The URL can be used to link to this page

Home My WebLink AboutGeology of Umnak & Bogoslof Islands Aleutian Islands AK 1959Alaska Energy Authority LIBRARY COPY ISSUED TO Z 2 = 2 Q = g = we r} obs 2 EN/ 038 | Geology of Umnak and Bogoslof Islands Aleutian Islands Alaska By F. M. BYERS, Jr. INVESTIGATIONS OF ALASKAN VOLCANOES GEOLOGICAL SURVEY BULLETIN 1028-L Prepared in cooperation with the Office, Chief of Engineers, U.S. Army ee UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1959 UNITED STATES DEPARTMENT OF THE INTERIOR FRED A. SEATON, Secretary GEOLOGICAL SURVEY Thomas B. Nolan, Director For sale by the Superintendent of Documents, U.S. Government Printing Office Washington 25, D.C. a ka PREFACE In October 1945 the War Department (now Department of the Army) requested the Geological Survey to undertake a program of volcano investigations in the Aleutian Islands—Alaska Peninsula area. Field studies under general direction of G. D. Robinson, were begun as soon as weather permitted in the spring of 1946. The results of the first year’s field, laboratory, and library work were assembled as two administrative reports. Part of the data was published in 1950 in Geological Survey Bulletin 974-B, “Volcanic Activity in the Aleutian Arc”, by Robert R. Coats. The rest of the data has been included in Bulletin 1028. The geologic investigations covered by this report were recon- naissance. The factual information presented is believed to be accu- rate, but many of the tentative interpretations and conclusions will be modified as the investigations continue and knowledge grows. The investigations of 1946 were supported almost entirely by the Military Intelligence Division of the Office, Chief of Engineers, U.S. Army. The Geological Survey is indebted to that Office for its early recognition of the value of geologic studies in the Aleutian region, which made this report possible, and for its continuing support. It re CONTENTS Abstract__...--.-- Introduction Fieldwork and acknowledgments. History and previous geologic work Russian period, 1741-1867____ United States Territorial period - Gangrentey 222) era Location, settlements, and accessibility_ Hydrography and topography. Climate, vegetation, and animal life__ General geologic setting The eastern Aleutian arc Area of this report. Rock names used in this report- Geology of Umnak__ General featu__-__ Albitized sedimentary and igneous complex Distribution and thickness__--__- Structure______.----------- Relation to younger rocks- - - Petrography --- Argillite___ Albitized intrusive rock. Keratophyre Age and correlation_-_ Plutonic rocks____..- Geologic relations. Petrography - Diorite__ Quartz diorite_ Quartz monzonite- Granophyre-_-_--- Altered contact rock __ Volcanic rocks of central Umnak- - Geologic relations____ Age and correlation. Petrography and chemical composition_-_ Volcanic rocks of Mount Recheschnoi General geologic features _- Hypersthene andesite Mafic phenocryst basalt of Kshaliuk Point-_.----- ‘Youngest hypersthene andesite flows. --- - Pyroclastic flow - ~~ Lava flows__ Page 267 268 269 270 270 271 272 272 272 275 276 276 278 279 285 285 285 285 286 286 286 286 287 287 288 289 289 289 VI CONTENTS Geology of Umnak—Continued Volcanic rocks of Mount Recheschnoi—Continued Rhyolite domes Quartz-olivine andesite flow_ Age of lavas of Mount Recheschnoi_ Petrography Hypersthene andesite _ Quartz-bearing olivine andesite___ Volcanic rocks of Mount Vsevidof- -- General geologic features _ -_ Age and correlation Petrography Basaltic andesite flows _ Andesite and latite flows. Andesitic and rhyodacitic pyroclastic bi Volcanic rocks of Okmok volcano_- Rocks of precaldera age--____ Ashishik basalt_.......------ Mafic phenocryst basalt__ Palagonitized pyroclastic rocks - Aphyric and feldspathic basalt Volcanic rocks of minor vents -_-- Vitreous andesite. Rhyolite. Plagioclase-olivine basalt of satellitic vents_ Crater Creek basalt. Age of precaldera rocks. Okmok volcanics of caldera age_ Agglomerate deposits __ Bedded ash_____.-------------------- Origin and emplacement of Okmok volcanics_ Age of the Okmok volcanics Basalt flows outside Okmok Caldera_ Mafic phenocryst basalt on caldera rim. Basalt flows on the arcuate ridges _-_ Basalt flow on Cape Aslik______.----_.------------------- Basalt flow near Cape Idak. Anorthite basalt flows near Inanudak Bay_ Rocks of postcaldera age within Okmok Caldera Rocks contemporaneous with caldera lake Basaltic flows later than caldera lake__ Petrography of Okmok Volcano. Olivine-rich basalt Plagioclase-olivine basalt __-_ Plagioclase-augite-olivine basalt Anorthite basalt__ Aphyric basalt and feldspathic basalt Andesite and latite__.__ Rhyolite. Palagonitized pyroclastic rocks_ Structure of Okmok Voleano-- Origin of Okmok Caldera CONTENTS vil Geology of Umnak—Continued Page Surficial deposits____------------------------------------------ 343 TW coos - 343 » Outivash - - 344 Beach deposits_ 344 Sand dunes - 346 Talus and other colluvium_ - 346 Alluvium - 346 Glacier ice. - 346 Erosional features _ - 3847 Crater Creek valley_- - 3847 Pedimentlike slopes on Okmok Voleano- - ---- _ 348 The Nikolski plain—Evidence of an older glaciation - _ 348 Geologic history_--- - 349 Geology of Bogoslof____- S 353 Geology of Bogoslof in 1947. - 353 Vent agglomerate of 1796__ _ 854 Hornblende andesite of Castle Rock-- -- 354 Hornblende basalt of Fire Island __ -- - 3855 Basaltic extrusive rocks of 1926-27 - 356 Beach deposits - 357 Structure. - 357 Geologic: history - 357 Literature cited = S6r Index... 2 2.2. 2 os a oe i epee ode teens 367 ILLUSTRATIONS Page Puate 39. Generalized geologic map of Umnak, Bogoslof, and part of 40. 41. 42. 43. 44, 45. 46. 47. 48. 49. Unalaska Islands, showing adjacent submarine topog- ering eee te et Se eee eee In pocket Geologic map and sections of southwestern Umnak Island. In pocket Geologic map and sections of northeastern Umnak Island. In pocket A, Photomicrograph of composite plagioclase crystal in al- bitized diorite. B, Photomicrograph of partly replaced albite phenocryst in hydrothermally altered keratophyre. Faces 290 Photomicrograph of plagioclase crystals veined and corroded by orthoclase in altered quartz diorite Follows 290 Mount Vsevidof from the south. Follows 290 View of east wall of Okmok Caldera_____.-------------- Faces 291 A, Nearly flat-lying, undisturbed flows of Crater Creek basalt exposed in the northwest wall of Crater Creek gorge. B, View northeast toward sea cliff northeast of Cape Chagak. Faces 306 A, Hand specimen of dark-gray welded andesitic agglomerate. B, Photomicrograph of small mass of basaltic andesite en- closed by rhyolite_...-.---- -- Faces 307 Section through cones C and D - In pocket Photomicrograph of anorthite phenocrysts in anorthite basalt flow_-- --- Faces 322 VIII CONTENTS Page Puate 50. Perspective diagram of Okmok Volcano viewed from the Reem b sc this tee ie es es eee ee ee ea Faces 323 51. Castle Rock, Bogoslof Island, from the southeast in 1947. Faces 354 Ficure 49. Index map of Alaska showing location of Umnak and Bogoslof et aia Laer ee Aceh ae ie pe eee at 269 50. Eastern part of Aleutian volcanic arc, showing positions of volcanoes and Aleutian Trench___--__.-.--_------------ 277 51. Stereographic projection of bisectrices and optic normals of anorthite phenocrysts in anorthite basalt__...___.__-__-_ 333 52. Profile through summit area of Okmok Volcano, showing hypothetical maximum and minimum profiles____-------- 342 53. Geologic map of Bogoslof in 1947__....._.._-.-.---.------ 353 54. Significant changes in Bogoslof Island between 1768 and 1947. 358 TaBLe 1, en 10. 11. 12. 13. 14, 15. TABLES Page Climatological data for two stations on Umnak Island, June 1942 to December 1948 - 275 Analyses of rocks from Umnak and Bogoslof Islands____-_____ 280 Petrographic summary of chemically analyzed rocks from Umnak and Bogoslof Islands_-_-_.-_------------------ In pocket . Petrographic analyses of plutonic rocks (modal analyses in Wokiine percenty tase nae a See 291 Partial analyses of hydrothermally altered (potassium felds- pathized) volcanic rocks near Bottleneck Pass_____-.------ 296 Petrographic analyses of specimens of hyperthene andesite of ROUT IGRECOSADMROL oOo WL rubra eevee wee arweate sre 300 Petrographic analyses of three mutually perpendicular thin sections of quartz-bearing olivine andesite___.._.._...----- 303 . Comparative petrographic data: rhyodacite pumice of Mount Vsevidof and pumice fragments in ash beds of archeological Cee ea a eee ee eee ee ee ee a a eee 305 . Petrographic analyses of basaltic andesite, andesite, and latite flows.of Mount: Veevidof 0 02 306 Petrographic analyses of rocks in dark andesitic scoria bed of Mount Vsevidof cone. 307 Petrographic analyses of agglomerate and ash marker She'Okmok volcanics 00000 tie ge 318 Petrographic analyses of mafic phenocryst basalt, Okmok Nae eee eee eNO ELE USHA aes LG LU 331 Optical data on anorthite phenocrysts in anorthite basalt near Doo eee ee OE My 334 Petrographic analyses of vitreous andesites and latite of Okmok eS 335 Petrographic analyses of rhyolite specimens and xenolithic inclusion, Okmok Voleano:-..02. 020010000 336 INVESTIGATIONS OF ALASKAN VOLCANOES GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS, ALASKA By F. M. Byers, Jr. ABSTRACT Umnak and Bogoslof Islands are in the eastern part of the Aleutian Island are, a geanticlinal uplift paralleled on the southerly (convex) side by the Aleutian Trench. Umnak Island, 675 square miles in area, is separated into northeastern and southwestern parts by a constriction in the central portion. The rocks consist of late Tertiary and Quaternary volcanic rocks that rest on a basement complex of probable early to middle Tertiary plutonic and low-grade meta- morphic rocks. Bogoslof, 25 miles north of Umnak, is composed almost entirely of historic lavas. - The oldest geologic unit, an albitized sedimentary and igneous complex probably largely of early to middle Tertiary age, is exposed near sea level on southwestern Umnak and includes folded argillite and tuff, keratophyre flows, and albitized intrusive rocks. Post-Oligocene plutonic rocks, largely quartz diorite, intrude the albitized sedimentary and igneous complex. Granophyre dikes cut the larger body of quartz diorite and are associated with a wide aureole of potassium feld- spathized quartz diorite. The volcanic rocks of central Umnak include both potassium feldspathized volcanic rocks that are probably prepluton in age and unaltered lavas of probable late Tertiary to early Quaternary age. The unaltered lavas rest unconformably on the albitized sedimentary and igneous complex. Two Quaternary volcanoes, Mount Recheschnoi and Mount Vsevidof, overlie with an erosional unconformity the older rocks on southwestern Umnak. Mount Recheschnoi is a deeply dissected ridge remnant of a composite stratovolcanic cone composed largely of hypersthene andesite. Satellitic lavas of Mount Recheschnoi comprise mafic phenocryst basalt, extruded early in the history of the volcano, rhyolite domes, and a quartz-olivine basalt flow of early Recent age. Mount Vsevidof is composed of an older sequence of basaltic andesite and sparsely hypersthene-bearing andesite flows, an intermediate sequence of pyroclastic rhyodacite pumice and andesite scoria beds of a culminating summit eruption, and younger andesite and basaltic andesite flows, including a historic(?) latite flow. Okmok Volcano on northeastern Umnak, in contrast to the andesitic strato- volcanoes of southwestern Umnak, consists largely of basalt flows and has a central caldera, Okmok Caldera, 6 miles in diameter. The rocks of Okmok Voleano are conveniently divided into precaldera rocks, Okmok volcanics, the deposits associated with the caldera-forming eruption, and postcaldera rocks. The precaldera rocks include the Ashishik basalt of latest Tertiary and Quater- nary age and volcanic rocks of minor vents, including vitreous andesite, rhyolite, 267 268 INVESTIGATIONS OF ALASKAN VOLCANOES and plagioclase-olivine basalt of satellitic vents. The youngest precaldera unit is the Crater Creek basalt of latest Pleistocene or early Recent age. The Okmok volcanics of early Recent age are composed of a few very thick beds of unsorted andesitic agglomerate, welded in the middle parts of beds, and an overlying unit, consisting of sorted ash, bomb, and lapilli beds. Postcaldera rocks within Okmok Caldera include early postcaldera pyroclastic rocks, plagioclase basalt flows of cones C and D, and bedded voleanic sediments. Subaerial basaltic flows later than the caldera lake include those from several vents within the caldera, the latest flow having been extruded as recently as 1945. The undissected condition of several cones and flows outside Okmok Caldera suggests that they are also postcaldera in age. The structure of Okmok Volcano is chiefly that of a low shield voleano with gentle outward dips. A fault is exposed for a short distance at the periphery of Okmok Caldera, and the measurable displacement is toward the caldera. The steep infacing cliff of Okmok Caldera is therefore inferred to be the scarp of a caldera ring fault. A few faults are radial to Okmok Caldera. The north and the east arcuate ridges may have originated in any one of three ways; they may be receded fault scarps of an older caldera ring fault, they may have formed concomitantly with the collapse of Okmok Caldera, though there are several objections to this hypothesis, or they may represent accumulations of volcanic debris erupted from vents localized along a fissure beneath the arcuate ridges them- selves with little or no fault movement. The first hypothesis, modified by: the third, seems most plausible. Regardless of which of two extreme assumptions is made in regard to size and shape of the missing summit of Okmok Volcano, the total volume of lithic frag- ments in the Okmok volcanics is not nearly sufficient to account for the missing summit. Hence, the missing portion of the summit must have collapsed inward to form Okmok Caldera. Surficial deposits of Umnak comprise older till, younger till, glacier ice, glacial outwash, alluvium, talus, beach deposits, and dune sand. Older till of latest Wisconsin or early Recent age forms a series of end moraines several miles in front of the present glaciers. Younger till is confined to moraines within a mile of the present glaciers and probably records the most recent glacial advance, 100 to 300 years ago. Beach deposits are in as many as 20 concentric beach ridges parallel to the present beach. The older till is covered in part by these beach ridges, which are believed to date from the high, 3-meter stand of sea level, generally ascribed to the post-Wisconsin thermal maximum. Hornblende andesite and basalt exposed in 1947 on Bogoslof Island are all of historic age. The oldest rock, representing the explosive phase of the 1796 erup- tion, is an andesitic vent agglomerate in fault contact with the hornblende andesite dome also extruded in 1796. The hornblende basalt dome of Fire Island was extruded in 1883 following a violent eruption. Domes were erupted in 1906, 1907, 1910, but no specimens of these domes were available for study, owing to ex- plosive disruption of the domes soon after they formed. Basaltic agglomeratic ash erupted in 1926 and a basalt dome extruded in early 1927 formed most of the area of the island in 1947. From 1927 to 1953 Bogoslof was steadily reduced in size by marine erosion. INTRODUCTION In June 1945, a spectacular eruption of Okmok Volcano on the northeast part of Umnak Island in the Aleutian Islands (fig. 49) threatened a military base on the eastern slope of the volcano. Al- GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 269 Fiaure 49.—Index map of Alaska showing location of Umnak and Bogoslof Islands, though this eruption did no damage, military authorities were alerted to the fact that personnel and installations in the Aleutian region are constantly exposed to possible danger and damage from its many active volcanoes. In October 1945, the United States Geological Survey at the request of the Alaskan Defense Command, United States Army, began a program of systematic geological and geophysical investiga- tion of the Aleutian Islands. As the site of a large base and of a cur- rently erupting voleano, Umnak was one of the first islands investi- gated. A restricted report to the Department of Defense in 1947 emphasized the possibilities of damage by volcanic eruptions to the local military base. The present report deals primarily with the stratigraphy and petrography of rocks on Umnak and its tiny but famous neighbor, Bogoslof Island; the glaciers and glacial geology also are briefly discussed. FIELDWORK AND ACKNOWLEDGMENTS This report is based on fieldwork during the summers of 1946 and 1947. The writer was assisted in 1946 by D. M. Hopkins and Bernard Fisher, geologists. Hopkins mapped most of the area within Okmok Caldera, (pl. 39) and part of the description and interpretation of the postcaldera deposits is based on his work. Fisher and two Army men 270 INVESTIGATIONS OF ALASKAN VOLCANOES were drowned in Umnak Pass on June 22, 1946, when they attempted to land their power dory on Ship Rock, a small island off the coast of Umnak. The loss of Fisher dealt a severe blow to the progress of the fieldwork. In 1946 the geologists were aided by Kenneth L. Wier, geophysicist, and by field assistants W. E. Hassman, O. L. Smith, G. A. Hill, and J. M. Borglum. Fieldwork in 1946 was confined to northeastern Umnak. In 1947, the writer, accompanied by W. S. Ogburn, field assistant, completed mapping of northeastern Umnak, made a 6-week reconnaissance of southwestern Umnak, and spent 1 day on Bogoslof. In 1948 the writer, accompanied by Howard A. Powers, supervising geologist, spent several days reexamining parts of northeastern Umnak. Ray E. Wilcox, presently geologist of the Geological Survey, kindly made available specimens and information he had collected on Umnak during 1945, while serving in the Army. Many officers and enlisted men, formerly stationed at Fort Glenn, cooperated fully in providing transportation and supplies. Maj. Jim McCall, commanding officer of Fort Glenn during the 1946 field season, Maj. Clarence J. Jackson, commanding officer during the 1947 field season, and Capt. Laverne E. Dye, adjutant during the 1947 season, were especially helpful. At Nikolski, H. Frank Barnett, and Harvey B. Bell gave material aid in August 1947, during two rainy periods. During 1947 and 1948, Harald Drewes gave valuable assistance in the office. The writer received many helpful suggestions in 1948 at the University of Chicago from Tom. F. W. Barth on optical study and chemical analysis of the lavas from Umnak. HISTORY AND PREVIOUS GEOLOGIC WORK RUSSIAN PERIOD, 1741-1867 Umnak Island was first sighted in 1741 by Capt. Alexei Chirikof of the St. Paul, a ship of Fleet-Captain Vitus Bering’s expedition to North America. The name, Umnak, was the native name (Baker, 1906, p. 650). No Russian investigations were made on Umnak during the latter part of the 18th century, although Stephen Glottof and other Russian fur traders visited Umnak during this period (Baker, 1906, p. 36). William Coxe (1787, p. 141-142 and 173-181) gives interesting accounts of the early Russian traders and of the natives. The most complete descriptions of Umnak and Bogoslof during the Russian period are those of Veniaminof (1840) and Grewingk (1850). The first detailed map of Umnak appeared in the marine atlas of Capt. M. D. Tebenkoff (1852). A rocky islet on the site of Bogoslof was first sighted in 1768 and named Sail Rock by the Russians; it was called Ship Rock by Captain Cook and other early navigators. The islet disappeared between GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 271 1884 and 1891. The name Joann Bogoslof (John the Apostle) was applied by the Russians to an island formed a little south of Ship Rock by an eruption of viscous lava in 1796 (Merriam, 1902, p. 291-294). UNITED STATES TERRITORIAL PERIOD In 1867, the United States purchased Alaska, but no survey or scientific study of Umnak by Americans was made until 1883, when the birth of a new volcanic dome, less than a mile north of Bogoslof Island, attracted interest. Dall (1884 and 1885) and Diller (1884 and 1885) used the names Grewingk and New Bogoslof, respectively, in describing the new dome, the remnant of which is now called Fire Island. Merrill (1885, p. 31-33) contributed petrographic descrip- tions and two analyses of specimens from Fire Island. Members of the Harriman Alaska Expedition stopped at Bogoslof in 1899 and collected specimens from Castle Rock. Merriam (1902), a member of that expedition, published the first complete paper in English on the history of Bogoslof during the period 1768-1900. Marcus Baker’s Geographic Dictionary of Alaska (1906) is a valuable source of infor- mation, not only about origin of names but also about earlier workers and their publications, especially maps and charts. Volcanic activity at Bogoslof was renewed during 1906-1907. Jaggar (1908) published a history of Bogoslof and an account. of his landing of August 7, 1907. Robert Dunn visited both Bogoslof and northeastern Umnak and was the first (Dunn 1908) to publish the name Okmok, supposedly the native name for this voleano. Sidney Powers (1916) gave an account of the 1910 eruption of Bogoslof and the extrusion of the Taboma Peak dome and Jaggar (1930) gave an account of the 1926-27 Bogoslof activity. Lukens (1936) gave a description of Bogoslof Island as it was in 1935. Paul A. Smith (1937) using Coast and Geodetic Survey data, described the submarine canyons near Bogoslof. On Umnak, Waldemar Jochelson, an anthropologist, made excava- tions at ancient village sites in connection with his studies on the origin of the Aleut (Jochelson, 1925, 1933). Hrdlitka (1945) also visited Umnak in connection with his anthropological studies of the Aleut. Umnak Island was briefly visited in 1931 by S. R. Capps (1934) of the U. S. Geological Survey, and later Hultén (1937) made botanical collections near Nikolski. An eruption in Okmok Caldera in 1931 was reported by Jaggar (1932). During World War II, from 1942 to 1945, Okmok Caldera became known to military personnel stationed at Fort Glenn. Dean Freiday (1945, p. 453-454) described Okmok Caldera by means of photographs and speculated on its origin. Ray E. Wilcox (1945 and 1946), while an officer at Fort Glenn, prepared file reports on the 1945 activity. 272 INVESTIGATIONS OF ALASKAN VOLCANOES G. D. Robinson and Howel Williams as consultants to the U.S. Army also observed the 1945 eruption and advised the Alaskan Defense Command that Umnak Airbase was not threatened. Byers, Hopkins, Wier, and Fisher (1947) prepared an administrative progress report on volcano investigations begun in 1946 on northeastern Umnak. Byers and Brannock (1949) sampled hot springs and fumaroles and described volcanic activity on Umnak Island during the period 1946-48. GEOGRAPHY LOCATION, SETTLEMENTS, AND ACCESSIBILITY Umnak and Bogoslof lie in the eastern Aleutian Islands, Alaska (fig. 49). Most of their area is enclosed by the meridians 168° and 169° W. and parallels 53° and 54° N. Umnak Island is bounded on the north by the Bering Sea and on the south by the Pacific Ocean. Bogoslof Island is entirely within the Bering Sea, 25 miles north of Umnak and slightly off the main trend of the eastern Aleutians. Umnak is 75 miles long and has an area of about 675 square miles. Bogoslof was only slightly more than a mile long in 1947 and had an area of less than a quarter of a square mile. The chief settlements are Cape Air Force Base (Fort Glenn prior to 1947) on northeastern Umnak Island and the Aleut village of Nikolski on the southwestern part (pl. 39). In 1947 a group of about 12 U. S. Coast Guardsmen operated the Cape Starr Loran Station, 4 miles west of Nikolski, a village with a population of several score Aleuts, a school teacher, and the manager and a few employees of a livestock company. No one has lived on Bogoslof since 1945, when a World War II outpost was abandoned. Umnak Island can be reached by Government vessel in about 1 week’s time from Seward, Alaska. The dock of Cape Air Force Base can accommodate small- to medium-sized vessels, but at Nikolski all supplies and passengers must be lightered ashore. Bogoslof Island is accessible only by small boat in calm weather. Roads, in bad repair, encircle northeast Umnak Island except on the northwest side. Southwest Umnak has only wagon roads and trails in the vicinity of Nikolski. HYDROGRAPHY AND TOPOGRAPHY Umnak, like nearly all the islands of the Aleutian arc, rests on a flat-topped submarine ridge, the Aleutian Ridge (Studds, 1950, p. 787). Bogoslof is the top of a submarine volcano rising from oceanic depths, 25 miles north of the Aleutian Ridge (pl. 39). The general submarine features of the Bering Sea, the Aleutian Ridge, and the parallel Aleutian Trench to the south have been described by Murray (1945). GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 273 The Aleutian Ridge near Umnak Island reaches a maximum width of 35 miles along a line through northeastern Umnak but tapers to 22 miles opposite the southwestern end of Umnak; its upper surface is almost entirely at depths of Jess than 100 fathoms (pl. 39). The break in slope between the ridge or shelf area and the north wall of the Aleutian Trench is at about 100 fathoms, 12 to 30 miles south of the coast of Umnak. The north wall of the Aleutian Trench is dissected by short submarine canyons, which extend to depths ranging from 1,500 to 2,000 fathoms. The average gradient of the canyons is about 600 feet per mile. Below 2,000 fathoms the north wall of the Aleutian Trench has a more gradual slope. The bottom of the Aleutian Trench is 4,030 fathoms at about 100 miles south of Umnak (Murray, 1945, fig. 9) or about 50 miles south of the southern boundary of the area shown in plate 39. Umnak Island is separated from Unalaska Island to the east by Umnak Pass, a strait 4 miles in minimum width and 26 fathoms in minimum depth. Samalga Pass, 20 miles wide and about 125 fathoms deep, separates Samalga Island off the southwestern tip of Umnak from Chuginadak, the nearest of the Islands of the Four Mountains, just west of the area shown in plate 39. The submarine slope at a depth of about 100 fathoms (pl. 39) steepens off the northwest coast of Umnak within a few miles ot the shoreline at the widest parts of the island. The gradient is about 600 feet per mile from 100 to about 600 fathoms and the submarine slope is incised by many steep V-shaped canyons as much as 1,000 feet deep. Below about the 600-fathom contour the submarine slope becomes less steep and also less regular. Submarine canyons sepa- rated by spurs extend northwesterly for several tens of miles to join a much larger submarine canyon with a westerly gradient (extreme top part of pl. 39). This canyon, 6,000 to 9,000 feet deep, is the largest submarine canyon of the eastern part of the Bering Sea (Murray, 1945, pl. 1; Smith, 1937). The largely submerged volcanic cone of Bogoslof rests on one of the northwest-trending intercanyon spurs. Umnak Island is divided into northeastern and southwestern parts at the narrow isthmus where Inanudak Bay makes a deep reentrant into the northwest coast. The two parts of Umnak have marked differences in topography. Southwestern Umnak is much more rugged, with two large volcanic mountains rising to nearly 7,000 feet. Northeastern Umnak Island is almost wholly occupied by Okmok Volcano, a low shieldlike volcano, less than 3,500 feet in altitude. Southwestern Umnak Island contains the glaciated voleanic moun- tains, Recheschnoi and Vsevidof, in the northern part and the Nikolski plain in the southern part (pls. 39 and 40). Mount Recheschnoi 274 INVESTIGATIONS OF ALASKAN VOLCANOES (altitude, 6,510 feet) is a deeply dissected ridge with radial flat-topped spurs separated by deep, glacier-filled valleys. Seven glaciers, 2 to 4 miles long, extended in 1947 from near the summit of the ridge to an altitude of a few hundred feet. Outwash plains in the valleys extend seaward from the glacier fronts. Mount Vsevidof, west of Mount Recheschnoi, is a symmetrical volcanic cone rising to about 7,000 feet above sea level. The crater of Mount Vsevidof, 0.8 mile across, is filled by a glacier with two distributaries, one north and the other east of the crater. The distributary glaciers have breached the crater wall and have incised gorges 200 to 400 feet deep. The larger, east distributary joins a glacier flowing southwest from Mount Rechesch- noi. The composite glacier then flows southward 2 to 3 miles to a large end moraine at an altitude of about 1,000 feet. This glacier drains to the two forks of Black Creek, the largest stream on south- western Umnak Island. The Nikolski plain southwest of Mount Recheschnoi and Mount Vsevidof is a rolling surface with a relief of a few hundred feet and about 300 feet in average altitude. Small lakes, less than a mile long, abound on the surface of the plain. Small islands, such as Ananiuliak, Samalga, and Vsevidof, off the coast of southwestern Umnak, are apparently remnants of the plain, which were isolated by marine erosion. Okmok Volcano on northeastern Umnak is a shield-type volcano with a large central caldera, 6 miles in diameter. The rim of the caldera ranges in altitude from 1,600 to about 3,200 feet (pl. 39). Several cinder cones and craters, designated on plate 41 by letters A through H, are on the floor of Okmok Caldera. The caldera is drained toward the northeast by Crater Creek through a gorge 600 feet deep. North and east of Okmok Caldera, two arcuate ridges, roughly concentric to the walls of Okmok Caldera, are herein referred to as the north arcuate ridge and the east arcuate ridge, respectively. Radial spurs separated by broad short valleys are typical of the slopes of Okmok Volcano above altitudes of about 1,000 feet. At lower levels, slightly dissected plains slope 100 to 200 feet per mile sea- ward, except along the northwest coast between Fox and Aguliuk Points (pl. 39), where slopes are steeper and more dissected. The slopes on the lower flanks of Okmok Volcano have been incised with V-shaped gullies, as much as 200 feet deep. In exposed places where the plains intersect the coast, the sea cliffs are 100 to 250 feet high. Northeast of Okmok Volcano, Mount Idak surmounts the south- western end of a small lava plateau that slopes northward toward the Bering Sea. On the south flanks of Okmok Volcano are several parasitic cones in different stages of dissection. The largest of these, and also the highest peak on northeastern Umnak Island, is Mount Tulik, with an altitude of 4,111 feet. Low, rugged volcanic moun- GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 275 tains, rising about 2,500 feet above sea level, occupy the narrow isthmus that joins the northeastern and southwestern parts of Umnak Island. This mountainous area is herein called the central Umnak area. CLIMATE, VEGETATION, AND ANIMAL LIFE The climate (U. S. Coast Pilot, 1947, p. 36-43) is marine subarctic, characterized by overcast skies, high winds, and small variations of diurnal temperature. Overcast skies and fog with considerable rain in the mountainous areas occur during the late spring and summer months, owing to warm southerly winds off the Pacific Ocean. Winds during the summer are usually accompanied by fog and drizzle and frequently exceed 30 miles per hour. Occasionally the winds reach velocities as high as 70 miles per hour, but winds of extreme velocity during the summer are probably williwaws, whose accelerated velocity is caused by local topographic conditions. Temperatures during the summer are cool, ranging from 46° to 68° F., on Umnak. After the September equinox, the average wind direction shifts from westerly to northwesterly, and winds of over 100 miles per hour are not un- common. Fog and low overcast give way to broken overcast with occasional clear days. Drifting snow that blocks roads becomes a problem in winter. Winter temperatures at sea level in the vicinity of Umnak and Bogoslof Islands average slightly above freezing and the seas are ice-free throughout the year. Table 1 summarizes climatological data at Cape Air Force Base and Nikolski from June 1942, to April 1948. Taste 1.