Guide to some volcanic terranes in Washington, Idaho, Oregon and Northern California

Material Information

Guide to some volcanic terranes in Washington, Idaho, Oregon and Northern California
Series Title:
USGS - Circular
Alternate Title:
USGS Circular: 838
Donnelly-Nolan, Julie M.
Johnston, David Alexander, 1949-1980
U.S. Geological Survey
Publication Date:


serial ( sobekcm )
United States


General Note:
This guidebook arose out of a series of field trips held in conjunction with the Pacific Northwest American Geophysical Union meeting held in Bend, Oregon, September 1979. The PNAGU meeting included special volcanology sessions planned by William I. Rose, Jr., Bruce A. Nolf, amd David A. Johnston. Publication of the guidebook volume was originally planned for early 1980 by the Oregon Department of Geology and Mineral Industries (DOGAMI). Inevitable delays, subsequent scheduling problems, and the death of Dave Johnston in the May 18 eruption of Mount St. Helens led to this publication as a USGS Circular. This circular differs from typical U.S. Geological Survey compilations in that not all these papers have been examined by the Geologic Names Committee of the Survey. This Committee is charged with ensuring consistent usage of formational and other stratigraphic names in U.S. Geological Survey publications. Because many of the contributions are from workers outside the Survey, review by the Geologic Names Committee would have been inappropriate. Each author provided camera-ready pages, and the articles have not been edited for uniformity of style or usage. The contributions are generally ordered so as to describe the areas from north to south. Typically, the roadlog comes after the descriptive article except in the case of the Medicine Lake Highland articles, for which the road log is first and several topical contributions follow.
Open Access
General Note:
See Extended description for more information.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
K26-01517 ( USFLDC DOI )
k26.1517 ( USFLDC Handle )
6295 ( karstportal - original NodeID )

USFLDC Membership

Karst Information Portal

Postcard Information



This item has the following downloads:

Full Text
This guidebook arose out of a series of field trips held
in conjunction with the Pacific Northwest American
Geophysical Union meeting held in Bend, Oregon, September
1979. The PNAGU meeting included special volcanology sessions
planned by William I. Rose, Jr., Bruce A. Nolf, amd David A.
Johnston. Publication of the guidebook volume was originally
planned for early 1980 by the Oregon Department of Geology
and Mineral Industries (DOGAMI). Inevitable delays,
subsequent scheduling problems, and the death of Dave
Johnston in the May 18 eruption of Mount St. Helens led to
this publication as a USGS Circular. This circular differs
from typical U.S. Geological Survey compilations in that not
all these papers have been examined by the Geologic Names
Committee of the Survey. This Committee is charged with
ensuring consistent usage of formational and other
stratigraphic names in U.S. Geological Survey publications.
Because many of the contributions are from workers outside
the Survey, review by the Geologic Names Committee would have
been inappropriate. Each author provided camera-ready pages,
and the articles have not been edited for uniformity of style
or usage. The contributions are generally ordered so as to
describe the areas from north to south. Typically, the
roadlog comes after the descriptive article except in the
case of the Medicine Lake Highland articles, for which the
road log is first and several topical contributions


Guides to some Volcanic Terranes in Washington, Oregon, and Northern California David A. Johnston and Julie Donnelly-Nolan Editors GEOLOOICAL SURVEY CIRCULAR 838 1981


UNITED STATES DEPARTMENT OF THE INTERIOR JAMES G WAIT, Secretory Geological Survey H William Menard Director Free o n application to Branch of D i str i b uti o n U.S Geological Survey 604 South Pickett Street Alexandri a VA 22304


CONTENTS Guide to geologic field trip between Lewiston, Idaho, and Kimberly, Oregon, emphasizing the Columbia River Basalt Group by D A Swanson U S Geological Surv ey, Menlo Pa r k California 9 4025 ; and T L. Wright, u s Geological Sur vey Reston, Virginia 22 09 2 -----------------------------1 Figure l. Index map with extent of Columbia River Basalt Group ---------2 2 Chart of stratigraphy, age, magnetic polarity, and chemical type, Colu mbia Ri ver Basalt Group 3 3 M a p s showing distribution, feeder dikes of four formations in Columbia River Basalt Grou p --------------------------------4 4 Stratigraphic section near Pomeroy Washin gton ---------------4 5 Maps showing distribution, feeder dikes of four members in two formations of Columbia River Basalt Group ------------------7 6 Idealized cros s section of flow in Yakim a Basalt Sub group ----9 Table l. Major-element compositions of chemical types Colu mbi a Ri ver Basalt Group -----------------------------------------------5 2 Trace-element compositions of chemical types, Yakima B asalt Subgroup ---------------------------------------------------6 Roadlog for geologic field trip between Lewiston, Idaho, and Kimberly, Oregon by D A. S wanson, U S Geological Survey, Menlo P a rk, C A 94025; and T L Wright, U S Geological Survey, Reston, VA 22092 -------------15 Figure 1 Map of field trip stops --------------------------------------Guide to geologic field trip between Kimberly and Bend, Oregon with emphasis on the John Day Formation by Paul T Robinson, Dep a rtment of Earth Sciences, University of California, Riverside, C A 92521; and Gerald F Brem Department of Earth Sciences, California State Uni v e rsity, 16 Fullerton, CA 92634-----------------------------------------------------29 Figure l. Index map of north-central Oregon with extent of John Day Formation -------------------------------------------------30 2. Stratigraphic columns, western and eastern John Day Formation-32 Table l. Modal analyses of John Day rocks -----------------------------35 2 Chemical composition, fresh and altered welded tuff ----------37 3 Chemical analysis, least altered John Day ash-flow tuffs -----38 4 Average composition of mafic lava flows ----------------------38 Roadlog for geologic field trip between Kimberly and Bend, Ore g on with emphasis on the John D a y Formation by P aul T Robinson Department of Earth Sciences, University of California, Riverside, C A 92521-----------41 Figure 3 Map of field trip stops -------------------------------------Central High Cascade roadside geology, by Edward M. Taylor, Department of Geology, Oregon State University ----------------------------------------Figure 1 Index map ----------------------------------------------------2 Diagrammatic cross section of Cascades and Deschutes Basin --Ill 42 55 56 58


1 11:.-.trtU U Roadlog f r Centr l High C cad ge l gy Bend ist rs, McK nzi P nd S ntiam Pass, Or gon by Edw rd M Tayl r, D p r v m nv of ology, Or gon State University--------------------------------------------------------' 9 Figure 3 4. 5 6 7 8 9 10 ll. 12 Geologic sketch map, mil Geologic sketch map, mile Geologic sketch map, mile Geologic sketch map, mile Geologic sketch map, mile Geologic sketch map, mil Geologic sketch map, mile Geologic sketch map, mile Geologic sketch map, mile Geologic sketch map, mil 0.0 vO 1 6 ------------------------16 6 to 7 1 -----------------------7 1 vO 34. 5 vO 44. 6 ---------------44. 6 vO 57. 9 -----------------------7 9 to 7 3 -----------------------7 2 3 to 83 ----------------------83 ? to 91. to 99. 8 ---------------99. 8 to 119 0 ---------------------119 0 to 1 9 1 ---------------------129.1 to 1 4 2 0 ---------------------Newberry Volcano, Oregon by Norman S MacLeod, David R h rrod, U Geological Survey, Menlo Park, C A 94025 ; Lawr nee A. Chitwood U .. Forest Service, Bend, OR 97701; and .dwin H McK e U S Geological 60 6 3 6) 69 71 74 76 78 80 8 Survey, Menlo Park, CA 94025--------------------------------------------8 Figure Table l Geologic sketch map of Newberry Volcano ----------------------2 Geologic sketch map of Newberry caldera ----------------------1 Representative chemical analyses of Newb rry rocks ----------Roadlog for Newberry Volcano, Oregon, by Norman S M cLeod David R Sherrod, U S Geological Survey, Menlo Park, CA 940 25 ; L wrence A. Chitwood, U.S. Forest Service, Bend, OR 97701 ; and .dwin H McKee, U 8 8 9 0 Geological Survey, Menlo Park, C A 94025 --------------------------------3 High Lava Plains, Brothers fault zone to Harney Basin, Oregon by George W. Walker, U S Geological Survey, Menlo Park, CA 940 25 and ruce o l f Central Oregon Community College, Bend, Oregon -------------------------1 0 ) Figure Table l Index map showing High Lava Plains and Harney asin ----------2A, B C Route map with major structural elements and rhyolit K/Ar ages ----------------------------------------------------3 Map of basalt fields -----------------------------------------l. Compositions of selected basalts and rhyolites ---------------Roadlog for High Lava Plains, Brothers fault zone to Harney B sin, Oreg on, by George W. Walker, U S Geological Surv ey Menlo Park, CA 0 2 a nd 10 0 0 1 0 Bruce Nolf, Central Oregon Community College, Bend, Oregon --------------1 3 A field trip to the maar volcanoes of the Fort Rock-Christmas La e Va e y basin, Oregon by G H Heiken, Geosciences Divis ion, Los Alamos c ien tific Laboratory, Los Alamos, NM 87545 ; R V isher, Departmen v o f Geology, University of California, Santa Barbara, CA 9 3106; an V Peterson, State of Oregon, Department of Geology and Min ral Indusvries, Grants Pass, OR 97526 ---------------------------------------------------Figure l Location map -------------------------------------------------2 Map of Fort Rock -Christmas Lake Valley Basin, showing route of field trip ------------------------------------------------IV l 9 1 0


CONTENTS--Con:inued Figure 3 Cross sections :hrough Moffitt Butte, Klamath County, Oregon 21 4. Photograph of maar deposits, Big Hole ------------------------123 5 Map of Hole-in-the-Ground ------------------------------------124 6 Geologic section through Hole -in-theGround ------------------125 7 Fort Rock tuff ring: map, photograph, cross section ---------27 8a. Map of Table Rock tuff ring complex --------------------------129 8b Stratigraphic section of sediments under Table Rock tuff ring complex -----------------------------------------------------3 0 9 Diagram of sediment-tuff -breccia contact, Table Rock tuff ring complex ------------------------------------------------------13 10 Sketch of cliff showing vent 8, Table Rock tuff ring complex 132 11. Diagram showing palagonitization of tuff, Table Rock maar complex ----------------------------------------------------133 12 Map and cross sections of vents 4 and 5 Table Rock tuff ring complex ----------------------------------------------------134 13 Sketch and map of ven t 5 Table Rock tuff ring complex -------135 14 Stratigraphic section, Table Rock tuff cone ------------------136 15 Cross section, Table Rock ------------------------------------137 16 Sketch and map of vent 9 Table Rock tuff ring complex -------138 17 Slump structure in tuff ring l Table Rock, and sketches showing its formation --------------------------------------1 3 Table l. Chemical analyses of basalts, Fort Rock -Christmas Lake Val l ey Basin ------------------------------------------------------137 Roadlog for field trip to Medicine Lake Highland by Julie M Donne ly-olan, U S Geological Survey, MS-18, 345 Middlefield Road, Menlo Park, CA 9 40 25 ; Eugene V Ciancanelli, Cascadia Exploration Corporation, 3358 Apostal Road, Escond ido, CA 92025 ; John C Eiche berger, Geologica Research G-6, Los Alamos Scientific Laboratory, Los Alamos, NM 87 5 45; Jon H Fink, Geology Department, Stanford University, Stanford, CA 9 305 ; and Grant Heiken, Geological Research G 6 Los A l amos Scientific Laboratory, Los Alamos, NM 87 5 4 5 ---------------------------------------l l F igure l Location map for field trip to Medicine Lake Highlana Captain Jack's Stronghold (The geologic events that createa a natural fortress) by Aaron C Waters, 308 Moore Street, Santa Cruz, C A 95060 l 2 Map by David Kimbrough and Aaron C Waters -----------------------------151 F igure 1 Map of Captain Jack's Stronghold -----------------------------152 Pre-Holocene silicic volcanism on the northern and western margins of the Medicine Lake Highland, California by Stan ey A. Mertzman, Department of Geology, Franklin and Marshall College, Lancaster, PA 17604 ------------163 Figure Table 1 Geologic map of selected units, Medicine Lake Highlana -------l resu o -Ar ----------------------------2 Chemical ana yses of the Andesite Tuff -----------------------v 16 165 166


Figure CONTENTS-ontinued 3 Chemical analyses of th lav flow equivalent of the And sit Tuff and of older ilicic rocks ----------------------------4 Modal analys s of the And site Tuff --------------------------5 Chemical analyses of three Holoc ne glass flow --------------167 168 168 Surface structure of Little Glass Mountain by John H. Fink, Geology Department, Stanford University, Stanford C A 4 3 0 5 (Pr sent addre s : Department of Geology, Arizona State Uni versity, Tempe, A Z 85281) ------171 Figure 1 Schematic cross section through 2 rhyolitic obsidi n flow ----171 2 Flow front, Little Glass Mountain: phov o foliations, interpretation ---------------------------------------------171 3 M a p of part of northeast lobe, Little Glass Mountain ---------172 4. Compressional folds in flow front, Little Glass Mountain -----173 5 Diagram of diapir rise in rhyolitic obsidian flow ------------173 6 Map of part of northwest lobe, Little Glass Mountain ---------174 7 Map of large ridge, northwest lobe ---------------------------175 8 Photo, part of northwest lobe, Little Glass M o untain ---------176 Holocene plinian tephra deposits of the Medicine Lake Highland, California, by Grant Heiken, Los Alamos Scienti fie Labor a tory, Geosciences Divis ion, Los Alamos, NM 87545 ---------------------------------------------------177 Figure 1. Map of deposits < 1100 years old, Medicine Lake Highland ------2 Isopach map of Little Glass Mountain and Glass Mountain tephra -----------------------------------------------------3. Correlation o f Little Glass Mountain tephra units ------------Table 1. Age relations of Holocene volcanic deposits, Medicine Lake Highland ---------------------------------------------------17 179 180 178 Mechan ism of magma m ixing at Glass Mountain, Medicine Lake Highland volcano, California, by John C Eichelberger, Geosciences Division, University of California, Los A lamos Sci en t i fie Labor a tory, Los Alamos, NM 87545 -----------------------------------------------------183 Figure l. Map of Glass Mountain sample locations and lithologic units 183 2 Map showing distribution of Holocene lavas, Medicine Lake Highland ---------------------------------------------------184 3 Cross section through Medicine Lake Highland volcano ---------85 4 Photomicrographs of basaltic xenolith in Glass Mountain rhyodacite, shown in reflected light, transmitted light, and crossed polars ---------------------------------------------186 5 Graph of density versus pressure, Glass Mountain materials ---187 6 Sketches of probable erupti ve sequence at Glass Mountain -----188 Table l. Densities of Glass Mountain samples --------------------------187 VI


PREFACE This guidebook arose out of a series of field trips held in conjunction with the pacific Northwest American Geophysical Union meeting held in Bend, Oregon, September l979. The PNAGU meeting included special volcanology sessions planned by William I Jr., Bruce A. Nolf, and David A Johnston. Publication of the guidebook volume was originally planned for early 1980 by the Oregon Department of Geology ana Mineral Industries (DOG AMI). I nevi table delays, subsequent scheduling problems, and the death of Dave Johnston in the May 18 eruption of Mount St. Helens led to this publication as a usGS Circular. This circular differs from typical U S Geological Survey compilations in that not all these papers have been examined by the Geologic Names Commit tee of the Survey. This Committee is charged with ensuring consistent usage of formational and other stratigraphic names in U S Geological Survey publications. Because many of the contributions are from workers outside the Survey, review by the Geologic Names Committee would have been inappropriate. Each author provided camera-ready pages, and the articles have not been edited for uniformity of style or usage. The contributions are generally ordered so as to describe the areas from north to south Typically, the roadlog comes after the descriptive article except in the case of the Medicine Lake Highland articles, for which the road log is first and several topical contributions follow. I wish to thank Wes memorial to Dave Johns ton. with publication. Hildreth for responding so quickly to my request for a Also, my thanks go to Beverly Vogt of DOGAMI for her help Most of all, we owe special posthumous thanks to Dave Johnston for conceiving the field-trip ideas, overseeing the planning and arrangements for the trips and maintaining his sense of humor in spite of the last-minute airline strike, and reminding us gently but firmly to produce the promised articles. Julie Donnelly-Nolan VII


Dav Johnston, relaxed and happy in th land h loved, Vall y of T n Thousand Smokes, Alaska, June 1978. Posng h r at th b of n, the rhyolite dome extruded at the clos great eruption of 1912, Dav had sampl fumaroles atop several nearby volcanoes and had undertaken a study of the emplacement, compac ion, and welding of the 1912 ash-flow sheet. Th initials ar those of petrologist C . F nn r and his assistant Charles Yori, who worked in th valley in 1919 and 1923. (Photo by Dan Kosco. ) VII I


DAVID ALEXANDER JOHNSTON 1949-1980 Dave Johnston, 30-year-old volcanologist with the U.S. Geological Survey, was swept awaY by the great directed blast of Mount St. Helens on May 18, 1980. Because Dave had been an organizer of the AGU meeting at Bend, an editor of its abstracts and field-trip guidebooks, and a champion of the Pacific Northwest, it is particularly appropriate to dedicate this volume to him, i n grateful remembrance of the effect that Dave's enthusiasm, diligence, and vitality had upon so many of us. As a working colleague and daily running companion, I was asked to try to summarize Dave's career and my impress ions of him. Dave Johnston was born and raised in Illinois, and it remained his custom to return there every Christmas season to visit his parents and sister, who survive him. At school and college, Dave became an outstanding scholar, runner, and photographer and, in 1971, he graduated from the University of Illinois with "Highest Honors and Distinction" in Geology. His first geologic project was on the Upper Peninsula of Michigan where, ten years ago, he studied the progressive metamorphism of a pile of Precambrian basaltic lavas and a differentiated gabbroic sill. Here was the inception of Dave's petrographic acuity and, here too, in an associated intrusion of gabbro and diorite, were the roots of Dave's first volcano. In 1971 and 1972, Dave assisted Pete Lipman in mapping the Platoro and Lake City calderas in the San Juan Mountains volcanic field of southwestern Colorado. This fortunate apprenticeship so confirmed his enthusiasm for field petrology that, as a graduate student at the University of Washington, Dave undertook a detailed mapping, dating, and geochemical study of the Cimarron Volcano, an andesitic complex of Oligocene age in the western San Juans. His intricate reconstruction of the eruptive style and structural history of this center from evidence provided by its intertonguing pyroclastic and alluvial deposits prepared Dave very well for his big jump to the study of active volcanism. The jump came in the summer of 1975 when, with JUrgen Kienle's group from Fairbanks, Dave took part in a geophysical survey of Augustine Volcano, an island in lower Cook Inlet. When Augustine erupted explosively in early 1976, Dave rushed back to study the sequence and mechanisms of its eruptions and the petrology of their ejecta. His careful field and microprobe work demonstrated convincingly (l) that the pyroclastic flows had become less pumiceous with time (correlating with a change in emplacement mechanism); (2) that H20, Cl, and S contents of the magmas were notably high; and (3) that basaltic magma, although at no time actually erupted, had mixed with the more silicic melts and had possibly triggered their outburst. Only 25 months later, Dave turned in his Ph.D. thesis, not on his nearly completed Cimarron project, but on Augustine. This was a remarkable achievement, since both studies were labor-intensive and of high quality. Every summer Dave returned to Augustine and, in 1978 and 1979, he went to Katmai as well. In the Valley of Ten Thousand Smokes, Dave was taking the lead in studying physical aspects of the emplacement, compaction, welding, and degassing of the ash-flow sheet as part of our comprehensive restudy of the 1912 eruption. The fundamental role of the gas phase in volcanic processes had, by 1978, led Dave increasingly to focus his work upon magmatic volatile components, the main evidence for which i s preserved in glass-vapor inclusions within phenocrysts, in quenched crystal-liquid equilibria, and in the fumarolic emissions of active volcanoes. Characteristically, Dave pursued all three lines of evidence vigorously He became an exceptionally rigorous analyst, whether with the microprobe, gas chromatograph, extraction line, or mass spectrometer, but Dave knew better than most not to waste his time on a lousy sample. Thus, he put enormous energy into improving the equipment and techniques necessary to collect m anin ful and repr sen ative samples, especially the least fractionated and least contaminated as samples possible. Dave's agility, nerve, patience, and determination around th jet-like summ fumaroles in the crater of Mt. Mageik were to me a spectacle of unfor ettable beauty. His work on vol anic ases brought Dave in 1978 to the U.S. Geological Survey, where he was assigned to expand our program for monitoring volcanic emissions in Alaska and in the Cascad s. He knew that such studies could improve our understanding of the nature and evolution of magma bodies, eruptiv e mechanisms, and contributions of magmatic volatiles to hydrothermal systems, but I think Dave's dearest hope was IX


that systematic monitoring of fumarolic mis ion might p rmit d t cton of chan g s characteristically precursory to eruptions. Behind th i mage conv yed by his e xceedingly solitary work was, i n fact, a g reat cone rn for makn g a contribution to society. Dave w anted to formulate a general mod 1 for the behavior of magmatic volatiles prior to explo i v e outbursts and to de velop a corollary rational for the evaluation of hazards. But Dave also held the conviction that, as a scientist, h e should devote time, energy, and imagination ufficient to communicate ff ctively to the non-scientific public the true range of potential volcanic hazard the geologic (or sometimes instrumental) reasons for our poor predictive capability, and some notion of the characteristic time scales upon which volcanic behavior proceeds. Dave s concern for the societal importance of his work was nowhere more e vident than i n the thoroughness and dedication with which he recently undertook an asse sment of the geothermal-energy resources of the Azores and mainland Portugal. Also during his last year, Dave developed growing i nterests i n volcanic contributi n s of halogens, sulfur, and co2 to the atmosphere and in the long-range effects on climate, health, and agriculture of both volcanic and anthropogenic emissions. Thus, it was only i n part by accident that when Mount St. Hel ns resumed its activity in March 1980, Dave Johnston was the first geologist on the mountain. From the earliest outbreak until the catastrophic event, Dave spent v irtually the entire seven weeks at Mount St. Helens, monitoring S02 emissions with a correlation spectrometer and coordinating the airborne sampli ng of gases and particulates. It deserves emphasis that Dave's work was an important contribution to a well-coordinated, scientifically mult-faceted, group effort that persuaded the authorities to resist public pressures to re-open the area around the volcano, thereby holding the May 18th death toll to a few dozen instead of thousands. Ironically, Dave was caught at an observation post thought to be relati vely safe, by an u nusual eruptiv e event that was largely unanticipated, in magnitude or style, except perhaps by Dave himself. Three years ago, Dave had p u blished this warning about Augustine: "High-temperature and high-velocity shock waves exte nded far beyond the limits of discernable deposits [of the 1976 eruption] Hazard zones defined on the basis of deposits [alone do not] reflect this more extensive hazard, which at Augustine extends many kilometers offshore." No one was more aware of the danger of directed blasts, pyroclastic surges, and the shock waves that sometimes accompany explosiv e eruptions than was Dave Johnston. Dave repeatedly voiced his convictions that studies of deposits left by old eruptions provide only part of the story and that informed hazards assessment requires accepting the dangers of on-site monitoring of activ e volcanic processes. Dave was unaffectedly genui ne in everything he ever said or did, so he gave his energy and vitality to advance our understandi ng and thus our predictive capability. Dave Johnston is gone now, and many of us who were c lose to him can still scarcely believe it. Dave was a natural scientist i n the finest sense, and to the extent that natural science is a collaborative effort, he was our sh ining example, and we are all diminished. His infectious curiosity and enthusiasm, his unaffected concern for the opinions and feelings of other people, and his joyful spontaneity uplifted his friends and touched people who barely knew him. His generosity was unsurpassed, and I am recurrently astonished whenever I recall that Dave voluntarily carried 40 kg of rocks out of Katmai for me. Dave Johnston was as self-aware and capable a person as I have ever known, a man who would learn to do and dared to do whatever might be necessary to get the job done. In his final year, he began taking a night course to impro ve his rna thema tics. And, fed up with the stage fright that plagued his speaking career (though he gave excellent talks), Dave took the bull by the horns and (though he otherwise rarely drank) solved the problem by taking a jug to the podium with him. But perhaps his most essential quality was the ability to dissipate cynicsm spontaneously and to uplift the mood all around him Dave looked for, saw, and thereby encouraged the best in all of us Dave Johnston would expect us to carry on the game without him, with all his wonderful enthusiasm. X Wes Hildreth July 1980


GUIDE TO GEX:>LOGIC FIELD TRIP BETWEEN LEWISTON, IDAHO AND KIMBERLY, OREGON, EMPHASIZING THE COLUMBIA RIVER BASALT GROUP D.A. Swanson, U.S. Geological Survey, Menlo Park, California 94025 and T.L. Wright, U.S. Geological Survey, Reston, Virginia 22092 The Columbia River Basalt Group comprises a tho l eiit i c flood-basalt province of moderate size (fig. 1), covering an area of about 2 x 105 km2 with an est i mated volume of about 2 x 105 km3 (Waters, 1 962). The group is the youngest assemblage of flood basalt known, with an age range from about 1 7 to 6 m .y. ago; most eruptions took place between about 17 14 m.y. ago (Watkins and Baksi, 1974; McKee and others 1977). It is the only flood basalt province o f Phanerozoic age i n North America. Wide-ranging regional studies of the basalt, underway for the last 1 0 years, have been devoted primarily to defining stratigraphic and chemical r e lations for use in unraveling the history of the province and comparing the Columb i a River Basalt Group with flood basalt elsewhere. Recently, these stu d ies have been accelerated because of the need to know more about the stratigraphy and structure of the basalt as related to potential storage of nuclear vaste within the basalt pile. A geologic map of the entire province is under preparation. Reconnaissance geologic maps of the basalt in most of Washington and northern Idaho have been comp leted ( Swanson and 1 979a) and eventually will be published in color (for example, Swanson and others, 1980 ) Studies in 1978 and 1 979 were conducted by full-time aoo temporary personnel of the U.S. Geo logical Survey Interagency Agreement No. EW-78-I-06-1978 w ith the U.S. Department of Energy i n support of the Basalt Waste Isolation Program, administered by Rock Hanford Operations, R ichland, Washington. Perronnel invo lved i n this mapping project have been: J.L. Anderson, R.D. Bentley, G.R. Byerly, V.E. Camp, J.N. Gardner, P.R. Hooper, D.A. Swanson, W.H. Taubeneck and T.L. Wright. GENERAL ASPECTS The Columbia R iver Basalt Group i s characterized by most features considered typical of flood-basal t provinces. Flows are vo luminous, typically 10-30 km3 w ith a maximum volume of 700 km3, and many cover large areas, as much as 40,000 km2. They generally advanced as sheetfloods, rather than as channelized or tub -fed flows, and form thick cooling units composed of one or more flows. Eruptions took place fr fissure systems tens of kilometers long. Erup tion rates wer high, generally greater by 2-3 orders of magnitude than those of other Cenozoic basalt pro vinces, as determined by 1) heoretical considerations based on the relations among eruption volumes, dimensions of lin ar vent systems, and distanc traveled without appreciable crystallization or coolin g breaks, and 2) the absence of constructional shields even though viscosities calculated at constant temp rature from dry-weight chemical analyses are equal to or higher than those for oceanic tho-leiites containing similar amounts of MgO. Small spatter ramparts formed along fissures but are poorly preserved owing to bulldozing and rafting by flows and to later erosion. Cinder cones are very rare. In these and other features, basalt of the Columbia River Basalt Group contrasts with that produced by basaltic plains or oceanic volcanism (Greeley, 1977), as in the Snake River Plain, Iceland, and Hawaii. The flows cover a diverse assemblage of rocks ranging in age from Precambrian to Miocene. The prebasalt topography had considerable local 1000 m or more in places, near the margin of the Plateau. Some of the prebasalt hills today stand high above the plateau surface, especially in the Spokane area. Little is known about the nature of the rocks or pre-basalt surface beneath the central part of the plateau. Sparse evidence from a drill hole 3.24 km-deep just west of the Pasco Basin suggests that a thick weathered or altered zone caps a sequence of lower Tertiary volcanic rocks of mafic and intermediate compos i t i on at least 1970 m thick ( Raymond and Tillson, 1968 ; Newman, 1970; Jackson, 1 975; Swanson and others, 1979b) STRATIGRAPHY Formal stratigraphic subdivision of the Columbia R iver Basal t Group has recently been made ( Swanson and others, 1979b) (fig. 2 ) Considerable effort was expended in doing this, in order to provide a strong framework for topical studies. Some additional distinctive units have been found in Idaho and Oregon since the nomenclature for the Columbia River Basalt Group was established, but they can easily be given member rank and assigned to one of the three formations in the Yakima Basalt Subgroup. The criteria used to recognize specific stratigraphic units include megascopic and less commonly microscopic petrography, magnetic polarity, and chemical composition. Megascopic petrography and magnetic polarity can be determined in the field and, taken together in over-all stratigraphic context, are generally sufficient to identify a particular unit. Ambiguities commonly arise, however, and chemical analyses provide an invaluable and independent guide for checking and correcting field identifications. In fact, use of chemical analyses for correlation purposes is so rewarding that no study of the basalt requiring identification of flows should be undertaken without provision for chemistry. Physical characteristics such as weathering color, size and shape of vesi c les, thickness, and type of columnar jointing have been used by some past workers as correlation criteria, but we have found them unreliable because of lateral variability except in some local areas.


CANADA -------UNITED STATES---I o WASHINGTON 47 % \.) a ... .. l> .. ....... 4... ....... l> Q., Seattle OREGON 0 150 KILOMETERS 123 COLUMBIA ow Crf!'t'k Warden y Othello 117 0 12 '0 \ Figure 1 Index map showing approximate outcrop extent of the Columbia River Basalt Group (from Waters, 1961) and locations mentioned in text. See figure 7 for details in southeast Washington and northeast Oregon. The formal stratigraphic units of the Columbia River Basalt Group (fig. 2) have been described in detail by Mackin (1961), Bingham and Grolier (1966), Schmincke (1967b), Swanson and others (1979b), Swan son and Wright (1978), and Caap and others (1979); only general statements are made here. Imnaha Basalt Outcrops of the Imnaha Basalt, the oldest formation in the group, are confined to extreme southeast Washington, northeast Oregon, and adjacent parts of Idaho, where feeder dikes are known (fig. 3A). Whether the Imnaha occurs farther west beneath younger rocks is conjectural. It covers a surface of ru99ed loca l relief and has an aggregate thickness of than 500 m. Most flows in the formation are 2 coarse grained and plagioclase phyric (Hooper, 1974). Five cheaica types have been distinguished (Bolden and Hooper, 1976; Kleck, 1976; Vallier and Hooper, 1976; Reidel, 1978; table 1). Trace element compositions are given by Nathan and Fruchter (1974). Many flows contain zeolite amygdules (Kleck, 1976), and smectitic alteration is widespread. The Imnaha Basalt conformably underlies the Grande Ronde Basalt,and future work may find areas in which the two form-ations interfinger. Most of the Imnaha has normal magnetic polarity, but the oldest and youngest flows known have reversed polarity, based on measurements with a portable fluxgate magnetometer. Two samples of the Imnaha have 87sr;86sr initial ratios of .7044 and .7043 ( McDougall, 1976).


0 ::> } ; .& 9 .8 "' ..'l Sodd ... UrnohUo .... Botoh 8o:toh of Shumol.e' ao.o o4 ilod9" beth of Movntro.n ( Bo.oh o4 a.,.,. ... Bcm.ohof &cr,olt of l woe\..enhom 'Cibl,e 1 fOf c t'M':m 811 1 3 lhe ond PIC G

0 WA n I d I US 4 I 12 4 1 2 11 11 1 m n a h aSJ P i c t u r e G o r g e o B a sa 1 t A I fee er Ike B 11 Wanapum Basalt c D F igure 3. Maps showing generalized d istribution and feeder d i k e s for the formations in the Columbia River Basal t Group. From Swanson and Wright (1978, fig. 3.4) lies the Imnaha Basalt, inter tongues with t .he Picture Gorge Basalt, and conformably underlies and locally interfingers with the Wanapum Basalt. Commonly a thick soil and, locally, weakly lithified clastic sediments occur on top of the Grande Ronde; they indicate a significant time break, although probably no longer than a few tens of thousands of years judging the local interbedded relations between the Grande Ronde and Wanapum Basalts in southeast Washington (fig. 4). The 87sr;86sR initial ratios may increase slightly upsection, from values of about 7046 in the older flows to about 7052 in the younger flows (McDougall, 1976); this interpretation is tentative, as the older flows were sampled at different locations than the younger flows, so that variation due to lateral isotopfc heterogeneity in the source rocks is also possible. Wanapum Basalt The Wanapum Basalt is the most extensive formation exposed at the surface of the Columb i a Plateau (fig. 3C) but is much less voluminous than the Grande Ronde, probably having a volume of less than 10,000 km3. On a local scale, the Wanapum conformably overlies the Grande Ronde, except for minor erosional unconformities or interbedded relations. On a regional scale, however, the Wanapum overlies pro gressively older basalt from the center toward the eastern margin of the plateau. Such onlap is not apparent along the northern and western.margins, however. These relations suggest that the plateau had tilted westward before Wanapum time. The oldest member of the Wanapum Basalt, the Eckler Mountain Member, occurs in the Blue Mountains and adjoining foothills of southeast Washington and northeast Oregon ( Swanson and others, 1979b, 1980). The oldest flow in the member, the basal t of Robin-4 EXPlANATION C\oysrone Oodg en m ocol rype Hoon -MgO Grond Ronde chemocol type I I I I Frenchmon Sprrnos che m1col rype F igure 4. Schematic stratigraphic section in Benjami n Gulch, 3 km south of Pomeroy, wash i ng ton, show i ng chemical types for 13 basal t f l ows. Note interbedded nature of chemicall y different f l ows. Uncor rected for north d i p of about 4 degrees. From Swanson and others (1979b)


(J\ Table 1. Average major-element compositions for chemical types in the Columbia River Basalt Group ( Averages include analyses available through March 1977; see Wright and others (1979) for more recent averages and further discussion of chemical types) Chemical t l 2 3 4 5 6 7 8 9 1 0 ll 12 13 14 15 (1J.V ( 44 ) (21) (68) (7) (8) (4) (13) (13) (8) (1 0) (9) (20) (4) (8) O xide ------_ -----Si02 A li'3 "FeO" Y MgO 50.99 15.42 12.24 5.94 51. 1 4 51.18 49.53 15.06 14.06 16.34 13.04 14.11 12.38 5.07 4.60 6.06 9 31 8.59 9.15 2 58 2.65 2.58 0.91 1.19 0.93 2.24 2.93 2.41 0.4 2 0.48 0.41 50.73 17.10 ll. 26 50.36 15.54 ll. 25 6.68 10.67 2.95 0.57 l. 56 0.22 51.46 15.39 12.46 4.86 9.45 3.29 0.74 l. 79 0.33 51.57 13.87 12.28 4.44 8.12 3.36 2.02 2. 7l l. 39 53.78 14.45 11.35 5.25 9.07 2.83 1. 05 l. 78 0.28 55.94 14.04 11.77 50.01 17.08 10.01 7.84 CaO Na20 K20 T i02 P2o5 MnO 10 11 2.55 0.53 l. 66 0.34 0.22 0.22 0.22 0.20 5.42 9.30 2.45 0.85 2.32 0.38 0.19 0.20 0.23 0.24 0.19 3.36 6.88 3.1 4 l. 99 2.27 0.43 0.19 54.37 15.28 9.46 5.91 9.79 2.80 0. 77 1.17 0.29 0.16 11.01 2.44 0.27 1. 00 0.19 0.14 100.00 99.99 100.00 100.00 100.00 100.00 100.00 100.00 100.01 100.01 100.00 99.99 Chemica l tvoe 1 7 18 19 20 21 22 23 -----:2:-::4:---2 5 -26 27 28 O xide Si02 Al203 MgO CaO Na2o K20 Ti02 P20s MnO TotaLY (15) 50.27 13.69 15.04 4.29 8.31 2.67 1.16 3.55 0.81 0.21 100.00 (55) 50.09 14.31 13 78 5 18 8.88 2.57 l. 07 3.15 0.78 0.1 9 100.00 (13) 54.70 14.10 12.63 2. 7l 6.14 3.20 2.68 2.80 0.88 0.17 100.01 (ll) 54.41 14.51 ll. 07 4.51 8.32 2.69 1.77 l. 95 0.56 0.21 100.00 l/Number of analyses used i n computing average 0.9Fe2o3 (6) 50.72 16.23 9.64 8.19 10.70 2.22 0.51 ( 3) 52.12 14.33 ll. 64 5.58 9.64 2.69 0.87 1.45 2.48 0.18 0.49 0.17 100.01 100.00 (2) 49.75 15.26 ll. 82 7.1 0 10.13 2.32 0.46 2.42 0.55 0.21 100.02 (12) 54.16 13.84 12.60 3.91 7. 7l 2.66 1. 70 2.82 0.41 0.19 100.00 l/oifference between total and 100 is due to rounding during normalization Chemical types (defined by method of Wright and Hamilton, 1978) l. Picture Gorge (Imnaha Basalt) 3. Frenchman Springs (Imnaha Basalt) 4. Rock Creek (3 0) 51.88 14.88 10.55 6.96 10.67 2.36 0.64 l. 62 0.25 0.17 99.98 (41) 51.08 13.54 14.75 4.28 8.34 2.45 1. 25 3.52 0.59 0.20 100.00 (8) 54.46 14.29 11. OS 4.85 8.54 2 75 1. 39 2.17 0.35 0.15 100.00 (12) 47.45 13.84 15. 22 5.99 9.71 2 31 0. 72 3.62 0.91 0.23 100.00 7. Low Mg-Picture Gorge (Wright and others, 1973) 8. Prineville (recalculated from Uppuluri 1974) 9. High Mg-Grande Ronde (one flow) 10. Low Mg-Grande Ronde (one flow) 11. Very high Mg-Grande Ronde (one flow) 12. Robinette Mountain 13. Dodge 14. Shumaker Creek 52 .13 15.41 10.66 5.92 10.18 3.00 0.68 l. 48 0.35 0.19 100.00 29 (13) 48.73 13.88 14.41 5.88 9. 72 2.42 0.73 3.30 0.73 0.20 100.00 54.80 13.86 13. 32 2.84 6.48 3.18 1. 87 2.46 0.93 0.26 100.00 30 (8) 47.50 12.50 17 53 4.41 8.80 2.44 1. 23 3.79 1. 54 0.27 100.01 18. Lolo 19. Umatilla 52.29 13.21 14.38 4.04 7.90 2.67 l. 41 3.17 o. 7l 0.22 100.00 31 (24) 50.44 14.07 13.78 5.01 8.67 2.79 1. 47 2.90 0.66 100.00 20. Wilbur Creek 21. Asotin 22. Slippery Creek 23. Lewiston Orchards 24. Esquatzel 25. Pomona 26. Elephant Mountain 27. Buford 28. Basin City 16 (35) 51.19 14.07 13.91 4.39 8.48 2. 72 l. 22 3.13 0.67 0.23 1oo-=o1 2 American Bar (equivalent to the high-Ti Picture Gorge chem ical type of Wright and others, 1973) 5. Fall Creek (Kleck, 1976) 15. Frenchman Springs (one flow) f9. Martindale (Ice Harbor l) 6. High Mg-Picture Gorge (Wright and others, 1973) 16. Roza 17. Rosalia 30. Goose Island (Ice Harbor 2) 31. Lower Monumental


Table 2. Av rage tra e -mpositi ns for h mi 1 types in h Yakim B s l Subgroup [Instrumental neutron a tivati analys s by L.J. s hwarz und r di r i n of J.J. Row 1 Y 2. 10 12 13 14 15 16 17 18 Ba 496.0 783.0 196 .o 319.0 1021.0 564.0 03.0 534.0 510.0 3t9s. 0 Co 41.3 36.9 45.6 38.1 23.2 39.4 7.5 37.7 41.1 28.o Cr 100.] 11.7 149.5 168.8 7.2 33.1 54.6 15.6 96.7 4.9 Cs 0.7 1.4 ]:_/ 0.9 0.95 1.2 1.0 0.7 0.) Rf 3.7 5.2 1.5 2.8 6.2 4.35 4.3 5.4 4.6 10.) Rb 28.0 46.0 44.0 28.5 25.0 28.0 19.5 47 o Ta 0.71 0.94 0.57 l. 41 1. 01 0.99 1.12 1.14 1.4( Th 3.5 6.1 0. 45 1.2 5.2 3.7 3.8 4.1 3.3 6.9 Zn 132.0 147.0 118.0 118.0 181.0 201.0 177.0 220.0 192.0 14S.o Zr 349.0 350.0 232.0 216.0 308.0 222.0 590.0 Sc 37.04 31.20 36.26 40.34 34.0 36.39 35.45 36.69 36.44 26.) La 18.2 28.7 8.0 17.0 37.0 26.5 27.0 34.0 29.0 46.) Ce 38.0 58.0 16.0 37.0 77 .o 52.5 55.0 69.0 59.0 88.0 Sm 5.4 7.7 3.0 5.4 11.5 7.2 7.8 8.7 10.) Eu 1. 68 2.18 1.04 1. 60 3.27 2.27 2.34 2.81 2.59 3. 9 3 Yb 2.6 3.6 2.0 5.7 3.4 3.2 4.2 3.7 4.4 Lu 0.50 0.60 0.37 0.52 0.81 0.63 0.61 0.73 0.63 0.6 20 21 24 25 26 27 28 29 30 Ba 823.0 284.0 614.0 235.0 517.0 415.0 579.0 505.0 808.0 Co 44.5 46.4 40.6 43.4 45.1 39.6 45.0 42.0 37.6 Cr 37.6 281.3 19.9 112.0 18.0 45.5 148.3 192.9 44.8 Cs 0.8 1.0 0.7 0.7 Hf 6.0 2.8 5.5 3.2 6.8 5.4 6.6 5.8 10.6 Rb 41.0 55.0 29.5 4.3 26.0 Ta 1.02 0.67 1.68 0.77 1. 73 1. 54 1. 82 1. 51 3.09 Th 6.2 2.1 8.5 2.5 6.0 7.0 2.2 2.2 4.4 zn 142.0 118.0 144.0 131.0 181.0 142.0 206.0 189.0 296.0 Zr 345.0 200.0 445.0 256.0 328.0 505.0 Sc 26.85 30.73 27.8 34.35 31.50 31.50 37.95 38.07 36.44 La 43.5 16.0 38.0 17.0 36.0 33.5 46.0 42.0 77.5 Ce 82.5 30.5 73.0 34.5 72.5 64.0 91.0 80.0 155.0 Sm 8.4 4.3 8.5 4.7 9.6 7.4 12.9 10.8 21.3 1. 92 1. 26 2.18 1. 43 2.63 1. 86 3.54 3.03 5.68 Yb 4.3 2.3 3.6 2. 7 4.9 4.1 4.9 4.0 8.5 3.1 Lu 0.63 0.34 0.52 0.39 0.68 0.57 0.86 0.73 1.40 o.r !/chemical types: 9. Grande Ronde (one flow) 15. Frenchman Springs 20. Wilbur Creek 27. Buford 10. Low-Mg Grande Ronde (one flow) 16. Roza 12. Robinette Mountain 17. Rosalia 13. Dodge 18. Lolo 14. Shumaker Creek 19. Umatilla J:./Dash means not determined ette Mountain, is diktytaxitic and contains the lowest KzO and incompatible trace element concentrations of any other flow in the Yakima Basalt Subgroup (tables 1 and 2); it was erupted from a long fissure south of Dayton, Washington. Several flows and dikes of the next youngest flow--the basalt of Dodge, a very coarse-grained plagioclase-phyric, grusy-weathering unit--form excellent markers in the Blue Mountains. The basalt of Dodge is chemically similar to but much coarser and more porphyritic than some high-Mg flows of the Grande Ronde Basalt (tables 1 and 2). The basalt of Shumaker Creek, the youngest unit in the member, is neither widespread nor easily recognized in the field, although its high KzO and PzOs are distinctive (tables 1 and 2). 6 21. Asotin 28. Basin Cit y 24. Esquatzel 29. Martindale 25. Pomona 30. Goose Islcml 26. Elephant Mountain 31. Lower Monu mental The Frenchman Springs Member overlies and toea:. interfingers with the Eckl r Mountain Member crops out widely in the central and western parts:! the plateau (fig. SA) Its volume is probably 3 : to 5000 km3. Generally three to six flows, in places as many as ten, occur in any one sectioo. Flows were erupted from north-northwest-trending dikes extending through the Walla Walla area southeast Washington (Swanson and Wright, 1978; son and others, 1979b, 1980). Highly porphyritic flows near Soap Lake at the southern end of Coulee may have erupted along the northward exte of the known feeder system. Most flows of the Frenchman Springs Member contain rare to abundant glomerocrysts of plagioclase, although some are


Frenchman Springs Member Roza Member A Priest Rapids Member Pomona Member c D Figur e 5 Maps show i n g distribution and feeder dikes for the Frenchman Springs, Roza, and Priest Rapids Member s of the Wanapum Basalt and the Pomona Member of the Saddle Mountains Basalt. From Swanson and Wright (1978, fig. 3.5) aph ric and indistinguishable i the field from som flows of Grande Ronde Basal The member h a s a h1g PeO and Ti02 composition known as Frenchman Springs chemic a l ype (tables and 2 ) The 87sr;86sr initial ratio is about 7053 ( McDougall, 976). The Frenchman Spring s is overlain by the Roz a Membe The Roza Member, a highly plagioclase-phyric unit with a volume of abou 1500 km3, is well known, and readers are referred to papers b Lefebvre (1970) and Swanson and others (1975; l979b) for details. The member was erupted from a 1near vent system more ha 165 km long in the easter par of the platea !fig. SB) The Roza consists pr1ncipall oft cooling units, a lthough more thin units occur near v n sys e Th compos1 10n o he Roza is s1 ilar o tha of h Fr nchman Spr"ngs M embe 1 hou gh on he av rage sligh ly rich r in MgO I ab es 1 and 2) Its 87sr;86s ini ial atio is abou 70 4 ( McDougall, 1976). The Pri st Rapids M mb r ov rlies he Roza M embe and 1s h young st b sal hroughou mos o he nor h rn par of the Columbia Plateau. The memb r oc urs as ar southw s as h Colu ia Go ge (fig. AlJ known fed r dik s are confined to the far as rn part of th provinc Several dik s occur near Orofino, Idaho, and along Sla Cre k abou 16 m e s of Fre dom, Idaho (W.H. Tauben ck, T.L. 7 Wright an d D.A. Swanson, unpub. chemical data, 1977 ; V.E. Camp i n Swanson and others, 1979a) and probabl vents are located near Emida, Idaho, and i n Pa louse, Washington. Other vents presumably exist in northern Idaho, as intracanyon flows of the member occur far up the ancestral St. Joe River valley. The estimated volume of the Priest Rapids is 2000-3000 km3. In and near Spokane, flows of the Priest Rapids Member fill valleys as much as 1 00 m deep e roded into the m ain part of the Latah Formation, a sequence of finegrained c lastic sediments interbedded with and ove l y 1 n g the Grande Ronde Basalt. Flows of the Pries t Rap ids invade sediments o f the Latah in many p lace near Spokane as well as other sediments near Orof1no Ida ho The Priest Rapids Member contains magnetically reversed flow s of two distinctly different compositions, a very high FeO and Ti02 type (Rosalia chemical and a high MgO type ( Lo l o chemica l type) (tables l and 2). The flows of Rosali a chemical type are found throughout most of the extent of he member and are consistentl y older than those of Lolo chemical type, which are confined to the southern two-thirds of the member's outcrop area. A few thin f o w s of different compositions occur near vent areas in northern Idaho and adjacent Wash ington. McDougall (1976) obtained an 8 7sr;86sr initial ratio of .7053 on a f ow of Rosalia chem ical type near Frenchman Springs Co ulee.


Saddle M ou ntains Basalt This formation, the younges i th Columbia River Basalt Gro up, is about 3 5 to m. 1 and contains flowc erupted sporadic 11 ur1ng period of wan1ng volcanism, deformation, canyon cutting, and devel opmen t of thick but local sedim ntary deposi t s between flows. The Saddle Mounta1ns Basalt has a volume of only about 3000 km3 less than one per cent of the total vo lume of basall, ye t conta1ns b far t he greatest chemical and d i versity of any formation in the group. The Umati la M ember is the oldest and on of the most extensive members in the forma 1on. It occurs in extreme southeast Washington and northwes Oregon (the Troy an d Lewiston bas1ns and Un1ontown Plateau) (Pr1ce, 1977; Ross, 1 978; S wanso an a o her 1 980 ) ; vent areas an d a feeder dike occur i the Puffer Butte area (fig. 1; Pr ice, 1977) Remnant s of the Umatill a fill a broad shallow paleovalley leading from the Troy basin across the presen -day Blue Mountains i n northern Oregon to the Milton-Freewater area, where the flow spread out o the paleovalley as a sheet flooci covering muc of south-central Washington (D.A. Swanson and T.L. Wrigh unpub map, 197 8; wanson and o hers, 1 979 a). Lav a w s charnell o dlong some canyons eroded dur 1ng po ... -Wanapurr. im the western part o the Columbta Plateau, as along Yak1ma Ridge (R. D Bentley 1n Swanson and o thers. l979a l The distribution ot the Umati la provi de s the e arliest evidence fo: extens1ve eros1on and o the Columo 1 R1ve Basal o the C o lumb1a Plateau, although eros1o w a suostantia i r the Columbi Gorq beror e Rap1ds time (Beeson and Mer n 197q has an unusual chemical composit1on n racter. z d lower contents of CaO and MgO an c hlQ ne r con ents o Na20, K20, and ::ac l ementr -:: .ar most other flow s in the roup ( taoles an T h content of Ba is 2000 ppm o r more, a l on to identify the member. P .R. Roope:: loral commun .. 1 979 ) has recognized two flows o f sUghtl,. differen: K20, Ti0 2 and Ba contents i the member Th 87sr;86sr initial ratio o the Umat 1 a 1< htg .7092 {McDougal, 19 7 6 Th urn unde lies the Wilbu r Creek Member. The Wilber Creek Member an h Aso 1r Member distinctly differen low!= o u possiD cnen> ically related, were apparentl e -up ed 1 the Clearwater embayment o f Idaho. From there, advanced down valleys an d gorges eadin g 'rom th Uniontown Plateau to the central oa of hP c o lumb1a Plateau; remnants of the valle -f l 1nq 1 n c r east anc wes t of lower Cow :ree l l near Warden a n d Othello, and elsewhere {Swanso anc others, 1 980 ) The flows c osseo h nortn e r pa o. the Pasc Basin ( Myers anc I ) tC(. .:. "1 n rr:ovecdow a c anyon along Yakima Fodgt pos:;iol "'!. f a!: w e., t as Yakima (R.D. Bentle i Swanso ana others. 1S79a ) The flows overli quartZltic o extra-plateau derivation alon9 Ya tma Ridge and rae a westward course of the ancestra Columb1 Rive from the Pasco Basin to Yak1ma. Th Wilbur Cree ha c a major element compositio t that of th lntermediate-Mg Grande Ronde cnemicd 4 an d Asotin is similar to the basalt of Robinett Mounta1r ftable 1); however, trace element compos "t1ons eas ; disc-lmlnate the flows (tabl The Weissenfels Ridge Member overlies t he Asotl Member and contains several flows con ine ma1nl to thP. Lew1ston Basin and presumaol er p ec :her. Th basall o Lewiston Orchards, on flow a veraging 10-15 m thick, is spars ly plagioclas -phyric a contains groundmass olivin visibl witl a hand n lens. I t r la i v ly rich in MgO n d poor i (tabl ) Th over ying basal S l t ppe r consis s of s veral lows, at eas on contain bundan roundm ss oliv 1 n . compos1t1on differs from o her flowc o Mounta1n s B sal 197 7) ha a maJor lemen Lolo c h mical ypc but i h e Th Esquatz 1 M embe r (fig. 2 ) occur. s isolate d remnants o an intracanyon flow along and us nor h of the mode r n Snak R iver upstream fro Devils Canyo ISwanson and others, 1980). It apparen ly was erupted within the ancestral Snak dr 1 nag flowed and en ered th ancestra l Columbia River valley 1n the cen ral part of th plat eau. Th Esquatzel then flow ed along Yakima Ridg simila to that of the W i bur Creek an T h Pomona Membe th ::1 of. cnm1n k ocat a flow a the C o lumb1a Go:ge. southwes Washln ton 1 an Nelson ana other ... 1976 fol low an ances eolumot a dra1na P omona w 0 th 2 ow d1 d Peperite t h m t f s an o':he. vo l Th E Mountal M embe ably e up:_ tr. par fro a dik 1978) tr he T r o bas1r of nortneas i i nowr a . the Wenah flow o f Flows o f th member advar.ced dow t R iver .anvon as the Pomona Member n y. ea The flows mout h o h :anyon red mucr o so t


wing volcanlclastic debris that had been erupted cascades, carried eastward by rivers, lahars, .n nd w inds, and depos1ted on the Pomona Member. :ecent mapping has defined the west and southwest argin of the member alon? a line extending approxiately southward from Yak1ma to the Horse Heaven plateau (Swanson and others, 1979a). The Elephant consists of several flows, all chemically similar of normal and transitional magnetic polar itY Its major element composition is similar to that of Rosalia chemical type except for lower PiJS (table 1), but its trace element content is distinct. The member has a 87sr;86sr initial r atio of about 7078 (McDougall, 1976; Nelson and o thers, 1976), similar to that of the Pomona. The Buford Member (fig. 2), a single magnetically r eversed flow 20-30 m thick, is the youngest known basalt on the plateau surface of extreme southeast and northeast Oregon, where it is confined and presumably was erupted. Its major-element compo sition shows distinctly higher light REE contents ( table 2). Ice Harbor Member (fig. 2), dated as about 8.5 m.y. old (McKee and others, 1977), was erupted the central part of the Columbia Plateau, where dikes and remnants of vent areas have been recognized. The last previous eruptions from the central part of the plateau were those of the Frenchman Member, about 6 m.y. before the Ice Harbor volcanism. Most flows are confined to the area of v enting, but at least one flow spread westward to the Richland area and southwestward to Wallula Gap. The Ice Harbor Member can be subdivided into three readily mappable units of different compositions (tables 1 and 2). The lowest unit--the basalt of Basin City--contains plagioclase and olivine phenocrysts, as normal magnetic polarity, and is chemically distinct. The middle unit--the basalt of Martindale-carriers clots and single crystals of clinopyroxene, plagi oc lase, and olivine and has reversed magnetic two related compositions characterize the Martindale (Helz, 1978), the dominant of which is is ted in table l. The upper unit--the basalt of Goose Island--contains sparse plagioclase and maget ite phenocrysts, has normal magnetic polarity, and has the most Feo-r ich composition of any known flow in the group (and one of the most Feo-r ich compositions o f any terrestrial basalt). The three informal units have similar 87sr;86sr initial ratio of aoout 7077 (Helz, 1978). Helz (197 8 ) has recently mpleted n e xhaustive experimental and petrogenetic s udy of h I Harbor Member. The Ice Harbor vent system is boo 0 km long and has s rong aeromagetic exprE:>ssi n ( Swanson and o hers, l979c). The Lower M onu m n al Member, about 6 m.y. old, is he mber in he Saddle Mountains Basalt. o h modern Snake River Canyon b e we nyon and s o in, Washington, a dis50 km. I s source as presumably bu has yet been omposition is -lmilar to liqht y higher i n g t REE ( ables \ i gh 110 9 'NE:>lson PHYSIC' L HARACTERtST CS F LOWS Gr nd t.ton c1e, Wan:tpum, 3 n d frns1stem. ::entatio, th n h ose i n he colonnade. t many n:a latures re 'undled into fans, e n t s ther un Jsu. ly haped arranaements. 1o;t olumns in an enta I l ure are highly eqmented n v Jross }Oln s so that the coltlll\T\s a r roadily 1 to fis size pieces. rhe en t a 1.-_; ,... :1ener 1 11 comprises a bou 70 o.. rr-pn '='


thickness of a flow but can make up 00 percent (one example that we know of) to zero percent. The upper part of the entablature is scoriaceous and commonly merges into a zone of short, wide, generally poorly defined co umns that some workers call the upper colonnade. A rubbly, clinkery zone occurs above the entablature of some f ows. Such a zone is absent from the base of flows. The origin of this rubbly zone is in question. The rubble is, in our experience, more common near vent areas than elsewhere; this observation suggests that the rubble may represent material thrown out of the vent near the end of eruption and modified during movement of the flow. Idealized jointing patterns can be satisfactorily explained by existi ng theory for the cooling of bodies of igneous rock {Jaeger, 1961), but such patterns are seldom found in nature. Acceptable thermo mechanical explanations for the typically complex jointing patterns, particularly in the entablature, are not available despite considerable descriptive information (Tomkeieff, 1 940; Waters, 1960; Mackin, 1961; Spry, 1962; Swanson, 1967; Schmincke, l967b; Long, 1978; Ryan and Sammis, 1978). Complications related to the mutual interference between columns growing inward from irregular contacts, pending of water on a flow surface and percolation down joint planes during solidification, the influence of chemical composition on tensile strengths and heat conduction, and inadequate knowledge of rock mechanics under high temperature-low pressure conditions are some of the difficulties that plague attempts at analysis of natural jointing habits. Some flows have a tiered appearance defined principally by alternating layers of vesicular and relatively nonvesicular rock rather than by joints. These layers may record separate gushes or thin flows that piled up and solidified as a single compound cooling unit. Tiers much more commonly occur in entablatures than colonnades. The upper surface of a flow is rarely exposed in plan view. Where seen, the surface is rather flat, smooth, filamented, and locally ropy--surface features characteristic of lava ponds at Kilauea. The surface of a flow with a rubbly upper zone is rough, has a relief of as much as 6 m, and otherwise appears unlike typical surfaces of ponded flows. Many flows entered water and formed pillows. Recent studies {Jones, 1968; Moore, 1975) have demonstrated conclusively that pillows are nothing more than the subaqueous equivalent of pahoehoe toes. Many of h pillowed flows occur near he margin of the plateau as it existed at the time of eruption, apparently because lakes were formed as flows ponded rivers draining from marginal highlands. Other pillowed flows are much more extensive, perhaps signifying entry into shallow lakes standing on the plateau surface. One such extensive flow, in the Priest Rapids Member, is pillowed throughout an area of tens of square kilometers in the Cheney-Palouse scabland southwest of Spokane. In places, lava deltas (Fuller, 1931; Moore and others, 1973) formed as lava poured into shallow lakes and ponded streams. The direction of dip of foreset "bedding" defined by elongate pillows and thin sheet flows in the lava deltas the local flow direction of the lava. Particularly good examples of lava deltas can be seen near Malden south of Spokane (Griggs, 1976), near the mouth of Moses Coulee {Fuller, 1931), and at the mouth of Sand Hollow south of Vantage. 10 Other cri eria for efining flow directions i nclude i ncUned pipe v sicles, which plunge rent, and bent spiracles (fig. 6), formed by steam blasts eneath flow, which tail out down-current Flow directional da a for basalts must be treated ln way as those for current-produced 1n sed1mentary rocks--carefully. A few data in a small area show the local direction but say little about regional patterns. Nonetheless, careful studies by Schmincke (l967a) and on-going work by othPrs are succeeding i n defining patterns of lava advance within the plateau. FEEDER DIKES, VENT SYSTEMS, AND ERUPTION One of the major results of recent mapping has been identification of sources for most stratigraph' c units and even single flows. Feeder dikes have been found for flows in all formations and members of the Colt:mbia River Basalt Group except the Wilbur Creek, Asotin, Esquatzel, Buford, and Lower Monumental Mem bers of the Saddle Mountains Basalt. The feeder dikes avfHage about 8 m wide but vary from a few centimeters to more than 60 m. They may tend to thin upward, but this is far from cer ain. The dikes generally trend north to north-northwest. They cannot be traced far along strike, in part because of exposur problems. Obviously related dike segments, offset a few meters to form an en echelon pattern, form systems extending tens of kilometers. Compound or multiple dikes, consisting of two or more pulses of magma re ated to the same intrusive event, are common, but composite dikes, containing two or more phases of contrasting compositions, have not been -reported. In other words, each fissure was used just once, not repeatedly. The chance of finding a dike connecting with a flow i fed is small, owing to problems of exposure. Nonetheless, several dikes have been found displaying such a connection. The top of most such dikes is rubbly, apparently consisting of slabs of crus once floating on a flow before it poured back into the fissure {for example, Plate la in Swanson and others, 1975). In one dike lacking such rubble, the dike merges imperceptibly with its flow of the Springs Member {fig. l, number 6, Swanson and others, 1975). Many other dikes can be inferred to correlate with particular flows, or at least sequences of flows, on the basis of chemical composition and mag netic polarity. In this way, feeders have been identified for most of the named stratigraphic units. The shape and extent of vent systems for specific flows or related sequences of flows can be reconstructed from the locations of feeder dikes, thick piles of degassed flows (presumably near their vent ) abundant collapsed pahoehoe (Swanson, 1973), and accumulations of basaltic tephra (in places occurring in still recognizable spatter cones and ramparts) Such reconstructions show that eruptions of single flows or related flows took place from fissures concentrated in long, narrow vent systems on the order of tens of kilometers long and several kilometers wide {Swanson and others, 1975, and later work). Vent systems for the Grande Ronde Basalt are distributed across the eastern half to two-thirds of the Columbia Plateau, but those for other units are more restricted { Swanson and Wright, 1 979). For example, feeder dikes for the Picture Gorge Basalt are confined within the John Day Basin and neighboring areas, those for he Frenchman Springs Member w i hin


e 6 o km w i de and probably more than 200 km long, 1 f or the Ice Harbor Member within a zone 15 !.. d and about 90 km long. On a still finer VI e vent systems for specific flows are nearly sea e as they occur along single dikes or closely near, ced related dikes. Examples are the vent Roza Member (probably less than 5 km w1de and t o be more than 165 km long Swanson and 5 1 975; P.R. Hooper, unpub. data, 1978 ) the : of Robinette Mountain in the Eckler Mountain r ( a singl e dike extending at least 25 km across :;es lueMountains), basalt.of Basin City in the Harbor Member (poss1bly a s1ngl e dike at least 50 ong) and several chemically. distinct flows in eGrande Ronde Basalt (T.L. Wnght and D.A. Swan unpub. data, 1978). o f t h e geometry of the vent systems not y allows predi ction of where vents for specific ,0115 should occur once one such vent i s found but ;so furth e r prov ides important constraints regarding generation, storage, and eruption mechan ics. s ing suc h knowledge, at tempts have been made to the rates of eruption and advance for single ; s. The estimates take into account the .. that flows, even those that advanced tens of kilometers from their sources, fJenched to a glass tey entered water; th1s 1nd1cates l1ttle coollng :Jring transport and hence rapid advance, since the ava apparently moved as sheet floods rather than rough insulating tube systems. Application of models, developed in part from this obser vation by Shaw and Swanson (1970), to vent systems of i dimensions suggests eruption rates of about l 3 j d a y pe r linear kilometer of active fissure for argest flow s s uch as those in the Roza Member, about lo-4km3/day /km for the smaller flows svanson a n d others, 1 9 7 5 ) For flows of "average" '10 ume, prob ably several tens of cubi c kilometers, rates of lo-1 to l o-2km3/day/km may be inferred. 3y comparison, sustai ned rates of eruption at Kilauea and Mauna Loa a r e lo-3 to l o-4km3 /day/km. : s ing observ e d dike widths, theoretical modeling s qqests tha t the high eruption rates coul d indeed ave bee n sustained by supply from depth ( Shaw and Sllanson, 1 9 7 0 ) Such eruptions probabl y lasted a few ays. Flow rate s of 5 to 1 5 km/hr down slopes of : 000 are calculate d from the model, adequate to a f lows to move far with little cooling. eruption rates do not necessarily imp l y rapid meltin g rates in the mantle. Flows were Hupte d only once every ten thousand years or so in area on the plateau during e ven t h e peak o f '10 c anic ac ivity (Grande Ronde t ime) as est imated the number of flows i n a magne t o s t rati ;raphic unit of assumed duration b ased on comparison h seafloor magnetic anomalies o f roughly compar e age. Calcula ion s s how that conti nu o u s melting at the presen t Hawaiian rate, lo-lkm3/yr ( Swanson, : 2 ) could accoun for t h e volum e o f t he Columbia ive r Basalt Group in the allote d t i m e Episod i c rap1 d melting ven ( "flash melting" ) c a n no t be !xcluded but are no r e qu ired If melting progressed at th Rawa iian rate, then deep storage reservoirs are required in order :D account for he 1 rge volum of single low s .... s contras s with th H waiian situation, where !ruptions ar much more frequen t and lava "leaks" t o e surface more or less continuously. The presence J f arg deep stor g r eservoirs, possibly in the .pper mantle, may be a princ1pal and distinguishi ng : arac eristic of flood-b salt provi n c s in general. 11 CHEMICAL PETROLOGY OF THE YAKIMA BASALT SUBGROUP Two major geochemical breaks occur within the Y ak ima Basal t Subgroup. The older break separates a igh-Si02, l ow-FeO and Ti02 sequence below (Grande Ronde Basalt) from a relatively low-Sio2 h i gh-FeO and Ti02 sequence (Wanapum Basalt). This chemical change took place over a short time, as magnetic stratigraphy is continuous across the break, K-Ar ages above and below agree within analytical error for these rocks (+ l m.y.), and flows of different chemical types locally interlayered (fig. 4). Trace-element levels are similar and Sr isotopic ratios are similar and relatively low ( 0 .704-0.706) i n both sequences. Major-and trace-element chemistry of the Grande Ronde Basalt, which makes up 75 percent of the volcanic pile, shows cyclic variation rather than an evolutionary trend. The younger geochemical break, separating the Wanapum and Sadd l e Mountains Basalts, i s marked by l) increase in Sr isotopi c ratios to 0.708-0.715, and 2) greater abundance of most incompatible elements and steeper chrondrite-normalized REE patterns in those flows of the Saddle Mountains Basalt whose major oxide chemistry is similar to older flows. These changes took place over a relatively long time (.5-l m.y.?) as there is evidence of erosion and deposition of sediments between the Wanapum and Saddle Mountains Basalts The Columbia River Basalt Group shows far less c9herent chemical variation than is typical for oceani c tholeiites. Ratios of incompatible elements vary widely among the various formations as well as among flows belonging to the same formation. Enrichment factors differ for different incompatible elements in units which show smooth chemical variation trends. For examp le, the least magnesian flows of the Grande Ronde Basalt ( Mg0r3 percent), have contents of PzD5, Hf, Ta, and LREE that are 2-3 times h igher t han i n the most magnesian f l ows of the Grande Ronde ( Mgor6 percent) whereas the enrichment of Th and K in these same flows is greater, about 3-4. We make the following inferences as a start toward a model to explain the generation of magmas in the subgroup. 1. We infer that magma compositions are controlled principally by partial melting. Enrichment factors in the Grande Ronde Basalt are not consistent with a fractionation model in which crystallization o f 55-70 percent of stored liquid would have to take p lace repeatedly. Mixing calculations fail to balance both major o x ides and incompatible trace elements usin g any reasonable fractionating mineral assemblage. 2 The scarcity of phenocrysts, absence of high-pressure "megacrysts", and absence of fractionation o r accumu lation t rends suggest that magmas accumulated near the site of melting and may have been s uperheat e d during transport to within a few kilomete r s of the surface. 3 w e i n fer littl e o r no high-level crustal storage f o r most of the magma, a s evidenced by l) abse n c e o f grabe ns o r calderas associated w i t h s ingle flows of large vo lume a n d 2 ) absence of phenocrystrela ed chemica l fractionation or accumulation trends e v e n betwe en phyri c a nd aphyric flows of the same strat i g r aph i c un it. We feel that p lagioclase pheno cryst s present in flows may have formed du r i ng


ascent of magma just prior to eruptio n a nd perhaps even after eruption. 4. If a partial melting model is accepted, melting must be relative l y "wet" to yield quartz-normative magmas. Some metasomatic enrichment of trace elements is required to explain the incoherent ratios of incompatible elements and possibly the highest 87sr;86sr ratios. 5. Bulk-lava chemistry suggests that the source for most of the magma was relativel y iron-rich and olivine-poor clinopyroxenite. The major chemical and stratigraphic units and many of the individual flows require chemically distinct source rocks. The overall source for magmas of the Col umbia R iver Basalt Group must be heterogeneous both in space and through time. INVASIVE FLOWS Weakly consolidated sedimentary rocks, generally medium-grained sandstone to siltstone, occur between many flows near the margin of the plateau and between some of the younger flows in local structural basins on the plateau. The sedimentary rocks rest depositionally on the underlying flow in some places, but in many other places the contact relations show that subjacent basalt intrudes or invades the sediment. Schmincke (1967c) was one of the first to recognize this, and recent work has demonstrated how common such invasive relations are. We estimate that more than half of the observed contacts between basalt and sedimentary rocks on the Columbia Plateau are invasive. How are such contacts interpreted? Do they s ignify "normal" intrusive relations in which basaltic magma never reached the surface before solidifying as in classic dikes and sills, or are they formed as lava flows burrow into unconsolidated sediments accumu lating on the ground surface (invasive flows of Byerly and Swanson, 1978)? Both processes produce similar results. The key to proper interpretation lies in the stratigraphy. If the basalt is at its proper stratigraphic position relative to overlying flows, it almost certainly was a flow that invaded sediments at the ground surface. This is because thin sedimentary deposits, generally less than 10 m thick on the plateau, are light and hence exert little confining pressure; vesiculating magma rising and encounteri ng such sediments would certainly blast through rather than spread laterally into h m. Mapping and chemical studies have shown that, in every exa.mple so far found on the p lateau, the invasive basalt is at its expected stratigraphic position relative to overlying flows and hence is in invasive flow (Schmincke, 1967c; Camp, 1976; Byerly and Swan son, 1978; D.A. Swanson, G.R. Byerly and T.L Wright, unpub. data, 1978). We include in this interpretation two thick sills previously interpreted conventionally, the Hammond sill of Hoyt (1 96 1) near Wenatchee, Washington and the "Whiskey Creek sills" of Bond (1963 ) Work by Byerly and Swanson (1978 and unpub. data) and V.E. Camp (unpub. data, 1 978 ) shows that even these thick sills, more than 120 m thick locally, are in proper stratigraphic position relative to overlying f l ows. These conclusions are significant, because they show that invasive contacts ptovide insufficient, i n fact totally m isleading, evidence E r he former 12 pres nee o m agma benea h h e ar a. For example invasive relations are particularly well displayed along the northw st margin of the plateau, where all other evidence negat s the former presence of magm no intrusive cont.cts are found h re in an; s1tuat1on other than o n e 1nvolving sediments, the invadi ng flows ar in heir proper stratigraphic position relative to other flows, and flow directions show that surface flows moved toward, not away from correlative sills. Invasive relations also demon' strat the unre iability of using sedimentary interbeds as stratigraphic guides on the Columbi a flows are always in their strat1graph1c pos1t1on, but the interbed s at least fine-grained ones are commonly no t owing to rafting by invasiv e flows. REFERENCES CITED Beeson, M.H., a nd Moran, M.R., 1979, Columbia River Basalt Group stratigraphy in western Oregon: Oregon Geology, v. 41, p. 11-14. Berggren, W.A., and van Couvering, J.A., 1974, The Late Neogene: B iostratigraphy, geochronology, and paleoclimatology of the last 15 million years i n marine continental sequences: Palaeogeography, Palaeocllmato1ogy, Palaeoecology, v. 16, p. l-216. Bingham, J.W., and Grolier, M.J., 1966, The Yakima Basalt and Ellensburg Formation of south-central Washington: U.S. Geol. Survey Bull. 122 4-G, 15 p. Bond, J.G., 1963, Geology of the Clearwater embayment: Idaho Bur. M ines and Geol., pamphlet 128, 83 p Brown, C.E., and Thayer, T.P., 1966, Geologic map of the Canyon City quadrangle, northeastern Oregon: U.S. Geo l Survey Misc. Geo l Invest. Map I-447, scale 1:250,000. Byerly, Gary, and Swanson, Don, 1978, Invasive bia River basalt flows along the northwestern margin of the Columbia Plateau, north-central Washington: Geol. Soc. America Abstracts with Programs, v. 10, no. 3, p. 98. Camp, V.E., 1976, Petrochemical stratigraphy and structure of he Columbia River basalt, Lewiston Basin area, Idaho-Washington: Washington State Univ., Pullman, Ph.D. Dissert., 201 p. Camp, V.E., Price, S.M., and Reidel, S.P., 1979, Descriptive summary of the Grande Ronde Basalt type section, Columbia River Basalt Group: Rockwell Hanford Operation s Report RRQ-BWI-LD-15, 24 p. Cockerham, R.S., and Bentley, R.D., 1973, Picture Gorge and Yakima Basalt between Clarno and Butte Creek, Oregon: Geol. Soc America Abstracts with Programs, v. 5, no. 2, p. 23. Fruchter, J.S., and Baldwin, S.F., 1975, Correlations between dikes of the M onument swarm central Or on, nd Pic ur Gorg Bas lt fl s: Geol. Soc. America Bull., v 86, p. 1 4-516. Fuller, R.E., 1 93 1 The a q ue ou s chilling of basaltic lava on the Columbia River Plateau: Am. Jour. Sci., v 21, p. 281-300. Greeley, Ronald, 1977, Basaltic "plains" volcanism, 1n Greeley, Ronald, and King, J.S., eds., Volcanism of the eastern Snake River Plain, Idaho, a Comparative Planetary Geology Gu1debook: Office of Planetary Geology, N.A.S.A., Washington, D.C., p. 23-44. Griggs, A.B., 1976, The River Basalt Group 1n the Spokane quadrangle, Washington, Idaho, and Mon tan a : U.S. Sur ve y Bull. 1 413, 39 p He1z, R .. 978, The e rogenesis o f he Ice arbor Member, Pate u. WashingtonA hem1ca and experimental tudy enn. State Univ., ntver sity Park, Ph.D 284 p. Helz, R T., Wognt, .... nd S wanson, D.A., 19 6. Petroqene :t aqntfic.H. ::e f chemical tren n h y ou nqes un1 {aklMa 1 n the Pl teau, .wesr J .. '). onzales .-d .


