. Plicatula- and Pecten-like pelycypods were found at 10,611-10,616 feet. Age.-—The formation has been correlated with the lower part of the Washita and Fredericksburg Groups of Texas. Exact correlations on the basis of diagnostic faunas, however, are unknown to the author. Cotton Valley Formation (16) is used instead of Cotton Valley Group because the formations making up the group in other States have not been differentiated in southeastern Mississippi. Haynesville Formation (17) includes the Buckner Member in the subsurface of southestern Mississippi. The Buckner is believed to include some salt beds in Alabama. The Louann Salt (18) (greatly contorted) has been intruded as diapirs into overlying sediments. One well in the report area is reported to have penetrated the bedded Louann Salt. Applin and Applin (1953) state that a core composed of 1 foot of anhydrite and 1 foot of rock salt was recovered in a basket core barrel from the lowermost 9-foot interval of George Vasen fee well No. 1, Stone County, Miss., drilled to a depth of 20,450 feet. The units listed here as occurring below the Louann have not been drilled in southeastern Mississippi and are, therefore, not known to be present there. They have been found, however, in one or more States adjoining Mississippi, where the salt occurs at shallower depths. The Eagle Mills (19) was originally described (Imlay, 1940) to include beds now included in the Norphlet, Louann, Werner, and the Eagle Mills, restricted. The exact age of this sequence is unknown, but has been considered by various authors as Per-mian(?), Triassic(?), or Jurassic(?). Scott, Hayes, and Fietz (1961) tentatively include the Eagle Mills in the Triassic, partly on the basis of identification by Erling Dorf of a plant fossil that is similar to one of the Chinle Formation of Arizona and the Newark Group of Virginia, and partly on lithologic similarities to the Newark Group. The structural and stratigraphic relations of the Morehouse Formation (20) to the Eagle Mills Formation, restricted, are not known. REFERENCES Andrews, D. I., 1960, The Louann salt and its relationship to Gulf Coast salt domes: Gulf Coast Assoc. Geol. Societies Trans., v. 10, p. 215-240. Applin, P. L., and Applin, E. R., 1953, The cored section in George Vasen’s fee well 1, Stone County, Mississippi: U.S. Geol. Survey Circ. 298, 29 p. Berry, E. W., 1916, The flora of the Citronelle Formation: U.S. Geol. Survey Prof. Paper 98-L, p. 193-208. Brown, G. F., and others, 1944, Geology and ground-water re sources of the coastal area in Mississippi: Mississippi Geol. Survey Bull. 60, 232 p. Carlston, C. W., 1951, Profile sections in Citronelle Formation in southwestern Alabama: Am. Assoc. Petroleum Geologists Bull., v. 35, no. 8, p. 1888-1892. Cole, W. S., and Applin, E. R., 1961, Stratigraphic and geographic distribution of larger Foraminifera occurring in a well in Coffee County, Georgia, in Contributions from the Cushman Foundation for Foraminiferal Research, v. 12, pt. 4: p. 127-135. Doering, J. A., 1956, Review of Quaternary surface formations of Gulf Coast region: Am. Assoc. Petroleum Geologists Bull., v. 40, no. 8, p. 1816-1862 Fisk, H. N., 1938, Geology of Grant and LaSalle Parishes: Louisiana Dept. Conserv., Geol. Bull. 10, 246 p. Hazzard, R. T., Blanpied, B. W., and Spooner, W. C., 1947, Notes on correlation of the Cretaceous of east Texas, south Arkansas, north Louisiana, Mississippi, and Alabama: Shreveport Geol. Soc. 1945 Reference Rept., v. 2, p. 472-481. Imlay, R. W., 1940, Lower Cretaceous and Jurassic Formations of southern Arkansas and their oil and gas possibilities: Arkansas Geol. Survey Inf. Circ. 12. MacNeil, F. S., 1944, Oligocene stratigraphy of southeastern United States: Am. Assoc. Petroleum Geologists Bull., v. 28, no. 9, p. 1313-1354.D48 STRATIGRAPHY AND STRUCTURAL GEOLOGY Matson, G. C., 1916, The Pliocene Citronelle Formation of the Gulf Coastal Plain: U.S. Geol. Survey Prof. Paper 98, p. 167-192. Nunnally, J. D., and Fowler, H D., 1954, Lower Cretaceous stratigraphy of Mississippi: Mississippi Geol. Survey Bull. 79, 45 p. Rainwater, E. H., 1960, Stratigraphy and its role in the future exploration for oil and gas in the Gulf Coast: Gulf Coast Assoc. Geol. Societies Trans., v. 10, p. 33-75. Rainwater, E. H., 1961, Outline of geological history of Mississippi: Gulf Coast Assoc. Geol. Societies Trans., v. 11, p. 43-45. Roy, C. J., 1939, Type locality of Citronelle Formation, Citro nelle, Alabama: Am. Assoc. Petroleum Geologists Bull., v. 23, no. 10, p. 1553-1559. Scott, K. R., Hayes, W. E. and Fietz, R. P., 1961, Geology of the Eagle Mills Formation: Gulf Coast Assoc. Geol. Societies Trans., v. 11, p. 1-14. Stringfield, V. T., and LaMoreaux, P. E., 1957, Age of Citronelle Formation in Gulf Coastal Plain: Am. Assoc. Petroleum Geologists Bull., v. 41, no. 4, p. 742-746.Article 131 MAGMATIC DIFFERENTIATION IN A VOLCANIC SEQUENCE RELATED TO THE CREEDE CALDERA, COLORADO By JAMES C. RATTE and THOMAS A. STEVEN, Denver, Colo. Work done in cooperation with the Colorado State Metal Mining Fund Board Abstract.—Arranged in order of age, the ash-flow sheets and interlayered lava flows related in origin to the Creede caldera display progressive changes in (a) the abundance and composition of phenocrysts, (b) chemical composition, and (c) mode of eruption. These changes are evidence of continuing differentiation of the source magma. A complex pile of welded ash-flow tuffs, lava flows, and intrusive rocks surrounds the Creede caldera, a nearly circular subsidence structure 10 to 12 miles in diameter in the central San Juan Mountains, Colo. The volcanic rocks were in part derived from a source area in the vicinity of the caldera and in part from other sources. Those from the caldera source area comprise 5 ash-flow units (terminology of Smith, 1960; Ross and Smith, 1961) and 3 lava-flow units. These units are listed according to order of eruption in figure 131.1, which also lists formational names defined by the authors in another article (Art. 132). When considered in order of age, the volcanic units show an irregular progression from early phenocryst-poor rhyolites to later phenocryst-rich quartz latites. Parallel with these changes in composition, the mode of erup- Formation Unit (in order of eruption) Volume (order of magnitude in cubic kilometers) Abundance of phenocrysts (percent) 0 10 20 30 40 50 60 70 Composition of phenocrysts Number of modes Plagioclase Sanidine Quartz Biotite Magnetite Pyroxene Amphibole Fisher Quartz 8 10-100 M O-m O-m m-M m O-M O-M 17 Latite Lava flows Snowshoe Mountain 7 100-1000 M m m M m m O-m 26 Quartz Latite Ash flows Lava flows 6 1-10 l • 1 M 0 O m-M m O-M O 6 Lava flows Wason Park 5 100-1000 M M O m m O 0 29 Rhyolite Ash flows Lava flows 4 1-10 M O O m-M m m-M O 27 Lava flows Mammoth Mountain 3b 1 r i • 1 M m-0 O M m O-m 0 53 Rhyolite ) 100-1000 J Ash flows (phenocryst-rich) Mammoth Mountain 3a 1 M M O m m O 0 13 Rhyolite J L Ash flows (phenocryst-poor) Farmers Creek 2 10-100 M M O m m O 0 12 Rhyolite Ash flows Bachelor Mountain 1 10-1000 1—•—1 M M O m m O 0 12 Rhyolite Ash flows Figure 131.1.—Abundance and composition of phenocrysts in the volcanic sequence related to the Creede caldera source area. These rocks have porosity ranging generally from 2 to 10 percent; thus compaction effects are negligible for purposes of comparison of abundance of phenocrysts. Lines represent range of abundance; dots, average abundance; M, generally greater than 10 percent of total phenocrysts; m, generally less than 10 percent of total phenocrysts; O, phenocrysts not observed, or present only in trace amounts. ART. 131 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D49-D53. 1964. D49D50 MINERALOGY AND PETROLOGY tion changed from predominantly ash-flow eruption through alternating ash-flow and lava-flow eruptions to a final predominance of lava-flow eruptions. We interpret these changes as evidence of differentiation in the source magma chamber throughout the general period of eruptivity. Progressive changes in the Mammoth Mountain Rhyolite, a single ash-flow sheet, illustrate the stratified nature of the magma in the chamber during a particular eruptive stage. We shall first discuss the evidence from the entire sequence of flows, and then the details displayed by the Mammoth Mountain. DIFFERENTIATION EXHIBITED BY FLOW SEQUENCE The early phenocryst-poor rhyolites consist of ash flows of the Bachelor Mountain Rhyolite, Farmers Creek Rhyolite, and the older ash flows of the Mammoth Mountain Rhyolite. The phenocryst content of these ash flows ranges from about 2 to 12 percent. The phenocrysts are predominantly sodic plagioclase and sanidine in nearly equal amounts and some biotite and magnetite. The younger ash flows of the Mammoth Mountain Rhyolite range in composition from rhyolitic quartz latite to quartz latite and contain 15 to nearly 60 percent phenocrysts. With the change to quartz latite, sanidine decreases relative to plagioclase, and mafic phenocrysts become more abundant. Beginning with the Mammoth Mountain, lava flows are interlayered with the ash flows and become more abundant upward. Near the top of the sequence, the ratio of ash flows to lava flows is about 1:10. The bulk of the lava flows are quartz latites with an average of 30 to 40 percent phenocrysts, predominantly plagioclase of intermediate composition, although mafic minerals, including biotite, magnetite, pyroxene, and hornblende are abundant. Sanidine and quartz are present in some flows of the Fisher Quartz Latite which range in composition from rhyolite to rhyodacite. The two ash-flow units in the upper part of the sequence, the Wason Park Rhyolite and the Snowshoe Mountain Quartz Latite, range in composition from rhyolite to quartz latite and are rich in phenocrysts. The Wason Park contains about 30 percent phenocrysts, mainly plagioclase and sanidine with some biotite, magnetite, and a trace of pyroxene, and is definitely less mafic than the ash flows of the underlying Mammoth Mountain. The Snowshoe Mountain, the uppermost ash-flow unit, contains 30 to 60 percent phenocrysts; plagioclase is dominant, but mafic crystals-— biotite, magnetite, pyroxene, and amphibole—are more abundant than in any of the earlier ash-flow units. Sanidine and quartz also are present in small amounts. The mineralogic changes outlined above express progressive changes in chemical composition. Analyses of samples practically free from alteration effects, other than hydration of the glass of vitrophyres, are given in table 131.1. The changes in composition with time, or with the position in the sequence, are shown graphically in figure 131.2 by plotting the differentiation index (D.I.) of Thornton and Tuttle (1960) against silica in the various rock units. The D.I. is simply the sum of normative quartz, orthoclase, and albite, and is described by Thornton and Tuttle as a measure of the basicity of a rock and as an indication of the position of a rock along the path toward petrogeny’s residua system (Bowen, 1937). The chemical analyses (table 131.1) and the differentiation diagram (fig. 131.2) indicate that: (a) the younger units of the sequence are generally more mafic in composition than older units; (b) the ash flows and lava flows, considered as separate series, show a similar trend toward more mafic rocks in the younger units; (c) the lava-flow series (units 4, 6, and 8) is slightly out of step with the ash-flow series, as each lava-flow unit is more mafic than the ash-flow unit that follows in the order of eruption; (d) the Wason Park Rhyolite (unit 5) plots out of the order of eruption in the ash-flow series on the differentiation diagram because these ash flows are considerably more rhyolitic than the lava flows (unit 4) and phenocryst-rich Mammoth Mountain ash flows (unit 3b) that precede the Wason Park. COMPOSITIONAL STRATIFICATION IN THE MAMMOTH MOUNTAIN RHYOLITE The Mammoth Mountain Rhyolite is a thick composite ash-flow sheet representing many successive ash flows. It is exposed for at least 300° around the margin of the Creede caldera, and it ranges in thickness from 0, where it wedges out against a local high in the underlying topography northwest of Creede, to about 1,700 feet on the north side of the caldera. Although the full extent of the sheet is incompletely known, its original volume was well in excess of 100 cubic kilometers. The composite sheet changes laterally from a simple cooling unit, near its probable source within the caldera, through a compound cooling unit, to separate cooling units farther from the source. The successive ash flows of the Mammoth Mountain Rhyolite were not all coextensive; early ash flows were confined by an irregular topography and had a more limited distribution than some of the later ash flows. Several representative sections of the ash-flow sheet are located by letters on figure 131.3, and approximate thicknesses of the sheet are given in parentheses. The formation is exposed almost continuously east of the caldera between sections A and F, but the area near Rat Creek and Miners Creek stood high whenRATTE AND STEVEN D51 Table 131.1.—Chemical analyses and differentiation index of volcanic rocks in the sequence related to the Creede caldera source area Units (in order of eruption) 1 2 3a 3b 4 6 6 7 8 Bachelor Farmers Mammoth Mammoth Lava flows Wason Park Lava flows Snowshoe Fisher Quartz Creek Mountain Mountain Rhyolite Mountain Latite. Rhyolite Rhyloite Rhyolite Rhyolite Quartz Latite Ash flow Ash flow Ash flow Ash flow Lava flow Ash flow Lava flow Ash flows Lava flow (vitro- (vitro- (pheno- (pheno- (Larsen, (vitro- (Larsen, (average of 3, (vitro- phyre) phyre) cryst poor) cryst rich) 1956, table phyre) 1956, table with range) phyre). 23, No. 14, 23, No. 6. p. 190) p. 190) C-769 R-40 C-985b C-539a Lag 1598 C-341b Lag 1027 S-54-13; S-56A; R-284 S-56B 153631 154777 C-1164 151484 D-1423 D-1422; 153105; 153630 153106 Chemical constitu- ents: Si02 71.7 72.5 72. 69 67.2 65.29 69. 24 60. 54 62.9 (61.7-66.0) 60.0 AI2O3 - 13.0 13.7 13. 85 16.8 15. 56 13.96 16. 01 15.9 (15.4-16.5) 15.5 .8 .6 1.51 2.2 4.13 1.15 3. 63 3.4 (2. 0-4. 6) 3.6 FeO .85 .52 .18 .32 .55 .72 2. 34 1.7 (1.0-2. 2) 2.8 Mg .22 .11 .40 .66 .78 .49 2. 49 1.5 (1.3-1. 8) 3.0 CaO.. .78 .67 .89 2.2 3. 41 1.18 5. 44 4.0 (3.2-4. 6) 5.1 Na20 3.4 3.6 3.11 4.0 3.66 3.19 3. 69 3.6 (3. 5-3. 6) 3.6 K20 4.7 5.1 5. 46 5.0 4. 61 6.06 3. 80 3.9 (3. 6-4. 3) 2.6 H20+—— 4.0 3.1 .49 1.3 .94 2.65 1.07 }l.l (.88-1.42) r 2.i Hi0— .37 .1 .70 .54 .60 .27 .23 \ .25 T102 .19 .18 .25 .38 .85 .35 1.10 .59 (.54-. 68) .78 P205 .00 .02 .05 .10 .26 .07 .51 .25 (.22-. 29) .25 MnO 11 . 08 .04 .05 . 08 .11 (.09-. 14) .16 C02 <. 05 <. 05 .00 <. 05 .04 (. 04-1. 4) <. 05 Cl .03 .12 F .05 .05 91.8 93.7 91 79.8 87 72.9 (69-76.8 j flows) D. I.i (lava 77.4 66.3 62.4 Analysts and date P. L. D. El- P. L. D. El- D. Taylor, P. L. D. El- J. G. Fair- M. Seerveld; R. E. Bailey; S. D. Botts, P. L. D. El- of analysts more, I. H. more, I. H. O. H. Neu- more, S. D. child; 1925 1957 1923. Alka- H. Jtl. more, I. H. Barlow, S. Barlow, S. erburg; 1957 Botts, M. lies by F. Thomas, M. Barlow, S. D. Botts; D. Botts, D. Mack; A. Gonyer; D. Mack; D. Botts; 1958 G. Chloe; 1957 1942 1958 1958. 1959 1 Differentiation index (Thornton and Tuttle, 1960). 80 r O cr CL Z 70 3a \ \ 5\ \. 3b \ \ 4*\ Z 60 NT 50 100 90 80 70 60 50 DIFFERENTIATION INDEX (normative Q+Or+Ab) Figure 131.2.—Differentiation diagram, showing differentiation index versus silica in the volcanic sequence derived from the Creede caldera source area. Numbers refer to units listed in table 131.1, and indicate order of eruption. Dots represent ash flows, and X’s lava flows. 107°00' Figure 131.3.—Location map for representative sections (letters in circles) of Mammoth Mountain Rhyolite. Thickness of sections, in feet, shown in parentheses.D52 MINERALOGY AND PETROLOGY Table 131.2.—Modes of Mammoth Mountain Rhyolite Representative section (fig. 131.3) Description Number of modes Rock constituents (percent) Total phenocrysts Sanidine Plagioclase Biotite (avg) Magnetite (avg) Pyroxene (avg) Amphibole (avg) Avg Range Avg Range Avg Range A, B, C Phenocryst poor 13 8 5-11 3 2-4 4 2-5 ‘Phoenix Park Member SUBSIDENCE OF LA GARITA CAULDRON Outlet Tunnel MemberSTEVEN AND RATTE D57 mates of the age of the formation based on the same fossils have ranged from very late Miocene or very early Pliocene (Brown, R. W., in Larsen and Cross, 1956, p. 167) to early to middle Pliocene (MacGinitie, 1953, p. 60-75). Pollen studies by Estella Leopold (written communication, 1961) suggest an age between late Miocene through middle Pliocene. The rock units in the central San Juan Mountains near Creede, and their relations to recurrent cauldron subsidence, are summarized in table 132.2. The relations of these units to those from other centers of eruption in the San Juan Mountains are not as yet fully established. LA GARITA QUARTZ LATITE La Garita Quartz Latite is a new name here given to a part of the assemblage of rocks formerly included in the Alboroto Rhyolite by Larsen and Cross (1956, p. 132). The La Garita Quartz Latite is well exposed along the upper canyon of East Willow Creek north of Creede, and forms the bulk of La Garita Mountain farther northeast (fig. 132.1), which serves as type area for the formation. The La Garita Quartz Latite consists largely of identical densely welded ash-flow tuffs that are equivalent to the former Outlet Tunnel and Phoenix Park Quartz Latites of Emmons and Larsen (1923, p. 18, 30). The Outlet Tunnel unit was described by them as underlying rocks here called Bachelor Mountain Rhyolite and the Phoenix Park unit as interbedded with and over-lying them. In addition to the difference in stratigraphic position noted by Emmons and Larsen, we have determined that accumulation of the two units was separated by subsidence of the La Garita cauldron (table 132.2). We are calling those rocks that accumulated before the Bachelor Mountain Rhyolite and before cauldron subsidence the Outlet Tunnel Member, and those that accumulated concurrently with and after the Bachelor Mountain Rhyolite and after cauldron subsidence the Phoenix Park Member. Emmons and Larsen (1923, p. 30) erroneously believed that the bulk of these rocks along upper East Willow Creek and La Garita Mountain belonged to their Phoenix Park unit, whereas we have found that they formed before cauldron subsidence and clearly belong to the redefined Outlet Tunnel Member. Typical rocks from both members of the La Garita Quartz Latite appear identical and consist of crystal-rich ash-flow tuff, largely densely welded, that consists of about half reddish, lithoidal matrix, and half phenocrysts and fragments of foreign rock types. Plagioclase phenocrysts constitute about 30 percent of the rock, sanidine 6 percent, quartz 4 percent, biotite 4 percent, hornblende 4 percent, and magnetite and accessory minerals 2 percent. The Outlet Tunnel Member is more than 2,000 feet thick in the type area, but the base of the member is nowhere exposed in this vicinity and the full thickness could not be determined. The Phoenix Park Member is between 350 and 500 feet thick along the east flank of East Willow Creek, and it wedges out a short distance northeast against steep underlying topography. RHYOLITE OF MINERS CREEK The oldest rock unit known in the western part of the Creede area in the central San Juan Mountains is an accumulation of rhyolite flows and pyroclastic rocks more than 1,200 feet thick exposed along the flanks of Miners Creek. It is overlapped on its southeast margin by the Bachelor Mountain Rhyolite, and the abrupt character of the overlap suggests that the main mass of the underlying rhyolite may be a bulbous volcanic dome or steep-sided flow. The rock is a gray to pink porphyry consisting of about 20 percent phenocrysts and 80 percent lithoidal groundmass. Plagioclase and sanidine predominate among the phenocrysts, but sparse, small biotite crystals are widespread. A formal name is not proposed for the rhyolite of Miners Creek because there is little evidence of its areal extent or of its relations with rock units other than the Bachelor Mountain Rhyolite. BACHELOR MOUNTAIN RHYOLITE Bachelor Mountain Rhyolite is a new name given here to all the rhyolitic rocks included by Emmons and Larsen (1923) in their Willow Creek Rhyolite, Campbell Mountain Rhyolite, Windy Gulch Rhyolite Breccia, and intrusive rhyolite porphyry. Emmons and Larsen (1923, p. 19-30) described the Willow Creek and Campbell Mountain Rhyolites as two sequences of rhyolitic lava flows and flow breccias separated by an irregular erosion surface, and included both in their Alboroto Group. They believed (1923, p. 38-40) that the Windy Gulch Rhyolite Breccia was separated from the Campbell Mountain Rhyolite by a major unconformity, and that it belonged to their younger Piedra Group. Instead, we have found the Willow Creek, Campbell Mountain, and Windy Gulch units to be members of an intergradational sequence of pumiceous pyroclastic rocks that warrants formation rank. These rocks range upward from densely welded tuff closely resembling fluidal rhyolite flow rock (Willow Creek Member), through compact welded tuff in the middle (Campbell Mountain Member), to a porous moderately welded to non welded tuff at the top (Windy Gulch Member). Breaks within this sequence occur only where the Bachelor Mountain rocks intertongue with other volcanic rocks at the margins of the deposit.D58 MINERALOGY AND PETROLOGY Bachelor Mountain Rhyolite is named for discontinuous but representative exposures in the type area of the formation on Bachelor Mountain, northwest of Creede. No single complete section of the formation is exposed anywhere in the central San Juan Mountains. The greatest thickness observed of the Willow Creek Member is more than 2,000 feet thick, and the base is not exposed. More than 1,000 feet of the Campbell Mountain Member and more than 900 feet of the Windy Gulch Member are exposed locally, but the full thickness of either unit could not be measured. The Campbell Mountain and Windy Gulch Members intertongue laterally with flows and breccias of Shallow Creek Quartz Latite to the west, and a lens of typical Campbell Mountain welded tuff is interlayered with the Phoenix Park Member of the La Garita Quartz Latite to the east. The Bachelor Mountain Rhyolite consists mainly of microcrystalline to cryptocrystalline matrix with 5 percent or less of phenocrysts and 20 percent or less of foreign rock fragments. The phenocrysts are largely sanidine and plagioclase, with sparse biotite and magnetite. Inclusions of foreign rock fragments range from a few percent in the Willow Creek Member, to as much as 20 percent in parts of the Campbell Mountain and Windy Gulch Members. Eruption of the Bachelor Mountain Rhyolite culminated in collapse of the Bachelor Mountain cauldron (table 132.2) near the town of Creede, and in the development of a series of normal faults that extend north to northwest out from the more strongly subsided area. SHALLOW CREEK QUARTZ LATITE Shallow Creek Quartz Latite is a new name here applied to a group of lava flows and volcanic breccias that crop out in the drainage basins of Rat Creek, Miners Creek, and Shallow Creek 2 to 4}{ miles west of Creede. The flows and breccias are irregularly mixed, and the unit varies so greatly from place to place that no single section is representative; the largest known body, along the northern flank of the Shallow Creek drainage basin, is designated as the type area of the formation. Maximum thickness in this vicinity is 600 to 700 feet. The rocks included in the Shallow Creek Quartz Latite were originally described by Emmons and Larsen (1923, p. 36-38) under the informal name hornblende quartz latite. This unit marked the base of their Piedra Group in the western part of the Creede district, and they believed that it was separated from the underlying Alboroto rocks by an irregular erosional unconformity. Larsen and Cross (1965, p. 143) called the same rocks hornblende rhyolitic latites, and agreed with the earlier stratigraphic interpretations. Our mapping, on the other hand, has shown that the Shallow Creek Quartz Latite intertongues laterally with the upper members of the gradational sequence of rocks in the Bachelor Mountain Rhyolite, and that the supposed unconformity between rocks included by Emmons and Larsen (1923) in their Alboroto and Piedra Groups does not exist. The Shallow Creek Quartz Latite consists largely of volcanic breccias; lava flows are thin and are estimated to make up less than 10 percent of the formation. The rock is porphyritic, with small phenocrysts, rarely exceeding a few millimeters in maximum dimensions, set in an aphanitic groundmass. The phenocrysts consists mainly of plagioclase, 15 percent; hornblende, 5 percent; and a few percent each of biotite and magnetite. Sanidine and clinopyroxene occur locally. Much of the Shallow Creek Quartz Latite is irregularly altered, and either is oxidized and burnt in appearance, Or is irregularly altered to clay and crisscrossed with veinlets of quartz and carbonate. FARMERS CREEK RHYOLITE Farmers Creek Rhyolite is a new name here given to a heterogeneous assemblage of pyroclastic rocks and minor flow rocks that was deposited locally on the rough topography left by subsidence of the Bachelor Mountain cauldron. Larsen and Cross (1956, p. 146-147) refer the rocks here called Farmers Creek Rhyolite to the basal part of the lower member of their Piedra Rhyolite. The different rock types in the Farmers Creek Rhyolite are well exposed in its type area along Farmers Creek, 2)4 to 4% miles east of the town of Creede where a thickness of more than 1,000 feet is indicated. The Farmers Creek Rhyolite can be divided into three subequal parts in the type area and along West Bellows Creek to the east. The lower third is a soft, cavernous-weathering pumiceous tuff breccia that forms massive, unsorted layers a few feet to a few tens of feet thick. Foreign rock fragments constitute 15 to 26 percent of the rock. Some of the rocks are moderately welded, but most are only slightly welded or are nonwelded. A few lenses of red porphyritic flow breccia are interlayered with the predominant tuff breccias. The middle third of the formation consists of a succession of layers of differentially welded tuffs whose harder units form prominent ledges along the hill slopes. Massiye dark-brown densely welded tuff is the most widely exposed rock type. This grades both vertically and laterally into a lighter brown obviously fragmental welded tuff with good eutaxitic texture. Phenocrysts constitute 5 to 10 percent of the rock andSTEVEN AND RATTE D59 consist of sanidine, plagioclase, biotite, and magnetite. Foreign rock fragments constitute 5 to 10 percent of the rock. Crystal-rich welded tuff forming the upper third of the Farmers Creek Rhyolite is exposed only along the canyon of West Bellows Creek. Phenocrysts of plagioclase, sanidine, biotite, hornblende, clino-pyroxene, sphene, quartz, and magnetite are so abundant that the matrix is largely obscured and most of the rock appears structureless in hand specimen. The composition of these rocks appears anomalous with respect to the rest of the formation, and its inclusion in the Farmers Creek Rhyolite is provisional. MAMMOTH MOUNTAIN RHYOLITE Mammoth Mountain Rhyolite was originally named by Emmons and Larsen (1923, p. 40-44), who believed it to be a single thick lava flow. An overlying unit consisting largely of partially welded to nonwelded tuffs was mapped separately. Larsen and Cross (1956, p. 147) later assigned all these rocks, as well as the underlying Farmers Creek Rhyolite, to the lower rhyolite member of their Piedra Rhyolite. As redefined here, the Mammoth Mountain Rhyolite includes the Mammoth Mountain and the overlying soft tuffs of Emmons and Larsen (1923), as well as all laterally equivalent rocks, which together form a great composite ash-flow sheet (Smith, 1960, p. 158). The Mammoth Mountain Rhyolite is widespread in the central San Juan Mountains, and good exposures have been recognized as far as 15 miles southeast, 10 miles southwest, and 18 miles south of Creede. It is absent along the Continental Divide north of Creede, where mountainous terrain stood above the level of accumulation at the time the Mammoth Mountain ash flows were erupted. The formation is locally as much as 2,000 feet thick, and is 500 to 1,000 feet thick over wide areas. The Mammoth Mountain Rhyolite ranges from crystal-poor welded ash-flow tuffs with 5 to 10 percent phenocrysts to crystal-rich welded ash-flow tuffs with 30 to 60 percent phenocrysts. The matrix of the densely welded tuff is typically lithoidal and reddish-brown except for the black glass of the basal vitrophyre layer. Plagioclase phenocrysts are predominant in all of the rock. Sanidine phenocrysts are common in the crystal-poor rocks, but are absent in most of the crystal-rich phase. Biotite is the most prevalent mafic pheno-cryst in the Mammoth Mountain Rhyolite; green clinopyroxene phenocrysts are characteristic of and are restricted to the crystal-rich rocks. The ash-flow tuffs range in composition from rhyolite to quartz latite. WASON PARK RHYOLITE Wason Park Rhyolite is a new name given to a distinctive sheet of rhyolitic welded tuff that forms prominent cliffs throughout much of the central San Juan area near Creede. It forms the floor of Wason Park, a high flat bench on the southern flank of the La Garita Mountains northeast of Creede, which is taken as the type area. Wason Park Rhyolite is equivalent to the tridymite latite unit of Emmons and Larsen (1923, p. 45), and in its type area and in most other places was included by Larsen and Cross (1956, p. 152-153) in the tridymite' rhyolitic latite member of their Piedra Rhyolite. The Wason Park Rhyolite is 600 to 700 feet thick over wide areas in the central San Juan Mountains; it wedges out northward toward the Continental Divide. The formation is largely coextensive with the underlying Mammoth Mountain Rhyolite, except that it extends somewhat farther laterally over the rough underlying topography. A tongue of Huerto Formation separates the Mammoth Mountain and Wason Park Rhyolites south of Bristol Head, about 10 miles southwest of Creede, and several quartz latite lava flows intervene between the formations east of Wagon Wheel Gap, 8 to 15 miles southeast of Creede. Typical Wason Park Rhyolite is a crystal-rich densely welded tuff with 25 to 30 percent phenocrysts set in a reddish-brown lithoidal matrix. Distinctive white to light-gray tridymitic streaks representing collapsed pumice fragments are widespread and are characteristic of the formation. The phenocrysts are predominantly plagioclase and sanidine, with minor biotite and magnetite. A dense black vitrophyre 10 to 15 feet thick is generally present at the base of the formation, and the top locally retains a few feet to a few tens of feet of less welded rock. HUERTO FORMATION The name Huerto Formation was originally used by Patton (1917, p. 20), ana later by Emmons and Larsen (1923, p. 12) and Knowlton (1923, p. 184). The same rocks were called Huerto Andesite by Cross and Larsen (1935, p. 82), and Huerto Quartz Latite by Larsen and Cross (1956, p. 143); they believed that the Huerto rocks accumulated during a separate period of volcanic activity between eruption of their A1 boro to and Piedra Rhyolites. In contrast, we have found that the Huerto rocks were erupted from a number of separate but areally associated centers in the central San Juans concurrently with many of the units included by Larsen and Cross in their Piedra Rhyolite. Also, a local assemblage of lavas and breccias 6 to 10 miles southwest of Creede, 716-195 0—64 -5D60 MINERALOGY AND PETROLOGY called Conejos Quartz Latite by Larsen and Cross (1956, p. 99), is now knowm to be part of the compound Huerto accumulation. Although it eventually may be possible to separate out individual mappable units, present limited data make it expedient to refer to the volcanic accumulation as a whole as the Huerto Formation. The main mass of the Huerto Formation lies southwest of Creede, where three main centers of accumulation were reported by Larsen and Cross (1956, p. 143). Marginal rocks of the Huerto Formation intertongue with the widespread welded tuff units north and northeast of these centers. The dark lavas and breccias underlie Mammoth Mountain Rhyolite; a thin tongue of breccias intervenes locally between the Mammoth Mountain and Wason Park Rhyolites; and a tongue of lavas and breccias above the Wason Park Rhyolite extends nearly to Creede. The Huerto Formation thickens and thins markedly from place to place, and the different flow and breccia accumulations may originally have ranged widely in thickness. More than 2,000 feet of rudely bedded Huerto breccias is exposed on the west face of Bristol Head, southwest of Creede; the local accumulation of Huerto underlying the Mammoth Mountain Rhyolite south of Bristol Head is more than 800 feet thick; and the maximum preserved thickness of Huerto rocks over-lying the Wason Park Rhyolite near Creede is about 450 feet. Most rocks in the Huerto Formation are dark, finegrained lavas and breccias. They range widely in composition, but most appear to be dacitic. Phenocrysts are generally small and inconspicuous, and consist of predominant plagioclase and minor associated clino-pyroxene, hornblende, and magnetite. INTRACALDERA LAVA FLOWS Local thick quartz latitic to rhyolitic lava flows are interlayered with the widespread ash-flow sheets in the central San Juan Mountains (table 132.2). Most of these flows and some related volcanic necks are exposed around the outer margin of the Creede caldera, suggesting that this margin served to localize many of the eruptions. Most of the rocks are drab gray, coarsely porphyritic quartz latites; some lighter colored rhyolitic flows are interspersed with the quartz latite flows or occur singly. The oldest of these local lava-flow accumulations underlies the Mammoth Mountain Rhyolite east of Goose Creek and 5 miles south of Wagon Wheel Gap. A number of flows intervene between the Mammoth Mountain and Wason Park Rhyolites east and northeast of Wagon Wheel Gap, and a distinctive white rhyolite flow occupies the same position near Shallow Creek, northwest of the caldera. A quartz latite flow overlies the Wason Park Rhyolite and forms the floor of Silver Park northeast of the Creede caldera. Similar flows and a related neck are exposed about 5)4 miles north of Creede; these rocks are believed to underlie the Rat Creek Quartz Latite, but available evidence is not conclusive. RAT CREEK QUARTZ LATITE Rat Creek Quartz Latite as here redefined is a widespread sequence of welded and nonwelded ash-flow tuffs, with both the base and top of the formation marked by surfaces that show evidence of erosion, but with no ero-sional break recognized within the sequence. A local near-vent accumulation of lavas and pyroclastic rocks along West Willow Creek intertongues laterally with the more widespread ash-flow tuffs. As originally defined by Emmons and Larsen (1923, p. 57-58), the name Rat Creek Quartz Latite was applied only to the upper part of this sequence. They believed the rocks to be interlayered lava flows and tuffs, but recognized that they were closely related to the underlying tuff to which they gave the informal name “quartz latite tuff” (Emmons and Larsen, 1923, p. 54-57). Larsen and Cross (1956, p. 153-156) generally followed Emmons and Larsen in separating the lower tuff from the upper mixed unit. The lower unit was called the tuff member of their Piedra Rhyolite, and the upper mixed unit was lumped with the over-lying Nelson Mountain Quartz Latite of Emmons and Larsen (1923, p. 59) into the rhyolitic latite member of their Piedra Rhyolite. Hence their upper rhyolitic member included an erosional surface representing a significant break in deposition of the pyroclastic rocks, and their upper and lower members met at a gradational contact. The Rat Creek Quartz Latite is now found only in the high mountains extending east from the Spring Creek Pass area, 13 miles northwest of Creede, through the Nelson Mountain area north of Creede, to the La Garita Mountains area northeast of Creede. It is excellently exposed at Wheeler Monument, just south of Half Moon Pass. The formation consists largely of soft pyroclastic rocks, which typically form steep debris-covered slopes beneath the highest cliff-forming welded tuff unit in the area. The formation is 600 to 800 feet thick over wide areas, but ranges in thickness from 0 where it wedges out against older hills to nearly 1,000 feet locally. Most of the Rat Creek rocks are slightly welded to nonwelded ash-flow tuffs. A distinctive pink welded tuff 50 to 100 feet thick forms a prominent ledge in the middle of the unit north of Creede. Thin layers of air-fall tuff or reworked sedimentary tuff intervene locally between the ash-flow layers. The welded tuffsSTEVEN AND RATTE D61 contain 20 to 30 percent phenocrysts, chiefly plagio-clase, minor biotite and clinopyroxene, and sparse sanidine, quartz, hornblende, and magnetite. NELSON MOUNTAIN QUARTZ LATITE As redefined in this report, Nelson Mountain Quartz Latite includes all rocks in the former Equity Quartz Latite (Alboroto Group), and Nelson Mountain Quartz Latite (Piedra Group) of Emmons and Larsen (1923, p. 32-34, 59-60). Larsen and Cross (1956) did not mention the former Equity unit specifically, but show the area underlain by it as Alboroto Rhyolite on their geologic map, and the former Nelson Mountain Quartz Latite was included in the rhyolitic latite member of their Piedra Rhyolite. We have found that these units intergrade and form a single genetic unit to which we have extended the name Nelson Mountain Quartz Latite. The name Equity Quartz Latite is herewith abandoned. Nelson Mountain Quartz Latite characteristically forms resistant capping ledges on flat-topped ridges from the Nelson Mountain area north of Creede to the Snow Mesa area, 12 miles northwest of Creede, and on the flat crest of the La Garita Mountains eastward from the Half Moon Pass area 9 miles northeast of Creede. In addition, a great mass of Nelson Mountain Quartz Latite underlies the San Luis Peak area 9 miles north of Creede, where it apparently fills a cauldron structure. The widespread sheet of Nelson Mountain Quartz Latite ranges generally from 0 to 500 feet in thickness, but according to Larsen and Cross (1956, p. 132) nearly 4,000 feet of the unit (which they called Alboroto Rhyolite) is exposed in the San Luis Peak area. The Nelson Mountain Quartz Latite is a composite ash-flow sheet that changes from a simple cooling unit (Smith, 1960, p. 157) of predominantly densely welded tuft in the Nelson Mountain area, north of Creede, laterally, both to the east and the west into a compound cooling unit, and eastward at least into several individual cooling units that range from predominantly densely welded to predominantly non welded. Phenocrysts range from 20 to 50 percent of the Nelson Mountain Quartz Latite, and consist for the most part of plagioclase, with lesser quantities of biotite, and minor or accessory clinopyroxene, quartz, sanidine, hornblende, and magnetite. SNOWSHOE MOUNTAIN QUARTZ LATITE Snowshoe Mountain Quartz Latite is a new name given to the mass of quartz latite ash-flow tufts, mostly densely welded, that constitutes most of the core of the Creede caldera. The type area of the formation is Snowshoe Mountain which forms the highest part of the caldera core, south of Creede. Larsen and Cross (1956, p. 132, 138) believed that the great mass of quartz latite under Snowshoe Mountain was a local accumulation, possibly a single large flow, within the upper rhyolitic latite member of their Alboroto Rhyolite. In contrast, we have found the Snowshoe Mountain Quartz Latite to be the youngest major ash-flow unit in the central San Juan Mountains, and it is thus more equivalent to the upper part of the Piedra Rhyolite of Larsen and Cross (1956, p. 144-155). The Snowshoe Mountain Quartz Latite is restricted to the core of the Creede caldera, and the sum of several partial sections indicates that the formation is at least 4,000 feet and perhaps more than 6,000 feet thick. Talus breccias are intertongued marginally with at least the upper 2,000 feet of this unit, indicating that it accumulated within an existing depression. Caldera subsidence and accumulation are believed to have been concurrent. The Snowshoe Mountain Quartz Latite consists entirely of crystal-rich ash-flow tuff. Most of the rock is densely welded, and toward the center of the caldera only a few local less welded partings indicate the compound cooling characteristics of the deposit. Toward the southern part of the caldera, however, at least two different segments of the formation, each several hundred feet thick, soften laterally from densely welded to slightly or moderately welded. The top of the formation is everywhere marked by less welded rocks that range up to several hundred feet in thickness. The Snowshoe Mountain Quartz Latite is a drab crystal rich welded tuff that closely resembles much of the more densely welded rock in the Nelson Mountain Quartz Latite. Snowshoe Mountain rocks consist of an average of nearly half phenocrysts, and have a distinctly granular aspect in hand specimen. As in the other quartz latitic ash-flow units, plagioclase predominates among the phenocrysts, followed in abundance by biotite, and by minor clinopyroxene, hornblende, quartz, sanidine, and magnetite. FISHER QUARTZ LATITE As used by Larsen and Cross (1956, p. 172), the Fisher Quartz Latite (or latite-andesite) was a sequence of local accumulations of coarsely porphyritic flows and breccias in the San Juan Mountains and was believed to be younger than the Potosi Volcanic Series and the Creede Formation, and older than the Hinsdale Formation. Larsen and Cross (1956, p. 172) state that these rocks have certain textural and mineral-ogical characteristics which distinguish them from rocks of the Potosi Volcanic Series, and that at different places they rest unconformably on the Silverton Volcanic Series, the Potosi Volcanic Series, and the Creede Formation. Rocks assigned to this formationD62 MINERALOGY AND PETROLOGY occur throughout the central San Juan Mountains. Our work has shown that this earlier concept must be revised in important aspects. The rocks in the type area of the formation on Fisher Mountain were erupted after the last main subsidence of the Creede caldera and followed doming of the caldera core. We therefore restrict the Fisher Quartz Latite to lavas of this age. Some Fisher flows around the Creede caldera underlie the Creede Formation, others intertongue with it, and still others appear somewhat younger; the bulk of the Fisher Quartz Latite, however, appears to have been erupted concurrently with Creede sedimentation. The evidence for the Fisher being equivalent in age to the Creede Formation directly contradicts Larsen and Cross (1956, p. 168, 172), who believed that the Fisher lavas flowed down deep valleys cut into the Creede. Many Fisher rocks appear identical with coarsely porphyritic flows interlayered with earlier widespread ash-flow sheets adjacent to the Creede caldera (Piedra Rhyolite of earlier terminology), and Larsen and Cross (1956, pi. 1) have included several of these older flows in their Fisher Quartz Latite. The only clear distinction between these different flows is in age relative to caldera subsidence. Thus the contention of Larsen and Cross (1956, p. 172) that Fisher Quartz Latite is distinct from all rocks in the Potosi Volcanic Series in time and petrologic character does not appear tenable, and in our opinion the Fisher eruptions represent merely a continuation of a related sequence of volcanic events. Most Fisher Quartz Latite near Creede was erupted around the margin of the Creede caldera, or along grabens extending outward from the caldera. The largest mass underlies Fisher Mountain and adjacent areas south of the caldera; another substantial body accumulated in the Wagon Wheel Gap area 5 to 8 miles southeast of Creede. A smaller accumulation is exposed on the west flank of the caldera core 7 miles southwest of Creede, and several flows cap the ridge between upper Rat and Miners Creeks northwest of Creede. Possibly some of the other relatively late accumulations of coarsely prophyritic lava through the central San Juan Mountains correlate with the Fisher Quartz Latite as we have redefined it, but we would prefer not to extend the name to these until their age relations can be established more definitely. Fisher Quartz Latite consists of coarsely porphyritic lava, predominantly quartz latite but with some rhyolite. Plagioclase, biotite, and clinopyroxene are the most abundant phenocrysts, and small quantities of magnetite are invariably present. In addition, some flows contain significant quantities of hornblende, sanidine, and quartz. CREEDE FORMATION The Creede Formation was originally named and described by Emmons and Larsen (1923, p. 61-70), who divided it into a lower member consisting largely of fine-grained well-bedded deposits with interbedded travertine, and an upper member consisting of coarser material, largely stream deposits, and a few thin lava flows. Larsen and Cross (1956, p. 167-172) closely followed the earlier description. We also include the same assemblage of rocks in the Creede Formation, but have found that the earlier recognized subdivisions are in a large part laterally equivalent facies not readily separated. The Creede Formation consists largely of sedimentary rocks that were deposited in a structural trough around the margin of the domed core of the Creede caldera. These deposits follow the western, northern and eastern margin of the caldera for nearly 270° of arc, from Lime Creek, down the Rio Grande to Wagon Wheel Gap, and up Goose Creek for nearly 5 miles. The original thickness of the formation is not known, but present remnants extend over a vertical range of at least 2,400 feet. The Creede Formation consists of several distinct facies that are so complexly intermixed that it is not practical to separate them. Thin-bedded tuffaceous lake beds, largely shale and sandstone with some tuff beds, intertongue marginally with fluviatile sandstone and conglomerate beds in the vicinity of old tributary valleys, and with fanglomerate and coarse sedimentary breccias of local derivation elsewhere along the margins. The edge of the formation against buried rough topography is commonly marked by accumulations of talus, landslide, slopewash, and similar debris. Travertine and calcareous tufa were deposited widely by mineral springs that were active concurrent with sedimentation. Some volcanic breccia and a few layers of welded tuff are interlayered locally. REFERENCES Cross, Whitman, and Purington, C. W., 1899, Description of the Telluride quadrangle [Colorado]: U.S. Geol. Survey Geol. Atlas, Folio 57. Cross, Whitman, Howe, Ernest, and Ransome, F. L., 1905, Description of the Silverton quadrangle [Colorado]: U.S. Geol. Survey Geol. Atlas, Folio 120. Cross, Whitman, and Larsen, E. S., 1935, A brief review of the geology of the San Juan region of southwestern Colorado: U.S. Geol. Survey Bull. 843, 138 p. Emmons, W. H., and Larsen, E. S., 1923, Geology and ore deposits of the Creede district, Colorado: U.S. Geol. Survey Bull. 718, 198 p.STEVEN AND RATTE D63 Knowlton, F. H., 1923, Fossil plants from the Tertiary lake beds of south-central Colorado: U.S. Geol. Survey Prof. Paper 131, p. 183-192. Larsen, E. S., Jr., and Cross, Whitman, 1956, Geology and petrology of the San Juan region, southwestern Colorado: U.S. Geol. Survey Prof. Paper 258, 303 p. Luedke, R. G., and Burbank, W. S., 1963, Tertiary volcanic stratigraphy in the western San Juan Mountains, Colorado: Art. 70 in U.S. Geol. Survey Prof. Paper 475-C, p. C39-C44. MacGinitie, H. D., 1953, Fossil plants of the Florissant beds, Colorado: Carnegie Inst. Washington Pub. 599, 198 p. Patton, H. B., 1917, Geology and ore deposits of the Platoro-Summitville mining district, Colorado: Colorado Geol. Survey Bull. 13, 122 p. Smith, R. L., 1960, Zones and zonal variations in welded ash flows: U.S. Geol. Survey Prof. Paper 354-F. Smith, R. L., and Bailey, R. A., 1962, Resurgent cauldrons— their relation to granitic ring complexes and large volume rhyolitic ash-flow fields, [abs.] in International Symposium on Volcanology, Japan: Internat. Assoc. Volcanology, 1962, p. 67-68. Steven, T. A., and Ratt6, J. C., 1960a, Geology and ore deposits of the Summitville district, San Juan Mountains, Colorado: U.S. Geol. Survey Prof. Paper 343, 70 p. ------- 1960b, Relation of mineralization to caldera subsidence in the Creede district, San Juan Mountains, Colorado: Art. 8 in U.S. Geol. Survey Prof. Paper 400-B, p. B14-B17.Article 133 VALLERIITE AND THE NEW IRON SULFIDE, MACKINAWITE By HOWARD T. EVANS, JR., CHARLES MILTON, E. C. T. CHAO, ISIDORE ADLER, CYNTHIA MEAD, BLANCHE INGRAM,- and RICHARD A. BERNER,-1 Washington, D C.,- La Jolla, Calif. Abstract.—Valleriite from Loolekop, South Africa, and Kavel-torp, Sweden, is rhombohedral, space group Rim or Rim, with a=3.792 A, and c=34.10 A, and its probable unit-cell content is 6CuFeS2. Mackinawite, a new copper-free iron sulfide from the Mackinaw mine, Snohomish County, Wash., is tetragonal, space group P4/nmm, with a = 3.679 A, c=5.047 A, and a unitcell content of 2FeS. Both minerals evidently have layer structures, in accord with their extreme optical anisotropy. Many occurrences of mackinawite have probably been mistaken for valleriite. Few minerals have caused mineralogists more justified uncertainty or unsound assurance than the sulfide valleriite. Since its discovery at Nya Kopparberg, Sweden, by Blomstrand (1870) almost a century ago, it has led an uneasy life in the mineralogical literature. Its first description was encumbered with a complex and implausible proposal for composition, which led Petren (1898) to reject the work of Blomstrand and consign the mineral to the limbo of species discredited as mixtures. It was revived by Ramdohr and Odman (1932) on the basis of a restudy of the Swedish material already described by Blomstrand and Petren. Only recently has the existence been suspected of another distinct sulfide phase occurring in mineral assemblages similar to those which contain valleriite (Milton and Milton, 1958; Kouvo and Vuorelainen, 1959). This new phase, which we designate mackinawite, is so similar in physical properties to valleriite that it has evidently been the object of a great deal of confusion. In this article we present new data for valleriite and descriptive data for a previously undefined mineral which we call mackinawite; and we take note of the difficulties that have arisen from confusion of the two minerals. NEW DATA FOR VALLERIITE We have not had available Blomstrand’s type material from Nya Kopparberg, but our specimen from 1 Scripps Institution of Oceanography; work done in preparation of a thesis at Harvard University, Cambridge, Mass. Kaveltorp, described in detail by Odman (1933), is closely similar. The apparently homogeneous bronzy graphitic mineral gives a characteristic powder pattern, but attempts to obtain single-crystal patterns from tiny flakes yielded such poor information that a unit cell could not be deduced. Fortunately, new material of much better quality from Loolekop in northeast Transvaal, South Africa, recently became available to us through the courtesy of L. R. Page, of the U.S. Geological Survey. The mineral occurrence, in small pea-sized masses in carbonate rock associated with other iron-copper sulfides, is quite like that of the Swedish mineral, but the crystal flakes are generally larger and better formed. The chemical composition is entirely analogous to that found for the Kaveltorp material, and to that given by Blomstrand (see table 133.1), showing again the anomalous presence of about Table 133.1.—Chemical analyses of valleriite Constituent Sample 1 £ 3 4 Cu 17.7 17. 6 19. 8 18. 6 Te 26. 3 21. 2 20. 0 21. 1 S 22.5 21. 4 21. 6 21. 3 5. 1 8. 1 8. 5 MgO 10. 6 16. 2 16. 0 CaO 0. 3 1. 7 1. 3 K20 0. 3 Na20 0. 6 H20 ... 10. 8 12. 2 10. 8 Insoluble + Si02 1. 8 3. 3 94 2 100 2 101. 3 1. Nya Kopparberg, Sweden; Blomstrand (1870), from 5 partial analyses. 2. Kaveltorp, Sweden; Blanche Ingram, analyst; MnO, 0.5 percent. 3. Loolekop, South Africa; Blanche Ingram, analyst; MnO, 0.5 percent. 4. Theoretical for CuFeSa, corresponding to 61.0 percent of total. the same amount of aluminum and magnesium hydroxides together with iron, copper, and sulfur. The characteristic X-ray powder-diffraction pattern, shown in table 133.2, conclusively establishes the identity ART. 133 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D64-D69. 1964. D64EVANS, MILTON, CHAO, ADLER, MEAD, INGRAM, AND BERNER D65 Table 133.2-—X-ray powder-diffraction data for valleriite Calculated 1 Loolekop 2 Kaveltorp 2 hkl d(A) d(A, obs.) J d(A, obs.) I 00. 3 11. 367 11. 39 10 11. 48 9 00. 6 5. 683 5. 71 10 5. 68 10 00. 9 3. 789 3. 80 5 10. 1 3. 269 3. 27 6 3. 27 7 10. 2 3. 225 3. 23 5 10. 4 3. 064 3. 07 2 10. 5 2. 959 2. 958 1 2. 974 i 00. 12 2. 842 2. 846 5 2. 842 5 10. 7 2. 723 10. 5 2. 601 2. 604 i 2. 531 3 10. 10 2. 365 2. 346 2 2. 364 1 00. 15 2. 273 2. 275 6 10. 11 2. 254 2. 259 2 10. 13 2. 050 2. 041 2 10. 14 1. 956 11. 0 1. 896 00. 18 1. 894 I. 885 5 1. 894 6 11. 3 1. 870 1. 860 5 1. 870 5 11. 6 1. 799 10. 16 1. 788 1. 780 i 1. 794 i 10. T7 1. 712 11. 9 1. 696 20. 1 1. 640 20. 2 1. 634 1. 636 l 00. 21 1. 624 1. 629 i 1. 620 l 20. 4 1. 612 20. 5 1. 596 11. 12 1. 577 10. 19 1. 575 20. 7 1. 556 20. 8 1. 532 L 526 i 1. 534 2 10. 20 1. 513 20. 10 1. 479 1. 485 i 11. 15 1. 456 20. 11 1. 451 -- — — — -- — -- Kaveltorp 3 * * * * d(A, obs.) I 11. 5 10 5. 75 10 3. 83 8 3. 29 10 3. 23 8 3. 08 2 2. 87 8 — -- 2. 48 2 2. 37 8 2. 29 8 2. 05 6 1. 91 10 1. 90 10 1. 87 10 1. 80 8 — -- — -- 1. 64 4 — -- 1. 59 6 — -- — -- 1. 54 6 1. 52 2 1. 50 2 1. 47 2 1. 23 2 1. 22 3 1. 14 4 1. 09 4 1. 06 6 1. 04 10 1. 03 10 1. 01 10 • Calculated for unit cell given in table 3. 2 Patterns made by the authors with FeKa radiation, camera diameter 114.6 mm. Intensities are on a geometric scale. 3 Data from Hiller (1939). of the South African and Swedish minerals. The powder-diffraction data given by Berry and Thompson (1962) for Kaveltorp valleriite agrees well with that obtained by us. Single flakes isolated from specimens from Loolekop, South Africa, were studied with the Buerger precession camera in an attempt to obtain single-crystal diffraction patterns. The best photographs were readily interpreted in terms of a rhombohedral unit cell with the properties shown in table 133.3. The similarity in symmetry and dimensions of the valleriite unit cell to that of the recently described iron sulfide, smythite, is striking and significant, although the character of diffraction-intensity distribution on the single-crystal patterns precludes any type of isostructural relationship. The crystallographic studies of Hiller (1939) are not supported by these findings. On the basis of powder Table 133.3.— Unit-cell data for valleriite, smythite, and mackinawite Mineral species Property Valleriite1 Smythite2 Mackinawite8 Crystal sys- Rhombohedral Rhombohedral Tetragonal tern. Space group . R3m or R3m i?3m P4/nmm a (A)_______ 3. 792 ±0.005 3.47 3. 673 ±0. 001 c (A)_______ 34. 10 ±0.05 34.5 5. 035 ±0.002 Cell content 6CuFeS2 3Fe3S4 2FeS Density (X- 4. 26 4. 09 4. 30 ray). Density 3. 14 4. 06 ----------- (measured). 1 Loolekop, South Africa; unit cell derived from least-squares analysis of first 12 d-spacings of data for Loolekop material, table 133.2; density from Blomstrand (1870). 2 Bloomington, Iijd.; Erd and others (1957). 3 Synthetic, Berner (1962); parameters derived by least-squares analysis of Berner’s diffraction data as shown in table 133.4. data alone, which correspond well with ours (table 133.2), he deduced an orthorhombic unit cell and even a crystal structure for valleriite. Evidently, his overextended hypotheses need not be further considered. The crystal structure of valleriite is not yet known, but its crystallography places severe restrictions on its constitution. Just as in smythite, the space-filling sulfur atoms must lie on the threefold symmetry axes, and the hexagonal unit cell is sufficiently large to contain three S4 units. Accordingly, the c axis length of 34.3 A corresponds to 12 layers of sulfur atoms in closest packing. Further, it is readily deduced that not less than 3 nor more than 4 cations also lie on the threefold axes. This reasoning leads us to the possible formulas: CuFe2S4, Cu2FeS4, and Cu2Fe2S4. The chemical analyses, apart from the magnesium and aluminum hydroxides, conform closely with the last. Thus it appears most probable that valleriite is a dimorph of chalcopyrite, CuFeS2. All single-crystal patterns of valleriite show the presence of a second, much weaker lattice originating from some other phase. This lattice is also rhombohedral, with a hexagonal a axis of about 3.0 A. Its smaller dimension corresponds well to that expected of an oxide compound of some sort. There is little doubt that this lattice is to be associated with the magnesium and aluminum hydroxides which have always appeared in chemical analyses of valleriite. Evidently these hydroxides correspond to a foreign mineral phase, hexagonal or rhombohedral in crystal character, interleaved in syntactical orientation on a submicroscopic scale with the valleriite. The identity of this foreign phase and its genetic relationship to valleriite remain a mystery. MACKINAWITE Recently, Milton and Milton (1958) described the occurrence of a sulfide mineral from the MackinawD66 MINERALOGY AND PETROLOGY mine in Snohomish County, Wash., which they identified as valleriite, mainly on the basis of its physical properties. The identification was tentative, and the doubt which arose late in their study was explained in a footnote: Preliminary studies in progress by C. Milton, E. C. T. Chao and H. T. Evans, Jr. indicate that the mineral identified as valleriite is this paper is, in fact, distinct from the type valleriite from Kaveltorp, Sweden. The Mackinaw mineral is probably an undescribed iron sulfide, the Kaveltorp valleriite is perhaps a copper-iron-magnesium sulfide. Probably both these phases, if not others also, are to be found among material from other localities that has been called valleriite. Both valleriite and mackinawite are extremely anisotropic and strongly reflection pleochroic. This feature is so outstanding that ore microscopists for many years have frequently labelled phases as “valleriite” on the basis of its anisotropism alone. We have observed that the pleochroism differs slightly between the two, valleriite appearing pale yellow to deep creamy brown (Skinner and Milton, 1955), while mackinawite shows a pale pink to pinkish gray color. The existence of a mineral different from valleriite was first definitely suspected when enough powder was successfully isolated from a chalcopyrite matrix by means of an ultrasonic vibrating needle (Kehl and others, 1957) to obtain a characteristic X-ray diffraction pattern. The pattern showed that a large amount of chalcopyrite and a small amount of cubanite were still present in the sample, but when the characteristic reflections for these minerals were discounted, 14 reflections remained to be associated with the new phase, as shown in table 133.4. These d-spacings bore no relation to those of valleriite. Examination of individual grains of the mineral in a polished section with an electron-microprobe apparatus by Birks and others (1959) showed that the copper content is very low and that the composition of the mineral is approximately FeS. This finding has been confirmed by new analyses with the electron-probe apparatus in our laboratory. Figure 133.1 (upper left) shows a polished section of ore from the Mackinaw mine containing feathered inclusions of mackinawite, which show as gray regions in the light-gray chalcopyrite matrix. Dark round spots indicate points probed by the electron beam and analyzed by X-ray spectroscopy. Copper was not detected, and on the basis of 16 determinations the following composition was determined by the authors: Measured, Theoretical (weight percent) (Feo.#«Nio.o4S) Fe 63±5 60. 5 Ni 3.1 ±0.5 3. 0 S 34±4 36. 4 determinations were made by comparison Table 133.4.—X-ray powder-diffraction data and unit-cell parameters for mackinawite Least- Synthetic 1 Mackinaw Outokumpo3 “Kansite”4 squares parameters: 3 mine -- 3. 673±0. 001 3.675±0. 002 3. 6773±0. 003 c__ 5. 035±0. 002 5.030±0.003 5.0217±0. 0007 hkl d(A,calc.)® d(A,obs.) I d(A,obs.) / d(A,obs.) I d(A,obs.) I 001.... ... 5.035 5.03 100 5.03 100 5.020 VS 5.05 50 101.... ... 2.967 2.97 80 2. 96 70 2. 966 s 2.99 50 110.... ... 2.597 2. 60 20 2.600 vw 002.... ... 2.518 Ill ... 2.308 2.305 80 2. 31 90 2. 309 s 2.32 100 102.... ... 2.077 200.... ... 1.8365 1.835 60 1.838 50 1.838 M 112.... ... 1.8077 1.805 80 1.809 80 1.806 s 1.80 100 201.... ... 1.7253 1.723 60 1.729 50 1.726 M 1. 73 50 003.... ... 1.6784 1.677 20 1.674 W 211.... ... 1.5616 1. 564 40 1.564 20 1.563 MW 1.54 10 103.... ... 1.5265 1.527 20 1.523 W 202 ... 1.4837 1.481 20 113.... ... 1.4096 1.410 30 1.410 30 1.408 W 1.42 30 212 ... 1.3757 220.... ... 1.2986 1. 298 50 1.300 MW 1.31 50 004.... ... 1.2588 1.258 50 1.257 35 1. 259 MW 1.26 50 221.... ... 1.2574 203.... ... 1.2389 1.239 30 1.237 30 1.238 W 104.... ... 1. 1908 1. 190 10 1.191 VW 301.... ... 1. 1897 213.... ... 1.1739 1. 174 20 1.173 VW 310.... ... 1.1615 222 ... 1.1541 114.... ... 1.1327 1.133 50 1. 132 35 1.133 MW 311.... ... 1.1318 302 ... 1.1010 312 ... 1.0547 1.055 80 1.055 40 1.055 MS 204.... ... 1.0383 1. 038 20 1.040 B 223.... ... 1.0271 1.027 B 005.... ... 1.0070 214.. .. 321.. .. 303.. .. ... .9991 ... .9985 ... .9891 --- --- .9994 B — — ::: —- 1 Berner (1962). 3 Measured by E. C. T. Chao on material removed from chalcopyrite matrix with an ultrasonic needle; 13 reflections due to chalcopyrite and 5 due to cubanite are omitted. FeKa radiation, 114.6-mm-diameter camera. 3 Kouvo and others (1963). V=very strong, S=strong, M=medium, W=weak, B = broad. 4 Artificial product, measured by Meyer and others (1958); ASTM powder data file No. 7-26. ! Unit-cell parameters and standard deviations calculated from the powder-diffraction data by least-squares analysis. 6 d-spac.ings calculated for the unit cell found for Berner’s synthetic FeS, assuming the space group P\\nmm. analyzed pyrite, pyrrhotite, and pentlandite as primary standards. Accuracy was limited by the small size of the grains, which are only a little larger than the electron beam, and the difficulty of obtaining quite plane surfaces on the relatively soft mackinawite. It is apparent that the determined composition is not significantly different from that corresponding to the formula, (Fe0.9SNi0 04)S. The key distinguishing characteristic of this mineral, namely, its X-ray powder-diffraction pattern, was soon found to resemble that of an artificial iron sulfide labelled “kansite” by Meyer and others (1958). This substance was identified by its X-ray pattern in the corrosion product caused by the action of hydrogen sulfide on pipeline steel. It could not be isolated for ordinary chemical and physical tests, but it did yield a rather diffuse X-ray pattern of 10 measurable lines (see table 133.4). These matched approximately the recorded pattern for pentlandite, and Meyer and others derived from it a cubic unit cell with a=10.1 A. Following this lead, they suggested a composition of Fe9S8 for this phase, by analogy with pentlandite. Our X-ray data (table 133.4) showed a close relationshipEVANS, MILTON, CHAO, ADLER, MEAD, INGRAM, AND BERNER D67 Figure 133.1.