—Climatological data for two stations on Umnak Island, June 1942 to December 1948 [Compiled from 11th Weather Squadron, 1950, Chart 19, and earlier files of U. 8. Air Force) Cape Air Nikolski ! Force Base ‘Temperature (degrees Fahrenheit): Mean annual --__. Maximum annual_ Minimum annual. Precipitation (inches) Mean annual -__ Annual snowfall, unmelted. Greatest depth on ground ---_ Number of days with at least Fog: Annual percentage of occurrence- Wind (miles per hour): Average velocity -- EA 228 3 3 ong coo g 8 2 Seo af. Bs SEESE Eee Bon ME Bicee wee ge. Be B coerce Cc. Scattered or broken Overcast or obscured. 1 Data cover only a limited period from October 1942 to April 1945. 489292—59—_2 276 INVESTIGATIONS OF ALASKAN VOLCANOES A general description of the vegetation of the Aleutian Islands is given by Walker (1945, p. 63-71). Grasses, sedges, and flowering plants cover the slopes below an altitude of 1,000 feet on northeastern Umnak. On southwestern Umnak, grasses and flowering plants grow in protected places at altitudes up to nearly 2,000 feet, and several types of edible berries grow at altitudes below 1,000 feet on the south- facing slopes of Mounts Recheschnoi and Vsevidof. In 1947, Bogoslof had very little vegetation other than a few grasses. Umnak and Bogoslof, like other Aleutian Islands, are characterized by a profusion of marine mammals and birds. (See Clark, 1946, p. 31-61.) Fish and marine invertebrates likewise are abundant. By comparison the terrestrial fauna is insignificant. No reptiles or amphibians are known in tbe Aleutian Islands and the most common land mammal is the red fox. Reindeer, imported from Siberia, were stocked on Umnak about 1900 (Lantis, 1950, p. 28-29), and in 1946 several hundred of these animals, then in the wild state, inhabited the northeastern part of the island. Sheep are raised on southwestern Umnak, where they produce a premium crop of wool. GENERAL GEOLOGIC SETTING THE EASTERN ALEUTIAN ARC Umnak and Bogoslof Islands are a small segment of the Aleutian Island arc, on the northern edge of the Pacific Ocean. Most Pacific island arcs (Umbgrove, 1947, p. 144-216; Hess, 1948, p. 417-445) are arcuate geanticlinal ridges with submarine trenches or downbuckles on their convex oceanward sides. Menard and Dietz (1951) provide hydro- graphic evidence that the Aleutian Trench has sunk, probably in late Tertiary and Quaternary time. The Aleutian arc is a single island arc at Umnak and adjacent is- lands, but becomes a double are to the northeast (Umbgrove, 1947, p. 189). In the Umnak portion of the arc, the geanticlinal ridge, marked by the line of Quaternary volcanoes, remains close to 100 miles from the axis of the Aleutian Trench (fig. 50). Bogoslof and Amak Islands are partly submerged volcanoes 30 miles north of the main Aleutian geanticlinal ridge and hence are exceptions to this general relationship. Northeast from Umnak, the line of Quaternary volcanoes diverges from the axis of the Aleutian Trench and is 200 miles away near Kodiak Island (fig. 50). The continental shelf south of the shoreline of the inner (northern) arc likewise widens from about 20 miles, south of Umnak, to 150 miles, southeast of the Alaska Peninsula near the Katmai Volcano. Where the shelf widens, an outer (southern) chain of nonvolcanic islands begins with Sanak, and extends northeastward. ~ N s -Pribilof Islands a % G 4 % $ % ste ‘Unimak Island ¢ ES EXPLANATION ‘Akutan Island. S oe Submarine entoars in fathome Alen at Vater Murrey, 13) = a Quaternary vleano 1 FiGuRe 50.— Eastern part of Aleutian voleanic are, showing positions of volcanoes and Aleutian Trench. Submarine contours selected to show positions of Aleutian Ridge and Aleutian Trench. 110 SANVISI NVILAGTIV ‘SANVISI JOTSODOd GNV MVNIWA JO XADNOTORD 278 INVESTIGATIONS OF ALASKAN VOLCANOES The geology of the outer, southeasterly islands of the double are is known chiefly from the work of S. R. Capps (1937) on Kodiak Island. Capps (1937, pl. 2) has mapped sedimentary rocks of four ages: an early Mesozoic and older, a late Mesozoic, an early Tertiary, and alate Tertiary. The rocks are progressively more deformed with increasing age. In general the beds of Tertiary age crop out nearer the Aleutian Trench than the rocks of Mesozoic age. These sedi- mentary rocks strike northeast, parallel to the Aleutian Trench, and are generally separated by great, northwesterly-dipping faults running the entire length of Kodiak Island (Capps, 1937, p. 136). The Alaska Peninsula, on the inner concave part of the Aleutian double arc, contains folded thick rocks of Mesozoic and Tertiary age (W. R. Smith, 1925, pl. 4; Atwood, 1911, p. 28; Knappen, 1929, pl. 6), which extend as far southwest as Pavlof Bay (P. S. Smith, 1939, pl. 1). Atwood (1911) mapped Jurassic to Tertiary sedimentary rocks, which have been bowed up into a broad anticlinorium. Thrust faults that strike northeast and dip northwest have been reported in the Aniakchak district (Knappen, 1929, p. 207). Southwest of Pavlof Bay (fig. 50) the thick sedimentary formations of Mesozoic and Tertiary age are not exposed. Between Pavlof Bay and the end of the Alaska Peninsula (fig. 50) only minor thicknesses of late Tertiary sedimentary rocks interbedded with basic tuffs crop out at low altitude beneath the Quaternary volcanoes (Kennedy and Waldron, 1955, p. 6-7; H. H. Waldron, 1947, oral communication). The Tertiary and older rocks probably are mostly concealed beneath the Quaternary volcanoes of Unimak, the easternmost island of the single arc. The islands of Akun, Akutan, and Unalaska, west of Unimak, are composed almost entirely of volcanic rocks of Cenozoic age and their sedimentary derivatives. On Unalaska, the next island east of Umnak, lower Miocene tuffaceous sedimentary rocks and associated volcanic rocks are cut by quartz dioritic plutonic rocks. AREA OF THIS REPORT Umnak and Bogoslof Islands are composed almost entirely of igneous or altered igneous rocks of Cenozoic age. The 2 islands are divisible into 3 areas that with few exceptions differ significantly in geologic, features and petrographic type. These areas are southwestern Umnak Island, northeastern Umnak Island, and Bogoslof Island (pl. 39), which are shown on separate geologic maps in plate 40, plate 41, and figure 53, respectively. Northeastern Umnak is arbitrarily separated from southwestern Umnak at the narrowest part of the island. South- western Umnak (pl. 40) contains the oldest rocks, a complex of albitized sedimentary and igneous rocks of possibly diverse geologic ages, mostly Tertiary, but possibly including still older rocks, Dioritic GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 279 plutonic rocks, intrusive into this complex, were also found only on southwestern Umnak. Altered Tertiary volcanic rocks that crop out in the central part of Umnak underlie Quaternary volcanoes on both northeastern and southwestern Umnak. The petrography, composi- tion, and physiographic expression of the volcanic rocks of Quaternary age in the northeastern part of Umnak differ from those of south- western Umnak; the volcanic rocks of Bogoslof Island, which are entirely of historic age, differ somewhat in composition from those of either southwestern or northeastern Umnak. Geologic dating of most of the rock units is uncertain and based chiefly on inferences from contact relationships, degrees of alteration, and, to less extent, degree of dissection of constructional surfaces, Tertiary fossils recently discovered on nearby Unalaska Island (pl. 39) have supplied some evidence, from which geologic ages of the older rocks may be inferred. ROCK NAMES USED IN THIS REPORT Names of volcanic rocks in the suite being investigated are based on the total content of silica in the rock, following in general (and extending to higher silica percentages), the system of Williams (1950 p. 234-235). As defined in this report, basalt contains less than 54 percent silica, equivalent to about 5 percent normative quartz; andesite contains more than 54 and less than 63 percent silica, approxi- mately equivalent to the range 5 to 14 percent of normative quartz; latite contains between 63 and 65 percent of silica, approximately equivalent to the range 14 to 17 percent of normative quartz; rhyoda- cite contains about 65 percent of silica, equivalent to about 20 percent of normative quartz; and rhyolite contains more than 72 percent of silica, or more than 25 percent of normative quartz. (For chemical analyses of the rocks see table 2.) Porphyritic rocks are named by applying the appropriate general rock name as determined by the silica content modified by mineral names, particularly those of mafic phenocrysts, to indicate the pheno- crysts in the rock. The names applied to many of the porphyritic rocks on the maps and in stratigraphic descriptions are shortened versions of names based on chemical and petrographic analyses; that is, hypersthene andesite for hypersthene-bearing labradorite andesite. One term, mafic phenocryst basalt, found useful in the field mapping, is retained in the report as a map unit and general rock group. It in- cludes porphyritic basalts that contain visible phenocrysts of olivine, augite, or both together in excess of about 2 percent by volume. Abundant plagioclase phenocrysts in association with mafic pheno- crysts are not generally considered in naming the rock, especially if less than 10 percent, because plagioclase phenocrysts are abundant in many of the rocks described. TaBiE 2.—Analyses of rocks from Umnak and Bogoslof Islands [Analyst indicated by initials: S. M. Berthold (SMB); W. W. Brannock (W WB); R. L. Hill (RLH), and Leonard M. Shapiro (LMS), of rapid-analysis laboratory (see Shapiro and Brannock, 1952), U. 8. Geological Survey; W. W. Hommel (W ‘H), of Ledoux and Co.; M. K. Carron (MKC) and E. J. Tomasi (EJT) of U. S. Geological Survey; F. M. Byers, Jr. (FMB), of U. 8. Geological Survey, on leave at University of Chicago laboratory; andPaul R. Barnett (PRB), Arthur Chodos (AC), K. J. Murata (KJM), A. T. Myers (ATM), spectrographers, U. 8. Geological Survey.) Northeastern Umnak 1 1A 2 3 4 5 6 7 8 9 10 ul 2 w | uw | 15 16 wv | 1s SMB SMB wws| FMB |rmB|sMB| FMB wws| FMB FMB | FMB | SMB | SMB RLH | Ey? |wwH| RLH| MKO | LMS | WWH|WWH| RLH| WWH| WWH | WWH| LMS | WWH|FMB| WWH|WWH| RLH | RL LMS LMS LM8s LMS | LMS Major oxides 45.7| 48.07 | 48.43] 48.7| 48.97 | 50.1 | 50.52 .4| 52.23] 62.43 | 62.58] 53.1 | 54.35] 54.41) 54.80 | 57.40 | 63.5] 65.6 12.9] 15.61] 18.94) 14.9] 16.27 | 23.0 | 19.71 ‘4 } 17,04] 15.27 | 14.97] 184 | 16.38] 15.77] 14:90] 15.49 | 161] 141 3.6| 234] 415) 47] 283 | 26 | 1.67 3{ 3.22] 5.06 | 3.86] 4:5 3.01] 4.27] 2.88] 1:97 14] 20 56] 6.98| 6.30] 34] 624 | 54 | 6.82 3] 6.60] 699 | 881] 32 8.00] 6.55] 9.00 | 7.86 50] 38 182| 1268] 6.55] 4:2] 962 | 56 | 4.64 5] 6.54) 4.13 | 4:00] 3.4 8.34] 4.02] 3.00 | 2:94 12] (149 9.6 10.76] 11.04 7.8] 12.86 | 116 | 11.96 4] 10.39) $70 | Bet] 93 8.08) 8.20] 7.15] 6.74 a8] 28 17} 1.86] 261) 21] 200 | 24 Jram :4| 297/350 | 3.47] 28 | 13.66) 3.50) 3.83] 3.96 46] 43 T52] 148 | 152] 192] lea | 44 | 1258 21 “73] 1.24 | 1.22] 14 | 11a 126) 1.23] 1.53 2.6] 3.0 fog | $12 $09 1108 112} 120 [20] 114] 1208} 09] 4218 | 1.20 3-164 [07} l1ilte10.1 ff 04 [a3 ft 01 3 |) 1206) 1 07] BA] FO] 502) #09 | 1-09 fr +08) 28.4 wy 173 [i 20o) 2} Tae |e | oT £93) 71024) 148 180] 0 1,25) 1.66| 2.48 | 1.54 -81] 64 So) 112] 27} 122} 218 | an | tlag | 1.28) tata] 125] .36 T38} 130] 4.83] 125] “265 | 1.32 523) 116 fio) a7] a7} Sao) a8 | 113 | alas | tte) S23} ea9} 617 tar] 14} 1215} 121) 125 | 1218 122) 118 Total.....| 98.62| 99.97 | 100.31| 98.44] 100.49 |100.12 |100.04 | 100.46] 98.66) 100.63] 100.54 | 100.21 | 99.85] 99.99] 100. 20] 100.44 100.22 | 99. 46] 100. 47 Powder sp. gr--| 3.02)... 297] 249] 430 | 280 | 286 | 296] 255) 294] 280 | 293] 278] 288] 428] 202) 283 2.58) 2.45 08% SHONVOTIOA NVMSVIV dO SNOLLVOILSAANT ne PD rarer 8 fs_....... yf oeB, pres No, PASE RONBNARS » WOANN! wow! © goren! BBx = ry = erpel BBx! PONYD 100.2 100.1 100.3 Niggli values 98 124 122 175 134 148 155 19 33. 28 29 23 26 25 54 2944 34 24 42 38 41 23 31. 31 31 24 24 22 4 6% 7 7 oo 12 12 il 19 +19) 37. 45 37 +35) 31 18 2 —6 7 —10 0 7 See footnotes on page 284, [8% SANVISI NVILOGIV ‘SANVISI 4O1SOD0d CNV MVNO JO ADOTORD TaBLe 2.—Analyses of rocks from Umnak and Bogoslof Islands—Continued [Analyst indicated by initials: S. M. Berthold My Ww. W. Brantooe OB) R. L. Hill (RLH), and Leonard M. Shapiro (LMS), of rapid-analysis laboratory (see Shapiro and Brannock, 1952), U. 8. Geological Survey; W. W. Hommel (WWH), of Ledoux and ©o.; M. K. Carron (MKC) and E. J. Tomasi (EJT) of U. S. Geological Survey; F. M. Byers, Jr., (FMB), of U. S. Geological Survey, on leave at University of Chicago laboratory; and Paul R. Barnett (PRB), Arthur Chodos (AC), K. J. Murata (KJM), A. T. Myers (ATM), spectrographers, U. S. Geological Survey.) Northeastern Southwestern Umnak Bogoslof Umnak—Con. 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 wws | wWB|WWB|WWB|WwWB| SMB | SMB | SMB | SMB | WWB| SMB | WWB/ SMB | WWB | WWB FMB | MKC| LMS LMS | LMS | LMS | LMS | RLH | RLH | RLH | RLH | LMS | RLH | LMS | RLH LMS LMS LMs | LMs | LMS | LMS LMS LMS Major oxides—Continued v4.3 | 56.0] 57.4 | 57.5 | 60.6 61.4] 644] 65.6] 65.6 | 682] 75.4 | 76.0 46.0 61.0 18.0 | 156/182 | 17.4 | 154 16.5] 160] 141] 156 | 114/180 | 128 18.5 18.6 L7 21] 20 | 22 | 27 26, 21] 45] 21 23|°.5 i) 6.1 3.9 7.8 48| 50 | 52 | 56 64] 41] 22] 34 60] '8 | 13 5.0 1 38 gs] 40 | 38 | 20 22] 1a) 30) .m| 88] 22] .09 5.4 i 7.9 7271 77 | 69 | 35 56| 42| 28] 34 294] 1.2 ‘a | 119 65 34 24] 29 | 386 | 54 43] 45] 61) 42 35] 35 | 35 26 4.2 10 ne) La | a7 | 22, Le] 26] 15) 28) +61] 39 | 4.7 16 2.9 20} 2.03} 2.20 | 2.24 { 231 s—29] s—.ai) 21.4 ]f 131 } 231/222 | 2.48 | 3—.29 2.08 207 107 12 66, .90 | 12 | 10 Li 82] .75] 178 nl |.18 |. .18 14 58 132 hy] fie | 14 148 718 23] 20) 127 :08} 106 | 102 136 128 118 f32] 110 | :10 | 215 “a7; 6) a} } aad 24} 102 | 102 a7 116 99.60 | 99.98} 99.96 [100.24 |100. 43 101.05} 100.31) 101.26| 100.17 | 100. 79/100. 96 |100. 0 99. 63 100. 45 2.79 | 2.88} 2.81 | 284 | 2.76 267, 259] 2.71) 248| 2.71] 246 | 261 2.86 2.60 Minor elements '—Continued -2> -BR° G8S SGONVOIOA NVMSVIV JO SNOILLVDILSHANI GEOLOGY OF UMNAK AND BUGOSLOF ISLANDS, ALEUTIAN ISLANDS 283 “¥8% OBed UO 8230300) deg emo ands ; ae “8&8 3s °° ° s “4 S 1 These constituents determined by W. W. Hommel (WWH), Ledoux and Co. 2Tgnition. Includes gain due to oxidation of FeO, i “cute due to oxidation of FeO greater than losses of volatile constituents; not included in totals. 4 Inferred from bulk specific gravity. 6 Also looked for but not found within the limits of spectrographic determination for the particular element: Ag (.0001), Cd (.005), Zn (.01), Sn (.002), As (.005), Sb (.01), Bi (.001), Ge (.00i), In (.001), TI (.005), and Re (.005). 1, Olivine sigh basalt flow (45-95) of Ashishik basalt at Cape Idak (lat 53°31.4’ N., jong .8’ E.). 1A, Olivinestich basalt flow (46 ABy 531) of Ashishik basalt 0.1 mile south of Cape lak. 2. Plagioclase-olivine basalt flow (6 ABy 92) of Mount Tulik. Lat 53°22.4’ N., long 167°04.3’ E., 3,000 feet S. 80° W. of summit. 3. Palagonitized tuff cone (46 ABy 132) of early postcaldera p; lastic rocks within Okmok caldera. Lat 53°26.4’ N., long 168°05.0’ E., 1 mile northeast of cone D. Specimen collected from palagonitized bed near summit. 4, Anorthite-augite-olivine basalt flow (46 AFr 28) of Ashishik basalt. Lat 53°25.7/ N., long 167°55.5’ E., 3 miles N. 25° W. of Umnak Airfield. Specimen collected from lower middle of flow. 5. Anorthite basalt flow (46 ABy 609) near Inanudak Bay. Specimen collected at lat 53°17.7’ N., long 168°21.3" E., from sea cliff 8. 25° E. of cinder cone. 6. Anorthite basalt flow (46 AHp 68) extruded in December 1945 from cone A. Lat 53°24.7' N., long 168°11.0’ E., 6,000 feet northwest of cone. 7. Bytownite-augite-olivine basalt flow (46 A By 29), a subaerial flow of cone D. Speci- a calected at lat 53°25.6’ N., long 168°05.8’ E., on south-southeast slope, alti- ude 1,900 feet. 8. Palagonitized scoria bed (45-73) from Ashishik basalt. Lat 53°21.2’ N., long 168° 9. Augite basalt flow (46 ABy 70) of Ashishik basalt in lower part of Okmok caldera wall. Lat 53°27.3’ N., long 168°05.5’ E. 10. Flow no. 2 (46 ABy 45) of Crater Creek basalt. Lat 53°28.5’ N., long 168°05.0’ E. in northwest wall of Crater Creek gorge. 11. Flow no. 11 (46 ABy 54) of Crater Creek basalt. Lat 53°28.7’ N., long 168°04.6’ E. in northwest wall of Crater Creek gorge. 12. Hydrothermally altered plagioclase basalt (47 ABy 44) from volcanic rocks of central Umnak. From cliff (lat 53°17.1’ N., long 168°19.8’ E.) just south of Air Force hut near Inanudak Bay. 13. Lower welded andesitic agglomerate bed (46 ABy 13) in Okmok volcanics exposed in upper part of north wall (lat 53°28.3’ N., long 168°08’ E.) of Okmok caldera. Sample taken to exclude most xenolithic fragments. 14, Older basaltic andesite flow (46 ABy 6) from cone Bin Okmok caldera. Near front of flow at lat 53°27.6’ N., long 168°07’ E. 15. Aphyric andesite pipe (46 ABY 63) of vitreous andesite. Near top of pipe (lat 53°28,0’ N., long 168°05.4’ E.) at intersection of southeast wall of Crater Creek gorge and Okmok caldera. 16. Upper welded andesitic agglomerate bed (46 ABy 15) in Okmok volcanics exposed in upper part of north wall (lat 53°28.3’ N., long 168°08’ E.) of Okmok caldera. Sample taken to exclude most xenolithic fragments. 17. Vitreous latite fragment (46 AFr 2) in Okmok volcanics. From bank of Crater Creek (lat 53°32.6’ N., long 167°59.2’ E.) near Cape Tanak. 18. Rhyodacite pumice lapilli bed (47 ABy 26) near base of Okmok pyar Sea cliff (lat 53°33.5’ N., long 168°04.9’ E.) 0.6 mile S. 10° E. of Ashishik Point. 19. Rhyolite obsidian band (46 ABy 30a) in rhyolite from northeastern Umnak (lat 53°29.5' N., long 168°10.3’ E.). 20. Felsitic rhyolite facies (48 Um 11) in east end of rhyolite from northeastern Umnak (lat 53°29.4 N., long 168°09 E.). 21. Basaltie andesite flow (47 ABy 73) of Mount Vsevidof from Twin Lava Point (lat 53°11.1’ N., long 168°47.7’ E.) 22. Quartz-bearing olivine andesite flow (47 ABy 10). North-facing cliff (lat 53°13.6’ N., long 168°21.9’ E.), south of Hot Springs Cove. Hypersthene-bearing labradorite andesite flow (47 ABy 9) of Mount Recheschnoi. t 63°13.8’ N., long 168°22.0’ E., south of Hot Springs Cove. Quartz diorite (47 ABy 41) facies of plutonic rocks (lat 53°16.1’ N., long 168°34.9’ E.), 1% miles 8. 70° E. of Cape Imalianuk. Albitized diorite (47 ABy 34) in albitized sedimentary and igneous complex. peste’ polpcted from shoreline (lat 52°51.8’ N., long 169°00.1’ E.) south of yw Hill. Andesitic glass clot (47 ABy 72) in dark scoria bed of Mount Vsevidof cone. Speci- men collected at an altitude of 4,000 feet on north slope of cone (lat 53°07.7’ N., long 168°41.3’ E.). 27. Hypersthene-bearing andesite flow (47 ABy 57) of Mount Recheschnoi, South slope at lat 53°05.3’ N., long 168°28.5 E. 28. Historic(?) latite flow (47 ABy 36) of Mount Vsevidof. Front of flow (lat 53°07.7’ N., long 168°46.6’ E.) on west slope of Mount Vsevidof. 29, Quartz keratophyre (47 ABy 29) in albitized sedimentary and igneous complex. From headland (lat 53°02.4’ N., ee E.) east of Amos Bay. 30. Rhyodacite pumice bed (47 ABy 67) that forms uppermost bed of Mount Vsevidof cone. Altitude, 1,700 feet on north slope of cone (lat 53°09.7’ N., long 168°41.7’ E.). 31. Argillite bed (47 ABy 54) in albitized ee and igneous complex. From bedded sequence (lat 53°07.7’ N., long 168°22.4’ E.), 1 mile southwest of Russian Bay. 32. Rhyolite dome (47 ABy 28). From middle dome (lat 53°10.2’ N., long 168°25.1’ E.) on west side of valley that drains into Russian Bay. 33. Granophyre dike (47 ABy 39A) that cuts plutonic rocks. From sea cliff (lat 53°16.2’ N., long 168°36.9’ E.) 34 mile southwest of Cape Ilmalianuk. 34, Bytownite-salite-hornblende basalt dome (47 ABy 100) extruded in 1927. Specimen collected at lat 53°56.1’ N., long 168°02.5’ E. 35. Andesine-hornblende andesite dome (47 ABy 103) (Old Bogoslof) extruded ia 1796. Specimen collected in saddle of Castle Rock (lat 53°55.8’ N., long 163°02.3’ E,) at an altitude of 100 feet. RE 8 8 F8% SHONVOIOA NVMSVIV JO SNOLLVOILSAANI ‘GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 285 GEOLOGY OF UMNAK GENERAL FEATURES The oldest rocks of the island, exposed only in the southwest part (pls. 39 and 40), include an albitized sedimentary and igneous complex intruded by plutonic rocks. These rocks have been deformed, and deeply eroded. The age of the sedimentary rocks of the complex is middle Tertiary or older; the plutonic rocks are post-Oligocene. Upon a basement of these rocks, volcanic edifices have been built around four main centers of activity. The oldest of the centers is in the isthmus region of Umnak; the rocks are called the volcanic rocks of central Umnak. The initial forms of the volcanoes have been completely de- stroyed by erosion. The volcanic rocks of central Umnak probably include lavas that range in age from middle Tertiary to early Quater- nary. The hypersthene andesite of Mount Recheschnoi, of probable Quaternary age, overlies the volcanic rocks of central Umnak on the southwest. The largely Recent symmetrical cone of Mount Vsevidof, consisting of basaltic andesite, andesite, and rhyodacite, buries the hypersthene andesite on the western slope of Mount Recheschnoi. Mount Vsevidof has been active in historic time. Okmok Volcano (pls. 39 and 41) lies northeast of the isthmus, and its rocks make up almost all of northeastern Umnak. Most lavas of Okmok Volcano are porphyritic to aphyric feldspathic basalts. The early activity of Okmok Volcano probably was simultaneous in part with that at Mount Recheschnoi but has persisted to the present. ALBITIZED SEDIMENTARY AND IGNEOT ~LEX DISTRIBUTION AND THICKNESS The oldest rocks of Umnak Island ‘include bedded argillite and tuff, keratophyre flows, albitized diabase sills, and irregular intrusive bodies exposed near sea level in the southwest tip and along the south-central coast of the island (pl. 40). In contrast to the younger rocks of the island, these rocks are deformed and characteristically show the effects of low grade metamorphic alteration. West and southwest of Kigul Island, the lake-studded rolling plain of southernmost Umnak is underlain by gently dipping flows of keratophyre, subordinate beds of argillite and tuff, and albitized gabbro or diorite. Sections of bedded argillite and tuff are exposed along the east shore of Nikolski Bay, in the north and east shore of Driftwood Bay, and in a small coastal out- crop 1% miles northeast of Lookout Point. Albitized intrusive rock is exposed in two small headlands at 2 and 3 miles southwest of Russian Bay, and bedded argillite and tuff with minor sills and dikes of albitized diabase crop out along the coast from the west shore of Russian Bay to the north shore of Partov Cove. 286 INVESTIGATIONS OF ALASKAN VOLCANOES The total thickness of the complex is unknown. About 5,000 feet of northeasterly dipping argillite, tuff, and albitized diabase sills are exposed along the shore northeast of Partov Cove, but parts of this sequence may be repeated by faulting. STRUCTURE The little that is known about the structure of the albitized sedi- mentary and igneous complex is included here with the stratigraphic descriptions, mainly because the structural features described below are confined to these oldest rocks and in part define them. The general strike of the sedimentary rocks is northwest in the largest area of their exposure near Russian Bay and Partov Cove (pl. 40), and dips range from 13° NE to vertical. Near Lookout Point, the sedimentary rocks dip 15° SE. Dips at the head of Driftwood Bay suggest an anticline plunging southeast. The sedimentary rocks north of Nikolski village, and the extensive keratophyre flows near Amos Bay, are nearly horizontal. No major faults were mapped in the field, but a few inferred faults in areas underlain by the albitized sedimentary and igneous complex have been mapped from well-defined alinements of swales, gullies, and lakes seen on the air photographs. The northwesterly strike of the sedimentary rocks suggests that axes of folds in these rocks cut across the northeasterly trend of the Aleutian arc, whereas the lineaments mapped from air photographs and the general elongation of the island parallel the trend of the arc. RELATION TO YOUNGER ROCKS In the few good exposures, the contact between albitized sedimentary and igneous complex and the other extrusive and sedimentary rocks of the island is one of angular erosional unconformity. That this relationship is general is indicated by the fact that major deformation is confined to the older rocks, and that the replacement of primary intermediate plagioclase by albite is confined to these rocks. PETROGRAPHY ARGILLITE The argillite varies from massive dark-gray bands 1 to 3 inches thick, to laminated argillite made up of randomly alternating thin beds of light-gray siltstone and laminae of medium-dark-gray argillite as much as one-eighth of an inch thick. The massive bands of argillite have small joints that form a crude rectangular pattern, parallel and perpendicular to the major fold axes. Thin calcite veinlets as much as one-sixteenth of an inch thick are found along many of the joints. The laminated argillite has a shaly parting, along which many dark carbonaceous patches suggestive of leaf fragments are found. One GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 287 specimen of argillite collected southwest of Russian Bay (pl. 40) contains a few small angular detrital fragments of plagioclase crystals_ and rare fragments of quartz in a fine-grained groundmass of cherty and feldspathic material whose index of refraction is near that of albite. Patches and veinlets of calcite are common. The chemical analysis of this specimen is given in column 31, table 2. TUFF Greenish water-laid tuff interbedded with the argillite ranges from moderate yellow green to slightly different shades of greenish gray. The grain size ranges from fine to coarse with a few grains as much as 1 millimeter. Three specimens from near Partov Cove are highly altered crystal tuffs in which fragments of the former pyroxene and plagioclase have been largely replaced by chlorite and albite respec- tively. A specimen from Driftwood Bay (pl. 40), however, contains fragments of unaltered pyroxene and labradorite, as well as fragments of quartz crystals. Calcite is common in scattered patches in all specimens. The groundmass is an aggregate of chlorite, calcite, albite, an unidentified zeolite mineral, and other unidentifiable light- colored, cloudy material. Pyrite is rare in the specimens from near Partov Cove, but absent from the Driftwood Bay specimen. ALBITIZED INTRUSIVE ROCK Three varieties of albitized intrusive rock include albitized diorite or gabbro, an albitized porphyritic rock, and albitized diabase. The albitized diorite is exposed near the southwestern tip of Umnak, chiefly at Elbow Hill (pl. 40). The albitized porphyritic rock is from a thick sill or flow near Amos Bay, and the albitized diabase occurs as sills, 2 to 10 feet thick, in the sedimentary rocks at Russian Bay and Partov Cove. Each is altered to a different degree from a rock probably of original gabbroic or dioritic composition. The albitized diorite near the southwest end of Umnak is the least altered of the three rocks. The plagioclase is mostly unzoned albite, but includes a few irregular-shaped relics of andesine (pl. 42A). Apatite needles are in both the andesine and the albite. The albite is clouded with aggregates of fibrous rosettes of sericite. A few small (less than 0.5 mm), clear crystals of plagioclase are zoned normally from a center of andesine-oligoclase to a rim of albite. Some of these are veined with albite, potassium feldspar, and a zeolite mineral. A few larger areas intersertal to the plagioclase contain cloudy potassium feldspar intergrown with quartz. The mafic minerals are largely intersertal to the feldspars. Anhedral patches of augite in places are intergrown with or replaced by green amphibole and greenish biotite dusted with opaque oxides. Chemical and petro- 288 INVESTIGATIONS OF ALASKAN VOLCANOES graphic analyses of this rock are given in column 25, tables 2 (p. 282 to 284) and 3 (in pocket), respectively. The albitized porphyritic rock near Amos Bay contains albite that has entirely replaced the former plagioclase of both phenocrysts and groundmass. Patches of quartz and potassium feldspar are similar to the Elbow Hill rock. Many subhedral and anhedral pyroxene grains are unaltered, but some augite is intergrown with, or replaced by, greenish biotite and high-iron chlorite. Small albite laths, shreds of green biotite and chlorite, and anhedral quartz and potassium feldspar, all dusted with opaque oxides, make up the groundmass. Chemical and petrographic analyses of the rock are listed in column 29, tables 2 and 3, respectively. The diabasic sills at Partov Cove show evidence of both albitization and potassium feldspathization, as might be expected from their proximity to areas of intensely potassium feldspathized rock near the narrow part of Umnak. ‘The feldspar in the most altered specimen is all clouded albite, in places cut by veinlets of chlorite and calcite. No pyroxene is present, but the space interstitial to the plagioclase is entirely occupied by calcite, chlorite, opaque oxides including some hematite, rare epidote, and pyrite. A few veinlets of quartz cut through all minerals. Optical data on the chlorite (alpha, 1.612, colorless; gamma, 1.618, light green) are indicative of high iron, suggesting prochlorite or ripidolite (Winchell, 1933, pt. 2, p. 284). KERATOPHYRE The keratophyre lava flows exposed between Amos and Driftwood Bays are greenish gray, specked with dark grains of opaque oxide, and more abundant, larger white phenocrysts of albite, 1 to 3 mm long. Some contain a few clear colorless calcite crystals of the same size and shape. Two specimens described in detail are representative of these flows. A specimen of fine-grained, greenish-gray keratophyre near