Inter. Assoc. Volcanol. Chern. Earth's Interior, Sym posium on Andean and Antarctic Volcanology, Proc., p 465-481. G.M., 1964, Geology and ground water of the umatilla River basin, Oregon: U .S. Geo Survey wate r -Supply Paper 1620, 16 2 p Hol den G. S., and Hooper, R., 197&, Petrology and hem1stry of a Columbia River basa l section, Rocky can on, west-central Idaho: Geo Soc. America aul ., v. 87, p. 215-225. Hcope r P.R. 1974, Petrology and chemistry of the Rock Creek flow, Columbia River Basalt, Idaho: Geol soc. America Bul ., v. 85, p. 15-26. aoyt, c.L., 1961, The Hammond sill-an in the Yakima Basalt near Wenatchee, Washington: Northwest Science, v. 3 5 p. 58-64. J a ckson, D.B., 1975, Description of the geoelectric section, Rattlesnake Hills unit 1 well, Washington: Jour. Research U.S. Geol. Survey, v. 3, p. 665-669. J.C., 1961, The cooling of irregularly shaped igneous bodies: Am. Jour Sci., v. 259, p 721-734. J ones, J.G., 1968, Pillow lava and pahoehoe: Jour. Geology, v. 76, p. 485-488. Kleck, W.O., 1976, Chemistry, petrography, and stra tigraphy of the Columbia R iver Group in the Imnaha Rtve r valley region, easter Oregon and western Idaho: Washington State Un1v., Pullman, Ph.D. 01ssert., 20 3 p. Lerebvre, R.H., 1970, C olumb1a R1ver basalts of the G r an Coulee area, in Gilmou r E.H and Strad ling, Dale, eds., Proc. Second Co lumbia R iver Basalt S vmposium: Cheney, E a ster Wash. State Colleg Pr ess, p. l-38. L1ndsle D H., Smi t Doug las, and Haggerty, S E., !971. etrography anc mineral chem1stry of a d iffer e ntiate f low of P 1cturc Gor g e Basal near Spray, Oregon: Carnegie Ins i ut o Washington, Yearbook 69 264-285. .E., 1978, Characterization and recognitio of structures, Grande Ronde Basalt: Rockwe l Hanford Operations In ormal RBO-BWI-LD-10, 7 l fMackin, .H., 196 A straoqraphlC. section 1n th Ya irna Basal and th Ellensbur Formation In sout Washington: Wash1nato. D Mines and Geolog} Report Inv. lS, 4 Ian, 1976, a orig1n of o sal o the Columb a R v e r ucou Orego ana Wasnington: Geo oc. Americ Bu :., v 87, p. 7-792. M cKee, E.H., Swanson, u.A., and Wrignt, T.L., 1977, Duration and volume ot C o lumbia R iver Basalt volcanlsm, Washington, Oregon, ana Idaho: Geol. Soc. America Abstracts Programs, 9, no. 4, p. 4 4 6 .:.G., 1975, Mechanisr tJf tormation of AA.. Scien is , ,-_:., t 269-277. G., ....... G::1G R.W. Peterson, a n d S wanson, .A., l FL of l ava 1nt o the sei. l 96 -197 K i a e a Vu1.cano. HawaiI: Geo Soc. Amenca Bull., v 8 4 J:-. jJi-:>4t.. M y r!., .. W., and Pr1c S .l'l . 1979, Compi ation ge olOCJlC map o the Pascc BasH . south-central Wash Ing on. Rockwell Hantor0 Opera 1ons, RHD-BWI-ST-4, seal. :62,500. Nathar .. Simon, and F ruchter .J. S 1974, Geochemical pa leomagnetic str tigraph ot the Picture Gorge ima Basalts (Columbia R1ver Group) 1 Geo l Soc. America Bul ., v. 85, N lsor, D O., Swanson 0.1.. and W r1ghl, T L., 1976, St c) ium isotopic c ompo s 1 ti o f 1ntracany o n flow o) 't, 1ma Basal southea t Wasn1ngton: Geo Soc. ca Abst r act with Pro ram no. 3 1. : 9. 13 Newman, K.R., 1970, Palynology of interflow sediments from the Standard Oil Company of California Rattle Snake Hills no. l well, Benton County, Washington, in Gilmour, E.H., and Stradling, Dale, eds., Proc. Second Columbia River Basalt Symposium: Cheney, East. Wash. State College Press, p. 201-207. Osawa, M., and Goles, G.G., 1970, Trace-element abundances in Columbia River basalts, in Gilmour, E.H., and Stradling, Dale, eds., Proc. Second Columbia River Basalt Symposium: Cheney, East. Wash. State College Press, p 173-175. Price, S.M., 1977, An evaluation of dike-flow correlations indicated by geochemistry, Chief Joseph Swarm, Columbia R iver basalt: Idaho Univ., Moscow, Ph.D. Dissert., 320 p. Raymond, J.R., and Tillson, D.O., 1968, Evaluation of a thick basalt sequence in south-central Washington: Atomic Energy Cormnis. Research and Development Rep BNWL-776, 126 p. Reidel, S.P., 1978, The stratigraphy and petrogenesis of the Grande Ronde Basalt i n the lower Salmon and adjacent Snake R iver canyons: Wash. State Univ., Pullman, Ph.D. thesis, 415 p. Ross, M.E., 1978, Stratigraphy, structure, and petrology of Columbia River basalt in a portion of the Grande Ronde-Blue Mountains area of Oregon and Washington: Idaho Univ., Moscow Ph.D. D issert., 413 p. Ryan, M.P., and Sammis, C.G., 1978. Cyclic fractur mechanisms in cooling basalt: Geol. Soc. America Bull., v. 89, p. 1295-1308. Schmincke, H.-U., 1967a, Flow directions in Columbic basalt flows and paleocurrents of interbedded sedimentary rocks, south-central Washington: Geo1 Rundschau, v. 56, p. 992-1020. 1967b, Stratigraphy and petrography of four upper --yaima Basalt flows in south-central Waghington: Geo l Soc. America Bull., v. 78, p. 1385-1422. l967c, Fused tuff and peperites in south-central Wash ington: Geo l Soc. America Bul l., v. 78, p. 319-330. Shaw, R.R., and Swanson, D.A., 1970, Eruption and flow rates of flood basalts, in Gilmour, E.H., and Stradling, Dale, eds., Proc. Second Columbia River Basal t Symposium: Cheney, Eastern. Wash. State College Press, p 271-299. Snavely, P.O., Jr., MacLeod, N.E., and Wagner, H.C., 1973, Miocene tholeiitic basalts of coastal Oregon and Washington and their relations to coeval basalts of the Columbia Plateau: Geo l Soc. America Bull., v. 84, p. 387-424. Spry, A., 1962, The origin of columnar jointing, particularly i n basalt flows: Geol. soc. Australia Jour., v. 8, p. 191-216. Swanson, O.A., 1972, Magma supply rate at Kilauea Volcano, 1952-1971: Science, v. 175, p. 169-170 1 973, Pahoehoe flows from the 1969-1971 Mauna Ulu --er-uption, Kilauea Volcano, Hawaii: Geol. Soc. America Bull., v. 83, p. 615-626. Swanson, D.A., and Wright, T.L., l976a, Guide to field trip between Pasco and Pullman, Washington, emphasizing stratigraphy, vent areas, and intracanyon flows of Yakima Basalt: Geol. Soc. America Cordilleran Sect. Mtg., Pullman, Washington, Field Guide no. l, 33 p. l976b, Magnetostratigraphic units in the Yakima --sa5alt, southeast Washington: Geol. Soc. America Abstracts with Programs, v. 8, no. 3, p. 413-414. 1978, Bedrock geology of the northern Columbia and adjacent areas, in The Channeled Scab land, Baker, V.R., and Nummedal, Dag, eds.: N.A.S.A. Office Space Sci., Planetary Geol. Program, Washington, D.C., p. 37-57. 1979, Source areas and distribution of major --units in the Columbia River Basalt Group: Geol.


soc. America Abstracts with Programs, v. 11, no. 3, p. 131. swanson D.A. Wright, T .L., and Helz, R.T., 1975, Linear vent systems and estimated rates of magma production and eruption for the Yakima Basalt on the Columbia Plateau: Am. Jour. Sc i v. 275, p. 877905. swanson, D.A., Wright, T.L., Camp, V.E. Gardner, J .N., Helz, R.T., Price, S.A., and Ross, M.E. 1977, Reconna issance geologic map of the Columbia River Basalt Group, Pullman and Walla Walla quadrangles, Washington and adjacent Idaho: U.S. Geol. Survey Open-fil e Rept. 77-100, scale 1:250,000. swanson, D.A., Anderson, J.L., Bentley, R.D., Byerly, G.R., Camp, V.E., Gardner, J.N., and Wright, T.L., 1979a, Reconnaissance geologic map of the Columbia River Basalt Group in eastern Washington and northern Idaho: U.S. Geol. Survey Open-file Rept. 79-1363, scale 1:250,000. Swanson, D.A., Wright, T.L., Hooper, P.R., and Bentley, R.D., l979b, Revisions in stratigraphic nomenclature of the Columbia River Basalt Group: U.S. Geol. Survey Bull. 1457-G, 59 p. Swanson, D.A., Wright, T.L., and Zietz, Isidore, l979c, Aeromagnetic map and geologic interpretation of the west-central Columbia Plateau, Washington and Oregon: u.s. Geol. Survey Geophys. Invest. Map. G.P. 917, scale 1:250,000. Swanson, D.A., Wright, T.L., Camp, V.E., Gardner, J.N., Helz, R.T., Price, S.M., Reidel, S.P., and Ross, M.E., 1980, Reconnaissance geologic map of the Columbia River Basalt Group, Pullman and Walla Walla quadrangles, southeast Washington and adjacent Idaho: U.S. Geo l Survey Misc. Geo l Invest. Map. I-1139, scale 1:250,000, in press. Thayer, T.P., and Brown, C.E., 1966, Local thickening of basalts and late Tertiary silicic volcanism in the Canyon City quadrangle, northeastern Oregon i n Geological Survey Research 1966: U.S. Geol. Survey Prof. Paper 550-c, p C73-C78. Tomkeieff, S.I., 1940, The basalt lavas of the Giant' s Causeway district of northern Ireland: Bull Volcanologique, v. 6, p. 90-143. Vallier, T.L., and Hooper, P.R., 1976, Geologic guide to Hells Canyon, Snake River: Geo l Soc. America, Cordilleran Sect. Mtg., Pullman, Wash. Field Guide no. 5, 38 p. Walker, G.W., l973a, Reconnaissance geologic map o f the Pendleton quadrangle, Oregon and Washington: U.S. Geol. Survey Misc. Geo l Invest. Map. I-727, scale 1:250,000. l973b, Contrasting compositions of the youngest Columbia River basalt flows in Union and Wallowa Counties, northeastern Oregon: Geo l Soc. America Bull., v. 84, p. 425-430. Waters, A.C., 1960, Determining direction of flow in basalts: Am. Jour. Sc i v. 258-A, p. 350-366. 1961, Stratigraphic and lithologic variations 1 R i ve r basalt: Am. Jour. sc ., v. 259, p. 583-61 1962, Basalt magma types and their tectonic asso --ciation--Pacific Northwest of the United States: Am. Geophys. Union, Mon. 6, p 158-170. Watkins, N.D., and Baksi A.K. 1 974, Magnetostratigraphy and oroclinal folding of the Columbia River, Steens, and Owyhee basalts i n Oregon, Washington, and Idaho: Am. Jour. Sci., v. 274, p 148-189. Wright, T.L., and Hamilton, M.S., 1978, A computerassisted graphical method for identificati on and correlation of igneous rock chemistries: Geo logy, v. 6, p 16-20. Wright., T.L., Grolier, M.J., and Swanson D.A., 19 73, Chemical variation related to the stratigraphy of the Col umbia R iver basalt: Geo l soc. America Bull., v. 84, p. 371-386. 14 Wright, T L. S wanson, D.A., Helz, R.T., and Byerl G.R., 1979, Major oxide, trace element, and chemistry of Columbia R i ve r basalt samples between 1971 and 1977: U.S Geol. Open-file Rept 0 79-711.


ROADLOG FOR GOOLOGIC FIELD TRIP BETWEEN LEWISTON, IDAHO AND K IMBERLY, OREGON MILES (0. 4) 0.4 (0.95) 1. 35 (0. 2) l. 55 (0.2) 1. 75 (1. 7) 3.45 (0.55) 4.0 (0. 95) 4.95 (1.1) 6.05 (0. 95) 7.0 (2. 55) 9 55 (0. 55) 10 (0.1) 10.2 (0. 3) 10.5 (2.3) 12.8 D.A. Swanson, U.S. Geological Survey, Menlo Park, CA 94025 and T.L. Wright, U.S. Geological Survey, Reston, VA 22092 Sacajawea Motor Lodge in Lewiston, Idaho (fig. 7). Drive west. Tapader a Inn. Continue west, following signs to u.s. 12. Turn left to west U.S. 12. Cross Snake River. Enter Clarkston, Washington. The Snake here is now a lake, impounded behind the recently completed Lower G ranite Dam. Continue straight ahead on U.S. 12 toward Pomeroy and Walla Walla. At 12:00, the Pomona Member of the Saddl e Mountains Basalt fills ancient canyon of Snake River eroded into flows of the Grande Ronde Basalt (R2 magnetostratigraphic unit). At 11:00, house on bluff rests on Lower Monumenta l Member, which was confined to an old Snake River canyon carved deeply into the older flows between about 10.5 and 6 m.y. ago. Roadcuts in Grande Ronde Basal t for several miles. On left, colonnade of the canyon-filling Pomona Member. Note underlying white ash, a probable correlative of which occurs near Ice Harbor Dam (stop 9). At 1:00, steeply dipping flatirons of the Grande Ronde Basalt near core of the Gaging Station anticline. This route passes through the Lewiston Basin, a structurall y comp lex area recently studied by Camp (1976). Flows at or near river level along the road also occur near the top of the escarpment north of the river. A major fault follows the west-flowing river, and a faulted anticline north o f the river exposes the lower three magnetostratigraphic units of the Grande Ronde Basal t as well as the upper part of the underlying Im naha Basalt. The latest tectonism occurred in post-Lower Monumental time, as canyon-filling remnants of this 6 m.y. member have been deformed (Camp, 1976). Quarry on left in the Pomona Member. Good exampl e o f c o lonnade and entab lature jointing habits, generally well developed in intr acanyon flows. Look back at 4:00-5:00 and see the Gaging Statio n a nticline i n core o f Lewiston Basin structure. Bluff on left is the Lower M onumental M ember At 2: 00, view down Snake River of the Grande Ronde Basalt Road leaves Snake River and follows Alpowa Creek. Road crosses approximate axis of Gaging Station anticline, rapidly dying out west of the Lewiston Basin. 15


SADDLE MOUNT A INS P A S C 0 8 AS I N 0Heppner 0Eitopia 11 7 \ L a Enterprise Grande 10 20 40 50 MILES r 1 r r1 0 10 20 30 40 5 0 60 KILOMETE R S ---Route of field trip 2 Stop Figure 1. Map showing route of field trip, scheduled stops, and locations of important geographic features. 16


(1. 3 ) 1 4. l ( l. 9) 16.0 ( 1. 25) 17.25 ( 0. 7) 17.95 ( l. 55) 19.5 (0. 55) 20.05 ( 0. 45) 20.5 ( 0. 35) 20.85 (0. 2) 21.05 (0. 3) 21.35 (0.15) 21.5 (0.15) 21.65 (1. 25) 22.9 (0. 3) 23.2 ( l. 5) 24.7 (0. 8) 25.5 ( 0. 85) 26.35 (l. 55) 27.9 (5.45) 33.35 ( 4. 05) 37.4 (0. 4) 37.8 Note sjgn on left commemor a t ing passage of Lewis and Clark on their return trip. Intersection with Clayton Gulch road. Continue on u.s. 12. Intersection with Howell Grade road. From here, u.s. 12 passes upsection across the upper two magnetostratigraphic units of the Grande Ronde Basalt capped by the Roza Member of the Wanapum Basalt near Alpowa Summit. An unknown but observant geologist-polluter painted yellow letters on each flow i n the section, with L the oldest and E (Roza Member) the youngest. The dip of the section is only slightly less than the gradient of the road. Base of flow L. R2. Base of flow K. R2. Base of flow J. N2. Base of flow I. N2. Base of flow H. N2. Base of flow G. N2. Base of flow F. N2. Road cut on left shows thick ancient soil (saprolite) developed on flow F. This saprolite, marking the top of the Grande Ronde Basalt, is overlain the Roza Member. Flow E (Roza Member) exposed from here to summit. Alpowa Surrmit, elevation = 2785 ft. Intersection with Ledgerwood Road. Turn right. Small outcrop of Roza spatter on right (near-vent locality 13 in Swanson and others, 1975, table 1) by Junction with Kirby-Mayview Road. Continue straight (northward) on pavement. Junction with Valentine Ridge Road. Continue straight. Valentine Ridge Road leads to stop 18 of Swanson and Wright (1967a), an interesting vent area for a flow in the Grande Ronde Basalt. Roadcut just beyond junction with Connell Hill Road. Thin flows of the Roza Member, with little or no spatter. Experience shows that, in this area, such thin flows are likely to be near a vent. Continue on Kirby Mayv iew Road. Intersection with Tramway Road. Turn right toward Tramway. Cliffs ahead expose a flow of LOlo chemical type in the Priest Rapids Member of the Wanapurn Basalt. Stop at white stile. STOP NO. 1. Gracefully cross stile and walk to lip of Snake River canyon for view of Grande Ronde flows and discussion of stratigraphy and mapping methods. Quarry is in a flow of Lolo chemical type in the Priest Rapids Member. 17


(0.45) 38.25 (0.9) 39.15 ( 0.]) 39.25 (2.1) 41.35 (0.6) 41.95 ( o. 55) 42.5 (5. 9) 48.4 ( 3. 35) 51.75 ( 4 .1) 55.85 (7 .9) 63.75 (0.15) 65.9 (8.4) 74.3 (2.65) 76.95 (2. 3) 79.25 (2.85) 82.1 (l.l) 83.2 (2.3) 85.'1 Continue north on Tramway Road. Keep left at inters ction with East Wawawai Grade road. Junction with Kirby-Mayvie w Road. Turn right. Junction with Wawawai Grade Road. Continue on pavement. Mayview. Continue straight. Roadcut exposing two thin flows of the Roza Member separated by welded spatter. Continue ahead. STOP NO. 2. Bus lets passengers off here and proceeds for 1.3 miles, turns around, and waits for passengers, who walk down road. Roadcuts show the Roza Member overlying thin, discontinuous silt resting on the sparsely plagioclase-phyric Frenchman Springs Member (2 flows). A thin saprolite and detrital(?) clay separate the Frenchman Springs from the Grande Ronde Basalt. The Frenchman Springs thickens west from here to 150 m in Devjls Canyon (Stop 6); it does not occur farther east. Eastward thinning of the Frenchman Springs Member is typical throughout southeast Washington and northeast Oregon and is interpreted to mean that the flows pinched out against a gentle westward slope. All known feeder dikes for the Frenchman Springs Member occur in the central part of the plateau; we will see some unusual examples tomorrow. The flows of the Grande Ronde Basalt, all in the N2 magnetostratigraphic unit, display a variety of primary structures, such as joints and rubbly tops. Many flows are thin and may have erupted locally. One thick and two thin feeder dikes for the Roza Member are exposed. Per haps the spatter at mileage 41.95 was erupted from this dike. This location is along the axis of the Roza vent system. One of the vent areas will be viewed at the next stop. Return to U.S. 12 via Kirby-Mayview, Ledgerwood, and Ledgerwood Spur Roads. Junction with Tramway Road. Keep right, on pavement. Junction with North Deadman Creek road, turn right, staying on KirbyMayview Road. Junction with Tramway Road. Continue straight. Junction with Ledgerwood Road. Turn left, leaving pavement. Junction with U.S. 12. Cross highway, continuing on Sweeney Gulch Road. Junction with State Highway 128. Continue straight, up a gentle dipslope related to the Blue Mountains uplift farther south. Road intersection. Continue straight on 128. Metropolis of Peola. Bear left on main road, avoiding road to Cold Springs. Continue east on Highway 128. Asotin-Garfield County Line. At 4:00, good view of dip slopes on north flank of Blue Mountains. Potter Hill ahead. 18


(1.1) 86.6 (12.35) 98.95 (0.1) 99.05 (0.15) 99.2 (1.25) 100.45 (0.9) 101.35 (0.6) 101.95 (4.05) 106.0 (0.65) 106.65 (0.1) 106.75 (1.25) 108.0 (1.5) 109.5 (0.4) 109.9 (6.0) 115.9 (7.2) 123.1 STOP NO. 3. Potter Hill vent area for Roza Member. Note pumice, welded spatter, dikes, and primary dips related to cone. The Roza Member vent system is more than 165 km long and less than 15 km (probably less than 5 km) wide. The vent system is defineq by dikes, such as those at stop 2, and vent areas such as this one. Big Butte, visible from here to south, is another vent area from which one of the latest eruptions took place. Turn around, and retrace route to Sweeney Gulch Road. Junction with Sweeney Gulch Road. Bear left, staying on Highway 128. Roadcut exposing a Dodge flow in the Eckler Mountain Member of the Wanapum Basalt. Base of Dodge over red saprolite developed on Grande Ronde Basalt. Roadcuts beyond expose flows in uppermost magnetostratigraphic unit (N2) of the Grande Ronde. Cross Pataha Creek at Columbia Center. Keep right across bridge. STOP NO. 4. Roadcut shows the basalt of Dodge overlain by dense aphyric flows of the Frenchman Springs Member. Walk back along road 30 m to see saprolite under Dodge. The basalt of Dodge is in places interleaved with saprolite, showing that it was erupted during the time of regional soil formation on the Grande Ronde Basalt. Note the coarse grain size, large plagioclase phenocrysts, and grusy nature of the Dodge, which make it a fine marker in the field. Turn right on pavement and continue down approximate dip slope on the Roza Member. Dirt road to right. Start of Benjamin Gulch section. This section, capped by the Roza Member, exposes interbedded flows of Grande Ronde, Dodge, and dominantly Frenchman Springs chemical types, and is the only such section known in Washington. This section is of great importance, as it demonstrates that eruptions of greatly different magma chemistries overlapped in time. Presumably this locality contains flows erupted throughout the time interval during which a soil was forming elsewhere in southeast Washington. Quarry in dense flow of the Frenchman Springs Member as at Stop 4. The basalt of Dodge under the Frenchman Springs Member. Next few cuts reveal interlayering; exposures are good but not eyecatching. All contacts are exposed. Approximate base of section. Pomeroy town limit. Intersection with U.S. 12. Turn left (west) on highway. Junction with State Highway 126. Continue ahead on U.S. 12. For the next few miles, a ledge-forming flow is quite prominent, particularly in about 3.5 miles. This f low, i n f o rmally called the Pataha flow, is of high-Mg Grande Ronde chemical type and is near the top of the Grande Ronde section i n thi s area. Junction with State H"ghway 127. Continue left (west) on U.S. 12, toward Walla Walla. Type locality of the basalt of Dodge is in roadcuts along S.H. 127, 0.95 miles northeast of junction. 19


(8.45) 131.55 (6. 85) 138.4 (3.3) 141.7 (0. 3) 142.0 (1. 4) 143.4 (1. 65) 145.05 (1.1) 146.15 (1. 3) 147.45 (2.4) 149.85 (2.2) 152. OS (2. 2) 154.25 (4.4) 158.65 Junction with State Highway 261. Turn right toward Palouse Falls State Park. Roadcuts in next few miles expose flows in N2 part of Grande Ronde Basalt. Road follows the Tucannon River. Junction with paved road to Little Goose Dam. Stay left, in valley. Road to dam is access to stops 10-14 of Swanson and Wright (1976a). For next mile, black cliff above road is in the Lower Monumenta l Member of the Saddle Mountains Basalt, which occupies a road valley eroded through the Wanapum Basalt into the Grande Ronde Basalt at the confluence of the Snake and Tucannon Rivers. Exposures of the Pomona Member in a canyon-filling r elation occur nearby. Good exposures of tiered flow of the Grande Ronde Basalt. White ash in pockets on hillside is Mazama ash. STOP NO. 5 (optional). Origin of tiered flows. Are tiers the result of some process completely internal to a ponded flow and related to its cooling history, or does each tier record a separate pulse of lava into a gradually deepening pond? In other words, is a tiered flow a single or multiple-flow cooling unit? Good exposures of vesicular layers defining boundaries of tiers are found above the railroad tracks to left of road. Dark mesas to left are remnants of the Lower Monumenta l Member resting erosionally on the Frenchman Springs Member. Intracanyon remnants of the Pomona and Esquatzel also occur near here. The prominent bench seen above both the Palouse and Snake Rivers is along the contact of the Grande Ronde and Wanapum Basalts. Cross Snake River near mouth of Palouse River. The renowned archeologic site, Marmes Rock Shelter, is drowned beneath lake waters just up the Palouse. Entrance to Lyons Ferry State Park. Continue straight. Poorly exposed contact of the Grande Ronde Basalt and the Frenchman Sprin gs Member of Wanapum Basalt. Base of the Roza Member. Junction with road to right to Palouse Falls State Park. Continue straight ahead. The junction is just below the contact of the Priest Rapids and Roza Members. The road to the park crosses the Roza and the Frenchman Springs Members. The lip of the falls is carved into the oldest flow of the Frenchman Springs. Flow of Lolo chemical type in the Priest Rapids Member crops out just above road; it is the youngest flow throughout the area. Good example of scabland topography (Scablandia) to east (right). Begin to cross upper reaches of H.U. Coulee; roadcuts expose the Roza and Frenchman Springs Members. View to left down H.U. Coulee, a scabland channel made famous by J. Harlan Bretz. Roadcuts beyond are the Roza and Frenchman Springs Members. Junction with Highway 260. Turn left toward Pasco. Highway follows dry Washtucna Coulee, former course of Palouse River. Erosion caused by catastrophic flooding (the Spokane floods) in the late Pleistocene diverted the river from its former course into a more direct link with the Snake River via Palouse Falls. Prior to the floods, the Palouse flowed down Washtucna Coulee to Connell and then south-southwestward to the Pasco area via Esquatzel Coulee. The highest continuous ledge along the coulee walls between here and Kahlotus is the Roza Member; the highest discontinuous exposures are of 20


(2. 85) 161.5 (5. 2) 166.7 (0.7) 167.4 ( 4. 55) 171.95 (0. 65) 172.6 {0. 65) 173. 25 (1. 5) 174.75 the Priest Rapids Member. View to right of gravel bar deposited by Spokane floods. Junction in Kahlotus with paved road to Pasco and Lower Monumental Dam. Turn left. Junction with road to Lower Monumental Dam. Turn left down Devils Canyon. All roadcuts expose flows of the Frenchman Springs Member, totalling about 150 m thick. Contrast with the 20 m of the Frenchman Springs seen at stop 2. Look ahead at remnant of the Lower Monumental Member forming cliff overlying river gravel across river beyond dam. Large parking area to right opposite turn toward dam. STOP 00. 6. Contact of the Frenchman Springs Member of the Wanapum Basalt and Grande Ronde Basalt. Lowest exposed flow is the Grande Ronde. It has an oxidized top and is overlain by a Frenchman Springs flow with sparse, small plagioclase phenocrysts. The oxidized zone is an incipient residual soil horizon. Traced eastward, this zone gradually thickens into a true saprolite, which is well exposed at stops 2 and 4. As the.soil zone thickens, the overlying Frenchman Springs thins and finally _dies out, leaving the Roza Member resting directly on saprolite. This is evidence that the Frenchman Springs flows were erupted in a basin, presumably created by earlier subsidence or westward tilting of the Columbia Plateau. STOP 00. 7. {Optional) From stop 6, drive across Lower Monumental Dam and park along road 1 mile beyond southeast end of dam, at base of the flow defining the Lower Monumental Member. This flow, at 6 m.y. the youngest flow in the Columbia River Basalt Group, extends at least as far east as Asotin, Washington. The Lower Monumental has not been recognized farther west than its type locality. The flow overlies river gravel consisting of basalt, metavolcanic rocks, and a very few plutonic rocks with lenses of quartz-feldspar-muscovite sand. This assemblage is clearly that of an ancestral Snake River, not a small stream heading on the plateau itself. Retrace route from stop 6. Looking up Devils Canyon from this point, a good cross-sectional view of flows filling a steep-walled gorge is exposed on the northeast wall of the canyon. Park in small turnout on right. STOP 00. 8. Intracanyon flows. Before Lower Monumental time, the ancestral Snake River apparently occupied a gorge 2-3 km north of the present canyon. Four different lava flows poured down this gorge, with renewed canyon cutting between each eruption, resulting in the compound lava fill exposed in cross section in the cliffs on either side of the road. East of the road, a "triple junction" is well exposed. Here the 12 m.y. Pomona Member {right) filled an almost vertically walled gorge cut into the oldest known intracanyon flow in this area, about 12.5 m.y. {left). Note the nearly horizontal columnar jointing in the Pomona along the contact. After a period of erosion, which beveled the Pomona and older intracanyon f l ow, the 10.5 my. Elephant Mountain Member was erupted. It forms the uppermost flow in contact with both of these older intracanyon flows. West of the road, relations are more involved. Four flows are present here, and they meet at a common point above the railroad grade. The same three flows in the cliff east of the road are also here, with a fourth, the Esquatzel Member, between the unnamed intracanyon flow and the Pomona. Eastward from a point 8 km east of Devils Canyon, the pre-Lower Monumental canyon of the Snake follows the present canyon rather closely. West of there, however, the older canyon continued due west, meeting Esquatzel Cou lee between Mesa and Connell. The course of this old canyon can be traced using detailed aeromagnetic data {Swanson and others, 21


(3.05) 177.8 (9.3) 187.1 (0. 75) 187.85 (9.4) 197.25 (0.65) 197.9 (10.65) 208.55 (5.05) 213.6 (2.75) 216.35 (0.3) 216.65 (0.4) 217.05 (0.2) 217.25 1979c), as the thick intracanyon flows of uniform polarity exhibit narrow, sinuous anomalies of greater amplitude than the thin, norma l and reversed flows forming the wallrock. We can speculate that the Elephant Mountain flow so thoroughly choked the canyon that the Snake spilled southward over a divide, located about 8 km east of Devils Canyon near Magallon ranch, into a tributary of the ancestral Columbia River leading southwest toward Pasco. Erosion of this divide proceeded comparatively rapidly, and the Snake has remained in this canyon since then. Continue up Devils Canyon. Roadcuts in the Roza Member. highest cut. Turn left at intersection, toward Pasco. Contact with the Priest Rapids exposed in Continue straight at junction with Wallace Walker road. Outcrop of the Priest Rapids Member below the Elephant Mountain Member in gully to right. Continue straight at intersection with Burr Canyon road. A feeder dike for a flow of the Frenchman Springs Member can be reached by following the Burr Canyon road to the lowest railroad right-of-way shown on the Lower Monumental Dam 7-1/2' quadrangle. The tracks have been removed from this right-ofway, and the route was drivable with care in 1977. From Burr Canyon, follow the abandoned railroad about 2.5 miles to near Abe triangulation station (Elwood 7-1/2' quad). Several small faults offset the red, oxidized contact of the Grande Ronde Basalt and Frenchman Springs Member exposed along the way. The widest plagioclase-phyric dike, evident in cuts, projects across the river and feeds the prominent flow exposed at top of dike. By climbing up hill (poor exposure) above Abe, the top of the dike can be seen in cuts along the relocated railroad. A shallow pool of columnar basalt fills the flared top of dike and merges laterally into an extensive porphyritic flow. Ghostly sand dunes to right. Leave loess hills and enter onto dip slope developed in the Elephant Mountain Member. Spokane floods stripped away the loess. Junction with road to Ice Harbor Dam. Turn left (south). Turn left, onto another paved road, just before dam. Sign points to Boat Landing. Junction with gravel road. Keep left on gravel road across wooden plank bridge. Junction of track roads. Both lead to same place. Righthand route was best in 1979. End of road. Bus may need to park before end. STOP NO. 9. Examine railroad cuts 500-1200 m upriver from end of road. Flows of the Pomona and Elephant Mountain Members are well exposed at railroad level. The basalt of Martindale in the Ice Harbor Member overlies the Elephant Mountain Member above the railroad. Examine invasive contact relation of the Pomona and a white vitric tuff 1200 m from end of road. Three dikes of Basin City chemical type (Helz and others, 19 76 ) intrude the Elephan t Mountain at locality 7 of Swanson and other s (1975, table 2). The dikes are difficult to spot. The nomally magnetized dikes occur along a positive linear aeromagnetic anomaly that extends 30 km south of the Snake River and 50 km north of the river (Swanson and others, 1979c). They form part of a linear vent system for the 8.5 m.y. Ice Harbor Member described by Swanso n and others (1975). 22


( 0 85) 218.1 (1. 05) 219.15 (0.15) 219. 3 (1.1) 220.4 (1.1) 221.5 (0. 2) 221.7 "(1. 3) 223.0 (0. 95) 223. 9 5 (5.3) 229. 25 (4. J) 232.55 ( 4. 6 ) 238 1 5 (2. 75) 240.9 (0. 75) 2 41. 6 5 (l. 7) 243.35 ( 1. 25) 244.6 Retrace route to intersection w ith road across dam. Turn left and drive across dam. Keep right on pavement at south end of dam. Turn right toward river at T-intersection. Bear left onto gravel road. Follow road toward west. Bluffs on left expose the basalt of Martindale overlying the Elephant Mountain Member. Park. STOP NO. 10. The craggy bluff near the parking area is the remnant of a tuff cone built over a vent for the basalt of Martindal e in the Ice Harbor Member (locality 16 of Swanson and others, 1975, table 2). Note the poorly bedded nature of the cone, the palagonitic alteration of most of the glassy tephra, and the angular unconformity visible on the northwest face. A Martindale flow overlies the cone. Walk 600 m downriver along trail to cuts exposing the Elephant Mountain Member and the basalts of Martindale and Goose Island (locality 12 of Swanson and others, 1975, table 2). The flows occupy a shallow northwest-trending syncline, also recognizable across the river. Note the plagioclase-clinopyroxene clots in the col umnar Martindal e flow. A dike of Goose Island chemical type cuts the Martindale flow. Breccia interpreted as drainback rubble occurs in upper part of dike; the rubble merges into thin vesicular flows, overlain by a thicker flow of Goose Island type. At this locality, the upper Goose Island flow and the columnar Martindale f low have both been dated by E. H. McKee, using whole rock K-Ar methods, at 8.6 0.3 m.y. Refer to Hel z (1978) for more information on the Ice Harbor Member. Retrace route back to paved road. Continue straight ahead (south) Martindale tephr a in roadcuts. Horse Heaven anticlinal uplift straight ahead in distance. Junction with road to Pasco. Turn right. Continue through Burbank. Junction with U.S. 395. Turn left (south). At 1:00, entrance to Wallula Gap, a water gap across the Horse Heaven uplift through whi ch the Columb i a River flows. Boise Cascade plant. Hold your nose. To right across Columbia River is a series of faulted, doubly plunging anticlines exposing the Frenchman Springs Member and younger flows. Junction with u.s. 12 Turn left, toward Walla Walla. Side trip to right for 4.2 miles into Wallula Gap show s e xcellent exposures of the thick Frenchman Springs Member overlai n by the Umatilla Member of Saddle Mountains Basalt; the Roza and Priest Rapids Members are m issing. A Martindale flow caps highest visible point west of river; it overlies imbricated Snake River gravel, compositionally similar to that at stop 7, show i ng that an ancestral Wallula Gap existed at least 8.5 m.y.b.p. Faceted ridge spurs to right reflect relatively young faulting along Horse Heaven front. Roadrut on right shows white Mt. St. Helens. "S" tephra layer in silts of the Touchet beds. The t ephra is about 12,000 years old and helps to date some of the youngest Spokane flooding, which produced the Touchet beds. 23


(1. 05 245.65 (l.l) 246.75 (0 .1) 246.85 (1. 9) 248.75 (1. 4) 250.15 ( 2 3) 252.45 (1. 3) 253. 75 (16 .15) 269.9 ( 0. 7) 270.6 ( 0. 3) 270.9 0 ( 4. 75) 4.75 (0. 75) 5.5 (0.15) 5.65 (0.15) 5.8 (1. 2) 7.0 (l. 45) 8.45 (0. 2) 8 65 (1. 55) 10.2 (0. 45) 10.65 ( 0 85) 11.5 Roadcuts in the Frenchman Springs Member. Flow of Martindale type in the Ice Harbor Member forms low mesa ahead. View left across Walla Walla River shows the thick Umatilla Member overlain by the basalt of Martindale and underlain by the Frenchman Springs Member. A small sheeted dike complex, consisting of 5 dikes of Goose Island (Ice Harbor Member) chemical type, cuts the Umatilla but is not evident from highway. Roadcuts for next mile in the Frenchman Springs Member. Approximate crest of north-south anticline forming west boundary of Walla Walla Basin. Easternmost basalt exposure (Frenchman Springs Member) in western part of Walla Walla basin. Bluffs to right are type locality of Touchet beds. From here to Walla Walla, roadcuts reveal Touchet beds and loess. Leave U.S. 12, turning right on road to Pendleton, Oregon. Continue straight at junction with Highway 125. Turn right on Highway 125. Stay on 125 through Walla Walla toward MiltonFreewater, Oregon. This route bypasses most motels. Trip l og i s resumed about 5.5 miles from downtown Walla Walla, at Oregon-Washington State line. Here, Washington Highway 125 becomes Oregon Highwa y 11 SECOND DA Oregon-Washington border between Walla Walla, Wash. and Milton-Freewat er, Oregon. From left lane in Milton Freewater, turn left toward Pendleton on Highwa y 11. Turn left off Highway 11 at bottom of hill just beyond Clifton Mote l toward upper Walla Walla River. Cross railroad tracks, keeping r ight. Follow pavement to left on SE 15th Ave. The Frenchman Springs Member is exposed on hillside ahead. Keep left at intersection. Approximate base of the Frenchman Springs in roadcut on left. Bridge across Walla Walla River. Road enters the Blue Mounta ins, an upljfted and faulted region extending from central Oregon to extreme southeast Washington. Route goes progressively downsection, as dips are about 2-3 degrees toward the Walla Walla basin. Bridge across South Fork Walla Walla River. Turn right toward South Fork. Roadcut on left exposes 10 m-wide, vertical feeder dike of the Frenchman Springs Member. Note nearby horizontal columns. Dike cuts vesicular flow top in N2 unit of the Grande Ronde Basalt. Route follows the straight canyon of the South Fork of the Walla Walla River, which forms a segment of the Olympic-Wallowa lineament, a straight 24


( 2. 4) 13.9 (l. 4) 15.3 (2. 3) l7.6 (0. 55) 18.15 ( 0. 4) 18.55 ( 0. 4) 18.95 (13. 4) 32.35 ( 0. 95) 33.3 (2.1) 35.4 ( 0. 5) 35.9 ( 0. 55) 36.45 (l. 5) 37.95 (19. 5) 57.45 (0. 2 ) 57.65 (0.4) 58.05 (0. 4 ) 58.45 Cl O S) 59.5 ( 0. 2) 59.7 (0. 25) 59.95 ( 0. 15) topographic feature reaching about 700 km from the northwest tip of washington to the Wallowa Mountains i n northeast Oregon. Origin of the lineament is subject of much debate. Evidence to be seen at stop 11 shows that this segment of the lineament has been structurall y inactive for the last 14-JS m.y. Approximate location of contact between R 2 and N 2 units of the Grande Ronde Basalt. Good exposure of rubbly top on a flow in the R 2 unit of the Grande Ronde. Exposure on left of gently dipping feeder dike of the Frenchman Springs Member. Cross river. Enter Harris Park. Continue straight. Bus parking. End of road. STOP NO. 11. Observe the 60 m-thick, gently dipping (15-20 degrees) dike of the Frenchman Springs Member on both sides of river. Note that it crosses canyon, which defines the Olympic-Wallowa lineament in this area, with no obvious offset. K-Ar date on dike by Ted McKee is about 14.5 m.y. Walk across footbridge, turn left on trail, dodge low-hanging branches, and walk to lower contact of columnar dike. Observe glassy selvedge. Continue on higher of two trails to upper glassy selvedge. Discuss criteria for distinguishing a flow from such a dike. What are conditions under which such gently dipping dikes are formed? Retrace route to Highway 11 in Milton-Freewater. Turn left on Highway 1 toward Pendleton. Good view to left of Blue Mountains front. Roadcut in the Umatilla Member. Start of series of roadcuts in the Frenchman Springs Member. Fault breccia in the Frenchman Springs Member. Rock throughout roadcut is shattered. Undated colluvium is offset by small fault at far end of cut. The breccia occurs along north-trending faults. Junction of road to Tollgate and Elgin across Blue Mountains. Continue toward Pendleton, more or less along axis of the Agency syncline, with poorly exposed flows of the Frenchman Springs Member in places. Narrow bridge across Umatilla River in Pendleton. Beyond bridge, keep right toward Portland. Junction with U.S. 30. Original Pendleton woolen mill to right. Continue straight ahead through town. Get in left lane opposite Pendleton Round-Up grounds. Stop sign. Continue on U.S. 395 across railroad tracks. Follow sign to Pilot Rock. Turn left at blinking stop light. 2 5


60.] (0.1) 60.2 (2.45) 63.6 (0. 95) 64.55 (0. 75) 65.3 ( 5. 5) 70.8 ( 2. 55) 73.35 ( 0. 6) 73.95 (0. 3) 74.25 ( 0. 2) 74.45 (7. 7) 82.15 ( 4. 55) 36.7 (7. 3) 94.0 (1. 5) 95.5 ( 0. 2) 95.7 ( l. 65) 97.35 (0. 25) 97.6 ( 0. 5) 98.1 (0.4) 98.5 (1. 25) 99.75 (5. 9) 105.65 ( 2. 85) 108.5 Turn right at stop sign. Dead ahead is cut s howing the Frenchman Springs Member overlying thick saprolite on the Grande Ronde Basalt. Continue on U.S. 395 toward Pilot Rock and John Day. Do not enter freeway On left, earth-fill McRay Reservoir Dam, in danger of collapsing until refurbished in 1978. Rounded hills on both sides of road are eroded in the McRay beds of Hogenson (1964), a Pliocene fanglomerate. Roadcut in the McRay beds. Many more such cuts between here and Pilot Rock. Good view of Blue Mountains front to left and ahead. The Frenchman Springs Member caps highest ridges to left but is absent ahead. Across valley to right, saprolite separating the Grande Ronde Basalt and the Frenchman Springs Member exposed in railroad cuts. Town of Pilot Rock. Bluff at 1:00 is the Pilot Rock, mostly the Grande Ronde Basalt but relay station rests on basal flow of the Frenchman Springs Member. Continue left on U.S. 395. Continue right on U.S. 395. Good view of unimposing Pilot Rock. Roadcuts for long distance are in N2 unit of the Grande Ronde Basalt. The Frenchman Springs does not occur south of this area. Junction with Highway 74. Keep left on U.S. 395. By fall of 1979, mapping of the basalt had not been finished beyond this point, so the following log is necessarily skimpier than it has been. Some information is taken from Walker (1973a) and Brown and Thayer (1966). Good views of Blue Mountains front. Dip slopes to right of road give clear sense of uplift. Roadcut on left exposes fault in the Grande Ronde Basalt. The Grande Ronde Basalt rests on soil developed on quartz diorite of Jurassic or Cretaceous age. Entrance to Battle Mountain State Park. Low-Mg flows of the Grande Ronde Basalt in roadcut on left (Nathan and Fruchter, 1974). Coarsely columnar, probably valley-filling flow of the Grande Ronde Basal t over granitoid in roadcut on left. Blue Mountains crest, 4270 ft. Structural crest 1-2 km south of here. Contact of the Grande Ronde Basalt and the John Day Formation poorly exposed. Terraces at 10:00 are developed in Pliocene sediments, trapped in structurally controlled basin (Walker, 1973a). Junction with Highway 344. Continue straight on U.S. 395. Ridge ahead i n the Grande Ronde Basalt. 26


(0. 85) 109.35 (0. 2) 109.5 5 (0. 95) no. 5 (0. 3) no. 8 (0. 35) 111. 15 (8. 9) 120.05 (1. 6) 121. 6 5 (1. 3) 122. 95 (1. OS) 124.0 (6. 95) 130. 95 ( 0. 45) 131.4 (0. 3) 131.7 (0. 8) 132.5 (0. 5) 133.0 (0. 5) 133.5 (0.1) 133.6 (l. OS) 134.65 (1.1) 135.75 (0. 85) 136.6 (1.1) 137.7 ( 3. 05) 140.75 (6. 35) 147.1 (8. 35) 155.45 (0. 3) 155.75 (0. 55) 156.3 Start of Camas Creek section through the Grande Ronde and Picture Gorge Basalts. Pillowed flow of the Grande Ronde Basalt in roadcut. Interbed between flows of the Grande Ronde Basalt in roadcut on right. Complexly jointed and faulted flows of the Grande Ronde. Some horizontal slickensides. STOP NO. 12. Roadcut near mile post 52 exposing contact of the Grande Ronde and Picture Gorge Basalts. Discovered and described by Nathan and Fruchter (1974). Aphyic, normally magnetized flow of the Grande Ronde (probably N1 but not known certainly) overlies thin organic-rich interbed resting on highly plagioclase-phyric, magnetically normal flow of the Picture Gorge. Chemistry by Nathan and Fruchter (1974) confirms stratigraphic assignments. Bridge at mouth of Camas Creek. North Fork of John Day River to right. Cross 45th parallel. Bridge over North Fork of John Day River. Dale. All roadcuts are in the Picture Gorge Basalt. Meadow Pass, 4127 ft. The Picture Gorge Basalt over the John Day Formation in roadcuts on right. Cretaceous diorite on left. Slopes to right are foliated diorite. Cross Middle Fork of John Day River. Hyaloclastite in the Picture Gorge Basalt in roadcut on left. Flow probably fills valley eroded into Paleozoic rocks. Paleozoic metasedimentary rocks for next 0.9 miles. At 3:00, scenic view of the Picture Gorge Basalt forming mesa. The Picture Gorge Basalt in roadcut on left. View straight ahead of the Picture Gorge Basalt dipping southeast from Blue Mountains uplift. Summit of Ritter Butte, 3991 ft. View ahead of Long Creek Mountain, underlain by the faulted Picture Gorge Basalt. Town of Long Creek. Turn right off U.S. 395 toward Monument and Kimberly. Roadcut on right exposes contact of the Picture Gorge Basalt and John Day Formation. Geologic marker on left, courtesy of T.P. Thayer, explains history of area. Scenic view down valley of the Picture Gorge Basalt over red tuffaceous 27


(0. 5) 156.8 (1. 7) 158.5 (0. 65) 159 15 (0. 5) 159.65 (0.6) 160 .25 (0. 35) 160.6 (1. 25) 161.85 (1. 8) 163.65 ( 3. 9) 167.55 (0. 75) 168.3 ( 0. 75) 169.05 (10. 35) 179.4 ( 3 .15) 182.55 rocks of the John Day Formation. Approximate contact of the Picture Gorge and the John Day in roadcuts on right. Hamilton. Tuffaceous rocks of the John Day Formation, here chiefly bioturbated airfall tuff. Landslides in the John Day Formation. Altered tuffs slide easily, especially when wet and overlain by basalt. Geologic marker describes Sunken Mountain landslide to left. This and other markers in the area were conceived and written by Tom Thayer, who worked for many years on the Canyon City 2 degree sheet. Thick section of the Picture Gorge Basalt ahead. Views ahead and to right of angular unconformity between the Picture Gorg e Basalt and John Day Formation. To right, dike and irregular intrusive body of the Picture Gorge Basalt irr the John Day Formation. The Picture Gorge was erupted wholly within this region. Its feeder dikes comprise the Monument dike swarm. Fruchter and Baldwin (1975) found good chemical matches between flows and dikes of the Picture Gorge, substantiating the notion that the two are indeed related. From here to Kimberly itself we will see many of these dikes; only a few are pointed out in the log. Dead ahead is Monument Mountain, one of the thickest exposed sections of the Picture Gorge Basalt. Cross the North Fork of John Day River and enter town of Monument. View to left of Picture Gorge intrusive bodies in the John Day Formation. Roadcut across thick dike. For next 10 miles, dikes are very common, cutting flows of the Picture Gorge as well as the older John Day Formation. Zeolites in pegmatoidal flow of the Picture Gorge Basalt in roadcut on right. General store in Kimberly. Start of geologic field trip summarized by P.T. Robinson (this volume). 28


GUIOE TO GEOLOGIC FIELD TRIP BETWF.E 0 THE JOH Kl RERLY A 0 B E 0 OREGO DAY FORNATIO WITH F. PHASIS Paul T Robinson, D partment of Earth Sciences, University of California, Riverside, California 92521 and Gerald F Brem, Department of Sciences, California State University, Fullerton, California 92634 I T ROD CTIO The John Day Formation of north-central O regon is a widespread, largely pyroclastic unit lying he ween the Eocene Clarno Formation and the Miocene Columbia River Basalt (FiRure 1) The bulk of t he unit consists of andesitic to dacitic tuffaceous claystone and air-fall tuff derived f rom vents west of the p resent-day outcrops. Th age of the forma tion based on vertebrate fossils and K-Ar dating ranges from approximately 36 to lR m B P Woodburne and Robinson, 1977). Hence t h e formation is believed to provide a well exposed depositonal record of middle Tertia r y Cascade volcanism no available elsewhere. Detailed studies of the formation have been c3rried out by a number o f worker s in the las 2 0 years. Studies in the vicinit o the type area near Picture Gorge, have been carried ou bv Fish e ( 1966a, 1966b, 1967, and 1968) fi her and Rensberger (1972), Fisher and Wilcox, (1960 ) and Ha (1962a, 1962b 1963). Outcrops in and west of t he Blue Mountains have been studied by Peck (1964) Robinson (1975), Robin on and S ensland (1979) Swanso n and Robinson (1 9 6 ), and and Robi nson (1977). Reconnaissance geologic maps have b een completed for most of the uni (Swanson, 1969 Robinson 197 5). De tailed chemica 1 and mine r a log i ca l studies of the ash-flow and air-al tuffs are in pro ress by the author s STRATIGRAPH':' General Asp cts The John Day Formation crop out widely over north-central Oregon, occurring in three geograrh ically separate a r e a s (Figure 1). The t p localit (eastern facies) lies east o f the Blue Mountain uplif a n d includ h Cor e dis rict and th i chell a r a the Ochoco o r nt in th ou h 0 tuffs and lava flo w rare astern anci sou h rn faci s Con sid rabl disagr ement xi ts as to th naure and stra !graphic position of th bounciar be twe n th John Day and Clarno Formations. In areas 2 9 east of the Blue Mountains the top of the Clarno formatio n is generally marked by a well developed red saprolite. Local incorporation of this material into the lowe r part of the John Day Formation account s in part for t he red colo r of the lower member (Big Basin member of fisher anci Rensberger, 1972). Waters (1954) recognized a simila r saprolite in the Horse Heaven area in the Blue !ountains (Figure 1) and suggested that it also markeci the top of the Clarno Formation at that locality. Howev r, Peck (1964) and Swanson and Robinson (1968) placed t he contact between the two formations in the western facies at the base of a widespreaci ash-flow tuff (membe r A of the western facies) which unconformabl overlies rocks of Clarno litholog and is several hundred feet stratigraphical! above the Horse Heaven saprolite. At most localities the John Da Formation is unconformabl o verlai n by basalts of the Columbia R i ver G r ou Where these a r e missing, the formation is unconformably overlain b younger un i such as the Madras Formation, the Hascall Formation, or Quaternar basalt. Easte r n Facies Merriam (190 1) defined three member s in the eastern facie of the John Day Formation based largely on col o r How e v er, mos o f the features used by Merr iam to distinguish the members f rom one another vary laterally and define alt e ration facies rather than stratigraphi units. H a y (1962a; 1963) p roposed a 3 -fold &ubdivision based on the presence o f a widespread ash-flow sheet (Picture Gorge Ignimb rite) near the middle of the Formation. Fisher and Rensber er (1972) divided the unit into four members, nearly com parable to tho e originally de f i ned by Merriam ( 19 0 1 ) From oldest to youngest, these are; a) t he Big Basin Member, a basal sequence of red clayston e ) the Turtle C ov e ember, characterized by green, buff, or pale r ed zeolitizeci tuff and uffaceous cla stone, c) the Kimberly charac erized by light g ray to buff unzeolitized tuff and tuff a ceous sedimen and d) the Haystack Valley Member composed of gra large! fluviatile volcaniclasti rocks. Several of the members, particularly the Kimber! and Big Basin members, are distinguished largely on the basis of d iagenetic features and, hence, the contacts often transgress time-stratigraphic boundaries. B i g Basin Memb r The Big Basin member is a thin, discontinous sequence of red and less abundant yellow tuffaceous claystone named for e xposures in Big Basin. It is typical! 15 to 4 0 met ers thick but r a n es from 0 t o 75 meters. Locally, thin, discontinuous olivine basalt flows occur near its base. The claysto ne are poorly bedded, and crop out in l o w,


' / ) r / "' / / I.J..J I .., I (!) I ''-.. :::-. ) c.; q /' Q: r_ "--' \ I.L.I I ,.. _ \ a I q I r-'-J ., u ( \ (/) J q ( u (, ( -;:::. I \ Q: ) / I.L.I J ( I-I (/) ....... I I.L.I ,.,._) \..' s 122 Figur l. <.. lnd x map Or 0 I.L.I (!) 2 q Q: I.L.I a q u (/) q u .:r (!) .:r I I I I h wing on Da pr b bl rounded hills mant 1 d by a thin "popcorn layer. The red color of mos of h unit incorporation of r d saprolite from local I \ \ \ 121 0 s of of the Clarno Formation and diagen tic al ra ion air-fall tuff in upland areas. coloration fo rmed in the tuffaceous cla ton drainage was poor and th re was l ss opportunit or i r on oxidation (Hay 1962a; Fisher and R nsb rg r, (1972). Turtle Cove Member The Turtle Cov m mber i th thickest and most widespread unit in th east rn facies of the Formation. Throughout mo t of the ar a it ranges from abou 120 to 50 met rs in hickn ss but it thins markedly northwar d against th Rlue Mountain uplift where it is less than 15 met rs thick (Fisher, 1967) It consists chiefly of varicolor d zeolitized, mostly fineg rained tuff and tuffaceous claystone with less important air-fall and ash-flow tuff. Outc rops in the lower part of the m mb r di -play a typical ribb d or pinnacled 'badland' topography; those higher in he unit ar smoo h and rounded. Colors vary widel re l c ing varying degrees of diagenetic alteration. os of the uni is light green with thin interbeds of light gray rna erial; however, b rown, ellow, and ligh red colors are common. The zeolitized tuffs and tuffaceous claystones were originally composed largel of vitric shards and pumice f ragments. Pu ice fragmen are generally less than l mm in length but range up to 4 mm. Sand -size cr stals and rock fragmen s compri e 5 -1 0 percent of most specimens. r.rystals are chiefly andesine feldspar with lesser amoun s of oligoclas labradorite, and sanidine. Pyroxen hornblende, biotite, magnetite, and ilmenite a r usually p resent 30 50 120 119 XL n of h John Da Formali n kn wn out rop, dot d lin r pr s n t in rh form Hills din uni. s g rain d, mod ra poor. Crystals consist o plagio c li nop yroxen mb r Th K imb llow grav, unz oli similar to tha of Cove member The boundary h n s drawn almos ntirel on a color change r a f w of he Pain d o e a l -ric s (Ha ash-flo .. cool in variations in diag n tic altera ion and chronologicall variable. Th Kimb rly m mber has a composit thickne o about 300 me rs bu th thi kn s vari s con id rabl becaus of h variabl lower boundarv and because erosion in lat ohn Oa ime s ripp d off much of th unit. Outcrops ar usuall smooth, ven slop s oft n cover d with ligh gra powd ry oil. Generally, h rocks form hick, poorl bedd d, featur less sections without dis inct marker uni