—Polished section showing mack-inawite from the Mackinaw mine, Snohomish County, Wash. The upper left view (plain light) shows cubanite (Cub in lower left drawing) as medium gray, chalcopyrite (Ch) as light gray, and the silicate gangue as black. The mackinawite in the chalcopyrite appears as dark-gray streaks that are brightly illuminated between crossed nicols, as shown in the upper right view. Four mackinawite areas probed by the electron beam are shown by p in the lower left drawing and as black-stained spots in the upper left view.D68 MINERALOGY AND PETROLOGY with “kansite,” but on the other hand it was certain that the Mackinaw mine iron sulfide could not be cubic because of its high anisotropy in polished section (fig. 133.1, upper right). Meanwhile, Kouvo and Vuorelainen (1959) published a short paper describing a new iron sulfide mineral resembling vallerite but distinct from it. They list the X-ray diffraction data obtained from this mineral, consisting of 23 lines (table 133.4). These contain the lines found by us for the Mackinaw mineral and also, as they noted, for “kansite,” as well as many other lines. They found the mineral in chalcopyrite assemblages, showing strong pleochroism and very strong optical anisotropy, quite analogous to the Mackinaw mineral. Evidently, Kouvo and Vuorelainen were able to prepare pure samples from which they could obtain sharper and stronger X-ray patterns than we have been able to get. Subsequently, evidence for the formation of an iron sulfide phase in the sediments of the Mystic River, Boston, Mass, was found by Berner (1962). He noted that this phase, as identified by its X-ray diffraction pattern, is the same as one which he prepared synthetically by the action at room temperature of an aqueous solution of H2S on reagent-grade iron wire in the absence of air. The pattern corresponds well with those of the Mackinaw mine and Outokumpo minerals, and “kansite.” Berner discovered that all these patterns can be quite satisfactorily indexed on the basis of a primitive tetragonal unit cell with parameters which he gave as a=3.679±0.002 A and c=5.0471 ±0.002 A. It was found that this crystallography closely resembled that described by Hagg and Kindstrom (1933) for FeSe, for which they determined the crystal structure. Berner further showed that a similar structure for tetragonal FeS yielded calculated diffraction intensities which agreed well with those observed for the natural and artificial iron sulfide phase. Thus, the key to the constitution of these phases is provided. Kouvo and others (1963) have now described their study of the new iron sulfide in detail, based on beautifully crystallized material associated with cubanite, pyrrhotite, and chalcopyrite from Outokumpo and other Finnish localities. Single crystals up to 1 mm in size allowed these authors to determine the unit-cell dimensions and symmetry by the Weissenberg method. They found a tetragonal unit cell in the probable space group P4/nmm, with parameters which they reported as a=3.676 ±0.002 A and c=5.032 A. Kouvo and others (1963) also were able to obtain good chemical and electron-probe analyses leading to a formulation (Fe0.9iNio.i3Coo,oo7)S, with no Cu present. The slight departure from stoichiometry is also reminiscent of the behavior of other iron sulfide phases. The close correspondence between the chemical and physical data given by Kouvo and others for their “tetragonal iron sulfide” and those found by us for the Mackinaw material leaves no doubt that the two are the same mineral. We have named the mineral “mackinawite” in allusion to the locality name. Dr. Olavi Kouvo, who gave the first clear description of this new mineral (Kouvo and others, 1963), has generously agreed to this choice of name. The X-ray powder-diffraction data given by Berner for synthetic iron sulfide and by Kouvo and others for the Mackinaw mine material are given in table 133.4. We have subjected all these data to least-squares analysis in order to find the best unit-cell parameters for each set. For this purpose we have used a new self-indexing computer program written for our Burroughs B220 digital computer, which extracts the standard deviations of the parameters from the inverse matrix of the normal equations based on the residuals of the measured 26 angles. These unit-cell parameters and standard deviations are shown at the beginning of table 133.4. The result for Berner’s data is given in table 133.3 to represent the pure FeS phase. The differences between the cell parameters as we found them and the published values are hardly significant. The variation in the three sets of cell parameters doubtless reflects the influence of the substitution of varying amounts of Ni for Fe, but the data available now are not sufficient to define this variation quantitatively. CONFUSION OF VALLERIITE AND MACKINAWITE To the ore microscopist, the outstanding character extreme their in istic of both valleriite and mackinawite is their anisotropism. Ramdohr and Odman (1932) restudy of Blomstrand’s material evidently considered the evidence of this striking anisotropism to Ibe more significant than the available chemical information and assumed on this basis that the material was identical with a mineral found by Schneiderhohn (1929) in the platinum-ore deposits of the Transvaal. In his careful studies, Schneiderhohn noticed in sections of chalcopyrite and pentlandite, lamellar or irregular inclusions of a highly birefringent mineral, which he termed an “unbekanntes Nickelerz.” Most important, he was able to isolate enough pure sample from a pentlandite matrix for a spectrographic analysis, which showed that the anisotropic mineral was primarily an iron-nickel sulfide, with no detectable copper. All aspects of his description match the properties of mackinawite as we now know them, and it seems entirely probable that mackinawite is actually the mineral he describes in “Lehrbuch der Erzmikroskopie” (Schneiderhohn and Ramdohr, 1931), without mentioning the Swedish mineral, and without giving it a name. Ramdohr, in subse-EVANS, MILTON, CHAO, ADLER, MEAD, INGRAM, AND BERNER D69 quent editions of his “Die Erzmineralien” (Ramdohr, 1960), combined the information about the South African and Swedish occurrences into one description under the name “Valleriit.” Consequently, most subsequent descriptions of a soft sulfide mineral having extreme anistropism and occurring as fine inclusions in chalcopyrite or pentlandite have given it the name “valleriite,” even though no other evidence was available to establish the identification. Thus, the numerous literature references to valleriite are ambiguous and could equally well apply to mackinawite. In conclusion, we believe that the occurrence in other sulfide crystals of tiny inclusions giving the appearance of an exsolution phase is quite characteristic of mackinawite and not valleriite, and that as this feature is usually mentioned in most literature references to “valleriite,” probably what these references describe is actually mackinawite. Authentic valleriite, on the other hand, has been established from only two localities, Loolekop in South Africa, and southern Sweden. A probable third locality has been reported by Skinner and Milton (1955) at the Elizabeth mine, South Stratford, Vt. X-ray powder patterns of a soft highly birefringent sulfide mineral from this mine show reflections for chalcopyrite, pyrrhotite, and a third phase. The d-spacings for this phase match those determined for Kaveltorp valleriite, and the data given by Hiller (1939). Kouvo and Vuorelainen (1959) also mention an occurrence of valleriite at Vihauti, Finland, verified by its diffraction pattern. The validity of any given occurrence of valleriite or mackinawite may evidently be established only by X-ray diffraction or electron-microprobe techniques. From what we know now, it appears that mackinawite is a common and widespread mineral, while valleriite is quite rare. REFERENCES Berner, R. A., 1962, Tetragonal iron sulfide: Science, v. 137, p. 669. Berry, L. G., and Thompson, R. M., 1962, X-ray powder data for ore minerals; The Peacock atlas: Geol. Soc. America Mem. 85, p. 61. Birks, L. S., Brooks, E. J., Adler, Isidore, and Milton, Charles, 1959, Electron probe analysis of minute inclusions of a copper-iron mineral: Am. Mineralogist, v. 44, p. 974-978. Blomstrand, C. W., 1870, On some new Swedish minerals and the composition of pyrrhotite: Ofversigt af Kungl. Vetenskaps-Akademiens Forhandlingar (Stockholm), v. 27, p. 19-27. [In Swedish] Erd, R. C., Evans, H. T., Jr., and Richter, D. H., 1957, Smythite, a new iron sulfide, and associated pyrrhotite from Indiana: Am. Mineralogist, v. 42, p. 309-333. Hagg, G., and Kindstrom, A.-L., 1933, X-ray study of the system iron-selenium: Zeitschr. fur physik. Chem., v. B22, p. 453-464. [In German] Hiller, J. E., 1939, On the crystal structure of valleriite: Zeitschr. Kristallographie, v. 101, p. 425-434. [In German] Kehl, G. L., Steinmetz, H., and McGonnagle, W. J., 1957, The removal of inclusions for analysis by an ultrasonic jackhammer: Metallurgia, v. 55, p. 151-154. Kouvo, Olavi, and Vuorelainen, Yrjo, 1959, On valleriite: Geologi, v. 11, p. 32-33. Kouvo, Olavi, Vuorelainen, Yrjo, and Long, J. V. P., 1963, A tetragonal iron sulfide: Am. Mineralogist, v. 48, p. 511-524. Meyer, F. H., Riggs, O. L., McGlasson, R. L., and Sudbury, J. D., 1958, Corrosion products of mild steel in hydrogen sulfide environment: Corrosion, v. 14, p. 69-75. Milton, Charles, and Milton, D. J., 1958, Nickel-gold ore of the Mackinaw mine, Snohomish County, Washington: Econ. Geology, v. 53, p. 426-447. Odman, O. H., 1933, Ore-microscopic study of the sulfide ores of Kaveltorp: Geol. Forenings Forhandlingar, v. 55, p. 563-611. [In German] Petr6n, J., 1898, On so-called valleriite: Geol. Forenings Forhandlingar, v. 187, p. 183-192. [In Swedish] Ramdohr, P., 1960, Die Erzmineralien: Berlin, Akademie Verlag, p. 527-531. Ramdohr, P., and Odman, O. H., 1932, Valleriite: Geol. Forenings Forhandlingar, v. 54, p. 89-98. [In German] Schneiderhohn, Hans, 1929, The mineragraphy, spectrography and genesis of the platinum-bearing nickel-pyrrhotite ores of the Bushveld igneous complex; chap. 17 in Wagner, P. A., Platinum deposits and mines of South Africa: London, Oliver and Boyd, p. 206-246. Schneiderhohn, Hans, and Ramdohr, P., 1931, Lehrbuch der Erzmikroskopie: Berlin, Verlag von Gebriider Borntraeger, v. 2, p. 127-130. Skinner, B. J., and Milton, D. J., 1955, The Elizabeth copper mine, Vermont—a discussion: Econ. Geology, v. 50, p. 751-752.Article 134 EDDIES AS INDICATORS OF LOCAL FLOW DIRECTION IN RHYOLITE By DAVID CUMMINGS, Denver, Colo. Work done in cooperation with the U.S. Atomic Energy Commission Abstract.—Eddies form on the downflow side of inclusions in rhyolite and indicate a single local direction of flow. The more commonly made observations of flow folds and lineations of phenocrysts in rhyolite indicate two possible directions of flow. Eddies or eddy zones develop on the downflow side of inclusions in rhyolite flows. Because a phenocryst or lithic inclusion has a greater inertia than the flowing lava, the inclusion moves more slowly than the surrounding fluid. Friction and drag result. The velocity of the lava is less on the downstream side of the inclusion than on the upstream side because part of the kinetic energy of the fluid particles is dissipated in overcoming resistance caused by shear stresses as the flow passes around the inclusion. As a result, the fluid particles are unable to flow completely around the inclusion. At some point beyond the center of the inclusion, fluid particles become momentarily static, accumulate, and are given a rotary motion by the surrounding flow (fig. 134.1). The pattern of the flow lines differs with differently shaped inclusions. If the long axis of a rectangular inclusion is not oriented parallel to the flow direction, the flow lines are offset slightly (fig. 134.2). The hypothetical outcrop shown on figure 134.3 illustrates the determination of flow direction from eddy Figure 134.1—Development of eddies on the downflow side of an inclusion. Figure 134.2—An inclusion 'and eddy zone in a rhyolite flow. Top, eddy zone and inclusion as seen in outcrop; bottom, diagrammatic view of outcrop. Eddy zone to the left of the lithic inclusion indicates that the flow moved to the left. ART. 134 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D70-D72. 1964. D70CUMMINGS D71 zones. As can be seen, the outcrop face need not be parallel to the direction of flow but should be no more than 45° from that direction. Eddy zones are most clearly seen in the devitrified parts of a lava flow but can also be seen in vitrophyres. The generalized geologic map (fig. 134.4) shows the inferred direction of movement of a rhyolite flow as determined by the direction of eddy zones. Flow foliation and lineation are also shown on the map. The eddy-zone directions are all northerly, ranging from about N. 45° W. to N. 30° E. The strike of the flow foliations is most commonly east-west, and the trend of the lineations is generally north-south. The foliations and lineations suggest either a north or a south direction of flow. The eddy zones, however, indicate that the flow direction is northerly. Figure 134.3—Determination of flow direction by use of eddy zones. Dip of flow layer, 40°; plunge of lineation, 20°. Lineation shows that direction of flow is either northeast or southwest. Face A is almost normal to direction of flow and shows no eddy zone. Faces B and C show east and north components of flow direction. Direction of flow is northeast.D72 MINERALOGY AND PETROLOGY EXPLANATION Contact Dashed where approximately located D U Fault Approximately located. Dotted where concealed; U, up-thrown side; D, downthrown side 20 Strike and dip of flow foliation Strike of vertical flow foliation + Horizontal flow foliation —• 30 Bearing and plunge of lineation Horizontal lineation Planar and linear symbols may be combined Flow direction of rhyolite, based on eddy zone Figure 134.4.—Flow direction of rhyolite, based on eddy zones, Scrugham Peak quadrangle, Nye County, Nev. Alluvium (Qal) is of Quaternary age; rhyolite (Tr), older volcanic rocks (Tvo), and younger volcanic rocks (Tvy) are of Tertiary age. Geology by David Cummings and F. M. Byers, Jr., 1962-63.Article 135 CADMIUM IN SAMPLES OF THE PIERRE SHALE AND SOME EQUIVALENT STRATIGRAPHIC UNITS, GREAT PLAINS REGION By HARRY A. TOURTELOT, CLAUDE HUFFMAN, JR., and LEWIS F. RADER, Denver, Colo. Abstract.—The cadmium content of 84 samples (10 of bentonite, 66 shale and claystone, and 8 marlstone) averages 1.4 ppm and ranges from <0.3 to 11.0 ppm. The median value of 0.8 ppm probably represents the cadmium content of the average shale. Some of the samples with high cadmium content contain relatively large amounts of organic matter which may serve as a collector for cadmium. Other high-cadmium samples, however, contain as little organic matter as most of the low-cadmium samples, suggesting that there may have been variation in the amount of available cadmium in the sea water in which these rocks were deposited. Data on the cadmium content of sedimentary rocks are very sparse. Preuss’ (1941, p. 19) spectrographic analysis showing 0.3 parts per million of cadmium in a composite sample of 36 shales from Germany is the basis for most estimates of the abundance of cadmium (Rankama and Sahama, 1950, p. 713; Turekian and Wedepohl, 1961, p. 184 and table 2). Mullin and Riley (1956) report cadmium contents ranging from 0.39 to 0.45 ppm for marine oozes. Cadmium is concentrated in zinc sulfides (Fleischer, 1955, p. 992) and has long been known to be concentrated in coal ashes and black marine carbonaceous shales (Krauskopf, 1955, p. 417). The cadmium content is reported here for 10 samples of bentonite, 8 of marlstone, and 66 of shale and claystone from the Pierre Shale, a thick and widespread marine unit of Late Cretaceous age in South Dakota and adjacent parts of Nebraska, Wyoming, and Montana. The samples come mostly from the Black Hills and Missouri River areas in these States. The samples were analyzed by Huffman using a method by which cadmium is separated and concentrated by ion exchange before its spectrophotometric determination with dithizone (Huffman, 1962). Under the conditions of the determinations, 1 division on the transmission scale of the spectrophotometer is equal to 0.12 ppm cadmium. The lower limit of detection is taken as 0.3 ppm, and the standard deviation for concentrations near 1 ppm is also about 0.3 ppm cadmium. Cadmium content was determined for two groups of samples. The first group consists of samples described by Tourtlelot (1962) in part of a reconnaissance investigation of the composition of the Pierre Shale. Table 135.1 shows the cadmium analyses for most of these samples for which other analytical data have been published. This group includes two sets of closely spaced samples representing weathering profiles (Tourtelot, 1962, p. 28-32); only the freshest samples are included in the statistical discussions that follow. The second group of samples analyzed for cadmium is from the Pierre Shale of South Dakota and parts of adjacent States and was selected for more detailed investigation than the first group. Stratigraphic, mineralogic, chemical, and spectrographic data for this group have been summarized by Tourtelot and others (1960). The cadmium content of both groups of samples is shown graphically in figure 135.1 and is included in the statistical treatment of all samples in which cadmium has been determined. The analytical methods, their precision, and the names of analysts determining the elements reported in this and other shale studies have been described by Rader and Grimaldi (1961) and Barnett (1961). The method for determining total carbon has been changed since some of the early analyses were made. Since the present method is more precise and has a lower limit for the total carbon determination of 0.05 percent C, all of the samples containing too little carbon for determination previously have now been analyzed, and certain other samples reanalyzed to make all the data comparable. The carbon data in table 135.1 should ART. 135 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D73-D78. 1964. D73D74 GEOCHEMISTRY Table 135.1.—Analyses for cadmium and forms of carbon [Cadmium analyses by Claude Huffman, Jr., carbon analyses by I. C. Frost. See Tourtelot (1962) for rock analyses and spectrographic and chemical analyses for minor constituents in these same samples] Location Description Cad- Carbon (percent) Sample Section Town- ship Range County State Forma- tion Member Material mium (ppm) Total carbon Car- bonate carbon Organic carbon C871 > NWM 31 18 N. 19 E. Fergus. . Montana.. Claggett.. 0. 5 1.3 1.4 .8 1.0 1.6 1.5 1.3 <0.03 <.03 <.03 <.03 1.3 <.03 . 5 1.2 1.4 C872 i 2__. NWM 31 18 N. 19 E C873 12 NW!4 31 18 N. 19 E. ...do .8 C874 i 2.... NWH 31 18 N. 19 E. Residual soillike weathering product... 1.0 C870 20 6 S. 23 E. C876 NEM 32 54 N. 67 W. Wyoming. South 1.5 .8 1.1 1.0 C886 NWH 18 18 N. 30 E. Corson Shale. ...do Upper part of Shale, dark-gray, hard, flaky .8 C887 SEVa 29 16 N. 31 E. Dakota Virgin Creek . 8 1.1 <.03 .05 <.03 .07 C881 SEH 36 6 N. 30 E. Stanley Claystone, dark-gray with brownish cast. 1 0 1.0 .9 C882 2 SE M 36 6 N. 30 E. .6 C883 2 SEM 36 6 N. 30 E. Residual soillike weathering product.. Claystone, dark-gray with brownish cast. Claystone, dark-gray to black; some silty layers. Claystone, black: some silt layers (core sample from 32.8 ft). Marlstone, yellowish-gray, slightly sandy (core sample from 51.0 ft). Claystone, black, bentonitic (core sample from 66.8 ft). .8 1. 5 1.4 1.5 1. 5 C884 SE]4 36 6 N. 30 E. .5 2.0 C885 SEH36 6 N. 30 E. ...do .8 2.2 .7 C878 1 111N. 80 W. .8 1.1 <.05 6.6 1 1 C879 1 Ill N. 80 W. .8 7.5 .9 C880 1 Ill N. 80 W. .8 1.1 .1 1.0 C888 17 104 N. 71 W. 1.4 .7 <.03 4.7 .7 C889 17 104 N. 71 W. 1.2 5.3 . 6 0890 17 104 N. 71 W. 1.5 1.1 .5 .6 C891 17 104 N. 71 W. Sharon Springs.. Shale, black, hard; rich in organic matter. 3.5 10.3 .03 10.3 1 Samples represent a weathering profile from the center of a roadcut (C871) into the coluvium and soil zone on the flank of the hill (C874). The samples were taken about a foot above a tuffaceous siltstone bed, 1 to 2 inches thick, that could be traced into the colluvium. 2 Not included in statistical discussion of cadmium. □ + o + + A A □ A AJ O ++« + , 0.3 0.4 0.5 1.0 2.0 3.0 4.0 5.0 10.0 CADMIUM, IN PARTS PER MILLION Figure 135.1.—Distribution of cadmium in 84 samples of Pierre Shale. Triangles, bentonite; squares, marlstone; circles, shale and claystone containing <0.5 percent organic carbon; crosses, shale and claystone containing 0.5 to 1.0 percent organic carbon; dots, shale and claystone containing >1.0 percent organic carbon. Such samples containing less than 0.3 ppm cadmium are plotted arbitrarily at 0.15 ppm, half of the lower limit of detection. be substituted for those shown previously by Tourtelot (1962, p. 57). Currently, total carbon is determined by burning a weighed sample at 1350°C in a stream of oxygen. Iron chips and tin metal act as burning accelerators and catalysts, the evolved gases being passed through manganese dioxide and hot copper oxide to remove sulfur and to convert carbon monoxide to carbon dioxide. The volume of the gas, collected in a closed system under standard conditions of temperature and pressure, is measured; then the gas is scrubbed with potassium hydroxide solution to remove the carbon dioxide, and the volume is redetermined. Total carbon is calculated, carbon dioxide from carbonates deter- mined separately is subtracted, and the remainder reported as organic carbon. The 10 samples of bentonite studied represent beds throughout the section of the Pierre Shale; 4 of these samples are of beds of practically the same age. The 8 marlstone samples include 3 from the Crow Creek Member in the lower-middle part of the Pierre Shale, and 5 from the Mobridge Member in the upper part of the Pierre Shale in the Missouri River area of South Dakota. The 66 samples of shale and claystone represent rocks throughout the Pierre Shale and equivalent stratigraphic units; these samples are grouped for convenience according to organic-carbon content: <0.5 percent (19 samples), 0.5 to 1.0 percent (26 samples), and >1.0 percent organic carbon (21 samples). Of the samples containing >1.0 percent organic carbon, 14 come from the Sharon Springs and Mitten Members in the lower part of the Pierre Shale and from the Claggett Shale, which make up a relatively thin unit of generally organic-rich shale of the same age from South Dakota to central Montana. CADMIUM CONTENT The statistical distribution of cadmium according to rock type is shown in figure 135.1 and, with other data, is summarized in table 135.2. The cadmium content of all samples ranges from <0.3 ppm to 11.0 ppm and averages 1.4 ppm. Samples reported with <0.3 ppm of cadmium, the lower limit of detection, are arbitrarilyTOURTELOT, HUFFMAN, AND RADER D75 assumed to contain 0.15 ppm. The graphic distribution (fig. 135.1) shows, however, that the average is too much influenced by the few high values to represent the expectable cadmium content of clayey rocks. The modal value of 0.8 ppm cadmium, which is the median value (table 135.2), seems to be a more reasonable figure. Table 135.2.—Cadmium, zinc, total sulfur as sulfur, organic carbon, and vanadium in groups of samples [Values reported as less than respective limits of detection are as counted half the limit of detection in calculating averages. Number of samples shown in parentheses] All Benton- Marl- Shale and claystone, grouped by organic carbon content samples ite stone <0.5 percent 0.5-1.0 precent >1.0 percent (84) (10) (8) (19) (26) (21) Cadmium (ppm): <0.3-11.0 0. 5-11.0 0. 4-10.0 <0.3-0. 8 <0.3-1. 8 0. 5-8. 4 1.4 2. 8 2.6 .35 .8 2.0 .8 .8-. 9 .9 .3 . 7-8 1.2 Zinc (ppm): 38-380 38-330 60-210 65-190 90-360 37-380 138 110 130 128 163 140 62-65 150 130 165 130 Total sulfur as sulfur (percent): Range <. 1-13. 5 .8 <. 1-3. 2 .9 <. 1-2.1 .99 <.l-.4 .2 <. 1-.9 .4 <. 1-13. 5 2.1 .4 .9 . 1 .4 1.2 Organic carbon (percent): <. 2-10.3 <•2 <. 2-2. 9 .2-. 4 .5-1.0 1.1-10.3 1.4 1.5 .3 .8 3.5 .9 .3 . 7-8 3.1 Vanadium (ppm): 50-740 20-580 4 70-380 2 50-300 3 310 3130-740 214 100 (4 42) 40 200 170 200 390 210 170 210 390 1 Excluding 2 samples for which only semiquantitative spectrographic analyses are available. 2 Excluding 1 sample for which only semiquantitative spectrograph ic analyses are available. 2 Excluding 5 samples for which only semiquantitative spectrographic analyses are available. 4 Average, excluding high value of 580 ppm. The average composition of 17 samples of Pierre Shale discussed by Tourtelot (1962, p. 37-44, 67) is very similar to that of several other average shales, mostly of pre-Cretaceous age (Clarke, 1915, p. 23; Vinogradov and Ronov, 1956, p. 6, 7; and others). The differences between these and the published averages are chiefly in alumina, silica, and potassium. The differences in silica and alumina reflect the variable admixtures of quartz and clay typical of such rocks; the lower potassium content of the Pierre Shale is attributable to the greater abundance in it of montmorillonite relative to potassium-rich clays. Data from the Pierre Shale thus are applicable to problems of average composition. These 17 samples provided 15 of the samples analyzed for cadmium (table 135.1). The average cadmium content of these 15 samples is 1.1 ppm and the median value is 0.8 ppm. Many uncertainties surround selection of an average value for the content of any element in shale. Clearly, there is much rock in the Pierre Shale that contains <0.3 ppm cadmium, and the analysis of additional samples of Pierre Shale and of other shales, particularly those with low organic content, no doubt would lower the modal value. It is probably unlikely that any rocks in the Pierre contain more than about 10 ppm cadmium, although other black shales for which analytical data were published by Davidson and Lakin (1961, 1962) contain as much as 300 ppm cadmium, as determined by semiquantitative spectrographic analysis (D. F. Davidson, oral communication, March 1963). Cissarz (1930, p. 64) also reports spectrographic analyses of 100 to 1,000 ppm cadmium in parts of the Mans-feld Kupferschiefer in Germany. Such high values, even though they are typical of a special kind of shale, must be taken into account as well as the low values. No rational basis for doing so, however, is apparent, and the median and modal value of 0.8 ppm cadmium derived from the Pierre samples seems to be an acceptable and usable average content for shale until much more work has been done. RELATION TO OTHER CONSTITUENTS All statistical measures of the cadmium content of the samples increase as the organic carbon content increases, but the relation is a general one (fig. 135.1 and table 135.2). Most of the bentonite and marlstone samples contain 1.0 ppm cadmium or less, but 2 of each contain >5.0 ppm cadmium. The organic-carbon content of bentonite is negligible, being less than the arbitrary cutoff of 0.2 percent for the peroxide-bomb analytical method (Rader and Grimaldi, 1961, p. 37). Hence, cadmium in the bentonite samples is independent of organic-carbon content. The 2 samples of marlstone that contain the most cadmium each contain 2.0 percent or more organic carbon, but 1 other sample of marlstone that contains >2.0 percent organic carbon contains only 0.7 ppm cadmium. Thus, cadmium in marlstone may be related to organic carbon, but the relation is not evidently a genetic one. The organic carbon in the Pierre samples is similar in composition to lignite or subbituminous coal and is inferred to be derived largely from land plants (Breger and others, 1960; see also Breger and Brown, 1962). Cadmium and other metals in the Pierre organic-rich samples may be original constituents of the plants, derived from their growing sites, or the metals may have been adsorbed by the plant material from the sea water during the transport, accumulation, and diagenesis of the plant material in the Pierre rocks. Each of these explanations has been advanced for metals in coal and marine carbonaceous rocks, and it seems likely that both could be operative. The data on the Pierre samples are not helpful in identifying 716-626 0—64 6D76 GEOCHEMISTRY which mode of origin may have played the largest role in controlling the accumulation of cadmium. The zinc/cadmium ratio plotted against the zinc content (fig. 135.2) has been used to divide the samples into 500 400 300 200 Z O -i _i 5 UJ 0. (/> H K 2 50 - 40 o' z (M . 100 10 1 1 1— 8.4 11.0A H0A -7.0 10.0° °6.5 '3.5 • 1 1 1 + + + • + + o+ • + + + + '’+* +* a+ 0 0 +< * 0* a 0 0 —1—1—1—1— >+ - a* □ 0 _ ■3.5 - □ ▲ - .2.1 A ■ 4- 1.6 Crustal - abundance 1 1 1 1 1 1 1 1 1 1 1 1 i-1 10 20 30 40 50 100 200 300 500 1000 ZINC/CADMIUM RATIO Figure 135.2.—Relation between the zinc/cadmium ratio and zinc content of the Pierre Shale. Numbers on symbols show cadmium content, in parts per million. Triangles, bentonite; squares, marlstone; circles, shale and claystone containing <0.5 percent organic carbon; crosses, shale and claystone containing 0.5 to 1.0 percent organic carbon; dots, shale and claystone containing >1.0 percent organic carbon. two groups—one with ratios of <50 and the other with ratios of >50. The zinc/cadmium ratio of the earth’s crust is 270, using 0.15 ppm for cadmium (Brooks and Ahrens, 1961, p. 112) and 40 ppm for zinc (Wedepohl, 1953, p. 94). Both zinc and cadmium are concentrated in most samples of Pierre Shale considerably above their crustal abundance. The group with ratios less than 50 includes the 9 samples of bentonite, marlstone, and organic-rich shale with a cadmium content of 2.0 ppm or more. The remaining sample contains 1.6 ppm cadmium. The ratio of zinc to cadmium is relatively constant, and the cadmium content is related directly to the zinc content. Results of semiquantitative spectrographic analyses show that zinc/cadmium ratios of 10-30 are common for black-shale samples which contain as much as 1,000 ppm zinc (D. F. Davidson, oral communication, March 1963). In the group with ratios of >50, the cadmium content ranges from <0.3 ppm to only 1.8 ppm, and the zinc/ cadmium ratio increases in direct proportion to the zinc content. A subgroup of nine samples lies somewhat above the main trend but parallel to it. The cadmium content of these 9 samples ranges from 1.5 to 1.8 ppm, whereas the main trend includes only a few samples that contain >1.0 ppm cadmium. These relations suggest that where the cadmium content is more than about 2 ppm, it is directly proportional to the zinc content at a relatively uniform ratio. Where the cadmium content is less than about 2 ppm, it does not vary systematically with zinc content and tends to be rather uniform in amount while the zinc content fluctuates. These relations may mean that (1) the proportions of cadmium and zinc in sea water varied at the deposition al sites, either because of different amounts delivered to the sea water from source areas or because of the chemical composition of the sea water itself, or that (2) the conditions under which the sediments accumulated varied in controlling the fixation of both cadmium and zinc. For example, the stronger chalcophile tendencies of cadmium at low and moderate temperatures in acid solutions, compared with those of zinc (Goldschmidt, 1954, p. 269), may account for the enrichment of cadmium in the organic-rich shales, which are also rich in syngenetic sulfide minerals. These tendencies may even account for enrichment of cadmium in marlstone. Samples that contain sulfide minerals, but only relatively small amounts of cadmium, may have accumulated where there was little cadmium in the environment, or the sulfide minerals may be late diagenetic to secondary in origin and thus have had no chance to reflect the cadmium content of the environment as a whole. The high cadmium content and low zinc/cadmium ratio of both bentonite and black shale in a single widespread stratigraphic unit suggest that these compositional features have a common origin. The generally low zinc/cadmium ratios of the bentonites studied here may be interpreted as reflecting the original relations of zinc and cadmium in the volcanic material that was altered to form the bentonite. The bentonites that accumulated on the bottom of the Pierre sea hundreds of miles east of the source of the volcanic material in Montana (Tourtelot, 1962, p. 10-11) must represent only a fraction of the total amount erupted. Bentonitic material is also abundant in the shales, as indicated by high montmorillonite content. Much of the total eruption product accumulated on land, but much also must have been moved aerially into the sea, along with whatever volatile substances accompanied the eruptions. The metal content of the sea water coidd be increased by solution of the volatile substances and the elements in the disseminated volcanic material. The elements then would be adsorbed by organic matter and by clay minerals that themselves were partly the result of alteration of the volcanic material. The volcanic material in discrete beds thatTOURTELOT, HUFFMAN, AND RADER D77 altered to bentonite would either retain its minor-element content, or lose it to sea water, depending on the nature of the alteration process. Cadmium and zinc thus were available in the sea water at a time when organic matter was abundant enough to participate directly in adsorption reactions or to make possible the formation of syngenetic sulfides by generally reducing the oxidation potential of the bottom sediments. The volcanic material did not necessarily contain more cadmium and zinc than volcanic material delivered at other times to the Pierre sea. Zinc, total sulfur as sulfur, organic carbon, and vanadium (table 135.2) have much the same distribution pattern among the rock types that cadmium has, and statistical correlations between many of them are positive and significant. This is not surprising, of course, because most past work has shown the co-enrichment of these elements in similar rocks. The mode of occurrence of these elements in such rocks still is largely unknown, so that the explicit factors that operated to produce at one place a sulfide-rich organic-rich rock with a relatively high content of cadmium and other elements, and at another place a similar rock with little cadmium, also are unknown. Variations from place to place of temperature, pressure, pH, and oxygen potential obviously would affect, directly or indirectly, the amount of cadmium that would be incorporated in clayey rocks. Such factors were different for a site where marlstone was deposited, compared to one where a sulfide-rich shale was deposited, yet the cadmium contents of some samples of such rocks are similar. At the same time, the cadmium contents of organic-rich shales differ considerably, and yet the enumerated factors, excepting oxygen potential, must have been of nearly the same order of magnitude during the deposition of the sediments represented by the samples. Consequently, other variable factors also must have been operating. RELATION TO STRATIGRAPHIC AND GEOGRAPHIC POSITION Eight of the 10 samples with the highest cadmium content and a zinc/cadmium ratio of <50 include 2 samples of bentonite and 6 of organic-rich shale from the Sharon Springs Member in the lower part of the Pierre Shale. These samples are from widely scattered localities in the lower part of the Missouri River valley in South Dakota and the southern part of the Black Hills in South Dakota and Wyoming. One sample comes from the Irish Creek well (Tourtelot and Schultz, 1961), almost halfway between the Black Hills and the Missouri River. The high cadmium content of the two bentonite samples may be related to the strati- graphic association of high-cadmium organic-rich shale and bentonite in the Sharon Springs Member. No regularity in relative position of samples with high cadmium content in the shale and in the bentonite can yet be seen—the high-cadmium shale may be either above or below the high-cadmium bentonite. Two additional samples of bentonite from the Sharon Springs have a cadmium content of 0.6 and 0.8 ppm and a zinc/cadmium ratio of only 70. Relatively low zinc/cadmium ratios seem characteristic of bentonites from other units, too, all having a ratio of less than about 100. The two marlstone samples with a high cadmium content and low zinc/cadmium ratio are from the Mobridge Member in the upper part of the Pierre Shale. They suggest the possibility that there were local areas in the Pierre sea in which the content of cadmium and other metals was higher than elsewhere during the same span of time. The samples are from localities about 35 miles apart, in southeastern Gregory County, S. Dak., and Knox County, Nebr. These two samples also contain relatively large amounts of molybdenum, selenium, and arsenic that, like cadmium, cannot be explained on the basis of what is known of the geochemistry of these elements in carbonate-rich rocks. Three other marlstone samples from the Mobridge Member at localities 50 miles or more northwest of Gregory County contain less than 1 ppm cadmium, and 6 samples of the same age but from a noncarbonate facies also have a low cadmium content. The mineral-ogical composition of these and other samples from the upper part of the Pierre Shale in this part of the Missouri River valley in southern South Dakota and adjacent parts of Nebraska differs from samples of rocks of the same age at other localities in a way that suggests that some of the materials were derived from land areas bordering the Pierre sea on the east. It is possible that the high cadmium content of the two marlstone samples reflects unusual amounts of cadmium and other elements in the sea water in this area during the latter part of Pierre time. The metals may have come from the eastern shore area, although no discrete sources for them are known. The metals could also have become locally concentrated in the water of the Pierre sea by physical and chemical processes of oceanic circulation after having been derived from sources west of the Pierre sea. Although they are unsatisfying, such speculations are about all that the present data permit. Bur’yanova (1960, p. 217) also postulates unknown metal-rich source areas for the cadmium minerals and camouflaged cadmium in Devonian sandstones in Tuva, south-central Siberia.D78 GEOCHEMISTRY SUMMARY The cadmium content of 84 samples (10 of bentonite, 66 shale, and 8 marlstone) averages 1.4 ppm and ranges from <0.3 to 11.0 ppm. The median value of 0.8 ppm probably represents the cadmium content of the average shale. Cadmium tends to be concentrated in the same pattern as sulfur and vanadium in rocks of the Pierre Shale rich in organic carbon. Much of the cadmium in such rocks may be in syngenetic sulfide minerals. Bentonite that contains no organic carbon itself but which is associated with organic-rich shale also may contain relatively large amounts of cadmium. High-cadmium samples of bentonite and organic-rich shale are from the Sharon Springs Member at widespread localities. The close geographic association of two high-cadmium samples from the Mobridge Member in southern South Dakota and adjacent Nebraska may indicate that locally the sea in this area was enriched in cadmium and other metals during the deposition of the Mobridge Member. REFERENCES Barnett, P. R., 1961, Spectrographic analysis for selected minor elements in Pierre shale: U.S. Geol. Survey Prof. Paper 391-B, p. B1-B10, 13 figs. Breger, I. A., and Brown, Andrew, 1962, Kerogen in the Chattanooga Shale: Science, v. 137, no. 3525, p. 221-224. Breger, I. A., Tourtelot, H. A., and Chandler, J. C., 1960, Geochemistry of kerogen from the Sharon Springs Member of the Pierre Shale: Geol. Soc. America Bull., v. 71, no. 12, pt. 2, p. 1832-1833. Brooks, R. R., and Ahrens, L. H., 1961, Some observations on the distribution of thallium, cadmium and bismuth in silicate rocks and the significance of covalency on their degree of association with other elements: Geochim. et Cosmochim. Acta, v. 23, p. 100-115. Bur’yanova, E. Z., 1960, Mineralogy and geochemistry of cadmium in the sedimentary rocks of Tuva: Geokhimija, no. 2, p. 177-182 [in Russian]: translation in Geochemistry, 1960, no. 2, p. 209-217. Cissarz, Arnold, 1930, Quantitative-spektralanalytische Untersuchung eines Mansfelder Kupferschieferprofiles: Chemie d. Erde, v. 5, (Linck-Festschr.) p. 48-75. Clarke, F. W., 1915, Analyses of rocks and minerals from the laboratory of the United States Geological Survey, 1880 to 1914: U.S. Geol. Survey Bull. 591, 376 p. Davidson, D. F., and Lakin, H. W., 1961, Metal content of some black shales of the Western United States: Art. 267 in U.S. Geol. Survey Prof. Paper 424-C, p. C329-C331. ------ 1962, Metal content of some black shales of the western conterminous United States, pt. 2: Art. 85 in U.S. Geol. Survey Prof. Paper 450-C, p. C74. Fleischer, Michael, 1955, Minor elements in some sulfide minerals, in pt. 2 of Bateman, A. M., ed., Economic Geology, 50th anniversary volume, 1905-55: Urbana, 111., Econ. Geology Pub. Co., p. 970-1024 Goldschmidt, V. M., 1954, Geochemistry: London, Oxford Univ. Press, 730 p. Huffman, Claude, Jr., 1962, Ion-exchange separation and spectro-photometric determination of cadmium: Art. 214 in U.S. Geol. Survey Prof. Paper 450-E, p. E126-E127 [1963]. Krauskopf, K. B., 1955, Sedimentary deposits of rare metals, in pt. 1 of Bateman, A. M., ed., Economic Geology, 50th anniversary volume, 1905-1955: Urbana, 111., Econ. Geology Pub. Co., p. 411-463. Mullin, J. B., and Riley, J. F., 1956, The occurrence of cadmium in seawater and in marine organisms and sediments: Jour. Marine Research, v. 15, no. 2, p. 103-122. Preuss, E., 1941, Beitrage zur spektralanlytichen Methodik. II. Bestimmung von Zn, Cd, Hg, In, Tl, Ge, Sn, Pb, Sb, Bi durch fraktionierte Destination: Zeitschr. angew. Min-eralogie, v. 3, p. 8-20. Rader, L. F., and Grimaldi, F. S., 1961, Chemical analyses for selected minor elements in Pierre Shale: U.S. Geol. Survey Prof. Paper 391-A, p. A1-A45, 19 figs. Rankama, Kalervo, and Sahama, Th. G., 1950, Geochemistry: Chicago, Chicago Univ. Press, 912 p. Tourtelot, H. A., 1962, Preliminary investigation of the geologic setting and chemical composition of the Pierre Shale, Great Plains region: U.S. Geol. Survey Prof. Paper 390, 74 p. Tourtelot, H. A., and Schultz, L. G., 1961, Core from the Irish Creek well, Ziebach County, South Dakota: U.S. Geol. Survey open-file report, 20 p. Tourtelot, H. A., Schultz, L. G., and Gill, J. R., 1960, Stratigraphic variations in mineralogy and chemical composition of the Pierre Shale in South Dakota and adjacent parts of North Dakota, Nebraska, Wyoming, and Montana: Art. 205 in U.S. Geol. Survey Prof. Paper 400-B, p. B447-B452. Turekian, K. K., and Wedepohl, K. H., 1961, Distribution of the elements in some major units of the Earth’s crust: Geol. Soc. America Bull., v. 72, p. 175-192. Vinogradov, A. P., and Ronov, A. B., 1956, Composition of sedimentary rocks of the Russian platform in relation to the history of their tectonic movements: Geokhimiya 1956, no. 6, p. 3-24 [In Russian; translation in Geochemistry, 1956, v. 6, p. 533-559]. Wedepohl, K. H., 1953, Untersuchungen zur Geochemie des Zinks: Geochim. et Cosmochim. Acta., v. 3, p. 93-142.Article 136 ANALYSIS OF GEOCHEMICAL PROSPECTING DATA FROM THE ROCKY RANGE, BEAVER COUNTY, UTAH By J- J- CONNOR and A. T. MIESCH, Denver, Colo. Abstract.—A spatial relation has been found between copper in alluvium and a copper ore body 20 to 225 feet beneath the surface. The relation became clear through statistical procedures used to reduce the effects of local variation and area-wide trends. Methods of trend analysis used in interpretations of geophysical data (Grant, 1957, p. 309; Oldham and Sutherland, 1955, p. 295) and applied to geologic problems by Krumbein (1956, p. 2163; 1959, p. 823), Whitten (1959, p. 835), and others have not been adequately tested for their application to data-interpretation problems in geochemical prospecting. This article presents the results of such an application. The geochemical data used in this study were collected by Erickson and Marranzino (1960, p. B98) in the Rocky Range of southwestern Utah and consist of 110 chemical analyses for trace amounts of copper in samples of pediment alluvium. The samples were col- lected on approximately 100-foot centers and were taken at a depth of 8 to 12 inches below the surface. The sampled area overlies a small copper ore body which lies at a depth of 20 to 225 feet below the surface. The alluvium over the ore body ranges from 6% to 45 feet in thickness. Trend analysis, and analysis of variance tests of the trends, show statistically significant lateral variations in the data that can be described by low-order polynomial surfaces, but small-scale local variation (probably due to sampling error) obscures subtle features of the data important to geochemical prospecting. These subtle features are brought out by applying a “moving average” operation (which serves to reduce random local variation) in conjunction with the ordinary methods of trend analysis. Figure 136.1 is a map of the sampled area showing deviations of the copper values from their mean and is equivalent to figure 47.1 of Erickson and Marranzion EXPLANATION Copper content More than 25 parts per million □ 25 to -25 parts per million Less than —25 parts per million Mineralized core hole Sampled locality Figure 136.1—Copper content of alluvium in the Rocky Range, Utah (contoured on deviations from the mean). Contour interval 25 ppm; values larger than 100 ppm and smaller than —100 ppm not contoured. Modified from Erickson and Marranzino (1960). ART. 136 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D79-D83. 1964. D79D80 GEOCHEMISTRY (1960, p. B99). The ore body is roughly outlined by the black dots, which show locations of mineralized drill holes. There is no obvious spatial relation of copper values at the surface to ore, a conclusion also reached by Erickson and Marranzino. Cursory examination of figure 136.1, however, reveals that, in general, the copper content tends to be high on the east side of the map and low on the west. Using methods of trend analysis, described fully in papers previously cited, seven polynomial surfaces were fitted by least-square techniques to the data of figure 136.1. Four of these surfaces or trends, of degree 1 through 4, and deviations from the trends, are shown in figures 136.2A-D. The deviations identify variance in the data that cannot be described by low-order polynomial equations, and is ascribed to local effects, in part geologic factors that do not act in a continuous fashion across the entire area of investigation. Some part of the variance of the deviations is due to analytical error and possibly other nongeologic factors. Analysis of variance tests of the computed trends are given in columns 1 through 4 of the accompanying table. Deviations of observed data from the trends were not tested for fit to a normal curve or for degree of independence, but the F ratios may be used in a relative sense, at least, to judge the degree of fit for various trends. Significance levels for the F ratios should be viewed cautiously. Analysis of variance of ■polynomial surfaces fitted to the data of figures 136.1 and 136.3 [E0.9S (2,107) =3.09; i?0.«s (3,104) =2.70; E0.95 (4,100) =2.46; Eo-95 (5,95) =2.31; E0.95 (6,89) = 2.20; F0.95 (7,82) =2.12; F0.95 (8,74) =2.06] Source of variation Original data Moving- average data Sum of squares Degrees of freedom Mean square F Sum of squares Linear terms 148,657 2 74,329 / 73,863 Deviations from linear trend 801,110 107 7,487 1137,800 Quadratic terms ... 71,120 3 23, 707 / 53,457 Deviations from quadratic trend 729, 990 104 7,019 l 84,343 Cubic terms 25, 880 4 6, 470 f 14,897 Deviations from cubic trend 704.110 100 7,041 l 69.446 Quartic terms 129, 630 5 25, 924 / 23,239 Deviations from quartic trend 574, 480 95 6,047 l 46,207 Quintic terms 53, 620 6 8, 937 1 10,210 Deviations from quintic trend 520, 860 89 5, 852 \ 35,997 Sextic terms - 41,620 7 5, 946 11 no / 3,345 Deviations from sextic trend. _ 479. 240 82 5, 844 1 32,652 Septic terms 48, 770 8 6,096 / 9,553 Deviations from septic trend 430, 470 74 5, 817 l 23,099 The proportions of the total sum of squares accounted for by the linear through quartic trends (see table, first column) amount to 16, 23, 26, and 40 percent, respectively. Though the F ratios suggest that the EXPLANATION Copper content More than 25 parts per million □ 25 to -25 parts per million Less than —25 parts per million • Mineralized core hole fi. QUADRATIC TREND AND DEVIATIONS FROM QUADRATIC TREND D. QUARTIC TREND AND DEVIATIONS FROM QUARTIC TREND Figure 136.2—Trends (dashed line), in parts per million, and trend deviations of copper in alluvium, Rocky Range, Utah. Contour interval 25 ppm; values larger than 100 ppm and smaller than —100 ppm not contoured.CONNOR AND MIESCH D81 linear, quadratic, and quartic trends may be statistically significant, the low-percentage sum of squares associated with the trends partially explains the similarities between the deviation maps (figs. 186.2/1 -D) and the map of the original data (fig. 136.1). Sampling error appears to be one major factor displayed by the deviation maps, and it gives rise to sharp local fluctuations of both the original data and the deviations of values about the trends. Comparison of replicate analyses made on the same sample indicates that analytical variation is comparatively small. Sampling error does not imply mistakes in selection of samples or in the sampling procedure, but refers only to the difference between the value obtained on a sample and the mean value that would have been obtained had all possible samples occurring at the sample site been taken. Such complete sampling, of course, is impossible and all samples contain error; the error is inversely proportional to the number of samples taken at the site and directly proportional to the natural local variation in the alluvium. The map in figure 136.3 shows the moving-average copper content of the alluvium across the area. In construction of this map, each original copper value was replaced by an unweighted average of that value and the four values closest to it on the map in figure 136.1. The effect of the moving average is to smooth out local variation—to reduce the effects of sampling error and other small-scale fluctuations in the data. A trend in the data across the area is now more apparent. Four trends computed on the moving-average data, and deviations of the averages from the trends, are shown in figures 136AA-D. Analysis of variance tests of the trends could not be made because of interdependence of data imposed by the moving-average method. The proportions of the total sums of squares accounted for by the linear through quartic trends, however, range from 35 to 78 percent. Deviations of moving-average values from trends fitted to the averages appear to clearly outline the major part of the copper ore body beneath the alluvial cover. Deviations from the quartic trend may be especially significant. Although “negative anomalies” are commonly not regarded as ore indicators by geochemical prospectors (Hawkes and Webb, 1962, p. 27), the presence of the negative deviations over the ore body in figures 136.3 and 136.4/1-71 necessitates consideration of their importance here. The possibility that the negative deviations over the ore body (figs. 136.3 and 136.4) are merely a result of the adjacent high values that elevate the trend in this region may be excluded for two reasons: (1) the negative deviations correspond closely to the ore body, except over the southern part, and (2) they occur on only one side of the high, the side where the ore body is located. It is assumed, therefore, that the copper low over the ore body is real and originated from natural processes. The copper high southwest of the ore body, though pronounced (figs. 136.3 and 136.4), is caused in very large part by one extreme value (600 ppm) in this area. Recomputation of the trends and trend deviations, excluding the extreme value, results in severe depression of the high, although the low deviations over the ore persist. Low copper contents in alluvium over the ore body might have resulted from circulation of sulfate solutions and consequent leaching of copper in this area; cores taken from the ore body contain tactite, abundant sulfide minerals, and secondary copper oxides. The positive deviations adjacent to the ore body on its southwest side (figs. 136.3 and 136.4/1-71) may have resulted from precipitation of leached copper downdip L EXPLANATION Copper content U111I1IU Mineralized More than 25 core hole parts per million Sampled 1------' locality 25 to -25 parts per million Less than —25 parts per million Figure 136.3.—Moving-average copper content of alluvium in soils, Rocky Range, Utah (contoured on deviations from the mean). Contour interval 25 ppm; values larger than 100 ppm and smaller than —100 ppm not contoured.D82 GEOCHEMISTRY A. LINEAR TREND AND DEVIATIONS FROM LINEAR TREND EXPLANATION Copper content More than 25 parts per million □ 25 to -25 parts per million Less than —25 parts per million • Mineralized core hole fi. QUADRATIC TREND AND DEVIATIONS FROM QUADRATIC TREND 0. QUARTIC TREND AND DEVIATIONS FROM QUARTIC TREND Figure 136.4.—Trends (dashed lines), in parts per million, and trend deviations of moving-average copper content in alluvium, Rocky Range, Utah. Contour interval 25 ppm; values larger than 100 ppm and smaller than —100 ppm not contoured. from its source, or from mechanical transportation of copper downhill during formation of the alluvial cover. This geologic interpretation of the trend-analysis results is only one of a number that could be advanced, but whether it is correct or not has no bearing on the value, or the potential value of these results or results from future work. There can be little doubt that deviations from progressively higher order trends, 1st through 4th degree, show an increasingly better spatial correspondence to the known ore body in the area, or that the moving-average trend deviations in general show a clearer correspondence to the ore body than do the original data. Polynomial terms of 5th, 6th, and 7th degree account for nonsignificant portions of the variance of the original data and, similarly, small portions of the total sum of squares in the moving averages (see table). Consequently, the quartic trend is regarded as the probable appropriate model to describe the trend (Ostle, 1954, p. 141-142). The question of selecting a model to describe the trend in a particular set of data (that is, how to decide whether the trend should be described by a 1st, 2d, or higher order equation in a particular problem) has not been completely resolved by previous workers, and a great deal of experimentation needs to be done. Trend analysis of geologic data is not yet developed to the stage where it can be applied with preset objective rules. Statistical tests for trend significance, employing the F ratio, are not strictly valid if local concentrations of positive or negative deviations are present. Such concentrations imply the absence of independence of the deviations, a fundamental assumption prerequisite to the test. Thus, the very property of the data being sought is a hindrance to an important part of the statistical analysis. The use of moving averages further reduces the independence of the deviations, to a point where F tests should not be presented as even approximately correct tests of significance. On the other hand, if the statistical significance determined for trends in the original data is accepted, there may be little need to test the almost identical trends in the moving-average data. The computation of trends in the moving-average data may be circumvented by applying the moving-average method to the deviations from trends in the original data. Contour maps constructed in this manner are given in figures 136.5A-D. In this partic-CONNOR AND MIESCH D83 A. MOVING AVERAGES OF DEVIATIONS FROM LINEAR TREND SHOWN IN FIGURE 136.2 A EXPLANATION Copper content More than 25 parts per million 25 to -25 parts per million Less than —25 parts per million • Mineralized core hole 8. MOVING AVERAGES OF DEVIATIONS FROM QUADRATIC TREND SHOWN IN FIGURE 136.2 8 C. MOVING AVERAGES OF DEVIATIONS FROM CUBIC TREND SHOWN IN FIGURE 136.2C D. MOVING AVERAGES OF DEVIATIONS FROM QUARTIC TREND SHOWN IN FIGURE 136.2D Figure 136.5.—Moving averages of trend deviations shown in figure 136.2. Contour interval 25 ppm. Values larger than 100 ppm and smaller than —100 ppm not contoured. ular study such maps do not exhibit as clear a reltioan of alluvia] copper to the ore body as observed in the moving-average deviations (figs. 136.3/1..D), but the patterns are similar. Which method, if either, will prove to be the better one on further application is unknown. One additional and highly pertinent matter deserves attention here; this is the problem of anomaly definition. Tn the trend-analysis method, a geochemical anomaly (a local feature of the data deserving of investigation) should be defined in terms of its magnitude, both in lateral extent and in intensity, but such definition must await additional experience with trend analysis in geochemical prospecting problems. REFERENCES Erickson, R. L., and Marranzino, A. P., 1960, Geochemical prospecting for copper in the Rocky Range, Beaver County, Utah: Art. 47 in U.S. Geol. Survey Prof. Paper 400-B, p. B98-B101. Grant, Fraser, 1957, A problem in the analysis of geophysical data: Geophysics, v. 22, p. 309-344. Hawkes, H. E., and Webb, J. S., 1962, Geochemistry in mineral exploration: New York, Harper and Row, 413 p. Krumbein, W. C., 1956, Regional and local components in facies maps: Am. Assoc. Petroleum Geologists Bull., v. 40, no. 9, p. 2163-2194. ------ 1959, Trend surface analysis of contour-type maps with irregular control-point spacing: Jour. Geophys. Research, v. 64, no. 7, p. 823-834. Oldham, C. H. G., and Sutherland, D. B., 1955, Orthogonal polynomials: their use in estimating the regional effect: Geophysics, v. 20, p. 295-306. Ostle, Bernard, 1954, Statistics in research: Ames, Iowa, The Iowa State College Press, 487 p. Whitten, E. H. T., 1959, Composition trends in a granite: modal variation and ghost stratigraphy in part of the Donegal granite, Eire: Jour. Geophys. Research, v. 64, no. 7, p. 835-848.Article 137 INVESTIGATION OF SAMPLING-ERROR EFFECTS IN GEOCHEMICAL PROSPECTING By A. T. MIESCH and J. J. CONNOR, Denver, Colo. Abstract.—Large local variation in soils and rocks causes large sampling error, and may obscure variation that is of interest in geochemical prospecting. A simulation experiment illustrates how the effects of sampling error may be overcome by dividing the area into cells and randomly selecting a number of samples per cell. Analysis of geochemical data from the Rocky Range, Utah, by Connor and Miesch (Art. 136) points to the probable importance of sampling error in obscuring subtle relations between minor elements in alluvium and buried ore deposits. Sampling error does not imply incorrect sampling procedures or mistakes in carrying out the procedures, but merely the difference between a sample and the total material that might have been sampled at the same locality. Locality here can mean a point on a sampling grid, a cell within a grid, an outcrop, or an entire rock body, depending on what the sample is intended to represent. In the following discussion, we shall assume that there is no sampling bias, that the average of a number of values representing a sampling locality tends toward the true average value for that locality, as the number of sample values is increased. In this sense, we are concerned here with sampling precision, one specific aspect of sampling error. Sampling precision obtained in any specific sampling program is determined entirely by local variation or local heterogeneity in the material being sampled. In the design of a geochemical prospecting sampling plan, the selection of the sampling interval is of particular importance. For the plan to be efficient, the interval must depend on the dimensions of the target being sought. An unnecessarily coarse sampling interval may prevent detection of the target and too fine an interval will result in sample and analytical waste. Actually, too little can be predicted about target sizes (that is, dimensions of dispersion halos) resulting from various sizes and types of ore deposits in different geologic environments to select the interval rigidly on this basis. The interval is commonly selected by consideration of target size gained through experience in specific areas or types of geologic environments. Once the sampling interval has been selected, on whatever basis, the area may be divided into a number of square cells, each cell having the dimensions of the sampling interval. The purpose of the sampling, then, is to examine variation among cells. Variation within cells is not of particular interest, except to evaluate the effects of sampling error. Due to natural local heterogeneity in many rocks and soils, the variation within cells may be significantly larger than that among cells. If this happens to be the case, any variation among cells will probably be obscure if one collects one or a number of samples per cell and plots the individual sample values on maps. Plotting average values for each cell will help clarify the pattern of variation among cells because the variation of an average is less than that of the individual values. The number of values on which each average should be computed in order to reduce the effects of within-cell variation sufficiently can be estimated from preliminary results, as shown in the following discussion. A simulation experiment has been conducted to illustrate the approach. A linear trend identical to that shown in figure 136.2A of Connor and Miesch (Art. 136), having the form Cu=90.4 + 0.88X— 0.31 Y, is assumed to represent a trace-copper trend (fig. 137.lA). Copper is measured in any arbitrary unit (for example, parts per million) and X and Y are units of distance from an arbitrarily selected origin. The map area measures 80 by 120 units and has been divided into cells 20 units on a side; values of the trend were calculated for 2 randomly selected positions in each cell. To each of these trend values was added a random D84 ART. 137 Iff U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D84-D8S. 1964.MIESCH AND CONNOR D85 X—-10 30 50 70 90 110 X—*■ 10 30 50 70 90 110 X—10 30 50 70 90 110 4* 2 i 3 \ +« \ 5 + 6 ~T- • 10 + 9 \ ' l +n \12 I +" ,1 + \ 15 \ L U + 18 1 +\ 19 \20 + 21 22 24 \ V' L \ 25 26 \+ 1 + 28 27 30 [ +29 1 32 w + ? \ 33 34 4- 71 -1 o 38 . v \ 40 + 39 V i 41 43 , \ 4“ 44 45 + 46 C. COMPUTED TREND DETERMINED FROM SAMPLE VALUES IN FIGURE 137.16 (Cu = 102.8+ 0.71X-0.16F) X—■ 10 30 50 70 90 110 TREND IN FIGURE 137.1C Figure 137.1.—Contour maps of original trend, sample values, computed trend, and trend deviations (contour interval equivalent to 25 parts per million). deviation drawn from a normal population with p=0 and cr2=7,487, providing 48 “sample values.” A sample value, V, is determined by: V=90.4+0.88 A- 0.31 Y+8. Instead of 2 samples per cell, 8 are needed to provide a cell-mean variance roughly equal to the estimate of between-cell variance. Accordingly, 6 additional samples were collected from each cell, using the same procedures used in obtaining the first 48. Figure 137.2(7 shows the distribution of copper based on means of eight values per cell. The strong highs and lows of figure 137.2A are reduced, but the imposed trend is still not particularly apparent. A regression surface fitted to the data of figure 137.2(7 and deviations from the surface are shown in figure 137.25. Analysis of variance (table 137.2, part C) indicates that the trend is statistically significant at the 95-percent confidence level and accounts for 27 percent of the total sum of squares in the average data. The trend deviations are becoming less impressive. Variance components within and between cells are estimated in table 137.1 (part C) for the 192 sample values used in figure 137.2(7. It is presumed that the new estimate of within-cell variance, based on 4 times as many samples per cell, is more stable than that previously estimated from the 48 samples (table 137.1, part B). Applying the technique of Youden and Mehlich again to the revised component estimates, a new estimate of samples needed per cell to reduce within-cell effects to an amount equal to or less thanMIESCH AND CONNOR D87 X—' 10 30 50 70 90 no A. CELL MEANS, EACH COMPUTED FROM 2 SAMPLE VALUES , USING THE 48 SAMPLE VALUES FROM FIGURE 137. 2B X—'10 30 50 70 90 110 X—' 10 30 50 70 90 110 VALUES . MAP BASED ON 192 SAMPLE VALUES X—'10 30 50 70 90 110 D. COMPUTED TREND (DASHED LINES) AND TREND DEVIATIONS USING CELL MEANS FROM FIGURE 137. 