Black Cape (pl. 40) is specked with grains of opaque oxide and mottled with white phenocrysts of albite and sparse clear, colorless crystals of calcite. In thin section, the phenocrysts are seen to be unzoned albite, flecked with wisps of sericite and minute rods of apatite. The groundmass is largely albite crystals (0.05 mm), patches and shreds of chlorite, opaque oxide euhedra, and rare anhedral potassium feldspar. Quartz and calcite are associated in patches and veinlets. A specimen of hydrothermally altered (potassium feldspathized) keratophyre collected east of Partov Cove contains evidence of two periods of alteration. The former plagioclase phenocrysts are now albite remnants largely replaced by calcite, sericite, and quartz (pl. 42 B). The groundmass is an aggregate of chlorite, sericite, quartz, GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 289 adularia(?), a low index zeolite and an opaque white mineral. The later type of alteration that has occurred is completely different from the regional albitization and is typical of later, locally intense hydro- thermal alteration of the volcanic rocks of central Umnak in the Inanudak Bay area (see p. 295 to 296). AGE AND CORRELATION Carbonized impressions suggestive of leaf fragments have been seen on bedding planes of the argillite. Botanist Eric Hultén (1937, p. 22) also reports fragments of coal in what he calls slate, presumably the bedded argillite of this investigation. Carbonaceous, coaly ar- gillite or mudstone deposits that have yielded plant fossils elsewhere in the Aleutian Islands and the Alaska Peninsula, have generally, though not always, proved to be early Tertiary. Bedded argillite and tuffaceous graywacke, similar in degree of metamorphism and deformation to the rocks of Umnak, are late Paleozoic in age on Adak Island (Coats, 1956) and early Tertiary on the Alaska Peninsula (Knappen, 1929, p. 182-198), Kodiak Island, and the Kenai Penin- sula (Capps, 1937, p. 138-153). A sequence of bedded argillite and tuff similar to that on Umnak has also been mapped by members of the Geological Survey party along the south coast of Unalaska about 30 miles to the northeast; carbonized remains were likewise found in these rocks but were unidentifiable (G. D. Fraser, oral communica- tion, 1955). Vertebrate-fossil remains, chiefly teeth, have been found in greenish volcanic graywacke exposed in a quarry on the north side of Unalaska. These teeth were identified by G. Edward Lewis (written communication, 1954) as those of Cornwallius sp., a marine mammal of early Miocene age. Both the fossiliferous graywacke and the argillite of Unalaska, like the argillite of Umnak, are intruded by plutonic rocks. The albitized sedimentary and igneous complex of Umnak is therefore probably early to middle Tertiary, but may possibly include older rocks. PLUTONIC ROCKS GEOLOGIC RELATIONS Coarse-grained dioritic rock is found in five small, widely separated areas around the margins of Mount Recheschnoi and Mount Vsevidof (pl. 40). The lava flows of these two volcanoes and the unaltered flows of the volcanic rocks of central Umnak rest on eroded surfaces in the dioritic rock. It intrudes the bedded argillite and tuff sequence and contains blocky xenoliths of these sediments in the area on the southeast slope of Mount Recheschnoi. Most of the rock is quartz diorite, but the composition ranges from diorite to quartz monzonite. Dikes and stringers of granophyre cut altered plutonic rock on the 290 INVESTIGATIONS OF ALASKAN VOLCANOES west side of Cape IImalianuk. The coarse, even grain of the quartz diorite and its similarity to that of batholithic quartz diorite on Unalaska indicate a rather large plutonic body. The distribution of its few outcrops suggests that it may underlie much of the area of the two younger volcanoes. The plutonic rocks on Umnak and Unalaska are so similar petro- graphically and so closely parallel in range of composition and in geologic relations that they must certainly belong to the same intru- sive episode. The plutonic rocks on Unalaska are intrusive into beds that, as noted above, are now known to be of early Miocene age. Thus the plutonic rocks can be no older than early Miocene. An upper age limit may be inferred from the fact that large areas of the plutonic rocks have been laid bare by erosion and subsequently covered by unaltered lava flows of central Umnak, which are prob- ably early Pleistocene but may be as old as Pliocene. The erosion of the plutonic rock, prior to extrusion of the Pliocene and Pleistocene lavas, must have taken part of the Pliocene epoch so that the youngest possible age of the plutonic rocks would be mid-Pliocene. On the basis of present evidence, the plutonic rocks were most likely intruded sometime during middle to late Tertiary time. PETROGRAPHY DIORITE A specimen of gray diorite from the small knob 4% miles south of Mount Vsevidof has an average grain size of about 1 mm. A few veinlets of light-gray granophyre cut the diorite, which is slightly altered adjacent to the veinlets. As seen in thin section the texture of the diorite (column 1, table 4) is hypidiomorphic granular except for a little interstitial micrographic intergrowth of quartz and ortho- clase. A few plagioclase grains are veined with orthoclase and to less extent with biotite. Orthoclase appears as rare irregular wisps in the plagioclase. Dusty fragments of biotite and sericite are common, and radiating sheaflike aggregates with the birefringence and relief of prehnite are rare as inclusions in the feldspar. Augite remnants are enclosed by the hornblende and biotite, and hornblende remnants in turn are enclosed and apparently replaced by biotite. Apatite is found as sparse tiny needlelike laths, as larger euhedra up to 0.3 mm long, and as anhedral grains 0.2 mm in diameter. QUARTZ DIORITE Fresh, granular quartz diorite is the most common plutonic rock type on southwestern Umnak. It is similar to the quartz diorite that forms the large intrusive body on nearby Unalaska Island (Becker, 1898, p. 42). The quartz diorite of southwestern Umnak has an GEOLOGICAL SURVEY BULLETIN 1028 PLATE 42 A, Photomicrograph of composite plagioclase cryst low birefringent (darker in picture original andesine of rock. Crossed nicols. in albitized diorite (column 2 are andesine remnants surround le 2). Discontinuous, y albite that replaced B. Photomicrograph of partly repla keratophyre. Quartz (g) and nicols. henocryst in hydrothermally altered (potassium feldspathized) calcite-service (light aggregates) have replaced earlier albite. Crossed GEOLOGICAL SURVEY BULLETIN 1028 PLATE 43 led by orthoclase in quartz diorite (specimen yre dikes. Crossed ni eined and ordering gra no. 47ABy GEOLOGICAL SURVEY BULLETIN 1028 PLATE 44 Mount Vsevidof from the south Steep-sided pyroclastic cone flows forming moderate slope at the base. Drained~ lava in center middle background apparently formed when lava flowed over a now-buried, eroded cliff older rocks. Glacial-outwash plain in foreground. GEOLOGICAL SURVEY BULLETIN 1028 PLATE 45 ORR View of east wall of Okmok Caldera, showing precaldera rocks, largely palagonitized rocks of Ashishik basalt, overlain by bedded Okmok volcanics. PLATE 46 GEOLOGICAL SURVEY BULLE’ |. Nearly flat-lying gorge. Uppe Arcuate ridg undisturbed Crater Creek basalt flows exposed in the northwest wall of Crater Creek st thin resistant bed is lower welded agglomerate bed near base of Okmok volcanics. orth of caldera is partly shrouded in cloud bank. B. View northeast toward sea cliff northeast of Cape Chagak. About 250 feet of Okmok volcanics rests on precaldera feldspathic basalt flows of Ashishik basalt (vertical sea cliff and foreground). Note welded palagonitized agglomerate beds in lower third of exposure. Middle of Okmok voleanics is eroded into grotesque forms. GEOLOGICAL SURVEY BULLETIN 1028 PLATE 47 A. Hand specimen of and plagioclase basalt. ay welded andesitic agglomerate with fragments of medium-gray aphy ted from headwaters of New Jersey Creek (lat. 53°29’, long. 1€8°O B. Photomicrograph of small mass of basaltic andesite enclosed by rhyolite (column 19, table 2). Olivine (ol), labradorite (J), andesitic groundmass (gr), contains microlites of p' clinopyroxene and opaque oxides in glass. Note flow banding in rhyolite (r). Plane-polarized 1 oclase. ht. GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 29] TaxixE 4.—Petrographic analyses of plutonic rocks [Modal analyses in volume percent) 2 3 4 Mh. foo ees| 33 4.0 10.2 15.8 23.6 9.7 5.9 28.1 43,7 64.5 64.9 44.5 24.7 Biotite (includes sparse chlorite). 7.4 8.4 14.7 +3 Hornblend 44 3.2 5.6 0 Augite._ 4.1 5.4 +4 12.6 Ey, +0 +2 -0 +0 Apatite. 8 3 al O01 Mi a 50 “ye ag eo icropegmatite_. . a Range in compositio1 Ansso| Anes | Aneus Ano Bulk specific gravity of rock 2.7 2. 82 2.72 2.58 1 Includes considerable chlorite, some epidote, and Eee ets ee largely micrographic intergrowth of quartz Potassium feldspar; minerals listed separately above. 3 Epidote. 4 Myrmekite, 1. Diorite from small knob surrounded by recent flow, 414 miles south of Mount Vsevidof, 2. Quartz diorite from sea cliff, 1}4 miles southeast of Cape Ilmalianuk. 3. Quartz monzonite from sea cliff, Cape Ilmalianuk. 4, Granophyre, from sea cliff, three-fourths of a mile, southwest of Cape Iimalianuk. average grain size of about 2 mm and consists of dominant, medium- dark-gray plagioclase as much as 8 mm long, minor light-gray to light-pinkish-gray orthoclase, inconspicuous quartz, and mafic min- erals, of which biotite is most abundant and easily recognizable (column 2, table 4). Quartz and-orthoclase occur interstitially to the subhedral to euhedral plagioclase. Myrmekite occurs on lobate ends of several large plagioclase crystals and also on irregular borders along the 010 face, where islands of myrmekitic plagioclase sur- rounded by orthoclase extinguish in sequence with the nearby main- land plagioclase. The plagioclase of the myrmekite is calcic oligoclase. Augite and rare hypersthene occur as corroded remnants enclosed by hornblende and biotite. Biotite occurs as individual grains ranging down to submicroscopic size, as an interstitial mineral with quartz and orthoclase, as reaction rims around pyroxene and hornblende, and as veinlets cutting pyroxene. Opaque oxides are chiefly associated witb biotite and in part appear to replace it. A minor amount of chlorite has replaced biotite, and sericite(?) is rare as tiny wisps in the feldspar. Apatite is scattered as rod-shaped inclusions in the other minerals. The chemical analysis of this specimen is given in table 2, column 24, QUARTZ MONZONITE The quartz monzonite is a facies of the quartz diorite and was observed only on Cape IImalianuk. In hand specimen the rock differs from the quartz diorite in the greater abundance of pink orthoclase and light-gray oligoclase. The hornblende and biotite are stained 489292—59——_-3 292 INVESTIGATIONS OF ALASKAN VOLCANOES green with chlorite and epidote. The hand specimen appears distinctly granitoid, although quartz is not conspicuous. The quartz monzonite (column 3, table 4) as seen in thin section resembles the quartz diorite except for differing proportionsof the minerals. Quartz and orthoclase appear interstitial, but show only a very slight tendency toward graphic intergrowth. Some oligoclase crystals are veined and contain irregular wisps of orthoclase as well as specks of calcite(?) and tiny flakes of sericite(?). A little of the orthoclase is perthitic. Augite remnants are sparse and contain reaction rims of hornblende and biotite. About half the biotite has altered to chlorite. Anhedral opaque oxides are associated with hornblende and biotite and appear to replace them. Minor epidote and rare prehnite(?) also appear as alteration products of the biotite. GRANOPHYRE Pinkish-gray granophyre (column 4, table 4), as dikes and narrow stringers, cuts the quartz diorite on the west side of Cape Timalianuk headland. The texture of the rock is largely micrographic superposed on an allotriomorphic texture. Each larger granular unit contains uniformly oriented quartz grains that differ in orientation from grains in a neighboring unit. A few larger grains of quartz, orthoclase and albite-oligoclase are not micrographically intergrown. Epidote is in aggregates interstitial to the other minerals. The biotite is largely altered to chlorite. Rare needlelike inclusions of apatite are in the quartz and feldspar. Actinolite also occurs as sparse aggregates. The chemical analysis of this specimen is listed in column 33, table 2. ALTERED CONTACT ROCK An altered monzonitic rock borders the granophyre dikes and ex- tends for some distance on either side. The altered rock is invaded by smaller stringers of granophyre at closely spaced intervals. The con- tact with the granophyre is gradational, and the alteration within the monzonitic contact rock is variable. Streaks of pinkish-gray rock with spots of mafic minerals alternate with streaks of mottled light- greenish-gray epidotized feldspar and pinkish-gray orthoclase. The streakiness in general parallels the approximate border of the gran- ophyre dike. Dark-greenish actinolitic hornblende as much as 3 mm in length shows slight segregation in bands parallel to the feldspathic streaks but shows no visible lineation. As seen in thin section, the rock has a variable texture. An allotriomorphic texture involves chiefly quartz and orthoclase with minor albite-oligoclase and grades into a micrographic intergrowth of quartz and potassium feldspar and also into poikilitic intergrowths of small quartz grains in potassium feldspar. A few formerly euhedral andesine and calcic oligoclase crys- GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 293 tals have relict outlines and are veined and corroded by potassium feldspar and albite-oligoclase (pl. 43). Much of the plagioclase con- tains fine to coarse aggregates of epidote, probably derived from the decomposition of the more calcic plagioclase. VOLCANIC ROCKS OF CENTRAL UMNAE GEOLOGIC RELATIONS Lava flows, vent breccias and associated irregular shallow intrusive bodies crop out over a large area in the narrow central part of Umnak Island. These rocks are herein grouped together as the volcanic rocks of central Umnak. The original constructional surface form of these volcanic rocks have been destroyed by erosion. Their original areal extent is not known because they are covered both to the north- east and to the southwest by Quaternary lavas of Okmok Volcano and Mount Recheschnoi, respectively. The volcanic rocks of central Umnak were in part erupted from several vents northeast of Hot Springs Cove (pl. 41). These rocks include both hydrothermally altered and fresh unaltered volcanic rocks in each of the two map units: bedded volcanic rocks (QTv) and vent and intrusive complex (QT vp), as shown on plates 40 and 41. No contacts between altered and fresh unaltered volcanic rocks within the units were seen. ‘The intensely silicified, potassium feldspathized rocks included as map unit QTva, (pls. 40 and 41), con- sists of rocks so completely replaced by silica, potassium feldspar, and in places tourmaline, that the original rock type cannot be identified in hand specimen or thin section, though field relations indicate that most grades into rock included in the map unit, QTv. Altered lava flows and vent complexes make up most of the isthmus terrain east and north of Hot Springs Cove (pl. 41). Breccias and irregular intrusive rocks, interpreted as old vent complexes, are found at several places in the isthmus. Volcanic necks of hydrothermally altered basaltic porphyry at Thumb Point (pl. 40), and High Hill, 2% miles north of Nikolski, are included here. Areas of most intense alteration, mapped under the symbol, QTvq, are along the coastline just west of Hot Springs Cove and in the southwest wall of Bottleneck Pass (pl. 41). The base of the altered volcanic rocks has not been observed. Unaltered hypersthene andesite flows considered part of this unit are found in several localities south and west of Hot Springs Cove (pl. 40). South of Hot Springs Cove these unaltered flows dip about 5° SW. under the similarly appearing hypersthene andesite flows from Mount Recheschnoi, which dip northeast and lie on an erosion surface cut on the volcanic rocks of central Umnak. Similar relations between unaltered lava flows of central Umnak and those of Recheschnoi are 294 INVESTIGATIONS OF ALASKAN VOLCANOES found at Broken Point, Geyser Bight, Partov Cove, Russian Bay, and in the west wall of the canyon 4 miles due south of the summit of Mount Recheschnoi. Along the Pacific coast at Partov Cove and Russian Bay, unaltered lava flows of this unit lie on an erosion surface cut on the albitized sedimentary rocks; on the Bering Sea coast at Broken Point and the west shore of Geyser Bight they lie on a plane erosion surface, possibly a raised marine bench, cut on the plutonic rocks; and west of Hot Springs Cove and along the south shore of Inanudak Bay they overlie hydrothermally altered lava flows and diorite in an irregular contact that was concealed or inaccessible for study in the field. The nature of the contact between the altered and unaltered lavas of the volcanic rocks of central Umnak is unknown. Near the head of the easternmost fork of Geyser Creek (pl. 40), a patch of intensely altered rock, about 0.1 square mile in area, is just upstream from an area of borate-bearing hot springs, geysers, and fumaroles (Byers and Brannock, 1949, p. 728). This altered rock consists of a very fine- grained aggregate of quartz, chlorite, pyrite, and colorless minerals, possibly including adularia. The altered rock may here be related to hot-spring activity in the recent geologic past. Elsewhere, however, as near the isthmus of Umnak, larger areas of tourmalinized, silicified, potassium-feldspathized rock are not associated with hot springs and are also somewhat coarser grained than the altered rocks on Geyser Creek. The relation between the unaltered and altered volcanic rocks may be gradational, inasmuch as gradation was observed in the field and confirmed by the laboratory study of specimens from the southeast wall of Bottleneck Pass (pl. 41). On the west side of Hot Springs Cove (pl. 40), however, intensely altered rock is overlain by fresh, unaltered hypersthene andesite flows. The overlying fresh lava could hardly have escaped some alteration during the intense silicification and potassium feldspatbization of the lower unit; hence, it is believed that a sharp contact, representing a significant period of geologic time, may be concealed between the altered and unaltered lavas at this place. Possibly the alteration is locally associated with buried plutonic rocks of the same age as those that now crop out at the surface. AGE AND CORRELATION The volcanic rocks of central Umnak, as shown on plates 40 and 41, include rocks that are probably older and unaltered lavas that are younger than the plutonic rocks. The altered volcanic rocks in many places on Umnak closely resemble rocks on Unalaska known to be intruded by the post-Oligocene plutonic rocks. The relations be- tween the altered volcanic rocks and the albitized sedimentary and GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 295 igneous complex are unknown, but from the petrographic evidence, the potash feldspathization of the altered volcanic rocks occurred later than the albitization of the complex; how much later remains an open question. The altered volcanic rocks therefore may range in age from early to middle Tertiary, though a still earlier age cannot be tuled out on present evidence. The unaltered lavas rest on the post- Oligocene plutonic rocks and are overlain by the Quaternary lavas of Recheschnoi and Okmok volcanoes; hence, the unaltered lavas of the volcanic rocks of central Umnak are late Tertiary or early Quaternary in age. on PETROGRAPHY AND CHEMICAL COMPOSITION A volcanic neck at Thumb Point intrudes the albitized rock of the basement. Examination of a specimen from the basement rock adjacent to the neck indicates that it was first albitized and later potassium feldspathized (see p. 288). The rock of the neck, however, shows only potassium feldspathization. The rock is a dense, medium- dark-gray, sparsely porphyritic lava with greenish-gray spots resem- bling amygdules, but as seen in thin section former labradorite pheno- crysts are much altered to chlorite and a zeolite. Former olivine phenocrysts are replaced by antigorite(?) and magnetite. The ground- mass consists of andesine, magnetite, quartz veinlets (secondary), chlorite, and a cloudy white feldspar (index of about 1.52), probably adularia. Fine-grained potassium feldspar was identified in several thin sections of the hydrothermally altered volcanic rocks of central Umnak. Partial chemical analyses were made of three specimens, progressively more altered, from the southwest wall of Bottleneck Pass (see pl. 41). The least altered specimen (column 12, tables 2 and 3) is representa- tive of most of the hydrothermally altered lavas. Quartz, magnetite, chlorite, and serpentine in various proportions have replaced former olivine phenocrysts and portions of the groundmass glass. The chlorite is penninite. The serpentine mineral, more abundant than chlorite, is probably antigorite («, colorless to faint yellowish green; y pale green and parallel to fibers; birefringence, 0.015-0.020; indices variable, ranging from about 1.53 to 1.58). Zeolites, potassium feldspar, and calcite embay the primary plagioclase. The introduced zeolite minerals, identified on the basis of optical properties, are heulandite(?), chabazite(?), and natrolite. Amygdaloidal fillings of quartz and natrolite(?) are common. A second specimen (12-A) collected about 0.7 mile southeast from the locality of the first specimen (column 12, table 2) and from appar- ently the same lavas, was originally a porphyritic lava. Secondary amphiboles, epidote, and sparse tourmaline have developed in addi- 306 INVESTIGATIONS OF ALASKAN VOLCANOES Tasie 9.—Petrographic analyses i basaltic grontte, andesite, and latite flows of fount Vsevii Mode in volume percent: Phenocrysts>0.1 or 0.2 mm (see notes): Plagioclase. 98.7 97.6 95.3 95.3 An | Ants | Anst-t0 Ans -| Anz Ans Anso Anu(?) Range of composit: Average composition of groundmass plagioclase. - an 1, Basaltic andesite flow at Twinlava Point older than pyroclastic beds. Chemical analysis, column 21, table 2, Crystals >0.2 mm are considered phenocrysts. 2. Youngest basaltic andesite flow on north slope of Mount Vsevidof; specimen from an altitude of 1,500 feet. Crystals >0.2 mm are considered phenocrysts. 3. Youngest andesite flow on south slope of Mount’ Vsevidof; specimen from flow front on Black Creek outwash plain. Crystals >0.1 mm are considered phenocrysts. 4. Latite flow, possibly historic in age, on west slope of Mount Vsevidof; specimen collected from flow ot at altitude of 650 feet. Chemical analysis, column 28 ,table 2. Crystals >0.1 mm are considered phenocrysts, average size is about 0.1 mm, comprise the bulk of the groundmass as defined in table 9, though a few, whose average size approaches 0.2 mm, have lengths as much as 0.4 mm parallel to axis @. The tabular plagioclase crystals grade downward to tiny microlites, about 0.02 mm long, that compare in size with microlites of clinopyroxene and opaque oxide. Olivine occurs as rare microphenocrysts as much as 0.4 mm and as rare microlites grading downward to 0.05 mm. Euhedral olivine is enclosed by a thin shell of clinopyroxene and ranges in forsterite content from Foss to Foz. ANDESITE AND LATITE FLOWS The andesite flows and the historic(?) latite flow contain light-gray plagioclase phenocrysts in a dark-gray groundmass; the groundmass of the latite is somewhat more glassy and darker in hand specimen than those of the andesites. Thin section analyses are presented in columns 3 and 4, respectively, table 9 (see also column 28, table 3). A size cutoff of 0.1 mm between phenocrysts and groundmass was assumed, because microphenocrysts, ranging in size upward from 0.1 mm, are set in a hyalopilitic groundmass. The microlites in the groundmass, with the exception of orthoclase, average from 0.002 to 0.005 mm. Plagioclase phenocrysts are normally zoned through the slight range indicated in the andesite (column 3, table 9). The plagioclase of the latite has a slight tendency toward oscillatory zoning with a reverse zoning associated with glass in the inner shells. A few of the plagioclase phenocrysts of the latite contain a small core of sodic andesine. Hypersthene occurs as small laths ranging from 0.2 to 0.5 mm and has a thin shell of augite or ferroaugite. Augite occurs as microphenocrysts ranging from 0.1 to 0.3 mm in average dimension. Opaque oxide euhedra average slightly over 0.1 mm in diameter. GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 307 The groundmass consists of very fine microlites of ferroaugite(?), andesine and sparse opaque oxide enclosed by glass and orthoclase. Orthoclase occurs in minor amounts in the groundmass of the andesite but is common in the latite groundmass. Part of the orthoclase in the latite occurs in euhedra, a few of which reach 0.2 mm in length but are considered part of the groundmass, because the fine microlites are poikilitically enclosed by the larger orthoclase euhedra. Ortho- clase thus crystallized slightly later than the microlites or at about the same time as the associated interstitial liquid was cooled to a glass. The glass in the latite groundmass (column 28, table 3) has an index of refraction comparable to glass of the analyzed rhyolites from Umnak. ANDESITIC AND RHYODACITIC PYROCLASTIC BEDS The beds that form the uppermost part of the cone consist of an underlying solid, welded, grayish-orange rhyodacite pumice bed (column 30, table 3), gradational with and overlain by a streakily banded, medium to dark-gray scoria bed. This contains minor solid clots of black andesitic glass (column 26, table 3) gradational in texture with andesitic scoria. Both the pumice and the scoria con- tain 10 to 15 percent of accessory blocks of andesitic lava and dioritic intrusive rock. Petrographic analyses of the nonvesicular andesite glass, the vesicular scoria, and a xenolith in the glass are presented in table 10. The plagioclase and augite of the limonite-stained xenolith average about 0.5 mm and are mutually intergrown, as in an intrusive rock. Euhedral opaque oxide grains average slightly larger than 0.1 mm. TasLE 10.—Petrographic analyses of rocks in dark andesitic scoria bed of Mount sevidof cone Mods dn volimas pereant (not determined in column 2): r} than 0.1 mm: Orthoclase (?) Plagioclase. Augite. Hypers' Opaque oxides. Groundmass (glass: 1,2) or mesostasis (3) Hydrated iron oxides (limonite?) -___.. Range of composition of low-An plagioclase Banep of composition of high-An plagioclase - 6 index of augite.........-...--.---.-. Range of composition of hypersthene- Refractive index of groundmass glass. 1 Cloudy orthoclase(?) is the dominant constituent of the mesostasis. 2 Minor augite with 6=1.703-05, probably from altered xenoliths? 1, Black andesitic glass clot (chemical analysis, column 26, table 2) within and ational into dark ande- sitic scoria erupted after rhyodacitic pumice during last major eruption of Mount Vsevidof. Collected at 4,000 feet on N. Slope of cone (lat 53°07.7’ N, long 168°41.3’ E). 2. Dark andesitic scoria that forms nearly all of bed. Same location as 1. 3. Limonite-stained dioritic xenolith in black andesitic glass. Same location as 1. 