Crystals comp rise about 5 moda l percent of these rocks and are chiefly andesine felcispar with minor amounts of clinopyroxene, hornblende biotite, and ilmenite The absence of quartz and sanidine suggests an andesitic to dacitic composition One distincti ve rhyolite tuff characterized by black obsidian sharcis occurs near the base of this member (Hay, 1963; Fisher and Rensb erger, 1972) Valley Member The Haystack V alley member comprises a sequence of interbedded fluviatile and lacustrine tuffaceous sediments near the top of the for mation It has a maximliDl thickness of about 150 eters. The member consists largely of unzeolitized conglomerate, sandstone, and siltstone with some lacustrine tuff. Massive air-fall tuffs are commonly interbedded with the r e work eci anci cross-bedded sediments. Indiv idual beds of conglomerate and sandstone are often lenticular, filling steep-walled channels cut into the Kimberly member. The composition of the reworked material that this unit was derived solely from erosion of the lwer John Day Formation (Fisher and 1972) Southern Facies Little detailed w ork has been done in the southern facies, which lies south of the Ochoco Mountains. studies indicate that this section is most like the eastern facies, consisting chiefly of fine-grained tuff and tuffaceous claystone but having prominent ash-flow sheets in the area south of Prineville. The southern facies has not been formally divided into members but lithologic variations are similar t o those of the eastern facies. A sequence of red clay stone occur s discontinuously the base of the formation and is similar to the Big Basin member of Fisher and Rensberger (1972). The overlying tuffs and tuffaceous claystones can be divideci into ze o litized and unzeolitized sections that correspond appro ximately with the Turtle Cov e and Ki mberly Members. The lower welded tuff south of Prineville Reservoir is a light gray to reddish-gray, stony rhyolite. It ranges f rom about 5 to 15 meters thick, generally thinning from north to south. The tuff i s densely welded w i t h a good eutaxtic structure, having flat tened pumice fragments up to 10 em long. Crystals co prise abou 1-2% and are chiefly plagioclase with traces of magnetit and partly altered biotite. The is a microcrys allin mixture of quartz and feldspar in which the original vitroclastic texture has been largel oblitera ed Many of the larger pumice fragm nts have been zeolitized. The upper a h-flow sheet is gen rally about 10 meters thick and consists of light g ray to reddishgray, den"el weldeci tuff. The tuff is relatively fine-grained and aphyric. Flattened pumice f rag ments, g nerall le s than 1 em long, define a crude eutaxtic structur These are replaced by a mixture of tridymi alkali f ldspar, and z olite. Small andesitic rock fragments omprise about 1 modal percent. 31 either of these ash-flow sheets can be correlated with welded tuffs in the eastern o r western facies and they are believed to have originated from separate vents south of the Ochoc o Mountains. Western Facies In the area west of the Blue Mountains and east of the Cascade Range a sequence of largely pyroclastic rocks, 600 to 1200 meters thick, lies between the Columbia River Basalt and the Clarno Formation. This seqeuence consists largely of tuff and tuffaceous sediment compositionally similar to that of the eastern and southern facies of the John Day Formation, hut it also includes numerous basaltic to rhyolitic lava flows and silicic ash-flow tuffs. These rocks have been correlated with the type John Day Formation in the eastern facies on the basis of stratigraphic position and general lithologic similarity (Hodge, 1 932; Waters, 1954; Peck, 1964; Swanson, 1969; Robinson, 1975; Woodburne and Robinson, 1977). In the a rea west o f the Blu e Mountains the John Day-Clarno contact is placed just below the basal ash-flow sheet of member A This ash-flow sheet lies unconformably on rocks of Clarno lithology and is locally separated from them by a thin, red saprolite. Peck (1964) subdivided this sequence into 9 conformable members (A-I) based largely on the presence of ash-flow sheets and lava flows (Figure 2). This subdivision, with minor additions, is valid over most of the western facies (Robinson, 1975) The only major exceptions are in the area between Fossil and Lonerock and in the Mutton Mountains. In the Fossil area course-grained lapilli tuffs and dacitic to basaltic lava flows occur near the base of the formation. The stratigraphic position of these units is uncertain but they may be correlative with the upper part of the western facies. In the Mutton Mountains the only ash-flow sheet recognized is that of member G Other ash-flow tuffs and lava flows are absent and the section cannot be further subdivided. Member A Member A is a widespread sequence of tuff, tuffaceous sediment, and welded tuff ranging in thickness from about 10 to 130 meters. At the base is a distinctive ash-flow sheet, up to 35 meters thick, that rests unconformably on the Clarno Formation. This is overlain by about 30 meters of poorly exposed lapilli tuff, tuff, and tuffaceous sediment which in turn is succeeded upward by a densely welded, fine-grained ash flow-tuff, usually less than 20 meters thick. The basal ash-flow tuff is densely welded, relatively coarse-grained, and sparsely porphyritic. A thin, light ray vitrophyre occurs locally at its base but most of the unit is devitrified and sparsely lithophysal. Two cooling units, each with a basal vitrophyre are exposed along the county road west of Ashwood; elsewhere only one cooling unit can be recognized. The tuff is characterized by a well developed eutaxitic structure with flattened pumice lapilli u p to R em long. Crystals average 5-8% and consist chiefly of quartz, sanidine, and plagioclase; some specimens also contain traces of green hornblende and opaque minerals. Rock fragments, mostly andesite and rhyolite, comprise 0.5 to 2 %



Member C A thick rhyolite flow and domal complex in the Ashwood-Antelope area comprises member C The rhyolite flows, which are up to 125 meters thick along Wilson Creek, were probably erupted from the domal complex along Trout Creek, just north of Ashwood. The domal complex has crude, nearly vertical columnar jointing, variable flow banding, and a rhyolite breccia interbedded with tuff along the southern edge. Both the flows and the dome consist of light gray to purplish-gray, massive to flow banded, very sparsely porphyritic rhyolite. Phenocrysts of plagio clase and quartz comprise less than 2 modal percent and lie in a matrix of microcrystalline quartz and feldspar. Member D Member D is a thin, areally restricted sequence of tuff and lapilli tuff that lies above rhyolite flows of member C and below ash-flow tuffs of member E The tuffs are light gray to yellowishgray poorly indurated, and moderately bedded. A maximum thickness of about 30 meters is exposed along Pony Creek between Ashwood and Willowdale. At most localities the unit is less than 5 meters thick. Member E A series of densely welded, highly lithophysal ash-flow tuffs, up to 120 meters thick, comprise member E These are best exposed along Pony Creek between Ashwood and Willowdale where they form steep, columnar-jointed outcrops. A light gray basal vitrophyre is exposed locally but most of the unit consists of stony rhyolite. Outcrops of crystallized tuff are characterized by abundant large lithophysae arranged in crude layers. The tuffs are very finegrained and generally lack any eutaxitic structure. Crystals of oligoclase and quartz comprise 1-2% and some specimens have trace amounts of altered pyroxene Rock fragments are rare or absent. In many places the large lithophysal cavities have been filled with opal or chalcedony to form round or oval masses known locally as 'thunder eggs'. Peck (1964) included the weakly welded ash-flow tuff at the Priday agate deposit in the base of member F but Robinson (1975) considered it to be the top of member E Member F Member F consists of 100 to 300 meters of varicolored tuff, lapilli tuff, and tuffaceous claystone with interlayered basalt flows near the base. The pyroclastic rocks are poorly indurated and poorly exposed, often being involved in small landslides or covered with talus from ash-flow sheets in overlying members At the base of the member is a coarse-rained, well bedded pumice lapilli tuff up to 3 meters thick. Overlying this is a 50to 70-meter-thick ection of brick-red, poorly bedded, tuffaceous claystone. The claystone is follow d upward by predominately yellow, gray, and green altered tuffs and lapilli tuffs. In Antelope Valle numerous alkali-olivine basalt flows are interbedd d with tuffs and tuffaceous claystones in the lower part of the member (Robinson, 1969). Individual flows are typically 5-10 meters thick and extend along strik for up to 20 km. Outcrops are generally low, rounded hills mantled with Small chips of weathered, yellow-hrown basalt. The basalts are usually holocrystalline, mediumto coarse-grained, and aphyric. They contain 10-15 33 percent olivine, mostly altered to smectite, abundant titaniferous augite, plagioclase, ilmenite, and minor interstitial alkali feldspar. Local pegmatitic segregations consist chiefly of titanaugite, ilmenite, and plagioclase. These highly titaniferous and alkalic basalts are chemically distinct from lava flows in the Clarno Formation, the Columbia River Basalt Group, and younger basalts in north-central Oregon (Robinson, 1969) Member G This member consist of a basal ash-flow sheet, about 15 meters thick, overlain by 30 to 120 meters of poorly bedded, poorly indurated tuff and lapilli tuff. The basal ash-flow sheet is typical! reddish-brown, densely welded, and completely cr stalized, forming resistant ridge-like outcrops. It is characterized by abundant large phenocrysts of soda-rich sanidine mantled by myrmekitic intergrowths of quartz and feldspar. It crops out chiefly in the Antelope-Ashwood-Willowdale area but a similar tuff occurs in the Mutton Mountains about 20 km west of Willowdale and an air-fall equivalent occurs in the eastern facies. The ash-flow sheet shows systematic variations in thickness, degree of welding, and crystallization from south to north. In the vicinity of Hay Creek Ranch, east of Madras the tuff ranges from about 15 to 30 meters thick and is densely welded, completely devitrified, and moderately lithophysal. Twenty kilometers to the north along Oregon Highway 206 between Willowdale and Antelope, the tuff is 5-6 meters thick, moderately welded, and completely glassy. Farther to the northeast, between Antelope and Clarno, an air-fall tuff containing similar large crystals of soda-rich sanidine mantled with myrmekite occurs in the same stratigraphic position. In this area the air-fall tuff is usually about 1 meter thick but, where extensively reworked it may attain a thickness up to 10 meters. This air-fall tuff has been correlated with a similar tuff in the Painted Hills area east of the Blue Mountains (Hay, 1963; Woodburne and Robinson, 1977) and is believed to provide a tie between the eastern and facies. The tuffs and lapilli tuffs above the basal ashflow sheet are light gray to light brownish-gray, poorly bedded, and moderately to weakly indurated. Outcrops are poorly exposed, often being covered with talus from overlying ash-flow sheets. At the base, there is usually a 2-to 10-meter-thick sequence .{)f lithic lapilli tuff, overlain by 20 to 30 meters of light brown, fine-grained tuff with about 5 % white pumice lapilli, which ive the rock a characteristic spotted apppearance. Locally, this is overlain by 5-6 meters of poorly bedded, light gray, fine-grained ash. At the top of the section is a sequence of poorly bedded, poorly indurated, light brown to light gray tuff, up to 50 meters thick. Most of the tuffs and lapilli tuffs are crystal poor, containing 2-5% pla iocase feldspar and traces of altered ferromagnesian minerals. Quartz and sanidine are rare to absent. Most specimens are diagenetically altered, containing abundant zeolite, chiefly clinoptilolite. Member H Member H consists of a 5 -to iS-meterthick basal ash-flow sheet overlain by about 30 to 40 meters of weakly indurated tuff and lapilli tuff. The


basal ash-flow sheet is v e r y wiclespr acl out almost continuously from th vicinit of Juniper Butt to the area northeast of Clarno 1) Generally, the tuff b comes thinn r and less d nsel wel ded to t he north and northeast. It is typically yellowish-to reddish-brown, v ry fineg rainecl, and very sparsely phyric. Crystals of sodic and quartz are present in trace amounts. At many localities t he r e is a light gray, layer at th base but most of the unit is devitrified. Outrops east and south of Madras contain n umerous lithophysal cavities but such features a r e sparse elsewhere. The overlying tuffs and lapilli tuff are poorly exposed, forming low, talus-mantled outcrops. They consist chiefly of light brown to huff, very poorly bedded, fine-grained tuff. Most rocks contain S 10 % of small white pumice lapilli producing a distin t speckled appearance. Outcrops also contain thin layers of light gray, slightly more resi tant tuff up to O S meters thick. Crystals of comprise 1-2% uf most tuffs and small lithic fragments are sometimes present. Member I -A basal ash-flow sheet, 20 to SO meters thick, and an overlying 200to 300-meter-thick sequence of tuff crops out extensively in the western facies from the area south of Madras to the vicinity of Fossil. Thicknesses vary considerably clue to erosion at the top of the formation. The basal ash-flow sheet is thickest and most densely welded in the area between Madras and Grizzly. Here it is consists of SO to 60 meters of densely welded tuff with a thick, black vitrophyre at the base. It thins and becomes less densely welded to the north, being largely unwelded near Willowdale. the tuff is coarse-grained, containing large flattened or irregular pumice fragments and pieces of black obsidian. Small oligoclase crystals comprise about 1% and rock frgments, chiefly andesite and rare shist, make up 1-2 % The overlying tuffs are mostly light t o yellowish-gray, poorly bedded, and massive, but a few layers are thinly bedded and cross-bedded. Outcrops are steep, even slopes often mantled with talus from the overlying Columbia River Rasalt. The tuffs contain 1-2% crystals, chiefly andesine, sparse lithic fragments, and sometimes small pieces of black, vesicular glass. Tuffs in the lower part of the member are usually zeolitized but those in the upper SO to 100 meters are still glassy but hyclrated. These tuffs often contain well preserved vertebrate fossils (Woodburne and Robinson, 1977; Dingus, 1979). Exposures of the John Day Formation between Fossil and Lonerock cannot be correlated directly with the facies but most of these rocks consist of fine-grained tuff similar to that in the upper parts of members H and I. At the base of the sequence is a cream-colored to gray, relatively coarse-grained, poorly sorted, very poorly bedded, pumice lapilli tuff. The tuff has a maximum thickness of about 2SO meters directly north of Fossil a n d thins to the northeast and southwest. No welding is observed in unit but the poor sorting ancl lack of bedding suggest an ash-flow origin. 34 In th vi inity of Fossil s v ral lav a flows are int rb dd d with th basal uff. W t and north of Fossil is a r ddish-brown rhyodacit (Robinso n, 1969), S to 10m ter s thi k, with rud platy j oi nting. Farth r ast at 1 ast two flow s of alkaliolivin -basalt occur a t the same leve This low r s quenc i lo ally ov erlain by a thin, discontinuous weld cl tuff, possibly orrelative with th ba al sh-flow sh t of m mb r H The tuf mod rat ly t o d nsely w lded with a weakly dev loped utaxitic structure, A thin lay r of light gra gla s o urs a t the base of the tuff, p assing upward into a w kly w lded, r ddish-gray, stony rock. Spars cry tals are chiefly oligo las Abov th weld d tuff is a 200-to 30 0-m ter-thic sequence of fine-grain d, moderately indurated, poorly b dd cl tuff imilar to that in th upper par of members H ancl I The uppermost tuffs ontain small bla k f ragments of fr sh v si ular glass; near the base are diagen ti ally alter d and zeolit ized. Correlation Of Eastern And W The w stern faci s of th John Day Formation differ s from the eastern fa ies in having generally coarser-rained air-fall tuffs and tuffaceous clay stones and more abundant lava flows and a h-flow sheets. These differences ar believ d to reflect a western source for most of the pyroclastic mate rial and the pres n c of topographic barr! r along the Blue Mountain axis thro ugh most of John Day time. Two specific correlations between the eastern a M western facies are proposed (Figure 2). A small outcrop of ash-flow tuff at th base of the eastern facies along Rowe Cr ek (Hay 196 3 ; Robi n on 1975) is correlated with the basa l ash-flow sheet o f member A in the western facies (Swanson and Robinson, 1968; Woodburne and Robinson, 1977). This orrelation is based largely on pheno ryst mineralogy, in.cluding the presence of a distinctive barium-rich sanidine (Woodburne and Robinson, 1977). If this correlation is correct it implies that uplift along this part of the Rlue Mountains took plac after about 36 m y A distinctive rhyolitic air-fall tuff in the ea tern facie (Hay, 1963) i s o r e l at-wi h the basal ash-flow tuff of member G and its air-fall equivalent in the western facies. The s e tuffs a r e mineralogically similar, having abundant phen oc r ysts 1-4 mm in diameter, of distinctive soda-rich sanidiM mantled with myrmekite. This correlation sugg ests that the bulk of t he western facies is equivalent t o about the lower 7S meters of the eastern facies. If thesP. correlations are correct, the Pic t u r e Gorge Ignimbrite has approximately the same s rati graphic position as t he basal ash-flow sheet o f member H of the western facies but these appear t o bt different units based on texture and mineralog y (Fisher, 1966). The general absence of ash-flow sheets in the eastern facies and thickness variations within the formation (Fisher, 1967) imply tha the Blue Mountains formed a topographic barrier t hrough out middle and late John Day time.


Table 1 odal Analyses of John nay Rocks 2 Groundmass 92. 7a 97 6b Rock Fragments 4.0 0.7 Sanidine tr Quartz tr 0.3 Plagioclase 2 3 l.4 (An -Content) 35-40 35 Clinopyroxene Hornblende Biotite tr Opaques 0 2 tr Zircon tr tr Apatite tr tr 3 98 4 0.4 tr 1.0 1 0 20 tr tr 0 1 t r tr 4 94.9 l.1 3 2 0 8 2 0-25 tr tr tr 5 84 4 3.6 1 0 .1 l.6 0 2 tr tr l. 2 3 4. 5. Tuffaceous claystone siltsto f ne rom western facies (average of 648 207 2 9 Vitric tuff from w stern facies (648-S7-6) -S' and -S2-l 1). Picture Gorge Ignimbrite, lower coolin ( 6 g un1t average of 48-458a, -458b, and -458c) Basal ash-flow tuff of memher A (Swanson ancl Robinson, 1968) Basal ash-flow tuff of member r. (average of 9 specimens) a) includes 7 2 percent pumice fragments larger than 0 5 mm b) includes 4 7 percent pumice fragments lar ge r than 0 5 mm = not recorc!ed CHEMICAL COMPOSITION OF THE JOH DAY FORMATIO Rocks of the John Day Formation form t hree distinct mineralogical and chemical groups. Tuffaceous claystones and vitric tuffs are chiefly andesitic to dacitic, ash-flow tuffs and minor silicic lava flows are rhyolitic, and mafic lava flows are alkali olivine basalt or trachyandesite. Tuffaceous claystone and vitric tuff The original composition of the tuffaceous claystone and vitric tuff cannot be determined directly by chemical analysis. Much of the original glassy material in these rocks has been replaced by montmorillonite o r clinoptilolite during weathering before burial or by later diagenetic alteration. Chemical analyses of altered rocks reported by Hay (1963) indicate extreme hydration accompanied by leaching of a20. Chemical analsyes of glassy material from the upper part of the formation are equally misleading because of extreme etchin bleaching, and hydration of the glass. The ini ial composition of the tuffaceous clayston and vitric tuff must therefore be inferred from the pyrog nic min ralogy. Crystals in these rocks are chiefly andesine feldspar with lesser amounts of labradorite, magnetite, pyroxene, hornblende, biotite, and ilmenite suggesting that the bulk of the material was originally andesitic t o dacitic in composition (Table 1 ; Hay, 1962a; 1963; Fisher and Rensberger, 1972). Less abundant oligoclase-bearing tuff and tuffaceous claystone was probably rhyodacitic and a few tuffs in the lower part of the form ation containing quartz, sanidine, and oligoclase were probably rh olitic in compostion. These inferred compositions are supported by the presence of accessory andesitic, dacitic, and rhyolitic fragments in many of the samples. 35 Silicic ash-flow tuff Ash-flow sheets in both the eastern and western facies are rhyolitic, based on phenocryst mineralogy and whole-rock chemical analyses. Phenocrysts are primarily oligoclase, sanidine, and quartz with only traces of ferromagnesian minerals (Table 1). Chemical analyses indicate generally silicic compositions (Tables 2 and 3), howev er, the tuffs show very wide compositional variations. These perplexing chemical variations within individual ash-flow sheets have been noted by previous workers (Hay, 1963; Swanson and Robinson, 1968). umerous chemical analyses of the Picture Gorge Ignimbrite and the basal ash-flow sheet of member H indicate a consistent chemical alteration pattern that is dependent on the porosity and crystallinity of the tuff (Table 2). Very densely welded, devitrified zones with little or no observed pri ary porosity are interpreted to be the least altered parts of the ash-flow sheets. Within these zones analyses are consistent, alkali elements are subequal to each other, water contents are consistently low, and petrographic characteristics do not suggest dissolution of glass or deposition of secondary minerals. Other zones exhibit increasingly severe effects of alteration. Glassy zones are hydrated as indicated by water contents or volatile losses of 3 to 6 weight percent. Typically, hydrated vitrophyres contain less Si02, Ti02, and a20 and more Fe203, CaO, and K20 than devitrified, densely welded tuffs. Alumina and 1g0 have no consistent pattern of loss or gain. With minor exceptions such as Fe2 03, the chemical differences between hydrated vitrophyres and devitrified, densely welded tuffs are similar to those reported for evada Test Site welded tuffs (Lipman, 1965).


Devitrified, mod rately to w akly w ld d or unwelded tuffs display different ch mical changes that become increasingly evere as th porosity increa es (Table 2; unpublished data). T pi all Si02 increases and Ti02, Al2 03 fgO, CaO, and a 2 0 all decrease with increasing porosity of the sample. Both iron and potassium are somewhat high r in devitrified, moderately welded tuffs hut d cr a in more thorough! silicified unwelded portions of the devitrified ash-flow tuffs. Chemical changes in devitrified samples an b partially correlated with petrographic f atures. Th loss of many elements is consistent with observ d leaching of pumice lapilli and phenocrysts from altered specimens. Iron cont nt i largely controlled by secondary i r on oxides that stain the rocks brick red or form liesegang banding. Very high potassium samples have minute euhedra of potassiumrich feldspar in the groundmass or in pumice lapilli, and very high silica rocks contain abundant prismatic quartz crystals in the groundmass and por e spaces. The ash-flow tuffs are believed to have been altered by the same diagenetic event that affected the air-fall tuffs and tuffaceous claystones (Hay 1962a; 1963). The very different alteration of airfall and ash-flow tuffs is due largely to differences in texture. Densely welded non-porous vitrophyres exhibit chemical alteration typical of leaching by groundwater during hydration, such as that reported by Lipman (1965). The chemical effects are generally small, as might be expected for rocks with no primary porosity and little induced porosity. Porous glassy material from unwelded basal zones of ash-flow sheets and enclosing tuffaceous claystones and vitric tuffs exhibit severe chemical changes induced by large volumes of groundwater acting on large surface areas. The low porosity and permeability of the vitrophyres precluded signifi cant chemical reaction with the groundwater except possibly for hydration and minor leaching. Where cut by fractures that increased permeability and porosity, the vitrophyres have thin (1-5 em), zeolitic alteration halos along the cracks (Hay, 1963). Porous, partially welded to unwelded zones in the ash-flow tuffs must have been devitrified prior to diagenetic alteration; otherwise, secondary minerals similar to those in surrounding claystones and tuffs would be expected. Devitrified tuffs contain alkali feldspar and quartz in the groundmass Whereas claystone and vitric tuff contain montmorillonite, celadonite, or clinoptilolite. The difference in secondary mineralogy of the welded tuffs surrounding claystones is probably not due to differences in initial rock compositons because Ha (1963) reported zeolitic alteration of silicic as well as intermediate composition tuff. The alteration of upper partially welded to unwelded zones is probably unrelated to vapor phase crystallization in the ash flows. This is inferred because typical vapor phase minerals such as tridymite are not commonly observed and qua rtz is the most abundant silica mineral. Minor chemical changes may have resulted from vapor phase activity, but these were overwhelmed by subsequent diagenetic alteration. 36 As uming tha ou r mod 1 of consist nt ch mical hang as a fun tion of sampl porosi y and crystaG linit is orr t, w can infer th original magma composition of alt r d unit by comparing rocks of imilar p trographic hara ter. For xampl a rhyolit compo iton is inf rr d for th ash-flow h et of m mb r E, whi h has no known fr sh material, by omparing its omposition to p trographically similar alt r d mat rial from th ld d uff of m mb r H or from th Pi tur Gorg Ign imbri t (Table 2). Chemi al analys of al t r d sampl s f rom other units sugg st that all of the ash-flow tuffs in the John Day Formation wer originally rh oli in composition (Tabl 3). C rtainl minor h mical differences betw heets must hav been initially pr s nt; how v r, th s ar masked by diag nP.tic alt ration eff ts. In gen ral, th ash-flow tuffs app ar to hav mod rately alkali rhyolite with sili a contents on th ord r of 75-76 % Th se rocks app a r to b ch mically di tin t from most of the vitric tuffs and tuffa ous claystones which mak up the bulk of th formation. Mafic lavas Although the ro ks are som wha alter d r liable chemical analys s hav b n obtained for most of th mafic lava flows in th formation (Table 4). Memb r B of th western faci s is a trachyand site with mod rat Si02 content. Equivalent rocks are not known in the astern facie outcrops. Basalt flows in the west rn facies are all low silica, silica-und rsaturat d alkali-olivine basalt with high titania. Minor basalt flow in the lower part of th eastern facies hav imilar compositions. The trachyandesites and alkali olivin basalts probably represent eparate magma puls s, unr lated to each other or to the silicic rocks. They form distinct groups and rocks of intermediate composition are not known In addition, th high FeO/MgO ratios and Fe203 contents of th alkaliolivine basalts indicate an evolution along high-Fe tholeiitic trends rather than along high silica lines toward trachyandesite compositions. ORIGI OF THE JOH DAY FORHATIO Tuffs, lapilli tuffs, and tuffaceous claystones in the western, southern, and eastern facies are texturally similar to rocks that represent accumulation of ash falls on the land surface (Hay, 1962a; 1963; Fisher, 1966b). These rocks become progressively thick r and coarser-grained from east to west (Waters, 1954; Peck, 1964; Robinson, 1975), suggesting that the source volcanoes lay to the west of the present outcrops. The source volcanoes were largely andesitic to dacitic in composition based on pyrogenic mineralogy. o andesitic or dacitic volcanoes of John Day age are known in the John Day basin, east of the present day Cascades. Previously suggested vents in the Horse Heaven Mining District (Waters and others, 1951) and at Smith Rock (Williams, 1957) are unconformably overlain by John Day rocks (Swanson and Robinson, 1968; Robinson and Stensland, 1979) and are here considered to be Clarno in age. The absence of andesitic or dacitic vents in the John Day outcrop area, the evidence for a western source, and the similarity of these rocks to younger units sugge t that the source volcanoes were located beneath the present day Cascade Range. This suggests that


w -.J able 2 Chemical Com1)osition of Fresh and Alter ed \.,Telded Tuff el'lhe r F. Member H 2 3 4 5 6 7 8 Si02 76.37 77. 5 1 83.23 7 5.94 77. 5 76.9 8 0 83 75.31 Ti02 0 .21 0 2 1 0 .17 0 .26 0 .17 0 1 6 0 .17 0 .39 Al2o3 12.03 11.39 8 .74 13. 0 7 12. 2 11.61 11.33 13. 0 8 Pe2o3 2 13 2 20 l. 21 1 20 2 .28 l. 79 0 .92 2 .22 nO 0 .02 0 09 0 .03 0 0 1 tr M g O 0 .11 0 06 0 .02 0 0 7 0 35 0 0 3 0 0 4 0 .25 CaO 1 00 0 .61 0 .31 0 50 0 .65 o 18 0 2 1 0 .68 a 2o 2 29 2 92 I. 43 4 .41 3 1 3 .33 2 .96 3.98 K20 5 0 9 4 .47 4 77 4 19 3 7 5 08 3 38 3 .95 r2o5 --0 .02 0 0 1 0 .03 0 .02 0 1 0 Total 99 .23 99 .37 99 88 99.66 99 .98 99 .14 99 .85 99 .96 LOI 3 .74 1.02 l. 1 2 0 86 6 0 + 0.42 0 .64 1.00 1. 648 86 648 -Sl3 648 -S7-1 0 anci DLP 5850 (Peck, 1964): Hydrated glass. 2 648-41, 648 -544 PTR-71-4b : Oevitrified, moderately welcied 3 648-S 1-4 : Devitrifieci, weak l y weldeci. 4 648 -129A and 648-189 : Oevitrified very ciensely welcied 5 DLP 58-39A from Peck (1964): llydrated g l a s s 6 648 311 : Oevitrified, mocier ately weldecl 7 648 -S41 0 and PTR 71-Sb : Oevitrified, weakly weldecl 8 113 Hay (1963): Oevitrified, v e r y rlensely weldecl, fresh. 9 62137 and 62376 from Fisher (l966a); 112 Hay (1963); 648 -4558: Hydrated glass. 1 0 1/4 Hay (1963) and 648-458A: Devitrifiecl, moderately welde d 11. 648-455C, 648 4550 ancl 648-455F.: Oevitrified, unwelded not cletectecl Picture Ignimbrite 9 1 0 11 74. 0 7 76.39 82. l 5 0 .25 0 35 0 ./.4 13.08 12.46 9.22 3 .23 1. 55 1.03 0 0 4 0 .02 0 03 0 19 0 1 0 0 0 4 l. 28 0 59 0 .55 3 60 3 .31 2 .41 4.28 5.14 4 .03 0 1 0 0 .14 0 0 5 1 00 .11 100 .05 99 .73 4.03 l. 0 4 0 .90


Ta hl Ch mi al Analvsis o 1. ;tS t Alt r d John Oa Formil ion A h-Flow Tuffs 1 mh r A 1 mhf>r I. 1 mh r F. 1 mh r \. nh r H mh r l P icture G org e Low r Tuff U pp r Tuff I g n i l'lhrite Basal Upper Cooling Unit Unit 1 2 3 5 0 7 R S i02 74. 68 73 .99 7 '). 17 Ti02 0 .25 0 .61 0 20 A l 203 1 2 5 7 12.78 13. 0 Fe203* 2 9 2 1 1 1.77 H n O 0 .03 0 .21 7 4 5 1 0 2 5 1 2 R O 2 .61) 0 06 7 6 0 76 .64 74. 0 6 7 5 94 0 .2C) n tR o 7 o n 12. 5 l l. 31 12. 4 5 11. 0 7 2 17 2 69 41 l. ? O 0 .12 0 03 0 0 2 1g0 0 .23 o 12 0 04 0 06 0 0 9 n s 4 0 0 7 n 19 CaO 0 .65 l. 01 o 9 0 46 0 .76 [.()C) 0 5 0 0 88 Na20 3 6 2 1 6 2 7 K20 4 38 6 .49 6 08 P 2o5 0 .20 2 .44 5 7 0 3 1 2 74 3 .64 4 41 5 I 4 71 4 .37 4 l q 0 02 0 II o 02 Total 99.26 99. 30 100 14 100 .12 99 I 0 99 .Q9 99. o6 99.46 LOI 0 .62 0 71 3 .22 l. 10 l. 54 l. 81 0 86 3 .46 *Total iron as Fe203 Sample Identification 1 648 -la: pa rtially devitrified, rlensel w ld d tuff 2 648 -lb: devitrified welded tuff with lea hed out pumi c e lapilli. and minut q u rtz and alkali f lclspar crystals in groundmass 3 648 4a : hyd r .ated, clensely welderl tuff 7 5 3 1 o 3 9 13. 0 R 2 2 2 tr 0 25 0 6 8 3 9 3 9 5 t o 99. 9 6 1. 0 0 4 648 27 and 648 -S11: crystallized rhyolite flow with 1 ach c1 ph nocrysts and quartz and lkali f ldsp a r in g r oundmass 5 648 -41: devitrified, densely weldecl, lithoph sal tuff crystal-lind caviti s; no hin e tion 6 648 34 : glassy, pa rtially hydrated vitrophyre 7 648 -129 A and 648 -1R9: devitrified, ver densely weldecl tuff with sph rulit s or v r y fin grain d q u artz alkali felclspar 8 6 4 8 5 1)48245 648 52 -12: hydratecl moderat to clensel welded tuf f with p rliti texture 9. Analysis 3 Hay (19o3): devitrified, very densel weldecl tuf f Table 4 Average l.hemical Composition of Mafic Lava Flow Mel'lber R Memhers F. and F L o w r Member I.Je t rn Facies F.a s ern Fac ies Si02 Ti02 Al203 Fe203 MnO MgO CaO N a20 K20 P2o5 55 .98 2 14 13.79 13 .84 0 21 2 13 o.26 3 .31 l. 7 5 0 6 0 100 0 0 l. 41 46 .78 .49 15.21 16. 0 1 0 2 1 5 17 .79 3 0 1 0 .78 0 .55 TOO. o o l. l R otes: Member R : average of three analyses from Robinson (1969) !embers F. and F: average of 11 analyses in western facies from Robin on (1969) Lower Member: average of 2 analyses, in eastern facies (Robinson, 196Q) Out crops are area below Picture Gorge Ignimbrite; however, corrrelation t o western facies members is uncertain All samples recalculated water free, all iron as Fe203 and totaled to 100 % 38 46 0 5 4.02 I 5 72 Jo.88 0 .15 3 73 7 3 2 3 R O 1.6 5 O n8 1 00 00 l. 98


ad the earli s phn ilar t o l nor h astt r nding Rlu uplift ( Swans on 1969). Tn h g n ral 0 h 39 pres n da Cascad s marin sedimen w r hein deposi ed along wi h mino r nsal ic rocks (P ck and o hers, 1 64). Sometime h we n 4 0 anct 3n m y ago, Clarno volanism rl c r a d in i n nsity, p r o bably nding ntirel b 36 m y Although the age of h u p permo s t Clarno rock i not known e xa tl no cal alkalin ande i ic roks of Clarno li holog a r e known in th John Day Forma ion. Initia ion of John Da v o lcanis about 36 m y ago signified t he emergene of a n e w vol anotec oni regime in north rn O regon \olcanoes in the vi init of the p r e ent day Cascades began rupting andesitic t o dacitic pyroclastic material tha t was deposited in the John Day basin a ash falls. This material r p r esents the earliest documented volcanic activit along the r.ascad r nd. Simultaneously, v ents ast of the High Cascades, between the Cascade Rang and the Blue lountain uplift, w e r e erupting rh elit e ashf low tuffs, lava flows, and min o r air-fall tuffs. ost of t he ash-flow tuffs were derived f rom vents we t of the p resent outcrops, whereas lava flows and dom s repres nt local eruptions within th John Oa basin. Other vents, mo tl east of the major rhyolitic v olcanoes, erupted alkaline b asalt and trachyandesite. This patt ern of volcanism continued until approximately 25 m y ago, at which time th r h olitic, basaltic and trach andesitic ruptions ceased. Andesitic t o dacit i c eruptions con inued in the Cascade Rang e leading to deposi ion o f th upper part of the John Day Formation Depositi on of the John Day Formation ceased about 1 8 20 m y ago (\Joodburne and Robinso n 1977) coin cident with a probable hiatus in Cas cade volcanism ( cBirney and o thers, 1974). Following a short peri od of folding and e rosion th Columbia River Rasalt was erupted from vents in the ea tern part of the John Day basin beginning about 1 o m ago (Watkins and Baksi, 1974 ; M Kee S1anson, and Wright, 1977). R e n ewed v olcanism in the Cascade Range is reflected in tuffaceou interbeds i n the Columbia River Basalt and in pos t -Columbia River dep sits such as t he Hascall and 1adras F ormation s Th regiona l ignificance of the patter n of Oligocene and earl ocene volcanism outlined above i s not clear The different magmas may reflect lateral variation in depth of magma gen eration, in crus tal thickness o r composition, o r in tectonic environment REFE R E CES CITED Col man, R G 1949 The John Day Format1on in the Pictur Gor g Quadrangle, O regon : I n pub S th sis, O r e on State niv., Corvallis, 211 p Brown C E and Thayer, T P., 1 966 Geologic map o f the Can on city quadrangle, northeastern Oregon: G ol. Survey isc. Geol. Inv Map I -447. Ding u s L w., 1979 The \Jarm Springs F auna (mammalia Hemingfordian) from the western facies of the John Day Formation, O regon: npub M S thesis, Univ. Calif. Riverside. E v ernd n J F., Savag n E. Curtis, G H and Jam s G T., 1 64, Pota siuM a r gon dates and th Cenozoi mammalian h ronolog of orth Ameri a : Amer. Jour. S i., v 2n2 p 145-1911


Fisher, R. 196 a \.eology of a lio n C' l a er, John Da Formation, a tern Or on: C:alif. Uni v Pub G ol. Sci. v. 67, 59 p Fisher, R \,, 1 96nh Tex t ural comparison of John Oa volcanic ilt ton s w1 h lo s and vol anic ash: Jour. Sed Pet., v 36 p 706 718 Fisher, R . 1967, Earl Tertiar d formation in n orth-c ntral Oregon: Am. A so Petrol. Geol. Bull., v 51, p 111-i23. Fisher, R V., 1968 P yrogeni minerals abilit lower member of the John Day Formation: Univ Calif. Pubs Geol. Sci., v 75, 36 p Fisher, R V and Rensb e rger, J H., 1972, Physi al stratigraphy of the John Oay Formation, c ntral Oregon: Calif. Univ Pubs. Geol. Sci., v 1 01, p 1 45 Fisher, R. V and Wilcox, R E 196 0 Th John Day Formation in the lonument quadrangle, Oregon: U S Geol. Survey Prof. Paper 4 00-R, p 302-304. Hay, R L., 1962a, Origin and diageneti alte ration of the lower part of the John Day Formation near Hitchell, Oregon : Geol. Soc America Ruddington Hemorial Hemoir, p 191-216. Hay, R L., 1962b Soda-rich sanidine of p rocla tic origin from the J ohn Oa Formation of Oregon: Am. Mineral., v 4 7, p 968 -971. Hay. R L., 1963, Stratigraph and zeolitic diagenesis of the John Oay Formation of O r gon : Univ. Calif. Pubs Geol. ci., v. 42, p 199-26 1. Hodge, E T., 1932, Geologic map of north-ntral O regon: Oregon lniv Pubs., Geology Ser. v 1 7 p Lindsley, D H., 1960 of the Spra Quadrangle, Oregon, w ith special emphasis on the petrography and magnetic p ropertie of h Pi tur Gor ge Basalt: npub P h D thesis, Johns Hopkin Univ. Balti mor e d., 236 p Lipman P W., 1.965, Chemical composition of glass and crystalline volcanic rocks: S Ceo Survey Bull. 1201 D p D 1-D 24 HcBirney, A R., Sutter, J F., aslund, H R Sutton, K. G and White, C H., 1974 Episodic volcanism in the central O regon Cascade Range: Geology, v 2 p 585 58 9 HcKee E H., Swans on, D A., and Wright, T L., 1977, Duration and volume of Colum ia River Basalt volcanism, Washington, Oregon, and Idaho: ol. Soc America Abstracts with P rogram v 9 no 4 p. 463 464 Herriam, J C., 19 0 1 A contribution t o the geology of the John Day Basin ( Oregon) : Calif. niv. Pubs. Dept. Geol. Bull. v 12 p 269 314 Peck, D L., 1964 Geologic reconnaissance of the AntelopeAshwood area, n o r h e n ral Or on, wi emphas i s on the John Day Formation of late Oligocene and early iocene a e : S Geol. Survey Bull. 1161 D p 01026 Peck, D L., G rig s, A. B., Schlicker, H G Wells, F G and Dole H M 196 4 Geology of the central and northern parts of the Western Cascade Range in Oregon: S Geol. Survey Prof. Paper, 449 56p Robinson, P T., 1969, High titania alkali-olivine basalt o f north-central O regon, U S A.: Contr. Mine ral. Petrol., v 22, p 349 360 Robinson, P. T., 1975 Reconnaissance geologic map o f the John Day Fomation in the southwestern part of the Blue Mountains and ad jacent areas, northcentral Oregon: U S Geol Survey 1ap I 872 40 P T .qncl S nsl nd, D., 1979, G ologtc lllap SJTti h Ro k a r a Oregon: t S G ol. Us. G ol. Inv. Map T-1142. 19 Law ni hlu s his from nor h-e n ral Or gon : in C:eologi 'al Surv y Re r h 19 69 l S Geol. Surv Prof. Pap 650-B p 8-11. Swan n D A., 1969, R t he half o f t h 1-rh J ff r on, co U S G ol. Surv Swan o n D A., and Robin on P T., 1968 Base of the John Day Formation in and n a r th Hors Heaven mining distri' north-c ntral Oregon; ..!.!.: G ologi cal Surv y R a r h 19 9: lJ. S G ol. Survey Prof Pap r 600-D p 0 154-0161. Waters, A C., 19 54 John Day Formation wes of its type localit (abst.), G ol. Soc Am r Bull., v 65 p 1320. Wat r A C., Brown R E., Compton R R Staples, L w., Walk r, G w., nd \.Jilliams, H 1951, Qui ksilv r d posit of the Hor H aven mining district O r gon: ,Geological Surv Bull.. 969 E pp 1 05 14 \-/atkins, D and Bak i A K., 1974 agn tostratigraphy and oro'linal f lding of th Columbia Riv r St n and Owyh e ha alts in O regon, Washington, and Idaho: Am. Jour . ci., v 274 p 148-1 89 Williflms, How 1 1957 A g ologi map of h B nd quadrangl Oregon, and a re onnaissan e geologk map of th ntral por ion of he Cascade ountains: Oregon D p G ol. Hn. Ind. in cooperation with U S G ological Survey, Scales 1 : 125 000 and 1 : 250 ,000. Woodburn 1 0 and Robinson P T 1977 A new late H mingfordian Mammal fauna from th John Da Formation, Or gon and its stratigraphic implication: Jour. Paleo., v 51, p 750 -757.


. ILES ( 0 35) 0 .35 ( 0 4 0 ) 0 7 (0 60) 1. 35 (0 05) 1. 40 (0.55) l. 95 ( 0 .75) 2 7 0 ( 0 45) 3 15 ( 0 4 0 ) 3 .55 ( 0 60) 4 15 (1.15) 5 .30 (0.40 ) 7 10 ( 1.40) 5 7 (0.55) 7 .65 1P K EPLY OR. 0 .'IT m OH '1' 0 Paul T Robinson, n rartl'l nt of Ear h Sience, lniv rsitv of la1ifornia R i v r i d C:a 1 i f o r n i a q 2 2 I Kil'lb rly. Turn 1 f on High\ av 1 oward and John nav Cross orth Fork of ohn nay Riv r High 1 vel r diPl nt at 12 o 'clok. Thes are oml'lon along h John na River b tween Kimberly and Pitur Gorg S op 1 (Op ional). Dike of ohn Day Formation. Dik It ha fin -grained margin and Rasalt cuts upp e r buff bed of the about 3 0\. and is es ntiallv v rtical. pl x joint patt rn. Although dik is about 1 0 Plet r wide th how littl baking and onl moderate induration. Stream gravels in roadcut on 1 ft. Dik eros es road at this poin Tuffac ous clay ton and tuffs of upp r John Da Formation in roadcut on 1 f From here to Pictur Gorg road uts ar Plostly in John Day rocks. High 1 vel pedil'l nts on both sid s of ohn na Riv r ahead. Pic ure l.o r g dike on hill ide ast of John nay R ver a t 2 o 'clo k l.r enish tuffs and tuffaceous la ston s of John nay Formation at o lock. Stop 2 (Optional). l.r en bed of John Formaton ov rl ing Picture Gorge Ignimbrite on 1 ft side of road. Th ignimbrite h r is about 1 0 et r thick and exhibit plat jointing near bas with Pla sive, rounded outcrop a ove. Ignimbrite is mod rately w lded with utaxitic structure defined b elongate pumic fragments. Quartz bip ramids and feldspar cr tals omprise 1 -2% and lithic fragment are sparse. Alter d glass chips and fragm nts are also present. Ignimbrite is alt red and zeolitized. l.ood 9 l.orge Good at 9 la iddl xpo ur s of o lo k ot Ignimbrit gr k r e n uff L dg is an tuffa ous 1 stone ro ion into pinna le of John Day Formation Ledge is Picture and tuffaceous a r light! mor lay ton la ston s of John Da Formation resi tant airfall tuffs inter-ohn na b d at 2 o 3 o'clo k with Pitur Gorg Ignimbrite forming 41


1'\.) 121 WARM SPRINGS 44 ( -v

( l 5() .I') ((). 5) q ,S(l ( O .R5) I 0 15 ( 0 5 ) I O .CJ(l ( l. 50) 12. 40) ( 0 50) 1 2 .CJO ) (0 9()) 1 .R0 (0 60) 14.40 (0 50) 14.90 ( 2 15) 17.05 (1. 60) 18. 6 ( 0 20) 18.R5 ( 0 9 0 ) 1 9 .75 ( 0 4 5 ) 20. 20 (1.20 ) 21.40 pro in n l clg in ro<'lclcu Th i gni.mhri on 1 f is r>Artlv slumpe d t owC'lrcl <" rivE>r S r <1m o For Ar <'l of John nav Fossil Reels g r ion of John For c1 ion wi h 1 clg on hoth side s o f riv r ote \ iclcll e T gniMhrit k o n w t Lm r r cl uffa ou lav on s of John n<'l Fomation <'lt CJ 0 lock. Larg lancl.licle hloks of Pi.tur Gorg Rasalt on both sicl s of the road. Ca h clral Rock 1 o k hows 1 dg of Pictur Gorge Ignimhrit int rhed d d \vith gr n tuff C'lnd tuffaceous la stone of John Day Formation. Ahead dip in Pictur Gorge flows defines ncline plunging south (upriv r). Sh ar cl s rpentinit pr su ahl of Tria sic age, R rown and Tha er (1966), 1 ft of road. F.n r<'ln of Turtl to Turtl Cover Area, John Day Fossil Red Cov e r embe r of John Da Formation. a tional Park. T pe ar a Rriclge aero s riv r F.nter Pi ture Gorge Rasalt, brought clown b plunge of syncline. Fault, south sicl up hrings C r taceous conglom rates into contact with Pictur Gorg Ra alt. Conglo e rat exposed on both icies of road at Goose Rock 0 mil s to the outh. Stop 3 p Ro k on ast id of river. Complete section of John Da Formation eel he r R cl heds on 1 ft (to the north) a r e the Rig Rasin on Cr tac ous Ov rlying green tuffs and ston s mak up m r, characterized h ext nsiv alt ration and z olitization. Gorge Igni brit occur in th up -per part of th Turtle Cov r ember. The uff to tan h ds in the upp r part of th formation comprise the Kimb rly Mh r charaterized b unze litized tuffs and tuffac ous lay tone This is ov rlain by two flows of Pi tur Gorg Ra alt. ote small normal fault offsetting Pictur e Gorge Ignim r 'te, downd ropp d to the no rth. Entrance to Pi tur e Gorge. Location of t pe section of the basalt. Th gorg is not named fo r th graffit i orne of it g ologic in the for m of painted number incorr tly designating the numbe r of flows. Lowest xpos d flow in roaclcut on right. Inter s ction with S 26 Continu straight ahec:ld through Gorg low on a ote many W 11 xpos d flow on at on right. L av org with tuffa ou dim ntary rocks of ioc ne 1a call Formation ov rl ing Pi ur Gorg Ra al Turn righ sha rpl onto C:r k Roaci. Pro d 0 1 mile, tur n right on dirt ra k and ontinu for hundr d f to overlook. Stop 4 Dip slop of Pi ur Gorg Ra al no r h p roj ts h n arth valley of John Oa Riv r Pic ur Gorg un'onformably ov rla in b las all formation with 4 3


(?. R O ) 2 n ( l. 6 0 ) 2 80 ( 1. 90) 27. 7 0 (1.00) 28 .70 ( 0 .90) 29 60 ( 0 90) 3 0 .50 ( 0 .40) 30 9 0 (0. 95) 31.85 (0 3 0 ) 32 15 (1.75) 33. 9 0 (2.05) 35 .95 (4.90) 40.85 (0.80) 41.65 ( 1. 00 ) 42 65 ( 1. 9 0 ) 44.55 s 8 n n r on 1. ur h o r g urn 1 f in Pi ur C o r e Hnci r o c ci Rasa l ci ure n o r 1\,(l rei Cll'k l Cl nci r r o ci ur ro


( ')) 4 7 0 (I. l ) 49 .05 (I 90) 50 q (I 70) 5 .65 (1 90) 54.55 ( 0 20) 54.75 ( 0 25) 55.00 (0 .35) 55.35 (0 95) 5 30 (0 25) 5 55 (0 75) 57.30 (2 00) 59.30 ( 1. 05) 60.35 ( 4 35) 64.70 (0. 70) 65.40 (2 65) 68.05 ( 0 8 0 ) 6 8 8 (3. 8 0 ) 72. 5 ( 1. 70) 74.35 (0 4 ) 74. 80 R i'l snake \ve cl cl u f a p s r i cl g e to l f This is w t rnmos oc urr n e of his m mh r in the John Dav draina Highlv alt r cl \.larno ancl si on rig sicl of r oacl Roacl uts \v.l] b in \.larno Formation f r approximat ly h n xt 7 miles. K yes Summi K Clarno a ncl it 1ountain, an Oligo ne volcano omposed of ancl breccias, can he s n at 2 o lo k Clarno lahari br 'ia with v r larg ancl ite lasts on righ id of road. Plat Clarno anclesit flow in road cut on righ t Turn off for 1ith 11 busin ss Con inue straight ahead on S 26 Contact of Clarno Formation and Cretaceous shales and conglom rats on right. Junction Oregon 206 with l S 26 Continue traight ahead on S 26. Turn r"ght on old highwa Stop 6 Examine Cr taeou sancl t on and conglomerate. Continu on old highwa for 0 .40 miles to Or gon 206 Turn sharply l ft on 206 oward Service Creek and Fo il. Contat of Clarno lava flow ov r l ing Cr ac ous rocks in road u on right. Ou crop of C r e a ous sands on conglom ra e and shal \.l rno dik can en at 9 o clock. Roadcut in C retaceou of Cr ta ou ro ks. Sutton ountain. shale Straigh But e at 2 o lo k i Clarno and it on top ahead on sk line i Pic ure G rge Ba alt of On left ar blueschi s P rman (?) metasedim nts. (Swanson, 196q) Rock nclud ph ,llit s, rna rbl s and Hills on left a r compo ed of Cretac ous conglomerate and sand tone. Clarno andesites ov rlie C:r ta ous rock on righ side of highway. Junction of Oregon 206 with Twickenha Road Turn left toward T\<.ickenham At right is narrow hand of John Da sediments betw en Clarno F rmation and Picture \.orge Basalt. For approximately next 7 miles e road traverses Sutton ountain, a shallow ncline of Picture Gorge Basalt. Hackl jointing in thick basalt flow on right. Ha kl ioin ing in hi k ha al low. 1'wick John Oa River. Thi i on he west side of the Su on and Pic ur Cor Rasa t dip as twa rd with John Da rocks s p, eli ff-forming flow at 12 o lock and 2 ar thi k hakl -jointed flows en along road. Hogba k on ri ht is Pi tur \.org Ignimhrit interlay r d with tuff and tuf-fa ous lay tons of John Oa Formation. urn rous andslid bloks of Pi tur Gorg Ra alt obsur th upper part of the John Oa Formation. Pi ur n mbri \.org Ign "mhrit xpo d on righ \{hit r si tant 1 dg is rh olit air-fall tuff \Vith large soda-rich anidine 45 below ig r tals


(I. OS 7 5 R (O.RS) 77.60 ( 0 .35) 77 .95 (1.15) 79 10 ( 0 .40) 79. 50 (1.25) R0.75 ( 0 .40) R 1.15 (2.35) R3. 50 (4.40) 87 9 0 ( 0 75) RR. 65 (5.60) 94 .25 (2 60 ) 9 6 85 ( O .RO) 97.65 THIRD OAY h li v cl John net o h

(0.35) 0 .90 ( 0 40) I. 30 (0. 30) I. 60 ( 0 .2 0 ) 1. RO ( 0 4 0 ) 2 .20 (0. 95) 3 .15 (0. 60) 3.75 (0.65) 4 40 ( 4 30) 8 7 0 (15. 75) 24. 45 (0.95) 25.4 0 ( 0.95) 26.35 ( 0.40) 26. 7 5 Rutte Cre k turnoff. Continue straight ahead on Highway 19. Good exposures of Riv r Rasalt occur along Rutte Creek. Picture Gorge and Yakima Rasalt 'nter.finger near mouth of Rutte Creek where it enters John Da River. Red tuffaceous cla stone in roadcut marks Clarno-John Da contact. Outcrops in roadcut are rhyodacite flow interlayed with lapilli tuffs of the John Day Formation. Junction of old highway. Turn right on Hoover Creek road. .top 8 Examine rhyodacite flow and lapilli tuff of John Day Formation. The lapilli tuff is probably an unwelded ash-flow deposit. It cannot be correlated with any other ash-flow sheets in the John nay Formation and is restricted to the area northwest of the Blue Mountain uplift. Road junction--stay on paved road toward Mayville. Note outcrops of Yakima Rasalt ahead and to the right, characterzied hy stone rings and stripes. ote difference in outcrop character between Yakima and Picture Gorge Rasalts. Stop 9 (Optional). Outcrop on right is light colored, poorly bedded tuffaceous claystone of upper John nay Formation. This sequence is t pical of the upper John nay in the area west and north of the Rlue Mountains. Junction with Oregon Highway 19 Turn left, back toward Fossil. Road crosses Rasalt-John Day contact and enters exposures of John Day Formation. Road junction Oregon Highways 19 and 218 Turn right on Highway 218 toward Antelope and Shaniko. Road cuts for next 19 will be in Clarno Formation. Stop 10 (Optional). Clarno Unit, John Day Fossil Beds ational Monument. Steep cliffs are Clarno mudflow brecia. This sequence is better bedded and more stratiform than most of the Clarno Formation. Entrance to Camp Hancock, run by Oregon Museum of Science and Industry. Locality of the famous Clarno 'nut' beds--numerous plant and vertebrate fossils in the upper Clarno Formation. Cliffs on both sides of highway are basal John Day welded tuff overlying red tuffaceous claystones of upper Clarno Formation. Stop 11. Rasal welded tuff of member A of the John Day Formation is exposed in road cut. From here the unit can be traced, in almost continuous outcrop, to Grizzly, a distance of nearly 50 miles. The unit extends 3-4 miles east of this point, where it is overlapped by younger of the John Day F ormatio n This unit is taken as the base of the John Day Formation in the Clarno-Ashwood-Antelope area, the type area of the western facies of the Formation. In this outcrop the tuff is fairly densely welded, devitrified, and omewhat richer in lithic fragments than elsewhere. Rock contains about cr tals--quartz, sandine, and plagioclase. High p ak to north on right hand side of John Day River is Ironside Mountain, underlain by Yakima Basalt with at least one flow of Picture Gorge Rasalt at base. The Columbia River Rasalt rests unconformably on the John Day Formation. The ledge extending toward the river from the base of Ironside ountains is composed of alkali-olivine bpsalt flows interbedded with tuffaceous sed im n ts of the John na Formation, member F The entire area west of the John Da River is a large landslide in which hlo ks of Columbia River Basalt ar mixed with tuffaceous of the J o hn nay On block of Colum ia River Rasalt, about 3 km long, 47


(0 .B0) 7 5 (0.40) 27.q5 (A. f\5) 34.60 (0 .70 35.30 (0.50) 35.RO ( 2.R5) 3R.65 (3.15) 41. R0 (0 55) 42.35 (0 55) 42.qo (1.40 ) 44.30 (1.35) 45 .65 ( 0 65) 46.30 (3.15) 49.45 (0.70) 50.15 (2.30) 52.45 (0.35) 52.80 oveci as Ct oh<>rent hlock. luMp blo ks. o t p n r 1 htiMfllockv topop r flhv nrl n11M r 1 1!': C'larno Briel Ct'ro.s John Dav Riv r BCtsal ashflow uff of John CtV FonnA ion in roAd u o n righ overl. ing tuffa ou s c im nt. of M mh r F Fr m h r road o te nci-lici ar for n xt 7 Mil Outcrop in liff on righ r has lt fl w o John Day Fom f m mb r II. ion (m mber F) o v rlain bv basal ash-low sh \"hit tuffa ous s cii.M nt of M mb r F John Dav ForfTI

(0. 0) 53. 10 (].35) 54.45 ( 0 95) 55. 40 (0. 40) 55. 80 (0 15) 55.95 (1.25) 57.20 (0 90) SR. 10 (0 80) 5R.90 (0 70) 59.60 ( 1. 90) 61.50 ( 0 7 ) 62.25 ( 1. 00 ) 63.25 (1. 95) 65. 20 ( 1. 1 ) .35 (0 .95) 67.30 (1.20) 6R.50 YakimA BAsalt for next 1 mile. Fault in Yakima Rasalt at 11 o 'clock; left side down. Hogbaks dipping to west (right) at 1 o'clock are John Day welded tuffs up lif ci h ma_ior fault. Road cuts on left are landslide blocks of Yakima Basalt. Low ledges on both sides of road are basal ash-flow sheet of member I John Day Format ion. Juntion with l .S. 96--turn right. Stop 14 Basal ash-flow tuff of member I; unwelded glassy tuff characterized b high percent of lithic fragments, abundant pumice lumps and glass chips and fragments. Contains 1-2% crystals, chiefly plagioclase. Unit becomes thicker and more densely welded to the south. Road junction--turn right on old highway. Road junction--turn right and return to U S 97. Jun tion U S 97 --turn left toward Willowdale. \ illowdale Road june ion--turn left toward Ashwood Ledge on right after turn is basal welded tuff of member H, involved in a small fold. Stop 15 (Optional). Examine ledge of welded tuff of member H Rock is red, ver fine-grained ash-flow tuff. Unit is not highly welded but is well indurated. Contains ver sparse plagioclase phenocrysts and rare rock frag ments. Locall glassy base is preserved. Ahead at 12 o 'clock is landslide lobe which moved down from outcrops on skyline. Ahead at 12 o'clock is dip slope of lithophysal welded tuff of E John na Formation, bounded by fault on far side. Hogback at 9 o 'clock is basal welded tuff of member H with a glassy base and crystallized top. This is overlain at 8 o'clock by basal welded tuff of member I with a thin zone of intervening sediments. Above basalt tuff of member I a r e tuffaceous sediments, mostly claystones with small pumice lapilli. Hill is capped by Yakima Rasalt. Road crosses major fault at this point--west (left) side up At 9 o'clock flat-1 ing welded tuffs of member E cap the rhyolite flow of member C separated by small amount of tuffaceous sediment of member D Just to left of his ou r p he ntir s q uence dips w stward about 40, uplifted by faul Fr m h r to A hwood the road will generall go down section through h John na Formation and into the Clarno. Road cuts for next 2 2 miles will b mber F with landslide blo ks of welded tuff from members G and H Pon ledg Rut at 4 o'clock is apped b basal is ba al weld d tuff of memb r G lded tuff of member H Lower Contact of uffa eous sediment of mem er F and welded tuff of member E For approxima 1 ne t 5 mil s road u will be in lithoph sal tuff of mem er E S op 16. Examin outcrop of lithoph sal tuff of m mber E Rock is fine1 weld d r stallized tuff containing sparse plagioclase and rar rok fragment ; lo all has glass base Abundant form la er alternating with massiv portions of the tuff but 49


(1.20) 69 .70 (0.85) 70. 5 (1.45) 72. 00 (0 50) 72.50 (0 .30) 72.80 (0. 10) 72.90 (0 10) 73. 00 ( 1. 50) 74.50 (0 .90) 75. 40 (0 90) 76. 30 (1.75) 78.05 (2.10) 80 .15 ( l. 20) 81.35 ( 1. 05) 82 40 (0 70) 83 .10 ( 0 60) 83.70 (0. 70) 84. 40 v o cur in v rti ;1l ars to fill a warms This tuff v, ri s g r a ly in thickopogrAphi ci pr ssion on rh olit flow 1aior out rop of tuff and tuffa' ous s dim nts of m mb r n b n a h w lded tuff of m mb r F.. F r n xt 0 mil s \v lei d tuff thin rapidly o ast anci north ast. Road now runs on top of m mb r F.. Low hill on 1 ft ar rhyoli member C fl w of Base of m mh r F. r ting on trah and sit flow of m mb r R, John Oa Formation. La er of s dim nt s parating tra h and it flows of m mb r B. Contat b of middl it of m mb r B and r deli h tuffac ous s diments m mb r A Basal w ld el tuff of member A Contact betwe n John Da and Clarno Formation Road uts will h in Cla rno to bottom of valley. ot r d zon at top of Clarno--thi aproli is dis-continuou and 1 ss well-dev lop d in this ar than in the ar a st of the Blue Mountain Road bends to right. Turn 1 ft on elirt road and pro road. el 0 90 mil o end o Stop 17. At 12 o'clock i steep-sid d rhyolit dom with flow x nding off to right. Oom marks the vent from which th rh olit flow of m mb r C, Jo Day Formation, was erupted. ot cruel olumnar jointing in dom Coarse boulder lahar on left is in Clarno Formation. Return to highway. Junction with paved road. Continue straight ahead toward Ashwood Ashwood Road junction--turn right on gravel road towarel Ha Cr k and Griz zl Contact of John Day and Clarno Formations in roadcut on right. S op 18. Bus will proceed 0 35 miles up road and park. Road traverses a ompl t section of member A of John Day Formation. Rasal ash-flow sheet exhibits two cooling units, mineralogically and petrographically similar. This is overlain by a sequence of poorly exposed lapilli tuffs and tuffa eous layston s which are capped by upper finegrained ash-flow tuff of member A. Entire s tion overlain by trachyandesite of member B Top of hill. Road junction--continue straight ahead and follow left hand fork. Tuffaceous sediments of member Dare exposed in roadcut. Here member lies between trachyandesite of member R and lithophysal tuff of member E From this point the road runs approximately south and roughly parallel the contact between members Band E crossing back and forth for next 5.5 miles. Contact of members Band E in roadcut on right. Contact of members Band E in roadcut on left. Contact of members Rand E in roadcut on left. Contact of members R and E in roaduct on righ L dge on righ is memh r F., much thinner here than in Pony Creek Canyon inor amounts of tuffaceous sediments of member D lie beneath member in roadcuts for the next 0 5 mil s 50


(1. ()()) R5. 40 ((). 20) R5.60 ((), 40) R6.00 (1.30) 87.30 (0 45) 87.75 (2. 10) 89.85 (0 50) 90.35 (1.35) 91.70 (2 85) 94.55 (0 .45) 95. 00 (0 95) 95.95 (1. 40) 97.35 (1.15) 98. 50 (0 65) 99.15 (1.15) 100.30 ( 1. 50) 101.80 (0.35) 102.1 (0 .70) 102 .85 F.x' 11 nt view of t igh Cas ade peaks on right. Oouhle hut te in foreground is Tell r P.utt --capped with weld d tuff of l'lel'lh r c.. Conta 't of l'lel'lh rs Rand F. in road ut on left S dil'lents h tw n trachyandesit flows of m mh r R Weld d tuff of member F overlies member R in roadcut on left. Welded tuff of memb r F. forms promin nt ledg dipping northwest on hill at 2 o'clock. Just ah ad, where road swings left, the hill at 1 o'clock is capped with welded tuff of m mher C.. Tuffac ous sediments of memh r F are exposed in roadcuts on right. Roadcuts are in memh r F for approximate! n xt 4 miles. \.Jhite pumic lapilli tuffs in roadcut on right are near the base of member F Good exposures of member F occur on hill at 8 o 'clock. These are mostly red to gray tuffaceous cla stones and tuffs. Ledg at top of hill at 3 o 'clock is Quaternary basalt. Small dam on left. From this point the road swings generall west and rises in John Oa Formation. ote landslide blocks of Quaternary basalt on right side of road for next 4 5 miles. Fault crosses road, uplifting welded tuff of member Eon west. Fault scarp is visible to left, across meadow Road juntion to Ha Creek Ranch--turn right toward adras. Quaternary basalt crops out on sk l'ne at 4 o'clock with man landslide blocks below the rim. The cliffs at 12 o 'clo k ware welded tuff of member C., John Da Forma -tion. Road junction--turn right toward Willowdale Stop 19. Examine outcrop of welded tuff of member C.. nit is densel welded, cr stallized, and contains sparse lithoph sae. ote abundant large soda-rich sanidine cr stals. A major fault runs along the road, tilting the welded tuff sheet to the west. The low cliff on the west (left) side of the road is the basal ash-flow sheet of member H This is overlain by Quaternary basalt, which at this point flowed down old channel of Hay C reek. Turn a r ound and return to jun tion of Hay Creek Road Junction of Hay Creek Road--turn right. Entrance to Hay Creek Ranch--take right hand for k of road. Lithoph sal w ld d tuff of member E on both sides of road. Small landslide a 10 o lock. Rr w r R s rvoir. Fault runs along front of liff on right, uplifting ledge of memb r E Road junction--turn right. After turn, flat-topped hill at 1 2 o 'clock i s basal welded tuff of m mher H overl ing member G. Hill at 10 o 'clock is w ld d tuff of m mber C.. R d outcrop at 2 o' lo k is etion of tuffa eous sediments of member F Road lil'lbs o top of a h-flow sheet at base of member C.. Here, welded tuff dip g n rall south but i folded into a series of broad, open folds. Out rop a ross fi ld at 11 o' is ba al welded tuff of member I and cliff 51


('0. 7)) 1 r (0.50 I 04. 10 (0 nS) 104.75 (0 .65) 105 40 (1.40) (0. 45) 107 .25 ( 0 55) 107 .RO (0 .35) 1 OR. 15 (0.70) lOR.RS (1.35) 110 .20 (0.40) 110 .60 (1 .55) 112 .15 ( 1. so) 113 .65 (0. 90) 114 .55 (0.25) 114 .RO (1.35) 116 15) (0. 25) 116 40 (1.20) 117 .60 At I o'clock i. "'eldc>rl tu f of I" r1h r l urlift cl AlonP ;t M;tjor nult Far r :1 1 :' o'l'lok is Rut in h Rok Ar :1. Stop "0. Tuff in roacl is rd ssv, Ash-flow tu f Cl ha e l';:llk 100 vet rrls up roacl to 1 ft flnci Xflmi n hASfll r I. r the unit is cl n lv w lrl rl with rt goorl vi.trophvre at rystalliz cl portion ahove. l 'eld rl tuff o r I ro. s s roacl. L dge on hoth irl of roetd i nuaternarv ha alt. Road JlMhs to top of hasalt flow. This flow cl f r m smetll hi lrl volano a nile to th east ancl flowecl rlmm old s r ar1 hann 1 towarcls 'unction guetrd--tt:rn 1 ft. Small mound on right i hi ld volano from whi h Ouat rnary ha al flow was erupted. Grizzlv P ak i at 1 2 o ]ok. f.dg of O uat hasalt. HE>r unit ov rli trahyancl it R John Dety Fomation, whih i expos d quarry at 1 o lok. ing clips of Grizzl Peetk and Cray Rut in eli tan at o lok. underlain h rhvolit flow prohethl folded into hroad vnlin of rn fllber o t oppos Ro h I elded tuff of meMb r A, John nay Formation, in ut. Thi L h southernmost outcrop of this unj t. To th outh, olcl Clarno hill ar onlappecl h ally hi h r units of th John nay Roacl junction--turn right towarrl Crizzly. Roacl iunction--turn right towarrl and Prinevill f.dge of Ouaternay hasalt flow exposed on right. Roacl has climh cl to top of flat-lying farlra Fomation. This uni mos ly of Pliocene age, consists of fluviatile silt tones, sanclston etncl onglomerates with interlayE>red air-fall and ash-flow tuffs anrl hasalt flows. SMall outcrop at 2 o 'clock is rh olit iMilar to that of Grizzl P ak, com-pletely surrounded by rocks of Madras Formation. Good view of Gray Rutte at 10 o 'clok. Low ridg just 1 ft of Cra Rutte is capped by saMe rhyolite flow. Low ridge at 9 o'clok is YakiMa Ba alt overlying tuffaceous sediments of John Da Formation. Low b nch to right of Gray Rutte underlain by Clarno-type flows dipping outh ast. Junction with t s Highway 26 Continue straight ahead on rlirt road towa r d Culver. Ridge at 3 o 'clock is hasalt ash-flow tuff of member H. ot tha strike of John Day is nearl f.W Hills on right at 1-2 o'clock ar also capped by welded tuff of memher H Road runs on surface of adras ForMation. Low hills at 3 o'clo k are capped hy welded tuff of member H Flat-topped hutte at 11 o'clock i Rutte, also capped by welded tuff of memher H Left of Ha stack Rutte this welded tuff laps onto Clarno-type rocks of Smith Rock. Road junction--turn left. Road junction--continue straight aheacl. Roarl .iunction--turn right. Goocl view of Crav Rutte on lef af er turn. 52