2C TREND EQUATION: Cu =103.1 + 0.61X-0.15y X—'10 30 50 70 90 no VALUES . MAP BASED ON 384 SAMPLE VALUES X—»10 30 50 70 90 no ATIONS USING CELL MEANS FROM FIGURE 137.2F TREND EQUATION: Cu = 93.6 + 0.66A’-0.13y Figure 137.2.—Contour maps of cell mean sample values, computed trends, and trend deviations (contour interval equivalent to 25 parts per million).D88 GEOCHEMISTRY those among cells is obtained: 8,865/n<638; n> 14. In an attempt to insure that within-cell effects are reduced below those between cells, 8 additional samples (rather than the indicated 6) were taken from each cell of figure 137.1.4 using the same procedures used in obtaining the previous 192. A total of 384 samples have now been collected. Figure 137.21? is a contour map of the 24 cell means based on 16 samples per cell. The regression surface fitted to the 24 cell means and the deviations from the surface are shown in figure 137.2F. The analysis of variance test of the trend is given in table 137.2 (part D). This trend is highly significant and accounts for 54 percent of the total sum of squares in the average data. A comparison of figure 137.2F with figure 137.ID indicates the large reduction of local or within-cell effects when cell means are computed from a larger number of samples. A new estimate of the number of samples per cell needed to reduce within-cell effects to an amount less than those between cells may be obtained from the components of table 137.1 (part D): 8,040/n<509; n>16. This estimate is equal to the actual number of samples used in the analysis and is twice that of the initial estimate (8 samples) computed from, the original 48 values. The initial estimate was based on too few samples to be stable, but did serve as a rough estimate or guide to the number of samples required from each cell. Increasing the number of samples per cell beyond 16 would clarify the trend and suppress the effects of within-cell variance on trend deviations even farther. In order to use the variance-component technique provided by Youden and Mehlich, it is not necessary to sample every cell in the preliminary survey. Cells to be examined in a preliminary survey might be picked randomly. Reliability of the results will in all cases depend on randomization procedures, the validity of other assumptions prerequisite to analysis of variance (Eisenhart, 1947), the number of cells sampled, and the number of samples taken per cell. Geochemical prospecting by soil analysis is well suited to the use of statistical designs. The restriction on sampling design imposed by lack of outcrops in most geologic sampling is, for the most part, not present. Generally, samples are taken over an area at more or less equal spacing, frequently on a square grid. Merits of grid samples, unalined stratified samples, and random samples and stratified random samples, are discussed by Quenouille (1945), Krumbein (1954), and Berry (1962), but the variance-component approach used here requires random sampling within each cell (stratified random samples). In the geochemical prospecting study in the Rocky Range (Connor and Miesch, Art. 136), the problem of sampling error was tentatively resolved by using a moving-average method to smooth out the data, but even though this procedure clarified the possible geologic relations shown by the data, it did introduce other problems. Variation among adjacent cells is suppressed, and independence of deviations from computed trends, important from statistical considerations, is partially destroyed. It is suggested that the estimation of variance components to determine within- and between-cell effects, and estimation of the number of samples required to sufficiently reduce the within-cell effects, is a more satisfactory method for reducing effects of local variation or sampling error in geochemical prospecting. REFERENCES Berry, B. J. L., 1962, Sampling, coding, and storing flood plain data: U.S. Dept. Agriculture, Agriculture Handb. 237, 27 p. Eisenhart, Churchill, 1947, the assumptions underlying the analysis of variance: Biometrics, v. 3, no. 1, p. 1-21. Krumbein, W. C., 1954, Statistical significance of beach sampling methods: Beach Erosion Board, Tech. Memo. 50, 33 p. ------- 1959, Trend surface analysis of contour-type maps with irregular control-point spacing: Jour. Geophys. Research, v. 64, no. 7, p. 823-834. Krumbein, W. C., and Slack, H. A., 1956, Statistical analysis of low-level radioactivity of Pennsylvanian black fissle shale in Illinois: Geol. Soc. America Bull., v. 67, no. 6, p. 739-761. Quenouille, M. H., 1945, Problems in plane sampling: Math. Statistics Annals, v. 20, p. 355-375. The Rand Corporation, 1955, A million random digits, with 100,000 normal deviates: Glencoe, 111., The Free Press, 600 p. Tippett, L. H. C., 1952, The methods of statistics, 4th ed.: New York, John Wiley and Sons, Inc., 395 p. Youden, W. S., and Mehlich, A., 1937, Selection of efficient methods for soil sampling: Boyce Thompson Inst. Contr., v. 9, p. 59-70.Article 138 EFFECT OF CATION EXCHANGE ON THE THERMAL BEHAVIOR OF HEULANDITE AND CUNOPTILOLITE By ANNA O. SHEPARD and HARRY C. STARKEY, Denver, Colo. Work done in cooperation with the U.S. Atomic Energy Commission Abstract.—Heulandite has the thermal stability of clinoptilolite when the calcium is largely replaced by potassium. Clinoptilolite has the thermal behavior of a heulandite-like mineral, associated with some natural clinoptilolites, when calcium replaces most of the alkalis. X-ray analysis was used in the experiments. Heulandite and clinoptilolite, zeolites with nearly identical X-ray patterns, differ in thermal behavior. The crystal structure of heulandite is changed on heating; that is, the mineral inverts. The higher temperature form has been referred to as heulandite B (Slawson, 1925; Milligan and Weiser, 1937; Koizumi and Kiriyama, 1953; and Mumpton, 1960). The temperature of inversion varies with the length of time that it is heated. Clinoptilolite does not invert on heating, and it withstands higher temperatures and longer heating than heulandite (Mumpton, 1960; Shepard, 1961). In chemical composition, these minerals differ in two respects: (1) the calcium content is higher than that of alkalis in heulandite, whereas alkalis exceed calcium in clinoptilolite, and (2) clinoptilolite has the higher silica-alumina ratio. Mumpton (1960, p. 363) considered that the essential difference between the two minerals is in the silica-alumina ratio: he reasoned that a mineral should not be defined by its exchangeable cation. Mason and Sand (1960, p. 348-349) emphasized the difference in cations and illustrated graphically that the two minerals differ more in content of CaO • Al20:i and (Na, K)2 O • A1203 than in silica content. Mumpton, Mason, and Sand, contrary to Hey and Bannister (1934), all considered heulandite and clinoptilolite distinct minerals because of the apparent absence of minerals of intermediate composition that would characterize an isomorphous series. Investigation of the effect of chemical composition on the thermal behavior of these two zeolites has been difficult because clinoptilolite, unlike heulandite, is not easily obtained in a pure state. A possible means of studying the effects of alkalis and alkaline earths is by the exchange of cations in the minerals. Results of preliminary tests that were limited largely to a single set of conditions are reported at this stage of the study because the properties of some of the products are similar to those of certain natural minerals that occur as intergrowths with clinoptilolite. The samples used in these cation-exchange experiments included heulandite from Summit and Paterson, N.J., and Berufiord, Iceland, and clinoptilolite from the Pierre Shale, Buffalo County, S. Dak. The heulandites were well crystallized. Clinoptilolite from the Pierre Shale had an exceptionally strong X-ray pattern that showed no impurities, but minute inclusions could be detected optically. Normal solutions of calcium, potassium, and sodium chlorides, neutralized with their respective hydroxides, were used to provide the exchangeable cations. Samples were of two grain sizes, less than 100 mesh and less than 2 microns. The process of exchange was accelerated by heating samples in a pressure cooker for 16 hours (Starkey, Art. 139), a method that might be classed as mildly hydrothermal. The thermal behavior of the minerals before and after exchange was determined by X-ray analysis of samples heated for 10 minutes at 50° C temperature intervals (Shepard, 1961). Samples were X-rayed as soon as cool enough to handle and scanned rapidly (2°/min) because heulandite B begins to change back to the low-temperature form 15 minutes after cooling. With the heating schedule adopted, the natural heulandite samples started to invert at 250 °C, the pattern of heulandite B attained maximum intensity between 300°C and 350 °C, and its structure was nearly ART. 13* IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D89-D92. 1964. D89D90 geochemistry xo to 5.0 10 d-SPACING, IN ANGSTROM UNITS Figure 138.1.—X-ray diffractograms illustrating thermal behavior of heulandite. A-C, untreated mineral; D, mineral treated with IN K61. destroyed at 500°C (fig. 138.1, A-C). When these samples were treated with liVKCl and heated, some did not invert at all and others contained only a minor constituent that inverted, forming a weak 020 heu-landite-B peak. Like clinoptilolite, the heulandite that did not invert could be heated above 800°C before its structure was destroyed (fig. 138.1, D). The effects of the length of time the sample was heated, grain size, and concentration of the solution used for cation exchange were tested in samples that showed a small amount of heulandite B after exchange with potassium chloride. Increase of the heating period from 16 to 48 hours did not eliminate the weak heu-landite-B 020 peak, but when grain size was reduced to less than 2 microns, a dimension still well above the limits of X-ray resolution, no trace of the peak remained. Increase of the concentration of the potassium chloride solution from IN to 2JV affected the inversion of the 020 heulandite-B peak without decreasing its intensity; contraction of the peak was more gradual than it is in typical heulandite (fig. 138.7, A, B). A similar behavior has been observed in some heulandite-like minerals from the Nevada Test Site. Clinoptilolite from the Pierre Shale (fig. 138.2, C) treated with a IN CaCl2 solution is rendered less stable; it inverts and the positions of nine strong peaks are similar to those of heulandite B (fig. 138.2, D). This modified clinoptilolite is not identical with heulandite, however; the temperatures of inversion and of destruction are both approximately 100°C higher than those of heulandite. The stabilizing effect of potassium on heulandite and the loss of stability of clinoptilolite when alkalis are replaced by calcium indicate that potassium is an important factor in determining the thermal stability of these two zeolites. Treatment with 1N NaCl did not affect the thermal behavior of either clinoptilolite or heulandite. The sodium content is usually higher than potassium in natural clinoptilolite, and the amount of exchange of __________i___________________i___________i______._____.___i—i—i— 3.0 4.0 5.0 10 d-SPACING, IN ANGSTROM UNITS Figure 138.2.— X-ray diffractograms illustrating thermal behavior of heulandite and clinoptilolite. A, B, heulandite treated with 2N KC1; C, D, clinoptilolite treated with IN CaClj.SHEPARD AND STARKEY D91 potassium is as yet undetermined. Lack of change in the thermal behavior of heulandite treated with 1N NaCl emphasizes the importance of determining the extent of exchange. If the cations of the heulandite after treatment with sodium chloride are comparable to those of natural clinoptilolite and the heulandite retains its original thermal behavior, the possibility that the silica-alumina structure also affects thermal behavior has to be considered. The composition of the minerals after the cation exchange has not yet been determined by wet chemical analysis. Estimations of the extent of alteration induced by these exchange procedures are based on results of analyses of leachates. The cations of the natural and the treated minerals were released by IN neutral ammonium chloride by exchange methods. Comparison of results is therefore limited to exchangeable cations. In the analysis of leachates, alkalis were determined with a flame photometer and calcium by versene titration. Table 138.1 .—Comparison of cations released by NH4+1 from natural and from previously heated minerals [Cations in milliequivalents per 100 g of sample] Heulandite from Summit, N.J. Clinoptilolite from Pierre Shale Cation Natural mineral Mineral treated with— Natural mineral Mineral treated with— KCl NaCl CaCli NaCl Ca Na 169.5 14.5 20.0 91.1 124.1 106.0 43.9 69.4 59.6 167.0 9.6 122.2 20.2 K 162.3 13.9 The analyses show high percentages of critical cations in the treated minerals. Expressed in meq/100 g of sample, potassium made up 63 percent of the exchanged cations in the leachate of the KCl-treated heulandite that did not invert, and calcium made up 93 percent of the exchanged cations in the leachate of the CaCl2-treated clinoptilolite that inverted. The calcium-potassium ratios of the determined cations (meq/100 g of sample) in the leachates from the natural and comparable treated minerals are in similar ranges: natural heulandite, 8.48, calcium-treated clinoptilolite, 12.01; natural clinoptilolite, 0.74; KCl-treated heulandite, 0.56. Analyses of the ammonium chloride leachate of the sodium-treated heulandite indicate a complete replacement of the potassium with sodium and depletion of the calcium; nevertheless, the amount (in milliequivalents) of calcium is still greater than the amount of sodium, a fact that may account for the typical heulandite-like thermal behavior of the mineral. 716-626 0—64---7 Sodium did not fully replace the potassium in the clinoptilolite, but it constitutes 80 percent of the sum of the three cations, expressed in milliequivalents, and calcium is reduced to 6 percent. The extent of exchange with ammonium chloride in these tests is indicated by comparison of the calcium, potassium, and sodium cations determined in leachates of heulandite and clinoptilolite with the determination of these ions by wet chemical analysis of these minerals, (table 138.2). Table 138.2.—Percentage of original cations in leachate from cation exchange [Cations in meq/100 g of sample] Heulandite from Summit, N.J. Clinoptilolite from Pierre Shale Cation Mineral1 Leachate Percent exchanged Mineral Leachate Percent exchanged Ca 244.0 169.5 69.5 51.8 43.8 84.7 Na 21.0 14.5 69.0 78.3 69.4 88.6 K 29.3 20.8 71.0 56.1 56.9 101.4 i Cations in original mineral determined by wet chemical analysis and converted to meq/100 g of sample. Approximately 70 percent of the cations expressed in meq/100 g in samples of heulandite and an average of 90 percent of those in samples of clinoptilolite were exchanged. The refractive indices of the minerals also show the effects of exchange, although they are less definitive than the thermal effects. The untreated Paterson heulandite with indices in white light of a, 1.494, /3, 1.495 + , and y, 1.500-1.501 appeared as a fine aggregate with an average index of 1.482 after treatment with potassium chloride. The average index of the untreated Pierre clinoptilolite, also an aggregate, was 1.470. Treatment with calcium chloride raised it to 1.480, not quite as high as that of the KCl-treated heulandite. Natural zeolites that behave thermally as though both clinoptilolite and heulandite or a “heulandite-like” mineral are present occur in tuffs of the Nevada Test Site (Shepard, 1961) and in other areas where clinoptilolite is the principal zeolite. Some of these zeolites are zoned, but most crystals are too fine to determine their optical homogeniety. The “heulandite-like” component of these samples does not invert at as low a temperature as typical heulandite. Its thermal behavior is similar to that of calcium-exchanged clinoptilolite. Natural zeolites with the thermal behavior of both clinoptilolite and a heulandite-like mineral may have been formed by continuous crystallization under a changing environment to produce zoning comparable to that of plagioclase or they may have been formed by partial natural exchange of cations of clinoptiloliteD92 GEOCHEMISTRY during some period after crystallization. The preliminary tests of the effects of cation exchange on the thermal behavior of heulandite and clinoptilolite suggests that partial natural exchange may have been the cause of the anomalous behavior of the heulandite-like mineral. Anomalous effects that have been reproduced in the laboratory include the higher temperatures as compared with those of heulandite at which the mineral inverts and is destroyed, the modification of the contraction of the 020 reflection during inversion, and the change in refractive indices. Cation size and charge must both affect the results obtained in the attempt to exchange cations of heulandite and clinoptilolite. It has seemed doubtful that a univalent cation larger than calcium could replace it and balance the negative charge of heulandite, but the thermal behavior of the KCl-treated heulandite clearly indicates that an effective amount of potassium was introduced. The replacement of potassium and sodium by bivalent calcium in the clinoptilolite structure should not be difficult, yet a heulandite-like mineral rather than typical heulandite was produced in the attempted exchange. The extent of exchange and the effect of the silica-alumina ratio of the two minerals are both factors to be considered in future experiments. REFERENCES Ames, L. L., Jr., Sand, L. B., and Goldich, S. S., 1958, A contribution on the Hector, California, bentonite deposit: Econ. Geology, v. 53, p. 22-37. Bramlette, M. N., and Posnjak, E., 1933, Zeolitic alteration of pyroclastics: Am. Mineralogist, v. 18, p. 167-171. Hay, R. L., 1963, Stratigraphy and zeolitic diagenesis of the John Day Formation of Oregon: Univ. Calif. Publ. in Geology, v. 42, p. 199-262. Hey, M. H., and Bannister, F. A., 1934, Studies on the zeolites, Part VIII, “Clinoptilolite,” a silica-rich variety of heulandite; Mineralogical Mag., v. 23, p. 556-559. Koizumi, M., and Kiriyama, R., 1953, Structural changes of some zeolites due to their thermal dehydrations: Univ. Osaka Sci. Rept. no. 2, p. 67-85. Mason, B., and Sand, L. B., 1960, Clinoptilolite from Patagonia; the relationship between clinoptilolite and heulandite: Am. Mineralogist, v. 45, p. 341-351. Milligan, W. D , and Weiser, H. B., 1937, The mechanism of the dehydration of heulandite: Jour. Phys. Chemistry, v., 41, p. 1029-1041. Mumpton, F. A., 1960, Clinoptilolite redefined: Am. Mineralogist, v. 45, p. 351-370. Piersson, L. V., 1890, On mordenite: Am. Jour. Sci., v. 40, p. 232-237. Shepard, A. O., 1961, A heulandite-like mineral associated with clinoptilolite in tuffs of Oak Spring Formation, Nevada Test Site, Nye County, Nevada: Art. 264 in U.S. Geol. Survey Prof. Paper 424-C, p. C320-C323. Slawson, C. B., 1925, The thermo-optic properties of heulandite: Am. Mineralogist, 10, p. 305-331.Article 139 DETERMINATION OF THE ION-EXCHANGE CAPACITY OF A ZEOLITIC TUFF By HARRY C. STARKEY, Denver, Colo. Abstract.—Determination of the exchange capacity of some zeolitic tuffs may require an extremely long time for the reaction rate to attain a constant value. A heating method is suggested whereby the rate of the reaction is increased, so that the ex-change-capacity determinations can be made within a period of time as short as 24 hours. Studies of the ion-exchange capacities of volcanic tuffs collected at the Nevada Test Site showed that the longer the samples were leached, the higher were the observed capacities. A method was sought which would enable a constant value to be attained within a reasonably short time. The sample used for most of the experimental work, sample B, had the following mineralogical composition, based on X-ray diffraction analysis using nickel-filtered copper radiation: Approximately two-thirds consisted of quartz and feldspar with a trace of montmorillonite; one-third was made up of a zeolite which appeared to be a heulandite-clinoptilolite type. The latter had a strong (004) peak at 22.3°20 which, according to Mumpton (1960), indicates clinoptilolite. Mumpton described clinoptilolite, however, as having a structure which, on heating, persisted to above 700 °C, whereas heulandite showed a phase change at 230°± 10°C. For sample B, the intensity of the diffraction peaks began to diminish at about 270°C and the peaks disappeared at about 600 °C. Mason and Sand (1960) state that heulandite and clinoptilolite have the same structure, but that heulandite is a calcium-rich zeolite and clinoptilolite is a sodium-potassium-rich mineral. The leachate of sample B contained about three times as much calcium as sodium and potassium. These data indicate that the zeolite is better described, following Mason and Sand (1960), as heulandite than as clinoptilolite. An ammonium chloride batch method was used to determine the exchange capacity of the experimental ART. 139 IN U.S. GEOL. SURVEY PROF material. The sample was crushed lightly to break up aggregates and then put through a 60-mesh sieve; no grinding was necessary except for some quartz and feldspar. One hundred fifty ml of 1.0 N NH4C1 solution neutralized with ammonium hydroxide was added to 1 gram of sample. After soaking without agitation for specified times in covered beakers at room temperatures which ranged from 17°C to 30 °C, the mixtures were centrifuged and the supernatant liquids retained for cation analysis. To samples which were leached longer than 24 hours, distilled water was added periodically to maintain the volume. The samples were then washed 4 times with ethyl alcohol, the first 2 washings being added to the supernatant liquid. The sodium and potassium in the supernatant liquid were analyzed by flame photometry; the calcium and magnesium by versene titration. The amount of ammonia adsorbed on the sample was determined by distillation. Sample A, a soil containing montmorillonite, mica, kaolinite, quartz, and feldspar, as determined by X-ray diffraction, was run for comparative purposes. As can be seen on figure 139.1, sample A reached a constant value almost immediately, but the exchange-capacity values of sample B increased with time. The values are plotted on semilog paper. Other tuff samples, not shown, exhibited the same characteristics. In an attempt to reduce the long leaching times necessary to achieve a constant value in the observed capacities of the tuff, a series of experiments was carried out at a temperature (95°C) just short of boiling, using the same leaching times as in the studies at room temperature. As can be seen in figure 139.2, the heated samples reached a constant value at 84 milliequivalents/100 grams in about 16 days. By extending both curves, and assuming no leveling off of the room-temperature curve, it appears that it would take at least 4 years to reach a constant value at room temperature. . PAPER 475-D, PAGES D93-D95. 1964. D93D94 GEOCHEMISTRY Figure 139.1—Exchange capacity of samples A (soil) and B (zeolitic tuff), determined at room temperature as a function of time. To see if the rate of exchange could be further increased, a series of runs was made in which the time remained constant at 24 hours but in which the temperatures were varied. The values produced a straight line when plotted on arithmetic graph paper (fig. 139.3). The highest temperature was obtained by heating the covered samples in an atmosphere of 100 percent relative humidity in a household pressure cooker, which produced approximately 2 atmospheres of pressure. The temperature, 118°C, was calculated from the pressure. The exchange capacity measured after 24 hours was 84 meq/100 g, identical with that determined at 95°C, indicating that a constant value had been obtained. The pH of the ammonium chloride was lowered from 7.0 to 6.4. Increasing the heating time at 118°C to 48 hours on another sample resulted in no change in the measured exchange capacity. Figure 139.3.—Determined exchange capacity of zeolitic tuff (heated for 24 hours) as a function of temperature. As the exchange reaction approached a constant value, the ratio of the magnesium to the sum of the cations in the leachate increased (see accompanying table). In the unheated series the ratio increased from 0.10 at half an hour to 0.20 after 128 days; in the series heated at 95°C the ratio increased from 0.14 to 0.26. Ratios of 0.26 were also obtained for samples heated at 118°C for 24 and 48 hours. As determined by visual estimation of X-ray diffraction pattern, which may be in error by ±10 percent, approximately a third of sample B is zeolite. If the determined exchange capacity is multiplied by 3 a value of 252 meq/100 g is obtained. This value is reasonable when compared with the exchange capacity of heuland-ite (330 meq/100 g) given by Barrer (1958), who obtained the value by calculation from the hydrated formula weight. Effect of time and temperature on exchange capacity and ratio of magnesium to sum of cations in zeolitic tuff Treatment Exchange capacity (meq/100 g) Ratio of magnesium to sum of cations Unheated (17°-30°C) lA hour. _ . . 12. 8 0. 10 128 days.. .. .. .... .. 59. 2 . 20 Heated (95°C) H hour. . . 19. 6 . 14 16 days ... 84. 0 . 26 Heated (118°C) 24 hours ... .. . . 84. 0 . 26 48 hours ------ 84. 0 . 26 Figure 139.2.—Comparison of rates of exchange in zeolitic tuff, as determined at room temperatures and at 95°C.STARKEY D95 Comparison of X-ray diffraction patterns made of samples both before and after treatment did not indicate any structural changes in the sample. In summary, the long period of time for the exchange reaction of some zeolites to reach a constant value at room temperatures may be shortened by heating. The identical values obtained at both 95°C and 118°C indicate that heating does not alter the exchange capacity or the ratios of the cations. The increase, with time, in the ratio of magnesium to the sum of the cations indicates that the magnesium is more difficult to exchange in this type of sample. REFERENCES Barrer, R. M., 1958, Crystalline ion-exchanges: Chem. Soc. (London) Proc., April 1958, p. 99-112. Mason, Brian, and Sand, L. B., 1960, Clinoptilolite from Patagonia; the relationship between clinoptilolite and heulandite: Am. Mineralogist, v. 45, p. 341-350. Mumpton, F. A., 1960, Clinoptilolite redefined: Am. Mineralogist, v. 45, p. 351-369.Article 140 GEOLOGICAL AND GEOCHEMICAL RECONNAISSANCE, SOUTHERN PART OF THE SMYRNA MILLS QUADRANGLE, AROOSTOOK COUNTY, MAINE By LOUIS PAVLIDES and F. C. CANNEY, Beltsville, Md., and Denver, Colo. Abstract.— Closely folded Paleozoic rocks, mostly impure calcic and ankeritic limestone, micaceous siltstone and quartzite, and slate, comprise the principal sedimentary rocks of the southern part of the Smyrna Mills quadrangle. Felsic plutons locally have intruded and thermally altered these regionally metamorphosed rocks (chlorite zone) to cordierite-biotite and calc-silicate hornfels. Thus, the plutons are enclosed by hornfels aureoles up to 2 miles wide. Stream-sediment sampling revealed a slightly higher-than-background molybdenum content in four streams draining an area underlain by felsic rocks and plutons. Tests also were made for antimony, zinc, and heavy metals. Seasonal variations of copper and molybdenum were detected. Closely folded sedimentary rocks of Paleozoic age in the chlorite zone of regional metamorphisrn have been locally intruded and thermally metamorphosed by felsic plutons in the southern part of the Smyrna Mills quadrangle. Thin glacial till mantles much of the upland areas, whereas stratified deposits of sand and gravel cover the lowlands. The topography is controlled by the bedrock except for local eskers and kames. High ridges of hornfels surround areas of rounded and subdued hills and ridges underlain by felsic plutons (fig. 140.1). The northwest part of the area is underlain by folded sedimentary rocks and is mostly a lowland with gently rolling hills, except for the ridge formed by a small felsic mass and its associated hornfels aureole. The ribbon rock member of the Meduxnekeag Formation (Pavlides, 1962, p. 11-12 and 21-23) of Middle Ordovician and probably Late Ordovician age contains the oldest rocks of the region. It consists chiefly of thin- to medium-bedded gray-blue impure calcic and ankeritic limestones, interlayered with each other and with slate. Fossils have not been found in the green and gray-green phyllite that lies between the Meduxnekeag Formation and the Silurian rocks, and these phyllites can only be dated as of Late Ordovician or Early Silurian age or both. Locally this phyllite is calcareous and contains minor amounts of conglomerate. The undifferentiated Silurian rocks span Early through Late Silurian time as determined from the graptolite faunas identified by W. B. N. Berry (written communications, 1961, 1962) from several localities within them. Gray and gray-green micaceous siltstone and quartzite, commonly interlayered with gray and gray-green slate, are the dominant rock types in this unit; conglomerate is locally abundant. Manganese-and iron-bearing lenses are also present at several horizons. Calcareous slate that contains brachiopods of Silurian or Devonian age (A. J. Boucot, oral communication, 1963) may be the youngest sedimentary unit of the area. The sedimentary rocks are closely folded along steep to vertical axial planes that trend east to northeast. Bedding normally dips 70° to vertical and well-developed fracture cleavage with equally steep dips trends east to the northeast, except in the northeast part of the area where it locally trends northwest. The final folding of the region occurred during the Acadian disturbance, following which the folded rocks were intruded by felsic plutons. The felsic plutons are not well exposed. The Hunt Ridge pluton appears to contain mostly fine to mediumgrained biotite-hornblende granodiorite, whereas in the Pleasant Lake pluton porphyritic medium- to coarsegrained hornblende-biotite quartz monzonite is present near the center and biotite granodiorite along the southwest margin. The two much smaller masses (fig. 140.1) are granitic. Thermal aureoles up to 2 miles wide rim the plutons and consist of cordierite-biotite and calc-silicate hornfels in which stratification is preserved but cleavage is obliterated. On the northeast side of the Hunt Ridge pluton, hornfels formed ART. 140 IN U.s. GEOL. SURVEY PROF. PAPER 475-D, PAGES D96-D99. 1964. D96PAVLIDES AND CANNEY D97 iOmm' 'Oakfield ■\MED UXNEKEA G M LAKE 0 Dyer Brook uton Hunt 1000 * ikitacook/ \ Lake ,r Area of report MAINE Pleasant Lake pluton^ \ “4 \ ' >» * 4 X ... > V V 68° 00' 1012 MILES I_____i_____I____________I___________I CONTOUR INTERVAL 200' EXPLANATION Di m ; Granodiorite and quartz monzonite EDSsE Calcareous slate mentary rocks USOpZ Phyllite Meduxnekeag Formation, ribbon rock member ■ > * ec o°=> . Su - _i - Sum/ ■ : V, '■/: c/> Thermally metamorphosed sedimentary rocks Diagonal dashes indicate injected zones iSOprX Thermally metamorphosed phyllite Thermally metamorphosed Meduxnekeag Formation, ribbon rock member Inferred contact Inferred fault _40 Sample locality Figure 140.1.—Reconnaissance bedrock geology and stream-sediment sample sites, southern part of Smyrna Mills quadrangle, Maine.D98 GEOCHEMISTRY from sedimentary rocks of Silurian age is intruded by sills and minor dikes of granodiorite, quartz monzonite, and granite of Devonian age. Many magnetic anomalies, some of considerable magnitude, are present in the hornfels aureoles (Dempsey, 1962). The formation of abundant magnetite by thermal metamorphism of tne ferruginous lenses present in the undifferentiated Silurian unit is believed to be the cause of these local magnetic anomalies. Small amounts of galena, chalcopyrite, and stibnite occur in quartz veins in hornfels on the east side of Meduxnekeag Lake (Houston, 1956, p. 109-110), just beyond the east margin of figure 140.1. A reconnaissance geochemical drainage survey was made in 1960 and 1961 of the felsic plutons and their aureoles. Samples of fine-grained active stream sediment were collected at the 41 sites shown on figure 140.1. The samples were dried, screened through a 50-mesh sieve (0.25-mm opening), and the fine fractions (lg to 25g) analyzed by geochemical prospecting field methods (Ward and others, 1963). All samples were analyzed for copper, lead, zinc, molybdenum, cold-extractable copper, and cold-extractable heavy metals (see accompanying table); the 1960 samples were also analyzed for tungsten and antimony. In general, the values are quite low and with a few exceptions fall within normal background ranges elsewhere in Maine in similar geologic terrane. Interpretation of these data, however, is complicated because the copper and molybdenum contents of the active alluvium in some of the streams appear to show time variations. The eight paired samples listed in the table were collected in 1960 and 1961. The data for these show that the copper and molybdenum contents of many of the 1961 samples are appreciably lower than the corresponding 1960 values. No significant differences were noted in the zinc content, however. Although the variability between the 1960 and 1961 copper and molybdenum values is such that no single normalizing factor would be entirely correct, all the data for these can be made more directly comparable by doubling the 1961 values wherever a definite figure is given. Variations in the metal content of stream sediments with time are known to exist. Barr and Hawkes (1963), who described time variations in the copper content of the stream sediments in two areas in British Columbia, also noted a weak inverse relationship between stream discharge and metal content of the sediments. In the Smyrna Mills area, too, both rainfall and average stream discharge were distinctly less in the summer of 1960 than in the corresponding period of 1961. The Partial chemical analyses of stream-sediment samples taken in 1960 and 1961 in the southern part of the Smyrna Mills quadrangle, Maine [Asterisk on sample number indicates 1961 sample. Analyses by O. A. Nowlan, Q. H. Van Sickle, and M. J. Bright, Jr.] Metal content (parts per million) 1 Site Sample Cu Zn Mo Sb Cold-ex- tractable Cu Cold-extractable heavy metals 2 1 A1580 15 100 <4 1 1 11 2 1581 20 100 4 1 ll N. P 32-33B 33 22 N. 48 E. 11,050 Pegmatite and amphibolite. 7 4679___ Amerada 1 Iverson-Nelson Unit- 2 155 N. 96 W. 13.603 Altered syenite. 8 BT-85— Amerada N.D. “A” Unit 9 16 156 N. 95 W. 14,825-14,827-.- Garnet-biotite gneiss. 9 4678 Amerada 1 Antelope Unit A __ 1 152 N. 95 W. 15,128 Hornblende-hypersthene gneiss. 10 BT-86 10 139 N. 101 W. 13,645 Biotite-quartz monzonite (?). 11 4680 Herman Hanson 1 Mueller . 20 140 N. 65 W. 3,305 Biotite-quartz monzonite. D100 ART. 141 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D100-D104. 1964.PETERMAN AND HEDGE D101 150 300 MILES Figure 141.1.—Structural provinces of the Canadian Shield showing age of major orogenies. Combinations of patterns show some principal areas of mixed isotopic ages resulting from overlap of orogenic belts. Boundaries between provinces are shown by dotted lines. called the Hudsonian orogeny (Stockwell, 1961). The western part of the Superior province includes the shield areas of Ontario and parts of Manitoba and northern Minnesota. Rocks from this province yield ages commonly ranging from 2.4 to 2.6 b.y. and represent the Algoman (Goldich and others, 1961) or Kenoran (Stock-well, 1961) orogeny. In the exposed Precambrian of northern Manitoba, the boundary between the Churchill and Superior provinces is marked by well-developed faulting, ultramafic intrusions, and a regional gravity low paralleled by gravity highs. On the basis of geologic and gravity data, Innes (1960) and Wilson and Brisbin (1961) have suggested that this zone is the remnant of an island arc or alpine mountain system. The gravity anomalies (Innes, 1960; Wilson and Brisbin, 1961) and the K-Ar ages of subsurface samples (Bur-wash and others, 1962) indicate a southward extension of the boundary zone in the subsurface to the international border. The location of the boundary in North Dakota, as shown in figure 141.2, is a projection from Canada fitted to the isotopic ages of basement-rock samples. The terminology used in this report is that of Goldich and others (1961).D102 GEOCHRONOLOGY Table 141.2.—Rb-Sr and K-Ar ages of samples of basement rocks from the Williston basin area [K-Ar ages from Burwash and others (1962), except Nos. 3 and 6 (hornblende), from Peterman (unpublished PhD. thesis, 1962)] Locality Rb-Sr K-Ar (flg. 141.1) Radiogenic Radiogenic Sr87 Age (b.y.) Age (b.y.) Mineral Material Rb87 (ppm) Sr87 (ppm) Total Sr87 1___ 184 4. 55 0. 52 1. 66 ±. 08 1. 20 Biotite. 2_ _ 52. 1 1. 33 . 12 1. 71 ±. 19 1. 49 Hornblende. 3 95. 8 2. 52 . 12 1. 77 ±. 19 1. 71 Biotite. 4 . 61. 8 2. 35 . 24 2. 54 ±. 15 1. 57 Do. 5 __ 64. 8 1. 63 . 16 1. 69 ±. 15 1. 51 Do. 6 75. 6 2. 02 . 10 1. 79 ±. 22 1. 72 Do. 1. 77 Hornblende. 7 65. 1 <. 90 <. 01 <. 93 8 128 . 992 . 35 . 52 ±. 03 128 . 988 . 37 . 52 ±. 03 9 K-feldspar — 131 3. 32 . 16 1. 70 ±. 15 2. 20 Hypersthene. 10 187 4. 32 . 77 1. 55 ±. 08 1 1 103 3. 29 . 25 2. 14±. 13 Decay constants, abundances, and normal isotopic ratios used in calculations: Rb87: 1.47X 10-" yi-1 K40: 0.585X 10-10 yr_1 (electron capture) 4.72X 10-10 yr_1 (beta decay) Rb87: 0.283 g/g Rb K40: 1.22X l()-4g/g K Sr87/Sr86 = 0.703 (assumed initial) Sr86/Sr88 = 0.1194 Exposed Precambrian Shield rocks Boundary zone between subsurface extension of Churchill province to west and Superior province to east ^Locality 6*1.7 K-Ar age, b.y. Rb Sr age, b.y. ▲ Outcrop samples, Little Rocky Mountains Figure 141.2.—Rb-Sr and K-Ar ages of samples of basement rocks from the Williston basin of North Dakota and of adjacent areas. Locality numbers 1-11 refer to this article (see tables 141.1 and 141.2); locality numbers 18, 19, 45, and 50-54 refer to Burwash and others (1962).PETERMAN AND HEDGE D103 DISCUSSION Isotopic ages, ranging from 0.5 to 2.5 b.y., are available for a total of 18 localities in an area of over 220,000 square miles (fig. 141.2). The age determinations, however, fall into a useful and interpretable pattern (fig. 141.3). Two major periods of metamorphism and orogeny, the Algoman approximately 2.5 b.y. ago and the Penokean approximately 1.7 b.y., are indicated and one or more younger events suggested. The pattern of ages, in part, shows the cumulative effects of two or more thermal events in which the Rb-Sr and K-Ar systems are variously affected in different minerals. For example, at locality 4 in Saskatchewan, the Rb-Sr age of 2.5 b.y. for feldspar indicates an Early Pre-cambrian rock, and the K-Ar age of 1.6 b.y. for biotite reflects the Penokean event at the close of Middle Precambrian time. Similarly, at locality 50 in Manitoba a K-Ar age of 1.93 for biotite was reported by Burwash and others (1962, p. 1621). The hornblende-biotite gneiss represented in the core is probably also of Early Precambrian age. If the lowering of the K-Ar age is attributed to the Penokean event 1.7-1.8 b.y. ago, as is suggested in figure 141.3, loss of the radiogenic argon was incomplete, and the somewhat older age of 1.9 b.y. for the biotite may be referred to as a survival value. The quartz inonzonite cored at locality 5 between localities 4 and 50 may have been emplaced during the Penokean orogeny. The Rb-Sr age for feldspar from this rock is 1.7 b.y., and the K-Ar age for biotite is 1.6 b.y. Early Precambrian . Middle Precambrian The pattern of the isotopic ages from east to west clearly indicates that two age provinces are present in the basement rocks. The buried Precambrian rocks of Manitoba are related to the Superior province of the Canadian Shield exposed to the east in Manitoba, Ontario, and Minnesota. The buried basement rocks of southestern Saskatchewan show isotopic ages that are characteristic of the Churchill province. There appears to be a broad transitional zone (fig. 141.2) between the Superior province age of 2.4-2.5 b.y. and the Churchill province age of 1.7-1.8 b.y. A similar broad zone of mixed ages, between 2.5 and 1.7 b.y., was found (Lowdon and others, 1963; Burwash and others, 1962; Moore and others, 1960) for rocks of the Superior province adjacent to the Churchill-Superior boundary zone in northern Manitoba (fig. 141.1). The subsurface extension of the Superior province in eastern North Dakota is indicated by K-Ar ages of 2.4-2.5 b.y. for biotite from gneissic rocks at localities 53 and 54 (Burwash and others, 1962). K-feldspar from a quartz monzonite cored at locality 11 was dated at 2.14 b.y. The analytical error is estimated at ±130 m.y. (table 141.2), but it cannot be assumed that the feldspar remained a closed system. It cannot be said, therefore, whether the determined age marks the time of origin of the quartz monzonite or represents a survival value. The ages for the samples from western North Dakota, Montana, and Saskatchewan give the peak at Late Precambrian . Paleozoic Algoman orogeny Penokean Unknown events orogeny AGE, IN B.Y. (109 YEARS) Figure 141.3.—Bar graph of Rb-Sr and K-Ar age determinations shown in figure 141.2. Locality numbers are indicated. Terminology is that of Goldicli and others (1961).D104 GEOCHRONOLOGY 1.7 b.y. shown in the histogram (fig. 141.3). The Rb-Sr age of 1.7 b.y. for the whole-rock sample of rhyolite (loc. 2) and 5 feldspar ages in the range 1.66 to 1.79 b.y. suggest a specific event, the Penokean. The range of 130 m.y. in the ages may represent differences in ages of the igneous rocks, but analytical errors must also be considered, and the time differences may be apparent rather than real. The Rb-Sr ages, however, are more consistent than the K-Ar ages for which the overall analytical precision of 5 percent is generally much better than that of the Rb-Sr method. The K-Ar age of 1.2 b.y. for biotite from the granite at locality 1 clearly indicates some disturbance in the area subsequent to the Penokean orogeny, and this is substantiated by Rb-Sr ages of 520 m.y. for biotite (loc. 8) and of approximately 900 m.y. for K-feldspar (loc. 7). Altered syenite was cored at a depth of 13,603 feet at locality 7. The age of 930 m.y. (table 141.2) computed for K-feldspar from the syenite is a maximum value, and the age calculation is uncertain because of the small content of radiogenic Sr87. The biotite with a Cambrian age of 520 m.y. is from a garnet-biotite gneiss cored at a depth of 14,825 feet. The low ages at localities 7 and 8 might be attributed to thermal effects due to the depth of burial except that K-feldspar from a hornblende-hypersthene gneiss at locality 9, cored at a depth of 15,128 feet, had an Rb-Sr age of 1.7 b.y. Hypersthene from this core was dated by K-Ar at 2.20 b.y., but this age is suspect because it has been shown that pyroxenes may contain excess argon (Hart and Dodd, 1962). It should be noted that the wells at localities 7 and 8 were drilled on a basement high. Localities 1 and 2, which yielded low K-Ar ages for biotite, are also a basement high. Possibly these high points of the buried basement were centers of later igneous or metamorphic activity. The Little Rocky Mountains in Montana (Iocs. 18 and 19) contain Tertiary intrusive rocks in Precambrian gneisses. The Penokean age of 1.7 b.y. (fig. 141.2) represents K-Ar ages for hornblende from the gneisses (Burwash and others, 1962). CONCLUSIONS The major part of the Williston basin of western North Dakota and adjacent areas is underlain by Precambrian rocks which yield isotopic ages of 1.7-1.8 b.y., indicating that the buried basement rocks are an extension of the Churchill province of the Canadian Shield. The eastern part of the basin is floored by a subsurface extension of the Superior province with rocks approximately 2.5 b.y. in age. The projection of the Churchill-Superior boundary trends southward through eastern Saskatchewan and western North Dakota. Mixed isotopic ages in the range 2.5 to 1.6 b.y. indicate that Early Precambrian rocks were affected in varying degree by the Penokean orogeny approximately 1.7 b.y. ago. The Churchill-Superior boundary in the subsurface, as delimited by isotopic age determinations, probably is a broad zone rather than a sharp line. Younger thermal events in the area are indicated by Rb-Sr and K-Ar ages in the range 0.5 to 1.5 b.y. The data, however, are too few for evaluation, and these ages may represent survival values via updating through Tertiary activity or one or more events in Late Precambrian or Cambrian time. REFERENCES Burwash, R. A., Baadsgaard, H., and Peterman, Z. E., 1962, Precambrian K-Ar dates from western Canada sedimentary basin: Jour. Geophys. Research, v. 67, no. 4, p. 1617-1625. Gill, J. E., 1949, Natural divisions of the Canadian shield: Royal Soc. Canada Trans., sec. 4, v. 43, p. 61-69. Goldich, S. S., Nier, A. O., Baadsgaard, H., Hoffman, J. H., and Krueger, H. W., 1961, The Precambrian geology and geochronology of Minnesota: Minnesota Geol. Survey, Bull. 41, 193 p. Hart, S. R., and Dodd, R. T., Jr., 1962, Excess radiogenic argon in pyroxenes: Jour. Geophys. Research, v. 67, no. 7, p. 2998-2999. Innes, M. J. S., 1960, Gravity and isostasy in northern Ontario and Manitoba: Canada Dominion Observatories Pub., v. 21, no. 6, p. 263-338. Lowdon, J. A., Stockwell, C. H., Tipper, H. W., and Wanless, R. H., 1965, Age determinations and geological studies: Canada Geol. Survey Pap. 62-17, 140 p. Moore, J. M., Jr., Hart, S. R., Barnett, C. C., and Hurley, P. M., 1960, Potassium-argon ages in northern Manitoba: Geol. Soc. America Bull., v. 71, no. 2, p. 225-230. Stockwell, C. H., 1961, Structural provinces, orogenies, and time classification of rocks of the Canadian Precambrian Shield: Canada Geol. Survey Pap. 61-17, p. 108-118. Wilson, H. D. B., and Brisbin, W. C., 1961, Regional structure of the Thompson-Moak Lake nickel belt: Canadian Mining and Metall. Bull., no. 595, p. 815-822. Wilson, J. T., 1949, The origin of continents and Precambrian history: Royal Soc. Canada Trans., sec. 4, v. 43, p. 157-184. Wilson, M. E., 1941, Pre-Cambrian; in Geology, 1888-1938 (Fiftieth anniversary volume); Geol. Soc. America, p. 269-305.Article 142 ISOTOPIC AGES OF GLAUCOPHANE SCHISTS FROM THE AREA OF CAZADERO, CALIFORNIA By D. E. LEE,1 H. H. THOMAS,2 R. F. MARVIN,2 and R. G. COLEMAN,1 'Menlo Park, Calif.,2 Washington, D.C. Abstract.—Glaucophane-bearing tectonic blocks are strati-graphically out of place with respect to the glaucophane-bearing bedrock terrain on which they rest. Five isotope age determinations ranging from 130 to 150 m.y. indicate that both the tectonic blocks and the bedrock schists were recrystallized as part of a Late Jurassic and Early Cretaceous metamorphic event. Field study has shown that muscovite-bearing glaucophane schists are present in two different structural positions within the Franciscan Formation in the area of Cazadero, Sonoma County, Calif., north of the San Francisco area (fig. 142.1) (Coleman and Lee, 1963). Glaucophane-bearing tectonic blocks are strati-graphically out of place with respect to the glaucophane-bearing bedrock terrain on which they rest. Laboratory study shows also that the tectonic blocks can be distinguished inineralogically from the bedrock terrain, even though glaucophane and muscovite are present in both (Coleman and Lee, 1963; Lee and others, 1963). This article presents isotope-age data determined for muscovites from three samples of tectonic blocks (type-IV rocks of Coleman and Lee) and two samples of bedrock schist (type-III rocks). Franciscan rocks designated types III and IV are described by Coleman and Lee (1963) as follows: Bedrock glaucophane schists [type III] as exemplified by a conformable sequence of metabasalts and metasediments exposed at the bottom of Ward Creek canyon near Cazadero, California. Some of the minerals present along with glaucophane and muscovite in these rocks are aragonite (Coleman and Lee, 1962), lawsonite, and almandine-spessartine-grossular garnet. ART. 142 IN U.S. GEOL. SURVEY PRO Coarsely crystalline gneissic schists [type IV] that occur as large isolated blocks; these blocks cannot be mapped as in place or as stratigraphically equivalent to the bedrock terrain on which they rest. Some of the minerals present along with glaucophane and muscovite in these rocks are epidote, rutile, and almandine-grossular garnet. The mineral assemblages present in these tectonic blocks indicate that they were recrystallized at higher pressure-temperature conditions than were the bedrock schists. Near Cazadero these tectonic blocks are concentrated in a band that conforms to some of the major faulting in the area. The K-Ar and Rb-Sr ages of muscovite from glaucophane schists near Cazadero (fig. 142.1) are listed in the accompanying table. Potassium was determined by a flame photometer using lithium as an internal standard. Rubidium and strontium were determined by isotope-dilution techniques described by Goldich and others (1961, p. 8-35). The overall analytical error is approximately ±5 percent of the quoted age value. Although type-III glaucophane schists may be distinguished from those of type IV, both on the basis of mineralogy and on the basis of field occurrence, the isotope-age results listed in the table indicate that both type-III and type-IV rocks were recrystallized as part of a metamorphic event that took place during Late Jurassic and Early Cretaceous time (Holmes, 1960). Pending further isotope-age work on these rocks, it appears that some of the glaucophane schists in this area represent the metamorphic equivalents of volcanic and sedimentary rocks that were laid down during pre-Cretaceous time. l. PAPER 475-0, PAGES, D105-D107. 1964. D1051 MILE EXPLANATION Geology by E. H. Bailey, M. C. Blake, J. 0. Berkland, J. W. Stewart, and C. Wentworth Landslide area Graywacke, shale, and conglomerate Sandstone Greenstone Serpentine % Gneissic glaucophane schist (tectonic blocks, type IV) Bedrock metamorphosed to O Jd glaucophane schist facies (includes type III) 3.5 (second figure) 1st quarter 2nd quarter Kilauea area ® (30 km depth) 42,4 36,3 Kalapana Trail © area 17,0 26,1 Kaoiki fault © system 52,3 Depth (km) Magnitude 2.0-3.5 >3.5 <10 • 0 10-20 + + 20-60 X X Figure 144.2.—Epicenter plot of earthquakes of magnitude 2.0 and greater for the first quarter of 1962 (top) and second quarter (bottom). The large number of quakes located 30 km beneath Kilauea and those located along the Kalapana Trail and Kaoiki fault system were counted and listed numerically. Epicenters which were located off the map are listed in the accompanying table. Geographic names are shown on figure 144.l.sKOYANAGI D115 Location of concentrations of quakes Numbers of quakes of magnitude 2.0 to 3.5 (first figure) and >3.5 (second figure) 3rd quarter 4th quarter Kilauea area ® (30 km depth) 23,2 20,1 Kalapana Trail (2) area 25,0 19,2 Kaoiki fault (3) system 26,0 15,0 Upper east ® rift zone 6,0 Depth (km) Magnitude 2.0-3.5 >3.5 <10 • 0 10-20 + + 20-60 X X Figure 144.3.—Epicenter plot of earthquakes of magnitude 2.0 and greater for the third quarter of 1962 (top) and fourth quarter (bottom). The large number of quakes located 30 km beneath Kilauea and those located along the Kalapana Trail, Kaoiki fault system, and upper east rift zone were counted and listed numerically. Epicenters which were located off the map are listed in the accompanying table. Geographic names are shown on figure 144.1.D116 GEOPHYSICS 336 186 NUMBER OF QUAKES FOR 1962 25 5 131 12 3 4 12 3 4 12 3 4 —rzaz^Esa 12 3 4 —'ey a—myy 12 3 4 < Ld < < < < -J 3 z < z < < J < 3 O D W I < < -i < * 0 3 * * I Active Dormant or extinct 1 2 3 4 Quarter ui °f the cr year 0 1 0) EXPLANATION □ 20-60 km deep (deep) 10-20 km deep (intermediate) 0-10 km deep (shallow) VOLCANOES Figure 144.4.—Bar graph showing earthquakes of magnitude 2.0 or greater on, and offshore from, the island of Hawaii in 1962. VOLCANOES AND SEISMICITY Seismic activity on the island of Hawaii during 1962 was greatest beneath Kilauea Volcano (fig. 144.4), although the number of quakes in the Mauna Loa region was also high. Seventy-five percent of all quakes were from the Mauna Loa and Kilauea systems. The large number of quakes originating from a highly seismic zone at a depth of 30 km beneath Kilauea caldera was responsible for the high count of deep quakes from the Kilauea system. However, 66 percent of the quakes from Mauna Loa and Kilauea were of shallow origin. Many shallow quakes occurred north of the Honuapo-Kaoiki fault system and therefore have been assigned to the Mauna Loa system (fig. 144.4). Such quakes, located near the surface trace separating two volcanic systems, have been assigned somewhat arbitrarily to one system or the other. The dormant or extinct volcanoes Hualalai, Mauna Kea, and Kohala accounted for few earthquakes in comparison with the active volcanoes Kilauea and Mauna Loa. A relatively high proportion of offshore quakes were assigned to intermediate-depth sources. Seismograms of earthquakes more than 50 km from the nearest station yield little information on focal depth; there- fore, “normal” depth travel-time curves are used to locate them. Relative to previous years, there was in 1962 a marked decrease in the number of quakes deeper than 30 km. This fact is not obvious on the maps and bar graph but it marks an important difference between 1962 and recent years since 1956. Those years were characterized by almost quarterly swarms of thousands of quakes beneath Kilauea and Mauna Loa from depths between 45 and 60 km. Such deeper events prior to 1962 (and their sudden cessation) are being studied further. REFERENCES Eaton, J. P., 1962, Crustal structures and volcanism in Hawaii, in Macdonald, G. A., and Kuno, Hisashi, eds., Crust of the Pacific Basin: Am. Geophys. Union Geophys. Mon. 6, p. 13-29. Koyanagi, R. Y., Krivoy, H. L., and Okamura, A. T., 1963, Hawaiian Volcano Observatory summary: U.S. Geol. Survey Hawaiian Volcano Observatory Summary 25 (Jan., Feb., and March 1962). [In press] Krivoy, H. L., Koyanagi, R. Y., Okamura, A., and Kojima, G., 1963, Hawaiian Volcano Observatory summary: U.S. Geol. Survey Hawaiian Volcano Observatory Summary 28 (Oct., Nov., and Dec. 1962). [In press] Macdonald, G. A., 1956, The structure of Hawaiian volcanoes: Gravenhage, K. Nederlandsch Geol.-Mijn. Genootschap Verh. Geol. Ser., Deel 16, p. 274-295.KOYANAGI D117 Okamura, A. T., Koyanagi, R. Y., and Krivoy, H. L., 1963, Hawaiian Volcano Observatory summary: U.S. Geol. Survey Hawaiian Volcano Observatory Summary 26 (April, May, and June 1962). [In press] Okamura, A. T., Kojima, G., and Yamamoto, A., 1963, Hawaiian Volcano Observatory summary: U.S. Geol. Survey Ha- waiian Volcano Observatory Summary 27 (July, Aug., and Sept. 1962). [In press] Richter, D. IL, Ault, W. U., Eaton, J. P., and Moore, J. G., 1964, The 1961 eruption of Klim* a Volcano, Hawaii: U.S. Geol. Survey Prof. Paper 474-D. [In press]Article 145 SEISMIC INVESTIGATIONS ON CAPE COD, MASSACHUSETTS By R. N. OLDALE and C. R. TUTTLE, Boston, Mass. Work done in cooperation with the Massachusetts Department of Public Works and U.S. Bureau of Public Roads Abstract.—Seismic studies on Cape Cod show that sedimentary deposits of post-Paleozoic age, ranging from 250 to possibly more than 960 feet in thickness, overlie crystalline basement rocks. A trough in the basement surface extending to about 900 feet below sea level was found on outer Cape Cod near Truro. Thirteen seismic-refraction traverses made on Cape Cod between 1958 and 1963 form the basis for interpretations presented here. Additional traverses in the Harwich and Dennis areas have been reported previously (Oldale and others, 1962). Three of the 13 traverses reported were made in the Falmouth area in October 1958 to determine the nature of the surficial deposits and the depth to bedrock along the proposed relocation of Massachusetts Route 28. The 10 other traverses, made for a similar purpose, were along the outer arm of Cape Cod between Orleans Beach and Race Point (fig. 145.1). Nine of these traverses were made in 1962 (Oldale and Tuttle, 1962); the other, an unreversed traverse (traverse 3, fig. 145.1), was made in 1960 at Pilgrim Spring State Park in Truro at a deep borehole drilled by the Woods Hole Oceanographic Institution (Zeigler and others, 1960). All seismic traverses were of the inline refraction type. Traverses before 1962 were made with a 12-channel portable refraction amplifier and oscillograph. The 1962 traverses were made using the same equipment coupled in “series” with a 12-channel refraction amplifier and oscillograph designed to record very low frequencies. The original equipment was used to record the first set of 12 geophones, and the very low frequency equipment recorded a second set of 12 geophones. Traverses in the Falmouth area were 1,100 feet long and those made in 1962 on outer Cape Cod between Orleans Beach and Race Point were 1,800 to 2,100 feet long. The unreversed traverse made at Pilgrim Spring State Park in 1960 was 1,555 feet long. Thickness of layers and the depth to bedrock were calculated for each traverse, using recorded velocities and critical distances. OUTER CAPE COD The upper two seismic layers (A, and L2) recorded in this area probably represent unconsolidated deposits of Pleistocene or, in places, possibly of Tertiary age. A third seismic layer (L3) recorded at traverse 4 may represent semiconsolidated or consolidated deposits of Cretaceous(?) age. The uppermost layer (L,), having velocities of 1,300 to 3,800 feet per second, was recorded at traverses 4, 7, and 8 (see accompanying table). This layer, which is thought to represent the unsaturated Pleistocene glacio- Thickness of the seismic layers, and depth, altitude, and average seismic velocity of the basement rocks at 13 traverses on Cape Cod [Minimum thickness, depth, and altitude indicated where traverse was not long enough to detect underlying material] Altitude of Average Thickness of layers (feet) Computed basement, seismic Traverse depth to mean velocity of (fig. 145.1) basement sea level basement (feet) datum (feet / ii D L, (feet) second) Outer Cape Cod , 510 510 -500 21,250 2 510 510 -500 21,250 3 >620 440 >620 >-580 4 60 >460 >960 390 >-910 -380 5 390 14, 750 6 400 400 -370 19, 500 7 30 470 500 -450 19,200 8. 50 480 530 -480 23. 500 9_ 470 470 -430 14,100 10 400 400 -390 20, 500 Falmouth Area 80 200 280 -150 15, 450 70 180 250 -150 17, 750 80 170 250 -160 12,150 D118 ART. 145 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D-118-D122. 1964.OLDALE AND TUTTLE D119 70° 30' 70° 15' 70° 00' 42° 00' 41 °45 Figure 145.1.