489292—59—_4 308 INVESTIGATIONS OF ALASKAN VOLCANOES The interstitial material is largely cloudy, limonite-stained potassium feldspar with abundant apatite needles. Sodic andesine occurs in the outermost shells of all the plagioclase of the xenolith, in some grains of the andesitic scoria and glass, and as mottled, irregular grains in all three rocks. The apatite-charged orthoclase(?) and mottled sodic andesine grains in the scoria and andesitic glass suggest that they are xenocrysts from the disaggregation of diorite xenoliths. VOLCANIC ROCKS OF OKMOK VOLCANO Erupted rocks of Okmok Volcano and satellitic vents occupy most of northeastern Umnak Island (pl. 41). The main cone of the central voleano is truncated by Okmok Caldera, an approximately circular depression 6 miles in diameter and 1,500 feet in maximum depth. Mount Tulik, Mount Idak, and Jag Peak are the most prominent secondary volcanic piles superimposed on the outer slopes of the main cone. The lava plateau northeast of Mount Idak and the ridge west of New Jersey Creek are made up of bedded volcanic rock with gentle dip away from the main center of Okmok Volcano and are probably faulted parts of the outer slopes. The rocks are divided into three stratigraphic groupings: precaldera rocks erupted prior to the formation of the caldera, pyroclastic rocks associated with the caldera-forming eruption, and postcaldera rocks erupted after the caldera-forming eruption. The precaldera rocks are mostly of basaltic composition and are by far in greatest volume. They include lava flows and pyroclastic beds and range from aphyric to porphyritic in texture. The pyroclastic rocks associated with the caldera-forming eruption consist of agglomerate, welded agglomerate, ash, and tuff-breccia. Basalt cones, plug-domes, and flows occur outside Okmok Caldera and cannot be dated precisely with respect to its formation. The known postcaldera deposits within Okmok Caldera include volcanic material erupted from different vents and also a sedimentary unit containing bedded volcanic material deposited in a caldera lake, which has now almost completely vanished. Several formal stratigraphic names were proposed in the 1947 administrative report describing Okmok Volcano (Byers, Hopkins, Wier, and Fisher, 1947, p. 19-53). These names in approximate order of age were: Ashishik basalt, Crater Creek basalt, Idak basalt, Tulik basalt, Tanak volcanics, and Okmok ash. The names, Ashishik basalt and Crater Creek basalt, are used but redefined in this report; Okmok ash is changed to Okmok volcanics and redefined in this report to include both the Tanak volcanics and Okmok ash, which are here- with abandoned. The names, Idak basalt and Tulik basalt are also abandoned and replaced by informal names. GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 309 ROCKS OF PRECALDERA AGE The precaldera rocks are subdivided on plate 41 into three main categories based on geologic and geomorphic relations. These cate- gories in order of relative age are the Ashishik basalt, volcanic rocks of minor vents, and the Crater Creek basalt. The three categories may in part overlap in time, but generally their ages are believed to be as indicated on the map explanation of plate 41. ASHISHIK BASALT The oldest precaldera rock of Okmok Volcano is named the Ashishik basalt after exposures in the east-facing cliff, 2 to 3 miles south of Ashishik Point, at the northernmost part of Umnak Island (pl. 41). The formation as redefined in this report includes most of the bulk of Okmok Volcano and comprises the Ashishik basalt of the 1947 admin- istrative report (Byers, Hopkins, Wier, and Fisher, 1947, p. 25), tuff-breccia and tuff, that part of the Crater Creek basalt on the outer flanks of Okmok Volcano, Idak basalt, excepting the mass of Mount Idak, and the “basalt flows of Hill 1200” (Byers and others, 1947, pl. 3). The Ashishik basalt is bere subdivided into three lithologic units, mafic phenocryst basalt, consisting of porphyritic basalt with more than 2 percent of mafic megaphenocrysts, palagonitized pyro- clastic rocks, and aphyric (nonporphyritic) and feldspathic basalt, made up mostly of aphyric basalt flows but with some porphyritic flows that contain phenocrysts of plagioclase and less than 2 percent of mafic phenocrysts. These units are listed in general order of age, although they complexly intertongue. The Ashishik basalt must rest on the volcanic rocks of central Umnak, although the contact is concealed by the extensive blanket of Okmok volcanics. The hydrothermal alteration, the topographi- cally low position, and the great amount of dissection of the volcanic rocks of central Umnak, in contrast to the constructional surfaces yet remaining on the Ashishik basalt, leaves little room for doubt that the Ashishik basalt is a generally younger pile of lavas superposed on the volcanic rocks of central Umnak. The volcanic rocks of minor vents intrude or overlie the Ashishik basalt. MAFIC PHENOCRYST BASALT At the type locality of the Ashishik basalt in the prominent ridge west of New Jersey Creek (pl. 41), mafic phenocryst basalt flows dip gently northward. The sequence is several hundred feet thick, with the base not exposed, and is overlain by approximately conformable aphyric and feldspathic basalt flows. A thickness of 500 feet of these flows was measured on the cliff, 1% miles upstream from the mouth of New Jersey Creek. Individual flows range from 20 to 60 feet in 310 INVESTIGATIONS OF ALASKAN VOLCANOES thickness. Most are massive without flow banding, but horizontal sheeting possibly representing flow structure, is conspicuous in a few. The upper 2 to 5 feet of many flows is reddish oxidized flow-breccia. The rocks are chiefly light to dark purplish gray, and range from aphanitic to coarsely porphyritic. Phenocrysts are in order of abundance calcic plagioclase, olivine, and monoclinic pyroxene. A group of six superimposed, nearly horizontal flows of porphyritic basalt with mafic phenocrysts is exposed in the base of the northeast wall of Okmok Caldera (pl. 41). The outcrop is about 1,000 feet long and 300 feet high. The lowest outcrops are about 1,100 feet above sea level, and do not expose the base. The rock is overlain by pala- gonitized pyroclastic rock, with intervening lenses of bedded conglom- erate as much as 15 feet thick. Mafic phenocryst basalt underlies or intertongues with aphyric and feldspathic basalt flows at several scattered places on the eastern outer slope of Okmok Volcano, but the base is nowhere exposed. Mafic phenocryst basalt flows, dipping gently northeast, are the oldest rocks exposed in the east face of the lava plateau east of Mount Idak. Similar flows crop out higher in the cliff section and inter- tongue with both feldspathic basalt flows and palagonitized pyroclastic rock. In part of the ridge, mafic phenocryst basalt is the uppermost rock in the section and directly underlies the Okmok volcanics. PALAGONITIZED PYROCLASTIC ROCKS The palagonitized pyroclastic rocks of the Ashishik basalt include bedded tuff, poorly bedded tuff-breccia, which crops out beneath the Okmok volcanics on the lower slopes of Okmok Volcano (pl. 41), and unbedded vent agglomerate exposed mainly in the walls of Okmok Caldera. These rocks were included in the tuff-breccia and tuff unit of the preliminary report (Byers and others, 1947, pl. 3). The color ranges from dull yellowish brown to bright yellowish orange; the breccias and agglomerates are mottled by darker lithic fragments against the bright-colored matrix of palagonitized tuff. Xenolithic fragments range in size from microscopic grains to 3-foot blocks, and are similar to the mafic phenocryst basalts and to the aphyric and feldspathic basalts exposed on the flanks of the volcano and in the caldera walls. These rocks are most abundant near the caldera, particularly in the east wall (pl. 45), where yellowish-orange agglomerate is exposed through a vertical distance of 1,000 feet. Comparable agglomerates more than 800 feet thick appear to make up a large part of the high cliff that truncates the north arcuate ridge on the east. Many small dikes and irregular bodies of intrusive rock cut the agglomerate here. At the southwest end of this same ridge, the caldera fault scarp GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 311 exposes 1,000 feet of palagonitized pyroclastic rocks cut by several vertical lava pipes or dikes as much as 300 feet wide. Many dikes are similarly associated with the palagonitized agglomerate elsewhere in the caldera wall. Palagonitized pyroclastic rocks are exposed in many places on the flanks of Okmok Volcano. In some places, chaotic dips and associ- ated irregular dikes suggest that they occupy old vents, as near the headwaters of Bering Creek east of the caldera. Most of the out- crops are of tuff-breccia or agglomerate with no well-defined bedding, but on the southwest flank of Okmok Volcano at the base of three of the satellitic mountains, bedded palagonitized tuff is exposed as part of the main mountain mass under the edges of plagioclase-olivine basalt flows of satellitic vents. Beds of the palagonitized pyroclastic rocks are exposed in the southeast face of the plateau northeast of Mount Idak: one bed between the lower mafic phenocryst basalt flow and the overlying aphyric and feldspathic basalt flows, and the other between the aphyric and feldspathic basalt flows and the upper mafic phenocryst basalt flow that tops the plateau. APHYRIC AND FELDSPATHIC BASALT The aphyric and feldspathic basalt unit of the Ashishik basalt includes abundant flows of aphyric basalt and less common feldspathic basalt rich in phenocrysts of calcic plagioclase. Phenocrysts of mafic minerals are rare. Aphyric and feldspathic basalt flows form much of the constructional surface of the flanks of Okmok Volcano and are exposed in the caldera walls, in the shallow stream cuts, and in much of the sea cliff around the periphery of Okmok. A thick unit of these rocks forms the middle of the sequence exposed in the lava plateau northeast of Mount Idak. These rocks cap the isolated ridge west of New Jersey Creek and form the bulk of Magazine Ridge south of Mount Idak. Except for those on the plateau northeast of Mount Idak, they were included in the Crater Creek basalt in the preliminary report (Byers, and others, 1947). Individual flows of these rocks are from 10 to 70 feet thick, and tops and bottoms commonly are reddish oxidized breccia and scoria from 5 to 20 feet thick. Columnar jointing is not well developed. The aphyric rocks are generally gray to light gray and commonly contain closely spaced joints, about one-eighth of an inch apart, parallel to the top or bottom. Parallel orientation of minute plagio- clase tablets gives a satinlike sheen to the joint surfaces. The porphyritic rocks are gray and conspicuously speckled with light-gray or white plagioclase phenocrysts. Flows have an irregular jointing which yields blocks with partly conchoidal surfaces. 312 INVESTIGATIONS OF ALASKAN VOLCANOES VOLCANIC ROCKS OF MINOR VENTS The volcanic rocks of minor vents are subdivided into three units: vitreous andesite in plugs, domes, and flows; rhyolite in the form of a dome, and the plagioclase-olivine basalt of satellitic vents. These rocks either rest on the Ashishik basalt or intrude it. They were emplaced after the major upbuilding of Okmok Volcano and yet were dissected prior to deposition of the Okmok volcanics. Because of similar geologic relations with the older Asbishik basalt, the three units are assumed to be about the same age, although nowhere are they in exposed contact. The mass of Mount Tulik may be slightly younger than the other plagioclase-olivine basalts, for it is little dissected. VITREOUS ANDESITE In the caldera wall, in Crater Creek gorge, and on the outer slopes of Okmok near the caldera (pl. 41) several pipes, plug-domes and necks, and a few thick stubby flows are of dense, vitreous andesitic rock. In the field, these rocks were distinguished from the more common aphyric basalts by vitreous lustre, darker color, and almost complete absence of vesicles. The largest exposed body is a neck in the northeast wall of Okmok Caldera about 500 feet across and exposed 1,700 feet vertically. The total volume of these rocks is insignificant compared to that of the Ashishik basalt. RHYOLITE Rhyolitic rocks crop out for a mile along the north side of the north arcuate ridge and also form a ridge that extends a mile north- west from the north arcuate ridge (pl. 41). The rhyolite of the ridge is made up of randomly alternating bands ranging from a fraction of an inch to several inches in thickness and consisting of dense black obsidian, vesicular gray and black obsidian, and light-gray felsitic rhyolite. The banding strikes parallel to the ridge and dips inward from each margin, suggesting that the ridge is an elongated narrow extrusion or dome from a fracture rather than an elongated flow. The eastern part of the mass along the north base of the north arcuate ridge contains black obsidian but is made up of gray vesicular glass and light-gray felsitic rhyolite in crude bands that dip southward into the ridge. The rhyolite presumably rests on the Ashishik basalt, although the contact is covered by talus from the dome. It is over- lain by the Okmok volcanics, and a Recent basaltic flow overlies it to the east. PLAGIOCLASE-OLIVINE BASALT OF SATELLITIC VENTS Flows of plagioclase-olivine basalt make up the bulk of satellitic piles on the outer slope of Okmok Volcano. These include volcanic rocks at the satellitic vents of Mount Idak on the east slope, Mount GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 313 Tulik on the southeast slope (pl. 41), and Jag Peak and three other volcanic mountains on the southwest slope. These satellitic moun- tains rise 1,000 to 2,000 feet above the adjacent slope of Okmok Volcano. The flows have quaquaversal dips as great as 30° away from their summits. Vent breccia or intrusive plug material is ex- posed at some of the summits. Thick ash and, in places, agglomerate of the Okmok volcanics blanket the eroded flanks of the satellitic mountains. CRATER CREEK BASALT The Crater Creek basalt, the youngest precaldera unit, was origi- nally defined in the 1947 administrative report (Byers, Hopkins, Wier, and Fisher, 1947, p. 25-26) to include the uppermost aphyric basalt flows on the outer slopes of Okmok Volcano, but it is restricted herein to the sequence of basalt flows exposed in Crater Creek gorge and in the north wall of Okmok Caldera to the west (pl. 464A). A fault separates the Crater Creek basalt in the north wall from the Ashishik basalt where the north arcuate ridge intersects Okmok Caldera. The Crater Creek basalt is separated from the Ashishik basalt by an erosional surface. The unit was mapped separately because it occupies a topographically low position between the arcu- ate ridges, composed largely of the older Ashishik basalt, and because it rests unconformably on the Ashishik basalt and vitreous andesite in the lower part of Crater Creek gorge (pl. 41); otherwise, the rocks are petrographically indistinguishable from the feldspathic basalts of the Ashishik basalt. The depression in which flows of the Crater Creek basalt lie may have included part of an earlier caldera on the concave side of the arcuate ridges. The Crater Creek basalt consists of at least 15 separate flows with a total exposed thickness of 500 feet. The north wall of Okmok Caldera exposes several with easterly dip components of 1° to 2° toward Crater Creek gorge. The base is not exposed, but the gorge of Crater Creek is cut through into older palagonitized pyroclastic rocks, vitreous andesite, and associated dikes. The dikes are trun- cated by an eroded surface upon which the basal basalt flows were extruded. The Crater Creek basalt is thickest in the northwest wall of Crater Creek gorge, appreciably thinner and less continuous in the southeast wall, and does not crop out at all in the east wall of the caldera south of the gorge. The basalt rocks are dark to light gray, range from aphanitic to porphyritic, and are rich in plagioclase phenocrysts. One to 10 feet of red ash or scoria separates several of the flows. Wherever this red scoria can be examined, it is seen to be the oxidized rubbly, scoriaceous surface facies of the underlying flow. Two specimens, whose chemical analyses are shown in columns 10 and 11, table 2, 314 INVESTIGATIONS OF ALASKAN VOLCANOES were taken from flows 2 and 11 (numbered from base) of the 15-flow sequence exposed in the northwest wall of Crater Creek gorge. The chemical analyses of these two flows are as closely alike, except for oxidation of the iron, as two replicate analyses of the same specimen (Michael Fleischer, written communication, 1948). AGE OF PRECALDERA ROCKS The precaldera rocks are probably mostly of Quaternary age, though some of the lowermost flows may be late Tertiary in age. None of the precaldera rocks show general hydrothermal alteration. This lack of alteration suggests a latest Tertiary and Quaternary age. Erosional features on these rocks are also youthful. Several moder- ately developed cirques are high on Mount Idak, the satellitic moun- tains, and the outer slopes of Okmok Volcano. Glacial striae were found as much as 2 miles downslope from some of these depressions. Of equal importance, much of the major slope surface is clearly only slightly modified from the original surface of the uppermost lava flow. Construction of the main volcanic structures was apparently keeping ahead of the destructive processes of erosion at the time of the latest glaciation, and some of the uppermost flows of the Crater Creek basalt probably were later than the last major glaciation, presumably at the end of the Pleistocene epoch. The precaldera rocks of Okmok Volcano are therefore considered to be mainly Pleistocene to early Recent in age though some of the oldest flows are possibly latest Tertiary. OKMOK VOLCANICS OF CALDERA AGE The thick deposit of pyroclastic material that covers much of the outer slopes of Okmok Volcano is herein named the Okmok volcanics. It includes the Tanak volcanics and the Okmok ash of the preliminary report (Byers and others, 1947), a subdivision abandoned herein, because it was based mainly on degree of sorting, consolidation and other field characteristics that are gradational both vertically and laterally and cannot be mapped consistently throughout the area. The Okmok volcanics reach their maximum thickness of about 400 feet at the type locality on the north rim of Okmok Caldera and thins outward to as little as 20 feet. It has smoothed irregularities of the pre-Okmok surface, filling valleys to depths of several hundred feet and covering thinly most bedrock spurs between valleys. Erosion has cut many V-shaped gullies in the formation, and has removed it entirely from some of the ridge areas. The amount removed by erosion appears negligible compared to the probable original GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 315 amount of material—not less than 7 cubic miles. An average thickness of 300 feet around the caldera rim and of 100 feet at the coast (an average of 6 miles distant from the rim) was used in this approximation. The Okmok volcanics consist of thick deposits of unsorted agglomer- ate in the lower part and beds of moderately sorted bombs and well sorted ash in the upper part. A few thin lava flows, seen in transverse lenticular section, are interbedded in the agglomerate near the caldera rim. Bedding and sorting are irregular in detail. Generally speaking, however, unsorted agglomerate predominates in the lower part at the caldera rim and persists to the seacoast, whereas agglomeratic beds of sorted ash predominate in the upper part at the rim and grade outward into thinner, well-sorted ash at the seacoast. A measured geologic section is given below. Measured section at type locality of Okmok volcanics in north wall, Okmok Caldera (lat 53°28.2'N., long 168°08.1'E.) Okmok volcanics: Feet Ash, loosely consolidated, bedded; contains spatter, teardrop bombs, dark scoria fragments, and very minor lithic fragments; beds appear nearly black from a distance____............-.----.--__._..._.- 120 Ash, loosely consolidated; contains scoria and bombs with many large angular blocks of basalt as much as several feet in diameter; beds appear medium light gray from a distance; gradational wontent With unit below..-.--===----<=----2o2c cece 160 Agglomerate, dense, medium-gray, andesitic; streaky appearance; contains common angular basalt fragments; forms upper welded bed; gradational contact with unit below___.-.......-.___________ 20 Agglomerate, medium-dark-gray, moderately consolidated; consists of spatter, bombs, and basaltic blocks as much as several feet in diameter, all in fine friable ash matrix } gradational contact with unit WOM ee ee ere | ee 40 Agglomerate, andesitic, welded, moderate- to dark-red; appears streaky; contains common angular basalt fragments; forms lower welded bed; gradational contact with unit below.._._______.._____ 30 Agglomerate, moderately consolidated, medium-dark-gray; contains spatter, bombs, and large blocks; upper 10 feet oxidized to moderate red; unit rests disconformably on Crater Creek basalt filling V- shaped gullies Crater Creek basalt _ AGGLOMERATE DEPOSITS The deposits of unsorted agglomerate in the lower part of the Okmok volcanics are in very thick beds as much as 100 feet in thickness. Two agglomerate beds are so completely indurated, or welded, that 316 INVESTIGATIONS OF ALASKAN VOLCANOES they were first mistaken for lava flows. No sorting of fragments by size was seen either vertically or horizontally. Around the caldera rim the agglomerate is locally more than 200 feet thick but is missing from radial ridges. Near Cape Chagak the agglomerate deposits are more than 200 feet thick (pl. 46 B), but just east of Ashishik Point they are only 20 feet thick. A short distance eastward at Cape Tanak they pass below sea level with an unknown thickness greater than 20 feet. Fragments of precaldera rock of all types commonly make up about 5 percent of the agglomerate but in places as much as 20 per- cent, and the agglomerate grades into tuff-breccia. Most of the material is juvenile lapilli and ash of very fine grained, partly glassy, erratically vesicular andesitic scoria. Fresh surfaces of the juvenile matrix are neutral gray, mottled with both lighter and darker colored lithic fragments, but at distances greater than about 2 miles from the caldera rim the fine-grained juvenile matrix is brownish gray, ap- proaching but not duplicating the yellowish brown to bright yellowish orange of the palagonitized pyroclastic rocks of the Ashishik basalt. In places a rhyodacite ash bed is interbedded within the agglomerate. WELDED AGGLOMERATE BEDS Near the rim of the caldera, the agglomerate contains two welded agglomerate beds that are nonvesicular and as completely indurated as dense lava flows. These two beds are exposed in nearly con- tinuous outcrop for 4 miles along the upper part of the north and east walls of the caldera (see pl. 45); one continues along the upper part of the walls of Crater Creek gorge. The two welded agglomerate beds range from 5 to 30 feet in thickness; the variation is apparently related to undulations of the underlying surface. These beds grade vertically into partly indurated agglomerate; one grades laterally into brownish-gray partly indurated (palagonitized) agglomerate, 2 to 3 miles from the edge of Okmok Caldera. The constituent juvenile lava particles of the welded beds were apparently welded at the time of their emplacement. Fine, streaky banding parallel to bedding, common in outcrops of the completely welded agglomerate, apparently is due to flattening of juvenile bombs that now differ in crystallinity. Dark streaks are extremely fine- grained and are charged with opaque oxides; lighter streaks are com- posed of pyroxene, plagioclase, and opaque oxide grains, whose aver- age size is about 0.01 mm. Aside from these differences, nothing in the groundmass of the completely indurated rock suggests boundaries between masses of the juvenile material. The xenolithic fragments are enclosed by the welded juvenile material (pl. 47). A few coarser grained fragments grade into the predominantly fine-grained surround- GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 317 ing groundmass and some patches appear to be ghosts of xenolithic fragments, whose ragged outlines suggest that the fragments have been partly assimilated into the matrix material. Though the micro- scopic texture of the completely indurated layers could be that of a rapidly cooled lava flow charged with xenoliths, the texture was more likely a result of complete welding of the juvenile lava masses, for these indurated layers grade vertically and laterally into unsorted agglomerates clearly composed of discrete bombs, lapilli, and xeno- lithic fragments. Two welded agglomerate beds are also exposed in less extensive outcrops in the north and east arcuate ridges. The following geologic section on the south slope of the north arcuate ridge includes these two welded agglomerate beds. This section differs in detail from the one at the type locality (page 315), in that the upper welded bed is basaltic, rather than andesitic in composition. Measured section of Okmok Volcanics on south slope, north arcuate ridge (lat §3°29.3'N., long 168°06.3’ E.) [Measured with altimeter and tape] Feet Vent agglomerate, oxidized, medium-red, scoriaceous; contains many fragments of mafic phenocryst basalt; weathers grayish red purple; lower 20 ft strongly indurated. Okmok voleaniecs: Agglomerate, pale- to medium-yellowish-brown; contains reddish scoria fragments; possible vent agglomerate________._-.-_------_-____ 60 Agglomerate, basaltic, welded, medium-gray; contains many frag- ments of aphyric basalt; gradational with agglomerates above and Doe) OY. perme perenne rats SSS SOT ene 10 Agglomerate, dark-yellowish-brown; contains large blocks of vitreous andesite as much as 10 ft thick and smaller angular fragments of basalt. Pantiy-coveredas © <Sss0 sei a 100 Agglomerate, andesitic, welded, medium-gray; stained with medium- red iron oxide; has streaky (eutaxitic) texture; gradational with beds above and below.°*- 5 ee 15 Ash, pumiceous lapilli, bedded, yellowish-gray; gradational with ash below ee Ash, rhyodacitic, grayish-yellow; sharp contact with agglomerat elo we Soe ey eee Le See es 75 Vitreous andesite flow__ The welded agglomerate beds were intensively studied both petro- graphically and chemically in the hope that they could be identified elsewhere, and thus be used as marker beds in structural studies of Okmok Volcano. Petrographic analyses (table 11) and silica analyses TasLe 11.—Petrographic analyses of agglomerate and ash marker beds of the Okmok volcanics [Modal analyses in volume percent] Chemical analysis number (table 2)-..------------------------------- 13 BG Nencccandactodn|anewsine-s-ofecocecncnnnclonseesocconst Mode in volume percent: Crystals larger than 0.2 mm: Plagioclase _ Age Smooworn Opaque oxide aggregat Zeolites (largely chabazite) Lithic fragments (largely b: Groundmass..-.... Range in composition e in composition of lower An plagioclase >0.2 mm. B index, higher Fe augite (phenocrysts?) - .---------------------------|-=------3-305-| B index, lower Fe augite Geena) Range in composition of high Fo olivine crystals. Range in composition of low Fo olivine crystal: a Refractive index of groundmass glass.. 2 g 3 s.. 2 8p F = z = s z 3 1. Lower welded agglomerate bed of type locality (p. 315) at north wall, Okmok caldera 6, Partly palagonitized agglomerate from bed at bottom of New Jersey Creek gorge (lat 53°28.3’ N., long 168°08.1’E.). (lat. 53°31.8’ N., long 168°04.5’ E.). 2. Upper welded agglomerate bed of type locality {. 315). 7. Pumice lapilli bed in northeast-facing sea cliff (lat 53°33.5' N., long 168°04.9 E.), 3. Lower welded jomerate bed of section (p. 317) at t south slope nora arcuate ridge 0.6 mile south 10° E. of Ashishik Point. (lat 53°29.2’ N., long 168°06.5’ E.). Composition of hype : Enss-s0. 4, Upper welded agglomerate bed of section {P. 317) at same locality as 3, 5, Partly palagonitized agglomerate from bed at road cut; 300 feet east of bridge across Crater Oreek (lat 53°32.8’ N., long 167°59.1’ E.). 8. Pumiceous ash bed in north side Camp Creek gully (lat 53°23.6’ N., long 167°59.1’ E.). 81g SHONVOIOA NVHSVIV JO SNOILLVOLLSAANI GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 319 (see below) suggest that the lower welded bed of the type locality is correlative with similar beds exposed several miles distant on the north slope. Silica analyses and bulk specific gravity determinations on welded agglomerate beds of the Okmok volcanics [Analysts: Leonard Shapiro, 3-8; author 1-2. Bulk specific gravity determinations by author) SiOz Bulk (weight specific percent) gravity Okmok Caldera (lat 53°28.3’ N., long 168°08.1’ E.): 1. Lower welded agglomerate bed of type section (p. 315) at north wall____ 54, 35 2. Upper welded agglomerate bed of type section (p.315)-. 57. 40 North arcuate ridge (lat 53°29.2’ N., long 168°06.5’ E.): 3. Lower welded agglomerate bed of section (p. 317) at south slope. fies ds S637: acecess 4, Upper welded agglomerate same locality as 3..-..------------------------- n.d. 2. 86 Crater Creek gorge (lat 58°28.0’ N., long 168°5.2’ E.): 5. Dense aphyric lava of welded bed near top of southeast Walloosocoseessscenseseece wenn se seen e- Hee e-- 54.0 2. 80 6. Dark finely scoriaceous lava of welded bed, 2}¢ mi. northeast of 5, in upper part of southeast wall_----- se | Crater Creek (lat 53°32.5’ N., long 167°59.2’ E.), a few hun- dred feet south of road bridge: 7, Black, scoria fragment in palagonite agglomerate bed, 53.7 ------- East arcuate ridge (lat 53°27.0’ N., long 2.8’ E.) at north end: 8. Welded, dense aphyric agglomerate bed at top of sec- i 52.6 2. 84 The composition of the phenocrysts (column 4, table 11) and the bulk specific gravity of the upper welded bed exposed in the north arcuate ridge section indicate a basaltic composition. The 2 welded beds of the section at the type locality and the 2 in the north arcuate ridge section are not correlative from their petrographic and chemical character, although the upper welded bed (column 2) at the type locality may possibly correlate with the lower welded bed (column 3) of the north arcuate ridge section. The welded basaltic beds (Nos. 4 and 8) are limited to the upper parts of the two stratigraphic sections at the north and east arcuate ridges. The present work does not demonstrate whether the section at the north arcuate ridge is partly correlative with that at the type locality. Possibly, the pyroclastic beds exposed in the north and east arcuate ridges were deposited by an earlier caldera-forming eruption related to the origin of the arcuate ridges. 320 INVESTIGATIONS OF ALASKAN VOLCANOES PARTLY PALAGONITIZED AGGLOMERATIC TUFF-BRECCIA No beds completely welded over a large area are found along the seacoast; there they vary from loosely consolidated to moderately in- durated. On the plain south of Cape Tanak, the brownish-gray ag- glomeratic tuff-breccia lies in huge remnant blocks over 2 square miles. These blocks are associated with old channel scars of Crater Creek and appear to have been undermined by removal of a loosely con- solidated agglomerate bed from beneath a thick bed of moderately indurated tuff-breccia. This tuff-breccia south of Cape Tanak contains as much as 20 percent of xenolithic fragments, which include medium- to dark-gray basalt, dark-gray to black vitreous andesite or latite, and yellowish- brown palagonite tuff. Abundant masses of juvenile grayish-black scoria are also enclosed in the palagonitic matrix which is light olive gray to dark yellowish brown. The vitreous latite (see column 17, table 2) is more siliceous than any known precaldera vitreous andesite and possibly represents chilled, unvesiculated juvenile lava of the Okmok volcanics. Less than 1 percent of euhedral to subhedral crystals of plagioclase, monoclinic pyroxene, and olivine are scattered through the grayish-black scoria clots and the matrix. The scoria clots include cryptocrystalline, nearly opaque amorphous material and clear, light-yellow to light-brown glass (see columns 5 and 6, table 11). The groundmass or matrix material enclosing the scoria clots consists mostly of a partly opaque, cellular cryptocrystalline aggregate that ranges from light to dark brown, less commonly greenish, in transmitted light. The refractive index of the aggregate is mostly slightly lower than that of balsam. Though this aggregate was not chemically analyzed it closely resembles the groundmass of the palagonite tuffs of the Ashisbik basalt both in outcrop and in thin section and is therefore presumed to be hydrated basaltic glass or palagonite. RHYODACITE PUMICE BEDS Light-yellowish-gray, unconsolidated pumice and pumiceous ash is associated in three places with agglomerate of the Okmok volcanics. On the south in-facing slope of the north arcuate ridge, 3% feet of gray ash rests in sharp contact on 40 feet of partly indurated ag- glomerate, and is overlain by a 15-foot layer of welded agglomerate (see section, p. 317). In a sea cliff 0.6 mile southeast of Ashishik Point, 1% feet of pumice lapilli ash rests on Ashishik basalt and is overlain by 20 feet of partly indurated agglomerate. On the east slope at an altitude of 1,000 feet, 2% feet of gray pumice ash in the north wall of Camp Creek gully rests on partly indurated agglomerate and is overlain by loosely consolidated agglomerate. A few fragments GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 321 of gray pumice were found in the partly palagonitized agglomerate at an altitude of 100 feet along Crater Creek near Cape Tanak. Petro- graphic analyses of two rhyodacitic pumice specimens are shown in columns 7 and 8, table 11. BEDDED ASH Massive, poorly sorted ash and bedded, sorted, unconsolidated ash in general forms the upper part of the Okmok volcanics. In the caldera walls, the thickness ranges from a few feet to more than 300 feet and averages about 170. (See page 315.) On the outer slopes the ash becomes distinctly sorted and thins rapidly with in- creasing distance from the caldera; less than 5 feet is exposed in the cliffs of the Bering Sea coast. The bedded deposits consist mostly of 1 to 10 foot beds of well- sorted ash, lapilli, and fragments of accessory materials. A con- solidated bomb bed, from 5 to 20 feet thick, is exposed near the bottom of the ash deposits in gullies on the upper slopes of the volcano and in places in the walls of the caldera. Abundant juvenile bombs, 1 to 2 feet long, consist of dark fine-grained aphyric lava, and xenolithic fragments of precaldera basaltic rocks locally as much as 3 feet in median diameter. Angular blocks, 1 to 3 feet across, litter the surface of the ash near the caldera wall and consist of fragments of Ashishik basalt, including palagonite tuff. With increasing distance from the caldera rim the xenolithic fragments become pro- gressively smaller, and at the seacoast they average only about 1 inch in diameter. ORIGIN AND EMPLACEMENT OF OKMOK VOLCANICS The Okmok volcanics are thickest around the rim of Okmok Caldera, yet no recognizable deposits assignable to them are found within the caldera. The thick blanket of Okmok volcanics must therefore be related to the formation of the caldera itself, as Williams (1941) has demonstrated for many calderas throughout the world. The unsorted agglomerate and welded agglomerate beds of the lower part of the Okmok volcanics certainly must have been emplaced as a catastrophic eruption. The lack of bedding and poor sorting of the agglomerate clearly indicate that little if any winnowing of material took place. Furthermore, the presence of welded agglomerate beds also suggests that heat loss due to air transport was at a minimum. The deposits are therefore believed to have been emplaced as a red- hot emulsion of volcanic gases, molten juvenile particles, and minor lithic fragments—the glowing avalanche or nuée ardente (glowing cloud) type of eruption. The absence of the unsorted agglomerate from most of the higher ridges radial to Okmok Caldera further supports this hypothesis of origin. 