(0 95) i!R .55 ( 0 85) !19 40 (0 40) 119.RO ( 0 75) !20 .55 (0 .95) 121. 50 ( 0 .65) 12 2 15 ( 1. 20) 123.35 ( 0 2 0 ) 123.55 (0 40 ) 123.95 (1.30) 125.25 ( 0 20) 125.45 ( 0 .50) 125.95 ( 1. 20 ) 12 7 15 (0 20) 127.35 (0 .50) 127.85 (0 .95) 12 8 80 (1.75) 130.55 (1.95) 132.50 (0 .40) 132.90 (0 20) 133. 1 0 (1 05 ) 134. 1 5 (1.15) l3 5 30 ( 0 .60) 135. 9 0 ( 2 5) 138 55 Contact of Madras Formation and John nay Formation John Day tuffaceous sedim ents of m m her C form small hill surrounded by Marl ras Format ion Contact of John Day sediments with Clarno Formation. Road junction--turn right. Good outcrop of Clarno Formation on right. Roadcuts are in Clarno Formation for next mile with John nay contact just ahove road on right. Road 'unction--turn left. Haystack Reservoir at 2 o'clock. Top of ridge just be yond reservoir is welded tuff of member H Just below is a local welded tuff that pinches out to the right. Welded tuff of member H crosses road. Ridge at 1 0 o 'clock is lava flow from Juniper Butte. Road climbs to top of Madras Formation which surrounds the older rocks. Rutte at 11 o 'clock is Round Butte, a small shield volcano from which some of the basalt flows in the Madras Formation were erupted. Road unction-turn left. Junction U S Highway 97--turn left toward Redmond and Rend Juniper Butte, a rhyolite dome of John Day age at 1 o 'clock. Lava flow f rom Juniper Rutte at 9 o'clock. Welded tuff of member H John Day Formation in roadcut. Three Sisters volcanoes in High Cascades visible at 2 o 'clock. Appproximate contact of Madras and John nay Formations. Haystack Butte at 9 o 'clock is capped with welded tuff of member H Gray Rutte at 10 o'clock, capped with rhyolite flow. Laharic breccias o f Smith rock at 11 o'clock. Road cuts with Madras basalt flow overlain by tuffaceous sediments. Madras basalt flow. a r o C r ook d R v r r. rg Stop 21 (optional). Turn right to overlook for Crooked River. Gorge was cut in Madras Forma ion, then filled in with Quaternary basalt. Present gorge was ut through intracanyon basalt which is visible in canyon walls. Not e olumnar join ing in intracan o n ba alt. Return to Highway 97 and turn right. Tuff of Smith Rock e xposed at 9 o 'clock. Basal flow of Madras Formation in roadcut on left. T rr bonn T h row Rutt This is a ries of inder ones marking vents from whic h som of the basalt flow of the Madras Formation were erupted. 53


(2 00) 140 .55 (0 90) 141.45 (0 20) 141.65 (2 05) 143.70 ( 5 70) 149 40 ( 5 75) 155.15 ( 0 85) 156 .00 (0 25) 156 .25 (1.55) 157.80 R dmond Juntion Or gon 126--ontinu traight ah ad on Highwa q7. Junction Or gon 126 --ontinu traight h ad n Highwa q7 Old hannel cut in Madras Formation nd fill d with Qu t rn ry ba al Low hill on right i age. Dome projet sil'ci dom of F rk d Horn Rutt probably John n through Ma cl r as F o rm a t i n Junction U S Highwa 20--continu str ight ah ad R nd. Road junction--turn right toward it ent r End of road log--corner of Franklin and Wall Str t 54


CE TRAL IIIGH CASCADE ROAD IDE GEOLOGY BE D, SISTERS, 1CKE Z I E PASS, D S TIN-1 PASS, OREGO Edward 1 Taylor Department of Geology Oregon Stat e University I TRODUCTIO This geologic field trip guid is a rev1s1on of "Roadside Geology Santiam a nd 1cKenzie Pass Hig hways, O regon" (Taylor, 19 68) w ith a dd itions pertinent to geology of the Bend-Sisters ar a. Checkpoint mileages express cumul a tive dista nc e from Cen ral Oregon Com muni y College in Bend, over a route whi c h leads to Sisters, around the 1cKenzie Pass-Clear Lake-Santiam Pass highway loop, and back to Bend (see Index Map, Figure 1 ) Person s using this guide should anti ipate one full day of at least 140 miles, and should check ac essibi li t y to fcKenzie Sununi t during winter, spri n g, an d late fall. GE ERAL REMARKS ABOUT CE TRAL CASCADE GEOLOGY A calc-alkalic volcanic arc has been in ermit-ent l y ac tive during the last 10-15 m.y along he eastern part of the central Cascade Range in O regon. The late P leistoce n e record of this vo lcanic activity is well p reserved on the crest of the High Cascades; he best exposed record of early P leis ocene, Plio cene, and late 1iocene Cascade volcan ism is foun d in olcanic a n d volcaniclastic deposits on the east flank of the range and i n the a djacent D schutes Basin. In the followin g discussion, structural, s ratigraphic, and magmatic features of the \estern Cascade, Hig h Cascade and Deschutes Basin subprovinc s are described and their interrelationships are briefly summarized. Centra l Hig h Cascade Province The central High Cascade Range i n Oreg on is chiefly a Pleistocene volcanic platform of overlapping basalt and basaltic an d esite lava flows whose aggregate thickness is generally unknown but probably exceeds 4, 000 feet locally. This platform is elonga t e north-south and is 20-30 miles wide. typical volcano of the platform is a broad shield of ligh t colored, vesicular basaltic andesite with a inder ha has e n in aded b plugs and radial is o en exam le e pos d in cross-se ion erosion include phinx Bu e sou h of posi c s rue ur r shie ld base 10 mile band ."or h Sis Finger d Jack. p la form w re a i duccd small c i nd r Deer Bu e nor h of Los k of om on a !Ius he es s, and n e 55 cinder cones are no less abundant but they are not as well preserved. Examples of glaciate d remnants of P leistocene cones include Bluegrass Butte Condon Butte, and Scott fuuntain near McKenzie Pass a nd w ell Butte, lloodoo Butte and Cache Mounta i n near Santiam Pass. systematic temporal inhomogeneity exists w ithin the Hig h Cascade p latform. Early Pleistocene lavas were predominantly high-alumina olivine tholeiites i n vesicular, thin, widespread units, commonly with pronounc ed diktytaxitic textures. Later P leis tocene lavas wer e predominantly high-alumi n a basaltic andesites i n thick, platy units, ge nerally with pilotaxitic textures. The arly basalts crop out i n greater abundance and ariet y along the w estern and eastern margins o f he platform and i n the w alls of deeply glaciated canyons however, identical basalts do occur at highe r levels. Later basaltic andesites cover most o f the p latform but they are also found at lower stratigraphic levels E xamples o f early ba salts are well exposed along the west margin of the platform at Cupola Rock in Lost Creek canyon, in cuts of Highway 126 north of Trailbri d ge R eservoir, and in the valley o f Hackleman Creek west of Fish Lake. Early basalts along he east margin o f h e p l a t orm are widespread in the upper 1 etolius Riv e r valley and in the vicinity of Sisters. A systematic spatial inhomogeneity also exists within the High Cascade platform. West of Bend, in the vicinity of South Sister, silicic volcanic rocks are interbedded with a nd r est u pon the mafic p l a t form rocks. A silicic highland of some 15 miles b readth was produced by the development of rhyolite rhyodaci e and dacite domes surrounded by andesite dacite, and rhyodacite lava s and ash-flow tuffs. of this highland is mantled by mafic cinder co nes and la as of the Triangle Hill group and by composit volcanoes such as Broken Top. E xamples o f inte rbedded andesites, daci t es, and rhyodacite lavas are common i n upper Squaw Creek a n d Tumalo Creek canyons. Two units of bla k andesitic ash-flow tuff (Century Drive Tu f and hevlin Park Tuff) and on e unit of pink, dei rified, da itic ash-flow tuff (Desert Sp r i n g Tuff) were erupted from the highland and are w ell exposed wes of Bend. Two rhyoda ash-flow tuffs (Tumalo Tuff and Lava Island Tuff) and one ex ensive lapillifall pumice deposit (Bend Pumice) of High Cascade origin are also exposed near Bend but were probabl y erupted from a vent south of he silicic highlan Sou h Sister volcano is chiefly andesite with minor dacite and rhyodacite. Broke n Top (east of South Sister) is basaltic a ndesit e with minor interedded dacite and rhyodacite lavas and small-volume ash-flow uffs. 1iddl e Sister ( north o f South Sister) is basal wi h minor basalti andesite, andesite, da i e and rhyodacite. South Sist e r and nearby volanoes are relatively l a t e products o f long-continued, ompositionally diverse, and localized silici magma-


c \' 0 0 30------4-----------r-----r-----------------+' X THREE FINGERED <) JACK I ( t;)) W \ err u l XM{}(JNT D OREGON MILES KILOMETERS ( WASHINGTON ( I ...._ SISTERS 0__.!..-126 lz6 \ 't: I 0 ( (j) \ 4.} l)l :::-NORTH MIOOLE X SISTER SISTER \ l SOUTHX SISTER ( ,/ ">

tism. In summa r y the not th simple P I i Rang is f a nd iL t xtbo k lavas sub rdinal d thi< in .iso-Deschut Basin Pr vin Th ast rn mar in platform is m rk d by a v r th late lioc n and Plio Early platform intra anyon lavas xt nd miles ast of the Cascad foothills a nd onta t with Formati n. as much as 5 isolat d Pl istoc n vol ano s of basalt and basaltic and sit rest on Plioc n ro k s of th D s hutes Basin. Examples of High Cas ad intra a ny o n lavas oc ur in low r M tolius Riv r can yon, D p Canyon, and n a r Squaw Creek, Bull Flat, and Tumalo Cr k b tw n Sist rs and B nd. Isolat d vol ano s of Hi gh Cas affinity includ Squaw Ba k Ridg Long Butt and Pilot Butt and Awbr y Butte n ar B nd Th D schutes Formation ontains str am-d posit d silt, sand, and grav l, and sitic-to-rhyodacitic ashflow and a h-fall tuffs, and interb dded basalt flows. Th basalti 1 vas w r rupt d from ind r o n s a nd fissure v nts within the D hut s Basin but th picla tic a nd vol anicla tic r o ks w r hiefly of Cascade prov nan Close to the Cas ad s, th formation b om s thi k r, basaltic a nd sit lavas pr dominat and th. vol ani lastic ro ks b com disconint rb d Indicators of transport dir tion volcani lastic s dim nts point astw ard. Deschutes Formation sourc vol a no s w r coincid nt with, or not far r mov d from the High Cascad axis. One d epl diss ct d r mnant of a n and sitic D schutes Formation sourc vole no is located at th bend of M tolius River, 12 mil s east of Mount Jeff rson Strata of th twe n Warm h of 57 by lavas f aul rpr tation can be summariz d 1 D schu Formati n ba lti lava and ash-flow f r m vol a n o s n ar th axis and flow d astward into n xl y in-lavas d Basin m r than 4 5 m.y ago. Rock units d posit d by this pro ess have be n trac d from Gr n Ridg to th D schut s Riv r. Th y r v al a continu ou gentle astward dip of onl l-2 d gr s Therefore, Green Ridg cannot represent a tilt d-up fault block. It is, instead, a r mnant of tabl ast rn Cas ad foothills and its rocks still r st on an initial pal a slope. 2 Th maximum ag y t obtain d by radiom tri dating of central High Cascad platform ro ks is 3 9 m.y. Th r fo r Deschutes Formation ro ks and sourc volcanoe must li b n ath the Cas cade platform and displac ment of the fault syst m probably xceeds (and might g r atly x d) 3,000 feet. 3. IE Gr en Rid ge has not be n displac d upward r lativ to the D schutes Basin, th Cascade axis has be n displa ed downward. This pr suma bly occurr d about 4.5 m.y ago, terminating d -position of D schut s Formation ro ks. Is i possibl that a whole rang of Plio n comp -sit vol ano s found e r e d and was buried ben ath the Pleisto n Hi gh Cas ad platform? Su h an int rpr tation is strongly suggest d by availabl vi-d nc \Hth appropriat informality, this hypoth ti-al s emblag of volcanic rocks might be all d th "Plio-Cascades." Ea t Margin of th Central West rn Cascade Provin Stratigraphi and structural r lationships at th w st rn margin of the central High Cascade platform ar obs ured by m o r extensiv rosion, thic k r alluvial ov r, and mor luxuriant v g tation than along th east rn margin. How ver, striking simila are vid nt. Isolat d Pleisto n volca n o s of basalt a nd basaltic and site o cur in th \ st rn Cascade at 1 ast 20 miles west of the platform. Examples nclud Harter Mountain, a cinder con and flow near Quartzvill and Battle Ax Mountain. Early High Cascad platform basalts on lavas in the adja nt \ s in-lud diktytaxitic ba can-y n of M K nzie Riv r imilar rocks in the canyon of orth S ntiam River. R o k of the W stern Cascad s adjac nt to th High C s ade platform ar pr dominantly late and Plio ene mafic lavas (Armstrong a nd o her 1975; Sutter, 1978) with subordinate ash-flow and sili ic vol anic domes. These rocks n included in the S rdin Formation by Peck and oth r (1964) but should be assign d to the Outerson F rmation of Thay r (1937). Thy ar equivalent in litholo y and ag to th D schutes Formation. They g n rall flat-lying x ept wher they are offs northw st-trending n nnal faults of small m nt. Although man units of th Outerson Formation w r v d along th ast rn dg of the West rn Cas ad s v r 1 and sitic ash-flow tuffs and one ba alti a nd sh-flow tuff app ar to have moved w stward from cad vol anoes. Exampl s of Out rson Form tion ro ks n b s en along th southto-north re tlin of mountains whic h includes Fris-


s ll Poin Bunchgr.1ss un .tin, 1 un ,in, E hl 1 untain, Crc Pyramids, and Coffin Butte. Out r he same age as lower (Arm trong and Bn\"d r Ridg L lro n I until in Thr 'l' 1 ngth, this aul s n ati n of the es arpm n and by lavas of th Cascade platfo r m deposit d a ain t th ba escarpm nt. Sardin e Little B utt e etc. Western Cascades Sj)

liLES ( 1. 2) 1.2 (0. 3) 1.5 (l 2) 2 7 (0. 3) o ROADLOG FOR CE TRAL HIGH C SCADE GEOLOGY BE D, SISTERS, !CKE ZIE PASS, A D SA TIAM PASS OR EGO Edward M Ta lor D partm nt of G ology Or go n Stat U niv rsity Start in B \.Jay and wport Ave. in valle b t\.J n Awbr a nd d W (out of town) on wport Av .-Shevlin Park Road. Ro d uts ahead ar fl in lavas asso iat d with inder co n s S.W. of A\vbr y Butt Awbr a shi ld volcano of diktytaxitic basalt that was probably last activ during arly Pleistoc ne. It r sts upon tuf ac ous dim nts and lavas of th upp r part of h D s hutes Formation. Four High Cas ad ash-flow tuffs mov d W-to-E b twe n Awbr y a nd Ov rturf Butt s and ar now s n in scatter d outcrops and in xcavations, eastward to th Deschute Riv r The ash-flow units ar : (1) D sert (black to gray, da itic, poorly w ld d, larg black pumice fr g m nts), ov rlain b (2) Tumal Tuff (oran rhyoda itic, poorly \.J lded), overlain by (3) Lava I land TuffTcr am to buff, rhyodacitic, complet ly d vitrifi d, ov Tuff (black, a nd sitic, mod ratel \.J lde d and rather well sorted along ash-flow ---m rgins). Right-hand urv ot pumic quarri s on left. They r veal welded Tumalo Tuff with a basal, thin whit intralay r r sting on Bend Pumice which has been xt nsiv ly quarri d for aggragate. ote a red of altered cind rs on right-hand road side. L ft-hand curv Th las 0 5-mile of ro dway has follow d the Tumalo which b hind this point xt nds 2 mils S.l2E. and ah ad of this point ext -nds i n a 43\..J. dire tion for 1 mile, th n ontinu s .38\..J. for 11 miles mor Rocks on th S W sid of th fault hav be n displac d downward alon g each s gm nt but not uniforml in amount Roadwa ah ad 1 ads ov r a narrow High Ca d mafi lava flow that mov d S W to E Th lava flow and und rlying Spring Tuff, Bend Pumi and Tumalo Tuff have all ben displaced b th Tumalo Fault. Th ov r l ing Sh vlin Park Tuff has not b n displac d her but has b n ffs t alan th S E segm nt of th fault. orthw t of th 1 va, th urfac i mantl d b th nonw ld d faci s of the Sh vlin Park Tuff. gr d J I 8 mil d h trip. h vlin Park 59 Follow park road to bridge id of r k where b dd d lapilli and ash of th k bl ck of d ns ly d pumi fragments. Road of Tumalo Cre k anyon, ut


High Cas d L v a Fl w eN Formation Lavas and Tuffs Tumalo High Cascad Lava Flow Cind r Con 2 7 Four Ash-Flow Tuffs FIGURE 3. GEOLOGIC SKETCH MAP FROM MILE 0.0 TO MILE 16.6 60 n Basalt Butt


( 2. 2) 5.2 (0. 7) 5 9 (1. 3) 7.2 (0. 4) 7.6 (0.6) 8.2 (1. 7) 9.9 (1. 8) 11.7 (0.4) 12.1 Roadway ah ad pass s W. of Tumalo Butte cinder one whi h is older than th ash-flow tuff uni Road uts on W flank of co n e expos fine-grain d marginal fa is of Sh vln Park Tuff. Lev 1 terrain of Tumalo Butte is und rlain by Pl isto en grav 1 deposits which occur stratigraphically between Tumalo and Shevlin Park Tuffs and hav b en ext nsiv ly quarried for construction aggr -gat Sharp turn right. Roadway for next mile is on top of a High Cascade mafic lava flow whi h mov d W. to E Lava is overlain by Pleistoc n e gravels. L ft turn leading to N.E downgrade, margin of lava flow. Roadcuts on righ expos basal contac t of lava resting on red, oxidized soil that develop d upon Tumalo Tuff. Nonweld d, pink upper zone of Tumalo Tuff has been remov d by erosion from lowlands to the revealing bedrock of orange, welded Tumalo Tuff between patches of Pleistocene gravel. Junction of Johnson Road and Tumalo Market Road. Desert Spring Tuff is exposed in vicinity of the junction and rests upon Deschutes Formation lavas. Turn left onto Tumalo Market Road STOP NO. 1 : Broad shoulder on left side of road opposite orange-colored road-ut. Examine vertical s ction through Desert Spring Tuff. From ere k bed to road 1 v 1: Partly collapsed black pumice fragments in pink, devitrified, weld d matrix are concentrated at base, but nearly absent above. From road 1 vel to top of unit: Hydrated, orange and yellow, poorly welded matrix with abundant fresh, noncollapsed pumice fragments. Alluvial gravels on tuff are overlain by Bend Pumice. Exposed in quarry a few 100-feet W : Bend Pumice overlain by very poorly welded and nonwelded Tumalo Tuff which contains a finegrained white intralayer at its base. Exposed in roadcut and quarries a f w 100-fe t E : A NW-SE normal fault which dropped the section about 12 feet on the .E. side. Bedded pumice deposits are exposed at roadway curve; they belong to the basal member of Bend Pumice and consist of ash-fall and reworked pumice ash and lapilli. Turn left onto Tumalo Reservoir Road; proceed west. Pleistocene aliuvial gravels mantl the nearby t rrain. Quarries intermittently visib1e to the north have b en xcavated wherever the overburden of alluvium and Tumal0 Tuff is sufficiently thin to permit extraction of Bend P umice. Tyl r Road R commended side trip to examine welded facies of Tumalo Tuff. Follow Tyler Road 0 3 -mile S stop at base of grade; outcrops are N E of the road. Thi wa n r t th margn of th Tumalo ash flow. Pumi e fragments s 1 rg 1 oot ontain ores of the only fresh, white, nonhydrated glass to b ound in this unit. Along roadway ahead, alluvial gravels give way to at r d out rops of D schut s Formation mafic lavas as the Tumalo Fault is approa h d End of Tum lo R s rvoir Road : turn right and cross brid e over o utlet str am from Upp r Tum lo R s rvoir. D sert Spring Tuff forms S. bridge abutm nt; D hut Form tion lavas ar xpos d along th N ba n k of the stream. This r ad foll w h Tumalo Fault for the next 7 miles. STOP NO. 2 : Examin Tuff exp s d along fin grain d mar inal faci s f and sitic Shevlin Park .E. sid of road. Bull Flat (W. of the road) is underlain 6 1


(1. 8) 13 9 (l. 3) 15 2 (l. 4) 16 6 (l. 2) 17. 8 (1. 1) 18.9 (l.l) 20. 0 (0 5) 20. 5 (4.4) 24. 9 (0. 7) 25. 6 b 0 vium h vi 1 n Boundar b tw n g Hill on th \ id and it lava low scarpm nt on th E Old Plainvi w School Highwa 20 In this an earl Hi gh Ca a d broadl to th E umi wn-d rc pp r Tuma l R s a d is th F rma ti 0 th r ad. Fi E. n ur l 3 nd 4. rop \ 0 s hut rk Tuf and allu-u1 R a iva ion of h tuff uni L s and I I u-in r to th whic h b s 1 i d dd d n-a ught i cti n with S E and pr d Int rs ction of Gist Ro d with Highw 20 Turn l ft on 20 tow rd Si t r Highwa curv l ft Sist rs, Pleistocen are common. Junction Hi ghwa s 20 r ound roadcut alluvium and in and 126 Pro Plei tocene alluvium, mo t of whic h alluvium is und rlain in this a r a basalt flows. b in 0 hu nspi uou Form tion lav s outer ps o Pli n d \ through Sist rs. Si ter has b n d posit d by Squ w Cr \vid pr ad High Cas ad dikt June t ion Highw a 20 and 24 2 Follmv 24 2 \v. toward K nzi P vi w of Thr Sist rs volcanoes o n S W skylin 62 From h r t m fie lava is built on k Th a xi i c Ex llen t


26. 8 Formation Lavas and Con s s Fo rrna ion La v High Cas ad Basaltic Lavas T rac High Ca ad B And i FIGURE 4 GEOLOGIC SKETCH MAP FROM MILE 16. 6 TO MILE 27. 1 63 d salt


(l. 4) 26. 8 (0 3) 27. 1 ( 2 7) 2 9 8 ( 2 3) 3 2 1 (0.9) 33. 0 (0.6) 33.6 (3. 4) 37.0 (3. 0) 40. 0 (0. 3) 40 3 L \ i diktyt X 1 b S Jt h highway as nd h' fly outw h Boundary b tw n g olo i sk h Figur s 4 and 5 L ft-hand curv at sit of Cold Spring. basaltic and it lava flow whos ur Lake, 9 mil s. \ of h r T h highway the next l.l miles, th n a nd othe r 1 1 miles. Fourmil Butte n abou t Pl isto ne mafic lava flows issu d from on 25 squar mil s T h w stern par c ne m oraine and outwa h d posits. Sharp right-hand turn. Th highwa al moraine d po it d alon g th s urf urv id of i from th margn oE a n r Sou h M thi u of t h f 1 tv or flatv for an-Fourm i i on of 10 1 te nd Bl ck Butte. So many ld now cov rs ppr xima ely buri d und r lat Pl i o -around th E. lat Pl ist nd of a 3-mil later n gla i r Li le Butte j ust of the highw is a t rminal mora'ne Secondary road This point mark basaltic lava from Belknap Crat r flow for th n xt 3 2 mil s th t rnmo t xt n ion of Holoc n Highway will follow th S margin of th's Windy Point. Glaciated promontory of ba altic nd it lavas and cind rs a the .W. base of Black Crater. Excellent vi w of th bro d shield o Mount Washington s urm ount d by glaciat e d summit on nd plug. In for ground i Holocen e basaltic andesite lava from Yapoah C n Farther W is th' lava field of Little Belknap. Highwa ahead will follot the margin of Yapoah lava for 3 miles. Outcrops of glaciat d basaltic and from Bla k Crater and basalt from smaller Pleistoc n e i nd r o n s n S.E of h highway. Highway cuts through E lev e of Y a poah flow S v ral flow units can b s n in whi c h thin lenses of dense lav a a r e separat d by thick r lay rs of rubbl STOP NO. 3 : Dee Wright Observatory. From th o bservatory roof th follow i n g landmarks are seen proceeding clockwise f rom true north in azimuthal d gr es. 0 Mount Jefferson: Andesitic composite volcan o 7 Cache Mountain: Glaciate d basaltic andesite volcano, 0 9 m y old, with late Pleistocen e basalt cinder cones and lavas on summit. 11 Bald Peter (far horizon): Basaltic andesite volcano, 2 1 m.y old, deeply glaciated constructed against the N. end of the Green Ridge fault escarpm e n t 20 Dugout Butte (forested foreground): Glaciated diktytaxitic basalt flows of the High Cascad e platform. 64


33 6 33 0 32. 1 High Cascade Basaltic Andesite High Cascade Basalt High Cascade Basalt Lat ral Moraine Terminus of Lava from Belknap Crater South Belknap Lava Lava From Belknap Crater Lav a Lav a FIGURE 5. GEOLOGIC SKETCH MAP FROM MILE 27.1 TO MILE 34.5 AND FROM MILE 34.5 TO MILE 44.6 65 I


30 Green R"dge (far horizon): N-S fault block moun ain, 2 0 miles long. Escarpment faces W. Lavas on cr st are ?-6 m y ol 40 Black Butte (background): B s ltic cinder cone located at S end of Gr en Ridge Older han any of he o her visible High Case e volcano s ; well preserved form is due to lack of glaciation east of the a cade Rang Bluegrass Butte (foreground): Gl ciated b saltic andesi e cinder cone. Ridge E of cone is a lateral moraine. 82 Black Crater (fills most of eastern sec or): Lat Pleistocene basaltic andesite volcano. Cra er" is actua ly a glaci l cirqu open o the N.E. 105 Unnamed Cascade summit ridge composed of glaciated basaltic andesi e to cinders, bombs, and lavas which issued from a -mile-long ch in of cones. 155 Probably was the site of spec acular lava fountains nd eruptions of unusually large and abundant volcanic bombs during the late Pleis ocen 168 North Sis er (elev. 10 085 fee). Basaltic and si e composi e vo cano on a broad shield. Central plug and dike systems exposed by glacial erosion. At base of North Sister stand Yapoah Cone (lef ) an Colier Cone (righ ) 174 Middle Sister (elev. 10,045 feet): Composite volcano supporting Co ier Glacier. Predominantly olivine basalt porphyry but also contains flows of basaltic andesite, andesite, dacite, and rhyodacite. Younger han Nor h Sister. 178 Summit of Little Brother and ridge W: asaltic composite volcano with exposed plug and dikes. Older than North ister. 188 Four-in-One cinder cone (below skyline): A ridge-cone breached in four places by andesite lava flows, 2600 years ago. Part of a N-S alignment of 19 vents. 197 The Husband (on skyline, partly obscured): Although of late Pleis ocene age, it i s one of the oldest and most extensively dissected volcanoes in this region. The exposed plug of basal ic andesite is 1/4-mile in diameter. 218 Condon Butte: Late Pleistocene cinder cone surrounded by glaciated lava field of basaltic andesite. Nested summit era ers. Knob visible at left base is an unnamed glaciated dome of rhyodacitic obsidian. 235 Horsepasture Mountain (far horizon): Western Cascade peak. 256 Scott Mountain: Small summit cone on a broad glaciated basa tic shield. 282 South belknap Cone: Cone was formed and breached 1800 years ago, then surrounded by basaltic andesite lava from a nearby vent about 1500 years ago 285 Unnamed twin steptoes (in foreground lava field): Glaciated basaltic anand desite volcanoes surrounded by lava from Little Belknap. 306 309 Belknap Crater (summit cone on skyline): Focal point of a long-continued and complex episode of Holocene basalt and basaltic andesite volcanism. The broad shield which fills the N W view is 5 miles in diameter; it is estimated to be 1700 feet in maximum thickness and 1 3 cubic miles in volume The volcano probably contains a core of cinders which inter-66


(0. 3) 0 6 (0 5) 41.1 (1. 3) 2 4 (0 6) 3 0 (1 0) 4 0 h ummit s \ main ir v nts. 321 Littl B lknap: ano, built 2900 ar ago on h E flank E th 340 Mount shingt f t): Gla iat d r mn nt of a larg basalti C n ral plug lank d b -S swarm of dik an ri \.Jh h ral 1 h Highw dm.;rn and it omposit in ummit on h d Th n The visitor aero tub collaps d and lava b y th rmal contraction and los onsolidation. On th E sid o the main hann 1 is visitor mi ght w 11 imagin the hann 1 in full flood, high a th 1 v cr st nd spilling ov r th ast rn flank. lava diminish d molt n int rior drain d and th subsid d Lacking support, large s gm nts of th tipp d causing d p, irr gular t nsion cracks to op n in f rd a vi w of on of th s cracks and o ve-t bas of th 1 v Yapoah flow W margin. Tr s grow r adil upon par l upon th old r Lit le Belknap flow. This r d with blocks, int rstiti 1 scori nd noti fly from oh r n t In g n ral, clj.. _ma te and highway 1 ves th dark soil. It is 3 cone. f a Littl B lkn p lava tongu which pour d S E r 0 in uth in i m r r nt and in its nbundan 67 he light-olor d b drock m rgimov d \ from h r tow rd a \vh r 1 v vall y has be n buried it ross s an alluvial from South B lknap B lkn p lav i distinct urf on 1 r g in g r in d glom r ph no -


L'f p l.1 g i h l ,lst' u l i vinL' .tnd tl in(l p rllx n Thi s l nv< l \vas vrup l d I ROO v ,;lt"S .l)!,l. Tn Lll' i s ;1n l':-;t' IIPnl vi. '\v o S. s i{'P' of lkl knnp v o lt'dn o '1\vin C r

eN Earl High Cascade Lava From Sims Butt Lava From South Belknap Lava From Twin Husband From Colli r Cone FIGURE 6. GEOLOGIC SKETCH MAP FROM MILE 44.6 TO MILE 57.9 69


(2.5) 56. 0 (1. 9) 57. 9 (0 9) 58. 8 (3 6) 62. 4 (l. 0) 63. 6 (1. 3) 64. 9 J ad fall b d r Boundar gl Cl ar a r it For th gla ad vol ani 1 b Road to th dar tuff is a ri s o intracanyon basalt flow whos x 11 ntly expos d. Road highway r v al m fi nd 70 rm i nu M n d b prings Pr xy FalL i h 7. During h hat a m Jl miL d \vn n w rom ri.n t r a r-o -d y d of


71.4 o Y 70. 9 67. 6 Late Pleistocene Terminal E rly High Cas d Ba lt Out rson F rm tion Mount i n ad nd Basaltic From Butt FIGURE 7. GEOLOGIC SKETCH MAP FROM MILE 57.9 TO MILE 72.3 71


( 2 7) 67.6 D r Cr High Ca (2 0) 69. 6 Oll (l. 3) 70. 9 STOP (0.5) 71.4 tur llO\ving oli vin (most i or of any oth r parti ulat the basal 6 fe t of the uni thick with sharp int rfac geneous interior. Additi (1) Right-hand curve; h droelectric powerhouse on opposite Water is co nducted to the facilit through a tunnel rom 2 miles W Highway cuts on the right how ev r 1 sprin aquifer of High Cascade pillow basalt. 72 d,


(0.9) 72. 3 ( 0 .5) 72.8 (2.4) 75. 2 (0 7) 75. 9 (l.l) 77. 0 (l. 5) 78. 5 (l. 7) 80.2 (0.5) 8 0 7 (l ) 8 1 B und r b tw n g olo High\vay fill r st f a th u pp r M n xt mil South m rgin 8 mil E nt r p ur d in a d o ub l s Cr k v 11 y B ginning of a lon g grade up onto th thin Hi g h Cascad basalt flows which ar intracan yon t o Riv r Th flow s a r w 11 xpos d in highway cuts for the basalt flow Lava f rom th bas of Belknap Cr a t er, M c K nzi R i v r can yo n h e r 1500 years ago. The flow liffs t o th \ . a nd spr a d out upon the can yo n s d ownstream During the Pleistocene, prior to ld volcano a nd r lated lava fields, the entire f r o m ic b twe n Mount Washington and Black r a t this p oint. Lava flowed through this valcan yo n many tim s ; t h intracanyon basalt flows last 2 miles w e r e so derived. S r i of high way cuts a nd cliffs t o th righ t xpos many thin lava flows dat d at 1 2 m . Th a r e probably a part of a larg early High Cascade basaltic a nd sit shi l d volcano whose central plug and flanking lavas can be s n i n th opposit wall of M c K nzi Riv r can yon, 1 mile W. Sahali Falls parking lot. Recomm nd d scenic Two thic k flows of basalt i a nd sit lava from th Sand Mountain chain of cinder cones (3 3 miles E of h r ) mov d into M c K nzie Riv r can yo n in this a r a, 3000 years ago. The terminus of th fir s t flow is marked b Koosah Falls, 0 4 miles downstream. The s o nd flow, whos t rminus is b y Sahali Falls, is seen in the highway cut ah a d Road on right to Cl ar Lak Cl ar Lak is 1 5 miles long, fed b large spring and is mor than 120 f et d ep in pla es. It was formed behind a dam of lava whi h issu d f rom a cind r con S of Sand Mountain and poured across th M K nzi Riv r Rising wat r s inundat d a standing forest; several dozen snags a r s till r oo d o n th lake floor. \ood f rom the snags and charcoal from b n ath h ast-sho r lavas a r both 3000 C 14 years old. South margin o Holoc n lava a nd outlet channel f r o m Fish Lake. The lava came from a r l v n of th Sand Mountain chain of cinder cones and dammed Hackleman Cr k t o form Fish Lak 3800 years a o The lava is basaltic and relative! thin with smooth rop surfa tr e molds are ammon. Road t o 73 ba-fl w from ash appar ntl was not th om tim s sensitiv i ndicators The jun n s p a r a t occasions: Crat r and Sand Mountain ) ( 2) ubs qu ntly j un tion), a nd (3) f a h rat r (bl k


8 1 Hi g h Cas a d B a alt 78.5 ilig h Cascad B a saltic Ande sit s 77 o L A V 80 7 Lava Fr m S,nd luntain h in Con s 75. 9 Lav From B lkn p r t r 75. 2 L t Pl is o n Intrac nyon B lt Hi gh Cascade Basalt i And s i s FIGURE 8. GEOLOGIC SKETCH MAP FROM MILE 72.3 TO MILE 83.5 74


( 1 4) 83. 5 (0.6) 84 J (0.6) 84. 7 (0.6) 85.3 (1. 0) 86.3 (1. 5) 87.8 (1. 5) 8 lava front ab ut 200 f t S f th jun tion). Boundar b tw n g sk t h maps, Figur s 8 and 9 nt to th highway on right, is a short lava tub r py rust on its st rn flo r. Sawy r's Cave adjawith a well-pres rved Gradual right-hand urv aero s a lava fi ld compos d of basaltic andesite fl ws (S.E. of th highway) from ash Crat rand ounger flows ( W of the highw y) fr m Littl ash Crat r. Fin ash from Crater fell on both of th flows, but th resulting deposit s thin and to be se n it must be reco-v r d from int rstic s in th lavas. R ad to quarri s in Little ash Crat r, left. R commended side trip, but not uitabl for a arg bus. Follow grav 1 road to top of cone. \ atch for trucks. Littl ash Crat r has be n breached on th \v. by lavas which extend from the bas of th breach d area, westward for 1. 7 miles. The quarried areas are litt red with discard d volcanic bombs and bomb fragments up to 4 feet in diam t r The top of the con affords a god view of the maturely dissected W st rn Ca cad escarpment and High Cas ade Holocene cones and lava fi lds. Follow a grav 1 road W and from th con e to Highway 22 and examine th lar road cut of 22. It r v als a later Pleistocene terminal moraine m ntl d with a thin la r of Mazama ash, overlain by a bed of fine mafic ash from ash Crat r This is ov rlain b coarse rock fragments blasted from bedrock and morain ov r th Little ash Crater vent. The top layer in the cut consists of mafic lapilli from Littl ash Crater. Turn E. on Highway 22. San iam Jun tion; Highways 20 and 22. Continu E on 20 Late Pleistocene lat ral morain s form ridges to th and S of th junction area, outlining th last advanc of gla ial ic Holoc n volcanic activity in the junction rain lud s : (1) arly basalt flows that moved W toward Lava Lake from v n s b ash Crat rand Sand ountain, (2) basalt flows from ash Crater, (3) d v o th Lost Lak chain of con s and associated basalt flows, and (4) and sit flows from th S. and \ bases of ash Crater. During phas of activity at ash Crater, Little ash Crater form d and br a h d by flows of basalti and site. All of these volcanic epis d s d posit d 1 va or ej eta or both in th vicinit of Santiam Junction. Broad 1 ft-hand curv thi part of th chain and th resulting ridg damm Th on f the highway on of Lost Lake chain of cinder cones. In w r constructed across a glacial canyon Cr k to form Lost Lake (one mile ahead). rater 1000 feet wide and 300 feet deep. In highwa uts and on hillsid of th highwa are several outcrops of colum-nar and plat -jointed basaltic and sit lava from a glaciated cone, 1 5 miles Lak Old r rocks in the S. wall of this canyon have been 0 Th 1 v o v rl in b a m ntl of reworked L k nd Li tle ash Crater. advancing 75


L va Fr m l.iltl c n La L r a l Mo r a i n Los L Lak ' n s !log 1 Rock And s i Dome High Cascade Basalts Blue Lake Crat r FIGURE 9. GEOLOGIC SKETCH MAPS FROM MILE 83.5 TO MILE 91.6 AND FROM MILE 91.6 TO MILE 99.8 76


(1. 9) 91. 2 ( l. ) 2 6 ( 2 5 ) 15. L (0 ) 9 5 5 (1 0) 9 5 (l. 3) 97. (2. 0) 99. 8 (1. 5 ) 101.3 ( 2 ) I 0 Summit of Santiam Pass. orth rn margin of a larg glaciated field of basaltic h spr ad E and W. from cones a nd lava v nts along the High Cascade from Highway 20, S to Mount Washington. It should be noted that ind r so pr val nt on road cuts between Little ash Crater and Suttle b n d p sited by snowplows and are not part of the natural strati-Trailh ad to Square L a k e left. From this point to the E. base of the High C s ad platform, basaltic a ndesite lavas of reverse paleomagnetic polarity d in highway cuts v rlook, right. Blu Lake occupies a Holocene crater surr o und e d by rim o cind r s spatter, a nd bombs of basaltic a ndesite, ejected 3500 years Although n o lava appeared, the eruption must have occurred with considerviol n for most of the crater was blasted out of solid bedrock and 1 rg fragm nts were scattered i n all Qjrections. T h e landscape is blanketed with cind r s fo r 3 mil s E and S E ha v th y curv Highway uts in thick, platy-jointed basaltic andesite flows n dat d by K-Ar at 0.54 m.y., but must be at least 0 7 m y because ss rev rs paleomagn tic polarity. Midpoint of long downgrade of Suttle Lake. Glacial drift of a late Pleisto e n lat ral moraine which forms a n E \ ridge 2 miles long and 600 feet high is xpo d in sid highway cuts. Suttle Lake rests in an elongate basin enclos d by lat ral and terminal moraines. Cache Mountain and Mount Washington ar o the S.W. Lak right. ext 0.5 miles of highway will lead thro ugh late t rminal morain s This represents the lowest elevation reached by an ext nsion of th ic sheet which accumulated between Mount -Washing-ton and Thr e Fing red Jack. Th moraines are covered with 0.5 t o 1.0 feet of fin ash from th Sand 1ountain chain of cind e r cones and by approximately 1 foot of younger cinders from Blue Lake Crater. Boundary b t w en g ologic sketch maps, Figures 9 and 10 Hi ghway leads over ir r gular surfac und e rlain by a widespread field of Pleistocene diktytaxitic basalt flows. Jun _____ Highway lQ, turn left o n road to Metolius River a nd Camp ahead will lead clockwise around the base of Black Butte. The t high and 4 miles in diame t r. It is a composite cone of nd 1 t d at th S. of the Green Ridge fault n Turn ri h n . F . RD 1 77 Bl Butte i older than many Thr e Fin gered Ja k a nd Mount dow f the C s ades. It flanks, rocks on its urlav s posse s reverse but th standard errors d lava f i ld of "Ha d f M toliu s


107 0 High Cascad Basalts High Cas ad Cr n Rid Faul of D s hu F nnat ion 109.4 V l n o L v s o D s hu s ).' 6 : Format ion , ::a.::-==::: .. .... I FIGURE 10. GEOLOGIC SKETCH MAP FROM MILE 99.8 TO MILE 119.0 78


(0. 2) 104. 1 ( 1. 9) 106. 0 (l. 0) 107. 0 (0 3) 107.3 (2 1) 109. 4 (4. 3) 113 7 ( 5 3) 119 0 (0 7) 11 7 ( ) 1 2 (3. 0) 1 27.6 S c d, U S F S . RD 1430, a nd follow thi road uphill f r Th Riv r m rg s from lar g springs at th b of Bla k Butt 1 mil his point. Bla k Butt a nd yo un g r Cascad 1 va hann 1 of a n rth-Elowing an tral M tolius Riv r Groundwat r still m v s through thi c h a nn J, m th margin of Black Butt 1 v a Larg b uld r s f Bl k Butt abundant along this road, but no utrop will b s n. L ft-hand urn f rom mitt nt out rops of n or th n xt 1 f Black Butt t o W lop of Green Rid g Int rs parat d by flow breccias will be find vi wpoin n ar top of grad Parts of one of the most signitural f atur s on th E flank of the c ntral Cascad n m ly the fault blo k, an b s n from this point. Th west-facing s arp for 15 mil s and rises 2000 f t above th Metolius River Basalt, basal ic a nd sit a nd and sit lavas with interbedded ashlow a nd airf 11 tuffs ranging in ag from 5 0 to 9 2 m y., o cur in the face of th scarpm nt; th y hav b n trac d E into th corr lative Deschutes Formation. M gmas of Bla k Butt probabl asc nd d through the Green Ridge fault z n and a w 11 pr rv d mafi cind r cone and lava flow of Pleistoc n e age also r sts on th fault trace, 3 0 mil s of h r Several High Cascade p aks can b s n t o th W Most promin nt ar Mount \ ashington (S l ) Three Fing r d Ja k (W. ) and Mount J fferson ( W ) J unction with Allingham Road, S F S o RD 1120. Turn right and follow RD 1120 south. Th hillsid visibl about 0.5 mil s E. is the S end of the main Gr n Ridg s a rpm nt the roadway leads across part of a ramp structure b w n two parall 1 faults. Large altered boulders of Deschutes Formation lava will b s n a ong the roadside for the next 5 miles. Jun tion with road 1 ading E. (RD 1139). Continu south on S F S RD 11 to rejoin Hi ghway 20 Jun tion with Highwa 20 Turn 1 ft toward Sisters. From her to Sisters the highwa cross s a n arly flat alluvial plain hi fly consisting of late Pleisto e n outwash ands and gravels from Hi gh Cas ad glacial s stems. The alluvium r sts up o n earl Pl istoc n lava flows and is overlain by late Pleisto c n cind r cones and a so iated lavas. Boundary b twe n G ologic Sk tch Map Figur s lOand 11 Junction of Highwa s 20 nd 242 ; Town of Sist rs. Pro eed E. through Sist rs. J un i n J un i n n x 2 w h i h pr h Hi hw y 2 0 nd 126. Pro 79 d E n 20 toward B nd. oward B nd on 20 For the High Cas ad basalt porph r in th D schut s Basin. hw 20 Low hills on lti and sit 1 vas n ar 1 v 1 p rt of highwa for 2 2


T r of Tumal 119 7 E rlv High as ad' / lnLr can / Porphvrv D hu s Form ion L 1V on B s, 1 L FIGURE 11. GEOLOGIC SKETCH MAP FROM MILE 119. 0 TO ILE 129.1 80


(1. 5) 129.1 (0. 2 ) 129. 3 (2 .6) 131 9 (1.1) 133 0 (0 5) 133 5 (1. 2 ) 134 7 (1. 3) 136 0 ( 3 3) 139 3 mil s ah ad is und rlain by D s rt Sprin Tuff, o b s ur d by alluvial cov r toward ll W. B und ary f G o J g i Sk t h Maps, Figur s 11 and 12 ully a h fl xpo d pink D s rt Spring Tuff in b oth banks. The this a r a t o w ard Dry Canyon Hill 0 7 miles E is an ind r o n Jun ti n of Cou h M rk t R oa d and H'ghwa y 20 Th low hill E of the highway a m 11 blo k o upp r D schut s Formatio n mafic lava bounded by northwestr nding n ormal faults. Doz ns of s u c h faults o cur E of the Tumalo Fault n Gr n Rid g and B th y are part of the Brothers-Sisters Fault Zone Laidl \.J Butt ctiv aft thi ar a th right. Laidlaw is a Pl istocen cinder cone which was rt Spring Tuff, Bend Pumi c and Tumalo Tuff were deposited in urv and d owngrad to floor of Des chutes River valley. Highway cuts along h grad xpos pink nonw ld d Tumalo Tuff in a position close to the E m a rgin of th Tumalo ash flow. The ash flow moved 3 miles farther E. and n unkn o wn, but probably much great r distanc eastward, where it is now buried b n ath Pl isto n lava fields. Highwa y 20 ross s Des chutes River. Alluvium on the valley floor has been quarri d for g r a v 1 Highway grade up the E valley wall exposes Bend Pumice and Tumalo Tuff (which are underlain b D sert Spring Tuff at water level), overlain by grav 1 b ds. At th top of the grad basaltic lava flows rest upon the grav ls. On of sev ral Highway cuts through basaltic lava which exposes pressure ridges in ross s ction. The lava was moving .W. in this area from late Pleistocene v nts on th flank of ewherry Volcano. The Deschutes River was forced westward to its present course along the west margin of the lava field. The city of Bend, E. of the river, has been built upon this lava field. Junction of Highway 97 and route to city center. Turn right and proceed to downtown Bend. Turn right again at Newport Avenue. Proceed to College Way and completion of field trip circuit at 142.0 miles. 81


Laidlaw Butt Cind r Con Q A L U nd e rlain by High Cascade Ash Flow T uffs eN Early Pleistocene of A w b r ey Butte pri n g Tuf Pl i s O<. n Bas 1ts From ort h Flan k of w b >rr Vol a n o FIGURE 12. GEOLOGIC SKETCH MAP FROM MILE 129.1 TO MILE 142.0 8 2


REFERE CES Armstr n g R. L., Tayl r E 1., Hal s P 0., nd Park r, D. J., 1975, K-Ar dat s for v 1 ni r ks, ntral Ca ad Rang of Or gon : Iso t, o 13, p S-10. Pe k' D L. G olog S G ol. r, H. G W lls, F G and Dol H M 1964, ntral a nd north rn parts of th W t rn Cascade Rang in Oregon: Pr f. Pap r 449, 56 p Cr h, E A Taylor, E. and St n land, D. E 1976, C ology and Pet rson, min r 1 r sour of D hut C unty, Or gon : Or D pt. Geol. Min. Ind. Bull. 89, 66 P Scott, \ E 1977, Qu t rnary gla iation and vol anism, etolius River area, Oregon: G 1 So Am. Bull. v 88, p. 133-134. Sutt r, J. F 1978, K-Ar ag s of C noz i v lcani ro ks from the Oregon Cascades west of 121 ': Iso hr n/W st, o 21, p 15-21. Ta lor, E M., 1968, Ro d id g ology, Santiam and McK nzie Pass Highways, Oregon: Ore D pt. G ol. Min. Ind. Bull. v 62 p 3-33. 1978, Fi ld g ology of B:ok n Top Quadrangl Oregon: Ore Dept. G ol. Min Ind., Sp cial Paper 2, SO p Thayer T P., 1937, P trology of lat r T rtiar y and Quaternary rocks of the north-central Cas ad Mountains in Or gon : G ol. So Am. Bull. 48, p 1611 1652 83




NEWBERRY VOLCA 0 ORECO orman s MacLeod, David R Sh rrod, U S G ological Survey, M nlo Park, California 94025 Lawr nee A Chitwood, U S Forest Service, Bend, O regon 97701 and Edwin H McK e U S Geological Survey, Menlo Park, California 94025 GEOLOGIC SUMMARY ewberry Volcano, centered a out 20 miles southeast of Bend, Ore on, is among th s Quaterna r y volcanoes in the conterlllinous States. It covers an area in xcess of 50() mi and lavas from it xtend nor hward many ns of miles beyond the volcano. The highest poin on the volcano Paulina P ak with an elevation of 7 ,984 ft, is about 4 ,000 ft higher than the errain surrounding the volcano. The gently sloping flanks, embellished by more than 400 cinder cones, cons is o f basalt and basaltic andesite flow s anclesitic to rhyolitic ash-flow and air-fall tuffs and other types of pyroclastic deposits, dacite to rhyolite domes and flows, and alluvial sediments produced during periods of rosion of the volcano. A Sewberry s summit is a 4 -to 5 -mile-wide caldera t hat contains sc nic Paul ina and Fast Lakes. The caldera has be n the site of numerous Holocene eruptions, mostly of rhyoli ic composition, tha occurred as r c ntly as ,400 years ago. Many geologists have tudied various a pects of Newberry Volcano wi h I. C Russell (1905) 11ho visited i durin a horseback reconnaissance of central and east rn Oregon in 1903. Howell Williams (1935, 1957) mapped the flanks of the volcano in re connaissanc and studi d the caldera in mor detail. His outstandin work forms the asis for subsequent investi ations, most of which have focused on caldera rocks o r oung flank asal f l ows. o comprehensive study has be n made of the g e o logy of ewberry' s forest-covered and azama ashcovered flanks even though they form more than 95 per cent of th a rea of the volcano. As part of a g e othermal resource investieation of central and eastern Oregon, the first three author s have mapped t h e sixteen 7 -l/2' quadrangles that cover the flanks o f ewberry at a scale of 1 :62,500, and r einterpre ed and par ly remapped the caldera. Highly g neralized geologic sketch maps are shown in igures 1 and 2 The new mapping and K Ar dating by he las r qui r ant ial r in erpr a i o n o f h of f o rma ion of i ewb rr l i 40 m i l Ca cade Rang in a R e Volcano in California vo ) Bo h b con s similA r he si s o rup i o n s obsidian flow s during th las c ivi within h Lak his y ars. e n and 85 ewberry lies at the west end of the High Lava Plains, a terrain formed of Miocene to Quaternary basalt flows and vents punctuated by rhyolitic domes and vent complexes and other s 1967; Greene and others, 1972; Walker and olf, this vol.). The rhyolitic rocks show a well-defined monotonic age progression starting at about 10 m y east of Harney basin and decreasing to less than 1 m y at ewberry s eastern border (Walker, 1974; MacLeod and others, 19 75; McKee and others, 19 76). ewberry rhyolites appear to be a continuation of these ageprogressive rhyolitic rocks. A complexly faulted terrain surrounds e wberry. The northeast-trending Rim fault zone impinges on e wberry' s southern flank but offsets only its older flow s A zone of faults that offsets older flows on the lower northern flank extends northwestward into the Cascade Range at Gre n Ridge. Although e wberry lavas obscure the relations of the Walker Rim and G reen Rid e fault zones, they likel join ben ath e wberry and represent but one curvin fault system. The Brother s fault zone, a major west-northwest-trending zone of faults, extends across the extreme northeastern flank but does not apparentl offset surficial ewberry flow s It probably extends a t depth to join or abut against the Green Rid e -Walker Rim fault zones. The north and south flanks of ewberry Volcano, which extend the g reatest distances f rom the summit calder a are almost exclusively venee red by basalt and basaltic andesite flows and associated vents. The basalt flows form much of the surface in a broad region extending far north of the volcano (Peterson and G roh, 1976) as well as southward to the Fort Rock basin. Individual flows are a few feet to more than 100 ft thick and cover areas of less than 1 square mile t o many tens of square miles. Flow margins are commonly well preserved even on older flows, but the flows are complexely interwoven and it is difficult and time consuming to trac& individual flow boundaries. Most flows are of block or aa type; pahoehoe surfaces occur locally on a few lower flank flow s Lava tubes are common and some ext nd uncollapsed for distances of 1 mile (Greeley, 1971); som lower flank flow s may have been fed by tube systems. Casts of trees occur in many flows, particularly the younger ones. The basalt and basaltic andesite flows can be readil divided into t w o groups on the basis of their a e relative to Mazama ash ( 14c age 6 600,700 years) derived from the volcan o at Crater Lake 70 miles dis ant. The t/,oungest lava flows over lie azama ash, and their 1 C ages range f rom 5, 800 to 6 380 ars c14c ages of this magnitude are en rally about 800 years younger than actual


10 MILES 8 8 E] 'o a lr ( Pl ANAl ION UndlletntJteO DJ\Jit and bd\JIIc anoe''' llo w $ (l avJ Bulle llo w I IJb I d) Jill Pie locene 10 H olotene Rn otne 10 dJ ne llows pumce and tnt compte earl 10 nolo cen Undrlleenuated pumc llow s and Jllural deoosns ol uppe IIJnk PI 1\IOcen ano t iulocen A s h !lo w tulls and aqgtutnat 0 p roc \ 0 1 upp w e 1 llan Pte\toc n l .tpllh lull ol west llan Pters1o ene UndllerenuJied Jllu al wnn nteDedded IJPIh lull (0111 Hh !lo w tull and pumc tall deP0\11\ Pltstocene A h !low lull ol Teepee OrJv. Plestocene flu al and lacustflne P1eslocene and Phocen 1'1 Ba Jlt ba IIIC andesne and llow s a h llo w and pumce tall deposns of 1he Ca caoe Rano Plestocene Bas all llo w s and mleroedded cnd rs and to11a deoosns 1a1 Mocene Rhyotne 10 noesne flow s dome ano pyocta s iC roc s of lhe Pme M ountam area earl Mocen C1nde1 cones and lr sue en1 / Con1ac1 Fault feld lfiP 10u1e 3 f 1eld HIP SlOP 1 o sopacn of oung pumce tall Figure 1 Geologic sketch map of ewberry Volcano. Geolo y of he cal ra is shown in fiqur 2 ages). Carbon for isotopic dating was obtained from carbonized root systems at the bases of lava tree casts (Peterson and Groh, 1969) and from beneath cinder deposits that extend as plumes leeward of cinder vents related to the flows (Chitwood and others, 1977). The young flows may have erupted during a much shorter period of time than the age spread indicates, perhaps as little as a few weeks or several years. All other flows are covered by Mazama ash and are older than 6,700 14c years; surface features on some flows suggest a relatively young age, perhaps 7,000 to 10,000 years, o hers ar likely several tens or hundred of thousands of years old. Seventy-three basaltic to andesitic flow rocks or bombs from associated vents analyzed by Higgins (1973) and Beyer (1973) contain 48 to 60 percent Sio2 The mean silica content of these rocks is about 54 percent, similar to that of analyzed rocks from the adjacent Cascade Range. Of these analyzed Newberry rocks, only about 20 percent contain less than 52 percent Si02 about 60 percent lie between 52 and 56 percent, and the remaining 20 percent have 56 to 60 percent. Thus the dominant analyzed rock type is basaltic andesite similar to many rocks in the Cascade Range. The analyzed rocks are generally biased toward younger rocks, however, and many of 86 the older voluminous flows on the lower flanks are basalt, mostly of high-alumina type (table 1 co 1). The 6 ,000-year-old flows contain 52 5 percent Si02 and are basaltic andesites col. 2). More than 400 cinder cones and have been identified on the f lanka of other volcanoes in the world contain so many are concentrated in three broad zones that join o the upper part of the volcano. The eastern zone is a continuation of the High Lava Plains zon e basaltic vents and parallels the Brothers aul zone; except for cones high o n the east flank, cones in this zone appear relatively old. northwestern zone of vents is collinea r with the zone of faults o n the lowermost flank that extends t o Green Ridge in the Cascade Range, and southwestern zone is collinear with the Walke r R! fault zone. Fissures and alined cinder generally parallel the belts in which they occur. The distribution of the vents, and particularly of alined vents and fissures, suggests that the northwest and southwest zones, and perhaps the faults that they parallel, are part of one b road arcuate zone that curves in the vicinity of Newberry's summit. Some ali ned cinder cone s and fissure vents near the s ummit occur in arcuate zones


cald r a rim a rallel to th P fractures; som occur ring caldera side is downd copped a n d likely lie along along faults whos Most of the cind r cones a r e well pr served oving to their high porosity and cons quent absorption rather than runoff of water. Larger cones a r e as much as 500 f t high, typical cones a r e .OO to 300 ft. Most a r marked by summit c r aters and flows em r f r o m their bases. Cind e r s dispersed by prevailing winds during eruptions form a rons extending le ward from some cones such as Butt e (Chitwood and others, 1977) (stop 1, road log) Fissure vents consist of long ridges or trenchlike depressions formed by cinders, spatter and agglutinate flows Small pit craters are developed along some fissure vents. Cinde r cones and fissure v ents on the lower flanks are generally dev oid of fragments of rhyolitic rock, whereas many of those higher o n the volcano contain rhyolitic inclusion s (stop 3, road log) Shieldlike vents occur at Spring and Green Butte 0 the southwest flank and Green Mountain t o the northwest of ewberry. They are 1 to 3 miles acr oss have gentle slopes, and are more faulted and older than most surficial ewberry flows. any of the h ills on Newberry's flanks are r h olitic domes In add it ion, pumice rings, obsidian flow s and s:nall rhyolite or obsidian p r otrusions occur in many places. Most of the domes form r ounded hills, such as McKay Butte on the west flank that are 100 to 500 ft high and up to 4,000 f t across. The largest dome, which forms Paulina Peak (s t op 6, road log), extends southwestward from the caldera wall for 3 miles. Its very elongate outcrop suggests that it was emplaced along a northeast-trending fissure or fault; an obsidian flo w c r ops out farther down the slope on a direct extension of the axis of the Paulina Peak dome and cay have been erupted from the same buried fissure o r fault. Several small rhyolitic outcrops may be the tops of rhyolite domes that are more extensive at depth An example is along Paulina Creek on the 11est flank of ewberry where obsidian irregularly invades basaltic andesite flows over a small a rea. n addition, the common occurrence of rhyolite as ragments in cinder cones on the upper flanks attests to the relative abundance of rhyolite at depth on the upper part of the volcano. K -Ar ages were determined on six rhyolite domes and flows. The ages range f rom 400,000 to 700,000 years, although many undated rhyolites are probably younger. Some small spinal protrusions, domes, and pumice rio s on the upp r sou h as flank may b less than 10 ,000 years old. In con cas t o he relativ ant iqui of many rhyoli es on h flanks, hose in h cald r a a r e commonly youn er han Mazama ash and as young as 1 400 years. Ash flow s pumice falls, mudflow s and oth r pyroclas i c d posits ac common on th w st and east flanks of e w b cry and ar lik ly present at d pth on he nor h and sou h flanks b low he v neer of basalt and basal ic andesit flow s The oldes t known unit occurs on th lower north ast and east flank. and consists of a widespread rhyolitic ash flow identifi d by G W Walk c (p rs. commun., 1973) during r connaissan c mappin of adjac nt areas to h ast. It has b en r ferr d to informally as h Teep Draw ash flow for ou crops 87 along Te pee D r aw (stop 14 road log). The Teepee Draw ash flow crops out along ravine walls for at least 6 miles toward the caldera befo r e being buried by younger rocks. Along some ravines it exceeds 70 ft in thickness, but it may be considerably thicker because its base is exposed only where kipukas of older basalt project t hrough it; also the upper part is eroded. The ash flow c rops out roughly over a 50 quadrant of the volcano, and, as it is apparently older than the surficial rocks on the other flanks, may be present at depth completely around the volcano. O riginal volume of the unit is difficult to estimate without information on distribution at depth on the other flanks, but it likely is much mor e than 10 cubic miles. The Teepee Draw ash flow probably relates to an early, perhaps the earliest, period of caldera collapse. Farther up the northeast flank the Teepee Draw ash flow is buried to progressively greater depth by alluvial sediments derived by erosion of rocks higher on the slopes. Interbedded in or under lying this alluvium are basalt and andesite flows, and several other ash-flow tuff units (stops 12 and 13, r oad log). Some of these ash flows are widespread, others occur in only a few scattered localities (commonly plastered o n the caldera side of cinde r cones). Most of these post-Teepee Draw ash flows are characterized by dark-colored, probably dacitic, pumice. The west flank of the volcano contains two major tephra units. The older unit forms most of the lower two-thirds of the slope and overlies basalt flows and vent deposits that in many places are deeply eroded. This unit also occurs on the northeast flank where it occurs higher in the section than most of the ash flows. Although locally over 200 ft thick, it is rarely exposed. Most of the scree-covered road cuts along the paved road on the west flank that leads to the caldera are in this unit (stop 2 road log). The unit consists of gray to black ash, lapilli, and small bombs and abundant accidental lithic fragments. The lapilli and bombs have characteristic cauliflowerlike surfaces and virtually all contain angular to r ounded inclusions of rhyolite, dacite, and andesite; some inclusions are fused and f rothed. Trenches dug in the deposit show that it is massive. In no place have w e seen any indication of collapse of pumiceous lapilli or welding. All of the lapilli and bombs, as well as the ashy matrix, have the same nonnal natural remnant magnetization, as measured by a field fluxgate magnetometer, which suggests emplacement temperatures above the Curie point for the entire unit. Most of the unit was likely deposited as pyroclastic flows. I n one area the unit is palagonitized and much more indurated and h s char act ris i s mor e like that of a lahar Lapilli and bombs id ntical to those in this unit a r e an ubiquitous and voluminous component of most alluvial deposits on the volcano. Furthermore, numerous gravel pits beyond the flanks of the volcano contain similar lapilli as a major component of the gravel. The original volume of the unit was probably several cubic miles. Its eruption could have been accompanied by caldera collapse. The second major tephra unit on the west flank forms the smooth and gentl dipping upper part of the flank e xtendin for about 2 miles from the caldera rim. At localities farthest from the rim (stop 4 road log) the unit consists of many thin ash-flow units, common! 3 to 20 ft thick. They are reddish to brownish i n color and consist of


andesitic sc ria and pumjc and ac id nt.ll 1i hie ra m n s in a p o r 1 ort d 1i hi -and cryst 1-r i h ash matrix. Bas s of i nd i vidual u nits are commonly w l d d Toward the cald r a rim th ash fl w s progress ivel chan chara t r a nd i n m a n y place n ar th rim, a s at P a u lina Falls ( t p 5 road lo ) thi k units hav the app aranc of agglutinate flows. Th se d posits probably represent hot co-ignimbrit la dep sits. Alluvial d posits occur ov r broad areas of th northeast, lower southeas and upper south flanks where they form round d slopes with virtually nonexistent exposure. Most of th deposits ar gravel and sand, but the occur rene of boulders as float at some horizons along sides of ravines indicates that boulder beds ar present also; pumic falls and ash flows are interbedded in the alluvial deposits in some areas. B cause most of th uppermost slopes of the volcan o in areas where th alluvium is present are veneered by young pumice falls, we were not able to determin th origin of the deposits near the caldera. They may be fluvial but equally well may be of glacio-fluvial or glacial origin. Farther down the slope they are probably entirely fluvial, representing broad alluvial fans. Much of the uppermost northeast, east, and southern flanks extending for about 1 to 2 miles outward from the caldera rim are formed of pumic and ash deposits derived from vents within the calder a Most of the deposits are younger than Mazama ash, but scattered holes dug through th m show that similar deposits underlie Mazama a h in a few places. Much of the east flank of the volcano is covered by an extensive pumice fall (stop 11, road log) derived from the vent for the Big Obsidian flow in the caldera (Sherrod and MacLeod, 1979) It extends as a plume oriented 80 E., is well over 10 ft thick near the caldera rim, and thins to about 10 in. at a distance of about 40 miles from the caldera. Williams (1935, 1957) first recognized that th 4 -to S -mile-wide depression at the summit of the volcano is a caldera; Russell (1905) had originally suggested that it was a large glacial cirque. Owing to the absence of known ash flows on the flanks, Williams (1957) interpreted the caldera as resulting f r o m "* * drainage of the underlying reservoir either by subterranean migration of magma, or, more likely, by copious eruptions of basalt from flank fissures * *," with summit collapse occurring along ring fractures. Higgins (1973), on the other hand, interpreted caldera formation as due t o tectonic-volcanic collapse along fault zones that supposedly intersetted at the summit. As noted by Peterson and Groh (1976), the main axis of the Brothers fault zone lies far north of the caldera, rather than crossing the summit, and faults within the zone do not appear to offset Newberry lavas. Also, most of the faults shown by Higgins on the upper part of Newberry that he .uses as part of his structural interpretation could not be corroborate d Ash flows and other. tephra units are now known to be common and voluminous on the flanks. Thus the caldera seems much more likely to be the result of v oluminous tephra eruptions from magma chambers below the forme r summit with concomitant collapse of the summit in a manner similar to that of most other calderas the size of ewberry' s or larger. As there were several major tephra eruptions, it seems likel y that collapse occurre d several times, each collapse areas smaller than that o f the present 88 a l d r a l n r p r ted a ly. h 1 n s p r s n Cfll d rn Th g o f c a ld ras and of consid r r La d t o th p orly known. Higgins (1973) a Holoc n of wid spr ad many rocks on th flank ar y ars old, so abs n e is not m aningful. volcano s on th ast whi h ar about th obvious glacial f n arly a million or The ring fractur s along which occurr d ar not expos d yet their g n ral locatio is indicated by arcuat zones (fi.g. 2), visible 0 aerial photographs. I is not possibl to rel a e individual ash-flow units to coil ps along specific ring fractur s Also it is possibl that collapse occurr d along parall 1 ring fractur s at the s a e time. Two par all 1 walls on th southeas side 0 the cald ra present an int resting problem i o interpreting th origin of th caldera '-'al sequences (describ d lat r). If th inn r wall i s the young r, th n the area b tw en the inner and outer wall may b f orm d of old caldera-fil deposits. Th inn r wall is thickly mantled young tephra d posits, and th only exposur s are a relatively small ar a of rhyolit in the lower p a r of the slop The remaind r of th slop on this wall dips 30 t o 40, n ar he angle of r pose, and has a very uniform smooth shape. If the youn tephra deposits were locally und rlain by indurated rocks, as are present in th other walls, it see.cs likely that th slope would b mor irr gular. Thus this caldera wall likely is dominantly or entirely form d of relatively unconsolidated fragments rocks, probably caldera-fill d e posits. Th rhyolit e exposed locally at the base of the w all m a y b e a faulted dome that was originally within the caldera. The upper part of the northeast above exposures of cald ra wall rocks may also be formed of caldera-fill depo its, the older c alder a wall bein farther north. The walls of the caldera are mostly cov e red by younger deposits (talus, pumic e falls, etc ) and the wall rocks are only locally expo s ed. The y are described by Williams (1935) and in mor e detail by Higgins (Higgins and Waters, 1968; Higgins 1 973). The most continuous exposure s ar a long h e n o r t h wall extending from the Red Slid north of P aulin a Lake, to the northeast obsidian flow, whic h d rape s the wall north a s t o f I.a Th bas of he wall sequence is f o rmed o f platy rhyolite Higher in the section are basaltic andesite flo palagonite tuff, cinder and agglutinate d spatter deposits, and at the t o p of the exposures are palagonite tuffs. Exposure s in the east wall are f ound only fr a b out midway o n East Lake northward t o the Sheeps Rump, a cinde r cone along the wall at the n ortheas corner o f the cald e r a The internittent exposure c onsist of basaltic andesite or andesite palagonite tuff, a n d pumice deposits. The last range from unw elded a nd uncollapsed pumice to densely welde d deposits


