Tertiary and of the Cook Inlet Region, Alaska

By JACK A. WOLFE, D. M. HOPKINS, and ESTELLA B. LEOPOLD

TERTIARY OF THE COOK INLET REGION, ALASKA

GEOLOGICAL SURVEY PROFESSIONAL PAPER 398-A

Discussion of stratigraphic signrJicance of plants from the Chichaloon, Kenai, and TsadzKa Formations

Property of District Library Alaska District

UNITED STATES GOVERNMENT PRINTING OFFICE. WASHINGTON : 1966 UNITED STATES DEPARTMENT OF THE INTERIOR STEWART L. UDALL, Secretary

GEOLOGICAL SURVEY William T. Pecora, Director

For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402 - Price 70 cents (paper cover) CONTENTS

Abstract--,-------Continued Introduction,------Neogene stratigraphy-Continued Previous geologic studies...... Kenai and Tsadaka Formations-Continued Acknowledgments------Provincial stages-Continued of stratigraphic nomenclature and age assign- Seldovian -Continued ments------Mollusks------.------Age ------..------Gwlogy------Homerian Stage-, ------ stratigraphy ...... Definition ---,------Chickaloon Formation -,,------Flora andage------Fresh-water mollusks-----,------, Age------Wishbone Formation------Clamgulchian Stage------Neogene stratigraphy------Definition ,---- - ,- - _------Kenai and Tsadaka Formations- --,-----,---- Flora -,------Structure and tectonics------Age------Provincial stages------,------Summary of the floral sequence------, Seldovian Stage- - -,------Fossil-plant localities------,------Definition--,------Fossil-mollusk localities ------Flora-,, ------References------

ILLUSTRATIONS

[Plates are in po&t] PLATE1. Map showing distribution of Neogene rocks. 2. Structure section. Page F~QURE1. Index map----^--,---,-,------,,- A2 2. Map of Cook Inlet Region ...... 3 3. Generalized geologic map of the Matanuska Valley ...... 8

TABLES

Page TABLE1. Partial checklist of flora from Chickaloon Formation in Alseka, showing occurrences of species in other selected areas ------,A9 2. Checklist of Seldovian megafossil flora ------,,------16 3. Checklist of Homerian megstossil flora------,------19 4. Check@khpf Clamgulchian megafossil flora------21 5. Stratigraphic distribution of significant species in Kenai and Tsadaka Fomations------22 6. Stratigraphic distribution of types in Kenai and Tsadaka Forma- tions------24

TERTIARY BIOSTRATICRAPHY OF THE COOK INLET REGION, ALASKA

TERTIARY STRATIGRAPHY AND PALEOBOTANY OF THE COOK INLET REGION, ALASKA

By JACKA. Worn, D. M. HOPKINS,and ESTELLAB. LEOPOLXI

ABSTRACT I but some paleobotanical evidence indicates that the Homerian Stage corresponds at least in part to the upper half of the The nonmarine sedimentary rocks of Tertiary age in the Cbok Miocene Series ; some lower Pliocene rocks may also be included. Inlet region, once thought to be entirely of Eocene age, are me Clamgulchian Stage is characterized by an extremely shown by paleobotanical evidence to be mostly of Paleocene, provincial flora that is depauperate in species of woody plants. Miocene, and Pliocene age. Our study of Clhickaloon floras Nearly all the warm-temperate exotic genera have disappeared. confirms the Paleocene age of the Chickaloon Formation, as Three species of willow and one species of alder seem to be suggested by Barnes and Payne (1956). The Wishbone Forma- restricted to the Clamgulchian Stage; all seem to be ancestral tion has not yielded fossil plants, but its conformable and grada- to living Alaskan species. The extreme provincialism of the tional relationship to the underlying Chickaloon Formation indi- flora makes correlations with deposits outside of Alaska im- cates that it is at least partly of Paleocene age, although some precise at the present time; we think, however, that the Clam- rocks of Eocene age may be included. Our study of Kenai floras, gulchian Stage corresponds to at least part of the Pliocene shown to be mostly of Miocene and probable Pliocene age, con- Series. firms the suggestion of Barnes and Payne that two different coal-bearing rock sequences of disparate age may be represented INTRODUCTION by the Chickaloon Formation of the Matanuska Valley and the Kenai Formation of the Cook Inlet-Susitna Lowlands. The The low-lying areas adjoining upper Cook Inlet and Tsadaka Formation, which rests unconformably upon the Chick- the lower courses of the Susitna and Matanuska Rivers aloon and Wishbone Formations, represents a marginal con- in southern Alaska (fig. 1) are underlain by a thick and glomeratic facies of the Kenai Formation; the fossil floras indi- cate that the Tsadaka Formation was deposited during the first complex sequence of nonmarine sedimentary rocks of half of the Miocene Epwh. Tertiary age that are of considerable economic impor- Three new provincial time-stratigraphic units-the Seldovian, tance because they contain coal, petroleum, and natural Homerian, and Clamgulchian Stages-are proposed. These gas. These nonmarine sedimentary rocks contain abun- units encompass all plant-bearing strata in Alaska and in dant fossil plants, but other kinds of are ex- adjoining parts of the same ancient floristic provinces that are of approximately the same age as those parts of the Kenai tremely rare. Any attempt to establish the ages of indi- Formation represented in the type and reference sections desig- vidual formational units in the Tertiary sequence of the nated in this report. Rocks belonging to these three stages are Cook Inlet-Susitna Lowlands and of the neighboring recognized and distinguished from one another primarily by Matanuska Valley must be based upon fossil plants. fossil plants. Because the fossil plants are abundant and varied, they The (Seldovian Stage is characterized by a rich and diversified warm-temperate flora containing many elements that are now can also provide the basis for understanding the time- exotic to Alaska but that were widespread during the Miocene. stratigraphic sequence of the Tertiary rocks. At least 23 fossil plant species appear to be restricted to the This report sets forth new conclusions concerning the Seldovian Stage. Comparisons of different floras suggest that ages of the Chickalwn, the Tsadaka, and the Kenai lower and upper subdivisions of the Seldovian Stage may be Formations of the Cook Inlet-Susitna Lowlands and recognized. Paleobotanical correlations indicate that the Sel- dovian (Stage corresponds approximately to the lower half of the Matanuska Valley, based on the fossil floras they the Miocene Series as recognized in northwestern conterminous contain, and it names and describes three time-strati- United States and Japan, but some upper Oligocene rocks may graphic units based on fossil floras that can be recog- also be included. nized within the Kenai Formation and correlative strati- The Homerian Stage is characterized by a less diversified and relatively provincial flora in which many of the exotic elements graphic units. are lacking. At least 11 fossil plant species appear to be re- The report is based primarily on a study by Wolfe stricted to the Homerian Stage. The provincialim of the flora of about 50 collections of fossil leaves and fruits ob- makes correlation in traditional Qmch-Series terms difficult, tained by Wolfe and Hopkins in various parts of the A1 TERTIARY BIOSTRATIGRAPW, COOK INLET REGION, ALASKA

FIGURE 1.-location of the Cook Inlet region (diagonal lines) and the Matanuska Valley (stippled), Alaska.