322 INVESTIGATIONS OF ALASKAN VOLCANOES The partly palagonitized agglomeratic tuff-breccia is not found within 2 to 3 miles of the rim of Okmok Caldera but occurs on the lower slopes of Okmok Volcano, where these partly palagonitized rocks occupy the same stratigraphic position as the nonpalagonitized agglomerate closer to the caldera. The change, moreover, from non- palagonitized agglomerate to palagonitized agglomerate is gradational through a distance of about a mile. Xenolithic fragments are also more abundant in the partly palagonitized agglomerate, which with in- creasing abundance of xenolithic fragments grades into tuff-breccia. These features of the partly palagonitized deposits suggest that as the glowing avalanche reached the lower slopes, farther from the source vents, considerable water from melting snow and ice became admixed with the primary constituents of the glowing avalanche so that when the mass finally came to rest the juvenile particles were hydrated to palagonite. The texture of the partly palagonitized tuff-breccia in many exposures superficially resembles glacial till or, more likely, a mudflow deposit. This general impression suggests that some of the partly palagonitized deposits may have been emplaced as hot mud- flows. The greater abundance of xenolithic fragments in the partly pala- gonitized deposits may be due in part to the glowing avalanche pick- ing up boulders enroute and in part to solidification of molten juvenile fragments with cooling. Part or all of the vitreous andesite-latite fragments in the tuff-breccia may actually represent chilled, unvesic- ulated juvenile lava. This same material closer to the source vents may have cooled more slowly into the finely microlitic lava of the welded agglomerate beds. The bedded ash and lapilli in the upper part of the Okmok volcanics were emplaced as ash falls from explosive or vulcanian eruptions, in part associated with the caldera-forming eviscerating eruptions but in part subsequent to them. Generally, the bedded deposits are the product of the caldera-forming eruptions but include in the upper part an indeterminate amount of ash from postcaldera eruptions. The part of the Okmok volcanics in the north and east arcuate ridges may possibly have been deposited as the result of an eruption earlier than that which deposited the pyroclastic material around the present caldera (see p. 319), but such deposits from an earlier caldera- forming eruption have not been recognized during the present study. AGE OF THE OKMOK VOLCANICS The Okmok volcanics postdate the last major glaciation of Okmok Volcano as neither they nor the wall of Okmok Caldera have been significantly glaciated. However, the carving of Crater Creek gorge 500 feet deep into basalt flows is all later than the Okmok volcanics. GEOLOGICAL SURVEY BULLETIN 1028 PLATE 49 ph of anorthite phenocrysts in anorthite basalt flow (column 5, table 2). Crossed nicols. GEOLOGICAL SURVEY BULLETIN 1028 PLATE 50 Perspective diagram of Okmok Volcano viewed from the northeast. GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 323 Hence, the Okmok volcanics were probably emplaced shortly after the waning of the last main glacial epoch on Umnak Island and are therefore considered early Recent in age, BASALT FLOWS OUTSIDE OKMOK CALDERA Many slightly dissected flows, plugs, and cones of basaltic compo- sition are outside Okmok Caldera and hence cannot be dated with respect to its formation. Vent lavas, such as plugs and cinder cone deposits, are discussed, where appropriate, with the flows with which the vent lavas are associated, though they are grouped together on plate 41 to reduce the number of color patterns. The relations of these youthful volcanic features to the Okmok volcanics could not always be ascertained, because ash from postcaldera eruptions has obscured critical contacts close to the caldera. At least one cone and its associated flow on the caldera rim are known to be precaldera in age. Some flows appear to be interbedded in the Okmok volcanics and are probably contemporaneous; two are shown on plate 41. One is in Missouri Creek about 2 miles north of Satellite field no. 2, and the other is near the mouth of Vermont Creek about 2 miles south of Satellite field no. 2. The unconsolidated pyroclastic debris of most cinder cones on the upper slopes of Okmok Volcano appears to grade laterally into bedded ash in the upper part of the Okmok volcanics, whereas some of the lava flows associated with the same cones pass under the bedded ash of the Okmok volcanics. Some of the cones and flows may predate the Okmok volcanics, but if the bedded ash in the upper part of the Okmok volcanics is actually postcaldera in age, some of these cones and flows may likewise be postcaldera in age. MAFIC PHENOCRYST BASALT ON CALDERA RIM Mafic phenocryst basalt dikes are exposed in the wall of the caldera, southeast of cone C. The dikes connect with a cinder cone and small lava flow on the caldera rim. The rough flow surface is barren of ash, suggesting that the flow may be younger than most ash eruptions in the vicinity, yet the dikes, cinder cone, and the upper edge of the flow are all truncated by the wall of Okmok Caldera, thereby proving that they are precaldera in age. The basalt is porphyritic with phenocrysts of calcic plagioclase, augite, and olivine. BASALT FLOWS ON THE ARCUATE RIDGES A number of small plug-domes, cinder cones, and associated small lava flows with surfaces barren of ash form part of the crest of the north arcuate ridge and the east arcuate ridge. A line of similar extrusive rocks branches from about the middle of the east arcuate ridge and extends southeast down the outer slope of the volcano. 489292—59—_5 324 INVESTIGATIONS OF ALASKAN VOLCANOES The rock in all these cinder cones and lava flows is dense, medium- gray, aphyric basalt. The volume is estimated at less than 0.1 cubic mile. BASALT FLOW ON CAPE ASLIK A large cinder cone, which has been slightly dissected by erosion, and an associated lava flow cover about 4 square miles of the lower western slope at Cape Aslik. The lava flow, which is 100 feet thick in the present sea cliff, extended a 3-mile section of the coastline nearly a mile seaward. There is no ash cover on most of the flow. The rock is aphyric basalt and its volume is estimated to be about 0.1 cubic mile. BASALT FLOW NEAR CAPE IDAK Two small cinder cones and one very small lava flow, with a surface barren of ash, overlie the ash deposit of the Okmok volcanics, which mantles most of the plateau between Mount Idak and Cape Idak. The rock is dark-gray dense basalt with a few small phenocrysts of plagioclase. These volcanic rocks are almost certainly of postcaldera age, for they rest on the Okmok volcanics. ANORTHITE BASALT FLOWS NEAR INANUDAK BAY Five cinder cones are alined easterly from Cinder Point in the head of Inanudak Bay. Anorthite basalt flows are associated with four of the cones. For convenience, these lavas are considered with those of Okmok Volcano even though these anorthite basalt flows near Inanudak Bay came from a separate alinement of vents closer to much earlier vents that extruded the volcanic rocks of central Umnak. The basalt flow at Cinder Point forms the shoreline including Cinder Point, and covers an area of about 0.6 square mile. Two miles inland, a narrow flow extends 1% miles northwest from the second cone. Lava from the third and fourth cones to the east flowed south and coalesced to form a lava field half a mile wide and over a mile long. No lava flow was extruded at the fifth and eastermost cone. No ash covers the surface of the flows, and the cones and flows have not been modified by erosion, although the basal contact at the lower end of the eastermost flow is covered by alluvium. The rock in the cones and flows is porphyritic anorthite basalt with common to abundant phenocrysts of translucent sodic anorthite with a maximum length of 6 mm. The total volume of anorthite basalt is estimated at 0.04 cubic mile. ROCKS OF POSTCALDERA AGE WITHIN OKMOK CALDERA All the rocks within the caldera are obviously younger than the caldera. The most recent deposits are subaerial, and therefore younger than the shrinkage of a lake which temporarily filled the GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 325 caldera to a depth of 500 feet above the present low point of the floor. Other deposits, contemporaneous with or older than the lake, show evi- dence of subaqueous eruption and are overlapped by lake beds or show erosional features associated with the changes of level of the lake. Major relations between some of these early deposits are shown in the diagrammatic profile (pl. 48 in pocket) taken approximately north through cones C and D in the southeastern part of the caldera. The oldest material in the profile is a chaotic assemblage of pyroclastic material that shows no systematic structure; elsewhere in the caldera the oldest deposits are relatively deeply eroded pyroclastic cones. Prominent benches ringing cones C and D are underlain by subacrial flows, which are between 1,350 and 1,400 feet in altitude. Below 1,350 feet they grade abruptly downward into brecciated pillow lavas, which were emplaced under water. The level of change at an altitude of 1,350 feet is consistent around both cones and indicates the altitude of the lake surface at the time of extrusion of the lava flows of cones C and D. The lake level rose subsequently to its highest altitude at 1,560 feet, which is marked by wave-cut scars on cones C0 and D and by the top of sediments deposited at or near the shore of the lake. This altitude is also the top of hard Crater Creek basalt at the low point of the caldera rim, into which the gorge of Crater Creek has been entrenched. Temporary halts in the lowering of the lake level, as the outlet stream deepened its gorge, are indicated by wave-cut scars nicked in cone D at altitudes of 1,140, 1,120, and 1,100 feet and by a prominent graded surface at 1,120 feet on lake sediments just north of cone D. A small, shallow remnant of the lake remains north of cone D at an altitude of about 1,075 feet. ROCKS CONTEMPORANEOUS WITH CALDERA LAKE EARLY POSTCALDERA PYROCLASTIC ROCKS This unit includes bedded deposits of cones G and Hin the northwest floor of Okmok Caldera (pl. 41), the material that comprises an un- named cone older than and due east of cone Cin the southeastern part of the caldera floor, and deposits of pyroclastic material whose source vents could not be determined. The pyroclastic debris not identified with recognized source vents may be part of a breccia that presumably formed the original floor of Okmok Caldera when it first subsided. The cones have very wide low craters with narrow serrate rims. The outer slopes consist of many radial ridges, fluted by gullies, and have the general appearance of badlands. The pyroclastic material of the cones is well sorted, fine-bedded unconsolidated ash and lapilli, ran- domly interbedded with layers of consolidated palagonitized tuff. Many xenolithic fragments ranging from a small fraction of an inch to 326 INVESTIGATIONS OF ALASKAN VOLCANOES blocks as much as 10 feet in diameter are randomly strewn through the sediments; most of them are of aphyric basalt but a few are of palag- onitized tuff, similar to palagonitized tuffs of the Ashishik basalt. The juvenile material is fragmented cryptocrystalline scoria, shards of vesicular glass and palagonite, and broken crystals of plagioclase and pyroxene. Some shards consist of a core of clear, unaltered glass surrounded by a green to yellow to dark-brown palagonitic aggregate of microcrystalline altered material; other shards are entirely the palagonitic aggregate. The eroded, wide-mouthed cones, containing as they do, bedded, well-sorted pyroclastic debris with a large amount of accessory xeno- lithic fragments, resemble Diamond Head on Oahu, Hawaiian Islands (Wentworth, 1926; Stearns, 1935), and are believed to be the products of phreatomagmatic eruptions caused when rising hot magma entered the water-saturated breccia beneath the caldera lake. PLAGIOCLASE BASALT FLOWS OF CONES C AND D Cones C and D are similar to each other but differ in form and rock type from all other cones in the caldera. Each consists of a cinder cone surmounting a terrace. The terrace is underlain by a lava flow that has structures and textures normal in subaerial flows in the upper part and pillow structures embedded in brecciated glass in the lower part (pl. 48). The physiographic form consisting of the cone, subaerial lava bench, and underlying pillow lava has been called a “tuya’” by Mathews (1947) after Mount Tuya, a similar feature in northern British Columbia. Cone D is symmetrical with a single crater at the summit, exposing a vent plug of coarse crystalline basalt. Cone Cis less symmetrical and near the summit has remnants of 5 different vents, one of which contains a plug of coarse-grained basalt. A line of active fumaroles crosses the lip of an asymmetric crater on the south flank, a hundred feet below the summit (Byers and Brannock, 1949, p. 724). Bedded palagonite tuff is exposed locally at one of the subsidiary vents. Most of the pyroclastic material of both cones C and D is dark ash and scoria, locally oxidized to brick red. The rock is porphyritic basalt, containing up to 25 percent of bytownite plagio- clase phenocrysts, and about 1 percent each of olivine and monoclinic pyroxene phenocrysts, all less than 2 mm in greatest dimension. The abundance of phenocrysts is about the same, whether in the chilled margin of the pillows with glass groundmass, in the subaerial flows with aphanitic but crystalline groundmass, or in the plugs with ganu- lar, gabbroid groundmass. The volume of material extruded from cones C and D is estimated at about 0.1 cubic mile. GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 327 BEDDED VOLCANIC SEDIMENTS Bedded volcanic sediments are composed mainly of water-laid volcanic ash and silt and subordinate older dissected alluvium, graded to the surface of the caldera lake. They are scattered in small areas over all except the southernmost part of the caldera floor. The beds range in altitude from nearly 1,900 feet along the wall of the caldera to about 1,140 feet in the northeastern part, where Crater Creek leaves the caldera (pl. 41). The greatest exposed thickness is about 300 feet with the base not exposed. The older dissected alluvium stands highest west of cones @ and H and is largely younger than these. The true lakebeds are of greatest extent in the center and southeastern part of the caldera floor below 1,560 feet, the altitude of a lakeshore scar on the flanks of cones C and D and the topmost horizontal beds in the lake terrace east of cone D. The lakebeds near the head of Crater Creek gorge are below about 1,120 feet, the altitude of a lake- shore scar in the pillow lavas of cone D. The bedded volcanic sediments contain well-bedded silt, sand, fine gravel and lapilli; the silt beds are as thin as a fraction of an inch and the coarser material is in beds as much as several feet in thickness. Light-gray to light-yellowish-gray silt beds form as much as 60 percent of the lakebed section in the northeastern part of the caldera, but coarser materials form most of the deposits, near cones @ and H, which may have contributed coarse debris to them. Blocks as much as several feet across constitute nearly 10 percent of the deposits near the caldera walls, but are less abundant away from the walls. BASALTIC FLOWS LATER THAN CALDERA LAKE BASALT FLOWS FROM CONE F Cone F, a symmetrical cinder cone, and its associated lava flow of aphyric basalt are southwest of cone C (pl. 41), almost against the south wall of the caldera. The lava flowed northward from the cinder cone, apparently following a shallow valley cut in lake beds and ma- terial from cone C after the lake shore had withdrawn to a lower level. A modern glacier encroaches on the south flank of the cinder cone, and the lower end of the lava flow has been covered by more recent alluvium. BASALT FLOWS FROM CONE E Cone E£ is a composite cone in the west-central part of the caldera floor. One of its youngest lava flows reaches over 2 miles eastward beyond the center of the caldera floor, beneath flows of cone A. The cone contains a pit crater, 3,000 feet in diameter at the rim. The lowest part is concealed by a crater lake 500 feet below the rim. Inter- 330 INVESTIGATIONS OF ALASKAN VOLCANOES OLIVINE-RICH BASALT Olivine-rich basalt is represented by 9 out of the 200 rock masses sampled and studied in detail, in the lower part of the Ashishik basalt near Cape Idak and the ridge west of New Jersey Creek. (As much as 25 percent of the rock is magnesian olivine (Fo7s-90) in euhedral to subhedral crystals (maximum size, 5 mm) in an intergranular groundmass of tabular plagioclase (An;;), euhedral to subhedral monoclinic pyroxene and olivine, and opaque oxides. Chemical and petrographic analyses of olivine-rich basalt are given in columns 1 and 1A, tables 2 and 3. PLAGIOCLASE-OLIVINE BASALT Porphyritic basalts containing phenocrysts of both plagioclase and olivine are much more abundant (48 samples) than those with only olivine. They are particularly common among the older lava flows of the plateau northeast of Mount Idak and of Mount Idak itself, and make up the bulk of the satellitic vents on the outer slopes of Okmok Volcano. A few are found in the main structure of Okmok Volcano. Euhedral to subhedral plagioclase crystals have the average composition of bytownite (labradorite in the analyzed sample from Mount Tulik), but the cores of most are anorthite. Magnesian olivine is in euhedral to subhedral phenocrysts, but some are embayed and have reaction rims of clinopyroxene. Augite is present as microphenocrysts as well as in euhedral to subhedral grains in the intergranular groundmass of tabular labradorite, clinopyroxene, olivine and opaque oxides. A chemical and petrographic analysis of a specimen from Mount Tulik is given in column 2, tables 2 and 3, respectively. The percentages of the groundmass constituents are clinopyroxene, 24 percent; tabular labradorite, 26 percent; opaque oxides, 5 percent; and a mesostasis of alkali feldspar and brown glass, 3 percent. Small grains of olivine are rare in the groundmass. PLAGIOCLASE-AUGITE-OLIVINE BASALT Porphyritic mafic phenocryst basalt with conspicuous phenocrysts of both olivine and augite as well as plagioclase are most abundant (20 samples) among the mafic phenocryst basalt flows of the Ashishik basalt. The plagioclase basalt flows of cones C and D within the caldera are described here, though they are on the borderline between mafic phenocryst basalt and feldspathic basalt. The upper 20 feet of a mafic phenocryst basalt flow on the south- east slope of Okmok Volcano is exposed 3 miles northwest of the hangar at Cape Air Force Base. Augite phenocrysts are much larger and more abundant in the lava 20 feet below the top of the flow than at the top. In the specimen collected near the top of the flow (columns 3 and 3a, table 12) diopsidic augite phenocrysts range Taste 12.—Petrographic analyses of matic phenocryst basalt, Okmok Volcano [All phenocrysts or crystals referred to are greater than 0.2 millimeters] 1 la 2 2a 3 3a 4 5 6 7 Chemical analysis number (table 2)- 1 1 4 & once loses 9 |--nncnennnce|-wnn-cnnnnen|-ne-nen- Mode in volume percent: Phenocrysts or crystals: Pi: jase .. 0.0 0.0 Augite_ .0 .0 Olivine. 24.2 20.2 Opaque oxides. 05 05 Groundmass or intersertal material & 75.7 79.7 Range in composition, plagioclase crystals. Average composition, plagioclase crystals. Average com) ition, groundmass plagioclase _ Average 8 index of augite crystals 2V of augite crystals__. sie Range in composition cr} Fore-s6 Average composition of olivine crystals. Fou |-- Fos? -- 1 The smallest ital is about 0.1 mm, 3a, Anorthits-apelie olivine basalt flow, same specimen, but different thin section cut 2 Data on groundmass augite. perpendicular to No. 3, choot Gal werg’ N. 1, Qlfvine baslt dow at Cape Tdak (at 58°14. long 1074787 B). 4 Angife besalt flow St base of northeast wall of Okmok Caldera (at 69°27.9" N., ae basalt flow; same specimen as above bu ' Section cut perpendicular ‘0.1. 5. Olivine-bearing bytownite basalt flow near Mount Idak. Specimen collected 700 2. Anorthite-augite-olivine basalt flow (lat 53°25.7’ N., long 167°65.5’ E.), 3 miles N. yards north of summit at an altitude of 1,000 feet (lat $358.4 N., long 167°53.6’ E.). 25° W. of Umnak Airfield. Specimen collected at nonvesicular base of 20-foot outerop. 6. Bytownite-augite basalt flow at top of Idak Plateau sea cliff facing Umnak Pass 2a, Anorthite-augite-olivine basalt flow; same specimen but different thin section cut perpendic to No. 2. 8, Anorthite-augite-olivine basalt flow, same locality as 2. Specimen collected 20 feet above No. 2 at vesicular top of flow. Note lower content of augite phenocrysts. (lat 53°29.1’ N., long 167°48.6’ E.); altitude 1,280 feet. Olivine-bearing bytownite basalt flow near base of Idak Plateau sea cliff facing Umnak Pass (lat 53°30.0’ N., long 167°47.5’ E.); altitude 240 feet. IS SANVISI NVLLOGTV ‘SANVISI JO1SOD0d GNV HVNIWA JO ADOTORD 334 INVESTIGATIONS OF ALASKAN VOLCANOES Tape 13.—Optical data on anorthite phenocrysis in anorthite basalt near Inanudak Bay No. on fig. 51 avi Pericline twin plane and 001 @) —10° “B) 8 o , gat ” 78° @ 1 @ is acute bisectrix. 2 Unable to determine. a composition of Ang. The same rock has some phenocrysts and microphenocrysts of olivine and a few of augite. The olivine crystals are all surrounded by a reaction rim of clinopyroxene, and those of augite by a reaction rim of opaque oxide and cryptocrystalline material. A sample from the similar flow, 2 miles east of Cinder Point contains less than 5 percent of anorthite phenocrysts similar in composition to those at Cinder Point, a few microphenocrysts ranging up to a com- position of Angs, and microlites of labradorite in a groundmass of glass with an index of refraction of 1.545. A few phenocrysts of olivine and augite have reaction rims comparable to those of the flow at Cinder Point. The 1945 lava from cone A contains nearly 30 percent of phenocrysts of anorthite, 0.5 to 2 mm in length, and about 1 percent each of inconspicuous augite and olivine microphenocrysts (column 6, tables 2 and 3). The plagioclase crystals are slightly zoned from Angg to Ag. Olivine and augite are in euhedral to subhedral crystals; no reaction rim surrounds the olivine. The groundmass is microlitic, ranging from about 0.01 to 0.02 mm in average grain size, and contains plagioclase, clinopyroxene, and opaque oxides. APHYRIC BASALT AND FELDSPATHIC BASALT Four-fifths of the lava flows of Okmok Volcano consist of aphyric basalt and feldspathic basalt. The aphyric or nearly aphyric basalts consist almost entirely of a microlitic intergrowth of labradorite, clinopyroxene (pigeonitic?), and opaque oxides (see column 11, table 3), but all gradations may be seen between aphyric and feldspathic basalts by increase of plagioclase phenocrysts. Most of both varieties have rare microphenocrysts of augite. A few relict microphenocrysts of olivine can usually be found. The groundmass in similar, but more coarsely crystalline rocks, is intergranular, and the plagioclase crystals grade in size from the small phenocrysts to those in the groundmass. The olivine is commonly embayed and surrounded by a reaction rim of clinopyroxene. One typical microgranular feldspathic basalt con- tains about 17 percent of normally zoned phenocrysts of bytownite, GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 335 a rare microphenocryst of augite, and a few embayed crystals of olivine surrounded by clinopyroxene reaction rims. The inter- granular groundmass is composed of 38 percent laboradorite, 35 percent monoclinic pyroxene, 6 percent opaque oxides, and 4 percent interstitial light-brown glass (index of refraction, 1.498-1.503) and alkali feldspar. ANDESITE AND LATITE Rocks of andesitic to latitic composition form a few dikes, plugs, and stubby lava flows of the volcanic rocks of minor vents near the walls of the caldera, and occur as fragments in the pyroclastic rocks of Okmok volcanics. They are dark-gray to black, dense to glassy, nonvesicular rocks, commonly aphyric, and many contain small xenoliths of basaltic rocks. Most specimens have hyalopilitic ground- masses with varying amounts of glass. Microphenocrysts of plagio- clase (Ang; to Ang) are common to rare, and some rocks contain amhedral and embayed fragments of calcic bytownite (Ang) to Angs). Anhedral fragments of augite, also embayed and commonly bent, are less abundant than those of calcic bytownite. Grains of olivine from several specimens have indices of refraction corresponding to the high fayalite olivines, hyalosiderite to ferrohortonolite (Fo;o to Fo.;) and contain no reaction rims of clinopyroxene. The index of refraction of the groundmass glass ranges from 1.545 in vitreous andesite to 1.511 in vitreous latite. Petrographic analyses of 4 specimens are given in table 14. Taste 14.—Petrographic analyses of vitreous andesites and latite of Okmok volcano [Phenocrysts are all greater than 0.2 millimeters] L 2 3 4 Chemical analysis number (table 2) 15 Silica content. (weight percent) .. 54.80 63.5 0.2 0.8 11 47 al al +2 +3 <01 <0 <1 <1 an -0 .2 +2 Groundmass.. a 99.6 99.1 98.5 94.7 Range in composition of high-An plagioclase phenocrysts.. -| Amgo-s5 |---------- Ang-ss |---------- Range in composition of low-An plagioclase phenocrysts. Range in composition of olivine phenocrysts. Refractive index of groundmass glass_ Bulk specific gravity of rock_...-...-- 1 $10; determination by Leonard Shapiro, using rapid methods. 2 Specimen contains a very fine microlitic groundmass charged with opaque oxides and hence the glass, if present, could not be determined. 1. Top of andesite pipe at intersection of Crater Creek gorge and Okmok Caldera (lat 53°28’ N., long 168°05.4’ E.) 2. Andesite flow exposed at 2,400 feet on south side, north arcuate ridge (lat 53°29.2’ N., long 168°06.6’ E.) 3. Top of small andesite plug or large block at 2,500 feet on north arcuate ridge (lat 53°29.3’ N., long 168°06’ E.) 4, Hand specimen of latite fragment in Okmok volcanics. Crater Creek canyon near Cape Tanak (lat 53°32.6’ N., long 167°59.2’ E.) 336 INVESTIGATIONS OF ALASKAN VOLCANOES RHYOLITE One apparently elongated dome of rhyolite was found on the outer northwest slope of Okmok Volcano. The rock ranges from light-gray felsitic vesicular rhyolite to nonvesicular black obsidian. Xenoliths of basaltic andesite are sparse to common in the obsidian but lacking in the felsite. The xenoliths apparently were being resorbed (pl. 47B) and fragments of laboradorite, augite and olivine in the rhyolite must be xenocrysts. Feldspar microlites in the glass have a composition of An,;, and the glass has an index of refraction of 1.497. Petrographic analyses are given in table 15. Taste 15.—Petrographic analyses of rhyolite specimens and xenolithic inclusion, Okmok Volcano Chemical analysis number (table 2) -.-.----.-------- ied cencdelcouasssstetowsspabene 20 on —— Percent: enocrysts >0.2 m Plagioclase. Augite-_ Olivine... Groundmass. - Range in composition of plagioclase phenycrysts 20.2 mm____-__-__...<,---.--.-----------.-----<--| Anss-70 Approximate composition, feldspar microlites in Angas groun 5. 6 of augite phenocrysts. ‘Range in composition of olivine phenocrysts. Refractive index of groundmass glass, Bulk specific gravity of rock. Foes-77 1,497 2.47 Fou-sa 1 Only one grain sampled. 1. Most vitreous obsidian band interlayered with less vitreous band and felsitic rhyolite near northwest end of northwestward extension of dome (lat 53°29.5’ N., long 168°10.3’ E.). 2. Most vitreous obsidian band interlayered with less vitreous band and felsitic rhyolite on northeast side of northwestward extension of dome (lat 53°29.2’ N., long 168°09.8’ E.). 3. Purplish basaltic andesite xenoliths in obsidian, same locality as 2. 4, Less vitreous obsidian band interlayered with more vitreous obsidian and felsitic rhyolite, same locality as 2. 