0 t:=-' 2 M ILES l'umottl IEJ Rh, It .1 an .lome 1n @"u Una 1!er p m ,. dOd d h ar o : PaiJ onott ull Condt cone' an sure n depo ,,, Flu oat ana latus 11nt aepo s B sal a allot anaesote and andes e loO>S llota:l buroed b pumoce and ash Ou) Caldera w all roc s west wall roc sho n s eparately Ool) Ash llo w ulls and agglutonaled pyroclas oc roc s ol upper wes1 llan 4 Foeld !lop sop ---foeld :11p rou1e Rom ol caldera Probabl e calder a ron lrac1ur e = Axos ol oun9 pumoce Ia I a
PAGE 100

Table 1. Repres ntativ ch mical a nalys s of N wb rry rocks. [n r not reported) Lava Butt Flank basalticbasalt andesite Paulina Falls Paulina andesiticP ak Si02 ------49.1 56. 0 60.43 71.07 Al 2o3 ----17.6 16 1 16 06 14.92 FeO+Fe 2o3 9 0 7 7 6 .51 2.81 MgO -------8 9 4 49 l. 71 .22 CaO -------10.0 8 2 4 58 1. 08 Na2o ------2.6 3 72 5 71 6 .04 K 2 0 -------.43 l. 25 1. 54 3.03 H 2 0+ .71 n.r. .53 17 H 20- 10 n.r. 07 .01 Ti02 1.0 1 13 1.30 .29 P205 31 n .r. .59 .05 MnO .16 .14 15 09 co2 < .05 n .r. .61 < .05 Column 1. Higgins (1973, table 6, col. 62) Column 2. Beyer (1973, table lc, LB-4) Column 3. Higgins ( 1973, table 4, col. 19). Column 4 Higgins (1973, table 4, col. 27) Big Obsidian flow 72.02 14.61 2 .43 164 85 5 .16 3 89 n r n.r. .24 n.r 064 n r Column 5 Laidley and McKay (1971), average of 66 analyses. 90

PAGE 101

Paulina Lakes, and th Interlak and Game Hu bsidian flow that c rop ou n r h and south of i are only sligh 1 youn r than Mazama ash ( 14c ag 6 006 ,700 years) Th East Lake obsidian flow are appar nt l ab u 3 00 y ars old, and the Bi Obsidian flow is about 1,400 y ars old. Widespr ad pumic ous t phra deposits cov r much of the east rn par of th caldera. They underlie the East Lake obsidian flows and probably are slightly older than the Cen ral Pumice cone and th obsidian flows on its n o r h and south sides. Th se tephra deposits may hav b n d rived from several different vents, but hol s dug through th m indicate that they cons is dominantly of 5 to 10 f t of oassive to poorly bedded pumic with accidental fragments (palagonit tuff, basalt, rhyolit etc. ) overlain by s veral fe t of w 11-bedded mud-armored puoice, accr tionary lapilli, ash and pumice (stop 9, road lo ) The oungest period of volcanism within the caldera was associat d with the vent for he Big Obsidian flow It began with ruptions that produced a wid spr ad pur:lice fall that cover s the southern part of the caldera and he astern flank of the volcano (Sherrod and MacLeod, 1979). 14c ages of 1, 72()_ (Higgins, 1969) and 1,550+120 (S \.' Ro inson, written commu n., 1978) years were obtained on carbon directly beneath the fall. On the basis of thickness measurements from 150 holes dug through the pumice fall, the axis of the fall extends 80 E a way from the vent for the Big Obsidian flow. At about 5-1/2 miles from the vent the fall is 12 ft thick, and at 40 miles it decreases to about 10 in. The pumice fall was followed by eruptions that produced an ash flow that extends over a broad area between the Big Obsidian flow and Paulina Lake. It is well exposed in roadcuts n ar the Big Obsidian flow (stop 7 road log). 14c a es of the ash flow are 1 ,270+60 and 1,390_ years (Pierson and others, 1966; Meyer Reubin, in F riedman, 1977) with an older age (2,054_ years) obtained many years ago by Libby (1952). Th final event was th eruption of the Big Obsidian flow and th domal protrusion that marks its vent. Slight collapse occurred over a one-half-mile-wide area around the vent before the flow was erupted. The flow ex tends northward f rom near the outer caldera wall to near he paved road in the caldera and, in its northern part, partly filled an older pumice ring (stop 8 r oad log). Considering the long time o ver which e ruptions took place on ewberr y the volcano should be considered dormant but capable of future eruptions eve n though about 1, 300-1,400 years hav e transpired since th last erup ions. ewberry is ideally suit d for thos who wish o div rse volcanic fea ur s I s rocks rang wid ly in composition, and exampl f rom it could be us d to illustrate a nearly compl a las of h yp s of produc s of volcanism. LIST OF REFERE CES Beyer, R L 1973, a rna di f rentiation a w b rry c r a t r in c ntral Or on: Eu en niversi y of Oregon, Ph. D th sis, 84 p Chitwood, L A., Jenson, R. A., and Groh, E 1977, Th of ava Butte: Th Or Bin, v A.' 39' no 10, p 157-164. Fri dman Irving, 19 77' Hydration datin of volcanism at wberr Crat r, Or s 91 Geological Surv y Journal of Research, v 5 3 p. 337-342 no. Gr le Ronald, 1971, G olo y of selected lava tubes in the Bend area, O regon: O regon Department of Geolo y and Mineral Industries Bulletin 71 47 P G r e ne, R C., Walker, G \.1. a nd Corcoran, R. E., 1972, Geologic map of the Burns quadrangle, Oregon: U S Geological Survey Miscellaneous G olo ic Investigations Map I -680, scale 1 :250,000 Higgins, W., 1969, Airfall ash and pumice lapilli deposits from Central Pumice cone, ewberry caldera, Or gon, Geological Survey Research 1969: U S Geological Survey Professional Paper 650-D, p D26-D32. ____ 1973, Petrology of ewberry volcano, central O regon: Geological Society of America Bulletin, v. 84 p 455-488. Higgins, M w., and Waters, A c 1968, ewberr y caldera field trip, Andesite Conference Guidebook; Oregon Department of Geology and Mineral Industries Bulletin, p 59-77. Laidley, R A., and McKay, D s., 1971, Geochemical examina ion of obsidians f rom ewberry caldera, Ore on: Contributions to Mineralogy and Petrology, v 30, p 336-342. Libby, w F., 1952, Chicago radiocarbon dates III: Scienc v 161, p 6 73 -681. acLeod, S Walker, G w., and McKee, E H., 1975, Geotherma l significance of eastward incr ase in age of upper Cenozoic rhyolitic domes in southeast Oregon: Second United ations Symposium on the Development and Use of Geothermal Resources Proceedings, v. 1 p 465-4 74 McKee E H acLeod, S., and Walker, G W., 1976, Potassium-argon ages of Late Cenozoic silicic volcanic rocks, southeast Oregon: Isochron/West, no. 15, p 37 -41. Peterson, V., and Groh, E. A., 1969, The ages of some Holocene volcanic eruptions in the ewberry volcano area, Ore on: The Ore Bin, v 31, P 73 -87. Geology and mineral resources of Deschutes County, Oregon: O regon Department of Geology and Mineral Industries Bulletin 89, 66 p Pierson, F J., Jr., Davis, E M and Tamers, M. A., 1966, University of Te xas radiocarbon dates I : Radiocarbon, v. 8, p 453-466. Russell, I c., 1905, Preliminary report on the geology and water resources of central Oregon: s Geological Survey Bulletin 252, 138 p. Sherrod, D R., and MacLeod, N S., 1979, The last eruptions at Newberry volcano, central Oregon [abs. ] : Geological Society of America, Abstracts with Programs, v. 11, no. 3, p 127. Walk r, G w., 1 9 74 So m e implications of Late Cenozoic volcanism t o geothermal potential in the High Lava Plains of south-central Oregon: The Ore Bin, v 36, n o 7, p 109-119. Walker, G w., Peterson, V., and Greene, R C., 1967, Reconnaissance geologic map of the east half of the Crescent quadrangle, Lake, Deschutes, and Crook Counties, Oregon: u s. Geological Survey Miscellaneous Geologic Investigations Map I-493, scale 1 :250,000. Williams, Howell, 1935, ewberry volcano of central Oregon: Geological Society of America Bulletin, v 46, p 253-304. 1957, A geologic map of the Bend quadrangle, and a reconnaissance geologic map of the c ntral portion of the High Cascade Mountains: Ore on Depar ment of Geology and Mineral Industries scales 1:125,000 and 1:250,000.

PAGE 102


PAGE 103

... RO DLOG FOR NEWBERRY VOLCANO, OREGON Norman S. MacLeod, David R. Sherrod U.S. Geological Survey, Menlo Park, California 94025 Lawrence A. Chitwood, U .S. Forest Service, Bend, Oregon 97701 and Edwin H. McKee, U.S. Geological Survey, Menlo Park, California 94025 The route of this field trip (figs. 1 2) leads southward from Bend across the basalt and basaltic-andesite flows of Newberry's northwest flank and eastward up the wesL flank where pyroclastic units are visited. Several stops in the caldera, including an overview stop provided by Paulina Peak, show the diversity of caldera rock units. The trip continues down the east and northeast flanks with stops to see several ash-flow tuffs and returns to Bend across the apron of basalt flows that extend far north of ewberry. Total mileage of the trip is about 118 miles, all on paved or improved gravel roads that are readily traversed by passenger cars or buses. Mileages are approximate, all turns and stops are underlined, and old and new (in parenthesis) U.S. Forest Service road numbers are shown where appropriate. The 14 c ages of young basalt flows are mostly from Peterson and Groh (1969). Miles o.o (0. 7) 0 7 (3.2) 3.9 (2 .8) 6 7 (3.3) 10 0 (0. 7) 10 7 Junction u.s. Highways 97 and 20 in Bend, Oregon. south on U S Highway 97. Elevation 3, 630. Head Diktytaxi tic olivine-phy ric high-alumina basalts exposed in rail road underpass and intermittently for the next 11 miles to south are typical of Newberry's lower flank flows. Northwest-trending faults that extend in a broad zone through the Bend area offset the flows. A typical analysis of a lower flank flow is shown in table 1, column 1 Road 1821 ( 18) leads eastward to interesting lava tubes (Boyd Cave, 9 miles; Skeleton Cave, 12 miles; Wind Cave, 13 miles; and Arnold Ice Cave, 13 miles) Ash-flow tuff exposed on east side of U.S. 97 near draw. Ash flow is widespread around the Bend area and was derived from the Cascade Range west of Bend. Outcrops and float near rail road west of road sho w a wide range i n p mice composition. Roadcuts expose cind r plume that extends northeast from La a Butte. Charcoal from base of cinders her e yielded 14c age of 6,160_ years (Chitwood and others, 1977). orthwest-trending faul exp sed on east side of high ay offsets diktytaxitic basal ote youn "Gas Line" basaltic-andesite flow south of fault (1 Cage to 6,160_ years, (S. w. Robinson, written commun 1978). 93

PAGE 104

( 0 .5) 11.2 ( 1. 8) 13.0 (1.2) 16.0 (2.4) 18.4 (2.4) 20.8 (3.2) 24.0 (0.6) 24.6 ( 1. 2) 25.8 (1.1) 26.9 Turn w e t t o Lava La nd s a nd dri v by and a ro s th Lava Butt ot that the lava f l o w m r g s from th Lav a Butt sou h sid low t o top of o f th butte. STOP 1 at top of Lava Butt (elevation 4, 970 ft). The bas ltic andesit flow derived from Lava Butt xtends northward more than 5 mil s and w e s ward 3 miles to th Deschut s River. A chemical analysis of th flow i s s hown in table 1, column 2 It is on of many basalti andesi flows on Newb rry that have 14c ages of about 6,100 years. S v ral, including th "Gas Line" and Mokst Butte flows, can be seen from this stop. No e also h many cinder cones on the north and northwest flank, an area formed mostly of basalt and basaltic-andesite flows and associated vents, most of which are youn r than the pyroclastic units that will be visited on the east and west flanks. A panorama of the Cascade Range, particularly of the Bachelor Butte, Brok n Top, and Three Sisters area, can be viewed from here. South of Lava Butte Pleistoc ne (and Pliocene?) lacustrine and fluvial sediments of the La Pine basin lie between Newberry Volcano and the Cascade Range. Return t o Highway 97. Turn south on U S Highway 97 Road on east leads to Lava River Cave, 1/4 mile from highway. The mile-long lava tube is now a state park and is open to the public (lanterns can be rented at the entrance). Sun River Junction. Road on east goes by Camp Abbot Cinder Pit (3/4 mile) t o Lava Cast Forest (9-1/2 miles). The cinder pit corzains brilliant blue, green, gold, and red cinders. Lava Cast Forest flow ( C ages, 6, 380_ and 6,150_ years) contains numerous well-developed tree casts. A mile-long paved trail leads past many casts. Other young flow s accessible by this road include Forest Road (5, 960 years), Mokst Butte, and Lava Cascade (5,800 years). Fall River Junction. 5,870 years). Road on east leads to Sugarpine Butte flow ( 14c age, Cross old logging railroad grade that went to the now abandoned town of Shevlin. Many hundreds of miles of roads on Newberry are on or along old railroads developed about 50 years ago when logging in the area was at a peak. Approximate boundary of Newberry lavas and sediments of the La Pine basin. Gravel in this area is mostly composed of black lapilli with rhyolitic inclusions like those that form the lapilli tuff at stop 2 Thes e can be seen in this area in several pits, including the garbage dump (Southwest Landfill), east of the highway. La Pine State Recreation Area junction. To west 5 miles are exc e llent exposures of fluvial and lacustrine sediments, including diatomite, along the scenic Deschutes River. Turn east 2129 (21). (elevation 4,205 ft) on Paulina Lake Road U.S. Forest Service R o ad Road crosses Mazama ash-covered alluvial sediments. 94 .......

PAGE 105

(2. 8) 29.7 (0. 2) 29.9 (0. 5) 30 4 (3. 2) 33.6 (2. 9) 36.5 (0. 4) 36.9 (0. 6) 37.5 (0 7) 38. 2 (0 .1) 38. 3 (0 .1) 38. 4 Cross Paulina Creek. Fed by Paulina Lake, it is the only stream on Newberry. Outcrops of Newberry lava flows near road. Farther along the road these flows are locally buried by lapilli tuffs like those at stop 2 as well as by gravel. At this site on September 19, 1979, The Cascade Rattlesnake Award was presented to J. P. Lockwood before a cheering audience of 80 geologists. Road 2045 provides access to three rhyolite domes at and near McKay Butte (KAr age, 0.6.1 m.y., may be reversely polarized). STOP 2 (elevation 5, 420) at large scree-covered slope on north side of road. The lapilli tuff at this stop is one of the most widespread tephra units on Newberry's flanks. The unit is rarely exposed but the lapilli in it are distinctive so that float from the unit can be recognized. On the west flank the lap ill i tuff covers an area of about 30 square miles, with basalt flows cropping out locally beneath it; higher on the west flank it is overlain by ash-flow tuffs like those at stop 4. The lapilli tuff on the northeast flank occurs higher in the section than the ash-flow tuffs to be seen at stops 12, 13, and 14. It is also a major constituent of gravel deposits on all sides of the volcano. The lapilli tuff is deeply eroded but locally exceeds 200 ft thick, and original volume was probably 5 to 10 cubic miles. It consists of dark gray to black lapilli, and less common blocks and bombs, in an ashy lithic-rich matrix. Sorting is very poor and in observed outcrops the unit is massive. Most lapilli are finely vesicular to frothy, have cauliflowerlike surfaces, and contain small, generally angular, rhyolite, and more mafic inclusions. The normal polarization of the lapilli and matrix suggests that they were above the Curie point temperature when deposited. The poor sorting and massiveness of the unit suggest that it was a hot pyroclastic flow or flows. The large volume of the unit permits speculation that its eruption may have been accompanied by caldera collapse. Continue upslope on road. Upper snowmobile parking lot. of here. Rhyolite domes or flows occur north and south Contact of lapilli tuff and cinder cones. These vents probably fed flows exposed farther down the road that under lie the lapilli tuff. Cinder cones are buried on their east (caldera) sides by ash-flow tuffs like at stop 4 Float of andesitic ash-flow tuff along road. Turn south to U .S. Forest Service pit F-17, also known as "Mixture Butte." STOP 3 (elevation 5 960 ft) at "Mixture Butte," a cinder cone with rhyolite and pumice inclusions. The pit is on the north side of a horseshoe-shaped cone that is buried by ash-flow tuffs on its east (caldera) side. Rhyolite and pumice, as well as basal and andesite, occur as inclusions in the cinders and bombs and as accidental fragments in the deposits. Some inclusions have been fused and have flowed. The pumice, in blocks to more than 2 ft wide, commonly shows bands of different color and probably composition. A mile-long rhyolite dome crops out about 1 mile to the southwest, contains phenocrysts similar to those in the pumice and rhyolite inclusions and may extend at depth 95

PAGE 106

(0.1) 38. 5 (0.6) 39.1 (0.9) 40. 0 (0.2) 40.2 (1.1) 41.3 (1. 0) 42. 3 (0.1) 42. 4 (0.1) 42.5 (0.3) 42.8 b neath th cind r con P ssi ly th pumic is d riv d from a buri d pumice rin associated with th dome. Many cind r con s a nd flow s on th upp r flank of the vol ano cont in rhyolit inclusions. Rhyolite domes and flow s crop out in many plac s and may b much more xtensiv at depth. R turn to Paulina Lake Road. Turn east on Paulina Lake Road. Area b twe n here and stop 4 is almost entirely underlain by andesitic ash-flow tuffs; found as large slabs of float and less commonly as outcrop. Turn south on Road 2131 (2121) (elevation 6 ,020 ft). West of road is eroded cinder cone with associated flow s that are partly buried by ash-flow tuff. STOP 4 (elevation 5950). Park at road intersection and walk 500 ft eastward along small side road to exposures of ash-flow tuff along side of ravine. Andesitic ash-flow tuffs crop out over a 5-square-mile area of the upper west flank of the volcano and grade near the rim to pyroclastic deposits that have characteristics more nearly like agglutinate deposits, as at stop 5 The unit consists of ash flows as little as 1 ft thick to as much as 30 ft. The ash flows are pumiceous, rich in lithic fragments and crystals, and reddish brown. Even thin units commonly show welding. At many localities collapsed pumices are brilliant blue and green. The ash flows are very near vent deposits, perhaps from a ring fracture near the west side of the caldera, and have relatively small volume. Ash-flow tuffs similar to these occur at scattered localities on the upper east flank. Return to Paulina Lake Road. Turn east on Paulina Lake Turn north into Paulina Falls parking lot. STOP 5 Paulina Falls (elevation 6,240 ft). Paulina Creek drains Paulina Lake, less than 1/2 mile to the east. The indurated rocks that form the cliffs of the falls are agglutinated andesitic pyroclastic deposits formed of many thin to thick units. Similar rocks occur along the caldera wall for about 1-1/2 miles north of the creek. They grade down the creek into ash flows like those at stop 4 and represent very near vent deposits, perhaps from a ring fracture bordering the west side of the caldera. Most of the deposit is probably co-ignimbrite lag, many units being entirely fall, others stubby agglutinated pyroclastic flows. A chemical analysis for this locality is shown in table 1 column 3. The less indurated rocks below the cliffs are formed of poorly sorted and rudely bedded ash, lapilli, and blocks with abundant accidental ithic fragments. Some beds low in the section contain accretionary and mudarmored lapilli. The contact betwee n the lower and upper unit is gradational. Probably the eruptions first deposited relatively cold material, perhaps from phreatic eruptions, and temperatures increased to the point where all fragments were agglutinated at time of deposit ion. Lap illi tuffs like those at stop 2 occur at creek level only a few hundred feet downstream from the falls. Return to Paulina Lake Road. Continue east on Paulina Lake Road. Turn south on road 2160 (500) to Paulina Peak. 96 ......

PAGE 107

(1. 3) 44. 1 (0.6) 44.7 ( 1.1) 45.8 ( 1. 3) 47.1 Buried contact between upper west flank pyroclastic flows and Paulina Peak rhyolite d ome. View northward of gently sloping upper west flank formed mostly of ash-flow tuffs like at stop 4 Note rhyolite domes at and near McKay Butte ( 60 w., 5-1/2 miles). View of numerous cinder cones on south flank and long narrow Surveyor flow c14c age 6080 years). STOP 6 Paulina Peak (elevation 7,984 ft). Before looking at the caldera, walk to peak for view southward. Note fault-bounded Walker Mountain (S. 42 w., 35 miles). Faults of the Walker Rim zone extend to Newberry's lower south flank. Some of the older rocks, such as the reversely polarized dacite flows on Indian-Amota Butte (S. 30 E., 11 miles), are offset several hundred feet by northeast-trending faults; faults that cut the nearby basalt flows have vertical offsets of less than 100 ft and the faults disappear upslope below younger flows. Other distant features include: Bald Mountain (S. 12 w., 28 miles), a rhyolite dome on the south side of a poorly preserved 4-or 5-million-year-old caldera with widespread ash-flow tuffs that crop out from Fort Rock Valley east of it, to Walker Mountain. (2) Cascade Range. From south to north on a clear day you can see Mount Shasta, Mount Scott, Crater Lake, Mount Thielson, Diamond Peak, The Three Sisters, and nearby Bachelor Butte and Broken Top, Mount Washington, Three Fingered Jack, Mount Jefferson, Mount Hood, and Mount Adams. (3) Fort Rock Valley southeast of Newberry with numerous palagonite tuff rings, cones, and maars. Included in these are Fort Rock in the middle of the valley, and beyond it about 39 miles from here is Table Rock, the site of a field trip by Heiken and Fisher (this vol.). ( 4) To the east are China Hat (N. 85 E., 11 miles) and East Butte next to it. They are 0.8and 0.9-m.y.-old rhyolite domes on the west end of belt of age transgressive rhyolite vents. Note the 5-m.y.-old rhyolites of the Glass Buttes dome complex on the skyline (S. 80 E., 60 miles). The High Lava Plains, a broad zone of Miocene to Quaternary basalts in addition to rhyolite domes and flows, extends from Newberry beyond Glass Butte. Now walk to the north along and beyond the fence bordering the cliffs of Paulina Peak to a viewpoint about 50 ft below the elevation of the parking lot. Please use caution in this are a--t he cliffs are 500 ft high. The generallized geologic map of the caldera (fig. 2) should be referred to for identifying features. From this location you can see the inner and outer (near pumice flats) walls of the south side of the caldera. The caldera wall rocks have been described by Higgins and Waters (1968) and Higgins (1973). The south wall near the Big Obsidian flow consists of platy rhyolite, basaltic andesite flows, scoria and cinder deposits, and an overlying obsidian flow. At the west end of the wall exposures is a thick sequence of near-vent pumice and ash deposits that have been fused near their contact with the obsidian flow. The north and east caldera walls contain rhyolite and basaltic andesite flows and pyroclastic rocks, as well as palagonite tuff; the east wall also contains n ear vent welded rhyolitic pumice deposits. Basaltic vents and flows occur on the north wall near Paulina Lake (Red Slide) and East Lake (East Lake fissure, Sheeps Rump), and a long fissure vent occurs near the top of the east wall. All of these vents and associated flows are pre-Mazama ash in age, 97

PAGE 108

(4.3) 51. (0 .1) 51.5 (0.3) 51.8 (0.5) 52.3 (0.3) 52.6 (0.1) 52.7 (0.2) 52.9 exc pt th E st L k Fi sur It is th uth rnm rift zone s quen of b s lt flow and v nts, ab contains abundant inclu ions of obsidian, rhyol it 1970). t v n t of t h n r h we 8 t 6,100 y ars old, and (Higgin and Wat rs, Youn rhyolitic flows, pumi on s, rin and other t phra d posits are widespread around East Lake. llydr tion-rind dating by Fri dman 0977) indicates that the East Lak obsidian flows ar bout 3,500 y ars old and that many of the other obsidian flows and pumi con s (Interlak flow, Gam hut flow, Central Pumi e cone) ar 4,500 to 6, 700 years old. The youngest period of volcani m was associat d with the v nt for th Big Obsidian flow. Eruptions b gan with a widespread pumice fall that now covers much of ewberry's east flank and which will be seen lat r a\4stop 11. Higgins (1969) obtained carbon from ben ath th fall that has a C age of 1, 720+250 years, and we found carbon at another locality with an age of 1,550j)20 years (S. w Robinson, writt n commun., 1978). Isopachs of the pumice fall clearly show that it was erupt d from at or very n ar the v nt for the Big Obsidian flow (Sherrod and MacL od, 1979). Later eruptions produced the Paulina Lake ash flow (stop 7) 1, 300 to 1, 400 y ars ago. Th final ev nt was the eruption of the Big Obsidian flow and the domal protrusion that marks its vent. Slight collapse occurred over a 1/2-mile-wide area around the vent before the flow was erupted. Paulina Peak dome extends about 3 miles southwest down the flank and is about 1 mile wide. It is marked by large rills parallel to its axis that formed during expansion of the surface of the dome much as era ks form on fr nch bread. The age of the dome is not yet known, but an obsidian flow that occurs on axis with the dome farther down the flank is 0 4 m y (E H McK e unpub. qata, 1979), and Paulina Peak dome may be similar in age. An analysis of rhyolite from the Paulina Peak dome is shown in table 1 Return t o Paulina Lake Road. Turn east on Paulina Lake Road. Paulina Lake campground provides a convenient lunch stop. Roadcuts expose Mazama ash covered by a few inches of fine ash from Newberry eruptions. Holes dug through the Mazama ash show that it overlies sand and gravel deposits. Road traverses south side of two rhyolite domes, over lain by Mazama ash and bordered on south by landslide deposits. Roadcut exposes rudely bedded ash and pumice deposits, perhaps the remnant of a pumice ring associated with the rhyolite domes. Westernlimit of Paulina Lake ash flow. STOP 7 Paulina Lake ash flow. Concrete boxes on south side of road preserve collecting site for carbon in ash flow. Originally dated by Libby ( 1952) as 2,054 14c years, subsequent dates indicate a younger age years, Pierson and others, 1966; and 1, 390, Meyer Reub in, l:.!!_ F riedman, 1977) The ash flow extends from the Big Obsidian flow to the shore of Paulina Lake. Ridges and furrows on the ash flow are apparently primary features, and 98

PAGE 109

(0. 4) 53 3 (0. 4) 53 7 (0 .1) 53 8 (0 .1) 53 9 (0 4) 54. 3 (0 5) 54 8 (0 5) 55.3 (0.3) 55 6 their orientation suggests that the vent is located beneath the southern part of the Big Obsidian flow (Higgins, 1 973). Although pumice is generally abundant, the ash flow in many places is composed almost entirely of ash. The absence of collapse and welding indicates the deposit had only a moderate temperature when emplaced, but sufficient for oxidation to give the deposit a slightly pinkish color. Continue eastwar d on road over ash flow. Road on north leads along shoreline of lak e deposits. to Little Crater Campground and interesting e xposures palagonite tuff and silicified pumiceous, fossiliferous, Turn Right to parking lot next to Big Obsidian flow. STOP 8 (elevation 6,370 ft) Big Obsidian flow. Before walking to the flow, note that the pinkish Paulina Lake ash flow overlies weathered pumice deposits of a pumice ring on the east side of the parking lot. Lost Lake pumice ring is partly covered by the Big Obsidian flow, but its northern part can be seen from the trail ascending the obsidian flow. The trail provides interestin g exposures of flow-banded obsidian, pumiceous obsidian, brown streaky obsidian that was formerly pumiceous before it collapsed, and various features indicating the flow behaved in both plastic and brittle manner during its emplacement. Laidley and McKay (1971) did extensive analytical work on the Big Obsidian flow a nd showed that is is essentially uniform in compo sition. Their average of 66 chemical analyses is shown in table 1 column 5 Return to parkin g lot a nd to main road; note exposures of palagoni te tuff of Little Crater north of road. Continue eastward on paved road. Road on south leads t o drill site of U S Geological Survey core hole. Note Central Pumice cone to north. Game Hut obsidian flow is exposed on north side of road. STOP 9 (elevation 6, 460 ft). Pumiceous tephra deposits of the East Lake area. Roadcuts expose massive t o rudely bedded pumice and ash with large accidental blocks overlain by mud-armored pumice, accretionary lapilli, pumice, and ash beds. Bomb sags occur in the upper bedded part and indicate that the beds were probably wet and cohesive. A hole dug vertically through the deposits indicates that they are about 14 ft thick and overlie Mazama ash, which rests on obsidian. Tephra deposits, similar to the exposures here, occur over most of the eastern part of the caldera except where buried by youn r units (East La k obsidian flow s etc. ) The vent or vents for the deposit ar not known, and conceivably may lie in East Lake. The obsidian that underlies the deposit is part of an extensive obsidian flow mostly on th basis of topographic expression, probably extends to the south (inner) wall of the caldera, its vent apparently being an obsidian dome near the wall. Con inue eastward along road. Turn left to East Lake Campground and walk eastward along shore. 99

PAGE 110

(0.1) 55. 7 (0 .1) 55.8 (0.6) 56.4 (0.4) 56.8 (0.3) 57.1 (0.4) 57.5 (0.2) 57.7 ( 1.1) 58. 8 (0.3) 59. 1 STOP 10 (elevation 6, 382 ft). P 1 onite tuff f East L k Th p lagonite tuff is overlain by Mazama ash and by pumic ous t phra d posits Jik those at stop 9. Many of the beds in th palagonite tuff ar formed of mud-armored lapilli and accretionary lapilli, presumably indi ating that th xpos d part of the unit was deposited above water. ote lso several bomb sags. About halfway along the shoreline exposures is a contact b tw n two palagonite tuff units; the western unit is youn er and locally is plastered on the eastern unit. The eastern palagonite tuff contains very abundant and large accidental blocks of palagonite tuff that are mor thoroughly palagonitized than is the tuff they occur in. From the shore you can look westward to see the Central Pumice cone, northward at the East Lake fissure, and northeastward at the Sheeps Rump and the obsidian flow that extends to the northeastern shoreline. Return to the main road. Turn east on paved road. Hot springs occur along shore near headland formed of palagonite tuff. Temperatures as high as 80C have been measured, but the hot spring water is diluted by lake water. Turn east on gravel road, U.S. Forest Service Road 2129 (21). Not exposures of andesite near road intersection. This flow has been traced southeastward to the fissure that occurs at the top of the east wall of the caldera. This fissure contains rocks of highly variable composition and phenocryst content, as does the flow. Cinders, bombs, and spatter associated with the fissure can be seen from several places along the road. Pumiceous tephra deposits exposed in roadcuts are similar to those seen at stop 9. Road here crosses over andesite flow, but it is buried by Mazama ash and pumiceous tephra like at stop 9. Pumice fall derived from vent for Big Obsidian flow forms a thin surficial unit that thickens as we progress up the road toward the south. Top of east rim of caldera (elevation 7,008 ft) covered by pumice fall about 8 ft thick. STOP 11 (elevation 7, 010). Pumice fall derived from vent for Big Obsidian flow. The fall is exposed in a large hole dug in roadcut on north side of road. The fall is about 9-1/2 f t thick here, but the axis of the fall is about 1/2 mile to the south. 14c ages of the fall are 1, 720_ (Higgins, 1969) and 1,550_ (S. W. Robinson written commun., 1978) years. One hundred and fifty holes dug through the pumice fall (and most not in roadcuts as at this site) show that it forms a narrow plume (at the 10-in. isopach it is 6 miles wide by 40 miles long) oriented N. 80 E., from the Big Obsidian flow vent. The orientation of the plume is the same as the current prevailing wind as indica ted by deformed trees on higher ridges and buttes (for instance, on The Dome, a cinder cone immediately south of here). Pumice lapilli and blocks constitute most of the unit, but accidental fragments of basalt, rhyolite, obsidian, etc., are common. Ash forms a conspicuously small part of the unit. Sizes of the pumice and lithic fragments decrease regularly eastward. 100

PAGE 111

(0.9) 60.0 (2. 0) 62.0 (4. 7) 66.7 (0. 9) 67.6 (1.5) 69.1 (0.8) 69.9 (0.5) 70.4 (0.1) 70.5 (3.4) 73.9 (0.4) 74 3 (4. 4) 78 7 (0. ) 79 6 Within about 6 miles of the vent the upper part of thin ash beds, generally 1 in. or less thick. They groundsurges, ash washed from the eruption cloud, pulses. Continue east. the unit contains a few may be products of small or decline of eruption View back to west of The Dome, a cinder cone partly covered by pumice fall. Road down east flank traverses the pumice fall. At the intersection of the road with Road 1821 (18) the pumice fall is about 3.3 ft thick. Note Red Hill, a young cinder cone to north; pumice fall once covered it but now is eroded. Turn south on Road 1821 (18) (elevation about 5,250 ft) which passes on west side of China Hat, an 0.8-m.y.-old rhyolite dome. Road intersection (Road 2228), keep to left on 1821 (18}. Note young basalt flow on west. pumice fall. Ash-flow tuff crops out beside road. Section line 23-26 marker on tree. It is over lain by Mazama ash and Newberry STOP 12. Ash-flow tuff (elevation about 5, 070 f t). Walk eastward on small dirt road to outcrops of ash-flow tuff. Note the exposures in roadcuts above the small cliffs. This ash flow crops out over a north-south distance of 5 miles but is overlain in places by alluvial deposits and basalt flows. It is buried on the west by similar rocks. The base of the ash flow is not exposed and the top is eroded, but it occurs in outcrops as much as 25 ft high. The ash flow is dark colored and contains abundant gray to black pumice in an ashy lithic-rich matrix. This ash flow may correlate with an ash flow that crops out locally for a distance of about 10 miles northward from this locality, and higher in the section than the ash flows to be visited at stops 1 3 and 14. Turn around and go back (northward) on Road 1821 (18). Cut on east side of road exposes pumice fall over Mazama ash. Old railroad grade to east leads to road on west side of China Hat where pumice falls and ash flows are plastered on the side of the rhyolite dome. Road 2129 (21) on west, continue northward on 1821 (18). Road on left leads to site of Brooks-Scanlon Camp No. 3, used in the heyday of log ing in the 1930's. Exposure of alluvium on left. The alluvium extends up the slope to the caldera rim. Ravine. South of the ravine are exposures of basalt with local windows of rhyoli t i ashlow tuff. orth of draw rhyolitic ash-flow tuff and a thin veneer of gravel form he surficial rocks to and beyond the next turn at 82. 2 miles. 101

PAGE 112

(2. 5) 82. 1 (0 .1) 82.2 (2.1) 84.3 (2.1) 86. 4 (1. 5) 87.9 1. 2) 89. 1 (0. 7) 89. 8 (0.6) 90. 4 Brook -Scanlon Camp No. 2 Form r log in it wa ast of road. Turn west on Road 2113 (1835) (elevation 4 885 f ) Road is along broad rid e covered by grav 1 deposits that overli ash-flow tuffs. STOP 13. Ash-flow tuff (elevation 5,170 ft). Park i n wid a r a on south side of road and walk southward to outcrops along side of ravine. This darkcolored ash-flow tuff, characterized by porphyritic obsidian as a common accidental lithic constituent, is formed of black to dark-gray porphyritic pumice in an ashy lithic-and crystal-rich matrix. It crops out along ravine walls at this general elevation for a distance of about 4 miles, beyond which it is buried by basal t flows. Farther upslope i t is buried b eneath alluvial deposits and basalt flows. The unit is similar to the ash flow visite d at stop 12 but has a different lithic component and is much more crystal rich. Ash flows, like those at stop 12, occur further north; some occur plastered on the sides of cinder cones (e.g. Orphan Butte). A thick sequence of lapilli tuff, identical to that at stop 2 on the west flank, occurs about 2 miles northeast o f this locality and higher in the section. Turn around and return eastward on road 2113 (1835). Continue on road 2113 across road 1821 (18). traverses area of rhyolitic ash-flow tuff like at stop 14, overlain by thin gravel veneer. Exposures of ash-flow tuff can be seen along many small ravines north and south of road. Kipuka of basalt projects through rhyolitic ash-flow tuff south of road. Turn north on Road 2013 (13) (elevation 4,665). Basalt flow on east was derived from a vent east of China Hat. STOP 14 (elevation 4 540 f t). Teepee Draw ash flow. Walk west from the border of the National Forest to outcrops of ash-flow tuff on sides of ravines. Note that basalt flows crop out on east side of nearest ravine and on west side of the far ravine immediately north of the ash-flow tuff. This rhyolitic ash-flow tuff covers a large area on the lower eastern and n ortheastern flank and is buried to the northwest and south by basalt flows. To ward t h e caldera gravel on the surface of the ash flow thickens to the point wher e float and outcrops of the ash flow can no longer be found. This is the oldest k n own tephra unit on the volcano. Plagioclase separated from pumice in it yielded an age of 0 7 m y (McKee and others, 1976), but unfortunately a high atmospheric a rgon content resulted in an uncertainty as large as the age. The a s h flow and overlying units are normally polarized. At this locality the lower par t of the ash flow is uncollapsed and consists of large pumices in an a shy lithic-and crystal-rich matrix. The pumices show progressive collapse upward and the top of the unit is welded. Note the scattered fossil fumaroles in the cliff face. Toward the caldera the unit becomes progressively more densely welded. The base of the unit is exposed only where kipukas of basalt project through it and the top is eroded. Along some ravines, however, the unit is more than 70 ft thick with the base not exposed. This ash-flow tuff probably was associated with an early, if n o t the earliest, period of caldera collapse. I t and the ash-flow tuff at stop 13 are older than the tephra deposits of the west flank and may occur at depth entirely around the volcano. Look to the east to see Pine Mountain, a complex 102

PAGE 113

(0.9) 91.3 (2.0) 93 3 (0 .4) 93 7 (1.1) 94.8 (2.5) 97.3 ( 1 7) 99.0 (2.0) 1 01.0 (16 .4) 117 4 (0. 7) 118 1 of flow s a n d intrusions mostly of dacitic to rhyolitic composition that are about 21 m y old and that probably correlate in part with the John Day Formation. Continue north on road. Turn n orthwestward on road 2043 ( 23). O utcrop of ash-flow tuff west of road. Outcrop of ash-flow tuff on subdued hill east of road is the most northeasterly known outcrop. Eroded alluvial fan deposits derived from Newberry and Pine Mountain cover the distal part of the ash flow and it is also buried on the west by basalt flows from Newberry. To the west is Horse Ridge, a highly faulted sequence of older basalt flows and near-vent deposits. The faults are part of the Brothers fault zone. Junction with U.S. Highway 20. Turn west. Viewpoint. Dry river canyon is 800 f t wide and 300 f t deep at this point. For next 0.8 mile roadcuts and canyon walls expose lavas of Horse Ridge. The basalt flows, with interbedded tuff and breccia, have yielded a 7 -m.y. K-Ar age. u s Highway 70 from here to Bend crosses diktytaxitic high-alumina basalts derived from Newberry that cover much of the low-lying region to the north. Road on north leads to Pilot Butte. Although young in appearance, this cinder cone is probably many hundreds of thousands years old, as it is locally overlain by an early Pleistocene pumice fall. End of trip at Junction of U.S. Highways 20 and 97 in Bend, Oreg. 103

PAGE 114


PAGE 115

HI H L V PL I BR THER F L T 1 E T HAl-( EY BA I I{E 1 zeorg Iker U . eologic I urv y M nlo Park, liforni 94025 nd Bruce olf Central r gon ommunity allege Bend, r egan RY The fi ld trip through the High L v PI ins province of c entral nd south-c ntr I regan (fig. I) provid s reconnatss nc ov rvi w of the Cenozoic volcanic geology long the Brothers fault zone from B nd to Harney Basin th site of vents for sever I sh-flow tuffs that cover thousands o f squar ktlom t r of southe t Oregon nd h ve volum s of hundreds of cubtc kilometer ost of this are has been mapped only in r conn iss nee (W lker and others 1967; Greene nd others 1972), and many inter sting det ils of the geolog have yet to be studi d From e st to :vest t h geol ogic record general! progr sses forw rd tn ttm wtth some of the youngest volcanic ro ks at the we t (or Bend) end o f the trip and the oldest at th ast end in and n ar Harney Bastn. This d scription is an oversimplific tion with respect to the basaltic rocks but the silicic rocks sho a ell-defined age progression in silicic volcanism described by lker (1974) a n d Macleod and others (1976) . <.."' ; I ; \ \)" ....... \ I ae.c I 0 R E G 0 N 119' Figure 1.-Ind x m p of r gon hawing the High Lava Plains and appro imate outlin of Harn y B sin (dott d line). 105 Idaho border ( alker, 1977), obscur s much of the older Cenozoic geology along the Brothers fault zone Th High Lava Plains are cont!guous ith nd gr da tiona! into the Basin and Range provtnce to the south, and many I te Cenozoic volcanic rocks nd fault structures are common to both provinces. For instance some oi the larger north-nd northeast-trending faults, characteristic of the Basin and R nge, ppear to change to a northwest trend and blend into th B rothers fault zone (Walker, 1977) A comparatively sharp boundary sep rates the High Lava Plains from the Blue ountain province to the north, here older Cenozoic and pre-Cenozoic rocks have been brought to the surface in the Blue Mountain-Ochoco ountains uplift. Older Cenozoic volcanic and tuffaceous sediment ry rocks are exposed along the northern margin of the High Lava Plains and in kipukas (inliers), such as Pine ountain, and include parts of the Columbia River Basalt Group nd the John Day and Clarno Formations (Walker and others, 1967) In most parts of the province, however the oldest rocks are aphyric and phyric basalt with plagioclase phenocrysts and minor andesite flows of middle(?) iocene ge. These flows, which have generally been referred to the teens Basalt, are exposed principally along the southern and southeastern margin of the province; sections hundreds of meters thick are beautifully exposed in the glaciated canyons high on Steens Mountain. An average chemic I composition of these mtddle(?) Miocene flows (table I, column l) indicates moder te to low silica content and a somewhat higher th n alumina content; this ver ge is not a f ir representation, however, of the chemical div rsi ty of th se rocks, hich v ry considerably in their proportions of pi gioclase, mostly labradorite, and olivine. Isotopic ges of the b salt flows indicate most ere e rupted bout 15 m y go (Baksi, and others, 1967; Laursen and Hammond, 1974; W lker, and others, 1974). T he basalts and andesites were over I pped by tuffaceous sedimentary rocks nd subsequently by widespread sheets of ash-flow tuff. The ash-flow tuffs, which re of rhyolitic composition (table I, columns 3, 4, nd 5), spre d laterally over tens of thousands of squ re kilometers of ancestral Harney Basin and form three principal units that r about 9, nd 6 5 m old. Th tot I volume of erupted tephra is unknown but is in the range of 1,000 2,000 km Eruption of this tephr was accomp nied by collapse into evacuated m gma chambers, which w s partly responsible for the development of the large structural d pression of Harney Basin The depression was subsequently filled w ith younger ash-flow tuffs, tuffaceous sedimentary rock and basalt flows and palagonitic sediments-all of middle(?) and late Cenozoic age (Piper, and other, 1939; aldwin, 1976; Greene and others, 1972). In places, the tuffaceous sedimentary rocks have been diagenetically altered to ben toni tic clay minerals, zeolites, and potassium feldspar (Walker and wanson, 1968), particularly where the sediments were deposited in ancient lakes that filled the lo er parts of the structural depression.

PAGE 116

120 10 I 1 0 L -, 10 2 0 1 0 2 0 3 0 40 30 MIL S 5 0 KILOM TEAS c R 0 ',L _____ 0 HORSE RIDG E ' 2 _,/ PI '\\ I ... EWBERRY VOLGA 0 DUN T A I \ ,_-.... ../ 3 I ,<,.....,...">.. (21 ) ._, HAMPTO BUTT CJ) '\ ...._ .Y CHI A HAT .EAST BUTTE \..,'-.'f-.. ( 0 .8) .) "{_! "'\ QUARTZ ''-f -...\FREDERIC \ \(.. ..Y / L A \ umbers in parentheses are selec ed K/Ar ages on rhyoli e dom s F igure 2A, B and C .--Map showing route of field trip long U .. High to Harney Basin, m jor tructural elem nts, nd select d K/, r domes. 106 y 20 from nd ges on rhyolt t 1. ... I

PAGE 117

c A 10 J. 20 R 120 30 MILES K G R A -.... .. -.. r .. / I .. : ., 2 107

PAGE 118

I 1 I ._J I L_. -. :-:-;--.-""" SOUA BUTTE TIRE TAl 7) gon t i r e ( 5.7) G R A ---./ T Buchanan 10__) ... 108 "!. . 3 (\). .. I I --2 DUC BU (10) 43'

PAGE 119

TJb C --:\,erJ'l' ( r t l r tc oJs .. dttc ...111 rh ---1 --2-I 2 7 -\ 2 .7 Fez 3 26 F 7 . 5 \lg .O dg t th t n .1r wbcrr} prob bly resurg nt dome related to v ry ob cur calderas tn Harney Basin or to ring-fracture zone Frederick Butte. Mo t of the older b salt flows in teens ount in re associat d ith major dike drms vi ibl on glaciat d c nyon walls on th east f c f the high t en s block; younger flows are associ a ted with broad shi ld volcanoes of lo relief, with cinder zones, or i th poorly exposed fi sure zones In pi ce the rising b s It magm encountered ground or surf a e water, which au ed fragmentation of the ba alt and violent ste m Where the explosions ere of madera te intensity b cones, tuff rings, and tuff ridges formed; th be t e of the e features are found bout 40 to 60 km south of th field trip route at Fort Rock, T bl Mountain, t. Patrick Mountain, and Seven Mile Ridge, all either within or margtn 1 to Fort Rock nd Christmas L ke Valley, and in the we t nd of Harn Basin, near Iron \ountain. More viol nt steam explosions bl sted out prominent craters, including uch fe tures as Hole-in-the-Ground. large t and most p c t cuJ r ruptive c nter wberry Volcano ne r the western end of the province. It is a comparatively young, Pleistocene and Holoc ne vole no, about 65 km long and 40 km wide surmount d by calder th t contains East nd Paulin Lakes nd v riety of volcanic features and products; t he walls of th c lder xpos s ver I varieties of both basaltic nd rhyolitic rocks, including sh-flo tuff, th t indic te it has h d a mor complex volcanic history th n most of the other vol e in th r gion (.VIac l od, and oth r this volume). of th High L v PI ins th northwe t-trending Brothers to be on of the fund mental gon. Th zone is continuou ly

PAGE 120

I /'.. \ \ ( ..... \ '-' --....., S 118 120 25 0 25 50 75 MILES I I I I I I I I I 25 0 50 1 00 KILOMET RS \ \-1 -----..... ------..... / _........., ) I ......__........-I 0 .Burns '\. Newberry Vole. I "--' .......... ----------D1amonu \ Craters / I Jordan Crater 0 R E G 0 ............ N .......... __ Cow Lake Craters 1170 Figure 3.--We t-northwest-tr nding belt of Middle(?) nd l te u t rn ry ba l t fi ld 110

PAGE 121

B ks1, K ork, lltr,.ll r'on. :\t d C od bsidi n vole nic R E J 972 u .. 1-680, r y of rocks, on : r t rs 111 P t r on , .. ( d .), 19 gu1d boo g I 0 p..1rt men 57 51 p \1n ral l ndu trte portion High Dep rtrnent eology and cooper tion 1 th U . ologi nd l : 250,000 of 1 r .11 t olcan1sm to I nd ks o F1eld c ntr reg on

PAGE 122


PAGE 123

ROADLOG FOR HIGH LAVA PLAINS, BROTHERS FAULT ZONE TO HARNEY BASIN, OREGON G org W. W lker, U .. G ologic 1 urvey, Menlo Park, aliforni 94025 and Bruce Nolf, Central Oregon Community College, Bend, Oregon This road log gives point by point information on the geology from Bend to Harne Basin and into the Diarr:ond Craters at the southea:t mar gin of Harney Basin. The route of the field trip and the locatiOns of some 1mportant geograph1c landmarks are shown on figures 2A, 2B, and 2C; the road log is keyed to these figures by mileages between principal points of interest and by field trip stops ILES (0.8) 0 .8 (3.5) 4.3 (16 .3) 20. 6 Junction U .. Hwys. 97 and 20 in Bend, Oregon, heading east on U Hwy. 20. Jet. with Pilot Butte Road. Pilot Butte, a late Pliocene or early Pleistocene cinder cone, is lapped by a pumiceous air-fall dated at between 1 and 2 m.y. For next few miles we travel on Pleistocene basalt cut by northwest-trending faults with displacements of up to 40 m; physiographic relief reduced o r eliminated by sedimentation in downthrown blocks. Jet. of road to Alfalfa. For several miles traverse young basalt partly covered by Mazama sh. Highway begins climb through older (7 m.y.) near vent basalt flows and flow breccias that characterize Horse Ridge. Northwest-trending faults, with up to 150 m vertical offset, cut these basalts. STOP NO. 1. At road jet. Geologic features that can be seen from this point include (from oldest to youngest): a. Pine Mountain to southeast, a kipuka of silicic to intermediate volcanic rocks dated at about 21 m.y. b. To west and north, faulted late Miocene m.y.) basalt of Horse Ridge. Note fault scarp north of and parallel to the highway. North of this fault are several additional and parallel normal faults which have localized basal tic vents in the form of small lava cones and cinder cones. c To southwest the large volcanic edifice is Newberry Volcano. (Refer to MacLeod and others, this volume.) d. To south, China Hat and East Butte represent silicic domes near west end of the age progr ssion of silicic domal rocks along the Brothers fault zone (MacLeod and others, 1976). Age of China Hat is 0 78 .!_0.20 m.y. and East Butte is 0 .85.!_ 0.05 m y e Low terr in to south and southeast is underlain by young basalt flows and lacustrine nd fluvi tile s dim nts of prob ble Pleistocene age. ot th t norm l f ulting along this p rt of the Brothers fault zone is bracketed by the 7 m.y. b 1t of Horse Ridge nd the basin fill. Faulting of Newberry Volcano and of flows in nd ba in indicate some displacements in middle Pleistocene to Holocene time. 113

PAGE 124

(3.8) 24. 4 (5.9) 30 3 (2.4) s e 1 highway f ul t th thin-b dd d 1 u trin dim nt in ro d ut on north id of nd 1 o on north id of highw y, h I on norm 1 Millican. ote c rp of noth r ch Ion f ult north of highw y econd flo from top of b it xpo d on s rp north of high y h b n d t d at 6.5 m.y. 32.7 Younger, mu h I s f ul t d flows int rb dded with gr v lly s dim nts dj nt to highway. ( 1.2) 33 9 Jet. with tate Hwy. 27. (7.2) 41.1 0 2 at Broth rs rest r Flat topp d ridg to northwe tis c pp d by basalt of (7. 0) late Miocene age, prob bly compar ble in g to b salt of Hor Ridg Irregul r tr ecovered hills to the north and ridg s forming north rn skyline r compos d of Eoc n to middle Miocene volcanic nd vole nicla tic rocks, including th Pictur org B s It nd the John ay and Clarno Form tions. Foreground is underlain by u tern ry b salt nd sediments. Source for some younger flows is low conical hill to thee st-north t with tank on top. Rounded hill to south of highw y is rhyoli t of unknown ge, prob bly older than the age p rogression. 48. 1 Jet. with road south to Frederick (Fredrick) Butte. Fault scarp north of highway is in late Pliocene o r Pleistocene basalt. (2.7) 50. 8 Jet. Camp Creek road. Turn northeast off highway on side trip. Go 3.6 miles to c ttle guard. STOP NO. 3 Rim to the northeast of cattle guard is 3 9 m.y. old ash-flow tuff which apparently came f rom vent area south of highway. Frederick Butte, to southsouthwest, is 3.9 m y old and composes one of several rhyolite domes on the east m rgin of a circular collapse area, probably the source of this ash-flow tuff. (7 4) Note that the ash-flow tuff and overlying basalt are cut by faults and here are downdropped south of the fault. South of highway both basal tic and rhyolite vents can be seen as low hill Hampton Butte, to east, is a volcanic pile that includes older rocks and intrusive rhyolite. The Clarno Formation crops out on west side and intrusiv rhyolite forms conspicuous hill on south side of Hampton Butte. Return to U.S. Hwy 20 and continue east. 58. 2 South of highway both north and south facing fault scarps are visible that cut Pliocene and Pleistocene basalts (younger than ash-flow tuff from Frederick Butte vent complex). Note ash-flow tuff rims no rth of highway. (3.4) 61.6 Hampton Station. As we approach Glass Buttes, the faulted basalts on both sides of highway are probably Pliocene in age. Some of these basalts appear to lap against the rhyolite domes of Glass Buttes. 11.0 72 6 Jet. Buck Creek Road, to G. I. Ranch. (2.4) 7 5 0 Jet. with BLM road to south. Side trip to outcrops of dated obsidian associated with Glass Buttes volcanic complex. Drive 2.7 miles to small reservoir. STOP NO. 4 Outcrops of obsidian, dated at 4 9 m y on both sides of wash. Glass Buttes complex is elongate in northwest-southeast direction, approximately parallel to the Brothers fault zone. At the east end of this group of rhyolite domes are opal! te-type mercury deposits and a thermal well. Return to highway. Note large mafic vent north of highway that fed Plio-Pleistocene basalt flows. Proceed east on U .S. Hwy. 20. 114

PAGE 125

p:z (6.4) 81.4 (3 .4) 84.8 ( 11.4) 96.2 (6.2) 102.4 ( 1.3) 103. 7 (5.3) 109.0 (0.7) 109.7 ( 1.9) 111.6 (0.5) 112. 1 (4.0) 116 1 (3.1) 1 19.2 ( 1.3) 120.5 (2.5) 123.0 (3 1) 126.1 (2 .8) 128 9 (2.7) 131. (3.5) 135. 1 Juniper-covered hills north of highw a y are another rhyolitic comple x e long ate northwestsoutheast parallel to the fault zone. Squaw Butte visible ahead. quaw Butte sits in the middle of another rhyolitic c o mplex, dated at 5 1 m y ., surrounded by b asalt whi c h in places laps onto the rhyolite Dry Mountain north of the highway is a large rounded volcanic pile of faulted middle Miocene hypersthene andesite lapped by 6-m.y.-old ash-flow tuffs. A few miles north of Dry Mountain, pre-Cenoz oi c rocks are exposed, some as old as Devonian. V alley north of highway and south of Dry Mountain is filled with Pleistocene gravel and a few thin Pliocene and Pleistocene basalt flows. Jet. U.S. Hwy 395 (Riley). Fault s carps cutting Pliocene or possibly late Miocene b asalt flows to the north and northe st of highway. Sharp peak on southern skyline is Iron Mountain, a rhyolite dome dated t 2-3 m.y. This appears to be the only exception in the well-defined age progression in sili c i c volcan i s m Enter area under l a in b y palagoni ti c (basal ti c ) bedded tuffs of Pliocene a nd l eistocene age. Ditch on north side of highway provides best exposures of these tuffs. STOP NO. 5 : Highway rest area. Palamino Buttes, south of highway, consists of 6 .5-m.y.-old rhyolite; adjacent to Buttes on east is eroded basaltic cinder cone. Buttes lapped by basalt flows dated elsewhere at 2-3-m.y. Both the basalt and under lying rhyolite are cut by northwest-trending normal faults with decreased offset of the basalt. South of highway for next several miles, rim of 2-3-m. y. basalt underlain by tuffaceous sedimentary rocks generally referred to as part of the Harney Formation. These sedimentary rocks rest on the Rattlesnake Ash-flow Tuff. Exposure of Rattlesnake Ash-flow Tuff in roadcut. Outcrop of rhyolite south of highway and addi tiona! outc rops to northeast and north of highway. These rhyolites are part of the Burns Butte volcanic complex which has been dated at 7.8 m.y Within the complex are some mafic cinder cones. Borrow pit on south side of highway exposes one of these mafic cinder cones. Clastic dikes cutting these cinder cones have recently been described by Peterson (1978) City limits of Hines, Oregon. J e t. of U S Hwy 20 (also U .S. Hwy 395) and State Hwy. 78 in the center of Burns, Oregon. Burns lies at the northwest margi n of a nearly circular are that represents central Harney Basin, an area characterized mostly by flat-lying younger fill and some dune deposits. Central Harney Basin lie s within a much larger structurally depressed r a that coincid s approximat e ly with the physiographic basin outlined on fig. 2 The s tr uc t u r I basin evolved i n part c on cu r rently with the eruption of late Miocene ash-flow tuffs, hich re tra v rse d b y t h e n e x t st g e o f the field trip. Turn no r t h on U .. Hwy s. 20 and 395 Jet. U .. H y 20 nd 3 9 5 turn north on 395. 6 : E xpos u r a bout 20 m t hic k, o f R attlesnake Ash-flow Tuff, which has been d t d t b ou t 6 4 m. y. o n t h b s i s of numerous analy es. A single flow and cooling unit of rhyolitic ash-fl o w t u ff that r sts on bedded sedimentary rocks. onsists of a well 115

PAGE 126

( 1.0) 136.1 (0.7) 136.8 ( 1.3) 138.1 ( 4 .1) 142.2 (0.0) 0.0 ( 1.7) 1.7 (10 .5) 12.2 d v l o p db a vitrophyr n Vents for thi nd underlying 1 t Miocen h flow tuff r buri db n th young r fill in th 1 wer central part of H rn y B in. nly indir ct nd m g r vid nee is available s to their location and chara t r. T P NO. 7: Prater Cr ek Ash-flow Tuff exposed in ro d cut. ingl flow nd ooling unit that is 8.4 m.y. old, a nd is slightly mor m fie than th Rattle nd D vine anyon sh-flow Tuffs. This unit is ch r terized by low cry t I cont nt nd bund ant 1 lithophysa Maximum thickness is bout 30 m; its xtent is poorly known b use it is largely buried by young r materials including th R ttlesn k Ash-flow Tuff. Unit is underlain by bedd d tuffaceous edim nt ry ro ks; int r -shflow tuff units thin w a y from central H rney asin. Cross railroad tr cks. STOP NO. 8: Upper non-welded part of vine anyon Ash-Flow Tuff. olumned s h flow tuff across creek to the north-northwest is welded nd compact d low r p rt of the same unit, samples of which have been d ted. from a number of widely separated localities at about 9.2 m.y. STOP N 9: Outcrops on both sides of road of Devine a nyon Tuff. A single cooling, probably multiple-flow unit that covers more than 18,000 km entered on the east half of Harney Basin and extending north, south, and e t of the basin. M ximum thickness is about 35 to 40 m, although most of the unit is less than 10 m thick Isopaching suggests a source caldera in the vicinity of Burns, perh ps with its west margin in the vicinity of Burns Butte. Contains s much s 30 percent crystals and crystal fragments, mostly sodic sanidine, ome quartz, and minor, ubi qui to us green ironrich clinopyroxene. Thinner distal ends tend to b glassy, where s thicker tions exhibit more de vi tr ifica tion and locally extensive vapor-phase a! ter tion. At this point turn around and return to Burns. LOG FROM BURNS TO DIAMOND CRATER Jet. of U.S. Hwys. 20 and 395 with State Hwy. 78 He d east on 78. Jet. State Hwys. 78 and 205. Turn south on 205 STOP NO. 10: Viewpoint on spur gravel road, south rim of Wrights Point. Wrights Point consists of a 2.4-m.y. valley-fill basalt flow which h s been erod d into positive relief. The point extends 6 5 km from here eastward; the vent is several kilometers to the west. Visible beneath the basalt flow are some of the best exposures of bedded tuffaceous sedimentary rocks of the Pliocene Harney Form tion. From this point looking to the south-southwest, the juniper-covered hill is Dog Mountain, original site of the type section of the Harney Formation. Inasmuch s Dog Mountain is a palagonite tuff ring, atypical of the formation, road cuts on Wrights Point have been recently de signa ted a reference section (Walker, in press). The tuff ring is late Pliocene. West of Dog Mountain are a number of additional young basaltic vents (tuff rings, cinder cones, lava cones, and dikes). East and southeast of Dog Mountain are ma n more young basaltic vents, including those at Malheur National Wildlife Refuge Headquarters, at Coyote Buttes, and Diamond Craters. Steens Mountain (elev. 9,733 feet; 2,967 m), a west-tilted fault block capped by the middle Miocene Steens Basalt dominates the southeast skyline. On the east-southeast skyline is Duck Butte, a large rhyolitic dome complex about 10 m.y. old, considered by 116

PAGE 127

.. ( !0.7) 22.9 (0.9) 23.8 ( !.6) 25. 4 (2.4) 27. 8 (3.1) 30. 9 (3.0) 33.9 (6.6) 40.5 (0.8) MacLeod and others (1976) as probably the oldest rhyolitic uni t within the progression. Excellent view to the west of the rhyolitic dome at Iron Mountain. ge Bridge t Narrows. Some geophysical data suggest that one or more source calderas for the late Miocene ash-flow tuffs are centered about here (H R. Blank, or 1 communication, 1974). Harney Lake, to west, is the current sump of the Harney Basin drainage system. Low hills to the east are young basaltic cinders, palagonite tuffs, and related flows. J et. with road to Malh ur National Wildlife Refuge Headquarters and to Princetor., Oregon. Along this road are numerous young basaltic vents, mostly characterized by low rounded hills consisting of red cinders, common cored bombs, and agglutinated basalt. STOP NO 11: Roadcut in a crystal-poor pumic i te that is thoroughly a! tered by zeoli tization (erioni te, clinoptiloli te, etc) and may represent either an ash-flow tuff or a pumice slurry deposit that flowed into shallow water. Pumice fragments are neither collapsed nor welded. Excavated basaltic cinder cone to east of highway. Rim on west side of road is welded pumiceous ash-flow tuff, probably Rattlesnake Ash-flow Tuff. Cap on Saddle Butte is basalt flow interstra tified with the Harney Basin deposi tiona! sequence. Exposures in low roadcut of non-compacted, non-welded, partly altered crystal-rich Devine Canyon Ash-flow Tuff. Alteration suggests that this also may represent depositi on of ash-flow in wet e nvironment. STOP NO. 12: On moderately well welded crystal-rich Devine Canyon Ash-flow Tuff (9.2 m.y.). This is overlain by pumiceous ash-flow tuff to the west and is underlain locally by as much as 150 m of bedded tuffaceous sediments which lap south and southeastward onto middle Miocene ( % 15 m y ) Steens Basalt. Th e o verlying pumiceous tuff also thins to the south. Note the several northwest-trending fault scarps in which units are stepped up to the south. This is a prominent northwest-trending zone which can be traced almost 100 km. 41.3 Exposures of densely welded Devine Canyon Ash-flow Tuff. (0.8) 42. 1 Turn left on paved road to Diamond Oregon. ( 1.5) 43. 6 Rimroc k o n both sides of road is Devine Canyon Ash-flow Tuff. View to south-southwest of large northeast-trending, east-facing scarp along the fault zone that sep rates the west-til ted High Steens block from the Jackass Mountain block. Th is i s one locality near the southern margin of the Brothers fault zone where a northeast-trending fault sy tem curves into a northwest-trending system. ( 1.2) 44.8 Excellent profile to the north of the Diamond Craters volcanic complex. This compl x has bee n studied by Peterson and Groh (1964). From here can be seen a few vents and loc 1 gentle doming of basalt flows ithin the complex. The valley to the north and east i und rlain by surfici 1 deposits resting on Devine Canyon Ash flo Tuff which is faulted ( 4 .I) and underlies th i mond Crater complex. 48.9 Turn north on gravel road to iamond Crat rs. ( 1.0) 49.9 L ft at "Y" on ro d to iamond Cr ters. Shortly beyond this point we see basalt flows of (2.2) th iamond Cr ter complex th t 1 p onto evin Cany n Ash-flow Tuff. 52.! o d jet. at borro pit in inder cone. The princip 1 p rt of th complex is west and north of this point. 117