—Map of Cape Cod showing the location of seismic traverses described. Symbol shows approximate azimuth, but not length, and number of seismic traverse. fluvial deposits above the water table, ranges from 30 to 60 feet in thickness. A second layer (L2) was recorded at all traverses and has average velocities in the range from 4,700 to 5,700 fps. The L2 layer presumably represents glaciofluvial sand and gravel of Pleistocene age underlain by sand, silt, and clay of Pleistocene and, in some places, of Tertiary age (Woodworth and Wiggles-worth, 1934, and Zeigler and others, 1960). At traverse 3 the L2 layer is made up of 2 distinct acoustic layers with velocities of 4,600 and 5,900 fps. This traverse was made at a deep borehole drilled by the Woods Hole Oceanographic Institution at Pilgrim Spring State Park that found Eocene deposits at a depth of only 86 feet, overlain by materials of Pleistocene age. However, the contact between the acoustic layers was at 175 feet, indicating no correlation between the acoustic boundaries and geologic-time boundaries at this site. Two seismic traverses (1 and 2, fig. 145.1) near Province town in the vicinity of a second borehole made by the Woods Hole Oceanographic Institution indicated average velocities of 5,250 and 5,500 fps, respectively,D120 GEOPHYSICS for the layer. The borehole penetrated 193 feet of glaciofluvial material above 10 feet of sand and silt of Eocene age (Zeigler and others, 1960). The seismic data gave no suggestion of this contact, indicating that materials of Pleistocene and Eocene age have similar acoustic characteristics. Therefore, it is impossible to tell from seismic data alone whether unconsolidated deposits of Tertiary age occur beneath the Pleistocene glaciofluvial deposits in this part of Cape Cod. The L2 layer ranges in thickness from 390 feet at traverse 5 to more than 620 feet at traverse 3, where the traverse was not long enough to detect the velocity and depth of the underlying material. The absence of velocities in the 6,000 to 9,000 fps range indicates that there is little or no compact till beneath the glaciofluvial deposits in this part of Cape Cod; compact till commonly has velocities between 6,000 and 9,000 fps in other parts of Massachusetts (Tuttle, 1961). Near Harwich, Mass., material having an average seismic velocity of 10,500 fps and underlying 313 feet of stratified drift was identified in a borehole as compact till (Koteff and Cotton, 1962). In seismic studies in the Harwich and Dennis areas, material having velocities of 6,000 to 11,000 fps and over-lain by 82 to 357 feet of stratified drift has been interpreted to be compact till (Oldale and others, 1962). A third seismic layer (i3) recognized at traverse 4 is thought to represent semi consolidated or consolidated deposits overlying crystalline basement rocks. The average velocity computed for the material below the L2 layer at this traverse is 12,100 fps. This velocity is about 7,000 fps slower than the average of the velocities of crystalline basement rocks in the area and more than 2,000 fps slower than the lowest basement velocity recorded in this area by the authors. Similar velocities have been recorded on Block Island, R.I. (Tuttle and others, 1961), and from offshore seismic-refraction studies in Cape Cod Bay (Hoskins and Knott, 1961), in Nantucket Sound south of Woods Hole, Mass. (Ewing and others, 1950), in the Gulf of Maine (Drake and others, 1954), and south of Nova Scotia (Officer and Ewing, 1954). In these investigations, velocities of 9,750 to 13,000 fps were interpreted to represent semiconsolidated or consolidated rock of Cretaceous age or rocks of Triassic age similar to the sediments found in the Connecticut Valley. These interpretations were based on other seismic measurements from areas where these lithologic units crop out or where seismic velocities and lithology can be correlated on the basis of borehole data. If the velocity recorded at traverse 4 for the L3 seismic layer actually represents a semiconsolidated or consolidated layer, the layer is more than 450 feet thick and overlies crystalline basement rocks at an altitude on the order of 900 feet below sea level. The L3 layer occurs in what appears to be an eastward extension of a trough found in Cape Cod Bay by Hoskins and Knott (1961), the bottom of which is more than 700 feet below sea level. This trough is in part filled with material having velocities of 10,000 to 12,500 fps, interpreted by Hoskins and Knott to be a semiconsolidated deposit of Cretaceous age. The crystalline basement in this area presumably consists of Paleozoic or Precambrian rocks similar to those exposed along the shore of Massachusetts Bay from Plymouth northward, because the average seismic velocities of 14,100 to 23,500 fps are comparable to the seismic velocities measured in the outcrop areas of the crystalline rocks. Comparison of variation in velocities with azimuth suggests that in the Wellfleet area the basement is composed of a bedded or foliated rock striking eastward, possibly similar to phyllite identified in a borehole near Harwich, Mass. (Koteff and Cotton, 1962). At traverse 8 the average of the rock velocities measured in an easterly direction was 23,500 fps, and at traverse 9 the average of the rock velocities measured in a northerly direction was 14,100 fps. Similar results were noted near Harwich, where seismic studies in the immediate vicinity of the borehole indicated an average velocity of 23,250 fps in an easterly direction, and a traverse 1.1 miles southeast of the borehole indicated an average velocity of 15,650 fps in a northerly direction (Oldale and others, 1962). As portrayed on figure 145.2 the basement surface is deepest at traverses 3 and 4, where the seismic lines were not long enough to detect basement rocks. Instead, minimum depths were computed on the assumption that the critical distance for the basement surface was equal to the length of the traverse. On this assumption the basement surface at traverse 3 was computed to be more than 580 feet below sea level. At traverse 4, two interpretations for the depth to the basement surface are possible, depending upon whether the 12,100 fps velocity recorded for the L3 layer is interpreted as representing crystalline basement rocks or a semiconsolidated to consolidated rock over-lying the basement. By the latter interpretation and on the assumption that the critical distance for the basement is equal to the total length of traverse 4, the altitude of the basement surface is computed to be 910 feet below sea level. However, if the 12,100-fps velocity actually represents the basement rocks, the surface of the basement in this area would be 440 feet below sea level and the maximum depth would then be at traverse 3. The authors’ preferred interpretation is that the basement surface dips steeply north of traverse 5 and south of traverse 2 to form a valley whose floor isOLDALE AND TUTTLE D121 SEA LEVEL- -200'- -400'- -600'- -800'- -1000'- Unconsolidated Pleistocene and Tertiary deposits Semiconsolidated or consolidated sediments (Cretaceous?) Basement (Paleozoic and Precambrian) rocks Figure 145.2—Generalized cross section of outer Cape Cod from Orleans Beach to Race Point, showing computed average altitude of the basement surface at 8 seismic traverses. Dashed lines indicate alternate interpretations in area of deepest basement rock. more than 900 feet below sea level in the vicinity of traverse 4 (fig. 145.2). The basement in this area is thought to be Paleozoic sedimentary and crystalline rock, possibly the Dedham Granodiorite that crops out along the west shore of Buzzards Bay. Average velocities measured on the basement rocks range from 12,150 to 17,750 fps, and seismic measurements on the bedrock in areas mapped as Dedham Granodiorite by Emerson in New Bedford, Dartmouth, and Westport, Mass. (Emerson, 1917), ranged from 11,900 to 16,600 fps. At traverse 13 the 12,150-fps velocity could represent a semiconsolidated or consolidated deposit of Mesozoic age similar to that recorded at traverse 4 in Truro, but is thought to more likely represent a low-velocity facies of the Dedham Granodiorite similar to those recorded in New Bedford, Dartmouth, and Westport. The high percentage of Dedham Granodiorite stones in the drift of the Buzzards Bay moraine (Mather and others, 1942) also suggests that the bedrock in this area is Dedham Granodiorite. REFERENCES FALMOUTH AREA At 3 traverses on the Buzzards Bay moraine in Falmouth (fig. 145.1), 2 seismic layers (Zi and L2) representing unconsolidated deposits of Pleistocene and possibly Tertiary and Cretaceous age were recorded over crystalline basement rocks. Velocities in the Lx layer ranged from 1,400 to 2,300 fps and are thought to represent an unconsolidated layer composed in part of unsaturated Pleistocene glaciofluvial sand and gravel and in part of unsaturated loose till similar to sandy till found in other areas of Massachusetts. The Li layer ranged in thickness from 70 to 80 feet. Velocities in the L2 layer ranged from 2,900 to 6,600 fps. The lower velocities in this layer probably represent loose sandy till. The intermediate and higher velocities are thought to represent a layer composed for the most part of compact till similar to high-velocity till in other parts of Massachusetts (Tuttle, 1961). On the other hand, it may be a complex layer composed of glaciofluvial deposits and till of Pleistocene age, and unconsolidated marine and lacustrine deposits of Pleistocene, Tertiary, and Cretaceous age. Unconsolidated marine and nonmarine deposits of sand, silt, and clay of Pleistocene, Tertiary, and Cretaceous age crop out 20 miles to the southwest on Marthas Vineyard and 7 miles to the south on Nonamesett Island. Similar deposits of Tertiary and possibly Cretaceous age are exposed 31 miles to the north at Duxbury and Marshfield (Woodworth and Wiggles-worth, 1934; Emerson, 1917). The L2 layer ranges in thickness from 170 feet at traverse 13 to 200 feet at traverse 11. Drake, C. L., Worzel, J. L., and Beckmann, W. C., 1954, Geophysical investigations in the emerged and submerged Atlantic Coastal Plain, Part IX, Gulf of Maine: Geol. Soc. America Bull., v. 65, p. 957-970. Emerson, B. K., 1917, Geology of Massachusetts and Rhode Island: U.S. Geol. Survey Bull. 597. Ewing, W. M., Worzel, J. L., Steenland, N. C., and Press, Frank, 1950, Geophysical investigations in the emerged and submerged Atlantic Coastal Plain, Part V, Woods Hole, New York, and Cape May sections: Geol. Soc. America Bull., v. 61, p. 877-892. Hoskins, Hartley, and Knott, S. T., 1961, Geophysical investigation of Cape Cod Bay, Massachusetts, using the continuous seismic profiler: Jour. Geology, v. 69, p. 330-340. Koteff, Carl, and Cotton, J. E., 1962, Preliminary results of recent deep drilling on Cape Cod, Massachusetts: Science, v. 137, p. 34. Mather, K. F., Goldthwait, R. P., and Thiesmeyer, L. R., 1942, Pleistocene geology of western Cape Cod, Massachusetts: Geol. Soc. America Bull., v. 53, p. 1127-1174. Officer, C. B., and Ewing, M., 1954, Geophysical investigations in the emerged and submerged Atlantic Coastal Plain, Part VII, Continental shelf, continental slope, and continental rise south of Nova Scotia: Geol. Soc. America Bull., v. 65, p. 653-670. Oldale, R. N., and Tuttle, C. R., 1962, Preliminary report on the seismic investigations in the Orleans, Wellfleet, North Truro, and Provincetown quadrangles, Massachusetts: U.S. Geol. Survey open-file report. Oldale, R. N., Tuttle, C. R., and Currier, L. W. 1962, Preliminary report on the seismic investigations in the Harwich and Dennis quadrangles, Massachusetts: U.S. Geol. Survey open-file report. Tuttle, C. R., 1961, Seismic high-speed till in Massachusetts [abs]: Geol. Soc. America Spec. Paper 68, p. 288-289. Tuttle, C. R., Allen, W. B., and Hahn, G. W., 1961, A seismic record of Mesozoic rocks on Block Island, Rhode Island: Art. 240 in U.S. Geol. Survey Prof. Paper 424-C, p. C254-C256.D122 GEOPHYSICS Woodworth, J. B., and Wigglesworth, Edward, 1934, Geography and geology of the region including Cape Cod, the Elizabeth Islands, Nantucket, Marthas Vineyard, No Mans Land and Block Island: Harvard Coll. Mus. Comp. Zoology Mem., v. 52, 322 p. Zeigler, J. M., HofTmeister, W. S., Geise, Graham, and Tasha, Herman, 1960, A discovery of Eocene sediments in the subsurface of Cape Cod, Massachusetts: Science, v. 132, p. 1397-1398.Article 146 GEOLOGIC SETTING OF THE SPAR CITY DISTRICT, SAN JUAN MOUNTAINS, COLORADO By THOMAS A. STEVEN, Denver, Colo. Work done in cooperation with the Colorado Metal Mining Fund Board Abstract.—The district is along the south margin of the Creede caldera, near the intersection of a narrow graben cutting the core of the caldera and a reactivated fault along the margin of the caldera. The exposed veins appear to have been deposited in a near-surface environment and may represent the upper parts of the original vertical range of ore deposition. The Spar City district, a small mineralized area along the south flank of the Creede caldera in the central San Juan Mountains, southwest Colorado (fig. 146.1), has been prospected sporadically since the early 1900’s. Several mine workings were driven in the early days to explore some of the exposed veins in the district, and vein fragments containing galena, sphalerite, barite, manganese oxides, and jaspery to amethystine quartz are strewn over several of the old dumps. Systematic exploration, however, has been discouraged by thick soil and dense forest cover, and by widespread landslides and glacial drift. Recent regional geologic studies in and adjacent to the Creede caldera have outlined the geologic framework of the Spar City district, and although many specific details remain obscure, the additional knowledge should help guide future prospecting and exploration programs. The mineralized area in the Spar City district is localized near the intersection of two differently trending structures that were active during volcanic activity, and the known veins appear to be along minor faults associated with this intersection. Mineralization took place late in the local sequence of volcanic and structural events, and was associated with or later than the last main period of faulting. Exposed veins were deposited in a near-surface environment. The Creede caldera (Steven and Ratt6, 1960) is a subcircular collapse structure 10 to 12 miles in diameter whose margin is clearly marked in the present topography for nearly 300° of arc by the valley of the Rio Grande and its tributaries (fig. 146.1). This caldera is the youngest of at least four separate overlapping cauldron structures that together caused an area at least 25 miles long and 5 to 15 miles wide in the central San Juan Mountains to subside now and again in response to voluminous ash-flow eruptions during middle or late Tertiary time (Steven and Ratte, 1963, and Art. 132). The last period of subsidence of the Creede caldera accompanied eruption of the Snowshoe Mountain Quartz Latite, a crystal-rich ash-flow tuff, largely densely welded, that consists of 40 to 60 percent phenocrysts in a drab gray to brownish matrix. The phenocrysts consist largely of intermediate plagio-clase, with lesser biotite, clinopyroxene, hornblende, sanidine, and quartz. The Snowshoe Mountain Quartz Latite accumulated to a thickness of more than 6,000 feet within the caldera, but has not been recognized outside. Talus and landslide breccias from the walls of the caldera intertongue marginally with the Snow-shoe Mountain and indicate that subsidence and accumulation were concurrent over an extended period of time. Final subsidence of the Creede caldera left a flat-floored basin at or below the level of the present Rio Grande, and talus and rockfall breccias spread widely over the floor. North west-trending tangential grabens, now followed by the Rio Grande both upstream and downstream from the caldera, formed late in the period of subsidence, and some faults in the recurrently active Creede graben extending north-northwest from the 716-626 0—64-----9 ART. 146 IN U.S. GEO I.. SURVEY PROF. PAPER 475-D, PAGES D123-D127. 1964. D123D124 ECONOMIC GEOLOGY 37°45' 0 4 _L 8 MILES J Figure 146.1—Geologic sketch map of the Creede caldera and adjacent areas. Snowshoe Mountain Quartz Latite (Ts) accumulated in core of caldera during subsidence. Sedimentary rocks of the Creede Formation (Tc) filled a structural moat around the domed core of the caldera, and viscous lava flows and breccias of Fisher Quartz Latite (Tf) were erupted locally around the margin of the caldera. Undifferentiated volcanic rocks that accumulated outside the caldera before and during subsidence are shown as blank. Contact lines indicate the general limit of distribution of the formations. Heavy lines are generalized faults, dashed where poorly known or hypothetical. northern margin of the caldera may have been active at the same time. The core of the Creede caldera was resurgently domed (Smith and Bailey, 1962) following final subsidence, and some parts near the center were uplifted more than 4,000 feet above a residual topographic moat left all around the margin. A complex keystone graben formed across the center of the core during doming; displacement in this graben was minor near the northern and southern margins of the caldera, but some blocks near the center are nearly 4,000 feet below equivalent rocks in the adjacent uplifted blocks. Volcanic activity broke out around the margin of the caldera following doming, and viscous flows and domes of Fisher Quartz Latite formed several local accumulations, particularly along the south margin of the caldera (fig. 146.1). Concurrently, stream and lake sediments, volcanic ash, and travertine from mineral springs were deposited elsewhere around the caldera to form the Creede Formation.STEVEN D125 The last major period of faulting known around the Creede caldera followed accumulation of the Fisher Quartz Latite and Creede Formation. Many of the faults that formed earlier during the period of recurrent volcanic eruptions and cauldron subsidence were reactivated at this time. Faults active at this time are economically significant, as the important ore deposits in the Creede district north of the caldera formed concurrently with or shortly following movement, and the active or recently active faults were favored sites for mineralization (Steven and Ratte, 1960). The mineral deposits in the Spar City district were formed at about the same time. Erosion of the Creede Formation has reexposed the margin of the Creede caldera. Although the adjacent slopes have been considerably dissected, the original form of the domed caldera core, as well as other constructional landforms in adjacent areas can still be recognized in many places. In detail, the Spar City district is localized near the intersection of a fault marking the southern extension of the core graben of the Creede caldera and a reactivated fault linking the northwest-trending tangential graben with the caldera margin (figs. 146.1 and 146.2). The exposed faults near the Bird Creek mine and Emma mine (fig. 146.2) all appear to be minor fractures subsidiary to larger faults that seem required under the adjacent landslide and moraine cover. At its southern end, the core graben changes southward from a wide zone of jumbled fault blocks between persistent marginal faults to a single fracture that can be followed closely north of Lime Creek, although it has not been seen in outcrop farther south. The position of the fracture shown in figure 146.2, adjacent to the mineralized area of the Spar City district, is hypothetical and depends strictly on projection from the well-defined trends farther north. The arcuate but generally east-trending fault postulated to separate the mineralized area of the Spar City district from the top of Fisher Mountain to the south (fig. 146.2) also has not been seen in outcrop. A fault south of the Emma mine seems certainly required, however, by the relations of the distinctive coarsely porphyritic quartz latite flow that caps Fisher Mountain. The base of this flow is clearly exposed along the west side of the spur ridge extending north from the summit of Fisher Mountain. The base ranges generally from an elevation of 11,250 feet near the north end of the spur, to 11,400 feet near the south end. The same capping flow comprises all the island of outcrop containing the Emma mine and the Denver Tunnel a short distance northwest across the postulated fault; the flow extends as low as 10,750 feet near the Denver Tunnel, but the base is not exposed. No bedrock is exposed where the hypothesized easttrending fault crosses the ridge extending northwest from the summit of Fisher Mountain, and no evidence either for or against the fault was seen in the soil or float that covers the ridge in this vicinity. Unbroken exposures of the capping flow, short distances north and south of the projected trend, limit the position of the fault to a span about 1,000 feet wide. The east-trending fault required between the Emma mine area and the top of Fisher Mountain projects northwesterly toward the southwest margin of the tangential graben that extends northwestward from the Creede caldera (fig. 146.1). Overlap relations at the southeastern end of the graben clearly indicate that the graben formed before the Creede Formation and Fisher Quartz Latite were deposited, and thus the arcuate east-trending fault in the Spar City district appears to represent post-Fisher reactivation of older trends related to caldera subsidence. The capping quartz latite flow on Fisher Mountain appears to have been the youngest lava flow deposited in this area. To the west, marginal breccias along the flank of this flow intertongue with sediments of the Creede Formation, and to the north a thin wedge of Creede Formation appears to separate the flow from the underlying Snowshoe Mountain Quartz Latite near the Bird Creek mine (fig. 146.2). All these younger rocks are cut by mineralized minor faults, indicating that late faulting and mineralization here as well as in the Creede district north of the caldera postdated accumulation of the Creede and Fisher. The exposed veins must have been deposited very near the surface, as the youngest Tertiary rocks deposited in the area form the host rocks. The common occurrence of barite, manganese oxides, and dense jaspery quartz in the gangues further suggests a near-surface environment of deposition. Deductively it would seem more probable that the present exposures represent the upper rather than the lower parts of the original vertical range of ore deposition, and that somewhat deeper levels may be more favorable for the discovery of ore deposits.D126 ECONOMIC GEOLOGY EXPLANATION f ■> Contact Fault <° Dashed, where approximately Dashed where approximately Strike and dip of compaction located; short dashed where located; dotted where con- foliation or bedding gradational or indefinite cealed; queried Where hypothetical. Bar and ball on downthrown side Figure 146.2—Geologic map of the Spar City district, Colorado. Tertiary rocks: Ts, Snowshoe Mountain Quartz Latite (in core of caldera); Tf, Fisher Quartz Latite; Tc, Creede Formation (in structural moat). Quaternary rocks: Qm, moraine; Qls, landslide debris; Qal, alluvium. Base from U.S. Geological Survey topographic quadrangle: Spar Citv, 1957.STEVEN D127 REFERENCES Smith, R. L., and Bailey, R. A., 1962, Resurgent cauldrons— their relation to granitic ring complexes and large volume rhyolite ash-flow fields [abs.], in International symposium on volcanology, Japan: Internat. Assoc. Volcanology, 1962, p. 67-68. Steven, T. A., and Ratte, J. C., 1960, Relation of mineralization to caldera subsidence in the Creede district, San Juan Mountains, Colorado: Art. 8 in U.S. Geol. Survey Prof. Paper 400-B, p. B14-B17. ------- 1963, Resurgent cauldrons in the Creede area, San Juan Mountains, Colorado [abs.j: Am. Geophys. Union 44th Ann. Mtg., April, 1963, Program, p. 112-113.Article 147 THORIUM AND URANIUM IN MONAZITE FROM SPOKANE COUNTY, WASHINGTON By JOHN W. HOSTERMAN, W. C. OVERSTREET; and J. J. WARR, JR., Beltsville, Md.; and Washington, D.C. Abstract.—Monazite occurs in residual clay derived from granodiorite and related rocks (saprolite) and sedimentary clay deposits of the Latah Formation. In three samples of monazite separates, thorium oxide (Th02) averages 3.6 percent and uranium oxide (U3C>8) averages 0.37 percent. The tenor of the monazite is not high enough to be considered commercial. OCCURRENCE Monazite, an anhydrous phosphate of the cerium earths containing variable amounts of thorium, has been known for a number of years to occur in saprolite of crystalline rocks in Spokane County (fig. 147.1) (Goodspeed and Weymouth, 1928, p. 687-695), but determinations of the amount of thorium and uranium in the monazite have not been reported heretofore. In this article, analyses are given of monazite collected by the senior author in 1962 during investigations of residual and sedimentary clay deposits in Spokane County. Analyses of two samples of monazite from residual clay deposits and one sample from sedimentary clay show that the amount of thorium is below the normal commercial specifications for monazite, but they add to information on the composition and resources of monazite in the Pacific Northwest. Monazite-bearing residual clay deposits occur on quartzose gneiss and granodiorite exposed in the Freeman clay pit, and on sillimanite gneiss reached in an auger hole located on the point of a hill 0.95 mile south of Saltese Flats. The deposits consist chiefly of kaolin-ite and halloysite formed by the weathering of alumina-silicate minerals, variable amounts of unaltered quartz, and minor quantities of resistate accessory minerals (Hosterman, 1960, p. 289-291). The residual deposits are soft compared to the underlying crystalline rocks. They preserve the textures and structures of the original gneiss and granodiorite even though much of the rock has been replaced by minerals formed during weathering. The term “saprolite” has been widely ART. 147 IN U.S. GEOL. SURVEY PROF. D128 used by geologists and soil scientists for such decomposed rocks since it was introduced by Becker (1895, p. 289-290), and it will be used here for these materials. The conclusion that the residual clay deposits are saprolites formed by weathering is supported by the wide distribution of clay deposits in eastern Washington and northern Idaho (Hosterman and others, 1960, fig. 1), and the observations that extensive kaolinitic and halloysitic saprolites are now known to develop during weathering in many parts of the world (Campbell, 1917, p. 67; Fox, 1923, p. 4; 1936, p. 419; Harrassowitz, 1926, p. 320-382; Gordon and others, 1958, p. 137-146). 117°15' Figure 147.1—Index map showing location of monazite deposits, Spokane County, Wash. PAPER 475-D, PAGES D128-D130. 1964.HOSTERMAN, OVERSTREET, AND WARR D129 Three monazite-bearing concentrates were panned from saprolite of the quartzose gneiss, granodiorite, and sillimanite gneiss, but only the concentrates from granodiorite saprolite (sample 160816, accompanying table) and sillimanite gneiss saprolite (sample 160817) contained a sufficient percentage of monazite for analysis. In these two samples monazite made up about 20 percent of the concentrate. Less than 5 percent of the concentrate from the quartzose gneiss saprolite was monazite. Monazite was estimated to make up from less than 0.001 percent to about 0.003 percent of the original rock from which the saprolite was derived. Thorium and uranium in monazite from Spokane County, Wash. [Analyses by J. J. Warr, Jr.] Laboratory sample Source of Location Percent No. monazite ThOi UjOs 160816 Granodioritic saprolite. Clay pit at Freeman; sec. 1, T. 23 N., R. 44 E. Auger hole on point of hill 0.95 mile south of Saltese Flats; sec. 32, T. 25 N., R. 44 E. Sommers clay pit; sec. 35, T. 25 N., R. 44 E. 2.74 0.65 160817 Sillimanite gneiss saprolite. 4.06 .20 160818 Sedimentary clay. 3. 92 .26 The mineralogic composition of the monazite-bearing concentrates varies with the type of rock from which the saprolite developed. Concentrates from saprolite of quartzose gneiss in the clay pit at Freeman were observed by Goodspeed and Weymouth (1928, p. 687-692) and the senior author to contain monazite, zircon, magnetite, ilmenite, cassiterite, rutile, tourmaline, and staurolite relict from the original rock. Concentrates from this saprolite also have accessory anatase, which is probably a secondary mineral formed by weathering. Concentrates from saprolite of granodiorite exposed at the same locality contain monazite (sample 160816), zircon, magnetite, ilmenite, cassiterite, rutile, tourmaline topaz, and lepidolite, but lack staurolite. They also have anatase of probable secondary origin. A concentrate panned from saprolite of sillimanite gneiss sampled with the aid of an auger at a locality west of the Saltese Flats consisted of monazite (sample 160817), zircon, ilmenite, spessartite, tourmaline, sillimanite, and actinolite. The resistate heavy minerals recovered from the saprolites are not abraded. The minerals occur mostly as discrete grains, some of which are euhedral, and as inclusions in quartz, or as intergrowths with quartz or other heavy minerals. Most monazite grains are subhedral to anhedral, and they are up to about 0.02 inch in maximum dimension. About 10 percent of the monazite grains from granodiorite saprolite are euhedral tabular crystals, but only a percent or two of the grains from the saprolite of gneiss are euhedral. The monazite is uniformly dark yellow with no per- ceptible difference in color relatable to kind of source rock. A monazite-bearing clayey sand bed occurs in a sedimentary clay deposit in the Tertiary age Latah Formation at the Sommers clay pit, about 6% miles north of Freeman and 3miles west of Saltese Flats. The sedimentary material is thought to have been derived from local sources inasmuch as the heavy minerals show few if any effects of transportation. Weathered gneiss seems to be the most likely source because the sedimentary rock contains detrital heavy minerals of the same species that occur as accessory minerals in the gneisses and because the detrital monazite in the clay has uranium and thorium contacts similar to the monazite in the sillimanite gneiss saprolite. Detrital monazite (sample 160818) was estimated to make up approximately 0.01 percent by weight of the total sample from the Sommers clay pit. This may not be appreciably greater than the amount of monazite in saprolite of the gneisses if there is a large loss of heavy minerals in panning clay (Theobald, 1957, p. 16-17). COMPOSITION OF THE MONAZITE Monazite for analysis was prepared from the panned concentrates by standard procedures of heavy liquids and magnetic separation, and the final monazite separate from each sample was handpicked using a miniature vacuum cleaner with a small orifice under a binocular microscope. J. J. Warr, Jr., determined the amount of thorium in the monazite (see table) by colorimetric methods slightly modified from the procedure described by Grimaldi and others (1957), and the amount of uranium was determined by the visual fluorimetric methods discussed by Grimaldi and Levine (1954). The analyzed samples of monazite contain an average of 3.6 percent thorium oxide (Th02). This is similar to the average of 3.3 percent Th02 observed for a large number of samples of monazite from placers in northern and central Idaho (Schrader, 1910, p. 188), but less than the average of 4 percent Th02 shown by 34 published analyses of detrital monazite in the area of the Idaho batholith (Staley, 1952, p. 306; Kauffman and Baber, 1956, p. 6). The amount of uranium in the three samples is similar to that found in monazite from central Idaho (Kauffman and Baber, 1956, p. 6). ECONOMIC ASPECTS The samples of monazite were collected primarily as part of an investigation of clays. They are neither large enough nor sufficiently numerous to furnish a basis for an estimate of the amount of monazite in the rocks. However, in the few samples taken, monazite was appreciably less than 0.01 percent of the rock, showing that at the sampled localities the amount ofD130 ECONOMIC GEOLOGY monazite in gneiss, granite, and clay is unacceptably low for mining. Moreover, the most thorium-rich monazite sample contained 4.06 percent Th02, an amount falling near the lower end of the range shown by monazite in sillimanite gneiss and associated granite elsewhere (Overstreet, 1960, table 27.1), and considerably less than the present commercially accepted minimum grade of 6 percent. On the basis of results of analyses of the three samples reported in this article, it is not to be expected that monazite containing significantly higher tenor of Th02 will prove to be of widespread occurrence in this part of Washington. REFERENCES Becker, G. F., 1895, Reconnaissance of the gold fields of the southern Appalachians: U.S. Geol. Survey 16th Ann. Rept., pt. 3, p. 251-319. Campbell, J. M., 1917, Laterite: its origin, structure, and minerals: Mining Mag. [London), v. 17, no. 2, p. 67-77: no. 3, p. 120-128; no. 4, p. 171-179; no. 5, p. 220-229. Fox, C. S., 1923, The bauxite and aluminous laterite occurrences of India: India Geol. Survey Mem., v. 49, pt. 1, p. 1-287. ------ 1936, Buchanan’s laterite of Malabar and Kanara: India Geol. Survey Rees., v. 69, pt. 4, p. 389-422. Goodspeed, G. E., and Weymouth, A. A., 1928, Mineral constituents and origin of a certain kaolin deposit near Spokane, Washington: Am. Ceramic Soc. Jour., v. 11, no. 9, p. 687-695. Gordon, Mackenzie, Jr., Tracey, J. I., Jr., and Ellis, M. W., 1958, Geology of the Arkansas bauxite region: U.S. Geol. Survey Prof. Paper 299, p. 1-268. Grimaldi, F. S., Jenkins, L., and Fletcher, M. H., 1957, Selective precipitation of thorium iodate from a tartaric acid-hydrogen peroxide medium: Anal. Chemistry, v. 29, p. 848-851. Grimaldi, F. S., and Levine, Harry, 1954, The visual fluorimetric determination of uranium in low-grade ores, pt. 6 in Grimaldi, F. S., May, Irving, Fletcher, M. H., and Titcomb, Jane, 1954, Collected papers on methods of analysis for uranium and thorium: U.S. Geol. Survey Bull. 1006, p. 43-48. Harrassowitz, Hermann, 1926, Laterit Material und Versuch erdgeschichtlicher Auswertung: Fortschr. Geol. Palaeont., v. 4, no. 14, p. 253-566. Hosterman, J. W., 1960, Clay deposits in parts of Washington and Idaho, in Clays and clay minerals: Proc. of the 7th Natl. Conf. on Clays and Clay Minerals, p. 285-292. Hosterman, J. W., Scheid, V. E., Allen, V. T., and Sohn, I. G., 1960, Investigations of some clay deposits in Washington and Idaho: U.S. Geol. Survey Bull. 1091, p. 1-147. Kauffman, A. J., Jr., and Baber, K. D., 1956, Potential heavy-mineral-bearing alluvial deposits in the Pacific Northwest: U.S. Bur. Mines Inf. Circ. 7767, 36 p. Overstreet, W. C., 1960, Metamorphic grade and the abundance of ThG2 in monazite: Art. 27 in U.S. Geol. Survey Prof. Paper 400-B, p. B55-B57. Schrader, F. C., 1910, An occurrence of monazite in northern Idaho, in Hayes, C. W., and Lingren, Waldemar, 1910, Contributions to economic geology, 1909: U.S. Geol. Survey Bull. 430, pt. 1, p. 184-191. Staley, W. W., 1952, Monazite in Idaho: The Compass, v. 29, no. 4, p. 303-312. Theobald, P. K., Jr., 1957, The gold pan as a quantitative geologic tool: U.S. Geol. Survey Bull. 1071-A, p. 1-54.Article 148 RICH OIL SHALE FROM NORTHERN ALASKA By IRVIN L. TAILLEUR, Menlo Park, Calif. Abstract.—Samples of oil shale of Jurassic(?) age from the foothills along the north edge of the Brooks Range assay 26-146 gallons of oil per ton of rock. The oil yield is greater than that of the Green River Formation, but owing to insufficient information the economic potential of the oil shale cannot be determined. Samples of an organic shale of Jurassic(?) age, which crops out in the foothills along the north edge of the Brooks Range (fig. 148.1), assay 26-146 gallons of oil per ton of rock, several times more than assays of the minable beds of oil shale in the Green River Formation in the Rocky Mountain region. This Alaskan oil shale may be a significant resource, but more stratigraphic and structural information will be required for its assessment. The oil shale seems to have been utilized for fuel in prehistoric times. Abnormal distribution of float, especially near old trailways and encampments, indicates that the Eskimos who lived in the interior before 1900 collected and transported the rock. It was probably carried for fuel. This is suggested by their identifica-of the material as wood in the times before the arrival of the white man (Stoney, 1900, p. 69). Material inferred to have been oil shale was observed on several early expeditions to northern Alaska, and oil shale was specifically identified along later geologic traverses. Dr. John Simpson, surgeon aboard the H.M.S. Plover wintering at Point Barrow during 1852 to 1854, noted that “there is strewed along the beach a quantity of coal, . . . bituminous enough to make an excellent fire for cooking. It is of the sort called candle-coal, and some of the pieces are sound enough to be carved by the natives into lip ornaments” (Collinson, 1875, p. 125). On his overland trek in 1886 Lt. Howard found a substance on the middle Etivluk River “called wood by the natives; it was hard, brittle, light brown in color, very light in weight and burned readily, giving out quantities of gas” (Stoney, 1900, p. 69). Specimens collected from the mouth of the Kukpuk River in 1904 that Collier (1906, p. 45) described as cannel coal were more likely to have been fragments of oil shale, for such fragments have been found on gravel bars upstream by the present writer. Smith (Smith and Mertie, 1930, p. 282-286) collected and identified oil shale from bedrock along the Kivalina River and from float along the Etivluk River in the 1920’s. The material was composed chiefly of megaspores. He speculated that the rock occurred at the base of the geosynclinal sedimentary deposits of Mesozoic age in northern Alaska and might be the source of the petroleum shows in the region. Geologic mapping of Naval Petroleum Reserve No. 4 (1948-53) indicated that oil shale similar to that from the Green River Formation occurs in the Tiglukpuk Formation (Jurassic) (W. W. Patton, Jr., written communication, 1959) and in rock units that appear to be correlative with the Tiglukpuk. Highly organic shales also occur in the Shublik Formation (Triassic) and in the Lisburne Group locally (Missis- ART. 148 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES DI3I-D133. 1934 D131D132 ECONOMIC GEOLOGY sippian) but lack the massive woody aspect of the Jurassic(?) shale (see accompanying table). The oil shale is known to crop out sporadically from the Ipewik River on the west to the Anaktuvuk River on the east, a distance of more than 300 miles. Owing to its low density and toughness, the shale is distributed widely as float and as outsized clasts in coarse-grained younger rocks. Exposures of oil-shale strata are sparse and incomplete, and most show strong deformation. Thickness and other stratigraphic details, therefore, are difficult to determine. Oil shale apparently is present in at least two distinct stratigraphic successions. One consists of oil shale and chert near the base of the Tiglukpuk Formation, which overlies the Shublik and Siksikpuk (Permian) Formations and the Lisburne Group; the Analyses and descriptions of organic shale samples [Specific gravity determined by author; Fischer analyses by .1. Budinsky] Fischer analyses Stratigraphic unit Laboratory No. and (field No.) Location 1 2 3 4 1 (coordinates) Specific gravity Oil (gal/ ton) Water (gal/ton) Oil (percent) W ater (percent) Gas, plus loss (percent) Ash at 900°C (percent) 1. Jurassic(?) 160 127 (51ATr220) Mid part of Kiligwa River (68°41'50" N., 158°28'00" W.) 1. 22 146 6. 3 53. 7 2. 6 8. 9 30. 0 2. Jurassic(?) 160 134 (50AKt261) East of mid part of Kuna River (68°40'05" N., 157°32'15" W.) 1. 27 144 17. 7 60. 6 7. 4 13. 7 23. 1 3. Jurassic(?) 160 129 (51 AKt67) Mid part of Kiligwa River (68°39'45" N., 158°28T0" W.) 1. 20 52. 6 25. 2 19. 7 10. 7 18. 7 12. 7 4. Jurassic(?) 160 128 (51ATr228) Mid part of Kiligwa River (68°42'25'' N., 158°27'45'' W.) 1. 51 40. 9 12. 9 17. 1 5. 3 8. 0 26. 0 5. Tiglukpuk Formation 160 133 (50AKt237) West of mid part of Ipnavik River (68°41'30" N., 157°17'40" W.) 1. 61 48. 8 14. 9 20. 3 6. 3 6. 1 48. 1 6. Tiglukpuk Formation 160 132 (50ATr60) West of mid part of Etivluk River (68°37'40" N., 156°43'00" W.) 1. 86 26. 6 9. 8 11. 1 4. 1 6. 0 57. 6 7. Shublik Formation 160 130 (51 ATr248) Mid part of Kiligwa River (68°44'50'' N., 158°24'50" W.) 1. 90 24. 7 8. 8 8. 2 3. 7 4. 1 59. 2 8. Lisburne Group 160 131 (50AKU06) East of mid part of Ipnavik River (68°40'05" N., 156°57'40" W.) 2. 4 6. 7 3. 4 2. 8 1. 4 1. 8 58. 3 1. Cutbank. Brownish-black woody shale; tough, compact, incipient very thin plates with conchoidal, resinous transverse fracture; weathers pale brown; thinly interbedded with varicolored chert; within 70 feet of unconformable basal contact of Okpikruak Formation on south flank of anticline. 2. Float along small stream. Dark-yellowish-brown woody shale; tough, compact; resinous, irregular fracture; crenu-lated varicolored chert and Okpikruak Formation nearby. 3. Cutbank. Grayish-black papery to woody shale; platy, weakly exfoliated; 12 feet thick; overlies moderately dipping, faulted, varicolored chert. Oil shale and oil-stained sandstone exposed in adjacent cutbacks. 4. Cutbank. Black organic shale; compact with incipient parting; didl with very fine resinous layers; grayish-yellow bloom; 20-foot thickness exposed, subordinate chert; associated with strongly deformed varicolored chert, shale, sandstone, and very finely stratified limestone. 5. Rubble bank. Grayish-black organic shale; compact, very fine, platy partings, dull luster; associated with chert. Apparently crenulated with underlying Shublik Formation and overlying Fortress Mountain Formation. Float of asphaltum also in rubble. 6. Outcrop. Dark brownish-gray papery shale; calcareous; belemnites; weathers light gray and pale brown; 15 feet of 2-3 inch layers, subordinate thin dark chert beds; overlies crenulated Shublik Formation. 7. Cutbank. Brownish-black papery shale; calcareous; abundant Halobia (pectenoid mollusk) imprints; 2-4-inch beds with interbeds of dark, very fine limestone and black chert. 8. Rubble. Grayish-brownish-black, platy shale; calcareous; associated with thin platy, brittle, very dark gray limestone. 1 From 1:63,360 manuscript maps for Howard Pass quadrangle, Alaskan Topographic Series.TAILLETJR D133 other consists of oil shale and chert lying between ill-defined stratigraphic units of Permian and Jurassic age and the Okpikruak Formation (lowermost Cretaceous).1 Oil-stained sandstone crops out near exposures of oil shale in the second succession on the Kiligwa River. Figure 148.2.—Location of oil-shale samples (x), other exposures of oil shale (o), and outcrops of oil-stained sandstone (s) on the middle Kiligwa River (from rectified twinplex photo GS TAL 18-104L). The oil shale has not been given special attention during current mapping because of the structural complexity and an apparent lack of appreciable thickness. However, samples of organic shale from the Nuka-Etivluk Rivers region were analyzed recently as part of a general compilation of oil-shale data (D. Duncan and V. Swanson, oral communication,-1962). Assayed yields of 26-146 gallons/ton of the Jurassic(?) samples are high in comparison with yields of the Green River Formation, in which reserves are computed in terms of 15 and 25 gallons/ton yields, with a maximum yield of 60-90 gallons/ton (Cushion, 1957, p. 135). Stratigraphic descriptions of the samples, location of outcrops, and Fischer analyses are given in the accompanying table. Additional study of the oil shale will be required to assess its potential. In the author’s experience, shale along the Kiligwa River offers the most promise. Known exposures of oil shale along this river are shown on figure 148.2. REFERENCES Cashion, W. B., 1957, Stratigraphic relations and oil shale of the Green River Formation in the eastern Uinta Basin, in Guidebook to the geology of the Uinta Basin: Intermountain Assoc. Petroleum Geologists, p. 131-135. Collier, A. J., 1906, Geology and coal resources of the Cape Lisburne region, Alaska: U.S. Geol. Survey Bull. 278, 54 p. Collinson, R., 1875, Notes on the state of the ice, and on the indications of open water from Behring Strait to Bellot Strait, along the coasts of Arctic America and Siberia, including the accounts of Anjou and Wrangell, in Royal Geographical Society, A selection of papers on Arctic geography and ethnology: London, John Murray, p. 105-162. Smith, P. S., and Mertie, J. B., Jr., 1930, Geology and mineral resources of northwestern Alaska: U.S. Geol. Survey Bull. 815, 351 p. Stoney, G. M., 1900, Naval explorations in Alaska: Annapolis, Md., U.S. Naval Inst., 105 p. 1 R. A. Scott (written communication, 1964) palynologically examined seven oil-shale samples and found that they contained only planktonic forms (chiefly dino-flagellates and hystrichosphaerids) and presumably reworked pollen of Permian or Early Triassic age. He inferred that the samples were of post-Triassic age and possibly were deposited far enough from shore to exclude contemporaneous pollen.Article 149 DISSIMILARITY BETWEEN SPATIAL AND VELOCITY-WEIGHTED SEDIMENT CONCENTRATIONS Rv H P. GUY and D. B. SIMONS, Fort Collins, Colo. Abstract.—Theoretical and measured differences between spatial and velocity-weighted concentrations show (1) that the spatial concentration is normally the greater and increases as the velocity and concentration gradients increase, (2) that velocity-weighted concentration must be used to compute sediment discharge, and (3) that spatial concentration must be used to compute the pressure or specific weight on the bed. Sediment-concentration data in natural streams are used mostly to determine the amount of solids, either by weight or by volume, moving with streamflow. Such concentration data must be discharge weighted; that is, they must be a mean of velocity-weighted concentrations at many points in the stream cross section. The depth-integrating suspended samplers give a velocity-weighted concentration when a uniform vertical transit rate is used at evenly spaced verticals in the cross section. A discharge-weighted concentration is also obtained by traversing the nappe of flow from a flume at a uniform transit rate with an interception device having a uniform width of slot. The sediment concentration computed from a spatial-collection procedure is defined as the relative quantity of sediment contained in an immobilized prism of water-sediment mixture over a specific area of the channel. The chief distinction between velocity-weighted and spatial concentrations is that one is based on sediment and water discharged through a cross section and the other on sediment and water in motion above an area of streambed at a particular instant. The difference between the two concentration measures (velocity-weighted and spatial) has been understood, to some degree, by a few sedimentologists for several years. Much confusion still exists, however, and the quantitative differences between the measures have seldom been determined even approximately. The purposes of this article are (1) to explain the differences between, and some uses of, these concentrations and (2) to show differences between the two concentrations as determined experimentally by use of plastic pellets transported by water in a small flume. THEORETICAL DIFFERENCES Differences between the velocity-weighted and spatial concentrations can be evaluated theoretically by consideration of the four equations that follow. Equation 1 is the Prandtl-von Karman relation used by Einstein (1950) ^-5.75 log 30.2 yx ~kT’ (i) where Uv= average point velocity at distance y above the streambed; TJ*=shear velocity, VyDS; g= acceleration due to gravity; D=depth of flow; S= slope of the energy grade line; x=& corrective parameter; and Ks =grain roughness. For the distribution of sediment concentration with respect to depth, Cv VD—y a T Ca L y ’D-a}’ (2) where Cv= concentration of particles at distance y above the bed, Ca=concentration of particles having the settling velocity w at distance a from the bed, and z=w/0A0 U*. The concentration, C, obtained by velocity-weighted samples, and the water discharge in the sampled zone, Qm, are used to compute the sediment discharge per unit of time, q„ through the sampled zone q_s=Qm.C. (3) D134 ART. 149 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D134-D137. 1964.GUY AND SIMONS D135 Equation 3 is also equivalent in concept to the sediment discharge, qs, determined by the relation qs= f CM, (4) J a where Cy and Uv are defined by equations 1 and 2. Figure 149.1 shows the approximate variation of velocity, sediment concentration, and sediment discharge with depth. This illustration and the preceding equations show that the mean concentration as weighted with velocity is considerably lower than the spatial concentration. The relatively low velocity-weighted concentration is due to the integration with depth ranging from low concentration and high velocity near the stream surface to high concentration and low velocity near the streambed. The spatial concentration, on the other hand, is the mean of the concentration from top to bottom and thus is considerably greater than the velocity-weighted concentration. Under the rare condition when all the sediment being transported is very fine, the concentration may be uniform from top to bottom. Under this condition, spatial and velocity-weighted concentrations are equal. Flow------------> Water surface Streamflow data show that neither Uv nor Cv are spatially or temporally consistent as defined in the equations; therefore, a sample of sediment or velocity at a point in a stream vertical or cross section cannot define the concentration or velocity distribution in such a vertical or cross section. To overcome this difficulty, the “precise method” was developed. This method, as described by the U.S. Inter-Agency Committee on Water Resources, Report 1 (1940, p. 63) involves collection of a relatively large number of point sediment samples simultaneously with velocity measurements. Sufficient data are collected to construct accurate vertical velocity and sediment distribution curves, the corresponding abscissas of which are multiplied to obtain a sediment-velocity curve. The area under this curve represents the sediment discharge in the vertical. Needless to say, the “precise method” was too laborious for routine investigations of sediment movement in streams. The velocity-weighting technique, commonly called the depth-integration method, was developed to facilitate sediment-data collection and analysis. The technique, as described by the U.S. Inter-Agency Committee on Water Resources, Report 6 (1952, p. 14) is as follows: The discharge of the depth-integrating sampler was predicated on the hypothesis that an integrated sample of the water-sediment mixture existing at the place and time of sampling would be obtained if the filling rate were such that the velocity at the point of intake is equal to the local stream velocity while the sampler is moved at a uniform vertical speed in the stream. The spatial concentration may be used to determine the actual load (weight per unit area) in transport over the streambed. It is also the correct concentration for determining the specific weight of the water-sediment complex. The spatial concentration can be obtained in streamflow by averaging the concentration of several equally spaced point samples in the stream cross section. In the laboratory, Bagnold (1955) used a mechanical device for isolating a rectangular slug of flow in a flume (fig. 149.2). A representative sample of spatial concentration is generally difficult to obtain because of the unsteady motion of particles near the bed. Most point samplers do not operate closer than within 0.3 or 0.4 foot of the bed. The mechanical-isolation device cannot distinguish between stationary and moving particles and thus cannot be used on an alluvial bed. EXPERIMENTAL RESULTS The difference between the velocity-weighted and spatial concentrations was studied by a series of runs in an 8-inch-wide recirculating flume. Plastic pellets with a specific gravity of 1.04 to 1.06 and with a median diameter of about 3 mm were used to represent the sediment grains. Samples were taken with a mechanical-isolation device after the velocity was sufficient to insure that all sediment over the bed was in motion, even though some particles were moving very slowly. The variation of sediment concentration and velocity with depth appeared to approximate that indicated by equations 1 and 2 and as illustrated in figure 149.1. Sediment concentration was sampled by both methods for six different series of runs whereby each series contained a different amount of sediment circulating in the flume. To maintain the same amount of sediment in circulation, an equivalent volume of sample was returned to the flume for each sample removed. Two of these series were sampled at three widely differentD136 SEDIMENTATION velocities. Velocity-weighted samples were taken by collecting in a nylon-mesh basket all particles discharged at the tail end of the flume. Sampling periods ranged from 8 to 15 seconds. Spatial-concentration samples were taken by quickly isolating with a sharp-edged frame having wire mesh ends, the water and sediment moving in a 1-foot length of the flume. This device is similar to the sampler illustrated in figure 149.2. Figure 149.2.—Side view of sampling device for determining the spatial concentration of sediment moving in an open channel (after Bagnold, 1955). The flow was stopped immediately after isolating the sample, and the particles were removed by siphoning from the known volume of water. The concentrations in parts per million by weight for both methods of sampling are summarized in the accompanying table. This shows the concentration mean, the Concentration of velocity-weighted and spatial samples of plastic pellets from 8-inch flume Flume run Mean velocity (ft/sec) Velocity-weighted concentration at tailbox (ppm) Spatial concentration on flume bed (ppm) Mean Standard deviation Limit of probable error 1 Mean Standard deviation Limit of probable errori 1 2. 09 85,000 1,350 2,290 110,600 6,920 6,830 2 1.72 63,200 6, 950 11,700 92,100 10.300 12,200 3... 1.66 41,300 1,900 3, 220 58,700 7,310 7,050 4 1.18 20.900 890 2, 610 40, 300 4,450 7,530 4 1.67 24. 800 1.140 2, 690 29,400 5,140 4.920 4 2 1.67 24,800 1.140 2, 690 21.400 950 1,161 4 2. 46 24,400 750 1,280 26,800 2,500 4,220 5 1.81 10,800 560 940 13,400 2, 760 3. 260 6 1.13 3, 500 130 220 5,760 890 850 6 1.42 3,440 260 440 5, 250 970 1,140 6 2.87 3,890 105 180 3,870 670 640 1 Limit of probable error of measurement at 90-percent confidence level. 2 Spatial samples collected near headbox of flume. standard deviation, and the limit of probable error of measurement at the 90-percent confidence level for each run. The sediment concentration ranges from a velocity-weighted value of 85,000 ppm and a spatial value of 110,600 ppm on the flume bed for run 1 to a velocity-weighted value of 3,440 ppm and a spatial value of 5,250 ppm on the flume bed for run 6. The error of measurement is appreciably less at the tailbox because of the greater time of sampling; hence, only 3 samples were taken at the tailbox for each run, whereas 5 were taken on the flume bed for most runs. The sample concentrations compared in figure 149.3 illustrate the differences between the two methods of determining concentration. As expected, the spatial concentration is greater than the velocity-weighted concentration. The limits of the probable error of concentration measurement (90-percent confidence level) are shown (fig. 149.3) in the form of a rectangle approximately centered over each concentration mean. These rectangles indicate that the variability in the movement of sediment, and consequently the measurement error, is greater for the spatial means than for the velocity-weighted means. Z o I- < ca i- z z o o o _l UJ > 2000 5000 10,000 20,000 50,000 100,000 SPATIAL CONCENTRATION, IN PARTS PER MILLION Figure 149.3.—Relation of velocity-weighted concentration to spatial concentration. The rectangle centered approximately around each concentration shows the limit of probable error of measurement at the 90-percent confidence level. Numbers correspond to flume runs described in table. The larger the variation of velocity and sediment concentration with respect to depth (comparatively high V near the surface and a comparatively large C near the bed) the greater the difference between the two concentrations. As indicated by the highest velocities sampled (runs 4 and 6) the two methods may be comparable when the sediment is thoroughly suspended. In fact, the reverse of the normal result was found by comparing spatial concentrations from near the headbox with the velocity-weighted concentrations at the tailbox. The lower mean concentration of the spatial samples (see table and fig. 149.3) is caused by an inversion of the normal concentration gradient due to centrifugal force on the sediment particles. CONCLUSIONS Velocity and sediment-concentration distributions in a cross section of streamflow can be predicted onlyGUY AND SIMONS D137 within rough limits. Theoretical and measured differences between spatial and velocity-weighted concentrations show that: 1. The dissimilarity between the two measures of con- centration widens as the velocity and concentration gradients increase with respect to depth. 2. Sediment discharge in a stream should be computed from a velocity-weighted concentration obtained by sampling at several points or verticals in the stream cross section. Routine measurements of velocity-weighted concentration are made with several kinds of depth-integrating samplers. 3. Spatial concentration is needed when either the actual load or amount of sediment exerting pressure on the bed or the average specific weight of the sediment-water mixture over the bed is desired. Measurement of spatial concentration is usually difficult because immobile particles cannot be distinguished from moving particles. REFERENCES Bagnold, R. A., 1955, Some flume experiments on large grains but little denser than the transporting fluid, and their implications: Civil Engineers Proc., Excerpt pt. Ill, London, p. 192. Einstein, H. A., 1950, The bed-load function for sediment transport in open channels: U.S. Dept. Agriculture Tech. Bull. 1026, p. 7-17. U.S. Inter-Agency Committee on Water Resources, 1940, Report 1—Field practice and equipment used in sampling suspended sediment: U.S. Army Corps of Engineers, Iowa City, Iowa. U.S. Inter-Agency Committee on Water Resources, 1952, Report 6—The design of improved types of suspended sediment samplers: U.S. Army Corps of Engineers, St. Paul, Minn.Article 150 TEMPORARY STORAGE OF FINE SEDIMENT IN ISLANDS AND POINT BARS OF ALLUVIAL CHANNELS OF THE RIO GRANDE, NEW MEXICO AND TEXAS By CARL F. NORDIN, JR., and JOSEPH P. BEVERAGE, Fort Collins, Colo., and Albuquerque, N. Mex. Abstract.—Islands and point bars are semipermanent features of alluvial channels. They contain as much as 20 percent sur-ficial material finer than 0.062 mm which is deposited during receding flows. Some of the scatter generally noted in discharge-transport relations probably derives from the transient storage of the fine material, which is flushed during higher discharges. The fine material transported by a stream is obtained from the soil of the watershed, from old flood plains by bank cutting and sloughing, and from deposits within the active channel of the stream. Deposits in the channel include islands, point bars, and other slack-water deposits. Although islands and point bars may be semipermanent channel features, they are periodically inundated and reworked in streams that have fairly wide variations in flow. The Rio Grande in New Mexico is such a stream. Figure 150.1 shows the particle-size distribution of channel bed material and bar material from the Rio Grande near Bernalillo, N. Mex., on June 1, 1962, at a sampling location designated section F. At the time of sampling, the bar projected 0.1-0.4 feet above the water surface and occupied about half of the high-flow channel width. Samples from the bar contained about 9 percent more fine material (diameter <0.062 mm) than samples from the wetted perimeter. Figure 150.2 shows similar curves for the Rio Grande, downstream, near Anthony, Tex., on April 16, 1962. Bar material at this site contained about 17 percent more fine material (diameter <0.062 mm) than channel material. These samples were collected from bars or islands formed in areas of decreased velocity during receding flows. At both sampling sites the bars occupied a third to a half of the bankfull channel width. Islands and point bars are generally considered to be deposits of coarser material, which is transported in appreciable quantities only at high flows. The samples Figure 150.1.—Particle-size analyses of bar and channel material from the Rio Grande near Bernalillo, N. Mex., on June 1, 1962. from the Rio Grande near Bernalillo, N. Mex., and Anthony, Tex., indicate, however, that the bars and islands of the river may include from 10 to 20 percent fine material. Wolman and Leopold (1957, p. 95) reported three point-bar samples from Watts Branch near Rockville, Md., that averaged 20 percent fine material. They concluded that part of the material eroded from the drainage basin “is stored temporarily in point bars and in the flood plain at various places in the channel system.” D138 ART. 150 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D138-D140. 1964.NORDIN AND BEVERAGE D139 Figure 150.2.—Particle-size analyses of bar and channel material from the Rio Grande near Anthony, Tex., on April 16, 1962. The fine material in islands and point bars of the Rio Grande is generally flushed out by high flows, such as spring snowmelt and sustained reservoir releases. The accompanying table lists discharge and concentration of various size classes of suspended sediment for the Rio Grande near Bernalillo during two spring runoff periods. In both runoff periods, the concentration of fine material decreased with time regardless of wide variations in discharge. Discharge and concentration of suspended sediment, by size class, in the Rio Grande near Bernalillo, N. Mex., during the spring runoff of 1952 and 1958 Date Discharge (cfs) Concentration, by size class (ppm) <0.062 mm 0.062-0.125 mm 0.125-0.250 mm 1952: April 25 2, 910 1, 270 959 423 May 12 6, 390 1, 160 646 753 June 17._ 6, 100 375 390 555 June 20__ 4, 720 321 336 555 June 26. 2, 850 222 264 273 1958: May 8 _ ... 6, 730 2, 020 1, 080 1, 360 May 13.. 8, 310 1, 650 1, 270 1, 060 May 21 . . 8, 700 1, 210 904 1, 070 May 27 9, 970 1, 040 796 1, 040 June 4 7, 790 750 575 675 June 10. — . 5, 450 621 476 724 June 13 . _ . . 4, 380 318 418 856 10 Figure 150.3.—Particle-size analyses of bed material from the Rio Grande near Bernalillo, N. Mex., May 8 and June 13, 1958. The apparent lack of relation between discharge and concentration of the sand sizes (>0.062 mm) suggests that these fractions, like the fine material, behave as “wash load.” Wash load is that part of suspended sediment “which is washed through the stream channel without any deposition” at a rate dependent only on its availability from the watershed (Einstein and Chien, 1953, p. 31). Although changes in availability of material from the watershed may explain some of the lack of relation for the sand fractions, changes in the size distribution of the bed material should be considered as well. Figure 150.3 indicates that the diameter of the bed material can increase markedly during several weeks of sustained high flow. As the size of bed material increases, the amount of the finer sand fractions available for suspension and transport by the stream decreases. During receding and low flows, clayey silt and fine sand are stored on islands and point bars and, to some extent, within the active channel (fig. 150.3). During rising stages and during sustained high flow, the bars and islands serve as a source for fine material. Throughout sustained high flows, the concentrations of the fine material and of the fine sand fractions seem to vary as a function of time rather than of discharge (see table). Both the magnitude of the concentration and the rate of change in concentration with time probably 716-62G O—64D140 SEDIMENTATION depend partly upon antecedent conditions, that is, upon how much fine material was stored during previous low flows. Storage time of fine material in bars and islands of the Rio Grande ranges from several months to a year and depends on subsequent climatic and hydrologic conditions, which determine when the channel features again become part of the wetted perimeter. When the stage rises, the stream begins to erode the deposits and to resuspend the fine material. This study indicates that some of the apparently random scatter in discharge-transport relations can be explained in terms of the time dependency of the concentration of the finer sand classes and of material finer than 0.062 mm. This time dependency is due, in part, to temporary storage within the active channel. REFERENCES Einstein, H. A., and Chien, N., 1953, Transport of sediment mixtures with large ranges of grain sizes: California Univ. Inst. Eng. Research, Missouri River Div. Sediment Ser. 2 Wolman, M. G., and Leopold, L. B., 1957, River flood plains— some observations on their formation: TJ.S. Geol. Survey Prof. Paper 282-C, p. 87-109.Article 151 FLUVIAL SEDIMENTATION IN MAMMOTH CAVE, KENTUCKY By CHARLES R. COLLIER and RUSSELL F FLINT, Columbus, Ohio Work done in cooperation with the National Park Service Abstract.—Alternating deposition and erosion occur in Mammoth Cave by flooding from the Green River. In 2% years, numerous low floods deposited 0.5 foot of sediment in the lowest levels. Three high floods removed that sediment, but caused thinner deposits at higher levels. The coarsest sediment is deposited in the lower parts of the cave. The sources of sediment and rate at which it is being deposited in Mammoth Cave, Ky., have long been of interest to geologists and speleologists. Detailed measurements by the U.S. Geological Survey of erosion and deposition between October 1959 and June 1962 indicate that sedimentation in the cave is closely related to flooding of the nearby Green River. The Green River, which is hydraulically connected to Mammoth Cave by Echo River spring and River Styx spring, is the chief source of sediment and floodwater to the cave. These springs are submerged during floods of the Green River, and water then enters the cave through them so that the water level in the cave corresponds closely to the level of the Green River (Hendrickson, G. E., 1961). W. E. Davies and E. C. T. Chao (written communication, 1959) found in their studies of Mammoth Cave that the sediments in the lower levels of the cave are similar in physical character and mineralogy to the sediments on the flood plain of the Green River. The turbulence and velocity of the floodwater entering the cave are sufficient to transport sediment into the cave, particularly during rapidly rising water levels. These same forces may cause a flushing of sediment from the cave during rapidly falling water levels. Thirteen lines were surveyed across cave passageways that are subject to flooding, and the elevations above gage datum of the sediment deposits on these lines were determined. These lines, or ranges, served as bases from which changes in deposition were determined by ART. 151 IN U.S. GEOL. SURVEY PROF later surveys. The approximate locations of the ranges are shown in figure 151.1, and the elevations of the cave floor at each range are shown in the accompanying table. Resurveys of the ranges in August 1961 and in January and June 1962 indicated that the greatest changes in deposition occurred at the lowest elevations (see table). Alternating deposition and erosion were observed at some ranges. Movement and redeposition of sediments were shown by recovery of ribbons placed vertically in the deposits of sediment. The lower and more frequent floods apparently caused deposition in the boat-ride section of Echo River. The higher and less frequent floods, however, tend to remove these deposits from the Echo River channel but cause deposition at the highest flooded and intervening levels. From October 1959 to January 1962, as much as 0.4 foot of sediment was deposited in the boat-ride section of Echo River (ranges 1, 2, 3, table). In the higher passages, at elevations of 5 to 25 feet above the gage datum, only small changes in deposition or erosion were observed. No floods exceeded 31.4 feet during this period, although 2 floods exceeded 21.5 feet and 8 exceeded 17.7 feet. Three floods exceeded 21.5 feet in January, February, and early March 1962. One of these floods reached 57.1 feet and is the highest flood of record. All three floods generally caused deposition in the passages above the 5-foot elevation, but they removed about 0.5 foot of sediment from the boat-ride section of Echo River. The particle size of the deposited sediment is somewhat related to the elevation of the deposit. Deposits at elevations below 16 feet are predominantly sand (0.062 to 2.0 mm diameter), and the deposits at higher elevations are predominantly silt (0.004 to 0.062 mm diameter). This correlation is shown by the grouping of points in figure 151.2, a plot of the median particle PAPER 475-D, PAGES D141-D143. 1964. D141OOOMCiCn^COtO D142 SEDIMENTATION Flooding and changes in deposition at sedimentation ranges in Mammoth Cave, October 1959 to June 1962 Range No. 10. 