5. Light-gray felsitic rhyolite, east end of dome on north side of north arcuate ridge (lat 53°29.4’ N., long 168°09’ E.). PALAGONITIZED PYROCLASTIC ROCKS Palagonitized pyroclastic rocks are most abundant and most com- pletely altered in the agglomerate and tuff of the Ashishik basalt ex- posed in the walls on the outer slopes of the caldera. Partly palagoni- tized rocks are in the agglomerate of the Okmok volcanics more than 2 miles from the caldera, and in some of the pyrcolastic cones within it. The palagonitized rocks of the Ashishik basalt contain fragments of vari-colored scoria, gray to black xenolithic fragments, and broken crystals of plagioclase in a yellowish-gray, earthy matrix. Fragments of light-brown scoria contain cores of unaltered clear light—yellow glass rimmed by palagonitized glass. In different specimens, the index of refraction of the unaltered glass ranges from 1.560 to 1.575. GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 337 Fragments of yellow to orange-brown scoria consist of isotropic palagonite glass that ranges in refractive index from 1.54 to 1.57. The scoria fragments are slightly to moderately cellular with vesicles 0.05 to 0.1 mm in diameter. The xenoliths probably are broken wall- rock fragments of Okmok Volcano. Fragments of plagioclase, augite and olivine crystals may be in part xenocrystic as their abundance seems to vary directly with the abundance of xenolithic fragments. The matrix is a cryptocrystalline aggregate, apparently of palagoni- tized glass shards, as shard shapes are discernible in parts of the ma- terial. The cryptocrystalline palagonite is green, yellowish, to dark brown, as seen in plane-polarized light in thin section, and is made up of birefringent microlites in both felted and radial-fibrous texture. Many vesicle fillings are compound and consist of a lining of concentric, moderately birefringent fibers, and a central filling of nonfibrous, low birefringent material with refractive indices 1.480 and 1.484 (chabazite ?). Some vesicles contain a yellowish to reddish- brown mineral in crudely spherulitic aggregates of curved concentric fibers with positive elongation, birefringence about 0.01, and index of refraction about 1.56 to 1.57. Another cavity-filling zeolite, in sheaflike aggregates of radiating fibers, has positive elongation, parallel extinction, and indices 1.467 and 1.470. Chemical analyses of two specimens of palagonitized rock are given in columns 3 and 8 in table 2. STRUCTURE OF OKMOK VOLCANO The major structural features of Okmok Volcano—the caldera ring faults and the probability of two calderas—are mostly concealed by the Okmok volcanics and other pyroclastic debris. Therefore, their existence is in large part inferred from physiographic features and vent alinements, and by comparison with similar calderas elsewhere in the world. First, however, the known or observable structural features of Okmok Volcano are considered, followed by a discussion of the origin of the arcuate ridges. Okmok Volcano is mainly composed of gently outward-dipping flows and interbedded pyroclastic rocks of the precaldera Ashishik basalt (see sections AA’ and BB’, pl. 41). Dips greater than 5° are restricted to local vent masses of relatively small volume. The welded agglomerate beds of the east-central part of the north arcuate ridge actually dip gently southward toward Okmok Caldera. The low plateau between Mount Idak and Cape Idak is underlain by Ashishik basalt and minor pyroclastic interbeds that dip 2° to 3° northeastward, except in the eroded cone of Mount Idak. Inasmuch as the average outward dip of the layered volcanic rocks is less than the topographic slope, progressively younger flows are found as one 338 INVESTIGATIONS OF ALASKAN VOLCANOES ascends Okmok Volcano. Most of the dips observed in the Okmok volcanics are essentially parallel to the present topographic slope. Okmok Caldera is in large part a feature of collapse and its infacing cliff walls are the slightly modified scarps of a ring fault. One fault, associated with the ring system, has been mapped. In the northeast wall of Okmok Caldera south of Crater Creek gorge (pl. 41), a fault that dips steeply southwest toward the caldera is exposed for about a mile. The southwestern block has subsided about 100 feet as shown by the displacement of a bed of welded agglomerate in the Okmok volcanics. The bedded volcanic sediments in the caldera have been displaced in a few places by postcaldera movement, probably involving the caldera ring fault. Along the northwest wall, north of cone H (pl. 41), the bedded volcanic sediments dip as much as 30° toward the center of the caldera but decrease in dip at greater distance from the wall. Near the caldera wall south of the entrance to Crater Creek gorge, the lake beds dip away from the wall, with dips as much as 25° and open folds, too small to show on plate 41, have produced local ridges and closed depressions with a relief of 20 to 40 feet. These features were formed by postcaldera movements, but their location with respect to the caldera wall suggests that the fault involved is the caldera ring fault. A few faults are radial to Okmok Caldera. Two north-striking faults are marked by scarps extending inland from the mouth of New York Creek on the north coast. The east block has dropped about 20 feet on each of these faults. A radial fault is suggested by the line of recent vents extending southeastward from the middle portion of the east arcuate ridge. Faultline control is suggested by the straightness of the cliffs bounding Mount Idak on the south and the northwest and by the parallel sea cliff along the southeast margin of the Mount Idak highland. Two hypothetical radial faults are also suggested by the north-trending cliff at the east end of the north arcuate ridge and by the northeast-trending cliff at the north end of the east arcuate ridge. Three hypotheses are considered relating to the origin of the arcuate ridges. As one hypothesis, the north and the east arcuate ridges (pl. 41) are possibly receding fault scarps controlled by a concealed underlying ring fault, between the caldera and the ridges. In the northwest wall of Okmok Caldera, north of cone H, a normal fault dips about 50° SE., toward the caldera. This fault separates Crater Creek basalt on the down-thrown side from the older Ashishik basalt (pl. 41) and may be a faulted depositional contact. About 100 feet of vertical displace- ment has occurred along this fault during or following the collapse of GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 339 Okmok Caldera, for the fault displaces the Okmok volcanics and can be followed as a recent scarp decreasiog toward the northeast (pl. 41). This fault, however, may have been localized along an earlier caldera fault, whose scarp is now the southerly slope of the north arcuate ridge. No trace of this hypothetical ring fault was found in the gap underlain by Crater Creek basalt between the north and the east arcuate ridges. Exposures of the Crater Creek basalt flows are con- tinuous on each side of the gorge of Crater Creek in the area where the ring fault should be found; furthermore, the Crater Creek basalt is not disturbed or even flexed in this area. (See pl. 464.) The ex- posure of older rocks of pre-Crater Creek age in the floor of the gorge is not continuous, so that this ring fault could be concealed. This fault or a similar fault intersects the east wall of Okmok Caldera northeast of cone D (pl. 41). Two sets of ring faults are therefore suggested by the known faults, volcanic vents, and the topography: a younger ring fault bounding Okmok Caldera and truncating the Crater Creek basalt, and an earlier ring fault farther to the northeast, which bounds a possible older caldera whose rim includes the north and east arcuate ridges. The two Umnak calderas slightly separated in space and time are not unlike two similarly related caldera subsidences at the island of Mull, Scotland (Richey, 1935, p. 61-73; 1937, p. 13-25). The arcuate ridges are in part analogous to the elliptical. arrangement of vent debris around Medicine Lake caldera in California (Anderson, 1941). As a second hypothesis, the two ring faults referred to above may possibly have formed at nearly the same time during a single caldera- forming episode (Byers, Hopkins, Wier, and Fisher, 1947, p. 37). The present author no longer favors this hypothesis, because the slope of the arcuate ridges that faces the caldera has a subdued topog- raphy, blanketed by Okmok volcanics, in contrast to the youthful precipitous cliff of Okmok Caldera, and also because the thick sequence of many flows of the Crater Creek basalt at low altitude seemingly have been localized prior to Okmok Caldera in a gap between the arcuate ridges (see p. 313). A third hypothesis is that the arcuate ridges are dominantly con- structional. The vents that supplied the constructional debris were probably controlled by an underlying ring fissure (not shown on pl. 41), directly beneath the vents along the crest of the arcuate ridges. The chief evidence in support of this hypothesis is as follows: First, the many recent vents situated along the crest suggests that the arcuate ridges may have resulted from similar volcanic accumulations earlier in their history; second, the gentle southerly dip of the Ashishik basalt on the south side of the north arcuate ridge is away from the 489292—59 6 340 INVESTIGATIONS OF ALASKAN VOLCANOES crest of the ridge and toward Okmok Caldera; and third, the fre- quency of volcanic pipes and dikes is greater where cross sections of the arcuate ridges are exposed in the walls of Okmok Caldera and in the high cliff at the east end of the north arcuate ridge (the many dikes are not shown separately on pl. 41). Under the third hypothesis the growth of the arcuate ridges took place largely prior to the extru- sion of the Crater Creek basalt, and no faults of large displacement would be required to account for the existence of the arcuate ridges. However, their origin is perhaps best explained by the first with subsequent modification under the third hypothesis. ORIGIN OF OKMOK CALDERA In the following discussion of the origin of Okmok Caldera an attempt is made to compare the missing summit mass of Okmok Volcano with the amount of xenolithic fragments in the known volume of Okmok volcanics, after the method used by Diller (1902) and Williams (1941 and 1942). Much and perhaps most of the original volume of the Okmok volcanics was deposited beyond the shores of Umnak, so the total volume of volcanic rocks cannot be calculated. The total volume of rock missing from the summit of Okmok Voleano could easily be computed if the form and height of the missing portion on the site of the present caldera were known. Two possi- bilities are considered. As one possibility the missing central portion might have included an earlier caldera with relatively low summit area. The infacing cliff of the north arcuate ridge and the topographically low position of the Crater Creek basalt in the gap between the north and east arcuate ridges suggest that an earlier caldera may have existed. The gentle, radially outward dips of the precaldera volcanic rocks likewise suggest a low shieldlike volcano. As the second possibility the central portion might have consisted of a siliceous steep- sided pyroclastic cone surmounting the pile of Ashishik basalt. Although the vitreous andesitic lava fragments in the Okmok volcanics may represent chilled, juvenile lava, it is also possible that fragments of andesite, latite, and rhyodacite in the Okmok volcanics may have come from thick deposits of siliceous volcanic rocks closer to the central vent. An analogy to a possible form of Okmok Volcano prior to the caldera-forming eruption is afforded by the pyroclastic cone of Mount Vsevidof on southwestern Umnak. Were Mount Vsevidof to experience a caldera-forming eruption similar to that at Okmok, little evidence of the former presence of the cone would remain after caldera collapse except xenolithic andesitic scoria and dacitic pumice fragments of the cone in the pyroclastic ejecta beyond the limits of the caldera. Andesitic scoria fragments as lapilli and bombs in the Okmok volcanics are probably juvenile, though some of the broken fragments could GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 341 conceivably have come from a precaldera cone. Hence, computations of the missing volume are made considering both possibilities as extremes. Assuming an earlier caldera as one extreme, the average altitude of the summit area of Okmok Volcano prior to the caldera-forming eruption is estimated as 2,500 feet or about the same as the present average altitude of the caldera rim (fig. 52). The average altitude of the present caldera floor beneath the postcaldera deposits is esti- mated to be about 1,000 feet (fig. 52). The diameter of the summit area is taken as 54% miles. The missing volume of rock calculated according to these assumptions would be about 7 cubic miles. As the other extreme, a pyroclastic cone is assumed with an average slope of 25° and a base, 5% miles in diameter, at 2,500 feet (fig. 52), on top of and in addition to the minimum portion already computed. Under these assumptions the missing central portion of Okmok Volcano would be about 16 cubic miles. The actual missing volume is prob- ably between 7 and 16 cubic miles. The xenolithic fragments in the Okmok volcanics are probably derived from the mechanical disintegration of the former summit portion of Okmok Volcano. The volume of xenolithic fragments observed in thin sections of the welded agglomerates averages less than one percent (see table 11), but the sections were taken from samples selected so as not to include the larger xenolithic fragments. The welded agglomerate beds as observed in outcrop were estimated to contain at least 10 percent of xenolithic fragments (see pl. 47A). The two specimens of partly palagonitized agglomerates studied (see table 11) contain 15 to 20 percent of xenolithic fragments, and from a study of many outcrops 20 percent is believed to represent nearly the maximum content of xenolithic fragments. If we assume that an overall average of 15 percent of xenolithic fragments is in the 7 cubic miles of the Okmok volcanics deposited on Umnak (see p. 315) only about 1 cubic mile of the missing summit portion of Okmok Volcano is represented in the known portion of the Okmok volcanics. Even if the minimum figure of 7 cubic miles is assumed for the volume of the missing summit, at least 49 cubic miles of Okmok volcanics would have been deposited in order to contain the missing summit as xenolithic fragments, or 42 cubic miles of the pyroclastic rocks deposited be- yond the present shoreline. Such a figure seems unreasonably large in view of the thinning of the Okmok volcanics from an average of about 300 feet at the caldera rim to an average of about 100 feet at the present coastline. Hence, most of the missing summit of Okmok Volcano is believed to have collapsed inward to form the present Okmok Caldera. 10,000' 1000'. Sea level! -| 10004 Sw Precaldera rocks Nt i NE Okmok volcanics =a - ta ‘Assumed floor of original caldera before postcaldera filling Postcaldera deposits Precaldera rocks Ls ‘534 miles > Fiaure 52.—Profile through the summit area of Okmok Volcano, showing idealized maximum (dotted) and minimum (dashed) profiles used in computing volumes of missing summit. ore SHONVOIOA NVMSVIV JO SNOILVDILSAANI GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 343 SURFICIAL DEPOSITS The surficial deposits on Umnak Island range in age from late Pleistocene to Recent and comprise till, glacial outwash, beach de- posits, dune sand, talus and alluvium. TILL Two tills have been differentiated on the geologic map. The older till is restricted to the southwestern part of the island, where it occurs on the slopes of Mounts Vsevidof and Recheschnoi (pl. 40). Near the base of these mountains the older till occurs as isolated patches of end moraines as high as 100 feet. Upslope from these end moraines the older till forms a discontinuous mantle of ground moraine on some of the gentler surfaces. Downslope from the moraines on the pied- mont surface south of Mount Vsevidof are a few patches of older till, which were impractical to map because of the thick tundra cover. It is not known whether these patches represent a farther advance of the same ice that deposited the end moraines or whether they date from an earlier and more extensive glaciation. (Geomorphic evi- dence for this earlier glaciation is discussed under the heading ‘‘Ni- kolski Plain.”) No soil profiles were measured on the older till but it has been exposed to weathering long enough so that it supports blueberry bushes. The older till unconformably overlies the bedrock volcanic rocks, including the older lava flows of Mount Vsevidof. It is in turn over- lain by the younger till, outwash, alluvium, and the oldest of the beach deposits. If, as is inferred below (see p. 345), the oldest beach ridges date from the post-Wisconsin thermal maximum of about 5,000 years ago, the till must antedate the thermal maximum and is therefore designated as Wisconsin(?) in this report. It may coincide with Sharp’s (1951, p. 100) “later glaciation” in the Wolf Creek area of the St. Elias range, Yukon territory. Sharp dates this glaciation as latest Wisconsin and post(?)-Wisconsin. The younger till occurs on the slopes of Mounts Vsevidof and Recheschnoi in the southwestern part of the island (pl. 40) and within Okmok Caldera (pl. 41) in the northeastern part. Its distribution is closely related to the pattern of existing glaciers, and it was evidently deposited in the very recent past when the glaciers were more exten- sive than they are today. In Okmok Caldera the younger till is represented by small patchy end moraines, many just below small rudimentary cirques in the caldera wall. One approximately half way between cones G and H represents the maximum advance of a postealdera glacier here. Another is at the base of the east wall 1% miles southeast of cone D. Some small moraines of the younger till probably have been buried by the recent lavas of cone A. The 344 INVESTIGATIONS OF ALASKAN VOLCANOES younger till consists entirely of unweathered bouldery debris. In contrast to the older till, which supports bushes, the exposed surfaces of the younger till are nearly bare. Some of the boulders in the end moraines of the younger till on Mount Recheschnoi are brightly stained by iron oxide that may have been developed by weathering prior to glacial transport. The younger till rests unconformably on the older till, the bedrock lavas, and the early postcaldera pyroclastic rocks in Okmok Caldera and is overlain only by some of the most recent alluvial deposits. The unweathered character of the younger till and its areal asso- ciation with existing glaciers indicates that it was deposited in late Recent time, possibly during Matthes’ “Little Ice Age,” (Matthes, 1941; Thorarinsson, 1939; Ahlmann, 1948; Lawrence, 1950; Sharp, 1951; and Mathews, 1951). OUTWASH Outwash consisting chiefly of boulder gravel fills most of the valleys that lead from existing glaciers on southwestern Umnak (pl. 40). The outwash ranges in age from that deposited penecon- temporaneously with the younger till to that which is being laid down today. No outwash correlative with the older till was identi- fied, although some remnants of older glaciofluvial deposits are probably preserved locally. BEACH DEPOSITS Beach ridges are found along both the northeast and southwest coasts of Umnak Island and are best developed behind those broad bights where the longshore currents and waves are generously supplied with glaciofluvial sediments. In several areas the supply of sediment has been sufficient to permit the seaward extension of the coastline more than one-half mile. Stages in this extension are marked by 12 or more concentric beach ridges subparallel to the present coastline. Along the Pacific side of Umnak, between Amos and Russian bays (pl. 40), 20 arcuate concentric beach ridges were deposited. No precise altitudes were measured, but the innermost or oldest ridge seemed to be roughly 10 feet higher than the outermost or youngest ridge. The material of the beach ridges ranges in size from sand up through boulders. For convenience in mapping, those beach deposits that consist primarily of sand and pebble gravel have been called sand beaches and those which consist of coarser material have been called boulder beaches. In general, the boulder beaches are small and are insignificant, and are found chiefly in the smaller coves exposed to GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 345 the open sea. Hypersthene andesite talus blocks from sea cliffs on some exposed headlands have been rounded by storm-wave action into nearly perfect spheres up to 6 feet in diameter. Concentric beach ridges subparallel with the present shoreline have been observed by the author on Unimak, Akun, Akutan, and Unalaska islands in the eastern Aleutians and on Attu in the western Aleutians. Bradley (1948, p. 226) describes concentric beach ridges on Adak and also describes a wave-cut terrace along Kulak Bay on Adak Island, about 10 feet above present sea level. Bradley ascribed the higher beach ridges and the wave-cut bench to tectonic uplift of Adak Island. However, the presence of beach ridges and wave-cut benches at about the same altitude above sea level on so many of the Aleutian Islands, suggests a small eustatic fall of the sea rather than tectonic uplift. Beach ridges, wave-cut benches, wave-cut cliffs and other shoreline features subparallel to the present coastline, record a recent stand of the sea roughly 10 feet (3 meters) above present sea level in many parts of the world. Many investigators have postulated that this higher stand of the sea coincided with the post-Wisconsin thermal maximum. (See Fairbridge, 1948, p. 62-63; Idem, 1950, p. 181-184; MacNeil, 1950, p. 1307-1308; Newell and others, 1951, p. 11; Ives, 1951, p. 220-221; Upson, 1951, p. 425-427.) The writer suggests that the oldest. beach ridges on Umnak Island were formed during the same high stand of the sea that was responsible for the 3-meter shoreline elsewhere in the world and, therefore, presumably date from the post-Wisconsin thermal maximum. According to Flint and Deevey (1951, p. 275-278), the thermal maximum occurred 4,000 to 6,000 years ago, and a reasonable check on the correlation of the oldest beach ridges of Umnak with this period is provided by a carbon-14 age determination of charcoal from the Nikolski archeological site. The Nikolski site is on top of one of the older gravel beach ridges (not shown on pl. 40) along the southern shore of Mueller Cove in Nikolski Bay. The material at the site consists of windblown silt, volcanic ash, and human occupa- tional debris (William S. Laughlin, written communication, 1951). A carbon-14 age determination of charcoal collected from a level about 150 centimeters above the top of the beach gravel gives an age of 3,018 +230 years, (Arnold and Libby, 1951, p. 118). By extrapolat- ing from the dated layer, Laughlin (written communication, 1951) estimates that the oldest deposits in this site are about 5,000 years old. Therefore, the beach ridge on which the site rests is probably older than 5,000 years, and may well have been formed during the thermal maximum. 346 INVESTIGATIONS OF ALASKAN VOLCANOES SAND DUNES Sand dunes derived principally from the beach deposits are encroach- ing on and burying their parent materials along several parts of the coastline, particularly on western and northern exposures, where on-shore winds frequently exceed 100 miles per hour during the winter months. The average height of the dunes in most areas is about 30 feet but some of the larger dunes at Geyser Bight, Russian Bay, and back of the bight south of Mount Vsevidof, are as high as 75 feet. Most of the dunes were active in 1947; a few were stabilized and partly covered with grass. TALUS AND OTHER COLLUVIUM Talus and other colluvium in the form of cones and aprons are at the base of all, or nearly all the steep slopes on the island, but only locally are the deposits large enough to warrant separate mapping. In most places they have been included with the pyroclastic debris, and on the slopes of Mount Recheschnoi they have in places been included with the older till. ALLUVIUM Recent alluvium consisting chiefly of reworked pyroclastic material has been deposited as alluvial fans by intermittent streams, and as flood plain deposits by the larger streams that are not fed directly by glaciers. The larger areas of alluvium have accumulated at low altitudes just inland from the oldest of the beach ridges. The largest area in the southwestern part of the island underlies the flood plain of the stream draining into Geyser Bight (pl. 40). In the northeast part of the island are several large areas of alluvium. One area covers about 5 square miles just northwest of Mount Idak (pl. 41), and another area of nearly 7 square miles has accumulated behind the beach ridges north of Cape Kigunak. Still another area is found at Izhiga Cove, where no beach ridges are present and the alluvium has been subject to marine erosion. The alluvium rests on all bedrock and surficial deposits except some of the most recent glacial outwash and some of the most recent flows within Okmok Caldera. These outwash deposits and flows locally overlap alluvial deposits. Although alluvium is currently being deposited by many streams, at several localities, especially near the coast, erosion is actively gullying older alluvial deposits. The gullied deposits may well be graded to the higher beach deposits and therefore may be as old as 5,000 years. GLACIER ICE For convenience, glacier ice has been included under the category of surficial deposits. Except for an arcuate patch of ice, probably stag- nant, along the south rim of Okmok Caldera (pl. 41), the existing gla- GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 347 ciers on Umnak are confined to the southwestern part of the island on the slopes of Mount Recheschnoi and Mount Vsevidof (pl. 40). Alto- gether, there are 9 major valley glaciers on these 2 mountains, and in addition several small glaciers and perennial snowfields. Seven of the 9 glaciers head from the central summit ridge of Mount Recheschnoi; 4 drain to the north and are in general longer and narrower than the 3 which drain to the south. The north-flowing glaciers extend down to altitudes of 200 to 300 feet, whereas the south-flowing glaciers, al- though bigger in volume, terminate at altitudes ranging from 500 to over 1,000 feet. Possibly the south slope of the mountain is both warmer and moister than the north slope; therefore, although the sup- ply of ice may be greater, the wastage is also more rapid and the fronts terminate at higher altitudes than those on the north side. Both glaciers on Mount Vsevidof head in the extinct summit crater. One drains to the north and divides into two branches about 1 mile from the crater. The other glacier flows east, then south, joining the westernmost glacier of Mount Recheschnoi. These two glaciers from Mount Recheschnoi and Mount Vsevidof are separated by a medial moraine. All 9 valley glaciers have excavated their valleys to depths of about 500 feet. The termini of 8 of the glaciers lie close to the end moraines of the younger till, the exception being the most westerly of the glaciers of the north slope of Mount Recheschnoi. The terminus of this glacier is more than 1 mile back of its recent end moraine, indicating that for some reason this glacier has retreated more rapidly than the others of Umnak Island. EROSIONAL FEATURES Three major physiographic features, not directly associated with either the volcanic deposits, or the surficial deposits, are described and interpreted in this section. These features are Crater Creek Valley, pedimentlike slopes on Okmok Volcano, and the Nikolski plain. CRATER CREEK VALLEY Crater Creek heads in Okmok Caldera, from which it flows northeast about 8 miles, entering the Bering Sea about 1% miles southeast of Cape Tanak (pl. 50). For a distance of about 2 miles from the rim of the volcano, Crater Creek flows in a deep gorge cut into the volcanic rocks of Okmok Volcano. The gorge is the result of downcutting of the outlet of the lake that once occupied Okmok Caldera. The gorge thus postdates the formation of the caldera and the 1,560-foot level of the caldera lake. Many small cascades 5 to 10 feet high, and one large waterfall 80 feet high indicate that rapid downcutting of the gorge is still in progress. Downstream from the gorge, Crater Creek crosses 348 INVESTIGATIONS OF ALASKAN VOLCANOES a broad gently sloping plain south of Cape Tanak. Many strath terraces and abandoned channels subparallel to the present valley of Crater Creek indicate that the stream has wandered widely during its dissection of this plain. Two principal sets of terraces were noted; one about 200 feet above the present valley floor and the other about 50 feet above the present stream bed (see pl. 50). PEDIMENTLIKE SLOPES ON OKMOK VOLCANO Most of the lower slopes of Okmok Volcano (pl. 50) are gentle, concave-upward erosional surfaces that in less humid climates would be called pediments. The surfaces start at altitudes ranging from 700 to 900 feet with gradients of 200 feet per mile that gradually flatten seaward to about 100 feet per mile. The slope of the volcano just above the heads of the pediments is 250 to 500 feet per mile. Recent gullies of the perennial streams have been incised about 100 to 200 feet into the pedimentlike slopes (pl. 50). The bedding in the Okmok volcanics parallels the lower pedimentlike slopes, which are in many places constructional surfaces. Whether the pedimentlike surface on the Okmok volcanics reflects a buried erosion surface or is merely the result of smoothing a buried irregular topography could not be determined. THE NIKOLSKEI PLAIN—EVIDENCE OF AN OLDER GLACIATION The southwestern part of Umnak Island (pl. 40) consists of a gently rolling piedmont plain with a relief of less than 300 feet, herein called the Nikolski plain. Except for a few scattered patches of questionable till and some flanking beach deposits, the Nikolski plain is free of surficial deposits, but its surface is dotted with many un- drained depressions, some of which serve as basins for small lakes. A few of the lakes near the coast have been drained by the cutting down of their outlets, and all of these former outlets are now youthful V-shaped gorges. The majority of the lakes have not been altered since they were formed. A study of the air photographs of the islands south of Umnak suggests that they also have been glaciated. Kigul and Ananiuliak Islands (see pl. 39) consist of rolling surfaces with undrained depres- sions similar to the Nikolski plain, suggesting that the glacier extended a considerable distance south of the present shores of Umnak. Comparison of the relatively flat level surface of the Nikolski plain with the flat submerged continental shelf area of the Aleutian Ridge (see pl. 39) suggests that the Nikolski plain is an uplifted portion of the shelf somewhat modified by extensive piedmont glaciation. Studies in the Near Islands of the western Aleutians (H. A. Powers, J. P. Shaefer, and others, oral communications, 1950) indicate similar GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 349 widespread glaciation of both the islands and the continental shelf areas. The overall lack of drainage integration of the Nikolski plain suggests that the glaciation of the area was late Pleistocene, possibly Wisconsin in age. Capps (1937, p. 161-168) has described and illus- trated similar features on Kodiak Island, and ascribes them to the latest Pleistocene or Wisconsin stage. Further work is necessary to substantiate this hypothesis. If the piedmont glacier that extended over the Nikolski plain and the island south of Umnak represented one of the late Wisconsin substages, it is possible that the older till of the end moraines (Qmo pl. 40) may represent some post-Wisconsin ad- vance of the ice. GEOLOGIC HISTORY The earliest known geologic event was the deposition of a thick sequence of waterlaid bedded tuffs and volcanic muds, whose meta- morphosed equivalents are now bedded tuff and argillite of the albitized sedimentary and igneous complex, exposed on southwestern Umnak Island. This event possibly took place as early as late Paleozoic time but more likely occurred sometime between the early and middle parts of the Tertiary period. Following deposition, the bedded tuff and argillite was folded along northwest axes, which are at a high angle to the present trend of the Aleutian arc. It is not known whether ex- trusion and intrusion of igneous rocks, now keratophyres and albitized dioritic rocks, occurred prior to or