PAGE 128

v il bl nd a. n hill top, approximately 1 km north of borrow pit j t., w th outh m rgin of a 2.3-km-long northwest-trending gr b n long th n longat dom 1 structure in the basalt. This gr ben w s pp rently the i te of li tt1 e or no eruptiv act1v1ty. ummit collapse prob bly result d from withdr w I of lav from high standing magm body beneath this up rch d tru tur The complex in ludes fi v of thes elongate domal tructure all of whi c h oriented in northw st-outh direction. Four of the five are ch racter iz d by various d grees of ummi t fr ture nd coil ps In the near foreground to the outh is a prominent pit rat r nd relat Many of these young t vent r localized t the margin of dom 1 stru tur s ( t point of maximum flexure) and represent points of lav egr s t times of ummit coli p e. \ The tendency to use degree of vegetative cover to d t rmine r 1 tiv be misleading. For example, in the foreground to the south, the sagebrush cover laps onto the I ss veget ted flow. Based on evaluation of weathering features in this dry environment, we b Jieve th t thi basal tic complex is late Pleistocene in age, rather th n Holocene. Thus f r no m ter ials suitable for isotopic dating have been found, but n age of 17 ,000+2,000 ye rs w s obtained by Friedman and Peterson (1971) on hydration rinds of rhyolitic m terial enclosed in basalt bombs. Jackass Mountain, to the west-southwest, is the high point on the J ck ss Mount in fault block which is bounded by prominent scarps. Most of the rims to the south are fault d and eroded Devine Canyon Ash-flow Tuff. Some patches of R ttlesn ke Ash-flow Tuff are present on the back (west) slope of teens Mountain. The giant west-dipping teen Mountain block is bounded on the east by precipitous northeast-trending, 1 ,800-m-high fault scarp, and is lapped on the west by sedimentary rock containing a B rstovi n vertebrate fauna and by late Miocene ash-flow tuffs and interbedded sediment ry rocks. NOTE: For those interested, a loop road between Frenchglen and northern atlow V lley traverses the dip slope and crest of the High Steens. The lower country to the north of the Diamond Craters complex is locally under! in by accumulations of young basalt, mostly valley-filling flows. Some of these flows effectively closed off the pluvial drainage from Harney Basin. On the skyline to the north, older rocks crop out north of the Brothers fault zone. b. Approximately 2 km west of the graben is a large vent area referred to by Peterson and Groh (1964) as the central crater complex. This complex is localized by a partly filled northwest-trending graben at the crest of another elongate domal structure. Probably graben development and eruptive activity were at least in part contemporaneous. More than 30 separate vents have been identified within this depression. Vent activity ranged from quiet emission of lava to highly explosive volcanophreatic events. Blocks of Devine Canyon Ash-flow Tuff and associ a ted units are found in the tephra; exotic blocks up to 50 m in maximum dimension are present. Other features within this central crater complex include exhumed conduits, driblet spires and spatter cones, and small-scale collapse structures. 118

PAGE 129

A FIELD TRIP TO THE MAAR VOLCANOES O F THE FORT ROCK-CHRISTMAS LAKE VALLEY BASIN, OREGO N G H Heiken, Geosciences Division, Los Alamos Scientific Laboratory, Los Alamos, NM 87545 R. v. Fisher, Dept. of Geology, University of California, Santa Barbara, CA 93106 N. V. Peterson, State of O regon, Dept. of Geology and Mineral Industries, Grants Pass, OR 97526 The Fort Rock Christmas Lake Valley basin is a former lake basin that e xisted from late Pliocene through late Pleistocene time. The basin is about 64 km long and 40 km wide (Fig. 1). Eruptions of basaltic magma occurr ed along faults that trend diagonally across the basin and adjacent highland, forming maar volcanoes within and o n lake margins and forming cinder cones with flow s beyond the lake margins (Peter son and Groh, 196 3 ; Heiken 1 97 1 ) The purpose of this field trip is to visit several of the maars and a maar complex in and near the basin. Road Log: (cumulative distance, in miles; stop-to-stop mileages are in pa -rentheses) 0 0 ( Fig. 2). La Pine Junction; intersection of Oregon State Highways 97 and 31. down Highway 31, to the southeast. 1: Location map. 119 Proceed

PAGE 130

_. I'\) 0 0 0 0 (] Cb Fl
PAGE 131

( 1. 0) 1.0 ( 1. 9) 2 9 (7.1) 10 0 (0. 3) 10 3 Large meadow ; good view of Newberry Volcano to left (north). Railroad crossing; as is typical of the area between La Pine Junction and Fort Rock Valley, the surface (except for steepest slopes) is mantled with Mazama pumice from Crater Lake. Bend in road; Moffitt Butte tuff ring is straight ahead. Moffitt Butte is on left; its base is about 50 m east of road in the woods Moffitt Butte Moffitt Butte is a dissected tuff ring, 1400 min diameter and 120m high. It is a praninent topographic feature, but is obscured by the forest from the road. Although not associated with a lake basin, as is the case for Big Hole and Hole-in-the-ground, Moffitt Butte is a tuff ring composed of hyaloclastic tuffs. Rising magma may have encountered permeable aquifers beneath the cone. A line of tuff rings between here and the Fort Rock Basin are along a topographic low between Fort Rock and the La Pine basins. The crater floor of Moffitt Butte is about 80 m above the surrounding plain. A parasitic vent, and small tuff ring, 510 m in diameter, is located on its southwestern flank. The deposits consist of sideromelane lapilli-tuff in graded and ungraded beds, 3 to 30 em thick. Near the main ring crest is an unconformity dipping 20 into the crater that truncates beds dipping outward at 35 (Fig. 3). Rocks above the unconformity consist of a 1m thick bed of angular basalt blocks and 18m of very wel l-bedded lapilli-tuff. The crater of the parasite vent is filled with lava that issued from a dike on its northwest edge. (2. 8) 13. 1 ( 1. 5) 14. 6 Road cut through pressure ridge or large tumuli in a basalt flow. This is typical of many road cuts between here and Fort Rock Valley; basalt flows are from vents on the southern flank of the Newberry Volcano; overlain by Mazama pumice. Junction of State Highway 31 with Rock Creek Road (gravel) Exposed in road cut south of highway are remnants of an unnamed tuff ring (Ridge 28 of Peterson and Groh, 1963); gray to orangish-brown (partly palagonitized) well-bedded hyaloclastites. Most units contain accretionary lapilli. NW 4sooF 44oo E NE 4soo E 44oo E? SE sw Figure 3 : Cross-sections through Moffitt Butte, Klamath County Oregon. The upper cross section is through the main body of the tuff ring. The lower cross-section is through a parasitic vent on the southwest flank of the ring; the lava-filled crater is close to the highway. 121

PAGE 132

(5.0) 19 6 Big Hole J nction, t t Highw y 3 1 Big Hol Ro ( so u t h) to en t r Big H o l e ( no i n cl u d lurn of'f in th Y 3 1 to right A g ravel road from this junction go circul r 1820 m t th cr st and o n h northeast iameter maar crater. Deposits o f the exten 1800 to 2500 m beyond th crater rim. si e along Highw a y 31 (Big Hole Butte). The tuff ring is composed o f moder te-to .J 11-b sid romelane lapilli-tuff n tuff breccia, in beds 5 e m to a met r thi k. Th ccia b ds i nclu porp hyritic basalt blocks up to 2 5 m in iamet r ; ther a r bund nt be ding pl n gs c us d by impact of these blocks into once w ter-satur t ash b ds during th ruption. Best exposures o f these deposits are along gullies on the east rn r i m o f th maar. Conv olute bedding within the rim deposits is w ll xposed along th east r i m (Fig. 4). Within the crater is a 152m-wide ledge that appe ars to beth top o f lar g b loc that slumped into the crat r, possibl y during ruption. Collaps into th c rater of such large blocks would explain the large volume o f the crat r an sm ll p rcentage of xenoliths within the ejecta. (0.6) 20. 2 ( 0 4) 20. 6 ( 4 5) 25. 1 O n the right are. well-bedde hyaloclastic tuff deposits o Big Hole tuff ring Deposits are thicker on the northeast rim, probably caused by prevailing winds from the southwest during the eruption. The thicker posit is called Big Hole Butte, although it is not a separate structure o r v ent. Good exposure of Big Hole tuffs on right side o f Highw a y 3 1 Junction of State Highw a y 3 1 with th roa o Hole-in-the g r oun (HIG) (Pro ceed along Boundary Road 11245). The side trip t o H IG will not be included in the overall mileage. Forest Road 245 is gravel; follow sign s or 2 8 miles (one left turn and one right turn t o the west rim and to HIG overlook). Hole-in-the-ground Hole-in-the-ground is described by Peterson and Groh ( 1961 196 (1971). Lorenz's (1971) abstract is as follows: and by Lorenz "Hole-in-the-Groun is a volcanic explosion crater o r maar located in Central Oregon on the edge of For Rock basin. At the time the crater was formed between 13 ,500 a n d 18 ,000 ago a lake occupied most o f the basin and the site o f the eruption was close to the water level near the shore. The crater is no w 112 to 156m below the original ground level and is surroun e by a rim that r ises another 35 to 65 m higher. The volume of the crater below the o riginal surface is only 60 percent of the volume of the ejecta. The latter contains only 10 percent juvenile basaltic material, mainly sideromelane produced by rapid quenching of the lava. Most o f the ejected mate rial is fine grained, but some of the blocks o f ol e r rocks r ach im nsio so 8 m. The argest blocks are concentrate in f our ho rizons and reached distances of 3. 7 km from the center of the era ter. Ace ret ionar y apilli, impact sags, and vesiculated tuffs are well develope The crater was formed in a few days or weeks by a series o f explosions that were triggered when basaltic magma rose along a northwest-trending fissure and came into contact with abundant ground water a t a depth of 300 to 500 m below the surface. After the initial e xplosion, repeated slumping and subsidence along a ring-fault l ed to intermittent closures of the vent, changes in the supply o f ground water, and repeated accumulations of pressure in the pipe. Four major e xplosive events resulted from pressures o f over 500 bars in the orifice of the vent. Ejection velocities during these periods reached 200 meters per second. The corresponding pressures 122

PAGE 133

1 23 his r gon. Con olute b ddd oll m nt-l.k mov m nt of

PAGE 134

. .. I ... CENTER FOR VOLGA, OLOGY 1'. .. ....... "":' .. : .. ..... +: ....... . I .. .... Figure 5 : Topographic and geologic map of Hole-in-the-Ground. feet. Field control by J W. Hawthorne. 124 - u .. a-.Jt ""'' Pyroclaatlo o j octo n.,-.toulo Co,.l_.roloe 1'11wu-l&a< llo.l Opp.r 1p1Jil>T1t e o f "-79rl T\off Contour interval is te

PAGE 135

.,..,. ..... Lake beds ,_.,,. .. ( rocks of H .G. "/JJ..JO'. Porphyrllrc basal!, lype 4 llllllr:tu:! !"'orphyrrllc basal! 1 ype 3 FrnOIJrarnod basal!, type 5 I,ll Drillho l e t\o. I No L: +++Intrusive basalt L : H 400 600 BOO IOOOm Figure 6: Geologic section through Hole-in-the-Ground based on surface exposures, drill holes and geophysical data. Interpretation by Lorenz (1971). and velocities during intervening, less violent stages were in the range of 200 to 250 bars and about 130 meters per second. The kinetic energy released during the most violent eruptions was approximately 9 x 1020 ergs and the seismic events that must have accompanied these explosions had a magnitude of about 5. Ejecta 10 centimeters in size were thrown to heights of 2 to 3 kilometers and the eruption cloud may have reached 5 kilometers or more The axis of eruption was slightly inclined toward the southeast; the form of the vent seems to have had a more important influence than wind. Base surges that accompanied some of the explosions left deposits of vesiculated tuff. The total energy derived from the basaltic magma was of the order of 5 7 x 102 3 ergs. Most of this energy went into heating of ground water and the enclosing country rocks; only a small part, possibly a tenth was released by expansion and vaporization of the water and mechanical processes, such as crushing, acceleration and ejection of debris. Geophysical measurements indicate a domical intrusion below the crater floor and extending upward as a ring dike around the margins of the crater." Return to State Highway 31. ( 1. 7) 26.8 26. 8 to 28. 2 Picnic grounds on right. Section (top to bot tom) through ( 1 ) basalt flow ( from flanks of Newberry Volcano?) and (2) Peyerl Tuff; a section consisting of 4.0 m.y. old pyroclastic flows and tuffaceous sediments. The section is exposed along Highway 31 in road cuts. I f you stop, please leave your auto at the top of the hill in the picnic grounds and stay on the edge o f the road; there are nl.ITlerous fastmoving logging truck s on this 125

PAGE 136

( 0. 9) 29 1 ( 6. 4) 35.5 ( 0 1) 35 6 (0.5) 36. 1 fort Rock Junction, Oregon State Highway 31 and Fort Rock Road. Turn left (east) on Fort Rock road. Before reaching the lake basin floor, the road cross s several north-south trending horsts and grabens th t cut through a section consisting of the Peyerl Tuff and Pliocene(?) age basalt flows. On the left (north) of Fort Rock road along an east-west line to Fort Rock re the r mnants of four deeply eroded basaltic vents. Onl y small remnants of lava lakes remain, but most are surrounded by aprons of rounded cobbles and pebbles consisting of palagonitized hyaloclastic tuffs. The vent closest to Fort Rock (Beggars Heel) has a cave where some of the oldest man-made artifacts in Oregon were found. Downtown Fort Rock and intersection with Fort Rock State Park Road (follow signs) Turn left to Fort Rock State Park (if you stay o n this road and follow th signs, you can reach Hole-in-the-Ground from the east side). Fort Rock State Park Fort Rock with its spectacular wave-cut cliffs, is an isolated maar volcano within a monotonous, flat lake basin (Peterson and Groh, 1963). The wave-cut remnant is 1360 m in diameter and 60 m high, and the present crater floor is 6 to 12 m above the floor of the lake basin. The south rim has been breached by waves of the former lake, providing easy access to the crater. A wave-cut terrace occurs 20m above the floor of Fort Rock Valley (fig. 7). The maar is composed of orange-brown lapilli-tuff in beds of 1 em to 1 m thick that can be traced from within the era ter to the outer flanks. Graded beds with accretionary lapilli are common Inward-dipping beds are parallel to the crater walls (fig. 7) and suggest that the crater is funnel shaped; the innermost beds dip inward at angles of 20 to 70 degrees. On the west side of Fort Rock is a distinct angular unconformity where the deposits, truncated by slunping into the crater, are plastered with younger beds. These younger beds are part of a continuous pyroclastic sequence on the outer flanks. An incipient slump is visible on the east flank where the sequence is in-place with quaquaversal dips a way from the rim crest; fracture planes associated with the slump, dip in ward at 40 degrees, the same angle as the surface of the unconformity on the opposite side of the crater. Fort Rock is typical of most of the smaller, isolated maars of the Fort Rock-Christmas Lake Valley basin, such as Table Mountain, Flat Top, Lost Forest, Green Mountain SW and Green MountainS (fig. 2). (7.00) 43 1 Return to junction of fort Rock Road and Highway 31 and turn left (south), down Highway 31 toward Silver Lake An alternate route, from the center of downtown fort Rock is south along Lake County Road 513. This is a gravel road that is a short-cut to Silver Lake, but the bridges will not support heavy trucks or buses. If you take this alternate route of 16 5 miles, you will go between a north-south trending normal fault (west) and the Connley Hills (east). As you leave the lake basin, near junction with County Road 510C you will see the northern end of the Connley Hills, a line of domes of intermediate composition. Remnants of a small maar are plastered onto the northernmost dome, near the former lake shoreline. South of the Connley Hills is Hayes Butte, a basaltic shield These volcanoes stood as an island in the lake that was present here during late Pliocene-Pleistocene time. To the west, several broad, older (Pliocene) maar volcanoes are exposed in a low ridge, partly buried by basalt flows. a gravel road to the west (turn right, 5 3 miles south of the tow n of Fort Rock) crosses one o f these maars. It is on private land, however. Please ask for permission to enter before doing so. 126

PAGE 137

0 / / / 1 as 300m. I I / / / / Qt '-,',,, '<. .... ... -----,, .. I I I I \ \ \ \ -I --; / \10 \ \ I ...,..... 30 30( 2?.-l-" -RO-\ 36 -"' I I I I I I I / Qs I I I I I I I I I ... ? C Figure 7: Fort Rock tuff ring A "MAP. Qs Pleistocene 1 ake sediments, mostly d iatomites. Qt-tuffs, lapilli-tuffs and tuff breccias. Scalloped line is an unconformity. Dot ted line is the cliff edge. B. O blique aerial photograph, from the southwest. C Cross-section. A

PAGE 138

(0.4) 43.5 ( 3. 1) 46.6 (9.4) 56.0 ( 2. 8) 58.8 ( 1 9) 60.7 (6.4) 67. 1 (2.0) 69. 1 (2.0) 71 1 ( 1. 1) 72.2 After crossing a small divide, you will enter Paulina Marsh; no outcrops until you reac h the intersection with Highway 31 n ar Silver Lake Road cut (Highway 31); park at base or top of road cut. Be careful; stay o n the side of the road. There are good exposures of part of the Peyerl Tuff, a sequence of pyroclastic flows and interbedded sediments, as much as 150 m thick, that crops out along the west side of Fort Rock Valley. Radiometric ages of 4.47 + 0.84 and 3 35 + 0.44 m.y. were determined for several of the pyroclastic flows (MacLeod et -al., 1976). The source for these pyroclastic units appears to be a 7 to 10 km wide caldera near Wart Peak, that is about 15 km WSW of this stop. Junction, Highway 31 and Wickiup Springs -Starns Well Road. The ridge on the horizon to the west is near the east rim of the Wart Peak Caldera, identified by N. MacLeod. Due south is a cinder cone of late Pliocene(?) age. On the east side of the road is a poorly exposed Pliocene(?) maar, the Wastina Maar of Peterson and Groh (1963). Straight ahead (south) is Hager Mountain, a 5 9 m. y. old silicic dome on the southwestern rim of the Fort Rock Christmas Lake Valley Basin (MacLeod et al., 1976). Exposure of lake sediments in road cut (Highway 31). Town limit; Silver Lake. As you drive along Highway 31 between the town of Silver Lake and Christmas Valley Road, note the wave-cut terraces and notche s along the basin rim, especially northward around the "island" formed by Hayes Butte, Connley Hills and the Table Rock Maar Complex. The maar s are straight ahead and north of the highway. The Silver Lake graben (south) is the keystone of a broad arch broken by normal faults; Pliocene age basalt flows exposed in this arch slope northward into and under the lake sediments of the Fort Rock Christmas Lake Valley Basin. Junction, Oregon State Highway 3 1 and Christmas Valley (Arrow Gap ) R oad. Turn left (north). Dunes south of the junction rim Silver Lake, the marshy rP.mnant of a much larger Pleistocene Lake. On the right (east) is the Table Rock maar volcano complex. The lowest cliff is a basalt flow from Hayes Butte that underlies the west edge o f the complex. Dirt road into the center of the Table Rock maar complex leaves the pav e d county .road (to the right). Continue east and return to this junction after visiting stop A. Stop A. Just b e y o nd the pass, a t t h e bottom o f gully c u t nto the maar complex, stop. Proceed on foot up a jeep trail to the right (east) TABLE ROCK MAAR COMPLEX (STOP 2) General Description Table Rock is a tuff cone located 14.5 km east of the village of Silver Lake, o n t h e shore of Silver Lake, one of the few remnants of a much larger Pleistocene lake that once filled the Fort Rock-Christmas Lake Valley Basin (Fig. 2). The Table Rock tuff cone is part of a maar complex consisting of the cone, two large tuff rings and six smaller tuff rings or eroded vents. This complex forms an elongate, NNW-trending oval 5 6 by 8 8 km. The highest point, about 395m above the basin floor, is the crest of Table Rock. Silver Lake graben, located immediately south of the tuff-ring complex, is the k eystone of a broad 25-km-wide structural arch that forms the southwest boundary of the lake basin. The normal fault defining the east wall of the graben (through which Highway 3 1 128

PAGE 139

. __.. .......... TABLE ROCK LAKE COUNTY,OREGON J \\ .. I EXPLANATION '-.......;..__ I CENTER of MAAR ';A CRATER ---RI M o f MAAR jt; A STOPS VENT NUMBER HORIZON TAL BEDS I I SYNCLINES 8 1:+ ANTICLINES OUE TO SLUMPING CHANNELS I i 1N [J GRAVEL BAR -D I KES ;:; AVA LAKE z w u g v D M AAR DEPOS ITS j D I TERBEDOEO c. SEDI M E NTS 0 I TERS E ODED BASA LT F"LO FROM HAYES BUTTE OOm I u ,..---, / / ..... \\--.. \ ..... \ R ANCH ' '--\ J I / + /H / / Figure Table Roc k tuff ring c omple>-.. 129

PAGE 140

continues o e r Picture Rock Pass) is p rallel to the long xis of th maar compl x and possibly lies beneath it. The fault, howev r, doe s not cut any of the tuff d posits The tuff ring complex o v e rlies a 220 m thick ection of lake sedim nts and interbedded tuffs, and sands and gravels derived from the Connley Hills. Th se predominantl y lacustrine sediments overlie the Pi cture Rock Basalt ( H mpton, 1964), feldspathic diktytaxitic basalt unit of Pl1occne age. The basalt is well-exposed along th south rn edge of the lake basin, especially along Highway 31, south of the T ble Roc k Complex The Picture Rock basalt, consisting of individu 1 flows up to 9 m thic k with an aggregate thickness of over 230m, dips into the basin wh r it is th main aquif r tapped for irrigation (Hampton, 1964). South and east of the basin, shallow lacustrine and flood plain deposits and maar eposits re int rbedded with the Picture Roc k bas lt ( Walker et al., 1967). Stop A, table Rock Complex Park on the county road, walk east along th gully south of vent 6 (bluff on north) for about 350 m Sediments underlying the Table Roc k Complex are well-exposed here (Fig. 8). In contrast with well-bedded diatomites found on the east side of the complex, these consist of interbedded volcanic litharenites, lithic arkoses, diatomaceous siltstones and lapilli-tuff that form an outwash apron around the Connley Hills. The Connley Hills, located northwest of Table Rock consist of a basaltic shield and intermediate t o THICKJ\ESS (1-IETERS) DESCRIPTION Massive tuff-breccia, w ith blocks of the underlying units. \.:....;1' ------::-------------------Gray pumicebe11ring s;,nd or.e ; \Olcanic 4.5+ litharenite. Well-bed e d 1\'h i te d i a ton ace ou5r. dStone ; Gray Well-bedded, subnature coarse sand tone; tuffaceous volcanic litha;cnitc. The sandstone contains abundant reworked pumice fragments. h'hite, well-bedded sandy clzystone; possibly diatomaceous. Poorly bedded su mature medium to coa:se pumice-bearing lithic subarkose. Pum1ce content increases n ear he base to a out 60\ of the rock. Well-bedded gray or whiter dium sand stone sub-lithite-arenite. Interbedded with ihe sandstone are a few thin layers of conglomerate, made up of well-rounded pumice fragw.ents. There i s some crossbedding. Talus covers the base of the section. ;;-!..Jure GJ: ..... cct. on o sediments exposed under the northwest corner of the Table Rock L f ring canplex. 130

PAGE 141

Fi.sur:_e 9 : Diagram -or -:Jor"rrna i.on pattern dL the tuff-breccia (stippled) sediment (pl "'.t.ll) .-< Htl-cict below tuff ring 112, table Rock tuff ring complex. Uppermost layers of sandstone ahd mudstone are intruded into the basalt tuff-breccia as dikes. Deformed bedding planes are preserved within the dikes, although nearly pinched off where the dike enters the breccia. Bedding below the contact is also deformed by small scale faults. silicic domes and flows. They w e r e an island, 6 4 km wide and 19 km long, throughout the late history of the lake basin. The sediments are higher than those of the basin and may have been deposited in a small depression between the large tuff ring (2) and the hilli to the north. The sediments are deformed beneath thick units of hyaloclastic tuff-breccia and penetrate the tuffs as mudstone dikes; the dikes vary from <2 em to 3 m thick. They have sharp, irregular boundaries and bulbous, ovoid and sheet-like shapes (Fig. 9). The entire sedimentary section is deformed here. Away from the contact with the tuff breccia these units are flat, but here they dip beneath the tuff ring at angles of 30 to 40 degrees. Similar features are visible in finely-lamnated diatomite beds under the 7-mile Ridge Comple x (see map Fig. 2). It isn't known i f the tilting is due to subsidence around the crater or to loading by the massive tuff-breccia. The plastically deformed sediments, many of which retain original bedding features within dikes, suggest that the sediments were water -saturated when buried by the overlying massive hyaloclastic breccias. The unstabl foundation upon whi c h he uff-ring compl ex rests may also account for s e large-scale deformation of h yaloclastic deposits o n its eastern flank. Broad anticlin sand synclines within the tuff ring d posits may in some instances, be caused by slumping. This is especially tr1..1e for vent 2 a broad ring with deposits up to 120 m thick. The bluff immediate y north of Stop A s part of vent 6 a small tuff ring about 240 m in d ameter with deposits 25 30 m thic k The outer slopes o f the ring have been eroded, leaving cliff s on all sides e xcept the S W edge, where its deposits lap onto tuff ring 2 Typically, the eposi s consist of well-bedded, yellow-brown sid eromel ane lapilli-tuff in 1 -60 em thick beds. About five p e rcent o f the deposit consists of angular basal block s Stop 8 Table Rock Compl x A small vent and tuff ring ( ent 8) exposed in cross-section along the cliff cuts h yaloclastic tuff and tuf -breccia o f the tuff ring 2 The vent is about 33 m in diamet r ; the tuff ring remnant has a radius o f 120m. The v ent has v ertical walls and is 1 3 1

PAGE 142

filled with near-v rtical, concentric b s o f yellow-brown em to 2 m thick. Near the cent r of th v nt some of th Some beds can be traced from within th vent into Lh ring 2 10) o f tuff V ertical beds a r in rprete s follows: uring Lh w ning ph s s o f ctivity, vertical vent walls were plastered with bed after b d of coh siv sh until th v nt ws filled. Unlike several other sm<:lll vents simil8r to this one, th r e is n o c ntr 1 cor of massive tuff-breccia. Between Stops Band C there are w ve-cut cliffs; t Stop C th w a v cut f e ture breaches the north end of tuff ring 2 The 1 20 m wide br ach through which th road passes, contains a gravel bar consisting of w 11-round d cobbles a nd p bbles o f palagonite tuff and tuff-breccia eroded from tuff ring 2 Stop C Table Rock Complex Visible here is a steep-walled, U-shaped channel filled with w ll-bedd d tuff. The channel trends and plunges north a way from the approximate center of v ent 2 This is one of 23 U -shaped channels cut into the rim o f vent 2 They rang from 1 to 2 1 m deep and 2 to 30m wide. Several channels are traceable for about 150m, but original lengths are not known The steep sides and U-shape of the channels re similar to those in deposits of maar volcanoes near Rome, Italy (Losacco and Parea, 1969; Mattson and Alvarez, 1973) and K o k o Craters, Hawaii ( fisher, 1977). Channels are interpreted to have been cut by base surges. Infilling beds, which thin and curve up against channel sides with some extending over the sides onto the rim beds of the tuff cone, w ere deposited by base-surge flows and by air-fall. The absence of consolidated palagonit tuff cobbles within the channels indicates that they were cut prior to palagonitization and induration of the tephra East of the channel, along the inside rim of vent 2 the effects of palagonitization on hyaloclastic tuffs are well shown Well-bedded sideromelane tuffs have been hydro thermally altered to massive, featureless orange-brown to dark brown, brittle palagonite tuff. The contact between slightly palagonitized bedded rocks and hydrothermall y palagonitized tuffs is sharp; it is irregular and crosses bedding planes (fig. 11). Only T I 1 8m I J_ Figure 10: Sketch o f the cliff face in which vent 8 Table Rock tuff ring e xposed. Heavy dots outline the ground surfac e prior to the eruption o f vent 8 A rem nant of a small tuff ring around the vent is visible. 132

PAGE 143

Figure 11: Bedding within the Table Rock maar complex is often d estro yed by palagoni tization of the well-bedded tuffs by late-stage hydrothermal activity. The dotted line t races the contact between partly palagonitized and completely palagonitized tuff. the most distinctly graded beds, with coarse lapilli at the base are preserved in the hydrothermally altered zones. Beds a bov e the contact are slightly altered sideromelane tuffs; individual sideromelane pyroclasts consist of a core o f brown glass, rimmed with o r a nge or yellow-orange pal agon i te. Sane pyroclasts, especially smaller ones, are completely altered, but their relict forms are preserved. There are also traces o f zeolite and calcite cement between pyroclasts. Below the h ydrothermal contact, sideromelane pyroclasts are completely altered to palagonite, and the rock is crossed by dessication(?) cracks that break it down into 10 to 40 particles. With the "homogenization" of the rock, grain size differences responsible for visibility of bedding and sedimentary structures are destroyed. The end product is a massive, brittle rock composed of clays, zeqlites, iron oxide and calcite cut by NW-trending toW-tre nding joints (vertical). Massive, altered areas within craters suggests that the alteration is due to reaction of sidercxnelane pyroclasts (basaltic glass) with steam seeping through ejecta. This process is being observed at Surtsey, Iceland (Jakobsson, 1978). Stop D Table Rock Complex Continuing east and then south to the rim of vent 2 then back down to the central crater area, many physical characteristics of maar volcanoes may be seen, including planar graded and reversely graded lapilli-tuffs, cross-bedding ( with current directions uphill, out of crater), bedding plane sags (block sags), and convolute beds. StopE, Table Rock Complex Park vehicles near the base of Table Rock and walk southeast about 500 m to the edge o f a depression about 45 m deep and 360 m in diameter and open to the lake basin on the east side. The area, designated as vent 4 formed a crater in tuffs of vents 1 and 2 with crater walls sloping inward at angles of 30 to 85 degrees. Parallel to and covering the crater walls are well-bedded, steeply dipping layers of lapilli-tuff, with an aggregate hie kness o f 6 0 m ( Fig. 1 2a). Yellow-brown lapill i -tuff and tuff-breccia are present in beds ranging in hickness from 1 mm to 1 m These beds dip steepl y w ithin the ven and may be traced out o f the v en onto the r emnants of a small maar deposit overlying tuffs o f vents 1 and 2 ( Fig. 12b). Wet, sticky h yaloclastite tephra were plastered onto crater walls during the waning phases o f the eruptive activity; many have remained i n place, but some sections have slumped and overturned. Lake sediments consisting of white diatcxnaceous mudstones occur within the crater o f ven 4 These deposits lie nearly 30 m abov e the lak e basin floor and were deposited in a crater lak e apart fr the larger basin. Within the southern part of the crater of vent 4 is a prominent oval column o f tuff ( 18 m high, 90 m long) that may have been the conduit for a small vent (5) ( F i g 13a). The column consists o f concentric, steeply dipping tuff beds that c u t across a nd there fore are younger than the crater lak e sediments o f vent 4 Within 2 e m o f the tuff-lake sedim e n t contact, the flat-lying lak e sediments are brittle, (affected thermally by the h y a 1 o c 1 a s tic t e p h r a o f the con d u i t?) Be s o f s i d e rom e 1 an e t u f f a nd 1 a pill i-t u f f d i p 133

PAGE 144

------,,-----------------------------. (/ I t-<---'00'Ind e x Hap u 20 J.N .. u' Figure 12a: Sketch map of ven areas 4 and 5 Table Roc tuff ring complex The map is represented on the i n dex map (inset) as a black square. An outline of the Tabl Rock uff ring complex is shown in he ind x llldp. Th pres nt crater edge of ven 4 is sho\'tn with a double line. Plain areas are bedded tuffs stippled areas are crater la e s ediments and alluvium diagonal lines represent l ake sedimen s of the For Rock-Chri s tmas L ke Valley B asin. Vent 5 is located at he south edge of the crater of e n 4 Figure 12b: Cross-sections U -U' and V-V' The alternating stipple-line pa tern r ep r e sents bed ded uffs of tuff ring 2 he cant i nuous l ine pa tern represents bedded tuffs of vent 4, alternating line-dash pa erns are era er lake sediments, and doted patterns represent lake sedimen s underl y ing the tuff ring complex 134

PAGE 145

.. T e e N m. 1 S EDS s Figure 13a: Sketch map of vent area 5 Table Rock tuff ring complex. Scalloped lines represent the most pronounced unconformities, with the scallops on the upper beds. The oval, concentric arrangement of the bedding is shown by the arrangement of attitudes. S'lf 1n'elltuffs Figure 13b : Field sketch of the southeastern end of vent area 5 inward from the outer edges of the conduit of vent 5 at angles of 30 to 80 degrees (Fig. 13) The center of the conduit consists of massive tuff breccia containing blocks of lake sediment. As at vent 8 (Stop B) and vent 4, concentric beds with steep dips suggests that cohesive tephra was plastered or1to crater walls during (waning phases?) of explosive activity. The vents may have been progressively clogged with tephra, somewhat analogous to clogging of pipes with grout. Stop F, Table Rock Complex Proceed up the dirt and gravel road to the top of Table Rock Table Rock is an erosional remnant of a tuff cone constructed above lake level on the southern rim of vent 2 At present it is a symmetrical cone about 1530 m in diameter at the base, tapering to a diameter o about 360 m at a height of 360 m above the surrounding plain. The cone is capped with flat-lying basalt which once filled the crater, but erosion has modified the original cone, exposing he once-ponded basalt lava lake (Fig. 14). Dikes extend north and SSE of the crater lake, parallel to the long axis of the tuff ring complex. On the lower flanks of the cone, the rocks are mostly hyaloclastic tuffs; yellowbrown or orange sideromelane and palogonite lapilli-tuff occurs in 1 mm to 2 m thick beds. ear the surrrnit, the uppermost hyaloclastites are overlain by 1.5 to 6 m of massive black or red cinders and bombs from fire-fountaining (Strombolian eruption) that preced d the filling of the crater with lava (Fig. 15). The lava lake is vertically jointed high-alumina basalt (Table 1) Blocks of the lava hav slumped toward the east, leaving 2-3 m high scarps along NNW-trending frac-tures. Why is the shape of Table Rock tuff cone different from the broad, low rnaars of most of the complex? A possible answer is the depth of explosive steam generation where 135

PAGE 146

T 6Jl 1 -r-----------------------------G?-AY, APHANITIC B AS.4LT TACHYLITE CI NDERS AND SPATTER -----------------SIDEROMELAN E TIJFF ; EAKE D )JAR CONTACT GRAY T FF PARTLY P ALAGON ITIZED SIDER 0 -1-!ELANE TUF F Figure 14: Section of tuff exposed below the lava lake on the west and south side of the Table Rock tuff cone. A zone of tuffs ( sideromelane) immediately belo\o.' I Jhe former lava lake is not altered to palagonite. Tuffs elsewhere in the tuff cone are partl y alterE:d to palagonite, implying that dehydration of the tuffs immediately below the lava lake decreases the instability and resistance to weathering. ascending magma interacted with water. The broad, low tuff rings may have resulted from shallow phreatanagmatic explosions where magma nearly reached lake level. Shallow steam e xplosions would likely produce broad explosion craters, low fragment trajectories and base surges resulting in broad, low rim deposits of the tuff rings. The base of the Table Rock tuff cone is, however, located on the rim o f vent 2 above former lake level. About 210 m below the cone base is a highl y permeable aquifer within basalt flows whi ch lies beneath essentially impermeable diatomite beds. In contrast with shallow explosions, phreatomagmatic explosions within the aquifer may have been guided by the conduit, resulting in higher angle trajectories and a higher, steep-sided con e There is an e xcellent view of the en tire complex and the Fort Rock-Christmas Lake Valley basin from the top of Table Rock: SouthThe edge of the basin, consisting of normally faulted Picture Rock Basalt of Pliocene age (Hampton, 1964). Silver Lake is located within a graben that is the k eystone of a broad arch. The Table Rock Complex is situated along one of these normal faults, but not cut by it. Also visible are wave -cut terraces and beach deposits that border the basin at an elevation of 1336 m The lake basin is about 64 km long and 40 km wid e with nl.ITierous islands such as Hayes Butte and the Connley Hills, NW of Table Rock The lak e probabl y e xisted f rom mid(?) Pliocene to late Pleistocene time. During that period, eruptions of basal tic magma along normal faults that cut across the basin produced maars within and immediately adjacent to the lake and cinder cones and flows beyond the lake margins. East13 km east of. Tab l e Rock and overlapping the basin margin is Seven-Mile Ridge, a NW-trending group of five overlapping maars. The complex is 12 km long and 3 2 to 4 8 km wide. The best preserved maars are at the basin edge above the former lake level; they drape over a 61 m high fault scarp. The two northernmost maars have been eroded by wave action to flat-topped mesas, 9 to 18 m high. Stop G table Rock Complex This stop, at the southern end of the table Rock Complex, is best reached by dirt roads fran near the junction of State Highway 31 and the southern route to the village of Christmas Valley. Vent area 9, a small, tuff-filled conduit similar in size to vent 5 crops out as a 6 to 9 m high knoll at the edge of a wave-cut terrace. It is about 30 m 136 .,

PAGE 147

Table 1. Chern ical analyses of basalts from the Fort RockChristmas Lake Valley Basin. X-ray fluorescence analyses b y G Heiken. Sample No. 2 3 4 Si02 50. 54 50. 34 50. 67 52. 39 Al2o3 16 18 15 67 17 28 16 02 FeO (total) 10.03 10 .82 10.87 8 49 MgO 6.63 8 08 7 05 6.48 CaO 9 .90 10.33 10. 12 9 .73 Na2o 2 70 3 09 2 42 3 33 K 2 0 0 .59 0 .39 0 39 0.86 H 2 0 + 0.2 0 0 0 0 0.8 H 2 0 -0.2 0.3 0.6 0.6 Ti02 1. 37 1. 68 1. 37 1. 2 3 MnO 0.19 0.19 0.19 0.16 Total 98.53 100 9 100. 95 100 09 1. Basalt flow in the lava lake of tuff cone 3 Table Rock tuff ring complex. 2 Lava Lake in the crater of the north tuff ring, Table Mountain. 3. Basalt fran a dike which crosses the largest tuff ring in the Seven-Mil e Ridge tuff ring complex. 4. Paho ho e flow f rom Lava Mountain. w ' TUFFS OF tlll11tluntu u o n Figure 15: C ross-section, Table Rock. 137 E I::

PAGE 148

P..assive / / I / / I J Figure 16a : Sketch of Jent 9 arrangemen t of tuff beds arou n d Table Rock tuff ring complex sho\ J ing t h cone n ric he outer edges and massive tuff in lhe c n r Stop G F i gure 16b: Sketch map of vent 9 Do ted lines r present are actually less t han a centimeter to several meters hick 30 meters long 138 he geoemetry of beds \vhich The scale on he right is

PAGE 149

in diameter. Tuff beds, 1 em to 75 em thick, are concentric and are plastered onto vent walls (walls now unexposed). The core of the vent is massive, palagonitize tuff (fig. 16) stop H Table Rock Complex About 4 50 m west o f stop G on the so u the a s t e r n fl an k o f v en t 1 on 1 y two r i d g e s remain as erosional remnants of vent 1 These ridges rise about 95 m above the lake basin floor. Originally thjs maar may have been nearly 3 km in diameter. The ring is composed of palagonitized sideromelane tuff, lapilli-tuff and tuff-breccia. The rocks form brownish-gray deposits on the outer flanks and ridge tops and orange-bro w n deposited within the crater. On the flanks, the tuff beds are in uniform 1 to 2m thick layers. Within the crater, ho wever, hydrothermal activity has completely palagonitized and "homogenized'' the tuff as is the case in vent 2 Deposits in some parts of ring 1 are deformed by slumping and are characterized by bedding. The 61 m thick section of tuffs at Stop Hare folded into a steep, overturned anticline in sharp contact with underlying, undeformed beds ( fig. 7). The glide plane for this slump is within the tuff sequence that dips outward (SE) from the vent at an angle o f 6 degrees. Figure 7a: Slumps ructure in bedded tuff on he east edge of Tuff Ring S op H able Rock. Figure 17b: (Below) Diagrams illus rating he his ory of even s leading to th tion of he slump illustra ed in Fig 1Gb. E::p:anat:.on: Solid lines represent tuff. DasJ-,ed line s r present l ake sediment. b. Forma ion of anti line-like stru lure typical of tuff rings by deposition of tuffs cruption(s). c. Mas es of tuff slum p 1nto the crat r. d. Uns tuff rem ining on the outer o f the r1ng slides the southeast a long a :-1 etent det c h hori::on. 139 forma-

PAGE 150

END O F FIELD TRIP Return to Bend via Oregon Highway 31. For more information on the r gion, see the enclosed references. REFERENCES Fisher, R V., 19 77 Erosion by volcanic base-surge density currents: U-shape d channels. Geol. Soc. Amer. Bull. 88, 1287-1297 Hampton, E. R., 1964 Geologic factors that control the occurre nce and availability o f ground water in the Fort Roc k basin, Lake County, O regon. U.S. G ol. Surv. Prof. Paper 383B B1B29 Heiken, G H., 1971 Tuff rings: examples from the Fort Rock Christmas Lake Valle y basin, south-central Oregon. Jour. Geophys. Res., 76, 5615-562 6 Jakobsson, S P., 1978 Environmental factors controlling the palagonitization of the Surtsey tephra, Iceland. Bull. Geol. Soc. Denmark, 27, 91-105. Lorenz, V., 1970 Some aspects of the eruption mechanism of the Big Hole maar central Oregon. Geol. Soc. Amer. Bull. 81B 1823-1830. Lorenz, V., 1971 An investigation of volcanic depressions. Part IV. Origin o f Hole-in-the-ground, a maar in central Oregon. NASA Progress Report, NGR38-003-012, 113 p Lusacco, U and Parea, G C., 1969 Saggio di un atlante di strutture sedimentarie e post-sedimentarie oss ervate nelle piroclastiti del Lazio. Atti Soc. Nat. e Matern. di Modena, 94, 30 p. MacLeod N S., Walker, G. W., and McKee, E H., 1976 Geothermal significance o f eastward increase in age of upper Cenozoic rhyolite dcxnes in southeastern Oregon. in Proc., Second United Nations S ymposium on the Development and Use of Geothermal Resources, 465-474. Mattson, P. A. and A lvarez, W., 1973 Base surge deposits in Pleistocene volcanic ash near Rome. Bull. Vol canol., 37, 553-572. Peterson, N V. and Groh, E. A., 1961 Hole-in-the-ground, central Oregon Meteorite Crater or volcanic explosion? The Ore Bin, 25, 73 -88. Peterson, N. V. and Groh, E. A., 1963, Maars of south-central O regon. The Ore Bin 25 73-89. 140

PAGE 151

... ROADLOG FOR FI LD TRIP TO MEDICI E LAKE HIGHLAND Julie M. Donnelly-Nolan, U.S Geological Survey, MS 18, 345 M iddlefield Rd., Menlo Park, CA 94025 Eugene V. cancanell., Cascadia Exploration Corporati on, 3358 Aposto l Rd. Escondido, CA 92025 John C. E ichelberger, Geol og ical Research G-6, Los Alamos Scientific Laboratory, Los A l amos, NM 87545 Jon H. Fjnk, GeoJogy Department, Stanford University, Stanford, CA 94305 Grant Heiken, Geological Research G-6, Los Alamos Scientific Los Alamos, NM 87545* MILES (intervaJ mileage in parentheses) 0 Leave Cimarron Mote l 3060 So. 6th St. Klamath Falls, and head east 3 1 At "Y" i n road, keep right to avoid going toward Winnemucca Cont i nue south on Oregon H i ghway 39. (14. 0 ) 17.1 (2.0) 19. 1 ( 1. 9 ) 21.0 (12.8) 33. 8 (3.3) 37.1 H i ghway turns east at Merrill. Pass through town and continue east to Malone Road Turn r ight. Proceed south to State Line Rd. Jog left to State L i ne Liquors, then r ight, and continue south toward F ish and Wildlife Headquarters. Continue south a l ong fault scarp toward Lava Beds National Monument Turn left at "T." STOP 1 Captain Jack's Stronghold. Geologically, this stop illustrates f l ow-end features in a high-alumina basalt flow. The stronghold is one of the major features at Lava Beds National Monument, and Aaron Waters has contributed h i s gee-historical study of the Stronghol d to the published gu idebook. One of the Park Rangers, Patti Easterla ( whose grandfather was born in the Stronghold during the Modoc War) has agreed to gu i de us through thi s natural fortress and explai n how so few Indians were ab e to hold off so many U.S soldiers for so long The unusua topography here, which was so critical to the Modoc s defense, is explained by Wa ers (this volume) Drive west from Captain Jack' s Stronghold, facing d irect y toward the north-south faul scarp known as Gillem's Bluff. Pre-Highland lavas are exposed jn he scarp, which is one of several stepping up to the west. As the road turns south, note he Big Horn Sheep Enclosure where some 3540 sheep currently 1 ve. The Park Service is attempting to reintroduce the sheep which were once nati ve but were killed off around 900. Authors are listed alphabetcally except f irst au hor. See A cknowledgements for further crE-dits. 141

PAGE 152

Dorris 0 Miles N 5 0 5 10 Kilometers Klamath Falls @ @). 142 Location map for f i e l d tri p to Medi c ine Lake Highland Q stop numbers OREGON CALIFORNIA

PAGE 153

(12. 9 ) 5 0 0 (1. 7 ) 51.7 ( 0 .2) 51.9 (0.3) 5 2 2 (2 8 ) 55. 0 (4 6 5 7 6 ( 6 3) 6 3.9 ( 0 1 A s yo u leav e the f a u l t scarp and hea d southeast, cross the D e vil' s Homestead f low whi ch is very y ou n g i n appearanc e It apparentl y vented a t F l eene C h imne y s spatte r con e s sittng astri de the s outhw ard c o ntinuation o f the Gillem's B luff fau t s c arp. V i s itor C e nter Turn lef t ont o Va entine C av e Road STOP 2 Valenti ne Cave This i s one of nearl y 300 l ava tube caves i n Lava Beds Nat ional Monument. Many o f t h e more important and accessib l e caves h ave be e n mapped i n d etail by Aaron Waters a n d h i s f i e l d assistants, at the request of the Nat i ona l Park Service. Waters has a lso written d e s c r ipti on s to accompany the maps, one of whi ch describes Valenti ne Cave (Waters, 1976 ) "Valenti ne is an interesti ng, c ean and vari ed cave. Here one can see a Java-tube system which i s almost undamaged by post-lava collapse. The cave shows most of the features to be found i n l ava tubes: among them pahoehoe floors, lava pools, and l ava cascades; well-devel oped and dripstone on ceilings and walls; extraordinarily interesti ng lava benches, one of whi ch marks a h ighl eve l stand of viscous lava that attempted to crust outward from the walls; and still another k ind of bench made by the penetration and "bull doz i ng" away of collapse blocks and rubbl e as molten l ava under hydraulic pressure forced its w ay through the tube. Large pillars ar ound whi ch the lava stream divi ded and reunited are present i n the upper part of the cave. The central part of the cave shows "nteresting l ava falls and cascades through whi ch the lava stream was transferred from a h igher level to a l ower level Downstream from thi s area of breakdowns, the Valenti ne tube subdivides into distributaries whi ch are gradually p hased out downstrea m by filling w ith lav a." Return to mai n roa d and turn r ight. Drive past V i s itor Center aga i n and turn left onto d irt road to Medicine Lake Mammoth Crater (opti ona l stop) P i t crater at s ummit o f broad, low, basal t i c s h i e l d Sou r c e of many of the lava s i n the Monument, probabl y includi ng the f low that made the Stronghol d (Waters, thi s volume. ) Conti nue south, f irst on paved and t h e n on d irt road. Turn sharp r ight and dri ve one -tenth of a mil e along west s i de of spatter cones. 64. 0 Park and wal k north a l ong spatter cones, then up onto flow for (0.1) 64. 1 ( 8) 65. 9 There are numerous ree molds in the area immediatel y north of t he last of the severa spa ter cones that a r e aligned appro x imately north-south. The lav a was a pparen ly v ery fluid when i t f i r s t emerged, flowing around the rees i n i s p a h M a ny o f he molds are large enough to c limb down into. Most a r e 1 0-1 5 fee t de e p a few h a d snow at t he bottom all summer About 200 feet north o f he last spatter cone the f low has a r ougher surface and there a ppear to b e n o m o r e molds. Return o main r oa d and turn r ight; conti nue south. Turn right o n good d irt road 143

PAGE 154

( l. 0) 66.9 (0 .8) 67.7 (0.2) 67. 9 (0 3) 68. 2 (2.1) 70 3 (2 4) 72. 7 ( 1. 4) 74. 1 (0 3) 74. 4 (0 .3) 74. 7 (1. 5) 76. 2 (7. 2) 83. 4 STOP 4 Cracks. Park and hike northwest about 200 feet to narrow d]rt track. Turn 1 ft and wa k about half a mile jnto an area of recent cracks. The larg stone s 30 to 40 feet deep and ha snow at the bottom. The exposed s quenc is pumice over scoria. bo h ove lying platy andesite that is well-displayed in the ragged edges of the cracks. Mos cracks trend north to northeast her with the west side down where offset can be determj ned. This is the same offset d"splayed at the next part of Stop 4. Fau ed c i nde cone ------Return to d"rt road track and wa k northwest to pumice-covered cinder cone Climb up on cone and stand on east side of fault scarp. The scarp is red and bare of pumice, thus fault movement followed deposjtion of the pumice. The fault trends north-northeast and at each end is covered by a very young rhyo i tic f ow. There is an excellent view to the northwest from this spot. Eat lunch and then proceed north along the fault to Crater Glass Climb up on the flow and examine the spnes on the surface. Note the scarcity of pumice on top of the flow despite the abuncance of pumice surrounding it. Return to cars and proceed northeast to junction with main road. Park for STOP 5 Glacial Striations. Andesite is polished and grooved in a northwesterly direction. See Anderson (1941) for a description of the glaciation of the H"ghland. Return to cars and proceed northeast to junction with main road. Turn right and drive south over rim of caldera as described by Anderson (1941). V iew of Medc i ne Lake basin and the young Medicine Lake flow (dacite). Fo low the road down and south past the dacite flow. At the Ranger Station, turn right. Proceed along mai n road, keeping to the r ight at next intersection w ith a good road. Turnoff to Little Mt. Hoffman Keep right. At Ljttle Glass Mounta i n, turn left and follow east edge of flow STOP 6 Park for f irst Little Glass Mountain stop. See accompan ing article F ink for description. This stop offers an oppor n i y o erv he d iffer nt units which make up the flow stratigraphy. First we note the layer of tephra which covers the ground surrounding the flow Isopach maps indicate that the source of this tephra is under L ittl e Glass Mountain (Heiken, 1978 Return to junction and turn left, driving around north side of Little Glass Mountain. STOP 7 Park for second Little G lass Mountain stop. See accompan ing article by J H Fink for descrption. Turn around and return to Medicine Lake Medici ne Campground 144

PAGE 155

AY 6 88 9 ( 2 ) 93. ( 9 7 02. 8 (0.6) 03. 4 (0.8) 04. 2 ( 2 ) 106. 6 ( 0 4 07. 0 ( 0 2 ) 07. 2 r::p_ .. ',... e a .; ::c :o.. "" ... :ef". t.e o a nc r _ef _ass M o _a::., e an mos as a and andes.:. e. Roac u c 0 T r ef Sharp ef r on 0 roac ( s i gn pon s o Lava Beds M 0 mi. Road u c : 0 Keep .J.ef rn ef on 0 -covered road. Road u ct'on. Keep rgh Tur sharp ef 0 0 arrow roa runn ng para e 0 f ow fron STOP 8. Park and wa k up bu ozed road on o f ow. This ocation was chosen speci y o show he mixng of magmas i nvo_ved n he geness of he G ass Moun an dac e -rhyo e f4ow (An erson, 933 Eichelberger, 974, 975 ) The north quarry ste affords an opportu i y o see he ex remes n ava ypes produced during a s g e erup on of e G ass Mountai n vo cano. r ee. vents a o g a northwes -southeas trend g issure were active during h's eruption which occurred within he _ast hou and years. Nine he nor h wes s and the sou heastern-mos vent produced domes, while lavas o f he e ven i g h ee ve s coa esced o form he G ass M oun ai f ow, km3 me. T e f o w s af th ee daci ic eas ern lobes which rade d over a n by rhyo e obes. W e are now a h where the of the rhyo ite obes stopped as it o v he a dac e obe The dacite o e f owed a 6 km ven our present loca ion s 3 1 / 2 km ven hern f low ven probably a so fed his lobe and the f o w ven may have con ributed ava as we 1 The rhyo i e lobe came from he nor hern f l o w ven 3 km is ant. Howeve here was no he eruption between the w o he lobes merge i n a con nuous surface west of h vents. or is any discontinuity in i ho ogy s i nce the aci e obe g ades ups ream o rhyoli e at the middle an d sou h v n s Vara io i n i hol ogy w ithi n o es occurs over a horter 145

PAGE 156

( l. 4) 108.6 ( 3 9) 112. 5 (9 8 ) 122. 3 ( 0 9 ) 123. 2 dis ance norma o s ream f ow, eflec ing the s rong velo i y gradien i n hat directon durng emp acem nt. H nee young r," nd h for mor rhyol ic, a v a occurs n the nner portion of each lobe Contrast be w een he ol er and youn app r n t i n flow morphology as we as itho_ og y The rhyoli e i m thicker than th dac i e obe on the sam e s o pe The dacte vare f om dense, massive 1 va w ith some f o w banding to ves cu ar, scori ceous, n brecciated rna ria par icu ar y abundan o n he flow surface. The dacite a l ways contains abundant basa tic xeno ths nd a larg porion o f th phenocry s within he dacte were fro m d i aggrega on of h se xenoli hs. The rhyolite varies from dense obsdan to very ight pumice whi ch is confin d to th surface. Obsidan has been forced upward hrough the surf ce pumice i n pressure ridges. Interiors of bo h he daci e and rhyolit lobes ar massiv e lav a Samples o f the basal tic xenol.ths ha v e a densi y of 2 15 .04 g/cm 3 ndependent of the vesicu arity or composi ion of the host lav a The mass density of the xenoliths (i. e density w i hout vescles) is 2 79 .01 g/cm3. Dens i t ies of non vesicular host a vas a r e as follow s : obsidian near midd l e vent--2 38 0 0 g/cm3 dacite at this stop--2 .49 g/cm3, dac ite at d istal end of this l obe--2 .55 g/cm 3 Development of h i s compositiona ly zoned f ow can be view ed as a consequence of "bubbling" o f basalt c magma hrough a rhyol ic magma cha m er. Bul k composition and phase assemblage of he dacite are attributable to addi i on of basaltic material o rhyol.tic magma. The rounded shape o f the xenoliths and d istinctive "q ue n ch textu e indicate that basa t was introduced into the chamber as a liquid, was chi led aga i ns rhyol.tic magma, and consequently underwent rapj d crysta. zation Vesicula ion accom panying crys allization produced a layer of low-density foam a he ase of the rhyolitic magma body Thi s foam was removed by f o a jon as i formed and became concentrated at the roof of the c hamber. The volume increase due to jnf ation of he foam triggered he erup on which first tapped the foam -rich hybrj d dacite at the top o f the cham er. Eruptj ons of similar avas prior to he G ass Mountain even suggest hat the chamber i s a largevolume, long i ved feature beneath he High a n d Known densi t ies of the materials evdence that flo ation occurred, and reasonabl e estimates of water content of basaltic magma cons ran the depth of his chamber to wjth i n about 1 0 km o f he surface. The road onto the f l o w provides easy access to the surface of the rhyoli e lobe. The climb takes 152 0 minutes and provides view s of he summt dome which was extruded from the north f low vent, the dstal e n d of he dac ite lobe, and many other interesti ng features of Glass Mountain and its environs. Drive east on narrow road back to mai n road. Turn right. Junction w jth paved road. Turn sharp right and return toward Medicine Lake Turnoff to Glass Mounta i n Keep left. STOP 9 (opti onal). A t second turnoff to Glass Mountain, park at intersection and hike north i n to explosion crater. This i s one i n a northeast-trendng series of explosion craters whi ch e xtend for about three miles. These craters e xpose the local stratigraphy which here is andesite over porphyriti c Lake Basalt. Conti nue on main gravelled road ke e ping f irst right, then left, at intersecti ons w ith other good roads, to four-w ay intersection. 146

PAGE 157

(3. 4) 126. 6 ( 6 .1) 132. 7 (0 7) 133. 4 (0 6) 134. 0 ( l. 4) 35. 4 ( 2 .l) 137. 5 ( 1. ) 138.6 ( 0 8) 139. 4 (0.9) 140. 3 ( 2 .1) 142. 4 (0. 8 153. 2 (0 9) 54. 1 Turn right toward Medicine Lake. Continue past Medicine Lake campgrounds and Ranger Stat1o n w st toward Little G lass Mountain. Little Mt. Hoffman turnoff. Turn left. STOP 1 0 L"tt1e Mt. Hoffman Lunch and panoramic view (weather permitting) of Mt. Shasta, Lassen Peak, and Mt. McLaughlin, with Little Glass Mountain directly below us. Bottom o f Little Mt. Hoffman Turn left at junction At Littl e Glass Mountain, keep right. Turn left at mai n (Harris Spring) road. Road to Lost Spring and Tennant. Turn right. At junction w ith road to Tennant, continue strajght ahead toward Lost Spring. STOP 11. Andesite Tuff Ths is nearly the southernmost mapped exposure of this ash-flow tuff. Here i t is nonwelded to partjally welded and you can see both the unoxidized buff-colored matrix having dark gray pumices, and the reddish to reddish-brown matri x co lor which is more typical of this widespread and distinctive unit. It extends at least 200 feet uphill to the west and is nonwe l ded to partially welded throughout. According to mapping by Hughes (1974), the tuff is exposed a mile and a half to the north at at elevation of 6400 feet, its highest known occurrence. It is also mapped by Weaver (1975) four miles to the southwest. Hart (1975) mapped large areas of it in the northeastern quarter of the Bray 15' quadrangle, just west of the Medicine Lake quadrangle. Hughes, Weaver, and Hart were students of S A. Mertzman at Franklin and Marshall College, and some of their work is summarized in Mertzman (1977) The thickest apparent exposure of the tuff is right here, but its poorly-w e l ded character suggests that the tuff may simply be coating the surface of the h1llside to the west To the north, the tuff is exposed on he south side of Mt. Dome, a mos 20 miles from here. To the northeast, we have mapped it east of the Callahan flow at about 5800 feet in elevation. In the areas to the north and east, i i s a very thn unit, always less than 20 feet thick and typically 2 5 feet thick. The source of the tuff is unknown but may be burjed under younger avas to he nor hwest of Medicine Lake If the tuff's average hickness is taken to be 20 feet and its areal extent to be 250 square miles, jts total erupted volume i s one cubic mile. Although its volume js m)nor i is very important as a stratigraphic marker. orve a ten h of a mile fur her down the road turn around, and return to Harris Spring road. Turn left. Proceed north on good road pas ca tle guard. Turn right on nex good dirt road. Follow oad east then south just pas junction. Park. 147

PAGE 158

(0.9) 155. 0 STOP 12. Andes ite tuff and Porphyrj ic oJjvne b a aJt overl s ndesjte and und rlie and site tuff. W a k up on top of tuff to g t v i w south toward th Highland. D jrectJy to he sou h near Dock Well, 5 mils from h re the tuff overlie these same two units as w e J two other andesi s nd two silicic domes. Anderson (1941) thought th t the domes had intruded into two -hundred-foot thjckn ss of and s e tuff, but detailed mapping show that the tuff flowed around the domes, lapping up onto the ea tern one. Th tuff is actual y very thi n, and it appears unstra if"ed ecause i ha flow d down a steep slope and because mo t blocks are v i ewed from the top. Samples col ected near Dock Well y ield a norma magnet i c polarity. Further to the east, but west of the Callahan flow, the tuff is expo ed in patches as high as 5600 feet. It overles an olivine andesite, two ba alts, and an o lder rhyolite flow. East of the Callahan flow i t overlies a glassy porphyritic dac ite and an andeste, as we 1 as an older sequenc of aphyric rhyolite and dacite on top of porphyrit c pheruliti c rhyolite. The pre-tuff sequence i n each case s complex, w ith al rock types from basa t through rhyoljte represented. The post-tuff sequence consists most y of silicic andesites pil ed up around Medic i ne Lake and young basalts on the flanks of the volcano. The youngest eruptjons are b imoda w ith basalts lower and rhyolites h igher on the volcano. Turn around and return to pav ed road. Turn r ight. Abou 4 l/2 miles further north, there s a good p aved road to he right. Keep left and proceed northwest into Red Rock Valley, then west and northwest to Macdoel and Highway 97. Turn right and go north through Dorri s toward Klamath Falls. For those returning to the Klamath Falls airport, turn r ight at Joe Wright Road (sign points to K i ngs ey F ield which i s the K lamath Falls airport). Acknowledgements Donnelly-Nol an was principall y responsible for the road l og, but Ciancane li, F i nk, a n d Heiken provided va l uable assistance in choosing fiel d trip stops. Eichelberger wrote the description of Glass Mounta i n (Stop 8), cancanelli measured the temperature at the Hot Spot. L Hose provided valuable assistance w ith f"eld tri p logstics and a lso discovered and mapped the exposures of andesite tuff east of th Callahan f low. We wish to thank P. Easterla of the staff at Lava Beds Nat ional M o nument f o r gu i ding us through Captain Jack's Stronghold and A. C Waters for contributing the unpu lished description of Valenti ne Cave We also thank the U S Forest Service and the U S. Park Service for their cooperation and assistance. References Anderson, C A., 1933, V o canc history of Gla s Moun ain, nor hern CaJ" fornia: American Journal of Science, vol 26, p 485 506 1941, Volcanoes of the Medicine Lake High and, California: Universi y of California Publications, Bulletin of the Department of Geologica Sciences vol 25, no. 7, p 347 422 Eichelberger, J. C., 1974, Magma contamination within the vo canic pile: origin of andesite and dacite: Geol ogy, vo l 2, p 2933 ______ 1975, Ori gin of andesite and dacite: ev idence of m i xing at GJass Mountan i n Californi a and at other circum-Pacifi c volcanoes: Geol og ical Soc iety of America Bulletin, vol. 86, p 13811391. Hart, W. K 1975, Geology of the northern one -half of the Bray Quadr angle: map to 148 -

PAGE 159

accompany unpublished B .A. thesis, Frankljn and Marsha l l College, scale 1:62,500. Hei ken, G 1978, PJin i an -type eruptions i n the Medic ine Lake High land, California, and the nature of the under.yng magma: Journal of Volcanology and Geothermal Research, vol 4, p. 375-402 Hughes, J., 1974, GeoJogy of the western Medicine Lake Highlands and Garner Mt. Area, map to accompany unpublished B.A. thess, Franklin and Marshall College, scale 1 :62,500. Mertzman, S. A., Jr. 1977, The petrology and geochemistry of the Medicine Lake VoJcano, California: Contributions to Mineralogy and Petrol ogy, v. 62, p. 221247 Waters, A. C., i n press, Captain Jack' s Stronghold: this volume ____ 1976, Valentine Cave : National Park Service unpublished document. Weaver, S 1975, Geology of the southeast Bray quadrangle, California: map to accom pany unpubljshed B A thesis, Frankl"n and Marshall College, scale 1:62,500. 149

PAGE 160


PAGE 161

CAPTAI JACK'S STRONGHOLD ( T h e Geolog i c Events that Create d a a t ural F o rtress) Aar o n C. Waters 308 Moore Street, Santa Cruz, Californi a 95060 Map b y David K i mbrough and Aaron C. Wate r s n t roduc tion .... reached by a side road, the v isitor comes w c aptai n Jack's stronghold, an area of natural rock t r e n c hes and rock shelters i n the lava flows, where a tiny band of Indians he l d hundreds of troops at bay f o r a mos t 5 months..... The record tells the tale o f a motley arm y of badl y trained sold iers, led by inept o fficers on a battlefield of the Indian's choosing about which h e army inte lligence had a bsolute y n o i nformation. G iven such condi tions, trage d y was inevitable for settle r s on t he north shore of TuleLake and for Jack's and Hooker J im's bands o f Modocs (Keith A. Murray, i n preface to Erwi n R Thompson, 1971 !iodoc War: Argus Books, sacramento, Calif., 188 p plus appendices) L a s t o f the Californi a Indi an uprisings, the M odoc war ( November 29, 1872 t o June 4, 1873) ha s been c h r o nicled by many newspape r writers, istorians, soci a l scientists and othe r s The writ ten record, however, is blurred and contradictory as to the causes mot i v es, heroism, and savagery of principa l pa t i c ipants on b oth sides. Parts of the record hav e been g reatly fict ionalized and roma n t i c ized, and s _ome even fal s i f ied as Jeff C Riddle scornfu ly notes: "I have read Cap William T. D rannan's 'Thirty Yea r s on the P l a ins,' where he wrote aoout t h e Modoc warri o r s According to what he says, h e captur ed and k illed more Modoc warriors than Capt. ack r eally had when he corranenced fighting. Jack had only f ifty-two warriors in a ll. I knew every one of them. ("Jeff C. R iddle, 191 4 (reprinted 1974) The Ind ian History of the Modoc War, Urion Press, Eugene, O regon, 295 p ) Our report and accompanying map makes no attempt "to set the record straight" w ith regard to what has been r eported about the histori c and sociological roots o f the Modoc War. Inste ad we investigate the ques ion repea dly asked by almos e v e r y wr i er: was i possibl ha 53 Modoc men, aided (or encumbered ?) by wic as many w o men and children, w i hs ood a sieg through h dead of win er, rou d 3!JO u s roops e ngaged i n h e first major assa u l wi hdre w und ec ed after repulsing secon d assault of 650 t oops suppor ed by mortars and h o w i zers, and hen only a few d ays later s aged a successful ambush and inflic ed 25 fa ali ies upon a patrol c ompos e d o f 59 en li s ed m n a n d 6 of f i c r s? One part o f he a n s wer o this ques ion is ha Indi n s chos a sup rb, b u by n o m e an s un i qu e natur I for ress i n which to make he i r stand. This for ress, h o w v e is o n l y o n e part of he t otal e q u a ion The v i al po i n is ha th M odocs knew horoughl y n d i n de ail h na ur o f h e terrai n 1 5 1 south of the shore line of Tule Lake; the army was totally i gn orant of this landscape's military advantages. Writers about the Modoc War clearly have not understood the real nature of the terrain in which the Indians holed up any more than did the U.S. troops and their officers. One h istorian writes vaguely about the Modocs "disappearing into the Schonchin flow" as if this barren patch of recent aa lava had some myth ical power t o swa l low up the Indians and h ide them f rom pursuing Army patrols. In fact thi s a lmost treeless expanse of sma l l and loose blocks o f lava would be the worst possible place for the Indians to try to h ide; and the Modocs avoided the Schonchin f low comp letely. Some writers have assumed that the Stronghold is "within the Schonchin f l ow," but the distal end of the Schonchin flow is 3.8 kilometers ( 2.4 miles) south of the Stronghold, and the source of the flow is at the east base of Schonchin Butte, another 6.4 kilometers (4 miles) further south. The Schonchi n f low played no part in the Modoc war, except i t seems possible that a few members of Captai n Jack's band may have staged the Thomas-Wright ambush out of fear of being trapped by this army patrol against the inhospitable west edge o f the Schonchin f l ow. We will f irst describe the terrain in and near the Stronghold, as seen through the eyes of a geol og ist. Then we return to the Modoc War and discuss, i n terms of terrain, the consequences of the first and second assaults by the Army upon the Stronghold, followed by an analysis of how the Modocs were able to withdraw from the Stronghold undetected, and a few days later stage the disastrous Thomas-Wright ambush. The Terrain Today the country in and adjacent to Captain Jack' s Stronghold consists of four kinds of topographic surfaces (see map) l The Tulelake p lain, which in 1872-1873 lay benea t h the waters of Tule Lake. Since 1906 over 3 / 4 o f the a r e a o f thi s large but shallow lake has been recl a imed for farmland. 2 Lowlands, of low relief, underlain by lobes and tongues of solidified lava, border the shoreline both to the east and west of the Stronghold. The lowlands are rough surfaced in places, but the variati on i n height is 2 meters or less. The lava tongues g rade at the shoreline into pillow lavas interspersed w ith sandy beaches of hyaloclastic debris (black glass sand and pillow-like ellipsoids o f basalti c lava, which formed where the molten lava tongues were quenched and granulated by entry into the waters of former Tule Lake).