11. 12 13 Elevation of cave floor (ft) Oct. 1959 to Aug. 1961 Aug. 1961 to Jan. 1962 Jan. 1962 to June 1962 Oct. 1959 to June 1962 Low Dct. 1959 High Mean Times flooded2 Percentage of days flooded2 Net change (ft) Times flooded2 Percentage of days flooded2 Net change (ft) Times flooded2 Percentage of days flooded2 Net change (ft) Times flooded2 Percentage of days flooded2 Net change (ft) -1. 3 + 5. 0 + 0. 3 100 + 0. 48 100 -0. 03 100 -0. 55 100 -0. 10 -4. 9 -. 3 -2. 1 100 + . 20 100 100 3 -. 45 100 -. 25 - 3. 2 + 1. 7 -1. 6 100 + . 34 100 100 3 -. 46 100 -. 12 -8. 0 + 1. 5 -5. 4 100 . 00 100 100 3 + . 04 100 + . 04 5. 2 18. 1 9. 8 18 34 + . 22 2 17 -. 15 6 58 -. 24 26 36 + .31 13. 4 18. 2 15. 3 19 12 -. 11 2 7 + . 08 6 20 -. 18 27 12 + . 15 11. 2 14. 9 11. 9 18 15 + . 01 2 7 + . 14 6 23 + . 42 26 15 + . 57 3. 5 11. 4 6. 6 15 77 -. 10 6 29 -. 13 100 + . 33 21 75 + . 10 3. 4 9. 5 5. 6 15 86 -. 29 6 44 + . 07 100 + . 13 21 83 -. 09 17. 8 24. 6 20. 0 6 4 + . 05 1 2 . 00 3 14 + . 13 10 6 + . 18 17. 7 26. 4 20. 3 7 4 + . 13 1 2 + . 06 3 14 -. 02 11 6 + . 17 21. 5 30. 9 25. 6 2 1 -. 11 0 3 10 + . 28 5 3 + • 17 31. 4 34. 5 32. 2 0 0 1 5 +. 20 1 1 + . 20 i Arbitrary gage datum. 2 Includes both complete flooding and partial flooding. 3 August 1961 to June 1962. Figure 151.1.—Map of lower level of Mammoth Cave, showing approximate location of sediment ranges. Pattern indicates area submerged at low water level. Base from National Park Service map, 1935. size against the average elevation of the floor at each range. A noticeable inconsistency in this grouping is the January 1962 sampling from ranges 7 and 8. The samples from these two ranges included only the surface material, whereas samples at the other ranges were taken at depths of 1 to 4 feet. The coarser particles tend to migrate by successive floods to higher elevations. This conclusion is supported by the position of the June 1962 triangles in figure 151.2 for elevations between 10 and 25 feet. Sediment deposited at these elevations generally had a larger median diameter after the floods of January to March 1962 than before. This increase in median diameter would occur if, during rising water level, there were sufficient turbulence and velocity to <5 < 0 25 iiiii 1 1—i—|—n~r V o 1 | III _ A V O — o = _ AD o V □ O A V An v □ % = 1 1 1 1 1 1 1 1 1 1 1 1 Avo — A xp '&= _J i i i 0.005 0.01 0.05 0.1 MEDIAN PARTICLE-SIZE DIAMETER. IN MILLIMETERS Figure 151.2.—Relation of median particle size of deposited sediment to mean elevation of floor, Mammoth Cave, Ky. Circles, samples taken October 1959; triangles (point up), August 1961; squares, January 1962; triangles (point down), June 1962.COLLIER AND FLINT D143 transport coarser particles through the passages to settle at higher elevations. Hydrographs of the water level in the cave show that during floods the water generally rises more rapidly than it falls. Turbulence and velocities probably are greatest during the more rapidly changing water levels during the rise of a flood. Clay particles are more easily transported than silt or sand and therefore are moved for greater distances by floodwater. After every flood, a thin layer of clay is deposited over all the submerged passageways. The record flood in February-March 1962 left an average deposit of 0.2 foot of clay and silt at range 13, elevation 32.2 feet. The three resurveys of the range lines in Mammoth Cave over a period of less than 3 years provide an indication of sediment movement in the cave. Deposition or erosion during a given flood depends upon (1) the rate at which the water rises, (2) the height of the flood, (3) the duration of the flood, and (4) the rate of fall. The rate of rise or fall and the height of the flood influence the turbulence and velocity of the water in the cave and, therefore, the capacity of the water to transport sediment. The duration of the flood, particularly the period of little change in water level at the flood crest, determines the time available for settling and deposition of the sediment. REFERENCE Hendrickson, G. E., 1961, Sources of water in Styx and Echo Rivers, Mammoth Cave, Kentucky: Art. 308 in U.S. Geol. Survey Prof. Paper 424-D, p. D41-D43.Article 152 QUATERNARY MUDFLOW DEPOSITS NEAR SANTIAGO, CHILE By KENNETH SEGERSTROM,- OCTAVIO CASTILLO U.,1 and EDUARDO FALCON M.,1 Denver, Colo.; Santiago, Chile Work done in cooperation with the Instituto de Invesligacidnes Geoldgicas de Chile under the auspices of the Agency for International Development, U.S. Department of State Abstract.—Mudflows appear to have played an important role in the accumulation of nonsorted, nonstratified Quaternary valley fill in central Chile. Much of the mudflow material in the vicinity of Santiago is pumiceous. A widely accepted hypothesis of glacial origin for these deposits is rejected. Santiago, the capital city of Chile, is in a tectonic depression or graben that forms the Valle Central of Chile (fig. 152.1). The depression is partly filled with unconsolidated sediments of Quaternary age. Large volumes of nonsorted and virtually nonstratified valley-fill sediments at or near Los Cerrillos, Pudahuel, Cerro Apoquindo, and Puente Alto (fig. 152.2), have been described by various workers as moraines (Briig-gen, 1950; Karzulovic, 1958; Munoz Cristi, I960). Brief observations by R. F. Flint in April 1959 of some of the deposits, particularly of those containing large volumes of pumiceous ash, cast considerable doubt on a glacial origin. Later and more detailed examination by the authors confirms the nonglacial origin of these deposits, which are now believed to result from mudflow deposition and creep. The general term “diamicton” is applicable to at least some of these nonsorted noncalcareous terrigenous deposits composed of sand and larger particles in a matrix of mud (Flint and others, 1960a, b). The Valle Central near Santiago is filled to a depth of more than 400 meters with sedimentary materials eroded from the surrounding highlands and transported by running water, mudflows, or as slides. Most of the 1 Instituto de Investigaciones Geologicas. materials are derived from the Cordillera de los Andes to the east; lesser amounts have been eroded from the Cordillera de la Costa to the west. The bedrock in both ranges is mainly igneous, including both volcanic and intrusive types, and is predominantly of Mesozoic age. The oldest valley-fill sediments known in the area are lacustrine silt, clay, and sand at a depth of as much as 250 m in study well 1, near Pudahuel, and 236 m in well E3-26, in Santiago (fig. 152.2). The age of these sediments is unknown, although in well 1, small freshwater gastropods were found in microlaminated clay at a depth of 157 m. The fine-grained lacustine sediments of the valley are overlain by coarse-grained fluvial materials which in well E3-26 extend to a depth of 150 m below the surface. Alluvial-fan deposits of the Rios Lampa, Colina, Mapocho, and Maipo make up the bulk of these materials. Gravels are exposed in terraces along the Rio Maipo and in gravel pits around the fringes of Santiago. Colluvium resulting from sheet erosion occurs in narrow zones bordering the valley, particularly north and west of Santiago. Knowledge of the origin and character of the valley fill is especially important in problems of ground-water supply, which are discussed in Article 169. Diamicton deposits overlie altered bedrock along the mountain front east of Santiago and interfinger with alluvium in the basin. Pumiceous deposits, which contain large blocks of andesite and granodiorite and cover the fluvial and lacustrine deposits west of Santiago, appear to be of similar origin. Other pumiceous deposits AKT. 152 Iff U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D144-D148. 1964. D144SEGERSTROM, CASTILLO U., AND FALCON M. D145 72” 71" 70” Figure 152.1.—Index map of part of central Chile. Quaternary valley-fill deposits outlined by dashed line. which do not contain large blocks of other rock, found in the valley of Estero Papeta and in the basin of the Rio Rapel, southwest of Santiago (fig. 152.1), appear to be lacustrine (Briiggen, 1950, p. 221). Within the area of figure 152.1, unquestioned till is limited to the high Cordillera de los Andes. Moraines are found at minimum altitudes of 1,600 m in one or two valleys (that of the Rfo Colorado, for example), but most till and all glacial cirques occur above 2,000 m at lat 33 °-34 ° S. The best exposed and most accessible moraine in the region extends 12 km along the Rio Yeso, between 1,670 and 2,700 m above sea level, and terminates on the south side of Laguna Negra (fig. 152.1). The till of this moraine has a matrix of gray unweathered rock flour. The included blocks are angular and are of two lithologic types, granodiorite and andesite. The granodiorite blocks are as much as 4 m across; those of andesite are as much as 1 m or a little more. Through much of the moraine the two kinds of boulders are segregated into alternate bands subparallel to the valley trend. The bands are boulder trains derived directly from intrusive and volcanic bedrock farther up the valley. DIAMICTON DEPOSITS The diamicton deposits of the region are of two types: one contains little or no pumice, and the other contains abundant pumice. Diamicton deposits of the first type form a hummock}7 topography along the foot of the steep Andean front at the east edge of the Valle Central between the Rfo Mapocho and the Rfo Maipo (fig. 152.2), in a belt 800 m to 4 km wide. An outstanding feature of this belt is a row of hills that extends 5 km from Loma de los Banos northwest to Cerro Calan. The material that makes up the hillocks is well exposed in a cut between Loma de los Banos and Cerro Apoquindo, where angular to subrounded blocks of granodiorite and andesite as much as 2 m or 3 m in diameter are found in a matrix of weathered clasts ranging from clay size to coarse gravel. Here and in other parts of the deposit are scattered lenses, 1 to 2 m thick and as much as 10 or 15 m long, of pumiceous ash and sand or gravel, generally consisting of subangular grains. A gravel lens on Cerro Calan contains rounded cobbles. Deposits resembling those exposed in the cut between Loma de los Banos and Cerro Apoquindo crop out on the southD146 SEDIMENTATION Figure 152.2.—Map of surficial deposits in the vicinity of Santiago. side of the Rio. Mapocho and between Loma de los Banos and the Rio Maipo. All these deposits are more than 30 m thick, and at Cerro Calan they are approximately 100 m thick. They overlie deeply weathered rock to the east and interfinger with fluvial deposits from the Cordillera de los Andes toward the west. East of Puente Alto, immediately north of where the Rid Maipo emerges from the Cordillera de los Andes, is an enormous fan-shaped diamicton deposit that extends downward and outward from an amphitheater-shaped hollow in the mountain front to the northeast. Angular blocks of granodiorite within and on top of the mass are so large and numerous that some of the diamicton deposits are easily confused with bedrock. Near the foot of the deposit a small lens of pumiceous sand is exposed in a highway cut 2 km north of the junction of the Santiago-El Volcan highway with the road to Puente Alto. The most extensive deposit of the pumiceous diamicton forms a hummocky surface in the lowest part of the valley (Pudahuel-Los Cerrillos), immediately west of the city of Santiago at 450 to 485 m above sea level (fig. 152.2). It is exposed to depths of 10 m or less in a dozen or more large quarries, and study of drill cuttings indicates that the deposit extends to depths of as much as 30 m. The material consists of white or grayish fragments of pumice, predominantly of sand size, but including also silt and gravel. The deposits are poorly sorted and without stratification. Irregularly distributed subangular to subrounded blocks of andesite and granodiorite as large as 80 cm in diameter, some of which are highly decomposed, make up, on an average, about 20 percent of the total volume. Other pumiceous diamicton sediments crop out on the south side of the valley of the Rio Yeso, along the Estero Collanco near its junction with the Rio Maipo and in the Melipilla area (fig. 152.1). On the summit and north flank of a ridge between the Rio Yeso and El Volcan, there are small bodies of sand, locally pumiceous and containing scoriaceous fragments as much as 20SEGERSTROM, CASTILLO U., AND FALCON M. D147 cm long. These deposits fill small depressions and form scattered patches on slopes 300 m to 450 m above the valley floor. Pumiceous float is found in colluvial deposits that consist chiefly of andesitic materials as much as 650 m above the valley. In a stream terrace 30 m above the lower course of the Estero Collanco a diamic-ton deposit is exposed that contains angular to sub-angular blocks as large as 2 m across in a light-gray ashy, partly pumiceous matrix. The blocks are of varied composition and are decomposed. The western slope of the Andes is very steep, dropping from 3,200 to 700 m above sea level in a horizontal distance of 12 km, just east of Santiago. Because of the great relief, mass movement of materials evidently has been a very active process along the Andean front and in the high Andes during Pleistocene and Recent time. A contributing factor is the semiarid climate, which has inhibited the growth of protective forest on the slopes. Mudflows are a common phenomenon on the western slopes of the Andes under present climatic conditions. Many mudflows were produced on the upper Rfo Maipo by an exceptionally heavy rainstorm on April 18, 1959. The canal that supplied a hydroelectric plant (fig. 152.1) was destroyed in several places, and the river upstream was dammed by a mudflow that descended a valley south of that of the Rfo Yeso and formed an unsorted deposit 10 km long and as much as 100 m thick. The resulting diamicton deposit, which contains blocks as large as 2 or 3 m in diameter (Pierre St. Amand, written communication, February 1963), closely resembles the diamictons of the Loma de los Banos-Cerro Calan area in lithology and thickness. Mass movement presumably produced the great tongue-shaped deposits, later modified by erosion, that extend from Loma de los Banos to Cerro Calan. Other deposits are fan shaped, like the deposit east of Puente Alto. The practically continuous belt of diamicton deposits along the Andean front between the mouths of the canyons of the Rfos Mapocho and Maipo (fig. 152.2) evidently formed by the coalescing of many mudflows. All degrees of transition between soupy mudflows and other types of mass movement can be recognized in central Chile, depending on steepness of the slope and proportion of water in the total mass as shown in the accompanying table. The authors agree with Borde (1955) that the pumice immediately west of Santiago was deposited as mudflows. Explosive eruptions of pumiceous ash are known to have taken place in Recent time in the Cordillera de los Andes. Heavy rains presumably washed the easily erodible ash down the slopes and into the valleys, Types of mass movement in the vicinity of Santiago [Classification after Sharpe, 1938] Movement Earth or rock, dry or with minor amounts of water Type Rate Earth or rock, with water Chiefly water Flow Usually imperceptible Creep Fluvial transpor- tation Perceptible to rapid Mudflow Slip (landslide) Debris avalanche carrying cobbles and boulders of the pre-existing colluvium with it. Lubricated by abundant water, some of the pumiceous mixture evidently flowed out across the Santiago basin, lost monentum, and stopped in the flat lowlands of the western part. Boulders were dropped enroute because of deceleration of the mudflow, but some of the blocks were carried to the terminus of the flow. The present authors reject the supposed glacial origin of the relatively unsorted deposits of the Valle Central near Santiago for the following reasons: 1. Similar diamicton deposits have been formed by mudflows in modern times in the same general area. 2. The bedrock underlying the deposits is deeply weathered; mountain glaciers scour deeply and their till commonly rests on fresh rock. 3. The fine matrix of the diamicton sediments is yel- lowish brown due to limonitization; nowhere is it gray and unweathered like the rock flour of the known glacial deposits of the Rfo Yeso moraine. The coarse material of the diamicton sediments shows no evidence of glacial transport. 4. Neither U-shaped valleys nor cirques, which charac- terize glacial sculpture, occur in or near the Santiago area. 5. At the latitude of Santiago, glacial moraines can be traced only as far downslope as 1,600 m above sea level, more than 1,000 m higher than the valley floor at Santiago. REFERENCES Borde, Jean, 1955, Las depresiones tectdnicas del Maipo inferior—glaciaciones y cenizas volcanicas: Informaciones Geo-graficas 1, Santiago, Chile, Editorial Universitaria, p. 6-16. Briiggen, Juan, 1950, Fundamentos de la geologia de Chile: Instituto Geografico Militar, Santiago, 374 p. Castillo, Octavio, Falcon, Eduardo, Doyel, W. W., and Valenzuela, Manuel, 1963, El agua subterrdnea de Santiago, Segundo informe: Instituto de Investigacitaes Geol6gicas de Chile. [In press]D148 SEDIMENTATION Flint, R. F., Sanders, J. E., and Rodgers, J., 1960a, Symmictite, a name for nonsorted terrigenous sedimentary rocks that contain a wide range of particle sizes: Geol. Soc. America Bull., v. 71, p. 507-510. ------- 1960b, Diamictite, a substitute term for symmictite: Geol. Soc. America Bull., v. 71, p. 1809-1810. Karzulovic, Juan, 1958, Sedimentos cuaternarios y aguas sub- terr&neas en la Cuenca de Santiago: Universidad de Chile, Instituto de Geologia, publ. 10, 120 p. Munoz Cristi, Jorge, 1960, Contribucion al conocimiento geolo-gico de la Cordillera de la Costa de la zona central: Miner-ales, no. 69, p. 28-47. Sharpe, C. F. S., 1938, Landslides and related phenomena: New York, Columbia Univ. Press, 137 p.Article 153 NEW OCCURRENCES OF THE RUGOSE CORAL RHIZOPHYLLUM IN NORTH AMERICA By WILLIAM A. OLIVER, JR., Washington, D.C. Abstract.—Rhizophyllum is known from Middle or Upper Silurian rocks in Tennessee, Kentucky, Indiana, California, Nevada, Alaska, and Maine; from Lower Devonian rocks in Nevada; and from probable Middle Devonian rocks in Alaska. Four of these occurrences are first noted here, and specimens from Alaska, California, Nevada, and Maine are illustrated. The externally similar Calceola is not known to occur in North America. The genus Rhizophyllum has worldwide distribution in rocks of Late Silurian and Early Devonian age; it occurs less commonly in rocks of probable Middle Silurian (Wenlock) and early Middle Devonian (Eifelian) age. In North America, Rhizophyllum is now known from coast to coast in rocks representing its full age range. Rhizophyllum is a near relative of the genus Calceola, with which it has often been confused because of external similarity. Calceola is apparently restricted to the Middle Devonian (Eifelian and Givetian) and is widely distributed on most continents but is not known from North America. Reported occurrences of Calceola in North America are numerous, but all so far have been based on the similar-looking Rhizophyllum. The purpose of this contribution is to illustrate and briefly describe specimens of Rhizophyllum from Alaska, California, Nevada, and Maine, where no specimens have previously been described or in some cases even noted, and to point out that the “Calceola” from Alaska listed by Kindle (1907) and later by Kirk (in Buddington and Chapin, 1929) is actually a Rhizophyllum. Helen Duncan and C. W. Merriam read the manuscript and made many helpful suggestions which were accepted in the final version; both provided specimens and information on occurrences. Thin sections were prepared by W. C. Pinckney, Jr.; photographs are by the author. ART. 153 IN U.S. GEOL. SURVEY PROF. PREVIOUSLY NOTED OCCURRENCES Middle or Upper Silurian (Niagaran)—Tennessee, Kentucky, and Indiana Seven nominal species of Rhizophyllum (as Calceola) were described from the Brownsport Formation in Tennessee and the Louisville Limestone in Kentucky and southern Indiana by Roemer (1854), Safford (1860), Lyon (1879), Hall (1882), and Davis (1887). Prior to this, Troost (1840) had described some Tennessee specimens as Calceola sandalina Lamarck (type species of Calceola). Bassler (1915, p. 157-158) grouped all of these in three species which he referred to “Calceola (Rhizophyllum)”. No review of these has since been attempted, but additional specimens of two of the three species were illustrated or described by Foerste (1931) and by Amsden (1949). The Brownsport and Louisville Formations are of Niagaran (Wenlock or early Ludlow) age. Middle or Upper Silurian—Great Basin, California, and Nevada Stauffer (1930, p. 107, pi. 12, figs. 2-3) described and illustrated a single specimen of “ Calceola sandalina Lamarck” from a section near Kearsarge, Calif, that he considered to be Devonian in age (1930, p. 85-89). Subsequently, Waite (1953, p. 1521) reidentified the coral as Rhizophyllum and dated the beds as “late Niagaran or early Cayugan”. Ross (1963) restudied the stratigraphy and named the formations involved. Merriam (in Ross, 1963, p. 83) listed Rhizophyllum among other corals from the Vaughan Gulch Limestone. Through the courtesy of C. W. Merriam, I have been able to examine the specimen he studied, and a brief description is included under Rhizophyllum sp. D. Apparently Merriam’s Rhizophyllum is from the same stratigraphic unit and locality as Stauffer’s. PAPER 475-D, PAGES D149-D158. 1964. D149D150 PALEONTOLOGY Merriam (1963, p. 37) listed Rhizophyllum from the Silurian at Tkes Canyon, Toquima Range, Nev. This is the first published report of Rhizophyllum from Nevada. Middle)?) Devonian—southeastern Alaska Kindle (1907, p. 325-327) listed fossils from a lower as well as an upper division of the “massive limestone of Long Island, Kasaan Bay”. From the upper division (Kindle collection 819) he listed Galceola cf. sandalina Lamarck with a few other corals and 52 species of brachiopods, mollusks, and arthropods. He regarded the upper fauna as Middle Devonian in age. Kirk (in Buddington and Chapin, 1929, p. 99-101) reexamined Kindle’s collections and made minor changes of identification, but repeated Calceola cf. sandalina. Kirk considered Kindle’s upper and lower faunas to be the same and thought that they might be Early Devonian but were more likely Middle Devonian in age. The identification of Calceola cf. sandalina in both reports was based on a single individual which was not sectioned. It is described and illustrated here as Rhizophyllum sp. B. Corals in the Kindle collection have been sectioned and restudied, and are listed in the following table. New identifications are in the left column, while the names assigned to the same specimens by Kindle and later by Kirk are in the right column. This report Kindle (1907) Amphiphora sp. (stromatopo-roid) Aulocystis sp. Favosites sp. Thamnopora sp. Eddastraeat sp. Rhizophyllum sp. B Pseudamplexus sp. cf. P. prin-ceps Etheridge stauriid coral indeterminate cyathophylloid Syringopora sp. Cyathophyllum sp. Favosites cf. F. radiciformi's Not identified Cyathophyllum sp. Calceola cf. C. sandalina Orthophylluml sp. Cyathophyllum sp. Zaphrentis sp. The coral assemblage is either Early or Middle Devonian in age. The genus Eddastraea and Pseudamplexus princeps are known from both Lower and Middle Devonian rocks of Australia (Hill, 1956, p. 306; Hill, 1950, p. 142). The stromatoporoid genus Amphiphora indicates a probable Middle Devonian age for the collection. Miscellaneous records North American examples of Rhizophyllum or “Calceola” have been formally described only from the Tennessee-Kentucky-Indiana area. Other listings and citations known to the author have been discussed in the preceding paragraphs but it is likely that some records have been missed since a comprehensive review of all literature dealing with the Silurian and Devonian of North America is impractical. Bassler (1950) presented extensive lists of corals by formation and geographic area but noted only the Tennessee-Kentucky-Indiana occurrences of “Calceola” in North America. Conrad (1840, 1841) in his preliminary reports on the paleontology of New York, twice mentioned “Calceola”. His Calceola plicata (Conrad 1840, p. 207) is a Platyceras (Hall, 1859, p. 334). His Calceola indenta (nomen nudum, Conrad, 1841, p. 37) was listed as a “univalve” along with gastropods, cephalopods, a conulariid, and a tentaculitid. It seems likely that the 1841 specimen was not a coral, but I have not found a later reference to it or attempted to trace the specimen. NEW OCCURRENCES Upper Silurian—southeastern Alaska USGS 1005-SD. Vermont marble prospect, south shore of Kosciusko Island, between Edna Bay and Holbrook. Collection of Edwin Kirk, 1917. Brachiopods from this locality were described by Kirk and Amsden (1952) with a brief summary of their occurrence and stratigraphy. On the basis of the brachiopods, Kirk and Amsden (1952, p. 53) suggested an “Upper Silurian” age for the collection. In addition to the brachiopods, collection 1005 includes tabulate and rugose corals and bryozoans. The following corals have been identified by the present author: Heliolites spp., Thamnopora sp., Cystiphyllum sp., Rhizophyllum sp. A, and Tryplasma spp. These are suggestive of a Late Silurian age but are not so restricted. The identification of Rhizophyllum is based on five specimens that are described in a following section. Silurian!?)—northwestern California USGS 5922-SD. Dislocated block of limestone, 4 by 5 by 8 feet, on west bank of Trinity River in the center of sec. 2, T. 32 N., R. 10 W., Weaverviile quadrangle, Trinity County. Collected by W. P. Irwin, 1961. A small collection of fossils, principally corals, was submitted by W. P. Irwin for examination and age determination. In a written communication to Irwin, Helen Duncan (October, 1961) listed “cystiphyllid corals (Rhizophyllum?)” among other forms and indicated that the age of the collection was Silurian or Devonian, more likely Silurian. With the help of additional sections, the presence of several specimens of Rhizophyllum has been confirmed. These are described below as R. sp. C. In addition, the associated corals have been restudied by Miss Duncan and myself, and the following list is the result of our joint efforts: lieliolitid coral fragments, Rhizophyllum sp.OLIVER D151 C, Spongophylloides? sp., streptelasmatoid corals, Tryplasmal sp., Zelophyllum? sp. The corals are badly recrystallized and somewhat distorted in shape due to metamorphism, and most of the identifications are queried. Tn spite of this, the apparent growth form of Rhizophyllum sp. C, the probable presence of Spongophylloides and Zelophyllum, and the internal morphology of the streptelasmatoids, are all strongly suggestive of Wenlock or Ludlow (Middle or Late Silurian) age. The loose block from which the corals were collected was in an area mapped as Bragdon Formation (in central metamorphic belt of Irwin, 1960, p. 18). The Bragdon is considered to be Mississippian in age but contains fragments of older fossiliferous limestones in some conglomerate beds (Irwin, 1960, p. 18). Lower Devonian—Nevada USGS 6267-SD. Unnamed limestone unit forming a prominent ledge on the east side of Coal Canyon, at an elevation of approximately 6,280 feet; SE% sec. 17, T. 25 N., R. 49 E., northern Simpson Park Mountains, Horse Creek Valley 15-minute quadrangle, Eureka County, Nev.; collected by M. R. Murphy, 1957 (J. G. Johnson, written communication, 1963). The following corals were present in a collection submitted for identification by J. G. Johnson, California Institute of Technology: Favosites sp., Endo-phyllum sp., Mucophyllum sp., Papiliophyllum ele-gantulum Stumm, Rhizophyllum sp. cf. R. enorme Etheridge, and cystiphylloid and chonophylloid corals. The identification of the Rhizophyllum is based on a single specimen described below. Papiliophyllum elegantulum suggests an Early Devonian age as the species is most common in the Spirijer kobehana zone of Merriam (1940, p. 52-53). Merriam considered the zone to be Early Devonian, Oriskany or later, in age, and subsequent workers have agreed with this evaluation. Endophyllum is principally a Devonian coral, but species have been described from the Upper Silurian of Europe (Prantl, 1952). Mucophyllum is a Middle and Upper Silurian genus that has not been described or illustrated from younger rocks although there are incidental references to Devonian occurrences (for example, Wang, 1950, p. 228). The ot her listed corals could be either Silurian or Devonian. As a whole, the corals support an Early Devonian age. The presence of Mucophyllum suggests that the fauna may represent an early part of the early Devonian. According to Johnson (written communication, 1963), who is describing the brachiopods, the unnamed limestone from which the corals were collected is of Gedin-nian (early Early Devonian) age and “is overlain by the Rabbit Hill Formation [Merriam, 1963, p. 42-44] with an upper Gedinnian brachiopod fauna.” Further, “the Rabbit Hill is in turn overlain by the Nevada Formation with the zone of Acrospirijer kobehana in which Papiliophyllum elegantulum is typical. Accordingly, the occurrence of P. elegantulum in the unnamed limestone extends the range of the species considerably downward”. Upper Silurian—Maine USGS 6525-SD. On the boundary line between the Spider Lake and Telos Lake quadrangles, on the shore of Third Lake Mattagamon. Collected by Bradford A. Hall, University of Maine, 1961. A collection of corals which was submitted for identification by B. A. Hall, includes the following: Cladoporad sp., kavosites sp., IPalysites sp., Heliolites sp., Syringo-pora sp., “Cystiphyllum" sp., Disphyllum sp., Entelo-phyllum sp., Rhizophyllum sp. cf. R. gotlandicum (Roemer), and Triplasma? sp. The coral assemblage is definitely Silurian, either Wenlock or Ludlow in age. Brachiopods from the same collection have been studied by A. J. Boucot, California Institute of Technology; they indicate a Ludlow (Late Silurian) age (Boucot, written communication, 1962). The identification of Rhizophyllum is based on a single specimen which is described and illustrated below. SYSTEMATIC DESCRIPTIONS Genus Rhizophyllum Lindstrom Rhizophyllum Lindstrom, 1866a, p. 287; 1866b, p. 411; 1883, p. 22; Sherzer, 1891, p. 296; Hill, 1940, p. 394; Wang, 1948, p. 1, 3; Hill, 1956, p. 314. Calceola of many authors (including all known descriptions or citations of North American specimens). Type species.—By monotypy, Calceola gotlandica Roemer, 1856, p. 798; Silurian, Island of Gotland, Sweden. Diagnosis.—Rhizophyllum includes calceoloid rugose corals with a semicircular operculum which articidates on the flat, counter side of the corallum. The interior is filled with arched dissepiments and tabellae arranged in an inverted cone pattern. Septa are acanthine or laminar and are limited to the flat side; the counter septum is commonly longer and thicker than the other septa. The wall on the flat side of the corallum is formed by the fusion of the marginal edges of the septa; the curved part of the wall is apparently formed in the same way, although the septa do not project into the lumen even as septal ridges. The axis is straight or curved. The corallum in some species was supported by radiciform processes. Discussion.—Lindstrom (1866a, b) established the genus Rhizophyllum to separate the “Calceola” withD152 PALEONTOLOGY cystiphylloid internal structure from typical Calceola with stereoplasm-filled interiors. He discussed opercu-lated corals at length (1866a, b, 1883) and placed all of the American “Calceola” in two species of Rhizophyl-lum. Bassler (1915) considered Rhizophyllum to be a subgenus of Calceola and was followed in this usage by Foerste (1931) and Shinier and Shrock (1944). Sherzer (1891) and Amsden (1949) are the only two American paleontologists who have published taxonomic papers using Rhizophyllum as a genus; European and Australian workers have recognized the genus since its first description. Rhizophyllum ranges from the Middle Silurian (Wenlock) to the lower Middle Devonian (Eifelian). Species have been described from rocks of both Silurian and Devonian age in Europe, Asia, Australia, and North America. Middle Devonian occurrences are uncommon but have been reported from the Eifelian of Mongolia (Spassky, 1960, p. 124-125) as well as from Alaska (this article). Hill (1942, p. 13-16) reported Rhizophyllum and Calceola from the same limestone in Queensland, but she concluded that the age was latest Early Devonian. Descriptions.—The following descriptions are of specimens rather than species, 1 being based on 5 specimens, 1 on 10 specimens, and the other 4 on one specimen each. Two of the “species” are compared with previously named species that are morphologically similar; at least one “species” is new but formal description must await the acquisition of more material. The dimensions of all specimens described are given in millimeters. Length was measured in a straight line from the counter side of the calice margin to the apex; since most specimens are incomplete at one or both ends, the figures represent minimum length. The diameters were measured at the calice margin or at the uppermost complete growth line. Rhizophyllum sp. A Figure 153.1a-e, 153.2a-,/, 153.36 Description.—Five specimens of solitary Rhizophyllum, lacking radiciform processes probably represent a new species. Two cycles of growth lines are prominent on the exteriors; where the outer surface is worn, the traces of the septa can be seen on both flat and convex sides. Tn transverse section the counter septum is two or more times the length of the other septa and is thickened to a spindle shape. The septum is formed of calcite fibers radiating upward and from the center of the spindle within an outer sheath of lamellar tissue. This gives an elongate rosette appearance in transverse section and a fan pattern in longitudinal section. Other septa are numerous and closely spaced; they are enveloped in lamellar tissue but the microstructure of their inner portions has not been observed. In longitudinal section, the dissepiments are very steep and elongate near the curved, cardinal side but almost horizontal near the counter margin. From the counter side, the dissepiments steepen toward the eccentric axis about which the tabellae are large and horizontal. Measurements are as follows: Diameter (mm) Specimen (USNM No.) Length (mm) Major Minor 121200 _____________ 14 12. 7 7. 1 121201 _____________ 19 17.2 9.6 121202 ________ 26 27. 2 11. 8 121203 _____________ 24 21.8 11.0 121204 ________ 18.5 10.1 Discussion.—R. sp. A is characterized by its inflated counter septum. A similar structure is seen in R. yotlandicum of Rozkowska (1946, p. 145-146, 154-155, pi. 5, figs. 4a-b) from the Silurian of Podolia. The Podolian material differs in having comparatively few, but large dissepiments and tabellae. R. yotlandicum of Soshkina (1937, p. 82, pi. 19, fig. 1-2) from the Upper Silurian of the Urals, lacks the prominent counter septum but is similar in other respects. Typical R. yotlandicum (Lindstrom, 1866a, p. 287, pi. 30, fig. 10-15, pi. 31, fig. 1-8; 1883, p. 23, pi. 3, fig. 1-12, 18, pi. 4, fig. 7) from the Middle and Upper Silurian of Gotland, has a less prominent counter septum and reduced (acanthine?) lateral septa; in addition, typical R. yotlandicum has numerous supporting radiciform processes. Material.—Five specimens from the Upper Silurian of Alaska (USGS 1005-SD; see stratigraphic discussion). Illustrated specimens, USNM 121200 to 121203. Unillustrated specimen, USNM 121204. Rhizophyllum sp. B Figure 153.1/,gr Description.—A single specimen of solitary Rhizophyllum from the Devonian of Alaska, has a curved, conical shape and lacks radiciform processes; dimensions are: length, 15+ mm; and diameter, 11.4 by 20.7 mm. The exterior is marked by growth rugae. In a transverse section, 26 discrete septa are present along 15 mm of the straight margin; the septa are vertically discontinuous and may be acanthine. The presumed counter septum is only slightly longer than the other septa. Dissepiments are small and elongate; large tabellae surround the eccentric axis. A neanic section indicates that the dissepimentarium was relatively narrow in early growth stages, but ephebic and neanic sections are similar in other respects.OLIVER D153 Figure 153.1.—Rhizophyllum sp. A and sp. B. The counter septum is toward the top of the plate in the transverse thin sections and to the left in the longitudinal thin section. All specimens X 4. a-e, Rhizophyllum sp. A; ephebic sections; a-b, USNM 121203; c-d, USNM 121202; e, USNM 121201; Upper Silurian, Alaska, f-g, R. sp. B; neanic and ephebic section; USNM 140899; Middle(?) Devonian, Alaska. Discussion.—The one available specimen is morphologically similar to R. robustum Shearsby (1906, p. 548, pi. 26, fig. 1-6; Hill, 1940, p. 396, pi. 11, fig. lla-b) from the Silurian of New South Wales. R. robustum differs in being more erect and in having finer tabellae and occasional radiciform processes. Material.— One specimen from the Middle? Devonian of Alaska (Kindle 897; see stratigraphic discussion). Illustrated specimen, USNM 140899; collected by E. M. Kindle, 1905. Rhizophyllum sp. C Figure 153.4 a-e Description.—'Pen or more specimens of a small Rhizophyllum are known from transverse sections only. The specimens were found in 2 groups, 1 with 3 or more individuals and the other with 7 or more; the presence of numerous radiciform processes and the spatial relationships of the individuals suggest that they occurred in clumps or small colonies. Individual diameters range from 2.5 by 3.0 mm to approximately 5 byD154 PALEONTOLOGY Figure 153.2.—Rhizophyllum sp. A; exterior views; Upper Silurian, Alaska, a-c, cardinal, counter, and side views; USNM 121201. d-e, counter and side views; USNM 121202. f-g, cardinal and side views; USNM 121203. h-j, cardinal, counter, and side views; USNM 121200. All specimens X 1%. Figure 153.3.—a, Rhizophyllum sp. cf. R. enorme; detail of figure 153.56; USNM 121205; Lower Devonian, Nevada. 6, R. sp. A; detail of figure 153.le; USNM 121201; Upper Silurian, Alaska. Transverse thin sections, X 10.OLIVER D155 Figure 153.4.—Rhizophyllum sp. C; USNM 121344; Silurian(?), northwestern California, a-b, Sections through group or clump of specimens, X 3; immature and mature transverse sections of individuals and sections of radiciform processes and probable opercula are shown, c, Detail of b, X 10; transverse sections through neanic portion of two corallites and partial section of ephebic portion of another, d, Detail of b, X 10; transverse section of neanic portion of a corallite. e, Detail of a, X 10; transverse section of early ephebic portion of a corallite. 9 mm (3 specimens). All specimens have a relatively thick wall (0.3-0.5 mm) with a prominent counter septum on the straight side. In smaller sections (neanic and brephic?; diameter less than 3.5 by 5.0 mm) other septa form the wall but do not project into the lumen; one or two dissepiments are present in some of the sections. In the larger sections (ephebic) the lumen is filled with dissepimental plates and septa project slightly from the straight margin. The calice is relatively deep and V-shaped. Several cresent-shaped objects in the thin sections may represent the opercula of this species. Discussion— In apparent growth form and general morphology, Rhizophyllum sp. C is similar to R. elon-gatum Lindstrom (1883, pi. 2) from the Middle or Upper Silurian of Gotland, Sweden, and to R. attenu-atum (Lyon) (Lindstrom, 1883, pi. 3, fig. 17) from the Middle or Upper Silurian, Louisville Limestone in 716-626 O—64-------11 Kentucky. R. sp. C is closer to R. attenuatum in size and number of radiciform processes, but adequate comparison is not possible because of the nature of the California specimens. R. sp. C may be allied to one of the two clump-forming species or may represent a third species but is, in either case, suggestive of Silurian age. Material.—Two groups totaling ten or more specimens from the Middle or Upper Silurian of northwestern California (USGS 5922-SD; see stratigraphic discussion). Illustrated group, USNM 121344; unillustrated group, USNM 121345. Rhizophyllum sp. D [?] Calceola sandalina Stauffer, 1930, p. 107, pi. 12, figures 2-3 (not Linnaeus) Description.—A single specimen of solitary Rhizo-phyttum from the Silurian of eastern California, has aD156 PALEONTOLOGY curved conical shape and apparently lacks radiciform processes; dimensions are: length, 43+ nun; and diameters, 15+ by 35 mm. The specimen is poorly preserved with beekitized exterior and normally silicified interior. The counter septum projects 2 to 3 mm in large sections; other septa are half as long, discrete, probably acanthine, and limited to the straight side. Dissepiments are steeply inclined on both sides of the eccentrically located axial zone of tabellae; dissepiments and tabellae are of approximately the same size. Discussion.—The specimen is too poorly preserved for comparison with other species of the genus, but it is probably conspecific with Stauffer’s specimen (see stratigraphic discussion). Material.•—One specimen from the Middle or Upper Silurian of California (USGS 6839-SD). Unillustrated specimen, USNM 121346. Rhizophyllum sp. cf. R. enorme Etheridge Figures 153.3a, 153.5a,6 Rhizophyllum enorme Etheridge, 1903, p. 232-233, pi. 47; Hill and Jones, 1940, p. 182, pi. 2, fig. 3^1. Description.—A single specimen with curved axis is 20+ mm in length, with diameters of 15.8 and 27.5 mm; the exterior is poorly preserved but there is no evidence of radiciform processes. In transverse section, 28? septa are present along 14 mm of the straight margin. The counter septum is twice the length of the adjacent septa and is thickened by a sheath of lamellar tissue. A longitudinal section in the cardinal-counter plane shows rather coarse tabellae at the eccentric axis with smaller, elongate, steeply inclined dissepiments on either side. The structure of the septa cannot be determined from available sections but they seem to be laminar. Discussion.—The described specimen is similar to R. enorme as illustrated by Hill and Jones (from the Lower Devonian of New South Wales) in internal structure but is much smaller and has relatively coarser tabellae and dissepiments. Material.—One specimen from the Lower Devonian of Nevada (USGS 6267-SD; see discussion). Illustrated specimen, USNM 121205. Rhizophyllum sp. cf. R. gotlandicum (Roemer) Figures 153.5 c-e Calceola gotlandica Roemer, 1856, p. 798. Rhizophyllum gotlandicum (Roemer), Lindstrom, 1866a, p. 287, pi. 30, fig. 10-15, pi. 31, fig. 1-8; 1866b, p. 406M11, pi. 14, fig. 8-18; 1883, p. 23, pi. 3, fig. 1-12, 18, pi. 4, fig. 7. Description.—One specimen of Rhizophyllum from the Upper Silurian of Maine is incomplete but represents a solitary, erect form with radiciform processes in the apical region; dimensions are: length, 45+ mm; and diameters, 26 by 16 mm. The appearance of the external surface is not known. In transverse section the counter septum is relatively long and inflated; there is a suggestion of a radiating structure and there is a e b Figure 153.5.—Rhizophyllum sp. cf. R. enorme and R. sp. cf. R. gotlandicum. The counter septum is toward the top of the plate in the transverse thin sections and to the left in the longitudinal thin section. All specimens X 4. a-b, Rhizophyllum sp. cf. R. enorme; ephebic longitudinal and transverse sections; USNM 121205; Lower Devonian, Nevada, c-e, R. sp. cf. R. gotlandicum; brephic?, neanic, and ephebic transverse sections; circular and oval bodies above and to the right of the specimen in e are sections of radiciform processes; USNM 121206; Upper Silurian Maine.OLIVER D157 sheath of lamellar tissue. Other septa are short and acan thine. The dissepiments are elongate and steeply inclined; the axial tabellae are more globose but not much larger than the dissepiments. Both dissepiments and tabellae are thickened with excess stereome deposits. A neanic section shows a few tabellae with an incomplete marginal row of dissepiments and a short, stumpy counter septum; no other septa are developed. An even earlier section (brephic?) shows only a bulge in the straight side representing the counter septum and a single tabella. Discussion.-—The Maine specimen is remarkably similar to specimens illustrated by Lindstrom (1883). A weathered longitudinal section is like Lindstrom’s figure 6 and the neanic section (my fig. 153.5d) is close to his figure 5. Lindstrom did not publish an illustration of an ephebic transverse section but his calice illustrations (1866b, pi. 14, fig. 8-11, 14) suggest that the septal development is similar to that described here for the Maine specimen. Material.—One individual from the Upper Silurian of Maine (USGS 6525-SD; see stratigraphic discussion). Illustrated specimen, USNM 121206. Collected by Bradford A. Hall, 1961. REFERENCES Amsden, T. W., 1949, Stratigraphy and paleontology of the Brownsport Formation (Silurian) of western Tennessee: Peabody Mus. Nat. Hist., Yale Univ., Bull. 5, 138 p., 34 pi. Bassler, R. S., 1915, Bibliographic index of American Ordovician and Silurian fossils, v. 1 and 2: Smithsonian Inst., U.S. National Mus. Bull. 92, 1521 p. ------ 1950, Faunal lists and descriptions of Paleozoic corals: Geol. Soc. America Mem. 44, 315 p. Buddington, A. F., and Chapin, Theodore, 1929, Geology and mineral deposits of southeastern Alaska: U.S. Geol. Survey Bull. 800, 398 p. Conrad, T. A., 1840, Third annual report on the palaeontological department of the Survey: New York Geol. Survey Ann. Rept. 4, p. 199-207. ------ 1841, Fifth Annual report on the paleontology of the State of New York: New York Geol. Survey Ann. Rept. 5, p. 25-57. Davis, W. J., 1887, Kentucky fossil corals, pt. 2: Kentucky Geol. Survey, 139 plates and expl. Etheridge, Robert, Jr., 1903, An unusually large form of Rhizo-phyllum lately discovered in New South Wales: Rees. Geol. Survey New South Wales, v. 7, p. 232-233, pi. 47. Foerste, A. F., 1931, The Silurian fauna: Kentucky Geol. Survey, ser. 6, Paleontology of Kentucky, p. 167-214. Hall, James, 1859, Descriptions and figures of the organic remains of the Lower Helderberg group and the Oriskany sandstone: New York Geol. Survey, Paleontology of New York, v. 3, text, 532 p. ------ 1882, Fossil corals of the Niagara and Upper Helderberg groups: New York State Mus., Ann. Rept. 35, advanced sheets, 59 p. Hill, Dorothy, 1940, The Silurian Rugosa of the Yass-Bowning district, N.S.W.: Linnean Soc. New South Wales Proc., v. 65, p. 388-420, pi. 11-13. ------- 1942, The Lower Devonian rugose corals from the Mt. Etna limestone, Qld.: Royal Soc. Queensland Proc., v. 54, p. 13-22, pi. 1. ------- 1950, Middle Devonian corals from the Buchan district, Victoria: Royal Soc. Victoria Proc., v. 62, p. 137-164, pi. 5-9. ------- 1956, Rugosa, in Moore, R. C., ed., Treatise on invertebrate paleontology, pt. F. Coelenterata: Geol. Soc. America and Kansas Univ. Press, Lawrence, Kans., p. 233-324. Hill, Dorothy, and Jones, O. A., 1940, The corals of the Garra beds, Molong district, New South Wales: Royal Soc. New South Wales Jour, and Proc., v. 74, p. 175-208. Irwin, W. P., 1960, Geo ogic reconnaissance of the northern Coast Ranges and Klamath Mountains, California: California Div. Mines Bull. 179, 80 p. Kindle, E. M., 1907, Notes on the Paleozoic faunas and stratigraphy of southeastern Alaska: Jour. Geology, v. 15, p. 314-337. Kirk, Edwin, and Amsden, T. W., 1952 Upper Silurian brachio-pods from southeastern Alaska: U.S. Geol. Survey Prof. Paper 233-C, p. 53-66, pi. 7-10. Lindstrom, Gustav, 1866a, Nagra iakttagelser ofver zoantharia rugosa: Ofvers. Kongl. Vetensk.-Akad. Forhandl., v. 22, p. 271-294, pi. 30-31. —------1866b, Some observations on the Zoantharia Rugosa: Geol. Mag., v. 3, p. 356-362, 406-414, pi. 14 (author translation of 1866a reference). -------1883, Om de Palaeozoiska formationernas Operkelbarande Koraller: Bihang Kongl. Svensk. Vetensk.-Akad. Handl., v. 7, 112 p., 9 pi. Lyon, V. W., 1879, Descriptions of three new species of Calceoli-dae from the upper Silurian rocks of Kentucky: Acad. Nat. Sci. Philadelphia Proc., p. 43-46. Merriam, C. W., 1940, Devonian stratigraphy and paleontology of the Roberts Mountains region, Nevada: Geol. Soc. America Spec. Paper. 25, 114 p., 16 pi. -------1963, Paleozoic rocks of Antelope Valley, Eureka and Nye Counties, Nevada: U.S. Geol. Survey Prof. Paper 423, 67 p. Prantl, Ferdinand, 1952, The genera Endophyllum Edwards and Haime and Spongophyllum Edwards and Haime in the Silurian and Devonian of Bohemia: Sbornik Geol. Survey Czechoslovakia, v. 18, p. 221-240. Roemer, F. A., 1852-1854, Erste Periode; Kohlen-Gebirge, in Bronn, H. G., Lethaea Geognostica, 3d ed.: Stuttgart, 788 p. ———• 1856, Bericht von einer geologisch-palaontologischen Reise nach Schweden: Neues Jahrb. Min., Geog., Geol., p. 794-815. Ross, D. C., 1963, New Cambrian, Ordovician, and Silurian formations in the Independence quadrangle, Inyo County, California: Art. 21 in U.S. Geol. Survey, Prof. Paper 475-B, p. B74-B85. Rozkowska, Maria, 1946, The Silurian rugose corals from Podolia: Polskiego Towarzystwa Geologicznego, Rocznik, T. 14, p. 139-157, pi. 4-5. Safford, J. M., 1860, On the species of Calceola found in Tennessee: Am. Jour. Sci., ser. 2, v. 29, p. 248-249. Shearsby, A. J., 1906, Operculate corals from New South Wales: Geol. Mag., new ser., Decade 5, v. 3, p. 547-552, pi. 26.D158 PALEONTOLOGY Sherzer, W. H., 1891, A chart of the rugose corals: Am. Geologist, v. 7, p. 273-301. Shimer, H. W., and Shrock, R. R., 1944, Index fossils of North America: New York, John Wiley and Sons, 837 p. Soshkina, E. D., 1937, [Corals of the upper Silurian and lower Devonian of the eastern and western slopes of the Urals]: USSR, Acad. Sci., Inst. Paleozool., Tr., v. 6, pt. 4, 112 p., 21 p. Spassky, N. Ya., 1960, [The Devonian rugose corals of the southern Altai and adjoining territories]: Leningrad, Gorny Inst. Records, v. 37, p. 108-131, pi. 1-10. Stauffer, C. R., 1930, The Devonian of California: California Univ. Pubs, in Geol. Sci., v. 19, no. 4, p. 81-118, pi. 10-14. Troost, Gerard, 1840, Organic remains discovered in the State of Tennessee: Fifth Geol. Rept. to the 23rd General Assembly of Tennessee, p. 45-75. Waite, R. H., 1953, Age of the “Devonian” of the Kearsarge area, California [abs.]: Geol. Soc. America Bull., v. 12, p. 1521. Wang, H. C., 1948, Note on a remarkable Rhizophyllum species from the middle Silurian of Hueitze, northern Yunnan: Palaeont. Soc. China, Palaeont. Novitates, no. 2, 4 p. ----—• 1950, A revision of the Zoantharia Rugosa in the light of their minute skeletal structures: Royal Soc. London, Philos-Trans., ser. B, no. 611, v. 234, p. 175-246, pi. 1-9.Article 154 GLACIAL CHRONOLOGY OF ULLSFJORD, NORTHERN NORWAY By G. WILLIAM HOLMES and BJORN G. ANDERSEN, Beltsville, Md., and University of Oslo Work done in cooperation arith the Air Force Cambridge Research Laboratories Abstract.—Quaternary features of Ullsfjord record Late Glacial fluctuations of the Fennoscandian ice sheet and of local alpine glaciers. The earliest known advance left erratics and subdued moraines near the coast. The subsequent Breidvika advance formed lateral moraines which continue into the sea as sharp submarine end moraines. Shortly thereafter the Skardmunken advance also resulted in moraines with both terrestrial and marine facies. Radiocarbon dates of marine shells place the Skardmunken maximum between 11,500 and 10,390 years B.P. (late Allerpd and Younger Dryas time). A nearly continuous beach level, probably the “main beach level” of northern Norway, is cut in bedrock and unconsolidated materials from the outer coast to the top of the deltaic sediments of the Skardmunken moraine. Shortly after the Skardmunken advance, the Stordal advance left small moraines and deltas at a level slightly below the projection of the “main beach level.” Lastly, cirque glaciers deposited steep unweathered moraines in historic time. Ullsfjord is a deep, narrow trough with a total length of 50 kilometers, a maximum width of 5 kilometers, and a maximum relief of nearly 1,700 meters. In places the sides are so steep that no surficial deposits remain. The fiord was eroded along the contact between gabbroic rocks that make up the high, spectacular Lyngen Alps, and schistose rocks that form the slightly lower alpine peaks to the west. Scores of mountain glaciers originated in these mountains during the Pleistocene, and many cirques are still filled with ice. Narrow structural valleys connect Ullsfjord with fiords to the east, west, and south. These troughs were access routes for outlet glaciers of the Fennoscandian ice sheet that invaded the fiord in Pleistocene time. The bottom of the fiord is very irregular, as a result of bedrock relief and glacial deposition. Depths range from about 11 meters over submarine moraines, to 268 meters in the outer fiord. ART. 154 IN U.S. GEOL. SURVEY PROF. Although Ullsfjord lies well above the Arctic Circle, its climate is humid temperate, very similar to that of southern Alaska (Haurwitz and Austin, 1944). The nearest weather station, at Tromsp, 25 kilometers to the west, has a mean annual temperature of 2.3°C and a mean annual precipitation of 940 mm (Hansen, 1960, p. 46-47). The glacial record in Ullsfjord is typical of northern Norway. Our studies show that the glacial sequence consists of (1) an unnamed early glaciation, (2) the Breidvika advance, (3) the Skardmunken advance, (4) the Stordal advance, and (5) a Recent advance (see accompanying table and fig. 154.1). One important basis for this chronology is the prominent elevated shoreline (informally referred to here as the “main beach level”) that was formed at the time of the Glacial chronology of Ullsfjord Glaciation Age Correlation Basis for correlation Recent 18th Century. “Little Ice Age.” Near or in contact with existing glaciers. Stordal Slightly younger than Skardmunken. Possibly Preboreal. Moraines are behind and (or) at lower levels than Skardmunken ice margins. Skardmunken 11,500-10,390 years. Equivalent to Troms0-Lyngen advance of northern Norway. Radiocarbon dates are approximately the same as the Troms0-Lyngen moraines. Moraines graded to “main beach level.” Breidvika Slightly older than Skardmunken. Possibly equivalent to Skarpnes advance of northern Norway. Moraine position is a short distance beyond Skardmunken moraines. Moraines truncated by “main beach level.” Early Somewhat older than Breidvika. Possibly equivalent to one or more older advances in southern Norway. Drift occurs seaward of Breidvika moraines. PAPER 475-D, PAGES D159-D163. 1964. D159D160 GLACIAL GEOLOGY Figure 154.1.—Glacial and marine deposits near Ullsfjord, northern Norway. A-A', line of profile shown in figure 154.2, large dots in Ullsfjord are projections of points on the shores where the “main beach level” was measured.HOLMES AND ANDERSEN D161 EXPLANATION Recent moraine Stordal moraine Skardmunken moraine Breidvika moraine Early moraine Alpine moraine, Glaciofluvial undifferentiated deposits Main beach level Inferred ice margin x T-333 Radiocarbon sample locality Figure 154.1.—Continued Skardmunken glaciation, between 11,500 and 10,390 years B.P., according to radiocarbon dates. Some moraines in the mountains are not clearly related to datable features and are designated only as alpine moraines, undifferentiated. SHORELINES Elevated, tilted Pleistocene shorelines in northern Norway were first recognized by Bravais in 1838 (Marthinussen, 1960, p. 416) in Altafjord, about 100 kilometers east of Ullsfjord. Their discovery created a basis for the theory of isostatic recovery of glaciated terrain. These shorelines were formed at times when sea level was nearly constant, which in late Pleistocene time probably corresponded to intervals when the sea and the land were rising at the same rate. The most distinct shoreline in Ullsfjord is informally named the “main beach level.” It is nearly continuous from the outer fiord, where it is about 35 meters above sea level, to a few kilometers south of Skardmunken, where it is about 70 meters above sea level. The “main beach level” is cut in bedrock, older moraines, marine till, and gravel deposits. It is about the same age as the Skardmunken moraine, for (1) it is graded to the top of that moraine at about the level that separates the terrestrial and marine portions; (2) it is graded to extensive Skardmunken outwash deposits west of Breidvika; and (3) it does not extend far beyond the Skardmunken moraine, indicating that ice of this advance filled the inner fiord while the shoreline was being cut. Shorelines related to other glaciations are limited in extent, and are marked primarily by deltas. Sea level was probably changing too rapidly during these in- tervals to create a persistent shoreline. A distinct lower shoreline formed after the “main beach level,” but it apparently is not related to a glacial advance and merely represents a second period of sea-level stability. GLACIATION Early glaciation The oldest recorded glaciation left scattered erratics, patches of till, a few indistinct moraines, and drainage channels in the northern part of the fiord district. The most distinct feature is a broad moraine that dams Trollvann in the northwestern part of the mapped area. The ice sheet that deposited this drift probably extended north of the present shoreline and covered the continental shelf. The age of this glaciation is not known, but the drift occupies a position analogous to that of the Lista substage, of Oldest Dryas age (more than 13,000 years old), or the Spangereid substage (dated about 13,000 years B.P.) (Andersen, report in preparation). Breidvika glaciation The oldest glaciation in Ullsfjord which left large well-preserved moraines is here named informally the Breidvika advance, for the cove on the west side of the fiord. The largest Fennoscandian outlet glacier moved northward and terminated in the sea just south of Breidvika, leaving a prominent lateral moraine on the east side of the fiord and a small segment on the steeper west side. Profiles of the lateral moraine on the east side of the fiord (fig. 154.2) show that the ice sloped steeply toward its terminus in the fiord. Echo soundings show a very sharp double ridge (fig. 154.2) on the fiord bottom with a relief of approximately 38 meters. A western arm of this glacier pushed westward out of the fiord, turned north, entered a parallel valley, and deposited a prominent terrestrial end moraine. A North A METERS South Figure 154.2.—Profiles of prominent moraines, shorelines, and submarine ridges in the main part of Ullsfjord. Large dots are projections of points on the shores where the “main beach level” was measured. Location of profile shown in figure 154.1.D162 GLACIAL GEOLOGY A second large glacier moved westward from Lyngen-fjord (east of the mapped area) and the Lyngen Alps down the east arm of the fiord, leaving a smaller pair of lateral ridges and a submarine ridge with a maximum relief of approximately 45 meters (fig. 154.1). The echo profile of this ridge shows many smaller ridges on the east (upglacier) side, but not a sharp reverse slope in that direction. The age of the Breidvika glaciation is not known. Its moraines appear to be only a little older than the succeeding (Skardmunken) glaciation, and related shorelines are only 2 to 4 meters above the “main beach level,” which is roughly the same age as the Skardmunken glaciation. Also, the Breidvika moraines occupy positions analogous to the Skarpnes moraine near Tromsp to the west, which is regarded to be somewhat older than Younger Dryas (Andersen, report in preparation), that is, the age of the Skardmunken glaciation. Skardmunken glaciation The best preserved moraine in Ullsfjord is here named informally for the settlement on the terminal section. This part of the moraine was deposited in the sea and consists of stratified drift displaying foreset bedding which was smoothly planed by subsequent wave action. The submarine part is a smooth ridge crossing the fiord; it is being reworked by strong tidal currents. The terminal section merges with a large terrestrial lateral moraine on the east side of the fiord. This lateral moraine, which in places branches into several ridges, extends southward for a distance of about 9 kilometers. The glacier that formed this moraine also pushed into the valley joining the fiord on the west. However, ice of this glaciation did not leave a clear record in the east arm of the fiord. A terrestrial moraine of this glaciation was also deposited in the broad trough west of Breidvika by ice which moved from adjoining Balsfjord. Glaciofluvial deposits spread down the valley from this moraine and merged with the “main beach level” along the shores of Ullsfjord. A minimum and probably close radiocarbon date of 10,390 years B.P. (T-333, Trondheim Radiological Dating Laboratory, 1962) for the Skardmunken moraine was obtained from mollusk shells (Macoma calcarea(?), Saxicava arctiea(‘i), and Mya tnmcata(T)). The shells were collected from varved north-dipping sandy marine clay on the proximal side of the end moraine, lying about 30 to 40 meters above sea level and overiying coarse stratified south-dipping drift. The date probably corresponds to the beginning of the retreat of the Skardmunken glacier. Shells of Portlandia arctica and Macoma calcarea (T-110) collected near the top of pebbly glaciomarine clay beds which underlie outwash of Skardmunken age in the valley west of Breidvika (fig. 154.1) were dated at 11,500±400 years B.P. The lithology and date of the shell-bearing material suggest it was deposited in a glacial environment immediately before the maximum of the Skardmunken advance. Marine till on the east side of Ullsfjord opposite Breidvika, although not demonstrably related to the Skardmunken moraine, is somewhat younger than the glaciomarine clay west of Breidvika. Shells in the former (Saxicava arctica, Macoma calcarea, and Mya truncata, T-332) have a radiocarbon date of 11,090 ± 190 years B.P. On the basis of the first two dates, the Skardmunken advance occurred between about 11,500 and 10,390 years B.P., or at approximately the same time as the Tromsp-Lyngen maximum in adjoining fiords (Andersen, report in preparation). These dates span late Allerpd and Younger Dryas time in southern Scandinavia (Flint, 1957, p. 397). Stordal glaciation Distinct end moraines occur behind Skardmunken moraines in several valleys, in some places in groups of two or more; none are associated with present glaciers. These minor readvances are here named informally for Stordal, a tributary valley on the west side of the inner fiord. Here is a small but distinct end moraine, at an elevation of about 70 meters, well below the projected level of the Skardmunken glacier and at about the same level as the oldest elevated shoreline in the inner part of the fiord. Similar moraines occur behind terrace-dated Skardmunken moraines in the northeastern corner of the mapped area. The outwash delta from the type Stordal moraine, graded to the oldest elevated marine terrace in the inner part of the fiord, is only about 5 meters below the “main beach level.” Thus this advance occurred shortly after the Skardmunken glaciation and is possibly Pre-Boreal in age. Recent glaciation Nearly all the existing glaciers are fronted by small, steep, bouldery, unweathered moraines. These are similar to modern moraines found in most alpine environments throughout the Northern Hemisphere. Although it is generally held that Recent moraines are not necessarily of precisely the same age, those in northern Scandinavia probably formed in the 18th century, as suggested by historical records in Norway (Liestpl, 1960, p. 487).HOLMES AND ANDERSEN D163 REFERENCES Flint, R. F., 1957, Glacial and Pleistocene geology: New York, John Wiley and Sons, 553 p. Hansen, S. W., 1960, The Climate, in Vorren, 0., ed., Norway north of 65: Oslo, Oslo University Press, 271 p. Haurwitz, B., and Austin, J. M., 1944, Climatology: New York, McGraw-Hill, 410 p. Liest0l, O., 1960, Glaciers of the present day, in Holtedahl, O., Geology of Norway: Norges Geol. Undersokelse, no. 208, 540 p. Marthinussen, M., 1960, Coast and fjord area of Finnmark, in Holtedahl, O., Geology of Norway: Norges Geol. Undersokelse, no. 208, 540 p.Article 155 EFFLUENT COLLECTOR FOR GAS CHROMATOGRAPHY By DONALD F. GOERLITZ and WILLIAM L. LAMAR, Menlo Park, Calif. Abstract.