Cook Inlet-Snstina Lowlands and the Matanuska Val- 1 they contain. Prior to the 1950'~~attention was focused ley during the summer of 1962 as well as upon examina- I chiefly on the coal-bearing Tertiary rocks exposed tion of many small collections obtained from the region along the east shore of Cook Inlet and in the Matanuska by other geologists. A companion paper (JVolfe, 1966) ' Valley because tl~eseare in areas relatively acc.rssible discusses the systematic relationships of these plant to water, rail, and higli~~aytransportation. Tlie results fossils and their floristic significance. The study of the of early studies in these areas are sun~rna~izcclin Martin, fossil leaves is supplemented by preliminary palynologi- Johnson, and Grant's account (1915) of the geology o E cal studies by Leopold and Wolfe of 17 outcrop samples tlie western part of the K~naiPeninsula and in Martin and several samples from well cuttings; the palynologi- and Katz' description (1912) of the geology and coal cal studies of the subsurface samples confirm the results of study of pollen samples and megafossil collections fields of the RIatanuska Valley (fig. 2). A more det tiled from surface exposures, but only the surface pollen description of the stratigraphy of the Tertiary i~ocks samples are discussed individually here. The strati- exposed in the 7Vishbone Hill district in the western graphic and geologic discussion is based primarily on part of the Matanuska Valley is given by Barnes ancl the publications cited in this report, but it is supple- Payne (1956) ; maps of various parts of the Matanuskn mented by observations made by Hopkins and Wolfe Valley and Cook Inlet area are given in VITaring (1936), while collecting fossil plants in 1962. Tuck (1937), Barnes (1962a, b; 1966), Barnes and Sokol (1959), Hazzard, Moran, Lian, and Simonson PREVIOUS GEOLOGIC STUDIES (1963), and Davison (1963). The stratigraphy of the The Tertiary sedimentary rocks of the Cook Inlet Tertiary rocks exposed along the north shore of Kache- region attracted interest very early in the geological mak Bay and the east shore of Cook Inlet is described exploration of Alaska because of the coal deposits that by Barnes and Cobb (1959). STRATIGRAPHY AND PALEOBOTANY A4 TERTIARY BIOSTRATIGRAPHY, COOK INLET REGION, ALASKA Aside from reconnaissance descriptions and maps pre- here and for their critical scrutiny of the supporting pared in the course of general regional geologic studies evidence. by Capps (1913,1935), the Tertiary sedimentary rocks of the western and northern parts of the Cook Inlet EVOLUTION OF STRATIGRAPHIC NOMENCLATURE AND region received little attention prior to the middle 1950's. AGE ASSIGNMENTS Field studies by P. F. Barnes in 1961 and 1962 (Barnes, The first fossil plants illustrated from western North 1966) were the first undertaken by the Geological Sur- America came from Tertiary in the (Cook vey to be devoted primarily to the Tertiary rocks of the Inlet area. Heer described and figured fossil plants ob- northwestern part of the Cook Inlet region. tained from sedimentary rocks now known as the Kenai New interest in the Tertiary rocks of the Cook Inlet Formation from Coal Cove at the northern entrance to region arose in the 1950's with the growth of commer- Port Graham (his "English Bay" locality) and from cial investigations of the petroleum possibilities there Ninilchik in "Flora fossilis alaskana" in 1869. Heer and with the discovery (from 1957 through 1965) of correctly diagnosed the "English Bay" flora as being of six oil fields and several gas fields in sedimentary rocks Miocene age, but elsewhere in the same work he diag- of the Kenai Formation. An aeromagnetic study of the nosed floras from Atanikerdluk, Greenland, and from Cook Inlet region by Grantz, Zietz, and Andreason the coal-bearing Tertiary sedimentary rocks of Spitz- (1963) and newly published information on total bergen as also being of Miocene age. Both the upper depths of stratigraphic test wells and production wells floras from Atanikerdluk, Greenland, and those of the in the Cook InletJSusitna Lowlands provided the first coal-bearing rocks of Spitzbergen are now known to information concerning the general configuration of be of Paleocene age. It should be noted, however, that the basin containing the Tertiary sedimentary rocks of the terms "Paleocene" and "Oligocene" did not come the Kenai Formation. Extensive field and subsurface into wide use until the first decades of the 20th century geological, mineralogical, palynologicaJ, and geophysi- and that the term V"1eocene" was not accepted by the cal studies have been undertaken during the 1950's and Geological Survey until 1939. 1960's by the staffs of several oil companies, but aside Gardner (in Gardner and Ettingshausen, 1879, p. 8), from brief accounts of the structure of the Swanson describing British Eocene floras, questioned the Miocene River oil field and the Kenai gas fields (Parkinson, age assigned by Heer to the Atanikerdluk beds and 1962; Hazzard and others, 1963; Davison, 1963) and Heer's Miocene age assignments for other floras, as brief accounts of the general geology of the Cook Inlet well, saying : region (Kelly, 1963; Hill, 1963 and in press), results There is no great break in passing from one to the other of these studies remain unpublished. [Eocene to Miocene floras] when we compare them over many latitudes, and but little change beyond that brought about by ACKNOWLEDGMENTS altered temperature or migration. But if Tertiary floras of different ages are met with in one area, great changes on the In this investigation, we have been greatly assisted by contrary are seen, and these are mainly due to progressive discussions with the geologists of several of the oil com- modifications in climate, and to altered distribution of land. panies that have been engaged in petroleum exploration Imperceptibly, too, the tropical members of the flora dimp peared; that is to say, they migrated, for most of their types, in the Cook Inlet region. We have also benefited by I think, actually survive at the present day, many but slightly the opportunity to examine a number of fossil plant altered. Then the mlbtropical members decreased, and the tem- collections that these geologists submitted to us from perate farms, never quite absent even in the Middle Eocenes, areas that we were unable to visit personally and espe- preponderated. As decreasing temperature drove the tropical cially from the opportunity to compare ideas and obser- forms south, the more northern must have pressed closely upon them. The Northern Eocene, or the temperate floras of that vations concerning the pdynology of the Tertiary rocks period, must have pushed, from their home in the far north, of the Cook Inlet region with Paul Wesendunk, of more and more south as climates chilled, and at least, in the Standard Oil Co. of California; with John Browning, Miocene time, occupied our latitudes. The relative preponder- of Bhell Oil Co.; and wit11 H. Tate Ames and Walter ence of these elements, I believe, will assist in determining tbe Riegel, of Pennsylvania State University. Dwight age of Tertiary deposits in Europe more than any minute COD- parisons of species. Thus it is useless to seek in the Arctic Taylor, of the U.S. Geological Survey, kindly furnished regions for Eocene floras, as we know them in our latitudes, determinations of fossil fresh-water mollusks and data for during the Tertiary period the climate conditions of the concerning their possible paleoecological and paleogeo- earth did not permit their growth there. Arctic floms of tem- graphic significance. porate, and therefore Miocene aspect, are in all probability of We are especially indebted to Arthur Grantz and Eocene age, and what has been recognized in them as a newer or Miocene facies is due to their having been first studied in Farrell F. Barnes, of the 1J.S. Geological Survey, for Europe in latitudes which only became fitted for them in Miocene their discussion and criticism of the concepts presented timee. STRATIGRAPHY AND PALEOBOTANY A5 Dall and Harris (1892, p. 234-249) named the beds at (in Hollick, 1936, p. 28) stated, for example, that "the Coal Cove and Ninilchik, Alaska, the Kenai Group name Kenai has been restricted to its more usual for- (and on p. 249, the Kenai Series) and proceeded to ex- mational sense and limited to beds directly connected tend this name to all beds of suspected Tertiary age with the Kenai beds in the type area [on Kenai throughout Alaska. Thus, the names "Kenai Group," Peninsula] ." and "Kenai Formation" came for some years to be syn- Hollick, in 1936 in his exhaustive monograph on the onymous with nonmarine beds of suspected Tertiary age Tertiary floras of Alaska, continued to treat all Alaskan in Alaska, and "Kenai flora" to be applied to any Alas- floral material as though it were synchronous in age, kan flora of suspected Tertiary age. Thus defined, the saying (p. 21,23) : name "Kenai" was applied to beds now known to range The most obvious fact in conllection with the general facies in age from to Pliocene. Although Dall and of the flora is its unmistakable identity with the so-called Arctic Harris discussed the Alaskan Tertiary sequence under Miocene flora of British America (Northwest Territory), the heading "Miocene of the Kenai Group," they ac- Greenland, Iceland, Svalbard (Spitsbergen), New Siberia, Sak- halin, and elsewhere in the holarctic region. This flora is now knowledged Gardner's view that the Atanikerdluk beds recognized as Eocene and is believed to be approximately equiva- were of Eocene age and concluded (p. 252) that "it must lent to the flora of the Fort Union and allied formations in me be conceded that the view that the [Kenai Group] is of United States and the Canadian Provinces. Eocene age does not seam unreasonable." * * * * Two years later, in a review of the fossil floras of In view of the facts above set forth the general similarity Alaska, Kowlton (1894), was prepared to agree that of the Alaska Tertiary flora to that of the Eocene in the States those floras to which he assigned a Tertiary age repre- proper would appear to be demonstrated ; but it may be objected sented the 'LAr~ti~Miocene" of Gardner (1879, p. 8) that certain d the listed species also occur in strata more recent than the Eocene, that this fact has been ignored in the and that they corresponded to the Eocene of Europe as discussion of distribution and stratigraphy, and that these well as to floras of Fort Union age in the United States. species might indicate a later than Eocene age for the flora. In 1904, he placed a fossil plant collection from Kukak Inasmuch, however, as such species are relatively few, and as Bay on the Alaska Peninsula (and by implication all certain of the Eocene species apparently persisted throughout other "Kenai floras") in the upper Eocene. Tertiary time and are represented in our existing flora, it would be logical to infer that many species persisted into later Tertiary The stratigraphic concept of the Kenai Formation time before becoming extinct and would therefore be recognized was refined by Martin, Johnson, and Grant (1915), who as elements in Oligocene, Miocene, and Pliocene floras. presented the first systematic description of the Tertiary Doubts were expressed as early as 1906 (Brooks, p. sedimentary rocks exposed on Kenai Peninsula. In 238) that rocks containing the so-called Kenai flora the meantime, Martin and Katz (1912) mapped the Ter- were all of the same age : tiary sedimentary rocks of part of the Matanuska Val- ley, subdividing them into the Chickaloon Formation At the type locality the Kenai is made up of only slightly indurated or entirely unconsolidated beds. The same terrane and the Eska lConglomerate (of former usage). Floras has, however, been identified in much more highly altered rocks from the Chickaloon Formation were diagnosed by which carry plant remains and also coal, such as those of Hollick (in Martin and Katz, 1912, p. 49-52) as 'LArctic Controller Bay, , and Cantwell River. These Miocene (Eocene)" and "probably Kenai." On this facts are difficult to reconcile, and it appears that the decipher- basis, Martin and Katz stated (p. 52) :"The Chickaloon ing of the Alaska Tertiary stratigraphy must await further investigations. It seems at least possible that the Kenai series Formation is shown by its flora to be certainly Tertiary of the Pacific littoral may include horizons younger or older and probably Eocene. It is the local equivalent of at than the Upper Eocene, and in any event that all the coal-bear- least part of the Kenai Formation of Cook Inlet and is ing beds of the Pacific coast province are not synchronous the approximate equivalent of the Tertiary coal-bearing deposits. beds which are present in many parts of Alaska.'? Brooks' doubts in 1906 were eventually confirmed by Martin and Katz found no diagnostic fossils in their Barnes and Payne (1956) in a detailed study of the Eska Conglomerate; they stated (p. 54) that "the only Tertiary sedimentary rocks exposed in the Wishbone conclusion that can be drawn regarding its age is that Hill district of the Matanuska Valley. Their careful it is certainly Tertiary and is possibly the equivalent mapping showed that the Eska Conglomerate of Martin of the Miocene conglomerates that have been recognized and Katz (1912) actually consists of two- units sep- at several places dong the Pacific coast of Alaska." arated by an angular unconformity and having con- The name "Kenai flora" continued for many years to trasting degrees of lithification and structural com- be applied to Tertiary floras far beyond the limits of plexity. Barnes and Payne (1956) proposed the name the Kenai Peninsula, but the name "Kenai Formation" Wishbone Formation for the lower unit, which rests gradually became restricted to Kenai Peninsula. Smith unconformably upon the Chickaloon Formation, and A6 TERTIARY BIOSTRATIGRAPHY, COOK INLET REGION, A~~ASKA the name Tsadaka Formation for the less consolidated studies by Paul Wesendunk were cited as demonstrat- and less deformed upper unit ; the name Eska Conglom- ing that the Tsadaka Formation is equivalent to some erate was abandoned. They reasoned on structural and part of the Kenai Formation. The Chickaloon For- paleogeographic grounds that the Chickaloon and mation was thought to be of middle and early Eocene Wishbone Formations are probably of Paleocene age. and possibly in part of Paleocene age. The prelimi- The Tsadaka Formation was thought to be Eocene or nary conclusions reached in 1961 are revised and re- younger and to be correlative with the Kenai fined in the present report. Formation. Recent exploration for petroleum and natural gas Knowlton evidently had come to have some reserva- has shown that the Tertiary sequence reaches thick- tions around the turn of the century about the homoge- nesses probably exceeding 18,000 feet and possibly neity of the so-called Kenai flora and about the equation reaching 25,000 feet in some parts of the Cook Inlet "Arctic Miocene=Eocene." This doubt is shown by lowland (Kelly, 1963; Hill, 1963, 1966; Hazzard and the presence of the word "Oligocene" in his handwrit- others, 1963). Petroleum geologists commonly apply ing on the labels of some of the Alaskan Tertiary fossil the name "Kenai Formation'' to the entire Tertiary plant collections in the U.S. National Museum. R. W. sequence in the Cook Inlet-Susitna Lowlands and the Brown, U.S. Geological Survey paleobotanist from part of the Matanuska Valley that lies west of Moose 1929 to 1959, also had occasionally expressed his belief, Creek, although the sequence is said to contain at least in conversations, that floras of other than Eocene age two unconformities (Hazzard and others, 1963), to probably were included in the collections. However, include at least five thick sedimentary sequences of con- Brown never undertook the thorough restudy of the trasting lithology (Kelly, 1963), and to include basal Alaskan Tertiary fossil floras that would have been beds thought to be of Paleocene age (Hazzard and required to revise and refine their age assignments. others, 1963 ; Hill, 1966). Consequently, the Kenai Formation was still assigned In this report, we shall follow the current practice to the upper part of the Eocene Series by Barnes and of applying the name Kenai Formation to all beds in Cobb (1959) in their careful restudy of the stratig- the Cook Inlet region that are or were probably once raphy of the Tertiary rocks in the Homer district. physically continuous with the beds of the Kenai For- Wolfe began a systematic reconsideration of the mation exposed in the type area. The name Tsadaka Alaskan Tertiary floras in 1960; some preliminary re- Formation is, however, retained for those beds in the sults were announced by MacNeil, Wolfe, Miller, and Wishbone Hill district to which the name Tsadaka Hopkins (1961). They noted that the dating of plant- Formation was given by Barnes and Payne (1956). bearing Tertiary beds in Alaska has been hampered by We wish to emphasize that other workers (Kelly, 1963 ; Gardner's old and still unproven concept of the "Arcto- Hill, 1963, 1966; Barnes, 1966) have recognized several Tertiary" flora and said, (p. 1802) : thick and distinctive lithologic units within the Kenai However, evaluation by Wolfe of new and old plant collections Formation. Further stratigraphic studies will prob- from Alaskan rocks whose age is determined by marine inverte- ably result in the naming of several of these lithologic brates clearly shows that floras of the same age in Alaska and unib as formations and thus in raising the Kenai For- in O~egonor Washington are similar on the specific level. In mation once again to group status. The paleobotanical fact, except for the ubiquitous and long-ranging conifers such as studies reported here demonstrate that the Tsadaka Metmequoia, there is not one species known in both the Eocene flora of Alaska and the Miocene flora of Oregon. Even on the Formation is equivalent to only a part of the present generic level, there is little resemblance ,between these two Kenai Formation; thus the Tsadaka Formation proba- floras. Nearly all of the so-called "Arcto-Tertiary" genera are bly will eventually become a formation within a "Kenai typical of the Oligocene and Miocene Alaskan floras rather than Group." the Eocene. The floras reported here from the Kenai and the The preliminary study of the floras reported by Mac- Tsadaka Formations indicate that the enclosing beds Neil, Wolfe, Miller, and Hopkins (1961) led to the con- are mostly of Miocene and Pliocene age, although some clusion that floras from Port Graham and Cache Creek beds of latest Oligocene age may be included. Floras (representing the Seldovian Stage of the present pa- from the Chickaloon Formation indicate that the en- per) are "no older than late Oligocene." Studies of closing beds are of Paleocene age. The Chickaloon Hollick's illustration led to the tentative conclusion and Wishbone Formations lie unconformably below the that floras from the Kenai Formation at Homer (repre- Tsadaka Formation and contrast sharply with the senting the Homerian Stage of this report) were corre- Tsadaka in degree of lithification and complexity of lative with those found in the Acila shumardi zone of structure. We suggest, therefore, that any beds in the the Alaska Peninsula and thus of Oligocene age. Pollen subsurface that can be demonstrated to be of Paleocene STRATIGRAPHY AND PALEOBOTANY A7 or Eocene age should be referred either to the Chicka- in many places in the Matanuska Valley between Hicks loon or the Wishbone Formations unless they can be Creek and Moose Creek (fig. 3). The formation is shown to lie conformably and gradationally beneath covered by younger Tertiary sediments west of Moose those parts of the Kenai Formation that contain Nw- Creek, and the western limits of its distribution are not gene (Miocene and Pliocene) floras. known. The Chickaloon Formation appears to rest uncon- GEOLOGY formably upon the Matanuska Formation of Creta- Most of the Cook Inlet region is underlain by igneous, ceous age in exposure on Anthracite Ridge (Waring, volcanic, and sedimentary rocks of Mefsozoic age; these 1936, p. 11-12,17) and in the valley of Wolverine Creek rocks constitute the basement upon which the Tertiary (Grantz and Wolfe, 1961, p. 1765) ; it may lie conform- sediments were depited. The Matanuska Formation, ably upon the Matanuska Formation near the mouth of the youngest part of the sequence, ranges in Moose Creek and in the Knob Creek area (Barnes, age from Albian to Maestrichtian(?) (Grantz and 1962b). The Chickaloon Formation is conformably Jones, 1960) ; it consists mostly of clastic marine sedi- overlain by the Wishbone Formation. ments. The Arkose Ridge Formation, which crops out Small stocks, large sill-like bodies, and narrow dikes in the southern Talkeetna Mountains just north of the intrude the Chickaloon Formation. These intrusive Matanuska Valley, consists of coarse-grained clastic bodies are especially abundant in the eastern part of the nonmarine sedimentary rocks of Albian or possibly Matanuska Valley; they are rare in the Moose Creek- age; it is possibly correlative with mollusk- Eska Creek area (Barnes, 1962a). dated marine beds in the lower part of the Matanuska The Chickaloon Formation consists of interbedded Formation (Grantz and Wolfe, 1961). claystone, siltstone, feldspathic sandstone, and con- The Tertiary sequence includes two older and more glomerate and includes many beds of bituminous coal in strongly lithified formational units-the Chickaloon the Moose Creek, Eska Creek, and Chickaloon areas. Formation and the Wishbone Formation-and two Most of the coal beds in the Moose Creek-Eska Creek younger and weakly lithified formational units-the area are in the upper 1,400 feet of the formation (Barnes Kenai Formation and the Tsadaka Formation. The and Payne, 1956, p. 14) ; those in the Chickaloon area Wishbone Formation rests conformably upon the appear to lie about midway between the top and bottom Chickaloon Formation ; both formations were deformed of the formation (Capps, 1927, p. 42). Carbonate con- prior to deposition of the Kenai and Chickaloon For- cretions and thin beds of fresh-water are mations. The Tsadaka Formation rests with angular scattered throughout the formation. unwnformity upon the Wishbone and Chickaloon For- Thick beds of conglomerate are a major component of mations, and it evidently interfingers laterally with beds exposures of the lower part of the Chickaloon Forma- that are low in the Kenai Formation. tion near the Chugach Mountain front in the vicinity of The Chickaloon Formation is intruded by numerous Wolverine Creek and along the base of Arkose Ridge dikes, stocks, and large sills in the eastern part of the northwest of Moose Creek, but conglomerate beds are Matanuska Valley. No intrusive rocks have been seen scarce in the lower part of the Chickaloon Formation in the Wishbone Formation, but a small dike intrudes adjoining the Castle Mountain fault northeast of Moose the overlying Tsadaka Formation at the eastern end of Creek (Barnes, 1962a, Barnes and Payne, 1956, p. 15). its outcrop area (Barnes and Payne, 1956, p. 23 and pl. The basal part of the formation in the Anthracite Ridge 2). At Castle Mountain basaltic lava flows rest unwn- area consists of thick-bedded greenish-gray sandstone formably upon conglomerate customarily correlated and a poorly defined pebbly sandstone which may repre- with the Wishbone Formation (Barnes, 1962a). sent a basal conglomerate containing well-rounded peb- The Tertiary sequence is covered throughout much of bles of quartz and chert in some places and angular the Cook Inlet region by glacial drift of late Pleistocene pebbles of shale presumably derived from the Mata- age. The glacial cover is so extensive that exposures nuska Formation in others (Waring, 1936, p. 12-13). of the Tertiary rocks are largely limited to sea bluffs Higher parts of the Chickaloon Formation contain only and to the walls of sharply cut stream valleys. a few lenses of pebble conglomerate that grade laterally into pebbly sandstone. Pebbles in the conglomerate PALEOGENE STRATIGRAPHY beds consist mostly of quartz, chert, and fine-grained CHICKALOON FORMATION igneous and metamorphic rocks. Pebbles of granitic The Chickaloon Formation, named by Martin and rocks are scarce or entirely lacking in most exposures; Katz (1912), consists of a sequence of nonmarine clastic however, some of the conglomerate beds exposed north- sedimentary rocks at least 5,000 feet thick that is exposed west of Wolverine Creek are rich in granitic pebbles J 0 Sedimentary, volcanic, and Intmsive rocks igneous rocks, undivided Fossil mollusk locality

Wishbone Formation 0 8 8 { Diamond-drill hole *@ I

Chickaloon Formation I I FICUBIQ3.-Generalized geologic map of the Matanuska Valley showing approximate extent of Tertiary sedimentary rocks and localities of fossil plants and moiiuaks. Modified from Capps (19271, Grants and Jones (1960)l. Barnes and Payne (1956), and Barnes (1962a). STRATIGRAPHY AND PALEOBOTANY A9

(F. F. Barnes, unpub. data), and the Tertiary rocks Knob Creek areas (locs? 9870-9873 and 9881, table 1and exposed in Fishhook Canyon of the Little Susitna River fig. 3) ; beds exposed in the bluffs along the north bank just south of the Castle Mountain fault consist mostly of of the Matanuska River just west of the mouth of Moose conglomerate rich in granitic pebbles and cobbles. Creek (loc. 5892) ; the canyon of Wolverine Creek (loc. Barnes (1962a) mapped the granite-rich conglomerate 9874) ; and beds exposed along the old alinement of the in Fishhook Canyon as representing a western extension just west of Packsaddle Gulch (1%. of the Chickaloon Formation. This interpretation is 9877). The flora from locality 9870 was obtained from adopted on figure 3, but the granite-bearing conglomer- the Burning Bed coal group, and the flora from locality ate in Fishhook Canyon may represent the basal part of 9881 is a composite collection from several horizons the Neogene sequence and may be an extension of the within and between the Premier and Jonesville coal Tsadaka Formation. groups; these floras represent the upper 800 feet of the Barnes and Payne (1956, p. 44) suggested that the Chiclialoon Formation (Barnes and Payne., 1956, pl. 6). Chickaloon Formation was derived from areas far to The floras obtained from strip-mining pits at localities the north of the present northern margin of the Mata- 9872-9874 are in a structurally complex area in which nuska Valley. The abundance of coarse-grained con- individual coal groups cannot be continually traced glomerate in exposures of the Chickaloon Formation (Barnes, 1962b), but the floras are thought to represent adjacent to the Cliugach Mountain front suggests that the same interval. The Mat,anuska Bluffs flora was the Cllugach Mountains may also have been a m:tjor obtained at the base of a 1,500-foot sequence that proba- source of during Chickaloon time. bly lies entirely below the coal-rich upper part of the FLORA AND AGE 1 All locality descriptions are given under sections on "~o~sil-plant We hav0 obtained megafossil floras from the follow- lo~alities'~and "~ossil-mollusklocalities." In this report 4-digit num- ing locations:the coal-bearing sequence exposed in bers are USGS Paleobotany localities (except locs. 10002 and 10003) ; 4-digit numbers preceded by the letter "D" are USGS localities, Denver strip-mining pits in the Moose Creek, Eska Creek, and catalog; and 8-digit numbers are USGS localities.