PAGE 162

DAVID L. KIMBROUGH AND AARON C WA TE R S 1976 SURFACE FEATURE S D D PLATEAU REMNANTS (Flat-topped remoons of o once contonuous 10\ o -llow surface. whoch on othe r ports of the mop has been IONered. broken and tilted onto baSins and schollendomes.l A R E A OF LARGE SCHOLLENDOMES (Tilted and broken solodofled crust of a former lava llow Molten lava. droonong out from beneath ots a lready solldofoed crust, caused collapse of the crust onto a very rough landscape of schollerdomes. small bosons collapse pot s deep crocks, and plies of talus) LOWLA NDS (South and East of former shoreline) (A lowe r lava-flow s urface between the schollendomed area and the former shorelone of Tul e lake Choroctero zed h:'={ shallow bosons low schollendomes. few crocks. FORMER LAKE BED (North and West of forme<" shoreline) (Mostly lake-beds slits. now under cultovotoon except near former shoreline lake bed recloomed tor farmland 1906-1918.) GEOLOGIC FEATtJRES $1 \C ENTRAL CRACK >TERS SCHOLLENDOME 8 0 8 COLLAPSE BASIN COLLAPSE P IT bet wee voews and nile cealed lndoon snoper Many outposts were rebuolt bv t Army after occupotoon. others hove tumbled nto ruons U S ARMY FORT I FICAT IONS G loose blocks of roc D > were poled onto I thiC wall s of voroous shapes Most ore 0 5 to 1 meter hogh The abundant Arrrry forto focot oons shOWn w thon the central area of thos mup were never used on bottl e they were buol t olte the Modocs hod abandoned the Strongold To guard ogoonst a OOSSibl e return attock the Army con structed rock walls. and man small emplacements that oro v oded shelter for 2 to 5 men Some fortoflcot oons outsode the c e n tral area, however. were has oly thrown up by t h e soldoers before and auro n g the Second Assault on the Stronghold Aprol 15-17 1873 I n the dosostrous First Assau l t January 17 t h e troops learned rom the Modocs to value shelter behond rocks SYMBOL S ....<-.__ TR A ILS ( W oth brodge ....... and stoors R OAD STRIKE AND DIP ( W o thon scholl endome areas. shOws slope of land, short li n e ondocotes h e d orectoon of tolt ) TRAIL NUMBERS (Stotoons on the sell-guo dong trool tnrough Coptoo n Jock's Stronghold) II FORMER --SHORELINE MOOOC _d 152 \..-0 G+-..J... 00 ROUG H HUMMOCKY SUR F AC E W I TH MANY SCHOL L ENDOMES LAKE oG+-..J... 0 PLATEAU BEO ..-. :1'? A
PAGE 163

I J ---.. (j I f\__.? I / /o 0 0 ... (<11\1. D s =-0 iJJ) ( ) L /J) PLATEAU P L ATEAU 0 @ 153 \. c ( &. D -ar----. 0 v' C) 0 L OWLANDS 0 c 0 0 D J ... "" \.. -I TECTONIC CRACKS CONTOURS (meters) (near former Lake Shore) --------10---------------5 ---------- --O(Lake Shore on 1873) -------s--------SC"LE 0 20 40 100 200 "ETERS 0 100 200 XlO 400

PAGE 164

3 where he is h map) out from he f ow fron and a so drained out from benea h i s solidified crus in a plexus of underground ava tubes. Therefore the plateau surface in the Stronghold is dimpled w i h smal sag basns and by vertica -wa led col apse pits (see map) Moreover, the plateau edges which were close to the original flow front have "turn-down margins" produced as the plateau surface sagged outward and was riven with deep cracks and fissures as it was rafted forward and let down o a lower elevation while molten materia was escaping from benea h i 4. Schollendomes. A marginal area of large schollendomes is present on three sides of the plateau remnant in the Stronghold area, and it continues to border other p lateau remnants for many miles to the east and southwest. These schollendomes are detached parts of the former plateau crust which have broken into e ongate domes characterized by a deep central clef along their top. Many deep fissures and minor cracks branch from he central cleft. (For description and ske ch of a schollendome see the Explanation accompanying the mapl W e now describe in more detai the geologic nature and origin of each of these four topographic units as they occur within the area of the accompanying map This will give the background and perspective necessary to understand how the topography was effectively used by the Modoc defenders of the Stronghold. Tulelake Plain, Contrasted With the Tule Lake of 1 872 In 1 87 2 the area north of Captain Jack's Stronghold looked very different than it does today. Instead of the grain fields laced with irrigation and drainage ditches, the waters of Tule Lake lapped against a steep slope of crazily tilted ava blocks that formed a schollendome field between the shoreline and a flat-topped plateau 12 to 20 meters above the lake. The change from lake to farms was brought about by diversion of Lost River to irriga e arid lands farther west. Over a period of about 12 years (1906-1 9 1 8 ) the lake shrank to less than l/4 of its former area, and the hyaloclastic sil s and sands on the ake bed, fortified with organic matter from the tule swamps that flourish in shallow water, became rich farm land. The position of the original shoreline i n 1 87 3 and the shape of the present land surface immediately adjacent o it, are shown by 5-meter contours on the accompanying map. Only over a stretch of about 500 meters (l/3 mile) did he shoreline of Tu e Lake come against the belt of large schollendomes hat rim the plateau remnant of he Stronghold. Both east and west the shoreline is against lava lowlands that rise only 0 to 3 meters above the level of the former lake. The site of a non-permanenl Indian village that occupied a rocky ledge just above lake level on the eastern lowland one kilometer (0 6 mile) northeast of the center of he Stronghold (se map) can be identified 154 y h b i r d 0-m vill -In rea above th level s art wi h pla eau nd th n proceed o areas of lower el vation This is the order in which th three opographic uni s (pla eau surface, schollendomed margin, and lowlands), wer e developed geologically. Each repres n s a phase ur ing h clima x and dying ou of a major episode of volcanic ac ivity. Tracing them in ord r of development makes i easi r to und rs and ho w h na ural fortress used by th Modocs was formed. The of h plat au (zippi em area on he op of an unusually hick l va flow whic h his area from he nor h umerous he margin of the pla eau penetra e 10 below its top; this dep h is p rhaps l / 3 thickness of the lava flow at the ime of its greates inflation during the climax of he erup ion The flow is one of a gr a number of lava flow s which spread north and eas from vents in and near Mammoth Crater, located 15 kilom ers (9.4 miles) airline to the south. The molten lava, ho wever, did not travel all this dis ance on he surface of the ground. Ins ead i was transmitted most of the way through underground lava tubes. (For an ana ysis of various parts of the lava-tube sys ems see he maps and reports on display a the Visi or Cen er in Lava B ds ational Monumen ) Small distributary tubes fanning out from one major branch of this intricate system of large lava tubes became active within he flow that forms the plateau at Captain Jack' s Stronghold during the last stages of eruption, as w e show later. The distribution of remnan s of he plateau surface indicates hat this hick flow reached its farthes north extent within the area of Captain Jack' s S ronqhold. From h r irr g 1ar1y 1 flow front trended about South 65 Eas to and beyond Fern Cave; and in the opposite direction it runs approxima ely South 30 Wes for abou 3 kilometers A most points, ho wever, he exact posi ion of he flow fron can b inferred only within a width of about 500 meters, because in almos all places the flow's margin was broken up and partly inunda ed by late movements of mol en lava wi hin the interior of the flow after i s fron and crus had partly solidified. After the flow front stopped moving and a 1-to 15-meter thick crus had congealed on op of the flow, a resurgence of volcanism sent large volume s of molten lava through the long syst m of tubes. Lava from distal branches of these tubes entered the s ill molten interior of the flow, lifting i and putting so much hydraulic stress agains he flow front hat

PAGE 165

> en Ia a brok hrough in num rous places. Dur ing maxi m u m p h a s o f ac vi ty he s i 11 en ava benea h h au crust began to creep oward hese new breaches, se i ng up skeins of flow toward he points of low st stress. During he waning s ages large par s of the flow fron and par s o f he crust of he flow began to break up nder hese stresses and were rafted slowly forward on he bac of the moving currents below. At he base of he flow front small lobes and tongues of mo ten lava escaped to h surface and flowed on to the nortf'l. After eruptive ac ivity a he dis ant vents had ceased the mol ten interior of the flow and the 1 ava f i l ing he underground tubes continued to drain out, eaving many parts of he plateau crus unsupported. sch parts sagg d downward rep acing the space lef b y he departing lava from the interio Large asses of he solidifi d flow fron and finger-like ongues f the crus extending hundreds of me ers os ream from the flow fron sagged down and were forward as the lava beneath them 1 eaked ou 1 ote on the map the many small oval collapse basins, and the long finger-like basins with schollendomes ... hich indent the plateau). wherever he edge of the plateau "turned down" to rep ace the escaping lava beneath, deep tensional cracks and fissures formed in the turned down fJap. n places parts of these fissured margins were rafted away, forming schol1endomes. Nearly all edges of the p ateau remnants became an intricate maze of deep Lssures which greatly impede travel across t is errain. Where shallow dis ributary lava tubes drained ou post-lava col apse of the plateau s urface into the ubes has left numerous vertical -wa led collapse pits and collapse trenches. Inspect he area of the map at the end of the p l ateau remnant where the Modocs took up residence. Each of the fain lines represents a crack or fissure too wide to jump over without extreme caution, and for each line on the map there are many more cracks i n the ground (omitted from the map) These fissures, especially the big ones at or near the top o f the turndown flaps, are the "natural trenches" u sed by the Modocs Note that hey are nearly con i nuous along he edge of he pla eau on three si des of the Modoc encampment These plateau-edge f issures, however only he last line of defense; the Modocs a so used rock clefts on the tops o f schollendomes beyond the plateau margin as sentry outposts. The Schollendomed Margin of the Lava Plateau. The map show s much better than words how this frontal p a r o f h hick flow ha form d h bro e n u p and ransform d i n o a o f s chqllendom s Scholl ndome s crus h and Jac s edge o end o f Only h larg s ind i ca d o n our map. also h urndown basins" 1n o unbrok n par s of he u, ar nd rs h 1 5 5 wi h smalle r schollend omes whi c h range down i n size t o schollen no 1 rg han a sma room and only a me er or w o in heigh Time did not permit s to map the sma ller schollen. Moreover, the cl t er of ines ha would result from plot ing all hese cracks and fissures would have made the map unreadable. The unbroken e xpanses of whi e paper i n this par of he map, however do give a false impression that some of the wider basins in the schollendomed area might be easy to cross. From a distance they a so look easily traversible, but actually they are deceptively cruel to a person who ries to cross them in a hurry. We have used the word hummocky in places on the map to indicate that such surfaces are quite irregu ar, rough with various sized schollendomes, and riven with hundreds of cracks and fissures. Several high schollendomes outside the plateau remnants provide excellent viewpoints over the shoreline and lowland areas. The Modocs developed them into well-camouflaged sniper positions by piling loose fretworks of rock in or around parts of the central crack. Lowlands Built of Pillow Lavas and Hyaloclastic Debris. Numerous outcrops where the lowlands meet the former shoreline of Tule Lake show that lava advancing from the south flowed into the lake. Molten lava changes on contact with a water body into pillow lavas and hyaloclastic deposits. These materials filled in the edges of the original Tule Lake pushing the shoreline northward much like a delta grows at he mouth of a large river. For details of the mechanics of formation of pillow lavas and various kinds of hyaloclastic deposits consult Fuller (1931) and Waters and Fisher (1971} In a quarry developed for road building materials near the site of the former Indian village (see map}, we can e xamine typical e xamples of pillow lavas with chilled glass rinds, pillow breccias, and granulated slag-like bits of hyaloclastic material (basalt glass) -the three kinds of deposits characteristically formed where molten basalt advances into a water body. In nearby areas molten lava also escaped directly into the lake through lava tubes. Good e xamples are on the Hovey Point peninsula wes of the Stronghold, and there are several additiona l vent areas and low maar craters to the east between the St:onghold and Hospital Rock In places molten lava also erupted directly into the lake, building larger underwater volcanoes at The Peninsula, Juniper Butte, and Prisoners Rock -localities several miles to the east and south of the Stronghold. Most of th lowland area probably was built into the lake by he quenching of lavas that are slightly older han he flow that formed the plateau. The a e ongues of lava that broke through the schollendomed front of the plateau flow, however, have also assis ed in pushing the lake front farther to the north. The lowland ar as are easily traversed. A l hough h lava surface is minu ely rough, and is in places diversifi d with small schollendomes about a meter high, h area show s non of the deep fissuring, arg schollendom s or broken areas of talus that make h adjacent scholl ndomed area so difficult to rav rse. Mor over, waves and wind have scattered hyaloclastic sands and silts over the surface of many Jow l nd ares, assuring exc ll n mobility for ravel. Soldi rs in raining on the lowland areas

PAGE 166

near Gillem's Camp or o n the hyalocl stic flats near Hospital Rock, probably had no concept of the ind of terrain they would meet in their assaults on the Stronghold. The Modoc War In late November of 1872, two groups of Modoc Indians were encamped at their winter villages, located about 1/2 mile apart on either side of Lost River a few miles upstream from where it entered Tule Lake. On November 29, a surprise confrontation just at dawn with a patrol of soldiers sent to put the Modocs on the Klamath Indian Reservation ended in a shootout, followed by withdrawal of the Indians from each village. One group, led by Captain Jack, and the other by Hooker Jim, each headed for their Stronghold on the opposite (south) shore of Tule Lake. Jack's group, which also included the women and children from both villages, paddled the 13 miles (21 kilometers) across the lake from the mouth of Lost River during the night of November 29-30. The following day they were joined by Hooker Jim's small group of warriors, who had ridden horses along the 35 mile route around the east end of the lake, and had taken revenge for the fatalities incurred in the early dawn confrontation by killing all men at isolated ranches and settlements along the north and east shore of Tule Lake. In the wake of the revenging Modocs, the widowed women and children of the settlers began the long walk across Stukel Mountain to the security of other white settlements. Men of two entirely different cultures, neither of whom fully comprehended the rights or the motives of the other, were now at war. The Stronghold. Examine the map and note the nature of the natural fortress which was to be the home of the Modocs for the next 5 months. Here the northern tip of the plateau surface overlooks a bay in the south shore of Tule Lake. The part of the plateau closest to the lake is a rounded table approximately 150 meters in diameter -about the size of a modern football stadium. It is bordered on three sides by a field of large schollendomes; to the south a neck about 50 meters wide connects it with a larger remnant of the plateau. Within the area chosen for the Modoc's living quarters the plateau surface is dimpled by 8 collapse pits, each a vertical-walled hole 2 to 15 meters in diameter and 3 to 8 meters deep. They formed where parts of the plateau's crust caved into small underground lava tubes. Floors of these pits are covered with large angular boulders tumbled from the roof ana walls. In places one can burrow around between these boulders, especially beneath the overhanging walls of parts of the pits, and find small chambers each of which would give protection for 1 to 5 individuals against the rain or snow and also from the strong cold winds that sweep the plateau in winter and spring. Three of these collapse pits are easily visited from the inner trail constructed by the Park Service: Captain Jack's Cave, Schonchin John's Cave, and Family Cave (see map) Nearby a small mound of pahoehoe lava served as a rostrum from which the Modoc leaders could address their people. Loose rocks picked up from the surface of the plateau were piled into low dry-walls, forming a partial breastworks around parts of the camp. (These fortifications were later rebuilt into much thicker and higher walls by Army sold'ers after the Modocs had w ithdrawn from the Stronghold). The main defense 1 56 positions used by th Modoc s however, are he dee natural crack s nd crevasses along th top of the turndown edges of the plat au, a nd similar along th tops of high schell ndome s which r1ng tr sid s of he Stronghold. The more strategic and important of these Modo c defense trenches are labe, on the map. Note that the y form a sinuous lineal: the entire northwest margin of th plateau, and tha.' they curve into a natural u-shaped ambush line whic: bars access to the Stronghold from the nearest po1 , of the lakeshore. The floors of these natural defense trenches wer e cleared of loose debris so the defenders could pass rapidly along them. Shor radial routes by which one can walk to various pan s of the trench system from the central Stronghold without encountering difficult crevasses were wel known to the Modoc defenders. Beyond this natural trench system lay Modoc outposts -high isolated overlooks with unimpaiced views of the surrounding country. Most such outposts are located in the central cracks of the highest schollendomes, and were further camouflaged by piling loose fretworks of rock thr which a sniper's rifle could be extended unobserve d : No doubt additional Modoc outposts within the area the map have gone unrecognized during our mapping (see the description of Modoc Fortifications on map). Contrary to the statements of several writers, these natural defense features of the stronghold an not unique. Many other table-like remnants of the lava plateau have even more formidable and deeply-crevassed turndown edges. The unique value the stronghold chosen by the Modocs was its proxia: to the shoreline of Tule Lake. A constant supply 0 : water, and of some food from wocus root, waterfow, fish, and fresh-water clams was thus assured. its location denied communication for an enemy usi; the easily traversible route along the lake shore. Moreover, the Modocs were well aware of an easy escape route to the south over the flat surfaces of scattered remnants of the lava plateau, whereas one unfamiliar with that terrain would flounder painfu: and slowly across the heavily fissured and schollendomed country that surrounds these plateau remnants. Still another unusual topographic feature, a natural cattle corral, was vital in helping the Modocs withstand the winter siege. Just west of e Stronghold encampment is a small and deep collapse basin, bounded on three sides by the steep and heavily crevassed sides of three large na on the fourth {east) side by the s eep and deep: fissured turndown flap of the plateau. Miraculous these bordering fissures along the edge of the plateau die out about 200 meters to the south. Hen a smooth and easily traversed slope leads down off the surface of the plateau and northward through a narrow "gate" across the end of the southern schollendome into the natural cattle corral. Stra y cattle on the southern plains, and others Modoc raids, wete driven north across the plateau remnants and into this natural corral, where they could be securely penned in by piling a wall of rOC1 and brush across the narrow gate (see map). Thus a adequate supply of beef was available throughout winter. Assault of January 17 1893. Meanwhile the Ar; stung by their lack of success in "rounding up" Indians and shocked by the murder of 1 4 settlers the retreating Modocs after the Lost River

PAGE 167

confrontation, wee preparing for a second attempt. Addi tiona troops wer ca led i n from other Army 5t5 and groups of Oregon and ( whos e enlistment per1od was only for 30 d a 5 w e e hastily organized and haphazardly ay d d l trained Also recru1 te were unorgan1ze vo unteers and abor support f near by towns and r By ear y January 400 "fJghting men" were "raring to go. Lieutenant Colonel Frank Wheaton set January 17 for the at tack on the Modoc' s encampment. At that time their natural fortress was unnamed; it soon became known as Captain Jack's stronghold. colonel Wheaton's strategy was a p incers movement from both west and east. Three hundred men were to to battle; 100 held in reserve. Captain Green's cavalry (but dismounted as foot troops) along contingents of Oregon and California vo lunteers to bear the brunt of the fighting and attack from the west. Captain Bernard's smaller command o advance from the east, primarily a s a holding force to keep the Modocs from escaping along the a eshore low lands. On January 1 6 the troops moved into position from thei r training areas. Green's forces marched 1 3 miles across rough country to a on the bluff overlooking the southwest corner of Tule Lake (near Gillem's camp on modern aps). On the same day Bernard's force advanced west over the low ands adjacent to the south shore of Tule Lake, but because of a heavy fog they probed too far, drew the Modoc's fire, and three men were wounded. Bernard's group then withdrew to an area near Hospita l Rock (about 3 kilometers east of the Stronghold ) and camped for the night. The morning of January 1 7 was cold, and a heavy encompassed the area around the lakeshore. Wheaton and Green's troops, although on the march at were slow i n making their way down Gillem's bluff and getti ng organized into a line of 5 irmishers as they advanced toward the Stronghold. o doubt Modoc scouts were fully aware of the troop oovements since early dawn, but it was after ll a .m. Modoc snipers opened fire. The debacle that followed has been chronicled by many; the most detai l ed account of number of men, their positions, and their movements during the "battle" is in Thompson (1971, Chapter 4, p 33-45) After wounding and killing several men the Modoc's relinquished a few of their outposts, thus leading the advanci ng t roops eastward until they were enmeshed in the chaos of deep cracks and crevasses within the schollendomed area Here the Modoc's fire from their natural defense trenches above was accurate and deadly; casualties mounted, and yet not an Indian had been seen by h befuddl d roops. B y midaf ernoon a hought of charging up on o the p lateau was abandoned; some pa r t s of he line were already in retreat leaving heir dead on he field. Captain Green personally led an attempt to round he Stronghold along he lakeshore a nd make contac with Bernard's command on the east. They suffered many casualties; mos retr ated or were killed, but a few en remain d con e aled behind boulders until darkness and then made heir way over to Bernard's position. Cap ain Gr e n w s among them. Bernard' s group had also ben in roubl during he day Casual i s w re inflic ed by unse n Modoc s nip rs. The hough n o as difficult as hat on x nsively schollendomed, crevass s assaul on re rea was reach rous groups of arning by signals that the sid had f iled, a mor orderly 10 p .m. o n h as sid and i 1 57 con inued through the night. The day after the rout the Modocs searched the battlefield and recovered much valuable booty ( Riddle, 1974, p. 56) They found the ground covered with ammuni ion, rifles and other kinds of guns where the Oregon volunteers had stampeded in bad order. In the area where mos t of the casual ties of the soldiers had occurred were 9 carbines and 6 belts filled with carbine cartridges. Also recovered was considerable field equipment, boots and clothing. The outcome of the January 17 assaul t was thus a spectacular victory for the Modocs. Casualties of the Army and the Oregon and California volunteers totaled 37, and 6 of the dead were left on the field. The Modocs had no casualties. In Thompson's words (1971, p. 43 ) "Three hundred men had been unabl e to make the slightest dent on the magnificient union of lava and Indian skills." In many written accounts the heavy fog is blamed for the Army's debacle, but a person thoroughly familiar with the terrain can argue effectively that the fog worked to the advantage of the Army, not the Modocs The Modocs early defense was accomplished entirely by snipers in Modoc outposts. From their secondary natural defense trenches at the top of the plateau, other Modoc defenders could not see through the fog and determine which parts of the Army's line were hung up on fissured ground, parts were advancing, and which were routed and in retreat. Therefore they could not concentrate their limited manpower to the points where it would be most effective -as they did so successfully later in the disastrous Thomas-Wright ambush. The Winter Program of "Gradual Compression". The events between January 18 and April 11, 1873 are well summarized in Murray (1968) and additional detail can be found in Thompson (1971, p. 45-65). The Modoc Shaman (Curley Headed Docter) had assured the warriors that they were invincible, and would not suffer a single fatality. After the easy repulse of the January 17 attack they had every reason to believe him The Army was humiliated. Changes in command took place: Brigadier General E R. S. Canby and Colonel Alvan C. Gillem were placed in charge, and steps were taken to greatly increase the number of soldiers. The Oregon and California volunteers, who had been so eager to do battle prior to the January 17 assault now had their fill of fighting the Modocs; most of them disappeared the moment their 30-day enlistment was up Irresolute and conflicting orders also came from the War Department in Wash i ng on, D.C. A strong feeli ng arose that it might be better to try to negotiate with the Modocs than to engage them in battle. A Peace Commission of 5 members was set up, but rapid changes occurred in its membership, and the Commission did not truly have the power to give answers with regard to the Indian s two chief concerns: l Establishment of a small reservation for the Modocs in the Lava Beds; 2 Escape from the hangman's noose for those who had murdered the Oregon settlers after the Lost River confrontation. Moreover the Indians were growing i ncreasingly concerned with the steady build up of roop s nearer and nearer to their Stronghold. Canby was engaged in a program of "gradual compression." By early April, Gillem's Camp, at the southwest corner of Tule Lake, and only 3 miles from the Stronghold, contained 3 50 soldiers. On April 6 Captai n Mason moved 5 companies of foot soldiers to Hospital Rock, only 2 miles to the east.

PAGE 168

Dissension bout fu ure ac a so arose in th Modoc's cam p F om heir poin he mee ings w i h he Peace Commission had b n compl e y frui less, and immediately af m eting Canby had mov d addi ional troops n o he S ronghold. This program of "gradual compression" after each meeting could only lead o frustration, and a final explosion of he Modocs. One of he M o ocs proposed ha on more P ace m eting be held a which General Canby, A. B M eacham he head of he Peace Commission, and the other members in attendance would be murdered. This audacious idea cap ured the fancy of many of the Modoc warriors. They reasoned that if the General in charge of operations, and the head of the Peace Commission were simultaneously put out of the way the soldiers might w ithdraw. Cap ain Jack kne w better, and vigorously opposed the plan, but som e of his col eagues taunted him as ? coward, pushed him to the ground, and placed women's clothing upon him His manh ood thus called into question, Jack agreed with the plan, and said that he would be the one to murder Canby. This treacherous p an was learned by Toby Riddle ( W inema ) a Modoc woman married to a white man, Frank R i ddle. The R iddles had served as in erpreters during the Peace negotiation s Sh e warned Canby and M eacham of the plan and begged them not to hold another meeting. They did refuse a meeting for the follow ing day, but after Captain Jack formally requested a meeti ng for April ll, to be held i n the "peace tent" located between Gil em's Camp and the Stronghold, Canby agreed. H e "assured the Commissioners that the Modocs would 'dare not molest us because (o ur) troops commanded the situation'" (Thompson, 1971 p. 59) The meeting held. Canby and Jack imme diately argued over peace terms. Jack gave the sign, pul ed his gun, and killed Canby Simultaneously Thomas was killed. Meacham was shot several times, but survived because Toby R i ddle deflected Schonchi n John's aim. She also stopped an attempt by Boston Charley to sea p Meacham. L S. Dyar and Frank Riddle ran the moment Jack pulled his gun, and each successfully evaded bullets fired by pursuing Modocs Assault of April 1 5-17, 1973. The tragedy of April 11 spurred both the War Depa rtment and the local A rmy command into action. Gillem's and Mason's soldiers were already poised on either side of the stronghold. Mason had 300 m en, and in addition 70 Warm Springs Indians were riding from their reservation to the north to join him as scouts in time or the battle. The western command under Colone l Gillem and Captai n Green already had 375 soldiers at Gillem's Camp. Surely 675 troops plus 70 Indi an scouts was sufficient to capture or exterminate 53 Modoc warriors! Wheaton had realized that Coehorn mortars would be more effective than the low-trajectory howitzers i n displacing the Modocs from caves and rock trenches. Many mortars had arrived and wer e ready to be moved into action. Wheaton, from his January 1 7 e xperience, had also passed on to his successors that the east side of the Stronghold was much more vulnerable to attack than the west side, but this e xcellent advice w a s totally disregarded. Gillem se Ap il 15 for the assaul I s planning, execution, and final resul s wer e almost a carbon copy of the assault of January e xcept hat his time he scenario was played in slow 158 mo ion. On he n ight o April 14-15 edged ging rly [orward i n occupi d roughly h s m of January 17 had re ch d fire. Bu i n hiss con d were prepared to s y nd hold o n roo s on he w s side wer moving into posi ion und r cov r of arkn ss soldi r los his footing mong he jagged rocks nd his rifl ccid n ally fired. Thus he Modoc' s were alerted, and th ir crie-s o warning were passed along hroughout th p rim ter of th Stronghold. The soldiers immediately bivouacked in plac throwing up rock forts as shelters. The soldiers on th east sid had alr ady con s ructed numerou s low forts of loose rock, and a rock wall lin of defense before he date for the assault was set. The lesson of th valu of rock s helters had been well-le rn d in the January 17 assault. Whe n the Modocs c ied oul their aler he troops on the east paused for the night i n these shelters. The ne x t day (April 15) mortar and ho witzer shells were poured i n o th Modoc' s e ncampm n and the troops began a cautious and slow ad vance. They immediately drew the Modoc' s fire. Some success was attained i n the early afternoon when Gre n's troops displaced a few M odoc snipers, who escaped safely to their defense trenches on the margin of the plateau, but not until they had taken a toll of casualties from the advancing troops. A s the night of April 15-16 approached, Green ordered his troops to straighten their line and build forts of loose rock, or else find shelter in rock crevices for the night. The advance o n the east side had been e ven more slow and cautious. Throughout this n ight the artillery sent bursts from their mortars and ho witzers toward the Modoc e ncampment. On the second day of fighting (April 16) the artillery continued to pour mortar and howitzer shells into the Modoc' s lair, and the foot soldiers advanced slowly. An attempt to push forward strongly on the south flank of the west side failed completely. On the eastern front Mason' s troops u nder sniper fire from their rear as well as in the direction of advance, and were pinned down for most of the day. It became obvious that it would be too costly to try to take the Stronghold by direct c ha r g e The strategy was changed to attempting a simultaneous advance along the lakeshore which w ould cut off the Modoc' s water supply. Reports disagree o n whether a real junction of the t w o forces was made but at least the soldiers approached close e nough to bring any Modoc seek i n wat r und r fir Throughout the night mortar shells were lobbed continuously at the Modoc's positions, and this activity was doubtless far more effective than the desultory and inaccurate rifle fire of the foot troops against an enemy that they could not see. At any rate it was during this night that the w omen, children and most of the Modoc warriors withdrew to the south, undoubtedly following the well-know n route across the flat-topped remnants of the lava plateau along which the Modocs had driven cattle into the natural corral. A few Modoc warriors remained through most of the night to harass and taunt the troops "in very plain, if not classical English" (Tho mpson, 1971 p 74) On the morning of April 17 the artillery stopped pouring shells into the Stronghold, a n d troops on both east and west sides began a cautious advance. N o sounds came from the Modoc's posi ion, no shots

PAGE 169

fired, and so soldiers i n par s of th Army ine advanced more apidly and soon w e on op of the p ateau with in the Stronghold area. They found it completely dese te Many soldiers, ho wever, delayed and wer still hiding behind their rock shelters. co onel Gillem, angry at their unwillingness to fight "got upon the highest rock availab e and ordered repeatedly 'Forward,' 'Forward," until finally Mason's Troop G u nder Captain Bernard came up to join Green in s weeping the area." (Thompson, 1971 p. 751 It was an empty s weep Over 650 Army regulars had spent 3 days and 3 nights i n "battle" and had suffered 23 casualties (6 killed, 17 wounded). Their attempt to "round up" or else e xterminate" the Modocs was a complete failure, even though they now occupied the Modoc' s Stronghold. The Modocs, ho wever, had suffered a few casualties as well, a nd so the Shaman's pronouncement of invincibility was discredited. It is reliably reported that one Modoc warrior was killed by a mortar shell, and very likely t w o additiona l Modocs d i ed from the shelling. Sergeant Fitzgerald reported that the severed head of a Modoc was be i ng kicked about by the troops after their entry into the stronghold, and W i liam Simpson, an English newspap r reporter, drew a picture of a soldier holding the sea p of a Modoc man, erroneously reported o be scar-Faced Charley ( w ho ho wever, was still alive and out of the Stronghold). One a ged Indian squaw, cra wling through the rocks in an attempt to get to water was captured, and then summarily shot through the head In this war savagery was not restricted to people of one culture or o n e color. The Modoc's Withdrawal Route. M uch nonsense has been written in A rmy Reports, as well as by historians and other writers, about the route by which 150-170 people mostly w omen and children, were able to escape from the Stronghold undetected. It was inconceivable to the Army command that they could slip a w ay so silently in the night without the soldiers' knowledge. The W arm Springs Indians, hired as mercenaries, were suspected of being traitors to their contract and of allowing the Modocs to "escape up a gulley, near the line that the Warm Springs held at the southeast e nd of the eastern front. Historians, as well, have appealed to t he finger-like collapse draw s south of the Stronghold, plus connivance with the W arm Springs scouts. Yet an observer w ho walks through these draws on the ground finds them so cluttered with schollendome s an d riv e n with deep cracks and crevasses that it would be quite impossible to get such a large group through, a ong with their dogs and horses in one night's tim Ev n more fanciful are h wri n sta m n s i n sam serious reports hat he Modocs "slipped pas the Warm Springs scouts in a large lava rench" (no such "trench" or "gulley" is th re), or the less ambiguous (bu quite impossibl ) s a ement ha they "disapp ared into he Schonchin Flow." To a opographic engineer equi pp d with mod rn airplan errain, the escape rou of They simply wal ked south and brisk pac on th remnants of plat au (see map), avoiding dimpl par s of h pl and Scholl nd o m s which pl au r mnan s Th h on long w hich h s ray and s tol n ca l or o a g ologis photographs of he he Modocs is obvious. basins which turndow n flaps bard rs of th h e sam s driv n ov r 10 0 n tur l corral from 15 h p ains and hi hl nds o the southeast. In the area of our map the northern end of this trail is shown by a broken line symbo ate that its closest point to the Warm Springs posi ion is about 1 / 2 mile (750 met rs) but hat i is no f r from the Modoc natura defense trench s at the top of the turndown flap on the west edge of the plateau. Here the Modocs had tota ly brok n an effort of th soldiers to "push forward strongly" during the previous day' s figh ing. Rock Fortifications Construe ed by the A rmy Where had the Modocs gone? Would they return? Was their disappearance only a ruse preliminary to a surprise attack ? Colonel Gillem and his troops did not know. It was obvious that a few Modocs were still around, for the occasional crack of a snip r s rifle was heard, and n o w a nd then an Indian could be seen out of rifle range walking o n the southern plateau remnants and others bathed in t h e lake. So the ne x t few days after the occupation were s pent i n hastily throwing up rock forts a nd rock walls for defense in case the Modocs decided to return. Over 200 of these fortifications have been located a n d are large e nough to b e shown by appropriate s ymbols on our map. Another 30 or 40, including a long but discontinuous rock wall, w re put up by Mason' s troop s before the April assault -they constitute the easternmost lin e of forts and walls show n on the map. During this activity the troops also discovered that the easiest travel in the area is o n the surfac of the lava plateau, and they selected and fortified with 1 0 larger and stronger "hollow-square" forts an east-west line of easy travel across the plateau remnants south of the Modoc' s former livin g quarters. The outer trail, constructed by the Park Service, follows this line of forts in its east-wes course (see map). The Thomas-Wright Ambush. Colonel Gillem decid d to send out patrols i n an attempt to find out where the Modocs had gone. Bernard headed a cavalry patrol to the east and northwest, and a second patrol under Perry rod e to the southeast o n the same day. Near Dry Lake som e Warm Springs Indians, riding with Perry, caught, killed and scalped two Modocs. When t he Warm Springs returned on April 21 they also brought information that som e Modoc warriors were in the Lava Bed s about 4 miles south of the Stronghold near a light colored butte (now known as Hardin Butte). Gillem first sent a small patrol to i nvestigate, and then decided to send a large patrol to see if it was feasible to set mortars on Hardin Butte to shell the Modocs in their new hideout. It se ms clear from other evidence that only a few Modoc warriors w re near Hardin Butte. The route of withdrawal of the Modocs was southeast toward Big Sand Butte and Dry Lake The large patrol started for Hardin Butte from Gillem's Camp o n April 26 Captain Evan Thomas, son the Adjutant General (retired) Lorenzo Thomas, was in charge. With Evan s were the sons of two other A rmy Generals: Lieutenant Thomas F Wright and Lieutenant Albion H o w e None of the t hree had any experience fighting Indians. The patrol moved along slowly fro m Gillem's Camp, and just before noon they entered an amphitheat r bounded by low ridges on the east, sou h and w st, a n d locat d near the west base of Hardin Bu It is reported that "old h a nd s a m o ng he foot soldi rs had ben worried on the march by th lack of x rience of th offic rs, a n d by the car 1 ss way in which the patrol trudge d along in a

PAGE 170

compac sides. becam and ac umn without deploying skirmish rs to th Ind d it is repor d tha Serg ant Rom r isgusted, climbed up the ridg on th w s ed as a Elan ing gu rd for pa r t of h e urn y When h pa rol reached he floor of h amphitheater Captain Thomas alled a h lt for lunch. The troops sa in s mall groups on the sandy ground, and many took th ir boots off. While the m e n w r resting and eating, Captain Thomas thr e o h r soldiers began o climb he o w ridge on he sou h t the head of the amphitheater. Before h e y had gon e far rifle fire cam e from the ve r y pile o f rocks that they had set out to reach. Lieutenant Wright react d at once by o r dering a "set of f ou rs" up the ridge to the southeast; these four men immediately were fired on; they turned and ran back to Wright's position in the middle of the amphitheater. Thomas then ordered Wright to advance with all of Company E onto the opposite ridge (west of the amphitheater}. Again th Modocs fired from rock piles i n the top of this r idge, but Wright and some of his men continued, only to die in the attempt. Others turned and ran. Meanwhile Lieutenant Cranston had taken 5 men to th northeast hoping to climb h i gh e n o ugh o nto Hardin Butte to signal Gillem's Camp. All six were killed when they reached the Butte. The soldiers i n the amphitheater were i n a state of confusion-many broke and ran toward Gillem's Camp. Those who remained were pinned down on the sandy floor with little or no protection by either rocks or bush. Most were killed or wounded. Although the first survivor of those who ran r eache d Gillem's Camp before 2 p.m., his story was not believed. Rescue parties were not organized or under way until nearly dark. It was the following day before they picked up the wounded, a nd buried the dead at the points were they found them. N o Modocs w e r e at hand to contest the operations. The return to camp was not attempted until after dark. It was shortly after dawn on April 28 whe n hey straggled into Gillem' s Camp. This stunning victory of the Modocs was accomplished by Scar-Faced Charley w ith only a few men. Most of the Modoc warriors were on the other side of the Schonchin Flow, far to the southeast. Charley's small band evidently remained behind to check on what was going on in the Stronghold and at Gillem's Camp, and also to protect the rear of the departing Modoc's. It is likely that he had less than a half dozen men, and they may h a ve s aged th ambush on the spur of the mom nt because of the E ar of being trapped against the west wall of the Schonchi n Flow. The only crossing of this flow that is negotiable withou t hours of slow plodding over tre cherous loose 1 va block s is just nor h of h base of Hard i n Butte From the standpoint of the Army the results of this battle are succinctly summed up in a report from Lieutenant Jocelyn, stationed at Camp Warner: "We have sicken ing n e w s a ga i n from the Lava Beds. Twenty-five men were killed, and 1 6 wounded -mos t of them seriously. In histori an Thompson's words (1971, p 91): "T wo-thirds of the patrol had fallen vic im to he Modocs accurate rifle fire. The rest h ad run." Again a ring of rock clefts and loose boulders in a schollendomed area had been used wth expert precision by Modoc snipers Dissension, W hy had he Modocs lef their natural fortress? W e 160 surmis rol s c au s th Obvi including w orn fact that th W O warriors. Th Shaman's claim rs would n ver die in c N o m tter what th reasons, the task o the Ar y in rounding up the Modoc s now becam one of movemen a nd fluidity, instead of a sieg In this n e w situation th Modocs, deprived of n arby sources of wate r and food, and e ncumb e red by worn n, children a the aged, could not for ve r withstand t h e probing 0 cavalr y patrols from a well-quipp d a nd w 11 fed army of over 1,000 men. N e verth less th retreatin Modocs were still a dangerous fore to cope with. May 7 t h e y raided a train of 4 wago n s drov off escort of 15 to 20m n 3 of whom suffer d w ounds a nd captur d 11 mul s and 3 hors s Th wagon s were e mpty -the train was on its way to pick up quartermas t r supplies -so t h Modocs burn d them. Three days later, at dawn on May 10, th Modocs s urrounded an Army patrol camp d for th nigh t nea r Dry Lake. This t i m however, t h e soldi rs stood t heir ground and drove th Mod cs off. Eight were w ou nd ed, 3 of whom lat r di d, bu h Modocs also lost two m n killed, on e of th m th popular leader Ellen's-M an George. Dissension and b trayal now b cam important factors in bringing the fighting to an n d The Modocs quarr l d bit e ly aft r th Dry L k ba tle, and split into t w o camps One group w ho form rly lived mainly along Hot Cr ek, n r Dorris, urned west toward th ir original hom land. Th o her and smaller group under Captain Jack remain d for h e present in the neighborhood of Dry Lak e a n d Big San d Butte. A rmy patrols continu d o reconnoi t r throuq .. this area for days, occasionall y s eing o n or more Modocs. Rock fortifications hat the soldiers put near Dry Lake and south of Big Sand But re num rous, and well preserv d o this day. Moreover, the visitor who will search out the highes points this nearly flat plain generally will b r ward d finding the op n fretwork of s on s w i h which a Modoc s n i p r had camouflaged his observa ion position. Th H o Springs group ran into roubl whil tr v eling west. Colon l J fferson C Davis, w ho succeeded to Canby s command arrived in the lava beds o n May 2 H e soon brok e up the large concentrations of soldiers at Gillem s Camp, the Stronghold, a n d Hospilal Rock into many roving patrols o f large size who would liv e in the Lava "where they could fight a the first oppor unity, or could rest and take things easy, like the Indians." Two such patrols pic ked up the rail of the Hot Cre e band O n e caught up with and gave chase to the I ndian s -who scattered in all directions hrough t rocks mountain mahogany n d junipers. Bu 5 Modocs were killed 3 w o men and 2m n -and sev ral women, children and horses were captured. Davis ordered hts Indian s bu chase som Modocs report d that th h n continue to harass hes e y could again tak up h e o the Fairchild Ranch and d to surren d r

PAGE 171

(Thompson, 1971, p. O S ; Riddle, 1974, p 2 8 oavis s n t w o I nd ian w o m e n hat he had employed o n other missions o seek ou the M o docs a n d g i v e h e m the t e m s of surrend er. John Fair c hild a nd his w ife also w e n t u p he mounta i n w i h them. Aft r a s hort discussion s ca r Faced Charley took he lead i n persuadi n g his peo p e to s u r n d e r w itho u fur her b oorls h e d E ventually 6 3 Modoc m n, w o m e n a nd children cam o f f the mou n tain a nd surr nd e r ed. They were a sorry-looking group: clothes i n tatters; m a n y lam e sick or i n j ured; a nd "half-na k e d childr n a n d a ged squa w s w ho could scarcely hobhl ". Hooker Jim, he Lost R i v e r murderer," was n o t a m o ng the m b u h came i n and s urrendere d a little l a er. Davis' a vowed purpose howeve r was to catch captain Jack and to hang h i m forthwith for the murder of canby. H e l o s t no time in accel e rati ng the search. Dav i s had earn d the val ue of hiring Ind i an s to catch Indians, so he immediately hired steamboat Frank, Hooker Jim, Bogus Charley, and Shacknasty Jim to f i nd Captain Jack (Thompson, 1971, p. 110; Riddle, 1974, p 1 64) These were the very same Modoc s who had taunted Captain Jack into killing Canby Davis als o sent along the t w o Indian women who had helped arrange the surrender of the Hot Creek group. W ithi n a w eek these Modocs now turned informer, and rid i ng w ell-fe d Army hors e s had located Captain Jack's band c amped a long Willow Creek in a wi l d and mounta inous area east of Cl ar Lake. A large contingent of troops moved in, and while the traitor Bogus Charley was talking w ith Captain Jack about surrender, the troops came upon the Indian camp a few hundred meters away and fired a volley of shots into it ( R iddle, 1974, p 1 4 7 ) The hunted Modocs scattered, leaving most of their guns, ammunition, and equipment. Eight Modoc women and some children were taken prisoner, among them Jack' s two wives a nd ch ild. Jack and a few others dashed into a thicket of willows, waited unti l darkness, and then escaped. The chase continued. Before it was over Jack and the few members stil l left in his band had raced approximatel y 200 m iles, f irst in the rough country a l ong W illow Creek, then to Langel l Valley and Bryant Mountain in Oregon, and back to C ear Lake and the l ower part of Will o w Creek again. But the odds against survival were too overwhelming; the mounted patrols, no w aided by skilled Indian trackers, had picked up his trail at many points, and on June l, 1 873 they closed in. Aga i n Scar-Faced Charley helped in pursuading both Modocs and soldiers to refrain from fur h e r bloodshed. Captain Jack, along with two men, three worn n, and s e v e ral children surrendered. Another Modo c warrior, B e n Lawve r w ith his wife, boy, a n d aged fa her and m o h e r cam e i n a few days later. The remai nd e r of h e Modoc' s s t ory i s a record tha contribut s littl of honor to ei her victor or vanquished. Davis inte n o n r e v ng i ng Canby, planned to hang C a p ain J ack publicly wi h i n a few d ays afte r cap ure but he was frus rate d by a ruling from Wash i ng o n h a t p nali t s mus b decided by th courts Th results of th trial w re predictable. Cap ain J ack, S c h o nch i n J ohn, Bos o n Charl y a nd Black J i m wer hang d Slolock a nd Barncho, who also were i n th vicin ity o f he p ac t n but d i d n o t a k pa r t i n h e actual k illings w r e s n e nc d t o tif a Alcatr az. F i v e w o und d M o doc prison r s wer killed b y w o unknown assasin s while b ing tran spor d t o j i l a F o r Klam ath. Ano her Modoc 1 6 1 committed suici d after cap ure. T h remaining Modocs, 55 m n, w o men, a n d children, w r ba nished o a barren tract o f land, 5-3/ 4 s quare mil s in re (2 5 miles o n a sid ) n ar the Q u a p a w I ndian Agency i n Ok ahom a Confin e ment, dise se, a n d poverty t h n a comp ished more than had th effort s o f o ver 1 000 troop s So m e e w survivors were giv e n permission to return t o t h K l m a h R e serv a t i o n i n O r gon aft r 1 909 The Modoc's e ternal manum nt is t h e natural fortres s o f t he S t r o n g hold, which remain s just the same today as when i t was formed b y th congealin g n d ere p i n g f orward of a large flow of lava mor e t h a n on million years a g o Ref e r ences Fuller, R. E., 193 1 The aqueous chilling of basal t lava on the Columb i a River Plateau; Ameri c a n Journal o f Science, v. 21, p. 281-300. Murray, Keith A., The Modocs and Their War: University of Okl ahoma Press, Norman, 1958, 3 4 6 p. (An objective, interesting, and well-written account by a historian and sociologist who has served as Park Ranger at Lava Beds National Monument). Riddl e Jeff C., The Indian History of the Modoc War: First printed in 1914; republished in 1974 by Urian Press, Eugene, Oregon, 1974 295 p (A fascinati ng book written i n s imp le, s raightforward language by the son of the Modoc woman Toby (Winema ) R iddl e who served as interpreter during the Peace meetings. Jeff Riddle knew all of the Modoc warriors and his reports of conversations among them in their own language give a vivid picture and a new dimension of many of the deliberations among the Indians, and also between Modocs and whites). Thompson, Erwin N., The Modoc War: Argus Books, Sacramento, Calif., 1971, 188 p. plus illustraions and maps. ( A comprehensive account with particular emphasis on numbers, distribution, personnel, and objectives of he various military units engaged in a struggle. Outstandi ng documentation by notes and references to many little known army, newspaper, and private records that are not easily available). Waters, A. C. and Fisher, R. V., 1971, Base surges and their deposits: Capelinhos and Taal volcanoes; Journal of Geophysical Research, v 76, p. 5596-5614.

PAGE 172


PAGE 173

PRE-HOLOCENE SILICIC VOLCANISM ON THE NORTHERN AND WESTERN MARGINS OF THE MEDICINE LAKE HIGHLAND, CALIFORNIA Stan ey A. Mer zman Department of Geo ogy, Franklin and Marshall Co lege, Lancaster PA 1 7604 A s rae -----edicine Lake H i gh and i s an area of diverse vocanism whose extrusi ves span he last 1 5 to 2 : lion years o f earth h istory. The enigmatic Andes e Tuff of C A. Anderson (1941) has been prov' s iona y p laced into he genera stratigraphic sequence as one uni w'thin the collage collec i ve l y known a s he "pre-Med icine Lake andesite" volcan ics. his proposed pos tioning results from f i e ld, etrographic, and chemical ev idence whi ch suggests the ex i s ence of a series of andesi tic 1 ava flows are coeval i n age or slight y younger han the Andes e Tuff. One of these equ ivalent l ava flows has g'ven an age of 0 .21 m.y. o ld. If this correlaUon s accu ate, then the conclusion that the A des e Tuff represents an event whi ch occurred r:or to he i n i ial e upti ons of he Medicine Lake and, therefore, he uff cannot be he deve opment of he Medici e Lake caldera. Additional s'l'cic volcan i sm, i n c uding rhyoli ic e s and flows, i s con s iderably o der ( 0 .43 to 0 95 y a fac which i ndicates no dire c genetic link eed e xst be ween the rhyolites and he tuff. ast y, he rhyolites can be d i vided into wo separate groups on the basis of S i02, ne i he of ove laps that of the Holocene glass f l o ws, a ea ure for which the Med cine Lake High and i s we known. Severa ecent summaries by Macdona l d (1972 and IIi liams and McBirney ( 979 concerning the o i gin of co lapse calderas suggest a number of these features formed concurrently w i h he eruption of ash-flow ffs oble 0 969) suggested tha such a may in fact exp a i n the ori g i n of the Medic 'ne Lake ca dera i n nor hern California. In articular he p oposed that a uni A PSi e ff may correl a e wi h he Med;c ne Lake caldera and her he ex rusionof h "old ,.. shie d-forming andesite" uni o C A Anderso n (1941). Thi i n erpre ation was a'so rei erat d by A. T And rson 0 976 ) On the ohe r hand, r zman ( 977) i n rpre d th s ol'v i ne andes e lavas a he in' ial e uptions of h : edici n Lake volcano and heref o neces a P"e-ca dera i n ag a conclusion s i milar o of r. A. Anderson (19 4 1 ) Thes con radic ory nterpreta ions ar abe ted by a ack o f outcrop n cri ic a areas a d h e nab i i y o obtain a reliab e K/Ar age da e f And Tuff, a di ffi cul ty a ising: from i t larg seal atmospheric argon contamina i on v r, f'e_ d mapping by Walter 1 975) and er zman ( unpublished da a) indicat he exi s nee of and si ic lavas which appear to be 163 coeva wi h the andesitic ash-flow tuff uni t Subsequent analytical studies concerni ng these tuff-equivalent avas has apparently resolved the contradictory interpretati ons previousl y outli ned and provides a rudimentary stratigraphy on which to build future research i n the Medic i ne Lake H ighland. Geologic Setting The Med'cine Lake H ighland, located approxi mate l y 50 km eas -northeast of Mount Shasta in northern California, is the southern extensi on of a discont'nuous belt of shield-like composite volcanoes whi ch lie 30-50 km east of the High Cascade stratovolcanoes. The northern extension of this belt i ncludes Newberry vo lcano in central Orego n and S i mcoe volcan o i n south-central Washington. Previous research i n the Medicine Lake High l and has been recent y summar i zed by Mertzman (1977) and Hei ken (197 8 ) and so the writer will present only the highlights of the geologc history. Over the past several years three dozen K-Ar analyses have been performed as part of an ongo i ng attempt to ascertai n the geochronologc development of the region surrounding the Medicin e Lak e H'ghl and and extendi ng toward Mount Shasta. To date, all he samples from w i hin the region covered by F igure l are less than wo m i lion years old. The oldest lavas are the basaltic rocks of Timber Mounta i n volcano, whi ch i s located jus off F i gure l approxiately 9 miles sou heast of Lava Beds Nat i ona Monument These lavas range from 1 7 to 1 9 million years old. Thin flow s o f low p o assium high-alumina olivi ne tholeiite a r e we l exposed along the porti on of Gillem's B luff fau t scarp w i hin Lava Beds National Monument and vary from 1 0 to 1 4 m.y old. In add ition, numerous olivi ne tholeii t e f lows of nearl y identical pe rography an d chemistry outcrop both around the periphery and intercalated with lavas of the Medicine Lake Highland and range i n age from approximately 1.4 m.y. old to Late P l e istocene to Recent. Moving from the H i gh l and westward there are approximately a dozen mos y andesi ic composite volcanoes within the Bray, Dorris and Moun Dome quadrangles whose ages fall etw n 0 3 and 1 3 m y old. It is within the contex o f th' s volcanic h i story that the silicic gneou s a tiv'ty on the northern and western margins o f he Highland must be viewed and its origin eventual y e xp l a i ned DATA K-A results The K / A r da a are shown i n Tab l e l w i h the e xact ample locations depicted i n Figure l Potassium was determi ned together w ith all he other major elements utilizing a lithium etraborate fusi on which results i n a one i nch g lass d 'sc. The actual ana ysis was

PAGE 174

MT DOME 6 / ....... .) I LAVA BEDS NATIONAL MONUMENT / ........ I i ,...i CONTACT INFERRED CONTACT 6 MOUNTAIN PEAK CALDERA RIM 5 km [] RECENT SILICIC . VOLCANICS ANDESITE TUFF i D i ------: 0 OLDER SILICIC t1@ VOLCANICS FLOWS EQUIVALENT OF ANDESITE TUFF 4140' LITTLE 6HORSE PEAK 6 BLACK MTN N 1 Figure 1. Geologic map of selected rock units from the Medicine Lake Highland area. The stratigraphic colum n is based in part on K/Ar dating. performed usi ng an automated vacuum X-ray fluorescence spectrometer. Argon was determined by pipette-spi k i ng with 38Ar; the isotopic measurements performed with a MS-1 0 mas s spectrometer. This instrument d isplays virtuall y no memory and no mass 36 background. The argon fusion was carri ed out on 3 6 grams o f 8-42 mesh fragments of whole rock whi ch had been ultrasonically c leaned i n distilled water and alcohol dried i n an o ven and pre-baked in a vacuum at 250C The uncertainty i n the age i s estimated utilizing a 0.3% uncertainty i n the argon spike, a 2% uncertainty in sample homogeneity, and an absolute uncertainty i n the potassium of 0.01% K20. The silicic rocks dated in this investigati on are from rhyolitic flows and domes. The contact relationships between the domes and the surrounding vo lcanics generally provide rather inconclusive information. These rhyolites range from 0 .43 to 0 95 m.y. old with relatively smal l uncertainties. This spread of ages falls within the range previously quoted for the composite volcanoes located generall y west of the Medicine Lake Highland. Any theory which attempts to explain the petrogenesis of thi s region over the past t w o million years must come to gri p s with this s pace-time-compositi on relationship whi ch ex ists between these two groups of volcanic rocks. 164 The Andesite Tuff i s the one major ignimbri t i c uni t i n the region and its areal d istributi on i s outlined i n f igure l In general it i s a poorl y welded ash-flow tuff that completely lacks any strati f icati on However, where the base is exposed ( e.g. i n the southern porti on of the area no w known as the "Bighorn Sheep Enclosure" on Gillem B luff) substanti a l f latteni ng of pumice lapilli has o ccurred together w ith s i gn i f icantl y more wel d i ng The wri te r has made three attempts t o radiometrically date this unit usi ng a moderatel y wel ded samp le, a poorl y w e l ded samp le, and a p ag i oc lase separate ( no potassium feldspar present) None o f these attempts were successful due to large scal e atmospheri c argon contaminati on Subsequent detailed f ield work in the area south of Dock W ell by Walter (1 97 5) and myse l f (See portion of f igure l delineated as "Flow s equivalent of Andesite Tuff") l ed to the isolation of several andesi t i c f l o w s w h i ch appeared to be the same age or s lightl y younger than the Andes ite Tuff. Petrographi c and chemical studies supported the mapp i ng interpretation whi ch has been further substantiated by reconnaissance m icroprobe mineral analysi s These data indicate that smooth compositi ona l gradati ons as wel l as a large measure of overl app i ng occur betw een the proposed equ i valent lava flows and the tuff w ith respect to the orthopyroxene, clinopyroxene and p lagioclase feldspar. In fact, the plagioclase phenocrysts whi c

PAGE 175

TABLE 1. Analytical results of K-Ar dating.a Sample Number Map Location Number Rhyolite domes ML49 18 17 5681 15 0 0269 0 0412 0 0443 Lava flow equivalent of Andesite Tuff 263Cl 12 0 0065 Earliest Medicine Lake volcano lavas 366W SM25c A 8 0 00115 11. 62 68.04 5 55 3.18 2.12 0.74 4 42 4 63 3.18 2 .14 0.85 0.85 K-Ar age (m.y.) 0.43.04 0.61. 03 0.95.14 0 .21.05 0.09. 05 0.08.04 aConstants used are: 4K/total K 1 167 x 10-4 ; x8 = 5.81 x lo-11yr-1 bRad i ogen i c argon, 10-lO moles. cThe age is slightly different from that reported in Mertzman (1977) as a result of recalculation with the new K-Ar age constants of Dalrymple and Mankinen (1979). occur in both the flows and the tuff have v irtually the same zon i ng pattern in addition to highly similar chemistries. One of the proposed tuff-equivalent ava flows g ives an age of 0.21 0.05 (Table 1). I submit that this age closely reflects the t ime of eruption of the Andes ite Tuff and, at the very worst, reflects its m i n imum age. The Andesite Tuff controls the topography on the north west side of the Medicine Lake Highland onto which the earliest of the Medicine Lake basalti c andesite flows were extruded. Two stratigraphically e q uivalent lava flows, from the northwest and the south west flanks of t he Medicine Lake volcano (B and A respectively n figur e 1) have given ages o f 0 .08.04 and 0.09. 05 m .y. old. Since these flows represent bo h the initial eruptions of the Lake volcano as wel l as the bulk of the extrusives which compose the Highland (equivalent to he shield-forming platy olivine andesite of C A. Anderson, 1941) the conclusion can be d awn tha the entire 900 000 m thickness of lavas related to the edicine Lake volcano (See Mertzman, 1977) has accumulated over the past 100,000 years. A corollary of this conclusion is that the caldera-forming event must be <100,000 years old; unfortunately no minimum age can yet be assigned since no rampart-forming andesite, which represen s syn-to post-caldera volcanism, has her tofore been radiometrically dated a void which will be filled over the next several months. Two additional points are worthy of no e : since the stratigraphic thickness of the tuff noticeably increases in a 165 southeasterly direction and yet it does not outcrop on the eastern and southern margins of the High land, i t seems safe in assuming that the vent area for the tuff lies buried beneath the younger andesitic lavas of the Medicine Lake volcano. Secondly, the gross age d istinctions which exist between the rhyolitic domes and flows and the tuff-equivalent andesitic lavas on the northwest margin of the Highland, which differ by a factor of 3 to 4, indicate separate pulses of vo lcanic activity rather than the rhyolite and the tuff being time correlative. The latter interpretation was suggested as a possibility by C. A Anderson (1941). Chemistry Major and trace element chemi s try was performed by X-ray fluorescence spectrometry and the resulting data together with CIP W norms are reported in Tables 2 and 3; the exact sample locations are depicted i n F igure 1 For trace element analysis one gram of dried whole rock powder was mixed with 0 5 g of microcrystalline cellulose and pressed into pellets using a course grade of c ellulose as a backing. Trace element data reduction was achieved by the mass absorption correction method outli ned by Hower (1 959 ) Analyses of the Andesite Tuff are f ou nd in Table 2 ; typically these samples have relatively large loss on i gnition values, ranging from 1.87 to 4 52 % by weight, a point which encouraged the writer to calculate the CIPW norms on t he basis of recalculated a nalyses derived by normalizing to 100 percent

PAGE 176

Major and trace elemen ch mis ry ogether with C .I.P. W norms of he Andesite Tuff. The exact sample locations are depicted i n figure l. TABLE 2 Map location number 1 2 5 6 7 8 Sample number SW236 SM107 SW253 SM97 SM72 27601 WH67 SM88 Si02 59.26 60.95 61. 0 2 61.06 61.29 61. A l 2 0 3 15.73 17.00 16.66 15.77 15.27 FeO 3 .10 0 9 1 l. 69 l. 1.92 l. l. Fe2o3 3 .28 3 .56 3 .88 T i 0 2 1.03 0 .90 0.95 0 .92 0 .90 0 .90 0 .86 MnO 0 .12 0 .11 0.11 0 .10 0.11 0 1 0 0 MgO l. 98 1.86 1.90 l. 89 l. 2 .28 l. 65 CaO 3 .83 4 .08 3 .71 08 3 .76 Na2o .60 72 71 K 2 0 1.53 2 .28 2 .17 2 .28 2 .36 2 82 2.46 P205 0.22 0 .20 0 .19 0 .20 0 .21 0 42 0 .20 LOI* 3 26 3 .32 2 .68 3 .19 l. 87 Tota l 98.99 100.41 100.29 100.07 100.00 98.79 ----------Q 16.17 17.08 15.58 73 16.26 c 2 .74 0 .69 Or 9 .31 13.12 13.88 16.98 15.17 Ab 39.63 37.77 1.10 30 A n 18.09 1 7 .79 1 7 .57 1 6 .59 17. 8 1 D i l. 57 2 71 0 7 1 3 .33 Hy 6 .63 3 .59 2 2 .88 MT 4.90 0 71 3 .22 2 2 34 2.65 2 .20 Ilm 2 .01 .79 1.85 1.80 l. 76 l. l. 70 l. 03 Hm 3 .05 0 .89 2 .73 2 .22 1.25 A p 0 .53 0.50 0 .99 0.49 0.32 ---TRACE ELEMENTS ( p p m ) R b 30 51 72 64 68 Sr 373 389 373 387 377 360 293 Ni 6 2 6 4 2 6 5 Zr 275 262 260 255 263 273 213 Ba 570 552 640 588 622 523 571 656 y 33 26 23 17 24 31 21 25 v 115 101 99 121 106 72 94 30 Rb/Sr 0 080 0 .139 0 123 0 .129 0 191 0 .178 0.232 K / R b 351 392 351 325 319 333 Loss on ignition 1 66

PAGE 177

TABLE 3 MaP lo c a t ion number Sampl e number Si02 A l 2o3 FeO Fe2o3 Ti0 2 MnO MgO CaO a2o K 2 0 P 2 0 5 LOI1 Tota l Q c Or A b A n Di H y Mt 1m Hm Ap Rb S r N i Zr Ba Rb/Sr K/Rb 'Loss on igni tion ND:not detected Major and trace element chemi s t ry together w ith C.I. P .W. norms of the lav a f low equ ivalent of the Andes ite Tuff (9-12) Also incl uded are s i x ana lyses ( 1 3-18 ) of o lder s i l icic vol canic rock s 9 5481 59. 55 16.50 4 36 l. 4 3 o 74 0. 1 0 3 .16 5 .56 3.58 2 .04 0 1 9 1.49 98. 70 1 3 .30 12.40 31.16 2 3 .59 2.89 12. 63 2 1 3 l. 45 0 .45 56 438 30 241 575 33 9 9 0.128 3 0 2 The exact sampl e locati ons are depicted i n figure l 1 0 ML57 60.50 16.7 1 1.88 4.72 0.91 0. 1 2 2 .66 4.88 4 1 8 2 .03 0.17 1.48 100 .24 15.15 12.15 35. 8 1 21. 1 0 l. 79 5 88 3 .86 l. 75 2 .12 0 .40 48 4 1 2 14 229 52 8 3 0 9 2 0 .117 351 11 259Cl 60.59 16. 4 5 3.74 1.68 0 .69 0.10 2 .89 5.25 3 .69 2 36 0.19 1.21 98.84 14.09 14.28 31.98 21.87 2.89 10. 60 2 .49 l. 34 0.45 59 415 19 212 523 2 4 9 6 0. 1 4 2 332 12 263Cl 60.84 16.98 4 .56 l. 41 0 .79 0 .11 3 .45 5 .74 3 .85 2.1 4 0 .24 0 .72 100 .83 11.59 12.63 32.54 22. 7 1 3 .43 13.01 2.04 1.50 0 .56 i3 ML55 71.08 1 4 .25 0 .70 l. 78 0 .32 0 07 0 53 l. 74 4 73 3 .19 0 .08 0.98 99. 4 5 27.48 0 .04 19.14 40.65 8 .24 l. 34 l. 58 0 62 0 .72 0.19 14 47Bl 71.30 14.32 1.07 l. 78 0.44 0 .08 0 .42 l. 92 5 .10 3 .02 0 .11 0 .95 100 .51 26.09 17.93 43.34 7 .30 1.21 0 .49 2 .44 0 .84 0 .10 0 .26 TRACE ELEMENTS (p p m ) 54 77 67 433 176 234 28 ND 3 245 5 4 2 19 106 0 .125 329 167 217 737 2 8 1 0 0 4 3 8 344 233 7 2 8 20 1 6 0.286 374 15 568 1 71.47 14.48 0 .91 1.63 0 .31 0.06 0 .51 l. 66 4.78 3.18 0 .09 0.87 99.95 27.68 0 37 18.97 40.82 7 .72 1.28 2 .25 0 .59 0 .09 0 .21 63 131 3 194 7 2 9 1 8 1 0 0 .481 4 1 9 16 SM38 71.99 14.11 0 .60 2 .23 0 .45 0 .05 0 .34 1.24 4.87 4 .28 0.08 0.36 100 .60 24.63 25.23 41.11 3 .99 1.27 0 .26 0 .79 0 85 l. 68 0 .19 123 91 ND 367 790 48 22 1 352 289 1 7 SM5l 71.96 12.93 0.82 0.42 0 .21 0 .04 0 .08 0 .86 4.06 4 .63 0.04 0 .29 99.34 31.81 27.62 34.68 3 .42 0 .55 0 .81 0 .61 0 .40 0 .09 128 70 ND 189 664 54 6 1.829 300 1 8 ML49 76.38 12.13 0 .70 0.38 0 .05 0 .04 <.02 0 .46 4 .46 4.42 0.01 0.44 99.47 32.92 26.38 38.11 0 .03 1.83 0.07 0 .?6 0 1 0 0.02 148 2 ND 130 85 99 ND 74. 0 248

PAGE 178

anhydrou s totals. C a s in this i gh all he tuff samp es become more d acitic than andesi ic i n bulk compos i i on. The only add i t i ona l face of their chemistry worthy of h ighlighting at thi s t i m i s that three of eight samples are peraluminous. Comparison of the chemistry of the tuffs with that of the four samples for the proposed equ ivalent andesitic lava flows y ields a pattern of l ower SiO and Na20 and higher MgO and CaO for the lavas. Also, the percentage of phenocrysts is higher in the lavas than in the tuff samples (See Table 4) These trends are analogous to those reported from numerous other ignimbritic areas (e.g., Lipman and Others, 1966; Ratte and Steven, 1967; Christiansen, 1979), further substantiating this writer's claim of a co-genetic relationship linking the lavas and the andesitic tuff. The rhyolitic volcanics form two distinct groups on the basis of their Si02 contents: a low Si02 TABLE 4 Modal analyses ( volume percent) The data reported are average values (8 and 4 samples, respectively) which cover the entire range of petrographic variation w ithin both the Andesite Tuff and the lava flows proposed as being equivalent to the tuff. Mineral Andesite Tuff Tuff equivalent lavas Olivine Trace Trace Orthopyroxene 3 2 3.9 Clinopyroxene 1.1 1.8 Amphibole Trace Opaques 2 3 1.2 Plagioclase 5 2 21.8 Groundmass 89.3 71.3 group (71-72% ) and a high Si02 group (>74. 5 % ) Exami n i ng he age and chem ical da t a concurrentl y produces a noteworthy correlati on be ween decreasi ng age and increasi ng S i 0 2 content. Further documentation of thi s emerg ing pattern i s presently underway. Whether or not thi s age-compos i t i on trend can be genetically linked to a v iabl e crysta fractionation model awaits m ineral compositi on data concerni ng phenocryst-forming phases, thus enabli ng some quantitative model l i ng to be performed. It i s a l so intrigui ng to note that the chemistry of these o lder rhyolitic extrusi ves does not mimic that of the Recent silici c g lass f low acti v ity i n the Medi c i ne Lake High l and ( E ichelberger, 1975; Heiken, 1978; Mertzman, i n press). Table 5 presents average chemical analyses for the three Recent g lass f lows. It i s ev ident that these data form a third group, intermed'ate between the two groups defi ned by the o lder silicic vo lcanism. Suffi c ient chem istry i s 168 cu r en l y a vailab l all p rform d w i h s m analy ical proc dures, o pro ably in ufficien s mplin g and a naly ic l u n c rtain t y as p o entia l f a tors producin g h clump-Jik rather h n a continuum o f l v a riation. Wh ther o r not this o lder (or e pisodes) of silicic volcanism is m o n o g ne ic o r poly n tic, s: esearch p r oblem currentl y und r study u ilizing micr opro b e and iso epic techn i qu TABLE 5. Major and trace element analyses of three Holocene glass flows from the Medi c ine Lake Highland. (n : number of samples analyzed). Little Glass Glass Northwes t Mountain Mountain Glass flaw n 5 4 3 Si02 73.41 73.57 7 3 .23 Al2o3 13 6 2 13.68 13. 5 6 FeOT 2 1 3 2.12 2 0 5 Ti02 0 .27 0.2 7 0.28 MnO 0 .03 0.04 0.04 MgO 0.35 0 2 2 0 .30 CaO 1.29 1.20 1.25 Na2o 4 10 4.11 4 .21 K 2 0 4.28 4 .29 4 .28 P205 0.03 0 05 0 .05 LOI* 0 .40 0.34 0.49 Total 99.91 99 .89 99.74 Rb 159 16 2 15 4 r 107 113 108 Ni ND ND ND Zr 229 215 223 Ba 741 728 7 2 2 v 6 8 3 Rb/Sr l. 486 l. 434 l. 426 K/Rb 223 2 2 0 231 *Loss on ignition. ND = not detected.