—The authors have designed a convenient effluent collector for gas chromatographic samples. This collector eliminates contamination of one component by another as well as contamination from other sources. The collector provides an individually complete assembly for each component, it pockets and magnifies the sample, and it has its own heat sink. Importantly, the sample is collected close to the end of the analytical column. Gas chromatography is being used increasingly for the analysis of volatile organic compounds and of those substances that can be made volatile. However, the collection of pure fractions of gas chromatographic samples from analytical columns presents a problem, particularly where several close-boiling components are involved. Contamination of one component by another can occur readily in the effluent line and at the exit port. The effluent line can also serve as a condensation chamber in which the column substrate may accumulate. Although several types of collectors are described in the literature (Grasselli and Snavely, 1962; Hajra and Radin, 1962; Haslam and others, 1961; and Lesser, 1959), all the collectors examined used the existing effluent line and exit port. Some collectors provided excellent trapping, but all were deficient in preventing contamination of one component by another. To avoid contamination of gas chromatographic samples, the samples must be collected at or near the the end of the analytical column and not at a common exit port supplied by a common effluent line. The collector should pocket the sample so that it can be readily withdrawn by a hypodermic needle. The collector should be relatively inexpensive, easily handled, individually cooled, and conveniently sealed from atmospheric moisture and coolant-bath vapor before and after collection. An especially designed collector that embodies these principles is shown in figure 155.1. The admission Figure 155.1.—Effluent collector. tube is 18-gage stainless-steel hypodermic tubing; it has a 17-degree bent point on one end and a square cut on the other. The part of the admission tube that extends beyond the rubber septum should be just less than the length of the common effluent line, so that when the tube is inserted, the septum will seal the exit port. Before and after use, the end of the admission tube can be sealed with an inverted septum if desired. D164 ART. 155 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D164-D166. 1964.GOERLITZ AND LAMAR D165 The carrier gas is vented through a hypodermic needle that is inserted in the discharge arm of the collector. The silicone-rubber septums are the standard half-drilled type that fit 6-mm ID tubing. The coolant-bath holder for each collector consists of a snap-on-cap polyethylene container 1%. inches in diameter by 2% inches high. Two holes and a connecting slit were made in each cap. An O-ring cut from rubber tubing positions the collector in the coolant-bath holder. The admission tube may be placed in the horizontal or vertical arm of the collector, depending upon the access to the detector vent. In order to reach the detector vent with an admission tube, the gas chromatograph must have a sufficiently straight common effluent line. In some gas chromatographs the end of the column is at the bottom of the instrument and the common effluent line is arranged so that it is not possible to reach the detector vent with an admission tube. However, these gas chromatographs can be modified so that the components can be collected from the bottom of the instrument. For example, a Perkin-Elmer Model 154-D gas chromatograph was modified by drilling a %-inch hole in the bottom of the oven cabinet. The hole was alined with the exit “tee” of the detector so that a 4-inch length of straight %-inch stainless-steel tubing attached to the exit “tee” just protruded outside the oven chamber. A 1%-inch length of %-inch copper tubing was inserted and crimped to line the hole and keep the insulation in place. The exit end of the K-inch stainless-steel tube was fashioned so that the entry was conical, to facilitate insertion of the admission tube. The inside of the exit tube was polished to a smooth finish with emery powder on a cotton-swab stick. The upstream end of the tube was swaged to the detector-outlet “tee.” Access to the new exit port was made by supporting the gas chromatograph cabinet on a frame. The original effluent line was removed. EFFLUENT COLLECTION PROCEDURE The collection apparatus is assembled and a suitable coolant is placed into the coolant receptacle. A sample is injected into the gas chromatograph, and just before the desired component emerges, the admission tube of the collector is started into the exit port of the chromatograph. At the time the recorder indicates the emergence of the component, the admission tube is inserted all the way to the rubber septum and held firmly in place. The uncontaminated component in the carrier gas passes through the individual admission tube and is condensed in the pocket of the glass collector. During waiting periods, the collectors and bath holders may be kept in a flat pan of crushed ice or on a slab of dry ice. EXPERIMENTAL RESULTS Methyl caproate was selected to demonstrate the suitability of the apparatus for collecting components with comparatively low boiling points, and the dimethyl esters of fumaric and malonic acids were used to show the effectiveness of the apparatus for the separate collection of two partially overlapping compounds. A 20-microliter sample containing 0.00023 g of methyl caproate in absolute methanol was injected into the gas chromatograph. Collection of the sample was begun when the recorder pen left the base line and ended when the pen returned to the line. The collected sample was transferred from the collector to a microcavity cell (0.2-0.3 /ul minimum volume) with a total of 5 yul of carbon disulfide by using a microliter syringe. Infrared analysis was made after the solvent was evaporated down to about 1 /ul under the gentle heat of an infrared heat lamp. Evaporation was halted by closing the cell with a drop of mercury. Another infrared scan was made after replacing the carbon disulfide with carbon tetrachloride by dilution and evaporation. A X 4 infrared beam condenser was used for these analyses. A 20-/nl aliquot of a solution containing 0.0124 g of dimethyl malonate and 0.0119 g of dimethyl fumarate in 1.00 ml of absolute methanol was injected into the gas chromatograph. As shown in figure 155.2 the resulting chromatograph is a good example of closely eluting, partially overlapping compounds. The collections were timed to eliminate the unseparated part of the elution. The overlapping part, comprising 20 percent of the dimethyl malonate and 40 percent of the dimethyl fumarate, was discarded by venting. The infrared spectrum of each of these components was obtained as described in the preceding paragraph except that only carbon tetrachloride was used with dimethyl fumarate. In addition, the dimethyl fumarate was collected for infrared analysis by the potassium bromide micropellet technique. RESULTS AND DISCUSSION Each component eluted by the gas chromatograph is collected close to the end of the analytical column in an individually complete assembly. An excellent infrared spectrum for methyl caproate was obtained from a single collection of an injection of only 0.00023 g of the ester. This demonstrated the application of this apparatus for the collection of components with comparatively low boiling points.D166 ANALYTICAL TECHNIQUES A separate collection of closely eluting and partially overlapping components was equally successful as demonstrated by the injection of a sample containing dimethyl malonate and dimethyl fumarate. Two injections were required because it was necessary to vent the overlapping part of the elution. Each of the 2 injections contained 0.00025 g of dimethyl malonate and 0.00024 g of dimethyl fumarate. From these injections, each component was collected separately and placed in a microcavity cell for infrared analysis. No contamination of one of these esters by the other could be observed in the infrared spectra. To test another handling technique, two collections of dimethyl fumarate were pressed in a micropellet of potassium bromide. Again, no contamination of this ester was observed in the infrared spectrum that was obtained with a X 4 beam condenser and X 5 scale expansion. An important feature of the apparatus described here is the completely individual system for the collection of each component very near the end of the column. Thus, contamination of one component by another as well as contamination from condensation of column substrate can be eliminated. This collector has several other distinct advantages. For example, the condensing surface can be easily washed with a small amount of solvent, and the thickened glass section magnifies and pockets the collected material. The collector can be completely sealed during cool-down or waiting periods, thus avoiding contamination by atmospheric moisture and cold-bath solvents. The collector is always maintained in its own heat sink that will not warm up before the collection is completed. This collector also enhances a collection and reinjection technique developed by the authors. With this technique the components are individually collected and reinjected into a different gas chromatographic column to obtain two characteristic retention volumes for each component. This provides a salient advantage in the identification and confirmation of the organic components in a sample. REFERENCES Grasselli, J. G., and Snavely, M. K., 1962, Analysis of organic reaction products by combined infrared-gas chromatography techniques: Appl. Spectroscopy, v. 16, p. 190-194. Hajra, A. K., and Radin,. N. S., 1962, Collection of gas-liquid chromatographic effluents: Lipid Research Jour., v. 3, p. 131-134. Haslam, J., Jeffs, A. R., and Willis, H. A., 1961, Applications of gas-liquid chromatography: Analyst, v. 86, p. 44-53. Lesser, J. M., 1959, Device for isolation of components separated by gas chromatography: Anal. Chemistry, v. 31, p. 484.Article 156 USE OF SODIUM-SENSITIVE GLASS ELECTRODES FOR SOLUBILITY DETERMINATIONS By A. H. TRUESDELL and C. L. CHRIST, Washington, D.C. Abstract.—A method applicable to concentrated solutions is described for converting activities, measured by cation-sensitive glass electrodes, to concentrations. The usual difficulty oi assigning accurate values to activity coefficients is avoided by the use of a dilution technique. The method is applied to the determinations of the solubilities of borax, ulexite, and sodium chloride. Following the initial discovery by Eisenman and others (1957) of practical sodium- and potassium-sensitive glasses, Garrels and others (1962) developed glasses specifically sensitive to divalent cations (Ca+2, Mg+2, Ba+2, Sr+2). An electrode fashioned from one of these glasses offers a convenient and rapid means of directly measuring cationic activity (conventionally defined), or of indirectly measuring concentration in aqueous solution. Although only the sodium-sensitive electrode was used to measure the solubilities of sodium salts, the same procedure could be used for any salt containing a cation for which an electrode having the proper response is available, for example, potassium or ammonium salts, or the alkaline-earth salts. Similarly, the solubility of a given salt in the presence of any ionic medium not affording a common cation could be determined. In general, the conversion of a measured ionic activity (ai) to a concentration (ct) requires the use of an accurate value of the appropriate ionic-activity coefficient (y(), where ci=ai/yt. The activity coefficient depends upon the total ionic strength of the solution. In saturated solutions of many readily soluble salts, such as the sodium borates, the concentration, and hence the ionic strength (I) is high, and the cation is complexed in an unknown way. As a result, there is no way to assign accurate values of cation-activity coefficients. We have evolved a technique that bypasses this difficulty. This technique involves making cation-activity measurements on successive ten-fold dilutions of a saturated solution of the salt whose solubility is to be measured. When the solution becomes very dilute, this fact is revealed by its electrochemical behavior. Because the solution is very dilute and any complexing will be vanishingly small, the cation-activity coefficient can be calculated from the Debye-Hiickel equation. The concentration of this dilute solution can be calculated, and because the number of dilutions is known, the original concentration can then be established. Using a sodium-sensitive electrode we have determined the solubilities of borax (at 25°C and 40°C), ulexite (at 25°C, incongruent solution), and sodium chloride (at 25°C). The values determined by the present method are in excellent agreement with those determined by conventional methods, as shown in table 156.1. PROCEDURE AND THEORETICAL BASIS The sodium-sensitive electrode used in this work was the Electronics Instruments Laboratory model Na GNA 23. This electrode was paired with a saturated-KC1 calomel reference electrode (wick type), and the developed potential was measured with a vibrating capacitor high-impedance electrometer, and a potentiometer. Potential measurements were made on solutions maintained at 25°C±0.05°C in a constant-temperature bath and stirred with a magnetic stirrer. The voltage measurements were generally accurate to ±0.1 mV, with an occasional reading doubtful to ±0.2 mV. Because the electrode is also sensitive to hydrogen ions, pH must be maintained >7; at sodium-ion concentrations below about 10~4 m (molal), the hydrogen-ion response of the electrode introduces error. The usual precautions were maintained with regard to adequate electrical shielding in working with the high resistance glass electrode (Mattock, 1961). The electrode was conditioned by soaking for several days in 0.1 m NaCl solution. ART. 156 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D167-D170. 1964. D167D168 ANALYTICAL TECHNIQUES Table 156.1.— Values of solubilities of sodium chloride, borax, and the sodium-ion concentration resulting from the incongruent solution of ulexite, determined by N^.-sensitive electrode, compared with literature values 1 This study Literature /* M 8 Weight Weight Reference percentage 4 percentage 4 Sodium chloride, 25°C 5. 40 5. 41 26. 4 26. 4 27. 0 26. 5 26. 48 Gillespie (1928, p. 105). 4. 62 4. 53 Density = 1.1978 g/cm3 (Gillspie, 1928, p. 105). Borax (as Na2B40:), 25°C 3. 04 3. 03 3. 05 3. 06 Menzel and Schulz (1940) 3. 13 Blasdale and Slansky (1939). Borax (as ^28407), 40°C 0. 308 6. 20 6. 00 Menzel and Schulz (1940), and Blasdale and Sian- sky. (1939). Ulexite (as NaCaBsO#), 25°C 0. 0187 5 0. 49 0. 0186 0. 49 1 The accuracy of the experimental results is best judged by internal consistency, and by comparison with the literature values. The limiting precision is determined by the ±0.1 mV reproducibility of the voltage measurements; an error of 0.1 mV leads to an error of about 0.4 percent in the solubility determination. 2 Weight dilution method. 8 Volume dilution method. 4 Grams of solute per 100 g of solution. 8 Incongruent solution; only sodium-ion concentration measured. The potential of the cell, E, is given by a Nernst-type equation E (volts)=E' + (RT/g’)ln[Nsi+1], (1) where R, T, and of have their usual significance. E' is a constant voltage and includes any liquid junction potential. E' has the value E at unit activity of sodium ion, [Na+1] = l. At 25°C, equation 1 becomes E=E' +0.0592 log [Na+1], (2) In practice, E' is not evaluated, but rather the difference in potential of the cell is determined with the unknown solution and with a solution whose sodium ion activity is known. This procedure minimizes the uncertainty due to the liquid-junction potential. From equation 2 we obtain the following operational equation: Ex-E,=0.0592 (log [Na+1L—log [Na+1],), (3) or (4) where x=unknown and s=standard. Equation 4 is of exactly the same form as the one on which the measurement of pH is based (Bates, 1954). The Na GNA 23 electrode was checked against a series of standard NaCl solutions (1 m to 10-4 m), and found to have exactly the theoretical slope predicted by equation 4, namely djE',/dpNas=0.0592 volts. The sodium molality of each standard solution was converted to sodium activity by the relation [Na+1]=7Ns+iXmNa+i, (5) where the activity coefficient yNa+i was obtained by the “mean-salt” method (Garrels, 1960), in which YNa+* —72±NaCl/7±KCl for ionic strength 7>0.1, and by the Debye-Huckel equation for /<0.1 (Klotz, 1950, p. 329). Values of Y±kci and 7±NaCi were taken from Lewis and Randall (rev. ed., 1961, p. 643), and Robinson and Stokes (1959, p. 492), respectively. The standard solution used in a given determination was either 0.1 m NaCl, with 7Na+i = 0.79 (mean-salt), pNa=1.102; or 0.01 m NaCl with 7Na+i=0.902 (Debye-Hiickel), pNa=2.045, or both. It was found that the solubility determinations were completely consistent with both of these single-ion activity coefficients, which attests to the validity of their value in practical problems. Saturated solutions of the salts were prepared by shaking the solutions in contact with excess solids, over a period of about 3 weeks. One method of doing this was to impart mechanically a rocking motion to a 125-ml rubber-stoppered flask containing solution plus solid, in a constant-temperature bath maintained at ± 1°C. In a second method the solution plus solids, contained in 125-ml Teflon bottles, were immersed in a constant-temperature bath held at ±0.1°C; the bottles were removed twice daily and vigorously shaken by hand. In the case of the mechanically shaken system, the grain size of the excess NaCl was sufficiently large that the supernatant liquid could be sampled directly with a pipet; this was not true with the less soluble borax and ulexite. In those experiments in which the bottles were shaken by hand all the solids remained finely crystalline, and the liquid was drawn by suction through a coarse-frit filter in order to remove small crystals. As previously stated, the voltage measurements were all made at 25°C. For the 25 °C solubility measurements, the potential of the cell was measured with the Ez—Es=0.0592 (pNas— pNa*),TRUESDELL AND CHRIST D169 saturated solution as the test solution. Following this, in one set of experiments, a 10-ml portion of the saturated solution was delivered to a calibrated 100-ml volumetric flask by means of a calibrated pipet, and diluted to 100 ml; the potential of this solution was then measured. This procedure was repeated stepwise to and including the 1:1,000 dilution. The volume-dilution procedure leads to concentrations in units of molarity (M, moles of solute per 1,000 ml solution). In a second set of experiments a weight-dilution technique was employed; this technique leads to concentrations in units of formality (/, moles of solute per 1,000 solution). Whether volume dilution or weight dilution was used in a particular determination is indicated in table 156.1. There is an internal formal inconsistency in the calculations because the values of the activity coefficients employed are based on the molal scale (m, moles of solute per 1,000 g solvent). However, for 7Na+i values from the mean-salt method, and especially for 7Na+i values calculated by the Debye-Hiickel method for the very dilute range, any changes resulting from a change in scale from molal to molar or formal would be outside the limits of error of the method. As the test solution becomes progressively dilute, with fewer sodium ions available to ensure stable equilibrium between the electrode surface and the solution, fluctuations in the voltage readings become larger. At the same time, the relative activity of hydrogen ions increases and the electrode begins to respond to these ions. For these reasons a decision must be made concerning the best dilution stage at which to calculate the concentration. This stage will depend upon the solubility of the salt, and will reveal itself through the magnitude of the measured activity. In general, it is best to use the highest concentration that shows the Nernst-type behavior of equation 2. In this study we choose the 1:1,000 dilution (of the saturated solution) of NaCl, the 1:100 dilution of borax, and the 1:10 dilution of ulexite. The tabulated values given by Klotz (1950, p. 332) for the Debye-Hiickel equation —log 7, Az2D 1+Ba,/*’ borate anion (Ingri and others, 1957), so that here also I=mNa+1. Ulexite, NaCaB509-8H20, in excess, in equilibrium with water at 25°C, yields a Na-rich borate solution and inyoite, Ca2B60nl3H20 (Kurna-kova and Nikolaev, 1948). Thus, the measurement here yields the sodium-ion concentration in equilibrium with excess ulexite and in the presence of inyoite. From the data of Kurnakova and Nikolaev (1948) on the Ca0-B203-H20 system, the solution in equilibrium with inyoite at 25°C is 1.32X10-2 M with respect to Ca+2. We assume as a rough guess that the presence of the common singly charged borate ion depresses this value to one-half in the case of the solution in equilibrium with both ulexite and inyoite, and calculate the ionic strength on this basis. In all cases, in calculating / it is first assumed that the measured activity is equal to the molality, and a provisional y, found. This provisional y( is divided into the activity to find the next approximation to the molality and a new yt is looked up, following which a new molality is calculated (see table 156.2); in this work, only two stages of iteration were necessary. Fortunately, yt changes relatively slowly as a function of I over the dilute-solution region, so that no appreciable error results in the calculated concentration in those cases where I cannot be evaluated precisely. Table 156.2.—Determination of solubility of NaCl at 25° C by volume-dilution method Ex-E,= 0.0592 (pNa„ —pNa*) E, = 41.0 mV for 0.1 m NaCl; pNa,= 1.102 1 E,— —14,9 mV for 0.01 m NaCl; pNas=2.045 1 Dilution stage Sat. sol. 1:10 1:100 1:1,000 Ex (mV) pNa* (0.1 m std.) 168. 2 — 1. 047 84. 3 0. 371 27. 0 1. 338 -30. 0 2. 301 For the 1:1,000 dilution „ -30.0-41.0 , oon, —pNa*=--------------1.102=-2.301 pNa*=2.301 1 -30.0 + 14.9 -pNa*=- 59.2 — 2.045=-2.300 pNa*=2.300 1 for 108«j=4, and for 25°C, were used to construct a plot of 7Na+i vs. 1; values of 7Na+i were read from this plot for the diluted solutions. The ionic strength, 7=22jm,2?, must be evaluated differently for each substance, depending upon the charges of the anions assumed to be present to balance the Na+1. For NaCl, /=mNa+i; for borax, in the absence of specific knowledge, it was assumed that each Na+1 is balanced by a singly charged [Na+1]=5.00X10-3; ti=0.927; ci = (5.00X10"3)/(0.927)=5.39 XIO-3; Y2 =0.925; c2 = (5.00 X 10-3)/(0.925) = 5.40 X 10~3; M= 5.40 » Third place after the decimal uncertain. RESULTS AND SUMMARY The final solubility values determined by the electrode method are listed in table 156.1. To show how the experimental procedure leads to a solubility value,D170 ANALYTICAL TECHNIQUES the details of one determination on NaCl are given in summary form in table 156.2. The investigation reported upon here was an exploratory one only, and had as its sole purpose the testing of the principle of the method; no real attempt was made to achieve the best possible accuracy. It seems clear that this dilution technique can be used to measure solubilities with a fair degree of confidence, even with simple techniques. To measure accurately the solubility of a substance at an elevated temperature, a better technique than that employed by us of separating solids from liquid would be needed. REFERENCES Bates, R. G., 1954, Electrometric pH determinations: New York, John Wiley and Sons, Inc., 331 p. Blasdale, W. C., and Slansky, C. M., 1939, The solubility curves of boric acid and the borates of sodium: Jour. Am. Chem. Soc., v. 61, p. 917-920. Eisenman, George, Rudin, D. O., and Casby, J. U., 1957, Glass electrode for measuring sodium ion: Science, v. 126, p. 831-834. Garrels, R. M., 1960, Mineral equilibria: New York, Harper and Brothers, 254 p. Garrels, R. M., Sato, M., Thompson, M. E., and Truesdell, A. H., 1962, Glass electrodes sensitive to divalent cations: Science, v. 135, p. 1045-1048. Gillespie, L. J., ed., 1928, Density (specific gravity) and thermal expansion (under atmospheric pressure) of aqueous solutions of inorganic substance and of strong electrolytes, in Washburn, E. W., and others, eds., International critical tables of numerical data, physics, chemistry and technology: New York, McGraw-Hill Book Co., Inc., v. 3, p. 51-111. Ingri, Nils, Langerstrom, G., Frydman, M., and Sillen, L. G., 1957, Equilibrium studies of polyanions. II. Polyborates in NaC104 medium: Acta Chem. Scand., v. 11, p. 1034-1058. Klotz, I. M., 1950, Chemical thermodynamics: Englewood Cliffs, N.J., Prentice-Hall, Inc., 369 p. Kurnakova, A. G., and Nikolaev, A. V., 1948, The solubility isotherm of the system Na20-Ca0-B203-H20 at 25°: Akad. Nauk. SSSR Izvestria, Otd. Khim. Nauk, no. 1, p. 377-382. [in Russian] Lewis, G. N., and Randall, Merle, 1961, Thermodynamics, revised by K. S. Pitzer and L. Brewer: New York, McGraw-Hill Book Co., 723 p. Mattock, G., 1961, pH measurement and titration: London, Heywood and Co., Ltd., 406 p. Menzel, Heinrich, and Schulz, Hans, 1940, Zur Kenntnis der Borsatiren und borsauren Alkalisalze. X. Der Kernit (Rasorit) Na2B407.4H20: Zeitschr. anorg. allg. Chem., v. 245, p. 157-220. Robinson, R. A., and Stokes, R. H., 1959, Electrolyte solutions: London, Butterworths Sci. Pubs.Article 157 SEMIMICRO X-RAY FLUORESCENCE ANALYSIS OF TEKTITES USING 50-MILLIGRAM SAMPLES By HARRY J. ROSE, JR., FRANK CUTTITTA, MAXWELL K. CARRON, and ROBENA BROWN, Washington, D.C. Abstract.—Semimicro determinations of the major constituents (Si02, AI2O3, total iron, K20, CaO, Ti02, and MnO) of tektites were made by X-ray fluorescence spectroscopy using 50-milligram samples. The X-ray analytical data are comparable to determinations obtained by conventional chemical techniques. X-ray fluorescence spectroscopy has been applied to the determination of the major constituents in relatively large samples of materials of geologic interest (Rose and others, 1962, 1963). This technique can also be used for the semimicro analysis of rare mineral specimens or of materials of astrogeologic interest where only very small quantities «100 mg) of material are available. The analysis of the light elements in tektites has been made possible by advances both in instrumentation and in techniques of sample preparation. It is well known that the intensity of the fluorescent radiation decreases sharply with a decrease in the atomic number (Z) of the element being analyzed. This decrease can be attributed to many factors such as diminishing fluorescence yield and absorption of the fluorescent radiation by air, by the sample, by the diffracting crystal, and by the window of the detector. Additionally, as the radiation becomes softer with decreasing atomic number, the depth from which the excited radiation emerges becomes shallower and shallower. Thus the effective radiation involves only those atoms at or near the surface of the sample, a situation that demands precise and reproducible sample preparation. The elimination of air from the spectrometer by either evacuation of the chamber or by flushing the chamber with helium provides the first necessary step to detecting the radiation. The use of supported thin-film windows (Balis and others, 1962) on the detectors has resulted in higher intensities by reducing absorption by the window and has extended the useful range of the technique to wavelengths beyond 10 angstroms. Accurate quantitative determination of the light elements depends primarily on sample preparation and ultimately on the surface of the specimen submitted to the X-ray beam. As mentioned above, the particles at or near the surface provide the effective signal for analysis. Infinite depth, the point beyond which no increase in signal can be observed regardless of increase in sample thickness, is reached for CaO in less than 100 microns and for MgO in less than 50n- It is evident, therefore, that the surface of the prepared specimen must be homogeneous and a true representation of all the layers beneath it. Fusion of the sample in a mixture of Li2B407 eliminates many of the problems inherent in the X-ray fluorescence analysis of powdered samples (Andermann, 1961; Rose and others, 1962), especially those related to particle size and crystal structure. The presence of a strongly absorbing element minimizes absorption differences among samples resulting from variations in matrix, thus obviating the need for absorption corrections or for standards matching the composition of the material under study. La203 was chosen because of its high absorption for the light elements. The method described initially (Rose and others, 1962) required a 250-milligram sample. Although this amount is considerably less than the quantity used for routine chemical analysis, there are instances when smaller quantities must be analyzed. The effort here was directed toward reducing the sample size to 50 mg by determining the minimum layer of fused sample necessary to maintain the desired signal for the elements being analyzed. 716-626 0—64 -12 ART. 157 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D171-D173. 1964. D171D172 ANALYTICAL TECHNIQUES X-RAY FLUORESCENCE METHOD The fusion mixture consists of 50 mg of sample, 50 mg La203, and 340 mg Li2B407. The components are mixed in a boron carbide mortar, transferred to a graphite crucible having a cone-shaped internal base, and fused at 1,100°C for 10 minutes. The bead is allowed to cool in the graphite crucible. Cooling may be accomplished more rapidly by placing the crucible on a large copper plate, which dissipates the heat more readily. Sufficient boric acid is added to bring the weight to 460 mg. The boric acid compensates for any losses during ignition and acts as a binder during preparation of the pellet. The glass bead is then ground. The grinding vial consists of a lucite cylinder provided with two caps containing tungsten carbide inserts for covering both ends of the cylinder. The bead, which must be crushed before grinding, is placed cone side up in the grinding vial with one of the caps and inserts in place. A %-inch drive pin punch is placed down the cylinder on top of the bead and is then tapped with a hammer. The weighed boric acid is added to the vial along with a %-inch tungsten carbide ball, and the upper cap is placed on the cylinder. The sample is then ground on a mixer grinder for 10 minutes, which reduces the sample to about 325 mesh. The ground powder is then pressed into a pellet 1 inch in diameter. For additional strength the pellet is prepared as a double layer, using boric acid as backing (Rose and Flanagan, 1962). It is essential that the sample layer be spread uniformly on the surface of the boric acid before final pressure is applied. To obtain maximum intensity for Si and Al, pressure in excess of 50,000 pounds per square inch must be used. The preparation of samples must be done consistently with attention to all details to insure the best results. A single-channel spectrometer was used for this study. The types of crystals, wavelengths, and detectors are given in table 157.1. Because the fluorescent radiation of elements Z < 22 is absorbed by air, the spectrometer chamber is flushed with helium for the determination of these elements. It is generally necessary to flush the chamber for about 1 minute between sample changes to allow the system to come to equilibrium. Granite G-l, diabase W-l, and an equal mixture of the granite and diabase are used as reference standards. National Bureau of Standards standard samples and other samples analyzed at the U.S. Geological Survey serve as additional reference materials to extend the range of the elements being determined. Several reference standards may be mixed in varying proportions to provide desired points on the calibration curve. The method has been used recently to analyze six samples of tektites from Java supplied by E. C. T. Chao. The results of both chemical and X-ray fluorescence determinations are given in table 157.2. Table 157.1.—Summary of crystals, detectors, and wavelengths Element A (A) Crystal Detector or counter 1 Path Si 7. 125 Gypsum __ Proportional . . He Al 8. 337 do do He Fe 1. 936 LiF 2 Scintillation . . Air Ca 3. 358 Eddt3 Proportional . He K 3. 741 Gypsum do._ He Mn 2. 102 LiF > Scintillation .. . Air Ti 2. 748 LiF2 . do He 1 Pulse-height analysis, channel width 12 V, base level 4.5 V. 2 Lithium fluoride. 3 Ethylene diamine ditartrate substituted for gypsum in samples with high K20 content. CHEMICAL ANALYTICAL METHODS Six javanites were carefully selected from a collection of about 80 specimens to represent the range of the indices of refraction and specific gravities of the Java tektite collection. The specimens are listed in table 157.2 in order of increasing index of refraction. The tektites were analyzed for Si02, A1203, total iron as Table 157.2.—Comparison of chemical and X-ray analyses of 6 tektites from Java [Index of refraction (IV) in sodium light; X-ray determination made on a single pellet] JS-12 JS-9 JSS JSS JSS JS-U Chem. X-ray Chem. X-ray Chem. X-ray Chem. X-ray Chem. X-ray Chem. X-ray Si02 74. 2 74. 4 73. 4 73. 5 73. 8 74. 2 74. 4 72. 7 72. 3 72. 3 72. 5 72. 3 11. 1 11. 6 11. 4 11. 3 11. 3 11. 5 11. 2 11. 3 11. 3 11. 3 11. 4 11. 6 Fe2(V 5. 67 5. 70 5. 63 5. 65 5. 28 5. 35 5. 52 5. 66 6. 14 6. 28 6. 17 6. 22 CaO_. 2. 25 2. 29 2. 70 2. 70 2. 75 2. 65 2. 36 2. 47 2. 91 2. 97 2. 79 2. 87 K20 2. 24 2. 23 2. 28 2. 27 2. 25 2. 25 2. 29 2. 32 2. 22 2. 22 2. 17 2. 15 MnO . 11 . 10 . 10 . 11 . 09 . 10 . 10 . 11 . 11 . 10 . 11 . 11 Ti02 . 67 . 68 . 68 . 68 . 67 . 66 . 69 . 69 . 70 . 70 . 69 . 68 N 2 1. 5073 1. 5089 1. 5102 1. 5112 1. 5127 1. 5141 Sp gr. 2 3_. . 2. 443 2. 464 2. 439 — 2. 426 ____ 2. 469 — 2. 483 1 Total Fe as Fe203. 2 By Janet Marteka and E. C. T. Chao. 3 Third place after the decimal not certain.ROSE, CUTTITTA, CARRON, AND BROWN D173 Fe203, CaO, K20, Ti02, and Mn02. The chemical results are averages of duplicate determinations by the various methods. Using these different techniques, the chemical determinations were closely monitored by similar determinations in granite G-l, diabase W-l, and selected National Bureau of Standards certified samples. Silica was determined spectrophotometrically using a molybdenum-blue method (Bunting, 1944), gravi-metrically by a volatilization-formaldehyde method described by Carron and Cuttitta (1962), and also by a combined gravimetric and photometric procedure (Jeffery and Wilson, 1960). Alumina (A1203) was determined by measuring the absorbance of the calcium aluminum alizarin red-S complex in a weakly acidic medium (pH 4.5) at 485 him (Parker and Goddard, 1950). Interference from iron was eliminated by use of potassium ferricyanide and thioglycolic acid as complexing agents. Total iron was determined spectrophotometrically with o-phenanthroline (Cuttitta, 1952, and Sandell, 1959) and by a magnetic-susceptibility method developed by Thorpe and others (1963). Calcium oxide was determined by flamephotometry (Kramer, 1957), and by a semimicro, automatic, photometric titration with EDTA in the pH range 12.1-12.3 at 590 m/i using murexide as the indicator. Potassium oxide was determined flamephotometrically at 766 ni/x using an instrument with a photomultiplier attachment. The sample was analyzed by bracketing between the closest potassium standards (Ray, 1956; Willgallis, 1957; Voimovitch and Debras, 1958). Titanium was determined spectrophotometrically with disodium-1, 2-dihydroxybenzene-3, 5-disulfonate (tiron) (Yoe and Armstrong, 1947). The interference of iron was overcome by reduction with ascorbic acid at pH 4.7. The purple permanganate color was utilized for the spectrophotometric determination of manganese. The oxidation was effected with ammonium persulfate (peroxysulfate) at the boiling point in a phosphoric-nitric acid medium in the presence of silver nitrate. Manganese was also determined by Janet D. Fletcher using a quantitative spectrographic method similar to that described by Bastron and others (1960). REFERENCES Andermann, George, 1961, Improvements in the X-ray emission analysis of cement raw mix: Anal. Chemistry, v. 33, 1689-1695. Balis, E. W., Bronk, L. B., Pfeiffer, H. G., Welborn, W. W., Winslow, E. H., and Zemany, P. D., 1962, Improved components for the X-ray emission analysis of the light elements: Anal. Chemistry, v. 34, p. 1731-1733. Bastron, H., Barnett, P. R., and Murata, K. J., 1960, Method for the quantitative spectrochemical analysis of rocks, minerals, ores, and other materials by a powder d-c arc technique: U.S. Geol. Survey Bull. 1084-B, p. 165-182. Bunting, W. E., 1944, The determination of soluble silica in very low concentration: Indus, and Eng. Chemistry, Anal. Ed., v. 16, p. 612-615. Carron, M. K. and Cuttitta, Frank, 1962, Determination of silica in tektites and similar glasses by volatilization: Art. 30 in U.S. Geol. Survey Prof. Paper 450-B, p. B78-B79. Cuttitta, Frank, 1952, The colorimetric determination of total iron with o-phenanthroline—A spectropotometric study: U.S. Geol. Survey TEIR-223 issued by U.S. Atomic Energy Comm. Tech. Info. Service, Oak Ridge, Tenn. Jeffery, P. G., and Wilson, A. D., 1960, A combined gravimetric and photometric procedure for determining silica in silicate rocks and minerals: Analyst, v. 85, p. 478-485. Kramer, H., 1957, Flame photometric determination of calcium in phosphate, carbonate, and silicate rocks: Anal. Chim. Acta, v. 17, p. 521. Parker, C. A., and Goddard, A. P., 1950, The reaction of aluminum ions with alizarin-3-sulphonate, with particular reference to the effect of calcium ions: Anal. Chim. Acta, v. 4, p. 517-536. Ray, Norman, 1956, Flame photometric determination of sodium, potassium, calcium, magnesium, and manganese in glass and raw materials. Anal. Chemistry, v. 28, p. 34. Rose, H. J., Adler, Isidore and Flanagan, F. J., 1962, Use of La203 as a heavy absorber in the X-ray fluorescence analysis of silicate rocks: Art. 31 in U.S. Geol. Survey Prof. Paper 450-B, p. B80-B82. ------- 1963, X-ray fluorescence analysis of the light elements in rocks and minerals: Appl. Spec., v. 17, no. 4, p. 81-85. Rose, H. J., and Flanagan, F. J., 1962, X-ray fluorescence determination of thallium in manganese ores: Art. 32 in U.S. Geol. Survey Prof. Paper 450-B, p. B82-B83. Sandell, E. B., 1959, Colorimetric determination of traces of metals, 3d ed.: New York, Interscience Publishers, Inc., p. 537-542. Thorpe, A. N., Senftle, F. E., and Cuttitta, Frank, 1963, Magnetic and chemical investigations of iron in tektites: Nature, v. 197, p. 836-840, March 2, 1963. Voimovitch, I. A., and Debras, J., 1958, Determination of sodium, potassium, and lithium in silicates by flame photometry: Indus. Ceramics, v. 502, p. 321-27. Willgallis, A., 1957, Application of flame photometry to alkali analysis of minerals: Zeitscht. anal. Chemie, v. 157, p. 249-257. Yoe, J. H., and Armstrong, A. R., 1947, Colorimetric determination of titanium with disodium-1, 2-dihydroxybenzene-3, 5-disulfonate: Anal. Chemistry, v. 19, p. 100-102.Article 158 DETERMINATION OF TOTAL IRON IN HEMATITIC IRON ORES BY X-RAY FLUORESCENCE SPECTROMETRY By WILLIAM W. NILES, Denver, Colo. Abstract.—A suite of hematitic iron-ore samples was analyzed by X-ray fluorescence spectrometry. Computations were made to derive an empirical equation from which the amount of iron in a sample can be calculated. The changes from true values caused by interfering elements are compensated for by simple calculation. Comparisons with wet chemical analyses are shown A suite of iron-ore samples was analyzed to cheek the reliability of a field method (Sheldon, 1964) for estimating iron content of hematitic iron ores. X-ray fluorescence spectrometry was selected because of its relative simplicity, rapidity, reasonable accuracy, and the wide range of elements that can be analyzed without much additional work on the samples. It was found that analyses for iron obtained by X-ray spectrometry were comparable in accuracy to the classical volumetric methods. The X-ray spectroinetric analyses were made using a Norelco 50 KV-50MA X-ray spectrometer with helium-path attachment and an FA-60 X-ray tube with molybdenum target. A lithium fluoride analyzing crystal, a primary collimator tube % inch square by 4 inches long, and a secondary collimator 4 inches long with 0.005-inch Soller slits were used exclusively. A flow proportional counter using P-10 gas and operated at 1,550 volts was used as a detector. All samples were crushed to inch fragments or smaller in a small jaw crusher and then were ground to 80-100 mesh with a mechanical grinder fitted with ceramic plates. A chart scan of each sample, before preparation, was made at the rate of 1° of arc per minute in order to estimate roughly the concentration of elements that might interfere with the iron determination through absorption effects. These estimates were made by comparison with an appropriate standard made up with an Si02 matrix. In these iron-ore samples, calcium and silicon were the only interfering elements present in appreciable quantities. None of the samples contained more than 25 percent calcium. All samples were prepared as follows: Enough CaC03 was added to a 0.1500-g sample to make the total calcium equal to approximately 25 percent of the sample by weight (0.0375 g calcium). Si02 was added to bring the total weight of the mixture of sample, CaC03 and Si02, to 0.3000 g. Boric acid as a noninterfering diluent was then added to obtain a total weight of 1.500 g. Each sample mixture was ground to an impalpable powder in an agate pestle and mortar under acetone, allowed to dry while grinding, and then pressed at 30,000 lbs. for 10 seconds into a pellet with a base of boric acid. Although the presence of silicon reduced the intensity of the Fe peak, Si02 was made the predominant matrix-component in the prepared sample. The Fe standards were made up as Fe203 in Si02. The effect of the small differences between the amounts of Si02 in the samples and the amount in the standards was negligible because of the predominance of Si02 as a matrix component. However, one must correct for absorption due to calcium and silicon. Lachances’ (written communication, 1960) correction factors, equations 1 and 2, and his expression for the iron content, equation 3, were modified as follows: r/ ______‘ i -f^Ca___ /. \ Ca BcaXpercent Ca F — /OX 81 BsiXpercent Si Percent Fe = 74'-|-(.BCaX/'caXpercent Ca) -f^(B'siXES1Xpercent Si), (3) where FCa=factor for absorption due to Ca, FSi=factor for absorption due to Si, A = actual percentage of Fe in standard, ART. 158 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D174-D175. 1964. D174NILES D175 B=apparent percentage of Fe in same standard with matrix added, j5Ca=apparent percentage of Fe in same standard with Ca added, 2?Si=apparent percentage of Fe in same standard with Si added, B' = apparent percentage of Fe in sample, effected by matrix, 5,ca= apparent percentage of Fe in sample, effected by Ca, and B'Si= apparent percentage of Fe in sample, effected by Si. By substituting equations 1 and 2 in 3, there is obtained Percent Fe=gH(go.X[A-fic,]^perc.nt Q) zfcaX percent Ca (.B'siXt-'d—-Bsi] Xpercent Si)^ -BSiXpercent Si and upon simplification this equation becomes Percent Fe=^+^[ii-^) + (^,,X[ilTgg]. •Oca B si Since the effect of the matrix is equal to the sum of the effects of each component in the matrix, this equation can be rewritten into the following form: Percent Fe=^ + (^X[^-B]) = i?, (1+Mzffl). n n __jg The factor —jy— is practically constant for all percentages of iron sought, and it is constant regardless of instrumental variations caused by normal fluctuations of voltage and temperature from day to day. Although this suite of iron-ore samples has a comparatively simple matrix composition, the method of computation has been used in analyzing samples containing as many as five interfering elements. The method has the advantage of allowing use of the same set of standards for rock and ores of widely varying matrices. Tables 158.1 and 158.2 show the accuracy and reproducibility of analytical results obtained by X-ray emission spectrometry of hematitic iron ores. The accuracy is based on the assumption that the chemical results are the absolute values. The chemical analyses were made by the standard dichromate volumetric oxidation-reduction method. Only one determination was made on each sample by each method. X-ray results are given to hundredths of a percent for comparison but should not be considered accurate beyond two figures. The reproducibility of results of the method was determined by analyzing each of the 10 samples 3 times. Table 158.1.—Comparison of results obtained by chemical and X-ray spectrographic methods (l) (2) (3) (4) Sample No. Fe (percent) by X-ray method Fe (percent) by wet chemical method 1 Absolute difference Id)-(2)1 Absolute percentage variation 100 (3)/(2) J12-13 2. 70 2. 65 0. 05 1. 89 J12-8 5. 25 5. 27 . 02 . 38 J12-20b 10. 86 10. 86 . 00 . 00 J3-30-1 15. 74 15. 79 . 05 . 32 J3-30-3 18. 15 18. 23 . 08 . 44 J12-28b 21. 47 21. 65 . 18 . 83 J3-33b 26. 93 26. 80 . 13 . 49 J12-40h _ ...... 31. 59 31. 50 . 09 . 29 J3-32f 34. 93 34. 26 . 67 1. 96 J3-32i 44. 11 44. 78 . 67 1. 50 Average. . . . . . 81 1 Analyst: Dwight L, Skinner, chemist, U.S. Geological Survey, Denver, Colo. Table 158.2.—Reproducibility of analytical results for iron by X-ray emmission spectrography (i) (2) (3) (4) (5) (6) Sample No. Fe (percent) 1st run Fe (percent) 2d run Fe (percent) 3d run Average Absolute mean deviation | (4) -(1)|+1(4) — (2) |+| (4) —(3)| /3 Absolute percent- age variation (5)/(4) J12-13 2.70 2.69 2.62 2.67 0.03 1.12 J12-8 5.25 5.44 5.37 5.35 .07 1.31 J12-20b 10. 86 10.91 11.02 10. 93 .06 .55 J3-30-1 15.74 15.31 15. 72 15. 59 .19 1.22 J3-30-3. 18.15 19.07 19.17 18.80 .43 2.29 J12-28b 21.47 22.78 22.66 22. 30 .56 2. 51 J3-33b 26.93 27.00 27.23 27. 05 .12 .44 J12-40h 31.59 31.86 32.04 31.83 .16 .50 J3-32f__. 34.93 35.16 34.35 34. 61 .38 1.10 J3-32i_ 44.11 43.49 44.09 43.90 .27 .62 1.17 REFERENCE Sheldon, R. P., 1964, Relation between specific gravity and iron content of rocks from the Red Mountain Formation, Alabama: U.S. Geol. Survey Bull. 1182-D. [In press]Article 159 A GAS JET FOR D-C ARC SPECTROSCOPY By A. W. HELZ, Washington, D.C. Abstract.-—A gas jet adaptable to enclosed arc-spark stands is described for use in d-c arc spectrochemical analysis. When used with a nitrogen-free gas, the jet is very effective for suppression of cyanogen bands. The gas jets for spectrochemical analysis described by Annell and Helz (1961) proved to be very effective for arc stabilization and cyanogen-band depression. However, they were designed for open-type arc stands and are difficult to use in spectrographic assemblies with conventional enclosed arc-spark stands. For enclosed stands it was considered desirable to use prefabricated electrodes of standard 1.5-inch length. Figure 159.1 is a drawing of the jet described herein. To minimize the required electrode length, the ceramic cap forming the nozzle, A (fig. 159.1 section XX') was shortened. To compensate for this loss of control of the gas flow and to try to insure lamellar flow, the internal construction was redesigned as described below. The control gas enters a lower chamber (fig. 159.1, C) surrounding the electrode, goes up through a ring of small holes (each 0.062 inch in diameter and 0.125 inch long) into an upper conically topped chamber (fig 159.1, B), and finally escapes through the annular space formed by the sample-bearing electrode and the ceramic cap. The jet was made in the U.S. Geological Survey analytical laboratory shop by J. B. Beasley. The main body of the jet is a ^-inch-thick brass plate supported on an electrically insulated post. Self alinement with the arc-stand jaws is augmented by freedom of the frame to swing on the post and by affixing the post to the arc-stand base loosely to permit some translational motion. The plate is effectively water cooled with a U-shaped channel. This jet is for use with standard electrodes 1% inches long and 0.242 inch in diameter. Thus, if the bottom hole of the brass plate and the central hole in the brass section that separates chambers B and C are each 0.250 inch in diameter, the electrode clearance will be 0.004 inch. At least this much clearance is advisable for convenience in inserting electrodes from above. The ceramic cap of the jet is a commercially available part used in tungsten-inert gas welding. The number 6 Linde Heliarc cup (Linde Co., division of Union Carbide Corp.) is suitable for use with the %-inch-diameter electrodes. The ceramic cup is shortened from 1.25 inches to 0.72 inch by cutting off the cylindrical tip at the outside circle where the cylindrical and conical parts of the cup meet. This leaves a short cylindrical section inside about 0.062 inch high and 0.375 inch in diameter. The annular space for the escaping gas is 0.066 inch thick. The assembled height of the jet is 1 inch. In use the lower electrode height is frequently adjusted to keep the burning rim of the crater a constant distance above the rim of the ceramic cup. About 2-millimeters seems optimum for this distance. This jet is much more compact than the earlier design of Annell and Helz (1961), but it is just as effective in producing a very steady arc and controlling cyanogen band interference. To show the effectiveness of the jet, 10 spectra of the cyanogen-band region are shown in figure 159.2. In addition to the spectrum of the arc in still air, spectra are also shown for the jet described herein, the Stallwood jet (Stallwood, 1954), and the enclosed Stallwood jet (see Spex Industries, Inc., 1962, for a summary status of Stallwood jets) at gas flows of 5, 10, and 15 cubic feet per hour. The control gas used was a mixture of 80 percent argon and 20 percent oxygen. For each spectrum shown in figure 159.2, 10 milligrams of a powdered rock was mixed with 20 mg of graphite and burned in a 15-ampere d-c arc. The spectra were recorded on an SA-1 plate using 38-percent transmission and 1-minute exposures. These first-order spectra have a reciprocal linear dispersion of approximately 5 A/mm. The difference between the type of gas flow for the jet described in this article and the gas flow of the Stallwood jet is strikingly illustrated by the spectra D176 ART. 159 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D176-D178. 1964.HELZ D177 Figure 159.1.—Jet design. Overall dimensions of the brass base plate, 3V/sX'A inch. In section XX' the control gas enters compartment C, passes through the ring of small holes to compartment B, and then out through the top of the ceramic cap A. D is the supporting post.D178 J IMI i i I Htl G III F III E III D III C III B III A II I I I o CTl ANALYTICAL TECHNIQUES I III >11 (III III II III III nil III III III! IIIHHHHIHMI M II H nii^ il M mi Hi ill ll If II nil min Min iiHiini!ii mi iiiiiiiniiiii n i in in ii i II 111 11IIIIIIIMHBBBB1 II 1 III 1 1 III! Ill 1 1- ! J J III III 1II II 1 III INI ill 1 • 1 1 II 1 III mi i 1 1 II 1 INI IIIII' 1 1 C\J Figure 159.2.—Spectra of 15-ampere d-c carbon arcs, cyanogen-band region. A, arc in air. B-D, enclosed Stallwood jet, 5, 10, and 15 cubic feet per hour, respectively, of a mixture of 80 percent argon and 20 percent oxygen. E-G, open Stallwood jet, 5, 10, and 15 cubic feet per hour, respectively, of argon and oxygen mixture. H-J, jet described in the report; 5, 10, and 15 cubic feet per hour, respectively, of argon and oxygen mixture. for different rates of flow. For the jet described herein the intensity of the cyanogen bands decreases with increasing gas flow (spectra H, I, J). For the Stall-wood jet, intensity increases with the gas flow (spectra E, F, and G). For the enclosed Stallwood jet, though at a much lower level, intensity decreases as the gas flow is increased from 5 to 15 cfh (spectra B, C, and D). Cyanogen bands are absent in spectra D and J; that is the gas flow is 15 cfh with either the enclosed Stallwood jet or the jet described in this article. Thus the effec- tive exclusion of cyanogen-band spectra in carbon arcs is obtainable without cumbersome enclosures of the arc. REFERENCES Annell, C. S., and Helz, A. W., 1961, A constant-feed direct-current arc: U.S. Geol. Survey Bull. 1084-J, p. 231-251. Spex Industries, Inc., 1962, Spex Speaker VII, no. 3: Scotch Plains, N.J. Stallwood, B. J., 1954, Air-cooled electrodes for the spectrochemi-cal analysis of powders: Optical Soc. America Jour., v. 44, 171 p.Article 160 DOLOMITE SOLUBILITY IN GROUND WATER By IVAN BARNES and WILLIAM BACK, Washington, D.C. Abstract.—The ion-activity product (IAP) d of dolomite has been computed for 87 samples of ground water from a variety of geologic environments. The upper limit of the (IAP)d for samples in or very near equilibrium with calcite agrees with the higher equilibrium constants reported in the literature. Detailed studies of the composition of ground water have been undertaken in an effort to determine the thermodynamic state of reactions between minerals and water. The approach has been to obtain chemical analyses reflecting as closely as possible the composition of the water in its native state by determining alkalinity, pH, and temperature at the time of sampling (Barnes, 1964). Ion activities of the chemical constituents can be calculated from the laboratory data and the Debye-Hiickel equation. From the activities, the ion-activity product (IAP) can be calculated for a particular reaction, using the law of mass action. Departure from equilibrium is determined by comparing the IAP calculated from the chemical analyses of water with the equilibrium constant, K (Back, 1960). If the IAP is greater than the K, the water is supersaturated with respect to the mineral of interest. For calcite, IAPc=ac a+2-<*co-2, (1) and for dolomite, IAPd=a Ca + 2-Q:Mg+2(Q!CO"2)2 (2) Figure 160.1 shows values of IAPd and the ratio IAP/K for calcite. The pH, temperature, and bicarbonate concentrations for all these samples were determined in the field; the ionic strength of all samples plotted is less than 0.1 the limit for the Debye-Hiickel equation. For samples in equilibrium with calcite, that is, those with an IAP/K ratio of 1.0±about 0.12, the data show that the maximum IAP for dolomite is 2.87X10-17 (see shaded area in fig. 160.1). The data indicate that the > Barnes (unpublished analyses, 1962). 2 J. W. Crooks (oral communication, 1962). s Back (1963). * Seaber (1962). »Jones (1961). Figure 160.1.—Graph showing relation ot ion-activity product for dolomite and the IAP/K ratio for calcite. Shaded area represents samples in equilibrium with calcite. upper limiting value for the IAP for dolomite in ground waters in equilibrium with calcite lies between approximately 1.5X10-17 and 3X10-17. Owing to the sluggishness of the dolomite precipitation, water supersaturated with calcite probably is supersaturated also with dolomite. These IAP data for dolomite may be compared with reported equilibrium constants for dolomite, as follows: Holland others 1___________________ 1X10 17 Kramer (1959)_______________________ 1.5X10_!7 Hsu (1964)____________________________ 2X10"17 Van Tassel (cited by Halla, 1962)__________________ 2.5X10 17 1 Holland, H. D., Kirsipu, T. V., Huebner, ,T. S., and Oxbough, V. M., 1962, On some aspects of the chemical evolutiDn of cave waters: Princeton Univ., technical report (duplicated). ART. 160 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D179-D180. 1964. D179D180 GEOCHEMISTRY OF WATER Halla and Ritter (cited by Kramer, 1959)______________________________ 3X10-17 Robie (cited by Halla, 1962)________ 6.6X1O-10 Garrels and others (1960)___________ 4.7X10-20 If the lower solubilities of Garrels and Robie are accepted, all water samples used in this study are supersaturated with dolomite. On the other hand, if we postulate that dolomite reacts with aqueous solutions and tends toward an equilibrium, the equilibrium constant would appear to be on the order of 2 to 3X10~17. In any event, the observed characteristics of the ground waters sampled are closer to the properties predicted at equilibrium with the higher solubilities reported than with the lower solubilities. REFERENCES Back, William, I960, Calcium carbonate saturation in ground water from routine analyses: U.S. Geol. Survey Water-Supply Paper 1535-D, 14 p. Back, William, 1963, Preliminary results of a study of the calcium carbonate saturation of ground water in central Florida: Internat. Assoc. Sci. Hydrology, v. 8, No. 3, p. 43-51. Barnes, Ivan, 1964, Field measurements of alkalinity and pH: U.S. Geol. Survey Water-Supply Paper 1535-H. [In press] Garrels, R. M., Thompson, M. E., and Siever, Raymond, 1960, Stability of some carbonates at 25° C and one atmosphere total pressure: Am. Jour. Sci., v. 258, p. 402. Halla, Franz, 1962, The free energy of the formation of dolomite from its carbonate components: Sedimentology, v. 1, p. 191. Hsu, K. ,1., 1964, Solubility of dolomite estimated on the basis of the chemical composition of Florida ground-waters, in Abstracts for 1963: Geol. Soc. America Spec. Paper. [In press] Jones, B. F., 1961, Zoning of saline minerals at Deep Spring Lake, California: Art. 83 in U.S. Geol. Survey Prof. Paper 424-B, p. B199-B202. Kramer, J. R., 1959, Correction of some earlier data on calcite and dolomite in sea water: Jour. Sed. Petrology, v. 29, p. 465. Seaber, P. R., 1962, Cation hydrochemical facies of ground water in the Englishtown formation, New Jersey: Art. 51 in U.S. Geol. Survey Prof. Paper 450-B, p. B124-B126.Article 161 EFFECT OF TREE LEAVES ON WATER QUALITY IN THE CACAPON RIVER, WEST VIRGINIA By KEITH V. SLACK, Washington, D.C. Abstract.—Accumulation of leaf litter in pools at low flow increases water color and results in marked changes in water composition. Temperature decreases rapidly from surface to bottom in the shallow pools. Dissolved-oxygen concentration decreases, but concentrations of many other solutes increase directly with water color. A film, believed to be ferric hydroxide, occurs on the surface of pools in which decomposition of litter is especially intense. Flushing of the pools by a rise in stream stage reduces the color and improves water quality. Changes in water composition resulting from natural additions of organic matter were investigated as part of a study of biological influences on water quality of streams. Previous work in Indiana indicated that the autumnal accumulation of tree leaves in streams coincident with the annual low-flow period occurs widely and is important in cycling of elements in nature. This accumulation is unusual in that it represents a natural environment in which decomposition greatly exceeds production of new organic matter. The environment is marked by low dissolved-oxygen concentration and considerable water color, which presumably is due to decomposition of plant material. Although of short duration, the condition seriously degrades water quality and can result in the death of aquatic organisms. The literature contains many references to the effects of leaf litter on the chemical quality of streams; for example, Schneller (1955) and Slack (1955) reported that the decay of leaves in pools generally resulted in greater water color and in the formation of free CO2 and lower dissolved-oxygen concentrations. Sylvester (1959) reported that replacement of native coniferous forests with deciduous trees in the Wenatchee River basin, Washington, resulted in increased water color. Hynes (1960) described Huet’s (1951) work, which indicated that a substance toxic to fish was derived from needles of spruce and red cedar. Chase and Ferullo (1958) showed that leaves deplete dissolved oxygen under ART. 161 IN U.S. GEOL. SURVEY PRO aerobic conditions in the laboratory; after 386 days, maple leaves had consumed a weight of oxygen equivalent to 75 percent of their initial dry weight. In the same length of time, oak leaves and pine needles exerted oxygen demands of about 50 percent of their dry weight. Myers (1961) suggested that manganese in San Clemente Reservoir, Calif., was leached from oak trees in the watershed. Robinson and others (1958) showed that leaves of hickory trees concentrated rare-earth elements to a remarkable degree. Thus, various species of trees may make widely differing chemical contributions to streams through leaf litter. Discolored water is neither a perennial phenomenon nor does it occur in all streams. Frequent rains during the peak leaf-fall period may flush the organic load downstream and dilute substances leached from the fresh leaves. The most favorable conditions for development of discolored water are a pooled stream with little or no surface flow and with stands of deciduous trees near the channel. Discoloring of water begins when tree leaves begin to drop, usually in September. The freshly shed leaves accumulate at the downstream end of pools, where they float on the surface at first but gradually sink to form a blanket over the streambed; later arrivals replenish the surface layer. Eventually the lower ends of pools become filled with a waterlogged mass of leaves that may inhibit flow over riffles. Unless there is a rise in stage and a flushing and exchange of water in the pool, a critical period is established. As the waterlogged leaves decompose by bacterial action, the supply of dissolved oxygen in the pool is depleted. Although decomposition of the leaves causes an increase in the mineral content of the water, the amount of the increase attributable to factors such as leaching of soluble substances from leaves, concentration by evaporation, mineral-matter release from sedi- PAPER 475—D, PAGES D181-D185. 1964. D181D182 GEOCHEMISTRY OF WATER ments by organic complexing and chemical reduction, and bacterial breakdown of leaf tissue is not known. The literature indicates that all these processes must be considered. In early stages of decomposition the color of the pool water closely resembles that of bog water. Under extreme conditions, the water acquires an inky appearance, which gives rise to the name “black water” in parts of the Midwest where discoloration is common. a moderately colored pool (d) was sampled as part of a temperature study on October 30, 1961 (table 161.3). The reach of river studied was shaded by a hardwood forest consisting of the following kinds of trees: maple, cottonwood, tulip tree, blue beech, sycamore, elm, dogwood, aspen, and oak. Many kinds of deciduous leaves were visible in the water on October 22, 1961, the date of the initial sampling, and most had sunk to the bottom. STUDY OF THE CACAPON RIVER, W. VA. Chemical composition A reach of the Cacapon River, Morgan County, W. Va., about 14 stream miles above the mouth where a series of elongate wooded islands extend parallel to the river bank, was selected for detailed study. At times of low flow, the channel which normally separates these islands from the right bank of the river becomes dry or is reduced to a few pools. Discolored water was general in the pools and in partly isolated marginal waters along this reach in October 1961. Three different environments, shown on figure 161.1, were sampled: (a) a riffle of the Cacapon River about 18 centimeters deep; (b) a slightly colored pool or backwater channel, which was connected with the river and in which the intensity of color increased with distance from the river, sampled at the upper end, about 5 meters from its junction with the main stream; and (c) an isolated highly colored pool about 6KX3 m in area, at the base of a steep bank. In addition, Figure 161.1—Sketch showing sampling sites on the Cacapon River, Morgan County, W. Va. A, riffle on the river; B, slightly colored pool; C, highly colored pool; and D, moderately colored pool. The chemical composition of the water of the river and the pools is shown in tables 161.1 and 161.2, which permit comparison between sampling sites on a single date and on different dates. The tables show that concentrations of many substances tend to increase as water color increases, although calcium, bicarbonate alkalinity, dissolved solids, and specific conductance, all were higher in the slightly colored pool (B) than in the highly colored pool ((7). The appearance of the ground on October 22 and 29 suggested that water from the main channel had recently flowed into the highly colored pool (C), although the composition of water of the channel and the pool differed greatly. Pool C was separated from the river by a low mound of wet leaves and rocks about 1 m in width. Wet masses of a green alga (Oedogonium) covered much of this barrier; however, the fresh appearance of the algae must have resulted from a light rain recorded on October 21, because streamflow records show no significant rise of the river during October. The low calcium and carbonate concentrations of the highly colored pool (C) could be explained in various ways; for example, much of its water could have been derived from local runoff or dilute ground water containing alkalies leached from the leaves. Another possibility is that ion-exchange reactions between the leaf material and the pool water produced water enriched in alkalies but depleted in calcium and bicarbonate (Hutchinson, 1957, p. 573). Still another explanation is that snails (Gyraulus) inhabiting this isolated pool extracted calcium carbonate from solution for shell formation, thus causing a lower calcium concentration. The volume of the highly colored pool was roughly 2,000 liters, indicating a deficit of 18.5 milligrams of calcium per liter compared with the calcium content of the other waters sampled, or a total loss of 37 grams. If this quantity were spread among an estimated population of 100 snails, the necessary uptake of 0.37 g calcium or 0.92 g calcium carbonate per animal would seem excessive for this thin-shelled form. However, without better population data the possibility of biological utilization cannot be excluded as a factor causing depletion of calcium carbonate in the highly coloredSLACK D183 Table 161.1.