TABLE1.-Partial checklist of flora from Chickaloon Fornation in Alaska, showing occurrences of species in other selected areas [Generic names in quotation marks indicate that the species are invalidly assigned to those genera]

Chickaloon Formation Hamilton Chignik Bay Fort Union Upper Bay, upr re ah of Forma Atanikerd- Species Locality Alaska Isiand, tion, High luk flora Peninsula southeast Plains Greenlaid Alaska 9871 ( 9874 ( 6892 ( 9881 ( 9870 ( 9873 1 9872 1 9871 ------Anemia elongata (Newb.) Knowl------X------Dennstaedtia americana Knowl------__------X ------Hymenophyllum confusum Lesq------X...... Onoclea hesperia R. W. Br...... ------XX X ------X Osmunda macrophylla Penh ------X------.------. Glyptostrobus nordenskioldi(Heer)R.W.Br------X X X X X ---- XX X Metasequoiaoccidentalis(Newb.)Chan------X XX X XX X X X Alismaphyllites grandifolius (Penh.) R. W. Br-----_ ------X------X ------. Carya antiquora Newb ...... ------X X---- XX cf. X Pterocarya sp------X ------Comptonia sp------,------X------X------Corylites fosteri (Ward) Bell------X X X------X cf. &uercophyllum groenlandicus (Heer) Koch------X ------X ------. X X "Planera" microphylla Newb ...... ------X X---- X---- X------X X ------Cocculusflabella (Newb.) Wolfe ------X ---- X XX X XX ------. X X Trochodendroides serrulata (Ward) Wolfe------.------X---- X---- X ------. X cf. Hamamelites inaequalis (Newb.) R. W. Br------_ ------,--X ------Sinowilsonia sp------__------X -2-.. X X---- X------. Maca~angasp------X------. sp------x ------. "Pterosperpites" sp. cf. "P." dentatus Heer ---____------,--X ------. ------. Melanolepzs sp------X------. "Sapindus" aflnis Newb ...... ------XX X X---- X------. Acer sp------X ------Decostea sp------XXXX------. pipe^" chapiniH~ll------~------X ...... ------Grewiopsis auriculaecordatus (Holl.) Wolfe ------X->-- X XX X---- X X X X Dicotylophyllum alaskanum (Holl.) Wolfe ------X X---- X------X _------cf. Dicotylophyllum flexuosa (Newb.) Wolfe------_ _ _ ------X X------X X X cf. DicotyEophyllum richardsoni (Heer) Wolfe --__-_---__------.- XX ---- X------X X A10 TERTIARY BIOSTRATIGRAPHY, OOK INLET REGION, ALASKA Chickaloon Formation represented by the floras from loon and Atanikerdluk floras is similar, and at least 10 the strip-mining pits (Barnes and Payne, 1956, pl. 6), species are common to both floras. The resemblance although the actual height of the horizon above the base between the Chickaloon and the Fort Union floras is of the Chickaloon Formation is unknown. The Wol- considerably stronger. Of the 36 well-defined species verine Creek flora was collected in a siltstone bed inter- reported here from the upper part of the Chickaloon stratified with conglomerate and lying about 1,000 feet Formation, 18 are also known from the Fort Union. above the base of the Chickaloon Formation. The None of the species that occur in the Chickaloon Forma- stratigraphic posikion of the Packsaddle Gulch flora is tion contradict a Paleocene age. unknown. In the latest correlation chart of the Alaskan Terti- The discussion that follows is based on a study of ary (MacNeil and others, 1961) rocks of Paleocene age some of the elements in these floras. We also obtained were definitely recognized only in the Sagavanirktok small megafossil floras from several localities in the Formation north of the Brooks Range, but the present Chickaloon River area, but these localities have not study has resulted in the recognition of the Paleocene added any species not recognized in the Matanuska Series in several other areas in Alaska. The fossil Bluffs and Wishbone Hill material and are not discussed floras from Hamilton Bay on Kupreanof Island in here. Pollen floras recovered from the Chickaloon southeastern Alaska (locs. 3652, 4389,4391,4392,7474, Formation thus far are small and poorly preserved ;they 7565) and from Chignik Bay (locs. 3519,3522,3523) on have been given only a cursory examination and are the south side of the Alaska Peninsula contain several not discussed here. species also found in the Chickaloon Formation (table The most diverse and best preserved floras in the 1). The Chignik Bay collections came from a series of Chickaloon Formation are those that occur in the upper rocks that rest with an angular unconformity on the part in the Moose Creek-Eska Creek-Knob Creek area. Campanian and Maestrichtian (Late ) These floras have many species in common, and some of Chignik Formation (C. A. Burk, oral commun., May these species are diagnostic of Paleocene floras in other 1963). Much of the Tertiary sequence in the Eagle- parts of the world. The Matanuska Bluffs and Wol- Circle district in east-central Alaska is also probably verine Creek floras do not differ greatly from floras in of Paleocene age (locs. 8680, 8681). the upper part of the Chickaloon Formation and also The Chickaloon floras probably grew in a subtropical contain some diagnostic Paleocene species. The Pack- or warm-temperate climate. Although the questionable saddle Gulch flora is considerably different from the familial relationships of many of the Chickaloon genera other floras, but it contains at least two ferns known make it difficult to draw conclusions from Recent dis- from the Fort Union Formation of Paleocene age in the tributions, almost 50 percent of the Chickalaon species northern Great Plains. Thus, the Chickaloon Forma- has entire margins, indicating a warm-temperate or a tion appears to be entirely of Paleocene age. subtropical climate. Dennstaedtia, Macaranga, Melano- Several of the most characteristic species of the lepis, and the fan palms grow today in warm climates. Chickaloon flora are listed in table 1. As can be seen The presence of Metasequoia, Carya, and Acer does not from that table, many of these forms are widely dis- contradict this conclusion ; the distributions of both liv- tributed in Paleocene rocks. Dennstaedtiu americana, ing and fossil species of these genera indicate that dl Onoclea hesperh, Trochodendroides serrmlata, "Plan- have been represented in subtropical and warm-temper- era" microphylla, Carya antipuora, Qtcercophylhm ate floras during some part of Cenozoic time. groedandicw, Dicotylophylltcm flemosa, Anemia elon- FRESH-WATER MOLLUSKS gata, and Hymenophyllzcm confusum are restricted to Fossil fresh-water mollusks were found in the Chick- beds of Paleocene age outside of Alaska and thus are aloon Formation in several places, but only those from valuable indices to this epoch. USGS Cenozoic locality 23367 on the Chickaloon Rivw The Paleocene floras that most resemble the Chicka- were preserved well enough for identification. D. W. loon flora are the Fort Union flora of the northern Taylor determined the following gastropods : Bellamya Great Plains (Brown, 1962) and the Upper Atanikerd- westoni (Tozer) 8; Campeloma edmontonense Tozer; luk flora of northwestern Greenland (Heer, 1868). Thiaridae?, indeterminate. The Paleocene age of the Fort Union Formation is well The two forms identified to species are previously established on the basis of fossil mammals (Wood and known only from Maestrichtian beds in western Alberta others, 1941), and the Atanikerdluk flora is well dated but, according to Taylor, could well have longer time by its stratigraphic relationship with beds containing a ranges. Ecologically, the mollusks add little informa- rich marine molluscan fauna of early Paleocene age tion ; both genera have extant species that live in warm- (Koch, 1959,1963). The general aspect of the Chicks- temperate to tropical climates. STRATIGRAPHY AND PALEOBOTANY A1 1 WISHBON18 FORMATION 1 siltstone, and claystone in the central part of the basin The Wishbone Formation, named by Barnes and and also in some marginal areas. Coal or lignite beds Payne (1956), consists of a sequence 2,000 to 3,000 feet 1 are atremely abundant in some parts of the sequence. thick of coarse-grained clastic nonmarine sedimentary Beds and lenses of pebbly sandstone or pebble ccmglom- rocks exposed at Wishbone Hill (between Moose Creek erate are scattered sparsely through the sequence and Eska Creek) (fig. 3). The Wishbone Formation ' throughout the basin. The Tsadaka Formation con- rests conformably and gradationally upon the Chicka- sists chiefly of conglomerate and coarse-grained sand- loon Formation. It "consists chiefly of conglomerate, stone in the Wishbone Hill district, but Barnes and but includes many interbeds of crossbedded feldspathic Payne (1956) also referred to the Tsadaka Formation sandstone, a few lenticular beds of siltstone, and some 1 a 1,000-foot sequence of predominantly sandy-textured claystone. The conglomerate consists of firmly ce- rocks containing only a few interbeds of conglomerate mented pebbles of fine-grained igneous and metamor- found in exploratory holes drilled by the Geological phic rocks, chert, vein quartz, and jasper in a sandy Survey in 1932 about a mile west of the head of Tsadaka matrix" (Barnes, 1962a). Pebbles of granitic rocks Canyon (Waring, 1934). The paleobotanical studies are scarce. Barnes (19&2a) assigned conglomerate beds described in this report confirm that the beds found in overlying the Chickaloon Formation at Castle Moun- the Geological Survey drill holes are approximately tain to the Wishbone Formation, although, because fos- the same age as the Tsadaka Formation. sils are lacking, the possibility cannot be excluded that The thick conglomerate sequence of the Tsadaka For- these beds are more nearly correlative with the Miocene mation constitutes a major part of the Neogene section Tsadaka Formation. At Castle Mountain, the conglom- at the margin of the basin in the Wishbone Hill district, erate assigned to the Wishbone consists of "alternating and thick unnamed sequences of conglomerate are pres- beds of conglomerate 5 to 50 feet or more in thickness ent in the Kenai Formation-possibly at several strati- and arkosic sandstones that range from a few inches to graphic levels-in a marginal belt extending along the 40 feet. The basal portion of the conglomerate is coarse west edge of the Cook Inlet region from the vicinity of and contains pebbles as much as a foot in diameter. The the Peters Hills southward to at least the vicinity of pebbles consist mainly of igneous materials, including Chinitna Bay. Conglomers~tebeds form a relatively acidic porphyry, fine-grained basic rocks, granite, and minor part of the Kenai Formation in exposures along diorite, as well as quartz, greenstone, and metamorphic the south shore of Kachemak Bay. No published infor- rocks of various typed' (Capps, 1927, p. 45). mation is available concerning the lithology of the The Wishbone Formation and the overlying Tsadaka Kenai Formation at the perimeter of the basin in other Formation were originally included in the Eska Con- areas. glomerate of Martin and Katz (1912), but Barnes and The best exposures of the marginal parts of the Kenai Payne (1956) showed that the two formations differ Formation are those found in several places along the in degree of lithification and structural complexity, that south shore of Kachemak Bay between Port Graham granitic pebbles are scarce in the Wishbone Formation and a point 5 miles northeast of Seldovia. Sedimen- at Wishbone Hill and abundant in the Tsadaka Forma- tary rocks of the Kenai Formation there bury a rugged tion, and that the two formations are separated by an topography carved in sedimentary and volcanic rocks angular unconformity. The Wishbone Formation has of age and having a local relief of at least 100 not yielded identifiable fossils; however, it is thought feet. The small exposures of the Kenai Formation to be at least partly of Paleocene age because it rests between Seldovia and Port Graham consist of places conformably and gradationally upon the Chickaloon where the present shoreline intersects a narrow, steep- walled northeast-trending canyon cut in the Jurassic Formation. Some Eocene rocks may also be included. rocks and subsequently filled with Nwgene clastic sedi- The Tsadaka Formation is of Miocene (Seldovian) age, ments (pl. 1) ; the exposures northeast of Seldovia ap- as is shown in later paragraphs. parently represent material filling a somewhat larger NEOGENS STRATIGRAPHY erosional valley or basin to which the buried canyon may have been tributary. The Neogene clastic sedi- EENAI AND TSADAKA FORMATION8 ments in these exposures along the south shore of The Kenai and Tsadaka Formations occupy a roughly Kachemak Bay consist of sandstone, siltstone, and clay- elliptical basin (the Shelikov trough of Payne, 1955) stone with a few thin lignite beds. These grade into extending from the vicinity of the Peters Hills south- masses of conglomerate and taluslike or colluviumlike ward to and beyond Seldovia and Cape Douglas (pl. 1). breccia in zones a few tens to a few hundreds of feet The Kwi Formation consists chiefly of sandstone, wide adjoining the buried slopes carved in the Jurassic A12 TERTIARY BIOSTRATIGRAPHY, COOK INLET. REGION, ALASKA