following the folding. Volcanism probably continued intermittently until sometime in the middle to late part of the Tertiary period, when the plutonic rocks were emplaced. The plutonic mass was largely quartz diorite varying to diorite and quartz monzonite. In the latest stages of consolida- tion of the plutonic mass, granophyre dikes were injected and portions of the dioritic mass adjacent to the granophyre were altered to a rock richer in quartz, orthoclase, and hydrous ferromagnesian minerals. A related and possibly contemporaneous alteration locally affected masses of the Tertiary lavas to produce quartz, chlorite, serpentine, zeolites, adularia(?), epidote, green amphibole, tourmaline, and rare iron sulfides. Following emplacement of the quartz diorite pluton, erosion exceeded volcanism during uplift in late Tertiary and early Quaternary time. Extensive erosion surfaces, at least in part marine, were cut on the plutonic and older rocks, although minor volcanism continued near the narrow part of Umnak. Volcanic activity possibly may have begun at the sites of Mount Recheschnoi and Okmok Volcano during the latter part of this erosional episode. : The building of the composite hypersthene andesite cone of 350 INVESTIGATIONS OF ALASKAN VOLCANOES Mount Recheschnoi probably had begun by early Quaternary time. The mafic phenocryst basalt of Kshaliuk Point was probably extruded during the same interval, though it might possibly have been ex- truded in latest Tertiary time. During the Pleistocene epoch, the central vent of Mount Recheschnoi moved westward. During Pleistocene glacial maxima, Mount Recheschnoi was deeply dissected by mountain glaciers, but cone-building activity predominated over glacial erosion. In latest Pleistocene or early Recent time a pyroclastic flow of hot wet mud and lava poured down a glacial valley 2 miles southwest of Russian Bay, and basaltic andesite was extruded at the site of Mount Vsevidof. Meanwhile on northeastern Umnak the upbuilding of Okmok Volcano had begun probably sometime during latest Tertiary or early Quaternary time. The earliest eruptions consisted largely of mafic phenocryst basalt flows of the Ashishik basalt. A few feldspathic basalt flows also were extruded in the early stages of the upbuilding of the voleano. As Okmok Volcano grew, explosive eruptions ex- truded gas-charged pyroclastic material that became palagonitized upon emplacement. As the extrusion of the basaltic rocks comprising the Ashishik basalt neared completion, feldspathic basalt flows formed the dominant extrusive product. During the latest stages of Ashishik activity and immediately following, domes and short viscous flows of vitreous andesite and subordinate rhyolite were extruded close to potential fissures along which subsidence later took place. Satellitic vents extruded viscous lavas charged with olivine and plagioclase crystals. These lavas accumulated into steep-sided piles, such as Mount Tulik. An early caldera block may have subsided just inward from the present site of the arcuate ridges shortly after the extrusion of the andesite-rhyolite association. A wedge-shaped volcanic sector graben may likewise have subsided at the gap between the north and east arcuate ridges. Whatever the origin of the gap between the arcuate ridges, a topographically low, seaward-sloping surface probably existed prior to the outpouring of the Crater Creek basalt. During the latter part of Pleistocene and early Recent time, flows of the Crater Creek basalt were extruded from vents situated some- where within what is now Okmok Caldera. At least 15 flows poured northeastward through the gap between the arcuate ridges. Toward the end of the period of extrusion of the Crater Creek basalt, a few small cones and flows of basalt were extruded on the outer flanks of Okmok Volcano. The stage was now set for the catastrophic caldera- forming eruption; GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 351 A catastrophic eruption at the summit of Okmok Volcano deposited the Okmok volcanics during early Recent time. The earlier phase of the eruption appears to have been peleean and consisted of gas- charged glowing avalanches (nuées ardentes). As the eruption in- creased in violence, red-hot glowing avalanches of molten bombs were welded together when they stopped moving to form continuous beds of welded agglomerate close to the source vents. The glowing avalanches were gas-charged and had great mobility as they moved swiftly down the slopes of Okmok Volcano. By the time they came to rest on the lower slopes, the avalanches had cooled and incorporated water from snow and ice so that they became partly palagonitized. The later phase of the eruption appears to have been violently vul- canian and consisted of showers of dark andesitic ash, bombs, lapilli, and large xenolithic blocks of earlier basalt and palagonitized rock to form a widespread blanket as much as 300 feet thick near the source vents. Very soon after the catastrophic eruption, the central part of the caldera collapsed to form Okmok Caldera inside a new ring fissure eccentric and 1 mile south with respect to the hypothetical earlier ring fissure suggested by the arcuate ridges. Immediately after collapse, the newly formed Okmok Caldera began to fill with water. Early postcaldera eruptions built bedded debris cones under water. Basalt flows of cones C and D were extruded when the water level of the lake had reached an altitude of about 1,350 feet. The highest recorded level of 1,560 feet was reached when the lake spilled over the lowest point on the hard upper surface of Crater Creek basalt near the present Crater Creek gorge. Further eruptions furnished material that was sorted into beds of silt and sandy ash by the lake. Foreset beds whose upper constructional surface was graded to the 1,560-foot lake level were formed. As the lake outlet was cut into the Crater Creek basalt, Crater Creek gorge was formed, and wave scars on the lava flows of cone D were recorded at 1,140, 1,120, and 1,100 feet. On southwestern Umnak, late or post-Wisconsin glaciers deposited the belt of older end moraines on the piedmont surface south of Mount Vsevidof and Mount Recheschnoi. The glaciers then receded to positions high in the mountains—some may have disappeared completely. At the same time the sea level rose in response to the world-wide deglaciation to a level 3 meters above present sea level, and Umnak enjoyed a slightly warmer climate during the post- Wisconsin thermal maximum than at present. Near the seacoast the older end moraines were partly buried by beach sand; older, now dead sea cliffs were cut by storm seas higher than present seas. 352 INVESTIGATIONS OF ALASKAN VOLCANOES A succession of beach ridges was deposited in favorable bights as the sea withdrew from its 3-meter stand. As the sea level continued to drop from the 3-meter level to its present stand, the glaciers on southwestern Umnak and those in Okmok Caldera advanced to the positions marked by the unweathered. terminal moraines within less than 1 mile of the present glacier fronts or cirque heads. Northwesterly winds built parabolic sand dunes behind beaches with westerly and northerly exposures. The date of the most recent glacial maximum recorded on Umnak probably coincided with that of the worldwide glacial maximum, 100 to 300 years ago. Following their maximum extent, the glaciers have receded. The greatest recession occurred in those glaciers that had previously been most successful in extending their fronts to a low altitude. While the glaciomarine events were taking place during the early part of the Recent epoch, the quartz-olivine andesite flow, the rhyolite domes, and the youngest hypersthene andesite flows of Mount Recheschnoi were extruded. Basaltic andesite flows continued to accumulate at the site of Mount Vsevidof. Toward the end of this period the cone of Mount Vsevidof attained its present height by a succession of frothy pumice and scoria eruptions. Lava flows, ranging in composition from basaltic andesite to latite poured out over the pumice and scoria on the north, south, and west sides of the volcano, probably beginning shortly after the culminating summit eruption and continuing into historic time. A chain of cinder cones was built along a rift extending down the west slope. Several lava flows were extruded from this west rift and formed Cape Kigushim- kada. The latest flow was latite from the west rift and probably was extruded during the recorded historic eruption of 1878. The most recent activity, probably on the west flank, was reported in 1957. On northeastern Umnak the most recent volcanic activity occurred within Okmok Caldera, although outside the caldera a few piles such as the two cinder cones northeast of Mount Idak and anorthite basalt flows near Inanudak Bay may have been extruded during the past few thousand years. In Okmok Caldera, cones Z and F and their flows were extruded after the caldera lake had receded from the sites of their vents. The pit crater of cone E was formed soon after the extrusion of flows from cone Z. Wall-fissure flows and those from cone B were extruded on what is essentially the present alluvial sur- face. Finally, the lava flows of cone A were extruded—probably since the first Russians sighted the island about 200 years ago. The most recent flows from cone A were extruded in 1945. GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 353 GEOLOGY OF BOGOSLOF GEOLOGY OF BOGOSLOF IN 1947 The rocks exposed on Bogoslof Island in 1947 (fig. 53) are all of historic age. Igneous rock units include: vent agglomerate of 1796 in fault contact with the andesine-hornblende andesite of Castle Rock 168°03" 168°02" Hornblende basal ee Contour interval 20 feet; 100 feet at steep cliffs Datum is mean sea level Zz of o se Basaltic dome of 1927¢ b : : ° y Na 4 53° Basaltic ash and ape \ ‘ erate of 1921 EXPLANATION oer etree 20 stle Roc! uv ae Fault showing dip; dashed where inferred U, upthrown side; D, downthrown side as Strike and dip of beds @e Horizontal beds 1 1 Base from U. S. Coast and Geodetic Survey, 1935, slightly "Geology by F.M. Byers, Jr. assisted by modified along coast to show island as of 1947 id Drewes FiGurE 53.—Geologic map of Bogoslof in 1947. 354 INVESTIGATIONS OF ALASKAN VOLCANOES extruded in 1796; hornblende basalt of 1883 on Fire Island; basaltic agglomeratic ash of 1926 and earlier explosive eruptions on Castle Rock and Fire Island; and bytownite-salite-hornblende basalt ex- truded in 1927 on Bogoslof Island. Sand and boulder beaches form most of the shoreline of Bogoslof. Two minor high angle faults are mapped; the attitudes of beds mapped in the agglomeratic ash are probably gentle initial dips. VENT AGGLOMERATE OF 1796 The vent agglomerate is exposed in the southeast end of the sea cliff that rises to Castle Rock (fig. 53). It occurs in fault contact with the andesine-hornblende andesite of Castle Rock, and is overlain by the basaltic agglomeratic ash of 1926 and earlier explosions. The vent agglomerate contains abundant angular fragments of limonite-stained lava in a grayish matrix of ash. The lava fragments are of olive-gray, dense andesite porphyry containing conspicuous phenocrysts of black hornblende, commonly 2 or 3 millimeters, but as much as a centimeter in length. As seen in thin section the rock contains phenocrysts of monoclinic pyroxene, hornblende, and sparse labradorite in a microlitic groundmass of andesine, clinopyroxene, opaque oxides, orthoclase, and a little antigorite(?). Remnants of the original hornblende are pleochroic (pale yellow to brownish green) and have a maximum extinction angle of 12°. The rims, cores, and in places entire crystals of the hornblende phenocrysts have been altered to an aggregate of magnetite, clinopyroxene, oligoclase(?), and antigorite(?). Calcite and chabazite occur in rare vesicles. The vent agglomerate is exposed well within the area originally occupied by the large dome of Old Bogoslof extruded in 1796 but has a composition slightly different from the material of Castle Rock, the remnant of Old Bogoslof. Although there is no written record of its emplacement, the agglomerate must be an explosive phase of the 1796 eruption, for no volcanic mass occupied its site prior to 1796 (see fig. 54). HORNBLENDE ANDESITE OF CASTLE ROCK The hornblende andesite of Castle Rock is an eroded remnant of the dome extruded in 1796. The twin-spired pinnacle of Castle Rock appears to have had approximately its present shape and size since about 1906 (Jaggar, 1931, p. 2-3). Chart 6433 of the U. S. Coast and Geodetic Survey, 1935, shows an altitude of 269 feet on the northwest pinnacle, and 333 feet on the southeast pinnacle. No change was observed in the shape and altitudes of the pinnacles in 1947 (pl. 51). Castle Rock is composed of a very light gray, fine-grained, holo- crystalline andesine-hornblende andesite with abundant black horn- GEOLOGICAL SURVEY BULLETIN 1028 PLATE 51 Castle Rock, Bogoslof Island, from the southeast in 1947. Agglomerate ash beds of 1926 and older eruptions overlie Castle Rock on inland slope. GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 355 blende laths ranging from a fraction of a millimeter to 3 millimeters in length. The abundant plagioclase phenocrysts about a millimeter in diameter are not conspicuous against the light-colored groundmass, Phenocrysts seen in thin section are plagioclase, oxyhornblende, ferrosalite and accessory minerals. Plagioclase phenocrysts are zoned andesine to calcic oligoclase (Ang;), but many contain a sharply defined core of anorthite-bytownite. The core boundary is generally sharp enough to be conspicuous in plane-polarized light, but is grada- tional in some crystals. A few of the andesine phenocrysts are veined with albite. Accessory microphenocrysts include magnetite (strongly magnetic grains), stubby prisms of apatite less than 0.3 millimeters in length, and sphene in both wedge-shaped and stubby prismatic forms. The groundmass is a fine-grained, xenomorphic aggregate of untwinned oligoclase-albite, orthoclase, analcite, and glass, with sparse subhedral microlites of clinopyroxene, apatite, and opaque oxides, The analcite is intergrown with other minerals in the ground- mass, but also occurs as cavity fillings. It is isotropic to weakly birefringent (0.001-0.002) with mottled extinction and in a few grains resembles the polysynthetic twinning of albite. The identification as analcite was confirmed by X-ray photography by Richard C. Erd, of the Geological Survey. Chemical and quantitative analyses of the rock are given in column 35, tables 2 and 3, respectively. The chemical analysis is similar to an earlier one of a specimen from Castle Rock (Emerson, 1904; C. N. Fenner 1926, p. 706-707). HORNBLENDE BASALT OF FIRE ISLAND As seen from a short distance in 1947, Fire Island was a flat-topped stack resembling a lighthouse, less than one-half the size mapped and photographed in 1935 (Lukens, 1936, fig. 1). Two specimens of lava collected in 1884 by Lt. G. M. Stoney of the U.S. Revenue Cutter Corwin are described by G. P. Merrill (1885, p. 31-33). The petrographic description of one of them fits the lava of Castle Rock, but the other one is a dark-gray hornblende basalt, probably from Fire Island. Hornblende makes up about 10 percent of the rock, and plagioclase microlites are abundant in an indeterminant ground- mass. This specimen was analyzed chemically by T. M. Chatard. Two specimens acquired by Diller (1885, p. 67) were from the lava of Fire Island and from the ash that fell at Unalaska village. Two determinations of silica by Chatard on the lava specimen were 51.54 and 51.65 percent; the silica content of the ash was 52.48 percent, significantly higher than the two silica determinations on the lava specimen. 489292—59—_7 356 INVESTIGATIONS OF ALASKAN VOLCANOES Chemical analysis and norm of hornblende basalt from Fire Island dome (New Bogoslof) {T. M. Chatard, analyst; in Merrill, 1885, p. 33] Analysis (weight-percent) Norm (weight-percent) - 51.54 20. 31 BASALTIC EXTRUSIVE ROCKS OF 1926-27 Basaltic ash and agglomerate of the explosions of 1926 are plastered against the steep north and east cliffs of Castle Rock (pl. 51) and form the extensive terrace that slopes away from Castle Rock to the northeast. None is on the dome-shaped western mass formed by the viscous extrusion of bytownite-salite-hornblende basalt in 1927. In an exposure at the north end of the east sea cliff the section of pyro- clastic material contains two thin ash beds separated by about 8 feet of dark-colored agglomeratic ash. The lower bed of fine ash is 3 inches thick and contains a grayish-yellow incrustation; the upper bed is 1 inch thick and contains a pale-yellowish-green incrustation, abun- dantly present as microbotryoidal coatings on the ash fragments. Gypsum, CaO-SO,-2H;0; paratacamite, CuCl,-3Cu(OH),; botryogen, 2Mg0-Fe,03-4S0;-15H,0; and caledonite, 2(Pb,Cu)O-SO;-H,0 were identified in the incrustation from the upper ash bed by Richard C. Erd of the Geological Survey. The lavas of both the dome and the larger fragments in the ag- glomeratic ash are medium-dark-gray, fine-grained basalts with con- spicuous phenocrysts of hornblende, and inconspicuous phenocrysts of plagioclase and clinopyroxene. Differences in mineralogy of the lava thrown out explosively in 1926 with the lava extruded as a large dome in 1927 can probably be correlated with the different cooling history of the two rock types. In the pyroclastic rock, phenocrysts of hornblende are unaltered, with extinction angles as much as 18°, whereas in the rock of the dome the hornblende has been partly to completely decomposed. The least altered hornblende has extinction angles as much as 14°, and is pleochroic in grayish yellow and brown. Clinopyroxene in clear, unaltered, subhedral to euhedral phenocrysts occurs in both rocks. In the rapidly chilled ejecta of the agglomerate, some of the larger crystals have grown around a core of unaltered hornblende. Both rocks contain inclusions composed of crystals of GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 357 plagioclase, clinopyroxene, and unaltered hornblende like that in the phenocrysts. The groundmass of the chilled ejecta of the agglomerate is a fine-grained, dark aggregate of microlitic plagioclase, clinopyrox- ene, and opaque oxides in a matrix of dark glass. In addition to these constituents a feldspar of low index, probably anorthoclase, is in the groundmass of the dome. Nepheline was not identified, though it appears as a theoretical mineral in the norm (column 34, table 2). BEACH DEPOSITS Beaches composed of rounded cobbles and boulders as much as 6 feet in diameter are found along three parts of the coast extending southeasterly from the rocky cliffs of Castle Rock, and both north- easterly and southeasterly from the rocky cliffs cut in the dome of 1927 lava. A sand beach forms the east coast of the ash and ag- glomerate terrace. Sand and gravel with scattered cobbles and boulders form a large deposit around the saltwater lagoon. STRUCTURE Two faults were mapped on Bogoslof Island (fig. 53). The longer is poorly exposed in the agglomeratic ash in the northeast sea cliff, but displaced beds exposed on opposite sides of the fault show that the northwest side has been dropped 5 feet relative to the southeast side. A prominent topographic scarp between altitudes of 60 and 80 feet marks the probable trace of the fault to the southwest across the raised terrace almost to Castle Rock. This scarp is a prominent feature on air photographs taken in 1934. However, Jaggar states (1930, p. 3) that explosion debris enclosing the lagoon in July 1927, reached only about 10 feet above high tide. If his description applies to the terrain of the faulted terrace, the uplift of the terrace and the formation of the fault scarp must have taken place between 1927 and 1934. The shorter southern fault separates the lava of Castle Rock from the vent agglomerate. The trace of the fault in the south sea cliff is nearly vertical. The direction of movement indicated on the map is inferred from the written record that the lava of Castle Rock is an extruded dome and presumably rose with respect to the vent agglomerate. GEOLOGIC HISTORY The first written record of the existence of Bogoslof is on the track chart of Krenitzin and Levashef, 1768-69 (Coxe, 1787, plate facing p. 205), on which a small unnamed island, later known as Ship Rock (Baker, 1906, p. 570), is shown on the approximate site of Bogoslof Island (see fig. 54.4). Captain Cook placed Ship Rock 17 miles north of the north shore of Umnak on the chart of his voyage in 1778-79 (Davidson, 1884, p. 282). A small island in the same position is also 358 INVESTIGATIONS OF ALASKAN VOLCANOES > SHIP ROCK IOANN 3 é Sess | SS/ METCALF we CONE SB FIRE ISLAND os ° | Noutical mile Se eeaeeaer eed Ficure 54.—Significant changes in Bogoslof Island between 1768 and 1947 (diagrams A-G after Jaggar, 1907 and 1931; diagram H after map in files of U.S. Coast and Geodetic Survey and after Lukens, 1936). shown on Billings’ track chart of 1802 (Baker, 1906, p. 570) and was Ly named Ship Island. Some of the earliest works to which references are made were not accessible to the author, but those included within parentheses were consulted. GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 359 In May of 1796, eruption of explosive debris and the extrusion of a viscous lava dome built an island about a quarter of a mile south of Ship Rock (Grewingk, 1850, p. 136; Dall, 1870, p. 467). The new island was named Bogoslof by the Russians at Unalaska. The mass steamed and fumed as it continued to rise in altitude. The steaming had ceased by 1804, but in many places the surface was still uncom- fortably hot. In 1806 lava flowed from the summit into the sea on the north side, and apparently the rising of the whole mass was renewed. This increase in size and altitude is said by Veniaminof (Grewingk, 1850, p. 140) to have ceased in 1823, when the dome was of pyramidal form (see fig. 54B). Estimates of the height of the new island ranged from 350 to 2,500 feet (Grewingk, 1850, p. 140), but later, more careful measurements by Dall (1884, p. 32) indicate that the lower estimate was probably closer to the true height. The lava dome later became known as Old Bogoslof (Merriam, 1902, p. 297), and the U. S. Coast and Geodetic Survey early in the present century first showed the official name, Castle Rock (Powers, 1916, p. 218), on the hydrographic charts. No further volcanic activity occurred until 1882, when it was reported that residents of Unalaska noticed steam rising from the ocean somewhat north of Ship Rock (Merriam, 1902, p. 305). On September 27, 1883, Capt. Anderson of the schooner Mathew Turner first observed a new volcanic dome rising just to the north of Ship Rock (Davidson, 1884, p. 283 ; Merriam, 1902, p.303). On October 20, 1883, a violent eruption showered ash on the village of Unalaska, 60 miles to the east. A week after the eruption the cone was seen to be still steaming by Capt. Hague of the schooner Dora. Capt. Hague suggested the name, New Bogoslof, for the new island; slightly later, Dall (1884, p. 92) proposed the name, Grewingk Island. The new island was visited several times during the following year by the U.S. Revenue Cutter Corwin, and the three Bogoslof islands were charted by Lieut. Cantwell of the Corwin in 1884 (Healy, 1889, p. 39- 41). (See fig. 54C.) The 3 Bogoslof islands were visited again by William C. Greenfield in 1887, at which time the 3 islands were con- nected as 1 by bars of volcanic debris and boulder and sand beaches. The new island to the north had a craggy profile, with pinnacles reaching to an estimated altitude of 500 feet. Old Bogoslof apparently had been greatly reduced in area and altitude, its three main crags appearing somewhat lower than the summit of New Bogoslof. Ship Rock was still in existence in 1887, but had begun to disintegrate. In 1890, Capt. Z. L. Tanner of the U. S. Fish Commission steamer Albatross reported that Ship Rock had fallen but that its original position was still marked by debris (Merriam, 1902, p. 312). In 1891, Merriam (1902, p. 313) visited the Bogoslof Islands and found 360 INVESTIGATIONS OF ALASKAN VOLCANOES steam and sulfur fumes escaping with a roaring noise from the principal fissure of New Bogoslof. An open channel separated Old and New Bogoslof. In 1895 (see fig. 54D) when Becker and Dall (Becker, 1898, p. 26; Merriam, 1902, p. 317) visited the Bogoslof Islands, New Bogoslof was still steaming ‘vigorously, though not violently” and had also been changed to a flat-topped island about 300 feet in alti- tude. Later reports in 1897 and 1899 by passing mariners indicate that New Bogoslof or Grewingk Island had finally cooled (Merriam, 1902, p. 319-320). In 1906 another extrusion of viscous lava appeared in the water about midway between old and new Bogoslof (see fig. 54£). The new island, called Metcalf Cone, was conical, showing at the summit a broken spine similar to the famous Pelee spine on Martinique (Jaggar, 1908, p. 388). An explosion early in 1907 destroyed the south half of Metcalf Cone, and another dome of lava grew into an island con- necting Metcalf Cone with Old Bogoslof. This 1907 dome was named McCulloch Peak (see fig. 547). A violent explosion on September 1, 1907, only 23 days after the Massachusetts Institute of Technology expedition had left Bogoslof, showered the village of Unalaska with a quarter inch of volcanic ash. McCulloch Peak disappeared in the explosion, and the debris mantled the surfaces of the remaining island. In 1909-10, a new conical island, Tahoma Peak, was pushed up in the bay that was formed by the destruction of McCulloch Peak (Powers, 1916, p. 218-221). A deep crater in the summit of Tahoma Peak was formed by explosions on September 18 and 19, 1910. Metcalf Cone and Tahoma Peak had a very short existence. By 1922, explosion, collapse, and marine erosion had removed all traces of them (see fig. 54@). Fire Island, a new name for New Bogoslof Island, was adopted by the Coast and Geodetic Survey on its published hydrographic charts. The island had been reduced further by erosion to asmall islet. Castle Rock (Old Bogoslof) had been reduced to two rocky horns, and a big accumulation of sand and gravel had been heaped against them, particularly on the northern and eastern sides. A wide channel of open water separated Fire Island and Castle Rock (Jaggar, 1930, p. 3). Submarine explosions between the two island remnants during the summer and fall of 1926 preceded the rise of another conical dome of lava during the winter of 1926-27. A ring of explosion debris, rising about 10 feet above high tide, surrounded the new dome and again completely connected Fire Island and Castle Rock into one elongate island. In July 1927, the new dome, about 200 feet high, was almost circular, a thousand feet in diameter, and surrounded by a shallow lagoon of warm water (Jaggar, 1930, p. 3). ‘Fire’ was still observed GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 361 at the 1927 dome as late as October 31, 1931, and quantities of pumice floated in the water south of the island (Jaggar, 1932, p. 2). In 1935, an open water channel again separated Fire Island from a larger southerly island that contained the 1927 dome and Castle Rock (see fig. 54H). Both islands were mapped by the Coast and Geodetic Survey at this time (Lukens, 1936, p. 205-206), and the name Bogoslof was officially applied to the larger island that included Castle Rock, pyroclastic rocks of the 1926 eruption, and the 1927 dome. The geology of Bogoslof Island was mapped in the summer of 1947 by the present author. Between 1935 and 1947 the size of both Fire island and Bogoslof Island was appreciably reduced by erosion (see fig. 547). A Navy quonset hut erected sometime during World War II was partly undermined by retreat of the east-facing sea cliff. On September 21, 1951, a coastwise vessel ran through “muddy water” for about 2 miles near Bogoslof; this may have been the result of a submarine eruption or landslide. Early in November, 1951, Bogoslof was again seen by G. R. MacCarthy of the Geological Survey, but no activity was observed (Volcano Letter 516, p. 8). Bogoslof was visited still later on August 13, 1953, by Adrian F. Richards (written communication, 1955), who reports further retreat southward and westward of the east sea cliff so that Bogoslof Island was reduced in area by about 25 percent. The Navy quonset hut was gone and must. have fallen over the east sea cliff as it was cut back by high waves sometime between 1947 and 1953. LITERATURE CITED Ahlmann, H. W., 1948, Glaciological research on the North Atlantic Coasts: Royal Geographic Society research series., no. 1, 83 p. Anderson, C. A., 1941, Volcanoes of the Medicine Lake Highland, California: California Univ. Dept. Geol. Sci. Bull., v. 25, p. 347-422. Arnold, J. R., and Libby, W. F., 1951, Radiocarbon dates: Science, v. 113, p. 111-120. Atwood, W. W., 1911, Geology and mineral resources of parts of the Alaska Peninsula: U. 8. Geol. Survey Bull. 467, 137 p. Baker, Marcus, 1906, Geographic dictionary of Alaska (2d. ed., prepared by James McCormick): U. S. Geol. Survey Bull. 299, 699 p. Becker, G. F., 1898, Reconnaissance of the gold fields of southern Alaska, with some notes on general geology: U. S. Geol. Survey 18th Ann. Rept., pt. 3, p. 1-86. Bradley, C. C., 1948, Geologic notes on Adak Island and the Aleutian chain, Alaska: Am. Jour. Sci., v. 246, no. 4, p. 214-240. Byers, F. M., Jr., and Brannock, W. W., 1949, Volcanic activity on Umnak and Great Sitkin Islands, 1946-1948: Trans. Am. Geophys. Union, v. 30, p. 719-734. 362 INVESTIGATIONS OF ALASKAN VOLCANOES Byers, F. M., Jr., Hopkins, D. M., Wier, K. L., and Fisher, Bernard, 1947, Volcano investigations on Umnak Island, 1946: U. 8. Geol. Survey Alaskan volcano inv. rept. no. 2, pt. 3, p. 19-53 (limited distribution). Capps, S. R., 1934, Notes on the geology of the Alaskan Peninsula and Aleutian Islands: U. S. Geol. Survey Bull. 857—D, p. 141-153. 1937, Kodiak and adjacent islands, Alaska: U. 8S. Geol. Survey Bull. 800-C, p. 111-184. Chayes, F., 1949, A simple point counter for thin-section analysis: Am. Miner- alogist, v. 34, p. 1-11. Clark, Austin H., 1945, Animal life of the Aleutian Islands: Smithsonian Inst. War Background Studies No. 21, p. 31-61. Coats, R. R., 1950, Volcanic activity in the Aleutian arc, U. 8. Geol. Survey Bull. 974-B, 47 p. 1956, Geology of northern Adak Island, Alaska: U. S. Geol. Survey Bull. 1028-C, p. 45-67. Coxe, William, 1787, Account of the Russian discoveries between Asia and America: London, T. Cadell, 3d ed., 454 p. Dall, W. H., 1870, Alaska and its resources: London, Sampson Low, Son, and Marston, 627 p. 1884, A new volcano island in Alaska: Science, v. 3, p. 89-93. 1885, Further notes on Bogoslof Island: Science, v. 5, p. 32-33. Davidson, George, 1884, The new Bogoslof Volcano in Bering Sea: Science, v. 3, p. 282-286. Diller, J. S., 1884, Volcanic sand that fell on Unalaska, Oct. 20, 1883, and some considerations concerning its composition: Science, v. 3, p. 651-654. 1885, Lava from the new volcano on Bogoslof Island: Science, v. 5, p. 66— 67. 