—Water properties at three sampling sites, Cacapon River, W. Va., during autumn 1961 1 [Analyses by H. R. Feltz, Washington, D.C.] October 22 October 29 November 19 Main stream (A) Slightly colored pool (B) Highly colored pool (C) Main stream (A) Slightly colored pool (B) Highly colored pool (C) Main stream (A) Slightly colored pool (B) Highly colored pool (C) Color (platinum-cobalt scale units) _ _ .. 7 20 120 4 60 90 2 3 3 Chemical constituents (ppm): Silica. _ _ . . . 2. 0 7. 3 6. 2 2. 6 8. 3 7. 9 4. 1 6. 1 7. 8 Iron.. . __ . 08 1. 1 1. 9 . 1 1. 1 2. 2 . 02 . 02 . 02 Manganese _ _ - _ . 00 . 93 2. 8 . 00 . 40 3. 0 . 02 . 08 . 01 Calcium 28 28 9. 5 31 30 9. 5 30 27 18 Magnesium 5. 8 7. 1 3. 0 5. 2 7. 8 4. 5 4. 0 4. 3 3. 9 Sodium 1. 7 5. 6 3. 1 1. 7 4. 8 3. 1 1. 9 2. 5 2. 2 Potassium 1. 5 7. 5 6. 8 1. 6 8. 2 6. 8 1. 8 2. 5 2. 2 Bicarbonate . .... 106 134 44 no 142 55 99 97 69 Sulfate. _ 10 4. 4 8. 2 n 4. 4 6. 0 11 9. 8 8. 0 Chloride. 1. 5 3. 0 3. 0 1. 5 3. 5 3. 0 2. 5 3. 0 3. 0 Fluoride _ _ _ . 1 . 2 . 2 . 1 . 2 . 1 . 0 . 1 . 1 Nitrate - ------- . 1 . 4 1. 2 . 1 . 6 . 8 . 1 . 1 . 2 Phosphate _ - . 16 . 08 . 12 . 06 . 05 . 12 . 01 . 01 . 01 Dissolved solids (ppm) (residue at 180°C).__ 107 143 78 115 155 80 103 105 77 Hardness (ppm) ______ 95 100 36 99 108 42 90 85 62 Free carbon dioxide (ppm) 4. 2 34 56 4. 4 28 55 4. 0 12 14 Specific conductance (micromhos at 25°C)__ 197 233 106 205 244 117 190 184 138 pH 7. 6 6. 8 6. 1 7. 6 6. 9 6. 2 7. 6 7. 1 6. 9 Ignition loss (ppm) 15 18 22 12 1 Sample locations shown on figure 161.1. Table 161.2.—Spectrographic analyses of minor-element content, in micrograms per liter, of three types of water, Cacapon River, W. Va., during autumn 1961 1 [Analyses by W. D. Silvey, Sacramento, Calif.] October 22 October 29 November 19 Main stream (A) Slightly colored pool(B) Highly colored pool (C) Main stream CA) Slightly colored pool (B) Highly colored pool (C) Main stream (A) Highly colored pool(C9 Aluminum 12 32 93 15 14 82 22 18 Cobalt <3.3 <3.3 ^3.3 <3.3 <3.3 <3.3 <3.3 <3.3 Copper 7.3 <3.3 <3.3 <3.3 <3.3 <3.3 <3.3 <3.3 Iron 35 >100 >100 28 >100 >100 47 >100 Manganese 11 >200 >200 31 >200 >200 9.3 33 Molybdenum <.67 <.67 <•67 <.67 2S.67 <6.7 <.67 <.67 Nickel <.67 3.6 7.0 <.67 4.5 6. 7 . 87 1.5 Lead 88 27 80 20 <3.3 32 13 8.0 Titanium <1.3 3.4 <1.3 <1.3 di.z <1.3 <1.3 <1.3 Vanadium <.67 <.67 ^.67 <.67 <.67 g.67 <.67 <.67 1 Sample locations shown on figure 161.1. pool. Both Slack (19551 and Schneller (1955) found that alkalinity increased with color in all their samples of leaf-colored waters in Indiana. Some differences between the highly colored pool (C) and the slightly colored pool (B), as noted in the October 22 and 29 samples, may have resulted from differences in redox potentials, which could explain the greater concentrations of iron, manganese, and free C02 in the more nearly anaerobic, highly colored pool (C). Just the opposite was observed for magnesium, sodium, and potassium, which were more concentrated in the slightly colored pool (B). In both colored waters (B, C), sulfate values were lower than in the river water (A). This may have resulted from reduction of sulfate to sulfide. Although one might expect the sulfate con- centration to be lowest in the highly colored pool (C), minimum values were found in the slightly colored pool (B). Concentrations of aluminum and nickel increased as water color increased (table 161.2). Similarly, iron and manganese concentrations were higher in the colored waters than in the main stream. The distribution of lead in the different environments resembled that of sulfate in that the slightly colored water (B) contained lower concentrations than either the main stream or the highly colored water (C). Data for other minor elements are inconclusive. It is tentatively concluded that the compositional differences observed between the river and the colored waters resulted from leaching and decomposition of leaf litter in isolated environments. Differences between the highly colored and slightly colored pools probably reflect differences in source of the water. Between October 22 and 29 the solute concentration in the highly colored water (C) increased somewhat, possibly due to evaporation. Although the pool was still isolated on November 19, its color had disappeared; the pool probably had been flooded and flushed by a rise in river stage on November 7 and 8. Discharge at the gaging station about 7 miles downstream from the sampling sites increased from 59 cubic feet per second on October 29 to 240 cfs on November 8, but decreased to 82 cfs on November 19. On October 22, small patches of a thin inflexible surface film were present on the highly colored pool.D184 GEOCHEMISTRY OF WATER A week later the entire surface was covered with a film in which there were only a few breaks or tears. The film probably was composed of ferric hydroxide, which formed where air oxidized ferrous iron in the water. Pearsall and Mortimer (1939) described such a condition on stagnant pools in bogs. The contribution of other metallic ions, for example manganese, to a surface film is unknown. In the colored pools under discussion, the films tended to remain in the center of an open space. Reducing conditions may have been so intense near the margins of pools that formation of the oxidized film was impossible. The effect of a surface film on reactions at the air-water interface may be important. For example, evaporation and reaeration probably are inhibited. Colored water resulting from leaf decay generally is accompanied by low concentrations of dissolved oxygen. Analysis, by the Alsterberg (azide) modification of the Winkler method, of two samples of highly colored pool water gave an average of 0.09 parts per million dissolved oxygen or about 0.8 percent of saturation on October 30. This method is not recommended for samples containing 1 mg or more of ferrous iron per liter or appreciable quantities of sulfite (Rainwater and Thatcher, 1960, p. 233). Because of the effects of interfering substances, the dissolved-oxygen values reported here may be too low, but they indicate very low concentrations. A few measurements made with a polarographic oxygen electrode (Kanwisher, 1959), on October 23, supported this conclusion. Readings in the highly colored pool ((7) were 2 percent of saturation at the bottom (24.5 cm), and a maximum of 6 percent at other places. On October 30, the slightly colored pool (B) had a dissolved-oxygen concentration of 1.9 ppm (18 percent of saturation) and the main stream of the river (A) had 10.3 ppm (99 percent of saturation). As with most other quality factors, water color correlated well with dissolved-oxygen content. The lowest oxygen concentration, 0.06 ppm, was measured near the upstream end of the highly colored pool (C), where decomposition of litter seemed to be especially far advanced. A further indication of low-dissolved oxygen content was the presence of many snails (Oyraulus) on leaves at the water surface along the margins of the highly colored pool {()). These snails are able to withstand unfavorable enrivonments better than most mollusks, owing to their ability to breathe air at the surface. Similar groups of snails were not observed in the slightly colored pool (B), which retained a connection with the oxygenated river water. Water temperature Temperature measurements made on two dates (table 161.3) indicated marked thermal stratification during the study period. On October 23, the temperature of the highly colored pool (C) increased from 6.6°C at a depth of 2.5 cm to 10°C at a depth of 11.5 cm. Below 11.5 cm the water was isothermal, although the thermometer used did not permit determination of the exact inflection points on the temperature-depth curve. On October 30, the water temperature of all colored pools decreased with increasing depth (table 161.3). The steepness of the thermal gradients in the three pools tested varied directly with intensity of water color. This relationship suggests that the mechanism involved was the rapid absorption of solar radiation near the surface by the colored water. No differences in temperature were noted in the main stream at comparable depths. The river-water temperature was highest and the temperature of the highly colored pool (C) was lowest on 4 of 5 days for which bottom-water temperatures were measured. Bottom-water temperature of the slightly colored pool (B) was always higher than that of the highly colored pool (C). This could have resulted from slow mixing with warmer river water from the downstream end of the slightly colored pool. It could also have resulted from a flow of cool ground water from the highly colored pool (C) through the moderately colored pool (D) into the slightly colored pool (B). The October 30 series of bottom temperatures seems to support this possibility. In summary, the effect of leaves on stream-water quality and temperature is a localized interaction be- Table 161.3.— Water temperature, in degrees centigrade, at three sampling sites, Cacapon River, W. Va., during autumn 1961 Depth (cm) October 23 October 30 Highly colored pool (C) Main stream (A) Slightly colored pool (B) Moderately colored pool(D) Highly colored pool (C) 2.5 11 5 6.6 10.0 13.2 14.9 14.1 14.1 16 5 10.2 (bot- tom) 18 5 10.8 (bot- tom) 19 0 13.2 (bot- tom) 20 0 - 8.5 (bot- tom) 24 5 9.8 (bot- tom) 0.14 Average thermal gradient, (degrees centigrade per cm) 0.00 0.26 0.28 0.32SLACK D185 tween the marginal vegetation and the water in low-flow pools, in backwater, or flood-plain pools. Leaf litter from other parts of the drainage area contributes also to the particulate and dissolved load of the stream. Probably this effect is delayed until heavy rains produce sufficient runoff to transport these materials downstream and out of the basin or into larger water bodies where their effects are less noticeable. REFERENCES Chase, E. S., and Ferullo, A. F., 1958, Oxygen demand of leaves in water: Water and Sewage Works, v. 105, no. 5, p. 204-205. Huet, M., 1951, Nocivite des boisements en Epieeas (Picea excelsa Link.) pour certains cours d’eaux de l’Ardenne Beige: Verh. int. Ver. Limnol., v. 11, p. 189-200. Hutchinson, G. E., 1957, Geography, physics, and chemistry, v. 1 of A treatise on limnology: New York, John Wiley and Sons, Inc., 1015 p. Hynes, H. B. N., 1960, The biology of polluted waters: Liverpool, England, Liverpool Univ. Press, 202 p. Kanwisher, John, 1959, Polarographic oxygen electrode: Limnology and Oceanography, v. 4, no. 2, p. 210-217. Myers, H. C., 1961, Manganese deposits in western reservoirs and distribution systems: Am. Water Works Assoc. Jour., v. 53, no. 5, p. 579-588. Pearsall, W. H., and Mortimer, C. H., 1939, Oxidation-reduction potentials in waterlogged soils, natural waters and muds: Jour. Ecology, v. 27, p. 483-501. Rainwater, F. H., and Thatcher, L. L., 1960, Methods for collection and analysis of water samples: U.S. Geol. Survey Water-Supply Paper 1454, 301 p. Robinson, W. O., Bastron, Harry, and Murata, K. J., 1958, Biogeochemistry of the rare-earth elements with particular reference to hickory trees: Geochim. et Cosmochim. Acta, v. 14, p. 55-67, 2 figs. Schneller, M. V., 1955, Oxygen depletion in Salt Creek, Indiana) in v. 4 of Investigations of Indiana Lakes and Streams: Indiana Univ. Dept. Zoology, p. 163-175. Slack, K. V., 1955, A study of the factors affecting stream productivity by the comparative method, in v. 4 of Investigations of Indiana Lakes and Streams: Indiana Univ. Dept. Zoology, p. 3-47. Sylvester, R. O., 1959, Water quality study of Wenatchee and Middle Columbia Rivers before dam construction: U.S. Fish and Wildlife Service, Spec. Sci. Rept.—Fisheries, no. 290.Article 162 RELATION OF PERCENT SODIUM TO SOURCE AND MOVEMENT OF GROUND WATER, NATIONAL REACTOR TESTING STATION, IDAHO By F. H. OLMSTED, Yuma, Ariz. Work done on behalf of the U.S. Atomic Energy Commission Abstract.—Most of the ground water in the National Reactor Testing Station is of the calcium magnesium bicarbonate type. The percent sodium decreases northwestward, where the source of recharge is a limestone and dolomite terrane, and increases southeastward, where the source is in silicic volcanic rocks and lacustrine deposits. The National Reactor Testing Station occupies an area of about 900 square miles in the north-central part of the eastern Snake River Plain in Idaho. Ground water occurs in a regional zone of saturation in flat-lying basalt flows and subordinate interbedded sediments of Quaternary age. The most important waterbearing openings are voids in the basalt adjacent to interflow contacts, openings in cinders, scoria, and blocky zones; and interstitial pores in the sedimentary deposits. The upper part of the regional ground-water body is unconfined in most places, but quasi-artesian conditions are common at depths of more than a few tens of feet below the water table. Depths to the water table within the station range from about 200 feet to more than 700 feet. Regional ground-water movement is south to southwest (fig. 162.1), although locally the water moves in other directions for short distances. A study of the chemical character of ground water in the station was based on analyses of 148 samples collected from 92 wells during the period 1949-61. Most of the water is of the calcium magnesium bicarbonate type, with calcium plus magnesium constituting more than 70 percent of the total cations, in equivalents per million, and bicarbonate more than 70 percent of the total anions. Other chemical types, chiefly chloride waters, are present in some areas as the result of contamination by liquid waste from the operation of atomic facilities, by return flow from irrigation in the Mud Lake area northeast of the station, and probably AKT. 162 IN U.S. GEOL. SURVEY PROF D186 by natural sources as yet not identified. A small body of ground water with a high sulfate content has been formed by seepage from a waste-disposal pond at one of the atomic facilities. As of 1961, none of this water, which is perched above a layer of basalt, had reached the regional ground-water body. Except for small bodies of water that are high in chloride or sulfate content and contain as much as 1,000 parts per million of dissolved solids, the range in chemical characteristics of the water is not large. In most of the bicarbonate water the concentration of dissolved solids ranges from about 180 to 240 ppm. The generally low mineralization of the ground water reflects the low solubility of the rocks with which the water has been in contact and the moderate to abundant precipitation in the mountainous areas that furnish the principal recharge. Recharge from precipitation on the Snake River Plain is believed to be minor. The chemical characteristics of the ground water do not appear to vary greatly with depth except near waste-discharge facilities and where deep wells penetrate thick extensive lacustrine deposits. Percent sodium as high as 80 in water from these wells probably results from cation exchange with the lacustrine clays. Areal differences in ionic ratios in the ground water beneath the station help to identify the sources of the water and to corroborate the inferred paths of water movement. In this respect, the variation in sodium plus potassium (chiefly sodium) in terms of percent of total cations (in equivalents per million) is particularly revealing. The available information indicates that at most places the chemical character of the ground water is fairly uniform to depths as great as 200 feet below the water table—that is, in the upper 200 feet of the PAPER 475-D, PAGES D186-D188. 1964.OLMSTED D187 113 °00' 112° 30' 716-626 0—64- 13D188 GEOCHEMISTRY OF WATER regional ground-water body. The chemical character of the water below that zone is not well known; accordingly, only analyses of ground water in the zone from the water table to 200 feet below the water table were used in the present study (fig. 162.1). In general, the percent sodium plus potassium increases gradually from about 8 percent along the northwest boundary of the station to more than 24 percent in the southeast. Exceptions to this general pattern are the two southernmost bodies of water high in percent sodium, which consist of contaminated water also high in chloride, and two small bodies of high-sodium water farther northeast, which are believed to be affected by upward flow in wells of sodium bicarbonate water from depths greater than 200 feet below the water table. The isograms in figure 162.1 have a general south-southwesterly trend—parallel to the regional ground-water flow lines inferred from water-level contour maps. The lower percent sodium plus potassium in the water beneath the northwestern part of the station reflects the extensively exposed limestone and dolomite in the Lost River and Lemhi Ranges to the northwest, and the higher percent sodium plus potassium in the water beneath the southeastern part of the station reflects the abundance of silicic volcanic rocks in the region to the north and northeast. Some of the water from the north and northeast also has moved through a considerable volume of fine-grained lacustrine deposits in the Mud Lake region and, therefore, may have been affected by cation exchange. The basalt, which forms the main aquifer system, appears to have little effect on the chemical character of the water moving through it. The gradual, rather than abrupt, southeastward increase in percent sodium plus potassium in the upper 200 feet of the ground-water body is believed to indicate a progressive decrease in the proportion of low-sodium water derived from the northwest and a progressive increase in the proportion of water higher in sodium content derived from the north and northeast. Variations in the rate of replenishment from these sources are indicated by substantial changes with time in the direction of ground-water flow, which are inferred from observed changes in the configuration of the water table in the northern part of the station.Article 163 RELATION OF FLUORIDE CONTENT TO RECHARGE AND MOVEMENT OF GROUND WATER IN OASIS VALLEY, SOUTHERN NEVADA By GLENN T. MALMBERG and T. E. EAKIN, Carson City, Nev. Work done in cooperation with the Nevada Department of Conservation and Natural Resources Abstract.—Although most of the ground water contains excessive fluoride, a local supply that meets Federal standards for public water supplies was located by means of a hydrogeochem-ical reconnaissance. Elsewhere in the Great Basin where water-quality problems exist, similar studies could be applied in locating usable ground water. An unusually high content of certain minor constituents, such as boron or fluoride, or of total dissolved solids renders ground water unfit for various uses in many valleys in the Great Basin. In Oasis Valley, Nev., about 100 miles northwest of Las Vegas, nearly all the ground water has a fluoride content higher than permissible limits established by the U.S. Public Health Service (1962) for public supply. A hydrogeocnemical reconnaissance by the authors (1962) indicated favorable possibilities for developing ground-water supplies of low fluoride content. Subsequent test drilling has confirmed the predictions. Oasis Valley is about 430 square miles in extent and is drained by the Amargosa River (fig! 163.1). The Amargosa is an intermittent stream which rises in a group of springs in Oasis Valley and drains southward toward the Amargosa Desert. The discharge from six springs in the alluvial fill adjacent to the flood plain of the river, about a mile northeast of Beatty, has long been the water supply for that town (population 500). As the fluoride content of this water is about 4.5 parts per million, most of the children, the native Indian population, and some adults who have resided in the community since childhood suffer from dental fluorosis. Water from wells and other springs along the Amargosa River also has high concentrations of fluoride. Most of the Oasis Valley drainage area is underlain by Tertiary and Quaternary volcanic rocks consisting of tuff and other pyroclastic rocks, welded tuff, and flows (fig. 163.1). Minor amounts of Paleozoic limestone, dolomite, shale, and sandstone make up the rest of the bedrock. Quaternary alluvial fill underlies about 65 square miles of the valley floor. The alluvial fill is saturated with ground water to within a foot or two of the land surface and forms the principal ground-water reservoir. Virtually all recharge to the alluvial fill is derived from precipitation in the drainage basin north and east of the Amargosa River and from ground-water underflow through bedrock from beyond the surficial drainage area to the north and northeast (Malmberg and Eakin, 1962). Only a minor amount of recharge is derived from infiltration of precipitation on the Bullfrog Hills, and virtually no recharge occurs from infiltration of precipitation on the valley floor and the low rolling hills adjacent to the valley floor. Analysis for fluoride and other mineral constituents of water samples throughout Oasis Valley indicated that ground water beneath the floor of the valley contained up to 5 ppm of fluoride and that throughout most of the drainage area, fluoride concentrations exceeded 1.6 ppm, the safe limit for public water supplies that would apply in the Beatty area (U.S. Public Health Service, 1962). Moreover, ground-water underflow toward the Amargosa River from the principal area of recharge north and east of the river contained higher concentrations of fluoride than ground-water underflow from the Bullfrog Hills. Fluoride in the ground water in Oasis Valley probably is derived from weathering of fluorite and fluoridebearing minerals which commonly occur as sublimation products associated with volcanic rocks. Chemical ART. 136 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D1S9-D191. 1964. D189D190 geochemistry of water EXPLANATION Qal Quaternary alluvium QTv Tertiary and Quaternary consolidated rocks of volcanic origin Composed principally of tuffs £*:ps# Paleozoic sedimentary rocks Composed principally of limestone, dolomite, and sandstone Fluoride content in ground water, in parts per million ® 0-1.5 3 1.6-3.0 3 3.1-4.5 # 4.6-6 O Well P Spring Figure 163.1.—Generalized geologic map of Oasis Valley, Nye County, Nev. Dashed outline enclosing patterned area is boundary of that part of the Amargosa River basin in the report area. decomposition and subsequent leaching of fluoridebearing minerals disseminated throughout the volcanic rocks probably are the major causes of fluoride in the ground water in the area. The high concentration of fluoride in the water samples from the Hot Springs and other thermal springs in Oasis Valley suggests that additional fluoride may be derived from hydrothermal solutions emanating from underlying magmatic rocks. All the large springs along the flood plain of the Amargosa River issue from volcanic rocks or from alluvial fill adjacent to volcanic rocks and yield thermal waters of high fluoride content. Leaching of lenses and veins of fluorite in carbonate rocks exposed in Fluorspar Canyon southeast of Beatty undoubtedly accounts for high concentrations of fluoride in the ground water downstream from that area. Chemical analyses of water from several small springs in the unnamed canyon about a mile north of Beatty and in Sober-Up Gulch, 4 miles north of Beatty indicated fluoride concentrations of less than 1.5 ppm (fig. 163.1). Analyses of water from a few wells in Oasis Valley suggested that fluoride concentrations in ground water near the base of the Bullfrog Hills is generally lower than in other parts of the valley butMALMBERG AND EAKIN D191 increases progressively toward the Amargosa River. This led to specific consideration of areas near the Bullfrog Hills for a possible supply of low-fluoride water. Because ground water in the volcanic rocks presumably would have a high fluoride content, only areas underlain by alluvium were considered. Accordingly, the alluvial fan at the mouth of the unnamed canyon north of Beatty and at the mouth of Sober-Up Gulch were considered the most likely places where ground-water supplies of relatively low fluoride content could be developed. The average annual recharge from precipitation in the catchment area of each of these basins was estimated to be about 20 acre-feet per year. Although there has been some hydrothermal alteration and mineral enrichment of the rocks in the Bullfrog Hills, underflow in the alluvium of these small tributary basins is in contact with the volcanic rock for a relatively short distance and short time and, consequently, has a minimum opportunity to dissolve fluoride. Furthermore, because the ground water in and near the areas of recharge is higher than that beneath the valley floor, there is little opportunity for deep circulation through the consolidated rock and mixing with ground water of high fluoride content. Thus, the chemical, geologic, and hydrologic data suggest that the alluvial fans at the mouth of these two canyons would be the most favorable location for developing ground water with a low fluoride content. Large withdrawals in either of the areas described above undoubtedly would cause ground-water levels to decline below the water level beneath the floor of Oasis Valley, and the resulting reversal of the hydraulic gradient would cause movement of water of high fluoride content from beneath the valley floor to the area of ground-water development. Withdrawals sufficient to meet the entire municipal demand could not be supplied entirely from this source on a perennial basis without depleting the low-fluoride water in storage and inducing inflow of high-fluoride water. However, if withdrawals in these areas were limited to the drinking-water requirements of the town of Beatty and if pumping was carefully managed to minimize drawdown, the low fluoride content might be maintained indefinitely. Since the publication of the report by Malmberg and Eakin (1962), a test well was drilled in the lower part of the alluvial fan at the mouth of the unnamed canyon north of Beatty (fig. 163.1). The well produced about 60 gallons per minute and the fluoride content of the water from the test well was 0.04 ppm after 36 hours of pumping (W. W. White, Director, Nevada Department of Environmental Health, oral communication, 1963). The success of the test drilling in locating low-fluoride ground water in this area is a good example of the value of the reconnaissance techniques used in this investigation. REFERENCES Malmberg, G. T., and Eakin, T. E., 1962, Ground-water appraisal of Sarcobatus Flat and Oasis Valley, Nye and Esmeralda Counties, Nev.: Nevada Dept. Conserv. and Nat. Resources, Ground-Water Resources Reconn. Ser. Rept. 10, 39 p. U.S. Public Health Service, 1962, Drinking water standards: Federal Register, March 6, p. 2152-2155. Webb, Barbara, and Wilson, Roland, 1962, Progress geologic map of Nevada: Nevada Bur. Mines Map 16.Article 164 RESERVOIR STORAGE ON STREAMS HAVING LOG-NORMAL DISTRIBUTIONS OF ANNUAL DISCHARGE By CLAYTON H. HARDISON, Washington, D.C. Abstract.—The standard deviation of the logarithms of annual discharge is used as an index of the required amount and frequency of carryover storage computed by probability routing of annual discharge. The amount of storage required to augment the yearly flow of streams during drought years depends on the variability of the annual discharge; for a given stream-gaging station this can be computed by probability routing as proposed by Langbein (1958, p. 1811-1817). For stations having the same type of distribution of annual discharge, the amount of required storage is related to the variability and thus can be estimated from standard curves for that distribution. Langbein (1961) used a three-step queuing model to compute curves of required storage for normal distributions of annual discharge and for a slightly skewed distribution. This article presents similar curves for log-normal distributions of annual discharge. The results presented here are based on probability routing of annual discharge. The method proposed by Langbein (1958) was modified by using a constant rate of draft, by substituting simultaneous equations for the trial-and-error solution, and by using an electronic computer to solve the equations (Hardison and Furness, 1963). The cumulative frequency-distribution curve for each variability index was drawn to cross the 50-percent line at a discharge that would preserve the theoretical relation between the arithmetic and geometric means in a log-normal distribution. The results, which show the percentage chance of deficiency for selected storage capacities and selected draft rates, were used to define curves from which the curves for selected probabilities of deficiency shown in figure 164.1 were obtained. ART. 164 IN U.S. GEOL. SURVEY PROF. D192 The results computed by probability routing assume uniform and independent annual discharges. No allowance is made for the within-the-year storage that would be required at the beginning and end of the critical period of drought years or for the effect of serial correlation between years. As both the seasonal storage and serial-correlation effect cause the total storage requirement to be greater than that shown, the curves in figure 164.1 serve primarily as a guide to the minimum amount of storage required at high draft rates where seasonal storage 'becomes relatively minor. The variability index used in figure 164.1 is the logarithmic standard deviation of the annual discharge similar to that, proposed by Lane and Lei (1950) for daily discharge. For log-normal distributions it can be estimated graphically by plotting arrayed figures of annual discharge on logarithmic probability paper at duration probabilities, in percent, computed by the formula 100 m / (n+1), in which m is the order number from highest to lowest and n is the number of items in the array. The variability index, may then be obtained by drawing a straight line through the points and expressing the vertical component of the distance between the 50-percent duration point and the 84.13-percent duration point as the proportional part of a log cycle. If the array of annual discharges on logarithmic probability paper tends not to lie on a straight line, the distribution should be replotted on arithmetic probability paper to see if the distribution is more nearly normal than log-normal. If the distribution is normal, the curves given by Langbein (1961) should be used instead of those in figure 164.1. If the distribution departs appreciably from both normal and log-normal, PAPER 475-D, PAGES D192-D193. 1964.HARDISON D193 VARIABILITY INDEX 1-percent chance of deficiency Figure 164.1.—Diagrams showing carryover storage requirements for log-normal distributions of annual discharge. Variability index is the logarithmic standard deviation of annual discharge. Parameter is constant-draft rate in percentage of mean flow. Percentage chance of deficiency is the percentage of years that the indicated reservoir capacity would be insufficient. probability routing of the actual distribution would be required. REFERENCES Hardison, C. H., and Furness, L. W., 1963, Discussion of “Reservoir Mass Analysis by a Low-Flow Series”, by J. B. Stall: Am. Soc. Civil Engineers Proc., v. 89, Sanitary Engineering Div., Jour. no. SA2, pt. 1, p. 119-122. Lane, E. W., and Lei, Kai, 1950, Stream flow variability: Am. Soc. Civil Engineers Trans., v. 115, p. 1084-1134. Langbein, W. B., 1958, Queuing theory and water storage: Am. Soc. Civil Engineers Proc., v. 84, Hydraulics Div. Jour., no. HY 5, pt. 1. ------ 1961, Reservoir storage—general solution of a queue model: Art. 298 in U.S. Geol. Survey Prof. Paper 424-D, p. D13-D17.Article 165 ENVIRONMENTAL FACTORS AFFECTING ATTACHED MACRO-ORGANISMS, PATUXENT RIVER ESTUARY, MARYLAND By ROBERT L. CORY, Washington, D.C. Abstract.—Periodic measurements of water temperature, salinity, and dissolved oxygen content were made at various depths during the period October through December 1962. Temperature was found to be the most significant variable affecting the growth of the organisms. Four principal organisms attached during October and November; growth was rapid in October but slight during the other months. An algal-tubeworm association, in which a yellow-green alga thrived on the waste from tubeworms, was noted and may proye useful as an indicator of organic pollution. A large steam-generating plant at Chalk Point, Md., on the Patuxent River, will begin operation in the spring of 1964. The plant is expected to discharge large amounts of heated water into the estuary which at times may be equal to or greater than the total freshwater inflow. To establish a basis for the future understanding of the effect of the heated water on the biota, the State of Maryland has initiated a comprehensive study of the entire river. Presently over 40 investigators representing State, Federal, and private organizations are studying a wide variety of problems. Although the estuary is close to a highly urbanized area, there has been little urban development along its shores. The value of oysters, fish, and crabs taken from the estuary annually is about half a million dollars, and the estuary is known as one of the best oystergrowing areas of the Chesapeake Bay. It is believed that any changes occuring in me estuary due to the heated water will be reflected in attached macro-organisms (chiefly barnacles, bryozoans, tubebuilding amphipods, and tubeworms) and marine borers (none were observed during the period reported on here). Study of these two types of organisms offers certain advantages because a dozen or more species representing at least eight phyla can be studied qualitatively and quantitatively at one time. Most of the mature organisms are fixed or only slightly mobile, so they are at the mercy of their immediate environment. They can be collected from uniform substrates, measured areas, and over known lengths of time, thus ruling out some of the questionable factors inherent in other quantitative biological studies. Preliminary results of the Geological Survey investigation, based on observations from October 1962 through March 1963, are reported here. Two types of test panels, wood, for marine borers, and asbestos-cement are to be exposed for periods of 1 month, 3 months, 1 year, and for the duration of the study before and after plant operation begins. The wood panels are 260 sq cm in area and the asbestos-cement panels are 290 sq cm. At each point of observation one of each type is exposed for the periods indicated. The panels are collected and preserved in alcohol, then analyzed in the laboratory. Monthly and quarterly panels have been set out and collected from the surface, middepth, and bottom (7 meters) at the center of Benedict Bridge; from middepth (2 meters) at the west end of Benedict Bridge; and from mid depth (1 meter) at the end of the pier of the Chesapeake Biological Laboratory at Solomons Island, Md. (fig. 165.1). The Benedict Bridge sampling site is in the region of expected high temperatures. The Solomons Island site was chosen as a control because it will be beyond the effect of unnatural heating. Accumulations of marine macro-organisms on the monthly panels for October and November and on the quarterly panels for October through December 1962 were measured and are reported in table 165.3. No attachment occurred during the cold months of December, January, and February, hence no accumulations are reported. ENVIRONMENTAL CHARACTERISTICS Water-quality data collected once a month at the Benedict Bridge site 1 indicate that temperature was * Data furnished by the State of Maryland Water Pollution Control Commission. Samples were taken at the center of the bridge. D194 ART. 165 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D194-D197. 1964.CORY D195 76°40' Figure 165.1.-—Index map of Patuxent River estuary. the most significant variable during the period of observations (October 1962 through February 1963). Temperatures decreased from a high of about 21.5°C in October 1962 (fig. 165.2) to a low of 0°C in January 1963. Salinity at the bridge was generally between 12 and 17 parts per thousand, and vertical variations were slight. However, the sampling program does not take into account the short-term variations caused by local weather changes and daily tidal fluctuations. Dissolved oxygen gradually increased as the temperature decreased and was high enough to support aquatic life throughout the period. Continuous recordings of temperature and daily salinity readings have been taken since 1938 at the laboratory pier at Solomons Island.2 During the study reported here, the mean daily surface temperatures gradually decreased from a high of 21°C in October 1962 to a low of 0.5°C when the quarterly panels were collected. Salinity ranged from 14.8 to 17.3 ppt during the same period. DISTRIBUTION OF ORGANISMS The density of the principal attached macro-organisms is shown in figure 165.3. Note, however, that the number of bryozoan colonies, rather than individuals, is shown, and that the number of organisms does not indicate the total mass contributed by each group. Microscopic organisms also covered much of the surface of the panels. These forms added 2 Data furnished by the Chesapeake Biological Laboratory, Solomons Island, Md. 30 co o u < 20 lu cr u tr 0 0 h Sgio o III III II ii 1 I TEMPERATURE 20 16 z o 1 10 111 111 111 - DISSOLVED OXYGEN October November December October November December Figure 165.2.—Monthly temperature, salinity, and dissolved oxygen content of water of Patuxent River estuary at Solomons Island and Benedict Bridge, October-December, 1962. *Data questionable. little to the total mass but are important in forming microscopic substrates, which often precede attachment of the larger forms. Attachment by marine macro-organisms occurred only during October and November—no macro-organisms attached to monthly test panels during December, January, or February. During October, barnacles, tubeworms, tube-building amphipods, and colonial bryozoans dominated the complex. At Benedict Bridge, barnacles were the dominant form during October and increased during November (fig. 165.3). At Solomons Island mud tubeworms were dominant in October and bryozoans in November. At Benedict Bridge, where test panels were placed at three depths, the populations varied with depth (fig. 165.3). The quarterly panels (October-December) for Benedict Bridge showed about the same organisms as the October panels, but those at Solomons Island did not have tubeworms and amphipods, which were attached to the October panels. Bryozoans were prominent and covered large areas of the panels, although the number of colonies was small. They were the dominant formNUMBER PER SQUARE DECIMETER D196 MARINE HYDROLOGY 30 20 10 0 100 50 0 200- TUBEWORMS 800 700 600 500 400 300 200 100 0 Oct. Nov. Oct.-Dec. (3 mos.) Oct. Nov. Oct.- Dec. (3 mos.) Figure 165.3.—Density of principal attachment organisms collected on test panels at Benedict Bridge and Solomons Island, October-December, 1962. BARNACLES 150 100 50 0 None at Solomons Island, where a single colony covered an area of 961 square millimeters. Many young barnacles were overgrown and killed by these bryozoan encrustations. GROWTH Even though attachment of new individuals ceased after November, the quarterly panels indicate that total accumulations of the organisms continued. Gravid tubeworms and their eggs were present, am-phipods were active on freshly collected panels, and new growth by the bryozoan colonies could be observed—evidence that all of these forms were still quite active when the panels were collected. Measurements of barnacle basal areas on the monthly panels show that growth of individuals was vigorous during October, diminished greatly in November, and was negligible the remainder of the winter. The growth varied both with depth and location. (See table 165.1.) Maximum sizes were attained at Solomons Island. The mean basal-area data are somewhat misleading. For example, although the October value for the bottom at Benedict Bridge is less than at middepth and surface, this panel not only had a great many more barnacles than those at the other two depths, but many appeared to have attached a day or two before collection. This difficulty is inherent in a monthly data-collecting program and can only be resolved by more frequent observations. Also, the high mean basal area for the Solomons Island quarterly panel was based on measurement of only 10 individuals, the total number on the panel. Note the large differences between mean and maximum basal areas for the quarterly and monthly panels. A probable explanation for this is that the organisms that attach the earliest have the advantage in gathering food and utilizing space. By the end of October all available space was occupied; Table 165.1.—Basal area of barnacles [Rostrocarinal-lateral measurements] Sampling site Mean basal area (square millimeters) Maximum basal area (square millimeters) Oct. Nov. Oct.- Dec. Oct. Nov. Oct- Dec. Benedict Bridge, center: 1.25 1.52 .88 0.22 .31 .44 0) 4.92 5.18 5.54 20.04 5.28 13.81 10.11 0.38 .49 .69 0) 16.49 42.88 28.59 45.43 Middepth Bottom (7 meters) Benedict Bridge, west side (2 Solomons Island (1 meter). 1.66 .43 11.31 .94 1 Panel scoured by ice.CORY D197 therefore, the size measurements are representative of individuals that attached during that month. Barnacle growth was slight in November and December. The mean basal area of barnacles on a special panel exposed for growth studies at middepth at Benedict Bridge from November 8 to March 8 was only 0.65 square millimeters as compared to a mean basal area of 4.92 square mm for barnacles on the panel exposed from October through December (table 165.1) Attachment and growth of the colonial bryozoans also varied with depth and location. Unlike barnacles, they grew actively in the colder winter water, and newly formed zooecia were observed on the colonies when the quarterly panels were collected. Mean and maximum areas of bryozoan colonies were considerably less at Benedict Bridge than at Solomons Island on both the monthly and quarterly panels (table 165.2). For some unknown reason, attachment and growth were less at middepth at the center of the bridge than at the surface or bottom. Table 165.2.—Area of bryozoan colonies, autum and winter, 1962 Sampling site Mean area (square millimeters) Maximum area (square millimeters) Oct. Nov. Oct- Dec. Oct. Nov. Oct- Dec. Benedict Bridge, center: 6.00 3.92 6.80 181.8 98.96 182.4 120.6 297.1 31.4 4.1 21.2 282.8 201.0 282.8 320.5 961.1 Bottom (7 meters). Benedict Bridge, west side (2 Solomons Island (1 meter) 40.6 380.2 9.4 WEIGHT OF ORGANIC MATERIAL The weight of organic carbon (dry, minus ash weight) produced by the attachment organisms varied with time, depth, and location (table 165.3). In order to obtain a figure to compare the standing crop at each location regardless of species, whether shelled or non-shelled, the material was scraped from the wooden panels, oven dried, weighed, ashed at 500°C, and reweighed. The weight was slight in November, compared to October, and negligible from December through February. Organic weight on the quarterly panels was highest at the bottom at Benedict Bridge and lowest at Solomons Island. Variations in organic weight undoubtedly reflect the relative abundance of the various species involved; however, measurements of single groups were not made. Table 165.3.—Comparison of dry and organic weights, autumn and winter, 1962 Benedict Bridge, center: Surface 9.3 7.08 76.1 2.22 23.9 30 Middepth 19.7 15.22 77.4 4. 48 22.6 30 Bottom (7 meters) 25.05 15.34 61.1 9.7 38.9 30 Solomons Island (1 meter) 4.72 3.17 67.1 1. 55 32.9 30 November Benedict Bridge, center: 2.99 2.13 71.2 0.86 28.8 30 2.46 1.67 68.0 .79 32.0 30 1.98 1.45 73.4 .53 26.6 30 Solomons Island (1 meter) 0) (■) (■) (0 « 30 Oc tober-December Benedict Bridge, center: Surface. _ (>) (!> (>> m 88 Middepth 117.36 91.56 78.0 25.80 22.0 88 Bottom (7 meters) 248.78 170.35 68.5 78. 43 31.5 88 Benedict Bridge, west side (2 meters) 114. 67 73. 45 64.1 41.22 35.9 93 Solomons Island (1 meter) 49. 92 31.09 62.4 18.83 37.6 88 1 Weight negligible. 2 Panel scoured by ice. ALGAE AND TUBEWORM ASSOCIATION The quarterly panel from the west side of Benedict Bridge showed an interesting association between the tubeworms and a yellow-green alga. Prominent tufts of the alga, 8-12 mm long, were scattered over the panel. Close examination revealed that these tufts were located at the anal end of the worm tubes. The algae appeared to profit from the organic enrichment of the water at this location. This relationship suggests that algae of this type may prove to be a useful indicator of other types of organic pollution.Article 166 ANOMALOUS STREAMFLOW-GROUND-WATER REGIMEN IN THE MAD RIVER BASIN, NEAR SPRINGFIELD, OHIO By WILLIAM P. CROSS and ALVIN J. FEULNER, Columbus, Ohio, and Anchorage, Alaska Work done in cooperation with the Ohio Department of Natural Resources, Division of Water Abstract.-—Streamflow measurements of the Mad River downstream from Springfield to the mouth of Mud Run indicated a loss in flow in this reach. Precise measurements of ground-water levels showed that ground-water flow had been diverted by drainage ditches, thereby reversing the normal ground-water gradient. The Mad River, in west-central Ohio, has the best sustained dry-weather flow of any major stream in the State, a situation indicative of large potential ground-water supplies in the basin. Because widespread and thick permeable gravels underlie most of the Mad River valley the river is an effluent, or gaining, stream throughout most of its length. However, spot measurements of streamflow indicated that, at least at some times under dry-weather conditions, there was a loss in flow downstream from Springfield. This anomaly suggested that for some reason the yield of aquifers in that reach was lower than would be expected from the known hydrogeology. As ground-water withdrawals, recharged by infiltration from the river, were being increased in both the Springfield and Dayton areas, an investigation was made during 1959 and 1960 with the purpose of accounting for the loss in streamflow. The Mad River rises on the eastern side of the Belle-fontaine outlier, a bedrock high in Logan County, the highest terrain in the State. The river flows southward across Champaign and northern Clark Counties over a gravel-filled buried valley as much as 3 miles wide. At Springfield the valley fill is constricted by a bedrock gorge approximately % mile wide. The drainage area above the stream-gaging station west of Springfield, at the lower end of the gorge, is 485 square miles. Downstream from Springfield the river flows west-southwest across a gravel-filled valley as much as 2 miles wide in Clark County and the northwestern corner of Greene County. Near the boundary between Greene and Montgomery Counties the river enters a second constriction, a bedrock gorge at Huffman Dam, one of five flood-control dams of the Miami Conservancy District. The dam forms a dry detention-type reservoir, and does ndt affect low and medium stages of flow. Between Huffman Dam and the mouth of the river at Dayton, the Mad River again flows over a gravel-filled valley. The drainage area above the gage just downstream from Huffman Dam is 632 square miles. Figure 166.1 shows the main features of the drainage basin and the inferred direction of ground-water movement. Except at the two gorge sections, the Mad River valley is underlain by large volumes of highly permeable gravels. Some of this highly permeable material is hydraulically connected to the stream, and as the hydraulic gradient is toward the stream, large inflows are contributed to the river. A series of flow measurements along such a stream, called a seepage investigation or seepage run, reveals the zones of ground-water contribution by showing large increases in flow between measuring points. Lack of gains indicated by the measurements does not eliminate the possibility that permeable water-bearing deposits are separated from the stream by relatively impermeable material or that the hydraulic gradient is not toward the stream. It was such a lack of gains in a seepage investigation in September 1948 that focused attention on the reach downstream from Springfield. The glacial features of the Mad River valley in Clark County have been described by Goldthwait ART. 166 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES 1)198-132(11. 1964. D198CROSS AND FEULNER D199 (Norris and others, 1952, p. 44-51), and those of Greene County by Norris and others (1950, p. 13-19). In its upper reaches the Mad River flows over a buried valley filled with sediments deposited in an inter-lobate zone between the Scioto (east) and the Miami (west) lobes of the Wisconsin glacier. This valley fill consists largely of permeable sand and gravel lain down as kames, kame terraces, high-level outwash, and valley-train deposits. The gorge through which the stream passes at Springfield is cut in limestone. This gorge, as well as the one which cuts through limestone and shale at Huffman Dam, has relatively impermeable bedrock walls. Downstream from each of the gorges the buried valley again widens, and is filled with highly permeable glaciofluvial deposits. Estimates of the permeability of the gravel deposits both upstream and downstream from the Springfield gaging station have been made at various times. An average for the permeability of all gravel deposits in the Mad River valley is between 2,500 and 4,000 gallons per day per square foot. This high value is in marked contrast to that of the relatively impermeable limestone and shale that form the valley walls in the two gorge sections. Walton and Scudder (1960, p. 38) estimate that the amount of water passing through the bedrock constriction as underflow beneath Huff- man Dam is about 0.6 million gallons per day. The ungaged underflow at the gage near Springfield probably is less than this figure. On the basis of extensive stream gaging at many selected locations during a period of low flow in September 1948, it is estimated that additions of ground water to the Mad River in Champaign County exceed 1 mgd per day per mile of stream. Farther south, in Clark County, more than 2 mgd per mile of stream was added during this low-flow period. Cross (1949, p. 563) estimated the amount of water stored in the gravel deposits contributing to streamflow north of the Spring-field gaging station to be as much as 2 to 3 billion cubic feet (15 to 22 billion gallons). The following low-flow measurements were made on September 15, 1948, west of Springfield: 204 cubic feet per second at the gage near Springfield, 191 cfs at Spangler Road (B-B', figs. 166.1 and 166.2), and 201 cfs at the gage downstream from Huffman Dam (Mad River near Dayton) (Norris, and others, 1952, p. 29). Because of the limitations in the accuracy of current-meter measurements, the flow loss in this reach cannot be reliably determined, but there undoubtedly was some loss. Flow-duration curves for the two stations indicate that on the average there are no losses, even at extreme low flows (Norris and others, 1952, p. 26). Increases in flow between Springfield and Dayton are less than the contributions north of Springfield, but generally they are substantial. The low flow at Midway (Spangler Road) suggested that the river was losing water within this area, and to study this condition the U.S. Geological Survey and the Division of Water, Ohio Department of Natural Resources, cooperated in constructing 18 observation wells, leveling to the wells, making observations of water levels, and obtaining streamflow measurements. All the observation wells except four, which are near the Springfield gaging station, are shown on figure 166.2. The wells averaged about 18 feet in depth, and the screened portion of the casing was placed in gravel deposits below the normal water level in the area. The holes were drilled along two cross-section lines, extending to both sides of the Mad River. Elevations of ground surface and measuring points were determined by a level traverse, including elevations of measuring points at four stream crossings. Twelve sets of well observations and four sets of streamflow measurements were made during 1959 and 1960. Two of the sets of observations are illustrated in the cross sections of figure 166.3. These are for May 21, 1959, when the flow in the river was about the mean, and for August 25, 1959, when the water levels were at the lowest point during the period of observation. The other 10 sets of observations areD200 SURFACE WATER Figure 166.2.—Map of the Mad River basin south of Springfield, showing observation-well locations. Figure 166.3.—Cross sections of the Mad River valley southwest of Springfield, showing water levels on May 21 and August 25, 1959. intermediate between the two extremes shown, but have the same general shape in cross section. The reason for the losses in the reach of the Mad River from Springfield to Huffman Dam is apparent from figure 166.3. The low area south of the river must have been very swampy in earlier times as the topographic map surveyed in 1904 shows an extensive system of drainage ditches. The map surveyed in 1955 shows that the channel now called Warden Ditch had been deepened and its drainage area more than doubled by capture of the upstream ends of other ditches between 1904 and 1955. As the ditches are lower than the river bottom, and the gravels of the valley fill are highly permeable, the direction of ground-water flow is from north to south, under the river, to Warden Ditch, which flows into the river above Huffman Dam. There is no indication of ground-water flow out of the drainage basin to the Little Miami River tributaries to the south. The accompanying table confirms these conclusions by indicating a loss (a slight gain occurred on August 30, 1960) between the Springfield gage and Spangler Road which is recovered at the Huffman Dam gage (Mad River near Dayton). Thus the Mad River is an influent or losing stream for all stages up to the mean, in the reach between Springfield and the mouth of Warden Ditch. Streamflow measurements for the Mad River between Springfield and Dayton, Ohio, 1959 and 1960 Location on Mad River Drainage area (sq mi) Discharge (cfc) 1959 1960 May 21 Aug. 25 Aug. 30 Oct. 6 Near Springfield (gage) 485 462 216 166 166 492 217 Spangler Road _ _ 539 214 171 157 Near Dayton (gage). -- - 632 670 263 195 184 The investigation illustrates the one-sided nature of a seepage run as a reconnaissance tool for locating potential ground-water supplies. Analogous to the null hypothesis of statistics, the hypothesis of potential ground-water supplies may be accepted if there is gain in the seepage run. On the other hand, in the event of a lack of gain, additional information is needed before a decision can be made. In this specific situation, the potential ground-water yield was there, but gains were not measured in the reach because the hydraulic gradient had been reversed locally, perhaps inadvertently, by drainage ditches. REFERENCES Cross, W. P., 1949, The relation of geology to dry-weather stream-flow in Ohio: Trans. Am. Geophysical Union, v. 30, p. 563-566. Kaser, Paul, 1962, Ground-water levels in the vicinity of Eagle City, Clark County, Ohio: Ohio Dept. Nat. Resources, Div. Water, Report to Ohio Water Commission, p. 1-82. Norris, S. E., Cross, W. P., and Goldthwait, R. P., 1950, The Water Resources of Greene County, Ohio: Ohio Dept. Nat. Resources, Div. Water Bull. 19, p. 1-52.CROSS AND FETJLNER D201 Norris, S. E. and others, 1952, The Water Resources of Clark Walton, W. C., and Scudder, G. D., 1960, Ground-water resources County, Ohio: Ohio Dept. Nat. Resources, Div. Water of the valley-train deposits in the Fairborn area, Ohio: Ohio Bull. 22, p. 1-82. Dept. Nat. Resources, Div. Water Tech. Rept. 3, p. 1-51.Article 167 HEIGHT-FREQUENCY RELATIONS FOR NEW JERSEY FLOODS By D. M. THOMAS, Trenton, N.J. Abstract.—Regional relations among height, discharge, and frequency are defined for New Jersey floods having annual recurrence intervals of 1.5 to 50 years. Separate relations are shown for Coastal Plain and non-Coastal Plain streams, and they can be used to predict flood heights at sites where mean annual flood discharge is known or can be estimated. Regional relations among height, discharge, and frequency have been defined for New Jersey floods. These relations, which were developed from a study of stream-gaging station records, can be used to predict flood heights at sites where mean annual flood discharge is known or can be estimated. The relations are similar to the depth-discharge-frequency relations found by Leopold and Maddock (1953). Leopold and Maddock (1953) determined for less than bankfull streamflow that a basin-wide relation exists between stream depth and discharge when discharge is of equal frequency of occurrence at all sites. They showed this relation as a simple power function of the form d=c Qf, where d is average cross-section depth, Q is discharge of a given frequency at the section, and c and / are constants for a given frequency. The simple power function also proved satisfactory for defining the New Jersey flood relations. However, for ease of field application, different measures of the two variables were used. Flood height (h) was used rather than average cross-section depth (d) because it is simply measured as a vertical distance. Flood height is defined as height of the water surface above the average channel bottom determined at time of median (50-percent duration) discharge. Mean annual flood discharge (Q2.33) was found to be a satisfactory index of flood sizes for the desired flood-frequency range, and it was used for all frequencies rather than actual discharge (Q). With these variables the equation of the New Jersey flood height-frequency relations is h=cQ*2.33. Records of 46 stream-gaging stations were used to define the flood-height-frequency relations by evaluating the c and / constants for annual flood recurrence intervals of 1.5 to 50 years. These gaging stations, which have mean annual flood discharges ranging from 117 to 140,000 cubic feet per second, are located at sites where natural flood flows and flood heights could be determined. Thirteen of the gaging stations are in the Coastal Plain physiographic province of southern New Jersey. Different c values were determined for these gaging stations than for the gaging stations outside the Coastal Plain. For the Coastal Plain, c ranged from 0.49 for a 1.5-year flood to 1.35 for a 50-year flood, while for the non-Coastal Plain sites c ranged from 0.33 for a 1.5-year flood to 0.90 for a 50-year flood. The / values for both Coastal Plain and non-Coastal Plain sites ranged from 0.360 for a 1.5-year flood to 0.314 for a 50-year flood. Figures 167.1 and 167.2 show graphically the defined relations. Accuracy of the defined relations was checked by using them to estimate flood heights at the 46 gaging stations and comparing the estimated and measured values. As a check on the accuracies expected at ungaged sites, the flood heights were estimated using mean annual flood discharges computed from drainage-basin characteristics by preliminary methods developed in another study. Two out of three estimates at gaging stations agreed with measured flood heights within ±23 percent for a 2.33-year flood, within ±16 percent for both a 10-year and a 25-year flood, and within ±21 percent for a 50-year flood. Maximum errors for a 2.33-year flood were ±54 percent and — 47 percent, and maximum errors for a 50-year flood were ±36 percent and —49 percent. No significant accuracy differences appeared between estimates for Coastal Plain and non-Coastal Plain streams. ART. 167 IN U.s. GEOL. SURVEY PROF. PAPER 475-D, PAGES D202-D203. 1964. D202FLOOD HEIGHT, IN FEET THOMAS D203 MEAN ANNUAL FLOOD DISCHARGE, IN CUBIC FEET PER SECOND Figure 167.1.—Flood height-discharge-frequency relations for non-Coastal Plain streams in New Jersey. MEAN ANNUAL FLOOD DISCHARGE, IN CUBIC FEET PER SECOND Figure 167.2.—Flood height-discharge-frequency relations for Coastal Plain streams in New Jersey. Although flood heights estimated by the areal relations are considered less reliable than those obtained from field surveys and hydraulic computations, the ease and simplicity of their determination are expected to make them useful for many purposes. REFERENCE Leopold, Luna B., and Maddock, Thomas, Jr., 1953, The hydraulic geometry of stream channels and some physiographic implications: U.S. Geol. Survey Prof. Paper 252. 716-626 0—64 -14Article 168 SEASONAL TEMPERATURE FLUCTUATIONS IN SURFICIAL SAND NEAR ALBANY, NEW YORK By RALPH C. HEATH, Albany, N.Y. Work done in cooperation with the New York Water Resources Commission Abstract.