rocks. These relationships are well illustrated in fig- Point along the north shore of Kachemak Bay and ures 1 and 2 of Martin, Johnson, and Grant (1915) al- another stratigraphic section approximately 2,000 feet though these authors interpreted as fault breccia the thick in exposures along the east shore of Cook Inlet fossil talus or colluvium shown near the left and right between Cape Starichkov and Clam Gulch. Their Cape margins of their figure 2. Starichkov-Clam Gulch measured section lies mostly In the western Matanuska Valley and along the west above their Kachemak Bay measured section, but the margin of the Cook Inlet region, thick masses of Neo- lower part may include some rocks correlative with the gene conglomerate adjoin modern steep mountain slopes highest part of the Kachemak Bay measured section. underlain by older rocks. The presence of this con- The Kachemak Bay measured section consists of a glomerat,e and of rugged topography buried beneath lower, predominantly fine-grained sequence in which sediments of the Kenai Formation near Seldovia sug- the clastic beds are strongly lithified and the organic gests that the original basin in which the Neogene rocks beds consist of subbituminous coal and an upper sandy were deposited may not have been much larger than sequence in which the clastic beds are only weakly lithi- the area presently underlain by the Kenai and Tsadaka fied and in which the organic beds consist of lignite. Formations. Only the weekly lithified sandy and lignitic sequence is Kelly (1963) distinguished five lithologic zones in represented in the Cape Starichkov-Clam Gulch meas- wells penetrating the Kenai Formation in the central ured section. Lithologic comparisons, comparisons in part of the basin. From top to bottom, these include his the rank of coal beds (Barnes, 1962c), and comparison zone 1, about 5,000 feet thick, composed of massive sand with the character of the pollen floras summarized by beds; zone 2, several thousand feet thick, composed of Hill (1963) all suggest that Barnes and Cobb's upper sandstone, shale, and lignite; zone 3, several thousand weakly lithified sandy and lignitic sequence corresponds feet thick, compmed of siltstone, shale, and low-rank to Kelly's zone 1 and Rill's "birch-alder zone." Barnes coal; zone 4, the "Hemloclc producing zone" of local and Cobb's lower well-lithified coaly sequence probably petroleum geologists, composed chiefly of conglomerate corresponds to Kelly's zone 2 and part of his zone 3 and sandstone and having a total thickness of about 700 and to part of Hill's "elm-hickory zone." feet; and zone 5, consisting of several hundred feet of STRUCTURE AND TECTONICS siltstone and shale. Hill (1963, p. 197) indicated that the upper massive The boundaries of the area underlain by Tertiary sand sequence (Kelly's zone 1) is charactei.ized by rocks in the Cook Inlet region (pl. 1) appear in most pollen assemblages rich in birch and alder and that the places to represent the outcrop of the unconformity that sandstone, shale, and lignite or coal sequence below separates the Neogene sequence from older rocks. (Kelly's zones 2 and 3) is characterized by pollen as- I-Iomever, the basin is bounded by the Castle Mountain semblages containing elm and hickory. Kelly's zone 1 fault along the north side of the western Matanuslta of massive sand beds contains several large accumula- Valley and by minor faults in several other localities. tions of methane gas on both the east and west sides of A major fault may also constitute the south boundary Cook Inlet (Kelly, 1963) ; gas has also been obtained in of the basin between Knik Arm and Kachemak Bay. a thick sandstone zone within Kelly's zone 2 (Hill, 1963, In that area, a nearly linear escarpment that is probably p. 19) Kelly's zone 4, the "Hemlock producing zone," a fault scarp separates the pre-Tertiary rocks exposed yields petroleum in the Swanson River oil field of north- in the rugged Kenai-Chugach Mountains from the low- western Kenai Peninsula and in several newly dis-. lands to the northwest which are apparently underlain covered oil fields on and near the northwest shore of by the Kenai Formation. Cook Inlet. Kelly suggested that the "Hemlock pro-, The basin containing the Kenai Formation is divided ducing zone" corresponds to the Chickalon Formaticm, ' into a deep southeastern segment and a shallow north- and Hill assigns a Paleocene( 1) age to some part of the western segment by a major structural discontinuity Kenai FormatJion found in the subsurface, but no ' consisting of the Bruin Bay fault, the Moquawlcie mag- paleontological substantiation for these age assignments netic contact, and the part of the Castle Mountain fault has been published. that lies east of Theodore River (Gmntz and others, Sections of thick sequences of Neogene rocks based on 1963). Paleobotanical correlations suggest that areas surface exposures have been measured for the Kenai south and east of this major structural discontinuity Formation only in the area between Swift Creek and were subsiding much more rapidly than areas to the Clam Gulch by Barnes and Cobb (1959). They have north and west while the Neogene sedimentary rocks described a stratigraphic section approximately 5,000 were accumulating because the individual provincial feet thick in the area extending eastward from Anchor stages defined in this report appear to be represented by STRATIGRAPHY AND PALEOBOTANY A13 much thicker stratigraphic sequences south and east of PROVINCIAL STAGES the discontinuity than to the north and west. The Surface exposures of the Kenai Formation contain Bruin Bay fault and the Moquawkie magnetic contact abundant fossil leaf floras of varied character. Some of may no longer be active structural features, but the part the differences among the local fossil leaf floras reflect of the Castle Mountain fault, that lies east of Theodore differences in the ecological conditions that prevailed River has undergone movement within late nearby when the enclosing sediments were deposited. time; its trace across the Susitna valley is marked in For example, fossil floras collected in the central part of many places by scarplets that transect morainal fea- the basin are generally poorer in species and genera than tures of the Naptowne ("classical" Wisconsin) Glacia- fossil floras collected at approximately the same strati- tion (Karlstrom, 1964). graphic level near the margin of the basin. Floras from Superimposed upon the major basinal structure are the central part of the basin consist largely of taxa several broad, gentle northeast-trending folds having a whose nearest living relatives thrive in pond, swamp, structural relief of several thousand feet. Dips on the and flood-plain enviroments; floras collected nearer the flanks of these folds as seen in surface exposures and in basin margin contain a better representation of the taxa published cross sections of subsurface structural fea- that grew best in the varied sites available in a well- tures (Parkinson, 1962; Hill, 1963) are generally less drained environment. However, certain consistent dif- than 10". The larger folds northwest of the Castle ferences among the local fossil leaf floras are clearly Mountain-Moquawkie-Bruin Bay structural discontinu- correlated with differences in stratigraphic position ity are expressed in the landscape by large, rounded within the Kenai Formation and are thought to reflect mountains such as Mount Susitna and the Peters Hills differences in the age of the enclosing beds. For (pl. 1) that are composed of pre-Tertiary rocks and that example, beds low in the exposed part of the Kenai represent the exhumed cores of anticlines involving the Formation contain leaves of species of Ahus and SaZh, Neogene rocks. Comparable folds are present in the among other genera, that appear to be ancestral to some deeper part of the basin to the southeast of the Castle of the species of AZnw and SaZix represented by leaves Mountain-Moquawkie-Bruin Bay structural disconti- in higher beds of the formation. Beds low in the nuity, but their cores remain deeply buried beneath a exposed part of the formation contain varied floras of a thick cover of Neogene and Quaternary sediments generally warm-temperate character, but beds high in (Kelly, 1963). Some of these folds were already grow- the formation contain floras closely related to the ing while the Kenai Formation was accumulating. De- present depauperate Hudsonian and Boreal forest of t,ailed correlations between development wells in the southern and central Alaska. Swanson River oil field show that individual strati- Pollen floras obtained from the Kenai Formation also graphic units thin over the crest of the low anticline show differences that are correlated with approximate there (Kelly, 1963, p. 296; Parkinson, 1962, p. 182). stratigraphic position and that therefore reflect evolu- Topographic anomalies on surfaces underlain by sedi- tionary changes in the regional vegetation, as well as ments of Quaternary age suggest that some of the anti- differences that can be correlated with geographic posi- clines may still be active (Kelly, 1963, fig. 9 ;Hill, 1963). tion within the basin. Differences due to geographic Smaller folds and small high-angle faults are super- position are less conspicuous in the pollen floras than in imposed upon the larger structural features in some the leaf floras because some pollen types can be trans- ported long distances by wind; consequently, upland areas. Small, steep, closely spaced folds in the Kenai plants are commonly represented by a few pollen grains Formation exposed in Cache Creek valley near localities in samples from positions in the central part of the basin 9867 and 9868 (pl. 1) may represent landslide masses. where leaves of upland plants are entirely lacking in the Barnes (1966) noted similar small areas of extremely megafossil collections. In spite of this fact, the pollen disturbed structural features along the walls of valleys floras are less useful than the leaf floras for age deter- cut deeply into the Kenai Formation northwest of Cook mination because pollen grains generally can only be Inlet. Elsewhere, the small faults that cut the Kenai distinguished at the generic level, whereas well- Formation typically have displacements of no more preserved leaves can generally be distinguished at the than a few tens of feet. Minor faults in the Swanson species level. Our age determinations based on pollen River oil field appear to show increasing displacements samples taken from the Kenai Formation during the with increasing depth (Parkinson, 1962, p. 182). This present study must depend more upon the abundance observation suggests that the faults were developing and variety of exotic genera than upon the presence or while the Kenai Formation was being deposited as was absence of taxa known to have a narrowly restricted the major anticline upon which they are superimposed. stratigraphic range. Furthermore; pollen grains are A14 TERTIARY BIOSTRATIGRAPHY, COOK INLET REGION, ALASKA readily eroded and redeposited, and reworked pollen I Seldovian Stage grains commonly are not distinguishable from grains DEFINITION derived from plants living at the time that the enclosing sediments were deposited. Consequently, the presence The Seldavian Stage is proposed as a provincial time- at a given stratigraphic level of single grains of pollen that encompasses all plant-bearing of exotic taxa cannot be interpreted confidently as evi- strata in Alaska and in adjoining parts of the same dence that those taxa were living in the area when the ancient floristic province that are of approximately the enclosing beds were laid down. In spite of these reser- same age as those parts of the Kenai Formation repre- vations, our preliminary palynological studies generally sented in the type section along the Chuitna River and have confirmed and amplified the stratigraphic results near Capps Glacier and in the reference section near of our studies of the megafossil floras. Seldovia Point, 2 miles north of Seldovia. Rocks Three time-stratigraphic units can be recognized belonging to the Seldovian Stage are recognized pri- within the Kenai Formation on the basis of the fossil marily on the basis of the fossil floras that they contain. leaf floras they contain. The lowest of these lthree The stage is named after Seldovia Point because strata units contains widespread floras much like the floras of the Kenai Formation exposed in sea cliffs % to 3 contained in beds of comparable age in Oregon, Wash- miles east of Seldovia Point on the south shore of ington, and Japan; we can state confidently that this Kachemak Bay contain an especially rich flora that unit includes beds of early or middle Miocene age be- includes most of the elements upon which recognition of cause its floras include taxa diagnostic of approxi- the Seldovian Stage is based. These exposures, briefly mately the first half of the Miocene Series. Floras described by Martin (in Martin and others, 1915, p. 82) within the two higher time-stratigraphic units become do not display either the top or the bottom of the Sel- increasingly provincial and are more difficult to place dovian Stage; they are therefore designated as a refer- within the framework of Pacific Basin upper Miocene ence section rather than a type section. The type see- and Pliocene stratigraphy. Nevertheless, Neogene beds tion of the Seldovian Stage is designated as the in many other areas in Alaska can be correlated with sequence of strata of the Kenai Formation that is ex- one or another of the three time-stratigraphic units posed on the flanks of the ridge south of Capps Glacier recognized within the Kenai Formation because they and along the walls of the upper valley of the Chuitna contain similar floras. Thus, these three units provide River downstream to the position of our locality Dl949 a useful standard on which to base correlations and (pls. 1and 2.) age assignnleiits withill Alaska and probably within The top of the Seldovian Stage is designated in the adjoining parts of Siberia and northwestern Canada. type section as lying at the level of the coal bed from For these reasons, the three time-stratigraphic units which our pollen specimen Dl949 was collected. This specimen contains a typical Seldovian pollen flora. A are defined and described on the following pages as the calcareous siltstone bed lying stratigraphically about 50 Seldovian (Oligocene? and Miocene), Homerian (Mio- feet above this coal bed and represented by our locality cene and Pliocene ? ) ,and Clamgulchian (bfiocene? and 9844 contains a leaf flora typical of those contained in Pliocene) provincial stages. rocks of the Homerian Stage. The Tsadaka Formation contains a flora diagnostic We are not prepared at this time to define the base of the Seldovian Stage, an indication that the forma- of the Seldovian Stage with precision. Recognizably tion is equivalent in age to the oldest part of the Kenai older floras are not known below the type section or in Formation found in surface exposures. The lowest any other surface exposures of the Kenai Formation part of the Kenai Formation penetrated in petroleum of which we are aware. On Sitltinak Island, however, exploration wells, however, may consist of beds that are fossil plants have been collected from a coal-bearing not present in any surface exposures from which we unit that conformably underlies marine rocks of earliest have floras and that may be appreciably older than the Miocene ("late Blakeley") age (G. W. Moore, written Seldovian Stage. Our paleobotanical studies indicate commun., Oct. 19, 1964; I?. S. MacNeil, written com- that beds belonging to the Seldovian Stage can be dis- mun., Sept. 12, 1962). A small collection in the posses- tinguished by their contained leaves from beds of Oli- sion of Mr. C. E. Nickles of King Salmon, Alaska, gocene age on Sitkinak Island and in the Gulf of obtained from very near the highest part of the coal- Alaska, as well as in Oregon and Washington (see bearing rocks, contains the characteristic lower( ?) Sel- Wolfe, 1966). Some of the lowest beds found in sub- dovian species Alnus evidem. Another collection (loc. surface penetrations of the Kenai Formation may be of 10002), from about 2,000 feet stratigraphically below Oligocene age or older. the Nickles locality, contains Alnus n.sp. aff. A. ewidens. STRATIGWHY AND PALEOBOTANY A15 A third collection (loc. 10003) comes from a locality However, we think that at least part of the variation that is thought to be at or below the horizon of locality among different Seldovian floras reflects difference in 10002; this third collection contains Alnus n.sp. aff. age. The floras from Tsadaka Canyon, Little Susitna A. evidens, Carpinus n,sp. aff. 0. cappsewis, and Cer- River, Redoubt Point, and Harriet Point have several cidiphyllum crenutum. The first two species are related similarities that set them somewhat apart from other and probably ancestral to species that we consider to Seldovian floras. These similarities include the lack be diapostic of the Seldovian Stage. Thus, the of m-hite oaks or other Fagaceae, a scarcity of maples Nickles collection is placed in the lower( ?) Seldovian, (only one species, Acer fatisiaefolia, is known in these but the two lower collections (10002, 10003) are wn- floras), and the occurrence of A7nu.3 esidens and Corylus sidered to be pre-Seldovian. A more precise definition harrimani. In contrast, the Seldovia Point flora, the of the position of the base of the Seldovian Stage must Capps Glacier localities, and the Cache Creek localities await additional collecting. have white oaks and other Fagaceae (abundant in the Seldovia Point flora), diverse maples, and Alnus cappsi. BLOBA In addition, species of Xalh that are closely related to Strata assigned to the Seldovian Stage in various Homerian or Clamgulchian forms are known mostly parts of the cbok Inlet region have thus far yielded 76 from floras containing white oaks and Ainus cappsi, species of plants based on fossil leaves (table 2) and and the predominantly Homerian species Spiraea weav- 25 pollen and forms representing at least 14 angio- eri is known from the Seldovian locality at Houston sperm genera and 19 families of tracheophytes. The along with white oaks and Atnzcs cappsi. Although we leaf and pollen floras are typical of the "Arcto- have not yet found the two types of floras in a single Tertiary" flora and are characterized by taxonomic continuous stratigraphic sequence, the floras containing richness, especially in the deciduous dicotyledon groups Atnus evidem or Coryk~harrimani and lacking such as Salicaceae, Juglmdaceae, Betulaceae, Fagaceae, Fagaceae or maples other than Acer fatisiaefotia are Ulmaceae, and Aceraceae. thought to be older than the floras containing Ainus The following significant megafossil species appear cappsi, Fagaceae, and diverse maples ; we therefore des- to be restricted to the Seldovian Stage: Salix in- ignate them tentatively in this paper as lower (a) and quirenda, Alnw J~ealyensis,A. barnesi, A. fairi, A. upper(?) Seldovian floras, respectively. kargei, A. evidens, A. cappsi, Carpintcs sezdowiana, C. The pollen and spore samples of the Seldovian Stage cappensis, Quercus (Leucobalanus) furuhjelmi, Q. that have been studied come from the type Seldovian (L.) bretzi, and Fagus antipofi. Also restricted in section in the Capps Glacier-Chuitna River area (seven Ahska to the ,'3eldovian Stage are Zetkova oregonianu samples) and from the referred Seldovian section near and Liquidan%barmioformosana, but these species are Seldovia Point (two samples). Grain counts are based known to hare longer time and stratigraphic ranges in on at least 100 grains except for one sample (~1719)in other regions. In addition, the following taxa are which only 36 gains were recovered. abundant or common in the Ekldovian Stage but are The most notable characteristic of all the Seldovian rare in the Homerian Stage: Carya bendirei, Pterocarya samplesis the large proportion of &coty- nigella, and Cercidi~hyllUmcrenatum- Ulmu.3 is repre- ledon pollen exclusive of Betulaceae. Most of these sellted by fossil leaves in Seldovian beds and by pollen dicotyledons represent genera now exotic to Alaska : in some Homerian beds. Carya, Juglans, Pterocarya, Ulms-Zelkova, Liquidam- The leaf floras of the Seldovian Stage show some bar, Zbz, Tilia, and Nyssa. We have not observed all individud variation from one locality to another. This these genera in any one sample, although in various floristic variation is undoubtedly due in part to former at least four are present in any given environmental differences between parts of the Cook sample. Ulmus-Zelko.va is present in all samples, Inlet region during Seldovian time. For example, the carya and Pterocarya in eight, and Nyssa in seven. richness of Seldovia Point flora (10~s.6061,9856-9858) Two have a low (1 and 1.5 percent of pains probably reflects the diversity of the ecological and tope- counted) of this exotic dicotyledon ele- graphic conditions that prevailed there during Seldo- ment, but the other seven samples have between 7 and vian time, and the strong dominance of Betulaceae in 92 percent; locality Dl949 (pl. 2), which is taken here Seldovian beds from many other parts of the region as the top of the Seldovian in the type seotion, has 8 probably reflects the rather monotonous topography and percent. the wide distribution of poorly drained sites that must Pimcs pollen is present in all samples and ranges in have prevailed in areas nearer the center of the basin abundance from 1to 18 percent. Picea is noted, some- of sedimentation. times in abundance (as much as 23 percent) in seven TERTIARY BIOSTRATIGRAPHY, COOK INLET REGION, ALASKA