PAGE 179

Analy ical d ata concerni ng the o der silic i c agma ic a c v i y on he nor h ern flanks o f Med i c i ne a k e H igh and is presented Available K /Ar r esu ts ive a 0 4 3 t o 0 95 m y range for the rhy o liti c canism w hich c an be div i ded che m ically i n t o l o w s i ic a ( 7 1 -72% ) and h i g h siUc a ( > 74.5 % ) group s Both of hese g oups are clearl y d isti n g u isha b l e from he H o l ocene g lass flow s w ith respect t o chemistry. h e area e x tent and chemical variab ility of t h e Andesite Tuf f i s o utlined. The past i nabilit y to radi o metrically d a t e the time o f erupt i o n of t h e tuff has tentati v e l y been c i r c umvent e d by a na l yz i ng o n e o f severa l lav a flows whi ch wer e erupted c o ncurrently w i h or very shortly after the tuff. T h e age of t he equivalen t flow i s 0 21.05 m .y. o ld, a date that cl e ar l y e stablishes the erupti on of the t uff as an event precedi ng the deve l opment of the Medi c i ne Lake andes i tic volcano and, therefore, not correlati ve with the formati on of its summit caldera. ACKNOWLEDGEMENTS I thank t he Research Corporati on for a Cottrell co l e ge S c ience Grant, the Franklin and Marshall committee o n Grants, and the American Philosoph ical Society for f inancial support of the f i e l d work. I a so thank Dr. James Aronson of Case Western Reserve Uni versity for access to h i s K / Ar laboratory as well as for h i s t imel y adv ice. The automated vacuum x-ray fluorescence spectrometer was purchased with grants froo the Pennsy lvani a Sc ience and Engineering foundation, the Nat i ona l Sc ience Foundation, the Gul f Oil C ompany, and the F l e i schmann Foundati on REFERENCES Anderson A T., 1976, Magma m i x ing, petrol og ical process and volcanol og ical tool ; Journal o f Vol c anology and Geotherma l Research, v. 1 p 3 3 3 Anderso n, C A., 1941, Volcanoes of the Medicine Lake California: University of Californi a Publications, Bulleti n of the Department of Geolog c a Sc iences, v 25, no. 7, p. 347 422 Christiansen, R. L., 1979, Cooling u n its and com p o s i t e s hee t s i n relation to caldera structure: Geol o g i c a l Soc iety of America Spec i a l Paper 180. D a rymp l e G. B and Manki ne n E.. A., 1979, Revised geom a gneti c po larity t ime scal e for the interval 0 5 m y B. P.: Journal of Geophysica l Research, v 84, p 615-626 E i chelberger, J C., 1 97 5 Ori g i n of andesite a n d dacite: e v idence o f mix i n g at G lass Mountai n in Cali f orni a and at other c i rcum Pac i f i c volcanoes: Geolog i c a l Society o f America Bulleti n v. 8 6 p 3 8 1 -139 1 Hei ken, G 1978 Plin i an-typ e erup t ions i n the Medicine Lake H i gh l and Calif o r n i a a n d the natur e o f the unde r l y i ng magma: Journal o f Volcan o l o gy a n d Geotherma l Research, v. 4, p 375 402 How e r J 195 9 Matri x corrections i n t h e x-ra y spect r ographic trace e lemen t a n a lysi s o f rock s a n d minerals: American M i neral o g ist, v 4 4 p 19 -32. 169 L i pman, P W C hris iansen, R L and O Connor, J. T 1966 A c omposi iona lly z o ned ashf l o w sheet i n s outher n ev a d a : U S Geologica Su vey Professional Paper 5 2 4 F, 4 7 p Macdona ld, G A 1972 Volcanoes Prentice-Hall, N e w Jersey, 5 1 0 p. Mertzm an S A 1 977 The petrol ogy and geochem istry o f the Med icin e Lak e volcano, Cal i fornia: Cont r i buti o n s to M i neralogy and Petrolog y v. 62 p. 2 2 1 2 4 7 Mertzman, S A i n p ress, The petrogeness o f Recent silicic magmatism i n the Medicin e Lake H i ghland: ev i d en c e from cognate i nc lusi ons found at L ittl e Glass Mounta in, Californi a : Geochim ica et C osmoch i m i ca Acta. Noble, D C 1979, Speculations on the ori g i n of the Medi c i ne Lake caldera: Oregon Department of Geology and Mineral Industries Bulleti n 65, p 1 93 Ratte, J. C., and Steven T A., 1967, Ash f lows and related vo lcani c rocks associated with the Creede caldera, San Juan Mounta ins, Colorado: U S. Geol og ical Survey Professi ona l Paper, 524 -H 58 p. Walter, R. C 1975, Geo l ogy and petrol ogy of the northwest portion of the Medic i ne Lake High land, California: Unpub. B. A. thesis, Franklin and Marshall Col lege. Williams H. and McBirney, A. R., 1979, Volcanology: Freeman Cooper and Co. San Francisco, 397 p.

PAGE 180


PAGE 181

S RFACE STR CT RE OF LITTLE GL SS ,tO TAL Jon H Fink Geology Department, Stanford niversity, Stanford, CA 94305 (Pr sent address: Department of Geology, Arizona Stat niversity, Tempe Arizona 85281) The purpose of these two stops is to observe and interpret the structural relations exposed at the sur face of a very young rhyolitic obsidian flow. Before surface structures can be discussed however, he flow stratigraphy and the origina l attitudes of foliations must be determined. Unfortunately these Holocene obsid'an flow s do not have any well e xposed cross sections so w e are limited to observations of now fronts and to comparisons with older, more d)ssected flows. A t these stop s we will first define he stratigraphy and discuss the attitudes o f foliations as seen in flow fronts; then w e will go on o the upper flow surface and see the variety of structural relations exposed there. l m SURFACE BRECCIA 5 m F I NELY VESICULAR PUMICE 15m 9m 2m 3m OBSIDIAN COARSELY VESICULAR PUMICE BASAL BRECCIA TEPHRA Figure 1 Schematic cross section through a 35 m hick rhyo i ic obs'dian flow a ed on observation of Lit le Glass Moun ain and dissect d flows in ew Mexico and Lipari, Italy. Stop Dl -Nor heast lobe, Lit le Glass Mountain Before w can interpret he deformed flow structure w must be able to recogniz he undeformed structure. This stop offers an oppor uni y to observe he different units which make up the flow 171 Figure 2 Flow fron Li tle Glass Mountain. Photo foliations; interpre ation. Comple x y fo ded coarsely vesicu ar pumice (c) overlain by obs'dian (o) and finely vesicular pumice (f). Base obscured by talus (T

PAGE 182

stra igraphy. Firs we no e he layer of ephra which covers he ground urrounding he flow Isopach maps indica e tha the source of this tephra is under Little G ass Mountain (Heiken, 1978) If we l'oo k a blocks w i hin he talus pile w can distinguish principal types based on r.olor nd vesicularity: black glassy brown to green'sh grey coarsely vesicu lar pumice, and whi ish grey finely vesicu ar pumice These hree rock types have uniform chemical composition (73% S i 02) and all exh'b't f o w ayering, but they differ in density, with the obsidian hav ing he highest. In the flow front above the talus pile these three rock types appear as coherent units, with foliations generally conforming to the contacts between units. [I] T alus Bo h i n he flow fron s and on the upper flow surfac he coarsely and f i ne l y vesicu ar pumice uni are nearly ah.rays separa d by obsidian; they do not appear in conformable contact. Furthermore he coars pumice nearly always underlies h obs'd'an, so hat the apparent stratigraphic seque, is ( upward ) : tephra coarsely vesicular pumice, obsidian, f i ne l y vesicular pumic A bo v e the and on op of he flow ar breccias comprised o f blocks o f he other three flow units (Figure 1) In older more dissected rhyolite and rhyolitic obsidian flows, foliations near the base are generally horizontal whereas those near the top are more nearly vertical. The depth of the transition from vertical to horizontal flow-layering varies f 0 25 50m D Coarse l y ves1cular pum1c e D Obsidian 0 F inel y vesicular pum ic e t N 172 Figure 3 Map of part of northeast lob Little Class Mountain. Compression al fold axes marked by 1 ines with arrows; axes of fractures mark d by lin s with on or two cross bars (two bars indicates separations across fracture of more than 5 m). Dotted lines indicate margins of fracture surfaces. -

PAGE 183

flow to f low and within a g i ven flow, however it generallY lies withi n the top 25% of the flow. Thus w i th i n the stratigraphy seen in L ittl e G lass Mounta i n the contact between the coarse pumice and obsi d i an layers may be considered to be ori g i nally horizontal. Mapping the orientations of this contact o n th e f low surface indicates the deformation. F i gur e 2 shows the three units exposed i n a fold w i th i n t he f low front. Thi s fol d and others like it F i gur e 4 Compress ional folds i n f low front, northwest lobe, L ittle G lass Mountain: a) overturned fold, plunging into flow. b) closer view of sam fold. tre ched vesicles show x ension d irection associa ed wi h foJdi ng 173 are also exposed on the upper f low surface. About 20 m south of thi s fold i s an anticl ine o f coarse pumice whi ch appears to be d issect ed by a valley 3-4 m deep. Later we will see that these va lleys whi ch trend perpendi cu lar to the f low front are fractures which commonly bisect anticlines cored by coarse pumice If we climb onto the flow through this fracture and climb the northern wall, we will be able to see much of the area covered by the map i n Figure 3 Figure 5 oapr rise accompa nied b y fracture o f surface. otice that fractur e a xis corresp o nds to earlier anticlinal a xis and hat f ne pumice be w een he diapirs form s tight s ynclines.

PAGE 184

.. 1 T a lus L Coarsely vesicula r pum1ce N f O bs1d1an D Fme l y vesicular pu m1ce Figure 6. Map of part of northwest lobe, Littl e G lass Mountain. Symbols s ame as F ig. 3. The most prominent structure i n thi s part of the map area i s a fracture about 2 m deep whi ch cuts through coarse pumice. One can easily see that the foliation patterns in the two oppos i ng walls correspond exactly. In addition, the entire coarse pumice area is enclosed by an area of obsidian, whi ch i n turn is surrounded by fine pumice The foliations within the coarse pumice and its contact w ith the obsi d i an defi ne an anti clina l structure whi ch p lunges to the northwest (into the flow) 20 m to the north is a series of several more of these p l ung ing, fractured, coarsely vesi cu lar pumice anticlines. The obs i d i an adjacent to each anticli ne gets pushed laterally away from the fracture p lane, and where two fractures parallel each other the obsidi an and fine pumice may get sandw i ched into a tight synclinal configuration. In F igure 3, these syncli na l axes paralle l to the fractures are marked by heavy dots. Not ice that the separati on or spreadi ng across each fracture increases toward the f low front, suggesting that these fractures form, at least i n part, i n response to circumferential spreadi ng near the front of an advanci ng flow lobe. 174 The coars e l y vesi cu lar pumice borderi ng each of the fractures i s folded, with fol d axes perpendi cular to the fracture axes. This creates a pattern of coarse pumice domes. In f igure 3, the fol d axes are indicated by heavy lines w ith arrows, whereas the fracture axes are marked by cross bars. These fo lds generally lie perpendi cu lar to the f low d irection a are interpreted t o b e surface folds caused b y compression, s i milar to the ropes on pahoehoe basal f lows ( F i nk and F letcher, 1 978 ) The basal posi t i on and relati ve l y low density of the coarsel y vesi cu lar pumice form the basi s f o r gravity instability w ithi n the f low structure. The instability can g i ve ris e to regularl y spaced d iapin of coarse pumice. On aeri a l photographs of both B i and Little Glass Mounta ins, such regularl y spaced areas can be clearly seen. Depend i ng on thei r positi on relati ve to the f low front these areas of coarse pumice may be subjected to subsequent folding or fracturi ng or both, resulti ng i n the compl ex structural pattern observed on thi s portion of L i t t e Glass Mounta in.

PAGE 185

I I N ', -,sf ( I t \ -+-30 \ 1 0 F i gure 7 Map of large ridge adjacen ured ar a on nor h w s \ \ \ \ ) / \. to lobe. Sop 1 2 -N orth west lob Li l e Glass M oun ain I I A his locality we will see s i milar structures as before but here fractures are more developed. A l hou h he entire stratigraphic sequence occurs here, w e will concentrate o n structures found within a larg e area o f coarsely vesicular pumice. Thi s area has a broad domal struc ure and is bordered by obsidian. 175 The firs s ructure is loca ed abou 2/3 up he f o w front, 20 m south of the small parki ng area Her e coa se pumice and obsidian w i h very large vesicles (over 10 em i n arne er) form an overturned fold, whose axia p ane plunges in o the f low ( F igure 4) S re ching of vesicles indicates he extension direc on associated w ith folding. M ineraliza i on o f he vesicles occurred prior to this folding as the coa i ngs are a lso stre ched. W ithin thi s fol d the obsid"an drapes o ve r the coarse pumic e and h igher up i n the front the obsidian forms a syncline cored by f i ne l y vesicular pumice. These folds w ith axes parallel o the f low front conti nue on the upper flow surface. Climbing to any high po int on the upper surface one can see several 3-5 m deep fractures trendi ng normal to the flow front. Here again, examination of the foliation patterns in the opposing walls shows the large amount of lateral separation associated with these fractures, up to 40 m i n some cases. The obsidian and f i ne pumice which stratigraphically overlie he coarse pumice are generally upturned around the marg ins of this area, but within the area they have been sandwiched into t ight synclines and even "subducted" between adjacent fractures. A lthough he cumulati ve extension associ a t ed with these fractures is over 100 m, this "subduction" could conserve the volume of the flow. F igure 5 i s an interpretation of the interaction between upward rising coarse pumic e and do wnwar d propagati ng fractures. Folds w ith axes normal to the flow d irection can also be s een i n this area, and these extend laterally into the higher ridges to the south. Detailed examination of the structure of he highest of these ridges (F"gure 7 ) shows that i t also is interrupted by a series of subparallel fractures, with separation increasing toward the flow front, again due to extension associated with radial expansi on of the flow. Aer i a l photographs of thi s lobe (Figure 8 ) show several other areas of coarse pumice and obsidian with irregular spacings along the flow direction. These have been interpreted as diapirs and the spacing has been related to the stratigraphic thicknesses and v iscosities of the d ifferent textural units. In summary, Little Glass Mounta in, being nearly erosion-free, offers a rare opportunity to observe the surface structure of a rhyolitic obsidian flow just as i t appeared upon coo i ng The structural relations seen here could only be detected on older flows w ith great difficulty because erosi on and vegetation prevent continuous examination of foliation patterns, and subsequent alteration obscures the textural differences originally present. References F ink, J H 1978, Surface structures on obsidian flows: PhD thesis, Stanford University, 175 p. F ink, J H Gravity instability in the Little Glass Mounta i n rhyolitic obsidian flow, northern California: Tectonophysics, in press.

PAGE 186

Figure 8. Photo, part of northwest lobe, L ittle G lass Mounta i n Lava moved from left to r i gh t D ark fractured areas are pri mar ily coarsel y vesicular pumice The locations of F igures 4 a n d 7 are indicated. Fink, J H., Surface folding on rhyolite flows: Geology, in press. Fink, J H and Fletcher, 1978, Ropy Pahoehoe: Surface folding of a viscous fluid: Journal o f Volcanol ogy and Geotherma l Research, v. 4, p 151 170. Heiken, G 19 78, Plinian type eruptions in the Medicine Lake Highland, California and the nature o f the underlying magma: Journal of Volcanology and Geothermal Research, v 4, p 375 402. 176

PAGE 187

H O L OCE E PLI IA TEPHRA DEPOSITS O F THE MEDICI E LAK E HI'HLA D C ALIF O R I A Grant Heiken Los Alamos Scientific L a b o ratory, Geosciences Division, Los A l a mos N e w M e xico 87545 Conden sed from : PLI N IAN-TYPE ERU P TIO N S I N THE MEDICI N E LAKE HIGHLAND C ALIF O R IA, AND THE NATURE O F T H E UNDERLYING MAGMA, Jou r Vole. and Geo t h e r m Res., 4 375 402 (1978). 0 Elsevie r S ci. P u b C o Amsterda m INTRODUCTIO N The Medicine Lake Highland i s a young volcano located 55 km east o f Mount Shasta, on the western margin o f the Modoc Plateau. The volcano is east of the north-south-trending chain of High Cascade vol canoes and within a 100 -km diameter depression filled with volcanic roc k s of Pliocene to Holocene age. It overlies a plateau consisting of late Tertiary tuffs, lava flows and sediments, cut by north-to north west -trending normal faults (Powers, 193 2 ; Anderson, 1941 ) The volcanic center is located at the intersection of s everal of these fault systems. The highland i s a 1 km thick, 25 km-diameter shield, composed of andesi tic a n d basal tic flows (Anderson 1941). There is an 8 km by 6 km, 100 200 m d e ep c a ldera at the shiel d summit. Noble ( 1969) believes that an andesitic pyroclastic flow, interb edded with andesi tic flow s of the shield, may have b een associated with caldera collapse. Andesitic, d acitic and rhyolitic flow s were erupte d along the caldera rim and partl y bury the walls. Construction o f the shiel d and c a ldera formation are believed to h ave occurred mainly during Pliocene to Pleistocene time (Anderso n 1941 ) During the time elapsed between the earlier eruptions of andesite dacite and r hyolite and more recent activity, there was g laciation of the highland and substantial soil developed (Ander son, 1941 ) Holocene a ctivity was bimodal, consisting of basa ltic volcanism, mainly on the shield flank s and eruption of tephra and lavas of intermediate t o rhyoliti c canposition within and close to the rim o f the caldera. The purpose o f thi s paper is to present data on the Holocene pumi c e deposits of the Medi cine Lake Highlan d and t o speculate o n their origin. These de posit s wer e erupted from Glass Mountain. a flo w of rh yolite-d a cite on the northeast rim o f the caldera a n d fran Little Glass Mountain, l o cated a few kilometers wes t o f the c a ldera (Fig. 1 ) Age o f the Holocene deposits Underlying Holocen e rhyolitic and ba saltic t ephra e xp osed in t he highlan d is a red or y ello w soil, d eveloped on both flow s and older tephra. This soil i s estimated t o be mor e than 1 5 ,000 years old ( D iller, person a l communication 1977 ) The oldest tephra o v e r l y ing the soil are fine-grai n e d basa l tic ash layers near Little Glass Mountain and n orth west o f Glass Mountain. The y r a ng e fro m 0 5 to 14 e m thick and a r e reason a b l y fresh ( sider o m e l a n e glass is well preserved). I t is possible that these t e p h r a are f rom Pain t Pot C r a t e r o r som e o f the y o ung e r *The 1 C measurements w e r e made by Isotopes. Inc. a n d the sample was collected by C D M ille r from the northwest slop e o f Little M t H offman (Fig. 1) The samp l e consists o f w ood leaves, b ark a n d con e s 177 cones in Lava Beds National Monument (Table 1 ) Immediately north of Glass Mountain, the soil has developed o n an older pumice bed (Cheste r m an 1955). Chesterman dated trees that were growing i n this soil and buried by younger pumice beds to indicate the age of the Holocene Glass Mountain tephra. Overlying the lowest Holocene basal tic tephra unit are the pumice deposits of Little Glass Mountain, with a radiocarbon date of 1 0 65 + 90 years*. As will be discussed later, the tephra are similar to those of Glass Mountai n Luckily, one of the more extensive tephra beds from Little Glass Mountain has a distinctive yellow hue. By following this bed it was determined that Little Glass Mountain tephra underlie tephra units erupted from Glass Mountain. The age of the Little Glass Mountain flow is not known. but is believed to be only slightly younger than its tephra deposits; they may have been part of the same eruptive sequence. Within the zone of overlapping Glass Mountain and Little Glass Mountain tephra, there is a thin (5 em) bed of basaltic ash, believed to be from one of the young cinder cones on the northwest flank of the shield. The uppermost tephra beds in the highland were erupted from Glass Mountain and have a radiocarbon age of 1360 + 240 years (Chesterman, 1955). The analyses were-of carbon from tree trunks buried by the Glass Mountain pumice falls. The Glass Mountain flows have been dated at 130 to 390 + 240 years by the 1'C method (Friedman, 1968); these dates were based on wood from standing trees damaged by the flow. There is an inconsistency here: field evidenc e indicates that little time elapsed between the pumice eruptions and extrusion of the flows at Glass Mounta i n whereas radiocarbon dates indicate a hiatus of 1 0 0 0 years. The older date of 1 3 6 0 + 2 40 years is of t ephra overlying tephra dated at 1065 + 90 years. Tephra from Glass Mountain also overlie the very well-preserved High Hole Crater cinde r cone and Burnt Lava Flow. The youth of these basaltic vents is also indicated by the presence of tree trunks burned by the flows, resti ng o n the flow m argins I n summary most Holocene silicic rocks near the h i ghland summit and the most youthful basal tic cones and flows on the highland flanks w e r e contemporary and may have erupted during the last 1100 years. RHYO LIT I C TEPHRA FROM THE HOLOCENE A CTIVIT Y Little Glass Mountain Tephra from Littl e Glass Mountain form a highly elongat e e l l ipse with its long a xis oriented southw est-nort h east (Fig. 2). Measurable tephra lay e r s are present up to a distanc e of 25 km and i n dividual rhyoliti c lapilli possibly from Little Gl a s s Mountain have been found 50 km west-southwest o f t he ven t ( The m ost distant lapilli were found by D Miller o n the s lopes o f Mt. Shasta) The area

PAGE 188

B a S LI rLO S W (.!111'1 0N0 ,,.. 1'(l1 c::=J ISlfiC CI.,.O(IIt CO,..[S M0L.0C(h( 0(P0$11S 0V(Iitl.YI ... G A tu-0( (L.OP(O SOil.. 8(l.I(Y(0 TO I( 1100 T ( A .. S Ol.O -<...,.... / C A LOflltA lt llll I AJilln(lt$0H, r ''"'' PUWICE STONE WTN 0 &@PAINT POT '""'" l A V A 8(0 $ N lt,f -----0 I I I I I I __ _) 121&' ,,..,,:; Fig. 1. Distribution of Holocene deposits believed to be <1100 years old in the Medicine Lake Hig land. The fissures near Crater Glass Flow are open and well preserved. Based on U.S. G S 15 topog aph quadrangle (Medicine Lake and Timber Mountain, Cali fornia), the geologic map o f Anderson (194) a fiel d work by the author. TABLE 1 Age a RhyollllC Tephra of Glass un a1n c date (b y F Lbby n: C hes ennan, 1955) of 1360 : 2 4 0 and 1107 : 380 flow s o f age? ephra o Li le Glass a1n "Cag e (by Isotopes. Inc ) of 065 : 90 years. Basaltic Lava Flo w and H1gh Hole a w e 11c 1 cone II ed basal 1c ephra o f Pa1n Po o by L le Glass unta 1 n So1l d eveloped on andes e flows and (La e 1 sconsn early Tioga; 0 Miller, C(lfTTilunic auon, 1977) ,. 1 0 000 -15, 000 years B.P 178

PAGE 189

> ._ M L -25 (PUM ICE PIT SECTION) 3 7 km NE @ = SfM -II (PUMICE STONE Mtn. Sec.) 1.8 km SW OTHER MEASURED SECTIONS ..../ / / -, // //// -... / / / / \ / / 'l I ,, / / / //, I / / / / (o / / /OE> I I / 1 / >-'f I I / /,1 ( (l) L'G ) / / _,.. I / I 1 I 1 I / 1 I / I \ ,..-1/ I I I fo '----;.." I I I / I / I o"? / I 1 1 1 o / I I I /' .-..--" I I / I / I I I I /I o ...... --o J l /// l I I I / / // 0 3km 0 "--!..._ 0 I I / / 1---'--L--i I \ -.t __ ., // { I \ ,' / ....... -' // / // fig. 2 Isopach map o f Little Glass Mountain (dashed) and Glass Mountain (solid) tephra Stippled areas are flows at the two vent areas. Dots are the locations of measured stratigraphic sections. The cross near Glass Mountain is the location of the thickest section measured. covered by measurable tephra layers is about 200 km2 ; the tephra have a volume of 0 .046 km1 ithin 3 km southwest of Little Glass Mountain, there are five tephra units, consisting of mainly reversely graded, pale gray pumice and lapill i -si ze b ocks in coarse gray ash. These units range in thickness from 20 to 50 em and each contains between 5 and 20% lithic fragments. The only evidence for a hiatus within these units is a thin (1-5 em) coating of orange-brown clay at the surface of the third tephra unit. Contacts be.tween all other units exhibit no evidence of weathering or ero sion. Between 3 and 8 km southwest and 6 km northeast of Little Glass Mountain, there are t w o tephra units; the lowest, overlying a well-developed soil, consists of a 0 24 em thick, mediLITl to coarse gray ash sometimes with a slight yellow hue; the upper unit consists of a 13-64 em thick, generally reversel y graded, lapilli-bearing coarse ash. It is possible that the break between units 3 and 4 near the vent corresponds to the break between the t w o e phra units farther fran the vent. Beyond 9 km, onl y o ne tephra layer is present. This tephra bed i s most readily correlated to the upper of the two ephra units, located closer to Little Glass t-1ountatn, on the basis of its greater thickness (Fig. 3). All of the tephra units erupted (rom Little Glass Mountain appear to have been deposited as air-fall. There appear to have been two periods o f explo sive activity at Little Glass Mountain; the only evidence for this, ho wever, is the clay coating developed on the top of a tephra layer ( fig. 3) The tephra are overlain by the flow s of Little Glass Mountai n a rhyolite flow w i Lh a volume of 0 3 km' The flow has buried th vent area. Glass Mountain Pumice deposits from Glass Mountain form an trr gular ellipse, with the long axis trending east-north ast (Fig. 2); the deposits with measurable 179 thicknesses cover an area of 320 km 2 and have a volume of 0 09 km1 Within an approximate range of 3 km east and northeast of the summit of Glass Mountain, there are multiple tephra b ed s Within a pumice pit located 2 8 km east of the summit, e even tephra beds, with a combined thickness of 3. 7 m, overlie a red-brown soil developed on cinders. The tephra units consist of reverse y graded or massive pumice block and lapilli-bearing coarse gray ash, 5-102 em thick. Several of the uppermost beds contain 10-30% lapilli and coarse ash-size lithic fragments (mainly angular obsidian fragments and cinders). Sharp depositional contacts e x ist between these beds; there was no erosion or weathering at the bedding plane surfaces. Several tephra units exhibit a pale orangepink discoloration near the top of each bed; the alteration may be due to mild vapor-phase activity that left a thin film of hematite stain on pyroclast surfaces. Between a distance from the vent of 3-11 km east, 3-11 km north and 1 ( ?) -3 km south and west, there are two or three tephra beds with com bined thicknesses of 22131 em. Both consist of reversely graded or massive beds of gray medium-ash to lap ill i-si z e tephra. Due to the remarkable uniformity of the Glass Mountain tephra, it was not possible to corela e any o f these beds with units described near the vent Units 1, 7 and 11 of the Pumic e Pit section are only tentatively correlated with the more distant units, because they are the thickest units. Beyond the range of multiple beds there is only one bed of massive coarse ash and lapilli; this rapidly thins t o zero. All tephra beds from Glass Mountain drape the underlying terrain and exhibit no current structures; there i s no ev ide nee for deposition by flow. After deposition of rhyolite tephra, the Glass Mountain dacite-rhyolite flows were erupted along a northnorthwest-trending fissure. In addition to the main body of Glass Mountain there are, along the fissure, nine small domes. Associated with the domes are

PAGE 190

ASH, E:=J MEDIUM to COARSE LAPILLI PUM ICE B LOCKS 20 km SW lm 0 8km SW ,VT7y-,"T. ,..,.. .. .-ORANGE CLAY 2km SW COATING on PUMICE Fig. 3 Tentative correlation of tephra units along a line southwest from Little Glass may have been two episodes of explosive activity at Little Glass Mountain. untain. There small, crescent-shaped pumice cones, consisting of angular, 2-45 em long pumice lapilli and blocks. Other silicic vents The Crater glass and Medicine Lake glass flows appear to be approximately the same age as Glass Mountain and Little Glass Mountain, but neither e xhibited much e xplosive activity. There were only a few lapill i or blocks erupted from these vents; there is no evidence for extensive tephra deposits asso ciated w ith them. DISCUSSIO N Vents for the Holocene eruptions of silicic tephra and lavas are located along the oval caldera rim and along fissures parallel to the rim. Glass Mountain and associated small domes were erupted along a fissure trending northwest-southeast. The r is no clear e vidence that Little Glass Mountain was erupted from a fissure. There is, ho wever, an arcuate trend, concentric to the caldera rim that includes the Crater Glass Flow (Fig. 1) and several en-echelon open fissures. The fissures, with a total length o f several kilometers, do not appear to be associated with an y of the normal faults of the region and are believed to be an expression of rhyolite dikes that a r e near the surface, but did not erupt. Crater Glass Flow was erupted f rom a segment of one of these fissures. It was suggested by Anderson (1941) that the caldera at Medicine Lake collapsed along cone fractures that later acted as conduits for Pleistocene and Holocene silicic magma This may indeed be the case; eruptions along the caldera rim and along fi ssures concentric to that rim may indicate the pres ence of several concentric, inward dipp'ng cone sheets located near the surrrnit o f the Medicine Lake shie d Cone fractur systems are developed over plutons during intrusion and have been observed in many intrusive-volcanic com plexes throughout the world (Anderson, 1937; Koide and Bhattacharji, 1976). Cone fractures dip inward, from the ground surface, at angles of 45 to near y vertical. If the models for cone sheet formation developed b y Anderson and 180 Koid e and Bhattachar j i are corr ct, then cone fractures may extend downward from the sur face xpression of calderas and cone ntric fissures to a magma body that is considerably narrower than the calder a The remarkable similarity and contemporaneity of the eruption at Little Glass Mountain and Glass Mountain, located 15 km apart, suggest that they were erupted from the same body of magma possibly along cone fractures. Eruptions at the two vents have the following in common: ( 1) the same eruption sequence, consisting of tephra and 92-93% flow; (2) identical major-element compositions for the tephra; (3) nearly identical tephra characteristics, as outlined earlier; (4) similar ages; (5) the lavas have identical trace element compositions (Condie and Hayslip, 1975). The uniqueness of trace element data for individual magma bodies has led to the use of these data to 'fingerprint' tephra deposits for th purpose of correlation (e. g Borchardt et al., 1971 Howorth and Rankin, 1975). The total volLJTle of Holocene silicic tephra and lavas is about 1.2 km3 On the basis of the smal volume of eruptive rocks, it is possible Lo mak e the inference, albeit weak, that the magma body is small. The caldera has a voll.llle of about 8 km3 (Anderson, 1941). If the caldera volLITle is a reflection of the approximate size of the magma chamber this is still a relatively small chamber. Also, the caldera volune rn y reflect the earli r phase of ndesitic volcanism and not that of the Holocene silic1c eruptions. The ratio of 1 : 9 for tephra to flows, observed at Glass Mountain and Little Glass Mountain also fits general model for small silicic magma chambers developed by Smith (1976). Interpretation of the structural setting and the petrology of the silicic tephra and lavas erupted in the Medicine Lake High land during Holocene time supports the hypothesis that there is a small silicic magma body below the highland. Concentric cone fractures may have developed over the apex of a body located at a depth of 4 8 km; this depth is consistent with the model presented by Loide and Bhattacharji (1976) and with observed intrusive-volcanic complexes and elsewhere in the Cascade Range (e. g Erikson, 1969) and Per u (Cobbing and Pitcher, 1972). If it is true that only

PAGE 191

small cooling body of magma is present below the shield, a reassessment of the area as a en ial geothermal area may be necessary. Th1 s pothesis should b e kept in min during he elopment. of the highland which has been classified as known geothermal resource area' (God win et al. 1971) The tephra blankets fit a general description of deposlts having been formed uring Plinian or sub-Plinian eruptions. A Plin1an eruption is exceptionallY violent and ejects copious pL.Illice (Escher, 1 933 ; Walker and Croasdale, 1970). Tephra were deposited as air fall with no ev .idence fo: deposi as pyroclastic flows. An erupt1on of th1s klnd f1ts the model outlined above of vesiculated rhyolite magma erupted from the volatile-ric h upper portion o f a small magma chamber. Deposition of tephra in beds may be due to indi idual eruptions relate to time of refilling of a fissure, vent blockage or variation in w ind intensity over the volcano. I n several tephra units from Glass Mountain and in all un1ts from Little Glass Mountain, concentrations of lithic fragments at the base of each bed indicates clearing of vents o r erosion of vent walls during e xplosive activity. After the volatile-rich top of the magma chamber was erupted as tephra, the bulk of the lavas with low (<20 0$) vesicularities were erupted. CONCLUSIONS (1) Holocene activity in the Medicine Lake Highland consisted o f basaltic eruptions on the flanks and eruptions o f silicic tephra and lavas near the s nmit. The rhyolitic eruptions at Class Moun ain and Little Glass Mountain occurred during the last 1100 years. (2) Air-fall tephra consists of very poor y sorted lapilli that are mostly pL.Illice pyroclasts. The tephra falls formed e ongate ovals, e xtending northeast of Glass Mountain and southwes of Little Glass Mountain. The tephra deposits are similar to those deposited during Plinian eruptions. (3) Pumice pyroclasts are remarkably homogeneous and consist of blocky, angular forms. Most have elongate simple and compound vesicles and vesicularities of 45 -60$. All of the pyroclasts of homogeneous rhyolite g lass. (4) The Medicin e Lake Highland is located at the intersection of several normal faults; a weak spot' that allowed pooling o f basal tic magmas needed for crustal mel't:.i n g or a 'conduit' for the rise of diapirs of silicic magma (5) The contemporaneity and physcal and chemical similarity of tephra a n d lavas erupted at Glass Mountain and Little Glass Mountain, located 15 k m a p a r t suggest t hat they were e r u pted from the sam e m agma bod y The y may h a v e e r u pted along con e fractures developed above the magma chamber. The relatively small vol ume of Holocene silicic eruptions, small caldera volL.Ille and a ratio of 1:9 for tephra and flows suggest that the magma body had an approximate volL.Ille of between 2 and 8 km3 If this hypothesis is correct, the highland might have less potential as a g eothermal resource than was previousl y believed. (6) The blocky, angular pu mice pyroclasts may have developed by: (a) vesiculation and elongation of vesicles by flow within 1-2 km o f the surface vent; ( b) and isruption of the brittle, vesicu lated magma b y an expansion, or 'relief' wave passing down into the vent from he magma-atmosphere interface. 18 1 A C K 0 LED ME TS Samuel Gallegos, Jr. was an able f1e d and laboratory assistant thro ghout the summer of 1976. wish to thank D Mann anJ T. Gregory for t.h ir help as well. I benefited from discussions of the Medicine Lake tephra with J Eiche berg r B Crow e and C D Miller. Howe Williams and an anonymou reviewer provide usef reviews of the manuscript. This research was s ppor ed by the Departmen of Energy, Division of Basic Energy Sciences under contract W-7405-EN G -36. REFER CES Anderson, E. M., 1937. Cone-sheets and ring-dykes: The dynamical explanation. Bull. Vo canol., 1 : 35-40. Borchardt, G. A., Harward, M E and Schmitt, R. A., 1971. Correlatio n of volcanic ash eposits by activation analysis of g ass separates. Quat. Res., 1 : 247-260. Chesterman, C W 1955. Age of an obsi ian flow at Glass Mountain, Siskiyou County, California. Am. J Sci., 253 : 418-424. Cob bing, E J and Pitcher W S 1972. The Coastal Batholith of central Peru. J Geol. Soc. London, 128 : 421 -460. Condie, K C and Hayslip, D. L 1975 Young b modal volcanism at Medicine Lake volcanic cent r northern California. Geochim. Cosmochim. A ta, 39: 1165-1178. Erikson, E. H. Jr., 1969 Petrology o f he composite Snoqualmie batholith, central Cascade Mountains, Washington. Geol. Soc. Am. Bull. 80: 2213-2239. Friedman, L. 1968. H ydrat10n rind dates rhyolite flows. Science, 159 : 878-880. Godwin, L. H Haigler, L. B., Rioux, R L., White, D E. Muffler, L J P. and Wayland, R G 1 9 7 1. Classification of public lands valuable for geo hermal steam and associated geothermal resources. U S Geol. Surv. Circ. 647. Howorth, R and Rankin, P C., 1975. M ti-element characterization of glass shards from stratigraphically correlated rhyolitic tephra units Chern. Geol., 15: 239-250. Koide, H and Bhattacharji, 1976. Formation of fractures around magmatic intrusions and their role in ore localization. Econ. G ol., 70: 781-799. Smith, R L 1976. Ash -flow magmatism (abstr.). Geol. Soc. Am. Abstr Progr. 8 : 63 3 -634.

PAGE 192


PAGE 193

MECHANISM OF MAGMA MIXING AT GLASS MOUNTAIN, MEDICINE LAKE HIGHLAND VOLCANO, CALIFORNI A John C Eichelberger Geosciences Division, University of California, Los Alamos Scientific Laboratory, Los Alamos, NM 87545 ABSTRACT Mixing of basal tic and rhyolitic magmas at Glass Ho ntain appears to have been driven by vesiculation 0 / basaltic magma as it intruded rhyolitic magma chamber. Rapid cool1ng of basal t1c magma formed a mafic foam which floated and concentrated at the roof of the chambe. r Foam-r 1ch lava emerged first during the erupt10n, and became the h ybrid dacite of the distal end of the flow. The c .hamber is robably a relatively large volLme, long-llved fea lying within 10 km of the surface beneath the caldera. Eruption of similar lava at Crater Lake shortlY before caldera collapse supports this inter-pretation. This mechanism of mixing between silicic magma stored in a crustal chamber and basal tic magma feeding the chamber is controlled b y initial water content of basaltic magma, and implies that dry basaltic magma would remain at the base of the chamber. The eastward change from andesi tic to bimodal volcanism in this portion of the Cascade Range may be due to an eastward decrease in water content of parental basaltic magmas INTRODUCTION The largest and most imposing product of Holocene activity at the Medicine Lake Highland Volcano is the Glass Mountain lava flow (Figure 1). The most perplexing feature of this flow is its lithologic variety. The 1 kms of lava which comprises the flow ranges from dull, stony, porphyritic dacite charged with fine-grained mafic xenoliths to shiny, black, phenocryst-free and xenolith-free obsidian. The range in silica content is from 66 to 73 Anderson (1933) described the geologic setting and general petrologic features of the Glass Mountain G l ass Mou n ta i n F l ow and Domes D RHYOLITE RHYOLITE WITH INCLUSIONS D BANDED LAVA (RHYODACITE) DACITE D BRECCIA ZONE SAMPLE LOCATION + VENT KILOMETER F1gure 1 Map of Glass Mountain sCnple locations and lithologic units. Three digit numbers refer to samp les discussed by Eichelberger ( 1975) Single digit numbers designate sample sites for density measurements discussed in this paper. Modified from Eichelberger (1975). 183

PAGE 194

I FERREO CALDERA C:J R(C S I I C I C A A -CROSS SECTION \ GLASS MOUNTAIN 4:1 I _..../ H I GH HOLE CRATER oil" ANOERSO (1941) Figure 2. Map showing distribution of Ho ocene avas of the Medicine Lake Highland. flow. Additiona data on its age and history were provided by Finch (1928), Anderson ( 94 ) Chesterman (1955), and Friedman (1968). More recently, Condie and Hayslip (1974) an Mertzman ( 977) presented gee chemica and petro ogic data for Highland avas, and I (Eichelberger, 1975) studed the compositiona zonation of the Glass Mountain flow as a means for understanding relationships among magmas o f the basalt-andesite-dacite-rhyolite family: Mapping and sampling of lava streams within the flow revealed that the eruption was continuous and, in genera proceeded from homogeneous dacite through highly heterogeneous banded rhyodacite to rhyo ite. This suggested a canpositional zonation of the parent magma body in which dacitic magma graded down ward to rhyolitic magma Electron microana ysis of phases within the dacite indicated that most of the phenocrysts, An75 plagioclase, Fo80 olivine, and o 1En Fs1& augite, were derived from disaggregation of the crystal-rich basaltic xenoliths. The dark acitic bands of the heterogeneous rhyodacite are streams of basaltic debris from disintegrating xeno iths. This process was more advanced in the dacite into which the rhyodacite grades, yie d ng an intermed'ate magma that was megascopic a ly homogeneous except f o r residual xenoliths. Since this ybr magma was concentrated at the top of t e c amber, and since the Highland shield through which the magma passed contains a hi g h pro po r t on o f m a f i c 1 a v a f o ws and c i n d e r cone s it was concluded that mafic material had s o ghed off the roof of the magma chamber and contam nated the top of the magma body. However, in subsequent studies of avas of similar composition from other Cascade volcanoes, found smilar xenoliths and recognized evidence that the xenoliths were iq i when they came in contact w t si ic c magma (Eiche berger, 978). This e 'dence ncl des: 184 1 1. Roun e shape o f xenoliths. 2 Texture indicating rapi crysta liz tion to jus above soli us. 3 Incorporation o f ph e no crysts from the host silicic magma in rinds on the outer portions of some large xenoliths. 4 Outward decrease in grain size in some arge xenoliths. 5 Reaction or resorption of phenocrysts in the host lava indicating heating o f the silic' c magma during m ixing. Although the parental rhyolitic magma of Glass Mountain lacked any phenocrysts to record a thermal ev en associated with mixing, significant c r ystalliza ion should have occurred if cool rock had been stirre into the chamber. Lac k of such crystallization suggested that mi xing at Glass Mountain might have in volved w o magmas rather than magma and roc k and that a reevaluation of the data was in order. The arrangement of vents on the Highland, sil cic vents associated with a caldera structure and flanked by numerous mafic vents (Figure 2), raises the pos sibility that mafic magma is supplying a sha lo silicic chamber (Figure 3) However such a mec hanism would introduce mafic magma to the base of the chamber and woul seem to produce a composi tiona zonation opposite to that observed in the Glass untain flow. Evi ence that mixing occurs by floatation o f basal tic foam in silicic chambers (Eichel berger, 1979) affords a so ution to this problem and prov' es a basis for un erstan ing the evol tion of the Medicine Lake magma chamber. VAPO R E XSOLUTION DURI G MIXING Mixing of arge batches of magma requires flow on a arge scale within a magma chamber. A lik l y mech anism of large sea e f ow and stirring is convection. Introduction of mafic magma into the base of a chamber would strong y heat silicic magma i n the ower portion o f the chamber and should induce re a tively rapid convection (Rice and Eichelberger, 1976 ; Sparks and others, 1977). But because the density of basa t is high, any mi xing of heated si icic magma with basaltic material would inhibit convection. For example, the ensity decrease due to heating silicic magma by 1 00C could be offset by addition of only 2 volume percent basaltic material. The s ystem old probab y remain stratified, with the silicic and basaltic layers convecting separately, and mix ng imited to diffusion at the interface. O f course, if the basal tic material were less dense than the si i cic magma heating of silicic magma and mixing w 'th basal tic material would both decrease density of magma in the lower portion o f the sil1c ic ayer, resu t ing in strong pward flow. Despite the h gh ens ty o f phases crysta zed by basa tic magma and present n abundance n the basaltic xenoliths c o ntained n andesitic and acitic avas, it appears that the b k ensity of these xenoliths s a ctua less than tha of the avas in w ich they occur. The reason f o r this apparen contradiction is that the xenoliths invariably contain ab n ant, fine, rou uniform y stributed vesicles. A though vesicu a tion can occur at very sha o w depth d ring eruption, these vesic es may be o f deep, primary origin. As state before, the texture of the xeno i ths suggests rapid crystallization o f basaltic melt in contac t with cooler silicic magma Our ing the six hours required for a chil wave to propagate 0 e m into basa tic magma corresponding to roughl y the 1 argest common iameter of he xenoli hs, a hydration front could advance o n y 0 -2 em into the s icic magma, based on a a of Shaw ( 97 ) Thus, heat s

PAGE 195

0 N 0 I RECENT MAFIC L AVAS MAFIC L AVAS s 5 km I Figure 3 A north-south cross section through the center of the Medicine Lake Highland volcano (line o f cross section shown in Figure 2), with major features projected onto the profile. Vertical exaggeration of surface topography is 2 X Silicic magma chamber is shown at depth discussed in text and intermediate in size between the minimlJTl case, a small body at intersection of cone sheet caldera fractures (Heiken, 1978 ) and the maximum possible e x tent marked by position of flanking mafic vents. extracted from the basalt orders of magnitude faster t han water, so that the basalt behaves nearl y as a closed system with respect to water. Under low to moderate crustal pressures a vapor phase exsolves be cause water is contained entirel y within the melt phase of the system and the melt fraction decreases w ith crystallization from almost 1.0 to about 0.1. The result is a basaltic foam layer at the base of the silicic magma body that can float and mix. Floatation of the foam exposes fresh basal tic magma to c oo l silicic magma s o that the process of foam formation continues. Dispersal of foam through the silicic magma is due t o convec t ive flow, driven b oth by thermal expansion and the positive !!.V phase change. Photomicrographs showing the Glass Mountain xenoliths to be a crystal-rich foam are 'presented in F igure 4 DENSITY MEASUREMENTS AT GLASS MOUNTAIN To test the h ypothesis that mixing is driven by vapor ex solution, samples of Glass Mountain lava a nd xenoliths were collected for density measurements. Several pairs of lava and xenolith samples wer e col lected fran sites ( Figur e 1) near the vent, middle, and distal end of the longest lava stream, representing "rhyolite with inclusions," dacite near the transition to rhyodacite, and the most mafic dacite lava of the flow, respectively (Eichelberger, 1975). Density was determined by measuring dry a nd submerged weight of samples which ranged in size from 300 g to 600 g. Results are given in Table 1. The obsidian near the vent has no porosity and a very narrow range of density (smaller than analytical error), 2 38 + 0 .01 g!cml. Dacite from mid way down the stream is 2.49 g/cml. Dacite from the distal end is 2 55 g/cml reflecting its more mafic and crystal-rich character. Bulk density o f the xenoliths fall in a surprisingly narrow range, 2 15 + 0 04 g/cml. This density appears little affected by the vesicularity o f the host lava, since a xenolith in a pliTliceous (1.49 g/cm3 ) sample has a density of 2 10 g /cml In order to determine th m ss density of xenolith material, xenolith samples w re ground to a fin sand 185 to destroy the vesicles. Grain size of the sand averaged 100 \lm and ranged from \lm to 400 \lm. Exanination of polished grain mounts revealed that grinding opened at least of the pores. The san d was then weighed dry and submerged. The submerged measurement was made after the sample was subjected to a vacuum for 20 minutes, and water added under vaculJTl The resulting density of 2 79 + 0 .01 g / cml was expected for thi s predominantly plagioclase and pyroxene material. Using the average values for bulk and mass density gives a porosity of Thus the xenolithic material is much denser than the lava, but because of high porosity, the xenoliths would have floated in the magma chamber. This explains why the most xenolith-rich lava emerged first. CONDITIONS OF MIXING In view of the evidence that the xenoliths floated, and because densities as a function of pressure and temperature of the materials involved are known or approximately known constraints can be placed on the depth o f the magma chamber (Figure 5). Following previous definitions (Eiche lberger, 1979 ) Ps and Px are the mass and bulk densities of the basaltic xenoliths, so that Px = p8 (1-Xv) + Pvxv, wher e Py and XV are the densi'!y ancr vollJTle fracti on o f the vapor phase respectively Bul k density of the foam (p ) versus pressure was calculated at T = 900C f o r bu1k water content i n basaltic magma of a and by weight, using t h e following assl.ITlp-: tions: l The vapor phase is water. 2 The xenolithic material (exclusive of vapor) i s 10% melt and 90% crystals. This estimate is based on inspection o f thin sections of xenoliths (grain s i z e i s too small for accurate modal ana lysis) and is consistent with the observ ed mass density. In the general cas e degree of crystallization is dependen t largely on temperature contrast between the basaltic and silicic magmas

PAGE 196

A Figure 4 Photomosaics of basaltic xenolith in banded rhyodacite lava from Glass Mountain. Views are reflected light (A), t ransmitted light (8), and transmitted light with crossed polars (C). Bar length is 1 0 mm. Note abundant, 0 1 rom-diameter vesicles in xenolith ( A ) and absence of vesicles in most of the host lav a except for a large vesicle at the lower left edge of the xenolith. Lava at the bottom is contaminated with basaltic material and contains some vesicles. In 8 the xenolith is a dark round object with prominent white plagioclase crystals. Pyroxene is also a major phase, with olivine, oxides, and glass present in lesser amounts. The large vesicle associated with the xenolith is an irregular white area. Lava contaminated with basaltic material show s up as dark streaks. InC, the crystal-rich character of the xenolith is apparent. Contaminated lava contains more crystals than uncontaminated lava. 3 The xenolithic material is incompressible. In fact, the melt becomes less dense with increasing pressure because mor e water is retained and plagioclase and pyro xene com press, but the effects are small, opposite and nearly equal. The measured value of Pg was used, correcting for thermal expansion of plagioclase and pyroxene (Skinner, 1966). Data of Burnham and Jahns (1962) were used for solubility of water in melt, and data of Burnham and others ( 1969) were used for density of the vapor phase ( pV). The resulting curves (Figure 5) show the density of the foam formed during mi xing at a specified pressure and water content of basal tic magma Portions of Px versus pressure curves for constant porosity of foam are also shown. Density of magma i n the chamber before mixing can be estimated for elevated temperature and pressure by assuming that rhyolitic glass and melt behave in a manner similar to glass and melt in the system albite + water (Burnham and Davis, 1971). The petrologic 186 data indicate that magma in the chamber was entirely rhyolitic melt. Hence the density measurements for rhyolite obsidian were used. Water content strongly affects both thermal e xpansion and compressibility of rhyolite glass and melt. Available analyses give water contents of Glass Mountain obsidian of a few tenths of a percent 1941). Therefore, PR versus pressure at 900 C was calculated for 0 and l wt $ water. Knowledge that the xenoliths floated constrains xenolith density to those portions of the Px curves left (lower density) of the dry p curve. Estimates of water content for high alumina magma of the type involved in mixing range up ward to 5 wt. $ (Anderson, 1974 ; Rose and others, 1978). This constrains pressure to a region above (lower pres sure) the 5 wt.$ water pX curve. Since the xenoliths have most likely experienced continuously decreasing pressure since formation, it is unlikely that they ever possessed higher than their present (surface) porosity (XV) which constrains xenolith density to the right (higher density) of the XV= 0 23 curve.

PAGE 197

TABLE 1 Densities of Glass Mountain Samples Sample (Fig. 1) 1A 1B 1B 1C 10 1E 2A 2B 3B 3B 3B Averages 1B 1D 2A Average w a::: ::::> (/) 2 4 (/) 6 w a::: a.. 8 Bulk Density of Xenolith(g/cm3l 2 14 2 13 2 14 2 .20 2 .18 2 10 2 .10 2 23 2 .15 2 13 2 11 Den s i t y o f ho s t lava(g/cm3l 2 40 2 37 2 .37 2 .38 2 .39 1 49 2 49 2 .49 2 25 2 .42 2 55 2 01 2 49 2 .55 Sample Description Rhyolite obsidian Pumice Nonvesicular dacite Vesicular dacite Nonvesicular dacite Rhyolite obsidian Nonvesicular dacite, site 02 Nonvesicular dacite, site 1/3 Mass Density, Xenolith (g/cml) --px AT CONSTANT XH2o px AT CONSTANT Xy ----PR AT CONSTANT XH20 + MEASURED p, T 20c 2 79 2 79 2 .78 2 .79 0 .01 0 23 10 I 201-a.. w 0 e CALCULATED p, T = 90oc __ L__L __ 2.0 2.2 2.4 2.6 2.8 DENSITY (9 /cm3 ) Figure 5 Density versus pressure (depth) for materials from Glass Mountain. p is bulk density o f basaltic f oam (xenoliths), pR is density of the reservoir magma and pB is mass density o f xenolithic material. The possible region for the Glass Mountain magma chamber is shaded. Arrows show path for ascending basaltic magma with 3 water. See text for further discussion. 187

PAGE 198

These constraints bound the shaded region in Figure 5. If we knew that the basalt initially contained 3 wt.$ water, which is probably more reasonable since Medicine Lake basalt eruptions are much less gaseous than the Fuego eruptions for which the 5 wt .$ water estimate wa s made, then the possible region would be a small triangle bounded below by the p X curve for 3 wt.$ H20 to the right by the appropriate pR curve, and above by a horizontal line extending rrom the intersection of the pX curves for 3 wt. $ H20 and 23$ porosity (foam with XH. O = 0.03 above this line would have XV >0. 23). Regrettably, water content of basaltic magma is a major unknown, but it seems likely frcxn these considerations that the Glass Mountain magma chamber lies within 15 km and probably within 10 km of the surface. A possible path of ascending basal tic magma is shown as arrows in Figure 5 XH 0 = 0 .03 for basalt and XH = 0 to 0.01 for The basalt ascends the crust, slowly beccxning less dense as indicated by the steep path from 10 kb to 2 kb. Densities along this path were calculated using average Medicine Lake basalt (Anderson, 1941) and the model of Bottinga and Weill (1970). Note that at 2 kb the magma is still vapor undersaturated. However, at this depth it enters the Glass Mountain magma chamber, undergoes rapid crystallization while exsolving a vapor phase, and forms a foam with P 2 .74 g/cm3 XV and'\ g/cm1 If t fo m remained rigid but leaked after formation it would follow a constant porosity curve to the sur face. If it maintained an equilibrium por pressure and did not leak, it would follow a constant water curve and blow up during ascent. The pat.h shown is intermediate: some leakage and some expansion to the observed XV = 0.23. The Glass 1-buntain xenoliths probably did leak, since most xenoliths have a large vesicle at the lava/xenolith interface, as shown in Figure 4 In flat ion of the foam does work, energy for which is supplied largely by crystallization. At 2 kb, energy available from crystallization is near y two orders of magnitude greater than energy required for inflation. THE GLASS MOUNTAIN ERUPTION Important implications of this discussion for Glass Mountain are that water played an important role in magma evolution and that the chamber lies at shallow depth. A probable reconstruction of the eruption sequence is shown in Figure 6 Basaltic magma intruded the chamber and formed a foam which floated and mixed (in a mechanical sense) with magma in the chamber. The subsequent eruption first Figure 6. Probable eruption sequence at Glass Mountain. (1)Mafic magma intrudes shallow rhyolitic magma chamber and erupts nearby at High Hole Crater. (2) Vesiculation of chilled mafic magma forms foam which floats and mixes in the convecting chamber. (3) Vesiculation in chamber triggers eruption, which first taps hybrid dacite concentrated near roof of chamber. (4) Eruption continues with extrusion of rhyolite magma. Rhyolitic tephra which underlies Glass 1-buntain may have erupted early in this sequence, as an illll1ediate response to intrusion of mafic magma into the chamber, before mixing was far advanced. Alternatively, low density of water-rich rhyolitic magma at the top of the chamber may have prevented further rise of foam-rich currents, so that a small rhyolitic cap remained near the chamber roof a:1d erupted as tephra just before the hybrid dacite. 188

PAGE 199

.. t.he hybri whi c h was concentrated at the top o f the chamber Evi e nce that vapor e xsolution occurred ithin the chamber suggests that mi xing trigger ed the eruption. Assumin g that 0 2 km3 of the flow is a 1 : 3 mixture of basalt and rhyolite, then the volume of vapor exsolved within the chamber was 0 .01 km1 Some of this almost certainly leaked f r o m the f oam and dissolved in the strongly water undersaturated rhyolitic melt. Nevertheless, a rapid volLme i ncrease within the chamber could most easily be accommodated by e xtrusion of material from the cham ber. Thus, the mixing and eruption e vents were probab l y closely spaced in time, and may correlate with the e xtrusion of mafic magma 10 km to the south on the flanks of the Highland at High Hole Crater a nd its associated Burnt Lava Flow. Similar rhyolite and hybrid dacite have erupted before near Glass Mountain and els e where near the summit of the Highland. These may b e pro-ucts of repeated intrusion of basal tic magma into a relatively large, shallow, long-liv ed magma chamber beneath the caldera (Figure 3). This would account for the liquidus or superheated condition of some of these lavas, as indicated b y their lack of phenocrysts. Although water content of basal tic magma is a major unkno wn, the high initial density contrast between basal tic and reservoir magmas (Eichelberger. 9 7 9) constrains the Glass Mountain chamber to shallow depth. Both silicic composition and scarcity of phenocrysts in the reservoir magma contribute to this high density contrast. A remarkably similar high density contrast mixture erupted from M t Mazama form the Llao Rock dacite flow, just prior to the Crater Lake even t The subsequent climactic tephra eruption shows that Mt. Mazama overlay a large shallow reservoir containing crystal-poor silicic magma with mafic magma at the base of the reservoir. The G lass Mountain data suggest that this condition no w exists beneath the Medicine Lake Highland. Another consequence of this hypothesis is that m ixing between magma stored in a crustal reservoir and mafic magma feeding the reservoir is controlled by the water content of the mafic magma Mixing uring replenishment of crustal magma reservoirs offsets compositional effects of fractionation or crustal melting, so that eruption products of reser v oirs fed by wet mafic magma will tend to be intermediate in canposition. Volcanism resulting from dry mafic magma will be bimodal because mafic magma remains at the base of the reservoir without mi xing. Anderson (1 9 74) suggested that Medicine Lake basalts are drier than those to the west of Mt. Shasta. This may ac count for the bimodal character of volcanism east o f the High Cascades, farther from the plate b oundary and the source of water. ACKNO WLEDGMENTS This work was supported by the Office of Basic Energ y Sciences, U S D epartment o f Energy. David Mann prepared the thin section s hown i n Figure 4 REFERE N C E S CIT E D Anderson, A. T., Jr., 1974 The b e f o r e eruptio n w ater content o f sane high-al Lmi n a magmas: Bull. V alcanol., V. 37. p. 530-552. Anderson, C A., 1933. Volcanic h istory o f G lass Mountain, northern California: Am. Jour. Sci., 226, p. 485 -506. 189 Anderson, C A., 1 941 Volcanoes of the Medicine Lake Highland, California: Univ California Pubs. Geol. Sci.. V 25 p 347 -422. Bottinga, Y and Weill, D F., 1970, Densities of liquid silicate s ystems calculated from partial molar volLmes of oxide components: Am. Jour. Sci V 269 p 169-1 82 Burnham C W and Davis, N. F., 1971 The role of H20 in silicate melts: I. P V T relations in the s ystem NaA1Si308 H20 to 10 kilobars and 1000VC: Am. Jour. Sci., V 270 p 54 -79. Burnham C W., Holloway J R., and Davis, N F., 1969 Thermod ynamic p roperties of water to 1000C and 10 ,000 bars: Geol. So c Amer. S pec. Paper 132 p 1 -96. Burnham C W and Jahns, R H., 1962 A method for determining the solubility of water in silicate melts: Am. Jour. Sci., V. 260 p 721 -745. Chesterman, C W., 1955, Age of the obsidian flow at Glass Mountain, Siskiyou County, California: Am. Jour. Sci., V 253. p 418 424 Condie, K. C and H a yslip, D L., 1975 Young bimodal volcanism at Medicine Lake volcanic center, northern California: Geochim. Cosmochim Acta, v 39. p 1165 -1178. Eichelberger, J C., 1975 O rigin of andesite and dacite: Evidence of mi xing at Glass Mountain in California and at other cir cum -Pacific volcanoes: Geol. Soc Amer. Bull., V. 86, p 1381 1391 Eichelberger, J. C., 1978 Andesitic volcanism and crustal evolution: Nature, V 275, p 21-27. Eichelberger, J C . 1979, Vapor e xsolution and mixing during replenishment of crustal magma reservoirs: Submitted to Nature. Finch, R H., 1928, Lassen r e port No. 14 Volcano Letter, V 161 p 1 Friedman, I., 1968 Hydration rind dates rhyolite flows: Science, V 159, p 878-880. Heiken, G., 1978, Pliniant ype e ruptions in the Medicine Lake Highland, California, and the nature of the underlying magma : Jour. Vol canol. Geo hermal Res., V 4 p 375 -402. Mer tzman, S A., Jr., 1977 The petrology and geochemistry of the Medicine Lake Volcano, Cali fornia: Contr ib. Mineral. Petrol., V. 62 p 221 -247. Rice, A and Eichelberger, J C., 1976 Convection in rhyolite magma (abstract): EOS, V 57, p 1024. Rose W. I., Jr., Anderson, A. T. Jr., Woodruff, L G., and Bonis, S B., 1978, The October 19 7 4 basal tic tephra from Fuego Volcano: Description and history of the magma body: Jour. Volcano!. Geothermal Res., V 4 p 3 54 Shaw H R., 1974, Diffusion of H 7 0 in granitic liquids: Part I. E xperimental data; Part II. Mass transfer in magma chambers: in Hofman A W., Giletti, B J., Yoder, H S., Jr. and Yund, R A., eds., Geochemical transport and kinetics. Carnegie Institution of Washington, V 634 p 1 39-170 Skinner, B J., 1966 Thermal e xpansion: in Clark, S P.. Jr . ed., Handbook of Physical Constants. Geo l Soc Amer. Mem., V 9 7 p 78 96 Sparks, R J S., Sigurdsson, H., and Wilson, L., 19 77 Magma mixing: A mechanism for tr'iggering a cid explosive eruptions: Nature, V 267. p. 315-318.