—Subsurface temperature fluctuations are due to thermal waves that originate at the land surface in response to changes in air temperature. The annual average temperature of the zone of saturation, 6 to 18 feet below land surface, is about 3°F higher than the annual average air temperature. The temperature of ground water at the top of the zone of saturation in most places corresponds closely to the mean annual air temperature. Collins (1925) pointed out that the temperature of ground water between depths of 30 and 60 feet is generally 2°F to 3°F above the mean annual air temperature. The temperature of shallow ground water under natural conditions is affected by (1) air temperature, (2) character of the land surface and ground cover (Pluhowski and Kantro-witz, 1963), (3) thermal conductivity of the zone of aeration, (4) temperature of recharge reaching the zone of saturation, and (5) heat from the earth’s interior. Of these, the first three are the most important. The temperature of recharge generally has a significant effect only where the water table is relatively close to the land surface or where the zone of aeration is composed of highly permeable material. Heat from the earth’s interior increases the temperature of ground water approximately 1°F for each 65 feet of increase in depth in northern New York. The effect of the earth’s heat on the temperature of shallow ground water cannot be determined until the direct and indirect effects of the other factors are better known. A study of the seasonal fluctuation of soil and ground-water temperatures was started in July 1959 in a residential area in the unincorporated village of McKnown-ville, N.Y., a suburb just west of Albany. The land surface is a gently rolling plain covered principally by ART. 168 IN U.S. GEOL. SURVEY PROF. D204 grass and scattered trees. The surficial deposit, about 20 feet thick, consists of fine to medium sand and mantles a series of interbedded silts and clays approximately 100 feet thick. The sand layer contains water under unconfined conditions. Because the sand is stratified, its horizontal permeability is substantially greater than its vertical permeability. Temperatures were measured at 5 levels within the surficial sand: 2 in the zone of aeration, and 3 in the zone of saturation. Temperature measurements were made in the zone of aeration with indoor-outdoor thermometers having steel-encased fluid reservoirs buried at 1.5 and 3 feet. The temperature within the zone of saturation was measured at depths of 6.4, 11.1, and 17.7 feet below land surface in 3 wells driven to depths of 8, 13, and 20 feet, respectively. The bulbs of the thermometers were imbedded in sand in the lower part of small open glass bottles that were left submerged between temperature readings. Two homes are located about 50 feet downgradient from the temperature-measurement site, and several others are within a few hundred feet downgradient. Other homes are located 100 to 200 feet upgradient. All domestic wastes from these homes, except those originating in the bathrooms, are disposed of through sewers. Bathroom wastes are disposed of through septic tanks, the nearest of which is 32 feet down-gradient from the site. The data collected during 1960 at the temperature-measurement site are summarized in the accompanying table. Also summarized in the table are the air temperatures measured by the U.S. Weather Bureau station at the Albany, N.Y., airport, 5 miles to the northeast. As the snow cover during the winter of 1959-60 was PAPER 475-D, PAGES D204-D208. 1964.HEATH D205 Summary of monthly air and subsurface temperatures, 1960, in degrees Fahrenheit Measuring point Monthly average maximum temperature Monthly average minimum temperature Range in monthly average Annual average Month Tem- pera- ture Month Tem- pera- ture tempera- tures temper- ature Air, at a height of July i 80.6 January... i 14.9 46.8 48.6 5 feet. Subsurface, at indicated depth (feet): 1.5... 2 70 ...do 27 43 48 3.0 . 2 69 35 34 52 6.4 September.. 59.5 ...do 43.2 16.3 51.5 11.1 55.9 46.4 3 9.5 3 51.3 17.7 November.. 53.1 May 49.5 3.6 51.3 1 U.S. Weather Bureau station at Albany, N.Y., airport. 2 Record incomplete. 3 November 1959 to October 1960. thin, and so did not have a detectable effect on subsurface temperatures, no data on it are included in the table. Each measurement of subsurface temperature made during the period July 1, 1959, to December 31, 1960, is plotted in the upper graph on figure 168.1. Also plotted in the upper graph are the monthly average maximum air temperatures and the monthly average minimum air temperatures. Measurements of the depth to water in the well used for making temperature measurements at a depth of 11.1 feet are plotted in the lower graph in figure 168.1. Subsurface temperatures respond principally to temperature waves originating at the land surface in response to changes in air temperature. There are at least three waves of different period and amplitude. The depth of penetration of the waves is proportional to their period, and the magnitude of the subsurface temperature fluctuations is proportional to both the amplitude and the period of the waves and to the thermal diffusivity of the subsurface materials. The shortest waves are generated by diurnal changes in air O Figure 168.1.—Air and subsurface temperatures and hydrograph of water-table fluctuations at the temperature-measurement site in McKnownville, N.Y.D206 GROUND WATER temperature and thus have a period of 1 day. Because of their short period, their effect can be detected only 1 to 2 feet below land surface. The next longest waves are those generated by the passage of frontal weather systems. The period of these waves is variable, ranging from a few days to a week or two. Their effect on the temperature of the soil is detectable to depths of a few feet. The longest waves are those generated by seasonal changes in air temperature and thus have a period of 1 year. Depending on geologic and hydro-logic conditions, these waves penetrate to depths of as much as 60 feet. The fluctuations of subsurface temperature in response to frontal weather systems and seasonal changes in air temperature are readily apparent at depths of 1.5 and 3.0 feet below land surface (fig. 168.1). During the passage of frontal weather systems, the daily average air temperature over a period of several days may change as much as 30 °F. The irregular, short-period fluctuations of soil temperature, which at a depth of 1.5 feet range from about 2°F or 3°F to as much as 10°F, show the fluctuations caused by these changes in air temperature. The fluctuations of soil temperature at a depth of 3.0 feet in response to frontal weather systems is much less than at the shallower depth. As may be seen from its consistent position about midway between the average maximum and average minimum air temperature for each month, the soil temperature at a depth of 1.5 feet closely approximates the annual average air temperature throughout the year. In 1960, the average air temperature was 48.6 °F and the average temperature at a depth of 1.5 feet was 48°F. From April to early September, when the flow of heat is downward from the land surface, the temperature at 1.5 feet generally was a degree or two above the temperature at 3.0 feet. During the fall, when the average air temperature declined rapidly, the temperature at a depth of 1.5 feet dropped progressively lower than the temperature at 3.0 feet. The divergence was greatest in December when the temperature at 1.5 feet averaged about 8 degrees lower than that at 3.0 feet. The difference remained relatively constant until the spring thaw at the end of March and then was completely eliminated within a few days. The average temperature at 3.0 feet was 52 °F, or 4 degrees higher than the average temperature at 1.5 feet and about the same as the average temperature of the ground water (see table). The ground-water temperature at three levels within the saturated zone is shown in figure 168.1. In 1960 the temperature at the uppermost leyel, 6.4 feet below land surface, fluctuated through a range of ! 6.3°F, whereas that at the lowest level, 17.7 feet below land surface, fluctuated only 3.6°F. The average ground-water temperature at all 3 levels was about 51.4°F, or 2.8°F above the annual average air temperature (see table). The graph of maximum and minimum temperatures in figure 168.2 suggests that there would be ,1___________CL Land surface CL Zone of water-table fluctuation ~l A /A i \_______z__. ui 12 u. 2 X A \ A e\ 3 I I 5 ' 1 3 I I f? 1 I 3 I I / A / / A I V) I 5> I 2 O Air temperature A Subsurface temperature A A \i U II II I 20 40 60 80 TEMPERATURE. IN DEGREES FAHRENHEIT Figure 168.2.—Maximum and minimum air and subsurface temperatures in 1960. Air temperatures are monthly average minimum temperatures in December 1959 and monthly average maximum temperatures in July 1960. no appreciable seasonal fluctuation below a depth of about 23 feet. This depth closely coincides with the base of the surficial sand, which the available data indicate is at a depth of about 20 to 21 feet. The relative lack of ground-water circulation in the silt and clay underlying the sand doubtless affects the depth of penetration of detectable seasonal temperature fluctuations in this area. The temperature of recharge reaching the zone of saturation had only a small effect on the ground-water temperature. The temperature at 6.4 feet declined about 1.5°F in February 1960 in response to recharge with cold water during a brief thaw (fig. 168.1). The cooling effect at the water table, which at the time was about 2 feet above the thermometer, doubtless was much greater.HEATH D207 Figure 168.3.—Isotherms of subsurface temperatures, in degrees Fahrenheit. (Interval) 5°F. The differences in the temperature gradient of soil between the depths of 1.5 and 3.0 feet in summer and winter is one of the most striking features of the graphs shown in figure 168.1. The downward gradient during the spring and summer was only about one-eighth the upward gradient during the fall and winter. However, the temperature gradient of water in the zone of saturation, as shown by the temperatures at depths of 6.4, 11.1, and 17.7 feet, was not significantly different in summer and winter. The explanation for the steeper thermal gradient across the zone of aeration during the fall and winter is not readily apparent. It suggests either a marked decrease in thermal diffusivity or the disturbing influence of septic-tank discharge. Figure 168.3 shows the differences in temperature with respect to time of year and depth. The diffusion ty can be computed from the relationship of the lag in time of either the maximum or the minimum temperature with depth (Singer and Brown, 1956, p. 747). Sloping straight lines representing the lag in temperature with depth are shown in the figure. Note that the slopes of the lines through the maximum isotherms and the slope of the line through the minimum isotherms in the zone of saturation (from 5 to 20 feet) are virtually the same. The diffusivity as determined from the maximum isotherms in the fall of 1960 is 0.007 cm2 per second. The slope of the line through the minimum isotherms in the zone of aeration is difficult to determine, but the line drawn in figure 168.3 seems to fit the data fairly well. The diffusivity based on this line is 0.005 cm2 per second. Because of the higher moisture content of the soil during the winter, as indicated by the intermittent rises in the water table starting in late October 1959 (fig. 168.1), and the presence of ice in the soil to a depth of 2 to 3 feet, the diffusivity during the winter should be substantially higher than during the summer. Therefore, the diffusivity of 0.005 cm2 per second computed for the zone of aeration during the winter probably is considerably less than the actual value. Although this anomaly cannot be definitely explained, it probably reflects the influence of septic-tank discharge. As noted above, a septic tank is located about 32 feet from the temperature-measurement site. However, the position of the septic-tank drain field with respect to the site is not known. Even if the position of the drain field were known, it would be difficult if not impossible to determine the effect of stratification of the soil zone on the movement of the septic-tank effluent. If the septic-tank effluent does affect the temperatures in the zone of aeration, the anomalous changes in gradient might be explained by assuming that the effluent acts as a weak heat source in the spring and summer and as a strong heat source in the fall and winter.D208 GROUND WATER The writer wishes to acknowledge the assistance of Charles O’Donnell, U.S. Geological Survey, in the computation of diffusivities. REFERENCES Collins, W. D., 1925, Temperature of water available for industrial use in the United States: U.S. Geol. Survey Water-Supply Paper 520, p. 97-104. Pluhowski, E. J., and Kantrowitz, I. H., 1963, Influence of land-surface conditions on ground-water temperatures in southwestern Suffolk County, Long Island, New York: Art. 51 in U.S. Geol. Survey Prof. Paper 475-B, p. B186-B188. Singer, I. A., and Brown, R. M., 1956, The annual variation of sub-soil temperatures about a 600-foot circle: Am. Geophys. Union Trans., v. 37, no. 6, p. 743-748.Article 169 HYDROGEOLOGY OF THE SANTIAGO AREA, CHILE By WILLIAM W. DOYEL, ROBERT J. DINGMAN, and OCTAVIO CASTILLO U.,1 * Washington, D.C., Lawrence, Kans., and Santiago, Chile Work done in cooperation with the Instituto de Investigacidnes Geoldgicas de Chile under the auspices of the Agency for International Development, U.S. Department of State Abstract.—Wells tapping unconsolidated valley fill in the Valle Central supply a large part of the water needs of Santiago. The east side of the valley is underlain by unconfined deposits that yield hard water whose mineral content increases westward in the direction of ground-water flow. Flowing wells on the west side of the valley tap a confined aquifer that yields bicarbonate water of lower mineral content and hardness. The Santiago area is a part of the Cuenca de Santiago (Santiago basin), the northernmost portion of the great Valle Central (Central Valley) of Chile. The Valle Central lies between the snow-capped Cordillera de los Andes and the lower Cordillera de la Costa (Coastal Range) and extends from the Cuesta de Chacabuco southward to Puerto Montt (fig. 169.1). The many streams that discharge into the Valle Central from the Cordillera de los Andes are used extensively for irrigation and, in the Santiago area, for a part of the water supply of the city. Santiago, with a population of 2,093,000 in 1961, uses approximately 8.0 cubic meters per second of water, 40 percent (3.2 m3/sec) of which is ground water. Approximately one-fourth of the ground water is obtained from a system of collection galleries in the Vitacura area (fig. 169.2); the remainder is obtained from 339 privately and publicly owned wells that range in depth from 12.6 to 236 m (Castillo and others 1963). The ground water is used for public water supply (74 percent), industry (23 percent), and irrigation (3 percent). The Valle Central is the topographic expression of a great north-trending graben that separates the Cordillera de los Andes from the geologically older Cordillera de la Costa (W. D. Carter, written communication, i Geologist, Instituto de Investigaciones Geologieas de Chile. AKT. 169 IN U.S. GEOL. SURVEY PROF 1962). Although the geologic age of ash-flow deposits in similar structural valleys in northern Chile indicates that downwarping in those valleys occurred in the early or middle Tertiary (Dingman, 1963), the major downward movement in the Valle Central probably occurred near the end of the Tertiary. The total thickness of unconsolidated sediments in the Cuenca de Santiago is not known because the few wells that completely penetrate the fill are located near rock outcrops. As the bedrock was subjected to erosion and tectonism after formation of the graben, its surface is characterized by high relief; Cerro Santa Lucia and Cerro Renca are the tops of bedrock hills whose lower slopes have been buried beneath valley fill. A gravity survey made by Edgar Kausel (written communication, 1960) showed the maximum depth to bedrock to be between 300 and 500 m, but until proved by test drilling these depths must be considered only approximate. In 1962, the first of a series of deep wells was being drilled in Santiago to explore for deeper aquifers and to determine the thickness of the fill. The surficial deposits in the basin are colluvial, fluvial, and fluvioglacial materials of Quaternary age. Some unconsolidated deposits in the vicinity of Santiago, which were described as glacial moraines by Briiggen (1950), Karzulovic (1958), and others, have since been examined by R. F. Flint, R. W. Lemke, Ernest Dobrovolny, Kenneth Segerstrom (Art. 152), and the authors, all of whom agree that the deposits were laid down by mudflows as a relatively thin slurry. At present, ground-water supplies are obtained from fluvial materials, possibly valley-train deposits, consisting of interconnected lenses of highly permeable sand and gravel interbedded with thick layers of clay-rich mudflows of low permeability. Most of the drilled PAPER 475—D, PAGES D209-D212. 1964. D209D210 GROUND WATER 71° 45' 30' 15' Figure 169.1—Map of the Santiago area, Chile. wells in the Santiago area produce ground water from depths of 50 to 100 m. The character of the deeper unconsolidated sediments in the valley is not known; they probably axe of continental origin and consist of detrital material derived chiefly from erosion of the Cordillera de los Andes and, to a lesser extent, from erosion of the Cordillera de la Costa. The deeper deposits may be largely of lacustrine origin. Approximately 50 kilometers north of Santiago, the Valle Central is terminated by bedrock hills that rise 200 m or more above the present valley floor. About 50 km south of Santiago is a bedrock hill that does not completely close the present valley but marks the southern end of Cuenca de Santiago and may have formed a barrier behind which a lake may have formed during Pleistocene time. Flowing wells with yields of up to 200 liters per second have been obtained from deltaic sediments that were deposited in a large glacial lake formed by glacial damming in the San Carlos area of the Valle Central, 300 km south of Santiago. It is conceivable that, during one of the Pleistocene glacial stages, similar ice tongues may have extended into the Cuenca de Santiago, blocking the drainage and forming a glacial lake. Sediments deposited in such a lake may have formed an areally extensive aquifer, which would be in contrast to the lenticular, fluvial deposits which now are the source of ground water for Santiago. At present almost all the sediments being deposited in the Valle Central are derived from the Cordillera de los Andes, where heavy precipitation and steep gradients combine to produce rapid erosion. The Cordillera de la Costa has a much slower rate of erosion because the rainfall is less and the gradients are gentler. Colluvium and some alluvial sediments derived from the Cordillera de la Costa are present, however, along the western side of the valley. Ground water presently used in and near Santiago is obtained from an unconfined aquifer in which the water table slopes generally southwestward. The average hydraulic gradient is about 10 m/km in the eastern part of Santiago but decreases to about 3 m/km in the southwestern part of the city. Although infiltrating rainfall and irrigation water are sources of some recharge in the central part of the basin, the principal source of recharge is runoff from the west slope of the Cordillera de los Andes. The Rio Colin a and Rio Lampa in the northern part of the basin, the Rio Mapocho, which flows through Santiago, and the Rio Maipo, approximately 25 km south of Santiago, contribute a large percentage of the recharge along the eastern side, as shown by the slope of the water table away from those rivers (Castillo and others, 1963). The remainder of the recharge infiltrates through the alluvial and colluvial sediments that lap up onto the flanks of the Andean foothills. Because the more permeable water-bearing materials are lenticular or narrowly enlongate, nearby wells may penetrate different bodies of water-bearing material; however, comparable water levels and similar responses to ground-water withdrawals demonstrate that the principal aquifer is hydraulically continuous throughout the central and eastern parts of the area. Perched bodies of ground water are found near the Rio Mapocho as a result of the lenticularity of the sediments and the relative impermeability of the underlying beds. Available information is insufficient to determine the area or areas of ground-water discharge. The relation of the streams in the central part of the SantiagoDOYEL, DINGMAN, AND CASTILLO U. D211 Edge of valley fill 0D4-8 Well 0 10 KILOMETERS 1 __i___i___I___i___i--1---1---1---1---1 500-700 900-1100 Q C/2 J 5 « O <1 W CQ OS H q23 W J ^ <300 300-500 700-900 >1100 EXPLANATION Figure 169.2.—Map showing dissolved-solids content in ground water of the Santiago area, Chile. Analyses made in the laboratory of the Instituto de Investigaciones Geoldgicas de Chile in 1960-61. area to the water table has not been studied, nor are data available regarding the movement of ground water south of the report area. The Rio Mapocho, however, probably serves as a ground-water drain in the western part of the basin. The relatively few wells drilled along the western side of the Cuenca de Santiago are flowing artesian wells. Although evidence of a geological separation, such as might result from faulting, is lacking, the difference in the chemical quality and temperature between the water from the flowing wells in the western part and from the water-table wells in the central and eastern part of the area indicates the existence of some sort of barrier, either geologic or hydraulic, the nature of which is not yet understood. Figure 169.2 shows the distribution of dissolved solids in the Santiago area, based on analyses reported by Castillo and others (1963). The ground water has a low dissolved-solids content, 200-300 milligrams per liter, near the Rio Mapocho in the east- ern part of the area. The dissolved-solids content increases in the direction of ground-water movement to as much as 700 mg/1 in the central part of the area and to more than 1,000 mg/1 to the southwest. The dissolved-solids content of water from flowing wells along the western side of the basin, however, is less than 300 mg/1. As shown in figures 169.2 and 169.3, the change is rather abrupt from the more highly mineralized water yielded by wells tapping the unconfined aquifer to the east. The hardness of the water presents the same general picture as the dissolved-solids content. Close to the Rid Mapocho in the eastern section of Santiago the hardness is relatively low, less than 400 mg/1 (computed as CaC03), but it increases to nearly 800 mg/1 south-westward. However, the water from flowing wells on the western side of the basin has a hardness of less than 200 mg/1. Moreover, the water from flowing wells on the west side of the basin is characterized by a higher proportion of sodium and a lower proportion of sulfateD212 GROUND WATER A CM CM a o r— 30 25 20 tn D 15 LiJ o 10 CE a ■10 Lo Figure 169.3.—Changes in dissolved-solids content, hardness, and temperature of ground water along line A-A' (fig. 169.2). and chloride as compared to the water from wells east of the Rid Mapocho. The low proportion of sulfate and chloride is reflected in the noncarbonate hardness, which commonly is very low in the well water of the west side of the basin. These differences in chemical quality suggest that there is little or no interconnection between the unconfined aquifers of the east and central parts of the area and the confined aquifers tapped in the western part. The problem of the marked differences in water quality cannot be resolved on the basis of geologic information now available. The differences may be due to geologic barriers such as faults or facies changes. On the other hand, the artesian water of the west side may be characteristic of a confined aquifer that passes beneath the unconfined aquifers to the east. Should such a deeper aquifer be present it would be of considerable economic importance to the area. Additional geological and hydrological data are necessary to define the control mechanism of the ground- water system in the Cuenca de Santiago, particularly the separation between the artesian and nonartesian waters. The test-drilling program now in progress (1963), as well as the continuing investigations by the Institute de Investigaciones Geoldgicas de Chile, should provide sufficient information for a better understanding of ground-water conditions and for a better utilization of ground-water resources. REFERENCES Briiggen, Juan, 1950, Fundamentos de la geologia de Chile: Instituto Geografico Militar, Santiago, 374 p. Castillo, Octavio, Falcdn, Eduardo, Doyel, W. W., and Valenzuela, Manuel, 1963, El agua subterranea de Santiago, segundo informe 1958-1962: Instituto de Investigacidnes Geol6gicas de Chile. [In press] Dingman, R. J., 1963, Geology of the Tulor quadrangle: Instituto de Investigaci6nes Geol6gicas de Chile. [In press] Karzulovic, Juan, 1958, Sediments cuaternarios de aguas sub-terraneas en la Cuenca de Santiago: Universidad de Chile, Instituto de Geologia, Pub. 10, 120 p.Article 170 GROUND WATER IN THE ARICA AREA, CHILE By WILLIAM W. DOYEL, Washington, D.C. Work done in cooperation with the Instituto de Investigacidnes Geol6gicas de Chile and the Corporation de Fomento de la Production under the auspices of the Agency for International Development, U.S. Department of State Abstract.—The water supply for the city of Arica presently is obtained from a water-table aquifer in the lower Valle de Azapa. Water levels are declining owing to overdevelopment; an additional supply is needed to meet present demands. An artesian aquifer underlying the Pleistocene coastal terrace north of Arica apparently is the only other source of large supplies of usable water in the area. The present (1962) water supply for Arica. the northernmost coastal city in Chile, is obtained from wells tapping a water-table aquifer in the lower Valle de Azapa1 (fig. 170.1). In 1960 an estimated 400 liters per second was being withdrawn for the' public supply and for irrigation in the part of the valley extending 40 kilometers inland from Arica (Kleiman and Torres, 1960). Water use both for the city and for irrigation has increased, and withdrawals have exceeded the capacity of the aquifer to yield water. The resulting rapid decline of the water table is forcing the city to search elsewhere for additional water supplies. Rainfall west of the foothills (Pre-Cordillera) of the Cordillera de los Andes is negligible; only traces of rainfall have been recorded in Arica at intervals of several years. One perennial stream, the Rfo Lluta, flows through the area, but the water is too highly mineralized for other than limited agricultural or industrial use. The Rfo San Jose, which flows through the Valle de Azapa, is perennial in its upper reaches and intermittent in the lower part of the valley. The river discharges into the ocean only every 4 or 5 years when it is in flood stage, which may last a few days or a few weeks. A small amount of the floodwater is used for irrigation, but most of it is lost into the Pacific Ocean. 1 Spanish terminology is used for geographic names; for example, Valle for valley, and Quebrada for valley containing dry or intermittent stream. Water from the Rfo Lauca, on the Altiplano approximately 100 km east of Arica, has been diverted by means of canals and a tunnel into the Valle de Azapa and will help alleviate the water shortage in the area. Arica lies at the northern end of the Cordillera de la Costa (Coast Range), at the mouth of the Rfo San Jose. Extending northward from Arica into Peru is La Concordia, a plain bounded on its eastern side by a north-south scarp and on its western side by the Pacific Ocean. The plain, which slopes gently westward to the ocean, widens northward and is approximately 8 km wide at the Peruvian border. The scarp marks the westernmost extent of a thick section of conglomerate, sand, and pyroclastic deposits of late Tertiary and early Pleistocene age, the upper surface of which forms a westward-sloping plain extending seaward from the Pre-Cordillera. The Rfos San Jose and Lluta, which issue onto La Concordia from steep-walled valleys, as well as the stream that formed the Quebrada Gallinazos and now is intermittent and flows only at widely spaced intervals, have dissected the highland in the Chilean territory. La Concordia was formed as a depositional terrace during the Pleistocene and is underlain by interbedded marine and continental sediments with a thickness of at least 200 meters. The surface is composed of at least three erosional terraces, each marking a stage in the recession of the sea. The Rfos Gallinazos (Quebrada Gallinazos) and San Jose have incised steep-walled channels into the surface of the plain, and the Rfo Lluta has cut a valley containing a series of terraces that can be correlated with successive stages in the lowering of the base level of the river. The Rfos Lluta and Gallinazos have built alluvial fans onto the most recent erosional terrace along the present shoreline, and the Rfo San Jose is building a delta. ART. 170 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D213-D215. 1964. D213D214 ground water Recent sediments are present in the form of colluvium, as alluvial deposits of the rivers, and as windblown sand covering much of the area. The face of the scarp that forms the eastern border of La Concordia and the surface of the high plain are covered in the Arica area with a concretelike layer of salt-cemented sandy or tuffaceous material. The only possibilities for developing substantial additional ground-water supplies in the Arica area are in the Valle de Azapa, which is presently overdeveloped, the Valle Lluta, and on La Concordia; that is, from the Pleistocene and Recent sediments. Some of the older sediments probably contain and could transmit water, but lack of recharge and the effects of diastrophism preclude the possibility of developing substantial water supplies from the pre-Pleistocene sediments. Ground water in the Valle de Azapa occurs in an unconfined aquifer in the alluvial deposits, which are less than 100 m thick. The aquifer has been exploited for several centuries, at least as far back as Inca times, but since the advent of drilled wells and large-capacity pumps it has been overdeveloped. As predicted by Taylor (1949), overdevelopment has resulted in a declining water table, which, in turn, has led to the abandonment of many hand-dug wells in the Arica area. In addition, all the springs have ceased to flow. Recharge to the aquifer comes from seepage and underflow from the upstream portion of the Rfo San Jose and also from floods and irrigation water. The water table slopes toward the Pacific Ocean, and fresh ground water probably is being discharged by evaporation and as seeps along the present shoreline nearDOYEL D215 Arica. Salt-water intrusion may be taking place but definitive data are not available. Test wells drilled in the Valle Lluta have shown that the valley fill will yield little ground water. About 10 years ago several successful water wells were drilled on La Concordia Peruana (the part of La Concordia in Peru), 7 km north of the Chile-Peru border. La Concordia Chilena (the part of La Concordia in Chile) is geographically and geologically similar to La Concordia Peruana, and 4 exploratory wells were drilled on it between the Quebrada Escritos and the border to determine whether the productive aquifer extended into Chile. All 4 wells struck artesian water that rose to within 15 m of the surface. The water occurs in sand and gravel strata between 75 and 130 m below the surface. Well 2 (fig. 170.1) was pumped at 150 liters per second (approximately 2,400 gallons/minute) for 120 hours, and as the maximum drawdown was 8.5 m, the specific capacity of the well was 17.6 liters per second per meter of drawdown. At the end of the test the water contained 835 milligrams per liter of total dissolved solids, 247 mg/1 of chloride, 172 mg/1 of sulfate, and 368 mg/1 of total hardness. The artesian head, the slope of the piezometric surface, and the water quality indicate that there is recharge to the artesian aquifer, although they give no information regarding the quantity. Test drilling has shown that there is no recharge reaching the aquifer from either the Valle Lluta or Valle de Azapa, and that recharge is probably coming from the northeast, from Peru. Ground water on La Concordia should not be developed without consideration of the possibility of intrusion of sea water, or possibly of inflow of salty marine connate water from unflushed parts of the aquifer as the artesian pressure declines. A shallow water-table aquifer, penetrated in wells 1, 2, and 3, about 18 km north-northwest of Arica (fig. 170.1), probably is recharged from the infrequent rains in the area and could not support large-scale sustained withdrawals. The artesian aquifer underlying La Concordia presents the best possibility for developing large additional water supplies in the Arica area. The diversion of water from the Rfo Lauca is increasing the ground water available in the Valle de Azapa, both by furnishing additional recharge and by replacing water supplies formerly withdrawn from the water-table aquifer. Artificial recharge by means of check dams and off-stream ponds also would increase the amount of recharge during periods of flood flow in the valley. Improvements in methods of irrigation could reduce the amount of water required per hectare, which would make additional water available in the Arica area. However, if the Arica area continues to develop, the Valle de Azapa, even with the Rfo Lauca diversions, cannot furnish sufficient water for irrigation, public supply, and industrial use. The artesian aquifer beneath La Concordia has been supplying water for irrigation on La Concordia Peruana for 10 years without any significant lowering of water levels. This fact, together with the information now available (1962) as a result of the test-drilling program, indicates that the artesian aquifer beneath La Concordia Chilena could be exploited to supply additional water to the Arica area. Even if the water could not be used for irrigation on La Concordia, owing to the salinity of the soil, it could be used for public and industrial supplies, thus releasing ground water in the Valle de Azapa for irrigation. Additional investigations, including drilling of test wells spaced throughout La Concordia Chilena, are being carried out by the Institute de Investigaciones Geologicas de Chile and the Corporacidn de Fomento de la Produccidn de Chile to determine the ground-water potential of the artesian aquifer beneath La Concordia, as well as the maximum amount of water that can be withdrawn from the water-table aquifer in the Valle de Azapa without causing progressive water-level declines. REFERENCES Kleiman, Pablo, and Torres, Juan, 1960, El agua subterranea en el Valle de Azapa: Ministerio de Obras Publicas, Direccion de Riego, mimeo. rept., 47 p. Taylor, G. C., Jr., 1949, Geology and ground water of the Azapa Valley, Province of Tarapaca, Chile: Econ. Geology, v. 44, no. 1, p. 40-62.Article 171 “LAZY” THERMOMETERS AND THEIR USE IN MEASURING GROUND-WATER TEMPERATURES By RALPH C. HEATH, Albany, N.Y. Work done in cooperation with the New York Water Resources Commission Abstract.—The temperature of ground water at selected depths in unpumped, nonflowing wells can be measured in place with high-lag “lazy” thermometers. These thermometers are constructed by enclosing the bulb of liquid-in-glass thermometers with material of high specific heat, a layer of insulation, and a protective case. Data on ground-water temperatures have been applied successfully in recent years to a variety of problems, and such data promise to become one of the most useful tools in ground-water hydrology. In most ground-water investigations, temperature measurements are obtained only for water being discharged from flowing or pumped wells. Such measurements are of limited usefulness because they represent a weighted average of the temperature of the producing zone open to the well, which in some places may be more than 100 feet thick. Measurements of water temperature in wells have been made with thermometers inserted in bottles and lowered down the wells. There are drawbacks to this method and, unless special precautions are taken, the measurements may be in error by several degrees. The most precise way to measure the temperature of water in place in wells is with an electrical thermometer. Such thermometers, accurate to 0.1 °F, are available from several manufacturers. However, because electrical thermometers are fairly expensive, and somewhat delicate, they are not practical for many field investigations. Where it is desirable to measure the temperature at the same depth in the same well periodically, satis-tory data can be obtained with so-called “lazy” thermometers. These thermometers are made by enclosing the bulb of an ordinary liquid-in-glass thermometer in such a manner that its response to change in temper- ature is delayed. Figure 171.1 shows the features of a lazy thermometer recently built by the author. It consists of four parts: 1. Standard liquid-in-glass thermometer in a metal case 2. Layer of material of high specific heat 3. Layer of insulation 4. Outer protective metal case. Standard field thermometer in metal case Plastic aluminum Thermometer in metal case Cork Brass “B-B" shot and water lVi-inch-diameter aluminum tubing Figure 171. 1—Detail of a “lazy” thermometer. ART. 171 l\ U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D216-D218. 1964. D216HEATH D217 Many materials are probably suitable for use in constructing lazy thermometers. The author uses aluminum tubing for the outer case. A layer of cork about 0.15 inch thick is used for insulation, and a layer of brass “B-B” shot is inserted between the thermometer case and the insulation. The top and bottom are sealed with plastic aluminum. However, because the annular space between the thermometer and its metal case is difficult to seal at the top, water displaces the air in the pores between the brass shot when the thermometer is submerged. This is, in fact, desirable because of the high specific heat of water. In order to determine whether a lazy thermometer is suitable for the use intended, its inertial characteristics must be determined. This is done by immersing the thermometer in water of a constant temperature, either above or below the air temperature. After the thermometer reaches the same temperature as the water, the thermometer is removed and a record is made of the change in temperature with time as the thermometer approaches the air temperature. The time is recorded to the nearest second and the temperature estimated to the nearest 0.1 degree. A graph is plotted of log of the time, in seconds, versus log of the change in temperature, in degrees Fahrenheit, divided by the difference between water and air temperatures. The time, in seconds, during which different percentage changes of temperature occur may be determined directly from the graph. The next step depends on the accuracy desired in the temperature measurements for which the lazy thermometers are to be used. Many field investigators read thermometers only to the nearest degree or nearest half degree. With care, however, the thermometer in common field use in the Water Resources Division of the U.S. Geological Survey can be read to 0.1 degree with an error of ±0.1 degree. (It is recognized, of course, that many field thermometers are not accurate to 0.1 degree. A part of this difficulty can be resolved by calibrating the field thermometer to a laboratory-tested master). Therefore, if it is desired to read the lazy thermometer to the nearest 0.1 degree, the problem is to determine the length of time required for a change of 0.1 degree at the maximum expected difference between water and air temperature at the locality in which the thermometer is to be used. Figure 171.2 shows the time required for a change of 0.1 °F in one of several lazy thermometers built by the author. For comparison, the inset in the upper right-hand corner of figure 171.2 shows the time required for the standard field thermometer in its metal case to change 1.0°F. It may be seen that its lag is one-fortieth that of the lazy thermometer. DIFFERENCE IN TEMPERATURE BETWEEN AIR AND WATER, IN DEGREES F q: Ld t— Ld O (/) O cr z Ld o X o 1- Id CO z — z Ld O Ld z CO < < X O o _l li- < h- o Ld m Z O LL. Ld h- Figure 171.2.—Inertial characteristics of a “lazy” thermometer. Whether a specific lazy thermometer can be used in a given well depends on (1) the inertial characteristics of the thermometer, (2) the depth at which temperatures are to be measured, and (3) the maximum expected difference in temperature between the water and the air. In normal practice the thermometer is left submerged at the desired depth between readings. (Parachute cord has been found to be highly satisfactory for holding the thermometers). Therefore, the time required to withdraw the thermometer from the well and read the temperature is the limiting factor in precision. Thermometers can be withdrawn from wells at a rate of about 3 feet per second and can be read in 5 to 10 seconds, depending on the skill of the observer. To determine the maximum depth at which a particular lazy thermometer can be used, it is necessary to determine the lag, in seconds, at the expected maximum difference between water and air temperature, subtract 10 seconds, and multiply the remainder by 3. For example, the expected maximum difference between water and air temperature at Albany, N.Y., is about 60°. The “lazy” thermometer for which data are plotted in figure 171.2 has a lag of 31 seconds at this difference. Therefore, by subtracting 10 seconds and multiplying by 3, it is found that this thermometer can be used at any depth down to about 60 feet. If measurements accurate to 0.2 degree were acceptable or if the expected maximum difference between water and air temperature was only 30°, the thermometer could be used to a depth of about 90 feet. As seasonal fluctuations of ground-water temperature, regardlessD218 HYDROLOGIC INSTRUMENTATION of whether they are caused by stream infiltration or changes in air temperature, occur at depths greater than about 60 feet only under unusual conditions, it is obvious that lazy thermometers are an economical and practical tool for monitoring fluctuations in ground-water temperatures.Article 172 TELEVISION APPARATUS FOR BOREHOLE EXPLORATION By J. E. EDDY, Washington, D.C. Abstract.-—A closed-circuit television system has been developed which operates in 8-inch-diameter and larger holes to depths of approximately 1,000 feet in either dry or water-filled holes. A closed-circuit television system for use in water wells has been developed that can be operated in boreholes 8 inches in diameter and larger. The equipment is well adapted for locating and examining fractures and solution openings, and for identifying changes in lithology in uncased holes drilled in consolidated rock. It is useful also for examining the position and condition of screens, casing, and obstructions in cased holes. The equipment was designed to be as simple as possible and to operate at a maximum hydrostatic pressure of 1,000 feet of water. The system uses a high-resolution Kin Tel camera and monitor modified for use in boreholes. The monitor and camera provide a full 500-line resolution, which is about twice that of home television receivers. A 14-inch aluminized picture tube with a Polaroid filter mounted in front provides a bright clear picture even in high ambient light. The control unit was designed to provide a high degree of flexibility for both camera and monitor. The unit includes a video amplifier of 8-megacycle bandwidth to furnish the 500-line resolution. Remote camera controls including horizontal and vertical centering controls may be used to shift the field of vision electronically. Fully automatic electronic target control provides self adjustment of the image intensity over a range of 4,000 to 1. Service and operation of the unit have been simplified by placing all the control adjustments on the front panel and building the chassis so that it may be removed readily from the front of the mount. Two cameras were used in the development of the present equipment. The first was a Kin Tel model 1990 enclosed in a metal tube 17% inches long by 6 inches in diameter. This camera was fitted with a lens with motor-driven focus and iris adjustments, and a mirror attached to the front of the case for viewing the side of the borehole. Several combinations of lenses and closeup attachments were used, but the results were generally unsatisfactory. The second camera was built from a Kin Tel model 1986-C studio camera, modified to fit into an aluminum tube 36 inches long with a 4%-inch outside and a 3%-inch inside diameter. The lower end is fitted with a quartz-glass plate a quarter of an inch thick, specially cut from glass having a tensile strength of 2,000 psi. Both the viewing end and the upper end of the camera case are fitted with O-rings and sealed with silastic rubber. A high-pressure coupling with neoprene sleeves joins the conductor cable to the camera. The camera (with attached cable) has been tested successfully in a pressure chamber to simulated depths of 950 feet of water. The camera assembly uses a very wide angle fixed-focus lens-a French-made Kinoptick “fish eye” lens with a focal length of 5.7 mm, a field of view of 117°, and an overall diameter of 3% inches. This lens has a depth of field of from 4 inches to infinity, thus making it possible to see the lighted area of the well bore very close to the camera and eliminating the problems involved in the motor-driven iris and focus controls. The camera is shown in figure 172.1. The most difficult problem of design has been adequate illumination in the viewed area when the camera is below water level in a well. Absorption and dispersion of the light in water reduce greatly both the contrast and intensity of the image received. The problem of lighting becomes increasingly difficult in designing lights for relatively small diameter holes. The most satisfactory lights used were 3-inch quartz iodide lamps rated at 600 watts each, wired in parallel. In larger holes, 3 lamps directed ahead at a 45° angle were placed 8 to 10 inches to the side of the camera. In small holes the best results were obtained with 3 lights positioned 120° apart and mounted on the lower 716-626 0—64- ■15 ART. 172 IN U.S. GEOL. SURVEY PROF. PAPER 475-D, PAGES D219-D220. 1964. D219D220 HYDROLOGIC INSTRUMENTATION Figure 172.1.—Camera with attached lights. A reel, wound with 1,300 feet of 24-conductor cable, is mounted in the truck. end of the camera case (fig. 172.1). With the lamps in this position, light is directed both outward and downward on the sidewall of the hole. In an 8-inch hole the lighted area of view extends from approximately 4 inches to 30 inches in front of the camera. Because of the heat generated, the quartz iodide lamp can be used for prolonged periods only when immersed in water. At present the lighting is only adequate; it is expected that future experimentation will improve this important part of the system. A municipal well at Hampstead, Md., was used for experimental tests of the equipment. The well, which is 8 inches in diameter and was drilled by the cable-tool method, penetrates fractured gneiss and schist of the Piedmont physiographic province. It is 400 feet deep and is cased to a depth of 60 feet. Tests were limited, however, to depths of about 200 feet because the reel assembly was not equipped with a motor. In the piedmont, ground water is produced largely from joints and fractures in the rock. Fractures, irregularities of the hole, changes from lighter to darker rock, and quartz veins were clearly distinguishable on the TV monitor. Figure 172.2 is a typical closeup photograph of the monitor showing some of the features observed in the uncased part of the borehole. Like home TV, the picture is sharper when viewed directly rather than from a photograph of the image. Figure 172.2.—Joints enlarged by drilling, in well at Hampstead. The experiments indicate the usefulness of the TV apparatus, particularly where ground water occurs in joints or solution openings. Although the present equipment is fully operational, additional improvements are being designed; for example, except for the vidicon tube, the camera circuit can be completely transistorized. This will make it feasible to provide a part of the camera power through self-contained batteries, and will reduce the number of conductors required from 24 to perhaps 12—permitting a sizable reduction in cable weight and cable-spool size. Transistorizing the camera also makes it feasible to reduce the diameter of the metal case containing the camera and accompanying circuit. It is expected that a modified camera assembly, without the light assembly, can be built with a 3%-inch outside diameter. Depending upon the lighting arrangement, this may make it possible to operate the equipment in 6-inch diameter holes. A number of lighting changes are under consideration, but most of these improvements will be based on experiments in boreholes. Proper lighting appears to be a question of correctly positioning and directing the lights. Commonly, if the lights are improperly positioned, a “hot spot” or intensely lighted area saturates the picture tube. Possibly, monochromatic light with a wavelength selected to coincide with the most sensitive range of the vidicon tube may reduce light dispersion and absorption. Experiments are planned to test the feasibility of using such a lighting system.SUBJECT INDEX [For major subject headings such as “Economic geology,” “Geophysics,” “Sedimentation,” see under State names or refer to table of contents.] Page Age determinations. See Potassium-argon, Radiocarbon, and Rubidium-strontium age determinations. Alaska, oil shale, Brooks Range foothills...D131 Andrew Formation, Mississippi, definition. __ 46 B Bachelor Mountain Rhyolite, Colorado, definition__________________________________________ 57 Barstow Formation, California, paleontology. 18 Boreholes, exploration by television......... 219 c Cadmium, in the Pierre Shale................... 73 California, geochronology, Sonoma County.— 105 paleontology, Mojave Desert................ 18 stratigraphy, Mojave Desert............... 18 San Joaquin Valley______*............. 14 structural geology, northern Coast Ranges. 7 Canadian Shield, geochronology................. 100 Carboniferous. See Mississippian, Pennsylvanian. Cation exchange, effect on thermal behavior of clinoptilolite and heulandite___ 89 Caves, fluvial sedimentation___________________ 141 Chile, geomorphology, Santiago................. 144 ground water, northern part.............. 213 Santiago............................... 209 stratigraphy, Santiago................... 144 Clay deposits, residual, thorium and uranium content__________________________ 128 Clinoptilolite, effect of cation exchange on thermal behavior................... 89 Colorado, petrology, Leadville area_____________ 28 petrology, San Juan Mountains______________ 49 stratigraphy, San Juan Mountains.......... 54 San Miguel County................... 33 structural geology, San Juan Mountains.. 123 San Miguel County______________________ 33 Copper, in alluvium, geochemical prospecting. 79.98 Corals, Silurian and Devonian, North America. 149 Creede Formation, Colorado, redefinition.... 62 Cretaceous, California, geochronology.......... 105 California, structural geology.............. 7 Great Plains, cadmium in Pierre Shale... 73 Massachusetts, stratigraphy............... 118 Mississippi, stratigraphy................. 43 Crustal studies, Nevada, Las Vegas area..... 108 Cyanogen bands, suppression................... 176 D Dantzler Formation, Mississippi, redefinition. 46 Devonian, North America, paleontology_______ 149 Diatons, California, San Joaquin Valley_____ 14 Differentiation, magmatic, in volcanic rocks.. 49 Dilution technique, for solubility determination....................................... 167 Dolomite, solubility in ground water___________ 179 E Eddies, as flow indicators in rhyolite.......... 70 Electrodes, cation-sensitive glass, for solubility determinations____________________ 167 Page Eocene, Mississippi, stratigraphy............. D43 Estuaries, factors affecting attached organisms________________________________________ 194 F Farmers Creek Rhyolite, Colorado, definition. 58 Fisher Quartz Latite, Colorado, redefinition.. 61 Floods, height-frequency relations........... 202 Flow direction, determination in rhyolite... 70 Fluoride, in ground water.................... 189 Franciscan rocks, California, geochronology.. 105 California, structural geology______________ 7 Friction carpet, at base of thrust sheets... 9 G Gas chromatography, effluent collector...... 164 Gas jet, d-c arc spectroscopy................. 176 Geochemical prospecting, copper in alluvium................................. 79,98 molybdenum in alluvium..................... 98 sampling-error effects................... 84 time variations in mineral abundance.... 98 Glaucophane, in schist, isotopic age determination_____________________________________ 105 Gravity studies, Nevada, Las Vegas area_____ 108 Ground water, effect of diversion by drainage ditches------------------------- 198 modified thermometer for temperature studies.......................... 216 reaction with dolomite....-............. 179 temperature variations____________________ 204 Gypsum Valley salt anticline, Colorado, late growth_____________________________ 33 H Hawaii, geophysics, island of Hawaii__________ 112 Hematite, determination of iron content_____ 174 Heulandite, effect of cation exchange on thermal behavior_________________________________ 89 Huerto Formation, Colorado, redefinition____ 59 i Idaho, ground water, National Reactor Testing Station___________________ 186 quality of water, National Reactor Testing Station_________________________ 186 stratigraphy, south-central part............ 1 structural geology, south-central part__ 1 Ion-activity product, dolomite................ 179 Ion-exchange capacity, improved determination__________________________________________ 93 Iron, determination in hematitic iron ore... 174 j Jurassic, Alaska, oil shale___________________ 131 California, geochronology________________ 105 structural geology_____________________ 7 Mississippi, stratigraphy.................. 43 K Kentucky, sedimentation, Mammoth Cave... 141 Page La Garita Quartz Latite, Colorado, definition. D57 M Mackinawite, new mineral, properties_______ 64 Maine, geochemical prospecting, Aroostook County.......................... 96 Mammoth Mountain Rhyolite, Colorado, redefinition...................... 59 Marine organisms, effect of temperature on... 194 Maryland, oceanography, Patuxent River estuary........................ 194 Massachusetts, geophysics, Cape Cod___________ 118 stratigraphy, Cape Cod.................... 118 Mineral solubility, in ground water........... 179 Miocene, California, paleontology.............. 18 Mississippi, stratigraphy__________________ 43 Mississippi, stratigraphy, southeastern part.. 43 Mississippian, Idaho, stratigraphy.............. 1 Idaho, structural geology................... 1 Nevada, stratigraphy........................ 1 structural geology..................... 1 Molybdenum, in aluvium, geochemical prospecting_______________________ 98 Monazite, thorium and uranium content...... 128 Montana, geochemistry, eastern part____________ 73 Mudflow deposits, Quaternary, in Chile_____ 144 N Nebraska, geochemistry, northwestern part... 73 Nelson Mountain Quartz Latite, Colorado, redefinition_______________________ 61 Nevada, geochemistry, Nevada Test Site..... 93 geophysics, Las Vegas area................ 108 ground water, Oasis Valley---------------- 189 quality of water, Oasis Valley____________ 189 stratigraphy, Esmeralda and Nye Counties___________________________________ 24 northern part___________________________ 1 structural geology, northern part----------- 1 New Jersey, floods........................... 202 New Mexico, sedimentation, Rio Grande...... 138 New York, ground water, Albany area________ 204 North Carolina, stratigraphy, Burke County. 38 North Dakota, geochronology, Williston basin_____________________________ 100 Norway, glacial chronology, Ullsfjord--------- 159 o Ohio, ground water, Springfield area.......... 198 surface water, Springfield area........... 198 Oil shale, Alaska, Jurassic................... 131 Oligocene, Mississippi, stratigraphy___________ 43 p Paleocene, Mississippi, stratigraphy........... 43 Pennsylvanian, Idaho, stratigraphy.........— 1 Idaho, structural geology__________________ 1 Mississippi, stratigraphy................. 43 Nevada, stratigraphy........................ 1 structural geology...................... 1 D221D222 SUBJECT INDEX Page Permian, Idaho, stratigraphy_________________ D1 Idaho, structural geology__________________ 1 Mississippi, stratigraphy________________ 43 Nevada, stratigraphy_______________________ 1 structural geology_____________________ 1 Pierre Shale, Great Plains, cadmium content. 73 Pleistocene, North Carolina, interglacial episode......................... 38 Pliocene, California, paleontology____________ 14 California, stratigraphy__________________ 14 Mississippi, stratigraphy................. 43 Nevada, stratigraphy..................... 24 Potassium-argon age determinations, California______________________________________ 105 Canadian Shield area_____________________ 100 Precambrian, Colorado, petrology_______________28 North Dakota, geochronology.............. 100 Q Quality of water, effect of tree leaves on. 181 Quaternary, Chile, geomorphology............. 144 Massachusetts, stratigraphy______________ 118 M ississippi, stratigraphy.............. 43 Norway, glacial chronology_______________ 159 See also Pleistocene. R Radiocarbon age determinations, Norway_____ 159 Rat Creek Quartz Latite, Colorado, redefinition........................................ 60 Reservoirs, storage-draft curves_____________ 192 Rhizophyllum, Silurian and Devonian, North America.......................... 149 Rhyolite flows, determination of flow direction__________________________________________ 70 Rubidium-strontium age determinations, California_____________________ 105 Canadian Shield area................... 100 Page St. Kevin Granite, Colorado, definition..... D28 Salt anticlines, Colorado, interpretation of growth.......................... 33 Sampling error, effects in geochemical prospecting.................................... 84 Saprolite, thorium and uranium in monazite.. 128 Sediment concentrations, spatial and velocity- weighted__________________________ 134 Sedimentation, storage of fine sediment in islands and point bars____________ 138 Seismic studies, island of Hawaii_____________ 112 Nevada, Las Vegas area.................... 108 Shallow Creek Quartz Latite, Colorado, definition...................................... 58 Silurian, North America, paleontology------ 149 Showshoe Mountain Quartz Latite, Colorado, definition_________________________ 61 Solubility determination, in concentrated solutions......................... 167 South Dakota, geochemistry, Black Hills area_______________________________ 73 Spectrochemical analysis, d-c arc, gas jet_ 176 Spectrometry, X-ray fluorescence, in iron determination..................... 174 Streamflow, regulation by storage............. 192 relation to ground-water movement......... 198 T Tatum Limestone Member, Catahoula Sandstone, Mississippi, definition............. 44 Tektites, semimicro X-ray fluorescence analysis......................................... 171 Television, use in borehole exploration...... 219 Temperature studies, ground water, New York................-.......... 204 modified thermometer for ground water.. 216 river water.....................-..... 184 Page Tertiary, Colorado, stratigraphy.............. D54 Colorado, structural geology------------r 123 Massachusetts, stratigraphy--------------- 118 See also Eocene, Miocene, Oligocene, Paleocene, Pliocene. Texas, sedimentation, Rio Grande.............. 138 Thermometer, “lazy,” for ground-water studies......................... 216 Thirsty Canyon Tuff, Nevada, definition_____ 24 Thorium, in monazite from saprolite........... 128 Tree leaves, effect on water quality........ 181 Tuff, zeolitic, ion-exchange capacity__________ 93 u Uranium, in monazite from saprolite........... 128 Utah, geochemical prospecting, Rocky Range. 79 v Valleriite, from Washington, properties----- 64 Vertebrate fossils, Miocene, California.....- 18 Volcanism, Colorado, San Juan Mountains... 49 w Washington, mineralogy, Snohomish County. 64 mineralogy, Spokane County............. 128 Wason Park Rhyolite, Colorado, definition... 59 West Virginia, quality of water, Cacapon River__________________________ 181 Wyoming, geochemistry, Black Hills area---- 73 XYZ X-ray fluorescence analysis, small samples- 171AUTHOR INDEX Page Adler, Isidore______________________________ D64 Andersen, B.G_______________________________ 159 Anderson, R. E_______________________________ 24 B Page Flint,E. F„............................... D141 G Goerlitz, D. F__________________________ 164 Guy, H. P.................................. 134 Back, William........................... 179 Barnes, Ivan......................... 179 Berner, R. A_______________________________ 64 Beverage, J. P__________________________ 138 Brown, R. D., Jr____________________________ 7 Brown, Robena_____________________________ 171 Bryant, Bruce--------------------------- 38 c H Hardison, C. H____________________________ 192 Heath, R. C............................... 204,216 Hedge, C. E______________________________ 100 Helz, A. W................................. 176 Holmes, G. W............................. 159 Hosterman, J. W____________________________ 128 Huffman, Claude, Jr________________________ 73 Canney, F. C_______________________________ 96 Carron, M. K________________________________ 171 Castillo U., Octavio_____________________ 144,209 Cater, F. W..........................—- 33 Chao, E. C. T.............................. 64 Christ, C.L.................-............ 167 Coleman, R. G_______________________________ 105 Collier, C. R............................... 141 Connor, J. J______________________________ 79,84 Cory, R. L__________________________________ 194 Cross, W. P................................ 198 Cummings, David----------------------------- 70 Outtitta, Frank----------------------— 171 D Dingman, R. J___________________________ 209 Doyel, W. W............................. 209,213 E Eakin, T. E............................... 189 Eargle, D.H........................... 43 Eddy, J. E................................. 219 Ekren, E. B............................. 24 Evans, H. T., Jr________________________ 64 F Falcon M., Eduardo......................... 144 Feulner, A. J........................... 198 i Ingram, Blanche_________________________ 64 K Klausing, R. L__......................... 14 Koyanagi, R. Y___________________________ 112 L Lamar, W. L............................. 164 Lee, D. E............................... 105 Leopold, E. B............................ 38 Lewis, G. E._........................... 18 Lohman, K. E............................. 14 M Malmberg, G. T__......................... 189 Marvin, R. F............................ 105 Mead, Cynthia...___________________________ 64 Miesch, A. T____________________________ 79,84 Milton, Charles____________________________ 64 N Niles, W. W................................... 174 Noble, D. C.................................... 24 Nordin, C. F., Jr............................ 138 Page O’Connor, J. T__............................. D24 Oldale, R. N................................. 118 Oliver, W. A., Jr............................. 149 Olmsted, F. H__.............................. 186 Overstreet, W. C______________________________ 128 p Pavlides, Louis............................ 96 Pearson, R. C_______________________________ 28 Peterman, Z. E______________________________ 100 R Rader, L. F_________________________________ 73 RattS, J. C._............................. 49,54 Reed, J. C., Jr........................... 38 Roberts, R. J________________________________ 1 Roller, J.C--.............................. 108 Rose, H. J., Jr_____________________________ 171 s Segerstrom, Kenneth........................... 144 Shepard, A. O-------------------------------- 89 Simons, D. B__----------------------------- 134 Slack, K. V................................. 181 Starkey, H. C-----------------------------— 89,93 Steven, T. A______________________________ 49,54,123 T Tailleur, I. L._.......................... 131 Thomas, D. M___..................... 202 Thomas, H. H...........................— 105 Thomasson, M.R____________________________ 1 Tourtelot, H. A.......................... 73 Trues dell, A.H___________________________ 167 Tuttle, C. R............................ 118 Tweto, Ogden_____________________________ 28 w Warr, J. J., Jr.......................... 128 Weiler, Louise...................-....... 38 D223 U. S. GOVERNMENT PRINTING OFFICE : 1964 O - 716-626 S.