TABLE2.-Checklist of Seldovian megafossil flora [Numbersindicate localities]

Lower(?) Seldovian -? Upper(?) Seldovian B V) a C1 ,. B 9 ..4 B U z P1 & t! 3 Species d 3 m rn Y 4 a 9 0 - 358 g 0 mM 5 ZZ 3 5 2 2 nga E W D la d U 63

------A ------9359 8380 9364 9865 9866 9863 98% 9701 9885 9886 9945 8887 9760 gs89 3605 9866 9857 6061 9858 9848 9849 9850 9845 9646 9937 9365 9867 6083 6066 Equketum sp ....------.------X ---- -.------X X ------Dryoptcris sp.....------.------..---..------. ------.------.--.. ---. ------.----. ---- X ---- .------.------. ---- -.---.-- -_-. Onodea sensibilk L ....-----.------.-.------..------...------.---.------.-.X ------..------.----_. -.-- ___- Osmunda sp -....------X ---- cf. --.-...------.------.. -.------.--.------..-.------.---_ ..-. --.- Ginkgo biloba L .-.------.---.-- ..-- -.-- .------. ------.------X X ---- X .-..------.------.----.- --.- .--- Glyptostrobvs europaevs (Brong.) Heer -..------X X ------.------.------.--- X X --.---.-----.-.-----X X -.------X ---...---.-- Mdasequoia glyptostroboidea Hu and Cheng...... X X ---- X X------X X X X X X----X X X ---.X ----.---X X X --.------X ---- Taxodiumdidtichum Rich--...------..------.----..- ---- .---..------.-.---.- .--- X .--.---- X --...------.-.---.------_-- Sequoia sem eroirena End1---.----.-- ---.------..------..--.-.------.------.-..------.----- .-.------.--.--. ------.------.---- -.-.---- X ~invssp.~?~.pondcrosa~aws.------.X -----.------.------X ---.---.------. Pseudolarw sp --.------X ------Chamaecyparia sp ...... ---..-- .------.-.--. .------X ---.------Potcrmogeton sp -----.--..------.-.------. ---- -.------.------.------.--.. X -..-.--- -.-- ---. ------.------.-_---- SmUax sp ------.------.------.------X --.-.--. ------..---.------CyprracUeesp --.---.--.------X ------.-.------X -----.------.-.. X .------X .---.------. X -.-- X X ----..._X Poacitccr tenuistriatua Heer --.-.------.---- -.-.. ------.-...------. ------.----- X ---- ..-.X ---.------.-.-.------.---.------PopzJw kenaians Wolfe--.------.... ------.--X cf. .-.------.------cf. ------X ------X X ---- X cf. X -_.-X X -.------sp. aft. P. latwr Braun ----.------..------.-- -.------.------.------.-.X ---.---. X ---- .-----.. X --..--.- --.- ---- x ---- Sal@ cappaensid Wolfe.----.--..------. ------.----. .------..------.--.---- -.------.---.--- -.-- .------.. X ---- -.------.------picroido (Heer) Wolfe------.-.-. ..----.- -.-- .--- -,------.------.------.------X ------.--X X ------.--.-. ------_-- inquirenda Know1...------.------.--..------X .---.-.. .--. -.------..---.-- X ---- X ------.---- X X X ---- sp. aff. S. confirmato (Holl.) Wolfe--.--.----.------.-..------.-- .-.------.------.---.-. cf. X ------...-.- X ------X .--. sp. aff. S. crasaijds Trautv..-.------.--.------.------...-- cf. ------.--.-- X X ---- -.------sp. aff.S. kachemakewis Wo1fe.-. ---- X ------.------.------..---- -.--X ..-.. ------.------X -.-.---. -.------Comptonla sp. ci. C. naumani (Nath.) Endo...------.-.------..----- ..------.------.------.------X ------.------.------Carya bendirei (Lesq.) Chsn. and Ax& ---.------.---. ------..------.------.--X ------..------X ------.-- Ptnocarya mizta (Knowl.) R. W. Br. ------.-.---. ------..--.- ---- .... ------.------.-X ------.X .------.----.-- ---. .-.. .--- ni~eUa(Heer) Wolfe----.-----.-- X X ------. .------.------.--.--- X X ---- X cf. ---- X X ------..-----. .------. (Cydoptera) sp----.------.------.------.----.- -.------+ X ------..---- -..- Alnw barneaf Wolfe----.----.------X ------.------X ---- X --L------cf. cap i (Holl.) Wolfe.------.--.. ------..---- -.-----. --.- -.------.------.------..-.-- -..- .... X ------X X X X X ---- m~(~oll.~~olfe------x cf. x d. x cf. ---- x x cf. ---- x x ...... jairf (Knowl.) Wolfe------..--.--- .------.--... ---- -.-----. ------.-...------...------.X --..------.---.------X ---- -.-..--- healyenaid WoUe....------_--. ------.------.--.------... X ------X ------X X X ---- X ------largei (Knowl.) Wolfe-.------..------.X -..- ..-- --.. ------.----. -.-- X ..-.---- -.----.- ---- X ------.------.-.---- Carpinuscappsenaia Wolfe..----.---- X X X ------..---- cf. ------X ---- X ------.------.------X X ------....-- .--.. seldomana Wolfe...------..- .... ---- .-..------.. ..------.------.---.-.- -.------.-.. X ------X ------.------. Corylus harrimani Knowl ..------X cf. ---. ------.------.--*------.------.------.+.- .-.------..-.---- .-..------Bet* subblutea Tan. and Suz .----.---..---- ..-.--.- ---- .--..------..-.--. ---- .--..------.-.------.-..- X ------..--.------.------.---.--. Fagus antipofi Abich ..--...-.------.---.------. ------.-.------cf. ---.X X X X ------.-.--.------.------sp. cf. F. paleocrenata Tan.--..------..-- .-.. ------.-----.------.--.------X ------.------.------.------.--.---- Quercus bretzi Chan---.-.------. --...--- .------.----.----.. ---. .--- .--. ---- .------X ------X .-.------.-.---. ---- .... X X ---- furuhjelmz Heer ------.------.------. -.-- .--. ---- -.------.--.--.------..---- X X ---- X ---- .---.------.X ---.---- .--- Ulmuslongi olio Ung ------..--.---- .--.---- ..-- --.------.----- .----.-- -.-. ..-. ---- .--.---- X .------X ------X ------..-. X --..------. sp. aft. 6. newberryi Knowl .----.X cf. ---- -.-.---. ------.----- .----.------..------.---- X ---.---- X .-.-X ---- X ---.---. ------.---- Zelkoworegoniana (Bowl.) R. W.Br. -.------.-.------.------..------X -..----- X .-.------.------X ---.---- .--- Nuphar sqi ----.---.------.------.--.------.------..----.------X ------.------.... ------Ceratoph~um sp ---.....------.--.--- -.-- ---. .--- .--. ---- .------.---.- -.------.--.-.-. .--. ------...----. ------X ---- .--- Cercidiphyllum crenatum (Ung.) R. W.Br...... -...---.------.------.-.-..----.--..-- X .------.------.------X X .---..----.-..------cf. X ----....------.-.-.. Coemhs auricu&ta (Heer) Wolfe....------_------.------.--.--.. ..-- .--- --.- --.- ---. .--. X ---- -.-.X ------.------.--.------Hydrangea sp. -..--.--....------.------.------...-..-- .--- --.. --.. ------.------..--. ------X ------..------.------Liquidambar mioformoaana Tan------.------.------.---.--. --.- .-.------.----.. .--- .--. ---- X X ..--X ------X ---- .-..------.--.-.- Platanus bendirei (Lesq.) Wolfe.------.-----.------.------...-. -.-- -.------X -.--.------.------.------Cidaegua sp- -.-----.-..------.------.----.------.-.-. .--. --.- .--. --.- --.------..------.----- .---X ------.------.--.X .-.------.-- Prunzls sp- .------.--.------....-. .------. .--- -.-- ..------.--..------...------X ------X .-.------..------Spiraea weaoeri Holl .. ------..---.------.... ---- ..-. ---. .--. -.-- .--- .--- .--- .-.- --.- -.-. ---. -.-- -.-- .------. ..-. .------.------X ------(7ladrastis japonica (Tan. and Suz.) Wolfe--.---. ------.------.------.-....-- ..------.------.-..---- -.-----. ---- .------.------.--- X .--.---. -- -- .-.. ------Akhorncal sp. (Populua delieata Chan.) --.-...-..-...------...--.~~~-~-~~.~..~--~--~~.~-----~--...---~.-~~ * -----.------.----.------X ------.---.- ---.----. ------.------.- Malldm s ...... x ------~--~-~------Rhus sp. &I------R. glabra .------....-.-.--.--- .--- cf. ---- ..-..--. ---. ------.------.-..------. .-.. --.- --.. .-.. ---- .---..------X ------.-- Acer enwnum Ois. and Huz--.-.-.--. .... ---- ..--.------.-..--.. ---.. .------.--.---. .------...----- .---X X ---- .---.------.------fatisiaefolia Huz.-----.----.------X .----.-- .--- .-.. ---- -.--.--. ---- -.--.--. ------.------..-.. ---- X ..-..--. --.. X .-.-.-.. ---- .... .----.-- glabrotdes R. W. Br .------.------.--..------.-.------.---.--. ---- .------.------..------X cf. ---- X X ---- macropterum Heer. ....-----.....------..-.. ---- .... .--..-.- -.-. .-.. ---. ..------.---X X -.-.X -.------.---...------sp. aff. A. crataegiJolium Sieb. and Zuce- ---.------.------.---- ..-...-.. -.-- .... ----. ------.--.---. -.-. ---. .------..--X ------.------.------sp. of. A. subpictum Sap ._...... ----. .--- -.-. ..------.... .--.------.---. ---- -.-.------.---.-.. X ------.---.--- X X ------Aesculzls majw (Nath.) Tan.-....------..--..-. ---- -.-..--. .------.-..--.- .--. ------..--.-.. .------.-.---.- ---- ..------X ------vitiasp --..-....-.-.------.------.---- ..-...------.------..-. ------.---x ------sp ---.---...-.-.------x ---- -.-.------.------.---. ---- .------.------Eli~sp ...--...... ------.------.-.------.--.-.. ------..--. ---- -.-.------..------X .-.------..------Nyssa sp. cf. N. knosltoni Berry .--.---.- .------.------.---.------.------..--. -.-- X -.-..------.------Hemerapa borealis (Heer) Miki ----.---.- ---. -.-- ..------.--.------. .-.. ------.------X ------.--..-.------.--..------Kalopanaz sp. ..------.-...------...--- -.-- -.-. ---- -.------.------.---- .-...--. ------X --.------.-.---. .------Vaccinium sp ....-.------.-.-.-.----.--.------.-..------.------.----.- X ---..------..-.. ------..------fiwinua n. sp. "-4". .-.------.-.--.-.---..-- .--- -.-- ---. X ------..------.-..------.-..------.------.------.------n. sp. "B" ..--.--.-.------.------..--- -... ------.--.--.------.------X ---- .------..-.------STRATIGRAPHY AND PALEOBOTANY A17 samples, but Abies and Tsuga are never abundant and The upper ( 1)Seldovian floras appear to be correlative occur in three and six samples, respectively. with the floras of either Wolfe7szone 2 or zone 3, which Pollen of Betulaceae is found in all the samples, are considered to be of late early and early middle typically in abundance (as much as 88 percent). Ex- Miocene age, respectively. cept in two samples, Alnus is more abundant than pollen The lower (8) Seldovian floras have not been as exten- of the Betula-type (BetuZa, Carpinm, Ostrya). sively collected as have the upper( ?) Seldovian floras, Ericales are represented in five samples but do not com- and thus their correlation is less certain. Several low- pose more than 4 percent of any tally. er ( ? ) Seldovian species are indicative of a Miocene age : Uhus sp. aff. U. newberry, Acer fatisiaefoZia, and MOLLUBKB Pterocarya nigella. P. nigella is especially significant With one exception, all the identifiable fresh-water because middle and upper Oligocene rocks in Oregon mollusks from the Kenai Formation are from beds and Washington contain a related and probably ances- definitely of Seldovian age. Mayer (in Heer, 1869) tral species. Similarly, on Big Sitkinak Island described three species of mollusks from the Seldovian Oarpinw cappsens& and Alnus evidem are probably beds exposed at Coal Cove, Port Graham. D. W. descended from closely related species found in the non- Taylor has reclassified two of these mollusks as Plesiel- marine beds that lie below beds containing an early liptio onariotis (Mayer) and Melanoides furuhjelmi Miocene marine molluscan fauna (Wolfe, 1966) and (Mayer) . Taylor considers that the third mollusk, below beds that contain A. evidens. An age of earliest "Pahdina" abavia Mayer, may be one of the Hydro- Miocene seems most probable for the lower(2) Seldo- biidae, perhaps Litbglyphus. Both Plesieltiptio and vian floras, but we cannot exclude the possibility that Melanoides also occur at Houston, Alaska (USGS some of them may be of latest Oligocene age. Cenozoic loc. 23343, same as USGS Paleobotany loc. Pollen of the Compositae is not certainly known in 9365; see (pl. 1). A fauna from the Eagle River other parts of the world in beds of pre-Miocene age, but locality (USGS Cenozoic loc. 23368, same as USGS it is abundant in some Miocene beds. We have not Paleobotany loc. 9864) may also contain Plesielliptio found Compositae pollen in any Seldovian flora, but onariotis, but the material is too poor to be certain. we do not feel that any age significance can be attached Fresh-water gastropods, including a new species of to its absence. Campelomn and a probable new species of Bellamya, Homerla Sage also were obtained at Eagle River.