1902, The geology of Crater Lake National Park: U.S. Geol. Survey Prof. Paper 3, Pt. 1, p. 11-61. Dunn, Robert, 1908, On the chase for volcanoes: Outing Magazine, v. 51, p. 540-550. Emerson, B. K., 1904, General geology; notes on the stratigraphy and igneous rocks [of Alaska], Harriman Alaska Expedition, v. 4, Smithsonian Inst., p. 11-56. Fairbridge, R. W., 1948, The geology and geomorphology of Point Peron, Western Australia: Jour. Royal Soc. of Western Australia, v. 34, p. 35-72. 1950, Recent advances in eustatie research: Report of committee for investigation and correlation of eustatic changes in sea level, Australia-New Zealand, Assoc. Adv. Sci., p. 181-184. Fenner, C. N., 1926, Katmai magmatic province: Jour. Geology, v. 34, p. 673-772. Flint, R. F., and Deevey, E. S., Jr., 1951, Radiocarbon dating of late Pleistocene events: Am. Jour. Sci., v. 249, p. 257-300. Freiday, Dean, 1945, The Aleutians, Island necklace of the North; Natural His- tory, v. 54, p. 444-455. Grewingk, Constantin, 1850, Beitrage zur Kenntniss der orographischen und geognostischen Beschaffenheit der nord-west Kiiste Amerikas, mit den anliegenden Inseln: Russ. K. min. Gesell. Verh., St. Petersburg, 1848-1849, 351 p. Hall, A. J., 1941a, The relation between color and chemical composition in the biotites: Am. Mineralogist, v. 26, p. 29-33. 1941b, The relation between chemical composition and refractive index in the biotites: Am. Mineralogist, v. 26, p. 34-41. GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 363 Healy, Capt. M. A., 1889, Report of the cruise of the Revenue Marine steamer Corwin in the Arctic Ocean in the year 1884: Washington Government Printing Office, 128 p. Hess, H. H., 1948, Major structural features on the Western North Pacific, an interpretation of H. O. 5485, Bathymetric Chart, Korea to New Guinea: Geol. Soc. America Bull., v. 59, p. 417-446. 1949, Chemical composition and optical properties of common clino- pyroxenes, part I: Am. Mineralogist, v. 34, p. 621-666. Hrdligka, Ales, 1945, The Aleutian and the Commander Islands and their inhab- itants: Wister Institute of Anatomy and Biology, Philadelphia, 630 p. Hultén, Eric, 1937, The flora of the Aleutian Islands, Stockholm, Bokforlags Aktiebolaget Thule, 397 p. Ives, R. L., 1951, High sea levels of the Sonoran Shore: Am, Jour. Sci., v. 249, p. 215-223. Jaggar, T. A., Jr., 1908, The evolution of Bogoslof Volcano: Bull. Am. Geog. Soc., v. 40, no. 7, p. 385-400. 1930, Recent Activity of Bogoslof Voleano: The Voleano Letter, no. 275, p. 1-3. 1932, Aleutian eruptions: 1930-1932: Volcano Letter, no. 375, Hawaiian Voleano Observatory, p. 1-3. Jochelson, Waldemar, 1925, Archaeological investigations in the Aleutian Islands: Carnegie Inst. Washington Pub. 367, 131 p. 1933, History, ethnology, and anthropology of the Aleutians: Carnegie Inst. Washington Pub. 432, 91 p. Kennedy, G. C., 1947, Charts for correlation of optical properties with chemical composition of some common rock-forming minerals: Am. Mineralogist, v. 32, p. 561-573. Kennedy, G. C., and Waldron, Howard H., 1955, Geology of Pavlof Volcano and vicinity, Alaska: U. 8. Geol. Survey Bull. 1028-A, p. 1-19. Knappen, R. S., 1929, Geology and mineral resources of the Aniakchak district. Alaska: U. S. Geol. Survey Bull. 797, p. 161-223. Kuno, Hisashi, 1950, Petrology of Hakone Voleano and the adjacent areas, Japan: Geol. Soc. America Bull., v. 61, p. 957-1020. Lantis, Margaret, 1950, The Reindeer Industry in Alaska: Arctic, v. 3, p. 27-44. Lawrence, D. B., 1950, Glacial fluctuation for 6 centuries in southeastern Alaska and its relation to solar activity, Geog. Review, v. 40, p. 191-223. Lukens, R. R., 1936, Bogoslof Volcano: The Military Engineer, v. 28, no. 159, p. 205-206. MacNeil, F. 8., 1950, Planation of recent reef flats on Okinawa: Geol. Soc. America Bull. 61, p. 1307-1308. Mathews, W. H., 1947, Tuyas, flat-topped volcanoes in northern British Columbia: Am. Jour. Sci., v. 245, p. 560-570. 1951, Historic and prehistoric fluctuations of Alpine Glaciers in the Mount Garibaldi Map-area, southwestern British Columbia: Jour. Geology, v. 59, p. 357-380. Matthes, F. E., 1941, Rebirth of the glaciers of the Sierra Nevada during late post- Pleistocene time [abstract]: Geol. Soc. America Bull., v. 52, p. 2030. Menard, H. W. and Deitz, R. S., 1951, Submarine geology of the Gulf of Alaska: Geol. Soc. America Bull., v. 62, p. 1263-1286. Merriam, C. H., 1902, Bogoslof, our newest volcano, Harriman Alaska Expedition, v. 2 (New York), p. 291-336. Merrill, G. P., 1885, On hornblende andesites from the new volcano on Bogoslof Island in Bering Sea: Proc. U. S. Natl. Museum, v. 8, (1886), p. 31-33. 364 INVESTIGATIONS OF ALASKAN VOLCANOES Murray, H. W., 1945, Profiles of the Aleutian Trench: Geol. Soc. America Bull., v. 56, p. 757-781. Newell, N. D., Rigby, J. K., Whiteman, A. J., and Bradley, J. S., 1951, Shoal- water geology and environments, Eastern Andros Island, Bahamas: Am. Mus. Nat. History Bull., v. 97, Art. 1, 29 p. Powers, Sidney, 1916, Recent changes in Bogoslof Volcano: Geog. Review, v. 2, (July—Dee.), p. 218-221. Richey, J. E., 1935, British regional geology: Scotland: The Tertiary volcanic districts: Dept. Sci. and Industrial Research, Geol. Survey and Museum; His Majesty’s Stationery office, 115 p. 1937, Some features of Tertiary voleanicity in Scotland and Ireland: Bull. Volcanologique, Serie II, Tome I, p. 11-34. Shapiro, Leonard, and Brannock, W. W., 1952, Rapid analysis of silicate rocks: U.S. Geol. Survey Circular 165, 19 p. Sharp, R. P., 1951, Glacial History of Wolf Creek, St. Elias Range, Canada: Jour. Geology, v. 59, p. 97-117. Smith, P. A., 1937, The submarine topography of Bogoslof: Geog. Review, v. 27, p. 630-636. Smith, P. S., 1939, Areal Geology of Alaska: U. 8. Geol. Survey Prof. Paper192, 100 p. Smith, W. R., 1925, Aniakchak Crater, Alaska Peninsula: U. 8. Geol. Survey Prof. Paper 132, p. 139-145. Stearns, H. T., and Vaksvik, K. N., 1935, Geology and ground water resources of the island of Oahu, Hawaii: Territory Hawaii Div. Hydrography Bull. 1, 479 p. Studds, R. F. A., 1950, Oceanographic activities of the U. 8. Coast and Geodetic Survey: Trans. Am. Geophys. Union, v. 31, p. 786-788. Tebenkoff, Capt., M. D., 1852, Atlas of the northwestern coast of America from Bering Strait to Cape Korrientes and Aleutian Islands with addition of certain places of the northwestern coast of Asia, 38 maps. Thorarinsson, Sigurdur, 1939, Vatnajokull; Scientific results of the Swedish— Icelandic investigations 1936-37-38; Chapter 9, The ice dammed lakes of Iceland with particular reference to their values as indicators of glacier oscillations: Geog. Annal., Stockholm, v. 21, p. 216-242. Tsuboi, Seitaro, 1925, A dispersion method of determining plagioclases in cleavage flakes: Mineralog. Mag., v. 20, p. 108-122. Umbgrove, J. H. F., 1947, The pulse of the Earth, The Hague, Martinus Nijhoff, 358 p. U.S. Coast and Geodetic Survey, 1947, United States Coast Pilot, Alaska: pt. 2, Yakutat Bay to Arctic Ocean. 5th edition [1948], 659 p. Upson, J. E., 1951, Former marine shore lines of the Gaviota quadrangle, Santa Barbara County, Calif.: Jour. Geology, v. 59, p. 415-446. Veniaminof, I., 1840, Notes on the islands of the Unalaska District, [in Russian], St. Petersburg, Book 1, 364 p., Book 2, 409 p., Book 3, 134 p. Walker, E. H., 1945, Plants of the Aleutian Islands: Smithsonian Inst. War Background Studies, no. 21, p. 63-71. Wentworth, C. K., 1926, Pyroclastic geology of Oahu: Bernice P. Bishop Museum Bull. 30, 121 p. Wilcox, R. E., 1945, 1946, Report on activity of voleano, Umnak Island, Alaska, June and July 1945, 1 August, 1945 through 4 January, 1946: U. 8. Geol. Survey files. GEOLOGY OF UMNAK AND BOGOSLOF ISLANDS, ALEUTIAN ISLANDS 365 Williams, Howel, 1941, Calderas and their origin: California Univ., Pub., Dept. Geol. Sci. Bull., v. 25, p. 239-346. 1942, The geology of Crater Lake Nationa] Park, Oregon: Carnegie Inst. Washington Pub. 540, 162 p. 1950, Voleanoes of the Parfcutin region, Mexico: U. S. Geol. Survey Bull. 965-B, p. 165-279. Winchell, A. N., 1933, Elements of optical mineralogy: Part 2, Descriptions of minerals, New York, John Wiley and Sons, Inc., 459 p. Actinolite. Adularia__ Age and correlation, albitized sedimentary and igneous complex... of rocks of central Umnak_ Agglomerate beds, welded. Agglomerate deposits____ Agglomeratic tuff-breccia, partly palagonitized. 320 Akun Island___-______. Akutan Island__. Alaska Peninsula. Andesite flows, basaltic, from cone B. quartz-olivine-___ Andesite and latite flows_ Arcuate ridges, hypotheses for origin... Area of report. Argillite..... Ash, bedded. Ashishik basalt_ Augite:......... 290, 291, 298, 300, 301, 303, 306, 307, 310, 318, 337 329, 330, 331, 332, 334, 335, 336, Basalt, anorthite... aphyric and feldspathi Ashishik___ olivine-rich___. Plagioclase-olivine, of satellitic vents_.__._ 312- Plagioclase-augite-olivine. Basalt flows, from cone A_ Jater than caldera Lake. near Cape Idak... on arcuate ridges. on Cape Aslik______... outside Okmok caldera. Plagioclase of cones C and D_ ~ 287, 295, 302, 303, 306, 307, 355 297-302, 312, 335, 354 324, 330, 332, 333, 334 Anorthite basalt flow, near Inanudak Bay. 287, 288, 290, 291, 292, 355 308, 309-311, 312, 313 308, 309-311, 312, 313 306-307, 335 ---- 275-276 287, ~ 332-333 311-312, 334 313 Basaltic andesite flows_ Beach deposits. Beach ridges.. Bering Sea_ Bering, Vitus, expedition of. Biotite.__ 270 287, 288, 290, 291, 301, 302 Black Creel 274 Botryoget 356 Bytownit 298, 300, 302, 303, 330, 334, 335, 355 Calcite... 286, 287, 288, 292, 295, 296, 354 Caledonite_ 356 273 324 320, 321 305, 345 Castle Rock. 271, 354-355, 357 Chabazite-_ 295, 337, 354 Chirikof, Captain Alexei, cited. - 270 Chlorite____. 287, 288, 291, 292, 294, 295 Chuginadak. 273 Cinder Point. 324 Climate. 275 Coal..... 289 Colluviam 346 Cornwallius sp... - 289 Corwin, U.S, Revenue Cutter_ 355, 359 Crater Creek basalt, use of name. --- 308 Crater Creek gorge. ~ 312, 338, 351 Crater Creek valley. - 347-348 Contact rock, altered. - 292-293 Diabase_ Geologic work, previous. Geography. 272-276 Geyser Bight. 294, 346 Geysers_ -- 294 Glaciers_ 346-347, 352 Granophyre. -- 289, 292 Graywacke 289 Gypsum... 356 Harriman, Alaska expedition of. 271 Hematite... 288 Heulandite. 295 367 368 Page History, geologic- of report area. Hornblende. - --. 290, 291, 292, 301, 354, 355, 356, 357 Hornblende andesite, of Castle Rock. 354-355 Hornblende basalt, of Fire Island_ Hot Springs----. 272, 204 Hot Springs Cove. Hyalosiderite.. Hyperthene... 298,300, 301, 303, 306, 307, 318, 329 Hypersthene andesite--... 297-298, 300-301 285-289 -- 273 - 268-269 Namen le aera ea acer tees eee 308, 313, Jochelson, Waldemar, anthropologist cited... 271 Keratophyre- ---= 285, 288-289 Labradorite...-...---------------. 287, 300,301, 302, 303, 330, 333, 335, 336, 354 a ---- 306,335 ~ 298, 299, 304-306, 323-329 Lavas, age of, Mount Recheschnoi.. Literature cited. McCulloch Peak. Mafic phenocryst basalt, of Kshaliuk Point. Mt. Idak_.--.- Mt. Recheschnoi. Mt. Tulik-... Mt. Vsevidot Nikolski, Aleut village of Nikolski plain... Nomenclature, of rocks, used in report. stratigraphic__.. Nuée ardente.......------------------------ Obsidian--.. Okmok Caldera. Okmok volcanics, age- ~ 322-323 origin and emplacement. - 321-322, 351 Okmok Volcano, petrography_ - 329-330 structure. - 337 Oligoclase. 291, 292, 293, 354, 355 Olivine. 295, 298, 300, 301, 302, 303, 306, 318, 320, 329, 330, 331, 332, 334, 335, 336, 337 Orthoclase. 290, 291, 292, 303, 306, 307, 308, 354, 355 Outwash_ 344 Oxyhornblende. 355 Paratacamite.. 356 Partov Cove_ Penninite.-.- INDEX Page Petrography...-..------------------ sae aaea 286-289, 290-293, 295-296, 299-304, 305-308, 329-330 Phenocrysts--.. .------ 297, 298, 302, 303, 305, 306, 310, 311, 319, 330, 331, 332, 333, 334, 354, 355, 356 ~ 287,290, 291, 300, 302, 303, 306, 311, 316, 318, 320, 323, 326, 329, 330, 331, 333, 334, 335, 336, 355, 356, 357 - 287, ~~ 287, 288, 290, 291, 292, 295, 296, 302, 303 289, 290-291 Quartz monzonite. 289, 291-292 Reindeer. 276 Rhyolite_ |-302, 312, 336 Rhyolite domes. Ripidolite. Roads, on Umnak 7 Rocks, contemporaneous with caldera lake. extrusive basaltic. of Central Umnak, chemical composition-. 295- 296 of postcaldera age, within Okmok caldera... 324- 325 palagonitized pyroclastic_ 310-811, 336-337 Precaldera, age of.. 314 Russian Bay- Russian period, in history of report area. 304 Structure_.... » Surficial deposits. -- 343-347 Tahoma Peak. 360 ‘Talus._.----- - 345, 346 ‘Temperature. 275 Thumb Point. ~ 293, 295 Till__.----. - 348, 344 296 Taft.. --- 287,289 Tuya. 326 Umnak Pass-- --- 270, 273 Unalaska Island. - 278, 290, 345 INDEX 369 Page Vegetation.--.... WWeeenen=---+---2----------- 275-276 | Vsevidof Island.. Volcanic rocks, of central Umnak. of minor vents_-_-.. of Mount Recheschno! of Mount Vsevidof. of Okmok volcano... Volcanics, Okmok, of caldera age. 314-315 — 7 we. UNITED STATES DEPARTMENT OF THE INTERIOR : GEOLOGICAL SURVEY : BULLETIN 1028 PLATE 40 169°00' 168°30" Cape IImalianuk (SS Broken Point INANUDAK BAY S$ Cemetery Cove Hot Springs 3 Cove G 53°15! é Kshaliuk Point SURFICIAL DEPOSITS & ) ae» aa & 8 $4 Younger till Glacier ice Glacial outwash ‘Allavinmn Talus and other Beach deposits Dune sand | > s gravel colluvium Qbs, sand aad gravel. fe Qob, boulder beach = 2 (summer 1947) & wi & : 5 = 390 S o = QTvp & Older till Thumb Point & Bad moraines shown x by hachures J IGNEOUS AND METAMORPHIC ROCKS Tyinlaya Polot ( One . < 7 & S , z arp Rhyolite domes Quartz-bearing olivine 3 &® 2 andesite flow g 2 os Cinder cone deposits g Youngest hypersthene oe : = ee undifferentiated e andesite flows z Derby Point 3 See oct i Gh mc reed . s Qr, lava flows. Qro, = j 3 Voleanic rocks of Mount Vsevidof = pyroclastic flows = Qvi, youngest flows, including s ci 3 hisiorie(2) latite flows. es Re S Qup, anileuils and rhyodacite 3 s : pyroclastic beds and vent g Compl: (suclion And 3 Qmb a a tie andesite wr . s Mee ties vad E Hypersthene andesite Mafie phenoenyst basalt of 3 S | ara, pypersthene-bearing labradorite aE =z > |" andesite and hypersthene-bearing Qmb, anorthite-olivine-augile ay aphyric andesite flows and minor basalt flows, Qmv, vent | pyroclastic interbeds. Qrv, pyro- complex | clastic beds and vent complex I l | UNCONFORMITY ! | | 7139304 | 1 Tdp Plutonie rocks i d = Divrite, quartz diorite, quartz Voleanie rocks of central Umnak = ny baw & mmpieniine, Une pranoplagre QTy, bedded voleanie racks fe Kigushimkada 2 ‘othermally altered in places. < ws OTvp, vent and intrusive co Ee e Tvs ltered and unalte ee 2 Albitized sedimentary and igneous complex Bediled urgillite and tuff, Keratophyre flows, and. albitized intrusive rocks eee ene eee DEGLINATION 1958 Contact Lony dashed where approimutely located, short dashed where indefinite & Probable fault Dotted where concealed 6 Anticline Showing trace of axial plane. Dashed where appronimately 9 located Showing trace of axial plane Dashed where approximately located 435 Strike und dip of layered rocks +90 Strike of vertical layered rocks Position of 90 indicates up side of beds e Horizontal layered rocks a Fumarole (Summer, 1947) on Thermal spring (Summer, 1947) 0 — Ananiuliak Island ——{53°00' A Mt Vsevidot 6000' 4 A 4000’ 4 & 8 L Kelp Point C eee & 20004 = 3 Qyba é Sea level + 2000 a wg Black Cape Zo ao _~ Loran Station _, (1947) "2 B Mt Recheshnoi 6000! 4 Qrv B y | 4000} [ a 2000'4 = : a Cape Udak Sea louat tees Tvs 2000" ‘ Idaliuk Point GEOLOGIC MAP AND SECTIONS OF SOUTHWESTERN UMNAK ISLAND, ALEUTIAN ISLANDS, ALASKA Scale 1:96 000 Go oe al 4 ee i 2 3 4 __) Miles LE Cae | Contour interval 200 feet QD Breadloat Datum is mean sea level af Island vs Cape Sagak Base by U. S. Geclogical Survey from official sources 169°00! 16°30" INTERIOR GEOLOGICAL SURVEY, WASHINGTO = ; Be MR 5508 = Geology by F. M. Byers, Jr., 1947 ERG EaTaT PTT TT TENT ee TY HELE REET TE UNITED STATES DEPARTMENT OF THE INTERIOR GEOLOGICAL SURVEY 16/° 30! oo (ZEgY Gof! Wie fae [ere y Los eat heen oe . Wt r \\SS ce - AC ve es Nea plea | ae ~ Priggt Rho EN. os a erKalete, \ Gn Staraya Bay *\ ane | } FO inence, Gs : et : ae as ee 166700" BULLETIN 1028 PLATE 75 es Subaerial topography by U.S. Coast and Geodetic Survey and U.S. Corps of Engineers, 1940-44; torm lines of Shaler Mountains and Makushin Volcano by Harald Drewes Submarine topography compiled from U.S. Coast and Geodetic Survey smooth sheets of 1934-39 by Harald Drewes GEOLOGIC MAP OF 5 0 5 10 15 167°00! UNALASKA ISLAND, ALASKA, AND ADJACENT SUBMARINE AREAS SCALE 1:250 000 20 MILES "mean a = 1 CONTOUR INTERVALS 200 AND 1000 FEET FORM LINES OR AREAS NOT SURVEYED IN DETAIL ARE INDICATED BY BROKEN LINES DATUM IS MEAN SEA LEVEL DEPTH CURVES IN FEET—DATUM IS MEAN LOWER LOW WATER BROKEN LINES ARF A HALF-INTERVAL AND DOTTED LINES ARE A QUARTER- INTERVAL 166° 00° INTERIOR. GEOLOGICAL SURVEY. WASHINGTON, D. ¢, 10163 53° Geology by H. F. Barnett, Jr., Harald Drewes, G. D. Fraser, E. H. Meitzner, and G. L. Snyder, 1953-54 Reeent Pleistocene and Recent Pliocene(?) and. Pleistocene Miocene?) Miocene wnd lower A Y pists EXPLANATION EES Surficial deposits Alluvial, beach, and colian deposits, Qa; till of terminal and ground moraines, Qt; glacier ice, Qi UNCONFORMITY Eider Point basalt Includes olivine, two pyroxene hasalt and minor rhyodacite porphyries. Dominantly pyroclastic rocks, Qep; dominantly lava, Qel UNCONFORMITY Makushin voleanies Basalt and andesite lana, pyroclastic rocks, and minor sedimentary rocks, Q\im; basalt plugs, sills, and dikes, QTi UNCONFORMITY Granodiorite batholith Includes contrasting border facies and as- sociated plutons. Granodiorite, Tg; diorite and minor granilic border facies and small plutons, Tgd; gabbro border facies and small plutons, Tgg UNCONFORMITY Unalaska formation Slightly altered andesite and basalt extrusive rocks, sills, and sedimentary rocks ranging in coarseness from argillite to conglomerute, Tu; andesite and basalt dikes, including dikes of undetermined age, Tui Lower Miocene and lower Contact, approximately located Inferred or gradational contact 7h Se oo Fault, approximately located, showing dip D, downthrown side; U, upthrown side Concealed fault Seaward margin of glaciated topography, probably moraines Shear zone ent Strike and dip of beds a Horizontal beds 2 Strike and dip of foliation aot Strike of vertical foliation Voleanic vent 3 Fumarole i? Collection site of chemically analyzed specimen = Glacial striae, showing probable direction of glacial movement TRUE NORTH MONET OS APPROXIMATE MEAN DECLINATION, 1960 QUATERNARY QUATERNARY TERTIARY AND TERTIARY (?) Jana] Das une 3. wn 480z OZ [VALOTUL noPUOG SOLA g ¥ & é T 0 iL 098 8971 e[eog ‘ € {vot pve 961 ‘meas previ8g pus oe VUSVIV ‘SANV'ISI NVLLOATV ‘GNVISI MVNO NUFLISVAHLYON JO SNOLLOWS UNV dVW OLN010R9 tone a. en bd eee SUIydOH "W PIAR “up ‘Siakg “W'4 Aq ABojcog D“@ 'NCLONIHSYM “AZAHNS T¥21901035—wolwaLNI sejBuespenb sreaugu3 Away 'S ‘| Wod pejlcuico oseg 0002 “0002 O01 | [roo 7 | AY 8 SES ae = sare yeseq y1ysiysy : SBAG] MOIlld age erapjeoeid Jo sy90) peqeiooeig, th \ Dee jaye FAO OdIOVd ras onraag f ° 8 3 5 42249 Linossapy fl na Nl J 0d vuagaityvos y OW -O espa g F i ayenoay YON 4IINL JUNOW GT 891 f aBe erapjeoeud Jo syoos payeisoa4g yeseq yiysiysy yeseq aaa aecenesniaie ao x f+/9A9/ bas yeseq 08 VaS ONTYRE weed Ser Woe a Od we x oO KM OC ¥ ed 3 5 adeg 9jHey e qUjod 9 A ajdeals | iy I 2 | . ep ©, es D ey | oC Fim neat ed nog s - = > one. eo o sa, dels RIORLS Jo uorpaatcy = < . oC plat z ‘i qulog uss T 5 / ABPUulO) y aa x 1 2 syooa posods] [eqUOZIO, LD vee @ 8 100 089. | 2 0 ea ; / te ay 890. - ; di Bo) 0 syaoa paaede] Jo dip pue ayiayg ‘ o . s = i “ Zz oe <q % : aang) eS <0 aynez jeoneujod.sq a0 pepeaou0; x \ ~ wbyeT | | apis unowyqumop ‘Q Sep umowyidn ‘1 fpaynaanos ‘avaym payop 5 | “paynou] Ayoyounxoaddn aay paysn(y : yeunsiy yoy o(Buv-y sy oe oe 2 ae adeg a 1eO ae See rere = f \ : ; : 10% | passafuy pun jruoyopvey : eas Teh ees 1 youquod onUyepuy fs : 2 : poynon; yoqvuicoaddy asoyn paysoy qauyu0g, A ee ald, led (a oe | x Ne sav m | V 4 | : 3 | aed. . rs < n \ 4 = eo = | | | 3% \ ( i} = | ALINAOINOINA | Ss — | : | sontengyn 20809 onpndsppaf pun aahydy 21d = | 1 | smoyf yyosng aypodspjay pun nckydy ‘qeLO syooa oysepoudd apuseq, oP LTE | yeseq orgudspjas pues o1dgdy peznmoseyeg qsdaoouayd oyeyw = | BE | GICLEE — eg CI | | = 8 | ou a Fe a yoo dius J youu avunajoa 40 wapduos qua, ‘Ach ea Sg ay smoy 27p8nq su2a20-asDp0%DIG ‘OFd Fos! eS a hi . SqUDA dTTITOTRS Jo TTesRq auLATO-osRPDOLAELY E3 3B & \ ee Ee : - Sal ae £3 e 3 6a q | qeseq Yaeay 109819, i ; Ea | sabpyt aymmoun g | yepa pun yptou ue syDo2 qysppouhd apjo S fyqssod sapnysus osyy ‘yore uf OL uny) $897 8 Bdayn UnoYys POAT ‘oUI}0A YowyQ fo nodded? a Buguncof-ocopyna fg poysodap sxo0u a8D)90.4h g be soluBaloA HOUAD z 2 4 sypo. axsmpooshd. vsapynoqsod. Apwog ‘dat Sq puv Q euoa mous smoyf yposn9 asnpoouing7 ‘pot ‘syuaunpas mwox0a pappoy ‘St sbngd any ‘xD oO OWT Blep/Bg ULM snosuBsoduIa}UOD SyD0y snsodap yusy adeg Cc auoo wopury ‘3d si = peynuezagipun Rape YOUHO ! m & BAR] [UO A, apisjno SMoy pEseg ES ww (FO a 2 8 Aap BO) 5 > < é ‘Aa 2 2 > i : Hf M09 Wout, & re a. sunt qosng 4) Sy 9u0d moss snort E = es = aN | eee : es yep . auoo most snout apsapun ayposog ‘08 é ‘96 7X S Hy owoo mous smoyf 770809 er6T-2.4 420 Gs cee Tht One Y SE {O76 UW VY au0d Wosf smoyf yospg ‘sjeO ad a x ‘ i AS OXVT] BAOP[RO UBYY 4oIB] BMOY OI[MBVET i a J F B COLL. : & | 2 E | 5 a = qd 1 ~ z A WOE oO SNDOU DIHdYOWVLAW GNV SNOANDI < = ee cS UINLAN [Oa Jay}0 S ofS zy pues aunq s}sodap yorag amNY pue snyry, ao] FATAR[y) m poe < . a —3E-- SS : aH -) zs E tere < SLISOdad TWIDLdNNs AAG y 2 == e == - => t J : S 2 A Ey : er Z / fe = f eS — NOILYNV1dX4 gel, = ke qwulog YAN 1 ON : PISIY BITAIVS 3 & a = a LV v quiog J2apulay 4 g o v 14 gh yIUSIUSY yeue, aceg : - (00681 1ST oT AWAYNS TVOTSOTORD 1’ ALVId 8201 NICATION MOINMMLNI AHL AO LNAWLY ddd SALYIS OTWTINA UNITED STATES DEPARTMENT OF THE INTERIOR GEOLOGICAL SURVEY BULLETIN 1028 PLATE 48 489292 © -59 (In pocket) EXPLANATION NORTH SOUTH E in ' 3000' - ee 3000 leo Qcoa Av ~a Alluvium Cinder cone deposits Cone C ed Li ==, Leh. ae Plagioclase basalt flows Eas ae of cones C and D weer at ; " ee ae eon Le ee oh he Qep 2000'- Nes AG oA int ote era Vent plug ; t- 2000 Bes ear ore ee ape Vou Na tay Early postcaldera pyroclastic Oal ee a 2 See Pe precaldera rocks : ee a ee ett Osa ees” Peet ea Sia MOC s ae. oes an Peo ee ae a cee ° Jae ae ve ea °. a BAS rae oe 3 2 E 4. ome o > aie aot a ee ae wos ae Le 1560-foot level (later stage) anh po Os. ge a a Pree a Fee age tan a. van Cae os ee ge < een as Bo Sea na Pe Ne a.) ee, aga ieee a < » = UU NU TU Hi : ——1350-faot level (early stage) I ON Sy MIMS MTT UR TTT Qep and precaldera rocks ee NS aN UE wt \ Qs CT ODI aNd, igs ED 55 SAN SON TOSS A ROSES Bees SS es GB Pees BIN GLE): TIE SSS Subaqueous pillow lava x * jes ~ <5 eos ne sd Ta Sei, a A Nos Ss Ry ISSRESSS SSS EGS L IA ILI NE ANIL os WSSROSS SSSA 1000! + =| Qep-and precaldera rocks Qep and precaldera rocks =| 1000 o 1900 2000 3000 4000 ue Feet ae ele | Es eae Sea lever 2 5 ~ - —— ~ Sea /eve/ SECTION THROUGH CONES C AND D SHOWING DIFFERENT LAKE LEVELS AND RELATIONS BETWEEN ROCKS CONTEMPORANEOUS WITH CALDERA LAKE,UMNAK ISLAND, ALEUTIAN ISLANDS , ALASKA Geological Survey TaBLE 3.—Quantitative petrographic summary of chemically analyzed rocks from Umnak and Bogoslof Islands [Numbers of columns correspond to those in table 2] Northeastern Umnak | 1 1A 2 3 4 5 \ 6 7 | 8 9 10 a | 12 | 13 14 15 16 17 18 19 20 Ieee ey peo a oe { 4 { yo eras He aa sales Mode (volume percent): ! ' PRCOOUGhIO PrAp ene Gel cue Wop aii ten HON men Ta Ga.) een Sane coe Sl ate [ee a Re Oe ee re A ae eS ES pia AO Gs Re POEM ae De ta amet an i Orb dean ec! HS tla BRL IIMs eae ure Bee ee Crystalszor phenocrysts >0.2 mm ! 1 Ee 58, 5 1.8 . 2 I . 6 Augite__ eis Nuts ee eles cen Il pal <6 ReneS a Tot 1. <oel Ao ist See 05 .O1 ad aque oxides. . Erode goa aad al aba oie as SY Ne Dra mand | C1 ee ot a oi eas or Perera plea 69.1 80.6 | #05. | 4. 6 97.8 99.8 589 95.9 | 98.9 99. 6 | 795.1 Somposition of less calcie plagioclase crys 5 66 | = Composition of more cal ie plagionlae crystal 7 re W Anoo~9s Atlga—93 Anzs-9s Anus—s0 Ansi-p3 ANngs-92 Anjs-s7 Anev—ss ae Ango~v4 { pies i pun pee Estimated average composition of less calcic plagioclase crystals_ An. n An, Estimated average composition of more calcic plagioclase erystals_ * Ang Ann Ans. Ana Anis Anss Ans Anus { Te Ann ae | i | Approximate composition of groundmass plagioclase (+ Ans) __ Average beta index of augite crystals 11_ Average 2V of augite crystals 4__ Composition of olivine crystals ¥_ Estimated average composition of olivine crystals_ Refractive index of groundmass or intersertal glass__ 1. 697 48°--54° ee | 7 Fos9—71 | Foro-75 Foz, |. aed Fogs | Foz Sel ee oe | BU eet Se 71, 560 | Mode (volume percent): Crystals or phenocrysts >0,2 mm: QUARGE Cal aa ‘ Potassium feldspar Less calcic plagioclase More ealcic plagioclas Biotite (and minor chloritic Amphibole (includes decomposed amphibole) _ Clinopyroxene (epidote in column 33)_ Orthopyroxene: hypersthene_ Olivine Apatite (includes 0.2 sphenc in column 35)_ Opaque oxides__.-.____.____ Groundmass or interserlal material_ Composition of less calcic plagioclase cry: i Composition of more calcic plagioclase crystals °_ Estimated average composition of less calcic plagioclase crysta! ‘ Estimated average composition of more calcie plagioclase erystals..__ Approximate composition of groundmass plagioclase (-: Ans)_. Gamma index of biotite crystals 4___ Beta index of amphibole crystals 1 i Beta index of ferriferous * clinopyroxene crystals 4_ Beta index of calciferous * clinopyroxene crystals 1_ 2V of ferriferous clinopyroxene crystals 4_ 2V of calciferous clinopyroxene crystals 4_ Composition of hypersthene crystals 25____ Average composition of hypersthene crystal Composition of olivine crystals ¥__.___-___ Average composition of olivine crystals ss Refractive index of groundmass or intersertal glass __ : Determined by point counter described by Chayes (1949). Southwestern Umnak * Minimum size of erystals about 0.1 mm; modes of intrusive rocks include somewhat smaller crystals. 3 Visual estimate. Rare hypersthene also present. “Ghosts” of olivine, now antigorite, ponninite, quartz, and magnetite. 5 Includes palagonitized glass, zeolites, ete. 5 Inchides 0.6 percent of calcite and 2 percent of chlorite, antigorite, quartz, and zcolites. 7 Includes 0.1 percent of zeolite-filled vesicles. 5 Includes felsic groundmass with microveinlets of quartz, zeolite, and magnetite. : (trace?) 50° 47° 1947, fig. 1) and by Tsuboi (1925, pl, 1). Most determinations within --Ang, limit, Ans. i Based on optical determination of refractive indices; partial analysis for alkalis (Na-K, 1.14 pers cent) indicates composition of Ang). ™ Refractive indices determined with sodium light at 25° C..12° C. Most measurements within 0.003, excepting those on harnblende, 18 Data on groundmass augite. +) Diopsidic augite (6=1.698) also present; probably xenocrysts. ‘42V measurements on grains in thin section, using 4-axis universal stage; average of several read- ings, except for No. 9, in which full range of 2V variation of augite is shown, ‘ions from reiractive index data, using curves by Calkins-Hess (Kennedy, (absent) 3 Ang Anas Angg 1947, fig. 6). Most determinations within +:Fo3. Foss to Fou. ” Data on unpalagonitized clear glass. 38 Bleached biotite; probably « chlorite. 4 Entirely decomposed to opaque oxide-clinopyroxene aggregate. % Largely decomposed to secondary mineral and amphibole, sodic plagioclase, brown glass ith apatite needles, and carpholite(?). ‘5 Olivine composition from refractive index data, using curves by Bowen and Schairer (Kennedy, 1@ Also at least two additional olivines, probably xenocrysts, with overall range of ‘composition from pague oxide rims, clinopyroxene, fibrous green mica 3! Determination by refractive indices (2=1.527—8; 7=1.638) indicates nearly pure albite. 2 A few plagioclase crystals are zaned through entire range, Ans3-ss. *s Includes ferriferous augite and ferroaugite following classification of Poldervaurt and Hess (1981, B. 474), % Includes diopsidic augite, salite, and ferrosalite following classification of Poldervaart and Hess (1951, p. 474). % Orthopyroxene composition fram refractive index data, using curves by Mess (Kennedy, 1947, fig. 3). 489292—59 (Im pocket) UNITED STATES DEPARTMENT OF THE INTERIOR GEOLOGICAL SURVEY 168°00' BULLETIN 1028 PLATE 39 167°30/ 169°30' 169°00/ EXPLANATION SURFICIAL DEPOSITS — ao Bos = 3 Ces = eS Glacier ice Surficial sedimentary deposits a Only larger areas shown = ao IGNEOUS AND METAMORPHIC ROCKS k Qra-D| S Voleanie rocks of Mt. Vsevidof Quartz-olivine andesite flow Youngest volcanic rocks S | Basaltic andesite and andesite on northeast flank of Mt. % flows; andesitic and rhyodacitic Recheshnoi pyroclastic beds S om | x= | Zz to : 3 3 Hypersthene andesite of Mafic phenocryst basalt of satellitic vents Voleanic rocks of Okmok Voleano So 2 Mt. Recheshnoi Flows of satellitic vents on flanks of Qov, Andesitic agglomerate and ash ‘8 major volcanoes; includes minor beds of the Okmok Volcanics, anda & areas of vent complex great quantity of older basaltic rocks, Also ineludes very minor vitreous andesite as small flows, 54° plugs and necks. Qor, Rhyolite 00’ : _ ] UNCONFORMITY Tmp Low grade métamorphie and plutonic complex Includes in order of age: (1) an albitized sedimentary and igneous complex that mag tain pre-Tertiary rocks, (2) middle to upper Tertiary diorit: lutonic rocks,and (3) the voleanie rocks of central Umnak, which are in a large part hydrothermally altered and potassium feldspathized rocks. Minor unaltered lavus thot may be as young as early Quaternary are also included under unit (3) Contact # 2 Probable fault 55 Strike and dip of layered rocks ee Strike of vertical layered rocks 8 Horizontal layered rocks TRUE NORTH Submarine contour interval 100 fathoms Numbers in Aleutian trench indicate depth to ocean floor, in fathoms. Data insuffi- cient for submarine contours APPROXIMATE MEAN DECLINATION, 1958 53° 54° TERTIARY 30' | Cape Kigunak oo" 52° 30° &, cee 4] saad N | ay eh, on ? > a Cy q De oe S / Voee ey i Otter Pt a Konets [7 cas > bs. — hod oN esyos\\ & s Bae Emerald @Tmp | x } : Ce U Lo ek Ly Island Le) BW A. . Kettle Cape 1. A ae a) > sumer ‘olivoni Rock = Oo urveyor Nay Bay Plate 41 ai 2 ‘The Pillars es _ Cape ve Kigushimkada Okee Point Ananilak rng SEA TMP in oa y Ta : 53° c aoe ~Tmp Ogchul =1 OO! eee Fs Island ee g a ~ Vsevidot Kelp Point Xx SS AN Island Nikolski “\ve ees Bay : fe a7 O. —y =. Black Cape Cape Starr e3, a ee, iG é Ld Pee ee. S, O ‘ mt 2, ane Se yy ees 2 - ay Cape Udak Adugak Island fy = “oF 5 : SS Idaliuk <7 7 fs - fo o iS a Cape Sagak Le © Plate 40 gs | gf” | fA & Island 1658 Wh 2025 / 1s A 1 2235 y 1668 oe , ee ae S705, JS76 oe seit 2180 Bye _ (Sit et 36 169°30" 169°00! 68°30! 163° 00! INTERIOR GEOLOGICAL SURVEY, WASHINGTON. D. © 167°30' Base adapted from U.S. Coast and Geodetic Survey chart 8861 MR 5508 Geology by F. M. Byers, Jr., 1947 GENERALIZED GEOLOGIC MAP OF UMNAK AND BOGOSLOF ISLAND AND VICINITY, ALEUTIAN ISLANDS, ALASKA, SHOWING SUBMARINE TOPOGRAPHY St Seale 1:300 000 5 Contour interval 500 feet Datum is mean sea level 10 15 Miles 1