AGE The Homerian Stage is proposed as a provincial time- The precise age limits of the Seldovian Stage are un- stratigraphic unit that encompasses all plant-bearing certain. Paleobotanical correlations indicate that the strata in Alaska and in adjoining parts of the same beds from which upper(?) Seldovian floras have been ancient floristic province that are of the same age as obtained are of late early and (or) middle Miocene age; those parts of the Kenai Formation represented in the beds that have yielded lower (a) Seldovian floras probn- type section near Homer and in the reference section in bly are of early Miocene age, but some beds of lato the valley of the Clluitna River. Rocks belonging to Oligocene age may also be included. the Homerian Stage are recognized primarily on the The upper( 8) Seldovian floras contain several species basis of the fossil floras that they contain. The stage that are restricted to beds of early and middle Miocene is named after the town of Homer. age in Oregon and Washington : SaZim inquirenda, Al- The type section of the Homerian Stage is designated nw fairri, A. healyemis, Qwrcus bretzi, CoccuZus as the sequence of strata of the Kenai Formation ap- aurieulata, Platanus bendirei, Alchornea? n. sp., and proximately 2,000 feet thick that is exposed in coastal Prminm n. sp. "A." Some Seldovian species have an bluffs and in steep gullies and canyons along the east early and middle Miocene age range in Japan: Fagus shore of Cook Inlet and the north shore of Kachemak antipofi, Cocculths au-ta, and Acer ezoanum. None Bay from Troublesome Gulch (loc. 4129, pl. 1) past the of the other species known from upper(?) Seldovian town of Homer to Fritz Creek (loc. 9853, pl. 1). A floras contradict an age assignment to the early half of geologic map, geologic sections, and correlated the Miocene Epoch. Moreover, these floras probably measured stratigraphic sections of the beds included in are not of earliest Miocene age; in Washington and the type section are given by Barnes and Cobb (1959. Oregon, the lobed oaks and Cocczclw do not appear in pls. 18 and 19). the floras assigned to Wolfe7szone 1 (Wolfe, 1962, table The base of the Homerian Stage cannot be recognized 89.1) representing the basal part of the Miocene Series. and may not be represented in the type section. The A18 TERTIARY BIOSTRATIGRAPHY, COOK ImET RE.GI,ON, ALASKA lowest flora that has been collected there comes from Homerian Stage; only Carya, Pterocarya, Carpltzcs, beds exposed at the mouth of Troublesome Gulch; a CZadrastis, and Taxodiaceae persist as rare elements in squence several hund&d feet thick of massive sand- the leaf floras. Ulmaceae occur rarely in pollen floras ; stone beds underlies this flora in blufts that extend morth- fossil leaves of this family have not yet been found in westward to Anchor Point, but ththese beds have yielded Homerian strata in the Cook Inlet region. Fagaceae no fossil plants. The base of the Homerian Stage is are lacking from Homerian floras of the Cook Inlet represented, however, in exposures at our localities region, and Acer, common in Seldovian floras, is rare Dl949 and 9844 on the lower Chuitna River (pl. 2). in Homerian floras. Leaves of Fagus sp. cf. F. sum- These exposures are designated as a reference section tieugeniensis Holl., are associated with a marine mollus- for the Hometian Stage, and the base of the stage is can fauna of late Miocene or early Pliocene age (F. S. defined 8s lying immediately above the coal bed at lo- MacNeil, oral com'mun., 1961) on Cenotaph Island in cality D1949, which contains a Seldovian pollen flora. Lituya Bay (fig. 1) ; these beds presumably must lie Locality 9844, which lies 50 feet higher in the reference within the Homerian or Clamgulchian stage. section, contains a leaf flora diagnostic of the Homerian Early in our investigation, Wolfe (in MacNeil and stage. others, 1961) erroneously placed the type Homerian The top of the Homerian Stage is represented in the beds below beds that we now assign to the Seldovian type section but oannot yet be closely defined because it Stage, basing his conclusions on study of published lies within a ~t~atigra.phicsequence about 1,500 feet illustrations of a few species from each of these floras. thick in which we have not yet sought fossil plants. However, study of new leaf and pollen floras from the The highest flora typical of the Homerian Stage was Capps Glacier and Chuitna River area (pl. 2) and of obtained from exposures just west of the mouth of Fritz pollen floras from wells penetrating strata of both Cmk (loc. 9853). Our next higher flora, collected stages indicates clearly that the Homerian Stage is about '734 miles to the northeast and just west of the younger than and lies stratigraphically above the Sel- mouth of Cottonwood Creek (loc. 9856) contains species dovian Stage. This conclusion is reinforced by Wolfe's that we consider diagnostic of the Clamgulchian Stage. systematic studies (Wolfe, 1966), which show that A more precise definition of the top of the Homerian Seldovian species are consistently morphologically Stage must await additional collecting within the strati- more similar to Oligocene species than are Homerian graphic interval between these two localities. species. Furthermore, Homerian pollen and leaf floras are more depauperate and more similar to Recent irLmh Alaskan floras than are Seldovian floras. Stmb &sighed to the Homerian Stage in vakious The stratigraphically highest flora in the type section p$&s of the Cbok Inlet region have yielded 4'7 species of the Homerian Stage is that obtained at locality 9853 of plants bltsad oh fossil leaves (table 3) and 26 pollen immediately west of Fritz Creek. No fossil plants ahd spoN types ~.epresentingat least 15 vascular genera have been collected from the approximately 1,500 feet and li families of seed plants. Homerian leaf and of strata between localities 9853 and 9855 ; the latter is pollen floras &re much less diversified than Seldovian the lowest locality assigned here to the Clamgulchian floras. Most of the %arm-temperate elements now Stage. We propose that all or part of the beds in this exotic to Aluka that characterize Seldovim floras are interval should be assigned to the Homerian Stage if lacking in Homerian floras. The Homerian leaf floras they contain A2nw adumbrata or to the Clamgulchian collected thus far in the Cook Inlet region are dom- Stage if they contain Alnus gchmidtae. inated by Salioacetle and Betulaceae, but some of the pollen floras are dominhted by Pinaceae. The pollen and spore flora from the Homerian type The following megafossil species appear to be re- section is known from five samples, and we have one stricted to the Homerian Stage: Salix abkam, 8. sample (D1'717, pl. 2) that lies several hundred feet chzcitemh, S. bachemkensb, S. tymkam, Ahw stratigraphically above the beds at locality 9844, which covyzjna, A. ad~llPzbraCa,Carpinus cobbi, Coryhs ohui- are considered to be the lowest Homerian rocks in the tensis, Spi~aea:hophimi, Rhododendron weaveri, and @huitnaRiver reference section (pl. 2). V~cinizcrn homerrmsis. Salicaceae and Betulaceae In contrast to the richness and abundance of the dominate the Homerian floras both in numbers of speci- exotic dicotyledon element in the Seldovian pollen mens and in numbers of species. In addition, Erica- floras, this element is poorly represented in the Homer- ceae form an important element in some individual ian pollen floras. Thus far, only three exotic dicotyle- floras. Most of the warm-temperate exotic elements don genera are known; Carya and Pterocaya each o-f the Seldovian flo- .are lacking in floras from the have been found in two samples, and UZmw, in three.

Prop* ?if District Library AInsIia District STRATIGRAPHY AND PALEOBOTANY A19

TABLE3.-Checklist of Homerian megafossil flora [Numbers indicate localities]

Cache Chuitna River Type section, Homerian Creek Bpecies ______---=xlz 9852 1 41ZQ 1 W I 5820 ( 5821 861 9361 I 4131 --9853 Osmunda sp_------X ...... Onoclea sp. cf. 0.sensibilis L ------X ------Glyptostrobus europaeus (Brong.) Heer----- X X ------X ------X ------X ------X MetasequoiaglyptostroboidesHuandCheng- X X X X X X ------X X ------Tazodium distichurn Rich------X ------X ------Typha sp------X ...... X ------Cyperacites sp------X ------X ...... X ------X Populus eotremuloides Knowl------X ...... kenaiana Wolfe------X ------X ------X ...... washoensis R. W. Br------X ...... Salix alaskana Holl------X ------X ------X chuitensis Wolfe------X X ...... X ------X wnjmata (Holl.) Wolfe------X X ------X X cookensis Wolfe------X kachemakensis Wolfe------X X ...... X ------X tyonekana Wolfe------X ------X ------X ...... picroides (Heer) Wolfe------X ------,------Myrica sp------X ------Carya bendirei (Lesq.) Chan, and Axelr ------_------X -..------Pterocarya sp. cf. P. ni ella (Heer) Wolfe..------X ------a- Alnus adumbrata (~0115Wolfe ------X of. X ------.------X X X ------X corylina Knowl. and Cock ------X X X cf. X ------X X X X sp. aff. A. healyensis Wolfe ------X ------Betula sp. cf. B. thor Knowl------X ...... Carpinus cobbi Wolfe------.. X ...... X ------Corylus chuitensis Wolfe------X X ...... X ------X ------Cercidiphyllum crenatum (Un .) R. W. Br- X ------Hydrangea bendirei (Ward) snow1---.--.. .----. ----. --.-- .------....- X ----.--.------Ribes sp------X X ------Prunus sp------X ------Rubus sp------X ...... Spiraea hopkinsi Wolfe ------X X ------X X ------X X X weaveri Hall------X ------X X X X Cladrustis japonica (Tan. and Suz.) Wolfe------X ------cf. ------Sophorasp-,------X ------Aeer sp. cf. A. glabroides R. W. Br------X ------Elaeagnus sp------X ...... X ------

Cornus sp------X X ------a- X ------sp-_------X ...... Aralia sp------X ...... Arbutus sp-,------X ------...... Rhododendronweaveri(Holl.)Wolfe------X ------...... X X Vaccinium homerensis Wolfe ------X ------X X X sp------x ------x ...... x ------Halesia ap------cf. ------X ------Diervillasp------X ...... X ------Symphoricarpos sp------X ......

Three of &hesamples contain two of these genera, one lava near McCarthy, Alaska (locs. 9933,9935) (fig. 1). sample has only Ulrnzcs, and two samples entirely lack In all samples, Alnua pollen is more abundant than that the exdio element. Oarya and Pterocarya are also of the Betula type. Total Betulaceae pollen represents represented by leaflets, but as yet Ulmncs is represented between 20 and 80 percent of pollen tallied in each sam- in the Homerian of the Cook Inlet area only by pollen. ple. Pinua occurs in each sample, and Pkea in five of The abundance of the exotic elements in the Homerian is the six samples studied. Abies and Tsz~gaare some- low and does not exceed 3 percent in any one sample. what better represented in Homerian than in Seldovian Betulaceae and Pinaceae dominate all the Homerian pollen floras ; each genus occurs in five Homerian Sam- microfossil floras. Although none of the megafossil ples. Ericales are also better represented in the floras of Homerian age in the Cook Inlet region are Homerian and were noted in five samples; in three sam- known to contain Pinaceae, leaves and seeds of several ples, Ericalean pollen represents at least 5 percent of the genera of this family have been found in late(?) pollen tallied. No Compositae pollen has been found as Seldovian or Homerian beds at the base of the Wrangdl yet in Homerian strata. Am TERTIARY BIOSTRATIGRAPHY, 7OOK INLET REGION, ALASKA

MOLLUSK8 the same age as those parts of the Kenai Formation that A fresh-water clam has been collected from USGS are represented in the type section on the east shore of Cenozoic locality 23396 on the southeast side of the Cook Inlet and in the reference section on the north Moquawkie magnetic contact along the Beluga River. shore of Kachemak Bay. Rocks belonging to the Clam- Because most exposures of the Kenai Formation on the gulchian Stage are recognized primarily on the basis of same side of the contact in the Chuitna Valley about '7 the fossil floras that they contain. The stage is named miles to the south are of Homerian age, this clam is for the village of Clam Gulch, near the north end of probably also of Homerian age. D. W. Taylor has the type section. determined the specimen as probably EypImop&; this The type section of the Clamgulchian Stage is desig- genus is today confined to tropical and warm-temper- nated as the sequence of strata of the Kenai Formation ate East Asia. at least 2,000 and possibly 3,000 feet thick that is exposed AGE in coastal bluffs along the east shore of Cook Inlet from Happy Creek (loc. 9883, pl. 1) northward to a point The age limits of the Homerian Stage are very uncer- 4 miles north of Clam Gulch (loc. 9860). The type tain. Although the stratigraphic relationships indicate section has yielded several good floras, but we do not clearly that the Homerian Stage is younger than the have large floras from either the top or the bottom of Seldovian Stage and hence younger than the early half the stage there. Floras from less than 1,500 feet above of the Miocene Epoch, there is little in the Homerim the base of the stage are present, however, in exposures floras themselves to strengthen this assertion. The gen- in sea bluffs and steep gullies and canyons along the eral provincialism of the Homerian flora makes correla- north shore of Kachemak Bay between Cottonwood tion difficult with floras in Oregon and Washington or Creek (loc. 9855) and Swift Creek (loc. 9859). Geo- in Japan. The occurrence in IIomerian floras of Ptero- logic maps, geologic sections, and correlated measured carya cf. P. ./LigelZa and Carya bendirei indicates that at sections are given for the type section and the reference least part of the Homerian Stage can be no younger section by Bmnes and Cobb (1959, pls. 17,18, and 19). than late Miocene in age; neither of these species is The strata definitely assigned to the Clamgulchian known above the Miocene in Washington or Oregon. Stage in the reference section on the north shore of More significantly, Ahus adumbrata is known from the Kachemak Bay include approximately 1,500 feet of beds late Miocene Stinking Water flora in Oregon (reported that lie above Barnes and Cobb's (1959, pl. 19) coal bed there as AZnus harneyana Chaney and Axelrod, 1959, G and that are exposed between Cottonwood Creek and p. 158). Swift Creek. The upper part of this sequence consists Unnamed beds on the west shore of Port Moller on of poorly lithified clastic beds and of lignitic organic the Alaska Peninsula (loc. 5182, fig. 1) have produced beds, as does the sequence that constitutes the type sec- a small Homerian flora containing Sdix kache?~u&ertsis tion. The lower part of the strata exposed in the refer- and S. azcc~kana. The age of these beds is not certainly ence section, however, is well lithified. The organic known, but on lithologic and stratigraphic evidence, beds consist of subbituminous coal and are similar in C. A. Burk (oral commun., May 1963) considers them general lithologic character to the strata making up the to be equivalent to marine beds exposed east of Port Hoinerian Stage in the type section. Moller which contain a molluscan fauna that F. S. Mac- The lowest flora typical of the Clamgulchian Stage Neil (oral commun., July 1963) places in the later half was obtained just above Barnes and Cobb's coal bed G of the Miocene Epoch. in the I

TABLE5.-8tratigraphio distribution of significant species [a,fossil found at indicated locality. 0, determination based on fragmentary material. Dashed range lines indicate a posalble

SPECIES

wa =n v, 3

9860 +(I) 9861 I z 9862 5 I 9360 1 53 I t(2' 9859 1 I 2 I I 9854 I 1 9855 I I 4131 9853 I I +(I) +(I) I I 9361 I +(I) I I I 1 9851 +(4) +(4) I 0 t(l) z I I I I ' 5820 - 1 I I I 5821 - 1 I I I 9366 I I I I 4129 1 1 I I 9852 I 1 I I 4130 I +(I) 41) I 0 9844 I I I I Z I I I I 5 Upper(?) ...... 2 4 :Lower(?) ! ! ! ! ! ! •

Summary of the Floral Sequence somewhat debatable; for purposes of the table they are Data obtained from a study of megafossils (table 5) broken into two groups, thought to represent the lower and from a preliminary study of (table 6) and upper parts of the Seldovian Stage. The Home- indicate that three time-stratigraphic units-the Sel- rian flora from locality 9844 on the Chuitna River is as- dovian, Homerian, and Clamgulchian provincial sumed to be at least as old if not older than the lowest stages-can be recognized in surface outcrops of the flora collected in the Homerian type section. The Kenai Formation. Megafossil plants are more clearly Clamgulchian localities on Kachemak Bay are assumed I diagnostic of the threa provincial stages at the present for purposes of placement in table 5 to be at least as time, but microfossils are more useful in determining old as the lowest locality in the Clamgulchian type stage assignments for subsurf ace samples. section. The stratigraphic ranges of individual plant species Table 6 indicates the known stratigraphic ranges of given in table 5 are abstracted from the checklists given various genera in the pollen floras of the Kenai Forma- earlier and illustrate graphically the bases for recogni- tion, and presents a generalized concept of the relative tion of the three provincial stages. The relative strati- abundance of the genera. Several forms are found in graphic positions of some of the Seldovian floras are all three stages, but the exotic broad-leaved element is STRATIGRAPHY AND PALEOBOTANY A23

in Kenai and Tsadaka Foornatiorc, Cook Inlet region, Alaeb. extension of range of species based on generic determination of pollen (+); number of grains given in parentbeseal

obviously more abundant and diverse in the Seldovian that the presence of this group is not consistent enough Stage. Juglam, Ilex, Nyssa, and possibly Liquidamber to be a helpful index to sediments of post-Oligocene age. appear to be restricted to the Seldovian pollen floras. Pollen-abundance diagrams were prepared for each Our somewhat limited sampling is probably responsible of the pollen floras discussed in this report, but they are for the apparent restriction of Coryks to the Homerian not reproduced here because they yielded little or no and of Compositae to the Clamgulchian; the former is information that is not already apparent from generic known from leaves in the Seldovian as well as the EIom- lists of the various assemblages. A progressive general 1 erian, and Compositae pollen should also be expected impoverishment of the Kenai pollen flora through time throughout the Clamgulchian. The first occurrence of is apparent even from the limited sampling of the pollen Compositae pollen might be taken by some pdynologists reported here and is fully confirmed by the study of the to mark the base of the Miocene Series, but our experi- megafossils. A general predominance of Betulaceae ence in Miocane floras in Oregon and the late Pliocene and Pinaceae is characteristic of all Kenai Formation or early Pleistocene "Submarine Beach" at Nome indi- samples examined thus far, except for two from the cates that pollen of Compositae is rare or lacking in Seldovian type section near Capps Glacier which are many floras of undoubted Neogene age. It would seem dominated by Ulmw-ZeZkova pollen. More significant A24 TERTIARY BIOSTRATIGRAPHY, COOK INLET REGION, ALASKA

TABLE6.-8tratigraphio diatributim of polh types t Kencvi an& Tsaaaba Fmatiorcs. [ +, present but not included in counts; 0, less than 2 percent, e, more than 2 percent]

I I I Pollen type

I Dl955 111 ll+ Ill I I I I I I I I I I I I I I I I t T T T T ? T T Ill 1 ti IIII II II Ill I I I 01776 1 I + I + I + 4 P l+l I 4 I Ill I I1 II I Ill I I I 111 4 I 11 + I +IlI I 1 I Ill I II I I I Ill 1 I I 111 41 1 Ail I d r IIII II II I Ill I I I 1 I i i 1 I Ill I I b Ill I II II I Ill I I I

d 1 d I L 1 BJIAI I I Ill I I/ I I I1 Ill11 I I Ill I I I I I Ill 11 ? I PtP?+ I I I1 1111 I I + I+++I II IIII Ill A4 111 .:I1 I I Ill/ I I ++I I i+I i + are minor changes in the frequency of the exotic broad- percent of the Homerian pollen tallies and by less than leaved element; this element represents 5 and 19 percent 1 percent of the Clamgulchian tallies. The parameter of the count in two samples collected near Seldovia of the pollen floras expressed by the percentage of now- Point and from 1 to 92 percent of samples collected in exotic forms appears to constitute a highly useful strati- the type section of the Seldovian Stage. In contrast, graphic tool for the recognition of the three stages broad-leaved trees are represented by no more than 5 erected here. STRATIGRAPHY AND PALEOBOTANY A25 Both the megafossil and the microfossil floras demon- Desrription of some fossiLpldlzt ZocaEtities +n Chhkaloola strate that the broad-leaved forest trees now most wm- Formation-Continued UBGB PaZeq- m0n in warm-temperate floras were not dominant in botany local+ty DeecrjptZon of locaUty, coCkwtor and year (Zt bnowm) the cmk ~~l~~mgion ahrthe seldovian R~~~-9873----.-.--- Lat 61°44.8' N., long 148'62.8' W. Hanging Wall of strip pit topographically lower than 9872 sentatives of typically warm-temperate families such as in Mrak mine. Stratiaraohicalls- above 91872. Juglandam, ~a~aceae,~lmaceae, and Aceraceae grad- Anchorage (C-6) quad. Hopkins and Wolfe, ually disappeared, whereas representatives of typically 1982. as 9874------Lat 81°38.3' W., long 148'57.5' W. West side of cool-temperate families such Betulaceae, Rosaceae, valley of Wolverine Creek. Anchorage (W) Ericaceae, and Caprifoliaceae proliferated and kame quad. Hopkins and Wolfe, 1962. steadilv more abundant in floras that gradually ap- ) W7------at 61°48.0' N., long 147'59.5' W. North side proach"ed the character of the modern boreal floras of of cut along old Glenn Highway. Anchorage (D-3) quad. Hopkins and Wolfe, 1962. southern Alaska. Although deteriorating climatic con- -- Lat 61044.4, N., long 148057.5, to 148058.4, ditions undoubtedly. played- a major role in dictating Ooilections from dumps of strip pits in Gvan the changing generic composition of successively Jones mine. Between Premier (No. 5) and younger Kenai floras, assumptions concerning past cli- Jonesville (No. 3) coal groups. Anchorage (C-6) quad. Hopkins and Wolfe, 1962. matic conditions do not provide a secure basis for dating individual fossil floras. ~~~~~~d,age assignments must Description of megafossi6phznt localities in Kenai ana Tsadaka Formations be based upon accumulated knowledge of the strati- Description of locality, etage wsignment, graphic ranges of the individual species represented in a,"I",zf Etii, and oo~~ectorand year ($1known) the megafossil florss md of the considerably longer 3505------.- Chinitna Bay, near entrance to bay on north local stratigraphic ranges of the individual genera con- side. From sandstone at top of exposure above conglomerate. Seldovian (?). Stan- stituting the microfossil floras. Heer (1869), in basing ton and Martin, 1904. his Miocene age assignment of certain Kenai floras upon 4129------At entrance to Troublesome Gulch. Seldovia the supposed presence in them of plant species found ((3-5) quad. Homerian. Weaver, 1906. 4130------0.5 mile south of town of Old Tyonek on sea in Miocene beds in Switzerland, was more nearly cor- cliff. Tyanelr (A4) quad. Homerian. rect than his critics, who named these floras "the Arctic Weaver, 1906. Miocene" and diagnosed them as Eocene t~l~oughoutthe 4131------Near entrance to Fritz Creek, Kachemak Bay. next 80 years because they believed that an Eocene flora Seldovia ((3-4) quad. I-Iomerian. Weaver, 1906. found in northern latitudes should resemble a ~~ocene5820------BIUE Point, 7 miles west of Homer, "30 ft. flora found in middle latitudes ! below Bradley coal" according to the speci- men label, but F. F. Barnes informs us that FOSSIL-PLANT LOCALITIES the Cooper coal bed is the only named coal bed present at BluE Point. Seldovia (G5) Description, of some fossil-plant localities in Chickaloon quad. Homerian. Stone and Stanton, 1904. Formation 5821------Talus on beach at Bluff Point about 1.5 miles UBGB PaZep- west of Cook Inlet Coal Field Company's botany looat+ty Deswipttom of ZomZjtv, cozleotor and year (it known) mine. Seldovia (C-5) quad. Homerian. 5892------Lat 61'40.3' N., long 149O03.5' W. North side Stone and Stanton, 1904. cut On side of 6081------2.5 miles southwest of Point Naskowhak. Sel- Matanuska River. About 1,500 ft above base dovia (B-5) quad. Seldovian. Martin, 1911. (C-6)quad. Mar- - From Cache Creek, 1.5 miles above Cache Creek tin, 1910; Hopkins and Wolfe, 1962. Mining Company's camp. Talkeetna (B-2) 9870------Lat 61°42.6' N., long 149'05' W. At new cut at quad. Seldovian. Capps, 1911. old Baxter mine on east side of Moose Creek @3&%------Mills Creek Basin, Chicago Gulch. Talkeetna valley. Premier coal group. Anchorage (B-4) quad. Seldovian. Capps, 1911. (C-6) quad. Hopkins and Wolfe, 1962. 8380------Lat 61°41' N., long 149"08' W. Core material, 9871 ------Lat 61'45.2' N., long 148'52.9' W. Hanging Anchorage (C-6) quad. Seldovian. Waring wall of strip pit topographically high in Mrak and Davidson, 1932. mine. Stratigraphically below 9872. An- 9358------Lat 61°42.1' N., long 149O05.6' W. West side of chorage (D-6) quad. Hopkins and Wolfe, Tsadaka Canyon. Anchorage (C-6) quad. 1962. Seldovian. Barnes, Bender, and Brown, 9872-,-,,----- Lat 61'44.9' N., long 148O53.5' W. Hanging wall I 1955 ; Hopkins and Wolfe, 1962. of strip pit topographically lower than 9871 9360------Lat 60°01.8' N., long 151°42.1' W. 0.75 mile in Mrak mine. istratigraphically above 9871. south of mouth of Deep Creek. Kenai (8-5) Anchorage (C-6) quad. Hopkins and Wolfe, quad. Olamgulchian. Bender and Brown, 1962. 1955. A26 TERTIARY BIOSTRATIGRAPHY, COOK INLET REGION, ALASKA

Deawiption of megafossil-plant locaMties in KenaZ and Tsadaka Description of megafossil-plant localities in Kenai and Tsaalaka Fmmaticms-Continued Pormaticms--Continued UBGB Paleo- Deacriptirm of locality, stage msignment, UBGB Paleo- Deacriptirm of locality, stage aasinnme.tct, botany locakty and collector ant3 year (of known) botany ZocaZity and collector and year (sf known) 93Q1-,------Lat 59'39.4' N., long 151'26.3' W. Sea cliff 9856------Lat 59'23.7' N., long 151'53.7' W. North side of about 1 mile south of Millers Landing. Sel- Coal Cove on Port Graham, probably the dovia ((2-4) quad. Homerian. Barnes, "Sinus Anglorum" (English Bay) locality of Bender, and Brown, 1955; Wolfe, 1962. Heer. Seldovia (B-6) quad. Seldovian. 9364 ------Lat 61'39.8' N., long 149O27.9' W. On Coal Hopkins, Schmidt, and Wolfe, 1962. Creek. Anchorage (C-7) quad. Seldovian. 9857------Lat 59'25.0' N., long 151'53.1' W. 0.6 mile Barnes, Bender, and Brown, 1955. south of Point Pogibshi. Seldovia (B-6) 9365------Same as USGS Cenozoic loc. 23343. Lat quad. Seldovian. Hopkins, Schmidt, and 61'38.4' N., long 149O50.8' W. In Houston Wolfe, 1962. strip pit. Anchorage (C-8) quad. Seldo- 9858-----,---- I~at5g028.3' N., long 151°40.6' W. 0.7 mile east viah. Barnes and Brown, 1955 ; Hopkins and of Seldovia Point. Seldovia (B-5) quad. Wolfe, 1962. Seldovian. Hopkins, Schmidt, and Wolfe, 9366----,----- Lat 59'40.3' N., long lfiL042.4' W. 0.25 mile 1982. northwest of mouth of Diamond Creek. Sel- 9859 --,------Lat 59O49.2' N. long 151°07.4' W. East bank of dovia ((3-5) quad. Homerian. Bender and Swift Greek. Seldovia (D4) quad. Clam- Brown, 1955. gulchian. Hopkins and Wolfe, 1962. 9760------0.25 mile west of southern tip of Redoubt 9880 ------Lat60°15.2'N.,long151"23.5' W. Seacliffs0.9 Point. Kenai (B-6) quad. Seldovian. Gulf mile north of Clam Gulch. Kenai (B-4) Oil Corp. quad. Clamgulchian. IIopkins and Wolfe, 9761------Cape Douglas. Afognak quad. Seldovian. 19f32. Gulf Oil Corp. 9861--,----,,, Lat 60°15.7' N., long 151°23.3' W. Sea cliffs 1.5 9763------Lat 60°11.5' N., long 15W28.5' W. Sea cliff miles north of Clam Gulch. Kenai (B4) north of Ninilchik. Clamgulchian. Gulf quad. Clamgulchian. Hopkills and Wolfe, Oil Corp. 1982. 9844 ------,--Lat 61°07.1' N., long 151°18.1' W. South bank 9862------Lat 60°12.5' N., long 151°25.5' W. Sea cliffs 2.4 of Chnitna River. Tyonek (8-4) quad. Ho- miles south of Clam Gulch. Kenai (A-4) merian. Barnes, 1961; Wolfe, 1962. quad. Clamgulchian. Hopkins and Wolfe, 9845 ------Lat 61°18.9' N., long 151'46.2' W. Cliffs on 1962. south side of Capps Glacier. Tyonek (B-5) 9863------Lat 61°19.1' N., long 149'36.5' W. South bank quad. Seldovian. Volfe, 1982. of Eagle River. Anchorage (B-7) quad. 9M6 ------11at 61 "16.7' N., long 18l045.1' W. West side of Seldovian. Hopkins and Wolfe, 1962. high hill. Tyonek (B-5) quad. Seldovian. 9864----,----- Same as USGS Cenozoic loc. 23368. Lat 61" Barnes, 1961 ;Wolfe, 1962. 18.7' N., long 149'34.8' W. South bank of 984------Lat 61°14.2' N., long 151°14.7' W. South bank Eagle River. Anchorage (B-7) qnad. Sel- of Beluga =ver. Tyonek (-4-4)quad. Sel- dovian. Hopkins and Wolfe, 1962. dovian. Wolfe, 1962. 9865 ------Lat 61°39.4' N., long 14Q027.8' W. North bank 9849--,------Lat 61°18.1' N., long 151°14.4' W. North bank of Little Susitna River. Anchorage (G7) of Beluga River. Tyonek (B-4) quad. Sel- quad. Seldovian. Hopkins ahd Wolfe, 1962. dovian. Wolfe, 1962. 9866 ------Lat 61'41.7' N., long 149"14.7' W. West bank of 9850---,------Lat 61°25.6' N., long 151°31.2' W. East bank Little Susitna River. Anchorage (G6) quad. of Coal Creek. Tyonek (B-5) quad. Sel- Seldovian. Hopkins and Wolfe, 1962. dorian. \Volfe, 1962. 9867------, Ilat 62'29.4' N., long 13O058.7' W. 8011th side of Cache Creek opposite mouth of Rambler 9851 ------Idat 50O38.6' N., long 151°35.1' W. In sea cliffs Creek. Talkeetna (B-2) quad. Seldorian. west of Homer. Seldovia (C-5) quad. Ho- Hopkins and Wolfe, 1%2. merian. Wolfe, 1962. 9868 ------Lat 62"B.Q' N., long 150'56.9' W. South side of 98.52 Lat 50O43.2' N., long 151'49.4' W. ------0.25 mile Cache Creek. Talkeetna (B-2) quad. south of ~MutnalaGulch in sea cliffs. Seldovia IIomerian. Hopkins and Wolfe, 1962. (G5) quad. Homerian. Wolfe, 1962. 9883------Near mouth of Happy Creek. Seldovia (D-5) 9853,---,----- Lat 69O40.9' N., long 151°22.6' W. Just west quad. Clamgulchian. Benninghof, 1955. of mouth of Fritz Creek. Seldovia (C4) / 9884,,------, North of Harriet Point. See. 13, T. 5 N., R. 18 quad. Homerian. Wolfe, 1962. W. Kenai (G6) qnad. Seldovian. Shell 9&51----,---,- Lat 59O45.1' N., long 151°10.2' W. 0.33 mile I. Oil 00. west of mouth of Eastland Creek. Seldovia 9885------Sorth of Harriet Point. Sec. 25, T. 5 N., R. 18 ((3-4)quad. Clamgulchian. Wolfe, 1962. W. Kenai (B-6) q11:ld. Se1dovi:in. Shell 9W------Lat 50O44.0' N., long 151°12.4' W. 0.25 mile Oil Co. west of mouth of Cottonwood Creek. Seldo- 9886 ------North of Harriet Point. Sec. 15, T. 4 N., R. 18 via ((3-4) quad. Clamgulchian. Wolfe, W. Kenai (B-6) qnad. Seldovian. Shell 1962. Oil Co. STRATIGRAPHY AND PALEOBOTANY A27

Desmiptbn of megafossil-plant localities in Kenai and Tsadaka Description of fossil-plant localities outside the Cook Inlct re- Formatiom-Continued gion mentioned in teat U8W Paleo- Description of local&ty, stage assignment, botany locality and collector and year (if known) USGB Paleo- Locality descrfption, 8tratigrapWc assignment, and botany locaUty collactor and year (4j know) 9887 Near Redoubt Point. Sec. 33, T. 3 N., R. 18 W. Kenai (B-G) quad. Seldovian. Shell Oil Co. 3519------About 1 mile east of north from Alaska Packers 0037------Lat 61°18.4' N., long 151°46.5' W. Cliffs on Association cannery at Chignik and about 200 south side of Gap~lsGlacier. Tyonek (B-5) yd south of native village. Paleocene. Ghig- qnad. Seldovian. British Petroleum Co., nil< quad. Stanton, 1904. 1962. 3522------B'rom steeply indined beds in ralley of creek 9945------On Harriet Creek, 21.85 niiles east and 30.15 about 1 mile northeast of Pacific Packing and iniles north of southwest corner of Kenai N. W. Company rannery, Anchorage Bay. quad. (Scale 1 : 250,000). Kenai (B-8) Paleocene. Ohignik quad. Stanton and quad. Seldovian. Mobil Oil Go. Stone, 1901. 3523------Description of microfossidplamt localities in Xcnai E'ormatio?~ Talus slopes on mountain about 1 mile south- east of Pacific Packing and N. W. Company UBGB Mcality, Desoription of locality, stage assignment, and Denver catalog colZector and year cannery, Anchorage Bay. Chignik quad. Dl717 Lat 61°06.8' N., long 151°20.4' W. Tyonek (A- Paleocene. Stanton and Stone, 1904. 4) quad. Homerian. Barnes, 1961. 3652------Head of Hamilton Bay, Knpreanof Island. Dl718------Lat 61°W.3' N., long 151°30'W. Tyonek (A-5) Petersburg quad. Paleocene. Kindle, 1905. quad. Seldovian. Barnes, 1961. 4389 ------,South side of Hamilton Bay, near head. High- Dl719 ------.Lat 61°9.4' N., long 151°30.4' W. Tyonek (A- est of three horizons. Petersburg quad. 5) quad. Seldovian. Barnes, 1961. Paleocene. Atwood, 1907. Dl720 Lat 61°10.2' N., long 151°34.1' W. Tyonek (A- 4391------. South side of Hamilton Bay, near head. Inter- 5) quad. Seldovian. Barnes, 1961. mediate of three horizons. Petersbnrg quad. Dl775 Sec. 34, T. 51 S., R. 14 W. Kenai (A-5) quad. Paleocene. Atwood, lW. Clamgulchian. Barnes, 1960. 4392 South side of Hamilton Bay, near head. Lowest Dl776 ------.See. 24, T. 6 S., R. 14 W. Seldovia (C4) quad. ------Homerian. Barnes, 1960. of three horizons. Petersburg quad. Paleo- Dl943-----,--. Same as loc. 9860. Kenai (B4) quad. Clam- cene. Atwood, 1907. gulchian. Hopkins and Wolfe, 1962. 5182------West side of Herendeen Bay, opposite Marble Dl944 -----,--. Same as loc. 9858. Seldovia (Bd) quad. Sel- Point. Port Moller quad. Homerian. At- dovian. Hopkins, Schmidt, and Wolfe, 1962. x~aodand Eakin, 1908. D1045-----,--. Carbonaceous claystone 30 ft stratigraphically 7474------. Hamilton Bay, Knpreanof Island. Petersburg above loc. 0859. Seldovia (D4) quad. Clam- quad. Paleocene. Wright, 1904. gulchian. Hoykins and Wolfe, 1962. 7565------Near head of Hamilton Bay, in middle of broad Dl946 Same as loc. 9845. Tyonek (B-5) quad. Sel- dovian. Wolfe, 1962 head land and on south side of bay. A little Dl947 Same as loc. 9366. Seldovia (G5) quad. south of 4.25 miles true east of Point Hamil- Homerian. Bender and Brown, 1955. ton, opposite 8.25 fathom mark. Petersburg D1948-----,--. Coal bed 10 ft stratigraphically below loc. m53. quad. Paleocene. Ruddington, 1922. Seldovia (C-4) quad. Homerian. Wolfe, 8680------Crooked River (tributary of Serentymile 1962. River), about 1 mile from mouth. Eagle quad. D1949-----,--. Coal bed about 30 ft stratigraphically below loc. Paleocene. Mertie, 1938. 9844. Tyonek (A-4) quad. Seldovian. Wolfe, 8681 Fourth of July Creek, about 7 miles from mouth. 1962. Oharley River qnad. Paleocene. Mertie, 1938. Dl950------_-. Coal bed 30 ft stratigraphically above loc. 9851. 9924 ------. Lat 63'52.6' N., long 14S040' W. From interval Seldovia (G5) quad. Homerian. Wolfe, 1962. of about No. 1 bed, on east side of Coal Creek. D1951------. Coal bed 130 ft stratigraphically above loc. 9851. Seldovian (C5) quad. Homerian. Wolfe, Healy (Dl) quad. Seldovian. Wolfe, 1963. 1962. 9933------. Lat 6l039.7' N., long 142O10.1' W. South side of D1952------. Coal bed about 125 ft stratigraphically lower Skolai Ckeek. McCarthy ((G-4) qnad. Up- than loc. 9846. Lat 61°15' N., long 151°45' per ( ?) Seldovian or Homerian (?). Wolfe, W. Tyonek (B-5) quad. Seldorian. WoIfe, 19G3. 1962. 99.35 ------.Lat 61°39.2' N., long 142'40.7' W. Northeast D1953-----,--. Coal bed about 150 ft stratigraphically above side of West Fork Glacier. McCarthy (C5) 1 loc. 9845. Tyonek (B-5) quad. Seldovian. quad. Upper ( ?) Seldovian. Wolfe, 1963. Wolfe, 1962. 10002 Lat 5G032' N., long 154°07' W. Sitkinak Is- Dl954 Same as loc. 9854. Seldovia (C-4) qurrd. land. Trinity Islands quad. Oligocene. Clamgulchian. Wolfe, 1962. Dl055 ------.Same as loc. 9855. Seldovia (C4) quad. Earle Taylor. Clamgulchian. Wolfe, 1962. 10003------. Lat 56"32' N., long 154°07' W. Sitkinak Island. D1973------. Same as loc. 9857. Seldovia (B-6) quad. Sel- Trinity Islands quad. Oligocene. G. W. dovian. Ffopkins, Schmidt, and Wolfe, 1962. Moore, 1962 A28 TERTIARY BIOSTRATIQRAPHY, COOK INLET REGION, ALASKA

FOSSIL-MOLLUSK LOCALITIES Grantz, Arthur, and Jones, D. L., 1960, Stratigraphy and age of the Matanuska Formation, south-central Alaska, in Damiptim of Zocalities i7~Kmi and ChtclcaZoort Fmatim Short papers in the geological sciences: U.S. Geol. Survey USGl Ueeqozoiu Locallty des-tion 8tratlpt~phi0 aes@nment, and Prof. Paper 4WB, p. B347-B350. locakty cotlector and year 23343------Same as USGS Paleobotany loc. 9365. Anchor- Grantz, Arthur, and Wolfe, J. A., 1961, Age of the Arkose Ridge age ((3-8) quad. Kenai Formation. Seldov- Formation, south-central Alaska: Am. Assoc. Petroleum ian. Barnes and Brown, 1955; Hopkins and Geologists Bull., v. 45, no. 10, p. 1762-1765. Wolfe, 1962. Grantz, Arthur, Zeitz, Isidore, and Andreasen, G. E., 1963, An 23367------Lat 61'48.2' N., long 148'26.5' W. Along west aeromagnetic reconnaissance of the Cook Inlet area, Alaska : side of Chickaloon River. Anchorage (D-4) U.S. Geol. Survey Prof. Paper 316-6, p. 117-134. quad. Chickaloon Formation. Paleocene. Hazzard, J. C., Moran, W. R., Lian, H. M., and Simonson, R. R., Hopkins and Wolfe, 1962. 1963, The search for oil in Alaska : an analysis of a success- 23368 ------Same as USGS Paleobotany loc. 9864. Anchor- ful exploration cycle, in Proceedings of the Second Sym- age (B-7) quad. Kenai Formation. Sel- posium on the development of petroleum resources of Asia dovian. Hopkins and Wolfe, 1962. and the Far East: United Nations Econ. Comm. Asia and 23396------Lat 61°13.4' N., long 151°11.3' W. Along north Far East, Mineral Resources Devel. ser. no. 18, v. 1, p. side of Beluga River. Tyonek (8-4) quad. 372-380. Kenai Formation. Homerian ( ?) . Barnes, Heer, Oswald, 1868, Flora fossilis arctica : Ziirich, J. Wurster 1962. and Co., v. 1, 192 p., 49 pls. 1869, Flora fossilis alaskana : K. Svenska Vetenskap- REFERENCES sakad. Handl., v. 8, no. 4, 41 p., 10 pls. Barnes, F. F., 1962a. Geologic map of lower Matanuska Valley, Hill, M. L., 1963, Here's how Alaska's search for oil is going Alaska: U.S. Geol. Survey Misc. Geol. Inv. Map 1-359. at Cook Inlet: Oil and Gas Jour., v. 61, no. 26, p. 194-198. 1962b, Topographic and geologic map of the Knob Creek 1963, Occurrences of petroleum in the Cook Inlet area, area of the Wishbone Hill district, Matanuska coal field, Alaska : World Petroleum Gong., 6th, Frankfurt, June 1963, Alaska: U.S. Geol. Survey Coal Inv. Map (3-51. Proc., see. 1, paper 39, p. 509-517. -1962c, Variation in rank of Tertiary coals in the Cook Hollick, Arthur, 1936, Tertiary floras of Alaska : U.S. Geol. Sur- Inlet basin, Alaska, in Short papers in geology and hy- vey Prof. Paper 182,185 p., 1.22 pls. drology: U.S. Geol. Survey Prof. Paper 4504, p. C14-C16. Karlstrom, T. N. V., 1964, Quaternary geology of the Kenai Low- -1966, Geology and coal resources of the Beluga-Yentna land and glacial history of the Cook Inlet region, Alaska: region, Alaska. U.S. Geol. Survey Bull. 1202-C, 54 p. U.S. Geol. Survey Prof. Paper 443, 69 p., 7 pls. Barnes, F. F., and Cobb, E. H., 1959, Geology and coal resources Kelly, T. E., 1963, Geology and hydrocarbons in Cook Inlet basin, of the Homer district, Kenai coal field, Alaska: U.S. Geol. Alaska: Am. Assoe. Petroleum Geologists, Mem. 2, p. Survey Bull. 1058-F, p. 217-260. 278-296. Barnes, F. F., and Payne, T. G., 1956, The Wishbone Hill district, Knowlton, F. H., 1894, A review of the fossil flora of Alaska, Matanuska coal Aeld, Alaska: U.S. Geol. Survey Bull. 1016, with descriptions of new species: U.S. Natl. Museum Proc., 88 p. V. 17, p. 207-240. Barnes, F. F., and Sokol, Daniel, 1959, Geology and coal re- 1904, Fossil plants from Kukak Bay : Harriman Alaska sources of the Little Susitna district, Matanuska coal field, Exped., v. 4, p. 149-162. Alaska: U.S. Geol. Survey Bull. 1058-D, p. 121-138. Koch, B. E., 1959, Contribution to the stratigraphy of the non- Brooks, A. H., 1908, The geography and geology of Alaska: marine Tertiary deposits: Medd. om Gr$nland, v. 162, no. 1, U.S. Geol. Survey Prof. Paper 45,327 p. p. 1-100. Brown, R. W., 1962, Paleocene flora of the Rocky Mountains and 1963, Fossil plants from the lower Paleocene of the Agat- Great Plains: U.S. Geol. Survey Prof. Paper 375, 119 p., dalen (Angmartussut) area, central Nugssuaq Peninsula, 69 pls. northwest Greenland : Medd. om Gr$nland, v. 172, no. 5, 120 Capps, S. R., 1913, The Yentna district, Alaska: U.S. Geol. p., 55 pls. Survey Bull. 534,75 p. MacNeil, F. S., Wolfe, J. A., Miller, D. J., and Hopkins, D. M., 1927, Geology of the upper Matanuska Valley, Alaska: 1961, Correlation of Tertiary formations of Alaska: Am. U.S. Geol. Survey Bull. 791, 92 p. Bssoc. Petroleum Geologists Bull., v. 45, no. 11, p. 1801-1809. 1935, The southern Alaska Range: U.S. Geol. Survey Martin, G. C., Johnson, B. L., and Grant, U. S., 1915, Geology and Bull. 862, 101 p. mineral resources of Kenai Peninsula, Alaska : U.S. Geol. 1940, Geology of the Alaska Railroad region : U.S. Geol. Survey Bull. 587,243 p. Survey Bull. 907,201 p. Martin, G. C., and Katz, F. J., 1912, Geology and coal fields of Chaney, R. W., and Axelrod, D. I., 1959, Miocene floras of the the lower Matanuska Valley, Alaska: U.S. Geol. Survey Columbia Plateau, Pt. 2, Systematic considerations : Carnegie Inst. Washington Pub. 617, p. 135-237,44 pls. Bull. 500,98 p. Moffitt, F. XI., 1927, The Iniskin-Chinitna Peninsula and the Dall, W. H., and Harris, G. D., 1892, Correlation papers: Neocene: U.S. Geol. Survey Bull. 84, p. 232-268. Snug Harbor district, Alaska: U.S. Geol. Survey Bull. 789, Davison, K., 1963, Cook Inlet gas finds are important: World 71 P. Oil, v. 157, no. 7, p. 92-98. Parkinson, L. J., 1962, One field, one giant . . . the story of Gardner, J. S., and Ettingshausen, Constantin, 1879, A mono- Swanson River: Oil and Gas Jour., v. 60, no. 13, p. 180-183. graph of the British Eocene flora: London, Palaeontogr. Payne, T. G., 1955, Mesozoic and Cenozoic tectonic elements of pot., 159 p. Alaska: U.S. Geol. Survey Misc. Geol. Inv. Map 1-84. STRATIGRAPHY AND PALEOBOTANY A29

Tuck, Ralph, 1934, The Curry district, Alaska : U.S. Geol. Survey Wolfe, J. A., 1962, A Miocene pollen sequence from the Cascade Bull. 862-C,p. 99-140. Range of northern Oregon, in Short papera in geology and -1937, The Eska Creek coal deposits, Matanuska Valley, hydrology : U.S. Geol. Survey Prof. Paper WC,p. C81-084. Alaska : U.S. Geol. Survey Bull. 880-D, p. 185-214. 1966, Tertiary plants from the Cook Inlet region, Alaska : Waring, G. A., 1934, Core drilling for coal in the Moose Creek U.S. Geol. Survey Prof. Paper 398-B., 31 p. area, Alaska : U.S. Geol. Survey Bull. 857-E, p. 155-173. Wood, H. E. and others, 1941, Nomenclature and correlation of 1936, Geology of the Anthracite Ridge coal district, the North American continental Tertiary: Geol. Soc. Alaska : U.S. Geol. Survey Bull. 861, 57 p. America Bull., v. 52,no. 1, p. 1-48.

UNITED STATES DEPARTMENT OF THE INTERIOR PROFESSIONAL PAPER 398-A

GEOLOGICAL SURVEY Propey!:I PLATE 2 DiigtcictI' ry + > -7 aaska l)istri@ EXPLANATION [1:

Landside deposits 3 0 - Kenai Formation Upper, coal-bearing member

000000 , Tkl bed oeooo Kenai Formation Lower, pebbly sandstone member

c .- 2 ---- D1717,, SEA LEVEL V) Contact Stratigraphic position of Homerian Dashed where appmximately located microf ossil locality ----- 9937. Fault Stratigraphic position of Sel- Dashed where apjmnSmately located dovian megafossil locality

9844@ Dl9490 Stratigraphic position of Homerian Stratigraphic position of Sel- megafossil locality dovian microfossil locality

A' 2000'

-.._...... covered 1500'

1000'

------500'

SEA LEVEL STRUCTURE SECTION BETWEEN CAPPS GLACIER AND LOWER CHUITNA RIVER, ALASKA

1 0 1 2 3 4 5 MILES

1 0 1 2 3 4 5 KILOMETERS IIIII I I I I

206-780 0 - 66 (In pocket) No. 2